mirror of https://github.com/go-gitea/gitea
Git with a cup of tea, painless self-hosted git service
Mirror for internal git.with.parts use
https://git.with.parts
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
42034 lines
1.3 MiB
42034 lines
1.3 MiB
/* Copyright 2012 Mozilla Foundation |
|
* |
|
* Licensed under the Apache License, Version 2.0 (the "License"); |
|
* you may not use this file except in compliance with the License. |
|
* You may obtain a copy of the License at |
|
* |
|
* http://www.apache.org/licenses/LICENSE-2.0 |
|
* |
|
* Unless required by applicable law or agreed to in writing, software |
|
* distributed under the License is distributed on an "AS IS" BASIS, |
|
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
|
* See the License for the specific language governing permissions and |
|
* limitations under the License. |
|
*/ |
|
/* jshint globalstrict: false */ |
|
/* umdutils ignore */ |
|
|
|
(function (root, factory) { |
|
'use strict'; |
|
if (typeof define === 'function' && define.amd) { |
|
define('pdfjs-dist/build/pdf.worker', ['exports'], factory); |
|
} else if (typeof exports !== 'undefined') { |
|
factory(exports); |
|
} else { |
|
factory((root.pdfjsDistBuildPdfWorker = {})); |
|
} |
|
}(this, function (exports) { |
|
// Use strict in our context only - users might not want it |
|
'use strict'; |
|
|
|
var pdfjsVersion = '1.4.20'; |
|
var pdfjsBuild = 'b15f335'; |
|
|
|
var pdfjsFilePath = |
|
typeof document !== 'undefined' && document.currentScript ? |
|
document.currentScript.src : null; |
|
|
|
var pdfjsLibs = {}; |
|
|
|
(function pdfjsWrapper() { |
|
|
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreArithmeticDecoder = {})); |
|
} |
|
}(this, function (exports) { |
|
|
|
/* This class implements the QM Coder decoding as defined in |
|
* JPEG 2000 Part I Final Committee Draft Version 1.0 |
|
* Annex C.3 Arithmetic decoding procedure |
|
* available at http://www.jpeg.org/public/fcd15444-1.pdf |
|
* |
|
* The arithmetic decoder is used in conjunction with context models to decode |
|
* JPEG2000 and JBIG2 streams. |
|
*/ |
|
var ArithmeticDecoder = (function ArithmeticDecoderClosure() { |
|
// Table C-2 |
|
var QeTable = [ |
|
{qe: 0x5601, nmps: 1, nlps: 1, switchFlag: 1}, |
|
{qe: 0x3401, nmps: 2, nlps: 6, switchFlag: 0}, |
|
{qe: 0x1801, nmps: 3, nlps: 9, switchFlag: 0}, |
|
{qe: 0x0AC1, nmps: 4, nlps: 12, switchFlag: 0}, |
|
{qe: 0x0521, nmps: 5, nlps: 29, switchFlag: 0}, |
|
{qe: 0x0221, nmps: 38, nlps: 33, switchFlag: 0}, |
|
{qe: 0x5601, nmps: 7, nlps: 6, switchFlag: 1}, |
|
{qe: 0x5401, nmps: 8, nlps: 14, switchFlag: 0}, |
|
{qe: 0x4801, nmps: 9, nlps: 14, switchFlag: 0}, |
|
{qe: 0x3801, nmps: 10, nlps: 14, switchFlag: 0}, |
|
{qe: 0x3001, nmps: 11, nlps: 17, switchFlag: 0}, |
|
{qe: 0x2401, nmps: 12, nlps: 18, switchFlag: 0}, |
|
{qe: 0x1C01, nmps: 13, nlps: 20, switchFlag: 0}, |
|
{qe: 0x1601, nmps: 29, nlps: 21, switchFlag: 0}, |
|
{qe: 0x5601, nmps: 15, nlps: 14, switchFlag: 1}, |
|
{qe: 0x5401, nmps: 16, nlps: 14, switchFlag: 0}, |
|
{qe: 0x5101, nmps: 17, nlps: 15, switchFlag: 0}, |
|
{qe: 0x4801, nmps: 18, nlps: 16, switchFlag: 0}, |
|
{qe: 0x3801, nmps: 19, nlps: 17, switchFlag: 0}, |
|
{qe: 0x3401, nmps: 20, nlps: 18, switchFlag: 0}, |
|
{qe: 0x3001, nmps: 21, nlps: 19, switchFlag: 0}, |
|
{qe: 0x2801, nmps: 22, nlps: 19, switchFlag: 0}, |
|
{qe: 0x2401, nmps: 23, nlps: 20, switchFlag: 0}, |
|
{qe: 0x2201, nmps: 24, nlps: 21, switchFlag: 0}, |
|
{qe: 0x1C01, nmps: 25, nlps: 22, switchFlag: 0}, |
|
{qe: 0x1801, nmps: 26, nlps: 23, switchFlag: 0}, |
|
{qe: 0x1601, nmps: 27, nlps: 24, switchFlag: 0}, |
|
{qe: 0x1401, nmps: 28, nlps: 25, switchFlag: 0}, |
|
{qe: 0x1201, nmps: 29, nlps: 26, switchFlag: 0}, |
|
{qe: 0x1101, nmps: 30, nlps: 27, switchFlag: 0}, |
|
{qe: 0x0AC1, nmps: 31, nlps: 28, switchFlag: 0}, |
|
{qe: 0x09C1, nmps: 32, nlps: 29, switchFlag: 0}, |
|
{qe: 0x08A1, nmps: 33, nlps: 30, switchFlag: 0}, |
|
{qe: 0x0521, nmps: 34, nlps: 31, switchFlag: 0}, |
|
{qe: 0x0441, nmps: 35, nlps: 32, switchFlag: 0}, |
|
{qe: 0x02A1, nmps: 36, nlps: 33, switchFlag: 0}, |
|
{qe: 0x0221, nmps: 37, nlps: 34, switchFlag: 0}, |
|
{qe: 0x0141, nmps: 38, nlps: 35, switchFlag: 0}, |
|
{qe: 0x0111, nmps: 39, nlps: 36, switchFlag: 0}, |
|
{qe: 0x0085, nmps: 40, nlps: 37, switchFlag: 0}, |
|
{qe: 0x0049, nmps: 41, nlps: 38, switchFlag: 0}, |
|
{qe: 0x0025, nmps: 42, nlps: 39, switchFlag: 0}, |
|
{qe: 0x0015, nmps: 43, nlps: 40, switchFlag: 0}, |
|
{qe: 0x0009, nmps: 44, nlps: 41, switchFlag: 0}, |
|
{qe: 0x0005, nmps: 45, nlps: 42, switchFlag: 0}, |
|
{qe: 0x0001, nmps: 45, nlps: 43, switchFlag: 0}, |
|
{qe: 0x5601, nmps: 46, nlps: 46, switchFlag: 0} |
|
]; |
|
|
|
// C.3.5 Initialisation of the decoder (INITDEC) |
|
function ArithmeticDecoder(data, start, end) { |
|
this.data = data; |
|
this.bp = start; |
|
this.dataEnd = end; |
|
|
|
this.chigh = data[start]; |
|
this.clow = 0; |
|
|
|
this.byteIn(); |
|
|
|
this.chigh = ((this.chigh << 7) & 0xFFFF) | ((this.clow >> 9) & 0x7F); |
|
this.clow = (this.clow << 7) & 0xFFFF; |
|
this.ct -= 7; |
|
this.a = 0x8000; |
|
} |
|
|
|
ArithmeticDecoder.prototype = { |
|
// C.3.4 Compressed data input (BYTEIN) |
|
byteIn: function ArithmeticDecoder_byteIn() { |
|
var data = this.data; |
|
var bp = this.bp; |
|
if (data[bp] === 0xFF) { |
|
var b1 = data[bp + 1]; |
|
if (b1 > 0x8F) { |
|
this.clow += 0xFF00; |
|
this.ct = 8; |
|
} else { |
|
bp++; |
|
this.clow += (data[bp] << 9); |
|
this.ct = 7; |
|
this.bp = bp; |
|
} |
|
} else { |
|
bp++; |
|
this.clow += bp < this.dataEnd ? (data[bp] << 8) : 0xFF00; |
|
this.ct = 8; |
|
this.bp = bp; |
|
} |
|
if (this.clow > 0xFFFF) { |
|
this.chigh += (this.clow >> 16); |
|
this.clow &= 0xFFFF; |
|
} |
|
}, |
|
// C.3.2 Decoding a decision (DECODE) |
|
readBit: function ArithmeticDecoder_readBit(contexts, pos) { |
|
// contexts are packed into 1 byte: |
|
// highest 7 bits carry cx.index, lowest bit carries cx.mps |
|
var cx_index = contexts[pos] >> 1, cx_mps = contexts[pos] & 1; |
|
var qeTableIcx = QeTable[cx_index]; |
|
var qeIcx = qeTableIcx.qe; |
|
var d; |
|
var a = this.a - qeIcx; |
|
|
|
if (this.chigh < qeIcx) { |
|
// exchangeLps |
|
if (a < qeIcx) { |
|
a = qeIcx; |
|
d = cx_mps; |
|
cx_index = qeTableIcx.nmps; |
|
} else { |
|
a = qeIcx; |
|
d = 1 ^ cx_mps; |
|
if (qeTableIcx.switchFlag === 1) { |
|
cx_mps = d; |
|
} |
|
cx_index = qeTableIcx.nlps; |
|
} |
|
} else { |
|
this.chigh -= qeIcx; |
|
if ((a & 0x8000) !== 0) { |
|
this.a = a; |
|
return cx_mps; |
|
} |
|
// exchangeMps |
|
if (a < qeIcx) { |
|
d = 1 ^ cx_mps; |
|
if (qeTableIcx.switchFlag === 1) { |
|
cx_mps = d; |
|
} |
|
cx_index = qeTableIcx.nlps; |
|
} else { |
|
d = cx_mps; |
|
cx_index = qeTableIcx.nmps; |
|
} |
|
} |
|
// C.3.3 renormD; |
|
do { |
|
if (this.ct === 0) { |
|
this.byteIn(); |
|
} |
|
|
|
a <<= 1; |
|
this.chigh = ((this.chigh << 1) & 0xFFFF) | ((this.clow >> 15) & 1); |
|
this.clow = (this.clow << 1) & 0xFFFF; |
|
this.ct--; |
|
} while ((a & 0x8000) === 0); |
|
this.a = a; |
|
|
|
contexts[pos] = cx_index << 1 | cx_mps; |
|
return d; |
|
} |
|
}; |
|
|
|
return ArithmeticDecoder; |
|
})(); |
|
|
|
exports.ArithmeticDecoder = ArithmeticDecoder; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreCharsets = {})); |
|
} |
|
}(this, function (exports) { |
|
|
|
var ISOAdobeCharset = [ |
|
'.notdef', 'space', 'exclam', 'quotedbl', 'numbersign', 'dollar', |
|
'percent', 'ampersand', 'quoteright', 'parenleft', 'parenright', |
|
'asterisk', 'plus', 'comma', 'hyphen', 'period', 'slash', 'zero', |
|
'one', 'two', 'three', 'four', 'five', 'six', 'seven', 'eight', |
|
'nine', 'colon', 'semicolon', 'less', 'equal', 'greater', 'question', |
|
'at', 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', |
|
'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', |
|
'bracketleft', 'backslash', 'bracketright', 'asciicircum', 'underscore', |
|
'quoteleft', 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', |
|
'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', |
|
'braceleft', 'bar', 'braceright', 'asciitilde', 'exclamdown', 'cent', |
|
'sterling', 'fraction', 'yen', 'florin', 'section', 'currency', |
|
'quotesingle', 'quotedblleft', 'guillemotleft', 'guilsinglleft', |
|
'guilsinglright', 'fi', 'fl', 'endash', 'dagger', 'daggerdbl', |
|
'periodcentered', 'paragraph', 'bullet', 'quotesinglbase', |
|
'quotedblbase', 'quotedblright', 'guillemotright', 'ellipsis', |
|
'perthousand', 'questiondown', 'grave', 'acute', 'circumflex', 'tilde', |
|
'macron', 'breve', 'dotaccent', 'dieresis', 'ring', 'cedilla', |
|
'hungarumlaut', 'ogonek', 'caron', 'emdash', 'AE', 'ordfeminine', |
|
'Lslash', 'Oslash', 'OE', 'ordmasculine', 'ae', 'dotlessi', 'lslash', |
|
'oslash', 'oe', 'germandbls', 'onesuperior', 'logicalnot', 'mu', |
|
'trademark', 'Eth', 'onehalf', 'plusminus', 'Thorn', 'onequarter', |
|
'divide', 'brokenbar', 'degree', 'thorn', 'threequarters', 'twosuperior', |
|
'registered', 'minus', 'eth', 'multiply', 'threesuperior', 'copyright', |
|
'Aacute', 'Acircumflex', 'Adieresis', 'Agrave', 'Aring', 'Atilde', |
|
'Ccedilla', 'Eacute', 'Ecircumflex', 'Edieresis', 'Egrave', 'Iacute', |
|
'Icircumflex', 'Idieresis', 'Igrave', 'Ntilde', 'Oacute', 'Ocircumflex', |
|
'Odieresis', 'Ograve', 'Otilde', 'Scaron', 'Uacute', 'Ucircumflex', |
|
'Udieresis', 'Ugrave', 'Yacute', 'Ydieresis', 'Zcaron', 'aacute', |
|
'acircumflex', 'adieresis', 'agrave', 'aring', 'atilde', 'ccedilla', |
|
'eacute', 'ecircumflex', 'edieresis', 'egrave', 'iacute', 'icircumflex', |
|
'idieresis', 'igrave', 'ntilde', 'oacute', 'ocircumflex', 'odieresis', |
|
'ograve', 'otilde', 'scaron', 'uacute', 'ucircumflex', 'udieresis', |
|
'ugrave', 'yacute', 'ydieresis', 'zcaron' |
|
]; |
|
|
|
var ExpertCharset = [ |
|
'.notdef', 'space', 'exclamsmall', 'Hungarumlautsmall', 'dollaroldstyle', |
|
'dollarsuperior', 'ampersandsmall', 'Acutesmall', 'parenleftsuperior', |
|
'parenrightsuperior', 'twodotenleader', 'onedotenleader', 'comma', |
|
'hyphen', 'period', 'fraction', 'zerooldstyle', 'oneoldstyle', |
|
'twooldstyle', 'threeoldstyle', 'fouroldstyle', 'fiveoldstyle', |
|
'sixoldstyle', 'sevenoldstyle', 'eightoldstyle', 'nineoldstyle', |
|
'colon', 'semicolon', 'commasuperior', 'threequartersemdash', |
|
'periodsuperior', 'questionsmall', 'asuperior', 'bsuperior', |
|
'centsuperior', 'dsuperior', 'esuperior', 'isuperior', 'lsuperior', |
|
'msuperior', 'nsuperior', 'osuperior', 'rsuperior', 'ssuperior', |
|
'tsuperior', 'ff', 'fi', 'fl', 'ffi', 'ffl', 'parenleftinferior', |
|
'parenrightinferior', 'Circumflexsmall', 'hyphensuperior', 'Gravesmall', |
|
'Asmall', 'Bsmall', 'Csmall', 'Dsmall', 'Esmall', 'Fsmall', 'Gsmall', |
|
'Hsmall', 'Ismall', 'Jsmall', 'Ksmall', 'Lsmall', 'Msmall', 'Nsmall', |
|
'Osmall', 'Psmall', 'Qsmall', 'Rsmall', 'Ssmall', 'Tsmall', 'Usmall', |
|
'Vsmall', 'Wsmall', 'Xsmall', 'Ysmall', 'Zsmall', 'colonmonetary', |
|
'onefitted', 'rupiah', 'Tildesmall', 'exclamdownsmall', 'centoldstyle', |
|
'Lslashsmall', 'Scaronsmall', 'Zcaronsmall', 'Dieresissmall', |
|
'Brevesmall', 'Caronsmall', 'Dotaccentsmall', 'Macronsmall', |
|
'figuredash', 'hypheninferior', 'Ogoneksmall', 'Ringsmall', |
|
'Cedillasmall', 'onequarter', 'onehalf', 'threequarters', |
|
'questiondownsmall', 'oneeighth', 'threeeighths', 'fiveeighths', |
|
'seveneighths', 'onethird', 'twothirds', 'zerosuperior', 'onesuperior', |
|
'twosuperior', 'threesuperior', 'foursuperior', 'fivesuperior', |
|
'sixsuperior', 'sevensuperior', 'eightsuperior', 'ninesuperior', |
|
'zeroinferior', 'oneinferior', 'twoinferior', 'threeinferior', |
|
'fourinferior', 'fiveinferior', 'sixinferior', 'seveninferior', |
|
'eightinferior', 'nineinferior', 'centinferior', 'dollarinferior', |
|
'periodinferior', 'commainferior', 'Agravesmall', 'Aacutesmall', |
|
'Acircumflexsmall', 'Atildesmall', 'Adieresissmall', 'Aringsmall', |
|
'AEsmall', 'Ccedillasmall', 'Egravesmall', 'Eacutesmall', |
|
'Ecircumflexsmall', 'Edieresissmall', 'Igravesmall', 'Iacutesmall', |
|
'Icircumflexsmall', 'Idieresissmall', 'Ethsmall', 'Ntildesmall', |
|
'Ogravesmall', 'Oacutesmall', 'Ocircumflexsmall', 'Otildesmall', |
|
'Odieresissmall', 'OEsmall', 'Oslashsmall', 'Ugravesmall', 'Uacutesmall', |
|
'Ucircumflexsmall', 'Udieresissmall', 'Yacutesmall', 'Thornsmall', |
|
'Ydieresissmall' |
|
]; |
|
|
|
var ExpertSubsetCharset = [ |
|
'.notdef', 'space', 'dollaroldstyle', 'dollarsuperior', |
|
'parenleftsuperior', 'parenrightsuperior', 'twodotenleader', |
|
'onedotenleader', 'comma', 'hyphen', 'period', 'fraction', |
|
'zerooldstyle', 'oneoldstyle', 'twooldstyle', 'threeoldstyle', |
|
'fouroldstyle', 'fiveoldstyle', 'sixoldstyle', 'sevenoldstyle', |
|
'eightoldstyle', 'nineoldstyle', 'colon', 'semicolon', 'commasuperior', |
|
'threequartersemdash', 'periodsuperior', 'asuperior', 'bsuperior', |
|
'centsuperior', 'dsuperior', 'esuperior', 'isuperior', 'lsuperior', |
|
'msuperior', 'nsuperior', 'osuperior', 'rsuperior', 'ssuperior', |
|
'tsuperior', 'ff', 'fi', 'fl', 'ffi', 'ffl', 'parenleftinferior', |
|
'parenrightinferior', 'hyphensuperior', 'colonmonetary', 'onefitted', |
|
'rupiah', 'centoldstyle', 'figuredash', 'hypheninferior', 'onequarter', |
|
'onehalf', 'threequarters', 'oneeighth', 'threeeighths', 'fiveeighths', |
|
'seveneighths', 'onethird', 'twothirds', 'zerosuperior', 'onesuperior', |
|
'twosuperior', 'threesuperior', 'foursuperior', 'fivesuperior', |
|
'sixsuperior', 'sevensuperior', 'eightsuperior', 'ninesuperior', |
|
'zeroinferior', 'oneinferior', 'twoinferior', 'threeinferior', |
|
'fourinferior', 'fiveinferior', 'sixinferior', 'seveninferior', |
|
'eightinferior', 'nineinferior', 'centinferior', 'dollarinferior', |
|
'periodinferior', 'commainferior' |
|
]; |
|
|
|
exports.ISOAdobeCharset = ISOAdobeCharset; |
|
exports.ExpertCharset = ExpertCharset; |
|
exports.ExpertSubsetCharset = ExpertSubsetCharset; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreGlyphList = {})); |
|
} |
|
}(this, function (exports) { |
|
|
|
var GlyphsUnicode = { |
|
A: 0x0041, |
|
AE: 0x00C6, |
|
AEacute: 0x01FC, |
|
AEmacron: 0x01E2, |
|
AEsmall: 0xF7E6, |
|
Aacute: 0x00C1, |
|
Aacutesmall: 0xF7E1, |
|
Abreve: 0x0102, |
|
Abreveacute: 0x1EAE, |
|
Abrevecyrillic: 0x04D0, |
|
Abrevedotbelow: 0x1EB6, |
|
Abrevegrave: 0x1EB0, |
|
Abrevehookabove: 0x1EB2, |
|
Abrevetilde: 0x1EB4, |
|
Acaron: 0x01CD, |
|
Acircle: 0x24B6, |
|
Acircumflex: 0x00C2, |
|
Acircumflexacute: 0x1EA4, |
|
Acircumflexdotbelow: 0x1EAC, |
|
Acircumflexgrave: 0x1EA6, |
|
Acircumflexhookabove: 0x1EA8, |
|
Acircumflexsmall: 0xF7E2, |
|
Acircumflextilde: 0x1EAA, |
|
Acute: 0xF6C9, |
|
Acutesmall: 0xF7B4, |
|
Acyrillic: 0x0410, |
|
Adblgrave: 0x0200, |
|
Adieresis: 0x00C4, |
|
Adieresiscyrillic: 0x04D2, |
|
Adieresismacron: 0x01DE, |
|
Adieresissmall: 0xF7E4, |
|
Adotbelow: 0x1EA0, |
|
Adotmacron: 0x01E0, |
|
Agrave: 0x00C0, |
|
Agravesmall: 0xF7E0, |
|
Ahookabove: 0x1EA2, |
|
Aiecyrillic: 0x04D4, |
|
Ainvertedbreve: 0x0202, |
|
Alpha: 0x0391, |
|
Alphatonos: 0x0386, |
|
Amacron: 0x0100, |
|
Amonospace: 0xFF21, |
|
Aogonek: 0x0104, |
|
Aring: 0x00C5, |
|
Aringacute: 0x01FA, |
|
Aringbelow: 0x1E00, |
|
Aringsmall: 0xF7E5, |
|
Asmall: 0xF761, |
|
Atilde: 0x00C3, |
|
Atildesmall: 0xF7E3, |
|
Aybarmenian: 0x0531, |
|
B: 0x0042, |
|
Bcircle: 0x24B7, |
|
Bdotaccent: 0x1E02, |
|
Bdotbelow: 0x1E04, |
|
Becyrillic: 0x0411, |
|
Benarmenian: 0x0532, |
|
Beta: 0x0392, |
|
Bhook: 0x0181, |
|
Blinebelow: 0x1E06, |
|
Bmonospace: 0xFF22, |
|
Brevesmall: 0xF6F4, |
|
Bsmall: 0xF762, |
|
Btopbar: 0x0182, |
|
C: 0x0043, |
|
Caarmenian: 0x053E, |
|
Cacute: 0x0106, |
|
Caron: 0xF6CA, |
|
Caronsmall: 0xF6F5, |
|
Ccaron: 0x010C, |
|
Ccedilla: 0x00C7, |
|
Ccedillaacute: 0x1E08, |
|
Ccedillasmall: 0xF7E7, |
|
Ccircle: 0x24B8, |
|
Ccircumflex: 0x0108, |
|
Cdot: 0x010A, |
|
Cdotaccent: 0x010A, |
|
Cedillasmall: 0xF7B8, |
|
Chaarmenian: 0x0549, |
|
Cheabkhasiancyrillic: 0x04BC, |
|
Checyrillic: 0x0427, |
|
Chedescenderabkhasiancyrillic: 0x04BE, |
|
Chedescendercyrillic: 0x04B6, |
|
Chedieresiscyrillic: 0x04F4, |
|
Cheharmenian: 0x0543, |
|
Chekhakassiancyrillic: 0x04CB, |
|
Cheverticalstrokecyrillic: 0x04B8, |
|
Chi: 0x03A7, |
|
Chook: 0x0187, |
|
Circumflexsmall: 0xF6F6, |
|
Cmonospace: 0xFF23, |
|
Coarmenian: 0x0551, |
|
Csmall: 0xF763, |
|
D: 0x0044, |
|
DZ: 0x01F1, |
|
DZcaron: 0x01C4, |
|
Daarmenian: 0x0534, |
|
Dafrican: 0x0189, |
|
Dcaron: 0x010E, |
|
Dcedilla: 0x1E10, |
|
Dcircle: 0x24B9, |
|
Dcircumflexbelow: 0x1E12, |
|
Dcroat: 0x0110, |
|
Ddotaccent: 0x1E0A, |
|
Ddotbelow: 0x1E0C, |
|
Decyrillic: 0x0414, |
|
Deicoptic: 0x03EE, |
|
Delta: 0x2206, |
|
Deltagreek: 0x0394, |
|
Dhook: 0x018A, |
|
Dieresis: 0xF6CB, |
|
DieresisAcute: 0xF6CC, |
|
DieresisGrave: 0xF6CD, |
|
Dieresissmall: 0xF7A8, |
|
Digammagreek: 0x03DC, |
|
Djecyrillic: 0x0402, |
|
Dlinebelow: 0x1E0E, |
|
Dmonospace: 0xFF24, |
|
Dotaccentsmall: 0xF6F7, |
|
Dslash: 0x0110, |
|
Dsmall: 0xF764, |
|
Dtopbar: 0x018B, |
|
Dz: 0x01F2, |
|
Dzcaron: 0x01C5, |
|
Dzeabkhasiancyrillic: 0x04E0, |
|
Dzecyrillic: 0x0405, |
|
Dzhecyrillic: 0x040F, |
|
E: 0x0045, |
|
Eacute: 0x00C9, |
|
Eacutesmall: 0xF7E9, |
|
Ebreve: 0x0114, |
|
Ecaron: 0x011A, |
|
Ecedillabreve: 0x1E1C, |
|
Echarmenian: 0x0535, |
|
Ecircle: 0x24BA, |
|
Ecircumflex: 0x00CA, |
|
Ecircumflexacute: 0x1EBE, |
|
Ecircumflexbelow: 0x1E18, |
|
Ecircumflexdotbelow: 0x1EC6, |
|
Ecircumflexgrave: 0x1EC0, |
|
Ecircumflexhookabove: 0x1EC2, |
|
Ecircumflexsmall: 0xF7EA, |
|
Ecircumflextilde: 0x1EC4, |
|
Ecyrillic: 0x0404, |
|
Edblgrave: 0x0204, |
|
Edieresis: 0x00CB, |
|
Edieresissmall: 0xF7EB, |
|
Edot: 0x0116, |
|
Edotaccent: 0x0116, |
|
Edotbelow: 0x1EB8, |
|
Efcyrillic: 0x0424, |
|
Egrave: 0x00C8, |
|
Egravesmall: 0xF7E8, |
|
Eharmenian: 0x0537, |
|
Ehookabove: 0x1EBA, |
|
Eightroman: 0x2167, |
|
Einvertedbreve: 0x0206, |
|
Eiotifiedcyrillic: 0x0464, |
|
Elcyrillic: 0x041B, |
|
Elevenroman: 0x216A, |
|
Emacron: 0x0112, |
|
Emacronacute: 0x1E16, |
|
Emacrongrave: 0x1E14, |
|
Emcyrillic: 0x041C, |
|
Emonospace: 0xFF25, |
|
Encyrillic: 0x041D, |
|
Endescendercyrillic: 0x04A2, |
|
Eng: 0x014A, |
|
Enghecyrillic: 0x04A4, |
|
Enhookcyrillic: 0x04C7, |
|
Eogonek: 0x0118, |
|
Eopen: 0x0190, |
|
Epsilon: 0x0395, |
|
Epsilontonos: 0x0388, |
|
Ercyrillic: 0x0420, |
|
Ereversed: 0x018E, |
|
Ereversedcyrillic: 0x042D, |
|
Escyrillic: 0x0421, |
|
Esdescendercyrillic: 0x04AA, |
|
Esh: 0x01A9, |
|
Esmall: 0xF765, |
|
Eta: 0x0397, |
|
Etarmenian: 0x0538, |
|
Etatonos: 0x0389, |
|
Eth: 0x00D0, |
|
Ethsmall: 0xF7F0, |
|
Etilde: 0x1EBC, |
|
Etildebelow: 0x1E1A, |
|
Euro: 0x20AC, |
|
Ezh: 0x01B7, |
|
Ezhcaron: 0x01EE, |
|
Ezhreversed: 0x01B8, |
|
F: 0x0046, |
|
Fcircle: 0x24BB, |
|
Fdotaccent: 0x1E1E, |
|
Feharmenian: 0x0556, |
|
Feicoptic: 0x03E4, |
|
Fhook: 0x0191, |
|
Fitacyrillic: 0x0472, |
|
Fiveroman: 0x2164, |
|
Fmonospace: 0xFF26, |
|
Fourroman: 0x2163, |
|
Fsmall: 0xF766, |
|
G: 0x0047, |
|
GBsquare: 0x3387, |
|
Gacute: 0x01F4, |
|
Gamma: 0x0393, |
|
Gammaafrican: 0x0194, |
|
Gangiacoptic: 0x03EA, |
|
Gbreve: 0x011E, |
|
Gcaron: 0x01E6, |
|
Gcedilla: 0x0122, |
|
Gcircle: 0x24BC, |
|
Gcircumflex: 0x011C, |
|
Gcommaaccent: 0x0122, |
|
Gdot: 0x0120, |
|
Gdotaccent: 0x0120, |
|
Gecyrillic: 0x0413, |
|
Ghadarmenian: 0x0542, |
|
Ghemiddlehookcyrillic: 0x0494, |
|
Ghestrokecyrillic: 0x0492, |
|
Gheupturncyrillic: 0x0490, |
|
Ghook: 0x0193, |
|
Gimarmenian: 0x0533, |
|
Gjecyrillic: 0x0403, |
|
Gmacron: 0x1E20, |
|
Gmonospace: 0xFF27, |
|
Grave: 0xF6CE, |
|
Gravesmall: 0xF760, |
|
Gsmall: 0xF767, |
|
Gsmallhook: 0x029B, |
|
Gstroke: 0x01E4, |
|
H: 0x0048, |
|
H18533: 0x25CF, |
|
H18543: 0x25AA, |
|
H18551: 0x25AB, |
|
H22073: 0x25A1, |
|
HPsquare: 0x33CB, |
|
Haabkhasiancyrillic: 0x04A8, |
|
Hadescendercyrillic: 0x04B2, |
|
Hardsigncyrillic: 0x042A, |
|
Hbar: 0x0126, |
|
Hbrevebelow: 0x1E2A, |
|
Hcedilla: 0x1E28, |
|
Hcircle: 0x24BD, |
|
Hcircumflex: 0x0124, |
|
Hdieresis: 0x1E26, |
|
Hdotaccent: 0x1E22, |
|
Hdotbelow: 0x1E24, |
|
Hmonospace: 0xFF28, |
|
Hoarmenian: 0x0540, |
|
Horicoptic: 0x03E8, |
|
Hsmall: 0xF768, |
|
Hungarumlaut: 0xF6CF, |
|
Hungarumlautsmall: 0xF6F8, |
|
Hzsquare: 0x3390, |
|
I: 0x0049, |
|
IAcyrillic: 0x042F, |
|
IJ: 0x0132, |
|
IUcyrillic: 0x042E, |
|
Iacute: 0x00CD, |
|
Iacutesmall: 0xF7ED, |
|
Ibreve: 0x012C, |
|
Icaron: 0x01CF, |
|
Icircle: 0x24BE, |
|
Icircumflex: 0x00CE, |
|
Icircumflexsmall: 0xF7EE, |
|
Icyrillic: 0x0406, |
|
Idblgrave: 0x0208, |
|
Idieresis: 0x00CF, |
|
Idieresisacute: 0x1E2E, |
|
Idieresiscyrillic: 0x04E4, |
|
Idieresissmall: 0xF7EF, |
|
Idot: 0x0130, |
|
Idotaccent: 0x0130, |
|
Idotbelow: 0x1ECA, |
|
Iebrevecyrillic: 0x04D6, |
|
Iecyrillic: 0x0415, |
|
Ifraktur: 0x2111, |
|
Igrave: 0x00CC, |
|
Igravesmall: 0xF7EC, |
|
Ihookabove: 0x1EC8, |
|
Iicyrillic: 0x0418, |
|
Iinvertedbreve: 0x020A, |
|
Iishortcyrillic: 0x0419, |
|
Imacron: 0x012A, |
|
Imacroncyrillic: 0x04E2, |
|
Imonospace: 0xFF29, |
|
Iniarmenian: 0x053B, |
|
Iocyrillic: 0x0401, |
|
Iogonek: 0x012E, |
|
Iota: 0x0399, |
|
Iotaafrican: 0x0196, |
|
Iotadieresis: 0x03AA, |
|
Iotatonos: 0x038A, |
|
Ismall: 0xF769, |
|
Istroke: 0x0197, |
|
Itilde: 0x0128, |
|
Itildebelow: 0x1E2C, |
|
Izhitsacyrillic: 0x0474, |
|
Izhitsadblgravecyrillic: 0x0476, |
|
J: 0x004A, |
|
Jaarmenian: 0x0541, |
|
Jcircle: 0x24BF, |
|
Jcircumflex: 0x0134, |
|
Jecyrillic: 0x0408, |
|
Jheharmenian: 0x054B, |
|
Jmonospace: 0xFF2A, |
|
Jsmall: 0xF76A, |
|
K: 0x004B, |
|
KBsquare: 0x3385, |
|
KKsquare: 0x33CD, |
|
Kabashkircyrillic: 0x04A0, |
|
Kacute: 0x1E30, |
|
Kacyrillic: 0x041A, |
|
Kadescendercyrillic: 0x049A, |
|
Kahookcyrillic: 0x04C3, |
|
Kappa: 0x039A, |
|
Kastrokecyrillic: 0x049E, |
|
Kaverticalstrokecyrillic: 0x049C, |
|
Kcaron: 0x01E8, |
|
Kcedilla: 0x0136, |
|
Kcircle: 0x24C0, |
|
Kcommaaccent: 0x0136, |
|
Kdotbelow: 0x1E32, |
|
Keharmenian: 0x0554, |
|
Kenarmenian: 0x053F, |
|
Khacyrillic: 0x0425, |
|
Kheicoptic: 0x03E6, |
|
Khook: 0x0198, |
|
Kjecyrillic: 0x040C, |
|
Klinebelow: 0x1E34, |
|
Kmonospace: 0xFF2B, |
|
Koppacyrillic: 0x0480, |
|
Koppagreek: 0x03DE, |
|
Ksicyrillic: 0x046E, |
|
Ksmall: 0xF76B, |
|
L: 0x004C, |
|
LJ: 0x01C7, |
|
LL: 0xF6BF, |
|
Lacute: 0x0139, |
|
Lambda: 0x039B, |
|
Lcaron: 0x013D, |
|
Lcedilla: 0x013B, |
|
Lcircle: 0x24C1, |
|
Lcircumflexbelow: 0x1E3C, |
|
Lcommaaccent: 0x013B, |
|
Ldot: 0x013F, |
|
Ldotaccent: 0x013F, |
|
Ldotbelow: 0x1E36, |
|
Ldotbelowmacron: 0x1E38, |
|
Liwnarmenian: 0x053C, |
|
Lj: 0x01C8, |
|
Ljecyrillic: 0x0409, |
|
Llinebelow: 0x1E3A, |
|
Lmonospace: 0xFF2C, |
|
Lslash: 0x0141, |
|
Lslashsmall: 0xF6F9, |
|
Lsmall: 0xF76C, |
|
M: 0x004D, |
|
MBsquare: 0x3386, |
|
Macron: 0xF6D0, |
|
Macronsmall: 0xF7AF, |
|
Macute: 0x1E3E, |
|
Mcircle: 0x24C2, |
|
Mdotaccent: 0x1E40, |
|
Mdotbelow: 0x1E42, |
|
Menarmenian: 0x0544, |
|
Mmonospace: 0xFF2D, |
|
Msmall: 0xF76D, |
|
Mturned: 0x019C, |
|
Mu: 0x039C, |
|
N: 0x004E, |
|
NJ: 0x01CA, |
|
Nacute: 0x0143, |
|
Ncaron: 0x0147, |
|
Ncedilla: 0x0145, |
|
Ncircle: 0x24C3, |
|
Ncircumflexbelow: 0x1E4A, |
|
Ncommaaccent: 0x0145, |
|
Ndotaccent: 0x1E44, |
|
Ndotbelow: 0x1E46, |
|
Nhookleft: 0x019D, |
|
Nineroman: 0x2168, |
|
Nj: 0x01CB, |
|
Njecyrillic: 0x040A, |
|
Nlinebelow: 0x1E48, |
|
Nmonospace: 0xFF2E, |
|
Nowarmenian: 0x0546, |
|
Nsmall: 0xF76E, |
|
Ntilde: 0x00D1, |
|
Ntildesmall: 0xF7F1, |
|
Nu: 0x039D, |
|
O: 0x004F, |
|
OE: 0x0152, |
|
OEsmall: 0xF6FA, |
|
Oacute: 0x00D3, |
|
Oacutesmall: 0xF7F3, |
|
Obarredcyrillic: 0x04E8, |
|
Obarreddieresiscyrillic: 0x04EA, |
|
Obreve: 0x014E, |
|
Ocaron: 0x01D1, |
|
Ocenteredtilde: 0x019F, |
|
Ocircle: 0x24C4, |
|
Ocircumflex: 0x00D4, |
|
Ocircumflexacute: 0x1ED0, |
|
Ocircumflexdotbelow: 0x1ED8, |
|
Ocircumflexgrave: 0x1ED2, |
|
Ocircumflexhookabove: 0x1ED4, |
|
Ocircumflexsmall: 0xF7F4, |
|
Ocircumflextilde: 0x1ED6, |
|
Ocyrillic: 0x041E, |
|
Odblacute: 0x0150, |
|
Odblgrave: 0x020C, |
|
Odieresis: 0x00D6, |
|
Odieresiscyrillic: 0x04E6, |
|
Odieresissmall: 0xF7F6, |
|
Odotbelow: 0x1ECC, |
|
Ogoneksmall: 0xF6FB, |
|
Ograve: 0x00D2, |
|
Ogravesmall: 0xF7F2, |
|
Oharmenian: 0x0555, |
|
Ohm: 0x2126, |
|
Ohookabove: 0x1ECE, |
|
Ohorn: 0x01A0, |
|
Ohornacute: 0x1EDA, |
|
Ohorndotbelow: 0x1EE2, |
|
Ohorngrave: 0x1EDC, |
|
Ohornhookabove: 0x1EDE, |
|
Ohorntilde: 0x1EE0, |
|
Ohungarumlaut: 0x0150, |
|
Oi: 0x01A2, |
|
Oinvertedbreve: 0x020E, |
|
Omacron: 0x014C, |
|
Omacronacute: 0x1E52, |
|
Omacrongrave: 0x1E50, |
|
Omega: 0x2126, |
|
Omegacyrillic: 0x0460, |
|
Omegagreek: 0x03A9, |
|
Omegaroundcyrillic: 0x047A, |
|
Omegatitlocyrillic: 0x047C, |
|
Omegatonos: 0x038F, |
|
Omicron: 0x039F, |
|
Omicrontonos: 0x038C, |
|
Omonospace: 0xFF2F, |
|
Oneroman: 0x2160, |
|
Oogonek: 0x01EA, |
|
Oogonekmacron: 0x01EC, |
|
Oopen: 0x0186, |
|
Oslash: 0x00D8, |
|
Oslashacute: 0x01FE, |
|
Oslashsmall: 0xF7F8, |
|
Osmall: 0xF76F, |
|
Ostrokeacute: 0x01FE, |
|
Otcyrillic: 0x047E, |
|
Otilde: 0x00D5, |
|
Otildeacute: 0x1E4C, |
|
Otildedieresis: 0x1E4E, |
|
Otildesmall: 0xF7F5, |
|
P: 0x0050, |
|
Pacute: 0x1E54, |
|
Pcircle: 0x24C5, |
|
Pdotaccent: 0x1E56, |
|
Pecyrillic: 0x041F, |
|
Peharmenian: 0x054A, |
|
Pemiddlehookcyrillic: 0x04A6, |
|
Phi: 0x03A6, |
|
Phook: 0x01A4, |
|
Pi: 0x03A0, |
|
Piwrarmenian: 0x0553, |
|
Pmonospace: 0xFF30, |
|
Psi: 0x03A8, |
|
Psicyrillic: 0x0470, |
|
Psmall: 0xF770, |
|
Q: 0x0051, |
|
Qcircle: 0x24C6, |
|
Qmonospace: 0xFF31, |
|
Qsmall: 0xF771, |
|
R: 0x0052, |
|
Raarmenian: 0x054C, |
|
Racute: 0x0154, |
|
Rcaron: 0x0158, |
|
Rcedilla: 0x0156, |
|
Rcircle: 0x24C7, |
|
Rcommaaccent: 0x0156, |
|
Rdblgrave: 0x0210, |
|
Rdotaccent: 0x1E58, |
|
Rdotbelow: 0x1E5A, |
|
Rdotbelowmacron: 0x1E5C, |
|
Reharmenian: 0x0550, |
|
Rfraktur: 0x211C, |
|
Rho: 0x03A1, |
|
Ringsmall: 0xF6FC, |
|
Rinvertedbreve: 0x0212, |
|
Rlinebelow: 0x1E5E, |
|
Rmonospace: 0xFF32, |
|
Rsmall: 0xF772, |
|
Rsmallinverted: 0x0281, |
|
Rsmallinvertedsuperior: 0x02B6, |
|
S: 0x0053, |
|
SF010000: 0x250C, |
|
SF020000: 0x2514, |
|
SF030000: 0x2510, |
|
SF040000: 0x2518, |
|
SF050000: 0x253C, |
|
SF060000: 0x252C, |
|
SF070000: 0x2534, |
|
SF080000: 0x251C, |
|
SF090000: 0x2524, |
|
SF100000: 0x2500, |
|
SF110000: 0x2502, |
|
SF190000: 0x2561, |
|
SF200000: 0x2562, |
|
SF210000: 0x2556, |
|
SF220000: 0x2555, |
|
SF230000: 0x2563, |
|
SF240000: 0x2551, |
|
SF250000: 0x2557, |
|
SF260000: 0x255D, |
|
SF270000: 0x255C, |
|
SF280000: 0x255B, |
|
SF360000: 0x255E, |
|
SF370000: 0x255F, |
|
SF380000: 0x255A, |
|
SF390000: 0x2554, |
|
SF400000: 0x2569, |
|
SF410000: 0x2566, |
|
SF420000: 0x2560, |
|
SF430000: 0x2550, |
|
SF440000: 0x256C, |
|
SF450000: 0x2567, |
|
SF460000: 0x2568, |
|
SF470000: 0x2564, |
|
SF480000: 0x2565, |
|
SF490000: 0x2559, |
|
SF500000: 0x2558, |
|
SF510000: 0x2552, |
|
SF520000: 0x2553, |
|
SF530000: 0x256B, |
|
SF540000: 0x256A, |
|
Sacute: 0x015A, |
|
Sacutedotaccent: 0x1E64, |
|
Sampigreek: 0x03E0, |
|
Scaron: 0x0160, |
|
Scarondotaccent: 0x1E66, |
|
Scaronsmall: 0xF6FD, |
|
Scedilla: 0x015E, |
|
Schwa: 0x018F, |
|
Schwacyrillic: 0x04D8, |
|
Schwadieresiscyrillic: 0x04DA, |
|
Scircle: 0x24C8, |
|
Scircumflex: 0x015C, |
|
Scommaaccent: 0x0218, |
|
Sdotaccent: 0x1E60, |
|
Sdotbelow: 0x1E62, |
|
Sdotbelowdotaccent: 0x1E68, |
|
Seharmenian: 0x054D, |
|
Sevenroman: 0x2166, |
|
Shaarmenian: 0x0547, |
|
Shacyrillic: 0x0428, |
|
Shchacyrillic: 0x0429, |
|
Sheicoptic: 0x03E2, |
|
Shhacyrillic: 0x04BA, |
|
Shimacoptic: 0x03EC, |
|
Sigma: 0x03A3, |
|
Sixroman: 0x2165, |
|
Smonospace: 0xFF33, |
|
Softsigncyrillic: 0x042C, |
|
Ssmall: 0xF773, |
|
Stigmagreek: 0x03DA, |
|
T: 0x0054, |
|
Tau: 0x03A4, |
|
Tbar: 0x0166, |
|
Tcaron: 0x0164, |
|
Tcedilla: 0x0162, |
|
Tcircle: 0x24C9, |
|
Tcircumflexbelow: 0x1E70, |
|
Tcommaaccent: 0x0162, |
|
Tdotaccent: 0x1E6A, |
|
Tdotbelow: 0x1E6C, |
|
Tecyrillic: 0x0422, |
|
Tedescendercyrillic: 0x04AC, |
|
Tenroman: 0x2169, |
|
Tetsecyrillic: 0x04B4, |
|
Theta: 0x0398, |
|
Thook: 0x01AC, |
|
Thorn: 0x00DE, |
|
Thornsmall: 0xF7FE, |
|
Threeroman: 0x2162, |
|
Tildesmall: 0xF6FE, |
|
Tiwnarmenian: 0x054F, |
|
Tlinebelow: 0x1E6E, |
|
Tmonospace: 0xFF34, |
|
Toarmenian: 0x0539, |
|
Tonefive: 0x01BC, |
|
Tonesix: 0x0184, |
|
Tonetwo: 0x01A7, |
|
Tretroflexhook: 0x01AE, |
|
Tsecyrillic: 0x0426, |
|
Tshecyrillic: 0x040B, |
|
Tsmall: 0xF774, |
|
Twelveroman: 0x216B, |
|
Tworoman: 0x2161, |
|
U: 0x0055, |
|
Uacute: 0x00DA, |
|
Uacutesmall: 0xF7FA, |
|
Ubreve: 0x016C, |
|
Ucaron: 0x01D3, |
|
Ucircle: 0x24CA, |
|
Ucircumflex: 0x00DB, |
|
Ucircumflexbelow: 0x1E76, |
|
Ucircumflexsmall: 0xF7FB, |
|
Ucyrillic: 0x0423, |
|
Udblacute: 0x0170, |
|
Udblgrave: 0x0214, |
|
Udieresis: 0x00DC, |
|
Udieresisacute: 0x01D7, |
|
Udieresisbelow: 0x1E72, |
|
Udieresiscaron: 0x01D9, |
|
Udieresiscyrillic: 0x04F0, |
|
Udieresisgrave: 0x01DB, |
|
Udieresismacron: 0x01D5, |
|
Udieresissmall: 0xF7FC, |
|
Udotbelow: 0x1EE4, |
|
Ugrave: 0x00D9, |
|
Ugravesmall: 0xF7F9, |
|
Uhookabove: 0x1EE6, |
|
Uhorn: 0x01AF, |
|
Uhornacute: 0x1EE8, |
|
Uhorndotbelow: 0x1EF0, |
|
Uhorngrave: 0x1EEA, |
|
Uhornhookabove: 0x1EEC, |
|
Uhorntilde: 0x1EEE, |
|
Uhungarumlaut: 0x0170, |
|
Uhungarumlautcyrillic: 0x04F2, |
|
Uinvertedbreve: 0x0216, |
|
Ukcyrillic: 0x0478, |
|
Umacron: 0x016A, |
|
Umacroncyrillic: 0x04EE, |
|
Umacrondieresis: 0x1E7A, |
|
Umonospace: 0xFF35, |
|
Uogonek: 0x0172, |
|
Upsilon: 0x03A5, |
|
Upsilon1: 0x03D2, |
|
Upsilonacutehooksymbolgreek: 0x03D3, |
|
Upsilonafrican: 0x01B1, |
|
Upsilondieresis: 0x03AB, |
|
Upsilondieresishooksymbolgreek: 0x03D4, |
|
Upsilonhooksymbol: 0x03D2, |
|
Upsilontonos: 0x038E, |
|
Uring: 0x016E, |
|
Ushortcyrillic: 0x040E, |
|
Usmall: 0xF775, |
|
Ustraightcyrillic: 0x04AE, |
|
Ustraightstrokecyrillic: 0x04B0, |
|
Utilde: 0x0168, |
|
Utildeacute: 0x1E78, |
|
Utildebelow: 0x1E74, |
|
V: 0x0056, |
|
Vcircle: 0x24CB, |
|
Vdotbelow: 0x1E7E, |
|
Vecyrillic: 0x0412, |
|
Vewarmenian: 0x054E, |
|
Vhook: 0x01B2, |
|
Vmonospace: 0xFF36, |
|
Voarmenian: 0x0548, |
|
Vsmall: 0xF776, |
|
Vtilde: 0x1E7C, |
|
W: 0x0057, |
|
Wacute: 0x1E82, |
|
Wcircle: 0x24CC, |
|
Wcircumflex: 0x0174, |
|
Wdieresis: 0x1E84, |
|
Wdotaccent: 0x1E86, |
|
Wdotbelow: 0x1E88, |
|
Wgrave: 0x1E80, |
|
Wmonospace: 0xFF37, |
|
Wsmall: 0xF777, |
|
X: 0x0058, |
|
Xcircle: 0x24CD, |
|
Xdieresis: 0x1E8C, |
|
Xdotaccent: 0x1E8A, |
|
Xeharmenian: 0x053D, |
|
Xi: 0x039E, |
|
Xmonospace: 0xFF38, |
|
Xsmall: 0xF778, |
|
Y: 0x0059, |
|
Yacute: 0x00DD, |
|
Yacutesmall: 0xF7FD, |
|
Yatcyrillic: 0x0462, |
|
Ycircle: 0x24CE, |
|
Ycircumflex: 0x0176, |
|
Ydieresis: 0x0178, |
|
Ydieresissmall: 0xF7FF, |
|
Ydotaccent: 0x1E8E, |
|
Ydotbelow: 0x1EF4, |
|
Yericyrillic: 0x042B, |
|
Yerudieresiscyrillic: 0x04F8, |
|
Ygrave: 0x1EF2, |
|
Yhook: 0x01B3, |
|
Yhookabove: 0x1EF6, |
|
Yiarmenian: 0x0545, |
|
Yicyrillic: 0x0407, |
|
Yiwnarmenian: 0x0552, |
|
Ymonospace: 0xFF39, |
|
Ysmall: 0xF779, |
|
Ytilde: 0x1EF8, |
|
Yusbigcyrillic: 0x046A, |
|
Yusbigiotifiedcyrillic: 0x046C, |
|
Yuslittlecyrillic: 0x0466, |
|
Yuslittleiotifiedcyrillic: 0x0468, |
|
Z: 0x005A, |
|
Zaarmenian: 0x0536, |
|
Zacute: 0x0179, |
|
Zcaron: 0x017D, |
|
Zcaronsmall: 0xF6FF, |
|
Zcircle: 0x24CF, |
|
Zcircumflex: 0x1E90, |
|
Zdot: 0x017B, |
|
Zdotaccent: 0x017B, |
|
Zdotbelow: 0x1E92, |
|
Zecyrillic: 0x0417, |
|
Zedescendercyrillic: 0x0498, |
|
Zedieresiscyrillic: 0x04DE, |
|
Zeta: 0x0396, |
|
Zhearmenian: 0x053A, |
|
Zhebrevecyrillic: 0x04C1, |
|
Zhecyrillic: 0x0416, |
|
Zhedescendercyrillic: 0x0496, |
|
Zhedieresiscyrillic: 0x04DC, |
|
Zlinebelow: 0x1E94, |
|
Zmonospace: 0xFF3A, |
|
Zsmall: 0xF77A, |
|
Zstroke: 0x01B5, |
|
a: 0x0061, |
|
aabengali: 0x0986, |
|
aacute: 0x00E1, |
|
aadeva: 0x0906, |
|
aagujarati: 0x0A86, |
|
aagurmukhi: 0x0A06, |
|
aamatragurmukhi: 0x0A3E, |
|
aarusquare: 0x3303, |
|
aavowelsignbengali: 0x09BE, |
|
aavowelsigndeva: 0x093E, |
|
aavowelsigngujarati: 0x0ABE, |
|
abbreviationmarkarmenian: 0x055F, |
|
abbreviationsigndeva: 0x0970, |
|
abengali: 0x0985, |
|
abopomofo: 0x311A, |
|
abreve: 0x0103, |
|
abreveacute: 0x1EAF, |
|
abrevecyrillic: 0x04D1, |
|
abrevedotbelow: 0x1EB7, |
|
abrevegrave: 0x1EB1, |
|
abrevehookabove: 0x1EB3, |
|
abrevetilde: 0x1EB5, |
|
acaron: 0x01CE, |
|
acircle: 0x24D0, |
|
acircumflex: 0x00E2, |
|
acircumflexacute: 0x1EA5, |
|
acircumflexdotbelow: 0x1EAD, |
|
acircumflexgrave: 0x1EA7, |
|
acircumflexhookabove: 0x1EA9, |
|
acircumflextilde: 0x1EAB, |
|
acute: 0x00B4, |
|
acutebelowcmb: 0x0317, |
|
acutecmb: 0x0301, |
|
acutecomb: 0x0301, |
|
acutedeva: 0x0954, |
|
acutelowmod: 0x02CF, |
|
acutetonecmb: 0x0341, |
|
acyrillic: 0x0430, |
|
adblgrave: 0x0201, |
|
addakgurmukhi: 0x0A71, |
|
adeva: 0x0905, |
|
adieresis: 0x00E4, |
|
adieresiscyrillic: 0x04D3, |
|
adieresismacron: 0x01DF, |
|
adotbelow: 0x1EA1, |
|
adotmacron: 0x01E1, |
|
ae: 0x00E6, |
|
aeacute: 0x01FD, |
|
aekorean: 0x3150, |
|
aemacron: 0x01E3, |
|
afii00208: 0x2015, |
|
afii08941: 0x20A4, |
|
afii10017: 0x0410, |
|
afii10018: 0x0411, |
|
afii10019: 0x0412, |
|
afii10020: 0x0413, |
|
afii10021: 0x0414, |
|
afii10022: 0x0415, |
|
afii10023: 0x0401, |
|
afii10024: 0x0416, |
|
afii10025: 0x0417, |
|
afii10026: 0x0418, |
|
afii10027: 0x0419, |
|
afii10028: 0x041A, |
|
afii10029: 0x041B, |
|
afii10030: 0x041C, |
|
afii10031: 0x041D, |
|
afii10032: 0x041E, |
|
afii10033: 0x041F, |
|
afii10034: 0x0420, |
|
afii10035: 0x0421, |
|
afii10036: 0x0422, |
|
afii10037: 0x0423, |
|
afii10038: 0x0424, |
|
afii10039: 0x0425, |
|
afii10040: 0x0426, |
|
afii10041: 0x0427, |
|
afii10042: 0x0428, |
|
afii10043: 0x0429, |
|
afii10044: 0x042A, |
|
afii10045: 0x042B, |
|
afii10046: 0x042C, |
|
afii10047: 0x042D, |
|
afii10048: 0x042E, |
|
afii10049: 0x042F, |
|
afii10050: 0x0490, |
|
afii10051: 0x0402, |
|
afii10052: 0x0403, |
|
afii10053: 0x0404, |
|
afii10054: 0x0405, |
|
afii10055: 0x0406, |
|
afii10056: 0x0407, |
|
afii10057: 0x0408, |
|
afii10058: 0x0409, |
|
afii10059: 0x040A, |
|
afii10060: 0x040B, |
|
afii10061: 0x040C, |
|
afii10062: 0x040E, |
|
afii10063: 0xF6C4, |
|
afii10064: 0xF6C5, |
|
afii10065: 0x0430, |
|
afii10066: 0x0431, |
|
afii10067: 0x0432, |
|
afii10068: 0x0433, |
|
afii10069: 0x0434, |
|
afii10070: 0x0435, |
|
afii10071: 0x0451, |
|
afii10072: 0x0436, |
|
afii10073: 0x0437, |
|
afii10074: 0x0438, |
|
afii10075: 0x0439, |
|
afii10076: 0x043A, |
|
afii10077: 0x043B, |
|
afii10078: 0x043C, |
|
afii10079: 0x043D, |
|
afii10080: 0x043E, |
|
afii10081: 0x043F, |
|
afii10082: 0x0440, |
|
afii10083: 0x0441, |
|
afii10084: 0x0442, |
|
afii10085: 0x0443, |
|
afii10086: 0x0444, |
|
afii10087: 0x0445, |
|
afii10088: 0x0446, |
|
afii10089: 0x0447, |
|
afii10090: 0x0448, |
|
afii10091: 0x0449, |
|
afii10092: 0x044A, |
|
afii10093: 0x044B, |
|
afii10094: 0x044C, |
|
afii10095: 0x044D, |
|
afii10096: 0x044E, |
|
afii10097: 0x044F, |
|
afii10098: 0x0491, |
|
afii10099: 0x0452, |
|
afii10100: 0x0453, |
|
afii10101: 0x0454, |
|
afii10102: 0x0455, |
|
afii10103: 0x0456, |
|
afii10104: 0x0457, |
|
afii10105: 0x0458, |
|
afii10106: 0x0459, |
|
afii10107: 0x045A, |
|
afii10108: 0x045B, |
|
afii10109: 0x045C, |
|
afii10110: 0x045E, |
|
afii10145: 0x040F, |
|
afii10146: 0x0462, |
|
afii10147: 0x0472, |
|
afii10148: 0x0474, |
|
afii10192: 0xF6C6, |
|
afii10193: 0x045F, |
|
afii10194: 0x0463, |
|
afii10195: 0x0473, |
|
afii10196: 0x0475, |
|
afii10831: 0xF6C7, |
|
afii10832: 0xF6C8, |
|
afii10846: 0x04D9, |
|
afii299: 0x200E, |
|
afii300: 0x200F, |
|
afii301: 0x200D, |
|
afii57381: 0x066A, |
|
afii57388: 0x060C, |
|
afii57392: 0x0660, |
|
afii57393: 0x0661, |
|
afii57394: 0x0662, |
|
afii57395: 0x0663, |
|
afii57396: 0x0664, |
|
afii57397: 0x0665, |
|
afii57398: 0x0666, |
|
afii57399: 0x0667, |
|
afii57400: 0x0668, |
|
afii57401: 0x0669, |
|
afii57403: 0x061B, |
|
afii57407: 0x061F, |
|
afii57409: 0x0621, |
|
afii57410: 0x0622, |
|
afii57411: 0x0623, |
|
afii57412: 0x0624, |
|
afii57413: 0x0625, |
|
afii57414: 0x0626, |
|
afii57415: 0x0627, |
|
afii57416: 0x0628, |
|
afii57417: 0x0629, |
|
afii57418: 0x062A, |
|
afii57419: 0x062B, |
|
afii57420: 0x062C, |
|
afii57421: 0x062D, |
|
afii57422: 0x062E, |
|
afii57423: 0x062F, |
|
afii57424: 0x0630, |
|
afii57425: 0x0631, |
|
afii57426: 0x0632, |
|
afii57427: 0x0633, |
|
afii57428: 0x0634, |
|
afii57429: 0x0635, |
|
afii57430: 0x0636, |
|
afii57431: 0x0637, |
|
afii57432: 0x0638, |
|
afii57433: 0x0639, |
|
afii57434: 0x063A, |
|
afii57440: 0x0640, |
|
afii57441: 0x0641, |
|
afii57442: 0x0642, |
|
afii57443: 0x0643, |
|
afii57444: 0x0644, |
|
afii57445: 0x0645, |
|
afii57446: 0x0646, |
|
afii57448: 0x0648, |
|
afii57449: 0x0649, |
|
afii57450: 0x064A, |
|
afii57451: 0x064B, |
|
afii57452: 0x064C, |
|
afii57453: 0x064D, |
|
afii57454: 0x064E, |
|
afii57455: 0x064F, |
|
afii57456: 0x0650, |
|
afii57457: 0x0651, |
|
afii57458: 0x0652, |
|
afii57470: 0x0647, |
|
afii57505: 0x06A4, |
|
afii57506: 0x067E, |
|
afii57507: 0x0686, |
|
afii57508: 0x0698, |
|
afii57509: 0x06AF, |
|
afii57511: 0x0679, |
|
afii57512: 0x0688, |
|
afii57513: 0x0691, |
|
afii57514: 0x06BA, |
|
afii57519: 0x06D2, |
|
afii57534: 0x06D5, |
|
afii57636: 0x20AA, |
|
afii57645: 0x05BE, |
|
afii57658: 0x05C3, |
|
afii57664: 0x05D0, |
|
afii57665: 0x05D1, |
|
afii57666: 0x05D2, |
|
afii57667: 0x05D3, |
|
afii57668: 0x05D4, |
|
afii57669: 0x05D5, |
|
afii57670: 0x05D6, |
|
afii57671: 0x05D7, |
|
afii57672: 0x05D8, |
|
afii57673: 0x05D9, |
|
afii57674: 0x05DA, |
|
afii57675: 0x05DB, |
|
afii57676: 0x05DC, |
|
afii57677: 0x05DD, |
|
afii57678: 0x05DE, |
|
afii57679: 0x05DF, |
|
afii57680: 0x05E0, |
|
afii57681: 0x05E1, |
|
afii57682: 0x05E2, |
|
afii57683: 0x05E3, |
|
afii57684: 0x05E4, |
|
afii57685: 0x05E5, |
|
afii57686: 0x05E6, |
|
afii57687: 0x05E7, |
|
afii57688: 0x05E8, |
|
afii57689: 0x05E9, |
|
afii57690: 0x05EA, |
|
afii57694: 0xFB2A, |
|
afii57695: 0xFB2B, |
|
afii57700: 0xFB4B, |
|
afii57705: 0xFB1F, |
|
afii57716: 0x05F0, |
|
afii57717: 0x05F1, |
|
afii57718: 0x05F2, |
|
afii57723: 0xFB35, |
|
afii57793: 0x05B4, |
|
afii57794: 0x05B5, |
|
afii57795: 0x05B6, |
|
afii57796: 0x05BB, |
|
afii57797: 0x05B8, |
|
afii57798: 0x05B7, |
|
afii57799: 0x05B0, |
|
afii57800: 0x05B2, |
|
afii57801: 0x05B1, |
|
afii57802: 0x05B3, |
|
afii57803: 0x05C2, |
|
afii57804: 0x05C1, |
|
afii57806: 0x05B9, |
|
afii57807: 0x05BC, |
|
afii57839: 0x05BD, |
|
afii57841: 0x05BF, |
|
afii57842: 0x05C0, |
|
afii57929: 0x02BC, |
|
afii61248: 0x2105, |
|
afii61289: 0x2113, |
|
afii61352: 0x2116, |
|
afii61573: 0x202C, |
|
afii61574: 0x202D, |
|
afii61575: 0x202E, |
|
afii61664: 0x200C, |
|
afii63167: 0x066D, |
|
afii64937: 0x02BD, |
|
agrave: 0x00E0, |
|
agujarati: 0x0A85, |
|
agurmukhi: 0x0A05, |
|
ahiragana: 0x3042, |
|
ahookabove: 0x1EA3, |
|
aibengali: 0x0990, |
|
aibopomofo: 0x311E, |
|
aideva: 0x0910, |
|
aiecyrillic: 0x04D5, |
|
aigujarati: 0x0A90, |
|
aigurmukhi: 0x0A10, |
|
aimatragurmukhi: 0x0A48, |
|
ainarabic: 0x0639, |
|
ainfinalarabic: 0xFECA, |
|
aininitialarabic: 0xFECB, |
|
ainmedialarabic: 0xFECC, |
|
ainvertedbreve: 0x0203, |
|
aivowelsignbengali: 0x09C8, |
|
aivowelsigndeva: 0x0948, |
|
aivowelsigngujarati: 0x0AC8, |
|
akatakana: 0x30A2, |
|
akatakanahalfwidth: 0xFF71, |
|
akorean: 0x314F, |
|
alef: 0x05D0, |
|
alefarabic: 0x0627, |
|
alefdageshhebrew: 0xFB30, |
|
aleffinalarabic: 0xFE8E, |
|
alefhamzaabovearabic: 0x0623, |
|
alefhamzaabovefinalarabic: 0xFE84, |
|
alefhamzabelowarabic: 0x0625, |
|
alefhamzabelowfinalarabic: 0xFE88, |
|
alefhebrew: 0x05D0, |
|
aleflamedhebrew: 0xFB4F, |
|
alefmaddaabovearabic: 0x0622, |
|
alefmaddaabovefinalarabic: 0xFE82, |
|
alefmaksuraarabic: 0x0649, |
|
alefmaksurafinalarabic: 0xFEF0, |
|
alefmaksurainitialarabic: 0xFEF3, |
|
alefmaksuramedialarabic: 0xFEF4, |
|
alefpatahhebrew: 0xFB2E, |
|
alefqamatshebrew: 0xFB2F, |
|
aleph: 0x2135, |
|
allequal: 0x224C, |
|
alpha: 0x03B1, |
|
alphatonos: 0x03AC, |
|
amacron: 0x0101, |
|
amonospace: 0xFF41, |
|
ampersand: 0x0026, |
|
ampersandmonospace: 0xFF06, |
|
ampersandsmall: 0xF726, |
|
amsquare: 0x33C2, |
|
anbopomofo: 0x3122, |
|
angbopomofo: 0x3124, |
|
angbracketleft: 0x3008, // This glyph is missing from Adobe's original list. |
|
angbracketright: 0x3009, // This glyph is missing from Adobe's original list. |
|
angkhankhuthai: 0x0E5A, |
|
angle: 0x2220, |
|
anglebracketleft: 0x3008, |
|
anglebracketleftvertical: 0xFE3F, |
|
anglebracketright: 0x3009, |
|
anglebracketrightvertical: 0xFE40, |
|
angleleft: 0x2329, |
|
angleright: 0x232A, |
|
angstrom: 0x212B, |
|
anoteleia: 0x0387, |
|
anudattadeva: 0x0952, |
|
anusvarabengali: 0x0982, |
|
anusvaradeva: 0x0902, |
|
anusvaragujarati: 0x0A82, |
|
aogonek: 0x0105, |
|
apaatosquare: 0x3300, |
|
aparen: 0x249C, |
|
apostrophearmenian: 0x055A, |
|
apostrophemod: 0x02BC, |
|
apple: 0xF8FF, |
|
approaches: 0x2250, |
|
approxequal: 0x2248, |
|
approxequalorimage: 0x2252, |
|
approximatelyequal: 0x2245, |
|
araeaekorean: 0x318E, |
|
araeakorean: 0x318D, |
|
arc: 0x2312, |
|
arighthalfring: 0x1E9A, |
|
aring: 0x00E5, |
|
aringacute: 0x01FB, |
|
aringbelow: 0x1E01, |
|
arrowboth: 0x2194, |
|
arrowdashdown: 0x21E3, |
|
arrowdashleft: 0x21E0, |
|
arrowdashright: 0x21E2, |
|
arrowdashup: 0x21E1, |
|
arrowdblboth: 0x21D4, |
|
arrowdbldown: 0x21D3, |
|
arrowdblleft: 0x21D0, |
|
arrowdblright: 0x21D2, |
|
arrowdblup: 0x21D1, |
|
arrowdown: 0x2193, |
|
arrowdownleft: 0x2199, |
|
arrowdownright: 0x2198, |
|
arrowdownwhite: 0x21E9, |
|
arrowheaddownmod: 0x02C5, |
|
arrowheadleftmod: 0x02C2, |
|
arrowheadrightmod: 0x02C3, |
|
arrowheadupmod: 0x02C4, |
|
arrowhorizex: 0xF8E7, |
|
arrowleft: 0x2190, |
|
arrowleftdbl: 0x21D0, |
|
arrowleftdblstroke: 0x21CD, |
|
arrowleftoverright: 0x21C6, |
|
arrowleftwhite: 0x21E6, |
|
arrowright: 0x2192, |
|
arrowrightdblstroke: 0x21CF, |
|
arrowrightheavy: 0x279E, |
|
arrowrightoverleft: 0x21C4, |
|
arrowrightwhite: 0x21E8, |
|
arrowtableft: 0x21E4, |
|
arrowtabright: 0x21E5, |
|
arrowup: 0x2191, |
|
arrowupdn: 0x2195, |
|
arrowupdnbse: 0x21A8, |
|
arrowupdownbase: 0x21A8, |
|
arrowupleft: 0x2196, |
|
arrowupleftofdown: 0x21C5, |
|
arrowupright: 0x2197, |
|
arrowupwhite: 0x21E7, |
|
arrowvertex: 0xF8E6, |
|
asciicircum: 0x005E, |
|
asciicircummonospace: 0xFF3E, |
|
asciitilde: 0x007E, |
|
asciitildemonospace: 0xFF5E, |
|
ascript: 0x0251, |
|
ascriptturned: 0x0252, |
|
asmallhiragana: 0x3041, |
|
asmallkatakana: 0x30A1, |
|
asmallkatakanahalfwidth: 0xFF67, |
|
asterisk: 0x002A, |
|
asteriskaltonearabic: 0x066D, |
|
asteriskarabic: 0x066D, |
|
asteriskmath: 0x2217, |
|
asteriskmonospace: 0xFF0A, |
|
asterisksmall: 0xFE61, |
|
asterism: 0x2042, |
|
asuperior: 0xF6E9, |
|
asymptoticallyequal: 0x2243, |
|
at: 0x0040, |
|
atilde: 0x00E3, |
|
atmonospace: 0xFF20, |
|
atsmall: 0xFE6B, |
|
aturned: 0x0250, |
|
aubengali: 0x0994, |
|
aubopomofo: 0x3120, |
|
audeva: 0x0914, |
|
augujarati: 0x0A94, |
|
augurmukhi: 0x0A14, |
|
aulengthmarkbengali: 0x09D7, |
|
aumatragurmukhi: 0x0A4C, |
|
auvowelsignbengali: 0x09CC, |
|
auvowelsigndeva: 0x094C, |
|
auvowelsigngujarati: 0x0ACC, |
|
avagrahadeva: 0x093D, |
|
aybarmenian: 0x0561, |
|
ayin: 0x05E2, |
|
ayinaltonehebrew: 0xFB20, |
|
ayinhebrew: 0x05E2, |
|
b: 0x0062, |
|
babengali: 0x09AC, |
|
backslash: 0x005C, |
|
backslashmonospace: 0xFF3C, |
|
badeva: 0x092C, |
|
bagujarati: 0x0AAC, |
|
bagurmukhi: 0x0A2C, |
|
bahiragana: 0x3070, |
|
bahtthai: 0x0E3F, |
|
bakatakana: 0x30D0, |
|
bar: 0x007C, |
|
barmonospace: 0xFF5C, |
|
bbopomofo: 0x3105, |
|
bcircle: 0x24D1, |
|
bdotaccent: 0x1E03, |
|
bdotbelow: 0x1E05, |
|
beamedsixteenthnotes: 0x266C, |
|
because: 0x2235, |
|
becyrillic: 0x0431, |
|
beharabic: 0x0628, |
|
behfinalarabic: 0xFE90, |
|
behinitialarabic: 0xFE91, |
|
behiragana: 0x3079, |
|
behmedialarabic: 0xFE92, |
|
behmeeminitialarabic: 0xFC9F, |
|
behmeemisolatedarabic: 0xFC08, |
|
behnoonfinalarabic: 0xFC6D, |
|
bekatakana: 0x30D9, |
|
benarmenian: 0x0562, |
|
bet: 0x05D1, |
|
beta: 0x03B2, |
|
betasymbolgreek: 0x03D0, |
|
betdagesh: 0xFB31, |
|
betdageshhebrew: 0xFB31, |
|
bethebrew: 0x05D1, |
|
betrafehebrew: 0xFB4C, |
|
bhabengali: 0x09AD, |
|
bhadeva: 0x092D, |
|
bhagujarati: 0x0AAD, |
|
bhagurmukhi: 0x0A2D, |
|
bhook: 0x0253, |
|
bihiragana: 0x3073, |
|
bikatakana: 0x30D3, |
|
bilabialclick: 0x0298, |
|
bindigurmukhi: 0x0A02, |
|
birusquare: 0x3331, |
|
blackcircle: 0x25CF, |
|
blackdiamond: 0x25C6, |
|
blackdownpointingtriangle: 0x25BC, |
|
blackleftpointingpointer: 0x25C4, |
|
blackleftpointingtriangle: 0x25C0, |
|
blacklenticularbracketleft: 0x3010, |
|
blacklenticularbracketleftvertical: 0xFE3B, |
|
blacklenticularbracketright: 0x3011, |
|
blacklenticularbracketrightvertical: 0xFE3C, |
|
blacklowerlefttriangle: 0x25E3, |
|
blacklowerrighttriangle: 0x25E2, |
|
blackrectangle: 0x25AC, |
|
blackrightpointingpointer: 0x25BA, |
|
blackrightpointingtriangle: 0x25B6, |
|
blacksmallsquare: 0x25AA, |
|
blacksmilingface: 0x263B, |
|
blacksquare: 0x25A0, |
|
blackstar: 0x2605, |
|
blackupperlefttriangle: 0x25E4, |
|
blackupperrighttriangle: 0x25E5, |
|
blackuppointingsmalltriangle: 0x25B4, |
|
blackuppointingtriangle: 0x25B2, |
|
blank: 0x2423, |
|
blinebelow: 0x1E07, |
|
block: 0x2588, |
|
bmonospace: 0xFF42, |
|
bobaimaithai: 0x0E1A, |
|
bohiragana: 0x307C, |
|
bokatakana: 0x30DC, |
|
bparen: 0x249D, |
|
bqsquare: 0x33C3, |
|
braceex: 0xF8F4, |
|
braceleft: 0x007B, |
|
braceleftbt: 0xF8F3, |
|
braceleftmid: 0xF8F2, |
|
braceleftmonospace: 0xFF5B, |
|
braceleftsmall: 0xFE5B, |
|
bracelefttp: 0xF8F1, |
|
braceleftvertical: 0xFE37, |
|
braceright: 0x007D, |
|
bracerightbt: 0xF8FE, |
|
bracerightmid: 0xF8FD, |
|
bracerightmonospace: 0xFF5D, |
|
bracerightsmall: 0xFE5C, |
|
bracerighttp: 0xF8FC, |
|
bracerightvertical: 0xFE38, |
|
bracketleft: 0x005B, |
|
bracketleftbt: 0xF8F0, |
|
bracketleftex: 0xF8EF, |
|
bracketleftmonospace: 0xFF3B, |
|
bracketlefttp: 0xF8EE, |
|
bracketright: 0x005D, |
|
bracketrightbt: 0xF8FB, |
|
bracketrightex: 0xF8FA, |
|
bracketrightmonospace: 0xFF3D, |
|
bracketrighttp: 0xF8F9, |
|
breve: 0x02D8, |
|
brevebelowcmb: 0x032E, |
|
brevecmb: 0x0306, |
|
breveinvertedbelowcmb: 0x032F, |
|
breveinvertedcmb: 0x0311, |
|
breveinverteddoublecmb: 0x0361, |
|
bridgebelowcmb: 0x032A, |
|
bridgeinvertedbelowcmb: 0x033A, |
|
brokenbar: 0x00A6, |
|
bstroke: 0x0180, |
|
bsuperior: 0xF6EA, |
|
btopbar: 0x0183, |
|
buhiragana: 0x3076, |
|
bukatakana: 0x30D6, |
|
bullet: 0x2022, |
|
bulletinverse: 0x25D8, |
|
bulletoperator: 0x2219, |
|
bullseye: 0x25CE, |
|
c: 0x0063, |
|
caarmenian: 0x056E, |
|
cabengali: 0x099A, |
|
cacute: 0x0107, |
|
cadeva: 0x091A, |
|
cagujarati: 0x0A9A, |
|
cagurmukhi: 0x0A1A, |
|
calsquare: 0x3388, |
|
candrabindubengali: 0x0981, |
|
candrabinducmb: 0x0310, |
|
candrabindudeva: 0x0901, |
|
candrabindugujarati: 0x0A81, |
|
capslock: 0x21EA, |
|
careof: 0x2105, |
|
caron: 0x02C7, |
|
caronbelowcmb: 0x032C, |
|
caroncmb: 0x030C, |
|
carriagereturn: 0x21B5, |
|
cbopomofo: 0x3118, |
|
ccaron: 0x010D, |
|
ccedilla: 0x00E7, |
|
ccedillaacute: 0x1E09, |
|
ccircle: 0x24D2, |
|
ccircumflex: 0x0109, |
|
ccurl: 0x0255, |
|
cdot: 0x010B, |
|
cdotaccent: 0x010B, |
|
cdsquare: 0x33C5, |
|
cedilla: 0x00B8, |
|
cedillacmb: 0x0327, |
|
cent: 0x00A2, |
|
centigrade: 0x2103, |
|
centinferior: 0xF6DF, |
|
centmonospace: 0xFFE0, |
|
centoldstyle: 0xF7A2, |
|
centsuperior: 0xF6E0, |
|
chaarmenian: 0x0579, |
|
chabengali: 0x099B, |
|
chadeva: 0x091B, |
|
chagujarati: 0x0A9B, |
|
chagurmukhi: 0x0A1B, |
|
chbopomofo: 0x3114, |
|
cheabkhasiancyrillic: 0x04BD, |
|
checkmark: 0x2713, |
|
checyrillic: 0x0447, |
|
chedescenderabkhasiancyrillic: 0x04BF, |
|
chedescendercyrillic: 0x04B7, |
|
chedieresiscyrillic: 0x04F5, |
|
cheharmenian: 0x0573, |
|
chekhakassiancyrillic: 0x04CC, |
|
cheverticalstrokecyrillic: 0x04B9, |
|
chi: 0x03C7, |
|
chieuchacirclekorean: 0x3277, |
|
chieuchaparenkorean: 0x3217, |
|
chieuchcirclekorean: 0x3269, |
|
chieuchkorean: 0x314A, |
|
chieuchparenkorean: 0x3209, |
|
chochangthai: 0x0E0A, |
|
chochanthai: 0x0E08, |
|
chochingthai: 0x0E09, |
|
chochoethai: 0x0E0C, |
|
chook: 0x0188, |
|
cieucacirclekorean: 0x3276, |
|
cieucaparenkorean: 0x3216, |
|
cieuccirclekorean: 0x3268, |
|
cieuckorean: 0x3148, |
|
cieucparenkorean: 0x3208, |
|
cieucuparenkorean: 0x321C, |
|
circle: 0x25CB, |
|
circlecopyrt: 0x00A9, // This glyph is missing from Adobe's original list. |
|
circlemultiply: 0x2297, |
|
circleot: 0x2299, |
|
circleplus: 0x2295, |
|
circlepostalmark: 0x3036, |
|
circlewithlefthalfblack: 0x25D0, |
|
circlewithrighthalfblack: 0x25D1, |
|
circumflex: 0x02C6, |
|
circumflexbelowcmb: 0x032D, |
|
circumflexcmb: 0x0302, |
|
clear: 0x2327, |
|
clickalveolar: 0x01C2, |
|
clickdental: 0x01C0, |
|
clicklateral: 0x01C1, |
|
clickretroflex: 0x01C3, |
|
club: 0x2663, |
|
clubsuitblack: 0x2663, |
|
clubsuitwhite: 0x2667, |
|
cmcubedsquare: 0x33A4, |
|
cmonospace: 0xFF43, |
|
cmsquaredsquare: 0x33A0, |
|
coarmenian: 0x0581, |
|
colon: 0x003A, |
|
colonmonetary: 0x20A1, |
|
colonmonospace: 0xFF1A, |
|
colonsign: 0x20A1, |
|
colonsmall: 0xFE55, |
|
colontriangularhalfmod: 0x02D1, |
|
colontriangularmod: 0x02D0, |
|
comma: 0x002C, |
|
commaabovecmb: 0x0313, |
|
commaaboverightcmb: 0x0315, |
|
commaaccent: 0xF6C3, |
|
commaarabic: 0x060C, |
|
commaarmenian: 0x055D, |
|
commainferior: 0xF6E1, |
|
commamonospace: 0xFF0C, |
|
commareversedabovecmb: 0x0314, |
|
commareversedmod: 0x02BD, |
|
commasmall: 0xFE50, |
|
commasuperior: 0xF6E2, |
|
commaturnedabovecmb: 0x0312, |
|
commaturnedmod: 0x02BB, |
|
compass: 0x263C, |
|
congruent: 0x2245, |
|
contourintegral: 0x222E, |
|
control: 0x2303, |
|
controlACK: 0x0006, |
|
controlBEL: 0x0007, |
|
controlBS: 0x0008, |
|
controlCAN: 0x0018, |
|
controlCR: 0x000D, |
|
controlDC1: 0x0011, |
|
controlDC2: 0x0012, |
|
controlDC3: 0x0013, |
|
controlDC4: 0x0014, |
|
controlDEL: 0x007F, |
|
controlDLE: 0x0010, |
|
controlEM: 0x0019, |
|
controlENQ: 0x0005, |
|
controlEOT: 0x0004, |
|
controlESC: 0x001B, |
|
controlETB: 0x0017, |
|
controlETX: 0x0003, |
|
controlFF: 0x000C, |
|
controlFS: 0x001C, |
|
controlGS: 0x001D, |
|
controlHT: 0x0009, |
|
controlLF: 0x000A, |
|
controlNAK: 0x0015, |
|
controlRS: 0x001E, |
|
controlSI: 0x000F, |
|
controlSO: 0x000E, |
|
controlSOT: 0x0002, |
|
controlSTX: 0x0001, |
|
controlSUB: 0x001A, |
|
controlSYN: 0x0016, |
|
controlUS: 0x001F, |
|
controlVT: 0x000B, |
|
copyright: 0x00A9, |
|
copyrightsans: 0xF8E9, |
|
copyrightserif: 0xF6D9, |
|
cornerbracketleft: 0x300C, |
|
cornerbracketlefthalfwidth: 0xFF62, |
|
cornerbracketleftvertical: 0xFE41, |
|
cornerbracketright: 0x300D, |
|
cornerbracketrighthalfwidth: 0xFF63, |
|
cornerbracketrightvertical: 0xFE42, |
|
corporationsquare: 0x337F, |
|
cosquare: 0x33C7, |
|
coverkgsquare: 0x33C6, |
|
cparen: 0x249E, |
|
cruzeiro: 0x20A2, |
|
cstretched: 0x0297, |
|
curlyand: 0x22CF, |
|
curlyor: 0x22CE, |
|
currency: 0x00A4, |
|
cyrBreve: 0xF6D1, |
|
cyrFlex: 0xF6D2, |
|
cyrbreve: 0xF6D4, |
|
cyrflex: 0xF6D5, |
|
d: 0x0064, |
|
daarmenian: 0x0564, |
|
dabengali: 0x09A6, |
|
dadarabic: 0x0636, |
|
dadeva: 0x0926, |
|
dadfinalarabic: 0xFEBE, |
|
dadinitialarabic: 0xFEBF, |
|
dadmedialarabic: 0xFEC0, |
|
dagesh: 0x05BC, |
|
dageshhebrew: 0x05BC, |
|
dagger: 0x2020, |
|
daggerdbl: 0x2021, |
|
dagujarati: 0x0AA6, |
|
dagurmukhi: 0x0A26, |
|
dahiragana: 0x3060, |
|
dakatakana: 0x30C0, |
|
dalarabic: 0x062F, |
|
dalet: 0x05D3, |
|
daletdagesh: 0xFB33, |
|
daletdageshhebrew: 0xFB33, |
|
dalethebrew: 0x05D3, |
|
dalfinalarabic: 0xFEAA, |
|
dammaarabic: 0x064F, |
|
dammalowarabic: 0x064F, |
|
dammatanaltonearabic: 0x064C, |
|
dammatanarabic: 0x064C, |
|
danda: 0x0964, |
|
dargahebrew: 0x05A7, |
|
dargalefthebrew: 0x05A7, |
|
dasiapneumatacyrilliccmb: 0x0485, |
|
dblGrave: 0xF6D3, |
|
dblanglebracketleft: 0x300A, |
|
dblanglebracketleftvertical: 0xFE3D, |
|
dblanglebracketright: 0x300B, |
|
dblanglebracketrightvertical: 0xFE3E, |
|
dblarchinvertedbelowcmb: 0x032B, |
|
dblarrowleft: 0x21D4, |
|
dblarrowright: 0x21D2, |
|
dbldanda: 0x0965, |
|
dblgrave: 0xF6D6, |
|
dblgravecmb: 0x030F, |
|
dblintegral: 0x222C, |
|
dbllowline: 0x2017, |
|
dbllowlinecmb: 0x0333, |
|
dbloverlinecmb: 0x033F, |
|
dblprimemod: 0x02BA, |
|
dblverticalbar: 0x2016, |
|
dblverticallineabovecmb: 0x030E, |
|
dbopomofo: 0x3109, |
|
dbsquare: 0x33C8, |
|
dcaron: 0x010F, |
|
dcedilla: 0x1E11, |
|
dcircle: 0x24D3, |
|
dcircumflexbelow: 0x1E13, |
|
dcroat: 0x0111, |
|
ddabengali: 0x09A1, |
|
ddadeva: 0x0921, |
|
ddagujarati: 0x0AA1, |
|
ddagurmukhi: 0x0A21, |
|
ddalarabic: 0x0688, |
|
ddalfinalarabic: 0xFB89, |
|
dddhadeva: 0x095C, |
|
ddhabengali: 0x09A2, |
|
ddhadeva: 0x0922, |
|
ddhagujarati: 0x0AA2, |
|
ddhagurmukhi: 0x0A22, |
|
ddotaccent: 0x1E0B, |
|
ddotbelow: 0x1E0D, |
|
decimalseparatorarabic: 0x066B, |
|
decimalseparatorpersian: 0x066B, |
|
decyrillic: 0x0434, |
|
degree: 0x00B0, |
|
dehihebrew: 0x05AD, |
|
dehiragana: 0x3067, |
|
deicoptic: 0x03EF, |
|
dekatakana: 0x30C7, |
|
deleteleft: 0x232B, |
|
deleteright: 0x2326, |
|
delta: 0x03B4, |
|
deltaturned: 0x018D, |
|
denominatorminusonenumeratorbengali: 0x09F8, |
|
dezh: 0x02A4, |
|
dhabengali: 0x09A7, |
|
dhadeva: 0x0927, |
|
dhagujarati: 0x0AA7, |
|
dhagurmukhi: 0x0A27, |
|
dhook: 0x0257, |
|
dialytikatonos: 0x0385, |
|
dialytikatonoscmb: 0x0344, |
|
diamond: 0x2666, |
|
diamondsuitwhite: 0x2662, |
|
dieresis: 0x00A8, |
|
dieresisacute: 0xF6D7, |
|
dieresisbelowcmb: 0x0324, |
|
dieresiscmb: 0x0308, |
|
dieresisgrave: 0xF6D8, |
|
dieresistonos: 0x0385, |
|
dihiragana: 0x3062, |
|
dikatakana: 0x30C2, |
|
dittomark: 0x3003, |
|
divide: 0x00F7, |
|
divides: 0x2223, |
|
divisionslash: 0x2215, |
|
djecyrillic: 0x0452, |
|
dkshade: 0x2593, |
|
dlinebelow: 0x1E0F, |
|
dlsquare: 0x3397, |
|
dmacron: 0x0111, |
|
dmonospace: 0xFF44, |
|
dnblock: 0x2584, |
|
dochadathai: 0x0E0E, |
|
dodekthai: 0x0E14, |
|
dohiragana: 0x3069, |
|
dokatakana: 0x30C9, |
|
dollar: 0x0024, |
|
dollarinferior: 0xF6E3, |
|
dollarmonospace: 0xFF04, |
|
dollaroldstyle: 0xF724, |
|
dollarsmall: 0xFE69, |
|
dollarsuperior: 0xF6E4, |
|
dong: 0x20AB, |
|
dorusquare: 0x3326, |
|
dotaccent: 0x02D9, |
|
dotaccentcmb: 0x0307, |
|
dotbelowcmb: 0x0323, |
|
dotbelowcomb: 0x0323, |
|
dotkatakana: 0x30FB, |
|
dotlessi: 0x0131, |
|
dotlessj: 0xF6BE, |
|
dotlessjstrokehook: 0x0284, |
|
dotmath: 0x22C5, |
|
dottedcircle: 0x25CC, |
|
doubleyodpatah: 0xFB1F, |
|
doubleyodpatahhebrew: 0xFB1F, |
|
downtackbelowcmb: 0x031E, |
|
downtackmod: 0x02D5, |
|
dparen: 0x249F, |
|
dsuperior: 0xF6EB, |
|
dtail: 0x0256, |
|
dtopbar: 0x018C, |
|
duhiragana: 0x3065, |
|
dukatakana: 0x30C5, |
|
dz: 0x01F3, |
|
dzaltone: 0x02A3, |
|
dzcaron: 0x01C6, |
|
dzcurl: 0x02A5, |
|
dzeabkhasiancyrillic: 0x04E1, |
|
dzecyrillic: 0x0455, |
|
dzhecyrillic: 0x045F, |
|
e: 0x0065, |
|
eacute: 0x00E9, |
|
earth: 0x2641, |
|
ebengali: 0x098F, |
|
ebopomofo: 0x311C, |
|
ebreve: 0x0115, |
|
ecandradeva: 0x090D, |
|
ecandragujarati: 0x0A8D, |
|
ecandravowelsigndeva: 0x0945, |
|
ecandravowelsigngujarati: 0x0AC5, |
|
ecaron: 0x011B, |
|
ecedillabreve: 0x1E1D, |
|
echarmenian: 0x0565, |
|
echyiwnarmenian: 0x0587, |
|
ecircle: 0x24D4, |
|
ecircumflex: 0x00EA, |
|
ecircumflexacute: 0x1EBF, |
|
ecircumflexbelow: 0x1E19, |
|
ecircumflexdotbelow: 0x1EC7, |
|
ecircumflexgrave: 0x1EC1, |
|
ecircumflexhookabove: 0x1EC3, |
|
ecircumflextilde: 0x1EC5, |
|
ecyrillic: 0x0454, |
|
edblgrave: 0x0205, |
|
edeva: 0x090F, |
|
edieresis: 0x00EB, |
|
edot: 0x0117, |
|
edotaccent: 0x0117, |
|
edotbelow: 0x1EB9, |
|
eegurmukhi: 0x0A0F, |
|
eematragurmukhi: 0x0A47, |
|
efcyrillic: 0x0444, |
|
egrave: 0x00E8, |
|
egujarati: 0x0A8F, |
|
eharmenian: 0x0567, |
|
ehbopomofo: 0x311D, |
|
ehiragana: 0x3048, |
|
ehookabove: 0x1EBB, |
|
eibopomofo: 0x311F, |
|
eight: 0x0038, |
|
eightarabic: 0x0668, |
|
eightbengali: 0x09EE, |
|
eightcircle: 0x2467, |
|
eightcircleinversesansserif: 0x2791, |
|
eightdeva: 0x096E, |
|
eighteencircle: 0x2471, |
|
eighteenparen: 0x2485, |
|
eighteenperiod: 0x2499, |
|
eightgujarati: 0x0AEE, |
|
eightgurmukhi: 0x0A6E, |
|
eighthackarabic: 0x0668, |
|
eighthangzhou: 0x3028, |
|
eighthnotebeamed: 0x266B, |
|
eightideographicparen: 0x3227, |
|
eightinferior: 0x2088, |
|
eightmonospace: 0xFF18, |
|
eightoldstyle: 0xF738, |
|
eightparen: 0x247B, |
|
eightperiod: 0x248F, |
|
eightpersian: 0x06F8, |
|
eightroman: 0x2177, |
|
eightsuperior: 0x2078, |
|
eightthai: 0x0E58, |
|
einvertedbreve: 0x0207, |
|
eiotifiedcyrillic: 0x0465, |
|
ekatakana: 0x30A8, |
|
ekatakanahalfwidth: 0xFF74, |
|
ekonkargurmukhi: 0x0A74, |
|
ekorean: 0x3154, |
|
elcyrillic: 0x043B, |
|
element: 0x2208, |
|
elevencircle: 0x246A, |
|
elevenparen: 0x247E, |
|
elevenperiod: 0x2492, |
|
elevenroman: 0x217A, |
|
ellipsis: 0x2026, |
|
ellipsisvertical: 0x22EE, |
|
emacron: 0x0113, |
|
emacronacute: 0x1E17, |
|
emacrongrave: 0x1E15, |
|
emcyrillic: 0x043C, |
|
emdash: 0x2014, |
|
emdashvertical: 0xFE31, |
|
emonospace: 0xFF45, |
|
emphasismarkarmenian: 0x055B, |
|
emptyset: 0x2205, |
|
enbopomofo: 0x3123, |
|
encyrillic: 0x043D, |
|
endash: 0x2013, |
|
endashvertical: 0xFE32, |
|
endescendercyrillic: 0x04A3, |
|
eng: 0x014B, |
|
engbopomofo: 0x3125, |
|
enghecyrillic: 0x04A5, |
|
enhookcyrillic: 0x04C8, |
|
enspace: 0x2002, |
|
eogonek: 0x0119, |
|
eokorean: 0x3153, |
|
eopen: 0x025B, |
|
eopenclosed: 0x029A, |
|
eopenreversed: 0x025C, |
|
eopenreversedclosed: 0x025E, |
|
eopenreversedhook: 0x025D, |
|
eparen: 0x24A0, |
|
epsilon: 0x03B5, |
|
epsilontonos: 0x03AD, |
|
equal: 0x003D, |
|
equalmonospace: 0xFF1D, |
|
equalsmall: 0xFE66, |
|
equalsuperior: 0x207C, |
|
equivalence: 0x2261, |
|
erbopomofo: 0x3126, |
|
ercyrillic: 0x0440, |
|
ereversed: 0x0258, |
|
ereversedcyrillic: 0x044D, |
|
escyrillic: 0x0441, |
|
esdescendercyrillic: 0x04AB, |
|
esh: 0x0283, |
|
eshcurl: 0x0286, |
|
eshortdeva: 0x090E, |
|
eshortvowelsigndeva: 0x0946, |
|
eshreversedloop: 0x01AA, |
|
eshsquatreversed: 0x0285, |
|
esmallhiragana: 0x3047, |
|
esmallkatakana: 0x30A7, |
|
esmallkatakanahalfwidth: 0xFF6A, |
|
estimated: 0x212E, |
|
esuperior: 0xF6EC, |
|
eta: 0x03B7, |
|
etarmenian: 0x0568, |
|
etatonos: 0x03AE, |
|
eth: 0x00F0, |
|
etilde: 0x1EBD, |
|
etildebelow: 0x1E1B, |
|
etnahtafoukhhebrew: 0x0591, |
|
etnahtafoukhlefthebrew: 0x0591, |
|
etnahtahebrew: 0x0591, |
|
etnahtalefthebrew: 0x0591, |
|
eturned: 0x01DD, |
|
eukorean: 0x3161, |
|
euro: 0x20AC, |
|
evowelsignbengali: 0x09C7, |
|
evowelsigndeva: 0x0947, |
|
evowelsigngujarati: 0x0AC7, |
|
exclam: 0x0021, |
|
exclamarmenian: 0x055C, |
|
exclamdbl: 0x203C, |
|
exclamdown: 0x00A1, |
|
exclamdownsmall: 0xF7A1, |
|
exclammonospace: 0xFF01, |
|
exclamsmall: 0xF721, |
|
existential: 0x2203, |
|
ezh: 0x0292, |
|
ezhcaron: 0x01EF, |
|
ezhcurl: 0x0293, |
|
ezhreversed: 0x01B9, |
|
ezhtail: 0x01BA, |
|
f: 0x0066, |
|
fadeva: 0x095E, |
|
fagurmukhi: 0x0A5E, |
|
fahrenheit: 0x2109, |
|
fathaarabic: 0x064E, |
|
fathalowarabic: 0x064E, |
|
fathatanarabic: 0x064B, |
|
fbopomofo: 0x3108, |
|
fcircle: 0x24D5, |
|
fdotaccent: 0x1E1F, |
|
feharabic: 0x0641, |
|
feharmenian: 0x0586, |
|
fehfinalarabic: 0xFED2, |
|
fehinitialarabic: 0xFED3, |
|
fehmedialarabic: 0xFED4, |
|
feicoptic: 0x03E5, |
|
female: 0x2640, |
|
ff: 0xFB00, |
|
ffi: 0xFB03, |
|
ffl: 0xFB04, |
|
fi: 0xFB01, |
|
fifteencircle: 0x246E, |
|
fifteenparen: 0x2482, |
|
fifteenperiod: 0x2496, |
|
figuredash: 0x2012, |
|
filledbox: 0x25A0, |
|
filledrect: 0x25AC, |
|
finalkaf: 0x05DA, |
|
finalkafdagesh: 0xFB3A, |
|
finalkafdageshhebrew: 0xFB3A, |
|
finalkafhebrew: 0x05DA, |
|
finalmem: 0x05DD, |
|
finalmemhebrew: 0x05DD, |
|
finalnun: 0x05DF, |
|
finalnunhebrew: 0x05DF, |
|
finalpe: 0x05E3, |
|
finalpehebrew: 0x05E3, |
|
finaltsadi: 0x05E5, |
|
finaltsadihebrew: 0x05E5, |
|
firsttonechinese: 0x02C9, |
|
fisheye: 0x25C9, |
|
fitacyrillic: 0x0473, |
|
five: 0x0035, |
|
fivearabic: 0x0665, |
|
fivebengali: 0x09EB, |
|
fivecircle: 0x2464, |
|
fivecircleinversesansserif: 0x278E, |
|
fivedeva: 0x096B, |
|
fiveeighths: 0x215D, |
|
fivegujarati: 0x0AEB, |
|
fivegurmukhi: 0x0A6B, |
|
fivehackarabic: 0x0665, |
|
fivehangzhou: 0x3025, |
|
fiveideographicparen: 0x3224, |
|
fiveinferior: 0x2085, |
|
fivemonospace: 0xFF15, |
|
fiveoldstyle: 0xF735, |
|
fiveparen: 0x2478, |
|
fiveperiod: 0x248C, |
|
fivepersian: 0x06F5, |
|
fiveroman: 0x2174, |
|
fivesuperior: 0x2075, |
|
fivethai: 0x0E55, |
|
fl: 0xFB02, |
|
florin: 0x0192, |
|
fmonospace: 0xFF46, |
|
fmsquare: 0x3399, |
|
fofanthai: 0x0E1F, |
|
fofathai: 0x0E1D, |
|
fongmanthai: 0x0E4F, |
|
forall: 0x2200, |
|
four: 0x0034, |
|
fourarabic: 0x0664, |
|
fourbengali: 0x09EA, |
|
fourcircle: 0x2463, |
|
fourcircleinversesansserif: 0x278D, |
|
fourdeva: 0x096A, |
|
fourgujarati: 0x0AEA, |
|
fourgurmukhi: 0x0A6A, |
|
fourhackarabic: 0x0664, |
|
fourhangzhou: 0x3024, |
|
fourideographicparen: 0x3223, |
|
fourinferior: 0x2084, |
|
fourmonospace: 0xFF14, |
|
fournumeratorbengali: 0x09F7, |
|
fouroldstyle: 0xF734, |
|
fourparen: 0x2477, |
|
fourperiod: 0x248B, |
|
fourpersian: 0x06F4, |
|
fourroman: 0x2173, |
|
foursuperior: 0x2074, |
|
fourteencircle: 0x246D, |
|
fourteenparen: 0x2481, |
|
fourteenperiod: 0x2495, |
|
fourthai: 0x0E54, |
|
fourthtonechinese: 0x02CB, |
|
fparen: 0x24A1, |
|
fraction: 0x2044, |
|
franc: 0x20A3, |
|
g: 0x0067, |
|
gabengali: 0x0997, |
|
gacute: 0x01F5, |
|
gadeva: 0x0917, |
|
gafarabic: 0x06AF, |
|
gaffinalarabic: 0xFB93, |
|
gafinitialarabic: 0xFB94, |
|
gafmedialarabic: 0xFB95, |
|
gagujarati: 0x0A97, |
|
gagurmukhi: 0x0A17, |
|
gahiragana: 0x304C, |
|
gakatakana: 0x30AC, |
|
gamma: 0x03B3, |
|
gammalatinsmall: 0x0263, |
|
gammasuperior: 0x02E0, |
|
gangiacoptic: 0x03EB, |
|
gbopomofo: 0x310D, |
|
gbreve: 0x011F, |
|
gcaron: 0x01E7, |
|
gcedilla: 0x0123, |
|
gcircle: 0x24D6, |
|
gcircumflex: 0x011D, |
|
gcommaaccent: 0x0123, |
|
gdot: 0x0121, |
|
gdotaccent: 0x0121, |
|
gecyrillic: 0x0433, |
|
gehiragana: 0x3052, |
|
gekatakana: 0x30B2, |
|
geometricallyequal: 0x2251, |
|
gereshaccenthebrew: 0x059C, |
|
gereshhebrew: 0x05F3, |
|
gereshmuqdamhebrew: 0x059D, |
|
germandbls: 0x00DF, |
|
gershayimaccenthebrew: 0x059E, |
|
gershayimhebrew: 0x05F4, |
|
getamark: 0x3013, |
|
ghabengali: 0x0998, |
|
ghadarmenian: 0x0572, |
|
ghadeva: 0x0918, |
|
ghagujarati: 0x0A98, |
|
ghagurmukhi: 0x0A18, |
|
ghainarabic: 0x063A, |
|
ghainfinalarabic: 0xFECE, |
|
ghaininitialarabic: 0xFECF, |
|
ghainmedialarabic: 0xFED0, |
|
ghemiddlehookcyrillic: 0x0495, |
|
ghestrokecyrillic: 0x0493, |
|
gheupturncyrillic: 0x0491, |
|
ghhadeva: 0x095A, |
|
ghhagurmukhi: 0x0A5A, |
|
ghook: 0x0260, |
|
ghzsquare: 0x3393, |
|
gihiragana: 0x304E, |
|
gikatakana: 0x30AE, |
|
gimarmenian: 0x0563, |
|
gimel: 0x05D2, |
|
gimeldagesh: 0xFB32, |
|
gimeldageshhebrew: 0xFB32, |
|
gimelhebrew: 0x05D2, |
|
gjecyrillic: 0x0453, |
|
glottalinvertedstroke: 0x01BE, |
|
glottalstop: 0x0294, |
|
glottalstopinverted: 0x0296, |
|
glottalstopmod: 0x02C0, |
|
glottalstopreversed: 0x0295, |
|
glottalstopreversedmod: 0x02C1, |
|
glottalstopreversedsuperior: 0x02E4, |
|
glottalstopstroke: 0x02A1, |
|
glottalstopstrokereversed: 0x02A2, |
|
gmacron: 0x1E21, |
|
gmonospace: 0xFF47, |
|
gohiragana: 0x3054, |
|
gokatakana: 0x30B4, |
|
gparen: 0x24A2, |
|
gpasquare: 0x33AC, |
|
gradient: 0x2207, |
|
grave: 0x0060, |
|
gravebelowcmb: 0x0316, |
|
gravecmb: 0x0300, |
|
gravecomb: 0x0300, |
|
gravedeva: 0x0953, |
|
gravelowmod: 0x02CE, |
|
gravemonospace: 0xFF40, |
|
gravetonecmb: 0x0340, |
|
greater: 0x003E, |
|
greaterequal: 0x2265, |
|
greaterequalorless: 0x22DB, |
|
greatermonospace: 0xFF1E, |
|
greaterorequivalent: 0x2273, |
|
greaterorless: 0x2277, |
|
greateroverequal: 0x2267, |
|
greatersmall: 0xFE65, |
|
gscript: 0x0261, |
|
gstroke: 0x01E5, |
|
guhiragana: 0x3050, |
|
guillemotleft: 0x00AB, |
|
guillemotright: 0x00BB, |
|
guilsinglleft: 0x2039, |
|
guilsinglright: 0x203A, |
|
gukatakana: 0x30B0, |
|
guramusquare: 0x3318, |
|
gysquare: 0x33C9, |
|
h: 0x0068, |
|
haabkhasiancyrillic: 0x04A9, |
|
haaltonearabic: 0x06C1, |
|
habengali: 0x09B9, |
|
hadescendercyrillic: 0x04B3, |
|
hadeva: 0x0939, |
|
hagujarati: 0x0AB9, |
|
hagurmukhi: 0x0A39, |
|
haharabic: 0x062D, |
|
hahfinalarabic: 0xFEA2, |
|
hahinitialarabic: 0xFEA3, |
|
hahiragana: 0x306F, |
|
hahmedialarabic: 0xFEA4, |
|
haitusquare: 0x332A, |
|
hakatakana: 0x30CF, |
|
hakatakanahalfwidth: 0xFF8A, |
|
halantgurmukhi: 0x0A4D, |
|
hamzaarabic: 0x0621, |
|
hamzalowarabic: 0x0621, |
|
hangulfiller: 0x3164, |
|
hardsigncyrillic: 0x044A, |
|
harpoonleftbarbup: 0x21BC, |
|
harpoonrightbarbup: 0x21C0, |
|
hasquare: 0x33CA, |
|
hatafpatah: 0x05B2, |
|
hatafpatah16: 0x05B2, |
|
hatafpatah23: 0x05B2, |
|
hatafpatah2f: 0x05B2, |
|
hatafpatahhebrew: 0x05B2, |
|
hatafpatahnarrowhebrew: 0x05B2, |
|
hatafpatahquarterhebrew: 0x05B2, |
|
hatafpatahwidehebrew: 0x05B2, |
|
hatafqamats: 0x05B3, |
|
hatafqamats1b: 0x05B3, |
|
hatafqamats28: 0x05B3, |
|
hatafqamats34: 0x05B3, |
|
hatafqamatshebrew: 0x05B3, |
|
hatafqamatsnarrowhebrew: 0x05B3, |
|
hatafqamatsquarterhebrew: 0x05B3, |
|
hatafqamatswidehebrew: 0x05B3, |
|
hatafsegol: 0x05B1, |
|
hatafsegol17: 0x05B1, |
|
hatafsegol24: 0x05B1, |
|
hatafsegol30: 0x05B1, |
|
hatafsegolhebrew: 0x05B1, |
|
hatafsegolnarrowhebrew: 0x05B1, |
|
hatafsegolquarterhebrew: 0x05B1, |
|
hatafsegolwidehebrew: 0x05B1, |
|
hbar: 0x0127, |
|
hbopomofo: 0x310F, |
|
hbrevebelow: 0x1E2B, |
|
hcedilla: 0x1E29, |
|
hcircle: 0x24D7, |
|
hcircumflex: 0x0125, |
|
hdieresis: 0x1E27, |
|
hdotaccent: 0x1E23, |
|
hdotbelow: 0x1E25, |
|
he: 0x05D4, |
|
heart: 0x2665, |
|
heartsuitblack: 0x2665, |
|
heartsuitwhite: 0x2661, |
|
hedagesh: 0xFB34, |
|
hedageshhebrew: 0xFB34, |
|
hehaltonearabic: 0x06C1, |
|
heharabic: 0x0647, |
|
hehebrew: 0x05D4, |
|
hehfinalaltonearabic: 0xFBA7, |
|
hehfinalalttwoarabic: 0xFEEA, |
|
hehfinalarabic: 0xFEEA, |
|
hehhamzaabovefinalarabic: 0xFBA5, |
|
hehhamzaaboveisolatedarabic: 0xFBA4, |
|
hehinitialaltonearabic: 0xFBA8, |
|
hehinitialarabic: 0xFEEB, |
|
hehiragana: 0x3078, |
|
hehmedialaltonearabic: 0xFBA9, |
|
hehmedialarabic: 0xFEEC, |
|
heiseierasquare: 0x337B, |
|
hekatakana: 0x30D8, |
|
hekatakanahalfwidth: 0xFF8D, |
|
hekutaarusquare: 0x3336, |
|
henghook: 0x0267, |
|
herutusquare: 0x3339, |
|
het: 0x05D7, |
|
hethebrew: 0x05D7, |
|
hhook: 0x0266, |
|
hhooksuperior: 0x02B1, |
|
hieuhacirclekorean: 0x327B, |
|
hieuhaparenkorean: 0x321B, |
|
hieuhcirclekorean: 0x326D, |
|
hieuhkorean: 0x314E, |
|
hieuhparenkorean: 0x320D, |
|
hihiragana: 0x3072, |
|
hikatakana: 0x30D2, |
|
hikatakanahalfwidth: 0xFF8B, |
|
hiriq: 0x05B4, |
|
hiriq14: 0x05B4, |
|
hiriq21: 0x05B4, |
|
hiriq2d: 0x05B4, |
|
hiriqhebrew: 0x05B4, |
|
hiriqnarrowhebrew: 0x05B4, |
|
hiriqquarterhebrew: 0x05B4, |
|
hiriqwidehebrew: 0x05B4, |
|
hlinebelow: 0x1E96, |
|
hmonospace: 0xFF48, |
|
hoarmenian: 0x0570, |
|
hohipthai: 0x0E2B, |
|
hohiragana: 0x307B, |
|
hokatakana: 0x30DB, |
|
hokatakanahalfwidth: 0xFF8E, |
|
holam: 0x05B9, |
|
holam19: 0x05B9, |
|
holam26: 0x05B9, |
|
holam32: 0x05B9, |
|
holamhebrew: 0x05B9, |
|
holamnarrowhebrew: 0x05B9, |
|
holamquarterhebrew: 0x05B9, |
|
holamwidehebrew: 0x05B9, |
|
honokhukthai: 0x0E2E, |
|
hookabovecomb: 0x0309, |
|
hookcmb: 0x0309, |
|
hookpalatalizedbelowcmb: 0x0321, |
|
hookretroflexbelowcmb: 0x0322, |
|
hoonsquare: 0x3342, |
|
horicoptic: 0x03E9, |
|
horizontalbar: 0x2015, |
|
horncmb: 0x031B, |
|
hotsprings: 0x2668, |
|
house: 0x2302, |
|
hparen: 0x24A3, |
|
hsuperior: 0x02B0, |
|
hturned: 0x0265, |
|
huhiragana: 0x3075, |
|
huiitosquare: 0x3333, |
|
hukatakana: 0x30D5, |
|
hukatakanahalfwidth: 0xFF8C, |
|
hungarumlaut: 0x02DD, |
|
hungarumlautcmb: 0x030B, |
|
hv: 0x0195, |
|
hyphen: 0x002D, |
|
hypheninferior: 0xF6E5, |
|
hyphenmonospace: 0xFF0D, |
|
hyphensmall: 0xFE63, |
|
hyphensuperior: 0xF6E6, |
|
hyphentwo: 0x2010, |
|
i: 0x0069, |
|
iacute: 0x00ED, |
|
iacyrillic: 0x044F, |
|
ibengali: 0x0987, |
|
ibopomofo: 0x3127, |
|
ibreve: 0x012D, |
|
icaron: 0x01D0, |
|
icircle: 0x24D8, |
|
icircumflex: 0x00EE, |
|
icyrillic: 0x0456, |
|
idblgrave: 0x0209, |
|
ideographearthcircle: 0x328F, |
|
ideographfirecircle: 0x328B, |
|
ideographicallianceparen: 0x323F, |
|
ideographiccallparen: 0x323A, |
|
ideographiccentrecircle: 0x32A5, |
|
ideographicclose: 0x3006, |
|
ideographiccomma: 0x3001, |
|
ideographiccommaleft: 0xFF64, |
|
ideographiccongratulationparen: 0x3237, |
|
ideographiccorrectcircle: 0x32A3, |
|
ideographicearthparen: 0x322F, |
|
ideographicenterpriseparen: 0x323D, |
|
ideographicexcellentcircle: 0x329D, |
|
ideographicfestivalparen: 0x3240, |
|
ideographicfinancialcircle: 0x3296, |
|
ideographicfinancialparen: 0x3236, |
|
ideographicfireparen: 0x322B, |
|
ideographichaveparen: 0x3232, |
|
ideographichighcircle: 0x32A4, |
|
ideographiciterationmark: 0x3005, |
|
ideographiclaborcircle: 0x3298, |
|
ideographiclaborparen: 0x3238, |
|
ideographicleftcircle: 0x32A7, |
|
ideographiclowcircle: 0x32A6, |
|
ideographicmedicinecircle: 0x32A9, |
|
ideographicmetalparen: 0x322E, |
|
ideographicmoonparen: 0x322A, |
|
ideographicnameparen: 0x3234, |
|
ideographicperiod: 0x3002, |
|
ideographicprintcircle: 0x329E, |
|
ideographicreachparen: 0x3243, |
|
ideographicrepresentparen: 0x3239, |
|
ideographicresourceparen: 0x323E, |
|
ideographicrightcircle: 0x32A8, |
|
ideographicsecretcircle: 0x3299, |
|
ideographicselfparen: 0x3242, |
|
ideographicsocietyparen: 0x3233, |
|
ideographicspace: 0x3000, |
|
ideographicspecialparen: 0x3235, |
|
ideographicstockparen: 0x3231, |
|
ideographicstudyparen: 0x323B, |
|
ideographicsunparen: 0x3230, |
|
ideographicsuperviseparen: 0x323C, |
|
ideographicwaterparen: 0x322C, |
|
ideographicwoodparen: 0x322D, |
|
ideographiczero: 0x3007, |
|
ideographmetalcircle: 0x328E, |
|
ideographmooncircle: 0x328A, |
|
ideographnamecircle: 0x3294, |
|
ideographsuncircle: 0x3290, |
|
ideographwatercircle: 0x328C, |
|
ideographwoodcircle: 0x328D, |
|
ideva: 0x0907, |
|
idieresis: 0x00EF, |
|
idieresisacute: 0x1E2F, |
|
idieresiscyrillic: 0x04E5, |
|
idotbelow: 0x1ECB, |
|
iebrevecyrillic: 0x04D7, |
|
iecyrillic: 0x0435, |
|
ieungacirclekorean: 0x3275, |
|
ieungaparenkorean: 0x3215, |
|
ieungcirclekorean: 0x3267, |
|
ieungkorean: 0x3147, |
|
ieungparenkorean: 0x3207, |
|
igrave: 0x00EC, |
|
igujarati: 0x0A87, |
|
igurmukhi: 0x0A07, |
|
ihiragana: 0x3044, |
|
ihookabove: 0x1EC9, |
|
iibengali: 0x0988, |
|
iicyrillic: 0x0438, |
|
iideva: 0x0908, |
|
iigujarati: 0x0A88, |
|
iigurmukhi: 0x0A08, |
|
iimatragurmukhi: 0x0A40, |
|
iinvertedbreve: 0x020B, |
|
iishortcyrillic: 0x0439, |
|
iivowelsignbengali: 0x09C0, |
|
iivowelsigndeva: 0x0940, |
|
iivowelsigngujarati: 0x0AC0, |
|
ij: 0x0133, |
|
ikatakana: 0x30A4, |
|
ikatakanahalfwidth: 0xFF72, |
|
ikorean: 0x3163, |
|
ilde: 0x02DC, |
|
iluyhebrew: 0x05AC, |
|
imacron: 0x012B, |
|
imacroncyrillic: 0x04E3, |
|
imageorapproximatelyequal: 0x2253, |
|
imatragurmukhi: 0x0A3F, |
|
imonospace: 0xFF49, |
|
increment: 0x2206, |
|
infinity: 0x221E, |
|
iniarmenian: 0x056B, |
|
integral: 0x222B, |
|
integralbottom: 0x2321, |
|
integralbt: 0x2321, |
|
integralex: 0xF8F5, |
|
integraltop: 0x2320, |
|
integraltp: 0x2320, |
|
intersection: 0x2229, |
|
intisquare: 0x3305, |
|
invbullet: 0x25D8, |
|
invcircle: 0x25D9, |
|
invsmileface: 0x263B, |
|
iocyrillic: 0x0451, |
|
iogonek: 0x012F, |
|
iota: 0x03B9, |
|
iotadieresis: 0x03CA, |
|
iotadieresistonos: 0x0390, |
|
iotalatin: 0x0269, |
|
iotatonos: 0x03AF, |
|
iparen: 0x24A4, |
|
irigurmukhi: 0x0A72, |
|
ismallhiragana: 0x3043, |
|
ismallkatakana: 0x30A3, |
|
ismallkatakanahalfwidth: 0xFF68, |
|
issharbengali: 0x09FA, |
|
istroke: 0x0268, |
|
isuperior: 0xF6ED, |
|
iterationhiragana: 0x309D, |
|
iterationkatakana: 0x30FD, |
|
itilde: 0x0129, |
|
itildebelow: 0x1E2D, |
|
iubopomofo: 0x3129, |
|
iucyrillic: 0x044E, |
|
ivowelsignbengali: 0x09BF, |
|
ivowelsigndeva: 0x093F, |
|
ivowelsigngujarati: 0x0ABF, |
|
izhitsacyrillic: 0x0475, |
|
izhitsadblgravecyrillic: 0x0477, |
|
j: 0x006A, |
|
jaarmenian: 0x0571, |
|
jabengali: 0x099C, |
|
jadeva: 0x091C, |
|
jagujarati: 0x0A9C, |
|
jagurmukhi: 0x0A1C, |
|
jbopomofo: 0x3110, |
|
jcaron: 0x01F0, |
|
jcircle: 0x24D9, |
|
jcircumflex: 0x0135, |
|
jcrossedtail: 0x029D, |
|
jdotlessstroke: 0x025F, |
|
jecyrillic: 0x0458, |
|
jeemarabic: 0x062C, |
|
jeemfinalarabic: 0xFE9E, |
|
jeeminitialarabic: 0xFE9F, |
|
jeemmedialarabic: 0xFEA0, |
|
jeharabic: 0x0698, |
|
jehfinalarabic: 0xFB8B, |
|
jhabengali: 0x099D, |
|
jhadeva: 0x091D, |
|
jhagujarati: 0x0A9D, |
|
jhagurmukhi: 0x0A1D, |
|
jheharmenian: 0x057B, |
|
jis: 0x3004, |
|
jmonospace: 0xFF4A, |
|
jparen: 0x24A5, |
|
jsuperior: 0x02B2, |
|
k: 0x006B, |
|
kabashkircyrillic: 0x04A1, |
|
kabengali: 0x0995, |
|
kacute: 0x1E31, |
|
kacyrillic: 0x043A, |
|
kadescendercyrillic: 0x049B, |
|
kadeva: 0x0915, |
|
kaf: 0x05DB, |
|
kafarabic: 0x0643, |
|
kafdagesh: 0xFB3B, |
|
kafdageshhebrew: 0xFB3B, |
|
kaffinalarabic: 0xFEDA, |
|
kafhebrew: 0x05DB, |
|
kafinitialarabic: 0xFEDB, |
|
kafmedialarabic: 0xFEDC, |
|
kafrafehebrew: 0xFB4D, |
|
kagujarati: 0x0A95, |
|
kagurmukhi: 0x0A15, |
|
kahiragana: 0x304B, |
|
kahookcyrillic: 0x04C4, |
|
kakatakana: 0x30AB, |
|
kakatakanahalfwidth: 0xFF76, |
|
kappa: 0x03BA, |
|
kappasymbolgreek: 0x03F0, |
|
kapyeounmieumkorean: 0x3171, |
|
kapyeounphieuphkorean: 0x3184, |
|
kapyeounpieupkorean: 0x3178, |
|
kapyeounssangpieupkorean: 0x3179, |
|
karoriisquare: 0x330D, |
|
kashidaautoarabic: 0x0640, |
|
kashidaautonosidebearingarabic: 0x0640, |
|
kasmallkatakana: 0x30F5, |
|
kasquare: 0x3384, |
|
kasraarabic: 0x0650, |
|
kasratanarabic: 0x064D, |
|
kastrokecyrillic: 0x049F, |
|
katahiraprolongmarkhalfwidth: 0xFF70, |
|
kaverticalstrokecyrillic: 0x049D, |
|
kbopomofo: 0x310E, |
|
kcalsquare: 0x3389, |
|
kcaron: 0x01E9, |
|
kcedilla: 0x0137, |
|
kcircle: 0x24DA, |
|
kcommaaccent: 0x0137, |
|
kdotbelow: 0x1E33, |
|
keharmenian: 0x0584, |
|
kehiragana: 0x3051, |
|
kekatakana: 0x30B1, |
|
kekatakanahalfwidth: 0xFF79, |
|
kenarmenian: 0x056F, |
|
kesmallkatakana: 0x30F6, |
|
kgreenlandic: 0x0138, |
|
khabengali: 0x0996, |
|
khacyrillic: 0x0445, |
|
khadeva: 0x0916, |
|
khagujarati: 0x0A96, |
|
khagurmukhi: 0x0A16, |
|
khaharabic: 0x062E, |
|
khahfinalarabic: 0xFEA6, |
|
khahinitialarabic: 0xFEA7, |
|
khahmedialarabic: 0xFEA8, |
|
kheicoptic: 0x03E7, |
|
khhadeva: 0x0959, |
|
khhagurmukhi: 0x0A59, |
|
khieukhacirclekorean: 0x3278, |
|
khieukhaparenkorean: 0x3218, |
|
khieukhcirclekorean: 0x326A, |
|
khieukhkorean: 0x314B, |
|
khieukhparenkorean: 0x320A, |
|
khokhaithai: 0x0E02, |
|
khokhonthai: 0x0E05, |
|
khokhuatthai: 0x0E03, |
|
khokhwaithai: 0x0E04, |
|
khomutthai: 0x0E5B, |
|
khook: 0x0199, |
|
khorakhangthai: 0x0E06, |
|
khzsquare: 0x3391, |
|
kihiragana: 0x304D, |
|
kikatakana: 0x30AD, |
|
kikatakanahalfwidth: 0xFF77, |
|
kiroguramusquare: 0x3315, |
|
kiromeetorusquare: 0x3316, |
|
kirosquare: 0x3314, |
|
kiyeokacirclekorean: 0x326E, |
|
kiyeokaparenkorean: 0x320E, |
|
kiyeokcirclekorean: 0x3260, |
|
kiyeokkorean: 0x3131, |
|
kiyeokparenkorean: 0x3200, |
|
kiyeoksioskorean: 0x3133, |
|
kjecyrillic: 0x045C, |
|
klinebelow: 0x1E35, |
|
klsquare: 0x3398, |
|
kmcubedsquare: 0x33A6, |
|
kmonospace: 0xFF4B, |
|
kmsquaredsquare: 0x33A2, |
|
kohiragana: 0x3053, |
|
kohmsquare: 0x33C0, |
|
kokaithai: 0x0E01, |
|
kokatakana: 0x30B3, |
|
kokatakanahalfwidth: 0xFF7A, |
|
kooposquare: 0x331E, |
|
koppacyrillic: 0x0481, |
|
koreanstandardsymbol: 0x327F, |
|
koroniscmb: 0x0343, |
|
kparen: 0x24A6, |
|
kpasquare: 0x33AA, |
|
ksicyrillic: 0x046F, |
|
ktsquare: 0x33CF, |
|
kturned: 0x029E, |
|
kuhiragana: 0x304F, |
|
kukatakana: 0x30AF, |
|
kukatakanahalfwidth: 0xFF78, |
|
kvsquare: 0x33B8, |
|
kwsquare: 0x33BE, |
|
l: 0x006C, |
|
labengali: 0x09B2, |
|
lacute: 0x013A, |
|
ladeva: 0x0932, |
|
lagujarati: 0x0AB2, |
|
lagurmukhi: 0x0A32, |
|
lakkhangyaothai: 0x0E45, |
|
lamaleffinalarabic: 0xFEFC, |
|
lamalefhamzaabovefinalarabic: 0xFEF8, |
|
lamalefhamzaaboveisolatedarabic: 0xFEF7, |
|
lamalefhamzabelowfinalarabic: 0xFEFA, |
|
lamalefhamzabelowisolatedarabic: 0xFEF9, |
|
lamalefisolatedarabic: 0xFEFB, |
|
lamalefmaddaabovefinalarabic: 0xFEF6, |
|
lamalefmaddaaboveisolatedarabic: 0xFEF5, |
|
lamarabic: 0x0644, |
|
lambda: 0x03BB, |
|
lambdastroke: 0x019B, |
|
lamed: 0x05DC, |
|
lameddagesh: 0xFB3C, |
|
lameddageshhebrew: 0xFB3C, |
|
lamedhebrew: 0x05DC, |
|
lamfinalarabic: 0xFEDE, |
|
lamhahinitialarabic: 0xFCCA, |
|
laminitialarabic: 0xFEDF, |
|
lamjeeminitialarabic: 0xFCC9, |
|
lamkhahinitialarabic: 0xFCCB, |
|
lamlamhehisolatedarabic: 0xFDF2, |
|
lammedialarabic: 0xFEE0, |
|
lammeemhahinitialarabic: 0xFD88, |
|
lammeeminitialarabic: 0xFCCC, |
|
largecircle: 0x25EF, |
|
lbar: 0x019A, |
|
lbelt: 0x026C, |
|
lbopomofo: 0x310C, |
|
lcaron: 0x013E, |
|
lcedilla: 0x013C, |
|
lcircle: 0x24DB, |
|
lcircumflexbelow: 0x1E3D, |
|
lcommaaccent: 0x013C, |
|
ldot: 0x0140, |
|
ldotaccent: 0x0140, |
|
ldotbelow: 0x1E37, |
|
ldotbelowmacron: 0x1E39, |
|
leftangleabovecmb: 0x031A, |
|
lefttackbelowcmb: 0x0318, |
|
less: 0x003C, |
|
lessequal: 0x2264, |
|
lessequalorgreater: 0x22DA, |
|
lessmonospace: 0xFF1C, |
|
lessorequivalent: 0x2272, |
|
lessorgreater: 0x2276, |
|
lessoverequal: 0x2266, |
|
lesssmall: 0xFE64, |
|
lezh: 0x026E, |
|
lfblock: 0x258C, |
|
lhookretroflex: 0x026D, |
|
lira: 0x20A4, |
|
liwnarmenian: 0x056C, |
|
lj: 0x01C9, |
|
ljecyrillic: 0x0459, |
|
ll: 0xF6C0, |
|
lladeva: 0x0933, |
|
llagujarati: 0x0AB3, |
|
llinebelow: 0x1E3B, |
|
llladeva: 0x0934, |
|
llvocalicbengali: 0x09E1, |
|
llvocalicdeva: 0x0961, |
|
llvocalicvowelsignbengali: 0x09E3, |
|
llvocalicvowelsigndeva: 0x0963, |
|
lmiddletilde: 0x026B, |
|
lmonospace: 0xFF4C, |
|
lmsquare: 0x33D0, |
|
lochulathai: 0x0E2C, |
|
logicaland: 0x2227, |
|
logicalnot: 0x00AC, |
|
logicalnotreversed: 0x2310, |
|
logicalor: 0x2228, |
|
lolingthai: 0x0E25, |
|
longs: 0x017F, |
|
lowlinecenterline: 0xFE4E, |
|
lowlinecmb: 0x0332, |
|
lowlinedashed: 0xFE4D, |
|
lozenge: 0x25CA, |
|
lparen: 0x24A7, |
|
lslash: 0x0142, |
|
lsquare: 0x2113, |
|
lsuperior: 0xF6EE, |
|
ltshade: 0x2591, |
|
luthai: 0x0E26, |
|
lvocalicbengali: 0x098C, |
|
lvocalicdeva: 0x090C, |
|
lvocalicvowelsignbengali: 0x09E2, |
|
lvocalicvowelsigndeva: 0x0962, |
|
lxsquare: 0x33D3, |
|
m: 0x006D, |
|
mabengali: 0x09AE, |
|
macron: 0x00AF, |
|
macronbelowcmb: 0x0331, |
|
macroncmb: 0x0304, |
|
macronlowmod: 0x02CD, |
|
macronmonospace: 0xFFE3, |
|
macute: 0x1E3F, |
|
madeva: 0x092E, |
|
magujarati: 0x0AAE, |
|
magurmukhi: 0x0A2E, |
|
mahapakhhebrew: 0x05A4, |
|
mahapakhlefthebrew: 0x05A4, |
|
mahiragana: 0x307E, |
|
maichattawalowleftthai: 0xF895, |
|
maichattawalowrightthai: 0xF894, |
|
maichattawathai: 0x0E4B, |
|
maichattawaupperleftthai: 0xF893, |
|
maieklowleftthai: 0xF88C, |
|
maieklowrightthai: 0xF88B, |
|
maiekthai: 0x0E48, |
|
maiekupperleftthai: 0xF88A, |
|
maihanakatleftthai: 0xF884, |
|
maihanakatthai: 0x0E31, |
|
maitaikhuleftthai: 0xF889, |
|
maitaikhuthai: 0x0E47, |
|
maitholowleftthai: 0xF88F, |
|
maitholowrightthai: 0xF88E, |
|
maithothai: 0x0E49, |
|
maithoupperleftthai: 0xF88D, |
|
maitrilowleftthai: 0xF892, |
|
maitrilowrightthai: 0xF891, |
|
maitrithai: 0x0E4A, |
|
maitriupperleftthai: 0xF890, |
|
maiyamokthai: 0x0E46, |
|
makatakana: 0x30DE, |
|
makatakanahalfwidth: 0xFF8F, |
|
male: 0x2642, |
|
mansyonsquare: 0x3347, |
|
maqafhebrew: 0x05BE, |
|
mars: 0x2642, |
|
masoracirclehebrew: 0x05AF, |
|
masquare: 0x3383, |
|
mbopomofo: 0x3107, |
|
mbsquare: 0x33D4, |
|
mcircle: 0x24DC, |
|
mcubedsquare: 0x33A5, |
|
mdotaccent: 0x1E41, |
|
mdotbelow: 0x1E43, |
|
meemarabic: 0x0645, |
|
meemfinalarabic: 0xFEE2, |
|
meeminitialarabic: 0xFEE3, |
|
meemmedialarabic: 0xFEE4, |
|
meemmeeminitialarabic: 0xFCD1, |
|
meemmeemisolatedarabic: 0xFC48, |
|
meetorusquare: 0x334D, |
|
mehiragana: 0x3081, |
|
meizierasquare: 0x337E, |
|
mekatakana: 0x30E1, |
|
mekatakanahalfwidth: 0xFF92, |
|
mem: 0x05DE, |
|
memdagesh: 0xFB3E, |
|
memdageshhebrew: 0xFB3E, |
|
memhebrew: 0x05DE, |
|
menarmenian: 0x0574, |
|
merkhahebrew: 0x05A5, |
|
merkhakefulahebrew: 0x05A6, |
|
merkhakefulalefthebrew: 0x05A6, |
|
merkhalefthebrew: 0x05A5, |
|
mhook: 0x0271, |
|
mhzsquare: 0x3392, |
|
middledotkatakanahalfwidth: 0xFF65, |
|
middot: 0x00B7, |
|
mieumacirclekorean: 0x3272, |
|
mieumaparenkorean: 0x3212, |
|
mieumcirclekorean: 0x3264, |
|
mieumkorean: 0x3141, |
|
mieumpansioskorean: 0x3170, |
|
mieumparenkorean: 0x3204, |
|
mieumpieupkorean: 0x316E, |
|
mieumsioskorean: 0x316F, |
|
mihiragana: 0x307F, |
|
mikatakana: 0x30DF, |
|
mikatakanahalfwidth: 0xFF90, |
|
minus: 0x2212, |
|
minusbelowcmb: 0x0320, |
|
minuscircle: 0x2296, |
|
minusmod: 0x02D7, |
|
minusplus: 0x2213, |
|
minute: 0x2032, |
|
miribaarusquare: 0x334A, |
|
mirisquare: 0x3349, |
|
mlonglegturned: 0x0270, |
|
mlsquare: 0x3396, |
|
mmcubedsquare: 0x33A3, |
|
mmonospace: 0xFF4D, |
|
mmsquaredsquare: 0x339F, |
|
mohiragana: 0x3082, |
|
mohmsquare: 0x33C1, |
|
mokatakana: 0x30E2, |
|
mokatakanahalfwidth: 0xFF93, |
|
molsquare: 0x33D6, |
|
momathai: 0x0E21, |
|
moverssquare: 0x33A7, |
|
moverssquaredsquare: 0x33A8, |
|
mparen: 0x24A8, |
|
mpasquare: 0x33AB, |
|
mssquare: 0x33B3, |
|
msuperior: 0xF6EF, |
|
mturned: 0x026F, |
|
mu: 0x00B5, |
|
mu1: 0x00B5, |
|
muasquare: 0x3382, |
|
muchgreater: 0x226B, |
|
muchless: 0x226A, |
|
mufsquare: 0x338C, |
|
mugreek: 0x03BC, |
|
mugsquare: 0x338D, |
|
muhiragana: 0x3080, |
|
mukatakana: 0x30E0, |
|
mukatakanahalfwidth: 0xFF91, |
|
mulsquare: 0x3395, |
|
multiply: 0x00D7, |
|
mumsquare: 0x339B, |
|
munahhebrew: 0x05A3, |
|
munahlefthebrew: 0x05A3, |
|
musicalnote: 0x266A, |
|
musicalnotedbl: 0x266B, |
|
musicflatsign: 0x266D, |
|
musicsharpsign: 0x266F, |
|
mussquare: 0x33B2, |
|
muvsquare: 0x33B6, |
|
muwsquare: 0x33BC, |
|
mvmegasquare: 0x33B9, |
|
mvsquare: 0x33B7, |
|
mwmegasquare: 0x33BF, |
|
mwsquare: 0x33BD, |
|
n: 0x006E, |
|
nabengali: 0x09A8, |
|
nabla: 0x2207, |
|
nacute: 0x0144, |
|
nadeva: 0x0928, |
|
nagujarati: 0x0AA8, |
|
nagurmukhi: 0x0A28, |
|
nahiragana: 0x306A, |
|
nakatakana: 0x30CA, |
|
nakatakanahalfwidth: 0xFF85, |
|
napostrophe: 0x0149, |
|
nasquare: 0x3381, |
|
nbopomofo: 0x310B, |
|
nbspace: 0x00A0, |
|
ncaron: 0x0148, |
|
ncedilla: 0x0146, |
|
ncircle: 0x24DD, |
|
ncircumflexbelow: 0x1E4B, |
|
ncommaaccent: 0x0146, |
|
ndotaccent: 0x1E45, |
|
ndotbelow: 0x1E47, |
|
nehiragana: 0x306D, |
|
nekatakana: 0x30CD, |
|
nekatakanahalfwidth: 0xFF88, |
|
newsheqelsign: 0x20AA, |
|
nfsquare: 0x338B, |
|
ngabengali: 0x0999, |
|
ngadeva: 0x0919, |
|
ngagujarati: 0x0A99, |
|
ngagurmukhi: 0x0A19, |
|
ngonguthai: 0x0E07, |
|
nhiragana: 0x3093, |
|
nhookleft: 0x0272, |
|
nhookretroflex: 0x0273, |
|
nieunacirclekorean: 0x326F, |
|
nieunaparenkorean: 0x320F, |
|
nieuncieuckorean: 0x3135, |
|
nieuncirclekorean: 0x3261, |
|
nieunhieuhkorean: 0x3136, |
|
nieunkorean: 0x3134, |
|
nieunpansioskorean: 0x3168, |
|
nieunparenkorean: 0x3201, |
|
nieunsioskorean: 0x3167, |
|
nieuntikeutkorean: 0x3166, |
|
nihiragana: 0x306B, |
|
nikatakana: 0x30CB, |
|
nikatakanahalfwidth: 0xFF86, |
|
nikhahitleftthai: 0xF899, |
|
nikhahitthai: 0x0E4D, |
|
nine: 0x0039, |
|
ninearabic: 0x0669, |
|
ninebengali: 0x09EF, |
|
ninecircle: 0x2468, |
|
ninecircleinversesansserif: 0x2792, |
|
ninedeva: 0x096F, |
|
ninegujarati: 0x0AEF, |
|
ninegurmukhi: 0x0A6F, |
|
ninehackarabic: 0x0669, |
|
ninehangzhou: 0x3029, |
|
nineideographicparen: 0x3228, |
|
nineinferior: 0x2089, |
|
ninemonospace: 0xFF19, |
|
nineoldstyle: 0xF739, |
|
nineparen: 0x247C, |
|
nineperiod: 0x2490, |
|
ninepersian: 0x06F9, |
|
nineroman: 0x2178, |
|
ninesuperior: 0x2079, |
|
nineteencircle: 0x2472, |
|
nineteenparen: 0x2486, |
|
nineteenperiod: 0x249A, |
|
ninethai: 0x0E59, |
|
nj: 0x01CC, |
|
njecyrillic: 0x045A, |
|
nkatakana: 0x30F3, |
|
nkatakanahalfwidth: 0xFF9D, |
|
nlegrightlong: 0x019E, |
|
nlinebelow: 0x1E49, |
|
nmonospace: 0xFF4E, |
|
nmsquare: 0x339A, |
|
nnabengali: 0x09A3, |
|
nnadeva: 0x0923, |
|
nnagujarati: 0x0AA3, |
|
nnagurmukhi: 0x0A23, |
|
nnnadeva: 0x0929, |
|
nohiragana: 0x306E, |
|
nokatakana: 0x30CE, |
|
nokatakanahalfwidth: 0xFF89, |
|
nonbreakingspace: 0x00A0, |
|
nonenthai: 0x0E13, |
|
nonuthai: 0x0E19, |
|
noonarabic: 0x0646, |
|
noonfinalarabic: 0xFEE6, |
|
noonghunnaarabic: 0x06BA, |
|
noonghunnafinalarabic: 0xFB9F, |
|
nooninitialarabic: 0xFEE7, |
|
noonjeeminitialarabic: 0xFCD2, |
|
noonjeemisolatedarabic: 0xFC4B, |
|
noonmedialarabic: 0xFEE8, |
|
noonmeeminitialarabic: 0xFCD5, |
|
noonmeemisolatedarabic: 0xFC4E, |
|
noonnoonfinalarabic: 0xFC8D, |
|
notcontains: 0x220C, |
|
notelement: 0x2209, |
|
notelementof: 0x2209, |
|
notequal: 0x2260, |
|
notgreater: 0x226F, |
|
notgreaternorequal: 0x2271, |
|
notgreaternorless: 0x2279, |
|
notidentical: 0x2262, |
|
notless: 0x226E, |
|
notlessnorequal: 0x2270, |
|
notparallel: 0x2226, |
|
notprecedes: 0x2280, |
|
notsubset: 0x2284, |
|
notsucceeds: 0x2281, |
|
notsuperset: 0x2285, |
|
nowarmenian: 0x0576, |
|
nparen: 0x24A9, |
|
nssquare: 0x33B1, |
|
nsuperior: 0x207F, |
|
ntilde: 0x00F1, |
|
nu: 0x03BD, |
|
nuhiragana: 0x306C, |
|
nukatakana: 0x30CC, |
|
nukatakanahalfwidth: 0xFF87, |
|
nuktabengali: 0x09BC, |
|
nuktadeva: 0x093C, |
|
nuktagujarati: 0x0ABC, |
|
nuktagurmukhi: 0x0A3C, |
|
numbersign: 0x0023, |
|
numbersignmonospace: 0xFF03, |
|
numbersignsmall: 0xFE5F, |
|
numeralsigngreek: 0x0374, |
|
numeralsignlowergreek: 0x0375, |
|
numero: 0x2116, |
|
nun: 0x05E0, |
|
nundagesh: 0xFB40, |
|
nundageshhebrew: 0xFB40, |
|
nunhebrew: 0x05E0, |
|
nvsquare: 0x33B5, |
|
nwsquare: 0x33BB, |
|
nyabengali: 0x099E, |
|
nyadeva: 0x091E, |
|
nyagujarati: 0x0A9E, |
|
nyagurmukhi: 0x0A1E, |
|
o: 0x006F, |
|
oacute: 0x00F3, |
|
oangthai: 0x0E2D, |
|
obarred: 0x0275, |
|
obarredcyrillic: 0x04E9, |
|
obarreddieresiscyrillic: 0x04EB, |
|
obengali: 0x0993, |
|
obopomofo: 0x311B, |
|
obreve: 0x014F, |
|
ocandradeva: 0x0911, |
|
ocandragujarati: 0x0A91, |
|
ocandravowelsigndeva: 0x0949, |
|
ocandravowelsigngujarati: 0x0AC9, |
|
ocaron: 0x01D2, |
|
ocircle: 0x24DE, |
|
ocircumflex: 0x00F4, |
|
ocircumflexacute: 0x1ED1, |
|
ocircumflexdotbelow: 0x1ED9, |
|
ocircumflexgrave: 0x1ED3, |
|
ocircumflexhookabove: 0x1ED5, |
|
ocircumflextilde: 0x1ED7, |
|
ocyrillic: 0x043E, |
|
odblacute: 0x0151, |
|
odblgrave: 0x020D, |
|
odeva: 0x0913, |
|
odieresis: 0x00F6, |
|
odieresiscyrillic: 0x04E7, |
|
odotbelow: 0x1ECD, |
|
oe: 0x0153, |
|
oekorean: 0x315A, |
|
ogonek: 0x02DB, |
|
ogonekcmb: 0x0328, |
|
ograve: 0x00F2, |
|
ogujarati: 0x0A93, |
|
oharmenian: 0x0585, |
|
ohiragana: 0x304A, |
|
ohookabove: 0x1ECF, |
|
ohorn: 0x01A1, |
|
ohornacute: 0x1EDB, |
|
ohorndotbelow: 0x1EE3, |
|
ohorngrave: 0x1EDD, |
|
ohornhookabove: 0x1EDF, |
|
ohorntilde: 0x1EE1, |
|
ohungarumlaut: 0x0151, |
|
oi: 0x01A3, |
|
oinvertedbreve: 0x020F, |
|
okatakana: 0x30AA, |
|
okatakanahalfwidth: 0xFF75, |
|
okorean: 0x3157, |
|
olehebrew: 0x05AB, |
|
omacron: 0x014D, |
|
omacronacute: 0x1E53, |
|
omacrongrave: 0x1E51, |
|
omdeva: 0x0950, |
|
omega: 0x03C9, |
|
omega1: 0x03D6, |
|
omegacyrillic: 0x0461, |
|
omegalatinclosed: 0x0277, |
|
omegaroundcyrillic: 0x047B, |
|
omegatitlocyrillic: 0x047D, |
|
omegatonos: 0x03CE, |
|
omgujarati: 0x0AD0, |
|
omicron: 0x03BF, |
|
omicrontonos: 0x03CC, |
|
omonospace: 0xFF4F, |
|
one: 0x0031, |
|
onearabic: 0x0661, |
|
onebengali: 0x09E7, |
|
onecircle: 0x2460, |
|
onecircleinversesansserif: 0x278A, |
|
onedeva: 0x0967, |
|
onedotenleader: 0x2024, |
|
oneeighth: 0x215B, |
|
onefitted: 0xF6DC, |
|
onegujarati: 0x0AE7, |
|
onegurmukhi: 0x0A67, |
|
onehackarabic: 0x0661, |
|
onehalf: 0x00BD, |
|
onehangzhou: 0x3021, |
|
oneideographicparen: 0x3220, |
|
oneinferior: 0x2081, |
|
onemonospace: 0xFF11, |
|
onenumeratorbengali: 0x09F4, |
|
oneoldstyle: 0xF731, |
|
oneparen: 0x2474, |
|
oneperiod: 0x2488, |
|
onepersian: 0x06F1, |
|
onequarter: 0x00BC, |
|
oneroman: 0x2170, |
|
onesuperior: 0x00B9, |
|
onethai: 0x0E51, |
|
onethird: 0x2153, |
|
oogonek: 0x01EB, |
|
oogonekmacron: 0x01ED, |
|
oogurmukhi: 0x0A13, |
|
oomatragurmukhi: 0x0A4B, |
|
oopen: 0x0254, |
|
oparen: 0x24AA, |
|
openbullet: 0x25E6, |
|
option: 0x2325, |
|
ordfeminine: 0x00AA, |
|
ordmasculine: 0x00BA, |
|
orthogonal: 0x221F, |
|
oshortdeva: 0x0912, |
|
oshortvowelsigndeva: 0x094A, |
|
oslash: 0x00F8, |
|
oslashacute: 0x01FF, |
|
osmallhiragana: 0x3049, |
|
osmallkatakana: 0x30A9, |
|
osmallkatakanahalfwidth: 0xFF6B, |
|
ostrokeacute: 0x01FF, |
|
osuperior: 0xF6F0, |
|
otcyrillic: 0x047F, |
|
otilde: 0x00F5, |
|
otildeacute: 0x1E4D, |
|
otildedieresis: 0x1E4F, |
|
oubopomofo: 0x3121, |
|
overline: 0x203E, |
|
overlinecenterline: 0xFE4A, |
|
overlinecmb: 0x0305, |
|
overlinedashed: 0xFE49, |
|
overlinedblwavy: 0xFE4C, |
|
overlinewavy: 0xFE4B, |
|
overscore: 0x00AF, |
|
ovowelsignbengali: 0x09CB, |
|
ovowelsigndeva: 0x094B, |
|
ovowelsigngujarati: 0x0ACB, |
|
p: 0x0070, |
|
paampssquare: 0x3380, |
|
paasentosquare: 0x332B, |
|
pabengali: 0x09AA, |
|
pacute: 0x1E55, |
|
padeva: 0x092A, |
|
pagedown: 0x21DF, |
|
pageup: 0x21DE, |
|
pagujarati: 0x0AAA, |
|
pagurmukhi: 0x0A2A, |
|
pahiragana: 0x3071, |
|
paiyannoithai: 0x0E2F, |
|
pakatakana: 0x30D1, |
|
palatalizationcyrilliccmb: 0x0484, |
|
palochkacyrillic: 0x04C0, |
|
pansioskorean: 0x317F, |
|
paragraph: 0x00B6, |
|
parallel: 0x2225, |
|
parenleft: 0x0028, |
|
parenleftaltonearabic: 0xFD3E, |
|
parenleftbt: 0xF8ED, |
|
parenleftex: 0xF8EC, |
|
parenleftinferior: 0x208D, |
|
parenleftmonospace: 0xFF08, |
|
parenleftsmall: 0xFE59, |
|
parenleftsuperior: 0x207D, |
|
parenlefttp: 0xF8EB, |
|
parenleftvertical: 0xFE35, |
|
parenright: 0x0029, |
|
parenrightaltonearabic: 0xFD3F, |
|
parenrightbt: 0xF8F8, |
|
parenrightex: 0xF8F7, |
|
parenrightinferior: 0x208E, |
|
parenrightmonospace: 0xFF09, |
|
parenrightsmall: 0xFE5A, |
|
parenrightsuperior: 0x207E, |
|
parenrighttp: 0xF8F6, |
|
parenrightvertical: 0xFE36, |
|
partialdiff: 0x2202, |
|
paseqhebrew: 0x05C0, |
|
pashtahebrew: 0x0599, |
|
pasquare: 0x33A9, |
|
patah: 0x05B7, |
|
patah11: 0x05B7, |
|
patah1d: 0x05B7, |
|
patah2a: 0x05B7, |
|
patahhebrew: 0x05B7, |
|
patahnarrowhebrew: 0x05B7, |
|
patahquarterhebrew: 0x05B7, |
|
patahwidehebrew: 0x05B7, |
|
pazerhebrew: 0x05A1, |
|
pbopomofo: 0x3106, |
|
pcircle: 0x24DF, |
|
pdotaccent: 0x1E57, |
|
pe: 0x05E4, |
|
pecyrillic: 0x043F, |
|
pedagesh: 0xFB44, |
|
pedageshhebrew: 0xFB44, |
|
peezisquare: 0x333B, |
|
pefinaldageshhebrew: 0xFB43, |
|
peharabic: 0x067E, |
|
peharmenian: 0x057A, |
|
pehebrew: 0x05E4, |
|
pehfinalarabic: 0xFB57, |
|
pehinitialarabic: 0xFB58, |
|
pehiragana: 0x307A, |
|
pehmedialarabic: 0xFB59, |
|
pekatakana: 0x30DA, |
|
pemiddlehookcyrillic: 0x04A7, |
|
perafehebrew: 0xFB4E, |
|
percent: 0x0025, |
|
percentarabic: 0x066A, |
|
percentmonospace: 0xFF05, |
|
percentsmall: 0xFE6A, |
|
period: 0x002E, |
|
periodarmenian: 0x0589, |
|
periodcentered: 0x00B7, |
|
periodhalfwidth: 0xFF61, |
|
periodinferior: 0xF6E7, |
|
periodmonospace: 0xFF0E, |
|
periodsmall: 0xFE52, |
|
periodsuperior: 0xF6E8, |
|
perispomenigreekcmb: 0x0342, |
|
perpendicular: 0x22A5, |
|
perthousand: 0x2030, |
|
peseta: 0x20A7, |
|
pfsquare: 0x338A, |
|
phabengali: 0x09AB, |
|
phadeva: 0x092B, |
|
phagujarati: 0x0AAB, |
|
phagurmukhi: 0x0A2B, |
|
phi: 0x03C6, |
|
phi1: 0x03D5, |
|
phieuphacirclekorean: 0x327A, |
|
phieuphaparenkorean: 0x321A, |
|
phieuphcirclekorean: 0x326C, |
|
phieuphkorean: 0x314D, |
|
phieuphparenkorean: 0x320C, |
|
philatin: 0x0278, |
|
phinthuthai: 0x0E3A, |
|
phisymbolgreek: 0x03D5, |
|
phook: 0x01A5, |
|
phophanthai: 0x0E1E, |
|
phophungthai: 0x0E1C, |
|
phosamphaothai: 0x0E20, |
|
pi: 0x03C0, |
|
pieupacirclekorean: 0x3273, |
|
pieupaparenkorean: 0x3213, |
|
pieupcieuckorean: 0x3176, |
|
pieupcirclekorean: 0x3265, |
|
pieupkiyeokkorean: 0x3172, |
|
pieupkorean: 0x3142, |
|
pieupparenkorean: 0x3205, |
|
pieupsioskiyeokkorean: 0x3174, |
|
pieupsioskorean: 0x3144, |
|
pieupsiostikeutkorean: 0x3175, |
|
pieupthieuthkorean: 0x3177, |
|
pieuptikeutkorean: 0x3173, |
|
pihiragana: 0x3074, |
|
pikatakana: 0x30D4, |
|
pisymbolgreek: 0x03D6, |
|
piwrarmenian: 0x0583, |
|
plus: 0x002B, |
|
plusbelowcmb: 0x031F, |
|
pluscircle: 0x2295, |
|
plusminus: 0x00B1, |
|
plusmod: 0x02D6, |
|
plusmonospace: 0xFF0B, |
|
plussmall: 0xFE62, |
|
plussuperior: 0x207A, |
|
pmonospace: 0xFF50, |
|
pmsquare: 0x33D8, |
|
pohiragana: 0x307D, |
|
pointingindexdownwhite: 0x261F, |
|
pointingindexleftwhite: 0x261C, |
|
pointingindexrightwhite: 0x261E, |
|
pointingindexupwhite: 0x261D, |
|
pokatakana: 0x30DD, |
|
poplathai: 0x0E1B, |
|
postalmark: 0x3012, |
|
postalmarkface: 0x3020, |
|
pparen: 0x24AB, |
|
precedes: 0x227A, |
|
prescription: 0x211E, |
|
primemod: 0x02B9, |
|
primereversed: 0x2035, |
|
product: 0x220F, |
|
projective: 0x2305, |
|
prolongedkana: 0x30FC, |
|
propellor: 0x2318, |
|
propersubset: 0x2282, |
|
propersuperset: 0x2283, |
|
proportion: 0x2237, |
|
proportional: 0x221D, |
|
psi: 0x03C8, |
|
psicyrillic: 0x0471, |
|
psilipneumatacyrilliccmb: 0x0486, |
|
pssquare: 0x33B0, |
|
puhiragana: 0x3077, |
|
pukatakana: 0x30D7, |
|
pvsquare: 0x33B4, |
|
pwsquare: 0x33BA, |
|
q: 0x0071, |
|
qadeva: 0x0958, |
|
qadmahebrew: 0x05A8, |
|
qafarabic: 0x0642, |
|
qaffinalarabic: 0xFED6, |
|
qafinitialarabic: 0xFED7, |
|
qafmedialarabic: 0xFED8, |
|
qamats: 0x05B8, |
|
qamats10: 0x05B8, |
|
qamats1a: 0x05B8, |
|
qamats1c: 0x05B8, |
|
qamats27: 0x05B8, |
|
qamats29: 0x05B8, |
|
qamats33: 0x05B8, |
|
qamatsde: 0x05B8, |
|
qamatshebrew: 0x05B8, |
|
qamatsnarrowhebrew: 0x05B8, |
|
qamatsqatanhebrew: 0x05B8, |
|
qamatsqatannarrowhebrew: 0x05B8, |
|
qamatsqatanquarterhebrew: 0x05B8, |
|
qamatsqatanwidehebrew: 0x05B8, |
|
qamatsquarterhebrew: 0x05B8, |
|
qamatswidehebrew: 0x05B8, |
|
qarneyparahebrew: 0x059F, |
|
qbopomofo: 0x3111, |
|
qcircle: 0x24E0, |
|
qhook: 0x02A0, |
|
qmonospace: 0xFF51, |
|
qof: 0x05E7, |
|
qofdagesh: 0xFB47, |
|
qofdageshhebrew: 0xFB47, |
|
qofhebrew: 0x05E7, |
|
qparen: 0x24AC, |
|
quarternote: 0x2669, |
|
qubuts: 0x05BB, |
|
qubuts18: 0x05BB, |
|
qubuts25: 0x05BB, |
|
qubuts31: 0x05BB, |
|
qubutshebrew: 0x05BB, |
|
qubutsnarrowhebrew: 0x05BB, |
|
qubutsquarterhebrew: 0x05BB, |
|
qubutswidehebrew: 0x05BB, |
|
question: 0x003F, |
|
questionarabic: 0x061F, |
|
questionarmenian: 0x055E, |
|
questiondown: 0x00BF, |
|
questiondownsmall: 0xF7BF, |
|
questiongreek: 0x037E, |
|
questionmonospace: 0xFF1F, |
|
questionsmall: 0xF73F, |
|
quotedbl: 0x0022, |
|
quotedblbase: 0x201E, |
|
quotedblleft: 0x201C, |
|
quotedblmonospace: 0xFF02, |
|
quotedblprime: 0x301E, |
|
quotedblprimereversed: 0x301D, |
|
quotedblright: 0x201D, |
|
quoteleft: 0x2018, |
|
quoteleftreversed: 0x201B, |
|
quotereversed: 0x201B, |
|
quoteright: 0x2019, |
|
quoterightn: 0x0149, |
|
quotesinglbase: 0x201A, |
|
quotesingle: 0x0027, |
|
quotesinglemonospace: 0xFF07, |
|
r: 0x0072, |
|
raarmenian: 0x057C, |
|
rabengali: 0x09B0, |
|
racute: 0x0155, |
|
radeva: 0x0930, |
|
radical: 0x221A, |
|
radicalex: 0xF8E5, |
|
radoverssquare: 0x33AE, |
|
radoverssquaredsquare: 0x33AF, |
|
radsquare: 0x33AD, |
|
rafe: 0x05BF, |
|
rafehebrew: 0x05BF, |
|
ragujarati: 0x0AB0, |
|
ragurmukhi: 0x0A30, |
|
rahiragana: 0x3089, |
|
rakatakana: 0x30E9, |
|
rakatakanahalfwidth: 0xFF97, |
|
ralowerdiagonalbengali: 0x09F1, |
|
ramiddlediagonalbengali: 0x09F0, |
|
ramshorn: 0x0264, |
|
ratio: 0x2236, |
|
rbopomofo: 0x3116, |
|
rcaron: 0x0159, |
|
rcedilla: 0x0157, |
|
rcircle: 0x24E1, |
|
rcommaaccent: 0x0157, |
|
rdblgrave: 0x0211, |
|
rdotaccent: 0x1E59, |
|
rdotbelow: 0x1E5B, |
|
rdotbelowmacron: 0x1E5D, |
|
referencemark: 0x203B, |
|
reflexsubset: 0x2286, |
|
reflexsuperset: 0x2287, |
|
registered: 0x00AE, |
|
registersans: 0xF8E8, |
|
registerserif: 0xF6DA, |
|
reharabic: 0x0631, |
|
reharmenian: 0x0580, |
|
rehfinalarabic: 0xFEAE, |
|
rehiragana: 0x308C, |
|
rekatakana: 0x30EC, |
|
rekatakanahalfwidth: 0xFF9A, |
|
resh: 0x05E8, |
|
reshdageshhebrew: 0xFB48, |
|
reshhebrew: 0x05E8, |
|
reversedtilde: 0x223D, |
|
reviahebrew: 0x0597, |
|
reviamugrashhebrew: 0x0597, |
|
revlogicalnot: 0x2310, |
|
rfishhook: 0x027E, |
|
rfishhookreversed: 0x027F, |
|
rhabengali: 0x09DD, |
|
rhadeva: 0x095D, |
|
rho: 0x03C1, |
|
rhook: 0x027D, |
|
rhookturned: 0x027B, |
|
rhookturnedsuperior: 0x02B5, |
|
rhosymbolgreek: 0x03F1, |
|
rhotichookmod: 0x02DE, |
|
rieulacirclekorean: 0x3271, |
|
rieulaparenkorean: 0x3211, |
|
rieulcirclekorean: 0x3263, |
|
rieulhieuhkorean: 0x3140, |
|
rieulkiyeokkorean: 0x313A, |
|
rieulkiyeoksioskorean: 0x3169, |
|
rieulkorean: 0x3139, |
|
rieulmieumkorean: 0x313B, |
|
rieulpansioskorean: 0x316C, |
|
rieulparenkorean: 0x3203, |
|
rieulphieuphkorean: 0x313F, |
|
rieulpieupkorean: 0x313C, |
|
rieulpieupsioskorean: 0x316B, |
|
rieulsioskorean: 0x313D, |
|
rieulthieuthkorean: 0x313E, |
|
rieultikeutkorean: 0x316A, |
|
rieulyeorinhieuhkorean: 0x316D, |
|
rightangle: 0x221F, |
|
righttackbelowcmb: 0x0319, |
|
righttriangle: 0x22BF, |
|
rihiragana: 0x308A, |
|
rikatakana: 0x30EA, |
|
rikatakanahalfwidth: 0xFF98, |
|
ring: 0x02DA, |
|
ringbelowcmb: 0x0325, |
|
ringcmb: 0x030A, |
|
ringhalfleft: 0x02BF, |
|
ringhalfleftarmenian: 0x0559, |
|
ringhalfleftbelowcmb: 0x031C, |
|
ringhalfleftcentered: 0x02D3, |
|
ringhalfright: 0x02BE, |
|
ringhalfrightbelowcmb: 0x0339, |
|
ringhalfrightcentered: 0x02D2, |
|
rinvertedbreve: 0x0213, |
|
rittorusquare: 0x3351, |
|
rlinebelow: 0x1E5F, |
|
rlongleg: 0x027C, |
|
rlonglegturned: 0x027A, |
|
rmonospace: 0xFF52, |
|
rohiragana: 0x308D, |
|
rokatakana: 0x30ED, |
|
rokatakanahalfwidth: 0xFF9B, |
|
roruathai: 0x0E23, |
|
rparen: 0x24AD, |
|
rrabengali: 0x09DC, |
|
rradeva: 0x0931, |
|
rragurmukhi: 0x0A5C, |
|
rreharabic: 0x0691, |
|
rrehfinalarabic: 0xFB8D, |
|
rrvocalicbengali: 0x09E0, |
|
rrvocalicdeva: 0x0960, |
|
rrvocalicgujarati: 0x0AE0, |
|
rrvocalicvowelsignbengali: 0x09C4, |
|
rrvocalicvowelsigndeva: 0x0944, |
|
rrvocalicvowelsigngujarati: 0x0AC4, |
|
rsuperior: 0xF6F1, |
|
rtblock: 0x2590, |
|
rturned: 0x0279, |
|
rturnedsuperior: 0x02B4, |
|
ruhiragana: 0x308B, |
|
rukatakana: 0x30EB, |
|
rukatakanahalfwidth: 0xFF99, |
|
rupeemarkbengali: 0x09F2, |
|
rupeesignbengali: 0x09F3, |
|
rupiah: 0xF6DD, |
|
ruthai: 0x0E24, |
|
rvocalicbengali: 0x098B, |
|
rvocalicdeva: 0x090B, |
|
rvocalicgujarati: 0x0A8B, |
|
rvocalicvowelsignbengali: 0x09C3, |
|
rvocalicvowelsigndeva: 0x0943, |
|
rvocalicvowelsigngujarati: 0x0AC3, |
|
s: 0x0073, |
|
sabengali: 0x09B8, |
|
sacute: 0x015B, |
|
sacutedotaccent: 0x1E65, |
|
sadarabic: 0x0635, |
|
sadeva: 0x0938, |
|
sadfinalarabic: 0xFEBA, |
|
sadinitialarabic: 0xFEBB, |
|
sadmedialarabic: 0xFEBC, |
|
sagujarati: 0x0AB8, |
|
sagurmukhi: 0x0A38, |
|
sahiragana: 0x3055, |
|
sakatakana: 0x30B5, |
|
sakatakanahalfwidth: 0xFF7B, |
|
sallallahoualayhewasallamarabic: 0xFDFA, |
|
samekh: 0x05E1, |
|
samekhdagesh: 0xFB41, |
|
samekhdageshhebrew: 0xFB41, |
|
samekhhebrew: 0x05E1, |
|
saraaathai: 0x0E32, |
|
saraaethai: 0x0E41, |
|
saraaimaimalaithai: 0x0E44, |
|
saraaimaimuanthai: 0x0E43, |
|
saraamthai: 0x0E33, |
|
saraathai: 0x0E30, |
|
saraethai: 0x0E40, |
|
saraiileftthai: 0xF886, |
|
saraiithai: 0x0E35, |
|
saraileftthai: 0xF885, |
|
saraithai: 0x0E34, |
|
saraothai: 0x0E42, |
|
saraueeleftthai: 0xF888, |
|
saraueethai: 0x0E37, |
|
saraueleftthai: 0xF887, |
|
sarauethai: 0x0E36, |
|
sarauthai: 0x0E38, |
|
sarauuthai: 0x0E39, |
|
sbopomofo: 0x3119, |
|
scaron: 0x0161, |
|
scarondotaccent: 0x1E67, |
|
scedilla: 0x015F, |
|
schwa: 0x0259, |
|
schwacyrillic: 0x04D9, |
|
schwadieresiscyrillic: 0x04DB, |
|
schwahook: 0x025A, |
|
scircle: 0x24E2, |
|
scircumflex: 0x015D, |
|
scommaaccent: 0x0219, |
|
sdotaccent: 0x1E61, |
|
sdotbelow: 0x1E63, |
|
sdotbelowdotaccent: 0x1E69, |
|
seagullbelowcmb: 0x033C, |
|
second: 0x2033, |
|
secondtonechinese: 0x02CA, |
|
section: 0x00A7, |
|
seenarabic: 0x0633, |
|
seenfinalarabic: 0xFEB2, |
|
seeninitialarabic: 0xFEB3, |
|
seenmedialarabic: 0xFEB4, |
|
segol: 0x05B6, |
|
segol13: 0x05B6, |
|
segol1f: 0x05B6, |
|
segol2c: 0x05B6, |
|
segolhebrew: 0x05B6, |
|
segolnarrowhebrew: 0x05B6, |
|
segolquarterhebrew: 0x05B6, |
|
segoltahebrew: 0x0592, |
|
segolwidehebrew: 0x05B6, |
|
seharmenian: 0x057D, |
|
sehiragana: 0x305B, |
|
sekatakana: 0x30BB, |
|
sekatakanahalfwidth: 0xFF7E, |
|
semicolon: 0x003B, |
|
semicolonarabic: 0x061B, |
|
semicolonmonospace: 0xFF1B, |
|
semicolonsmall: 0xFE54, |
|
semivoicedmarkkana: 0x309C, |
|
semivoicedmarkkanahalfwidth: 0xFF9F, |
|
sentisquare: 0x3322, |
|
sentosquare: 0x3323, |
|
seven: 0x0037, |
|
sevenarabic: 0x0667, |
|
sevenbengali: 0x09ED, |
|
sevencircle: 0x2466, |
|
sevencircleinversesansserif: 0x2790, |
|
sevendeva: 0x096D, |
|
seveneighths: 0x215E, |
|
sevengujarati: 0x0AED, |
|
sevengurmukhi: 0x0A6D, |
|
sevenhackarabic: 0x0667, |
|
sevenhangzhou: 0x3027, |
|
sevenideographicparen: 0x3226, |
|
seveninferior: 0x2087, |
|
sevenmonospace: 0xFF17, |
|
sevenoldstyle: 0xF737, |
|
sevenparen: 0x247A, |
|
sevenperiod: 0x248E, |
|
sevenpersian: 0x06F7, |
|
sevenroman: 0x2176, |
|
sevensuperior: 0x2077, |
|
seventeencircle: 0x2470, |
|
seventeenparen: 0x2484, |
|
seventeenperiod: 0x2498, |
|
seventhai: 0x0E57, |
|
sfthyphen: 0x00AD, |
|
shaarmenian: 0x0577, |
|
shabengali: 0x09B6, |
|
shacyrillic: 0x0448, |
|
shaddaarabic: 0x0651, |
|
shaddadammaarabic: 0xFC61, |
|
shaddadammatanarabic: 0xFC5E, |
|
shaddafathaarabic: 0xFC60, |
|
shaddakasraarabic: 0xFC62, |
|
shaddakasratanarabic: 0xFC5F, |
|
shade: 0x2592, |
|
shadedark: 0x2593, |
|
shadelight: 0x2591, |
|
shademedium: 0x2592, |
|
shadeva: 0x0936, |
|
shagujarati: 0x0AB6, |
|
shagurmukhi: 0x0A36, |
|
shalshelethebrew: 0x0593, |
|
shbopomofo: 0x3115, |
|
shchacyrillic: 0x0449, |
|
sheenarabic: 0x0634, |
|
sheenfinalarabic: 0xFEB6, |
|
sheeninitialarabic: 0xFEB7, |
|
sheenmedialarabic: 0xFEB8, |
|
sheicoptic: 0x03E3, |
|
sheqel: 0x20AA, |
|
sheqelhebrew: 0x20AA, |
|
sheva: 0x05B0, |
|
sheva115: 0x05B0, |
|
sheva15: 0x05B0, |
|
sheva22: 0x05B0, |
|
sheva2e: 0x05B0, |
|
shevahebrew: 0x05B0, |
|
shevanarrowhebrew: 0x05B0, |
|
shevaquarterhebrew: 0x05B0, |
|
shevawidehebrew: 0x05B0, |
|
shhacyrillic: 0x04BB, |
|
shimacoptic: 0x03ED, |
|
shin: 0x05E9, |
|
shindagesh: 0xFB49, |
|
shindageshhebrew: 0xFB49, |
|
shindageshshindot: 0xFB2C, |
|
shindageshshindothebrew: 0xFB2C, |
|
shindageshsindot: 0xFB2D, |
|
shindageshsindothebrew: 0xFB2D, |
|
shindothebrew: 0x05C1, |
|
shinhebrew: 0x05E9, |
|
shinshindot: 0xFB2A, |
|
shinshindothebrew: 0xFB2A, |
|
shinsindot: 0xFB2B, |
|
shinsindothebrew: 0xFB2B, |
|
shook: 0x0282, |
|
sigma: 0x03C3, |
|
sigma1: 0x03C2, |
|
sigmafinal: 0x03C2, |
|
sigmalunatesymbolgreek: 0x03F2, |
|
sihiragana: 0x3057, |
|
sikatakana: 0x30B7, |
|
sikatakanahalfwidth: 0xFF7C, |
|
siluqhebrew: 0x05BD, |
|
siluqlefthebrew: 0x05BD, |
|
similar: 0x223C, |
|
sindothebrew: 0x05C2, |
|
siosacirclekorean: 0x3274, |
|
siosaparenkorean: 0x3214, |
|
sioscieuckorean: 0x317E, |
|
sioscirclekorean: 0x3266, |
|
sioskiyeokkorean: 0x317A, |
|
sioskorean: 0x3145, |
|
siosnieunkorean: 0x317B, |
|
siosparenkorean: 0x3206, |
|
siospieupkorean: 0x317D, |
|
siostikeutkorean: 0x317C, |
|
six: 0x0036, |
|
sixarabic: 0x0666, |
|
sixbengali: 0x09EC, |
|
sixcircle: 0x2465, |
|
sixcircleinversesansserif: 0x278F, |
|
sixdeva: 0x096C, |
|
sixgujarati: 0x0AEC, |
|
sixgurmukhi: 0x0A6C, |
|
sixhackarabic: 0x0666, |
|
sixhangzhou: 0x3026, |
|
sixideographicparen: 0x3225, |
|
sixinferior: 0x2086, |
|
sixmonospace: 0xFF16, |
|
sixoldstyle: 0xF736, |
|
sixparen: 0x2479, |
|
sixperiod: 0x248D, |
|
sixpersian: 0x06F6, |
|
sixroman: 0x2175, |
|
sixsuperior: 0x2076, |
|
sixteencircle: 0x246F, |
|
sixteencurrencydenominatorbengali: 0x09F9, |
|
sixteenparen: 0x2483, |
|
sixteenperiod: 0x2497, |
|
sixthai: 0x0E56, |
|
slash: 0x002F, |
|
slashmonospace: 0xFF0F, |
|
slong: 0x017F, |
|
slongdotaccent: 0x1E9B, |
|
smileface: 0x263A, |
|
smonospace: 0xFF53, |
|
sofpasuqhebrew: 0x05C3, |
|
softhyphen: 0x00AD, |
|
softsigncyrillic: 0x044C, |
|
sohiragana: 0x305D, |
|
sokatakana: 0x30BD, |
|
sokatakanahalfwidth: 0xFF7F, |
|
soliduslongoverlaycmb: 0x0338, |
|
solidusshortoverlaycmb: 0x0337, |
|
sorusithai: 0x0E29, |
|
sosalathai: 0x0E28, |
|
sosothai: 0x0E0B, |
|
sosuathai: 0x0E2A, |
|
space: 0x0020, |
|
spacehackarabic: 0x0020, |
|
spade: 0x2660, |
|
spadesuitblack: 0x2660, |
|
spadesuitwhite: 0x2664, |
|
sparen: 0x24AE, |
|
squarebelowcmb: 0x033B, |
|
squarecc: 0x33C4, |
|
squarecm: 0x339D, |
|
squarediagonalcrosshatchfill: 0x25A9, |
|
squarehorizontalfill: 0x25A4, |
|
squarekg: 0x338F, |
|
squarekm: 0x339E, |
|
squarekmcapital: 0x33CE, |
|
squareln: 0x33D1, |
|
squarelog: 0x33D2, |
|
squaremg: 0x338E, |
|
squaremil: 0x33D5, |
|
squaremm: 0x339C, |
|
squaremsquared: 0x33A1, |
|
squareorthogonalcrosshatchfill: 0x25A6, |
|
squareupperlefttolowerrightfill: 0x25A7, |
|
squareupperrighttolowerleftfill: 0x25A8, |
|
squareverticalfill: 0x25A5, |
|
squarewhitewithsmallblack: 0x25A3, |
|
srsquare: 0x33DB, |
|
ssabengali: 0x09B7, |
|
ssadeva: 0x0937, |
|
ssagujarati: 0x0AB7, |
|
ssangcieuckorean: 0x3149, |
|
ssanghieuhkorean: 0x3185, |
|
ssangieungkorean: 0x3180, |
|
ssangkiyeokkorean: 0x3132, |
|
ssangnieunkorean: 0x3165, |
|
ssangpieupkorean: 0x3143, |
|
ssangsioskorean: 0x3146, |
|
ssangtikeutkorean: 0x3138, |
|
ssuperior: 0xF6F2, |
|
sterling: 0x00A3, |
|
sterlingmonospace: 0xFFE1, |
|
strokelongoverlaycmb: 0x0336, |
|
strokeshortoverlaycmb: 0x0335, |
|
subset: 0x2282, |
|
subsetnotequal: 0x228A, |
|
subsetorequal: 0x2286, |
|
succeeds: 0x227B, |
|
suchthat: 0x220B, |
|
suhiragana: 0x3059, |
|
sukatakana: 0x30B9, |
|
sukatakanahalfwidth: 0xFF7D, |
|
sukunarabic: 0x0652, |
|
summation: 0x2211, |
|
sun: 0x263C, |
|
superset: 0x2283, |
|
supersetnotequal: 0x228B, |
|
supersetorequal: 0x2287, |
|
svsquare: 0x33DC, |
|
syouwaerasquare: 0x337C, |
|
t: 0x0074, |
|
tabengali: 0x09A4, |
|
tackdown: 0x22A4, |
|
tackleft: 0x22A3, |
|
tadeva: 0x0924, |
|
tagujarati: 0x0AA4, |
|
tagurmukhi: 0x0A24, |
|
taharabic: 0x0637, |
|
tahfinalarabic: 0xFEC2, |
|
tahinitialarabic: 0xFEC3, |
|
tahiragana: 0x305F, |
|
tahmedialarabic: 0xFEC4, |
|
taisyouerasquare: 0x337D, |
|
takatakana: 0x30BF, |
|
takatakanahalfwidth: 0xFF80, |
|
tatweelarabic: 0x0640, |
|
tau: 0x03C4, |
|
tav: 0x05EA, |
|
tavdages: 0xFB4A, |
|
tavdagesh: 0xFB4A, |
|
tavdageshhebrew: 0xFB4A, |
|
tavhebrew: 0x05EA, |
|
tbar: 0x0167, |
|
tbopomofo: 0x310A, |
|
tcaron: 0x0165, |
|
tccurl: 0x02A8, |
|
tcedilla: 0x0163, |
|
tcheharabic: 0x0686, |
|
tchehfinalarabic: 0xFB7B, |
|
tchehinitialarabic: 0xFB7C, |
|
tchehmedialarabic: 0xFB7D, |
|
tcircle: 0x24E3, |
|
tcircumflexbelow: 0x1E71, |
|
tcommaaccent: 0x0163, |
|
tdieresis: 0x1E97, |
|
tdotaccent: 0x1E6B, |
|
tdotbelow: 0x1E6D, |
|
tecyrillic: 0x0442, |
|
tedescendercyrillic: 0x04AD, |
|
teharabic: 0x062A, |
|
tehfinalarabic: 0xFE96, |
|
tehhahinitialarabic: 0xFCA2, |
|
tehhahisolatedarabic: 0xFC0C, |
|
tehinitialarabic: 0xFE97, |
|
tehiragana: 0x3066, |
|
tehjeeminitialarabic: 0xFCA1, |
|
tehjeemisolatedarabic: 0xFC0B, |
|
tehmarbutaarabic: 0x0629, |
|
tehmarbutafinalarabic: 0xFE94, |
|
tehmedialarabic: 0xFE98, |
|
tehmeeminitialarabic: 0xFCA4, |
|
tehmeemisolatedarabic: 0xFC0E, |
|
tehnoonfinalarabic: 0xFC73, |
|
tekatakana: 0x30C6, |
|
tekatakanahalfwidth: 0xFF83, |
|
telephone: 0x2121, |
|
telephoneblack: 0x260E, |
|
telishagedolahebrew: 0x05A0, |
|
telishaqetanahebrew: 0x05A9, |
|
tencircle: 0x2469, |
|
tenideographicparen: 0x3229, |
|
tenparen: 0x247D, |
|
tenperiod: 0x2491, |
|
tenroman: 0x2179, |
|
tesh: 0x02A7, |
|
tet: 0x05D8, |
|
tetdagesh: 0xFB38, |
|
tetdageshhebrew: 0xFB38, |
|
tethebrew: 0x05D8, |
|
tetsecyrillic: 0x04B5, |
|
tevirhebrew: 0x059B, |
|
tevirlefthebrew: 0x059B, |
|
thabengali: 0x09A5, |
|
thadeva: 0x0925, |
|
thagujarati: 0x0AA5, |
|
thagurmukhi: 0x0A25, |
|
thalarabic: 0x0630, |
|
thalfinalarabic: 0xFEAC, |
|
thanthakhatlowleftthai: 0xF898, |
|
thanthakhatlowrightthai: 0xF897, |
|
thanthakhatthai: 0x0E4C, |
|
thanthakhatupperleftthai: 0xF896, |
|
theharabic: 0x062B, |
|
thehfinalarabic: 0xFE9A, |
|
thehinitialarabic: 0xFE9B, |
|
thehmedialarabic: 0xFE9C, |
|
thereexists: 0x2203, |
|
therefore: 0x2234, |
|
theta: 0x03B8, |
|
theta1: 0x03D1, |
|
thetasymbolgreek: 0x03D1, |
|
thieuthacirclekorean: 0x3279, |
|
thieuthaparenkorean: 0x3219, |
|
thieuthcirclekorean: 0x326B, |
|
thieuthkorean: 0x314C, |
|
thieuthparenkorean: 0x320B, |
|
thirteencircle: 0x246C, |
|
thirteenparen: 0x2480, |
|
thirteenperiod: 0x2494, |
|
thonangmonthothai: 0x0E11, |
|
thook: 0x01AD, |
|
thophuthaothai: 0x0E12, |
|
thorn: 0x00FE, |
|
thothahanthai: 0x0E17, |
|
thothanthai: 0x0E10, |
|
thothongthai: 0x0E18, |
|
thothungthai: 0x0E16, |
|
thousandcyrillic: 0x0482, |
|
thousandsseparatorarabic: 0x066C, |
|
thousandsseparatorpersian: 0x066C, |
|
three: 0x0033, |
|
threearabic: 0x0663, |
|
threebengali: 0x09E9, |
|
threecircle: 0x2462, |
|
threecircleinversesansserif: 0x278C, |
|
threedeva: 0x0969, |
|
threeeighths: 0x215C, |
|
threegujarati: 0x0AE9, |
|
threegurmukhi: 0x0A69, |
|
threehackarabic: 0x0663, |
|
threehangzhou: 0x3023, |
|
threeideographicparen: 0x3222, |
|
threeinferior: 0x2083, |
|
threemonospace: 0xFF13, |
|
threenumeratorbengali: 0x09F6, |
|
threeoldstyle: 0xF733, |
|
threeparen: 0x2476, |
|
threeperiod: 0x248A, |
|
threepersian: 0x06F3, |
|
threequarters: 0x00BE, |
|
threequartersemdash: 0xF6DE, |
|
threeroman: 0x2172, |
|
threesuperior: 0x00B3, |
|
threethai: 0x0E53, |
|
thzsquare: 0x3394, |
|
tihiragana: 0x3061, |
|
tikatakana: 0x30C1, |
|
tikatakanahalfwidth: 0xFF81, |
|
tikeutacirclekorean: 0x3270, |
|
tikeutaparenkorean: 0x3210, |
|
tikeutcirclekorean: 0x3262, |
|
tikeutkorean: 0x3137, |
|
tikeutparenkorean: 0x3202, |
|
tilde: 0x02DC, |
|
tildebelowcmb: 0x0330, |
|
tildecmb: 0x0303, |
|
tildecomb: 0x0303, |
|
tildedoublecmb: 0x0360, |
|
tildeoperator: 0x223C, |
|
tildeoverlaycmb: 0x0334, |
|
tildeverticalcmb: 0x033E, |
|
timescircle: 0x2297, |
|
tipehahebrew: 0x0596, |
|
tipehalefthebrew: 0x0596, |
|
tippigurmukhi: 0x0A70, |
|
titlocyrilliccmb: 0x0483, |
|
tiwnarmenian: 0x057F, |
|
tlinebelow: 0x1E6F, |
|
tmonospace: 0xFF54, |
|
toarmenian: 0x0569, |
|
tohiragana: 0x3068, |
|
tokatakana: 0x30C8, |
|
tokatakanahalfwidth: 0xFF84, |
|
tonebarextrahighmod: 0x02E5, |
|
tonebarextralowmod: 0x02E9, |
|
tonebarhighmod: 0x02E6, |
|
tonebarlowmod: 0x02E8, |
|
tonebarmidmod: 0x02E7, |
|
tonefive: 0x01BD, |
|
tonesix: 0x0185, |
|
tonetwo: 0x01A8, |
|
tonos: 0x0384, |
|
tonsquare: 0x3327, |
|
topatakthai: 0x0E0F, |
|
tortoiseshellbracketleft: 0x3014, |
|
tortoiseshellbracketleftsmall: 0xFE5D, |
|
tortoiseshellbracketleftvertical: 0xFE39, |
|
tortoiseshellbracketright: 0x3015, |
|
tortoiseshellbracketrightsmall: 0xFE5E, |
|
tortoiseshellbracketrightvertical: 0xFE3A, |
|
totaothai: 0x0E15, |
|
tpalatalhook: 0x01AB, |
|
tparen: 0x24AF, |
|
trademark: 0x2122, |
|
trademarksans: 0xF8EA, |
|
trademarkserif: 0xF6DB, |
|
tretroflexhook: 0x0288, |
|
triagdn: 0x25BC, |
|
triaglf: 0x25C4, |
|
triagrt: 0x25BA, |
|
triagup: 0x25B2, |
|
ts: 0x02A6, |
|
tsadi: 0x05E6, |
|
tsadidagesh: 0xFB46, |
|
tsadidageshhebrew: 0xFB46, |
|
tsadihebrew: 0x05E6, |
|
tsecyrillic: 0x0446, |
|
tsere: 0x05B5, |
|
tsere12: 0x05B5, |
|
tsere1e: 0x05B5, |
|
tsere2b: 0x05B5, |
|
tserehebrew: 0x05B5, |
|
tserenarrowhebrew: 0x05B5, |
|
tserequarterhebrew: 0x05B5, |
|
tserewidehebrew: 0x05B5, |
|
tshecyrillic: 0x045B, |
|
tsuperior: 0xF6F3, |
|
ttabengali: 0x099F, |
|
ttadeva: 0x091F, |
|
ttagujarati: 0x0A9F, |
|
ttagurmukhi: 0x0A1F, |
|
tteharabic: 0x0679, |
|
ttehfinalarabic: 0xFB67, |
|
ttehinitialarabic: 0xFB68, |
|
ttehmedialarabic: 0xFB69, |
|
tthabengali: 0x09A0, |
|
tthadeva: 0x0920, |
|
tthagujarati: 0x0AA0, |
|
tthagurmukhi: 0x0A20, |
|
tturned: 0x0287, |
|
tuhiragana: 0x3064, |
|
tukatakana: 0x30C4, |
|
tukatakanahalfwidth: 0xFF82, |
|
tusmallhiragana: 0x3063, |
|
tusmallkatakana: 0x30C3, |
|
tusmallkatakanahalfwidth: 0xFF6F, |
|
twelvecircle: 0x246B, |
|
twelveparen: 0x247F, |
|
twelveperiod: 0x2493, |
|
twelveroman: 0x217B, |
|
twentycircle: 0x2473, |
|
twentyhangzhou: 0x5344, |
|
twentyparen: 0x2487, |
|
twentyperiod: 0x249B, |
|
two: 0x0032, |
|
twoarabic: 0x0662, |
|
twobengali: 0x09E8, |
|
twocircle: 0x2461, |
|
twocircleinversesansserif: 0x278B, |
|
twodeva: 0x0968, |
|
twodotenleader: 0x2025, |
|
twodotleader: 0x2025, |
|
twodotleadervertical: 0xFE30, |
|
twogujarati: 0x0AE8, |
|
twogurmukhi: 0x0A68, |
|
twohackarabic: 0x0662, |
|
twohangzhou: 0x3022, |
|
twoideographicparen: 0x3221, |
|
twoinferior: 0x2082, |
|
twomonospace: 0xFF12, |
|
twonumeratorbengali: 0x09F5, |
|
twooldstyle: 0xF732, |
|
twoparen: 0x2475, |
|
twoperiod: 0x2489, |
|
twopersian: 0x06F2, |
|
tworoman: 0x2171, |
|
twostroke: 0x01BB, |
|
twosuperior: 0x00B2, |
|
twothai: 0x0E52, |
|
twothirds: 0x2154, |
|
u: 0x0075, |
|
uacute: 0x00FA, |
|
ubar: 0x0289, |
|
ubengali: 0x0989, |
|
ubopomofo: 0x3128, |
|
ubreve: 0x016D, |
|
ucaron: 0x01D4, |
|
ucircle: 0x24E4, |
|
ucircumflex: 0x00FB, |
|
ucircumflexbelow: 0x1E77, |
|
ucyrillic: 0x0443, |
|
udattadeva: 0x0951, |
|
udblacute: 0x0171, |
|
udblgrave: 0x0215, |
|
udeva: 0x0909, |
|
udieresis: 0x00FC, |
|
udieresisacute: 0x01D8, |
|
udieresisbelow: 0x1E73, |
|
udieresiscaron: 0x01DA, |
|
udieresiscyrillic: 0x04F1, |
|
udieresisgrave: 0x01DC, |
|
udieresismacron: 0x01D6, |
|
udotbelow: 0x1EE5, |
|
ugrave: 0x00F9, |
|
ugujarati: 0x0A89, |
|
ugurmukhi: 0x0A09, |
|
uhiragana: 0x3046, |
|
uhookabove: 0x1EE7, |
|
uhorn: 0x01B0, |
|
uhornacute: 0x1EE9, |
|
uhorndotbelow: 0x1EF1, |
|
uhorngrave: 0x1EEB, |
|
uhornhookabove: 0x1EED, |
|
uhorntilde: 0x1EEF, |
|
uhungarumlaut: 0x0171, |
|
uhungarumlautcyrillic: 0x04F3, |
|
uinvertedbreve: 0x0217, |
|
ukatakana: 0x30A6, |
|
ukatakanahalfwidth: 0xFF73, |
|
ukcyrillic: 0x0479, |
|
ukorean: 0x315C, |
|
umacron: 0x016B, |
|
umacroncyrillic: 0x04EF, |
|
umacrondieresis: 0x1E7B, |
|
umatragurmukhi: 0x0A41, |
|
umonospace: 0xFF55, |
|
underscore: 0x005F, |
|
underscoredbl: 0x2017, |
|
underscoremonospace: 0xFF3F, |
|
underscorevertical: 0xFE33, |
|
underscorewavy: 0xFE4F, |
|
union: 0x222A, |
|
universal: 0x2200, |
|
uogonek: 0x0173, |
|
uparen: 0x24B0, |
|
upblock: 0x2580, |
|
upperdothebrew: 0x05C4, |
|
upsilon: 0x03C5, |
|
upsilondieresis: 0x03CB, |
|
upsilondieresistonos: 0x03B0, |
|
upsilonlatin: 0x028A, |
|
upsilontonos: 0x03CD, |
|
uptackbelowcmb: 0x031D, |
|
uptackmod: 0x02D4, |
|
uragurmukhi: 0x0A73, |
|
uring: 0x016F, |
|
ushortcyrillic: 0x045E, |
|
usmallhiragana: 0x3045, |
|
usmallkatakana: 0x30A5, |
|
usmallkatakanahalfwidth: 0xFF69, |
|
ustraightcyrillic: 0x04AF, |
|
ustraightstrokecyrillic: 0x04B1, |
|
utilde: 0x0169, |
|
utildeacute: 0x1E79, |
|
utildebelow: 0x1E75, |
|
uubengali: 0x098A, |
|
uudeva: 0x090A, |
|
uugujarati: 0x0A8A, |
|
uugurmukhi: 0x0A0A, |
|
uumatragurmukhi: 0x0A42, |
|
uuvowelsignbengali: 0x09C2, |
|
uuvowelsigndeva: 0x0942, |
|
uuvowelsigngujarati: 0x0AC2, |
|
uvowelsignbengali: 0x09C1, |
|
uvowelsigndeva: 0x0941, |
|
uvowelsigngujarati: 0x0AC1, |
|
v: 0x0076, |
|
vadeva: 0x0935, |
|
vagujarati: 0x0AB5, |
|
vagurmukhi: 0x0A35, |
|
vakatakana: 0x30F7, |
|
vav: 0x05D5, |
|
vavdagesh: 0xFB35, |
|
vavdagesh65: 0xFB35, |
|
vavdageshhebrew: 0xFB35, |
|
vavhebrew: 0x05D5, |
|
vavholam: 0xFB4B, |
|
vavholamhebrew: 0xFB4B, |
|
vavvavhebrew: 0x05F0, |
|
vavyodhebrew: 0x05F1, |
|
vcircle: 0x24E5, |
|
vdotbelow: 0x1E7F, |
|
vecyrillic: 0x0432, |
|
veharabic: 0x06A4, |
|
vehfinalarabic: 0xFB6B, |
|
vehinitialarabic: 0xFB6C, |
|
vehmedialarabic: 0xFB6D, |
|
vekatakana: 0x30F9, |
|
venus: 0x2640, |
|
verticalbar: 0x007C, |
|
verticallineabovecmb: 0x030D, |
|
verticallinebelowcmb: 0x0329, |
|
verticallinelowmod: 0x02CC, |
|
verticallinemod: 0x02C8, |
|
vewarmenian: 0x057E, |
|
vhook: 0x028B, |
|
vikatakana: 0x30F8, |
|
viramabengali: 0x09CD, |
|
viramadeva: 0x094D, |
|
viramagujarati: 0x0ACD, |
|
visargabengali: 0x0983, |
|
visargadeva: 0x0903, |
|
visargagujarati: 0x0A83, |
|
vmonospace: 0xFF56, |
|
voarmenian: 0x0578, |
|
voicediterationhiragana: 0x309E, |
|
voicediterationkatakana: 0x30FE, |
|
voicedmarkkana: 0x309B, |
|
voicedmarkkanahalfwidth: 0xFF9E, |
|
vokatakana: 0x30FA, |
|
vparen: 0x24B1, |
|
vtilde: 0x1E7D, |
|
vturned: 0x028C, |
|
vuhiragana: 0x3094, |
|
vukatakana: 0x30F4, |
|
w: 0x0077, |
|
wacute: 0x1E83, |
|
waekorean: 0x3159, |
|
wahiragana: 0x308F, |
|
wakatakana: 0x30EF, |
|
wakatakanahalfwidth: 0xFF9C, |
|
wakorean: 0x3158, |
|
wasmallhiragana: 0x308E, |
|
wasmallkatakana: 0x30EE, |
|
wattosquare: 0x3357, |
|
wavedash: 0x301C, |
|
wavyunderscorevertical: 0xFE34, |
|
wawarabic: 0x0648, |
|
wawfinalarabic: 0xFEEE, |
|
wawhamzaabovearabic: 0x0624, |
|
wawhamzaabovefinalarabic: 0xFE86, |
|
wbsquare: 0x33DD, |
|
wcircle: 0x24E6, |
|
wcircumflex: 0x0175, |
|
wdieresis: 0x1E85, |
|
wdotaccent: 0x1E87, |
|
wdotbelow: 0x1E89, |
|
wehiragana: 0x3091, |
|
weierstrass: 0x2118, |
|
wekatakana: 0x30F1, |
|
wekorean: 0x315E, |
|
weokorean: 0x315D, |
|
wgrave: 0x1E81, |
|
whitebullet: 0x25E6, |
|
whitecircle: 0x25CB, |
|
whitecircleinverse: 0x25D9, |
|
whitecornerbracketleft: 0x300E, |
|
whitecornerbracketleftvertical: 0xFE43, |
|
whitecornerbracketright: 0x300F, |
|
whitecornerbracketrightvertical: 0xFE44, |
|
whitediamond: 0x25C7, |
|
whitediamondcontainingblacksmalldiamond: 0x25C8, |
|
whitedownpointingsmalltriangle: 0x25BF, |
|
whitedownpointingtriangle: 0x25BD, |
|
whiteleftpointingsmalltriangle: 0x25C3, |
|
whiteleftpointingtriangle: 0x25C1, |
|
whitelenticularbracketleft: 0x3016, |
|
whitelenticularbracketright: 0x3017, |
|
whiterightpointingsmalltriangle: 0x25B9, |
|
whiterightpointingtriangle: 0x25B7, |
|
whitesmallsquare: 0x25AB, |
|
whitesmilingface: 0x263A, |
|
whitesquare: 0x25A1, |
|
whitestar: 0x2606, |
|
whitetelephone: 0x260F, |
|
whitetortoiseshellbracketleft: 0x3018, |
|
whitetortoiseshellbracketright: 0x3019, |
|
whiteuppointingsmalltriangle: 0x25B5, |
|
whiteuppointingtriangle: 0x25B3, |
|
wihiragana: 0x3090, |
|
wikatakana: 0x30F0, |
|
wikorean: 0x315F, |
|
wmonospace: 0xFF57, |
|
wohiragana: 0x3092, |
|
wokatakana: 0x30F2, |
|
wokatakanahalfwidth: 0xFF66, |
|
won: 0x20A9, |
|
wonmonospace: 0xFFE6, |
|
wowaenthai: 0x0E27, |
|
wparen: 0x24B2, |
|
wring: 0x1E98, |
|
wsuperior: 0x02B7, |
|
wturned: 0x028D, |
|
wynn: 0x01BF, |
|
x: 0x0078, |
|
xabovecmb: 0x033D, |
|
xbopomofo: 0x3112, |
|
xcircle: 0x24E7, |
|
xdieresis: 0x1E8D, |
|
xdotaccent: 0x1E8B, |
|
xeharmenian: 0x056D, |
|
xi: 0x03BE, |
|
xmonospace: 0xFF58, |
|
xparen: 0x24B3, |
|
xsuperior: 0x02E3, |
|
y: 0x0079, |
|
yaadosquare: 0x334E, |
|
yabengali: 0x09AF, |
|
yacute: 0x00FD, |
|
yadeva: 0x092F, |
|
yaekorean: 0x3152, |
|
yagujarati: 0x0AAF, |
|
yagurmukhi: 0x0A2F, |
|
yahiragana: 0x3084, |
|
yakatakana: 0x30E4, |
|
yakatakanahalfwidth: 0xFF94, |
|
yakorean: 0x3151, |
|
yamakkanthai: 0x0E4E, |
|
yasmallhiragana: 0x3083, |
|
yasmallkatakana: 0x30E3, |
|
yasmallkatakanahalfwidth: 0xFF6C, |
|
yatcyrillic: 0x0463, |
|
ycircle: 0x24E8, |
|
ycircumflex: 0x0177, |
|
ydieresis: 0x00FF, |
|
ydotaccent: 0x1E8F, |
|
ydotbelow: 0x1EF5, |
|
yeharabic: 0x064A, |
|
yehbarreearabic: 0x06D2, |
|
yehbarreefinalarabic: 0xFBAF, |
|
yehfinalarabic: 0xFEF2, |
|
yehhamzaabovearabic: 0x0626, |
|
yehhamzaabovefinalarabic: 0xFE8A, |
|
yehhamzaaboveinitialarabic: 0xFE8B, |
|
yehhamzaabovemedialarabic: 0xFE8C, |
|
yehinitialarabic: 0xFEF3, |
|
yehmedialarabic: 0xFEF4, |
|
yehmeeminitialarabic: 0xFCDD, |
|
yehmeemisolatedarabic: 0xFC58, |
|
yehnoonfinalarabic: 0xFC94, |
|
yehthreedotsbelowarabic: 0x06D1, |
|
yekorean: 0x3156, |
|
yen: 0x00A5, |
|
yenmonospace: 0xFFE5, |
|
yeokorean: 0x3155, |
|
yeorinhieuhkorean: 0x3186, |
|
yerahbenyomohebrew: 0x05AA, |
|
yerahbenyomolefthebrew: 0x05AA, |
|
yericyrillic: 0x044B, |
|
yerudieresiscyrillic: 0x04F9, |
|
yesieungkorean: 0x3181, |
|
yesieungpansioskorean: 0x3183, |
|
yesieungsioskorean: 0x3182, |
|
yetivhebrew: 0x059A, |
|
ygrave: 0x1EF3, |
|
yhook: 0x01B4, |
|
yhookabove: 0x1EF7, |
|
yiarmenian: 0x0575, |
|
yicyrillic: 0x0457, |
|
yikorean: 0x3162, |
|
yinyang: 0x262F, |
|
yiwnarmenian: 0x0582, |
|
ymonospace: 0xFF59, |
|
yod: 0x05D9, |
|
yoddagesh: 0xFB39, |
|
yoddageshhebrew: 0xFB39, |
|
yodhebrew: 0x05D9, |
|
yodyodhebrew: 0x05F2, |
|
yodyodpatahhebrew: 0xFB1F, |
|
yohiragana: 0x3088, |
|
yoikorean: 0x3189, |
|
yokatakana: 0x30E8, |
|
yokatakanahalfwidth: 0xFF96, |
|
yokorean: 0x315B, |
|
yosmallhiragana: 0x3087, |
|
yosmallkatakana: 0x30E7, |
|
yosmallkatakanahalfwidth: 0xFF6E, |
|
yotgreek: 0x03F3, |
|
yoyaekorean: 0x3188, |
|
yoyakorean: 0x3187, |
|
yoyakthai: 0x0E22, |
|
yoyingthai: 0x0E0D, |
|
yparen: 0x24B4, |
|
ypogegrammeni: 0x037A, |
|
ypogegrammenigreekcmb: 0x0345, |
|
yr: 0x01A6, |
|
yring: 0x1E99, |
|
ysuperior: 0x02B8, |
|
ytilde: 0x1EF9, |
|
yturned: 0x028E, |
|
yuhiragana: 0x3086, |
|
yuikorean: 0x318C, |
|
yukatakana: 0x30E6, |
|
yukatakanahalfwidth: 0xFF95, |
|
yukorean: 0x3160, |
|
yusbigcyrillic: 0x046B, |
|
yusbigiotifiedcyrillic: 0x046D, |
|
yuslittlecyrillic: 0x0467, |
|
yuslittleiotifiedcyrillic: 0x0469, |
|
yusmallhiragana: 0x3085, |
|
yusmallkatakana: 0x30E5, |
|
yusmallkatakanahalfwidth: 0xFF6D, |
|
yuyekorean: 0x318B, |
|
yuyeokorean: 0x318A, |
|
yyabengali: 0x09DF, |
|
yyadeva: 0x095F, |
|
z: 0x007A, |
|
zaarmenian: 0x0566, |
|
zacute: 0x017A, |
|
zadeva: 0x095B, |
|
zagurmukhi: 0x0A5B, |
|
zaharabic: 0x0638, |
|
zahfinalarabic: 0xFEC6, |
|
zahinitialarabic: 0xFEC7, |
|
zahiragana: 0x3056, |
|
zahmedialarabic: 0xFEC8, |
|
zainarabic: 0x0632, |
|
zainfinalarabic: 0xFEB0, |
|
zakatakana: 0x30B6, |
|
zaqefgadolhebrew: 0x0595, |
|
zaqefqatanhebrew: 0x0594, |
|
zarqahebrew: 0x0598, |
|
zayin: 0x05D6, |
|
zayindagesh: 0xFB36, |
|
zayindageshhebrew: 0xFB36, |
|
zayinhebrew: 0x05D6, |
|
zbopomofo: 0x3117, |
|
zcaron: 0x017E, |
|
zcircle: 0x24E9, |
|
zcircumflex: 0x1E91, |
|
zcurl: 0x0291, |
|
zdot: 0x017C, |
|
zdotaccent: 0x017C, |
|
zdotbelow: 0x1E93, |
|
zecyrillic: 0x0437, |
|
zedescendercyrillic: 0x0499, |
|
zedieresiscyrillic: 0x04DF, |
|
zehiragana: 0x305C, |
|
zekatakana: 0x30BC, |
|
zero: 0x0030, |
|
zeroarabic: 0x0660, |
|
zerobengali: 0x09E6, |
|
zerodeva: 0x0966, |
|
zerogujarati: 0x0AE6, |
|
zerogurmukhi: 0x0A66, |
|
zerohackarabic: 0x0660, |
|
zeroinferior: 0x2080, |
|
zeromonospace: 0xFF10, |
|
zerooldstyle: 0xF730, |
|
zeropersian: 0x06F0, |
|
zerosuperior: 0x2070, |
|
zerothai: 0x0E50, |
|
zerowidthjoiner: 0xFEFF, |
|
zerowidthnonjoiner: 0x200C, |
|
zerowidthspace: 0x200B, |
|
zeta: 0x03B6, |
|
zhbopomofo: 0x3113, |
|
zhearmenian: 0x056A, |
|
zhebrevecyrillic: 0x04C2, |
|
zhecyrillic: 0x0436, |
|
zhedescendercyrillic: 0x0497, |
|
zhedieresiscyrillic: 0x04DD, |
|
zihiragana: 0x3058, |
|
zikatakana: 0x30B8, |
|
zinorhebrew: 0x05AE, |
|
zlinebelow: 0x1E95, |
|
zmonospace: 0xFF5A, |
|
zohiragana: 0x305E, |
|
zokatakana: 0x30BE, |
|
zparen: 0x24B5, |
|
zretroflexhook: 0x0290, |
|
zstroke: 0x01B6, |
|
zuhiragana: 0x305A, |
|
zukatakana: 0x30BA, |
|
'.notdef': 0x0000 |
|
}; |
|
|
|
var DingbatsGlyphsUnicode = { |
|
space: 0x0020, |
|
a1: 0x2701, |
|
a2: 0x2702, |
|
a202: 0x2703, |
|
a3: 0x2704, |
|
a4: 0x260E, |
|
a5: 0x2706, |
|
a119: 0x2707, |
|
a118: 0x2708, |
|
a117: 0x2709, |
|
a11: 0x261B, |
|
a12: 0x261E, |
|
a13: 0x270C, |
|
a14: 0x270D, |
|
a15: 0x270E, |
|
a16: 0x270F, |
|
a105: 0x2710, |
|
a17: 0x2711, |
|
a18: 0x2712, |
|
a19: 0x2713, |
|
a20: 0x2714, |
|
a21: 0x2715, |
|
a22: 0x2716, |
|
a23: 0x2717, |
|
a24: 0x2718, |
|
a25: 0x2719, |
|
a26: 0x271A, |
|
a27: 0x271B, |
|
a28: 0x271C, |
|
a6: 0x271D, |
|
a7: 0x271E, |
|
a8: 0x271F, |
|
a9: 0x2720, |
|
a10: 0x2721, |
|
a29: 0x2722, |
|
a30: 0x2723, |
|
a31: 0x2724, |
|
a32: 0x2725, |
|
a33: 0x2726, |
|
a34: 0x2727, |
|
a35: 0x2605, |
|
a36: 0x2729, |
|
a37: 0x272A, |
|
a38: 0x272B, |
|
a39: 0x272C, |
|
a40: 0x272D, |
|
a41: 0x272E, |
|
a42: 0x272F, |
|
a43: 0x2730, |
|
a44: 0x2731, |
|
a45: 0x2732, |
|
a46: 0x2733, |
|
a47: 0x2734, |
|
a48: 0x2735, |
|
a49: 0x2736, |
|
a50: 0x2737, |
|
a51: 0x2738, |
|
a52: 0x2739, |
|
a53: 0x273A, |
|
a54: 0x273B, |
|
a55: 0x273C, |
|
a56: 0x273D, |
|
a57: 0x273E, |
|
a58: 0x273F, |
|
a59: 0x2740, |
|
a60: 0x2741, |
|
a61: 0x2742, |
|
a62: 0x2743, |
|
a63: 0x2744, |
|
a64: 0x2745, |
|
a65: 0x2746, |
|
a66: 0x2747, |
|
a67: 0x2748, |
|
a68: 0x2749, |
|
a69: 0x274A, |
|
a70: 0x274B, |
|
a71: 0x25CF, |
|
a72: 0x274D, |
|
a73: 0x25A0, |
|
a74: 0x274F, |
|
a203: 0x2750, |
|
a75: 0x2751, |
|
a204: 0x2752, |
|
a76: 0x25B2, |
|
a77: 0x25BC, |
|
a78: 0x25C6, |
|
a79: 0x2756, |
|
a81: 0x25D7, |
|
a82: 0x2758, |
|
a83: 0x2759, |
|
a84: 0x275A, |
|
a97: 0x275B, |
|
a98: 0x275C, |
|
a99: 0x275D, |
|
a100: 0x275E, |
|
a101: 0x2761, |
|
a102: 0x2762, |
|
a103: 0x2763, |
|
a104: 0x2764, |
|
a106: 0x2765, |
|
a107: 0x2766, |
|
a108: 0x2767, |
|
a112: 0x2663, |
|
a111: 0x2666, |
|
a110: 0x2665, |
|
a109: 0x2660, |
|
a120: 0x2460, |
|
a121: 0x2461, |
|
a122: 0x2462, |
|
a123: 0x2463, |
|
a124: 0x2464, |
|
a125: 0x2465, |
|
a126: 0x2466, |
|
a127: 0x2467, |
|
a128: 0x2468, |
|
a129: 0x2469, |
|
a130: 0x2776, |
|
a131: 0x2777, |
|
a132: 0x2778, |
|
a133: 0x2779, |
|
a134: 0x277A, |
|
a135: 0x277B, |
|
a136: 0x277C, |
|
a137: 0x277D, |
|
a138: 0x277E, |
|
a139: 0x277F, |
|
a140: 0x2780, |
|
a141: 0x2781, |
|
a142: 0x2782, |
|
a143: 0x2783, |
|
a144: 0x2784, |
|
a145: 0x2785, |
|
a146: 0x2786, |
|
a147: 0x2787, |
|
a148: 0x2788, |
|
a149: 0x2789, |
|
a150: 0x278A, |
|
a151: 0x278B, |
|
a152: 0x278C, |
|
a153: 0x278D, |
|
a154: 0x278E, |
|
a155: 0x278F, |
|
a156: 0x2790, |
|
a157: 0x2791, |
|
a158: 0x2792, |
|
a159: 0x2793, |
|
a160: 0x2794, |
|
a161: 0x2192, |
|
a163: 0x2194, |
|
a164: 0x2195, |
|
a196: 0x2798, |
|
a165: 0x2799, |
|
a192: 0x279A, |
|
a166: 0x279B, |
|
a167: 0x279C, |
|
a168: 0x279D, |
|
a169: 0x279E, |
|
a170: 0x279F, |
|
a171: 0x27A0, |
|
a172: 0x27A1, |
|
a173: 0x27A2, |
|
a162: 0x27A3, |
|
a174: 0x27A4, |
|
a175: 0x27A5, |
|
a176: 0x27A6, |
|
a177: 0x27A7, |
|
a178: 0x27A8, |
|
a179: 0x27A9, |
|
a193: 0x27AA, |
|
a180: 0x27AB, |
|
a199: 0x27AC, |
|
a181: 0x27AD, |
|
a200: 0x27AE, |
|
a182: 0x27AF, |
|
a201: 0x27B1, |
|
a183: 0x27B2, |
|
a184: 0x27B3, |
|
a197: 0x27B4, |
|
a185: 0x27B5, |
|
a194: 0x27B6, |
|
a198: 0x27B7, |
|
a186: 0x27B8, |
|
a195: 0x27B9, |
|
a187: 0x27BA, |
|
a188: 0x27BB, |
|
a189: 0x27BC, |
|
a190: 0x27BD, |
|
a191: 0x27BE, |
|
a89: 0x2768, // 0xF8D7 |
|
a90: 0x2769, // 0xF8D8 |
|
a93: 0x276A, // 0xF8D9 |
|
a94: 0x276B, // 0xF8DA |
|
a91: 0x276C, // 0xF8DB |
|
a92: 0x276D, // 0xF8DC |
|
a205: 0x276E, // 0xF8DD |
|
a85: 0x276F, // 0xF8DE |
|
a206: 0x2770, // 0xF8DF |
|
a86: 0x2771, // 0xF8E0 |
|
a87: 0x2772, // 0xF8E1 |
|
a88: 0x2773, // 0xF8E2 |
|
a95: 0x2774, // 0xF8E3 |
|
a96: 0x2775, // 0xF8E4 |
|
'.notdef': 0x0000 |
|
}; |
|
|
|
exports.GlyphsUnicode = GlyphsUnicode; |
|
exports.DingbatsGlyphsUnicode = DingbatsGlyphsUnicode; |
|
})); |
|
|
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreJpg = {})); |
|
} |
|
}(this, function (exports) { |
|
|
|
/* |
|
This code was forked from https://github.com/notmasteryet/jpgjs. The original |
|
version was created by github user notmasteryet |
|
|
|
- The JPEG specification can be found in the ITU CCITT Recommendation T.81 |
|
(www.w3.org/Graphics/JPEG/itu-t81.pdf) |
|
- The JFIF specification can be found in the JPEG File Interchange Format |
|
(www.w3.org/Graphics/JPEG/jfif3.pdf) |
|
- The Adobe Application-Specific JPEG markers in the Supporting the DCT Filters |
|
in PostScript Level 2, Technical Note #5116 |
|
(partners.adobe.com/public/developer/en/ps/sdk/5116.DCT_Filter.pdf) |
|
*/ |
|
|
|
var JpegImage = (function jpegImage() { |
|
var dctZigZag = new Uint8Array([ |
|
0, |
|
1, 8, |
|
16, 9, 2, |
|
3, 10, 17, 24, |
|
32, 25, 18, 11, 4, |
|
5, 12, 19, 26, 33, 40, |
|
48, 41, 34, 27, 20, 13, 6, |
|
7, 14, 21, 28, 35, 42, 49, 56, |
|
57, 50, 43, 36, 29, 22, 15, |
|
23, 30, 37, 44, 51, 58, |
|
59, 52, 45, 38, 31, |
|
39, 46, 53, 60, |
|
61, 54, 47, |
|
55, 62, |
|
63 |
|
]); |
|
|
|
var dctCos1 = 4017; // cos(pi/16) |
|
var dctSin1 = 799; // sin(pi/16) |
|
var dctCos3 = 3406; // cos(3*pi/16) |
|
var dctSin3 = 2276; // sin(3*pi/16) |
|
var dctCos6 = 1567; // cos(6*pi/16) |
|
var dctSin6 = 3784; // sin(6*pi/16) |
|
var dctSqrt2 = 5793; // sqrt(2) |
|
var dctSqrt1d2 = 2896; // sqrt(2) / 2 |
|
|
|
function constructor() { |
|
} |
|
|
|
function buildHuffmanTable(codeLengths, values) { |
|
var k = 0, code = [], i, j, length = 16; |
|
while (length > 0 && !codeLengths[length - 1]) { |
|
length--; |
|
} |
|
code.push({children: [], index: 0}); |
|
var p = code[0], q; |
|
for (i = 0; i < length; i++) { |
|
for (j = 0; j < codeLengths[i]; j++) { |
|
p = code.pop(); |
|
p.children[p.index] = values[k]; |
|
while (p.index > 0) { |
|
p = code.pop(); |
|
} |
|
p.index++; |
|
code.push(p); |
|
while (code.length <= i) { |
|
code.push(q = {children: [], index: 0}); |
|
p.children[p.index] = q.children; |
|
p = q; |
|
} |
|
k++; |
|
} |
|
if (i + 1 < length) { |
|
// p here points to last code |
|
code.push(q = {children: [], index: 0}); |
|
p.children[p.index] = q.children; |
|
p = q; |
|
} |
|
} |
|
return code[0].children; |
|
} |
|
|
|
function getBlockBufferOffset(component, row, col) { |
|
return 64 * ((component.blocksPerLine + 1) * row + col); |
|
} |
|
|
|
function decodeScan(data, offset, frame, components, resetInterval, |
|
spectralStart, spectralEnd, successivePrev, successive) { |
|
var mcusPerLine = frame.mcusPerLine; |
|
var progressive = frame.progressive; |
|
|
|
var startOffset = offset, bitsData = 0, bitsCount = 0; |
|
|
|
function readBit() { |
|
if (bitsCount > 0) { |
|
bitsCount--; |
|
return (bitsData >> bitsCount) & 1; |
|
} |
|
bitsData = data[offset++]; |
|
if (bitsData === 0xFF) { |
|
var nextByte = data[offset++]; |
|
if (nextByte) { |
|
throw 'unexpected marker: ' + |
|
((bitsData << 8) | nextByte).toString(16); |
|
} |
|
// unstuff 0 |
|
} |
|
bitsCount = 7; |
|
return bitsData >>> 7; |
|
} |
|
|
|
function decodeHuffman(tree) { |
|
var node = tree; |
|
while (true) { |
|
node = node[readBit()]; |
|
if (typeof node === 'number') { |
|
return node; |
|
} |
|
if (typeof node !== 'object') { |
|
throw 'invalid huffman sequence'; |
|
} |
|
} |
|
} |
|
|
|
function receive(length) { |
|
var n = 0; |
|
while (length > 0) { |
|
n = (n << 1) | readBit(); |
|
length--; |
|
} |
|
return n; |
|
} |
|
|
|
function receiveAndExtend(length) { |
|
if (length === 1) { |
|
return readBit() === 1 ? 1 : -1; |
|
} |
|
var n = receive(length); |
|
if (n >= 1 << (length - 1)) { |
|
return n; |
|
} |
|
return n + (-1 << length) + 1; |
|
} |
|
|
|
function decodeBaseline(component, offset) { |
|
var t = decodeHuffman(component.huffmanTableDC); |
|
var diff = t === 0 ? 0 : receiveAndExtend(t); |
|
component.blockData[offset] = (component.pred += diff); |
|
var k = 1; |
|
while (k < 64) { |
|
var rs = decodeHuffman(component.huffmanTableAC); |
|
var s = rs & 15, r = rs >> 4; |
|
if (s === 0) { |
|
if (r < 15) { |
|
break; |
|
} |
|
k += 16; |
|
continue; |
|
} |
|
k += r; |
|
var z = dctZigZag[k]; |
|
component.blockData[offset + z] = receiveAndExtend(s); |
|
k++; |
|
} |
|
} |
|
|
|
function decodeDCFirst(component, offset) { |
|
var t = decodeHuffman(component.huffmanTableDC); |
|
var diff = t === 0 ? 0 : (receiveAndExtend(t) << successive); |
|
component.blockData[offset] = (component.pred += diff); |
|
} |
|
|
|
function decodeDCSuccessive(component, offset) { |
|
component.blockData[offset] |= readBit() << successive; |
|
} |
|
|
|
var eobrun = 0; |
|
function decodeACFirst(component, offset) { |
|
if (eobrun > 0) { |
|
eobrun--; |
|
return; |
|
} |
|
var k = spectralStart, e = spectralEnd; |
|
while (k <= e) { |
|
var rs = decodeHuffman(component.huffmanTableAC); |
|
var s = rs & 15, r = rs >> 4; |
|
if (s === 0) { |
|
if (r < 15) { |
|
eobrun = receive(r) + (1 << r) - 1; |
|
break; |
|
} |
|
k += 16; |
|
continue; |
|
} |
|
k += r; |
|
var z = dctZigZag[k]; |
|
component.blockData[offset + z] = |
|
receiveAndExtend(s) * (1 << successive); |
|
k++; |
|
} |
|
} |
|
|
|
var successiveACState = 0, successiveACNextValue; |
|
function decodeACSuccessive(component, offset) { |
|
var k = spectralStart; |
|
var e = spectralEnd; |
|
var r = 0; |
|
var s; |
|
var rs; |
|
while (k <= e) { |
|
var z = dctZigZag[k]; |
|
switch (successiveACState) { |
|
case 0: // initial state |
|
rs = decodeHuffman(component.huffmanTableAC); |
|
s = rs & 15; |
|
r = rs >> 4; |
|
if (s === 0) { |
|
if (r < 15) { |
|
eobrun = receive(r) + (1 << r); |
|
successiveACState = 4; |
|
} else { |
|
r = 16; |
|
successiveACState = 1; |
|
} |
|
} else { |
|
if (s !== 1) { |
|
throw 'invalid ACn encoding'; |
|
} |
|
successiveACNextValue = receiveAndExtend(s); |
|
successiveACState = r ? 2 : 3; |
|
} |
|
continue; |
|
case 1: // skipping r zero items |
|
case 2: |
|
if (component.blockData[offset + z]) { |
|
component.blockData[offset + z] += (readBit() << successive); |
|
} else { |
|
r--; |
|
if (r === 0) { |
|
successiveACState = successiveACState === 2 ? 3 : 0; |
|
} |
|
} |
|
break; |
|
case 3: // set value for a zero item |
|
if (component.blockData[offset + z]) { |
|
component.blockData[offset + z] += (readBit() << successive); |
|
} else { |
|
component.blockData[offset + z] = |
|
successiveACNextValue << successive; |
|
successiveACState = 0; |
|
} |
|
break; |
|
case 4: // eob |
|
if (component.blockData[offset + z]) { |
|
component.blockData[offset + z] += (readBit() << successive); |
|
} |
|
break; |
|
} |
|
k++; |
|
} |
|
if (successiveACState === 4) { |
|
eobrun--; |
|
if (eobrun === 0) { |
|
successiveACState = 0; |
|
} |
|
} |
|
} |
|
|
|
function decodeMcu(component, decode, mcu, row, col) { |
|
var mcuRow = (mcu / mcusPerLine) | 0; |
|
var mcuCol = mcu % mcusPerLine; |
|
var blockRow = mcuRow * component.v + row; |
|
var blockCol = mcuCol * component.h + col; |
|
var offset = getBlockBufferOffset(component, blockRow, blockCol); |
|
decode(component, offset); |
|
} |
|
|
|
function decodeBlock(component, decode, mcu) { |
|
var blockRow = (mcu / component.blocksPerLine) | 0; |
|
var blockCol = mcu % component.blocksPerLine; |
|
var offset = getBlockBufferOffset(component, blockRow, blockCol); |
|
decode(component, offset); |
|
} |
|
|
|
var componentsLength = components.length; |
|
var component, i, j, k, n; |
|
var decodeFn; |
|
if (progressive) { |
|
if (spectralStart === 0) { |
|
decodeFn = successivePrev === 0 ? decodeDCFirst : decodeDCSuccessive; |
|
} else { |
|
decodeFn = successivePrev === 0 ? decodeACFirst : decodeACSuccessive; |
|
} |
|
} else { |
|
decodeFn = decodeBaseline; |
|
} |
|
|
|
var mcu = 0, marker; |
|
var mcuExpected; |
|
if (componentsLength === 1) { |
|
mcuExpected = components[0].blocksPerLine * components[0].blocksPerColumn; |
|
} else { |
|
mcuExpected = mcusPerLine * frame.mcusPerColumn; |
|
} |
|
if (!resetInterval) { |
|
resetInterval = mcuExpected; |
|
} |
|
|
|
var h, v; |
|
while (mcu < mcuExpected) { |
|
// reset interval stuff |
|
for (i = 0; i < componentsLength; i++) { |
|
components[i].pred = 0; |
|
} |
|
eobrun = 0; |
|
|
|
if (componentsLength === 1) { |
|
component = components[0]; |
|
for (n = 0; n < resetInterval; n++) { |
|
decodeBlock(component, decodeFn, mcu); |
|
mcu++; |
|
} |
|
} else { |
|
for (n = 0; n < resetInterval; n++) { |
|
for (i = 0; i < componentsLength; i++) { |
|
component = components[i]; |
|
h = component.h; |
|
v = component.v; |
|
for (j = 0; j < v; j++) { |
|
for (k = 0; k < h; k++) { |
|
decodeMcu(component, decodeFn, mcu, j, k); |
|
} |
|
} |
|
} |
|
mcu++; |
|
} |
|
} |
|
|
|
// find marker |
|
bitsCount = 0; |
|
marker = (data[offset] << 8) | data[offset + 1]; |
|
if (marker <= 0xFF00) { |
|
throw 'marker was not found'; |
|
} |
|
|
|
if (marker >= 0xFFD0 && marker <= 0xFFD7) { // RSTx |
|
offset += 2; |
|
} else { |
|
break; |
|
} |
|
} |
|
|
|
return offset - startOffset; |
|
} |
|
|
|
// A port of poppler's IDCT method which in turn is taken from: |
|
// Christoph Loeffler, Adriaan Ligtenberg, George S. Moschytz, |
|
// 'Practical Fast 1-D DCT Algorithms with 11 Multiplications', |
|
// IEEE Intl. Conf. on Acoustics, Speech & Signal Processing, 1989, |
|
// 988-991. |
|
function quantizeAndInverse(component, blockBufferOffset, p) { |
|
var qt = component.quantizationTable, blockData = component.blockData; |
|
var v0, v1, v2, v3, v4, v5, v6, v7; |
|
var p0, p1, p2, p3, p4, p5, p6, p7; |
|
var t; |
|
|
|
// inverse DCT on rows |
|
for (var row = 0; row < 64; row += 8) { |
|
// gather block data |
|
p0 = blockData[blockBufferOffset + row]; |
|
p1 = blockData[blockBufferOffset + row + 1]; |
|
p2 = blockData[blockBufferOffset + row + 2]; |
|
p3 = blockData[blockBufferOffset + row + 3]; |
|
p4 = blockData[blockBufferOffset + row + 4]; |
|
p5 = blockData[blockBufferOffset + row + 5]; |
|
p6 = blockData[blockBufferOffset + row + 6]; |
|
p7 = blockData[blockBufferOffset + row + 7]; |
|
|
|
// dequant p0 |
|
p0 *= qt[row]; |
|
|
|
// check for all-zero AC coefficients |
|
if ((p1 | p2 | p3 | p4 | p5 | p6 | p7) === 0) { |
|
t = (dctSqrt2 * p0 + 512) >> 10; |
|
p[row] = t; |
|
p[row + 1] = t; |
|
p[row + 2] = t; |
|
p[row + 3] = t; |
|
p[row + 4] = t; |
|
p[row + 5] = t; |
|
p[row + 6] = t; |
|
p[row + 7] = t; |
|
continue; |
|
} |
|
// dequant p1 ... p7 |
|
p1 *= qt[row + 1]; |
|
p2 *= qt[row + 2]; |
|
p3 *= qt[row + 3]; |
|
p4 *= qt[row + 4]; |
|
p5 *= qt[row + 5]; |
|
p6 *= qt[row + 6]; |
|
p7 *= qt[row + 7]; |
|
|
|
// stage 4 |
|
v0 = (dctSqrt2 * p0 + 128) >> 8; |
|
v1 = (dctSqrt2 * p4 + 128) >> 8; |
|
v2 = p2; |
|
v3 = p6; |
|
v4 = (dctSqrt1d2 * (p1 - p7) + 128) >> 8; |
|
v7 = (dctSqrt1d2 * (p1 + p7) + 128) >> 8; |
|
v5 = p3 << 4; |
|
v6 = p5 << 4; |
|
|
|
// stage 3 |
|
v0 = (v0 + v1 + 1) >> 1; |
|
v1 = v0 - v1; |
|
t = (v2 * dctSin6 + v3 * dctCos6 + 128) >> 8; |
|
v2 = (v2 * dctCos6 - v3 * dctSin6 + 128) >> 8; |
|
v3 = t; |
|
v4 = (v4 + v6 + 1) >> 1; |
|
v6 = v4 - v6; |
|
v7 = (v7 + v5 + 1) >> 1; |
|
v5 = v7 - v5; |
|
|
|
// stage 2 |
|
v0 = (v0 + v3 + 1) >> 1; |
|
v3 = v0 - v3; |
|
v1 = (v1 + v2 + 1) >> 1; |
|
v2 = v1 - v2; |
|
t = (v4 * dctSin3 + v7 * dctCos3 + 2048) >> 12; |
|
v4 = (v4 * dctCos3 - v7 * dctSin3 + 2048) >> 12; |
|
v7 = t; |
|
t = (v5 * dctSin1 + v6 * dctCos1 + 2048) >> 12; |
|
v5 = (v5 * dctCos1 - v6 * dctSin1 + 2048) >> 12; |
|
v6 = t; |
|
|
|
// stage 1 |
|
p[row] = v0 + v7; |
|
p[row + 7] = v0 - v7; |
|
p[row + 1] = v1 + v6; |
|
p[row + 6] = v1 - v6; |
|
p[row + 2] = v2 + v5; |
|
p[row + 5] = v2 - v5; |
|
p[row + 3] = v3 + v4; |
|
p[row + 4] = v3 - v4; |
|
} |
|
|
|
// inverse DCT on columns |
|
for (var col = 0; col < 8; ++col) { |
|
p0 = p[col]; |
|
p1 = p[col + 8]; |
|
p2 = p[col + 16]; |
|
p3 = p[col + 24]; |
|
p4 = p[col + 32]; |
|
p5 = p[col + 40]; |
|
p6 = p[col + 48]; |
|
p7 = p[col + 56]; |
|
|
|
// check for all-zero AC coefficients |
|
if ((p1 | p2 | p3 | p4 | p5 | p6 | p7) === 0) { |
|
t = (dctSqrt2 * p0 + 8192) >> 14; |
|
// convert to 8 bit |
|
t = (t < -2040) ? 0 : (t >= 2024) ? 255 : (t + 2056) >> 4; |
|
blockData[blockBufferOffset + col] = t; |
|
blockData[blockBufferOffset + col + 8] = t; |
|
blockData[blockBufferOffset + col + 16] = t; |
|
blockData[blockBufferOffset + col + 24] = t; |
|
blockData[blockBufferOffset + col + 32] = t; |
|
blockData[blockBufferOffset + col + 40] = t; |
|
blockData[blockBufferOffset + col + 48] = t; |
|
blockData[blockBufferOffset + col + 56] = t; |
|
continue; |
|
} |
|
|
|
// stage 4 |
|
v0 = (dctSqrt2 * p0 + 2048) >> 12; |
|
v1 = (dctSqrt2 * p4 + 2048) >> 12; |
|
v2 = p2; |
|
v3 = p6; |
|
v4 = (dctSqrt1d2 * (p1 - p7) + 2048) >> 12; |
|
v7 = (dctSqrt1d2 * (p1 + p7) + 2048) >> 12; |
|
v5 = p3; |
|
v6 = p5; |
|
|
|
// stage 3 |
|
// Shift v0 by 128.5 << 5 here, so we don't need to shift p0...p7 when |
|
// converting to UInt8 range later. |
|
v0 = ((v0 + v1 + 1) >> 1) + 4112; |
|
v1 = v0 - v1; |
|
t = (v2 * dctSin6 + v3 * dctCos6 + 2048) >> 12; |
|
v2 = (v2 * dctCos6 - v3 * dctSin6 + 2048) >> 12; |
|
v3 = t; |
|
v4 = (v4 + v6 + 1) >> 1; |
|
v6 = v4 - v6; |
|
v7 = (v7 + v5 + 1) >> 1; |
|
v5 = v7 - v5; |
|
|
|
// stage 2 |
|
v0 = (v0 + v3 + 1) >> 1; |
|
v3 = v0 - v3; |
|
v1 = (v1 + v2 + 1) >> 1; |
|
v2 = v1 - v2; |
|
t = (v4 * dctSin3 + v7 * dctCos3 + 2048) >> 12; |
|
v4 = (v4 * dctCos3 - v7 * dctSin3 + 2048) >> 12; |
|
v7 = t; |
|
t = (v5 * dctSin1 + v6 * dctCos1 + 2048) >> 12; |
|
v5 = (v5 * dctCos1 - v6 * dctSin1 + 2048) >> 12; |
|
v6 = t; |
|
|
|
// stage 1 |
|
p0 = v0 + v7; |
|
p7 = v0 - v7; |
|
p1 = v1 + v6; |
|
p6 = v1 - v6; |
|
p2 = v2 + v5; |
|
p5 = v2 - v5; |
|
p3 = v3 + v4; |
|
p4 = v3 - v4; |
|
|
|
// convert to 8-bit integers |
|
p0 = (p0 < 16) ? 0 : (p0 >= 4080) ? 255 : p0 >> 4; |
|
p1 = (p1 < 16) ? 0 : (p1 >= 4080) ? 255 : p1 >> 4; |
|
p2 = (p2 < 16) ? 0 : (p2 >= 4080) ? 255 : p2 >> 4; |
|
p3 = (p3 < 16) ? 0 : (p3 >= 4080) ? 255 : p3 >> 4; |
|
p4 = (p4 < 16) ? 0 : (p4 >= 4080) ? 255 : p4 >> 4; |
|
p5 = (p5 < 16) ? 0 : (p5 >= 4080) ? 255 : p5 >> 4; |
|
p6 = (p6 < 16) ? 0 : (p6 >= 4080) ? 255 : p6 >> 4; |
|
p7 = (p7 < 16) ? 0 : (p7 >= 4080) ? 255 : p7 >> 4; |
|
|
|
// store block data |
|
blockData[blockBufferOffset + col] = p0; |
|
blockData[blockBufferOffset + col + 8] = p1; |
|
blockData[blockBufferOffset + col + 16] = p2; |
|
blockData[blockBufferOffset + col + 24] = p3; |
|
blockData[blockBufferOffset + col + 32] = p4; |
|
blockData[blockBufferOffset + col + 40] = p5; |
|
blockData[blockBufferOffset + col + 48] = p6; |
|
blockData[blockBufferOffset + col + 56] = p7; |
|
} |
|
} |
|
|
|
function buildComponentData(frame, component) { |
|
var blocksPerLine = component.blocksPerLine; |
|
var blocksPerColumn = component.blocksPerColumn; |
|
var computationBuffer = new Int16Array(64); |
|
|
|
for (var blockRow = 0; blockRow < blocksPerColumn; blockRow++) { |
|
for (var blockCol = 0; blockCol < blocksPerLine; blockCol++) { |
|
var offset = getBlockBufferOffset(component, blockRow, blockCol); |
|
quantizeAndInverse(component, offset, computationBuffer); |
|
} |
|
} |
|
return component.blockData; |
|
} |
|
|
|
function clamp0to255(a) { |
|
return a <= 0 ? 0 : a >= 255 ? 255 : a; |
|
} |
|
|
|
constructor.prototype = { |
|
parse: function parse(data) { |
|
|
|
function readUint16() { |
|
var value = (data[offset] << 8) | data[offset + 1]; |
|
offset += 2; |
|
return value; |
|
} |
|
|
|
function readDataBlock() { |
|
var length = readUint16(); |
|
var array = data.subarray(offset, offset + length - 2); |
|
offset += array.length; |
|
return array; |
|
} |
|
|
|
function prepareComponents(frame) { |
|
var mcusPerLine = Math.ceil(frame.samplesPerLine / 8 / frame.maxH); |
|
var mcusPerColumn = Math.ceil(frame.scanLines / 8 / frame.maxV); |
|
for (var i = 0; i < frame.components.length; i++) { |
|
component = frame.components[i]; |
|
var blocksPerLine = Math.ceil(Math.ceil(frame.samplesPerLine / 8) * |
|
component.h / frame.maxH); |
|
var blocksPerColumn = Math.ceil(Math.ceil(frame.scanLines / 8) * |
|
component.v / frame.maxV); |
|
var blocksPerLineForMcu = mcusPerLine * component.h; |
|
var blocksPerColumnForMcu = mcusPerColumn * component.v; |
|
|
|
var blocksBufferSize = 64 * blocksPerColumnForMcu * |
|
(blocksPerLineForMcu + 1); |
|
component.blockData = new Int16Array(blocksBufferSize); |
|
component.blocksPerLine = blocksPerLine; |
|
component.blocksPerColumn = blocksPerColumn; |
|
} |
|
frame.mcusPerLine = mcusPerLine; |
|
frame.mcusPerColumn = mcusPerColumn; |
|
} |
|
|
|
var offset = 0; |
|
var jfif = null; |
|
var adobe = null; |
|
var frame, resetInterval; |
|
var quantizationTables = []; |
|
var huffmanTablesAC = [], huffmanTablesDC = []; |
|
var fileMarker = readUint16(); |
|
if (fileMarker !== 0xFFD8) { // SOI (Start of Image) |
|
throw 'SOI not found'; |
|
} |
|
|
|
fileMarker = readUint16(); |
|
while (fileMarker !== 0xFFD9) { // EOI (End of image) |
|
var i, j, l; |
|
switch(fileMarker) { |
|
case 0xFFE0: // APP0 (Application Specific) |
|
case 0xFFE1: // APP1 |
|
case 0xFFE2: // APP2 |
|
case 0xFFE3: // APP3 |
|
case 0xFFE4: // APP4 |
|
case 0xFFE5: // APP5 |
|
case 0xFFE6: // APP6 |
|
case 0xFFE7: // APP7 |
|
case 0xFFE8: // APP8 |
|
case 0xFFE9: // APP9 |
|
case 0xFFEA: // APP10 |
|
case 0xFFEB: // APP11 |
|
case 0xFFEC: // APP12 |
|
case 0xFFED: // APP13 |
|
case 0xFFEE: // APP14 |
|
case 0xFFEF: // APP15 |
|
case 0xFFFE: // COM (Comment) |
|
var appData = readDataBlock(); |
|
|
|
if (fileMarker === 0xFFE0) { |
|
if (appData[0] === 0x4A && appData[1] === 0x46 && |
|
appData[2] === 0x49 && appData[3] === 0x46 && |
|
appData[4] === 0) { // 'JFIF\x00' |
|
jfif = { |
|
version: { major: appData[5], minor: appData[6] }, |
|
densityUnits: appData[7], |
|
xDensity: (appData[8] << 8) | appData[9], |
|
yDensity: (appData[10] << 8) | appData[11], |
|
thumbWidth: appData[12], |
|
thumbHeight: appData[13], |
|
thumbData: appData.subarray(14, 14 + |
|
3 * appData[12] * appData[13]) |
|
}; |
|
} |
|
} |
|
// TODO APP1 - Exif |
|
if (fileMarker === 0xFFEE) { |
|
if (appData[0] === 0x41 && appData[1] === 0x64 && |
|
appData[2] === 0x6F && appData[3] === 0x62 && |
|
appData[4] === 0x65) { // 'Adobe' |
|
adobe = { |
|
version: (appData[5] << 8) | appData[6], |
|
flags0: (appData[7] << 8) | appData[8], |
|
flags1: (appData[9] << 8) | appData[10], |
|
transformCode: appData[11] |
|
}; |
|
} |
|
} |
|
break; |
|
|
|
case 0xFFDB: // DQT (Define Quantization Tables) |
|
var quantizationTablesLength = readUint16(); |
|
var quantizationTablesEnd = quantizationTablesLength + offset - 2; |
|
var z; |
|
while (offset < quantizationTablesEnd) { |
|
var quantizationTableSpec = data[offset++]; |
|
var tableData = new Uint16Array(64); |
|
if ((quantizationTableSpec >> 4) === 0) { // 8 bit values |
|
for (j = 0; j < 64; j++) { |
|
z = dctZigZag[j]; |
|
tableData[z] = data[offset++]; |
|
} |
|
} else if ((quantizationTableSpec >> 4) === 1) { //16 bit |
|
for (j = 0; j < 64; j++) { |
|
z = dctZigZag[j]; |
|
tableData[z] = readUint16(); |
|
} |
|
} else { |
|
throw 'DQT: invalid table spec'; |
|
} |
|
quantizationTables[quantizationTableSpec & 15] = tableData; |
|
} |
|
break; |
|
|
|
case 0xFFC0: // SOF0 (Start of Frame, Baseline DCT) |
|
case 0xFFC1: // SOF1 (Start of Frame, Extended DCT) |
|
case 0xFFC2: // SOF2 (Start of Frame, Progressive DCT) |
|
if (frame) { |
|
throw 'Only single frame JPEGs supported'; |
|
} |
|
readUint16(); // skip data length |
|
frame = {}; |
|
frame.extended = (fileMarker === 0xFFC1); |
|
frame.progressive = (fileMarker === 0xFFC2); |
|
frame.precision = data[offset++]; |
|
frame.scanLines = readUint16(); |
|
frame.samplesPerLine = readUint16(); |
|
frame.components = []; |
|
frame.componentIds = {}; |
|
var componentsCount = data[offset++], componentId; |
|
var maxH = 0, maxV = 0; |
|
for (i = 0; i < componentsCount; i++) { |
|
componentId = data[offset]; |
|
var h = data[offset + 1] >> 4; |
|
var v = data[offset + 1] & 15; |
|
if (maxH < h) { |
|
maxH = h; |
|
} |
|
if (maxV < v) { |
|
maxV = v; |
|
} |
|
var qId = data[offset + 2]; |
|
l = frame.components.push({ |
|
h: h, |
|
v: v, |
|
quantizationTable: quantizationTables[qId] |
|
}); |
|
frame.componentIds[componentId] = l - 1; |
|
offset += 3; |
|
} |
|
frame.maxH = maxH; |
|
frame.maxV = maxV; |
|
prepareComponents(frame); |
|
break; |
|
|
|
case 0xFFC4: // DHT (Define Huffman Tables) |
|
var huffmanLength = readUint16(); |
|
for (i = 2; i < huffmanLength;) { |
|
var huffmanTableSpec = data[offset++]; |
|
var codeLengths = new Uint8Array(16); |
|
var codeLengthSum = 0; |
|
for (j = 0; j < 16; j++, offset++) { |
|
codeLengthSum += (codeLengths[j] = data[offset]); |
|
} |
|
var huffmanValues = new Uint8Array(codeLengthSum); |
|
for (j = 0; j < codeLengthSum; j++, offset++) { |
|
huffmanValues[j] = data[offset]; |
|
} |
|
i += 17 + codeLengthSum; |
|
|
|
((huffmanTableSpec >> 4) === 0 ? |
|
huffmanTablesDC : huffmanTablesAC)[huffmanTableSpec & 15] = |
|
buildHuffmanTable(codeLengths, huffmanValues); |
|
} |
|
break; |
|
|
|
case 0xFFDD: // DRI (Define Restart Interval) |
|
readUint16(); // skip data length |
|
resetInterval = readUint16(); |
|
break; |
|
|
|
case 0xFFDA: // SOS (Start of Scan) |
|
var scanLength = readUint16(); |
|
var selectorsCount = data[offset++]; |
|
var components = [], component; |
|
for (i = 0; i < selectorsCount; i++) { |
|
var componentIndex = frame.componentIds[data[offset++]]; |
|
component = frame.components[componentIndex]; |
|
var tableSpec = data[offset++]; |
|
component.huffmanTableDC = huffmanTablesDC[tableSpec >> 4]; |
|
component.huffmanTableAC = huffmanTablesAC[tableSpec & 15]; |
|
components.push(component); |
|
} |
|
var spectralStart = data[offset++]; |
|
var spectralEnd = data[offset++]; |
|
var successiveApproximation = data[offset++]; |
|
var processed = decodeScan(data, offset, |
|
frame, components, resetInterval, |
|
spectralStart, spectralEnd, |
|
successiveApproximation >> 4, successiveApproximation & 15); |
|
offset += processed; |
|
break; |
|
|
|
case 0xFFFF: // Fill bytes |
|
if (data[offset] !== 0xFF) { // Avoid skipping a valid marker. |
|
offset--; |
|
} |
|
break; |
|
|
|
default: |
|
if (data[offset - 3] === 0xFF && |
|
data[offset - 2] >= 0xC0 && data[offset - 2] <= 0xFE) { |
|
// could be incorrect encoding -- last 0xFF byte of the previous |
|
// block was eaten by the encoder |
|
offset -= 3; |
|
break; |
|
} |
|
throw 'unknown JPEG marker ' + fileMarker.toString(16); |
|
} |
|
fileMarker = readUint16(); |
|
} |
|
|
|
this.width = frame.samplesPerLine; |
|
this.height = frame.scanLines; |
|
this.jfif = jfif; |
|
this.adobe = adobe; |
|
this.components = []; |
|
for (i = 0; i < frame.components.length; i++) { |
|
component = frame.components[i]; |
|
this.components.push({ |
|
output: buildComponentData(frame, component), |
|
scaleX: component.h / frame.maxH, |
|
scaleY: component.v / frame.maxV, |
|
blocksPerLine: component.blocksPerLine, |
|
blocksPerColumn: component.blocksPerColumn |
|
}); |
|
} |
|
this.numComponents = this.components.length; |
|
}, |
|
|
|
_getLinearizedBlockData: function getLinearizedBlockData(width, height) { |
|
var scaleX = this.width / width, scaleY = this.height / height; |
|
|
|
var component, componentScaleX, componentScaleY, blocksPerScanline; |
|
var x, y, i, j, k; |
|
var index; |
|
var offset = 0; |
|
var output; |
|
var numComponents = this.components.length; |
|
var dataLength = width * height * numComponents; |
|
var data = new Uint8Array(dataLength); |
|
var xScaleBlockOffset = new Uint32Array(width); |
|
var mask3LSB = 0xfffffff8; // used to clear the 3 LSBs |
|
|
|
for (i = 0; i < numComponents; i++) { |
|
component = this.components[i]; |
|
componentScaleX = component.scaleX * scaleX; |
|
componentScaleY = component.scaleY * scaleY; |
|
offset = i; |
|
output = component.output; |
|
blocksPerScanline = (component.blocksPerLine + 1) << 3; |
|
// precalculate the xScaleBlockOffset |
|
for (x = 0; x < width; x++) { |
|
j = 0 | (x * componentScaleX); |
|
xScaleBlockOffset[x] = ((j & mask3LSB) << 3) | (j & 7); |
|
} |
|
// linearize the blocks of the component |
|
for (y = 0; y < height; y++) { |
|
j = 0 | (y * componentScaleY); |
|
index = blocksPerScanline * (j & mask3LSB) | ((j & 7) << 3); |
|
for (x = 0; x < width; x++) { |
|
data[offset] = output[index + xScaleBlockOffset[x]]; |
|
offset += numComponents; |
|
} |
|
} |
|
} |
|
|
|
// decodeTransform contains pairs of multiplier (-256..256) and additive |
|
var transform = this.decodeTransform; |
|
if (transform) { |
|
for (i = 0; i < dataLength;) { |
|
for (j = 0, k = 0; j < numComponents; j++, i++, k += 2) { |
|
data[i] = ((data[i] * transform[k]) >> 8) + transform[k + 1]; |
|
} |
|
} |
|
} |
|
return data; |
|
}, |
|
|
|
_isColorConversionNeeded: function isColorConversionNeeded() { |
|
if (this.adobe && this.adobe.transformCode) { |
|
// The adobe transform marker overrides any previous setting |
|
return true; |
|
} else if (this.numComponents === 3) { |
|
return true; |
|
} else { |
|
return false; |
|
} |
|
}, |
|
|
|
_convertYccToRgb: function convertYccToRgb(data) { |
|
var Y, Cb, Cr; |
|
for (var i = 0, length = data.length; i < length; i += 3) { |
|
Y = data[i ]; |
|
Cb = data[i + 1]; |
|
Cr = data[i + 2]; |
|
data[i ] = clamp0to255(Y - 179.456 + 1.402 * Cr); |
|
data[i + 1] = clamp0to255(Y + 135.459 - 0.344 * Cb - 0.714 * Cr); |
|
data[i + 2] = clamp0to255(Y - 226.816 + 1.772 * Cb); |
|
} |
|
return data; |
|
}, |
|
|
|
_convertYcckToRgb: function convertYcckToRgb(data) { |
|
var Y, Cb, Cr, k; |
|
var offset = 0; |
|
for (var i = 0, length = data.length; i < length; i += 4) { |
|
Y = data[i]; |
|
Cb = data[i + 1]; |
|
Cr = data[i + 2]; |
|
k = data[i + 3]; |
|
|
|
var r = -122.67195406894 + |
|
Cb * (-6.60635669420364e-5 * Cb + 0.000437130475926232 * Cr - |
|
5.4080610064599e-5 * Y + 0.00048449797120281 * k - |
|
0.154362151871126) + |
|
Cr * (-0.000957964378445773 * Cr + 0.000817076911346625 * Y - |
|
0.00477271405408747 * k + 1.53380253221734) + |
|
Y * (0.000961250184130688 * Y - 0.00266257332283933 * k + |
|
0.48357088451265) + |
|
k * (-0.000336197177618394 * k + 0.484791561490776); |
|
|
|
var g = 107.268039397724 + |
|
Cb * (2.19927104525741e-5 * Cb - 0.000640992018297945 * Cr + |
|
0.000659397001245577 * Y + 0.000426105652938837 * k - |
|
0.176491792462875) + |
|
Cr * (-0.000778269941513683 * Cr + 0.00130872261408275 * Y + |
|
0.000770482631801132 * k - 0.151051492775562) + |
|
Y * (0.00126935368114843 * Y - 0.00265090189010898 * k + |
|
0.25802910206845) + |
|
k * (-0.000318913117588328 * k - 0.213742400323665); |
|
|
|
var b = -20.810012546947 + |
|
Cb * (-0.000570115196973677 * Cb - 2.63409051004589e-5 * Cr + |
|
0.0020741088115012 * Y - 0.00288260236853442 * k + |
|
0.814272968359295) + |
|
Cr * (-1.53496057440975e-5 * Cr - 0.000132689043961446 * Y + |
|
0.000560833691242812 * k - 0.195152027534049) + |
|
Y * (0.00174418132927582 * Y - 0.00255243321439347 * k + |
|
0.116935020465145) + |
|
k * (-0.000343531996510555 * k + 0.24165260232407); |
|
|
|
data[offset++] = clamp0to255(r); |
|
data[offset++] = clamp0to255(g); |
|
data[offset++] = clamp0to255(b); |
|
} |
|
return data; |
|
}, |
|
|
|
_convertYcckToCmyk: function convertYcckToCmyk(data) { |
|
var Y, Cb, Cr; |
|
for (var i = 0, length = data.length; i < length; i += 4) { |
|
Y = data[i]; |
|
Cb = data[i + 1]; |
|
Cr = data[i + 2]; |
|
data[i ] = clamp0to255(434.456 - Y - 1.402 * Cr); |
|
data[i + 1] = clamp0to255(119.541 - Y + 0.344 * Cb + 0.714 * Cr); |
|
data[i + 2] = clamp0to255(481.816 - Y - 1.772 * Cb); |
|
// K in data[i + 3] is unchanged |
|
} |
|
return data; |
|
}, |
|
|
|
_convertCmykToRgb: function convertCmykToRgb(data) { |
|
var c, m, y, k; |
|
var offset = 0; |
|
var min = -255 * 255 * 255; |
|
var scale = 1 / 255 / 255; |
|
for (var i = 0, length = data.length; i < length; i += 4) { |
|
c = data[i]; |
|
m = data[i + 1]; |
|
y = data[i + 2]; |
|
k = data[i + 3]; |
|
|
|
var r = |
|
c * (-4.387332384609988 * c + 54.48615194189176 * m + |
|
18.82290502165302 * y + 212.25662451639585 * k - |
|
72734.4411664936) + |
|
m * (1.7149763477362134 * m - 5.6096736904047315 * y - |
|
17.873870861415444 * k - 1401.7366389350734) + |
|
y * (-2.5217340131683033 * y - 21.248923337353073 * k + |
|
4465.541406466231) - |
|
k * (21.86122147463605 * k + 48317.86113160301); |
|
var g = |
|
c * (8.841041422036149 * c + 60.118027045597366 * m + |
|
6.871425592049007 * y + 31.159100130055922 * k - |
|
20220.756542821975) + |
|
m * (-15.310361306967817 * m + 17.575251261109482 * y + |
|
131.35250912493976 * k - 48691.05921601825) + |
|
y * (4.444339102852739 * y + 9.8632861493405 * k - |
|
6341.191035517494) - |
|
k * (20.737325471181034 * k + 47890.15695978492); |
|
var b = |
|
c * (0.8842522430003296 * c + 8.078677503112928 * m + |
|
30.89978309703729 * y - 0.23883238689178934 * k - |
|
3616.812083916688) + |
|
m * (10.49593273432072 * m + 63.02378494754052 * y + |
|
50.606957656360734 * k - 28620.90484698408) + |
|
y * (0.03296041114873217 * y + 115.60384449646641 * k - |
|
49363.43385999684) - |
|
k * (22.33816807309886 * k + 45932.16563550634); |
|
|
|
data[offset++] = r >= 0 ? 255 : r <= min ? 0 : 255 + r * scale | 0; |
|
data[offset++] = g >= 0 ? 255 : g <= min ? 0 : 255 + g * scale | 0; |
|
data[offset++] = b >= 0 ? 255 : b <= min ? 0 : 255 + b * scale | 0; |
|
} |
|
return data; |
|
}, |
|
|
|
getData: function getData(width, height, forceRGBoutput) { |
|
if (this.numComponents > 4) { |
|
throw 'Unsupported color mode'; |
|
} |
|
// type of data: Uint8Array(width * height * numComponents) |
|
var data = this._getLinearizedBlockData(width, height); |
|
|
|
if (this.numComponents === 3) { |
|
return this._convertYccToRgb(data); |
|
} else if (this.numComponents === 4) { |
|
if (this._isColorConversionNeeded()) { |
|
if (forceRGBoutput) { |
|
return this._convertYcckToRgb(data); |
|
} else { |
|
return this._convertYcckToCmyk(data); |
|
} |
|
} else if (forceRGBoutput) { |
|
return this._convertCmykToRgb(data); |
|
} |
|
} |
|
return data; |
|
} |
|
}; |
|
|
|
return constructor; |
|
})(); |
|
|
|
exports.JpegImage = JpegImage; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreMetrics = {})); |
|
} |
|
}(this, function (exports) { |
|
|
|
// The Metrics object contains glyph widths (in glyph space units). |
|
// As per PDF spec, for most fonts (Type 3 being an exception) a glyph |
|
// space unit corresponds to 1/1000th of text space unit. |
|
var Metrics = { |
|
'Courier': 600, |
|
'Courier-Bold': 600, |
|
'Courier-BoldOblique': 600, |
|
'Courier-Oblique': 600, |
|
'Helvetica' : { |
|
'space': 278, |
|
'exclam': 278, |
|
'quotedbl': 355, |
|
'numbersign': 556, |
|
'dollar': 556, |
|
'percent': 889, |
|
'ampersand': 667, |
|
'quoteright': 222, |
|
'parenleft': 333, |
|
'parenright': 333, |
|
'asterisk': 389, |
|
'plus': 584, |
|
'comma': 278, |
|
'hyphen': 333, |
|
'period': 278, |
|
'slash': 278, |
|
'zero': 556, |
|
'one': 556, |
|
'two': 556, |
|
'three': 556, |
|
'four': 556, |
|
'five': 556, |
|
'six': 556, |
|
'seven': 556, |
|
'eight': 556, |
|
'nine': 556, |
|
'colon': 278, |
|
'semicolon': 278, |
|
'less': 584, |
|
'equal': 584, |
|
'greater': 584, |
|
'question': 556, |
|
'at': 1015, |
|
'A': 667, |
|
'B': 667, |
|
'C': 722, |
|
'D': 722, |
|
'E': 667, |
|
'F': 611, |
|
'G': 778, |
|
'H': 722, |
|
'I': 278, |
|
'J': 500, |
|
'K': 667, |
|
'L': 556, |
|
'M': 833, |
|
'N': 722, |
|
'O': 778, |
|
'P': 667, |
|
'Q': 778, |
|
'R': 722, |
|
'S': 667, |
|
'T': 611, |
|
'U': 722, |
|
'V': 667, |
|
'W': 944, |
|
'X': 667, |
|
'Y': 667, |
|
'Z': 611, |
|
'bracketleft': 278, |
|
'backslash': 278, |
|
'bracketright': 278, |
|
'asciicircum': 469, |
|
'underscore': 556, |
|
'quoteleft': 222, |
|
'a': 556, |
|
'b': 556, |
|
'c': 500, |
|
'd': 556, |
|
'e': 556, |
|
'f': 278, |
|
'g': 556, |
|
'h': 556, |
|
'i': 222, |
|
'j': 222, |
|
'k': 500, |
|
'l': 222, |
|
'm': 833, |
|
'n': 556, |
|
'o': 556, |
|
'p': 556, |
|
'q': 556, |
|
'r': 333, |
|
's': 500, |
|
't': 278, |
|
'u': 556, |
|
'v': 500, |
|
'w': 722, |
|
'x': 500, |
|
'y': 500, |
|
'z': 500, |
|
'braceleft': 334, |
|
'bar': 260, |
|
'braceright': 334, |
|
'asciitilde': 584, |
|
'exclamdown': 333, |
|
'cent': 556, |
|
'sterling': 556, |
|
'fraction': 167, |
|
'yen': 556, |
|
'florin': 556, |
|
'section': 556, |
|
'currency': 556, |
|
'quotesingle': 191, |
|
'quotedblleft': 333, |
|
'guillemotleft': 556, |
|
'guilsinglleft': 333, |
|
'guilsinglright': 333, |
|
'fi': 500, |
|
'fl': 500, |
|
'endash': 556, |
|
'dagger': 556, |
|
'daggerdbl': 556, |
|
'periodcentered': 278, |
|
'paragraph': 537, |
|
'bullet': 350, |
|
'quotesinglbase': 222, |
|
'quotedblbase': 333, |
|
'quotedblright': 333, |
|
'guillemotright': 556, |
|
'ellipsis': 1000, |
|
'perthousand': 1000, |
|
'questiondown': 611, |
|
'grave': 333, |
|
'acute': 333, |
|
'circumflex': 333, |
|
'tilde': 333, |
|
'macron': 333, |
|
'breve': 333, |
|
'dotaccent': 333, |
|
'dieresis': 333, |
|
'ring': 333, |
|
'cedilla': 333, |
|
'hungarumlaut': 333, |
|
'ogonek': 333, |
|
'caron': 333, |
|
'emdash': 1000, |
|
'AE': 1000, |
|
'ordfeminine': 370, |
|
'Lslash': 556, |
|
'Oslash': 778, |
|
'OE': 1000, |
|
'ordmasculine': 365, |
|
'ae': 889, |
|
'dotlessi': 278, |
|
'lslash': 222, |
|
'oslash': 611, |
|
'oe': 944, |
|
'germandbls': 611, |
|
'Idieresis': 278, |
|
'eacute': 556, |
|
'abreve': 556, |
|
'uhungarumlaut': 556, |
|
'ecaron': 556, |
|
'Ydieresis': 667, |
|
'divide': 584, |
|
'Yacute': 667, |
|
'Acircumflex': 667, |
|
'aacute': 556, |
|
'Ucircumflex': 722, |
|
'yacute': 500, |
|
'scommaaccent': 500, |
|
'ecircumflex': 556, |
|
'Uring': 722, |
|
'Udieresis': 722, |
|
'aogonek': 556, |
|
'Uacute': 722, |
|
'uogonek': 556, |
|
'Edieresis': 667, |
|
'Dcroat': 722, |
|
'commaaccent': 250, |
|
'copyright': 737, |
|
'Emacron': 667, |
|
'ccaron': 500, |
|
'aring': 556, |
|
'Ncommaaccent': 722, |
|
'lacute': 222, |
|
'agrave': 556, |
|
'Tcommaaccent': 611, |
|
'Cacute': 722, |
|
'atilde': 556, |
|
'Edotaccent': 667, |
|
'scaron': 500, |
|
'scedilla': 500, |
|
'iacute': 278, |
|
'lozenge': 471, |
|
'Rcaron': 722, |
|
'Gcommaaccent': 778, |
|
'ucircumflex': 556, |
|
'acircumflex': 556, |
|
'Amacron': 667, |
|
'rcaron': 333, |
|
'ccedilla': 500, |
|
'Zdotaccent': 611, |
|
'Thorn': 667, |
|
'Omacron': 778, |
|
'Racute': 722, |
|
'Sacute': 667, |
|
'dcaron': 643, |
|
'Umacron': 722, |
|
'uring': 556, |
|
'threesuperior': 333, |
|
'Ograve': 778, |
|
'Agrave': 667, |
|
'Abreve': 667, |
|
'multiply': 584, |
|
'uacute': 556, |
|
'Tcaron': 611, |
|
'partialdiff': 476, |
|
'ydieresis': 500, |
|
'Nacute': 722, |
|
'icircumflex': 278, |
|
'Ecircumflex': 667, |
|
'adieresis': 556, |
|
'edieresis': 556, |
|
'cacute': 500, |
|
'nacute': 556, |
|
'umacron': 556, |
|
'Ncaron': 722, |
|
'Iacute': 278, |
|
'plusminus': 584, |
|
'brokenbar': 260, |
|
'registered': 737, |
|
'Gbreve': 778, |
|
'Idotaccent': 278, |
|
'summation': 600, |
|
'Egrave': 667, |
|
'racute': 333, |
|
'omacron': 556, |
|
'Zacute': 611, |
|
'Zcaron': 611, |
|
'greaterequal': 549, |
|
'Eth': 722, |
|
'Ccedilla': 722, |
|
'lcommaaccent': 222, |
|
'tcaron': 317, |
|
'eogonek': 556, |
|
'Uogonek': 722, |
|
'Aacute': 667, |
|
'Adieresis': 667, |
|
'egrave': 556, |
|
'zacute': 500, |
|
'iogonek': 222, |
|
'Oacute': 778, |
|
'oacute': 556, |
|
'amacron': 556, |
|
'sacute': 500, |
|
'idieresis': 278, |
|
'Ocircumflex': 778, |
|
'Ugrave': 722, |
|
'Delta': 612, |
|
'thorn': 556, |
|
'twosuperior': 333, |
|
'Odieresis': 778, |
|
'mu': 556, |
|
'igrave': 278, |
|
'ohungarumlaut': 556, |
|
'Eogonek': 667, |
|
'dcroat': 556, |
|
'threequarters': 834, |
|
'Scedilla': 667, |
|
'lcaron': 299, |
|
'Kcommaaccent': 667, |
|
'Lacute': 556, |
|
'trademark': 1000, |
|
'edotaccent': 556, |
|
'Igrave': 278, |
|
'Imacron': 278, |
|
'Lcaron': 556, |
|
'onehalf': 834, |
|
'lessequal': 549, |
|
'ocircumflex': 556, |
|
'ntilde': 556, |
|
'Uhungarumlaut': 722, |
|
'Eacute': 667, |
|
'emacron': 556, |
|
'gbreve': 556, |
|
'onequarter': 834, |
|
'Scaron': 667, |
|
'Scommaaccent': 667, |
|
'Ohungarumlaut': 778, |
|
'degree': 400, |
|
'ograve': 556, |
|
'Ccaron': 722, |
|
'ugrave': 556, |
|
'radical': 453, |
|
'Dcaron': 722, |
|
'rcommaaccent': 333, |
|
'Ntilde': 722, |
|
'otilde': 556, |
|
'Rcommaaccent': 722, |
|
'Lcommaaccent': 556, |
|
'Atilde': 667, |
|
'Aogonek': 667, |
|
'Aring': 667, |
|
'Otilde': 778, |
|
'zdotaccent': 500, |
|
'Ecaron': 667, |
|
'Iogonek': 278, |
|
'kcommaaccent': 500, |
|
'minus': 584, |
|
'Icircumflex': 278, |
|
'ncaron': 556, |
|
'tcommaaccent': 278, |
|
'logicalnot': 584, |
|
'odieresis': 556, |
|
'udieresis': 556, |
|
'notequal': 549, |
|
'gcommaaccent': 556, |
|
'eth': 556, |
|
'zcaron': 500, |
|
'ncommaaccent': 556, |
|
'onesuperior': 333, |
|
'imacron': 278, |
|
'Euro': 556 |
|
}, |
|
'Helvetica-Bold': { |
|
'space': 278, |
|
'exclam': 333, |
|
'quotedbl': 474, |
|
'numbersign': 556, |
|
'dollar': 556, |
|
'percent': 889, |
|
'ampersand': 722, |
|
'quoteright': 278, |
|
'parenleft': 333, |
|
'parenright': 333, |
|
'asterisk': 389, |
|
'plus': 584, |
|
'comma': 278, |
|
'hyphen': 333, |
|
'period': 278, |
|
'slash': 278, |
|
'zero': 556, |
|
'one': 556, |
|
'two': 556, |
|
'three': 556, |
|
'four': 556, |
|
'five': 556, |
|
'six': 556, |
|
'seven': 556, |
|
'eight': 556, |
|
'nine': 556, |
|
'colon': 333, |
|
'semicolon': 333, |
|
'less': 584, |
|
'equal': 584, |
|
'greater': 584, |
|
'question': 611, |
|
'at': 975, |
|
'A': 722, |
|
'B': 722, |
|
'C': 722, |
|
'D': 722, |
|
'E': 667, |
|
'F': 611, |
|
'G': 778, |
|
'H': 722, |
|
'I': 278, |
|
'J': 556, |
|
'K': 722, |
|
'L': 611, |
|
'M': 833, |
|
'N': 722, |
|
'O': 778, |
|
'P': 667, |
|
'Q': 778, |
|
'R': 722, |
|
'S': 667, |
|
'T': 611, |
|
'U': 722, |
|
'V': 667, |
|
'W': 944, |
|
'X': 667, |
|
'Y': 667, |
|
'Z': 611, |
|
'bracketleft': 333, |
|
'backslash': 278, |
|
'bracketright': 333, |
|
'asciicircum': 584, |
|
'underscore': 556, |
|
'quoteleft': 278, |
|
'a': 556, |
|
'b': 611, |
|
'c': 556, |
|
'd': 611, |
|
'e': 556, |
|
'f': 333, |
|
'g': 611, |
|
'h': 611, |
|
'i': 278, |
|
'j': 278, |
|
'k': 556, |
|
'l': 278, |
|
'm': 889, |
|
'n': 611, |
|
'o': 611, |
|
'p': 611, |
|
'q': 611, |
|
'r': 389, |
|
's': 556, |
|
't': 333, |
|
'u': 611, |
|
'v': 556, |
|
'w': 778, |
|
'x': 556, |
|
'y': 556, |
|
'z': 500, |
|
'braceleft': 389, |
|
'bar': 280, |
|
'braceright': 389, |
|
'asciitilde': 584, |
|
'exclamdown': 333, |
|
'cent': 556, |
|
'sterling': 556, |
|
'fraction': 167, |
|
'yen': 556, |
|
'florin': 556, |
|
'section': 556, |
|
'currency': 556, |
|
'quotesingle': 238, |
|
'quotedblleft': 500, |
|
'guillemotleft': 556, |
|
'guilsinglleft': 333, |
|
'guilsinglright': 333, |
|
'fi': 611, |
|
'fl': 611, |
|
'endash': 556, |
|
'dagger': 556, |
|
'daggerdbl': 556, |
|
'periodcentered': 278, |
|
'paragraph': 556, |
|
'bullet': 350, |
|
'quotesinglbase': 278, |
|
'quotedblbase': 500, |
|
'quotedblright': 500, |
|
'guillemotright': 556, |
|
'ellipsis': 1000, |
|
'perthousand': 1000, |
|
'questiondown': 611, |
|
'grave': 333, |
|
'acute': 333, |
|
'circumflex': 333, |
|
'tilde': 333, |
|
'macron': 333, |
|
'breve': 333, |
|
'dotaccent': 333, |
|
'dieresis': 333, |
|
'ring': 333, |
|
'cedilla': 333, |
|
'hungarumlaut': 333, |
|
'ogonek': 333, |
|
'caron': 333, |
|
'emdash': 1000, |
|
'AE': 1000, |
|
'ordfeminine': 370, |
|
'Lslash': 611, |
|
'Oslash': 778, |
|
'OE': 1000, |
|
'ordmasculine': 365, |
|
'ae': 889, |
|
'dotlessi': 278, |
|
'lslash': 278, |
|
'oslash': 611, |
|
'oe': 944, |
|
'germandbls': 611, |
|
'Idieresis': 278, |
|
'eacute': 556, |
|
'abreve': 556, |
|
'uhungarumlaut': 611, |
|
'ecaron': 556, |
|
'Ydieresis': 667, |
|
'divide': 584, |
|
'Yacute': 667, |
|
'Acircumflex': 722, |
|
'aacute': 556, |
|
'Ucircumflex': 722, |
|
'yacute': 556, |
|
'scommaaccent': 556, |
|
'ecircumflex': 556, |
|
'Uring': 722, |
|
'Udieresis': 722, |
|
'aogonek': 556, |
|
'Uacute': 722, |
|
'uogonek': 611, |
|
'Edieresis': 667, |
|
'Dcroat': 722, |
|
'commaaccent': 250, |
|
'copyright': 737, |
|
'Emacron': 667, |
|
'ccaron': 556, |
|
'aring': 556, |
|
'Ncommaaccent': 722, |
|
'lacute': 278, |
|
'agrave': 556, |
|
'Tcommaaccent': 611, |
|
'Cacute': 722, |
|
'atilde': 556, |
|
'Edotaccent': 667, |
|
'scaron': 556, |
|
'scedilla': 556, |
|
'iacute': 278, |
|
'lozenge': 494, |
|
'Rcaron': 722, |
|
'Gcommaaccent': 778, |
|
'ucircumflex': 611, |
|
'acircumflex': 556, |
|
'Amacron': 722, |
|
'rcaron': 389, |
|
'ccedilla': 556, |
|
'Zdotaccent': 611, |
|
'Thorn': 667, |
|
'Omacron': 778, |
|
'Racute': 722, |
|
'Sacute': 667, |
|
'dcaron': 743, |
|
'Umacron': 722, |
|
'uring': 611, |
|
'threesuperior': 333, |
|
'Ograve': 778, |
|
'Agrave': 722, |
|
'Abreve': 722, |
|
'multiply': 584, |
|
'uacute': 611, |
|
'Tcaron': 611, |
|
'partialdiff': 494, |
|
'ydieresis': 556, |
|
'Nacute': 722, |
|
'icircumflex': 278, |
|
'Ecircumflex': 667, |
|
'adieresis': 556, |
|
'edieresis': 556, |
|
'cacute': 556, |
|
'nacute': 611, |
|
'umacron': 611, |
|
'Ncaron': 722, |
|
'Iacute': 278, |
|
'plusminus': 584, |
|
'brokenbar': 280, |
|
'registered': 737, |
|
'Gbreve': 778, |
|
'Idotaccent': 278, |
|
'summation': 600, |
|
'Egrave': 667, |
|
'racute': 389, |
|
'omacron': 611, |
|
'Zacute': 611, |
|
'Zcaron': 611, |
|
'greaterequal': 549, |
|
'Eth': 722, |
|
'Ccedilla': 722, |
|
'lcommaaccent': 278, |
|
'tcaron': 389, |
|
'eogonek': 556, |
|
'Uogonek': 722, |
|
'Aacute': 722, |
|
'Adieresis': 722, |
|
'egrave': 556, |
|
'zacute': 500, |
|
'iogonek': 278, |
|
'Oacute': 778, |
|
'oacute': 611, |
|
'amacron': 556, |
|
'sacute': 556, |
|
'idieresis': 278, |
|
'Ocircumflex': 778, |
|
'Ugrave': 722, |
|
'Delta': 612, |
|
'thorn': 611, |
|
'twosuperior': 333, |
|
'Odieresis': 778, |
|
'mu': 611, |
|
'igrave': 278, |
|
'ohungarumlaut': 611, |
|
'Eogonek': 667, |
|
'dcroat': 611, |
|
'threequarters': 834, |
|
'Scedilla': 667, |
|
'lcaron': 400, |
|
'Kcommaaccent': 722, |
|
'Lacute': 611, |
|
'trademark': 1000, |
|
'edotaccent': 556, |
|
'Igrave': 278, |
|
'Imacron': 278, |
|
'Lcaron': 611, |
|
'onehalf': 834, |
|
'lessequal': 549, |
|
'ocircumflex': 611, |
|
'ntilde': 611, |
|
'Uhungarumlaut': 722, |
|
'Eacute': 667, |
|
'emacron': 556, |
|
'gbreve': 611, |
|
'onequarter': 834, |
|
'Scaron': 667, |
|
'Scommaaccent': 667, |
|
'Ohungarumlaut': 778, |
|
'degree': 400, |
|
'ograve': 611, |
|
'Ccaron': 722, |
|
'ugrave': 611, |
|
'radical': 549, |
|
'Dcaron': 722, |
|
'rcommaaccent': 389, |
|
'Ntilde': 722, |
|
'otilde': 611, |
|
'Rcommaaccent': 722, |
|
'Lcommaaccent': 611, |
|
'Atilde': 722, |
|
'Aogonek': 722, |
|
'Aring': 722, |
|
'Otilde': 778, |
|
'zdotaccent': 500, |
|
'Ecaron': 667, |
|
'Iogonek': 278, |
|
'kcommaaccent': 556, |
|
'minus': 584, |
|
'Icircumflex': 278, |
|
'ncaron': 611, |
|
'tcommaaccent': 333, |
|
'logicalnot': 584, |
|
'odieresis': 611, |
|
'udieresis': 611, |
|
'notequal': 549, |
|
'gcommaaccent': 611, |
|
'eth': 611, |
|
'zcaron': 500, |
|
'ncommaaccent': 611, |
|
'onesuperior': 333, |
|
'imacron': 278, |
|
'Euro': 556 |
|
}, |
|
'Helvetica-BoldOblique': { |
|
'space': 278, |
|
'exclam': 333, |
|
'quotedbl': 474, |
|
'numbersign': 556, |
|
'dollar': 556, |
|
'percent': 889, |
|
'ampersand': 722, |
|
'quoteright': 278, |
|
'parenleft': 333, |
|
'parenright': 333, |
|
'asterisk': 389, |
|
'plus': 584, |
|
'comma': 278, |
|
'hyphen': 333, |
|
'period': 278, |
|
'slash': 278, |
|
'zero': 556, |
|
'one': 556, |
|
'two': 556, |
|
'three': 556, |
|
'four': 556, |
|
'five': 556, |
|
'six': 556, |
|
'seven': 556, |
|
'eight': 556, |
|
'nine': 556, |
|
'colon': 333, |
|
'semicolon': 333, |
|
'less': 584, |
|
'equal': 584, |
|
'greater': 584, |
|
'question': 611, |
|
'at': 975, |
|
'A': 722, |
|
'B': 722, |
|
'C': 722, |
|
'D': 722, |
|
'E': 667, |
|
'F': 611, |
|
'G': 778, |
|
'H': 722, |
|
'I': 278, |
|
'J': 556, |
|
'K': 722, |
|
'L': 611, |
|
'M': 833, |
|
'N': 722, |
|
'O': 778, |
|
'P': 667, |
|
'Q': 778, |
|
'R': 722, |
|
'S': 667, |
|
'T': 611, |
|
'U': 722, |
|
'V': 667, |
|
'W': 944, |
|
'X': 667, |
|
'Y': 667, |
|
'Z': 611, |
|
'bracketleft': 333, |
|
'backslash': 278, |
|
'bracketright': 333, |
|
'asciicircum': 584, |
|
'underscore': 556, |
|
'quoteleft': 278, |
|
'a': 556, |
|
'b': 611, |
|
'c': 556, |
|
'd': 611, |
|
'e': 556, |
|
'f': 333, |
|
'g': 611, |
|
'h': 611, |
|
'i': 278, |
|
'j': 278, |
|
'k': 556, |
|
'l': 278, |
|
'm': 889, |
|
'n': 611, |
|
'o': 611, |
|
'p': 611, |
|
'q': 611, |
|
'r': 389, |
|
's': 556, |
|
't': 333, |
|
'u': 611, |
|
'v': 556, |
|
'w': 778, |
|
'x': 556, |
|
'y': 556, |
|
'z': 500, |
|
'braceleft': 389, |
|
'bar': 280, |
|
'braceright': 389, |
|
'asciitilde': 584, |
|
'exclamdown': 333, |
|
'cent': 556, |
|
'sterling': 556, |
|
'fraction': 167, |
|
'yen': 556, |
|
'florin': 556, |
|
'section': 556, |
|
'currency': 556, |
|
'quotesingle': 238, |
|
'quotedblleft': 500, |
|
'guillemotleft': 556, |
|
'guilsinglleft': 333, |
|
'guilsinglright': 333, |
|
'fi': 611, |
|
'fl': 611, |
|
'endash': 556, |
|
'dagger': 556, |
|
'daggerdbl': 556, |
|
'periodcentered': 278, |
|
'paragraph': 556, |
|
'bullet': 350, |
|
'quotesinglbase': 278, |
|
'quotedblbase': 500, |
|
'quotedblright': 500, |
|
'guillemotright': 556, |
|
'ellipsis': 1000, |
|
'perthousand': 1000, |
|
'questiondown': 611, |
|
'grave': 333, |
|
'acute': 333, |
|
'circumflex': 333, |
|
'tilde': 333, |
|
'macron': 333, |
|
'breve': 333, |
|
'dotaccent': 333, |
|
'dieresis': 333, |
|
'ring': 333, |
|
'cedilla': 333, |
|
'hungarumlaut': 333, |
|
'ogonek': 333, |
|
'caron': 333, |
|
'emdash': 1000, |
|
'AE': 1000, |
|
'ordfeminine': 370, |
|
'Lslash': 611, |
|
'Oslash': 778, |
|
'OE': 1000, |
|
'ordmasculine': 365, |
|
'ae': 889, |
|
'dotlessi': 278, |
|
'lslash': 278, |
|
'oslash': 611, |
|
'oe': 944, |
|
'germandbls': 611, |
|
'Idieresis': 278, |
|
'eacute': 556, |
|
'abreve': 556, |
|
'uhungarumlaut': 611, |
|
'ecaron': 556, |
|
'Ydieresis': 667, |
|
'divide': 584, |
|
'Yacute': 667, |
|
'Acircumflex': 722, |
|
'aacute': 556, |
|
'Ucircumflex': 722, |
|
'yacute': 556, |
|
'scommaaccent': 556, |
|
'ecircumflex': 556, |
|
'Uring': 722, |
|
'Udieresis': 722, |
|
'aogonek': 556, |
|
'Uacute': 722, |
|
'uogonek': 611, |
|
'Edieresis': 667, |
|
'Dcroat': 722, |
|
'commaaccent': 250, |
|
'copyright': 737, |
|
'Emacron': 667, |
|
'ccaron': 556, |
|
'aring': 556, |
|
'Ncommaaccent': 722, |
|
'lacute': 278, |
|
'agrave': 556, |
|
'Tcommaaccent': 611, |
|
'Cacute': 722, |
|
'atilde': 556, |
|
'Edotaccent': 667, |
|
'scaron': 556, |
|
'scedilla': 556, |
|
'iacute': 278, |
|
'lozenge': 494, |
|
'Rcaron': 722, |
|
'Gcommaaccent': 778, |
|
'ucircumflex': 611, |
|
'acircumflex': 556, |
|
'Amacron': 722, |
|
'rcaron': 389, |
|
'ccedilla': 556, |
|
'Zdotaccent': 611, |
|
'Thorn': 667, |
|
'Omacron': 778, |
|
'Racute': 722, |
|
'Sacute': 667, |
|
'dcaron': 743, |
|
'Umacron': 722, |
|
'uring': 611, |
|
'threesuperior': 333, |
|
'Ograve': 778, |
|
'Agrave': 722, |
|
'Abreve': 722, |
|
'multiply': 584, |
|
'uacute': 611, |
|
'Tcaron': 611, |
|
'partialdiff': 494, |
|
'ydieresis': 556, |
|
'Nacute': 722, |
|
'icircumflex': 278, |
|
'Ecircumflex': 667, |
|
'adieresis': 556, |
|
'edieresis': 556, |
|
'cacute': 556, |
|
'nacute': 611, |
|
'umacron': 611, |
|
'Ncaron': 722, |
|
'Iacute': 278, |
|
'plusminus': 584, |
|
'brokenbar': 280, |
|
'registered': 737, |
|
'Gbreve': 778, |
|
'Idotaccent': 278, |
|
'summation': 600, |
|
'Egrave': 667, |
|
'racute': 389, |
|
'omacron': 611, |
|
'Zacute': 611, |
|
'Zcaron': 611, |
|
'greaterequal': 549, |
|
'Eth': 722, |
|
'Ccedilla': 722, |
|
'lcommaaccent': 278, |
|
'tcaron': 389, |
|
'eogonek': 556, |
|
'Uogonek': 722, |
|
'Aacute': 722, |
|
'Adieresis': 722, |
|
'egrave': 556, |
|
'zacute': 500, |
|
'iogonek': 278, |
|
'Oacute': 778, |
|
'oacute': 611, |
|
'amacron': 556, |
|
'sacute': 556, |
|
'idieresis': 278, |
|
'Ocircumflex': 778, |
|
'Ugrave': 722, |
|
'Delta': 612, |
|
'thorn': 611, |
|
'twosuperior': 333, |
|
'Odieresis': 778, |
|
'mu': 611, |
|
'igrave': 278, |
|
'ohungarumlaut': 611, |
|
'Eogonek': 667, |
|
'dcroat': 611, |
|
'threequarters': 834, |
|
'Scedilla': 667, |
|
'lcaron': 400, |
|
'Kcommaaccent': 722, |
|
'Lacute': 611, |
|
'trademark': 1000, |
|
'edotaccent': 556, |
|
'Igrave': 278, |
|
'Imacron': 278, |
|
'Lcaron': 611, |
|
'onehalf': 834, |
|
'lessequal': 549, |
|
'ocircumflex': 611, |
|
'ntilde': 611, |
|
'Uhungarumlaut': 722, |
|
'Eacute': 667, |
|
'emacron': 556, |
|
'gbreve': 611, |
|
'onequarter': 834, |
|
'Scaron': 667, |
|
'Scommaaccent': 667, |
|
'Ohungarumlaut': 778, |
|
'degree': 400, |
|
'ograve': 611, |
|
'Ccaron': 722, |
|
'ugrave': 611, |
|
'radical': 549, |
|
'Dcaron': 722, |
|
'rcommaaccent': 389, |
|
'Ntilde': 722, |
|
'otilde': 611, |
|
'Rcommaaccent': 722, |
|
'Lcommaaccent': 611, |
|
'Atilde': 722, |
|
'Aogonek': 722, |
|
'Aring': 722, |
|
'Otilde': 778, |
|
'zdotaccent': 500, |
|
'Ecaron': 667, |
|
'Iogonek': 278, |
|
'kcommaaccent': 556, |
|
'minus': 584, |
|
'Icircumflex': 278, |
|
'ncaron': 611, |
|
'tcommaaccent': 333, |
|
'logicalnot': 584, |
|
'odieresis': 611, |
|
'udieresis': 611, |
|
'notequal': 549, |
|
'gcommaaccent': 611, |
|
'eth': 611, |
|
'zcaron': 500, |
|
'ncommaaccent': 611, |
|
'onesuperior': 333, |
|
'imacron': 278, |
|
'Euro': 556 |
|
}, |
|
'Helvetica-Oblique' : { |
|
'space': 278, |
|
'exclam': 278, |
|
'quotedbl': 355, |
|
'numbersign': 556, |
|
'dollar': 556, |
|
'percent': 889, |
|
'ampersand': 667, |
|
'quoteright': 222, |
|
'parenleft': 333, |
|
'parenright': 333, |
|
'asterisk': 389, |
|
'plus': 584, |
|
'comma': 278, |
|
'hyphen': 333, |
|
'period': 278, |
|
'slash': 278, |
|
'zero': 556, |
|
'one': 556, |
|
'two': 556, |
|
'three': 556, |
|
'four': 556, |
|
'five': 556, |
|
'six': 556, |
|
'seven': 556, |
|
'eight': 556, |
|
'nine': 556, |
|
'colon': 278, |
|
'semicolon': 278, |
|
'less': 584, |
|
'equal': 584, |
|
'greater': 584, |
|
'question': 556, |
|
'at': 1015, |
|
'A': 667, |
|
'B': 667, |
|
'C': 722, |
|
'D': 722, |
|
'E': 667, |
|
'F': 611, |
|
'G': 778, |
|
'H': 722, |
|
'I': 278, |
|
'J': 500, |
|
'K': 667, |
|
'L': 556, |
|
'M': 833, |
|
'N': 722, |
|
'O': 778, |
|
'P': 667, |
|
'Q': 778, |
|
'R': 722, |
|
'S': 667, |
|
'T': 611, |
|
'U': 722, |
|
'V': 667, |
|
'W': 944, |
|
'X': 667, |
|
'Y': 667, |
|
'Z': 611, |
|
'bracketleft': 278, |
|
'backslash': 278, |
|
'bracketright': 278, |
|
'asciicircum': 469, |
|
'underscore': 556, |
|
'quoteleft': 222, |
|
'a': 556, |
|
'b': 556, |
|
'c': 500, |
|
'd': 556, |
|
'e': 556, |
|
'f': 278, |
|
'g': 556, |
|
'h': 556, |
|
'i': 222, |
|
'j': 222, |
|
'k': 500, |
|
'l': 222, |
|
'm': 833, |
|
'n': 556, |
|
'o': 556, |
|
'p': 556, |
|
'q': 556, |
|
'r': 333, |
|
's': 500, |
|
't': 278, |
|
'u': 556, |
|
'v': 500, |
|
'w': 722, |
|
'x': 500, |
|
'y': 500, |
|
'z': 500, |
|
'braceleft': 334, |
|
'bar': 260, |
|
'braceright': 334, |
|
'asciitilde': 584, |
|
'exclamdown': 333, |
|
'cent': 556, |
|
'sterling': 556, |
|
'fraction': 167, |
|
'yen': 556, |
|
'florin': 556, |
|
'section': 556, |
|
'currency': 556, |
|
'quotesingle': 191, |
|
'quotedblleft': 333, |
|
'guillemotleft': 556, |
|
'guilsinglleft': 333, |
|
'guilsinglright': 333, |
|
'fi': 500, |
|
'fl': 500, |
|
'endash': 556, |
|
'dagger': 556, |
|
'daggerdbl': 556, |
|
'periodcentered': 278, |
|
'paragraph': 537, |
|
'bullet': 350, |
|
'quotesinglbase': 222, |
|
'quotedblbase': 333, |
|
'quotedblright': 333, |
|
'guillemotright': 556, |
|
'ellipsis': 1000, |
|
'perthousand': 1000, |
|
'questiondown': 611, |
|
'grave': 333, |
|
'acute': 333, |
|
'circumflex': 333, |
|
'tilde': 333, |
|
'macron': 333, |
|
'breve': 333, |
|
'dotaccent': 333, |
|
'dieresis': 333, |
|
'ring': 333, |
|
'cedilla': 333, |
|
'hungarumlaut': 333, |
|
'ogonek': 333, |
|
'caron': 333, |
|
'emdash': 1000, |
|
'AE': 1000, |
|
'ordfeminine': 370, |
|
'Lslash': 556, |
|
'Oslash': 778, |
|
'OE': 1000, |
|
'ordmasculine': 365, |
|
'ae': 889, |
|
'dotlessi': 278, |
|
'lslash': 222, |
|
'oslash': 611, |
|
'oe': 944, |
|
'germandbls': 611, |
|
'Idieresis': 278, |
|
'eacute': 556, |
|
'abreve': 556, |
|
'uhungarumlaut': 556, |
|
'ecaron': 556, |
|
'Ydieresis': 667, |
|
'divide': 584, |
|
'Yacute': 667, |
|
'Acircumflex': 667, |
|
'aacute': 556, |
|
'Ucircumflex': 722, |
|
'yacute': 500, |
|
'scommaaccent': 500, |
|
'ecircumflex': 556, |
|
'Uring': 722, |
|
'Udieresis': 722, |
|
'aogonek': 556, |
|
'Uacute': 722, |
|
'uogonek': 556, |
|
'Edieresis': 667, |
|
'Dcroat': 722, |
|
'commaaccent': 250, |
|
'copyright': 737, |
|
'Emacron': 667, |
|
'ccaron': 500, |
|
'aring': 556, |
|
'Ncommaaccent': 722, |
|
'lacute': 222, |
|
'agrave': 556, |
|
'Tcommaaccent': 611, |
|
'Cacute': 722, |
|
'atilde': 556, |
|
'Edotaccent': 667, |
|
'scaron': 500, |
|
'scedilla': 500, |
|
'iacute': 278, |
|
'lozenge': 471, |
|
'Rcaron': 722, |
|
'Gcommaaccent': 778, |
|
'ucircumflex': 556, |
|
'acircumflex': 556, |
|
'Amacron': 667, |
|
'rcaron': 333, |
|
'ccedilla': 500, |
|
'Zdotaccent': 611, |
|
'Thorn': 667, |
|
'Omacron': 778, |
|
'Racute': 722, |
|
'Sacute': 667, |
|
'dcaron': 643, |
|
'Umacron': 722, |
|
'uring': 556, |
|
'threesuperior': 333, |
|
'Ograve': 778, |
|
'Agrave': 667, |
|
'Abreve': 667, |
|
'multiply': 584, |
|
'uacute': 556, |
|
'Tcaron': 611, |
|
'partialdiff': 476, |
|
'ydieresis': 500, |
|
'Nacute': 722, |
|
'icircumflex': 278, |
|
'Ecircumflex': 667, |
|
'adieresis': 556, |
|
'edieresis': 556, |
|
'cacute': 500, |
|
'nacute': 556, |
|
'umacron': 556, |
|
'Ncaron': 722, |
|
'Iacute': 278, |
|
'plusminus': 584, |
|
'brokenbar': 260, |
|
'registered': 737, |
|
'Gbreve': 778, |
|
'Idotaccent': 278, |
|
'summation': 600, |
|
'Egrave': 667, |
|
'racute': 333, |
|
'omacron': 556, |
|
'Zacute': 611, |
|
'Zcaron': 611, |
|
'greaterequal': 549, |
|
'Eth': 722, |
|
'Ccedilla': 722, |
|
'lcommaaccent': 222, |
|
'tcaron': 317, |
|
'eogonek': 556, |
|
'Uogonek': 722, |
|
'Aacute': 667, |
|
'Adieresis': 667, |
|
'egrave': 556, |
|
'zacute': 500, |
|
'iogonek': 222, |
|
'Oacute': 778, |
|
'oacute': 556, |
|
'amacron': 556, |
|
'sacute': 500, |
|
'idieresis': 278, |
|
'Ocircumflex': 778, |
|
'Ugrave': 722, |
|
'Delta': 612, |
|
'thorn': 556, |
|
'twosuperior': 333, |
|
'Odieresis': 778, |
|
'mu': 556, |
|
'igrave': 278, |
|
'ohungarumlaut': 556, |
|
'Eogonek': 667, |
|
'dcroat': 556, |
|
'threequarters': 834, |
|
'Scedilla': 667, |
|
'lcaron': 299, |
|
'Kcommaaccent': 667, |
|
'Lacute': 556, |
|
'trademark': 1000, |
|
'edotaccent': 556, |
|
'Igrave': 278, |
|
'Imacron': 278, |
|
'Lcaron': 556, |
|
'onehalf': 834, |
|
'lessequal': 549, |
|
'ocircumflex': 556, |
|
'ntilde': 556, |
|
'Uhungarumlaut': 722, |
|
'Eacute': 667, |
|
'emacron': 556, |
|
'gbreve': 556, |
|
'onequarter': 834, |
|
'Scaron': 667, |
|
'Scommaaccent': 667, |
|
'Ohungarumlaut': 778, |
|
'degree': 400, |
|
'ograve': 556, |
|
'Ccaron': 722, |
|
'ugrave': 556, |
|
'radical': 453, |
|
'Dcaron': 722, |
|
'rcommaaccent': 333, |
|
'Ntilde': 722, |
|
'otilde': 556, |
|
'Rcommaaccent': 722, |
|
'Lcommaaccent': 556, |
|
'Atilde': 667, |
|
'Aogonek': 667, |
|
'Aring': 667, |
|
'Otilde': 778, |
|
'zdotaccent': 500, |
|
'Ecaron': 667, |
|
'Iogonek': 278, |
|
'kcommaaccent': 500, |
|
'minus': 584, |
|
'Icircumflex': 278, |
|
'ncaron': 556, |
|
'tcommaaccent': 278, |
|
'logicalnot': 584, |
|
'odieresis': 556, |
|
'udieresis': 556, |
|
'notequal': 549, |
|
'gcommaaccent': 556, |
|
'eth': 556, |
|
'zcaron': 500, |
|
'ncommaaccent': 556, |
|
'onesuperior': 333, |
|
'imacron': 278, |
|
'Euro': 556 |
|
}, |
|
'Symbol': { |
|
'space': 250, |
|
'exclam': 333, |
|
'universal': 713, |
|
'numbersign': 500, |
|
'existential': 549, |
|
'percent': 833, |
|
'ampersand': 778, |
|
'suchthat': 439, |
|
'parenleft': 333, |
|
'parenright': 333, |
|
'asteriskmath': 500, |
|
'plus': 549, |
|
'comma': 250, |
|
'minus': 549, |
|
'period': 250, |
|
'slash': 278, |
|
'zero': 500, |
|
'one': 500, |
|
'two': 500, |
|
'three': 500, |
|
'four': 500, |
|
'five': 500, |
|
'six': 500, |
|
'seven': 500, |
|
'eight': 500, |
|
'nine': 500, |
|
'colon': 278, |
|
'semicolon': 278, |
|
'less': 549, |
|
'equal': 549, |
|
'greater': 549, |
|
'question': 444, |
|
'congruent': 549, |
|
'Alpha': 722, |
|
'Beta': 667, |
|
'Chi': 722, |
|
'Delta': 612, |
|
'Epsilon': 611, |
|
'Phi': 763, |
|
'Gamma': 603, |
|
'Eta': 722, |
|
'Iota': 333, |
|
'theta1': 631, |
|
'Kappa': 722, |
|
'Lambda': 686, |
|
'Mu': 889, |
|
'Nu': 722, |
|
'Omicron': 722, |
|
'Pi': 768, |
|
'Theta': 741, |
|
'Rho': 556, |
|
'Sigma': 592, |
|
'Tau': 611, |
|
'Upsilon': 690, |
|
'sigma1': 439, |
|
'Omega': 768, |
|
'Xi': 645, |
|
'Psi': 795, |
|
'Zeta': 611, |
|
'bracketleft': 333, |
|
'therefore': 863, |
|
'bracketright': 333, |
|
'perpendicular': 658, |
|
'underscore': 500, |
|
'radicalex': 500, |
|
'alpha': 631, |
|
'beta': 549, |
|
'chi': 549, |
|
'delta': 494, |
|
'epsilon': 439, |
|
'phi': 521, |
|
'gamma': 411, |
|
'eta': 603, |
|
'iota': 329, |
|
'phi1': 603, |
|
'kappa': 549, |
|
'lambda': 549, |
|
'mu': 576, |
|
'nu': 521, |
|
'omicron': 549, |
|
'pi': 549, |
|
'theta': 521, |
|
'rho': 549, |
|
'sigma': 603, |
|
'tau': 439, |
|
'upsilon': 576, |
|
'omega1': 713, |
|
'omega': 686, |
|
'xi': 493, |
|
'psi': 686, |
|
'zeta': 494, |
|
'braceleft': 480, |
|
'bar': 200, |
|
'braceright': 480, |
|
'similar': 549, |
|
'Euro': 750, |
|
'Upsilon1': 620, |
|
'minute': 247, |
|
'lessequal': 549, |
|
'fraction': 167, |
|
'infinity': 713, |
|
'florin': 500, |
|
'club': 753, |
|
'diamond': 753, |
|
'heart': 753, |
|
'spade': 753, |
|
'arrowboth': 1042, |
|
'arrowleft': 987, |
|
'arrowup': 603, |
|
'arrowright': 987, |
|
'arrowdown': 603, |
|
'degree': 400, |
|
'plusminus': 549, |
|
'second': 411, |
|
'greaterequal': 549, |
|
'multiply': 549, |
|
'proportional': 713, |
|
'partialdiff': 494, |
|
'bullet': 460, |
|
'divide': 549, |
|
'notequal': 549, |
|
'equivalence': 549, |
|
'approxequal': 549, |
|
'ellipsis': 1000, |
|
'arrowvertex': 603, |
|
'arrowhorizex': 1000, |
|
'carriagereturn': 658, |
|
'aleph': 823, |
|
'Ifraktur': 686, |
|
'Rfraktur': 795, |
|
'weierstrass': 987, |
|
'circlemultiply': 768, |
|
'circleplus': 768, |
|
'emptyset': 823, |
|
'intersection': 768, |
|
'union': 768, |
|
'propersuperset': 713, |
|
'reflexsuperset': 713, |
|
'notsubset': 713, |
|
'propersubset': 713, |
|
'reflexsubset': 713, |
|
'element': 713, |
|
'notelement': 713, |
|
'angle': 768, |
|
'gradient': 713, |
|
'registerserif': 790, |
|
'copyrightserif': 790, |
|
'trademarkserif': 890, |
|
'product': 823, |
|
'radical': 549, |
|
'dotmath': 250, |
|
'logicalnot': 713, |
|
'logicaland': 603, |
|
'logicalor': 603, |
|
'arrowdblboth': 1042, |
|
'arrowdblleft': 987, |
|
'arrowdblup': 603, |
|
'arrowdblright': 987, |
|
'arrowdbldown': 603, |
|
'lozenge': 494, |
|
'angleleft': 329, |
|
'registersans': 790, |
|
'copyrightsans': 790, |
|
'trademarksans': 786, |
|
'summation': 713, |
|
'parenlefttp': 384, |
|
'parenleftex': 384, |
|
'parenleftbt': 384, |
|
'bracketlefttp': 384, |
|
'bracketleftex': 384, |
|
'bracketleftbt': 384, |
|
'bracelefttp': 494, |
|
'braceleftmid': 494, |
|
'braceleftbt': 494, |
|
'braceex': 494, |
|
'angleright': 329, |
|
'integral': 274, |
|
'integraltp': 686, |
|
'integralex': 686, |
|
'integralbt': 686, |
|
'parenrighttp': 384, |
|
'parenrightex': 384, |
|
'parenrightbt': 384, |
|
'bracketrighttp': 384, |
|
'bracketrightex': 384, |
|
'bracketrightbt': 384, |
|
'bracerighttp': 494, |
|
'bracerightmid': 494, |
|
'bracerightbt': 494, |
|
'apple': 790 |
|
}, |
|
'Times-Roman': { |
|
'space': 250, |
|
'exclam': 333, |
|
'quotedbl': 408, |
|
'numbersign': 500, |
|
'dollar': 500, |
|
'percent': 833, |
|
'ampersand': 778, |
|
'quoteright': 333, |
|
'parenleft': 333, |
|
'parenright': 333, |
|
'asterisk': 500, |
|
'plus': 564, |
|
'comma': 250, |
|
'hyphen': 333, |
|
'period': 250, |
|
'slash': 278, |
|
'zero': 500, |
|
'one': 500, |
|
'two': 500, |
|
'three': 500, |
|
'four': 500, |
|
'five': 500, |
|
'six': 500, |
|
'seven': 500, |
|
'eight': 500, |
|
'nine': 500, |
|
'colon': 278, |
|
'semicolon': 278, |
|
'less': 564, |
|
'equal': 564, |
|
'greater': 564, |
|
'question': 444, |
|
'at': 921, |
|
'A': 722, |
|
'B': 667, |
|
'C': 667, |
|
'D': 722, |
|
'E': 611, |
|
'F': 556, |
|
'G': 722, |
|
'H': 722, |
|
'I': 333, |
|
'J': 389, |
|
'K': 722, |
|
'L': 611, |
|
'M': 889, |
|
'N': 722, |
|
'O': 722, |
|
'P': 556, |
|
'Q': 722, |
|
'R': 667, |
|
'S': 556, |
|
'T': 611, |
|
'U': 722, |
|
'V': 722, |
|
'W': 944, |
|
'X': 722, |
|
'Y': 722, |
|
'Z': 611, |
|
'bracketleft': 333, |
|
'backslash': 278, |
|
'bracketright': 333, |
|
'asciicircum': 469, |
|
'underscore': 500, |
|
'quoteleft': 333, |
|
'a': 444, |
|
'b': 500, |
|
'c': 444, |
|
'd': 500, |
|
'e': 444, |
|
'f': 333, |
|
'g': 500, |
|
'h': 500, |
|
'i': 278, |
|
'j': 278, |
|
'k': 500, |
|
'l': 278, |
|
'm': 778, |
|
'n': 500, |
|
'o': 500, |
|
'p': 500, |
|
'q': 500, |
|
'r': 333, |
|
's': 389, |
|
't': 278, |
|
'u': 500, |
|
'v': 500, |
|
'w': 722, |
|
'x': 500, |
|
'y': 500, |
|
'z': 444, |
|
'braceleft': 480, |
|
'bar': 200, |
|
'braceright': 480, |
|
'asciitilde': 541, |
|
'exclamdown': 333, |
|
'cent': 500, |
|
'sterling': 500, |
|
'fraction': 167, |
|
'yen': 500, |
|
'florin': 500, |
|
'section': 500, |
|
'currency': 500, |
|
'quotesingle': 180, |
|
'quotedblleft': 444, |
|
'guillemotleft': 500, |
|
'guilsinglleft': 333, |
|
'guilsinglright': 333, |
|
'fi': 556, |
|
'fl': 556, |
|
'endash': 500, |
|
'dagger': 500, |
|
'daggerdbl': 500, |
|
'periodcentered': 250, |
|
'paragraph': 453, |
|
'bullet': 350, |
|
'quotesinglbase': 333, |
|
'quotedblbase': 444, |
|
'quotedblright': 444, |
|
'guillemotright': 500, |
|
'ellipsis': 1000, |
|
'perthousand': 1000, |
|
'questiondown': 444, |
|
'grave': 333, |
|
'acute': 333, |
|
'circumflex': 333, |
|
'tilde': 333, |
|
'macron': 333, |
|
'breve': 333, |
|
'dotaccent': 333, |
|
'dieresis': 333, |
|
'ring': 333, |
|
'cedilla': 333, |
|
'hungarumlaut': 333, |
|
'ogonek': 333, |
|
'caron': 333, |
|
'emdash': 1000, |
|
'AE': 889, |
|
'ordfeminine': 276, |
|
'Lslash': 611, |
|
'Oslash': 722, |
|
'OE': 889, |
|
'ordmasculine': 310, |
|
'ae': 667, |
|
'dotlessi': 278, |
|
'lslash': 278, |
|
'oslash': 500, |
|
'oe': 722, |
|
'germandbls': 500, |
|
'Idieresis': 333, |
|
'eacute': 444, |
|
'abreve': 444, |
|
'uhungarumlaut': 500, |
|
'ecaron': 444, |
|
'Ydieresis': 722, |
|
'divide': 564, |
|
'Yacute': 722, |
|
'Acircumflex': 722, |
|
'aacute': 444, |
|
'Ucircumflex': 722, |
|
'yacute': 500, |
|
'scommaaccent': 389, |
|
'ecircumflex': 444, |
|
'Uring': 722, |
|
'Udieresis': 722, |
|
'aogonek': 444, |
|
'Uacute': 722, |
|
'uogonek': 500, |
|
'Edieresis': 611, |
|
'Dcroat': 722, |
|
'commaaccent': 250, |
|
'copyright': 760, |
|
'Emacron': 611, |
|
'ccaron': 444, |
|
'aring': 444, |
|
'Ncommaaccent': 722, |
|
'lacute': 278, |
|
'agrave': 444, |
|
'Tcommaaccent': 611, |
|
'Cacute': 667, |
|
'atilde': 444, |
|
'Edotaccent': 611, |
|
'scaron': 389, |
|
'scedilla': 389, |
|
'iacute': 278, |
|
'lozenge': 471, |
|
'Rcaron': 667, |
|
'Gcommaaccent': 722, |
|
'ucircumflex': 500, |
|
'acircumflex': 444, |
|
'Amacron': 722, |
|
'rcaron': 333, |
|
'ccedilla': 444, |
|
'Zdotaccent': 611, |
|
'Thorn': 556, |
|
'Omacron': 722, |
|
'Racute': 667, |
|
'Sacute': 556, |
|
'dcaron': 588, |
|
'Umacron': 722, |
|
'uring': 500, |
|
'threesuperior': 300, |
|
'Ograve': 722, |
|
'Agrave': 722, |
|
'Abreve': 722, |
|
'multiply': 564, |
|
'uacute': 500, |
|
'Tcaron': 611, |
|
'partialdiff': 476, |
|
'ydieresis': 500, |
|
'Nacute': 722, |
|
'icircumflex': 278, |
|
'Ecircumflex': 611, |
|
'adieresis': 444, |
|
'edieresis': 444, |
|
'cacute': 444, |
|
'nacute': 500, |
|
'umacron': 500, |
|
'Ncaron': 722, |
|
'Iacute': 333, |
|
'plusminus': 564, |
|
'brokenbar': 200, |
|
'registered': 760, |
|
'Gbreve': 722, |
|
'Idotaccent': 333, |
|
'summation': 600, |
|
'Egrave': 611, |
|
'racute': 333, |
|
'omacron': 500, |
|
'Zacute': 611, |
|
'Zcaron': 611, |
|
'greaterequal': 549, |
|
'Eth': 722, |
|
'Ccedilla': 667, |
|
'lcommaaccent': 278, |
|
'tcaron': 326, |
|
'eogonek': 444, |
|
'Uogonek': 722, |
|
'Aacute': 722, |
|
'Adieresis': 722, |
|
'egrave': 444, |
|
'zacute': 444, |
|
'iogonek': 278, |
|
'Oacute': 722, |
|
'oacute': 500, |
|
'amacron': 444, |
|
'sacute': 389, |
|
'idieresis': 278, |
|
'Ocircumflex': 722, |
|
'Ugrave': 722, |
|
'Delta': 612, |
|
'thorn': 500, |
|
'twosuperior': 300, |
|
'Odieresis': 722, |
|
'mu': 500, |
|
'igrave': 278, |
|
'ohungarumlaut': 500, |
|
'Eogonek': 611, |
|
'dcroat': 500, |
|
'threequarters': 750, |
|
'Scedilla': 556, |
|
'lcaron': 344, |
|
'Kcommaaccent': 722, |
|
'Lacute': 611, |
|
'trademark': 980, |
|
'edotaccent': 444, |
|
'Igrave': 333, |
|
'Imacron': 333, |
|
'Lcaron': 611, |
|
'onehalf': 750, |
|
'lessequal': 549, |
|
'ocircumflex': 500, |
|
'ntilde': 500, |
|
'Uhungarumlaut': 722, |
|
'Eacute': 611, |
|
'emacron': 444, |
|
'gbreve': 500, |
|
'onequarter': 750, |
|
'Scaron': 556, |
|
'Scommaaccent': 556, |
|
'Ohungarumlaut': 722, |
|
'degree': 400, |
|
'ograve': 500, |
|
'Ccaron': 667, |
|
'ugrave': 500, |
|
'radical': 453, |
|
'Dcaron': 722, |
|
'rcommaaccent': 333, |
|
'Ntilde': 722, |
|
'otilde': 500, |
|
'Rcommaaccent': 667, |
|
'Lcommaaccent': 611, |
|
'Atilde': 722, |
|
'Aogonek': 722, |
|
'Aring': 722, |
|
'Otilde': 722, |
|
'zdotaccent': 444, |
|
'Ecaron': 611, |
|
'Iogonek': 333, |
|
'kcommaaccent': 500, |
|
'minus': 564, |
|
'Icircumflex': 333, |
|
'ncaron': 500, |
|
'tcommaaccent': 278, |
|
'logicalnot': 564, |
|
'odieresis': 500, |
|
'udieresis': 500, |
|
'notequal': 549, |
|
'gcommaaccent': 500, |
|
'eth': 500, |
|
'zcaron': 444, |
|
'ncommaaccent': 500, |
|
'onesuperior': 300, |
|
'imacron': 278, |
|
'Euro': 500 |
|
}, |
|
'Times-Bold': { |
|
'space': 250, |
|
'exclam': 333, |
|
'quotedbl': 555, |
|
'numbersign': 500, |
|
'dollar': 500, |
|
'percent': 1000, |
|
'ampersand': 833, |
|
'quoteright': 333, |
|
'parenleft': 333, |
|
'parenright': 333, |
|
'asterisk': 500, |
|
'plus': 570, |
|
'comma': 250, |
|
'hyphen': 333, |
|
'period': 250, |
|
'slash': 278, |
|
'zero': 500, |
|
'one': 500, |
|
'two': 500, |
|
'three': 500, |
|
'four': 500, |
|
'five': 500, |
|
'six': 500, |
|
'seven': 500, |
|
'eight': 500, |
|
'nine': 500, |
|
'colon': 333, |
|
'semicolon': 333, |
|
'less': 570, |
|
'equal': 570, |
|
'greater': 570, |
|
'question': 500, |
|
'at': 930, |
|
'A': 722, |
|
'B': 667, |
|
'C': 722, |
|
'D': 722, |
|
'E': 667, |
|
'F': 611, |
|
'G': 778, |
|
'H': 778, |
|
'I': 389, |
|
'J': 500, |
|
'K': 778, |
|
'L': 667, |
|
'M': 944, |
|
'N': 722, |
|
'O': 778, |
|
'P': 611, |
|
'Q': 778, |
|
'R': 722, |
|
'S': 556, |
|
'T': 667, |
|
'U': 722, |
|
'V': 722, |
|
'W': 1000, |
|
'X': 722, |
|
'Y': 722, |
|
'Z': 667, |
|
'bracketleft': 333, |
|
'backslash': 278, |
|
'bracketright': 333, |
|
'asciicircum': 581, |
|
'underscore': 500, |
|
'quoteleft': 333, |
|
'a': 500, |
|
'b': 556, |
|
'c': 444, |
|
'd': 556, |
|
'e': 444, |
|
'f': 333, |
|
'g': 500, |
|
'h': 556, |
|
'i': 278, |
|
'j': 333, |
|
'k': 556, |
|
'l': 278, |
|
'm': 833, |
|
'n': 556, |
|
'o': 500, |
|
'p': 556, |
|
'q': 556, |
|
'r': 444, |
|
's': 389, |
|
't': 333, |
|
'u': 556, |
|
'v': 500, |
|
'w': 722, |
|
'x': 500, |
|
'y': 500, |
|
'z': 444, |
|
'braceleft': 394, |
|
'bar': 220, |
|
'braceright': 394, |
|
'asciitilde': 520, |
|
'exclamdown': 333, |
|
'cent': 500, |
|
'sterling': 500, |
|
'fraction': 167, |
|
'yen': 500, |
|
'florin': 500, |
|
'section': 500, |
|
'currency': 500, |
|
'quotesingle': 278, |
|
'quotedblleft': 500, |
|
'guillemotleft': 500, |
|
'guilsinglleft': 333, |
|
'guilsinglright': 333, |
|
'fi': 556, |
|
'fl': 556, |
|
'endash': 500, |
|
'dagger': 500, |
|
'daggerdbl': 500, |
|
'periodcentered': 250, |
|
'paragraph': 540, |
|
'bullet': 350, |
|
'quotesinglbase': 333, |
|
'quotedblbase': 500, |
|
'quotedblright': 500, |
|
'guillemotright': 500, |
|
'ellipsis': 1000, |
|
'perthousand': 1000, |
|
'questiondown': 500, |
|
'grave': 333, |
|
'acute': 333, |
|
'circumflex': 333, |
|
'tilde': 333, |
|
'macron': 333, |
|
'breve': 333, |
|
'dotaccent': 333, |
|
'dieresis': 333, |
|
'ring': 333, |
|
'cedilla': 333, |
|
'hungarumlaut': 333, |
|
'ogonek': 333, |
|
'caron': 333, |
|
'emdash': 1000, |
|
'AE': 1000, |
|
'ordfeminine': 300, |
|
'Lslash': 667, |
|
'Oslash': 778, |
|
'OE': 1000, |
|
'ordmasculine': 330, |
|
'ae': 722, |
|
'dotlessi': 278, |
|
'lslash': 278, |
|
'oslash': 500, |
|
'oe': 722, |
|
'germandbls': 556, |
|
'Idieresis': 389, |
|
'eacute': 444, |
|
'abreve': 500, |
|
'uhungarumlaut': 556, |
|
'ecaron': 444, |
|
'Ydieresis': 722, |
|
'divide': 570, |
|
'Yacute': 722, |
|
'Acircumflex': 722, |
|
'aacute': 500, |
|
'Ucircumflex': 722, |
|
'yacute': 500, |
|
'scommaaccent': 389, |
|
'ecircumflex': 444, |
|
'Uring': 722, |
|
'Udieresis': 722, |
|
'aogonek': 500, |
|
'Uacute': 722, |
|
'uogonek': 556, |
|
'Edieresis': 667, |
|
'Dcroat': 722, |
|
'commaaccent': 250, |
|
'copyright': 747, |
|
'Emacron': 667, |
|
'ccaron': 444, |
|
'aring': 500, |
|
'Ncommaaccent': 722, |
|
'lacute': 278, |
|
'agrave': 500, |
|
'Tcommaaccent': 667, |
|
'Cacute': 722, |
|
'atilde': 500, |
|
'Edotaccent': 667, |
|
'scaron': 389, |
|
'scedilla': 389, |
|
'iacute': 278, |
|
'lozenge': 494, |
|
'Rcaron': 722, |
|
'Gcommaaccent': 778, |
|
'ucircumflex': 556, |
|
'acircumflex': 500, |
|
'Amacron': 722, |
|
'rcaron': 444, |
|
'ccedilla': 444, |
|
'Zdotaccent': 667, |
|
'Thorn': 611, |
|
'Omacron': 778, |
|
'Racute': 722, |
|
'Sacute': 556, |
|
'dcaron': 672, |
|
'Umacron': 722, |
|
'uring': 556, |
|
'threesuperior': 300, |
|
'Ograve': 778, |
|
'Agrave': 722, |
|
'Abreve': 722, |
|
'multiply': 570, |
|
'uacute': 556, |
|
'Tcaron': 667, |
|
'partialdiff': 494, |
|
'ydieresis': 500, |
|
'Nacute': 722, |
|
'icircumflex': 278, |
|
'Ecircumflex': 667, |
|
'adieresis': 500, |
|
'edieresis': 444, |
|
'cacute': 444, |
|
'nacute': 556, |
|
'umacron': 556, |
|
'Ncaron': 722, |
|
'Iacute': 389, |
|
'plusminus': 570, |
|
'brokenbar': 220, |
|
'registered': 747, |
|
'Gbreve': 778, |
|
'Idotaccent': 389, |
|
'summation': 600, |
|
'Egrave': 667, |
|
'racute': 444, |
|
'omacron': 500, |
|
'Zacute': 667, |
|
'Zcaron': 667, |
|
'greaterequal': 549, |
|
'Eth': 722, |
|
'Ccedilla': 722, |
|
'lcommaaccent': 278, |
|
'tcaron': 416, |
|
'eogonek': 444, |
|
'Uogonek': 722, |
|
'Aacute': 722, |
|
'Adieresis': 722, |
|
'egrave': 444, |
|
'zacute': 444, |
|
'iogonek': 278, |
|
'Oacute': 778, |
|
'oacute': 500, |
|
'amacron': 500, |
|
'sacute': 389, |
|
'idieresis': 278, |
|
'Ocircumflex': 778, |
|
'Ugrave': 722, |
|
'Delta': 612, |
|
'thorn': 556, |
|
'twosuperior': 300, |
|
'Odieresis': 778, |
|
'mu': 556, |
|
'igrave': 278, |
|
'ohungarumlaut': 500, |
|
'Eogonek': 667, |
|
'dcroat': 556, |
|
'threequarters': 750, |
|
'Scedilla': 556, |
|
'lcaron': 394, |
|
'Kcommaaccent': 778, |
|
'Lacute': 667, |
|
'trademark': 1000, |
|
'edotaccent': 444, |
|
'Igrave': 389, |
|
'Imacron': 389, |
|
'Lcaron': 667, |
|
'onehalf': 750, |
|
'lessequal': 549, |
|
'ocircumflex': 500, |
|
'ntilde': 556, |
|
'Uhungarumlaut': 722, |
|
'Eacute': 667, |
|
'emacron': 444, |
|
'gbreve': 500, |
|
'onequarter': 750, |
|
'Scaron': 556, |
|
'Scommaaccent': 556, |
|
'Ohungarumlaut': 778, |
|
'degree': 400, |
|
'ograve': 500, |
|
'Ccaron': 722, |
|
'ugrave': 556, |
|
'radical': 549, |
|
'Dcaron': 722, |
|
'rcommaaccent': 444, |
|
'Ntilde': 722, |
|
'otilde': 500, |
|
'Rcommaaccent': 722, |
|
'Lcommaaccent': 667, |
|
'Atilde': 722, |
|
'Aogonek': 722, |
|
'Aring': 722, |
|
'Otilde': 778, |
|
'zdotaccent': 444, |
|
'Ecaron': 667, |
|
'Iogonek': 389, |
|
'kcommaaccent': 556, |
|
'minus': 570, |
|
'Icircumflex': 389, |
|
'ncaron': 556, |
|
'tcommaaccent': 333, |
|
'logicalnot': 570, |
|
'odieresis': 500, |
|
'udieresis': 556, |
|
'notequal': 549, |
|
'gcommaaccent': 500, |
|
'eth': 500, |
|
'zcaron': 444, |
|
'ncommaaccent': 556, |
|
'onesuperior': 300, |
|
'imacron': 278, |
|
'Euro': 500 |
|
}, |
|
'Times-BoldItalic': { |
|
'space': 250, |
|
'exclam': 389, |
|
'quotedbl': 555, |
|
'numbersign': 500, |
|
'dollar': 500, |
|
'percent': 833, |
|
'ampersand': 778, |
|
'quoteright': 333, |
|
'parenleft': 333, |
|
'parenright': 333, |
|
'asterisk': 500, |
|
'plus': 570, |
|
'comma': 250, |
|
'hyphen': 333, |
|
'period': 250, |
|
'slash': 278, |
|
'zero': 500, |
|
'one': 500, |
|
'two': 500, |
|
'three': 500, |
|
'four': 500, |
|
'five': 500, |
|
'six': 500, |
|
'seven': 500, |
|
'eight': 500, |
|
'nine': 500, |
|
'colon': 333, |
|
'semicolon': 333, |
|
'less': 570, |
|
'equal': 570, |
|
'greater': 570, |
|
'question': 500, |
|
'at': 832, |
|
'A': 667, |
|
'B': 667, |
|
'C': 667, |
|
'D': 722, |
|
'E': 667, |
|
'F': 667, |
|
'G': 722, |
|
'H': 778, |
|
'I': 389, |
|
'J': 500, |
|
'K': 667, |
|
'L': 611, |
|
'M': 889, |
|
'N': 722, |
|
'O': 722, |
|
'P': 611, |
|
'Q': 722, |
|
'R': 667, |
|
'S': 556, |
|
'T': 611, |
|
'U': 722, |
|
'V': 667, |
|
'W': 889, |
|
'X': 667, |
|
'Y': 611, |
|
'Z': 611, |
|
'bracketleft': 333, |
|
'backslash': 278, |
|
'bracketright': 333, |
|
'asciicircum': 570, |
|
'underscore': 500, |
|
'quoteleft': 333, |
|
'a': 500, |
|
'b': 500, |
|
'c': 444, |
|
'd': 500, |
|
'e': 444, |
|
'f': 333, |
|
'g': 500, |
|
'h': 556, |
|
'i': 278, |
|
'j': 278, |
|
'k': 500, |
|
'l': 278, |
|
'm': 778, |
|
'n': 556, |
|
'o': 500, |
|
'p': 500, |
|
'q': 500, |
|
'r': 389, |
|
's': 389, |
|
't': 278, |
|
'u': 556, |
|
'v': 444, |
|
'w': 667, |
|
'x': 500, |
|
'y': 444, |
|
'z': 389, |
|
'braceleft': 348, |
|
'bar': 220, |
|
'braceright': 348, |
|
'asciitilde': 570, |
|
'exclamdown': 389, |
|
'cent': 500, |
|
'sterling': 500, |
|
'fraction': 167, |
|
'yen': 500, |
|
'florin': 500, |
|
'section': 500, |
|
'currency': 500, |
|
'quotesingle': 278, |
|
'quotedblleft': 500, |
|
'guillemotleft': 500, |
|
'guilsinglleft': 333, |
|
'guilsinglright': 333, |
|
'fi': 556, |
|
'fl': 556, |
|
'endash': 500, |
|
'dagger': 500, |
|
'daggerdbl': 500, |
|
'periodcentered': 250, |
|
'paragraph': 500, |
|
'bullet': 350, |
|
'quotesinglbase': 333, |
|
'quotedblbase': 500, |
|
'quotedblright': 500, |
|
'guillemotright': 500, |
|
'ellipsis': 1000, |
|
'perthousand': 1000, |
|
'questiondown': 500, |
|
'grave': 333, |
|
'acute': 333, |
|
'circumflex': 333, |
|
'tilde': 333, |
|
'macron': 333, |
|
'breve': 333, |
|
'dotaccent': 333, |
|
'dieresis': 333, |
|
'ring': 333, |
|
'cedilla': 333, |
|
'hungarumlaut': 333, |
|
'ogonek': 333, |
|
'caron': 333, |
|
'emdash': 1000, |
|
'AE': 944, |
|
'ordfeminine': 266, |
|
'Lslash': 611, |
|
'Oslash': 722, |
|
'OE': 944, |
|
'ordmasculine': 300, |
|
'ae': 722, |
|
'dotlessi': 278, |
|
'lslash': 278, |
|
'oslash': 500, |
|
'oe': 722, |
|
'germandbls': 500, |
|
'Idieresis': 389, |
|
'eacute': 444, |
|
'abreve': 500, |
|
'uhungarumlaut': 556, |
|
'ecaron': 444, |
|
'Ydieresis': 611, |
|
'divide': 570, |
|
'Yacute': 611, |
|
'Acircumflex': 667, |
|
'aacute': 500, |
|
'Ucircumflex': 722, |
|
'yacute': 444, |
|
'scommaaccent': 389, |
|
'ecircumflex': 444, |
|
'Uring': 722, |
|
'Udieresis': 722, |
|
'aogonek': 500, |
|
'Uacute': 722, |
|
'uogonek': 556, |
|
'Edieresis': 667, |
|
'Dcroat': 722, |
|
'commaaccent': 250, |
|
'copyright': 747, |
|
'Emacron': 667, |
|
'ccaron': 444, |
|
'aring': 500, |
|
'Ncommaaccent': 722, |
|
'lacute': 278, |
|
'agrave': 500, |
|
'Tcommaaccent': 611, |
|
'Cacute': 667, |
|
'atilde': 500, |
|
'Edotaccent': 667, |
|
'scaron': 389, |
|
'scedilla': 389, |
|
'iacute': 278, |
|
'lozenge': 494, |
|
'Rcaron': 667, |
|
'Gcommaaccent': 722, |
|
'ucircumflex': 556, |
|
'acircumflex': 500, |
|
'Amacron': 667, |
|
'rcaron': 389, |
|
'ccedilla': 444, |
|
'Zdotaccent': 611, |
|
'Thorn': 611, |
|
'Omacron': 722, |
|
'Racute': 667, |
|
'Sacute': 556, |
|
'dcaron': 608, |
|
'Umacron': 722, |
|
'uring': 556, |
|
'threesuperior': 300, |
|
'Ograve': 722, |
|
'Agrave': 667, |
|
'Abreve': 667, |
|
'multiply': 570, |
|
'uacute': 556, |
|
'Tcaron': 611, |
|
'partialdiff': 494, |
|
'ydieresis': 444, |
|
'Nacute': 722, |
|
'icircumflex': 278, |
|
'Ecircumflex': 667, |
|
'adieresis': 500, |
|
'edieresis': 444, |
|
'cacute': 444, |
|
'nacute': 556, |
|
'umacron': 556, |
|
'Ncaron': 722, |
|
'Iacute': 389, |
|
'plusminus': 570, |
|
'brokenbar': 220, |
|
'registered': 747, |
|
'Gbreve': 722, |
|
'Idotaccent': 389, |
|
'summation': 600, |
|
'Egrave': 667, |
|
'racute': 389, |
|
'omacron': 500, |
|
'Zacute': 611, |
|
'Zcaron': 611, |
|
'greaterequal': 549, |
|
'Eth': 722, |
|
'Ccedilla': 667, |
|
'lcommaaccent': 278, |
|
'tcaron': 366, |
|
'eogonek': 444, |
|
'Uogonek': 722, |
|
'Aacute': 667, |
|
'Adieresis': 667, |
|
'egrave': 444, |
|
'zacute': 389, |
|
'iogonek': 278, |
|
'Oacute': 722, |
|
'oacute': 500, |
|
'amacron': 500, |
|
'sacute': 389, |
|
'idieresis': 278, |
|
'Ocircumflex': 722, |
|
'Ugrave': 722, |
|
'Delta': 612, |
|
'thorn': 500, |
|
'twosuperior': 300, |
|
'Odieresis': 722, |
|
'mu': 576, |
|
'igrave': 278, |
|
'ohungarumlaut': 500, |
|
'Eogonek': 667, |
|
'dcroat': 500, |
|
'threequarters': 750, |
|
'Scedilla': 556, |
|
'lcaron': 382, |
|
'Kcommaaccent': 667, |
|
'Lacute': 611, |
|
'trademark': 1000, |
|
'edotaccent': 444, |
|
'Igrave': 389, |
|
'Imacron': 389, |
|
'Lcaron': 611, |
|
'onehalf': 750, |
|
'lessequal': 549, |
|
'ocircumflex': 500, |
|
'ntilde': 556, |
|
'Uhungarumlaut': 722, |
|
'Eacute': 667, |
|
'emacron': 444, |
|
'gbreve': 500, |
|
'onequarter': 750, |
|
'Scaron': 556, |
|
'Scommaaccent': 556, |
|
'Ohungarumlaut': 722, |
|
'degree': 400, |
|
'ograve': 500, |
|
'Ccaron': 667, |
|
'ugrave': 556, |
|
'radical': 549, |
|
'Dcaron': 722, |
|
'rcommaaccent': 389, |
|
'Ntilde': 722, |
|
'otilde': 500, |
|
'Rcommaaccent': 667, |
|
'Lcommaaccent': 611, |
|
'Atilde': 667, |
|
'Aogonek': 667, |
|
'Aring': 667, |
|
'Otilde': 722, |
|
'zdotaccent': 389, |
|
'Ecaron': 667, |
|
'Iogonek': 389, |
|
'kcommaaccent': 500, |
|
'minus': 606, |
|
'Icircumflex': 389, |
|
'ncaron': 556, |
|
'tcommaaccent': 278, |
|
'logicalnot': 606, |
|
'odieresis': 500, |
|
'udieresis': 556, |
|
'notequal': 549, |
|
'gcommaaccent': 500, |
|
'eth': 500, |
|
'zcaron': 389, |
|
'ncommaaccent': 556, |
|
'onesuperior': 300, |
|
'imacron': 278, |
|
'Euro': 500 |
|
}, |
|
'Times-Italic': { |
|
'space': 250, |
|
'exclam': 333, |
|
'quotedbl': 420, |
|
'numbersign': 500, |
|
'dollar': 500, |
|
'percent': 833, |
|
'ampersand': 778, |
|
'quoteright': 333, |
|
'parenleft': 333, |
|
'parenright': 333, |
|
'asterisk': 500, |
|
'plus': 675, |
|
'comma': 250, |
|
'hyphen': 333, |
|
'period': 250, |
|
'slash': 278, |
|
'zero': 500, |
|
'one': 500, |
|
'two': 500, |
|
'three': 500, |
|
'four': 500, |
|
'five': 500, |
|
'six': 500, |
|
'seven': 500, |
|
'eight': 500, |
|
'nine': 500, |
|
'colon': 333, |
|
'semicolon': 333, |
|
'less': 675, |
|
'equal': 675, |
|
'greater': 675, |
|
'question': 500, |
|
'at': 920, |
|
'A': 611, |
|
'B': 611, |
|
'C': 667, |
|
'D': 722, |
|
'E': 611, |
|
'F': 611, |
|
'G': 722, |
|
'H': 722, |
|
'I': 333, |
|
'J': 444, |
|
'K': 667, |
|
'L': 556, |
|
'M': 833, |
|
'N': 667, |
|
'O': 722, |
|
'P': 611, |
|
'Q': 722, |
|
'R': 611, |
|
'S': 500, |
|
'T': 556, |
|
'U': 722, |
|
'V': 611, |
|
'W': 833, |
|
'X': 611, |
|
'Y': 556, |
|
'Z': 556, |
|
'bracketleft': 389, |
|
'backslash': 278, |
|
'bracketright': 389, |
|
'asciicircum': 422, |
|
'underscore': 500, |
|
'quoteleft': 333, |
|
'a': 500, |
|
'b': 500, |
|
'c': 444, |
|
'd': 500, |
|
'e': 444, |
|
'f': 278, |
|
'g': 500, |
|
'h': 500, |
|
'i': 278, |
|
'j': 278, |
|
'k': 444, |
|
'l': 278, |
|
'm': 722, |
|
'n': 500, |
|
'o': 500, |
|
'p': 500, |
|
'q': 500, |
|
'r': 389, |
|
's': 389, |
|
't': 278, |
|
'u': 500, |
|
'v': 444, |
|
'w': 667, |
|
'x': 444, |
|
'y': 444, |
|
'z': 389, |
|
'braceleft': 400, |
|
'bar': 275, |
|
'braceright': 400, |
|
'asciitilde': 541, |
|
'exclamdown': 389, |
|
'cent': 500, |
|
'sterling': 500, |
|
'fraction': 167, |
|
'yen': 500, |
|
'florin': 500, |
|
'section': 500, |
|
'currency': 500, |
|
'quotesingle': 214, |
|
'quotedblleft': 556, |
|
'guillemotleft': 500, |
|
'guilsinglleft': 333, |
|
'guilsinglright': 333, |
|
'fi': 500, |
|
'fl': 500, |
|
'endash': 500, |
|
'dagger': 500, |
|
'daggerdbl': 500, |
|
'periodcentered': 250, |
|
'paragraph': 523, |
|
'bullet': 350, |
|
'quotesinglbase': 333, |
|
'quotedblbase': 556, |
|
'quotedblright': 556, |
|
'guillemotright': 500, |
|
'ellipsis': 889, |
|
'perthousand': 1000, |
|
'questiondown': 500, |
|
'grave': 333, |
|
'acute': 333, |
|
'circumflex': 333, |
|
'tilde': 333, |
|
'macron': 333, |
|
'breve': 333, |
|
'dotaccent': 333, |
|
'dieresis': 333, |
|
'ring': 333, |
|
'cedilla': 333, |
|
'hungarumlaut': 333, |
|
'ogonek': 333, |
|
'caron': 333, |
|
'emdash': 889, |
|
'AE': 889, |
|
'ordfeminine': 276, |
|
'Lslash': 556, |
|
'Oslash': 722, |
|
'OE': 944, |
|
'ordmasculine': 310, |
|
'ae': 667, |
|
'dotlessi': 278, |
|
'lslash': 278, |
|
'oslash': 500, |
|
'oe': 667, |
|
'germandbls': 500, |
|
'Idieresis': 333, |
|
'eacute': 444, |
|
'abreve': 500, |
|
'uhungarumlaut': 500, |
|
'ecaron': 444, |
|
'Ydieresis': 556, |
|
'divide': 675, |
|
'Yacute': 556, |
|
'Acircumflex': 611, |
|
'aacute': 500, |
|
'Ucircumflex': 722, |
|
'yacute': 444, |
|
'scommaaccent': 389, |
|
'ecircumflex': 444, |
|
'Uring': 722, |
|
'Udieresis': 722, |
|
'aogonek': 500, |
|
'Uacute': 722, |
|
'uogonek': 500, |
|
'Edieresis': 611, |
|
'Dcroat': 722, |
|
'commaaccent': 250, |
|
'copyright': 760, |
|
'Emacron': 611, |
|
'ccaron': 444, |
|
'aring': 500, |
|
'Ncommaaccent': 667, |
|
'lacute': 278, |
|
'agrave': 500, |
|
'Tcommaaccent': 556, |
|
'Cacute': 667, |
|
'atilde': 500, |
|
'Edotaccent': 611, |
|
'scaron': 389, |
|
'scedilla': 389, |
|
'iacute': 278, |
|
'lozenge': 471, |
|
'Rcaron': 611, |
|
'Gcommaaccent': 722, |
|
'ucircumflex': 500, |
|
'acircumflex': 500, |
|
'Amacron': 611, |
|
'rcaron': 389, |
|
'ccedilla': 444, |
|
'Zdotaccent': 556, |
|
'Thorn': 611, |
|
'Omacron': 722, |
|
'Racute': 611, |
|
'Sacute': 500, |
|
'dcaron': 544, |
|
'Umacron': 722, |
|
'uring': 500, |
|
'threesuperior': 300, |
|
'Ograve': 722, |
|
'Agrave': 611, |
|
'Abreve': 611, |
|
'multiply': 675, |
|
'uacute': 500, |
|
'Tcaron': 556, |
|
'partialdiff': 476, |
|
'ydieresis': 444, |
|
'Nacute': 667, |
|
'icircumflex': 278, |
|
'Ecircumflex': 611, |
|
'adieresis': 500, |
|
'edieresis': 444, |
|
'cacute': 444, |
|
'nacute': 500, |
|
'umacron': 500, |
|
'Ncaron': 667, |
|
'Iacute': 333, |
|
'plusminus': 675, |
|
'brokenbar': 275, |
|
'registered': 760, |
|
'Gbreve': 722, |
|
'Idotaccent': 333, |
|
'summation': 600, |
|
'Egrave': 611, |
|
'racute': 389, |
|
'omacron': 500, |
|
'Zacute': 556, |
|
'Zcaron': 556, |
|
'greaterequal': 549, |
|
'Eth': 722, |
|
'Ccedilla': 667, |
|
'lcommaaccent': 278, |
|
'tcaron': 300, |
|
'eogonek': 444, |
|
'Uogonek': 722, |
|
'Aacute': 611, |
|
'Adieresis': 611, |
|
'egrave': 444, |
|
'zacute': 389, |
|
'iogonek': 278, |
|
'Oacute': 722, |
|
'oacute': 500, |
|
'amacron': 500, |
|
'sacute': 389, |
|
'idieresis': 278, |
|
'Ocircumflex': 722, |
|
'Ugrave': 722, |
|
'Delta': 612, |
|
'thorn': 500, |
|
'twosuperior': 300, |
|
'Odieresis': 722, |
|
'mu': 500, |
|
'igrave': 278, |
|
'ohungarumlaut': 500, |
|
'Eogonek': 611, |
|
'dcroat': 500, |
|
'threequarters': 750, |
|
'Scedilla': 500, |
|
'lcaron': 300, |
|
'Kcommaaccent': 667, |
|
'Lacute': 556, |
|
'trademark': 980, |
|
'edotaccent': 444, |
|
'Igrave': 333, |
|
'Imacron': 333, |
|
'Lcaron': 611, |
|
'onehalf': 750, |
|
'lessequal': 549, |
|
'ocircumflex': 500, |
|
'ntilde': 500, |
|
'Uhungarumlaut': 722, |
|
'Eacute': 611, |
|
'emacron': 444, |
|
'gbreve': 500, |
|
'onequarter': 750, |
|
'Scaron': 500, |
|
'Scommaaccent': 500, |
|
'Ohungarumlaut': 722, |
|
'degree': 400, |
|
'ograve': 500, |
|
'Ccaron': 667, |
|
'ugrave': 500, |
|
'radical': 453, |
|
'Dcaron': 722, |
|
'rcommaaccent': 389, |
|
'Ntilde': 667, |
|
'otilde': 500, |
|
'Rcommaaccent': 611, |
|
'Lcommaaccent': 556, |
|
'Atilde': 611, |
|
'Aogonek': 611, |
|
'Aring': 611, |
|
'Otilde': 722, |
|
'zdotaccent': 389, |
|
'Ecaron': 611, |
|
'Iogonek': 333, |
|
'kcommaaccent': 444, |
|
'minus': 675, |
|
'Icircumflex': 333, |
|
'ncaron': 500, |
|
'tcommaaccent': 278, |
|
'logicalnot': 675, |
|
'odieresis': 500, |
|
'udieresis': 500, |
|
'notequal': 549, |
|
'gcommaaccent': 500, |
|
'eth': 500, |
|
'zcaron': 389, |
|
'ncommaaccent': 500, |
|
'onesuperior': 300, |
|
'imacron': 278, |
|
'Euro': 500 |
|
}, |
|
'ZapfDingbats': { |
|
'space': 278, |
|
'a1': 974, |
|
'a2': 961, |
|
'a202': 974, |
|
'a3': 980, |
|
'a4': 719, |
|
'a5': 789, |
|
'a119': 790, |
|
'a118': 791, |
|
'a117': 690, |
|
'a11': 960, |
|
'a12': 939, |
|
'a13': 549, |
|
'a14': 855, |
|
'a15': 911, |
|
'a16': 933, |
|
'a105': 911, |
|
'a17': 945, |
|
'a18': 974, |
|
'a19': 755, |
|
'a20': 846, |
|
'a21': 762, |
|
'a22': 761, |
|
'a23': 571, |
|
'a24': 677, |
|
'a25': 763, |
|
'a26': 760, |
|
'a27': 759, |
|
'a28': 754, |
|
'a6': 494, |
|
'a7': 552, |
|
'a8': 537, |
|
'a9': 577, |
|
'a10': 692, |
|
'a29': 786, |
|
'a30': 788, |
|
'a31': 788, |
|
'a32': 790, |
|
'a33': 793, |
|
'a34': 794, |
|
'a35': 816, |
|
'a36': 823, |
|
'a37': 789, |
|
'a38': 841, |
|
'a39': 823, |
|
'a40': 833, |
|
'a41': 816, |
|
'a42': 831, |
|
'a43': 923, |
|
'a44': 744, |
|
'a45': 723, |
|
'a46': 749, |
|
'a47': 790, |
|
'a48': 792, |
|
'a49': 695, |
|
'a50': 776, |
|
'a51': 768, |
|
'a52': 792, |
|
'a53': 759, |
|
'a54': 707, |
|
'a55': 708, |
|
'a56': 682, |
|
'a57': 701, |
|
'a58': 826, |
|
'a59': 815, |
|
'a60': 789, |
|
'a61': 789, |
|
'a62': 707, |
|
'a63': 687, |
|
'a64': 696, |
|
'a65': 689, |
|
'a66': 786, |
|
'a67': 787, |
|
'a68': 713, |
|
'a69': 791, |
|
'a70': 785, |
|
'a71': 791, |
|
'a72': 873, |
|
'a73': 761, |
|
'a74': 762, |
|
'a203': 762, |
|
'a75': 759, |
|
'a204': 759, |
|
'a76': 892, |
|
'a77': 892, |
|
'a78': 788, |
|
'a79': 784, |
|
'a81': 438, |
|
'a82': 138, |
|
'a83': 277, |
|
'a84': 415, |
|
'a97': 392, |
|
'a98': 392, |
|
'a99': 668, |
|
'a100': 668, |
|
'a89': 390, |
|
'a90': 390, |
|
'a93': 317, |
|
'a94': 317, |
|
'a91': 276, |
|
'a92': 276, |
|
'a205': 509, |
|
'a85': 509, |
|
'a206': 410, |
|
'a86': 410, |
|
'a87': 234, |
|
'a88': 234, |
|
'a95': 334, |
|
'a96': 334, |
|
'a101': 732, |
|
'a102': 544, |
|
'a103': 544, |
|
'a104': 910, |
|
'a106': 667, |
|
'a107': 760, |
|
'a108': 760, |
|
'a112': 776, |
|
'a111': 595, |
|
'a110': 694, |
|
'a109': 626, |
|
'a120': 788, |
|
'a121': 788, |
|
'a122': 788, |
|
'a123': 788, |
|
'a124': 788, |
|
'a125': 788, |
|
'a126': 788, |
|
'a127': 788, |
|
'a128': 788, |
|
'a129': 788, |
|
'a130': 788, |
|
'a131': 788, |
|
'a132': 788, |
|
'a133': 788, |
|
'a134': 788, |
|
'a135': 788, |
|
'a136': 788, |
|
'a137': 788, |
|
'a138': 788, |
|
'a139': 788, |
|
'a140': 788, |
|
'a141': 788, |
|
'a142': 788, |
|
'a143': 788, |
|
'a144': 788, |
|
'a145': 788, |
|
'a146': 788, |
|
'a147': 788, |
|
'a148': 788, |
|
'a149': 788, |
|
'a150': 788, |
|
'a151': 788, |
|
'a152': 788, |
|
'a153': 788, |
|
'a154': 788, |
|
'a155': 788, |
|
'a156': 788, |
|
'a157': 788, |
|
'a158': 788, |
|
'a159': 788, |
|
'a160': 894, |
|
'a161': 838, |
|
'a163': 1016, |
|
'a164': 458, |
|
'a196': 748, |
|
'a165': 924, |
|
'a192': 748, |
|
'a166': 918, |
|
'a167': 927, |
|
'a168': 928, |
|
'a169': 928, |
|
'a170': 834, |
|
'a171': 873, |
|
'a172': 828, |
|
'a173': 924, |
|
'a162': 924, |
|
'a174': 917, |
|
'a175': 930, |
|
'a176': 931, |
|
'a177': 463, |
|
'a178': 883, |
|
'a179': 836, |
|
'a193': 836, |
|
'a180': 867, |
|
'a199': 867, |
|
'a181': 696, |
|
'a200': 696, |
|
'a182': 874, |
|
'a201': 874, |
|
'a183': 760, |
|
'a184': 946, |
|
'a197': 771, |
|
'a185': 865, |
|
'a194': 771, |
|
'a198': 888, |
|
'a186': 967, |
|
'a195': 888, |
|
'a187': 831, |
|
'a188': 873, |
|
'a189': 927, |
|
'a190': 970, |
|
'a191': 918 |
|
} |
|
}; |
|
|
|
exports.Metrics = Metrics; |
|
})); |
|
|
|
|
|
|
|
|
|
var NetworkManager = (function NetworkManagerClosure() { |
|
|
|
var OK_RESPONSE = 200; |
|
var PARTIAL_CONTENT_RESPONSE = 206; |
|
|
|
function NetworkManager(url, args) { |
|
this.url = url; |
|
args = args || {}; |
|
this.isHttp = /^https?:/i.test(url); |
|
this.httpHeaders = (this.isHttp && args.httpHeaders) || {}; |
|
this.withCredentials = args.withCredentials || false; |
|
this.getXhr = args.getXhr || |
|
function NetworkManager_getXhr() { |
|
return new XMLHttpRequest(); |
|
}; |
|
|
|
this.currXhrId = 0; |
|
this.pendingRequests = {}; |
|
this.loadedRequests = {}; |
|
} |
|
|
|
function getArrayBuffer(xhr) { |
|
var data = xhr.response; |
|
if (typeof data !== 'string') { |
|
return data; |
|
} |
|
var length = data.length; |
|
var array = new Uint8Array(length); |
|
for (var i = 0; i < length; i++) { |
|
array[i] = data.charCodeAt(i) & 0xFF; |
|
} |
|
return array.buffer; |
|
} |
|
|
|
var supportsMozChunked = (function supportsMozChunkedClosure() { |
|
try { |
|
var x = new XMLHttpRequest(); |
|
// Firefox 37- required .open() to be called before setting responseType. |
|
// https://bugzilla.mozilla.org/show_bug.cgi?id=707484 |
|
// Even though the URL is not visited, .open() could fail if the URL is |
|
// blocked, e.g. via the connect-src CSP directive or the NoScript addon. |
|
// When this error occurs, this feature detection method will mistakenly |
|
// report that moz-chunked-arraybuffer is not supported in Firefox 37-. |
|
x.open('GET', 'https://example.com'); |
|
x.responseType = 'moz-chunked-arraybuffer'; |
|
return x.responseType === 'moz-chunked-arraybuffer'; |
|
} catch (e) { |
|
return false; |
|
} |
|
})(); |
|
|
|
NetworkManager.prototype = { |
|
requestRange: function NetworkManager_requestRange(begin, end, listeners) { |
|
var args = { |
|
begin: begin, |
|
end: end |
|
}; |
|
for (var prop in listeners) { |
|
args[prop] = listeners[prop]; |
|
} |
|
return this.request(args); |
|
}, |
|
|
|
requestFull: function NetworkManager_requestFull(listeners) { |
|
return this.request(listeners); |
|
}, |
|
|
|
request: function NetworkManager_request(args) { |
|
var xhr = this.getXhr(); |
|
var xhrId = this.currXhrId++; |
|
var pendingRequest = this.pendingRequests[xhrId] = { |
|
xhr: xhr |
|
}; |
|
|
|
xhr.open('GET', this.url); |
|
xhr.withCredentials = this.withCredentials; |
|
for (var property in this.httpHeaders) { |
|
var value = this.httpHeaders[property]; |
|
if (typeof value === 'undefined') { |
|
continue; |
|
} |
|
xhr.setRequestHeader(property, value); |
|
} |
|
if (this.isHttp && 'begin' in args && 'end' in args) { |
|
var rangeStr = args.begin + '-' + (args.end - 1); |
|
xhr.setRequestHeader('Range', 'bytes=' + rangeStr); |
|
pendingRequest.expectedStatus = 206; |
|
} else { |
|
pendingRequest.expectedStatus = 200; |
|
} |
|
|
|
var useMozChunkedLoading = supportsMozChunked && !!args.onProgressiveData; |
|
if (useMozChunkedLoading) { |
|
xhr.responseType = 'moz-chunked-arraybuffer'; |
|
pendingRequest.onProgressiveData = args.onProgressiveData; |
|
pendingRequest.mozChunked = true; |
|
} else { |
|
xhr.responseType = 'arraybuffer'; |
|
} |
|
|
|
if (args.onError) { |
|
xhr.onerror = function(evt) { |
|
args.onError(xhr.status); |
|
}; |
|
} |
|
xhr.onreadystatechange = this.onStateChange.bind(this, xhrId); |
|
xhr.onprogress = this.onProgress.bind(this, xhrId); |
|
|
|
pendingRequest.onHeadersReceived = args.onHeadersReceived; |
|
pendingRequest.onDone = args.onDone; |
|
pendingRequest.onError = args.onError; |
|
pendingRequest.onProgress = args.onProgress; |
|
|
|
xhr.send(null); |
|
|
|
return xhrId; |
|
}, |
|
|
|
onProgress: function NetworkManager_onProgress(xhrId, evt) { |
|
var pendingRequest = this.pendingRequests[xhrId]; |
|
if (!pendingRequest) { |
|
// Maybe abortRequest was called... |
|
return; |
|
} |
|
|
|
if (pendingRequest.mozChunked) { |
|
var chunk = getArrayBuffer(pendingRequest.xhr); |
|
pendingRequest.onProgressiveData(chunk); |
|
} |
|
|
|
var onProgress = pendingRequest.onProgress; |
|
if (onProgress) { |
|
onProgress(evt); |
|
} |
|
}, |
|
|
|
onStateChange: function NetworkManager_onStateChange(xhrId, evt) { |
|
var pendingRequest = this.pendingRequests[xhrId]; |
|
if (!pendingRequest) { |
|
// Maybe abortRequest was called... |
|
return; |
|
} |
|
|
|
var xhr = pendingRequest.xhr; |
|
if (xhr.readyState >= 2 && pendingRequest.onHeadersReceived) { |
|
pendingRequest.onHeadersReceived(); |
|
delete pendingRequest.onHeadersReceived; |
|
} |
|
|
|
if (xhr.readyState !== 4) { |
|
return; |
|
} |
|
|
|
if (!(xhrId in this.pendingRequests)) { |
|
// The XHR request might have been aborted in onHeadersReceived() |
|
// callback, in which case we should abort request |
|
return; |
|
} |
|
|
|
delete this.pendingRequests[xhrId]; |
|
|
|
// success status == 0 can be on ftp, file and other protocols |
|
if (xhr.status === 0 && this.isHttp) { |
|
if (pendingRequest.onError) { |
|
pendingRequest.onError(xhr.status); |
|
} |
|
return; |
|
} |
|
var xhrStatus = xhr.status || OK_RESPONSE; |
|
|
|
// From http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.35.2: |
|
// "A server MAY ignore the Range header". This means it's possible to |
|
// get a 200 rather than a 206 response from a range request. |
|
var ok_response_on_range_request = |
|
xhrStatus === OK_RESPONSE && |
|
pendingRequest.expectedStatus === PARTIAL_CONTENT_RESPONSE; |
|
|
|
if (!ok_response_on_range_request && |
|
xhrStatus !== pendingRequest.expectedStatus) { |
|
if (pendingRequest.onError) { |
|
pendingRequest.onError(xhr.status); |
|
} |
|
return; |
|
} |
|
|
|
this.loadedRequests[xhrId] = true; |
|
|
|
var chunk = getArrayBuffer(xhr); |
|
if (xhrStatus === PARTIAL_CONTENT_RESPONSE) { |
|
var rangeHeader = xhr.getResponseHeader('Content-Range'); |
|
var matches = /bytes (\d+)-(\d+)\/(\d+)/.exec(rangeHeader); |
|
var begin = parseInt(matches[1], 10); |
|
pendingRequest.onDone({ |
|
begin: begin, |
|
chunk: chunk |
|
}); |
|
} else if (pendingRequest.onProgressiveData) { |
|
pendingRequest.onDone(null); |
|
} else if (chunk) { |
|
pendingRequest.onDone({ |
|
begin: 0, |
|
chunk: chunk |
|
}); |
|
} else if (pendingRequest.onError) { |
|
pendingRequest.onError(xhr.status); |
|
} |
|
}, |
|
|
|
hasPendingRequests: function NetworkManager_hasPendingRequests() { |
|
for (var xhrId in this.pendingRequests) { |
|
return true; |
|
} |
|
return false; |
|
}, |
|
|
|
getRequestXhr: function NetworkManager_getXhr(xhrId) { |
|
return this.pendingRequests[xhrId].xhr; |
|
}, |
|
|
|
isStreamingRequest: function NetworkManager_isStreamingRequest(xhrId) { |
|
return !!(this.pendingRequests[xhrId].onProgressiveData); |
|
}, |
|
|
|
isPendingRequest: function NetworkManager_isPendingRequest(xhrId) { |
|
return xhrId in this.pendingRequests; |
|
}, |
|
|
|
isLoadedRequest: function NetworkManager_isLoadedRequest(xhrId) { |
|
return xhrId in this.loadedRequests; |
|
}, |
|
|
|
abortAllRequests: function NetworkManager_abortAllRequests() { |
|
for (var xhrId in this.pendingRequests) { |
|
this.abortRequest(xhrId | 0); |
|
} |
|
}, |
|
|
|
abortRequest: function NetworkManager_abortRequest(xhrId) { |
|
var xhr = this.pendingRequests[xhrId].xhr; |
|
delete this.pendingRequests[xhrId]; |
|
xhr.abort(); |
|
} |
|
}; |
|
|
|
return NetworkManager; |
|
})(); |
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreNetwork = {})); |
|
} |
|
}(this, function (exports) { |
|
exports.NetworkManager = NetworkManager; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsSharedGlobal = {})); |
|
} |
|
}(this, function (exports) { |
|
|
|
var globalScope = (typeof window !== 'undefined') ? window : |
|
(typeof global !== 'undefined') ? global : |
|
(typeof self !== 'undefined') ? self : this; |
|
|
|
var isWorker = (typeof window === 'undefined'); |
|
|
|
// The global PDFJS object exposes the API |
|
// In production, it will be declared outside a global wrapper |
|
// In development, it will be declared here |
|
if (!globalScope.PDFJS) { |
|
globalScope.PDFJS = {}; |
|
} |
|
|
|
if (typeof pdfjsVersion !== 'undefined') { |
|
globalScope.PDFJS.version = pdfjsVersion; |
|
} |
|
if (typeof pdfjsVersion !== 'undefined') { |
|
globalScope.PDFJS.build = pdfjsBuild; |
|
} |
|
|
|
globalScope.PDFJS.pdfBug = false; |
|
|
|
exports.globalScope = globalScope; |
|
exports.isWorker = isWorker; |
|
exports.PDFJS = globalScope.PDFJS; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreBidi = {}), root.pdfjsSharedGlobal); |
|
} |
|
}(this, function (exports, sharedGlobal) { |
|
|
|
var PDFJS = sharedGlobal.PDFJS; |
|
|
|
var bidi = PDFJS.bidi = (function bidiClosure() { |
|
// Character types for symbols from 0000 to 00FF. |
|
var baseTypes = [ |
|
'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'S', 'B', 'S', 'WS', |
|
'B', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', |
|
'BN', 'BN', 'B', 'B', 'B', 'S', 'WS', 'ON', 'ON', 'ET', 'ET', 'ET', 'ON', |
|
'ON', 'ON', 'ON', 'ON', 'ON', 'CS', 'ON', 'CS', 'ON', 'EN', 'EN', 'EN', |
|
'EN', 'EN', 'EN', 'EN', 'EN', 'EN', 'EN', 'ON', 'ON', 'ON', 'ON', 'ON', |
|
'ON', 'ON', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', |
|
'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'ON', 'ON', |
|
'ON', 'ON', 'ON', 'ON', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', |
|
'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', |
|
'L', 'ON', 'ON', 'ON', 'ON', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'B', 'BN', |
|
'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', |
|
'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', 'BN', |
|
'BN', 'CS', 'ON', 'ET', 'ET', 'ET', 'ET', 'ON', 'ON', 'ON', 'ON', 'L', 'ON', |
|
'ON', 'ON', 'ON', 'ON', 'ET', 'ET', 'EN', 'EN', 'ON', 'L', 'ON', 'ON', 'ON', |
|
'EN', 'L', 'ON', 'ON', 'ON', 'ON', 'ON', 'L', 'L', 'L', 'L', 'L', 'L', 'L', |
|
'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', |
|
'L', 'ON', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', |
|
'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L', |
|
'L', 'L', 'L', 'ON', 'L', 'L', 'L', 'L', 'L', 'L', 'L', 'L' |
|
]; |
|
|
|
// Character types for symbols from 0600 to 06FF |
|
var arabicTypes = [ |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'CS', 'AL', 'ON', 'ON', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', |
|
'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AN', 'AN', 'AN', 'AN', 'AN', 'AN', 'AN', 'AN', 'AN', |
|
'AN', 'ET', 'AN', 'AN', 'AL', 'AL', 'AL', 'NSM', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', |
|
'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'NSM', 'ON', 'NSM', |
|
'NSM', 'NSM', 'NSM', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', |
|
'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL', 'AL' |
|
]; |
|
|
|
function isOdd(i) { |
|
return (i & 1) !== 0; |
|
} |
|
|
|
function isEven(i) { |
|
return (i & 1) === 0; |
|
} |
|
|
|
function findUnequal(arr, start, value) { |
|
for (var j = start, jj = arr.length; j < jj; ++j) { |
|
if (arr[j] !== value) { |
|
return j; |
|
} |
|
} |
|
return j; |
|
} |
|
|
|
function setValues(arr, start, end, value) { |
|
for (var j = start; j < end; ++j) { |
|
arr[j] = value; |
|
} |
|
} |
|
|
|
function reverseValues(arr, start, end) { |
|
for (var i = start, j = end - 1; i < j; ++i, --j) { |
|
var temp = arr[i]; |
|
arr[i] = arr[j]; |
|
arr[j] = temp; |
|
} |
|
} |
|
|
|
function createBidiText(str, isLTR, vertical) { |
|
return { |
|
str: str, |
|
dir: (vertical ? 'ttb' : (isLTR ? 'ltr' : 'rtl')) |
|
}; |
|
} |
|
|
|
// These are used in bidi(), which is called frequently. We re-use them on |
|
// each call to avoid unnecessary allocations. |
|
var chars = []; |
|
var types = []; |
|
|
|
function bidi(str, startLevel, vertical) { |
|
var isLTR = true; |
|
var strLength = str.length; |
|
if (strLength === 0 || vertical) { |
|
return createBidiText(str, isLTR, vertical); |
|
} |
|
|
|
// Get types and fill arrays |
|
chars.length = strLength; |
|
types.length = strLength; |
|
var numBidi = 0; |
|
|
|
var i, ii; |
|
for (i = 0; i < strLength; ++i) { |
|
chars[i] = str.charAt(i); |
|
|
|
var charCode = str.charCodeAt(i); |
|
var charType = 'L'; |
|
if (charCode <= 0x00ff) { |
|
charType = baseTypes[charCode]; |
|
} else if (0x0590 <= charCode && charCode <= 0x05f4) { |
|
charType = 'R'; |
|
} else if (0x0600 <= charCode && charCode <= 0x06ff) { |
|
charType = arabicTypes[charCode & 0xff]; |
|
} else if (0x0700 <= charCode && charCode <= 0x08AC) { |
|
charType = 'AL'; |
|
} |
|
if (charType === 'R' || charType === 'AL' || charType === 'AN') { |
|
numBidi++; |
|
} |
|
types[i] = charType; |
|
} |
|
|
|
// Detect the bidi method |
|
// - If there are no rtl characters then no bidi needed |
|
// - If less than 30% chars are rtl then string is primarily ltr |
|
// - If more than 30% chars are rtl then string is primarily rtl |
|
if (numBidi === 0) { |
|
isLTR = true; |
|
return createBidiText(str, isLTR); |
|
} |
|
|
|
if (startLevel === -1) { |
|
if ((strLength / numBidi) < 0.3) { |
|
isLTR = true; |
|
startLevel = 0; |
|
} else { |
|
isLTR = false; |
|
startLevel = 1; |
|
} |
|
} |
|
|
|
var levels = []; |
|
for (i = 0; i < strLength; ++i) { |
|
levels[i] = startLevel; |
|
} |
|
|
|
/* |
|
X1-X10: skip most of this, since we are NOT doing the embeddings. |
|
*/ |
|
var e = (isOdd(startLevel) ? 'R' : 'L'); |
|
var sor = e; |
|
var eor = sor; |
|
|
|
/* |
|
W1. Examine each non-spacing mark (NSM) in the level run, and change the |
|
type of the NSM to the type of the previous character. If the NSM is at the |
|
start of the level run, it will get the type of sor. |
|
*/ |
|
var lastType = sor; |
|
for (i = 0; i < strLength; ++i) { |
|
if (types[i] === 'NSM') { |
|
types[i] = lastType; |
|
} else { |
|
lastType = types[i]; |
|
} |
|
} |
|
|
|
/* |
|
W2. Search backwards from each instance of a European number until the |
|
first strong type (R, L, AL, or sor) is found. If an AL is found, change |
|
the type of the European number to Arabic number. |
|
*/ |
|
lastType = sor; |
|
var t; |
|
for (i = 0; i < strLength; ++i) { |
|
t = types[i]; |
|
if (t === 'EN') { |
|
types[i] = (lastType === 'AL') ? 'AN' : 'EN'; |
|
} else if (t === 'R' || t === 'L' || t === 'AL') { |
|
lastType = t; |
|
} |
|
} |
|
|
|
/* |
|
W3. Change all ALs to R. |
|
*/ |
|
for (i = 0; i < strLength; ++i) { |
|
t = types[i]; |
|
if (t === 'AL') { |
|
types[i] = 'R'; |
|
} |
|
} |
|
|
|
/* |
|
W4. A single European separator between two European numbers changes to a |
|
European number. A single common separator between two numbers of the same |
|
type changes to that type: |
|
*/ |
|
for (i = 1; i < strLength - 1; ++i) { |
|
if (types[i] === 'ES' && types[i - 1] === 'EN' && types[i + 1] === 'EN') { |
|
types[i] = 'EN'; |
|
} |
|
if (types[i] === 'CS' && |
|
(types[i - 1] === 'EN' || types[i - 1] === 'AN') && |
|
types[i + 1] === types[i - 1]) { |
|
types[i] = types[i - 1]; |
|
} |
|
} |
|
|
|
/* |
|
W5. A sequence of European terminators adjacent to European numbers changes |
|
to all European numbers: |
|
*/ |
|
for (i = 0; i < strLength; ++i) { |
|
if (types[i] === 'EN') { |
|
// do before |
|
var j; |
|
for (j = i - 1; j >= 0; --j) { |
|
if (types[j] !== 'ET') { |
|
break; |
|
} |
|
types[j] = 'EN'; |
|
} |
|
// do after |
|
for (j = i + 1; j < strLength; --j) { |
|
if (types[j] !== 'ET') { |
|
break; |
|
} |
|
types[j] = 'EN'; |
|
} |
|
} |
|
} |
|
|
|
/* |
|
W6. Otherwise, separators and terminators change to Other Neutral: |
|
*/ |
|
for (i = 0; i < strLength; ++i) { |
|
t = types[i]; |
|
if (t === 'WS' || t === 'ES' || t === 'ET' || t === 'CS') { |
|
types[i] = 'ON'; |
|
} |
|
} |
|
|
|
/* |
|
W7. Search backwards from each instance of a European number until the |
|
first strong type (R, L, or sor) is found. If an L is found, then change |
|
the type of the European number to L. |
|
*/ |
|
lastType = sor; |
|
for (i = 0; i < strLength; ++i) { |
|
t = types[i]; |
|
if (t === 'EN') { |
|
types[i] = ((lastType === 'L') ? 'L' : 'EN'); |
|
} else if (t === 'R' || t === 'L') { |
|
lastType = t; |
|
} |
|
} |
|
|
|
/* |
|
N1. A sequence of neutrals takes the direction of the surrounding strong |
|
text if the text on both sides has the same direction. European and Arabic |
|
numbers are treated as though they were R. Start-of-level-run (sor) and |
|
end-of-level-run (eor) are used at level run boundaries. |
|
*/ |
|
for (i = 0; i < strLength; ++i) { |
|
if (types[i] === 'ON') { |
|
var end = findUnequal(types, i + 1, 'ON'); |
|
var before = sor; |
|
if (i > 0) { |
|
before = types[i - 1]; |
|
} |
|
|
|
var after = eor; |
|
if (end + 1 < strLength) { |
|
after = types[end + 1]; |
|
} |
|
if (before !== 'L') { |
|
before = 'R'; |
|
} |
|
if (after !== 'L') { |
|
after = 'R'; |
|
} |
|
if (before === after) { |
|
setValues(types, i, end, before); |
|
} |
|
i = end - 1; // reset to end (-1 so next iteration is ok) |
|
} |
|
} |
|
|
|
/* |
|
N2. Any remaining neutrals take the embedding direction. |
|
*/ |
|
for (i = 0; i < strLength; ++i) { |
|
if (types[i] === 'ON') { |
|
types[i] = e; |
|
} |
|
} |
|
|
|
/* |
|
I1. For all characters with an even (left-to-right) embedding direction, |
|
those of type R go up one level and those of type AN or EN go up two |
|
levels. |
|
I2. For all characters with an odd (right-to-left) embedding direction, |
|
those of type L, EN or AN go up one level. |
|
*/ |
|
for (i = 0; i < strLength; ++i) { |
|
t = types[i]; |
|
if (isEven(levels[i])) { |
|
if (t === 'R') { |
|
levels[i] += 1; |
|
} else if (t === 'AN' || t === 'EN') { |
|
levels[i] += 2; |
|
} |
|
} else { // isOdd |
|
if (t === 'L' || t === 'AN' || t === 'EN') { |
|
levels[i] += 1; |
|
} |
|
} |
|
} |
|
|
|
/* |
|
L1. On each line, reset the embedding level of the following characters to |
|
the paragraph embedding level: |
|
|
|
segment separators, |
|
paragraph separators, |
|
any sequence of whitespace characters preceding a segment separator or |
|
paragraph separator, and any sequence of white space characters at the end |
|
of the line. |
|
*/ |
|
|
|
// don't bother as text is only single line |
|
|
|
/* |
|
L2. From the highest level found in the text to the lowest odd level on |
|
each line, reverse any contiguous sequence of characters that are at that |
|
level or higher. |
|
*/ |
|
|
|
// find highest level & lowest odd level |
|
var highestLevel = -1; |
|
var lowestOddLevel = 99; |
|
var level; |
|
for (i = 0, ii = levels.length; i < ii; ++i) { |
|
level = levels[i]; |
|
if (highestLevel < level) { |
|
highestLevel = level; |
|
} |
|
if (lowestOddLevel > level && isOdd(level)) { |
|
lowestOddLevel = level; |
|
} |
|
} |
|
|
|
// now reverse between those limits |
|
for (level = highestLevel; level >= lowestOddLevel; --level) { |
|
// find segments to reverse |
|
var start = -1; |
|
for (i = 0, ii = levels.length; i < ii; ++i) { |
|
if (levels[i] < level) { |
|
if (start >= 0) { |
|
reverseValues(chars, start, i); |
|
start = -1; |
|
} |
|
} else if (start < 0) { |
|
start = i; |
|
} |
|
} |
|
if (start >= 0) { |
|
reverseValues(chars, start, levels.length); |
|
} |
|
} |
|
|
|
/* |
|
L3. Combining marks applied to a right-to-left base character will at this |
|
point precede their base character. If the rendering engine expects them to |
|
follow the base characters in the final display process, then the ordering |
|
of the marks and the base character must be reversed. |
|
*/ |
|
|
|
// don't bother for now |
|
|
|
/* |
|
L4. A character that possesses the mirrored property as specified by |
|
Section 4.7, Mirrored, must be depicted by a mirrored glyph if the resolved |
|
directionality of that character is R. |
|
*/ |
|
|
|
// don't mirror as characters are already mirrored in the pdf |
|
|
|
// Finally, return string |
|
for (i = 0, ii = chars.length; i < ii; ++i) { |
|
var ch = chars[i]; |
|
if (ch === '<' || ch === '>') { |
|
chars[i] = ''; |
|
} |
|
} |
|
return createBidiText(chars.join(''), isLTR); |
|
} |
|
|
|
return bidi; |
|
})(); |
|
|
|
exports.bidi = bidi; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsSharedUtil = {}), root.pdfjsSharedGlobal); |
|
} |
|
}(this, function (exports, sharedGlobal) { |
|
|
|
var PDFJS = sharedGlobal.PDFJS; |
|
var globalScope = sharedGlobal.globalScope; |
|
|
|
var FONT_IDENTITY_MATRIX = [0.001, 0, 0, 0.001, 0, 0]; |
|
|
|
var TextRenderingMode = { |
|
FILL: 0, |
|
STROKE: 1, |
|
FILL_STROKE: 2, |
|
INVISIBLE: 3, |
|
FILL_ADD_TO_PATH: 4, |
|
STROKE_ADD_TO_PATH: 5, |
|
FILL_STROKE_ADD_TO_PATH: 6, |
|
ADD_TO_PATH: 7, |
|
FILL_STROKE_MASK: 3, |
|
ADD_TO_PATH_FLAG: 4 |
|
}; |
|
|
|
var ImageKind = { |
|
GRAYSCALE_1BPP: 1, |
|
RGB_24BPP: 2, |
|
RGBA_32BPP: 3 |
|
}; |
|
|
|
var AnnotationType = { |
|
TEXT: 1, |
|
LINK: 2, |
|
FREETEXT: 3, |
|
LINE: 4, |
|
SQUARE: 5, |
|
CIRCLE: 6, |
|
POLYGON: 7, |
|
POLYLINE: 8, |
|
HIGHLIGHT: 9, |
|
UNDERLINE: 10, |
|
SQUIGGLY: 11, |
|
STRIKEOUT: 12, |
|
STAMP: 13, |
|
CARET: 14, |
|
INK: 15, |
|
POPUP: 16, |
|
FILEATTACHMENT: 17, |
|
SOUND: 18, |
|
MOVIE: 19, |
|
WIDGET: 20, |
|
SCREEN: 21, |
|
PRINTERMARK: 22, |
|
TRAPNET: 23, |
|
WATERMARK: 24, |
|
THREED: 25, |
|
REDACT: 26 |
|
}; |
|
|
|
var AnnotationFlag = { |
|
INVISIBLE: 0x01, |
|
HIDDEN: 0x02, |
|
PRINT: 0x04, |
|
NOZOOM: 0x08, |
|
NOROTATE: 0x10, |
|
NOVIEW: 0x20, |
|
READONLY: 0x40, |
|
LOCKED: 0x80, |
|
TOGGLENOVIEW: 0x100, |
|
LOCKEDCONTENTS: 0x200 |
|
}; |
|
|
|
var AnnotationBorderStyleType = { |
|
SOLID: 1, |
|
DASHED: 2, |
|
BEVELED: 3, |
|
INSET: 4, |
|
UNDERLINE: 5 |
|
}; |
|
|
|
var StreamType = { |
|
UNKNOWN: 0, |
|
FLATE: 1, |
|
LZW: 2, |
|
DCT: 3, |
|
JPX: 4, |
|
JBIG: 5, |
|
A85: 6, |
|
AHX: 7, |
|
CCF: 8, |
|
RL: 9 |
|
}; |
|
|
|
var FontType = { |
|
UNKNOWN: 0, |
|
TYPE1: 1, |
|
TYPE1C: 2, |
|
CIDFONTTYPE0: 3, |
|
CIDFONTTYPE0C: 4, |
|
TRUETYPE: 5, |
|
CIDFONTTYPE2: 6, |
|
TYPE3: 7, |
|
OPENTYPE: 8, |
|
TYPE0: 9, |
|
MMTYPE1: 10 |
|
}; |
|
|
|
PDFJS.VERBOSITY_LEVELS = { |
|
errors: 0, |
|
warnings: 1, |
|
infos: 5 |
|
}; |
|
|
|
// All the possible operations for an operator list. |
|
var OPS = PDFJS.OPS = { |
|
// Intentionally start from 1 so it is easy to spot bad operators that will be |
|
// 0's. |
|
dependency: 1, |
|
setLineWidth: 2, |
|
setLineCap: 3, |
|
setLineJoin: 4, |
|
setMiterLimit: 5, |
|
setDash: 6, |
|
setRenderingIntent: 7, |
|
setFlatness: 8, |
|
setGState: 9, |
|
save: 10, |
|
restore: 11, |
|
transform: 12, |
|
moveTo: 13, |
|
lineTo: 14, |
|
curveTo: 15, |
|
curveTo2: 16, |
|
curveTo3: 17, |
|
closePath: 18, |
|
rectangle: 19, |
|
stroke: 20, |
|
closeStroke: 21, |
|
fill: 22, |
|
eoFill: 23, |
|
fillStroke: 24, |
|
eoFillStroke: 25, |
|
closeFillStroke: 26, |
|
closeEOFillStroke: 27, |
|
endPath: 28, |
|
clip: 29, |
|
eoClip: 30, |
|
beginText: 31, |
|
endText: 32, |
|
setCharSpacing: 33, |
|
setWordSpacing: 34, |
|
setHScale: 35, |
|
setLeading: 36, |
|
setFont: 37, |
|
setTextRenderingMode: 38, |
|
setTextRise: 39, |
|
moveText: 40, |
|
setLeadingMoveText: 41, |
|
setTextMatrix: 42, |
|
nextLine: 43, |
|
showText: 44, |
|
showSpacedText: 45, |
|
nextLineShowText: 46, |
|
nextLineSetSpacingShowText: 47, |
|
setCharWidth: 48, |
|
setCharWidthAndBounds: 49, |
|
setStrokeColorSpace: 50, |
|
setFillColorSpace: 51, |
|
setStrokeColor: 52, |
|
setStrokeColorN: 53, |
|
setFillColor: 54, |
|
setFillColorN: 55, |
|
setStrokeGray: 56, |
|
setFillGray: 57, |
|
setStrokeRGBColor: 58, |
|
setFillRGBColor: 59, |
|
setStrokeCMYKColor: 60, |
|
setFillCMYKColor: 61, |
|
shadingFill: 62, |
|
beginInlineImage: 63, |
|
beginImageData: 64, |
|
endInlineImage: 65, |
|
paintXObject: 66, |
|
markPoint: 67, |
|
markPointProps: 68, |
|
beginMarkedContent: 69, |
|
beginMarkedContentProps: 70, |
|
endMarkedContent: 71, |
|
beginCompat: 72, |
|
endCompat: 73, |
|
paintFormXObjectBegin: 74, |
|
paintFormXObjectEnd: 75, |
|
beginGroup: 76, |
|
endGroup: 77, |
|
beginAnnotations: 78, |
|
endAnnotations: 79, |
|
beginAnnotation: 80, |
|
endAnnotation: 81, |
|
paintJpegXObject: 82, |
|
paintImageMaskXObject: 83, |
|
paintImageMaskXObjectGroup: 84, |
|
paintImageXObject: 85, |
|
paintInlineImageXObject: 86, |
|
paintInlineImageXObjectGroup: 87, |
|
paintImageXObjectRepeat: 88, |
|
paintImageMaskXObjectRepeat: 89, |
|
paintSolidColorImageMask: 90, |
|
constructPath: 91 |
|
}; |
|
|
|
// A notice for devs. These are good for things that are helpful to devs, such |
|
// as warning that Workers were disabled, which is important to devs but not |
|
// end users. |
|
function info(msg) { |
|
if (PDFJS.verbosity >= PDFJS.VERBOSITY_LEVELS.infos) { |
|
console.log('Info: ' + msg); |
|
} |
|
} |
|
|
|
// Non-fatal warnings. |
|
function warn(msg) { |
|
if (PDFJS.verbosity >= PDFJS.VERBOSITY_LEVELS.warnings) { |
|
console.log('Warning: ' + msg); |
|
} |
|
} |
|
|
|
// Deprecated API function -- treated as warnings. |
|
function deprecated(details) { |
|
warn('Deprecated API usage: ' + details); |
|
} |
|
|
|
// Fatal errors that should trigger the fallback UI and halt execution by |
|
// throwing an exception. |
|
function error(msg) { |
|
if (PDFJS.verbosity >= PDFJS.VERBOSITY_LEVELS.errors) { |
|
console.log('Error: ' + msg); |
|
console.log(backtrace()); |
|
} |
|
throw new Error(msg); |
|
} |
|
|
|
function backtrace() { |
|
try { |
|
throw new Error(); |
|
} catch (e) { |
|
return e.stack ? e.stack.split('\n').slice(2).join('\n') : ''; |
|
} |
|
} |
|
|
|
function assert(cond, msg) { |
|
if (!cond) { |
|
error(msg); |
|
} |
|
} |
|
|
|
var UNSUPPORTED_FEATURES = PDFJS.UNSUPPORTED_FEATURES = { |
|
unknown: 'unknown', |
|
forms: 'forms', |
|
javaScript: 'javaScript', |
|
smask: 'smask', |
|
shadingPattern: 'shadingPattern', |
|
font: 'font' |
|
}; |
|
|
|
// Combines two URLs. The baseUrl shall be absolute URL. If the url is an |
|
// absolute URL, it will be returned as is. |
|
function combineUrl(baseUrl, url) { |
|
if (!url) { |
|
return baseUrl; |
|
} |
|
return new URL(url, baseUrl).href; |
|
} |
|
|
|
// Validates if URL is safe and allowed, e.g. to avoid XSS. |
|
function isValidUrl(url, allowRelative) { |
|
if (!url) { |
|
return false; |
|
} |
|
// RFC 3986 (http://tools.ietf.org/html/rfc3986#section-3.1) |
|
// scheme = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." ) |
|
var protocol = /^[a-z][a-z0-9+\-.]*(?=:)/i.exec(url); |
|
if (!protocol) { |
|
return allowRelative; |
|
} |
|
protocol = protocol[0].toLowerCase(); |
|
switch (protocol) { |
|
case 'http': |
|
case 'https': |
|
case 'ftp': |
|
case 'mailto': |
|
case 'tel': |
|
return true; |
|
default: |
|
return false; |
|
} |
|
} |
|
PDFJS.isValidUrl = isValidUrl; |
|
|
|
/** |
|
* Adds various attributes (href, title, target, rel) to hyperlinks. |
|
* @param {HTMLLinkElement} link - The link element. |
|
* @param {Object} params - An object with the properties: |
|
* @param {string} params.url - An absolute URL. |
|
*/ |
|
function addLinkAttributes(link, params) { |
|
var url = params && params.url; |
|
link.href = link.title = (url ? removeNullCharacters(url) : ''); |
|
|
|
if (url) { |
|
if (isExternalLinkTargetSet()) { |
|
link.target = LinkTargetStringMap[PDFJS.externalLinkTarget]; |
|
} |
|
// Strip referrer from the URL. |
|
link.rel = PDFJS.externalLinkRel; |
|
} |
|
} |
|
PDFJS.addLinkAttributes = addLinkAttributes; |
|
|
|
function shadow(obj, prop, value) { |
|
Object.defineProperty(obj, prop, { value: value, |
|
enumerable: true, |
|
configurable: true, |
|
writable: false }); |
|
return value; |
|
} |
|
PDFJS.shadow = shadow; |
|
|
|
var LinkTarget = PDFJS.LinkTarget = { |
|
NONE: 0, // Default value. |
|
SELF: 1, |
|
BLANK: 2, |
|
PARENT: 3, |
|
TOP: 4, |
|
}; |
|
var LinkTargetStringMap = [ |
|
'', |
|
'_self', |
|
'_blank', |
|
'_parent', |
|
'_top' |
|
]; |
|
|
|
function isExternalLinkTargetSet() { |
|
if (PDFJS.openExternalLinksInNewWindow) { |
|
deprecated('PDFJS.openExternalLinksInNewWindow, please use ' + |
|
'"PDFJS.externalLinkTarget = PDFJS.LinkTarget.BLANK" instead.'); |
|
if (PDFJS.externalLinkTarget === LinkTarget.NONE) { |
|
PDFJS.externalLinkTarget = LinkTarget.BLANK; |
|
} |
|
// Reset the deprecated parameter, to suppress further warnings. |
|
PDFJS.openExternalLinksInNewWindow = false; |
|
} |
|
switch (PDFJS.externalLinkTarget) { |
|
case LinkTarget.NONE: |
|
return false; |
|
case LinkTarget.SELF: |
|
case LinkTarget.BLANK: |
|
case LinkTarget.PARENT: |
|
case LinkTarget.TOP: |
|
return true; |
|
} |
|
warn('PDFJS.externalLinkTarget is invalid: ' + PDFJS.externalLinkTarget); |
|
// Reset the external link target, to suppress further warnings. |
|
PDFJS.externalLinkTarget = LinkTarget.NONE; |
|
return false; |
|
} |
|
PDFJS.isExternalLinkTargetSet = isExternalLinkTargetSet; |
|
|
|
var PasswordResponses = PDFJS.PasswordResponses = { |
|
NEED_PASSWORD: 1, |
|
INCORRECT_PASSWORD: 2 |
|
}; |
|
|
|
var PasswordException = (function PasswordExceptionClosure() { |
|
function PasswordException(msg, code) { |
|
this.name = 'PasswordException'; |
|
this.message = msg; |
|
this.code = code; |
|
} |
|
|
|
PasswordException.prototype = new Error(); |
|
PasswordException.constructor = PasswordException; |
|
|
|
return PasswordException; |
|
})(); |
|
PDFJS.PasswordException = PasswordException; |
|
|
|
var UnknownErrorException = (function UnknownErrorExceptionClosure() { |
|
function UnknownErrorException(msg, details) { |
|
this.name = 'UnknownErrorException'; |
|
this.message = msg; |
|
this.details = details; |
|
} |
|
|
|
UnknownErrorException.prototype = new Error(); |
|
UnknownErrorException.constructor = UnknownErrorException; |
|
|
|
return UnknownErrorException; |
|
})(); |
|
PDFJS.UnknownErrorException = UnknownErrorException; |
|
|
|
var InvalidPDFException = (function InvalidPDFExceptionClosure() { |
|
function InvalidPDFException(msg) { |
|
this.name = 'InvalidPDFException'; |
|
this.message = msg; |
|
} |
|
|
|
InvalidPDFException.prototype = new Error(); |
|
InvalidPDFException.constructor = InvalidPDFException; |
|
|
|
return InvalidPDFException; |
|
})(); |
|
PDFJS.InvalidPDFException = InvalidPDFException; |
|
|
|
var MissingPDFException = (function MissingPDFExceptionClosure() { |
|
function MissingPDFException(msg) { |
|
this.name = 'MissingPDFException'; |
|
this.message = msg; |
|
} |
|
|
|
MissingPDFException.prototype = new Error(); |
|
MissingPDFException.constructor = MissingPDFException; |
|
|
|
return MissingPDFException; |
|
})(); |
|
PDFJS.MissingPDFException = MissingPDFException; |
|
|
|
var UnexpectedResponseException = |
|
(function UnexpectedResponseExceptionClosure() { |
|
function UnexpectedResponseException(msg, status) { |
|
this.name = 'UnexpectedResponseException'; |
|
this.message = msg; |
|
this.status = status; |
|
} |
|
|
|
UnexpectedResponseException.prototype = new Error(); |
|
UnexpectedResponseException.constructor = UnexpectedResponseException; |
|
|
|
return UnexpectedResponseException; |
|
})(); |
|
PDFJS.UnexpectedResponseException = UnexpectedResponseException; |
|
|
|
var NotImplementedException = (function NotImplementedExceptionClosure() { |
|
function NotImplementedException(msg) { |
|
this.message = msg; |
|
} |
|
|
|
NotImplementedException.prototype = new Error(); |
|
NotImplementedException.prototype.name = 'NotImplementedException'; |
|
NotImplementedException.constructor = NotImplementedException; |
|
|
|
return NotImplementedException; |
|
})(); |
|
|
|
var MissingDataException = (function MissingDataExceptionClosure() { |
|
function MissingDataException(begin, end) { |
|
this.begin = begin; |
|
this.end = end; |
|
this.message = 'Missing data [' + begin + ', ' + end + ')'; |
|
} |
|
|
|
MissingDataException.prototype = new Error(); |
|
MissingDataException.prototype.name = 'MissingDataException'; |
|
MissingDataException.constructor = MissingDataException; |
|
|
|
return MissingDataException; |
|
})(); |
|
|
|
var XRefParseException = (function XRefParseExceptionClosure() { |
|
function XRefParseException(msg) { |
|
this.message = msg; |
|
} |
|
|
|
XRefParseException.prototype = new Error(); |
|
XRefParseException.prototype.name = 'XRefParseException'; |
|
XRefParseException.constructor = XRefParseException; |
|
|
|
return XRefParseException; |
|
})(); |
|
|
|
var NullCharactersRegExp = /\x00/g; |
|
|
|
function removeNullCharacters(str) { |
|
if (typeof str !== 'string') { |
|
warn('The argument for removeNullCharacters must be a string.'); |
|
return str; |
|
} |
|
return str.replace(NullCharactersRegExp, ''); |
|
} |
|
PDFJS.removeNullCharacters = removeNullCharacters; |
|
|
|
function bytesToString(bytes) { |
|
assert(bytes !== null && typeof bytes === 'object' && |
|
bytes.length !== undefined, 'Invalid argument for bytesToString'); |
|
var length = bytes.length; |
|
var MAX_ARGUMENT_COUNT = 8192; |
|
if (length < MAX_ARGUMENT_COUNT) { |
|
return String.fromCharCode.apply(null, bytes); |
|
} |
|
var strBuf = []; |
|
for (var i = 0; i < length; i += MAX_ARGUMENT_COUNT) { |
|
var chunkEnd = Math.min(i + MAX_ARGUMENT_COUNT, length); |
|
var chunk = bytes.subarray(i, chunkEnd); |
|
strBuf.push(String.fromCharCode.apply(null, chunk)); |
|
} |
|
return strBuf.join(''); |
|
} |
|
|
|
function stringToBytes(str) { |
|
assert(typeof str === 'string', 'Invalid argument for stringToBytes'); |
|
var length = str.length; |
|
var bytes = new Uint8Array(length); |
|
for (var i = 0; i < length; ++i) { |
|
bytes[i] = str.charCodeAt(i) & 0xFF; |
|
} |
|
return bytes; |
|
} |
|
|
|
function string32(value) { |
|
return String.fromCharCode((value >> 24) & 0xff, (value >> 16) & 0xff, |
|
(value >> 8) & 0xff, value & 0xff); |
|
} |
|
|
|
function log2(x) { |
|
var n = 1, i = 0; |
|
while (x > n) { |
|
n <<= 1; |
|
i++; |
|
} |
|
return i; |
|
} |
|
|
|
function readInt8(data, start) { |
|
return (data[start] << 24) >> 24; |
|
} |
|
|
|
function readUint16(data, offset) { |
|
return (data[offset] << 8) | data[offset + 1]; |
|
} |
|
|
|
function readUint32(data, offset) { |
|
return ((data[offset] << 24) | (data[offset + 1] << 16) | |
|
(data[offset + 2] << 8) | data[offset + 3]) >>> 0; |
|
} |
|
|
|
// Lazy test the endianness of the platform |
|
// NOTE: This will be 'true' for simulated TypedArrays |
|
function isLittleEndian() { |
|
var buffer8 = new Uint8Array(2); |
|
buffer8[0] = 1; |
|
var buffer16 = new Uint16Array(buffer8.buffer); |
|
return (buffer16[0] === 1); |
|
} |
|
|
|
Object.defineProperty(PDFJS, 'isLittleEndian', { |
|
configurable: true, |
|
get: function PDFJS_isLittleEndian() { |
|
return shadow(PDFJS, 'isLittleEndian', isLittleEndian()); |
|
} |
|
}); |
|
|
|
// Lazy test if the userAgent support CanvasTypedArrays |
|
function hasCanvasTypedArrays() { |
|
var canvas = document.createElement('canvas'); |
|
canvas.width = canvas.height = 1; |
|
var ctx = canvas.getContext('2d'); |
|
var imageData = ctx.createImageData(1, 1); |
|
return (typeof imageData.data.buffer !== 'undefined'); |
|
} |
|
|
|
Object.defineProperty(PDFJS, 'hasCanvasTypedArrays', { |
|
configurable: true, |
|
get: function PDFJS_hasCanvasTypedArrays() { |
|
return shadow(PDFJS, 'hasCanvasTypedArrays', hasCanvasTypedArrays()); |
|
} |
|
}); |
|
|
|
var Uint32ArrayView = (function Uint32ArrayViewClosure() { |
|
|
|
function Uint32ArrayView(buffer, length) { |
|
this.buffer = buffer; |
|
this.byteLength = buffer.length; |
|
this.length = length === undefined ? (this.byteLength >> 2) : length; |
|
ensureUint32ArrayViewProps(this.length); |
|
} |
|
Uint32ArrayView.prototype = Object.create(null); |
|
|
|
var uint32ArrayViewSetters = 0; |
|
function createUint32ArrayProp(index) { |
|
return { |
|
get: function () { |
|
var buffer = this.buffer, offset = index << 2; |
|
return (buffer[offset] | (buffer[offset + 1] << 8) | |
|
(buffer[offset + 2] << 16) | (buffer[offset + 3] << 24)) >>> 0; |
|
}, |
|
set: function (value) { |
|
var buffer = this.buffer, offset = index << 2; |
|
buffer[offset] = value & 255; |
|
buffer[offset + 1] = (value >> 8) & 255; |
|
buffer[offset + 2] = (value >> 16) & 255; |
|
buffer[offset + 3] = (value >>> 24) & 255; |
|
} |
|
}; |
|
} |
|
|
|
function ensureUint32ArrayViewProps(length) { |
|
while (uint32ArrayViewSetters < length) { |
|
Object.defineProperty(Uint32ArrayView.prototype, |
|
uint32ArrayViewSetters, |
|
createUint32ArrayProp(uint32ArrayViewSetters)); |
|
uint32ArrayViewSetters++; |
|
} |
|
} |
|
|
|
return Uint32ArrayView; |
|
})(); |
|
|
|
exports.Uint32ArrayView = Uint32ArrayView; |
|
|
|
var IDENTITY_MATRIX = [1, 0, 0, 1, 0, 0]; |
|
|
|
var Util = PDFJS.Util = (function UtilClosure() { |
|
function Util() {} |
|
|
|
var rgbBuf = ['rgb(', 0, ',', 0, ',', 0, ')']; |
|
|
|
// makeCssRgb() can be called thousands of times. Using |rgbBuf| avoids |
|
// creating many intermediate strings. |
|
Util.makeCssRgb = function Util_makeCssRgb(r, g, b) { |
|
rgbBuf[1] = r; |
|
rgbBuf[3] = g; |
|
rgbBuf[5] = b; |
|
return rgbBuf.join(''); |
|
}; |
|
|
|
// Concatenates two transformation matrices together and returns the result. |
|
Util.transform = function Util_transform(m1, m2) { |
|
return [ |
|
m1[0] * m2[0] + m1[2] * m2[1], |
|
m1[1] * m2[0] + m1[3] * m2[1], |
|
m1[0] * m2[2] + m1[2] * m2[3], |
|
m1[1] * m2[2] + m1[3] * m2[3], |
|
m1[0] * m2[4] + m1[2] * m2[5] + m1[4], |
|
m1[1] * m2[4] + m1[3] * m2[5] + m1[5] |
|
]; |
|
}; |
|
|
|
// For 2d affine transforms |
|
Util.applyTransform = function Util_applyTransform(p, m) { |
|
var xt = p[0] * m[0] + p[1] * m[2] + m[4]; |
|
var yt = p[0] * m[1] + p[1] * m[3] + m[5]; |
|
return [xt, yt]; |
|
}; |
|
|
|
Util.applyInverseTransform = function Util_applyInverseTransform(p, m) { |
|
var d = m[0] * m[3] - m[1] * m[2]; |
|
var xt = (p[0] * m[3] - p[1] * m[2] + m[2] * m[5] - m[4] * m[3]) / d; |
|
var yt = (-p[0] * m[1] + p[1] * m[0] + m[4] * m[1] - m[5] * m[0]) / d; |
|
return [xt, yt]; |
|
}; |
|
|
|
// Applies the transform to the rectangle and finds the minimum axially |
|
// aligned bounding box. |
|
Util.getAxialAlignedBoundingBox = |
|
function Util_getAxialAlignedBoundingBox(r, m) { |
|
|
|
var p1 = Util.applyTransform(r, m); |
|
var p2 = Util.applyTransform(r.slice(2, 4), m); |
|
var p3 = Util.applyTransform([r[0], r[3]], m); |
|
var p4 = Util.applyTransform([r[2], r[1]], m); |
|
return [ |
|
Math.min(p1[0], p2[0], p3[0], p4[0]), |
|
Math.min(p1[1], p2[1], p3[1], p4[1]), |
|
Math.max(p1[0], p2[0], p3[0], p4[0]), |
|
Math.max(p1[1], p2[1], p3[1], p4[1]) |
|
]; |
|
}; |
|
|
|
Util.inverseTransform = function Util_inverseTransform(m) { |
|
var d = m[0] * m[3] - m[1] * m[2]; |
|
return [m[3] / d, -m[1] / d, -m[2] / d, m[0] / d, |
|
(m[2] * m[5] - m[4] * m[3]) / d, (m[4] * m[1] - m[5] * m[0]) / d]; |
|
}; |
|
|
|
// Apply a generic 3d matrix M on a 3-vector v: |
|
// | a b c | | X | |
|
// | d e f | x | Y | |
|
// | g h i | | Z | |
|
// M is assumed to be serialized as [a,b,c,d,e,f,g,h,i], |
|
// with v as [X,Y,Z] |
|
Util.apply3dTransform = function Util_apply3dTransform(m, v) { |
|
return [ |
|
m[0] * v[0] + m[1] * v[1] + m[2] * v[2], |
|
m[3] * v[0] + m[4] * v[1] + m[5] * v[2], |
|
m[6] * v[0] + m[7] * v[1] + m[8] * v[2] |
|
]; |
|
}; |
|
|
|
// This calculation uses Singular Value Decomposition. |
|
// The SVD can be represented with formula A = USV. We are interested in the |
|
// matrix S here because it represents the scale values. |
|
Util.singularValueDecompose2dScale = |
|
function Util_singularValueDecompose2dScale(m) { |
|
|
|
var transpose = [m[0], m[2], m[1], m[3]]; |
|
|
|
// Multiply matrix m with its transpose. |
|
var a = m[0] * transpose[0] + m[1] * transpose[2]; |
|
var b = m[0] * transpose[1] + m[1] * transpose[3]; |
|
var c = m[2] * transpose[0] + m[3] * transpose[2]; |
|
var d = m[2] * transpose[1] + m[3] * transpose[3]; |
|
|
|
// Solve the second degree polynomial to get roots. |
|
var first = (a + d) / 2; |
|
var second = Math.sqrt((a + d) * (a + d) - 4 * (a * d - c * b)) / 2; |
|
var sx = first + second || 1; |
|
var sy = first - second || 1; |
|
|
|
// Scale values are the square roots of the eigenvalues. |
|
return [Math.sqrt(sx), Math.sqrt(sy)]; |
|
}; |
|
|
|
// Normalize rectangle rect=[x1, y1, x2, y2] so that (x1,y1) < (x2,y2) |
|
// For coordinate systems whose origin lies in the bottom-left, this |
|
// means normalization to (BL,TR) ordering. For systems with origin in the |
|
// top-left, this means (TL,BR) ordering. |
|
Util.normalizeRect = function Util_normalizeRect(rect) { |
|
var r = rect.slice(0); // clone rect |
|
if (rect[0] > rect[2]) { |
|
r[0] = rect[2]; |
|
r[2] = rect[0]; |
|
} |
|
if (rect[1] > rect[3]) { |
|
r[1] = rect[3]; |
|
r[3] = rect[1]; |
|
} |
|
return r; |
|
}; |
|
|
|
// Returns a rectangle [x1, y1, x2, y2] corresponding to the |
|
// intersection of rect1 and rect2. If no intersection, returns 'false' |
|
// The rectangle coordinates of rect1, rect2 should be [x1, y1, x2, y2] |
|
Util.intersect = function Util_intersect(rect1, rect2) { |
|
function compare(a, b) { |
|
return a - b; |
|
} |
|
|
|
// Order points along the axes |
|
var orderedX = [rect1[0], rect1[2], rect2[0], rect2[2]].sort(compare), |
|
orderedY = [rect1[1], rect1[3], rect2[1], rect2[3]].sort(compare), |
|
result = []; |
|
|
|
rect1 = Util.normalizeRect(rect1); |
|
rect2 = Util.normalizeRect(rect2); |
|
|
|
// X: first and second points belong to different rectangles? |
|
if ((orderedX[0] === rect1[0] && orderedX[1] === rect2[0]) || |
|
(orderedX[0] === rect2[0] && orderedX[1] === rect1[0])) { |
|
// Intersection must be between second and third points |
|
result[0] = orderedX[1]; |
|
result[2] = orderedX[2]; |
|
} else { |
|
return false; |
|
} |
|
|
|
// Y: first and second points belong to different rectangles? |
|
if ((orderedY[0] === rect1[1] && orderedY[1] === rect2[1]) || |
|
(orderedY[0] === rect2[1] && orderedY[1] === rect1[1])) { |
|
// Intersection must be between second and third points |
|
result[1] = orderedY[1]; |
|
result[3] = orderedY[2]; |
|
} else { |
|
return false; |
|
} |
|
|
|
return result; |
|
}; |
|
|
|
Util.sign = function Util_sign(num) { |
|
return num < 0 ? -1 : 1; |
|
}; |
|
|
|
var ROMAN_NUMBER_MAP = [ |
|
'', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', |
|
'', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', |
|
'', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX' |
|
]; |
|
/** |
|
* Converts positive integers to (upper case) Roman numerals. |
|
* @param {integer} number - The number that should be converted. |
|
* @param {boolean} lowerCase - Indicates if the result should be converted |
|
* to lower case letters. The default is false. |
|
* @return {string} The resulting Roman number. |
|
*/ |
|
Util.toRoman = function Util_toRoman(number, lowerCase) { |
|
assert(isInt(number) && number > 0, |
|
'The number should be a positive integer.'); |
|
var pos, romanBuf = []; |
|
// Thousands |
|
while (number >= 1000) { |
|
number -= 1000; |
|
romanBuf.push('M'); |
|
} |
|
// Hundreds |
|
pos = (number / 100) | 0; |
|
number %= 100; |
|
romanBuf.push(ROMAN_NUMBER_MAP[pos]); |
|
// Tens |
|
pos = (number / 10) | 0; |
|
number %= 10; |
|
romanBuf.push(ROMAN_NUMBER_MAP[10 + pos]); |
|
// Ones |
|
romanBuf.push(ROMAN_NUMBER_MAP[20 + number]); |
|
|
|
var romanStr = romanBuf.join(''); |
|
return (lowerCase ? romanStr.toLowerCase() : romanStr); |
|
}; |
|
|
|
Util.appendToArray = function Util_appendToArray(arr1, arr2) { |
|
Array.prototype.push.apply(arr1, arr2); |
|
}; |
|
|
|
Util.prependToArray = function Util_prependToArray(arr1, arr2) { |
|
Array.prototype.unshift.apply(arr1, arr2); |
|
}; |
|
|
|
Util.extendObj = function extendObj(obj1, obj2) { |
|
for (var key in obj2) { |
|
obj1[key] = obj2[key]; |
|
} |
|
}; |
|
|
|
Util.getInheritableProperty = function Util_getInheritableProperty(dict, |
|
name) { |
|
while (dict && !dict.has(name)) { |
|
dict = dict.get('Parent'); |
|
} |
|
if (!dict) { |
|
return null; |
|
} |
|
return dict.get(name); |
|
}; |
|
|
|
Util.inherit = function Util_inherit(sub, base, prototype) { |
|
sub.prototype = Object.create(base.prototype); |
|
sub.prototype.constructor = sub; |
|
for (var prop in prototype) { |
|
sub.prototype[prop] = prototype[prop]; |
|
} |
|
}; |
|
|
|
Util.loadScript = function Util_loadScript(src, callback) { |
|
var script = document.createElement('script'); |
|
var loaded = false; |
|
script.setAttribute('src', src); |
|
if (callback) { |
|
script.onload = function() { |
|
if (!loaded) { |
|
callback(); |
|
} |
|
loaded = true; |
|
}; |
|
} |
|
document.getElementsByTagName('head')[0].appendChild(script); |
|
}; |
|
|
|
return Util; |
|
})(); |
|
|
|
/** |
|
* PDF page viewport created based on scale, rotation and offset. |
|
* @class |
|
* @alias PDFJS.PageViewport |
|
*/ |
|
var PageViewport = PDFJS.PageViewport = (function PageViewportClosure() { |
|
/** |
|
* @constructor |
|
* @private |
|
* @param viewBox {Array} xMin, yMin, xMax and yMax coordinates. |
|
* @param scale {number} scale of the viewport. |
|
* @param rotation {number} rotations of the viewport in degrees. |
|
* @param offsetX {number} offset X |
|
* @param offsetY {number} offset Y |
|
* @param dontFlip {boolean} if true, axis Y will not be flipped. |
|
*/ |
|
function PageViewport(viewBox, scale, rotation, offsetX, offsetY, dontFlip) { |
|
this.viewBox = viewBox; |
|
this.scale = scale; |
|
this.rotation = rotation; |
|
this.offsetX = offsetX; |
|
this.offsetY = offsetY; |
|
|
|
// creating transform to convert pdf coordinate system to the normal |
|
// canvas like coordinates taking in account scale and rotation |
|
var centerX = (viewBox[2] + viewBox[0]) / 2; |
|
var centerY = (viewBox[3] + viewBox[1]) / 2; |
|
var rotateA, rotateB, rotateC, rotateD; |
|
rotation = rotation % 360; |
|
rotation = rotation < 0 ? rotation + 360 : rotation; |
|
switch (rotation) { |
|
case 180: |
|
rotateA = -1; rotateB = 0; rotateC = 0; rotateD = 1; |
|
break; |
|
case 90: |
|
rotateA = 0; rotateB = 1; rotateC = 1; rotateD = 0; |
|
break; |
|
case 270: |
|
rotateA = 0; rotateB = -1; rotateC = -1; rotateD = 0; |
|
break; |
|
//case 0: |
|
default: |
|
rotateA = 1; rotateB = 0; rotateC = 0; rotateD = -1; |
|
break; |
|
} |
|
|
|
if (dontFlip) { |
|
rotateC = -rotateC; rotateD = -rotateD; |
|
} |
|
|
|
var offsetCanvasX, offsetCanvasY; |
|
var width, height; |
|
if (rotateA === 0) { |
|
offsetCanvasX = Math.abs(centerY - viewBox[1]) * scale + offsetX; |
|
offsetCanvasY = Math.abs(centerX - viewBox[0]) * scale + offsetY; |
|
width = Math.abs(viewBox[3] - viewBox[1]) * scale; |
|
height = Math.abs(viewBox[2] - viewBox[0]) * scale; |
|
} else { |
|
offsetCanvasX = Math.abs(centerX - viewBox[0]) * scale + offsetX; |
|
offsetCanvasY = Math.abs(centerY - viewBox[1]) * scale + offsetY; |
|
width = Math.abs(viewBox[2] - viewBox[0]) * scale; |
|
height = Math.abs(viewBox[3] - viewBox[1]) * scale; |
|
} |
|
// creating transform for the following operations: |
|
// translate(-centerX, -centerY), rotate and flip vertically, |
|
// scale, and translate(offsetCanvasX, offsetCanvasY) |
|
this.transform = [ |
|
rotateA * scale, |
|
rotateB * scale, |
|
rotateC * scale, |
|
rotateD * scale, |
|
offsetCanvasX - rotateA * scale * centerX - rotateC * scale * centerY, |
|
offsetCanvasY - rotateB * scale * centerX - rotateD * scale * centerY |
|
]; |
|
|
|
this.width = width; |
|
this.height = height; |
|
this.fontScale = scale; |
|
} |
|
PageViewport.prototype = /** @lends PDFJS.PageViewport.prototype */ { |
|
/** |
|
* Clones viewport with additional properties. |
|
* @param args {Object} (optional) If specified, may contain the 'scale' or |
|
* 'rotation' properties to override the corresponding properties in |
|
* the cloned viewport. |
|
* @returns {PDFJS.PageViewport} Cloned viewport. |
|
*/ |
|
clone: function PageViewPort_clone(args) { |
|
args = args || {}; |
|
var scale = 'scale' in args ? args.scale : this.scale; |
|
var rotation = 'rotation' in args ? args.rotation : this.rotation; |
|
return new PageViewport(this.viewBox.slice(), scale, rotation, |
|
this.offsetX, this.offsetY, args.dontFlip); |
|
}, |
|
/** |
|
* Converts PDF point to the viewport coordinates. For examples, useful for |
|
* converting PDF location into canvas pixel coordinates. |
|
* @param x {number} X coordinate. |
|
* @param y {number} Y coordinate. |
|
* @returns {Object} Object that contains 'x' and 'y' properties of the |
|
* point in the viewport coordinate space. |
|
* @see {@link convertToPdfPoint} |
|
* @see {@link convertToViewportRectangle} |
|
*/ |
|
convertToViewportPoint: function PageViewport_convertToViewportPoint(x, y) { |
|
return Util.applyTransform([x, y], this.transform); |
|
}, |
|
/** |
|
* Converts PDF rectangle to the viewport coordinates. |
|
* @param rect {Array} xMin, yMin, xMax and yMax coordinates. |
|
* @returns {Array} Contains corresponding coordinates of the rectangle |
|
* in the viewport coordinate space. |
|
* @see {@link convertToViewportPoint} |
|
*/ |
|
convertToViewportRectangle: |
|
function PageViewport_convertToViewportRectangle(rect) { |
|
var tl = Util.applyTransform([rect[0], rect[1]], this.transform); |
|
var br = Util.applyTransform([rect[2], rect[3]], this.transform); |
|
return [tl[0], tl[1], br[0], br[1]]; |
|
}, |
|
/** |
|
* Converts viewport coordinates to the PDF location. For examples, useful |
|
* for converting canvas pixel location into PDF one. |
|
* @param x {number} X coordinate. |
|
* @param y {number} Y coordinate. |
|
* @returns {Object} Object that contains 'x' and 'y' properties of the |
|
* point in the PDF coordinate space. |
|
* @see {@link convertToViewportPoint} |
|
*/ |
|
convertToPdfPoint: function PageViewport_convertToPdfPoint(x, y) { |
|
return Util.applyInverseTransform([x, y], this.transform); |
|
} |
|
}; |
|
return PageViewport; |
|
})(); |
|
|
|
var PDFStringTranslateTable = [ |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
|
0x2D8, 0x2C7, 0x2C6, 0x2D9, 0x2DD, 0x2DB, 0x2DA, 0x2DC, 0, 0, 0, 0, 0, 0, 0, |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0x2022, 0x2020, 0x2021, 0x2026, 0x2014, |
|
0x2013, 0x192, 0x2044, 0x2039, 0x203A, 0x2212, 0x2030, 0x201E, 0x201C, |
|
0x201D, 0x2018, 0x2019, 0x201A, 0x2122, 0xFB01, 0xFB02, 0x141, 0x152, 0x160, |
|
0x178, 0x17D, 0x131, 0x142, 0x153, 0x161, 0x17E, 0, 0x20AC |
|
]; |
|
|
|
function stringToPDFString(str) { |
|
var i, n = str.length, strBuf = []; |
|
if (str[0] === '\xFE' && str[1] === '\xFF') { |
|
// UTF16BE BOM |
|
for (i = 2; i < n; i += 2) { |
|
strBuf.push(String.fromCharCode( |
|
(str.charCodeAt(i) << 8) | str.charCodeAt(i + 1))); |
|
} |
|
} else { |
|
for (i = 0; i < n; ++i) { |
|
var code = PDFStringTranslateTable[str.charCodeAt(i)]; |
|
strBuf.push(code ? String.fromCharCode(code) : str.charAt(i)); |
|
} |
|
} |
|
return strBuf.join(''); |
|
} |
|
|
|
function stringToUTF8String(str) { |
|
return decodeURIComponent(escape(str)); |
|
} |
|
|
|
function utf8StringToString(str) { |
|
return unescape(encodeURIComponent(str)); |
|
} |
|
|
|
function isEmptyObj(obj) { |
|
for (var key in obj) { |
|
return false; |
|
} |
|
return true; |
|
} |
|
|
|
function isBool(v) { |
|
return typeof v === 'boolean'; |
|
} |
|
|
|
function isInt(v) { |
|
return typeof v === 'number' && ((v | 0) === v); |
|
} |
|
|
|
function isNum(v) { |
|
return typeof v === 'number'; |
|
} |
|
|
|
function isString(v) { |
|
return typeof v === 'string'; |
|
} |
|
|
|
function isArray(v) { |
|
return v instanceof Array; |
|
} |
|
|
|
function isArrayBuffer(v) { |
|
return typeof v === 'object' && v !== null && v.byteLength !== undefined; |
|
} |
|
|
|
/** |
|
* Promise Capability object. |
|
* |
|
* @typedef {Object} PromiseCapability |
|
* @property {Promise} promise - A promise object. |
|
* @property {function} resolve - Fullfills the promise. |
|
* @property {function} reject - Rejects the promise. |
|
*/ |
|
|
|
/** |
|
* Creates a promise capability object. |
|
* @alias PDFJS.createPromiseCapability |
|
* |
|
* @return {PromiseCapability} A capability object contains: |
|
* - a Promise, resolve and reject methods. |
|
*/ |
|
function createPromiseCapability() { |
|
var capability = {}; |
|
capability.promise = new Promise(function (resolve, reject) { |
|
capability.resolve = resolve; |
|
capability.reject = reject; |
|
}); |
|
return capability; |
|
} |
|
|
|
PDFJS.createPromiseCapability = createPromiseCapability; |
|
|
|
/** |
|
* Polyfill for Promises: |
|
* The following promise implementation tries to generally implement the |
|
* Promise/A+ spec. Some notable differences from other promise libaries are: |
|
* - There currently isn't a seperate deferred and promise object. |
|
* - Unhandled rejections eventually show an error if they aren't handled. |
|
* |
|
* Based off of the work in: |
|
* https://bugzilla.mozilla.org/show_bug.cgi?id=810490 |
|
*/ |
|
(function PromiseClosure() { |
|
if (globalScope.Promise) { |
|
// Promises existing in the DOM/Worker, checking presence of all/resolve |
|
if (typeof globalScope.Promise.all !== 'function') { |
|
globalScope.Promise.all = function (iterable) { |
|
var count = 0, results = [], resolve, reject; |
|
var promise = new globalScope.Promise(function (resolve_, reject_) { |
|
resolve = resolve_; |
|
reject = reject_; |
|
}); |
|
iterable.forEach(function (p, i) { |
|
count++; |
|
p.then(function (result) { |
|
results[i] = result; |
|
count--; |
|
if (count === 0) { |
|
resolve(results); |
|
} |
|
}, reject); |
|
}); |
|
if (count === 0) { |
|
resolve(results); |
|
} |
|
return promise; |
|
}; |
|
} |
|
if (typeof globalScope.Promise.resolve !== 'function') { |
|
globalScope.Promise.resolve = function (value) { |
|
return new globalScope.Promise(function (resolve) { resolve(value); }); |
|
}; |
|
} |
|
if (typeof globalScope.Promise.reject !== 'function') { |
|
globalScope.Promise.reject = function (reason) { |
|
return new globalScope.Promise(function (resolve, reject) { |
|
reject(reason); |
|
}); |
|
}; |
|
} |
|
if (typeof globalScope.Promise.prototype.catch !== 'function') { |
|
globalScope.Promise.prototype.catch = function (onReject) { |
|
return globalScope.Promise.prototype.then(undefined, onReject); |
|
}; |
|
} |
|
return; |
|
} |
|
var STATUS_PENDING = 0; |
|
var STATUS_RESOLVED = 1; |
|
var STATUS_REJECTED = 2; |
|
|
|
// In an attempt to avoid silent exceptions, unhandled rejections are |
|
// tracked and if they aren't handled in a certain amount of time an |
|
// error is logged. |
|
var REJECTION_TIMEOUT = 500; |
|
|
|
var HandlerManager = { |
|
handlers: [], |
|
running: false, |
|
unhandledRejections: [], |
|
pendingRejectionCheck: false, |
|
|
|
scheduleHandlers: function scheduleHandlers(promise) { |
|
if (promise._status === STATUS_PENDING) { |
|
return; |
|
} |
|
|
|
this.handlers = this.handlers.concat(promise._handlers); |
|
promise._handlers = []; |
|
|
|
if (this.running) { |
|
return; |
|
} |
|
this.running = true; |
|
|
|
setTimeout(this.runHandlers.bind(this), 0); |
|
}, |
|
|
|
runHandlers: function runHandlers() { |
|
var RUN_TIMEOUT = 1; // ms |
|
var timeoutAt = Date.now() + RUN_TIMEOUT; |
|
while (this.handlers.length > 0) { |
|
var handler = this.handlers.shift(); |
|
|
|
var nextStatus = handler.thisPromise._status; |
|
var nextValue = handler.thisPromise._value; |
|
|
|
try { |
|
if (nextStatus === STATUS_RESOLVED) { |
|
if (typeof handler.onResolve === 'function') { |
|
nextValue = handler.onResolve(nextValue); |
|
} |
|
} else if (typeof handler.onReject === 'function') { |
|
nextValue = handler.onReject(nextValue); |
|
nextStatus = STATUS_RESOLVED; |
|
|
|
if (handler.thisPromise._unhandledRejection) { |
|
this.removeUnhandeledRejection(handler.thisPromise); |
|
} |
|
} |
|
} catch (ex) { |
|
nextStatus = STATUS_REJECTED; |
|
nextValue = ex; |
|
} |
|
|
|
handler.nextPromise._updateStatus(nextStatus, nextValue); |
|
if (Date.now() >= timeoutAt) { |
|
break; |
|
} |
|
} |
|
|
|
if (this.handlers.length > 0) { |
|
setTimeout(this.runHandlers.bind(this), 0); |
|
return; |
|
} |
|
|
|
this.running = false; |
|
}, |
|
|
|
addUnhandledRejection: function addUnhandledRejection(promise) { |
|
this.unhandledRejections.push({ |
|
promise: promise, |
|
time: Date.now() |
|
}); |
|
this.scheduleRejectionCheck(); |
|
}, |
|
|
|
removeUnhandeledRejection: function removeUnhandeledRejection(promise) { |
|
promise._unhandledRejection = false; |
|
for (var i = 0; i < this.unhandledRejections.length; i++) { |
|
if (this.unhandledRejections[i].promise === promise) { |
|
this.unhandledRejections.splice(i); |
|
i--; |
|
} |
|
} |
|
}, |
|
|
|
scheduleRejectionCheck: function scheduleRejectionCheck() { |
|
if (this.pendingRejectionCheck) { |
|
return; |
|
} |
|
this.pendingRejectionCheck = true; |
|
setTimeout(function rejectionCheck() { |
|
this.pendingRejectionCheck = false; |
|
var now = Date.now(); |
|
for (var i = 0; i < this.unhandledRejections.length; i++) { |
|
if (now - this.unhandledRejections[i].time > REJECTION_TIMEOUT) { |
|
var unhandled = this.unhandledRejections[i].promise._value; |
|
var msg = 'Unhandled rejection: ' + unhandled; |
|
if (unhandled.stack) { |
|
msg += '\n' + unhandled.stack; |
|
} |
|
warn(msg); |
|
this.unhandledRejections.splice(i); |
|
i--; |
|
} |
|
} |
|
if (this.unhandledRejections.length) { |
|
this.scheduleRejectionCheck(); |
|
} |
|
}.bind(this), REJECTION_TIMEOUT); |
|
} |
|
}; |
|
|
|
function Promise(resolver) { |
|
this._status = STATUS_PENDING; |
|
this._handlers = []; |
|
try { |
|
resolver.call(this, this._resolve.bind(this), this._reject.bind(this)); |
|
} catch (e) { |
|
this._reject(e); |
|
} |
|
} |
|
/** |
|
* Builds a promise that is resolved when all the passed in promises are |
|
* resolved. |
|
* @param {array} array of data and/or promises to wait for. |
|
* @return {Promise} New dependant promise. |
|
*/ |
|
Promise.all = function Promise_all(promises) { |
|
var resolveAll, rejectAll; |
|
var deferred = new Promise(function (resolve, reject) { |
|
resolveAll = resolve; |
|
rejectAll = reject; |
|
}); |
|
var unresolved = promises.length; |
|
var results = []; |
|
if (unresolved === 0) { |
|
resolveAll(results); |
|
return deferred; |
|
} |
|
function reject(reason) { |
|
if (deferred._status === STATUS_REJECTED) { |
|
return; |
|
} |
|
results = []; |
|
rejectAll(reason); |
|
} |
|
for (var i = 0, ii = promises.length; i < ii; ++i) { |
|
var promise = promises[i]; |
|
var resolve = (function(i) { |
|
return function(value) { |
|
if (deferred._status === STATUS_REJECTED) { |
|
return; |
|
} |
|
results[i] = value; |
|
unresolved--; |
|
if (unresolved === 0) { |
|
resolveAll(results); |
|
} |
|
}; |
|
})(i); |
|
if (Promise.isPromise(promise)) { |
|
promise.then(resolve, reject); |
|
} else { |
|
resolve(promise); |
|
} |
|
} |
|
return deferred; |
|
}; |
|
|
|
/** |
|
* Checks if the value is likely a promise (has a 'then' function). |
|
* @return {boolean} true if value is thenable |
|
*/ |
|
Promise.isPromise = function Promise_isPromise(value) { |
|
return value && typeof value.then === 'function'; |
|
}; |
|
|
|
/** |
|
* Creates resolved promise |
|
* @param value resolve value |
|
* @returns {Promise} |
|
*/ |
|
Promise.resolve = function Promise_resolve(value) { |
|
return new Promise(function (resolve) { resolve(value); }); |
|
}; |
|
|
|
/** |
|
* Creates rejected promise |
|
* @param reason rejection value |
|
* @returns {Promise} |
|
*/ |
|
Promise.reject = function Promise_reject(reason) { |
|
return new Promise(function (resolve, reject) { reject(reason); }); |
|
}; |
|
|
|
Promise.prototype = { |
|
_status: null, |
|
_value: null, |
|
_handlers: null, |
|
_unhandledRejection: null, |
|
|
|
_updateStatus: function Promise__updateStatus(status, value) { |
|
if (this._status === STATUS_RESOLVED || |
|
this._status === STATUS_REJECTED) { |
|
return; |
|
} |
|
|
|
if (status === STATUS_RESOLVED && |
|
Promise.isPromise(value)) { |
|
value.then(this._updateStatus.bind(this, STATUS_RESOLVED), |
|
this._updateStatus.bind(this, STATUS_REJECTED)); |
|
return; |
|
} |
|
|
|
this._status = status; |
|
this._value = value; |
|
|
|
if (status === STATUS_REJECTED && this._handlers.length === 0) { |
|
this._unhandledRejection = true; |
|
HandlerManager.addUnhandledRejection(this); |
|
} |
|
|
|
HandlerManager.scheduleHandlers(this); |
|
}, |
|
|
|
_resolve: function Promise_resolve(value) { |
|
this._updateStatus(STATUS_RESOLVED, value); |
|
}, |
|
|
|
_reject: function Promise_reject(reason) { |
|
this._updateStatus(STATUS_REJECTED, reason); |
|
}, |
|
|
|
then: function Promise_then(onResolve, onReject) { |
|
var nextPromise = new Promise(function (resolve, reject) { |
|
this.resolve = resolve; |
|
this.reject = reject; |
|
}); |
|
this._handlers.push({ |
|
thisPromise: this, |
|
onResolve: onResolve, |
|
onReject: onReject, |
|
nextPromise: nextPromise |
|
}); |
|
HandlerManager.scheduleHandlers(this); |
|
return nextPromise; |
|
}, |
|
|
|
catch: function Promise_catch(onReject) { |
|
return this.then(undefined, onReject); |
|
} |
|
}; |
|
|
|
globalScope.Promise = Promise; |
|
})(); |
|
|
|
var StatTimer = (function StatTimerClosure() { |
|
function rpad(str, pad, length) { |
|
while (str.length < length) { |
|
str += pad; |
|
} |
|
return str; |
|
} |
|
function StatTimer() { |
|
this.started = {}; |
|
this.times = []; |
|
this.enabled = true; |
|
} |
|
StatTimer.prototype = { |
|
time: function StatTimer_time(name) { |
|
if (!this.enabled) { |
|
return; |
|
} |
|
if (name in this.started) { |
|
warn('Timer is already running for ' + name); |
|
} |
|
this.started[name] = Date.now(); |
|
}, |
|
timeEnd: function StatTimer_timeEnd(name) { |
|
if (!this.enabled) { |
|
return; |
|
} |
|
if (!(name in this.started)) { |
|
warn('Timer has not been started for ' + name); |
|
} |
|
this.times.push({ |
|
'name': name, |
|
'start': this.started[name], |
|
'end': Date.now() |
|
}); |
|
// Remove timer from started so it can be called again. |
|
delete this.started[name]; |
|
}, |
|
toString: function StatTimer_toString() { |
|
var i, ii; |
|
var times = this.times; |
|
var out = ''; |
|
// Find the longest name for padding purposes. |
|
var longest = 0; |
|
for (i = 0, ii = times.length; i < ii; ++i) { |
|
var name = times[i]['name']; |
|
if (name.length > longest) { |
|
longest = name.length; |
|
} |
|
} |
|
for (i = 0, ii = times.length; i < ii; ++i) { |
|
var span = times[i]; |
|
var duration = span.end - span.start; |
|
out += rpad(span['name'], ' ', longest) + ' ' + duration + 'ms\n'; |
|
} |
|
return out; |
|
} |
|
}; |
|
return StatTimer; |
|
})(); |
|
|
|
PDFJS.createBlob = function createBlob(data, contentType) { |
|
if (typeof Blob !== 'undefined') { |
|
return new Blob([data], { type: contentType }); |
|
} |
|
// Blob builder is deprecated in FF14 and removed in FF18. |
|
var bb = new MozBlobBuilder(); |
|
bb.append(data); |
|
return bb.getBlob(contentType); |
|
}; |
|
|
|
PDFJS.createObjectURL = (function createObjectURLClosure() { |
|
// Blob/createObjectURL is not available, falling back to data schema. |
|
var digits = |
|
'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/='; |
|
|
|
return function createObjectURL(data, contentType) { |
|
if (!PDFJS.disableCreateObjectURL && |
|
typeof URL !== 'undefined' && URL.createObjectURL) { |
|
var blob = PDFJS.createBlob(data, contentType); |
|
return URL.createObjectURL(blob); |
|
} |
|
|
|
var buffer = 'data:' + contentType + ';base64,'; |
|
for (var i = 0, ii = data.length; i < ii; i += 3) { |
|
var b1 = data[i] & 0xFF; |
|
var b2 = data[i + 1] & 0xFF; |
|
var b3 = data[i + 2] & 0xFF; |
|
var d1 = b1 >> 2, d2 = ((b1 & 3) << 4) | (b2 >> 4); |
|
var d3 = i + 1 < ii ? ((b2 & 0xF) << 2) | (b3 >> 6) : 64; |
|
var d4 = i + 2 < ii ? (b3 & 0x3F) : 64; |
|
buffer += digits[d1] + digits[d2] + digits[d3] + digits[d4]; |
|
} |
|
return buffer; |
|
}; |
|
})(); |
|
|
|
function MessageHandler(sourceName, targetName, comObj) { |
|
this.sourceName = sourceName; |
|
this.targetName = targetName; |
|
this.comObj = comObj; |
|
this.callbackIndex = 1; |
|
this.postMessageTransfers = true; |
|
var callbacksCapabilities = this.callbacksCapabilities = {}; |
|
var ah = this.actionHandler = {}; |
|
|
|
this._onComObjOnMessage = function messageHandlerComObjOnMessage(event) { |
|
var data = event.data; |
|
if (data.targetName !== this.sourceName) { |
|
return; |
|
} |
|
if (data.isReply) { |
|
var callbackId = data.callbackId; |
|
if (data.callbackId in callbacksCapabilities) { |
|
var callback = callbacksCapabilities[callbackId]; |
|
delete callbacksCapabilities[callbackId]; |
|
if ('error' in data) { |
|
callback.reject(data.error); |
|
} else { |
|
callback.resolve(data.data); |
|
} |
|
} else { |
|
error('Cannot resolve callback ' + callbackId); |
|
} |
|
} else if (data.action in ah) { |
|
var action = ah[data.action]; |
|
if (data.callbackId) { |
|
var sourceName = this.sourceName; |
|
var targetName = data.sourceName; |
|
Promise.resolve().then(function () { |
|
return action[0].call(action[1], data.data); |
|
}).then(function (result) { |
|
comObj.postMessage({ |
|
sourceName: sourceName, |
|
targetName: targetName, |
|
isReply: true, |
|
callbackId: data.callbackId, |
|
data: result |
|
}); |
|
}, function (reason) { |
|
if (reason instanceof Error) { |
|
// Serialize error to avoid "DataCloneError" |
|
reason = reason + ''; |
|
} |
|
comObj.postMessage({ |
|
sourceName: sourceName, |
|
targetName: targetName, |
|
isReply: true, |
|
callbackId: data.callbackId, |
|
error: reason |
|
}); |
|
}); |
|
} else { |
|
action[0].call(action[1], data.data); |
|
} |
|
} else { |
|
error('Unknown action from worker: ' + data.action); |
|
} |
|
}.bind(this); |
|
comObj.addEventListener('message', this._onComObjOnMessage); |
|
} |
|
|
|
MessageHandler.prototype = { |
|
on: function messageHandlerOn(actionName, handler, scope) { |
|
var ah = this.actionHandler; |
|
if (ah[actionName]) { |
|
error('There is already an actionName called "' + actionName + '"'); |
|
} |
|
ah[actionName] = [handler, scope]; |
|
}, |
|
/** |
|
* Sends a message to the comObj to invoke the action with the supplied data. |
|
* @param {String} actionName Action to call. |
|
* @param {JSON} data JSON data to send. |
|
* @param {Array} [transfers] Optional list of transfers/ArrayBuffers |
|
*/ |
|
send: function messageHandlerSend(actionName, data, transfers) { |
|
var message = { |
|
sourceName: this.sourceName, |
|
targetName: this.targetName, |
|
action: actionName, |
|
data: data |
|
}; |
|
this.postMessage(message, transfers); |
|
}, |
|
/** |
|
* Sends a message to the comObj to invoke the action with the supplied data. |
|
* Expects that other side will callback with the response. |
|
* @param {String} actionName Action to call. |
|
* @param {JSON} data JSON data to send. |
|
* @param {Array} [transfers] Optional list of transfers/ArrayBuffers. |
|
* @returns {Promise} Promise to be resolved with response data. |
|
*/ |
|
sendWithPromise: |
|
function messageHandlerSendWithPromise(actionName, data, transfers) { |
|
var callbackId = this.callbackIndex++; |
|
var message = { |
|
sourceName: this.sourceName, |
|
targetName: this.targetName, |
|
action: actionName, |
|
data: data, |
|
callbackId: callbackId |
|
}; |
|
var capability = createPromiseCapability(); |
|
this.callbacksCapabilities[callbackId] = capability; |
|
try { |
|
this.postMessage(message, transfers); |
|
} catch (e) { |
|
capability.reject(e); |
|
} |
|
return capability.promise; |
|
}, |
|
/** |
|
* Sends raw message to the comObj. |
|
* @private |
|
* @param message {Object} Raw message. |
|
* @param transfers List of transfers/ArrayBuffers, or undefined. |
|
*/ |
|
postMessage: function (message, transfers) { |
|
if (transfers && this.postMessageTransfers) { |
|
this.comObj.postMessage(message, transfers); |
|
} else { |
|
this.comObj.postMessage(message); |
|
} |
|
}, |
|
|
|
destroy: function () { |
|
this.comObj.removeEventListener('message', this._onComObjOnMessage); |
|
} |
|
}; |
|
|
|
function loadJpegStream(id, imageUrl, objs) { |
|
var img = new Image(); |
|
img.onload = (function loadJpegStream_onloadClosure() { |
|
objs.resolve(id, img); |
|
}); |
|
img.onerror = (function loadJpegStream_onerrorClosure() { |
|
objs.resolve(id, null); |
|
warn('Error during JPEG image loading'); |
|
}); |
|
img.src = imageUrl; |
|
} |
|
|
|
// Polyfill from https://github.com/Polymer/URL |
|
/* Any copyright is dedicated to the Public Domain. |
|
* http://creativecommons.org/publicdomain/zero/1.0/ */ |
|
(function checkURLConstructor(scope) { |
|
/* jshint ignore:start */ |
|
|
|
// feature detect for URL constructor |
|
var hasWorkingUrl = false; |
|
try { |
|
if (typeof URL === 'function' && |
|
typeof URL.prototype === 'object' && |
|
('origin' in URL.prototype)) { |
|
var u = new URL('b', 'http://a'); |
|
u.pathname = 'c%20d'; |
|
hasWorkingUrl = u.href === 'http://a/c%20d'; |
|
} |
|
} catch(e) { } |
|
|
|
if (hasWorkingUrl) |
|
return; |
|
|
|
var relative = Object.create(null); |
|
relative['ftp'] = 21; |
|
relative['file'] = 0; |
|
relative['gopher'] = 70; |
|
relative['http'] = 80; |
|
relative['https'] = 443; |
|
relative['ws'] = 80; |
|
relative['wss'] = 443; |
|
|
|
var relativePathDotMapping = Object.create(null); |
|
relativePathDotMapping['%2e'] = '.'; |
|
relativePathDotMapping['.%2e'] = '..'; |
|
relativePathDotMapping['%2e.'] = '..'; |
|
relativePathDotMapping['%2e%2e'] = '..'; |
|
|
|
function isRelativeScheme(scheme) { |
|
return relative[scheme] !== undefined; |
|
} |
|
|
|
function invalid() { |
|
clear.call(this); |
|
this._isInvalid = true; |
|
} |
|
|
|
function IDNAToASCII(h) { |
|
if ('' == h) { |
|
invalid.call(this) |
|
} |
|
// XXX |
|
return h.toLowerCase() |
|
} |
|
|
|
function percentEscape(c) { |
|
var unicode = c.charCodeAt(0); |
|
if (unicode > 0x20 && |
|
unicode < 0x7F && |
|
// " # < > ? ` |
|
[0x22, 0x23, 0x3C, 0x3E, 0x3F, 0x60].indexOf(unicode) == -1 |
|
) { |
|
return c; |
|
} |
|
return encodeURIComponent(c); |
|
} |
|
|
|
function percentEscapeQuery(c) { |
|
// XXX This actually needs to encode c using encoding and then |
|
// convert the bytes one-by-one. |
|
|
|
var unicode = c.charCodeAt(0); |
|
if (unicode > 0x20 && |
|
unicode < 0x7F && |
|
// " # < > ` (do not escape '?') |
|
[0x22, 0x23, 0x3C, 0x3E, 0x60].indexOf(unicode) == -1 |
|
) { |
|
return c; |
|
} |
|
return encodeURIComponent(c); |
|
} |
|
|
|
var EOF = undefined, |
|
ALPHA = /[a-zA-Z]/, |
|
ALPHANUMERIC = /[a-zA-Z0-9\+\-\.]/; |
|
|
|
function parse(input, stateOverride, base) { |
|
function err(message) { |
|
errors.push(message) |
|
} |
|
|
|
var state = stateOverride || 'scheme start', |
|
cursor = 0, |
|
buffer = '', |
|
seenAt = false, |
|
seenBracket = false, |
|
errors = []; |
|
|
|
loop: while ((input[cursor - 1] != EOF || cursor == 0) && !this._isInvalid) { |
|
var c = input[cursor]; |
|
switch (state) { |
|
case 'scheme start': |
|
if (c && ALPHA.test(c)) { |
|
buffer += c.toLowerCase(); // ASCII-safe |
|
state = 'scheme'; |
|
} else if (!stateOverride) { |
|
buffer = ''; |
|
state = 'no scheme'; |
|
continue; |
|
} else { |
|
err('Invalid scheme.'); |
|
break loop; |
|
} |
|
break; |
|
|
|
case 'scheme': |
|
if (c && ALPHANUMERIC.test(c)) { |
|
buffer += c.toLowerCase(); // ASCII-safe |
|
} else if (':' == c) { |
|
this._scheme = buffer; |
|
buffer = ''; |
|
if (stateOverride) { |
|
break loop; |
|
} |
|
if (isRelativeScheme(this._scheme)) { |
|
this._isRelative = true; |
|
} |
|
if ('file' == this._scheme) { |
|
state = 'relative'; |
|
} else if (this._isRelative && base && base._scheme == this._scheme) { |
|
state = 'relative or authority'; |
|
} else if (this._isRelative) { |
|
state = 'authority first slash'; |
|
} else { |
|
state = 'scheme data'; |
|
} |
|
} else if (!stateOverride) { |
|
buffer = ''; |
|
cursor = 0; |
|
state = 'no scheme'; |
|
continue; |
|
} else if (EOF == c) { |
|
break loop; |
|
} else { |
|
err('Code point not allowed in scheme: ' + c) |
|
break loop; |
|
} |
|
break; |
|
|
|
case 'scheme data': |
|
if ('?' == c) { |
|
this._query = '?'; |
|
state = 'query'; |
|
} else if ('#' == c) { |
|
this._fragment = '#'; |
|
state = 'fragment'; |
|
} else { |
|
// XXX error handling |
|
if (EOF != c && '\t' != c && '\n' != c && '\r' != c) { |
|
this._schemeData += percentEscape(c); |
|
} |
|
} |
|
break; |
|
|
|
case 'no scheme': |
|
if (!base || !(isRelativeScheme(base._scheme))) { |
|
err('Missing scheme.'); |
|
invalid.call(this); |
|
} else { |
|
state = 'relative'; |
|
continue; |
|
} |
|
break; |
|
|
|
case 'relative or authority': |
|
if ('/' == c && '/' == input[cursor+1]) { |
|
state = 'authority ignore slashes'; |
|
} else { |
|
err('Expected /, got: ' + c); |
|
state = 'relative'; |
|
continue |
|
} |
|
break; |
|
|
|
case 'relative': |
|
this._isRelative = true; |
|
if ('file' != this._scheme) |
|
this._scheme = base._scheme; |
|
if (EOF == c) { |
|
this._host = base._host; |
|
this._port = base._port; |
|
this._path = base._path.slice(); |
|
this._query = base._query; |
|
this._username = base._username; |
|
this._password = base._password; |
|
break loop; |
|
} else if ('/' == c || '\\' == c) { |
|
if ('\\' == c) |
|
err('\\ is an invalid code point.'); |
|
state = 'relative slash'; |
|
} else if ('?' == c) { |
|
this._host = base._host; |
|
this._port = base._port; |
|
this._path = base._path.slice(); |
|
this._query = '?'; |
|
this._username = base._username; |
|
this._password = base._password; |
|
state = 'query'; |
|
} else if ('#' == c) { |
|
this._host = base._host; |
|
this._port = base._port; |
|
this._path = base._path.slice(); |
|
this._query = base._query; |
|
this._fragment = '#'; |
|
this._username = base._username; |
|
this._password = base._password; |
|
state = 'fragment'; |
|
} else { |
|
var nextC = input[cursor+1] |
|
var nextNextC = input[cursor+2] |
|
if ( |
|
'file' != this._scheme || !ALPHA.test(c) || |
|
(nextC != ':' && nextC != '|') || |
|
(EOF != nextNextC && '/' != nextNextC && '\\' != nextNextC && '?' != nextNextC && '#' != nextNextC)) { |
|
this._host = base._host; |
|
this._port = base._port; |
|
this._username = base._username; |
|
this._password = base._password; |
|
this._path = base._path.slice(); |
|
this._path.pop(); |
|
} |
|
state = 'relative path'; |
|
continue; |
|
} |
|
break; |
|
|
|
case 'relative slash': |
|
if ('/' == c || '\\' == c) { |
|
if ('\\' == c) { |
|
err('\\ is an invalid code point.'); |
|
} |
|
if ('file' == this._scheme) { |
|
state = 'file host'; |
|
} else { |
|
state = 'authority ignore slashes'; |
|
} |
|
} else { |
|
if ('file' != this._scheme) { |
|
this._host = base._host; |
|
this._port = base._port; |
|
this._username = base._username; |
|
this._password = base._password; |
|
} |
|
state = 'relative path'; |
|
continue; |
|
} |
|
break; |
|
|
|
case 'authority first slash': |
|
if ('/' == c) { |
|
state = 'authority second slash'; |
|
} else { |
|
err("Expected '/', got: " + c); |
|
state = 'authority ignore slashes'; |
|
continue; |
|
} |
|
break; |
|
|
|
case 'authority second slash': |
|
state = 'authority ignore slashes'; |
|
if ('/' != c) { |
|
err("Expected '/', got: " + c); |
|
continue; |
|
} |
|
break; |
|
|
|
case 'authority ignore slashes': |
|
if ('/' != c && '\\' != c) { |
|
state = 'authority'; |
|
continue; |
|
} else { |
|
err('Expected authority, got: ' + c); |
|
} |
|
break; |
|
|
|
case 'authority': |
|
if ('@' == c) { |
|
if (seenAt) { |
|
err('@ already seen.'); |
|
buffer += '%40'; |
|
} |
|
seenAt = true; |
|
for (var i = 0; i < buffer.length; i++) { |
|
var cp = buffer[i]; |
|
if ('\t' == cp || '\n' == cp || '\r' == cp) { |
|
err('Invalid whitespace in authority.'); |
|
continue; |
|
} |
|
// XXX check URL code points |
|
if (':' == cp && null === this._password) { |
|
this._password = ''; |
|
continue; |
|
} |
|
var tempC = percentEscape(cp); |
|
(null !== this._password) ? this._password += tempC : this._username += tempC; |
|
} |
|
buffer = ''; |
|
} else if (EOF == c || '/' == c || '\\' == c || '?' == c || '#' == c) { |
|
cursor -= buffer.length; |
|
buffer = ''; |
|
state = 'host'; |
|
continue; |
|
} else { |
|
buffer += c; |
|
} |
|
break; |
|
|
|
case 'file host': |
|
if (EOF == c || '/' == c || '\\' == c || '?' == c || '#' == c) { |
|
if (buffer.length == 2 && ALPHA.test(buffer[0]) && (buffer[1] == ':' || buffer[1] == '|')) { |
|
state = 'relative path'; |
|
} else if (buffer.length == 0) { |
|
state = 'relative path start'; |
|
} else { |
|
this._host = IDNAToASCII.call(this, buffer); |
|
buffer = ''; |
|
state = 'relative path start'; |
|
} |
|
continue; |
|
} else if ('\t' == c || '\n' == c || '\r' == c) { |
|
err('Invalid whitespace in file host.'); |
|
} else { |
|
buffer += c; |
|
} |
|
break; |
|
|
|
case 'host': |
|
case 'hostname': |
|
if (':' == c && !seenBracket) { |
|
// XXX host parsing |
|
this._host = IDNAToASCII.call(this, buffer); |
|
buffer = ''; |
|
state = 'port'; |
|
if ('hostname' == stateOverride) { |
|
break loop; |
|
} |
|
} else if (EOF == c || '/' == c || '\\' == c || '?' == c || '#' == c) { |
|
this._host = IDNAToASCII.call(this, buffer); |
|
buffer = ''; |
|
state = 'relative path start'; |
|
if (stateOverride) { |
|
break loop; |
|
} |
|
continue; |
|
} else if ('\t' != c && '\n' != c && '\r' != c) { |
|
if ('[' == c) { |
|
seenBracket = true; |
|
} else if (']' == c) { |
|
seenBracket = false; |
|
} |
|
buffer += c; |
|
} else { |
|
err('Invalid code point in host/hostname: ' + c); |
|
} |
|
break; |
|
|
|
case 'port': |
|
if (/[0-9]/.test(c)) { |
|
buffer += c; |
|
} else if (EOF == c || '/' == c || '\\' == c || '?' == c || '#' == c || stateOverride) { |
|
if ('' != buffer) { |
|
var temp = parseInt(buffer, 10); |
|
if (temp != relative[this._scheme]) { |
|
this._port = temp + ''; |
|
} |
|
buffer = ''; |
|
} |
|
if (stateOverride) { |
|
break loop; |
|
} |
|
state = 'relative path start'; |
|
continue; |
|
} else if ('\t' == c || '\n' == c || '\r' == c) { |
|
err('Invalid code point in port: ' + c); |
|
} else { |
|
invalid.call(this); |
|
} |
|
break; |
|
|
|
case 'relative path start': |
|
if ('\\' == c) |
|
err("'\\' not allowed in path."); |
|
state = 'relative path'; |
|
if ('/' != c && '\\' != c) { |
|
continue; |
|
} |
|
break; |
|
|
|
case 'relative path': |
|
if (EOF == c || '/' == c || '\\' == c || (!stateOverride && ('?' == c || '#' == c))) { |
|
if ('\\' == c) { |
|
err('\\ not allowed in relative path.'); |
|
} |
|
var tmp; |
|
if (tmp = relativePathDotMapping[buffer.toLowerCase()]) { |
|
buffer = tmp; |
|
} |
|
if ('..' == buffer) { |
|
this._path.pop(); |
|
if ('/' != c && '\\' != c) { |
|
this._path.push(''); |
|
} |
|
} else if ('.' == buffer && '/' != c && '\\' != c) { |
|
this._path.push(''); |
|
} else if ('.' != buffer) { |
|
if ('file' == this._scheme && this._path.length == 0 && buffer.length == 2 && ALPHA.test(buffer[0]) && buffer[1] == '|') { |
|
buffer = buffer[0] + ':'; |
|
} |
|
this._path.push(buffer); |
|
} |
|
buffer = ''; |
|
if ('?' == c) { |
|
this._query = '?'; |
|
state = 'query'; |
|
} else if ('#' == c) { |
|
this._fragment = '#'; |
|
state = 'fragment'; |
|
} |
|
} else if ('\t' != c && '\n' != c && '\r' != c) { |
|
buffer += percentEscape(c); |
|
} |
|
break; |
|
|
|
case 'query': |
|
if (!stateOverride && '#' == c) { |
|
this._fragment = '#'; |
|
state = 'fragment'; |
|
} else if (EOF != c && '\t' != c && '\n' != c && '\r' != c) { |
|
this._query += percentEscapeQuery(c); |
|
} |
|
break; |
|
|
|
case 'fragment': |
|
if (EOF != c && '\t' != c && '\n' != c && '\r' != c) { |
|
this._fragment += c; |
|
} |
|
break; |
|
} |
|
|
|
cursor++; |
|
} |
|
} |
|
|
|
function clear() { |
|
this._scheme = ''; |
|
this._schemeData = ''; |
|
this._username = ''; |
|
this._password = null; |
|
this._host = ''; |
|
this._port = ''; |
|
this._path = []; |
|
this._query = ''; |
|
this._fragment = ''; |
|
this._isInvalid = false; |
|
this._isRelative = false; |
|
} |
|
|
|
// Does not process domain names or IP addresses. |
|
// Does not handle encoding for the query parameter. |
|
function jURL(url, base /* , encoding */) { |
|
if (base !== undefined && !(base instanceof jURL)) |
|
base = new jURL(String(base)); |
|
|
|
this._url = url; |
|
clear.call(this); |
|
|
|
var input = url.replace(/^[ \t\r\n\f]+|[ \t\r\n\f]+$/g, ''); |
|
// encoding = encoding || 'utf-8' |
|
|
|
parse.call(this, input, null, base); |
|
} |
|
|
|
jURL.prototype = { |
|
toString: function() { |
|
return this.href; |
|
}, |
|
get href() { |
|
if (this._isInvalid) |
|
return this._url; |
|
|
|
var authority = ''; |
|
if ('' != this._username || null != this._password) { |
|
authority = this._username + |
|
(null != this._password ? ':' + this._password : '') + '@'; |
|
} |
|
|
|
return this.protocol + |
|
(this._isRelative ? '//' + authority + this.host : '') + |
|
this.pathname + this._query + this._fragment; |
|
}, |
|
set href(href) { |
|
clear.call(this); |
|
parse.call(this, href); |
|
}, |
|
|
|
get protocol() { |
|
return this._scheme + ':'; |
|
}, |
|
set protocol(protocol) { |
|
if (this._isInvalid) |
|
return; |
|
parse.call(this, protocol + ':', 'scheme start'); |
|
}, |
|
|
|
get host() { |
|
return this._isInvalid ? '' : this._port ? |
|
this._host + ':' + this._port : this._host; |
|
}, |
|
set host(host) { |
|
if (this._isInvalid || !this._isRelative) |
|
return; |
|
parse.call(this, host, 'host'); |
|
}, |
|
|
|
get hostname() { |
|
return this._host; |
|
}, |
|
set hostname(hostname) { |
|
if (this._isInvalid || !this._isRelative) |
|
return; |
|
parse.call(this, hostname, 'hostname'); |
|
}, |
|
|
|
get port() { |
|
return this._port; |
|
}, |
|
set port(port) { |
|
if (this._isInvalid || !this._isRelative) |
|
return; |
|
parse.call(this, port, 'port'); |
|
}, |
|
|
|
get pathname() { |
|
return this._isInvalid ? '' : this._isRelative ? |
|
'/' + this._path.join('/') : this._schemeData; |
|
}, |
|
set pathname(pathname) { |
|
if (this._isInvalid || !this._isRelative) |
|
return; |
|
this._path = []; |
|
parse.call(this, pathname, 'relative path start'); |
|
}, |
|
|
|
get search() { |
|
return this._isInvalid || !this._query || '?' == this._query ? |
|
'' : this._query; |
|
}, |
|
set search(search) { |
|
if (this._isInvalid || !this._isRelative) |
|
return; |
|
this._query = '?'; |
|
if ('?' == search[0]) |
|
search = search.slice(1); |
|
parse.call(this, search, 'query'); |
|
}, |
|
|
|
get hash() { |
|
return this._isInvalid || !this._fragment || '#' == this._fragment ? |
|
'' : this._fragment; |
|
}, |
|
set hash(hash) { |
|
if (this._isInvalid) |
|
return; |
|
this._fragment = '#'; |
|
if ('#' == hash[0]) |
|
hash = hash.slice(1); |
|
parse.call(this, hash, 'fragment'); |
|
}, |
|
|
|
get origin() { |
|
var host; |
|
if (this._isInvalid || !this._scheme) { |
|
return ''; |
|
} |
|
// javascript: Gecko returns String(""), WebKit/Blink String("null") |
|
// Gecko throws error for "data://" |
|
// data: Gecko returns "", Blink returns "data://", WebKit returns "null" |
|
// Gecko returns String("") for file: mailto: |
|
// WebKit/Blink returns String("SCHEME://") for file: mailto: |
|
switch (this._scheme) { |
|
case 'data': |
|
case 'file': |
|
case 'javascript': |
|
case 'mailto': |
|
return 'null'; |
|
} |
|
host = this.host; |
|
if (!host) { |
|
return ''; |
|
} |
|
return this._scheme + '://' + host; |
|
} |
|
}; |
|
|
|
// Copy over the static methods |
|
var OriginalURL = scope.URL; |
|
if (OriginalURL) { |
|
jURL.createObjectURL = function(blob) { |
|
// IE extension allows a second optional options argument. |
|
// http://msdn.microsoft.com/en-us/library/ie/hh772302(v=vs.85).aspx |
|
return OriginalURL.createObjectURL.apply(OriginalURL, arguments); |
|
}; |
|
jURL.revokeObjectURL = function(url) { |
|
OriginalURL.revokeObjectURL(url); |
|
}; |
|
} |
|
|
|
scope.URL = jURL; |
|
/* jshint ignore:end */ |
|
})(globalScope); |
|
|
|
exports.FONT_IDENTITY_MATRIX = FONT_IDENTITY_MATRIX; |
|
exports.IDENTITY_MATRIX = IDENTITY_MATRIX; |
|
exports.OPS = OPS; |
|
exports.UNSUPPORTED_FEATURES = UNSUPPORTED_FEATURES; |
|
exports.AnnotationBorderStyleType = AnnotationBorderStyleType; |
|
exports.AnnotationFlag = AnnotationFlag; |
|
exports.AnnotationType = AnnotationType; |
|
exports.FontType = FontType; |
|
exports.ImageKind = ImageKind; |
|
exports.InvalidPDFException = InvalidPDFException; |
|
exports.LinkTarget = LinkTarget; |
|
exports.LinkTargetStringMap = LinkTargetStringMap; |
|
exports.MessageHandler = MessageHandler; |
|
exports.MissingDataException = MissingDataException; |
|
exports.MissingPDFException = MissingPDFException; |
|
exports.NotImplementedException = NotImplementedException; |
|
exports.PasswordException = PasswordException; |
|
exports.PasswordResponses = PasswordResponses; |
|
exports.StatTimer = StatTimer; |
|
exports.StreamType = StreamType; |
|
exports.TextRenderingMode = TextRenderingMode; |
|
exports.UnexpectedResponseException = UnexpectedResponseException; |
|
exports.UnknownErrorException = UnknownErrorException; |
|
exports.Util = Util; |
|
exports.XRefParseException = XRefParseException; |
|
exports.assert = assert; |
|
exports.bytesToString = bytesToString; |
|
exports.combineUrl = combineUrl; |
|
exports.createPromiseCapability = createPromiseCapability; |
|
exports.deprecated = deprecated; |
|
exports.error = error; |
|
exports.info = info; |
|
exports.isArray = isArray; |
|
exports.isArrayBuffer = isArrayBuffer; |
|
exports.isBool = isBool; |
|
exports.isEmptyObj = isEmptyObj; |
|
exports.isExternalLinkTargetSet = isExternalLinkTargetSet; |
|
exports.isInt = isInt; |
|
exports.isNum = isNum; |
|
exports.isString = isString; |
|
exports.isValidUrl = isValidUrl; |
|
exports.addLinkAttributes = addLinkAttributes; |
|
exports.loadJpegStream = loadJpegStream; |
|
exports.log2 = log2; |
|
exports.readInt8 = readInt8; |
|
exports.readUint16 = readUint16; |
|
exports.readUint32 = readUint32; |
|
exports.removeNullCharacters = removeNullCharacters; |
|
exports.shadow = shadow; |
|
exports.string32 = string32; |
|
exports.stringToBytes = stringToBytes; |
|
exports.stringToPDFString = stringToPDFString; |
|
exports.stringToUTF8String = stringToUTF8String; |
|
exports.utf8StringToString = utf8StringToString; |
|
exports.warn = warn; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreChunkedStream = {}), root.pdfjsSharedUtil); |
|
} |
|
}(this, function (exports, sharedUtil) { |
|
|
|
var MissingDataException = sharedUtil.MissingDataException; |
|
var assert = sharedUtil.assert; |
|
var createPromiseCapability = sharedUtil.createPromiseCapability; |
|
var isInt = sharedUtil.isInt; |
|
var isEmptyObj = sharedUtil.isEmptyObj; |
|
|
|
var ChunkedStream = (function ChunkedStreamClosure() { |
|
function ChunkedStream(length, chunkSize, manager) { |
|
this.bytes = new Uint8Array(length); |
|
this.start = 0; |
|
this.pos = 0; |
|
this.end = length; |
|
this.chunkSize = chunkSize; |
|
this.loadedChunks = []; |
|
this.numChunksLoaded = 0; |
|
this.numChunks = Math.ceil(length / chunkSize); |
|
this.manager = manager; |
|
this.progressiveDataLength = 0; |
|
this.lastSuccessfulEnsureByteChunk = -1; // a single-entry cache |
|
} |
|
|
|
// required methods for a stream. if a particular stream does not |
|
// implement these, an error should be thrown |
|
ChunkedStream.prototype = { |
|
|
|
getMissingChunks: function ChunkedStream_getMissingChunks() { |
|
var chunks = []; |
|
for (var chunk = 0, n = this.numChunks; chunk < n; ++chunk) { |
|
if (!this.loadedChunks[chunk]) { |
|
chunks.push(chunk); |
|
} |
|
} |
|
return chunks; |
|
}, |
|
|
|
getBaseStreams: function ChunkedStream_getBaseStreams() { |
|
return [this]; |
|
}, |
|
|
|
allChunksLoaded: function ChunkedStream_allChunksLoaded() { |
|
return this.numChunksLoaded === this.numChunks; |
|
}, |
|
|
|
onReceiveData: function ChunkedStream_onReceiveData(begin, chunk) { |
|
var end = begin + chunk.byteLength; |
|
|
|
assert(begin % this.chunkSize === 0, 'Bad begin offset: ' + begin); |
|
// Using this.length is inaccurate here since this.start can be moved |
|
// See ChunkedStream.moveStart() |
|
var length = this.bytes.length; |
|
assert(end % this.chunkSize === 0 || end === length, |
|
'Bad end offset: ' + end); |
|
|
|
this.bytes.set(new Uint8Array(chunk), begin); |
|
var chunkSize = this.chunkSize; |
|
var beginChunk = Math.floor(begin / chunkSize); |
|
var endChunk = Math.floor((end - 1) / chunkSize) + 1; |
|
var curChunk; |
|
|
|
for (curChunk = beginChunk; curChunk < endChunk; ++curChunk) { |
|
if (!this.loadedChunks[curChunk]) { |
|
this.loadedChunks[curChunk] = true; |
|
++this.numChunksLoaded; |
|
} |
|
} |
|
}, |
|
|
|
onReceiveProgressiveData: |
|
function ChunkedStream_onReceiveProgressiveData(data) { |
|
var position = this.progressiveDataLength; |
|
var beginChunk = Math.floor(position / this.chunkSize); |
|
|
|
this.bytes.set(new Uint8Array(data), position); |
|
position += data.byteLength; |
|
this.progressiveDataLength = position; |
|
var endChunk = position >= this.end ? this.numChunks : |
|
Math.floor(position / this.chunkSize); |
|
var curChunk; |
|
for (curChunk = beginChunk; curChunk < endChunk; ++curChunk) { |
|
if (!this.loadedChunks[curChunk]) { |
|
this.loadedChunks[curChunk] = true; |
|
++this.numChunksLoaded; |
|
} |
|
} |
|
}, |
|
|
|
ensureByte: function ChunkedStream_ensureByte(pos) { |
|
var chunk = Math.floor(pos / this.chunkSize); |
|
if (chunk === this.lastSuccessfulEnsureByteChunk) { |
|
return; |
|
} |
|
|
|
if (!this.loadedChunks[chunk]) { |
|
throw new MissingDataException(pos, pos + 1); |
|
} |
|
this.lastSuccessfulEnsureByteChunk = chunk; |
|
}, |
|
|
|
ensureRange: function ChunkedStream_ensureRange(begin, end) { |
|
if (begin >= end) { |
|
return; |
|
} |
|
|
|
if (end <= this.progressiveDataLength) { |
|
return; |
|
} |
|
|
|
var chunkSize = this.chunkSize; |
|
var beginChunk = Math.floor(begin / chunkSize); |
|
var endChunk = Math.floor((end - 1) / chunkSize) + 1; |
|
for (var chunk = beginChunk; chunk < endChunk; ++chunk) { |
|
if (!this.loadedChunks[chunk]) { |
|
throw new MissingDataException(begin, end); |
|
} |
|
} |
|
}, |
|
|
|
nextEmptyChunk: function ChunkedStream_nextEmptyChunk(beginChunk) { |
|
var chunk, numChunks = this.numChunks; |
|
for (var i = 0; i < numChunks; ++i) { |
|
chunk = (beginChunk + i) % numChunks; // Wrap around to beginning |
|
if (!this.loadedChunks[chunk]) { |
|
return chunk; |
|
} |
|
} |
|
return null; |
|
}, |
|
|
|
hasChunk: function ChunkedStream_hasChunk(chunk) { |
|
return !!this.loadedChunks[chunk]; |
|
}, |
|
|
|
get length() { |
|
return this.end - this.start; |
|
}, |
|
|
|
get isEmpty() { |
|
return this.length === 0; |
|
}, |
|
|
|
getByte: function ChunkedStream_getByte() { |
|
var pos = this.pos; |
|
if (pos >= this.end) { |
|
return -1; |
|
} |
|
this.ensureByte(pos); |
|
return this.bytes[this.pos++]; |
|
}, |
|
|
|
getUint16: function ChunkedStream_getUint16() { |
|
var b0 = this.getByte(); |
|
var b1 = this.getByte(); |
|
if (b0 === -1 || b1 === -1) { |
|
return -1; |
|
} |
|
return (b0 << 8) + b1; |
|
}, |
|
|
|
getInt32: function ChunkedStream_getInt32() { |
|
var b0 = this.getByte(); |
|
var b1 = this.getByte(); |
|
var b2 = this.getByte(); |
|
var b3 = this.getByte(); |
|
return (b0 << 24) + (b1 << 16) + (b2 << 8) + b3; |
|
}, |
|
|
|
// returns subarray of original buffer |
|
// should only be read |
|
getBytes: function ChunkedStream_getBytes(length) { |
|
var bytes = this.bytes; |
|
var pos = this.pos; |
|
var strEnd = this.end; |
|
|
|
if (!length) { |
|
this.ensureRange(pos, strEnd); |
|
return bytes.subarray(pos, strEnd); |
|
} |
|
|
|
var end = pos + length; |
|
if (end > strEnd) { |
|
end = strEnd; |
|
} |
|
this.ensureRange(pos, end); |
|
|
|
this.pos = end; |
|
return bytes.subarray(pos, end); |
|
}, |
|
|
|
peekByte: function ChunkedStream_peekByte() { |
|
var peekedByte = this.getByte(); |
|
this.pos--; |
|
return peekedByte; |
|
}, |
|
|
|
peekBytes: function ChunkedStream_peekBytes(length) { |
|
var bytes = this.getBytes(length); |
|
this.pos -= bytes.length; |
|
return bytes; |
|
}, |
|
|
|
getByteRange: function ChunkedStream_getBytes(begin, end) { |
|
this.ensureRange(begin, end); |
|
return this.bytes.subarray(begin, end); |
|
}, |
|
|
|
skip: function ChunkedStream_skip(n) { |
|
if (!n) { |
|
n = 1; |
|
} |
|
this.pos += n; |
|
}, |
|
|
|
reset: function ChunkedStream_reset() { |
|
this.pos = this.start; |
|
}, |
|
|
|
moveStart: function ChunkedStream_moveStart() { |
|
this.start = this.pos; |
|
}, |
|
|
|
makeSubStream: function ChunkedStream_makeSubStream(start, length, dict) { |
|
this.ensureRange(start, start + length); |
|
|
|
function ChunkedStreamSubstream() {} |
|
ChunkedStreamSubstream.prototype = Object.create(this); |
|
ChunkedStreamSubstream.prototype.getMissingChunks = function() { |
|
var chunkSize = this.chunkSize; |
|
var beginChunk = Math.floor(this.start / chunkSize); |
|
var endChunk = Math.floor((this.end - 1) / chunkSize) + 1; |
|
var missingChunks = []; |
|
for (var chunk = beginChunk; chunk < endChunk; ++chunk) { |
|
if (!this.loadedChunks[chunk]) { |
|
missingChunks.push(chunk); |
|
} |
|
} |
|
return missingChunks; |
|
}; |
|
var subStream = new ChunkedStreamSubstream(); |
|
subStream.pos = subStream.start = start; |
|
subStream.end = start + length || this.end; |
|
subStream.dict = dict; |
|
return subStream; |
|
}, |
|
|
|
isStream: true |
|
}; |
|
|
|
return ChunkedStream; |
|
})(); |
|
|
|
var ChunkedStreamManager = (function ChunkedStreamManagerClosure() { |
|
|
|
function ChunkedStreamManager(length, chunkSize, url, args) { |
|
this.stream = new ChunkedStream(length, chunkSize, this); |
|
this.length = length; |
|
this.chunkSize = chunkSize; |
|
this.url = url; |
|
this.disableAutoFetch = args.disableAutoFetch; |
|
var msgHandler = this.msgHandler = args.msgHandler; |
|
|
|
if (args.chunkedViewerLoading) { |
|
msgHandler.on('OnDataRange', this.onReceiveData.bind(this)); |
|
msgHandler.on('OnDataProgress', this.onProgress.bind(this)); |
|
this.sendRequest = function ChunkedStreamManager_sendRequest(begin, end) { |
|
msgHandler.send('RequestDataRange', { begin: begin, end: end }); |
|
}; |
|
} else { |
|
|
|
var getXhr = function getXhr() { |
|
return new XMLHttpRequest(); |
|
}; |
|
this.networkManager = new NetworkManager(this.url, { |
|
getXhr: getXhr, |
|
httpHeaders: args.httpHeaders, |
|
withCredentials: args.withCredentials |
|
}); |
|
this.sendRequest = function ChunkedStreamManager_sendRequest(begin, end) { |
|
this.networkManager.requestRange(begin, end, { |
|
onDone: this.onReceiveData.bind(this), |
|
onProgress: this.onProgress.bind(this) |
|
}); |
|
}; |
|
} |
|
|
|
this.currRequestId = 0; |
|
|
|
this.chunksNeededByRequest = {}; |
|
this.requestsByChunk = {}; |
|
this.promisesByRequest = {}; |
|
this.progressiveDataLength = 0; |
|
|
|
this._loadedStreamCapability = createPromiseCapability(); |
|
|
|
if (args.initialData) { |
|
this.onReceiveData({chunk: args.initialData}); |
|
} |
|
} |
|
|
|
ChunkedStreamManager.prototype = { |
|
onLoadedStream: function ChunkedStreamManager_getLoadedStream() { |
|
return this._loadedStreamCapability.promise; |
|
}, |
|
|
|
// Get all the chunks that are not yet loaded and groups them into |
|
// contiguous ranges to load in as few requests as possible |
|
requestAllChunks: function ChunkedStreamManager_requestAllChunks() { |
|
var missingChunks = this.stream.getMissingChunks(); |
|
this._requestChunks(missingChunks); |
|
return this._loadedStreamCapability.promise; |
|
}, |
|
|
|
_requestChunks: function ChunkedStreamManager_requestChunks(chunks) { |
|
var requestId = this.currRequestId++; |
|
|
|
var chunksNeeded; |
|
var i, ii; |
|
this.chunksNeededByRequest[requestId] = chunksNeeded = {}; |
|
for (i = 0, ii = chunks.length; i < ii; i++) { |
|
if (!this.stream.hasChunk(chunks[i])) { |
|
chunksNeeded[chunks[i]] = true; |
|
} |
|
} |
|
|
|
if (isEmptyObj(chunksNeeded)) { |
|
return Promise.resolve(); |
|
} |
|
|
|
var capability = createPromiseCapability(); |
|
this.promisesByRequest[requestId] = capability; |
|
|
|
var chunksToRequest = []; |
|
for (var chunk in chunksNeeded) { |
|
chunk = chunk | 0; |
|
if (!(chunk in this.requestsByChunk)) { |
|
this.requestsByChunk[chunk] = []; |
|
chunksToRequest.push(chunk); |
|
} |
|
this.requestsByChunk[chunk].push(requestId); |
|
} |
|
|
|
if (!chunksToRequest.length) { |
|
return capability.promise; |
|
} |
|
|
|
var groupedChunksToRequest = this.groupChunks(chunksToRequest); |
|
|
|
for (i = 0; i < groupedChunksToRequest.length; ++i) { |
|
var groupedChunk = groupedChunksToRequest[i]; |
|
var begin = groupedChunk.beginChunk * this.chunkSize; |
|
var end = Math.min(groupedChunk.endChunk * this.chunkSize, this.length); |
|
this.sendRequest(begin, end); |
|
} |
|
|
|
return capability.promise; |
|
}, |
|
|
|
getStream: function ChunkedStreamManager_getStream() { |
|
return this.stream; |
|
}, |
|
|
|
// Loads any chunks in the requested range that are not yet loaded |
|
requestRange: function ChunkedStreamManager_requestRange(begin, end) { |
|
|
|
end = Math.min(end, this.length); |
|
|
|
var beginChunk = this.getBeginChunk(begin); |
|
var endChunk = this.getEndChunk(end); |
|
|
|
var chunks = []; |
|
for (var chunk = beginChunk; chunk < endChunk; ++chunk) { |
|
chunks.push(chunk); |
|
} |
|
|
|
return this._requestChunks(chunks); |
|
}, |
|
|
|
requestRanges: function ChunkedStreamManager_requestRanges(ranges) { |
|
ranges = ranges || []; |
|
var chunksToRequest = []; |
|
|
|
for (var i = 0; i < ranges.length; i++) { |
|
var beginChunk = this.getBeginChunk(ranges[i].begin); |
|
var endChunk = this.getEndChunk(ranges[i].end); |
|
for (var chunk = beginChunk; chunk < endChunk; ++chunk) { |
|
if (chunksToRequest.indexOf(chunk) < 0) { |
|
chunksToRequest.push(chunk); |
|
} |
|
} |
|
} |
|
|
|
chunksToRequest.sort(function(a, b) { return a - b; }); |
|
return this._requestChunks(chunksToRequest); |
|
}, |
|
|
|
// Groups a sorted array of chunks into as few contiguous larger |
|
// chunks as possible |
|
groupChunks: function ChunkedStreamManager_groupChunks(chunks) { |
|
var groupedChunks = []; |
|
var beginChunk = -1; |
|
var prevChunk = -1; |
|
for (var i = 0; i < chunks.length; ++i) { |
|
var chunk = chunks[i]; |
|
|
|
if (beginChunk < 0) { |
|
beginChunk = chunk; |
|
} |
|
|
|
if (prevChunk >= 0 && prevChunk + 1 !== chunk) { |
|
groupedChunks.push({ beginChunk: beginChunk, |
|
endChunk: prevChunk + 1 }); |
|
beginChunk = chunk; |
|
} |
|
if (i + 1 === chunks.length) { |
|
groupedChunks.push({ beginChunk: beginChunk, |
|
endChunk: chunk + 1 }); |
|
} |
|
|
|
prevChunk = chunk; |
|
} |
|
return groupedChunks; |
|
}, |
|
|
|
onProgress: function ChunkedStreamManager_onProgress(args) { |
|
var bytesLoaded = (this.stream.numChunksLoaded * this.chunkSize + |
|
args.loaded); |
|
this.msgHandler.send('DocProgress', { |
|
loaded: bytesLoaded, |
|
total: this.length |
|
}); |
|
}, |
|
|
|
onReceiveData: function ChunkedStreamManager_onReceiveData(args) { |
|
var chunk = args.chunk; |
|
var isProgressive = args.begin === undefined; |
|
var begin = isProgressive ? this.progressiveDataLength : args.begin; |
|
var end = begin + chunk.byteLength; |
|
|
|
var beginChunk = Math.floor(begin / this.chunkSize); |
|
var endChunk = end < this.length ? Math.floor(end / this.chunkSize) : |
|
Math.ceil(end / this.chunkSize); |
|
|
|
if (isProgressive) { |
|
this.stream.onReceiveProgressiveData(chunk); |
|
this.progressiveDataLength = end; |
|
} else { |
|
this.stream.onReceiveData(begin, chunk); |
|
} |
|
|
|
if (this.stream.allChunksLoaded()) { |
|
this._loadedStreamCapability.resolve(this.stream); |
|
} |
|
|
|
var loadedRequests = []; |
|
var i, requestId; |
|
for (chunk = beginChunk; chunk < endChunk; ++chunk) { |
|
// The server might return more chunks than requested |
|
var requestIds = this.requestsByChunk[chunk] || []; |
|
delete this.requestsByChunk[chunk]; |
|
|
|
for (i = 0; i < requestIds.length; ++i) { |
|
requestId = requestIds[i]; |
|
var chunksNeeded = this.chunksNeededByRequest[requestId]; |
|
if (chunk in chunksNeeded) { |
|
delete chunksNeeded[chunk]; |
|
} |
|
|
|
if (!isEmptyObj(chunksNeeded)) { |
|
continue; |
|
} |
|
|
|
loadedRequests.push(requestId); |
|
} |
|
} |
|
|
|
// If there are no pending requests, automatically fetch the next |
|
// unfetched chunk of the PDF |
|
if (!this.disableAutoFetch && isEmptyObj(this.requestsByChunk)) { |
|
var nextEmptyChunk; |
|
if (this.stream.numChunksLoaded === 1) { |
|
// This is a special optimization so that after fetching the first |
|
// chunk, rather than fetching the second chunk, we fetch the last |
|
// chunk. |
|
var lastChunk = this.stream.numChunks - 1; |
|
if (!this.stream.hasChunk(lastChunk)) { |
|
nextEmptyChunk = lastChunk; |
|
} |
|
} else { |
|
nextEmptyChunk = this.stream.nextEmptyChunk(endChunk); |
|
} |
|
if (isInt(nextEmptyChunk)) { |
|
this._requestChunks([nextEmptyChunk]); |
|
} |
|
} |
|
|
|
for (i = 0; i < loadedRequests.length; ++i) { |
|
requestId = loadedRequests[i]; |
|
var capability = this.promisesByRequest[requestId]; |
|
delete this.promisesByRequest[requestId]; |
|
capability.resolve(); |
|
} |
|
|
|
this.msgHandler.send('DocProgress', { |
|
loaded: this.stream.numChunksLoaded * this.chunkSize, |
|
total: this.length |
|
}); |
|
}, |
|
|
|
onError: function ChunkedStreamManager_onError(err) { |
|
this._loadedStreamCapability.reject(err); |
|
}, |
|
|
|
getBeginChunk: function ChunkedStreamManager_getBeginChunk(begin) { |
|
var chunk = Math.floor(begin / this.chunkSize); |
|
return chunk; |
|
}, |
|
|
|
getEndChunk: function ChunkedStreamManager_getEndChunk(end) { |
|
var chunk = Math.floor((end - 1) / this.chunkSize) + 1; |
|
return chunk; |
|
}, |
|
|
|
abort: function ChunkedStreamManager_abort() { |
|
if (this.networkManager) { |
|
this.networkManager.abortAllRequests(); |
|
} |
|
for(var requestId in this.promisesByRequest) { |
|
var capability = this.promisesByRequest[requestId]; |
|
capability.reject(new Error('Request was aborted')); |
|
} |
|
} |
|
}; |
|
|
|
return ChunkedStreamManager; |
|
})(); |
|
|
|
exports.ChunkedStream = ChunkedStream; |
|
exports.ChunkedStreamManager = ChunkedStreamManager; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreJbig2 = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCoreArithmeticDecoder); |
|
} |
|
}(this, function (exports, sharedUtil, coreArithmeticDecoder) { |
|
|
|
var error = sharedUtil.error; |
|
var log2 = sharedUtil.log2; |
|
var readInt8 = sharedUtil.readInt8; |
|
var readUint16 = sharedUtil.readUint16; |
|
var readUint32 = sharedUtil.readUint32; |
|
var shadow = sharedUtil.shadow; |
|
var ArithmeticDecoder = coreArithmeticDecoder.ArithmeticDecoder; |
|
|
|
var Jbig2Image = (function Jbig2ImageClosure() { |
|
// Utility data structures |
|
function ContextCache() {} |
|
|
|
ContextCache.prototype = { |
|
getContexts: function(id) { |
|
if (id in this) { |
|
return this[id]; |
|
} |
|
return (this[id] = new Int8Array(1 << 16)); |
|
} |
|
}; |
|
|
|
function DecodingContext(data, start, end) { |
|
this.data = data; |
|
this.start = start; |
|
this.end = end; |
|
} |
|
|
|
DecodingContext.prototype = { |
|
get decoder() { |
|
var decoder = new ArithmeticDecoder(this.data, this.start, this.end); |
|
return shadow(this, 'decoder', decoder); |
|
}, |
|
get contextCache() { |
|
var cache = new ContextCache(); |
|
return shadow(this, 'contextCache', cache); |
|
} |
|
}; |
|
|
|
// Annex A. Arithmetic Integer Decoding Procedure |
|
// A.2 Procedure for decoding values |
|
function decodeInteger(contextCache, procedure, decoder) { |
|
var contexts = contextCache.getContexts(procedure); |
|
var prev = 1; |
|
|
|
function readBits(length) { |
|
var v = 0; |
|
for (var i = 0; i < length; i++) { |
|
var bit = decoder.readBit(contexts, prev); |
|
prev = (prev < 256 ? (prev << 1) | bit : |
|
(((prev << 1) | bit) & 511) | 256); |
|
v = (v << 1) | bit; |
|
} |
|
return v >>> 0; |
|
} |
|
|
|
var sign = readBits(1); |
|
var value = readBits(1) ? |
|
(readBits(1) ? |
|
(readBits(1) ? |
|
(readBits(1) ? |
|
(readBits(1) ? |
|
(readBits(32) + 4436) : |
|
readBits(12) + 340) : |
|
readBits(8) + 84) : |
|
readBits(6) + 20) : |
|
readBits(4) + 4) : |
|
readBits(2); |
|
return (sign === 0 ? value : (value > 0 ? -value : null)); |
|
} |
|
|
|
// A.3 The IAID decoding procedure |
|
function decodeIAID(contextCache, decoder, codeLength) { |
|
var contexts = contextCache.getContexts('IAID'); |
|
|
|
var prev = 1; |
|
for (var i = 0; i < codeLength; i++) { |
|
var bit = decoder.readBit(contexts, prev); |
|
prev = (prev << 1) | bit; |
|
} |
|
if (codeLength < 31) { |
|
return prev & ((1 << codeLength) - 1); |
|
} |
|
return prev & 0x7FFFFFFF; |
|
} |
|
|
|
// 7.3 Segment types |
|
var SegmentTypes = [ |
|
'SymbolDictionary', null, null, null, 'IntermediateTextRegion', null, |
|
'ImmediateTextRegion', 'ImmediateLosslessTextRegion', null, null, null, |
|
null, null, null, null, null, 'patternDictionary', null, null, null, |
|
'IntermediateHalftoneRegion', null, 'ImmediateHalftoneRegion', |
|
'ImmediateLosslessHalftoneRegion', null, null, null, null, null, null, null, |
|
null, null, null, null, null, 'IntermediateGenericRegion', null, |
|
'ImmediateGenericRegion', 'ImmediateLosslessGenericRegion', |
|
'IntermediateGenericRefinementRegion', null, |
|
'ImmediateGenericRefinementRegion', |
|
'ImmediateLosslessGenericRefinementRegion', null, null, null, null, |
|
'PageInformation', 'EndOfPage', 'EndOfStripe', 'EndOfFile', 'Profiles', |
|
'Tables', null, null, null, null, null, null, null, null, |
|
'Extension' |
|
]; |
|
|
|
var CodingTemplates = [ |
|
[{x: -1, y: -2}, {x: 0, y: -2}, {x: 1, y: -2}, {x: -2, y: -1}, |
|
{x: -1, y: -1}, {x: 0, y: -1}, {x: 1, y: -1}, {x: 2, y: -1}, |
|
{x: -4, y: 0}, {x: -3, y: 0}, {x: -2, y: 0}, {x: -1, y: 0}], |
|
[{x: -1, y: -2}, {x: 0, y: -2}, {x: 1, y: -2}, {x: 2, y: -2}, |
|
{x: -2, y: -1}, {x: -1, y: -1}, {x: 0, y: -1}, {x: 1, y: -1}, |
|
{x: 2, y: -1}, {x: -3, y: 0}, {x: -2, y: 0}, {x: -1, y: 0}], |
|
[{x: -1, y: -2}, {x: 0, y: -2}, {x: 1, y: -2}, {x: -2, y: -1}, |
|
{x: -1, y: -1}, {x: 0, y: -1}, {x: 1, y: -1}, {x: -2, y: 0}, |
|
{x: -1, y: 0}], |
|
[{x: -3, y: -1}, {x: -2, y: -1}, {x: -1, y: -1}, {x: 0, y: -1}, |
|
{x: 1, y: -1}, {x: -4, y: 0}, {x: -3, y: 0}, {x: -2, y: 0}, {x: -1, y: 0}] |
|
]; |
|
|
|
var RefinementTemplates = [ |
|
{ |
|
coding: [{x: 0, y: -1}, {x: 1, y: -1}, {x: -1, y: 0}], |
|
reference: [{x: 0, y: -1}, {x: 1, y: -1}, {x: -1, y: 0}, {x: 0, y: 0}, |
|
{x: 1, y: 0}, {x: -1, y: 1}, {x: 0, y: 1}, {x: 1, y: 1}] |
|
}, |
|
{ |
|
coding: [{x: -1, y: -1}, {x: 0, y: -1}, {x: 1, y: -1}, {x: -1, y: 0}], |
|
reference: [{x: 0, y: -1}, {x: -1, y: 0}, {x: 0, y: 0}, {x: 1, y: 0}, |
|
{x: 0, y: 1}, {x: 1, y: 1}] |
|
} |
|
]; |
|
|
|
// See 6.2.5.7 Decoding the bitmap. |
|
var ReusedContexts = [ |
|
0x9B25, // 10011 0110010 0101 |
|
0x0795, // 0011 110010 101 |
|
0x00E5, // 001 11001 01 |
|
0x0195 // 011001 0101 |
|
]; |
|
|
|
var RefinementReusedContexts = [ |
|
0x0020, // '000' + '0' (coding) + '00010000' + '0' (reference) |
|
0x0008 // '0000' + '001000' |
|
]; |
|
|
|
function decodeBitmapTemplate0(width, height, decodingContext) { |
|
var decoder = decodingContext.decoder; |
|
var contexts = decodingContext.contextCache.getContexts('GB'); |
|
var contextLabel, i, j, pixel, row, row1, row2, bitmap = []; |
|
|
|
// ...ooooo.... |
|
// ..ooooooo... Context template for current pixel (X) |
|
// .ooooX...... (concatenate values of 'o'-pixels to get contextLabel) |
|
var OLD_PIXEL_MASK = 0x7BF7; // 01111 0111111 0111 |
|
|
|
for (i = 0; i < height; i++) { |
|
row = bitmap[i] = new Uint8Array(width); |
|
row1 = (i < 1) ? row : bitmap[i - 1]; |
|
row2 = (i < 2) ? row : bitmap[i - 2]; |
|
|
|
// At the beginning of each row: |
|
// Fill contextLabel with pixels that are above/right of (X) |
|
contextLabel = (row2[0] << 13) | (row2[1] << 12) | (row2[2] << 11) | |
|
(row1[0] << 7) | (row1[1] << 6) | (row1[2] << 5) | |
|
(row1[3] << 4); |
|
|
|
for (j = 0; j < width; j++) { |
|
row[j] = pixel = decoder.readBit(contexts, contextLabel); |
|
|
|
// At each pixel: Clear contextLabel pixels that are shifted |
|
// out of the context, then add new ones. |
|
contextLabel = ((contextLabel & OLD_PIXEL_MASK) << 1) | |
|
(j + 3 < width ? row2[j + 3] << 11 : 0) | |
|
(j + 4 < width ? row1[j + 4] << 4 : 0) | pixel; |
|
} |
|
} |
|
|
|
return bitmap; |
|
} |
|
|
|
// 6.2 Generic Region Decoding Procedure |
|
function decodeBitmap(mmr, width, height, templateIndex, prediction, skip, at, |
|
decodingContext) { |
|
if (mmr) { |
|
error('JBIG2 error: MMR encoding is not supported'); |
|
} |
|
|
|
// Use optimized version for the most common case |
|
if (templateIndex === 0 && !skip && !prediction && at.length === 4 && |
|
at[0].x === 3 && at[0].y === -1 && at[1].x === -3 && at[1].y === -1 && |
|
at[2].x === 2 && at[2].y === -2 && at[3].x === -2 && at[3].y === -2) { |
|
return decodeBitmapTemplate0(width, height, decodingContext); |
|
} |
|
|
|
var useskip = !!skip; |
|
var template = CodingTemplates[templateIndex].concat(at); |
|
|
|
// Sorting is non-standard, and it is not required. But sorting increases |
|
// the number of template bits that can be reused from the previous |
|
// contextLabel in the main loop. |
|
template.sort(function (a, b) { |
|
return (a.y - b.y) || (a.x - b.x); |
|
}); |
|
|
|
var templateLength = template.length; |
|
var templateX = new Int8Array(templateLength); |
|
var templateY = new Int8Array(templateLength); |
|
var changingTemplateEntries = []; |
|
var reuseMask = 0, minX = 0, maxX = 0, minY = 0; |
|
var c, k; |
|
|
|
for (k = 0; k < templateLength; k++) { |
|
templateX[k] = template[k].x; |
|
templateY[k] = template[k].y; |
|
minX = Math.min(minX, template[k].x); |
|
maxX = Math.max(maxX, template[k].x); |
|
minY = Math.min(minY, template[k].y); |
|
// Check if the template pixel appears in two consecutive context labels, |
|
// so it can be reused. Otherwise, we add it to the list of changing |
|
// template entries. |
|
if (k < templateLength - 1 && |
|
template[k].y === template[k + 1].y && |
|
template[k].x === template[k + 1].x - 1) { |
|
reuseMask |= 1 << (templateLength - 1 - k); |
|
} else { |
|
changingTemplateEntries.push(k); |
|
} |
|
} |
|
var changingEntriesLength = changingTemplateEntries.length; |
|
|
|
var changingTemplateX = new Int8Array(changingEntriesLength); |
|
var changingTemplateY = new Int8Array(changingEntriesLength); |
|
var changingTemplateBit = new Uint16Array(changingEntriesLength); |
|
for (c = 0; c < changingEntriesLength; c++) { |
|
k = changingTemplateEntries[c]; |
|
changingTemplateX[c] = template[k].x; |
|
changingTemplateY[c] = template[k].y; |
|
changingTemplateBit[c] = 1 << (templateLength - 1 - k); |
|
} |
|
|
|
// Get the safe bounding box edges from the width, height, minX, maxX, minY |
|
var sbb_left = -minX; |
|
var sbb_top = -minY; |
|
var sbb_right = width - maxX; |
|
|
|
var pseudoPixelContext = ReusedContexts[templateIndex]; |
|
var row = new Uint8Array(width); |
|
var bitmap = []; |
|
|
|
var decoder = decodingContext.decoder; |
|
var contexts = decodingContext.contextCache.getContexts('GB'); |
|
|
|
var ltp = 0, j, i0, j0, contextLabel = 0, bit, shift; |
|
for (var i = 0; i < height; i++) { |
|
if (prediction) { |
|
var sltp = decoder.readBit(contexts, pseudoPixelContext); |
|
ltp ^= sltp; |
|
if (ltp) { |
|
bitmap.push(row); // duplicate previous row |
|
continue; |
|
} |
|
} |
|
row = new Uint8Array(row); |
|
bitmap.push(row); |
|
for (j = 0; j < width; j++) { |
|
if (useskip && skip[i][j]) { |
|
row[j] = 0; |
|
continue; |
|
} |
|
// Are we in the middle of a scanline, so we can reuse contextLabel |
|
// bits? |
|
if (j >= sbb_left && j < sbb_right && i >= sbb_top) { |
|
// If yes, we can just shift the bits that are reusable and only |
|
// fetch the remaining ones. |
|
contextLabel = (contextLabel << 1) & reuseMask; |
|
for (k = 0; k < changingEntriesLength; k++) { |
|
i0 = i + changingTemplateY[k]; |
|
j0 = j + changingTemplateX[k]; |
|
bit = bitmap[i0][j0]; |
|
if (bit) { |
|
bit = changingTemplateBit[k]; |
|
contextLabel |= bit; |
|
} |
|
} |
|
} else { |
|
// compute the contextLabel from scratch |
|
contextLabel = 0; |
|
shift = templateLength - 1; |
|
for (k = 0; k < templateLength; k++, shift--) { |
|
j0 = j + templateX[k]; |
|
if (j0 >= 0 && j0 < width) { |
|
i0 = i + templateY[k]; |
|
if (i0 >= 0) { |
|
bit = bitmap[i0][j0]; |
|
if (bit) { |
|
contextLabel |= bit << shift; |
|
} |
|
} |
|
} |
|
} |
|
} |
|
var pixel = decoder.readBit(contexts, contextLabel); |
|
row[j] = pixel; |
|
} |
|
} |
|
return bitmap; |
|
} |
|
|
|
// 6.3.2 Generic Refinement Region Decoding Procedure |
|
function decodeRefinement(width, height, templateIndex, referenceBitmap, |
|
offsetX, offsetY, prediction, at, |
|
decodingContext) { |
|
var codingTemplate = RefinementTemplates[templateIndex].coding; |
|
if (templateIndex === 0) { |
|
codingTemplate = codingTemplate.concat([at[0]]); |
|
} |
|
var codingTemplateLength = codingTemplate.length; |
|
var codingTemplateX = new Int32Array(codingTemplateLength); |
|
var codingTemplateY = new Int32Array(codingTemplateLength); |
|
var k; |
|
for (k = 0; k < codingTemplateLength; k++) { |
|
codingTemplateX[k] = codingTemplate[k].x; |
|
codingTemplateY[k] = codingTemplate[k].y; |
|
} |
|
|
|
var referenceTemplate = RefinementTemplates[templateIndex].reference; |
|
if (templateIndex === 0) { |
|
referenceTemplate = referenceTemplate.concat([at[1]]); |
|
} |
|
var referenceTemplateLength = referenceTemplate.length; |
|
var referenceTemplateX = new Int32Array(referenceTemplateLength); |
|
var referenceTemplateY = new Int32Array(referenceTemplateLength); |
|
for (k = 0; k < referenceTemplateLength; k++) { |
|
referenceTemplateX[k] = referenceTemplate[k].x; |
|
referenceTemplateY[k] = referenceTemplate[k].y; |
|
} |
|
var referenceWidth = referenceBitmap[0].length; |
|
var referenceHeight = referenceBitmap.length; |
|
|
|
var pseudoPixelContext = RefinementReusedContexts[templateIndex]; |
|
var bitmap = []; |
|
|
|
var decoder = decodingContext.decoder; |
|
var contexts = decodingContext.contextCache.getContexts('GR'); |
|
|
|
var ltp = 0; |
|
for (var i = 0; i < height; i++) { |
|
if (prediction) { |
|
var sltp = decoder.readBit(contexts, pseudoPixelContext); |
|
ltp ^= sltp; |
|
if (ltp) { |
|
error('JBIG2 error: prediction is not supported'); |
|
} |
|
} |
|
var row = new Uint8Array(width); |
|
bitmap.push(row); |
|
for (var j = 0; j < width; j++) { |
|
var i0, j0; |
|
var contextLabel = 0; |
|
for (k = 0; k < codingTemplateLength; k++) { |
|
i0 = i + codingTemplateY[k]; |
|
j0 = j + codingTemplateX[k]; |
|
if (i0 < 0 || j0 < 0 || j0 >= width) { |
|
contextLabel <<= 1; // out of bound pixel |
|
} else { |
|
contextLabel = (contextLabel << 1) | bitmap[i0][j0]; |
|
} |
|
} |
|
for (k = 0; k < referenceTemplateLength; k++) { |
|
i0 = i + referenceTemplateY[k] + offsetY; |
|
j0 = j + referenceTemplateX[k] + offsetX; |
|
if (i0 < 0 || i0 >= referenceHeight || j0 < 0 || |
|
j0 >= referenceWidth) { |
|
contextLabel <<= 1; // out of bound pixel |
|
} else { |
|
contextLabel = (contextLabel << 1) | referenceBitmap[i0][j0]; |
|
} |
|
} |
|
var pixel = decoder.readBit(contexts, contextLabel); |
|
row[j] = pixel; |
|
} |
|
} |
|
|
|
return bitmap; |
|
} |
|
|
|
// 6.5.5 Decoding the symbol dictionary |
|
function decodeSymbolDictionary(huffman, refinement, symbols, |
|
numberOfNewSymbols, numberOfExportedSymbols, |
|
huffmanTables, templateIndex, at, |
|
refinementTemplateIndex, refinementAt, |
|
decodingContext) { |
|
if (huffman) { |
|
error('JBIG2 error: huffman is not supported'); |
|
} |
|
|
|
var newSymbols = []; |
|
var currentHeight = 0; |
|
var symbolCodeLength = log2(symbols.length + numberOfNewSymbols); |
|
|
|
var decoder = decodingContext.decoder; |
|
var contextCache = decodingContext.contextCache; |
|
|
|
while (newSymbols.length < numberOfNewSymbols) { |
|
var deltaHeight = decodeInteger(contextCache, 'IADH', decoder); // 6.5.6 |
|
currentHeight += deltaHeight; |
|
var currentWidth = 0; |
|
var totalWidth = 0; |
|
while (true) { |
|
var deltaWidth = decodeInteger(contextCache, 'IADW', decoder); // 6.5.7 |
|
if (deltaWidth === null) { |
|
break; // OOB |
|
} |
|
currentWidth += deltaWidth; |
|
totalWidth += currentWidth; |
|
var bitmap; |
|
if (refinement) { |
|
// 6.5.8.2 Refinement/aggregate-coded symbol bitmap |
|
var numberOfInstances = decodeInteger(contextCache, 'IAAI', decoder); |
|
if (numberOfInstances > 1) { |
|
bitmap = decodeTextRegion(huffman, refinement, |
|
currentWidth, currentHeight, 0, |
|
numberOfInstances, 1, //strip size |
|
symbols.concat(newSymbols), |
|
symbolCodeLength, |
|
0, //transposed |
|
0, //ds offset |
|
1, //top left 7.4.3.1.1 |
|
0, //OR operator |
|
huffmanTables, |
|
refinementTemplateIndex, refinementAt, |
|
decodingContext); |
|
} else { |
|
var symbolId = decodeIAID(contextCache, decoder, symbolCodeLength); |
|
var rdx = decodeInteger(contextCache, 'IARDX', decoder); // 6.4.11.3 |
|
var rdy = decodeInteger(contextCache, 'IARDY', decoder); // 6.4.11.4 |
|
var symbol = (symbolId < symbols.length ? symbols[symbolId] : |
|
newSymbols[symbolId - symbols.length]); |
|
bitmap = decodeRefinement(currentWidth, currentHeight, |
|
refinementTemplateIndex, symbol, rdx, rdy, false, refinementAt, |
|
decodingContext); |
|
} |
|
} else { |
|
// 6.5.8.1 Direct-coded symbol bitmap |
|
bitmap = decodeBitmap(false, currentWidth, currentHeight, |
|
templateIndex, false, null, at, decodingContext); |
|
} |
|
newSymbols.push(bitmap); |
|
} |
|
} |
|
// 6.5.10 Exported symbols |
|
var exportedSymbols = []; |
|
var flags = [], currentFlag = false; |
|
var totalSymbolsLength = symbols.length + numberOfNewSymbols; |
|
while (flags.length < totalSymbolsLength) { |
|
var runLength = decodeInteger(contextCache, 'IAEX', decoder); |
|
while (runLength--) { |
|
flags.push(currentFlag); |
|
} |
|
currentFlag = !currentFlag; |
|
} |
|
for (var i = 0, ii = symbols.length; i < ii; i++) { |
|
if (flags[i]) { |
|
exportedSymbols.push(symbols[i]); |
|
} |
|
} |
|
for (var j = 0; j < numberOfNewSymbols; i++, j++) { |
|
if (flags[i]) { |
|
exportedSymbols.push(newSymbols[j]); |
|
} |
|
} |
|
return exportedSymbols; |
|
} |
|
|
|
function decodeTextRegion(huffman, refinement, width, height, |
|
defaultPixelValue, numberOfSymbolInstances, |
|
stripSize, inputSymbols, symbolCodeLength, |
|
transposed, dsOffset, referenceCorner, |
|
combinationOperator, huffmanTables, |
|
refinementTemplateIndex, refinementAt, |
|
decodingContext) { |
|
if (huffman) { |
|
error('JBIG2 error: huffman is not supported'); |
|
} |
|
|
|
// Prepare bitmap |
|
var bitmap = []; |
|
var i, row; |
|
for (i = 0; i < height; i++) { |
|
row = new Uint8Array(width); |
|
if (defaultPixelValue) { |
|
for (var j = 0; j < width; j++) { |
|
row[j] = defaultPixelValue; |
|
} |
|
} |
|
bitmap.push(row); |
|
} |
|
|
|
var decoder = decodingContext.decoder; |
|
var contextCache = decodingContext.contextCache; |
|
var stripT = -decodeInteger(contextCache, 'IADT', decoder); // 6.4.6 |
|
var firstS = 0; |
|
i = 0; |
|
while (i < numberOfSymbolInstances) { |
|
var deltaT = decodeInteger(contextCache, 'IADT', decoder); // 6.4.6 |
|
stripT += deltaT; |
|
|
|
var deltaFirstS = decodeInteger(contextCache, 'IAFS', decoder); // 6.4.7 |
|
firstS += deltaFirstS; |
|
var currentS = firstS; |
|
do { |
|
var currentT = (stripSize === 1 ? 0 : |
|
decodeInteger(contextCache, 'IAIT', decoder)); // 6.4.9 |
|
var t = stripSize * stripT + currentT; |
|
var symbolId = decodeIAID(contextCache, decoder, symbolCodeLength); |
|
var applyRefinement = (refinement && |
|
decodeInteger(contextCache, 'IARI', decoder)); |
|
var symbolBitmap = inputSymbols[symbolId]; |
|
var symbolWidth = symbolBitmap[0].length; |
|
var symbolHeight = symbolBitmap.length; |
|
if (applyRefinement) { |
|
var rdw = decodeInteger(contextCache, 'IARDW', decoder); // 6.4.11.1 |
|
var rdh = decodeInteger(contextCache, 'IARDH', decoder); // 6.4.11.2 |
|
var rdx = decodeInteger(contextCache, 'IARDX', decoder); // 6.4.11.3 |
|
var rdy = decodeInteger(contextCache, 'IARDY', decoder); // 6.4.11.4 |
|
symbolWidth += rdw; |
|
symbolHeight += rdh; |
|
symbolBitmap = decodeRefinement(symbolWidth, symbolHeight, |
|
refinementTemplateIndex, symbolBitmap, (rdw >> 1) + rdx, |
|
(rdh >> 1) + rdy, false, refinementAt, |
|
decodingContext); |
|
} |
|
var offsetT = t - ((referenceCorner & 1) ? 0 : symbolHeight); |
|
var offsetS = currentS - ((referenceCorner & 2) ? symbolWidth : 0); |
|
var s2, t2, symbolRow; |
|
if (transposed) { |
|
// Place Symbol Bitmap from T1,S1 |
|
for (s2 = 0; s2 < symbolHeight; s2++) { |
|
row = bitmap[offsetS + s2]; |
|
if (!row) { |
|
continue; |
|
} |
|
symbolRow = symbolBitmap[s2]; |
|
// To ignore Parts of Symbol bitmap which goes |
|
// outside bitmap region |
|
var maxWidth = Math.min(width - offsetT, symbolWidth); |
|
switch (combinationOperator) { |
|
case 0: // OR |
|
for (t2 = 0; t2 < maxWidth; t2++) { |
|
row[offsetT + t2] |= symbolRow[t2]; |
|
} |
|
break; |
|
case 2: // XOR |
|
for (t2 = 0; t2 < maxWidth; t2++) { |
|
row[offsetT + t2] ^= symbolRow[t2]; |
|
} |
|
break; |
|
default: |
|
error('JBIG2 error: operator ' + combinationOperator + |
|
' is not supported'); |
|
} |
|
} |
|
currentS += symbolHeight - 1; |
|
} else { |
|
for (t2 = 0; t2 < symbolHeight; t2++) { |
|
row = bitmap[offsetT + t2]; |
|
if (!row) { |
|
continue; |
|
} |
|
symbolRow = symbolBitmap[t2]; |
|
switch (combinationOperator) { |
|
case 0: // OR |
|
for (s2 = 0; s2 < symbolWidth; s2++) { |
|
row[offsetS + s2] |= symbolRow[s2]; |
|
} |
|
break; |
|
case 2: // XOR |
|
for (s2 = 0; s2 < symbolWidth; s2++) { |
|
row[offsetS + s2] ^= symbolRow[s2]; |
|
} |
|
break; |
|
default: |
|
error('JBIG2 error: operator ' + combinationOperator + |
|
' is not supported'); |
|
} |
|
} |
|
currentS += symbolWidth - 1; |
|
} |
|
i++; |
|
var deltaS = decodeInteger(contextCache, 'IADS', decoder); // 6.4.8 |
|
if (deltaS === null) { |
|
break; // OOB |
|
} |
|
currentS += deltaS + dsOffset; |
|
} while (true); |
|
} |
|
return bitmap; |
|
} |
|
|
|
function readSegmentHeader(data, start) { |
|
var segmentHeader = {}; |
|
segmentHeader.number = readUint32(data, start); |
|
var flags = data[start + 4]; |
|
var segmentType = flags & 0x3F; |
|
if (!SegmentTypes[segmentType]) { |
|
error('JBIG2 error: invalid segment type: ' + segmentType); |
|
} |
|
segmentHeader.type = segmentType; |
|
segmentHeader.typeName = SegmentTypes[segmentType]; |
|
segmentHeader.deferredNonRetain = !!(flags & 0x80); |
|
|
|
var pageAssociationFieldSize = !!(flags & 0x40); |
|
var referredFlags = data[start + 5]; |
|
var referredToCount = (referredFlags >> 5) & 7; |
|
var retainBits = [referredFlags & 31]; |
|
var position = start + 6; |
|
if (referredFlags === 7) { |
|
referredToCount = readUint32(data, position - 1) & 0x1FFFFFFF; |
|
position += 3; |
|
var bytes = (referredToCount + 7) >> 3; |
|
retainBits[0] = data[position++]; |
|
while (--bytes > 0) { |
|
retainBits.push(data[position++]); |
|
} |
|
} else if (referredFlags === 5 || referredFlags === 6) { |
|
error('JBIG2 error: invalid referred-to flags'); |
|
} |
|
|
|
segmentHeader.retainBits = retainBits; |
|
var referredToSegmentNumberSize = (segmentHeader.number <= 256 ? 1 : |
|
(segmentHeader.number <= 65536 ? 2 : 4)); |
|
var referredTo = []; |
|
var i, ii; |
|
for (i = 0; i < referredToCount; i++) { |
|
var number = (referredToSegmentNumberSize === 1 ? data[position] : |
|
(referredToSegmentNumberSize === 2 ? readUint16(data, position) : |
|
readUint32(data, position))); |
|
referredTo.push(number); |
|
position += referredToSegmentNumberSize; |
|
} |
|
segmentHeader.referredTo = referredTo; |
|
if (!pageAssociationFieldSize) { |
|
segmentHeader.pageAssociation = data[position++]; |
|
} else { |
|
segmentHeader.pageAssociation = readUint32(data, position); |
|
position += 4; |
|
} |
|
segmentHeader.length = readUint32(data, position); |
|
position += 4; |
|
|
|
if (segmentHeader.length === 0xFFFFFFFF) { |
|
// 7.2.7 Segment data length, unknown segment length |
|
if (segmentType === 38) { // ImmediateGenericRegion |
|
var genericRegionInfo = readRegionSegmentInformation(data, position); |
|
var genericRegionSegmentFlags = data[position + |
|
RegionSegmentInformationFieldLength]; |
|
var genericRegionMmr = !!(genericRegionSegmentFlags & 1); |
|
// searching for the segment end |
|
var searchPatternLength = 6; |
|
var searchPattern = new Uint8Array(searchPatternLength); |
|
if (!genericRegionMmr) { |
|
searchPattern[0] = 0xFF; |
|
searchPattern[1] = 0xAC; |
|
} |
|
searchPattern[2] = (genericRegionInfo.height >>> 24) & 0xFF; |
|
searchPattern[3] = (genericRegionInfo.height >> 16) & 0xFF; |
|
searchPattern[4] = (genericRegionInfo.height >> 8) & 0xFF; |
|
searchPattern[5] = genericRegionInfo.height & 0xFF; |
|
for (i = position, ii = data.length; i < ii; i++) { |
|
var j = 0; |
|
while (j < searchPatternLength && searchPattern[j] === data[i + j]) { |
|
j++; |
|
} |
|
if (j === searchPatternLength) { |
|
segmentHeader.length = i + searchPatternLength; |
|
break; |
|
} |
|
} |
|
if (segmentHeader.length === 0xFFFFFFFF) { |
|
error('JBIG2 error: segment end was not found'); |
|
} |
|
} else { |
|
error('JBIG2 error: invalid unknown segment length'); |
|
} |
|
} |
|
segmentHeader.headerEnd = position; |
|
return segmentHeader; |
|
} |
|
|
|
function readSegments(header, data, start, end) { |
|
var segments = []; |
|
var position = start; |
|
while (position < end) { |
|
var segmentHeader = readSegmentHeader(data, position); |
|
position = segmentHeader.headerEnd; |
|
var segment = { |
|
header: segmentHeader, |
|
data: data |
|
}; |
|
if (!header.randomAccess) { |
|
segment.start = position; |
|
position += segmentHeader.length; |
|
segment.end = position; |
|
} |
|
segments.push(segment); |
|
if (segmentHeader.type === 51) { |
|
break; // end of file is found |
|
} |
|
} |
|
if (header.randomAccess) { |
|
for (var i = 0, ii = segments.length; i < ii; i++) { |
|
segments[i].start = position; |
|
position += segments[i].header.length; |
|
segments[i].end = position; |
|
} |
|
} |
|
return segments; |
|
} |
|
|
|
// 7.4.1 Region segment information field |
|
function readRegionSegmentInformation(data, start) { |
|
return { |
|
width: readUint32(data, start), |
|
height: readUint32(data, start + 4), |
|
x: readUint32(data, start + 8), |
|
y: readUint32(data, start + 12), |
|
combinationOperator: data[start + 16] & 7 |
|
}; |
|
} |
|
var RegionSegmentInformationFieldLength = 17; |
|
|
|
function processSegment(segment, visitor) { |
|
var header = segment.header; |
|
|
|
var data = segment.data, position = segment.start, end = segment.end; |
|
var args, at, i, atLength; |
|
switch (header.type) { |
|
case 0: // SymbolDictionary |
|
// 7.4.2 Symbol dictionary segment syntax |
|
var dictionary = {}; |
|
var dictionaryFlags = readUint16(data, position); // 7.4.2.1.1 |
|
dictionary.huffman = !!(dictionaryFlags & 1); |
|
dictionary.refinement = !!(dictionaryFlags & 2); |
|
dictionary.huffmanDHSelector = (dictionaryFlags >> 2) & 3; |
|
dictionary.huffmanDWSelector = (dictionaryFlags >> 4) & 3; |
|
dictionary.bitmapSizeSelector = (dictionaryFlags >> 6) & 1; |
|
dictionary.aggregationInstancesSelector = (dictionaryFlags >> 7) & 1; |
|
dictionary.bitmapCodingContextUsed = !!(dictionaryFlags & 256); |
|
dictionary.bitmapCodingContextRetained = !!(dictionaryFlags & 512); |
|
dictionary.template = (dictionaryFlags >> 10) & 3; |
|
dictionary.refinementTemplate = (dictionaryFlags >> 12) & 1; |
|
position += 2; |
|
if (!dictionary.huffman) { |
|
atLength = dictionary.template === 0 ? 4 : 1; |
|
at = []; |
|
for (i = 0; i < atLength; i++) { |
|
at.push({ |
|
x: readInt8(data, position), |
|
y: readInt8(data, position + 1) |
|
}); |
|
position += 2; |
|
} |
|
dictionary.at = at; |
|
} |
|
if (dictionary.refinement && !dictionary.refinementTemplate) { |
|
at = []; |
|
for (i = 0; i < 2; i++) { |
|
at.push({ |
|
x: readInt8(data, position), |
|
y: readInt8(data, position + 1) |
|
}); |
|
position += 2; |
|
} |
|
dictionary.refinementAt = at; |
|
} |
|
dictionary.numberOfExportedSymbols = readUint32(data, position); |
|
position += 4; |
|
dictionary.numberOfNewSymbols = readUint32(data, position); |
|
position += 4; |
|
args = [dictionary, header.number, header.referredTo, |
|
data, position, end]; |
|
break; |
|
case 6: // ImmediateTextRegion |
|
case 7: // ImmediateLosslessTextRegion |
|
var textRegion = {}; |
|
textRegion.info = readRegionSegmentInformation(data, position); |
|
position += RegionSegmentInformationFieldLength; |
|
var textRegionSegmentFlags = readUint16(data, position); |
|
position += 2; |
|
textRegion.huffman = !!(textRegionSegmentFlags & 1); |
|
textRegion.refinement = !!(textRegionSegmentFlags & 2); |
|
textRegion.stripSize = 1 << ((textRegionSegmentFlags >> 2) & 3); |
|
textRegion.referenceCorner = (textRegionSegmentFlags >> 4) & 3; |
|
textRegion.transposed = !!(textRegionSegmentFlags & 64); |
|
textRegion.combinationOperator = (textRegionSegmentFlags >> 7) & 3; |
|
textRegion.defaultPixelValue = (textRegionSegmentFlags >> 9) & 1; |
|
textRegion.dsOffset = (textRegionSegmentFlags << 17) >> 27; |
|
textRegion.refinementTemplate = (textRegionSegmentFlags >> 15) & 1; |
|
if (textRegion.huffman) { |
|
var textRegionHuffmanFlags = readUint16(data, position); |
|
position += 2; |
|
textRegion.huffmanFS = (textRegionHuffmanFlags) & 3; |
|
textRegion.huffmanDS = (textRegionHuffmanFlags >> 2) & 3; |
|
textRegion.huffmanDT = (textRegionHuffmanFlags >> 4) & 3; |
|
textRegion.huffmanRefinementDW = (textRegionHuffmanFlags >> 6) & 3; |
|
textRegion.huffmanRefinementDH = (textRegionHuffmanFlags >> 8) & 3; |
|
textRegion.huffmanRefinementDX = (textRegionHuffmanFlags >> 10) & 3; |
|
textRegion.huffmanRefinementDY = (textRegionHuffmanFlags >> 12) & 3; |
|
textRegion.huffmanRefinementSizeSelector = |
|
!!(textRegionHuffmanFlags & 14); |
|
} |
|
if (textRegion.refinement && !textRegion.refinementTemplate) { |
|
at = []; |
|
for (i = 0; i < 2; i++) { |
|
at.push({ |
|
x: readInt8(data, position), |
|
y: readInt8(data, position + 1) |
|
}); |
|
position += 2; |
|
} |
|
textRegion.refinementAt = at; |
|
} |
|
textRegion.numberOfSymbolInstances = readUint32(data, position); |
|
position += 4; |
|
// TODO 7.4.3.1.7 Symbol ID Huffman table decoding |
|
if (textRegion.huffman) { |
|
error('JBIG2 error: huffman is not supported'); |
|
} |
|
args = [textRegion, header.referredTo, data, position, end]; |
|
break; |
|
case 38: // ImmediateGenericRegion |
|
case 39: // ImmediateLosslessGenericRegion |
|
var genericRegion = {}; |
|
genericRegion.info = readRegionSegmentInformation(data, position); |
|
position += RegionSegmentInformationFieldLength; |
|
var genericRegionSegmentFlags = data[position++]; |
|
genericRegion.mmr = !!(genericRegionSegmentFlags & 1); |
|
genericRegion.template = (genericRegionSegmentFlags >> 1) & 3; |
|
genericRegion.prediction = !!(genericRegionSegmentFlags & 8); |
|
if (!genericRegion.mmr) { |
|
atLength = genericRegion.template === 0 ? 4 : 1; |
|
at = []; |
|
for (i = 0; i < atLength; i++) { |
|
at.push({ |
|
x: readInt8(data, position), |
|
y: readInt8(data, position + 1) |
|
}); |
|
position += 2; |
|
} |
|
genericRegion.at = at; |
|
} |
|
args = [genericRegion, data, position, end]; |
|
break; |
|
case 48: // PageInformation |
|
var pageInfo = { |
|
width: readUint32(data, position), |
|
height: readUint32(data, position + 4), |
|
resolutionX: readUint32(data, position + 8), |
|
resolutionY: readUint32(data, position + 12) |
|
}; |
|
if (pageInfo.height === 0xFFFFFFFF) { |
|
delete pageInfo.height; |
|
} |
|
var pageSegmentFlags = data[position + 16]; |
|
var pageStripingInformation = readUint16(data, position + 17); |
|
pageInfo.lossless = !!(pageSegmentFlags & 1); |
|
pageInfo.refinement = !!(pageSegmentFlags & 2); |
|
pageInfo.defaultPixelValue = (pageSegmentFlags >> 2) & 1; |
|
pageInfo.combinationOperator = (pageSegmentFlags >> 3) & 3; |
|
pageInfo.requiresBuffer = !!(pageSegmentFlags & 32); |
|
pageInfo.combinationOperatorOverride = !!(pageSegmentFlags & 64); |
|
args = [pageInfo]; |
|
break; |
|
case 49: // EndOfPage |
|
break; |
|
case 50: // EndOfStripe |
|
break; |
|
case 51: // EndOfFile |
|
break; |
|
case 62: // 7.4.15 defines 2 extension types which |
|
// are comments and can be ignored. |
|
break; |
|
default: |
|
error('JBIG2 error: segment type ' + header.typeName + '(' + |
|
header.type + ') is not implemented'); |
|
} |
|
var callbackName = 'on' + header.typeName; |
|
if (callbackName in visitor) { |
|
visitor[callbackName].apply(visitor, args); |
|
} |
|
} |
|
|
|
function processSegments(segments, visitor) { |
|
for (var i = 0, ii = segments.length; i < ii; i++) { |
|
processSegment(segments[i], visitor); |
|
} |
|
} |
|
|
|
function parseJbig2(data, start, end) { |
|
var position = start; |
|
if (data[position] !== 0x97 || data[position + 1] !== 0x4A || |
|
data[position + 2] !== 0x42 || data[position + 3] !== 0x32 || |
|
data[position + 4] !== 0x0D || data[position + 5] !== 0x0A || |
|
data[position + 6] !== 0x1A || data[position + 7] !== 0x0A) { |
|
error('JBIG2 error: invalid header'); |
|
} |
|
var header = {}; |
|
position += 8; |
|
var flags = data[position++]; |
|
header.randomAccess = !(flags & 1); |
|
if (!(flags & 2)) { |
|
header.numberOfPages = readUint32(data, position); |
|
position += 4; |
|
} |
|
var segments = readSegments(header, data, position, end); |
|
error('Not implemented'); |
|
// processSegments(segments, new SimpleSegmentVisitor()); |
|
} |
|
|
|
function parseJbig2Chunks(chunks) { |
|
var visitor = new SimpleSegmentVisitor(); |
|
for (var i = 0, ii = chunks.length; i < ii; i++) { |
|
var chunk = chunks[i]; |
|
var segments = readSegments({}, chunk.data, chunk.start, chunk.end); |
|
processSegments(segments, visitor); |
|
} |
|
return visitor.buffer; |
|
} |
|
|
|
function SimpleSegmentVisitor() {} |
|
|
|
SimpleSegmentVisitor.prototype = { |
|
onPageInformation: function SimpleSegmentVisitor_onPageInformation(info) { |
|
this.currentPageInfo = info; |
|
var rowSize = (info.width + 7) >> 3; |
|
var buffer = new Uint8Array(rowSize * info.height); |
|
// The contents of ArrayBuffers are initialized to 0. |
|
// Fill the buffer with 0xFF only if info.defaultPixelValue is set |
|
if (info.defaultPixelValue) { |
|
for (var i = 0, ii = buffer.length; i < ii; i++) { |
|
buffer[i] = 0xFF; |
|
} |
|
} |
|
this.buffer = buffer; |
|
}, |
|
drawBitmap: function SimpleSegmentVisitor_drawBitmap(regionInfo, bitmap) { |
|
var pageInfo = this.currentPageInfo; |
|
var width = regionInfo.width, height = regionInfo.height; |
|
var rowSize = (pageInfo.width + 7) >> 3; |
|
var combinationOperator = pageInfo.combinationOperatorOverride ? |
|
regionInfo.combinationOperator : pageInfo.combinationOperator; |
|
var buffer = this.buffer; |
|
var mask0 = 128 >> (regionInfo.x & 7); |
|
var offset0 = regionInfo.y * rowSize + (regionInfo.x >> 3); |
|
var i, j, mask, offset; |
|
switch (combinationOperator) { |
|
case 0: // OR |
|
for (i = 0; i < height; i++) { |
|
mask = mask0; |
|
offset = offset0; |
|
for (j = 0; j < width; j++) { |
|
if (bitmap[i][j]) { |
|
buffer[offset] |= mask; |
|
} |
|
mask >>= 1; |
|
if (!mask) { |
|
mask = 128; |
|
offset++; |
|
} |
|
} |
|
offset0 += rowSize; |
|
} |
|
break; |
|
case 2: // XOR |
|
for (i = 0; i < height; i++) { |
|
mask = mask0; |
|
offset = offset0; |
|
for (j = 0; j < width; j++) { |
|
if (bitmap[i][j]) { |
|
buffer[offset] ^= mask; |
|
} |
|
mask >>= 1; |
|
if (!mask) { |
|
mask = 128; |
|
offset++; |
|
} |
|
} |
|
offset0 += rowSize; |
|
} |
|
break; |
|
default: |
|
error('JBIG2 error: operator ' + combinationOperator + |
|
' is not supported'); |
|
} |
|
}, |
|
onImmediateGenericRegion: |
|
function SimpleSegmentVisitor_onImmediateGenericRegion(region, data, |
|
start, end) { |
|
var regionInfo = region.info; |
|
var decodingContext = new DecodingContext(data, start, end); |
|
var bitmap = decodeBitmap(region.mmr, regionInfo.width, regionInfo.height, |
|
region.template, region.prediction, null, |
|
region.at, decodingContext); |
|
this.drawBitmap(regionInfo, bitmap); |
|
}, |
|
onImmediateLosslessGenericRegion: |
|
function SimpleSegmentVisitor_onImmediateLosslessGenericRegion() { |
|
this.onImmediateGenericRegion.apply(this, arguments); |
|
}, |
|
onSymbolDictionary: |
|
function SimpleSegmentVisitor_onSymbolDictionary(dictionary, |
|
currentSegment, |
|
referredSegments, |
|
data, start, end) { |
|
var huffmanTables; |
|
if (dictionary.huffman) { |
|
error('JBIG2 error: huffman is not supported'); |
|
} |
|
|
|
// Combines exported symbols from all referred segments |
|
var symbols = this.symbols; |
|
if (!symbols) { |
|
this.symbols = symbols = {}; |
|
} |
|
|
|
var inputSymbols = []; |
|
for (var i = 0, ii = referredSegments.length; i < ii; i++) { |
|
inputSymbols = inputSymbols.concat(symbols[referredSegments[i]]); |
|
} |
|
|
|
var decodingContext = new DecodingContext(data, start, end); |
|
symbols[currentSegment] = decodeSymbolDictionary(dictionary.huffman, |
|
dictionary.refinement, inputSymbols, dictionary.numberOfNewSymbols, |
|
dictionary.numberOfExportedSymbols, huffmanTables, |
|
dictionary.template, dictionary.at, |
|
dictionary.refinementTemplate, dictionary.refinementAt, |
|
decodingContext); |
|
}, |
|
onImmediateTextRegion: |
|
function SimpleSegmentVisitor_onImmediateTextRegion(region, |
|
referredSegments, |
|
data, start, end) { |
|
var regionInfo = region.info; |
|
var huffmanTables; |
|
|
|
// Combines exported symbols from all referred segments |
|
var symbols = this.symbols; |
|
var inputSymbols = []; |
|
for (var i = 0, ii = referredSegments.length; i < ii; i++) { |
|
inputSymbols = inputSymbols.concat(symbols[referredSegments[i]]); |
|
} |
|
var symbolCodeLength = log2(inputSymbols.length); |
|
|
|
var decodingContext = new DecodingContext(data, start, end); |
|
var bitmap = decodeTextRegion(region.huffman, region.refinement, |
|
regionInfo.width, regionInfo.height, region.defaultPixelValue, |
|
region.numberOfSymbolInstances, region.stripSize, inputSymbols, |
|
symbolCodeLength, region.transposed, region.dsOffset, |
|
region.referenceCorner, region.combinationOperator, huffmanTables, |
|
region.refinementTemplate, region.refinementAt, decodingContext); |
|
this.drawBitmap(regionInfo, bitmap); |
|
}, |
|
onImmediateLosslessTextRegion: |
|
function SimpleSegmentVisitor_onImmediateLosslessTextRegion() { |
|
this.onImmediateTextRegion.apply(this, arguments); |
|
} |
|
}; |
|
|
|
function Jbig2Image() {} |
|
|
|
Jbig2Image.prototype = { |
|
parseChunks: function Jbig2Image_parseChunks(chunks) { |
|
return parseJbig2Chunks(chunks); |
|
} |
|
}; |
|
|
|
return Jbig2Image; |
|
})(); |
|
|
|
exports.Jbig2Image = Jbig2Image; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreJpx = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCoreArithmeticDecoder); |
|
} |
|
}(this, function (exports, sharedUtil, coreArithmeticDecoder) { |
|
|
|
var info = sharedUtil.info; |
|
var log2 = sharedUtil.log2; |
|
var readUint16 = sharedUtil.readUint16; |
|
var readUint32 = sharedUtil.readUint32; |
|
var warn = sharedUtil.warn; |
|
var ArithmeticDecoder = coreArithmeticDecoder.ArithmeticDecoder; |
|
|
|
var JpxImage = (function JpxImageClosure() { |
|
// Table E.1 |
|
var SubbandsGainLog2 = { |
|
'LL': 0, |
|
'LH': 1, |
|
'HL': 1, |
|
'HH': 2 |
|
}; |
|
function JpxImage() { |
|
this.failOnCorruptedImage = false; |
|
} |
|
JpxImage.prototype = { |
|
parse: function JpxImage_parse(data) { |
|
|
|
var head = readUint16(data, 0); |
|
// No box header, immediate start of codestream (SOC) |
|
if (head === 0xFF4F) { |
|
this.parseCodestream(data, 0, data.length); |
|
return; |
|
} |
|
|
|
var position = 0, length = data.length; |
|
while (position < length) { |
|
var headerSize = 8; |
|
var lbox = readUint32(data, position); |
|
var tbox = readUint32(data, position + 4); |
|
position += headerSize; |
|
if (lbox === 1) { |
|
// XLBox: read UInt64 according to spec. |
|
// JavaScript's int precision of 53 bit should be sufficient here. |
|
lbox = readUint32(data, position) * 4294967296 + |
|
readUint32(data, position + 4); |
|
position += 8; |
|
headerSize += 8; |
|
} |
|
if (lbox === 0) { |
|
lbox = length - position + headerSize; |
|
} |
|
if (lbox < headerSize) { |
|
throw new Error('JPX Error: Invalid box field size'); |
|
} |
|
var dataLength = lbox - headerSize; |
|
var jumpDataLength = true; |
|
switch (tbox) { |
|
case 0x6A703268: // 'jp2h' |
|
jumpDataLength = false; // parsing child boxes |
|
break; |
|
case 0x636F6C72: // 'colr' |
|
// Colorspaces are not used, the CS from the PDF is used. |
|
var method = data[position]; |
|
if (method === 1) { |
|
// enumerated colorspace |
|
var colorspace = readUint32(data, position + 3); |
|
switch (colorspace) { |
|
case 16: // this indicates a sRGB colorspace |
|
case 17: // this indicates a grayscale colorspace |
|
case 18: // this indicates a YUV colorspace |
|
break; |
|
default: |
|
warn('Unknown colorspace ' + colorspace); |
|
break; |
|
} |
|
} else if (method === 2) { |
|
info('ICC profile not supported'); |
|
} |
|
break; |
|
case 0x6A703263: // 'jp2c' |
|
this.parseCodestream(data, position, position + dataLength); |
|
break; |
|
case 0x6A502020: // 'jP\024\024' |
|
if (0x0d0a870a !== readUint32(data, position)) { |
|
warn('Invalid JP2 signature'); |
|
} |
|
break; |
|
// The following header types are valid but currently not used: |
|
case 0x6A501A1A: // 'jP\032\032' |
|
case 0x66747970: // 'ftyp' |
|
case 0x72726571: // 'rreq' |
|
case 0x72657320: // 'res ' |
|
case 0x69686472: // 'ihdr' |
|
break; |
|
default: |
|
var headerType = String.fromCharCode((tbox >> 24) & 0xFF, |
|
(tbox >> 16) & 0xFF, |
|
(tbox >> 8) & 0xFF, |
|
tbox & 0xFF); |
|
warn('Unsupported header type ' + tbox + ' (' + headerType + ')'); |
|
break; |
|
} |
|
if (jumpDataLength) { |
|
position += dataLength; |
|
} |
|
} |
|
}, |
|
parseImageProperties: function JpxImage_parseImageProperties(stream) { |
|
var newByte = stream.getByte(); |
|
while (newByte >= 0) { |
|
var oldByte = newByte; |
|
newByte = stream.getByte(); |
|
var code = (oldByte << 8) | newByte; |
|
// Image and tile size (SIZ) |
|
if (code === 0xFF51) { |
|
stream.skip(4); |
|
var Xsiz = stream.getInt32() >>> 0; // Byte 4 |
|
var Ysiz = stream.getInt32() >>> 0; // Byte 8 |
|
var XOsiz = stream.getInt32() >>> 0; // Byte 12 |
|
var YOsiz = stream.getInt32() >>> 0; // Byte 16 |
|
stream.skip(16); |
|
var Csiz = stream.getUint16(); // Byte 36 |
|
this.width = Xsiz - XOsiz; |
|
this.height = Ysiz - YOsiz; |
|
this.componentsCount = Csiz; |
|
// Results are always returned as Uint8Arrays |
|
this.bitsPerComponent = 8; |
|
return; |
|
} |
|
} |
|
throw new Error('JPX Error: No size marker found in JPX stream'); |
|
}, |
|
parseCodestream: function JpxImage_parseCodestream(data, start, end) { |
|
var context = {}; |
|
try { |
|
var doNotRecover = false; |
|
var position = start; |
|
while (position + 1 < end) { |
|
var code = readUint16(data, position); |
|
position += 2; |
|
|
|
var length = 0, j, sqcd, spqcds, spqcdSize, scalarExpounded, tile; |
|
switch (code) { |
|
case 0xFF4F: // Start of codestream (SOC) |
|
context.mainHeader = true; |
|
break; |
|
case 0xFFD9: // End of codestream (EOC) |
|
break; |
|
case 0xFF51: // Image and tile size (SIZ) |
|
length = readUint16(data, position); |
|
var siz = {}; |
|
siz.Xsiz = readUint32(data, position + 4); |
|
siz.Ysiz = readUint32(data, position + 8); |
|
siz.XOsiz = readUint32(data, position + 12); |
|
siz.YOsiz = readUint32(data, position + 16); |
|
siz.XTsiz = readUint32(data, position + 20); |
|
siz.YTsiz = readUint32(data, position + 24); |
|
siz.XTOsiz = readUint32(data, position + 28); |
|
siz.YTOsiz = readUint32(data, position + 32); |
|
var componentsCount = readUint16(data, position + 36); |
|
siz.Csiz = componentsCount; |
|
var components = []; |
|
j = position + 38; |
|
for (var i = 0; i < componentsCount; i++) { |
|
var component = { |
|
precision: (data[j] & 0x7F) + 1, |
|
isSigned: !!(data[j] & 0x80), |
|
XRsiz: data[j + 1], |
|
YRsiz: data[j + 1] |
|
}; |
|
calculateComponentDimensions(component, siz); |
|
components.push(component); |
|
} |
|
context.SIZ = siz; |
|
context.components = components; |
|
calculateTileGrids(context, components); |
|
context.QCC = []; |
|
context.COC = []; |
|
break; |
|
case 0xFF5C: // Quantization default (QCD) |
|
length = readUint16(data, position); |
|
var qcd = {}; |
|
j = position + 2; |
|
sqcd = data[j++]; |
|
switch (sqcd & 0x1F) { |
|
case 0: |
|
spqcdSize = 8; |
|
scalarExpounded = true; |
|
break; |
|
case 1: |
|
spqcdSize = 16; |
|
scalarExpounded = false; |
|
break; |
|
case 2: |
|
spqcdSize = 16; |
|
scalarExpounded = true; |
|
break; |
|
default: |
|
throw new Error('JPX Error: Invalid SQcd value ' + sqcd); |
|
} |
|
qcd.noQuantization = (spqcdSize === 8); |
|
qcd.scalarExpounded = scalarExpounded; |
|
qcd.guardBits = sqcd >> 5; |
|
spqcds = []; |
|
while (j < length + position) { |
|
var spqcd = {}; |
|
if (spqcdSize === 8) { |
|
spqcd.epsilon = data[j++] >> 3; |
|
spqcd.mu = 0; |
|
} else { |
|
spqcd.epsilon = data[j] >> 3; |
|
spqcd.mu = ((data[j] & 0x7) << 8) | data[j + 1]; |
|
j += 2; |
|
} |
|
spqcds.push(spqcd); |
|
} |
|
qcd.SPqcds = spqcds; |
|
if (context.mainHeader) { |
|
context.QCD = qcd; |
|
} else { |
|
context.currentTile.QCD = qcd; |
|
context.currentTile.QCC = []; |
|
} |
|
break; |
|
case 0xFF5D: // Quantization component (QCC) |
|
length = readUint16(data, position); |
|
var qcc = {}; |
|
j = position + 2; |
|
var cqcc; |
|
if (context.SIZ.Csiz < 257) { |
|
cqcc = data[j++]; |
|
} else { |
|
cqcc = readUint16(data, j); |
|
j += 2; |
|
} |
|
sqcd = data[j++]; |
|
switch (sqcd & 0x1F) { |
|
case 0: |
|
spqcdSize = 8; |
|
scalarExpounded = true; |
|
break; |
|
case 1: |
|
spqcdSize = 16; |
|
scalarExpounded = false; |
|
break; |
|
case 2: |
|
spqcdSize = 16; |
|
scalarExpounded = true; |
|
break; |
|
default: |
|
throw new Error('JPX Error: Invalid SQcd value ' + sqcd); |
|
} |
|
qcc.noQuantization = (spqcdSize === 8); |
|
qcc.scalarExpounded = scalarExpounded; |
|
qcc.guardBits = sqcd >> 5; |
|
spqcds = []; |
|
while (j < (length + position)) { |
|
spqcd = {}; |
|
if (spqcdSize === 8) { |
|
spqcd.epsilon = data[j++] >> 3; |
|
spqcd.mu = 0; |
|
} else { |
|
spqcd.epsilon = data[j] >> 3; |
|
spqcd.mu = ((data[j] & 0x7) << 8) | data[j + 1]; |
|
j += 2; |
|
} |
|
spqcds.push(spqcd); |
|
} |
|
qcc.SPqcds = spqcds; |
|
if (context.mainHeader) { |
|
context.QCC[cqcc] = qcc; |
|
} else { |
|
context.currentTile.QCC[cqcc] = qcc; |
|
} |
|
break; |
|
case 0xFF52: // Coding style default (COD) |
|
length = readUint16(data, position); |
|
var cod = {}; |
|
j = position + 2; |
|
var scod = data[j++]; |
|
cod.entropyCoderWithCustomPrecincts = !!(scod & 1); |
|
cod.sopMarkerUsed = !!(scod & 2); |
|
cod.ephMarkerUsed = !!(scod & 4); |
|
cod.progressionOrder = data[j++]; |
|
cod.layersCount = readUint16(data, j); |
|
j += 2; |
|
cod.multipleComponentTransform = data[j++]; |
|
|
|
cod.decompositionLevelsCount = data[j++]; |
|
cod.xcb = (data[j++] & 0xF) + 2; |
|
cod.ycb = (data[j++] & 0xF) + 2; |
|
var blockStyle = data[j++]; |
|
cod.selectiveArithmeticCodingBypass = !!(blockStyle & 1); |
|
cod.resetContextProbabilities = !!(blockStyle & 2); |
|
cod.terminationOnEachCodingPass = !!(blockStyle & 4); |
|
cod.verticalyStripe = !!(blockStyle & 8); |
|
cod.predictableTermination = !!(blockStyle & 16); |
|
cod.segmentationSymbolUsed = !!(blockStyle & 32); |
|
cod.reversibleTransformation = data[j++]; |
|
if (cod.entropyCoderWithCustomPrecincts) { |
|
var precinctsSizes = []; |
|
while (j < length + position) { |
|
var precinctsSize = data[j++]; |
|
precinctsSizes.push({ |
|
PPx: precinctsSize & 0xF, |
|
PPy: precinctsSize >> 4 |
|
}); |
|
} |
|
cod.precinctsSizes = precinctsSizes; |
|
} |
|
var unsupported = []; |
|
if (cod.selectiveArithmeticCodingBypass) { |
|
unsupported.push('selectiveArithmeticCodingBypass'); |
|
} |
|
if (cod.resetContextProbabilities) { |
|
unsupported.push('resetContextProbabilities'); |
|
} |
|
if (cod.terminationOnEachCodingPass) { |
|
unsupported.push('terminationOnEachCodingPass'); |
|
} |
|
if (cod.verticalyStripe) { |
|
unsupported.push('verticalyStripe'); |
|
} |
|
if (cod.predictableTermination) { |
|
unsupported.push('predictableTermination'); |
|
} |
|
if (unsupported.length > 0) { |
|
doNotRecover = true; |
|
throw new Error('JPX Error: Unsupported COD options (' + |
|
unsupported.join(', ') + ')'); |
|
} |
|
if (context.mainHeader) { |
|
context.COD = cod; |
|
} else { |
|
context.currentTile.COD = cod; |
|
context.currentTile.COC = []; |
|
} |
|
break; |
|
case 0xFF90: // Start of tile-part (SOT) |
|
length = readUint16(data, position); |
|
tile = {}; |
|
tile.index = readUint16(data, position + 2); |
|
tile.length = readUint32(data, position + 4); |
|
tile.dataEnd = tile.length + position - 2; |
|
tile.partIndex = data[position + 8]; |
|
tile.partsCount = data[position + 9]; |
|
|
|
context.mainHeader = false; |
|
if (tile.partIndex === 0) { |
|
// reset component specific settings |
|
tile.COD = context.COD; |
|
tile.COC = context.COC.slice(0); // clone of the global COC |
|
tile.QCD = context.QCD; |
|
tile.QCC = context.QCC.slice(0); // clone of the global COC |
|
} |
|
context.currentTile = tile; |
|
break; |
|
case 0xFF93: // Start of data (SOD) |
|
tile = context.currentTile; |
|
if (tile.partIndex === 0) { |
|
initializeTile(context, tile.index); |
|
buildPackets(context); |
|
} |
|
|
|
// moving to the end of the data |
|
length = tile.dataEnd - position; |
|
parseTilePackets(context, data, position, length); |
|
break; |
|
case 0xFF55: // Tile-part lengths, main header (TLM) |
|
case 0xFF57: // Packet length, main header (PLM) |
|
case 0xFF58: // Packet length, tile-part header (PLT) |
|
case 0xFF64: // Comment (COM) |
|
length = readUint16(data, position); |
|
// skipping content |
|
break; |
|
case 0xFF53: // Coding style component (COC) |
|
throw new Error('JPX Error: Codestream code 0xFF53 (COC) is ' + |
|
'not implemented'); |
|
default: |
|
throw new Error('JPX Error: Unknown codestream code: ' + |
|
code.toString(16)); |
|
} |
|
position += length; |
|
} |
|
} catch (e) { |
|
if (doNotRecover || this.failOnCorruptedImage) { |
|
throw e; |
|
} else { |
|
warn('Trying to recover from ' + e.message); |
|
} |
|
} |
|
this.tiles = transformComponents(context); |
|
this.width = context.SIZ.Xsiz - context.SIZ.XOsiz; |
|
this.height = context.SIZ.Ysiz - context.SIZ.YOsiz; |
|
this.componentsCount = context.SIZ.Csiz; |
|
} |
|
}; |
|
function calculateComponentDimensions(component, siz) { |
|
// Section B.2 Component mapping |
|
component.x0 = Math.ceil(siz.XOsiz / component.XRsiz); |
|
component.x1 = Math.ceil(siz.Xsiz / component.XRsiz); |
|
component.y0 = Math.ceil(siz.YOsiz / component.YRsiz); |
|
component.y1 = Math.ceil(siz.Ysiz / component.YRsiz); |
|
component.width = component.x1 - component.x0; |
|
component.height = component.y1 - component.y0; |
|
} |
|
function calculateTileGrids(context, components) { |
|
var siz = context.SIZ; |
|
// Section B.3 Division into tile and tile-components |
|
var tile, tiles = []; |
|
var numXtiles = Math.ceil((siz.Xsiz - siz.XTOsiz) / siz.XTsiz); |
|
var numYtiles = Math.ceil((siz.Ysiz - siz.YTOsiz) / siz.YTsiz); |
|
for (var q = 0; q < numYtiles; q++) { |
|
for (var p = 0; p < numXtiles; p++) { |
|
tile = {}; |
|
tile.tx0 = Math.max(siz.XTOsiz + p * siz.XTsiz, siz.XOsiz); |
|
tile.ty0 = Math.max(siz.YTOsiz + q * siz.YTsiz, siz.YOsiz); |
|
tile.tx1 = Math.min(siz.XTOsiz + (p + 1) * siz.XTsiz, siz.Xsiz); |
|
tile.ty1 = Math.min(siz.YTOsiz + (q + 1) * siz.YTsiz, siz.Ysiz); |
|
tile.width = tile.tx1 - tile.tx0; |
|
tile.height = tile.ty1 - tile.ty0; |
|
tile.components = []; |
|
tiles.push(tile); |
|
} |
|
} |
|
context.tiles = tiles; |
|
|
|
var componentsCount = siz.Csiz; |
|
for (var i = 0, ii = componentsCount; i < ii; i++) { |
|
var component = components[i]; |
|
for (var j = 0, jj = tiles.length; j < jj; j++) { |
|
var tileComponent = {}; |
|
tile = tiles[j]; |
|
tileComponent.tcx0 = Math.ceil(tile.tx0 / component.XRsiz); |
|
tileComponent.tcy0 = Math.ceil(tile.ty0 / component.YRsiz); |
|
tileComponent.tcx1 = Math.ceil(tile.tx1 / component.XRsiz); |
|
tileComponent.tcy1 = Math.ceil(tile.ty1 / component.YRsiz); |
|
tileComponent.width = tileComponent.tcx1 - tileComponent.tcx0; |
|
tileComponent.height = tileComponent.tcy1 - tileComponent.tcy0; |
|
tile.components[i] = tileComponent; |
|
} |
|
} |
|
} |
|
function getBlocksDimensions(context, component, r) { |
|
var codOrCoc = component.codingStyleParameters; |
|
var result = {}; |
|
if (!codOrCoc.entropyCoderWithCustomPrecincts) { |
|
result.PPx = 15; |
|
result.PPy = 15; |
|
} else { |
|
result.PPx = codOrCoc.precinctsSizes[r].PPx; |
|
result.PPy = codOrCoc.precinctsSizes[r].PPy; |
|
} |
|
// calculate codeblock size as described in section B.7 |
|
result.xcb_ = (r > 0 ? Math.min(codOrCoc.xcb, result.PPx - 1) : |
|
Math.min(codOrCoc.xcb, result.PPx)); |
|
result.ycb_ = (r > 0 ? Math.min(codOrCoc.ycb, result.PPy - 1) : |
|
Math.min(codOrCoc.ycb, result.PPy)); |
|
return result; |
|
} |
|
function buildPrecincts(context, resolution, dimensions) { |
|
// Section B.6 Division resolution to precincts |
|
var precinctWidth = 1 << dimensions.PPx; |
|
var precinctHeight = 1 << dimensions.PPy; |
|
// Jasper introduces codeblock groups for mapping each subband codeblocks |
|
// to precincts. Precinct partition divides a resolution according to width |
|
// and height parameters. The subband that belongs to the resolution level |
|
// has a different size than the level, unless it is the zero resolution. |
|
|
|
// From Jasper documentation: jpeg2000.pdf, section K: Tier-2 coding: |
|
// The precinct partitioning for a particular subband is derived from a |
|
// partitioning of its parent LL band (i.e., the LL band at the next higher |
|
// resolution level)... The LL band associated with each resolution level is |
|
// divided into precincts... Each of the resulting precinct regions is then |
|
// mapped into its child subbands (if any) at the next lower resolution |
|
// level. This is accomplished by using the coordinate transformation |
|
// (u, v) = (ceil(x/2), ceil(y/2)) where (x, y) and (u, v) are the |
|
// coordinates of a point in the LL band and child subband, respectively. |
|
var isZeroRes = resolution.resLevel === 0; |
|
var precinctWidthInSubband = 1 << (dimensions.PPx + (isZeroRes ? 0 : -1)); |
|
var precinctHeightInSubband = 1 << (dimensions.PPy + (isZeroRes ? 0 : -1)); |
|
var numprecinctswide = (resolution.trx1 > resolution.trx0 ? |
|
Math.ceil(resolution.trx1 / precinctWidth) - |
|
Math.floor(resolution.trx0 / precinctWidth) : 0); |
|
var numprecinctshigh = (resolution.try1 > resolution.try0 ? |
|
Math.ceil(resolution.try1 / precinctHeight) - |
|
Math.floor(resolution.try0 / precinctHeight) : 0); |
|
var numprecincts = numprecinctswide * numprecinctshigh; |
|
|
|
resolution.precinctParameters = { |
|
precinctWidth: precinctWidth, |
|
precinctHeight: precinctHeight, |
|
numprecinctswide: numprecinctswide, |
|
numprecinctshigh: numprecinctshigh, |
|
numprecincts: numprecincts, |
|
precinctWidthInSubband: precinctWidthInSubband, |
|
precinctHeightInSubband: precinctHeightInSubband |
|
}; |
|
} |
|
function buildCodeblocks(context, subband, dimensions) { |
|
// Section B.7 Division sub-band into code-blocks |
|
var xcb_ = dimensions.xcb_; |
|
var ycb_ = dimensions.ycb_; |
|
var codeblockWidth = 1 << xcb_; |
|
var codeblockHeight = 1 << ycb_; |
|
var cbx0 = subband.tbx0 >> xcb_; |
|
var cby0 = subband.tby0 >> ycb_; |
|
var cbx1 = (subband.tbx1 + codeblockWidth - 1) >> xcb_; |
|
var cby1 = (subband.tby1 + codeblockHeight - 1) >> ycb_; |
|
var precinctParameters = subband.resolution.precinctParameters; |
|
var codeblocks = []; |
|
var precincts = []; |
|
var i, j, codeblock, precinctNumber; |
|
for (j = cby0; j < cby1; j++) { |
|
for (i = cbx0; i < cbx1; i++) { |
|
codeblock = { |
|
cbx: i, |
|
cby: j, |
|
tbx0: codeblockWidth * i, |
|
tby0: codeblockHeight * j, |
|
tbx1: codeblockWidth * (i + 1), |
|
tby1: codeblockHeight * (j + 1) |
|
}; |
|
|
|
codeblock.tbx0_ = Math.max(subband.tbx0, codeblock.tbx0); |
|
codeblock.tby0_ = Math.max(subband.tby0, codeblock.tby0); |
|
codeblock.tbx1_ = Math.min(subband.tbx1, codeblock.tbx1); |
|
codeblock.tby1_ = Math.min(subband.tby1, codeblock.tby1); |
|
|
|
// Calculate precinct number for this codeblock, codeblock position |
|
// should be relative to its subband, use actual dimension and position |
|
// See comment about codeblock group width and height |
|
var pi = Math.floor((codeblock.tbx0_ - subband.tbx0) / |
|
precinctParameters.precinctWidthInSubband); |
|
var pj = Math.floor((codeblock.tby0_ - subband.tby0) / |
|
precinctParameters.precinctHeightInSubband); |
|
precinctNumber = pi + (pj * precinctParameters.numprecinctswide); |
|
|
|
codeblock.precinctNumber = precinctNumber; |
|
codeblock.subbandType = subband.type; |
|
codeblock.Lblock = 3; |
|
|
|
if (codeblock.tbx1_ <= codeblock.tbx0_ || |
|
codeblock.tby1_ <= codeblock.tby0_) { |
|
continue; |
|
} |
|
codeblocks.push(codeblock); |
|
// building precinct for the sub-band |
|
var precinct = precincts[precinctNumber]; |
|
if (precinct !== undefined) { |
|
if (i < precinct.cbxMin) { |
|
precinct.cbxMin = i; |
|
} else if (i > precinct.cbxMax) { |
|
precinct.cbxMax = i; |
|
} |
|
if (j < precinct.cbyMin) { |
|
precinct.cbxMin = j; |
|
} else if (j > precinct.cbyMax) { |
|
precinct.cbyMax = j; |
|
} |
|
} else { |
|
precincts[precinctNumber] = precinct = { |
|
cbxMin: i, |
|
cbyMin: j, |
|
cbxMax: i, |
|
cbyMax: j |
|
}; |
|
} |
|
codeblock.precinct = precinct; |
|
} |
|
} |
|
subband.codeblockParameters = { |
|
codeblockWidth: xcb_, |
|
codeblockHeight: ycb_, |
|
numcodeblockwide: cbx1 - cbx0 + 1, |
|
numcodeblockhigh: cby1 - cby0 + 1 |
|
}; |
|
subband.codeblocks = codeblocks; |
|
subband.precincts = precincts; |
|
} |
|
function createPacket(resolution, precinctNumber, layerNumber) { |
|
var precinctCodeblocks = []; |
|
// Section B.10.8 Order of info in packet |
|
var subbands = resolution.subbands; |
|
// sub-bands already ordered in 'LL', 'HL', 'LH', and 'HH' sequence |
|
for (var i = 0, ii = subbands.length; i < ii; i++) { |
|
var subband = subbands[i]; |
|
var codeblocks = subband.codeblocks; |
|
for (var j = 0, jj = codeblocks.length; j < jj; j++) { |
|
var codeblock = codeblocks[j]; |
|
if (codeblock.precinctNumber !== precinctNumber) { |
|
continue; |
|
} |
|
precinctCodeblocks.push(codeblock); |
|
} |
|
} |
|
return { |
|
layerNumber: layerNumber, |
|
codeblocks: precinctCodeblocks |
|
}; |
|
} |
|
function LayerResolutionComponentPositionIterator(context) { |
|
var siz = context.SIZ; |
|
var tileIndex = context.currentTile.index; |
|
var tile = context.tiles[tileIndex]; |
|
var layersCount = tile.codingStyleDefaultParameters.layersCount; |
|
var componentsCount = siz.Csiz; |
|
var maxDecompositionLevelsCount = 0; |
|
for (var q = 0; q < componentsCount; q++) { |
|
maxDecompositionLevelsCount = Math.max(maxDecompositionLevelsCount, |
|
tile.components[q].codingStyleParameters.decompositionLevelsCount); |
|
} |
|
|
|
var l = 0, r = 0, i = 0, k = 0; |
|
|
|
this.nextPacket = function JpxImage_nextPacket() { |
|
// Section B.12.1.1 Layer-resolution-component-position |
|
for (; l < layersCount; l++) { |
|
for (; r <= maxDecompositionLevelsCount; r++) { |
|
for (; i < componentsCount; i++) { |
|
var component = tile.components[i]; |
|
if (r > component.codingStyleParameters.decompositionLevelsCount) { |
|
continue; |
|
} |
|
|
|
var resolution = component.resolutions[r]; |
|
var numprecincts = resolution.precinctParameters.numprecincts; |
|
for (; k < numprecincts;) { |
|
var packet = createPacket(resolution, k, l); |
|
k++; |
|
return packet; |
|
} |
|
k = 0; |
|
} |
|
i = 0; |
|
} |
|
r = 0; |
|
} |
|
throw new Error('JPX Error: Out of packets'); |
|
}; |
|
} |
|
function ResolutionLayerComponentPositionIterator(context) { |
|
var siz = context.SIZ; |
|
var tileIndex = context.currentTile.index; |
|
var tile = context.tiles[tileIndex]; |
|
var layersCount = tile.codingStyleDefaultParameters.layersCount; |
|
var componentsCount = siz.Csiz; |
|
var maxDecompositionLevelsCount = 0; |
|
for (var q = 0; q < componentsCount; q++) { |
|
maxDecompositionLevelsCount = Math.max(maxDecompositionLevelsCount, |
|
tile.components[q].codingStyleParameters.decompositionLevelsCount); |
|
} |
|
|
|
var r = 0, l = 0, i = 0, k = 0; |
|
|
|
this.nextPacket = function JpxImage_nextPacket() { |
|
// Section B.12.1.2 Resolution-layer-component-position |
|
for (; r <= maxDecompositionLevelsCount; r++) { |
|
for (; l < layersCount; l++) { |
|
for (; i < componentsCount; i++) { |
|
var component = tile.components[i]; |
|
if (r > component.codingStyleParameters.decompositionLevelsCount) { |
|
continue; |
|
} |
|
|
|
var resolution = component.resolutions[r]; |
|
var numprecincts = resolution.precinctParameters.numprecincts; |
|
for (; k < numprecincts;) { |
|
var packet = createPacket(resolution, k, l); |
|
k++; |
|
return packet; |
|
} |
|
k = 0; |
|
} |
|
i = 0; |
|
} |
|
l = 0; |
|
} |
|
throw new Error('JPX Error: Out of packets'); |
|
}; |
|
} |
|
function ResolutionPositionComponentLayerIterator(context) { |
|
var siz = context.SIZ; |
|
var tileIndex = context.currentTile.index; |
|
var tile = context.tiles[tileIndex]; |
|
var layersCount = tile.codingStyleDefaultParameters.layersCount; |
|
var componentsCount = siz.Csiz; |
|
var l, r, c, p; |
|
var maxDecompositionLevelsCount = 0; |
|
for (c = 0; c < componentsCount; c++) { |
|
var component = tile.components[c]; |
|
maxDecompositionLevelsCount = Math.max(maxDecompositionLevelsCount, |
|
component.codingStyleParameters.decompositionLevelsCount); |
|
} |
|
var maxNumPrecinctsInLevel = new Int32Array( |
|
maxDecompositionLevelsCount + 1); |
|
for (r = 0; r <= maxDecompositionLevelsCount; ++r) { |
|
var maxNumPrecincts = 0; |
|
for (c = 0; c < componentsCount; ++c) { |
|
var resolutions = tile.components[c].resolutions; |
|
if (r < resolutions.length) { |
|
maxNumPrecincts = Math.max(maxNumPrecincts, |
|
resolutions[r].precinctParameters.numprecincts); |
|
} |
|
} |
|
maxNumPrecinctsInLevel[r] = maxNumPrecincts; |
|
} |
|
l = 0; |
|
r = 0; |
|
c = 0; |
|
p = 0; |
|
|
|
this.nextPacket = function JpxImage_nextPacket() { |
|
// Section B.12.1.3 Resolution-position-component-layer |
|
for (; r <= maxDecompositionLevelsCount; r++) { |
|
for (; p < maxNumPrecinctsInLevel[r]; p++) { |
|
for (; c < componentsCount; c++) { |
|
var component = tile.components[c]; |
|
if (r > component.codingStyleParameters.decompositionLevelsCount) { |
|
continue; |
|
} |
|
var resolution = component.resolutions[r]; |
|
var numprecincts = resolution.precinctParameters.numprecincts; |
|
if (p >= numprecincts) { |
|
continue; |
|
} |
|
for (; l < layersCount;) { |
|
var packet = createPacket(resolution, p, l); |
|
l++; |
|
return packet; |
|
} |
|
l = 0; |
|
} |
|
c = 0; |
|
} |
|
p = 0; |
|
} |
|
throw new Error('JPX Error: Out of packets'); |
|
}; |
|
} |
|
function PositionComponentResolutionLayerIterator(context) { |
|
var siz = context.SIZ; |
|
var tileIndex = context.currentTile.index; |
|
var tile = context.tiles[tileIndex]; |
|
var layersCount = tile.codingStyleDefaultParameters.layersCount; |
|
var componentsCount = siz.Csiz; |
|
var precinctsSizes = getPrecinctSizesInImageScale(tile); |
|
var precinctsIterationSizes = precinctsSizes; |
|
var l = 0, r = 0, c = 0, px = 0, py = 0; |
|
|
|
this.nextPacket = function JpxImage_nextPacket() { |
|
// Section B.12.1.4 Position-component-resolution-layer |
|
for (; py < precinctsIterationSizes.maxNumHigh; py++) { |
|
for (; px < precinctsIterationSizes.maxNumWide; px++) { |
|
for (; c < componentsCount; c++) { |
|
var component = tile.components[c]; |
|
var decompositionLevelsCount = |
|
component.codingStyleParameters.decompositionLevelsCount; |
|
for (; r <= decompositionLevelsCount; r++) { |
|
var resolution = component.resolutions[r]; |
|
var sizeInImageScale = |
|
precinctsSizes.components[c].resolutions[r]; |
|
var k = getPrecinctIndexIfExist( |
|
px, |
|
py, |
|
sizeInImageScale, |
|
precinctsIterationSizes, |
|
resolution); |
|
if (k === null) { |
|
continue; |
|
} |
|
for (; l < layersCount;) { |
|
var packet = createPacket(resolution, k, l); |
|
l++; |
|
return packet; |
|
} |
|
l = 0; |
|
} |
|
r = 0; |
|
} |
|
c = 0; |
|
} |
|
px = 0; |
|
} |
|
throw new Error('JPX Error: Out of packets'); |
|
}; |
|
} |
|
function ComponentPositionResolutionLayerIterator(context) { |
|
var siz = context.SIZ; |
|
var tileIndex = context.currentTile.index; |
|
var tile = context.tiles[tileIndex]; |
|
var layersCount = tile.codingStyleDefaultParameters.layersCount; |
|
var componentsCount = siz.Csiz; |
|
var precinctsSizes = getPrecinctSizesInImageScale(tile); |
|
var l = 0, r = 0, c = 0, px = 0, py = 0; |
|
|
|
this.nextPacket = function JpxImage_nextPacket() { |
|
// Section B.12.1.5 Component-position-resolution-layer |
|
for (; c < componentsCount; ++c) { |
|
var component = tile.components[c]; |
|
var precinctsIterationSizes = precinctsSizes.components[c]; |
|
var decompositionLevelsCount = |
|
component.codingStyleParameters.decompositionLevelsCount; |
|
for (; py < precinctsIterationSizes.maxNumHigh; py++) { |
|
for (; px < precinctsIterationSizes.maxNumWide; px++) { |
|
for (; r <= decompositionLevelsCount; r++) { |
|
var resolution = component.resolutions[r]; |
|
var sizeInImageScale = precinctsIterationSizes.resolutions[r]; |
|
var k = getPrecinctIndexIfExist( |
|
px, |
|
py, |
|
sizeInImageScale, |
|
precinctsIterationSizes, |
|
resolution); |
|
if (k === null) { |
|
continue; |
|
} |
|
for (; l < layersCount;) { |
|
var packet = createPacket(resolution, k, l); |
|
l++; |
|
return packet; |
|
} |
|
l = 0; |
|
} |
|
r = 0; |
|
} |
|
px = 0; |
|
} |
|
py = 0; |
|
} |
|
throw new Error('JPX Error: Out of packets'); |
|
}; |
|
} |
|
function getPrecinctIndexIfExist( |
|
pxIndex, pyIndex, sizeInImageScale, precinctIterationSizes, resolution) { |
|
var posX = pxIndex * precinctIterationSizes.minWidth; |
|
var posY = pyIndex * precinctIterationSizes.minHeight; |
|
if (posX % sizeInImageScale.width !== 0 || |
|
posY % sizeInImageScale.height !== 0) { |
|
return null; |
|
} |
|
var startPrecinctRowIndex = |
|
(posY / sizeInImageScale.width) * |
|
resolution.precinctParameters.numprecinctswide; |
|
return (posX / sizeInImageScale.height) + startPrecinctRowIndex; |
|
} |
|
function getPrecinctSizesInImageScale(tile) { |
|
var componentsCount = tile.components.length; |
|
var minWidth = Number.MAX_VALUE; |
|
var minHeight = Number.MAX_VALUE; |
|
var maxNumWide = 0; |
|
var maxNumHigh = 0; |
|
var sizePerComponent = new Array(componentsCount); |
|
for (var c = 0; c < componentsCount; c++) { |
|
var component = tile.components[c]; |
|
var decompositionLevelsCount = |
|
component.codingStyleParameters.decompositionLevelsCount; |
|
var sizePerResolution = new Array(decompositionLevelsCount + 1); |
|
var minWidthCurrentComponent = Number.MAX_VALUE; |
|
var minHeightCurrentComponent = Number.MAX_VALUE; |
|
var maxNumWideCurrentComponent = 0; |
|
var maxNumHighCurrentComponent = 0; |
|
var scale = 1; |
|
for (var r = decompositionLevelsCount; r >= 0; --r) { |
|
var resolution = component.resolutions[r]; |
|
var widthCurrentResolution = |
|
scale * resolution.precinctParameters.precinctWidth; |
|
var heightCurrentResolution = |
|
scale * resolution.precinctParameters.precinctHeight; |
|
minWidthCurrentComponent = Math.min( |
|
minWidthCurrentComponent, |
|
widthCurrentResolution); |
|
minHeightCurrentComponent = Math.min( |
|
minHeightCurrentComponent, |
|
heightCurrentResolution); |
|
maxNumWideCurrentComponent = Math.max(maxNumWideCurrentComponent, |
|
resolution.precinctParameters.numprecinctswide); |
|
maxNumHighCurrentComponent = Math.max(maxNumHighCurrentComponent, |
|
resolution.precinctParameters.numprecinctshigh); |
|
sizePerResolution[r] = { |
|
width: widthCurrentResolution, |
|
height: heightCurrentResolution |
|
}; |
|
scale <<= 1; |
|
} |
|
minWidth = Math.min(minWidth, minWidthCurrentComponent); |
|
minHeight = Math.min(minHeight, minHeightCurrentComponent); |
|
maxNumWide = Math.max(maxNumWide, maxNumWideCurrentComponent); |
|
maxNumHigh = Math.max(maxNumHigh, maxNumHighCurrentComponent); |
|
sizePerComponent[c] = { |
|
resolutions: sizePerResolution, |
|
minWidth: minWidthCurrentComponent, |
|
minHeight: minHeightCurrentComponent, |
|
maxNumWide: maxNumWideCurrentComponent, |
|
maxNumHigh: maxNumHighCurrentComponent |
|
}; |
|
} |
|
return { |
|
components: sizePerComponent, |
|
minWidth: minWidth, |
|
minHeight: minHeight, |
|
maxNumWide: maxNumWide, |
|
maxNumHigh: maxNumHigh |
|
}; |
|
} |
|
function buildPackets(context) { |
|
var siz = context.SIZ; |
|
var tileIndex = context.currentTile.index; |
|
var tile = context.tiles[tileIndex]; |
|
var componentsCount = siz.Csiz; |
|
// Creating resolutions and sub-bands for each component |
|
for (var c = 0; c < componentsCount; c++) { |
|
var component = tile.components[c]; |
|
var decompositionLevelsCount = |
|
component.codingStyleParameters.decompositionLevelsCount; |
|
// Section B.5 Resolution levels and sub-bands |
|
var resolutions = []; |
|
var subbands = []; |
|
for (var r = 0; r <= decompositionLevelsCount; r++) { |
|
var blocksDimensions = getBlocksDimensions(context, component, r); |
|
var resolution = {}; |
|
var scale = 1 << (decompositionLevelsCount - r); |
|
resolution.trx0 = Math.ceil(component.tcx0 / scale); |
|
resolution.try0 = Math.ceil(component.tcy0 / scale); |
|
resolution.trx1 = Math.ceil(component.tcx1 / scale); |
|
resolution.try1 = Math.ceil(component.tcy1 / scale); |
|
resolution.resLevel = r; |
|
buildPrecincts(context, resolution, blocksDimensions); |
|
resolutions.push(resolution); |
|
|
|
var subband; |
|
if (r === 0) { |
|
// one sub-band (LL) with last decomposition |
|
subband = {}; |
|
subband.type = 'LL'; |
|
subband.tbx0 = Math.ceil(component.tcx0 / scale); |
|
subband.tby0 = Math.ceil(component.tcy0 / scale); |
|
subband.tbx1 = Math.ceil(component.tcx1 / scale); |
|
subband.tby1 = Math.ceil(component.tcy1 / scale); |
|
subband.resolution = resolution; |
|
buildCodeblocks(context, subband, blocksDimensions); |
|
subbands.push(subband); |
|
resolution.subbands = [subband]; |
|
} else { |
|
var bscale = 1 << (decompositionLevelsCount - r + 1); |
|
var resolutionSubbands = []; |
|
// three sub-bands (HL, LH and HH) with rest of decompositions |
|
subband = {}; |
|
subband.type = 'HL'; |
|
subband.tbx0 = Math.ceil(component.tcx0 / bscale - 0.5); |
|
subband.tby0 = Math.ceil(component.tcy0 / bscale); |
|
subband.tbx1 = Math.ceil(component.tcx1 / bscale - 0.5); |
|
subband.tby1 = Math.ceil(component.tcy1 / bscale); |
|
subband.resolution = resolution; |
|
buildCodeblocks(context, subband, blocksDimensions); |
|
subbands.push(subband); |
|
resolutionSubbands.push(subband); |
|
|
|
subband = {}; |
|
subband.type = 'LH'; |
|
subband.tbx0 = Math.ceil(component.tcx0 / bscale); |
|
subband.tby0 = Math.ceil(component.tcy0 / bscale - 0.5); |
|
subband.tbx1 = Math.ceil(component.tcx1 / bscale); |
|
subband.tby1 = Math.ceil(component.tcy1 / bscale - 0.5); |
|
subband.resolution = resolution; |
|
buildCodeblocks(context, subband, blocksDimensions); |
|
subbands.push(subband); |
|
resolutionSubbands.push(subband); |
|
|
|
subband = {}; |
|
subband.type = 'HH'; |
|
subband.tbx0 = Math.ceil(component.tcx0 / bscale - 0.5); |
|
subband.tby0 = Math.ceil(component.tcy0 / bscale - 0.5); |
|
subband.tbx1 = Math.ceil(component.tcx1 / bscale - 0.5); |
|
subband.tby1 = Math.ceil(component.tcy1 / bscale - 0.5); |
|
subband.resolution = resolution; |
|
buildCodeblocks(context, subband, blocksDimensions); |
|
subbands.push(subband); |
|
resolutionSubbands.push(subband); |
|
|
|
resolution.subbands = resolutionSubbands; |
|
} |
|
} |
|
component.resolutions = resolutions; |
|
component.subbands = subbands; |
|
} |
|
// Generate the packets sequence |
|
var progressionOrder = tile.codingStyleDefaultParameters.progressionOrder; |
|
switch (progressionOrder) { |
|
case 0: |
|
tile.packetsIterator = |
|
new LayerResolutionComponentPositionIterator(context); |
|
break; |
|
case 1: |
|
tile.packetsIterator = |
|
new ResolutionLayerComponentPositionIterator(context); |
|
break; |
|
case 2: |
|
tile.packetsIterator = |
|
new ResolutionPositionComponentLayerIterator(context); |
|
break; |
|
case 3: |
|
tile.packetsIterator = |
|
new PositionComponentResolutionLayerIterator(context); |
|
break; |
|
case 4: |
|
tile.packetsIterator = |
|
new ComponentPositionResolutionLayerIterator(context); |
|
break; |
|
default: |
|
throw new Error('JPX Error: Unsupported progression order ' + |
|
progressionOrder); |
|
} |
|
} |
|
function parseTilePackets(context, data, offset, dataLength) { |
|
var position = 0; |
|
var buffer, bufferSize = 0, skipNextBit = false; |
|
function readBits(count) { |
|
while (bufferSize < count) { |
|
var b = data[offset + position]; |
|
position++; |
|
if (skipNextBit) { |
|
buffer = (buffer << 7) | b; |
|
bufferSize += 7; |
|
skipNextBit = false; |
|
} else { |
|
buffer = (buffer << 8) | b; |
|
bufferSize += 8; |
|
} |
|
if (b === 0xFF) { |
|
skipNextBit = true; |
|
} |
|
} |
|
bufferSize -= count; |
|
return (buffer >>> bufferSize) & ((1 << count) - 1); |
|
} |
|
function skipMarkerIfEqual(value) { |
|
if (data[offset + position - 1] === 0xFF && |
|
data[offset + position] === value) { |
|
skipBytes(1); |
|
return true; |
|
} else if (data[offset + position] === 0xFF && |
|
data[offset + position + 1] === value) { |
|
skipBytes(2); |
|
return true; |
|
} |
|
return false; |
|
} |
|
function skipBytes(count) { |
|
position += count; |
|
} |
|
function alignToByte() { |
|
bufferSize = 0; |
|
if (skipNextBit) { |
|
position++; |
|
skipNextBit = false; |
|
} |
|
} |
|
function readCodingpasses() { |
|
if (readBits(1) === 0) { |
|
return 1; |
|
} |
|
if (readBits(1) === 0) { |
|
return 2; |
|
} |
|
var value = readBits(2); |
|
if (value < 3) { |
|
return value + 3; |
|
} |
|
value = readBits(5); |
|
if (value < 31) { |
|
return value + 6; |
|
} |
|
value = readBits(7); |
|
return value + 37; |
|
} |
|
var tileIndex = context.currentTile.index; |
|
var tile = context.tiles[tileIndex]; |
|
var sopMarkerUsed = context.COD.sopMarkerUsed; |
|
var ephMarkerUsed = context.COD.ephMarkerUsed; |
|
var packetsIterator = tile.packetsIterator; |
|
while (position < dataLength) { |
|
alignToByte(); |
|
if (sopMarkerUsed && skipMarkerIfEqual(0x91)) { |
|
// Skip also marker segment length and packet sequence ID |
|
skipBytes(4); |
|
} |
|
var packet = packetsIterator.nextPacket(); |
|
if (!readBits(1)) { |
|
continue; |
|
} |
|
var layerNumber = packet.layerNumber; |
|
var queue = [], codeblock; |
|
for (var i = 0, ii = packet.codeblocks.length; i < ii; i++) { |
|
codeblock = packet.codeblocks[i]; |
|
var precinct = codeblock.precinct; |
|
var codeblockColumn = codeblock.cbx - precinct.cbxMin; |
|
var codeblockRow = codeblock.cby - precinct.cbyMin; |
|
var codeblockIncluded = false; |
|
var firstTimeInclusion = false; |
|
var valueReady; |
|
if (codeblock['included'] !== undefined) { |
|
codeblockIncluded = !!readBits(1); |
|
} else { |
|
// reading inclusion tree |
|
precinct = codeblock.precinct; |
|
var inclusionTree, zeroBitPlanesTree; |
|
if (precinct['inclusionTree'] !== undefined) { |
|
inclusionTree = precinct.inclusionTree; |
|
} else { |
|
// building inclusion and zero bit-planes trees |
|
var width = precinct.cbxMax - precinct.cbxMin + 1; |
|
var height = precinct.cbyMax - precinct.cbyMin + 1; |
|
inclusionTree = new InclusionTree(width, height, layerNumber); |
|
zeroBitPlanesTree = new TagTree(width, height); |
|
precinct.inclusionTree = inclusionTree; |
|
precinct.zeroBitPlanesTree = zeroBitPlanesTree; |
|
} |
|
|
|
if (inclusionTree.reset(codeblockColumn, codeblockRow, layerNumber)) { |
|
while (true) { |
|
if (readBits(1)) { |
|
valueReady = !inclusionTree.nextLevel(); |
|
if (valueReady) { |
|
codeblock.included = true; |
|
codeblockIncluded = firstTimeInclusion = true; |
|
break; |
|
} |
|
} else { |
|
inclusionTree.incrementValue(layerNumber); |
|
break; |
|
} |
|
} |
|
} |
|
} |
|
if (!codeblockIncluded) { |
|
continue; |
|
} |
|
if (firstTimeInclusion) { |
|
zeroBitPlanesTree = precinct.zeroBitPlanesTree; |
|
zeroBitPlanesTree.reset(codeblockColumn, codeblockRow); |
|
while (true) { |
|
if (readBits(1)) { |
|
valueReady = !zeroBitPlanesTree.nextLevel(); |
|
if (valueReady) { |
|
break; |
|
} |
|
} else { |
|
zeroBitPlanesTree.incrementValue(); |
|
} |
|
} |
|
codeblock.zeroBitPlanes = zeroBitPlanesTree.value; |
|
} |
|
var codingpasses = readCodingpasses(); |
|
while (readBits(1)) { |
|
codeblock.Lblock++; |
|
} |
|
var codingpassesLog2 = log2(codingpasses); |
|
// rounding down log2 |
|
var bits = ((codingpasses < (1 << codingpassesLog2)) ? |
|
codingpassesLog2 - 1 : codingpassesLog2) + codeblock.Lblock; |
|
var codedDataLength = readBits(bits); |
|
queue.push({ |
|
codeblock: codeblock, |
|
codingpasses: codingpasses, |
|
dataLength: codedDataLength |
|
}); |
|
} |
|
alignToByte(); |
|
if (ephMarkerUsed) { |
|
skipMarkerIfEqual(0x92); |
|
} |
|
while (queue.length > 0) { |
|
var packetItem = queue.shift(); |
|
codeblock = packetItem.codeblock; |
|
if (codeblock['data'] === undefined) { |
|
codeblock.data = []; |
|
} |
|
codeblock.data.push({ |
|
data: data, |
|
start: offset + position, |
|
end: offset + position + packetItem.dataLength, |
|
codingpasses: packetItem.codingpasses |
|
}); |
|
position += packetItem.dataLength; |
|
} |
|
} |
|
return position; |
|
} |
|
function copyCoefficients(coefficients, levelWidth, levelHeight, subband, |
|
delta, mb, reversible, segmentationSymbolUsed) { |
|
var x0 = subband.tbx0; |
|
var y0 = subband.tby0; |
|
var width = subband.tbx1 - subband.tbx0; |
|
var codeblocks = subband.codeblocks; |
|
var right = subband.type.charAt(0) === 'H' ? 1 : 0; |
|
var bottom = subband.type.charAt(1) === 'H' ? levelWidth : 0; |
|
|
|
for (var i = 0, ii = codeblocks.length; i < ii; ++i) { |
|
var codeblock = codeblocks[i]; |
|
var blockWidth = codeblock.tbx1_ - codeblock.tbx0_; |
|
var blockHeight = codeblock.tby1_ - codeblock.tby0_; |
|
if (blockWidth === 0 || blockHeight === 0) { |
|
continue; |
|
} |
|
if (codeblock['data'] === undefined) { |
|
continue; |
|
} |
|
|
|
var bitModel, currentCodingpassType; |
|
bitModel = new BitModel(blockWidth, blockHeight, codeblock.subbandType, |
|
codeblock.zeroBitPlanes, mb); |
|
currentCodingpassType = 2; // first bit plane starts from cleanup |
|
|
|
// collect data |
|
var data = codeblock.data, totalLength = 0, codingpasses = 0; |
|
var j, jj, dataItem; |
|
for (j = 0, jj = data.length; j < jj; j++) { |
|
dataItem = data[j]; |
|
totalLength += dataItem.end - dataItem.start; |
|
codingpasses += dataItem.codingpasses; |
|
} |
|
var encodedData = new Uint8Array(totalLength); |
|
var position = 0; |
|
for (j = 0, jj = data.length; j < jj; j++) { |
|
dataItem = data[j]; |
|
var chunk = dataItem.data.subarray(dataItem.start, dataItem.end); |
|
encodedData.set(chunk, position); |
|
position += chunk.length; |
|
} |
|
// decoding the item |
|
var decoder = new ArithmeticDecoder(encodedData, 0, totalLength); |
|
bitModel.setDecoder(decoder); |
|
|
|
for (j = 0; j < codingpasses; j++) { |
|
switch (currentCodingpassType) { |
|
case 0: |
|
bitModel.runSignificancePropogationPass(); |
|
break; |
|
case 1: |
|
bitModel.runMagnitudeRefinementPass(); |
|
break; |
|
case 2: |
|
bitModel.runCleanupPass(); |
|
if (segmentationSymbolUsed) { |
|
bitModel.checkSegmentationSymbol(); |
|
} |
|
break; |
|
} |
|
currentCodingpassType = (currentCodingpassType + 1) % 3; |
|
} |
|
|
|
var offset = (codeblock.tbx0_ - x0) + (codeblock.tby0_ - y0) * width; |
|
var sign = bitModel.coefficentsSign; |
|
var magnitude = bitModel.coefficentsMagnitude; |
|
var bitsDecoded = bitModel.bitsDecoded; |
|
var magnitudeCorrection = reversible ? 0 : 0.5; |
|
var k, n, nb; |
|
position = 0; |
|
// Do the interleaving of Section F.3.3 here, so we do not need |
|
// to copy later. LL level is not interleaved, just copied. |
|
var interleave = (subband.type !== 'LL'); |
|
for (j = 0; j < blockHeight; j++) { |
|
var row = (offset / width) | 0; // row in the non-interleaved subband |
|
var levelOffset = 2 * row * (levelWidth - width) + right + bottom; |
|
for (k = 0; k < blockWidth; k++) { |
|
n = magnitude[position]; |
|
if (n !== 0) { |
|
n = (n + magnitudeCorrection) * delta; |
|
if (sign[position] !== 0) { |
|
n = -n; |
|
} |
|
nb = bitsDecoded[position]; |
|
var pos = interleave ? (levelOffset + (offset << 1)) : offset; |
|
if (reversible && (nb >= mb)) { |
|
coefficients[pos] = n; |
|
} else { |
|
coefficients[pos] = n * (1 << (mb - nb)); |
|
} |
|
} |
|
offset++; |
|
position++; |
|
} |
|
offset += width - blockWidth; |
|
} |
|
} |
|
} |
|
function transformTile(context, tile, c) { |
|
var component = tile.components[c]; |
|
var codingStyleParameters = component.codingStyleParameters; |
|
var quantizationParameters = component.quantizationParameters; |
|
var decompositionLevelsCount = |
|
codingStyleParameters.decompositionLevelsCount; |
|
var spqcds = quantizationParameters.SPqcds; |
|
var scalarExpounded = quantizationParameters.scalarExpounded; |
|
var guardBits = quantizationParameters.guardBits; |
|
var segmentationSymbolUsed = codingStyleParameters.segmentationSymbolUsed; |
|
var precision = context.components[c].precision; |
|
|
|
var reversible = codingStyleParameters.reversibleTransformation; |
|
var transform = (reversible ? new ReversibleTransform() : |
|
new IrreversibleTransform()); |
|
|
|
var subbandCoefficients = []; |
|
var b = 0; |
|
for (var i = 0; i <= decompositionLevelsCount; i++) { |
|
var resolution = component.resolutions[i]; |
|
|
|
var width = resolution.trx1 - resolution.trx0; |
|
var height = resolution.try1 - resolution.try0; |
|
// Allocate space for the whole sublevel. |
|
var coefficients = new Float32Array(width * height); |
|
|
|
for (var j = 0, jj = resolution.subbands.length; j < jj; j++) { |
|
var mu, epsilon; |
|
if (!scalarExpounded) { |
|
// formula E-5 |
|
mu = spqcds[0].mu; |
|
epsilon = spqcds[0].epsilon + (i > 0 ? 1 - i : 0); |
|
} else { |
|
mu = spqcds[b].mu; |
|
epsilon = spqcds[b].epsilon; |
|
b++; |
|
} |
|
|
|
var subband = resolution.subbands[j]; |
|
var gainLog2 = SubbandsGainLog2[subband.type]; |
|
|
|
// calulate quantization coefficient (Section E.1.1.1) |
|
var delta = (reversible ? 1 : |
|
Math.pow(2, precision + gainLog2 - epsilon) * (1 + mu / 2048)); |
|
var mb = (guardBits + epsilon - 1); |
|
|
|
// In the first resolution level, copyCoefficients will fill the |
|
// whole array with coefficients. In the succeding passes, |
|
// copyCoefficients will consecutively fill in the values that belong |
|
// to the interleaved positions of the HL, LH, and HH coefficients. |
|
// The LL coefficients will then be interleaved in Transform.iterate(). |
|
copyCoefficients(coefficients, width, height, subband, delta, mb, |
|
reversible, segmentationSymbolUsed); |
|
} |
|
subbandCoefficients.push({ |
|
width: width, |
|
height: height, |
|
items: coefficients |
|
}); |
|
} |
|
|
|
var result = transform.calculate(subbandCoefficients, |
|
component.tcx0, component.tcy0); |
|
return { |
|
left: component.tcx0, |
|
top: component.tcy0, |
|
width: result.width, |
|
height: result.height, |
|
items: result.items |
|
}; |
|
} |
|
function transformComponents(context) { |
|
var siz = context.SIZ; |
|
var components = context.components; |
|
var componentsCount = siz.Csiz; |
|
var resultImages = []; |
|
for (var i = 0, ii = context.tiles.length; i < ii; i++) { |
|
var tile = context.tiles[i]; |
|
var transformedTiles = []; |
|
var c; |
|
for (c = 0; c < componentsCount; c++) { |
|
transformedTiles[c] = transformTile(context, tile, c); |
|
} |
|
var tile0 = transformedTiles[0]; |
|
var out = new Uint8Array(tile0.items.length * componentsCount); |
|
var result = { |
|
left: tile0.left, |
|
top: tile0.top, |
|
width: tile0.width, |
|
height: tile0.height, |
|
items: out |
|
}; |
|
|
|
// Section G.2.2 Inverse multi component transform |
|
var shift, offset, max, min, maxK; |
|
var pos = 0, j, jj, y0, y1, y2, r, g, b, k, val; |
|
if (tile.codingStyleDefaultParameters.multipleComponentTransform) { |
|
var fourComponents = componentsCount === 4; |
|
var y0items = transformedTiles[0].items; |
|
var y1items = transformedTiles[1].items; |
|
var y2items = transformedTiles[2].items; |
|
var y3items = fourComponents ? transformedTiles[3].items : null; |
|
|
|
// HACK: The multiple component transform formulas below assume that |
|
// all components have the same precision. With this in mind, we |
|
// compute shift and offset only once. |
|
shift = components[0].precision - 8; |
|
offset = (128 << shift) + 0.5; |
|
max = 255 * (1 << shift); |
|
maxK = max * 0.5; |
|
min = -maxK; |
|
|
|
var component0 = tile.components[0]; |
|
var alpha01 = componentsCount - 3; |
|
jj = y0items.length; |
|
if (!component0.codingStyleParameters.reversibleTransformation) { |
|
// inverse irreversible multiple component transform |
|
for (j = 0; j < jj; j++, pos += alpha01) { |
|
y0 = y0items[j] + offset; |
|
y1 = y1items[j]; |
|
y2 = y2items[j]; |
|
r = y0 + 1.402 * y2; |
|
g = y0 - 0.34413 * y1 - 0.71414 * y2; |
|
b = y0 + 1.772 * y1; |
|
out[pos++] = r <= 0 ? 0 : r >= max ? 255 : r >> shift; |
|
out[pos++] = g <= 0 ? 0 : g >= max ? 255 : g >> shift; |
|
out[pos++] = b <= 0 ? 0 : b >= max ? 255 : b >> shift; |
|
} |
|
} else { |
|
// inverse reversible multiple component transform |
|
for (j = 0; j < jj; j++, pos += alpha01) { |
|
y0 = y0items[j] + offset; |
|
y1 = y1items[j]; |
|
y2 = y2items[j]; |
|
g = y0 - ((y2 + y1) >> 2); |
|
r = g + y2; |
|
b = g + y1; |
|
out[pos++] = r <= 0 ? 0 : r >= max ? 255 : r >> shift; |
|
out[pos++] = g <= 0 ? 0 : g >= max ? 255 : g >> shift; |
|
out[pos++] = b <= 0 ? 0 : b >= max ? 255 : b >> shift; |
|
} |
|
} |
|
if (fourComponents) { |
|
for (j = 0, pos = 3; j < jj; j++, pos += 4) { |
|
k = y3items[j]; |
|
out[pos] = k <= min ? 0 : k >= maxK ? 255 : (k + offset) >> shift; |
|
} |
|
} |
|
} else { // no multi-component transform |
|
for (c = 0; c < componentsCount; c++) { |
|
var items = transformedTiles[c].items; |
|
shift = components[c].precision - 8; |
|
offset = (128 << shift) + 0.5; |
|
max = (127.5 * (1 << shift)); |
|
min = -max; |
|
for (pos = c, j = 0, jj = items.length; j < jj; j++) { |
|
val = items[j]; |
|
out[pos] = val <= min ? 0 : |
|
val >= max ? 255 : (val + offset) >> shift; |
|
pos += componentsCount; |
|
} |
|
} |
|
} |
|
resultImages.push(result); |
|
} |
|
return resultImages; |
|
} |
|
function initializeTile(context, tileIndex) { |
|
var siz = context.SIZ; |
|
var componentsCount = siz.Csiz; |
|
var tile = context.tiles[tileIndex]; |
|
for (var c = 0; c < componentsCount; c++) { |
|
var component = tile.components[c]; |
|
var qcdOrQcc = (context.currentTile.QCC[c] !== undefined ? |
|
context.currentTile.QCC[c] : context.currentTile.QCD); |
|
component.quantizationParameters = qcdOrQcc; |
|
var codOrCoc = (context.currentTile.COC[c] !== undefined ? |
|
context.currentTile.COC[c] : context.currentTile.COD); |
|
component.codingStyleParameters = codOrCoc; |
|
} |
|
tile.codingStyleDefaultParameters = context.currentTile.COD; |
|
} |
|
|
|
// Section B.10.2 Tag trees |
|
var TagTree = (function TagTreeClosure() { |
|
function TagTree(width, height) { |
|
var levelsLength = log2(Math.max(width, height)) + 1; |
|
this.levels = []; |
|
for (var i = 0; i < levelsLength; i++) { |
|
var level = { |
|
width: width, |
|
height: height, |
|
items: [] |
|
}; |
|
this.levels.push(level); |
|
width = Math.ceil(width / 2); |
|
height = Math.ceil(height / 2); |
|
} |
|
} |
|
TagTree.prototype = { |
|
reset: function TagTree_reset(i, j) { |
|
var currentLevel = 0, value = 0, level; |
|
while (currentLevel < this.levels.length) { |
|
level = this.levels[currentLevel]; |
|
var index = i + j * level.width; |
|
if (level.items[index] !== undefined) { |
|
value = level.items[index]; |
|
break; |
|
} |
|
level.index = index; |
|
i >>= 1; |
|
j >>= 1; |
|
currentLevel++; |
|
} |
|
currentLevel--; |
|
level = this.levels[currentLevel]; |
|
level.items[level.index] = value; |
|
this.currentLevel = currentLevel; |
|
delete this.value; |
|
}, |
|
incrementValue: function TagTree_incrementValue() { |
|
var level = this.levels[this.currentLevel]; |
|
level.items[level.index]++; |
|
}, |
|
nextLevel: function TagTree_nextLevel() { |
|
var currentLevel = this.currentLevel; |
|
var level = this.levels[currentLevel]; |
|
var value = level.items[level.index]; |
|
currentLevel--; |
|
if (currentLevel < 0) { |
|
this.value = value; |
|
return false; |
|
} |
|
|
|
this.currentLevel = currentLevel; |
|
level = this.levels[currentLevel]; |
|
level.items[level.index] = value; |
|
return true; |
|
} |
|
}; |
|
return TagTree; |
|
})(); |
|
|
|
var InclusionTree = (function InclusionTreeClosure() { |
|
function InclusionTree(width, height, defaultValue) { |
|
var levelsLength = log2(Math.max(width, height)) + 1; |
|
this.levels = []; |
|
for (var i = 0; i < levelsLength; i++) { |
|
var items = new Uint8Array(width * height); |
|
for (var j = 0, jj = items.length; j < jj; j++) { |
|
items[j] = defaultValue; |
|
} |
|
|
|
var level = { |
|
width: width, |
|
height: height, |
|
items: items |
|
}; |
|
this.levels.push(level); |
|
|
|
width = Math.ceil(width / 2); |
|
height = Math.ceil(height / 2); |
|
} |
|
} |
|
InclusionTree.prototype = { |
|
reset: function InclusionTree_reset(i, j, stopValue) { |
|
var currentLevel = 0; |
|
while (currentLevel < this.levels.length) { |
|
var level = this.levels[currentLevel]; |
|
var index = i + j * level.width; |
|
level.index = index; |
|
var value = level.items[index]; |
|
|
|
if (value === 0xFF) { |
|
break; |
|
} |
|
|
|
if (value > stopValue) { |
|
this.currentLevel = currentLevel; |
|
// already know about this one, propagating the value to top levels |
|
this.propagateValues(); |
|
return false; |
|
} |
|
|
|
i >>= 1; |
|
j >>= 1; |
|
currentLevel++; |
|
} |
|
this.currentLevel = currentLevel - 1; |
|
return true; |
|
}, |
|
incrementValue: function InclusionTree_incrementValue(stopValue) { |
|
var level = this.levels[this.currentLevel]; |
|
level.items[level.index] = stopValue + 1; |
|
this.propagateValues(); |
|
}, |
|
propagateValues: function InclusionTree_propagateValues() { |
|
var levelIndex = this.currentLevel; |
|
var level = this.levels[levelIndex]; |
|
var currentValue = level.items[level.index]; |
|
while (--levelIndex >= 0) { |
|
level = this.levels[levelIndex]; |
|
level.items[level.index] = currentValue; |
|
} |
|
}, |
|
nextLevel: function InclusionTree_nextLevel() { |
|
var currentLevel = this.currentLevel; |
|
var level = this.levels[currentLevel]; |
|
var value = level.items[level.index]; |
|
level.items[level.index] = 0xFF; |
|
currentLevel--; |
|
if (currentLevel < 0) { |
|
return false; |
|
} |
|
|
|
this.currentLevel = currentLevel; |
|
level = this.levels[currentLevel]; |
|
level.items[level.index] = value; |
|
return true; |
|
} |
|
}; |
|
return InclusionTree; |
|
})(); |
|
|
|
// Section D. Coefficient bit modeling |
|
var BitModel = (function BitModelClosure() { |
|
var UNIFORM_CONTEXT = 17; |
|
var RUNLENGTH_CONTEXT = 18; |
|
// Table D-1 |
|
// The index is binary presentation: 0dddvvhh, ddd - sum of Di (0..4), |
|
// vv - sum of Vi (0..2), and hh - sum of Hi (0..2) |
|
var LLAndLHContextsLabel = new Uint8Array([ |
|
0, 5, 8, 0, 3, 7, 8, 0, 4, 7, 8, 0, 0, 0, 0, 0, 1, 6, 8, 0, 3, 7, 8, 0, 4, |
|
7, 8, 0, 0, 0, 0, 0, 2, 6, 8, 0, 3, 7, 8, 0, 4, 7, 8, 0, 0, 0, 0, 0, 2, 6, |
|
8, 0, 3, 7, 8, 0, 4, 7, 8, 0, 0, 0, 0, 0, 2, 6, 8, 0, 3, 7, 8, 0, 4, 7, 8 |
|
]); |
|
var HLContextLabel = new Uint8Array([ |
|
0, 3, 4, 0, 5, 7, 7, 0, 8, 8, 8, 0, 0, 0, 0, 0, 1, 3, 4, 0, 6, 7, 7, 0, 8, |
|
8, 8, 0, 0, 0, 0, 0, 2, 3, 4, 0, 6, 7, 7, 0, 8, 8, 8, 0, 0, 0, 0, 0, 2, 3, |
|
4, 0, 6, 7, 7, 0, 8, 8, 8, 0, 0, 0, 0, 0, 2, 3, 4, 0, 6, 7, 7, 0, 8, 8, 8 |
|
]); |
|
var HHContextLabel = new Uint8Array([ |
|
0, 1, 2, 0, 1, 2, 2, 0, 2, 2, 2, 0, 0, 0, 0, 0, 3, 4, 5, 0, 4, 5, 5, 0, 5, |
|
5, 5, 0, 0, 0, 0, 0, 6, 7, 7, 0, 7, 7, 7, 0, 7, 7, 7, 0, 0, 0, 0, 0, 8, 8, |
|
8, 0, 8, 8, 8, 0, 8, 8, 8, 0, 0, 0, 0, 0, 8, 8, 8, 0, 8, 8, 8, 0, 8, 8, 8 |
|
]); |
|
|
|
function BitModel(width, height, subband, zeroBitPlanes, mb) { |
|
this.width = width; |
|
this.height = height; |
|
|
|
this.contextLabelTable = (subband === 'HH' ? HHContextLabel : |
|
(subband === 'HL' ? HLContextLabel : LLAndLHContextsLabel)); |
|
|
|
var coefficientCount = width * height; |
|
|
|
// coefficients outside the encoding region treated as insignificant |
|
// add border state cells for significanceState |
|
this.neighborsSignificance = new Uint8Array(coefficientCount); |
|
this.coefficentsSign = new Uint8Array(coefficientCount); |
|
this.coefficentsMagnitude = mb > 14 ? new Uint32Array(coefficientCount) : |
|
mb > 6 ? new Uint16Array(coefficientCount) : |
|
new Uint8Array(coefficientCount); |
|
this.processingFlags = new Uint8Array(coefficientCount); |
|
|
|
var bitsDecoded = new Uint8Array(coefficientCount); |
|
if (zeroBitPlanes !== 0) { |
|
for (var i = 0; i < coefficientCount; i++) { |
|
bitsDecoded[i] = zeroBitPlanes; |
|
} |
|
} |
|
this.bitsDecoded = bitsDecoded; |
|
|
|
this.reset(); |
|
} |
|
|
|
BitModel.prototype = { |
|
setDecoder: function BitModel_setDecoder(decoder) { |
|
this.decoder = decoder; |
|
}, |
|
reset: function BitModel_reset() { |
|
// We have 17 contexts that are accessed via context labels, |
|
// plus the uniform and runlength context. |
|
this.contexts = new Int8Array(19); |
|
|
|
// Contexts are packed into 1 byte: |
|
// highest 7 bits carry the index, lowest bit carries mps |
|
this.contexts[0] = (4 << 1) | 0; |
|
this.contexts[UNIFORM_CONTEXT] = (46 << 1) | 0; |
|
this.contexts[RUNLENGTH_CONTEXT] = (3 << 1) | 0; |
|
}, |
|
setNeighborsSignificance: |
|
function BitModel_setNeighborsSignificance(row, column, index) { |
|
var neighborsSignificance = this.neighborsSignificance; |
|
var width = this.width, height = this.height; |
|
var left = (column > 0); |
|
var right = (column + 1 < width); |
|
var i; |
|
|
|
if (row > 0) { |
|
i = index - width; |
|
if (left) { |
|
neighborsSignificance[i - 1] += 0x10; |
|
} |
|
if (right) { |
|
neighborsSignificance[i + 1] += 0x10; |
|
} |
|
neighborsSignificance[i] += 0x04; |
|
} |
|
|
|
if (row + 1 < height) { |
|
i = index + width; |
|
if (left) { |
|
neighborsSignificance[i - 1] += 0x10; |
|
} |
|
if (right) { |
|
neighborsSignificance[i + 1] += 0x10; |
|
} |
|
neighborsSignificance[i] += 0x04; |
|
} |
|
|
|
if (left) { |
|
neighborsSignificance[index - 1] += 0x01; |
|
} |
|
if (right) { |
|
neighborsSignificance[index + 1] += 0x01; |
|
} |
|
neighborsSignificance[index] |= 0x80; |
|
}, |
|
runSignificancePropogationPass: |
|
function BitModel_runSignificancePropogationPass() { |
|
var decoder = this.decoder; |
|
var width = this.width, height = this.height; |
|
var coefficentsMagnitude = this.coefficentsMagnitude; |
|
var coefficentsSign = this.coefficentsSign; |
|
var neighborsSignificance = this.neighborsSignificance; |
|
var processingFlags = this.processingFlags; |
|
var contexts = this.contexts; |
|
var labels = this.contextLabelTable; |
|
var bitsDecoded = this.bitsDecoded; |
|
var processedInverseMask = ~1; |
|
var processedMask = 1; |
|
var firstMagnitudeBitMask = 2; |
|
|
|
for (var i0 = 0; i0 < height; i0 += 4) { |
|
for (var j = 0; j < width; j++) { |
|
var index = i0 * width + j; |
|
for (var i1 = 0; i1 < 4; i1++, index += width) { |
|
var i = i0 + i1; |
|
if (i >= height) { |
|
break; |
|
} |
|
// clear processed flag first |
|
processingFlags[index] &= processedInverseMask; |
|
|
|
if (coefficentsMagnitude[index] || |
|
!neighborsSignificance[index]) { |
|
continue; |
|
} |
|
|
|
var contextLabel = labels[neighborsSignificance[index]]; |
|
var decision = decoder.readBit(contexts, contextLabel); |
|
if (decision) { |
|
var sign = this.decodeSignBit(i, j, index); |
|
coefficentsSign[index] = sign; |
|
coefficentsMagnitude[index] = 1; |
|
this.setNeighborsSignificance(i, j, index); |
|
processingFlags[index] |= firstMagnitudeBitMask; |
|
} |
|
bitsDecoded[index]++; |
|
processingFlags[index] |= processedMask; |
|
} |
|
} |
|
} |
|
}, |
|
decodeSignBit: function BitModel_decodeSignBit(row, column, index) { |
|
var width = this.width, height = this.height; |
|
var coefficentsMagnitude = this.coefficentsMagnitude; |
|
var coefficentsSign = this.coefficentsSign; |
|
var contribution, sign0, sign1, significance1; |
|
var contextLabel, decoded; |
|
|
|
// calculate horizontal contribution |
|
significance1 = (column > 0 && coefficentsMagnitude[index - 1] !== 0); |
|
if (column + 1 < width && coefficentsMagnitude[index + 1] !== 0) { |
|
sign1 = coefficentsSign[index + 1]; |
|
if (significance1) { |
|
sign0 = coefficentsSign[index - 1]; |
|
contribution = 1 - sign1 - sign0; |
|
} else { |
|
contribution = 1 - sign1 - sign1; |
|
} |
|
} else if (significance1) { |
|
sign0 = coefficentsSign[index - 1]; |
|
contribution = 1 - sign0 - sign0; |
|
} else { |
|
contribution = 0; |
|
} |
|
var horizontalContribution = 3 * contribution; |
|
|
|
// calculate vertical contribution and combine with the horizontal |
|
significance1 = (row > 0 && coefficentsMagnitude[index - width] !== 0); |
|
if (row + 1 < height && coefficentsMagnitude[index + width] !== 0) { |
|
sign1 = coefficentsSign[index + width]; |
|
if (significance1) { |
|
sign0 = coefficentsSign[index - width]; |
|
contribution = 1 - sign1 - sign0 + horizontalContribution; |
|
} else { |
|
contribution = 1 - sign1 - sign1 + horizontalContribution; |
|
} |
|
} else if (significance1) { |
|
sign0 = coefficentsSign[index - width]; |
|
contribution = 1 - sign0 - sign0 + horizontalContribution; |
|
} else { |
|
contribution = horizontalContribution; |
|
} |
|
|
|
if (contribution >= 0) { |
|
contextLabel = 9 + contribution; |
|
decoded = this.decoder.readBit(this.contexts, contextLabel); |
|
} else { |
|
contextLabel = 9 - contribution; |
|
decoded = this.decoder.readBit(this.contexts, contextLabel) ^ 1; |
|
} |
|
return decoded; |
|
}, |
|
runMagnitudeRefinementPass: |
|
function BitModel_runMagnitudeRefinementPass() { |
|
var decoder = this.decoder; |
|
var width = this.width, height = this.height; |
|
var coefficentsMagnitude = this.coefficentsMagnitude; |
|
var neighborsSignificance = this.neighborsSignificance; |
|
var contexts = this.contexts; |
|
var bitsDecoded = this.bitsDecoded; |
|
var processingFlags = this.processingFlags; |
|
var processedMask = 1; |
|
var firstMagnitudeBitMask = 2; |
|
var length = width * height; |
|
var width4 = width * 4; |
|
|
|
for (var index0 = 0, indexNext; index0 < length; index0 = indexNext) { |
|
indexNext = Math.min(length, index0 + width4); |
|
for (var j = 0; j < width; j++) { |
|
for (var index = index0 + j; index < indexNext; index += width) { |
|
|
|
// significant but not those that have just become |
|
if (!coefficentsMagnitude[index] || |
|
(processingFlags[index] & processedMask) !== 0) { |
|
continue; |
|
} |
|
|
|
var contextLabel = 16; |
|
if ((processingFlags[index] & firstMagnitudeBitMask) !== 0) { |
|
processingFlags[index] ^= firstMagnitudeBitMask; |
|
// first refinement |
|
var significance = neighborsSignificance[index] & 127; |
|
contextLabel = significance === 0 ? 15 : 14; |
|
} |
|
|
|
var bit = decoder.readBit(contexts, contextLabel); |
|
coefficentsMagnitude[index] = |
|
(coefficentsMagnitude[index] << 1) | bit; |
|
bitsDecoded[index]++; |
|
processingFlags[index] |= processedMask; |
|
} |
|
} |
|
} |
|
}, |
|
runCleanupPass: function BitModel_runCleanupPass() { |
|
var decoder = this.decoder; |
|
var width = this.width, height = this.height; |
|
var neighborsSignificance = this.neighborsSignificance; |
|
var coefficentsMagnitude = this.coefficentsMagnitude; |
|
var coefficentsSign = this.coefficentsSign; |
|
var contexts = this.contexts; |
|
var labels = this.contextLabelTable; |
|
var bitsDecoded = this.bitsDecoded; |
|
var processingFlags = this.processingFlags; |
|
var processedMask = 1; |
|
var firstMagnitudeBitMask = 2; |
|
var oneRowDown = width; |
|
var twoRowsDown = width * 2; |
|
var threeRowsDown = width * 3; |
|
var iNext; |
|
for (var i0 = 0; i0 < height; i0 = iNext) { |
|
iNext = Math.min(i0 + 4, height); |
|
var indexBase = i0 * width; |
|
var checkAllEmpty = i0 + 3 < height; |
|
for (var j = 0; j < width; j++) { |
|
var index0 = indexBase + j; |
|
// using the property: labels[neighborsSignificance[index]] === 0 |
|
// when neighborsSignificance[index] === 0 |
|
var allEmpty = (checkAllEmpty && |
|
processingFlags[index0] === 0 && |
|
processingFlags[index0 + oneRowDown] === 0 && |
|
processingFlags[index0 + twoRowsDown] === 0 && |
|
processingFlags[index0 + threeRowsDown] === 0 && |
|
neighborsSignificance[index0] === 0 && |
|
neighborsSignificance[index0 + oneRowDown] === 0 && |
|
neighborsSignificance[index0 + twoRowsDown] === 0 && |
|
neighborsSignificance[index0 + threeRowsDown] === 0); |
|
var i1 = 0, index = index0; |
|
var i = i0, sign; |
|
if (allEmpty) { |
|
var hasSignificantCoefficent = |
|
decoder.readBit(contexts, RUNLENGTH_CONTEXT); |
|
if (!hasSignificantCoefficent) { |
|
bitsDecoded[index0]++; |
|
bitsDecoded[index0 + oneRowDown]++; |
|
bitsDecoded[index0 + twoRowsDown]++; |
|
bitsDecoded[index0 + threeRowsDown]++; |
|
continue; // next column |
|
} |
|
i1 = (decoder.readBit(contexts, UNIFORM_CONTEXT) << 1) | |
|
decoder.readBit(contexts, UNIFORM_CONTEXT); |
|
if (i1 !== 0) { |
|
i = i0 + i1; |
|
index += i1 * width; |
|
} |
|
|
|
sign = this.decodeSignBit(i, j, index); |
|
coefficentsSign[index] = sign; |
|
coefficentsMagnitude[index] = 1; |
|
this.setNeighborsSignificance(i, j, index); |
|
processingFlags[index] |= firstMagnitudeBitMask; |
|
|
|
index = index0; |
|
for (var i2 = i0; i2 <= i; i2++, index += width) { |
|
bitsDecoded[index]++; |
|
} |
|
|
|
i1++; |
|
} |
|
for (i = i0 + i1; i < iNext; i++, index += width) { |
|
if (coefficentsMagnitude[index] || |
|
(processingFlags[index] & processedMask) !== 0) { |
|
continue; |
|
} |
|
|
|
var contextLabel = labels[neighborsSignificance[index]]; |
|
var decision = decoder.readBit(contexts, contextLabel); |
|
if (decision === 1) { |
|
sign = this.decodeSignBit(i, j, index); |
|
coefficentsSign[index] = sign; |
|
coefficentsMagnitude[index] = 1; |
|
this.setNeighborsSignificance(i, j, index); |
|
processingFlags[index] |= firstMagnitudeBitMask; |
|
} |
|
bitsDecoded[index]++; |
|
} |
|
} |
|
} |
|
}, |
|
checkSegmentationSymbol: function BitModel_checkSegmentationSymbol() { |
|
var decoder = this.decoder; |
|
var contexts = this.contexts; |
|
var symbol = (decoder.readBit(contexts, UNIFORM_CONTEXT) << 3) | |
|
(decoder.readBit(contexts, UNIFORM_CONTEXT) << 2) | |
|
(decoder.readBit(contexts, UNIFORM_CONTEXT) << 1) | |
|
decoder.readBit(contexts, UNIFORM_CONTEXT); |
|
if (symbol !== 0xA) { |
|
throw new Error('JPX Error: Invalid segmentation symbol'); |
|
} |
|
} |
|
}; |
|
|
|
return BitModel; |
|
})(); |
|
|
|
// Section F, Discrete wavelet transformation |
|
var Transform = (function TransformClosure() { |
|
function Transform() {} |
|
|
|
Transform.prototype.calculate = |
|
function transformCalculate(subbands, u0, v0) { |
|
var ll = subbands[0]; |
|
for (var i = 1, ii = subbands.length; i < ii; i++) { |
|
ll = this.iterate(ll, subbands[i], u0, v0); |
|
} |
|
return ll; |
|
}; |
|
Transform.prototype.extend = function extend(buffer, offset, size) { |
|
// Section F.3.7 extending... using max extension of 4 |
|
var i1 = offset - 1, j1 = offset + 1; |
|
var i2 = offset + size - 2, j2 = offset + size; |
|
buffer[i1--] = buffer[j1++]; |
|
buffer[j2++] = buffer[i2--]; |
|
buffer[i1--] = buffer[j1++]; |
|
buffer[j2++] = buffer[i2--]; |
|
buffer[i1--] = buffer[j1++]; |
|
buffer[j2++] = buffer[i2--]; |
|
buffer[i1] = buffer[j1]; |
|
buffer[j2] = buffer[i2]; |
|
}; |
|
Transform.prototype.iterate = function Transform_iterate(ll, hl_lh_hh, |
|
u0, v0) { |
|
var llWidth = ll.width, llHeight = ll.height, llItems = ll.items; |
|
var width = hl_lh_hh.width; |
|
var height = hl_lh_hh.height; |
|
var items = hl_lh_hh.items; |
|
var i, j, k, l, u, v; |
|
|
|
// Interleave LL according to Section F.3.3 |
|
for (k = 0, i = 0; i < llHeight; i++) { |
|
l = i * 2 * width; |
|
for (j = 0; j < llWidth; j++, k++, l += 2) { |
|
items[l] = llItems[k]; |
|
} |
|
} |
|
// The LL band is not needed anymore. |
|
llItems = ll.items = null; |
|
|
|
var bufferPadding = 4; |
|
var rowBuffer = new Float32Array(width + 2 * bufferPadding); |
|
|
|
// Section F.3.4 HOR_SR |
|
if (width === 1) { |
|
// if width = 1, when u0 even keep items as is, when odd divide by 2 |
|
if ((u0 & 1) !== 0) { |
|
for (v = 0, k = 0; v < height; v++, k += width) { |
|
items[k] *= 0.5; |
|
} |
|
} |
|
} else { |
|
for (v = 0, k = 0; v < height; v++, k += width) { |
|
rowBuffer.set(items.subarray(k, k + width), bufferPadding); |
|
|
|
this.extend(rowBuffer, bufferPadding, width); |
|
this.filter(rowBuffer, bufferPadding, width); |
|
|
|
items.set( |
|
rowBuffer.subarray(bufferPadding, bufferPadding + width), |
|
k); |
|
} |
|
} |
|
|
|
// Accesses to the items array can take long, because it may not fit into |
|
// CPU cache and has to be fetched from main memory. Since subsequent |
|
// accesses to the items array are not local when reading columns, we |
|
// have a cache miss every time. To reduce cache misses, get up to |
|
// 'numBuffers' items at a time and store them into the individual |
|
// buffers. The colBuffers should be small enough to fit into CPU cache. |
|
var numBuffers = 16; |
|
var colBuffers = []; |
|
for (i = 0; i < numBuffers; i++) { |
|
colBuffers.push(new Float32Array(height + 2 * bufferPadding)); |
|
} |
|
var b, currentBuffer = 0; |
|
ll = bufferPadding + height; |
|
|
|
// Section F.3.5 VER_SR |
|
if (height === 1) { |
|
// if height = 1, when v0 even keep items as is, when odd divide by 2 |
|
if ((v0 & 1) !== 0) { |
|
for (u = 0; u < width; u++) { |
|
items[u] *= 0.5; |
|
} |
|
} |
|
} else { |
|
for (u = 0; u < width; u++) { |
|
// if we ran out of buffers, copy several image columns at once |
|
if (currentBuffer === 0) { |
|
numBuffers = Math.min(width - u, numBuffers); |
|
for (k = u, l = bufferPadding; l < ll; k += width, l++) { |
|
for (b = 0; b < numBuffers; b++) { |
|
colBuffers[b][l] = items[k + b]; |
|
} |
|
} |
|
currentBuffer = numBuffers; |
|
} |
|
|
|
currentBuffer--; |
|
var buffer = colBuffers[currentBuffer]; |
|
this.extend(buffer, bufferPadding, height); |
|
this.filter(buffer, bufferPadding, height); |
|
|
|
// If this is last buffer in this group of buffers, flush all buffers. |
|
if (currentBuffer === 0) { |
|
k = u - numBuffers + 1; |
|
for (l = bufferPadding; l < ll; k += width, l++) { |
|
for (b = 0; b < numBuffers; b++) { |
|
items[k + b] = colBuffers[b][l]; |
|
} |
|
} |
|
} |
|
} |
|
} |
|
|
|
return { |
|
width: width, |
|
height: height, |
|
items: items |
|
}; |
|
}; |
|
return Transform; |
|
})(); |
|
|
|
// Section 3.8.2 Irreversible 9-7 filter |
|
var IrreversibleTransform = (function IrreversibleTransformClosure() { |
|
function IrreversibleTransform() { |
|
Transform.call(this); |
|
} |
|
|
|
IrreversibleTransform.prototype = Object.create(Transform.prototype); |
|
IrreversibleTransform.prototype.filter = |
|
function irreversibleTransformFilter(x, offset, length) { |
|
var len = length >> 1; |
|
offset = offset | 0; |
|
var j, n, current, next; |
|
|
|
var alpha = -1.586134342059924; |
|
var beta = -0.052980118572961; |
|
var gamma = 0.882911075530934; |
|
var delta = 0.443506852043971; |
|
var K = 1.230174104914001; |
|
var K_ = 1 / K; |
|
|
|
// step 1 is combined with step 3 |
|
|
|
// step 2 |
|
j = offset - 3; |
|
for (n = len + 4; n--; j += 2) { |
|
x[j] *= K_; |
|
} |
|
|
|
// step 1 & 3 |
|
j = offset - 2; |
|
current = delta * x[j -1]; |
|
for (n = len + 3; n--; j += 2) { |
|
next = delta * x[j + 1]; |
|
x[j] = K * x[j] - current - next; |
|
if (n--) { |
|
j += 2; |
|
current = delta * x[j + 1]; |
|
x[j] = K * x[j] - current - next; |
|
} else { |
|
break; |
|
} |
|
} |
|
|
|
// step 4 |
|
j = offset - 1; |
|
current = gamma * x[j - 1]; |
|
for (n = len + 2; n--; j += 2) { |
|
next = gamma * x[j + 1]; |
|
x[j] -= current + next; |
|
if (n--) { |
|
j += 2; |
|
current = gamma * x[j + 1]; |
|
x[j] -= current + next; |
|
} else { |
|
break; |
|
} |
|
} |
|
|
|
// step 5 |
|
j = offset; |
|
current = beta * x[j - 1]; |
|
for (n = len + 1; n--; j += 2) { |
|
next = beta * x[j + 1]; |
|
x[j] -= current + next; |
|
if (n--) { |
|
j += 2; |
|
current = beta * x[j + 1]; |
|
x[j] -= current + next; |
|
} else { |
|
break; |
|
} |
|
} |
|
|
|
// step 6 |
|
if (len !== 0) { |
|
j = offset + 1; |
|
current = alpha * x[j - 1]; |
|
for (n = len; n--; j += 2) { |
|
next = alpha * x[j + 1]; |
|
x[j] -= current + next; |
|
if (n--) { |
|
j += 2; |
|
current = alpha * x[j + 1]; |
|
x[j] -= current + next; |
|
} else { |
|
break; |
|
} |
|
} |
|
} |
|
}; |
|
|
|
return IrreversibleTransform; |
|
})(); |
|
|
|
// Section 3.8.1 Reversible 5-3 filter |
|
var ReversibleTransform = (function ReversibleTransformClosure() { |
|
function ReversibleTransform() { |
|
Transform.call(this); |
|
} |
|
|
|
ReversibleTransform.prototype = Object.create(Transform.prototype); |
|
ReversibleTransform.prototype.filter = |
|
function reversibleTransformFilter(x, offset, length) { |
|
var len = length >> 1; |
|
offset = offset | 0; |
|
var j, n; |
|
|
|
for (j = offset, n = len + 1; n--; j += 2) { |
|
x[j] -= (x[j - 1] + x[j + 1] + 2) >> 2; |
|
} |
|
|
|
for (j = offset + 1, n = len; n--; j += 2) { |
|
x[j] += (x[j - 1] + x[j + 1]) >> 1; |
|
} |
|
}; |
|
|
|
return ReversibleTransform; |
|
})(); |
|
|
|
return JpxImage; |
|
})(); |
|
|
|
exports.JpxImage = JpxImage; |
|
})); |
|
|
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreMurmurHash3 = {}), root.pdfjsSharedUtil); |
|
} |
|
}(this, function (exports, sharedUtil) { |
|
|
|
var Uint32ArrayView = sharedUtil.Uint32ArrayView; |
|
|
|
var MurmurHash3_64 = (function MurmurHash3_64Closure (seed) { |
|
// Workaround for missing math precison in JS. |
|
var MASK_HIGH = 0xffff0000; |
|
var MASK_LOW = 0xffff; |
|
|
|
function MurmurHash3_64 (seed) { |
|
var SEED = 0xc3d2e1f0; |
|
this.h1 = seed ? seed & 0xffffffff : SEED; |
|
this.h2 = seed ? seed & 0xffffffff : SEED; |
|
} |
|
|
|
var alwaysUseUint32ArrayView = false; |
|
// old webkits have issues with non-aligned arrays |
|
try { |
|
new Uint32Array(new Uint8Array(5).buffer, 0, 1); |
|
} catch (e) { |
|
alwaysUseUint32ArrayView = true; |
|
} |
|
|
|
MurmurHash3_64.prototype = { |
|
update: function MurmurHash3_64_update(input) { |
|
var useUint32ArrayView = alwaysUseUint32ArrayView; |
|
var i; |
|
if (typeof input === 'string') { |
|
var data = new Uint8Array(input.length * 2); |
|
var length = 0; |
|
for (i = 0; i < input.length; i++) { |
|
var code = input.charCodeAt(i); |
|
if (code <= 0xff) { |
|
data[length++] = code; |
|
} |
|
else { |
|
data[length++] = code >>> 8; |
|
data[length++] = code & 0xff; |
|
} |
|
} |
|
} else if (input instanceof Uint8Array) { |
|
data = input; |
|
length = data.length; |
|
} else if (typeof input === 'object' && ('length' in input)) { |
|
// processing regular arrays as well, e.g. for IE9 |
|
data = input; |
|
length = data.length; |
|
useUint32ArrayView = true; |
|
} else { |
|
throw new Error('Wrong data format in MurmurHash3_64_update. ' + |
|
'Input must be a string or array.'); |
|
} |
|
|
|
var blockCounts = length >> 2; |
|
var tailLength = length - blockCounts * 4; |
|
// we don't care about endianness here |
|
var dataUint32 = useUint32ArrayView ? |
|
new Uint32ArrayView(data, blockCounts) : |
|
new Uint32Array(data.buffer, 0, blockCounts); |
|
var k1 = 0; |
|
var k2 = 0; |
|
var h1 = this.h1; |
|
var h2 = this.h2; |
|
var C1 = 0xcc9e2d51; |
|
var C2 = 0x1b873593; |
|
var C1_LOW = C1 & MASK_LOW; |
|
var C2_LOW = C2 & MASK_LOW; |
|
|
|
for (i = 0; i < blockCounts; i++) { |
|
if (i & 1) { |
|
k1 = dataUint32[i]; |
|
k1 = (k1 * C1 & MASK_HIGH) | (k1 * C1_LOW & MASK_LOW); |
|
k1 = k1 << 15 | k1 >>> 17; |
|
k1 = (k1 * C2 & MASK_HIGH) | (k1 * C2_LOW & MASK_LOW); |
|
h1 ^= k1; |
|
h1 = h1 << 13 | h1 >>> 19; |
|
h1 = h1 * 5 + 0xe6546b64; |
|
} else { |
|
k2 = dataUint32[i]; |
|
k2 = (k2 * C1 & MASK_HIGH) | (k2 * C1_LOW & MASK_LOW); |
|
k2 = k2 << 15 | k2 >>> 17; |
|
k2 = (k2 * C2 & MASK_HIGH) | (k2 * C2_LOW & MASK_LOW); |
|
h2 ^= k2; |
|
h2 = h2 << 13 | h2 >>> 19; |
|
h2 = h2 * 5 + 0xe6546b64; |
|
} |
|
} |
|
|
|
k1 = 0; |
|
|
|
switch (tailLength) { |
|
case 3: |
|
k1 ^= data[blockCounts * 4 + 2] << 16; |
|
/* falls through */ |
|
case 2: |
|
k1 ^= data[blockCounts * 4 + 1] << 8; |
|
/* falls through */ |
|
case 1: |
|
k1 ^= data[blockCounts * 4]; |
|
/* falls through */ |
|
k1 = (k1 * C1 & MASK_HIGH) | (k1 * C1_LOW & MASK_LOW); |
|
k1 = k1 << 15 | k1 >>> 17; |
|
k1 = (k1 * C2 & MASK_HIGH) | (k1 * C2_LOW & MASK_LOW); |
|
if (blockCounts & 1) { |
|
h1 ^= k1; |
|
} else { |
|
h2 ^= k1; |
|
} |
|
} |
|
|
|
this.h1 = h1; |
|
this.h2 = h2; |
|
return this; |
|
}, |
|
|
|
hexdigest: function MurmurHash3_64_hexdigest () { |
|
var h1 = this.h1; |
|
var h2 = this.h2; |
|
|
|
h1 ^= h2 >>> 1; |
|
h1 = (h1 * 0xed558ccd & MASK_HIGH) | (h1 * 0x8ccd & MASK_LOW); |
|
h2 = (h2 * 0xff51afd7 & MASK_HIGH) | |
|
(((h2 << 16 | h1 >>> 16) * 0xafd7ed55 & MASK_HIGH) >>> 16); |
|
h1 ^= h2 >>> 1; |
|
h1 = (h1 * 0x1a85ec53 & MASK_HIGH) | (h1 * 0xec53 & MASK_LOW); |
|
h2 = (h2 * 0xc4ceb9fe & MASK_HIGH) | |
|
(((h2 << 16 | h1 >>> 16) * 0xb9fe1a85 & MASK_HIGH) >>> 16); |
|
h1 ^= h2 >>> 1; |
|
|
|
for (var i = 0, arr = [h1, h2], str = ''; i < arr.length; i++) { |
|
var hex = (arr[i] >>> 0).toString(16); |
|
while (hex.length < 8) { |
|
hex = '0' + hex; |
|
} |
|
str += hex; |
|
} |
|
|
|
return str; |
|
} |
|
}; |
|
|
|
return MurmurHash3_64; |
|
})(); |
|
|
|
exports.MurmurHash3_64 = MurmurHash3_64; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCorePrimitives = {}), root.pdfjsSharedUtil); |
|
} |
|
}(this, function (exports, sharedUtil) { |
|
|
|
var isArray = sharedUtil.isArray; |
|
|
|
var Name = (function NameClosure() { |
|
function Name(name) { |
|
this.name = name; |
|
} |
|
|
|
Name.prototype = {}; |
|
|
|
var nameCache = {}; |
|
|
|
Name.get = function Name_get(name) { |
|
var nameValue = nameCache[name]; |
|
return (nameValue ? nameValue : (nameCache[name] = new Name(name))); |
|
}; |
|
|
|
return Name; |
|
})(); |
|
|
|
var Cmd = (function CmdClosure() { |
|
function Cmd(cmd) { |
|
this.cmd = cmd; |
|
} |
|
|
|
Cmd.prototype = {}; |
|
|
|
var cmdCache = {}; |
|
|
|
Cmd.get = function Cmd_get(cmd) { |
|
var cmdValue = cmdCache[cmd]; |
|
return (cmdValue ? cmdValue : (cmdCache[cmd] = new Cmd(cmd))); |
|
}; |
|
|
|
return Cmd; |
|
})(); |
|
|
|
var Dict = (function DictClosure() { |
|
var nonSerializable = function nonSerializableClosure() { |
|
return nonSerializable; // creating closure on some variable |
|
}; |
|
|
|
var GETALL_DICTIONARY_TYPES_WHITELIST = { |
|
'Background': true, |
|
'ExtGState': true, |
|
'Halftone': true, |
|
'Layout': true, |
|
'Mask': true, |
|
'Pagination': true, |
|
'Printing': true |
|
}; |
|
|
|
function isRecursionAllowedFor(dict) { |
|
if (!isName(dict.Type)) { |
|
return true; |
|
} |
|
var dictType = dict.Type.name; |
|
return GETALL_DICTIONARY_TYPES_WHITELIST[dictType] === true; |
|
} |
|
|
|
// xref is optional |
|
function Dict(xref) { |
|
// Map should only be used internally, use functions below to access. |
|
this.map = Object.create(null); |
|
this.xref = xref; |
|
this.objId = null; |
|
this.__nonSerializable__ = nonSerializable; // disable cloning of the Dict |
|
} |
|
|
|
Dict.prototype = { |
|
assignXref: function Dict_assignXref(newXref) { |
|
this.xref = newXref; |
|
}, |
|
|
|
// automatically dereferences Ref objects |
|
get: function Dict_get(key1, key2, key3) { |
|
var value; |
|
var xref = this.xref; |
|
if (typeof (value = this.map[key1]) !== 'undefined' || key1 in this.map || |
|
typeof key2 === 'undefined') { |
|
return xref ? xref.fetchIfRef(value) : value; |
|
} |
|
if (typeof (value = this.map[key2]) !== 'undefined' || key2 in this.map || |
|
typeof key3 === 'undefined') { |
|
return xref ? xref.fetchIfRef(value) : value; |
|
} |
|
value = this.map[key3] || null; |
|
return xref ? xref.fetchIfRef(value) : value; |
|
}, |
|
|
|
// Same as get(), but returns a promise and uses fetchIfRefAsync(). |
|
getAsync: function Dict_getAsync(key1, key2, key3) { |
|
var value; |
|
var xref = this.xref; |
|
if (typeof (value = this.map[key1]) !== 'undefined' || key1 in this.map || |
|
typeof key2 === 'undefined') { |
|
if (xref) { |
|
return xref.fetchIfRefAsync(value); |
|
} |
|
return Promise.resolve(value); |
|
} |
|
if (typeof (value = this.map[key2]) !== 'undefined' || key2 in this.map || |
|
typeof key3 === 'undefined') { |
|
if (xref) { |
|
return xref.fetchIfRefAsync(value); |
|
} |
|
return Promise.resolve(value); |
|
} |
|
value = this.map[key3] || null; |
|
if (xref) { |
|
return xref.fetchIfRefAsync(value); |
|
} |
|
return Promise.resolve(value); |
|
}, |
|
|
|
// Same as get(), but dereferences all elements if the result is an Array. |
|
getArray: function Dict_getArray(key1, key2, key3) { |
|
var value = this.get(key1, key2, key3); |
|
var xref = this.xref; |
|
if (!isArray(value) || !xref) { |
|
return value; |
|
} |
|
value = value.slice(); // Ensure that we don't modify the Dict data. |
|
for (var i = 0, ii = value.length; i < ii; i++) { |
|
if (!isRef(value[i])) { |
|
continue; |
|
} |
|
value[i] = xref.fetch(value[i]); |
|
} |
|
return value; |
|
}, |
|
|
|
// no dereferencing |
|
getRaw: function Dict_getRaw(key) { |
|
return this.map[key]; |
|
}, |
|
|
|
// creates new map and dereferences all Refs |
|
getAll: function Dict_getAll() { |
|
var all = Object.create(null); |
|
var queue = null; |
|
var key, obj; |
|
for (key in this.map) { |
|
obj = this.get(key); |
|
if (obj instanceof Dict) { |
|
if (isRecursionAllowedFor(obj)) { |
|
(queue || (queue = [])).push({target: all, key: key, obj: obj}); |
|
} else { |
|
all[key] = this.getRaw(key); |
|
} |
|
} else { |
|
all[key] = obj; |
|
} |
|
} |
|
if (!queue) { |
|
return all; |
|
} |
|
|
|
// trying to take cyclic references into the account |
|
var processed = Object.create(null); |
|
while (queue.length > 0) { |
|
var item = queue.shift(); |
|
var itemObj = item.obj; |
|
var objId = itemObj.objId; |
|
if (objId && objId in processed) { |
|
item.target[item.key] = processed[objId]; |
|
continue; |
|
} |
|
var dereferenced = Object.create(null); |
|
for (key in itemObj.map) { |
|
obj = itemObj.get(key); |
|
if (obj instanceof Dict) { |
|
if (isRecursionAllowedFor(obj)) { |
|
queue.push({target: dereferenced, key: key, obj: obj}); |
|
} else { |
|
dereferenced[key] = itemObj.getRaw(key); |
|
} |
|
} else { |
|
dereferenced[key] = obj; |
|
} |
|
} |
|
if (objId) { |
|
processed[objId] = dereferenced; |
|
} |
|
item.target[item.key] = dereferenced; |
|
} |
|
return all; |
|
}, |
|
|
|
getKeys: function Dict_getKeys() { |
|
return Object.keys(this.map); |
|
}, |
|
|
|
set: function Dict_set(key, value) { |
|
this.map[key] = value; |
|
}, |
|
|
|
has: function Dict_has(key) { |
|
return key in this.map; |
|
}, |
|
|
|
forEach: function Dict_forEach(callback) { |
|
for (var key in this.map) { |
|
callback(key, this.get(key)); |
|
} |
|
} |
|
}; |
|
|
|
Dict.empty = new Dict(null); |
|
|
|
Dict.merge = function Dict_merge(xref, dictArray) { |
|
var mergedDict = new Dict(xref); |
|
|
|
for (var i = 0, ii = dictArray.length; i < ii; i++) { |
|
var dict = dictArray[i]; |
|
if (!isDict(dict)) { |
|
continue; |
|
} |
|
for (var keyName in dict.map) { |
|
if (mergedDict.map[keyName]) { |
|
continue; |
|
} |
|
mergedDict.map[keyName] = dict.map[keyName]; |
|
} |
|
} |
|
return mergedDict; |
|
}; |
|
|
|
return Dict; |
|
})(); |
|
|
|
var Ref = (function RefClosure() { |
|
function Ref(num, gen) { |
|
this.num = num; |
|
this.gen = gen; |
|
} |
|
|
|
Ref.prototype = { |
|
toString: function Ref_toString() { |
|
// This function is hot, so we make the string as compact as possible. |
|
// |this.gen| is almost always zero, so we treat that case specially. |
|
var str = this.num + 'R'; |
|
if (this.gen !== 0) { |
|
str += this.gen; |
|
} |
|
return str; |
|
} |
|
}; |
|
|
|
return Ref; |
|
})(); |
|
|
|
// The reference is identified by number and generation. |
|
// This structure stores only one instance of the reference. |
|
var RefSet = (function RefSetClosure() { |
|
function RefSet() { |
|
this.dict = {}; |
|
} |
|
|
|
RefSet.prototype = { |
|
has: function RefSet_has(ref) { |
|
return ref.toString() in this.dict; |
|
}, |
|
|
|
put: function RefSet_put(ref) { |
|
this.dict[ref.toString()] = true; |
|
}, |
|
|
|
remove: function RefSet_remove(ref) { |
|
delete this.dict[ref.toString()]; |
|
} |
|
}; |
|
|
|
return RefSet; |
|
})(); |
|
|
|
var RefSetCache = (function RefSetCacheClosure() { |
|
function RefSetCache() { |
|
this.dict = Object.create(null); |
|
} |
|
|
|
RefSetCache.prototype = { |
|
get: function RefSetCache_get(ref) { |
|
return this.dict[ref.toString()]; |
|
}, |
|
|
|
has: function RefSetCache_has(ref) { |
|
return ref.toString() in this.dict; |
|
}, |
|
|
|
put: function RefSetCache_put(ref, obj) { |
|
this.dict[ref.toString()] = obj; |
|
}, |
|
|
|
putAlias: function RefSetCache_putAlias(ref, aliasRef) { |
|
this.dict[ref.toString()] = this.get(aliasRef); |
|
}, |
|
|
|
forEach: function RefSetCache_forEach(fn, thisArg) { |
|
for (var i in this.dict) { |
|
fn.call(thisArg, this.dict[i]); |
|
} |
|
}, |
|
|
|
clear: function RefSetCache_clear() { |
|
this.dict = Object.create(null); |
|
} |
|
}; |
|
|
|
return RefSetCache; |
|
})(); |
|
|
|
function isName(v) { |
|
return v instanceof Name; |
|
} |
|
|
|
function isCmd(v, cmd) { |
|
return v instanceof Cmd && (cmd === undefined || v.cmd === cmd); |
|
} |
|
|
|
function isDict(v, type) { |
|
if (!(v instanceof Dict)) { |
|
return false; |
|
} |
|
if (!type) { |
|
return true; |
|
} |
|
var dictType = v.get('Type'); |
|
return isName(dictType) && dictType.name === type; |
|
} |
|
|
|
function isRef(v) { |
|
return v instanceof Ref; |
|
} |
|
|
|
function isStream(v) { |
|
return typeof v === 'object' && v !== null && v.getBytes !== undefined; |
|
} |
|
|
|
exports.Cmd = Cmd; |
|
exports.Dict = Dict; |
|
exports.Name = Name; |
|
exports.Ref = Ref; |
|
exports.RefSet = RefSet; |
|
exports.RefSetCache = RefSetCache; |
|
exports.isCmd = isCmd; |
|
exports.isDict = isDict; |
|
exports.isName = isName; |
|
exports.isRef = isRef; |
|
exports.isStream = isStream; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreStream = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreJbig2, root.pdfjsCoreJpg, |
|
root.pdfjsCoreJpx); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreJbig2, coreJpg, |
|
coreJpx) { |
|
|
|
var Util = sharedUtil.Util; |
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isArray = sharedUtil.isArray; |
|
var shadow = sharedUtil.shadow; |
|
var warn = sharedUtil.warn; |
|
var Dict = corePrimitives.Dict; |
|
var Jbig2Image = coreJbig2.Jbig2Image; |
|
var JpegImage = coreJpg.JpegImage; |
|
var JpxImage = coreJpx.JpxImage; |
|
|
|
var coreParser; // see _setCoreParser below |
|
var EOF; // = coreParser.EOF; |
|
var Lexer; // = coreParser.Lexer; |
|
|
|
var coreColorSpace; // see _setCoreColorSpace below |
|
var ColorSpace; // = coreColorSpace.ColorSpace; |
|
|
|
var Stream = (function StreamClosure() { |
|
function Stream(arrayBuffer, start, length, dict) { |
|
this.bytes = (arrayBuffer instanceof Uint8Array ? |
|
arrayBuffer : new Uint8Array(arrayBuffer)); |
|
this.start = start || 0; |
|
this.pos = this.start; |
|
this.end = (start + length) || this.bytes.length; |
|
this.dict = dict; |
|
} |
|
|
|
// required methods for a stream. if a particular stream does not |
|
// implement these, an error should be thrown |
|
Stream.prototype = { |
|
get length() { |
|
return this.end - this.start; |
|
}, |
|
get isEmpty() { |
|
return this.length === 0; |
|
}, |
|
getByte: function Stream_getByte() { |
|
if (this.pos >= this.end) { |
|
return -1; |
|
} |
|
return this.bytes[this.pos++]; |
|
}, |
|
getUint16: function Stream_getUint16() { |
|
var b0 = this.getByte(); |
|
var b1 = this.getByte(); |
|
if (b0 === -1 || b1 === -1) { |
|
return -1; |
|
} |
|
return (b0 << 8) + b1; |
|
}, |
|
getInt32: function Stream_getInt32() { |
|
var b0 = this.getByte(); |
|
var b1 = this.getByte(); |
|
var b2 = this.getByte(); |
|
var b3 = this.getByte(); |
|
return (b0 << 24) + (b1 << 16) + (b2 << 8) + b3; |
|
}, |
|
// returns subarray of original buffer |
|
// should only be read |
|
getBytes: function Stream_getBytes(length) { |
|
var bytes = this.bytes; |
|
var pos = this.pos; |
|
var strEnd = this.end; |
|
|
|
if (!length) { |
|
return bytes.subarray(pos, strEnd); |
|
} |
|
var end = pos + length; |
|
if (end > strEnd) { |
|
end = strEnd; |
|
} |
|
this.pos = end; |
|
return bytes.subarray(pos, end); |
|
}, |
|
peekByte: function Stream_peekByte() { |
|
var peekedByte = this.getByte(); |
|
this.pos--; |
|
return peekedByte; |
|
}, |
|
peekBytes: function Stream_peekBytes(length) { |
|
var bytes = this.getBytes(length); |
|
this.pos -= bytes.length; |
|
return bytes; |
|
}, |
|
skip: function Stream_skip(n) { |
|
if (!n) { |
|
n = 1; |
|
} |
|
this.pos += n; |
|
}, |
|
reset: function Stream_reset() { |
|
this.pos = this.start; |
|
}, |
|
moveStart: function Stream_moveStart() { |
|
this.start = this.pos; |
|
}, |
|
makeSubStream: function Stream_makeSubStream(start, length, dict) { |
|
return new Stream(this.bytes.buffer, start, length, dict); |
|
}, |
|
isStream: true |
|
}; |
|
|
|
return Stream; |
|
})(); |
|
|
|
var StringStream = (function StringStreamClosure() { |
|
function StringStream(str) { |
|
var length = str.length; |
|
var bytes = new Uint8Array(length); |
|
for (var n = 0; n < length; ++n) { |
|
bytes[n] = str.charCodeAt(n); |
|
} |
|
Stream.call(this, bytes); |
|
} |
|
|
|
StringStream.prototype = Stream.prototype; |
|
|
|
return StringStream; |
|
})(); |
|
|
|
// super class for the decoding streams |
|
var DecodeStream = (function DecodeStreamClosure() { |
|
// Lots of DecodeStreams are created whose buffers are never used. For these |
|
// we share a single empty buffer. This is (a) space-efficient and (b) avoids |
|
// having special cases that would be required if we used |null| for an empty |
|
// buffer. |
|
var emptyBuffer = new Uint8Array(0); |
|
|
|
function DecodeStream(maybeMinBufferLength) { |
|
this.pos = 0; |
|
this.bufferLength = 0; |
|
this.eof = false; |
|
this.buffer = emptyBuffer; |
|
this.minBufferLength = 512; |
|
if (maybeMinBufferLength) { |
|
// Compute the first power of two that is as big as maybeMinBufferLength. |
|
while (this.minBufferLength < maybeMinBufferLength) { |
|
this.minBufferLength *= 2; |
|
} |
|
} |
|
} |
|
|
|
DecodeStream.prototype = { |
|
get isEmpty() { |
|
while (!this.eof && this.bufferLength === 0) { |
|
this.readBlock(); |
|
} |
|
return this.bufferLength === 0; |
|
}, |
|
ensureBuffer: function DecodeStream_ensureBuffer(requested) { |
|
var buffer = this.buffer; |
|
if (requested <= buffer.byteLength) { |
|
return buffer; |
|
} |
|
var size = this.minBufferLength; |
|
while (size < requested) { |
|
size *= 2; |
|
} |
|
var buffer2 = new Uint8Array(size); |
|
buffer2.set(buffer); |
|
return (this.buffer = buffer2); |
|
}, |
|
getByte: function DecodeStream_getByte() { |
|
var pos = this.pos; |
|
while (this.bufferLength <= pos) { |
|
if (this.eof) { |
|
return -1; |
|
} |
|
this.readBlock(); |
|
} |
|
return this.buffer[this.pos++]; |
|
}, |
|
getUint16: function DecodeStream_getUint16() { |
|
var b0 = this.getByte(); |
|
var b1 = this.getByte(); |
|
if (b0 === -1 || b1 === -1) { |
|
return -1; |
|
} |
|
return (b0 << 8) + b1; |
|
}, |
|
getInt32: function DecodeStream_getInt32() { |
|
var b0 = this.getByte(); |
|
var b1 = this.getByte(); |
|
var b2 = this.getByte(); |
|
var b3 = this.getByte(); |
|
return (b0 << 24) + (b1 << 16) + (b2 << 8) + b3; |
|
}, |
|
getBytes: function DecodeStream_getBytes(length) { |
|
var end, pos = this.pos; |
|
|
|
if (length) { |
|
this.ensureBuffer(pos + length); |
|
end = pos + length; |
|
|
|
while (!this.eof && this.bufferLength < end) { |
|
this.readBlock(); |
|
} |
|
var bufEnd = this.bufferLength; |
|
if (end > bufEnd) { |
|
end = bufEnd; |
|
} |
|
} else { |
|
while (!this.eof) { |
|
this.readBlock(); |
|
} |
|
end = this.bufferLength; |
|
} |
|
|
|
this.pos = end; |
|
return this.buffer.subarray(pos, end); |
|
}, |
|
peekByte: function DecodeStream_peekByte() { |
|
var peekedByte = this.getByte(); |
|
this.pos--; |
|
return peekedByte; |
|
}, |
|
peekBytes: function DecodeStream_peekBytes(length) { |
|
var bytes = this.getBytes(length); |
|
this.pos -= bytes.length; |
|
return bytes; |
|
}, |
|
makeSubStream: function DecodeStream_makeSubStream(start, length, dict) { |
|
var end = start + length; |
|
while (this.bufferLength <= end && !this.eof) { |
|
this.readBlock(); |
|
} |
|
return new Stream(this.buffer, start, length, dict); |
|
}, |
|
skip: function DecodeStream_skip(n) { |
|
if (!n) { |
|
n = 1; |
|
} |
|
this.pos += n; |
|
}, |
|
reset: function DecodeStream_reset() { |
|
this.pos = 0; |
|
}, |
|
getBaseStreams: function DecodeStream_getBaseStreams() { |
|
if (this.str && this.str.getBaseStreams) { |
|
return this.str.getBaseStreams(); |
|
} |
|
return []; |
|
} |
|
}; |
|
|
|
return DecodeStream; |
|
})(); |
|
|
|
var StreamsSequenceStream = (function StreamsSequenceStreamClosure() { |
|
function StreamsSequenceStream(streams) { |
|
this.streams = streams; |
|
DecodeStream.call(this, /* maybeLength = */ null); |
|
} |
|
|
|
StreamsSequenceStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
StreamsSequenceStream.prototype.readBlock = |
|
function streamSequenceStreamReadBlock() { |
|
|
|
var streams = this.streams; |
|
if (streams.length === 0) { |
|
this.eof = true; |
|
return; |
|
} |
|
var stream = streams.shift(); |
|
var chunk = stream.getBytes(); |
|
var bufferLength = this.bufferLength; |
|
var newLength = bufferLength + chunk.length; |
|
var buffer = this.ensureBuffer(newLength); |
|
buffer.set(chunk, bufferLength); |
|
this.bufferLength = newLength; |
|
}; |
|
|
|
StreamsSequenceStream.prototype.getBaseStreams = |
|
function StreamsSequenceStream_getBaseStreams() { |
|
|
|
var baseStreams = []; |
|
for (var i = 0, ii = this.streams.length; i < ii; i++) { |
|
var stream = this.streams[i]; |
|
if (stream.getBaseStreams) { |
|
Util.appendToArray(baseStreams, stream.getBaseStreams()); |
|
} |
|
} |
|
return baseStreams; |
|
}; |
|
|
|
return StreamsSequenceStream; |
|
})(); |
|
|
|
var FlateStream = (function FlateStreamClosure() { |
|
var codeLenCodeMap = new Int32Array([ |
|
16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15 |
|
]); |
|
|
|
var lengthDecode = new Int32Array([ |
|
0x00003, 0x00004, 0x00005, 0x00006, 0x00007, 0x00008, 0x00009, 0x0000a, |
|
0x1000b, 0x1000d, 0x1000f, 0x10011, 0x20013, 0x20017, 0x2001b, 0x2001f, |
|
0x30023, 0x3002b, 0x30033, 0x3003b, 0x40043, 0x40053, 0x40063, 0x40073, |
|
0x50083, 0x500a3, 0x500c3, 0x500e3, 0x00102, 0x00102, 0x00102 |
|
]); |
|
|
|
var distDecode = new Int32Array([ |
|
0x00001, 0x00002, 0x00003, 0x00004, 0x10005, 0x10007, 0x20009, 0x2000d, |
|
0x30011, 0x30019, 0x40021, 0x40031, 0x50041, 0x50061, 0x60081, 0x600c1, |
|
0x70101, 0x70181, 0x80201, 0x80301, 0x90401, 0x90601, 0xa0801, 0xa0c01, |
|
0xb1001, 0xb1801, 0xc2001, 0xc3001, 0xd4001, 0xd6001 |
|
]); |
|
|
|
var fixedLitCodeTab = [new Int32Array([ |
|
0x70100, 0x80050, 0x80010, 0x80118, 0x70110, 0x80070, 0x80030, 0x900c0, |
|
0x70108, 0x80060, 0x80020, 0x900a0, 0x80000, 0x80080, 0x80040, 0x900e0, |
|
0x70104, 0x80058, 0x80018, 0x90090, 0x70114, 0x80078, 0x80038, 0x900d0, |
|
0x7010c, 0x80068, 0x80028, 0x900b0, 0x80008, 0x80088, 0x80048, 0x900f0, |
|
0x70102, 0x80054, 0x80014, 0x8011c, 0x70112, 0x80074, 0x80034, 0x900c8, |
|
0x7010a, 0x80064, 0x80024, 0x900a8, 0x80004, 0x80084, 0x80044, 0x900e8, |
|
0x70106, 0x8005c, 0x8001c, 0x90098, 0x70116, 0x8007c, 0x8003c, 0x900d8, |
|
0x7010e, 0x8006c, 0x8002c, 0x900b8, 0x8000c, 0x8008c, 0x8004c, 0x900f8, |
|
0x70101, 0x80052, 0x80012, 0x8011a, 0x70111, 0x80072, 0x80032, 0x900c4, |
|
0x70109, 0x80062, 0x80022, 0x900a4, 0x80002, 0x80082, 0x80042, 0x900e4, |
|
0x70105, 0x8005a, 0x8001a, 0x90094, 0x70115, 0x8007a, 0x8003a, 0x900d4, |
|
0x7010d, 0x8006a, 0x8002a, 0x900b4, 0x8000a, 0x8008a, 0x8004a, 0x900f4, |
|
0x70103, 0x80056, 0x80016, 0x8011e, 0x70113, 0x80076, 0x80036, 0x900cc, |
|
0x7010b, 0x80066, 0x80026, 0x900ac, 0x80006, 0x80086, 0x80046, 0x900ec, |
|
0x70107, 0x8005e, 0x8001e, 0x9009c, 0x70117, 0x8007e, 0x8003e, 0x900dc, |
|
0x7010f, 0x8006e, 0x8002e, 0x900bc, 0x8000e, 0x8008e, 0x8004e, 0x900fc, |
|
0x70100, 0x80051, 0x80011, 0x80119, 0x70110, 0x80071, 0x80031, 0x900c2, |
|
0x70108, 0x80061, 0x80021, 0x900a2, 0x80001, 0x80081, 0x80041, 0x900e2, |
|
0x70104, 0x80059, 0x80019, 0x90092, 0x70114, 0x80079, 0x80039, 0x900d2, |
|
0x7010c, 0x80069, 0x80029, 0x900b2, 0x80009, 0x80089, 0x80049, 0x900f2, |
|
0x70102, 0x80055, 0x80015, 0x8011d, 0x70112, 0x80075, 0x80035, 0x900ca, |
|
0x7010a, 0x80065, 0x80025, 0x900aa, 0x80005, 0x80085, 0x80045, 0x900ea, |
|
0x70106, 0x8005d, 0x8001d, 0x9009a, 0x70116, 0x8007d, 0x8003d, 0x900da, |
|
0x7010e, 0x8006d, 0x8002d, 0x900ba, 0x8000d, 0x8008d, 0x8004d, 0x900fa, |
|
0x70101, 0x80053, 0x80013, 0x8011b, 0x70111, 0x80073, 0x80033, 0x900c6, |
|
0x70109, 0x80063, 0x80023, 0x900a6, 0x80003, 0x80083, 0x80043, 0x900e6, |
|
0x70105, 0x8005b, 0x8001b, 0x90096, 0x70115, 0x8007b, 0x8003b, 0x900d6, |
|
0x7010d, 0x8006b, 0x8002b, 0x900b6, 0x8000b, 0x8008b, 0x8004b, 0x900f6, |
|
0x70103, 0x80057, 0x80017, 0x8011f, 0x70113, 0x80077, 0x80037, 0x900ce, |
|
0x7010b, 0x80067, 0x80027, 0x900ae, 0x80007, 0x80087, 0x80047, 0x900ee, |
|
0x70107, 0x8005f, 0x8001f, 0x9009e, 0x70117, 0x8007f, 0x8003f, 0x900de, |
|
0x7010f, 0x8006f, 0x8002f, 0x900be, 0x8000f, 0x8008f, 0x8004f, 0x900fe, |
|
0x70100, 0x80050, 0x80010, 0x80118, 0x70110, 0x80070, 0x80030, 0x900c1, |
|
0x70108, 0x80060, 0x80020, 0x900a1, 0x80000, 0x80080, 0x80040, 0x900e1, |
|
0x70104, 0x80058, 0x80018, 0x90091, 0x70114, 0x80078, 0x80038, 0x900d1, |
|
0x7010c, 0x80068, 0x80028, 0x900b1, 0x80008, 0x80088, 0x80048, 0x900f1, |
|
0x70102, 0x80054, 0x80014, 0x8011c, 0x70112, 0x80074, 0x80034, 0x900c9, |
|
0x7010a, 0x80064, 0x80024, 0x900a9, 0x80004, 0x80084, 0x80044, 0x900e9, |
|
0x70106, 0x8005c, 0x8001c, 0x90099, 0x70116, 0x8007c, 0x8003c, 0x900d9, |
|
0x7010e, 0x8006c, 0x8002c, 0x900b9, 0x8000c, 0x8008c, 0x8004c, 0x900f9, |
|
0x70101, 0x80052, 0x80012, 0x8011a, 0x70111, 0x80072, 0x80032, 0x900c5, |
|
0x70109, 0x80062, 0x80022, 0x900a5, 0x80002, 0x80082, 0x80042, 0x900e5, |
|
0x70105, 0x8005a, 0x8001a, 0x90095, 0x70115, 0x8007a, 0x8003a, 0x900d5, |
|
0x7010d, 0x8006a, 0x8002a, 0x900b5, 0x8000a, 0x8008a, 0x8004a, 0x900f5, |
|
0x70103, 0x80056, 0x80016, 0x8011e, 0x70113, 0x80076, 0x80036, 0x900cd, |
|
0x7010b, 0x80066, 0x80026, 0x900ad, 0x80006, 0x80086, 0x80046, 0x900ed, |
|
0x70107, 0x8005e, 0x8001e, 0x9009d, 0x70117, 0x8007e, 0x8003e, 0x900dd, |
|
0x7010f, 0x8006e, 0x8002e, 0x900bd, 0x8000e, 0x8008e, 0x8004e, 0x900fd, |
|
0x70100, 0x80051, 0x80011, 0x80119, 0x70110, 0x80071, 0x80031, 0x900c3, |
|
0x70108, 0x80061, 0x80021, 0x900a3, 0x80001, 0x80081, 0x80041, 0x900e3, |
|
0x70104, 0x80059, 0x80019, 0x90093, 0x70114, 0x80079, 0x80039, 0x900d3, |
|
0x7010c, 0x80069, 0x80029, 0x900b3, 0x80009, 0x80089, 0x80049, 0x900f3, |
|
0x70102, 0x80055, 0x80015, 0x8011d, 0x70112, 0x80075, 0x80035, 0x900cb, |
|
0x7010a, 0x80065, 0x80025, 0x900ab, 0x80005, 0x80085, 0x80045, 0x900eb, |
|
0x70106, 0x8005d, 0x8001d, 0x9009b, 0x70116, 0x8007d, 0x8003d, 0x900db, |
|
0x7010e, 0x8006d, 0x8002d, 0x900bb, 0x8000d, 0x8008d, 0x8004d, 0x900fb, |
|
0x70101, 0x80053, 0x80013, 0x8011b, 0x70111, 0x80073, 0x80033, 0x900c7, |
|
0x70109, 0x80063, 0x80023, 0x900a7, 0x80003, 0x80083, 0x80043, 0x900e7, |
|
0x70105, 0x8005b, 0x8001b, 0x90097, 0x70115, 0x8007b, 0x8003b, 0x900d7, |
|
0x7010d, 0x8006b, 0x8002b, 0x900b7, 0x8000b, 0x8008b, 0x8004b, 0x900f7, |
|
0x70103, 0x80057, 0x80017, 0x8011f, 0x70113, 0x80077, 0x80037, 0x900cf, |
|
0x7010b, 0x80067, 0x80027, 0x900af, 0x80007, 0x80087, 0x80047, 0x900ef, |
|
0x70107, 0x8005f, 0x8001f, 0x9009f, 0x70117, 0x8007f, 0x8003f, 0x900df, |
|
0x7010f, 0x8006f, 0x8002f, 0x900bf, 0x8000f, 0x8008f, 0x8004f, 0x900ff |
|
]), 9]; |
|
|
|
var fixedDistCodeTab = [new Int32Array([ |
|
0x50000, 0x50010, 0x50008, 0x50018, 0x50004, 0x50014, 0x5000c, 0x5001c, |
|
0x50002, 0x50012, 0x5000a, 0x5001a, 0x50006, 0x50016, 0x5000e, 0x00000, |
|
0x50001, 0x50011, 0x50009, 0x50019, 0x50005, 0x50015, 0x5000d, 0x5001d, |
|
0x50003, 0x50013, 0x5000b, 0x5001b, 0x50007, 0x50017, 0x5000f, 0x00000 |
|
]), 5]; |
|
|
|
function FlateStream(str, maybeLength) { |
|
this.str = str; |
|
this.dict = str.dict; |
|
|
|
var cmf = str.getByte(); |
|
var flg = str.getByte(); |
|
if (cmf === -1 || flg === -1) { |
|
error('Invalid header in flate stream: ' + cmf + ', ' + flg); |
|
} |
|
if ((cmf & 0x0f) !== 0x08) { |
|
error('Unknown compression method in flate stream: ' + cmf + ', ' + flg); |
|
} |
|
if ((((cmf << 8) + flg) % 31) !== 0) { |
|
error('Bad FCHECK in flate stream: ' + cmf + ', ' + flg); |
|
} |
|
if (flg & 0x20) { |
|
error('FDICT bit set in flate stream: ' + cmf + ', ' + flg); |
|
} |
|
|
|
this.codeSize = 0; |
|
this.codeBuf = 0; |
|
|
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
FlateStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
FlateStream.prototype.getBits = function FlateStream_getBits(bits) { |
|
var str = this.str; |
|
var codeSize = this.codeSize; |
|
var codeBuf = this.codeBuf; |
|
|
|
var b; |
|
while (codeSize < bits) { |
|
if ((b = str.getByte()) === -1) { |
|
error('Bad encoding in flate stream'); |
|
} |
|
codeBuf |= b << codeSize; |
|
codeSize += 8; |
|
} |
|
b = codeBuf & ((1 << bits) - 1); |
|
this.codeBuf = codeBuf >> bits; |
|
this.codeSize = codeSize -= bits; |
|
|
|
return b; |
|
}; |
|
|
|
FlateStream.prototype.getCode = function FlateStream_getCode(table) { |
|
var str = this.str; |
|
var codes = table[0]; |
|
var maxLen = table[1]; |
|
var codeSize = this.codeSize; |
|
var codeBuf = this.codeBuf; |
|
|
|
var b; |
|
while (codeSize < maxLen) { |
|
if ((b = str.getByte()) === -1) { |
|
// premature end of stream. code might however still be valid. |
|
// codeSize < codeLen check below guards against incomplete codeVal. |
|
break; |
|
} |
|
codeBuf |= (b << codeSize); |
|
codeSize += 8; |
|
} |
|
var code = codes[codeBuf & ((1 << maxLen) - 1)]; |
|
var codeLen = code >> 16; |
|
var codeVal = code & 0xffff; |
|
if (codeLen < 1 || codeSize < codeLen) { |
|
error('Bad encoding in flate stream'); |
|
} |
|
this.codeBuf = (codeBuf >> codeLen); |
|
this.codeSize = (codeSize - codeLen); |
|
return codeVal; |
|
}; |
|
|
|
FlateStream.prototype.generateHuffmanTable = |
|
function flateStreamGenerateHuffmanTable(lengths) { |
|
var n = lengths.length; |
|
|
|
// find max code length |
|
var maxLen = 0; |
|
var i; |
|
for (i = 0; i < n; ++i) { |
|
if (lengths[i] > maxLen) { |
|
maxLen = lengths[i]; |
|
} |
|
} |
|
|
|
// build the table |
|
var size = 1 << maxLen; |
|
var codes = new Int32Array(size); |
|
for (var len = 1, code = 0, skip = 2; |
|
len <= maxLen; |
|
++len, code <<= 1, skip <<= 1) { |
|
for (var val = 0; val < n; ++val) { |
|
if (lengths[val] === len) { |
|
// bit-reverse the code |
|
var code2 = 0; |
|
var t = code; |
|
for (i = 0; i < len; ++i) { |
|
code2 = (code2 << 1) | (t & 1); |
|
t >>= 1; |
|
} |
|
|
|
// fill the table entries |
|
for (i = code2; i < size; i += skip) { |
|
codes[i] = (len << 16) | val; |
|
} |
|
++code; |
|
} |
|
} |
|
} |
|
|
|
return [codes, maxLen]; |
|
}; |
|
|
|
FlateStream.prototype.readBlock = function FlateStream_readBlock() { |
|
var buffer, len; |
|
var str = this.str; |
|
// read block header |
|
var hdr = this.getBits(3); |
|
if (hdr & 1) { |
|
this.eof = true; |
|
} |
|
hdr >>= 1; |
|
|
|
if (hdr === 0) { // uncompressed block |
|
var b; |
|
|
|
if ((b = str.getByte()) === -1) { |
|
error('Bad block header in flate stream'); |
|
} |
|
var blockLen = b; |
|
if ((b = str.getByte()) === -1) { |
|
error('Bad block header in flate stream'); |
|
} |
|
blockLen |= (b << 8); |
|
if ((b = str.getByte()) === -1) { |
|
error('Bad block header in flate stream'); |
|
} |
|
var check = b; |
|
if ((b = str.getByte()) === -1) { |
|
error('Bad block header in flate stream'); |
|
} |
|
check |= (b << 8); |
|
if (check !== (~blockLen & 0xffff) && |
|
(blockLen !== 0 || check !== 0)) { |
|
// Ignoring error for bad "empty" block (see issue 1277) |
|
error('Bad uncompressed block length in flate stream'); |
|
} |
|
|
|
this.codeBuf = 0; |
|
this.codeSize = 0; |
|
|
|
var bufferLength = this.bufferLength; |
|
buffer = this.ensureBuffer(bufferLength + blockLen); |
|
var end = bufferLength + blockLen; |
|
this.bufferLength = end; |
|
if (blockLen === 0) { |
|
if (str.peekByte() === -1) { |
|
this.eof = true; |
|
} |
|
} else { |
|
for (var n = bufferLength; n < end; ++n) { |
|
if ((b = str.getByte()) === -1) { |
|
this.eof = true; |
|
break; |
|
} |
|
buffer[n] = b; |
|
} |
|
} |
|
return; |
|
} |
|
|
|
var litCodeTable; |
|
var distCodeTable; |
|
if (hdr === 1) { // compressed block, fixed codes |
|
litCodeTable = fixedLitCodeTab; |
|
distCodeTable = fixedDistCodeTab; |
|
} else if (hdr === 2) { // compressed block, dynamic codes |
|
var numLitCodes = this.getBits(5) + 257; |
|
var numDistCodes = this.getBits(5) + 1; |
|
var numCodeLenCodes = this.getBits(4) + 4; |
|
|
|
// build the code lengths code table |
|
var codeLenCodeLengths = new Uint8Array(codeLenCodeMap.length); |
|
|
|
var i; |
|
for (i = 0; i < numCodeLenCodes; ++i) { |
|
codeLenCodeLengths[codeLenCodeMap[i]] = this.getBits(3); |
|
} |
|
var codeLenCodeTab = this.generateHuffmanTable(codeLenCodeLengths); |
|
|
|
// build the literal and distance code tables |
|
len = 0; |
|
i = 0; |
|
var codes = numLitCodes + numDistCodes; |
|
var codeLengths = new Uint8Array(codes); |
|
var bitsLength, bitsOffset, what; |
|
while (i < codes) { |
|
var code = this.getCode(codeLenCodeTab); |
|
if (code === 16) { |
|
bitsLength = 2; bitsOffset = 3; what = len; |
|
} else if (code === 17) { |
|
bitsLength = 3; bitsOffset = 3; what = (len = 0); |
|
} else if (code === 18) { |
|
bitsLength = 7; bitsOffset = 11; what = (len = 0); |
|
} else { |
|
codeLengths[i++] = len = code; |
|
continue; |
|
} |
|
|
|
var repeatLength = this.getBits(bitsLength) + bitsOffset; |
|
while (repeatLength-- > 0) { |
|
codeLengths[i++] = what; |
|
} |
|
} |
|
|
|
litCodeTable = |
|
this.generateHuffmanTable(codeLengths.subarray(0, numLitCodes)); |
|
distCodeTable = |
|
this.generateHuffmanTable(codeLengths.subarray(numLitCodes, codes)); |
|
} else { |
|
error('Unknown block type in flate stream'); |
|
} |
|
|
|
buffer = this.buffer; |
|
var limit = buffer ? buffer.length : 0; |
|
var pos = this.bufferLength; |
|
while (true) { |
|
var code1 = this.getCode(litCodeTable); |
|
if (code1 < 256) { |
|
if (pos + 1 >= limit) { |
|
buffer = this.ensureBuffer(pos + 1); |
|
limit = buffer.length; |
|
} |
|
buffer[pos++] = code1; |
|
continue; |
|
} |
|
if (code1 === 256) { |
|
this.bufferLength = pos; |
|
return; |
|
} |
|
code1 -= 257; |
|
code1 = lengthDecode[code1]; |
|
var code2 = code1 >> 16; |
|
if (code2 > 0) { |
|
code2 = this.getBits(code2); |
|
} |
|
len = (code1 & 0xffff) + code2; |
|
code1 = this.getCode(distCodeTable); |
|
code1 = distDecode[code1]; |
|
code2 = code1 >> 16; |
|
if (code2 > 0) { |
|
code2 = this.getBits(code2); |
|
} |
|
var dist = (code1 & 0xffff) + code2; |
|
if (pos + len >= limit) { |
|
buffer = this.ensureBuffer(pos + len); |
|
limit = buffer.length; |
|
} |
|
for (var k = 0; k < len; ++k, ++pos) { |
|
buffer[pos] = buffer[pos - dist]; |
|
} |
|
} |
|
}; |
|
|
|
return FlateStream; |
|
})(); |
|
|
|
var PredictorStream = (function PredictorStreamClosure() { |
|
function PredictorStream(str, maybeLength, params) { |
|
var predictor = this.predictor = params.get('Predictor') || 1; |
|
|
|
if (predictor <= 1) { |
|
return str; // no prediction |
|
} |
|
if (predictor !== 2 && (predictor < 10 || predictor > 15)) { |
|
error('Unsupported predictor: ' + predictor); |
|
} |
|
|
|
if (predictor === 2) { |
|
this.readBlock = this.readBlockTiff; |
|
} else { |
|
this.readBlock = this.readBlockPng; |
|
} |
|
|
|
this.str = str; |
|
this.dict = str.dict; |
|
|
|
var colors = this.colors = params.get('Colors') || 1; |
|
var bits = this.bits = params.get('BitsPerComponent') || 8; |
|
var columns = this.columns = params.get('Columns') || 1; |
|
|
|
this.pixBytes = (colors * bits + 7) >> 3; |
|
this.rowBytes = (columns * colors * bits + 7) >> 3; |
|
|
|
DecodeStream.call(this, maybeLength); |
|
return this; |
|
} |
|
|
|
PredictorStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
PredictorStream.prototype.readBlockTiff = |
|
function predictorStreamReadBlockTiff() { |
|
var rowBytes = this.rowBytes; |
|
|
|
var bufferLength = this.bufferLength; |
|
var buffer = this.ensureBuffer(bufferLength + rowBytes); |
|
|
|
var bits = this.bits; |
|
var colors = this.colors; |
|
|
|
var rawBytes = this.str.getBytes(rowBytes); |
|
this.eof = !rawBytes.length; |
|
if (this.eof) { |
|
return; |
|
} |
|
|
|
var inbuf = 0, outbuf = 0; |
|
var inbits = 0, outbits = 0; |
|
var pos = bufferLength; |
|
var i; |
|
|
|
if (bits === 1) { |
|
for (i = 0; i < rowBytes; ++i) { |
|
var c = rawBytes[i]; |
|
inbuf = (inbuf << 8) | c; |
|
// bitwise addition is exclusive or |
|
// first shift inbuf and then add |
|
buffer[pos++] = (c ^ (inbuf >> colors)) & 0xFF; |
|
// truncate inbuf (assumes colors < 16) |
|
inbuf &= 0xFFFF; |
|
} |
|
} else if (bits === 8) { |
|
for (i = 0; i < colors; ++i) { |
|
buffer[pos++] = rawBytes[i]; |
|
} |
|
for (; i < rowBytes; ++i) { |
|
buffer[pos] = buffer[pos - colors] + rawBytes[i]; |
|
pos++; |
|
} |
|
} else { |
|
var compArray = new Uint8Array(colors + 1); |
|
var bitMask = (1 << bits) - 1; |
|
var j = 0, k = bufferLength; |
|
var columns = this.columns; |
|
for (i = 0; i < columns; ++i) { |
|
for (var kk = 0; kk < colors; ++kk) { |
|
if (inbits < bits) { |
|
inbuf = (inbuf << 8) | (rawBytes[j++] & 0xFF); |
|
inbits += 8; |
|
} |
|
compArray[kk] = (compArray[kk] + |
|
(inbuf >> (inbits - bits))) & bitMask; |
|
inbits -= bits; |
|
outbuf = (outbuf << bits) | compArray[kk]; |
|
outbits += bits; |
|
if (outbits >= 8) { |
|
buffer[k++] = (outbuf >> (outbits - 8)) & 0xFF; |
|
outbits -= 8; |
|
} |
|
} |
|
} |
|
if (outbits > 0) { |
|
buffer[k++] = (outbuf << (8 - outbits)) + |
|
(inbuf & ((1 << (8 - outbits)) - 1)); |
|
} |
|
} |
|
this.bufferLength += rowBytes; |
|
}; |
|
|
|
PredictorStream.prototype.readBlockPng = |
|
function predictorStreamReadBlockPng() { |
|
|
|
var rowBytes = this.rowBytes; |
|
var pixBytes = this.pixBytes; |
|
|
|
var predictor = this.str.getByte(); |
|
var rawBytes = this.str.getBytes(rowBytes); |
|
this.eof = !rawBytes.length; |
|
if (this.eof) { |
|
return; |
|
} |
|
|
|
var bufferLength = this.bufferLength; |
|
var buffer = this.ensureBuffer(bufferLength + rowBytes); |
|
|
|
var prevRow = buffer.subarray(bufferLength - rowBytes, bufferLength); |
|
if (prevRow.length === 0) { |
|
prevRow = new Uint8Array(rowBytes); |
|
} |
|
|
|
var i, j = bufferLength, up, c; |
|
switch (predictor) { |
|
case 0: |
|
for (i = 0; i < rowBytes; ++i) { |
|
buffer[j++] = rawBytes[i]; |
|
} |
|
break; |
|
case 1: |
|
for (i = 0; i < pixBytes; ++i) { |
|
buffer[j++] = rawBytes[i]; |
|
} |
|
for (; i < rowBytes; ++i) { |
|
buffer[j] = (buffer[j - pixBytes] + rawBytes[i]) & 0xFF; |
|
j++; |
|
} |
|
break; |
|
case 2: |
|
for (i = 0; i < rowBytes; ++i) { |
|
buffer[j++] = (prevRow[i] + rawBytes[i]) & 0xFF; |
|
} |
|
break; |
|
case 3: |
|
for (i = 0; i < pixBytes; ++i) { |
|
buffer[j++] = (prevRow[i] >> 1) + rawBytes[i]; |
|
} |
|
for (; i < rowBytes; ++i) { |
|
buffer[j] = (((prevRow[i] + buffer[j - pixBytes]) >> 1) + |
|
rawBytes[i]) & 0xFF; |
|
j++; |
|
} |
|
break; |
|
case 4: |
|
// we need to save the up left pixels values. the simplest way |
|
// is to create a new buffer |
|
for (i = 0; i < pixBytes; ++i) { |
|
up = prevRow[i]; |
|
c = rawBytes[i]; |
|
buffer[j++] = up + c; |
|
} |
|
for (; i < rowBytes; ++i) { |
|
up = prevRow[i]; |
|
var upLeft = prevRow[i - pixBytes]; |
|
var left = buffer[j - pixBytes]; |
|
var p = left + up - upLeft; |
|
|
|
var pa = p - left; |
|
if (pa < 0) { |
|
pa = -pa; |
|
} |
|
var pb = p - up; |
|
if (pb < 0) { |
|
pb = -pb; |
|
} |
|
var pc = p - upLeft; |
|
if (pc < 0) { |
|
pc = -pc; |
|
} |
|
|
|
c = rawBytes[i]; |
|
if (pa <= pb && pa <= pc) { |
|
buffer[j++] = left + c; |
|
} else if (pb <= pc) { |
|
buffer[j++] = up + c; |
|
} else { |
|
buffer[j++] = upLeft + c; |
|
} |
|
} |
|
break; |
|
default: |
|
error('Unsupported predictor: ' + predictor); |
|
} |
|
this.bufferLength += rowBytes; |
|
}; |
|
|
|
return PredictorStream; |
|
})(); |
|
|
|
/** |
|
* Depending on the type of JPEG a JpegStream is handled in different ways. For |
|
* JPEG's that are supported natively such as DeviceGray and DeviceRGB the image |
|
* data is stored and then loaded by the browser. For unsupported JPEG's we use |
|
* a library to decode these images and the stream behaves like all the other |
|
* DecodeStreams. |
|
*/ |
|
var JpegStream = (function JpegStreamClosure() { |
|
function JpegStream(stream, maybeLength, dict, xref) { |
|
// Some images may contain 'junk' before the SOI (start-of-image) marker. |
|
// Note: this seems to mainly affect inline images. |
|
var ch; |
|
while ((ch = stream.getByte()) !== -1) { |
|
if (ch === 0xFF) { // Find the first byte of the SOI marker (0xFFD8). |
|
stream.skip(-1); // Reset the stream position to the SOI. |
|
break; |
|
} |
|
} |
|
this.stream = stream; |
|
this.maybeLength = maybeLength; |
|
this.dict = dict; |
|
|
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
JpegStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
Object.defineProperty(JpegStream.prototype, 'bytes', { |
|
get: function JpegStream_bytes() { |
|
// If this.maybeLength is null, we'll get the entire stream. |
|
return shadow(this, 'bytes', this.stream.getBytes(this.maybeLength)); |
|
}, |
|
configurable: true |
|
}); |
|
|
|
JpegStream.prototype.ensureBuffer = function JpegStream_ensureBuffer(req) { |
|
if (this.bufferLength) { |
|
return; |
|
} |
|
try { |
|
var jpegImage = new JpegImage(); |
|
|
|
// checking if values needs to be transformed before conversion |
|
if (this.forceRGB && this.dict && isArray(this.dict.get('Decode'))) { |
|
var decodeArr = this.dict.get('Decode'); |
|
var bitsPerComponent = this.dict.get('BitsPerComponent') || 8; |
|
var decodeArrLength = decodeArr.length; |
|
var transform = new Int32Array(decodeArrLength); |
|
var transformNeeded = false; |
|
var maxValue = (1 << bitsPerComponent) - 1; |
|
for (var i = 0; i < decodeArrLength; i += 2) { |
|
transform[i] = ((decodeArr[i + 1] - decodeArr[i]) * 256) | 0; |
|
transform[i + 1] = (decodeArr[i] * maxValue) | 0; |
|
if (transform[i] !== 256 || transform[i + 1] !== 0) { |
|
transformNeeded = true; |
|
} |
|
} |
|
if (transformNeeded) { |
|
jpegImage.decodeTransform = transform; |
|
} |
|
} |
|
|
|
jpegImage.parse(this.bytes); |
|
var data = jpegImage.getData(this.drawWidth, this.drawHeight, |
|
this.forceRGB); |
|
this.buffer = data; |
|
this.bufferLength = data.length; |
|
this.eof = true; |
|
} catch (e) { |
|
error('JPEG error: ' + e); |
|
} |
|
}; |
|
|
|
JpegStream.prototype.getBytes = function JpegStream_getBytes(length) { |
|
this.ensureBuffer(); |
|
return this.buffer; |
|
}; |
|
|
|
JpegStream.prototype.getIR = function JpegStream_getIR() { |
|
return PDFJS.createObjectURL(this.bytes, 'image/jpeg'); |
|
}; |
|
/** |
|
* Checks if the image can be decoded and displayed by the browser without any |
|
* further processing such as color space conversions. |
|
*/ |
|
JpegStream.prototype.isNativelySupported = |
|
function JpegStream_isNativelySupported(xref, res) { |
|
var cs = ColorSpace.parse(this.dict.get('ColorSpace', 'CS'), xref, res); |
|
return (cs.name === 'DeviceGray' || cs.name === 'DeviceRGB') && |
|
cs.isDefaultDecode(this.dict.get('Decode', 'D')); |
|
}; |
|
/** |
|
* Checks if the image can be decoded by the browser. |
|
*/ |
|
JpegStream.prototype.isNativelyDecodable = |
|
function JpegStream_isNativelyDecodable(xref, res) { |
|
var cs = ColorSpace.parse(this.dict.get('ColorSpace', 'CS'), xref, res); |
|
return (cs.numComps === 1 || cs.numComps === 3) && |
|
cs.isDefaultDecode(this.dict.get('Decode', 'D')); |
|
}; |
|
|
|
return JpegStream; |
|
})(); |
|
|
|
/** |
|
* For JPEG 2000's we use a library to decode these images and |
|
* the stream behaves like all the other DecodeStreams. |
|
*/ |
|
var JpxStream = (function JpxStreamClosure() { |
|
function JpxStream(stream, maybeLength, dict) { |
|
this.stream = stream; |
|
this.maybeLength = maybeLength; |
|
this.dict = dict; |
|
|
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
JpxStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
Object.defineProperty(JpxStream.prototype, 'bytes', { |
|
get: function JpxStream_bytes() { |
|
// If this.maybeLength is null, we'll get the entire stream. |
|
return shadow(this, 'bytes', this.stream.getBytes(this.maybeLength)); |
|
}, |
|
configurable: true |
|
}); |
|
|
|
JpxStream.prototype.ensureBuffer = function JpxStream_ensureBuffer(req) { |
|
if (this.bufferLength) { |
|
return; |
|
} |
|
|
|
var jpxImage = new JpxImage(); |
|
jpxImage.parse(this.bytes); |
|
|
|
var width = jpxImage.width; |
|
var height = jpxImage.height; |
|
var componentsCount = jpxImage.componentsCount; |
|
var tileCount = jpxImage.tiles.length; |
|
if (tileCount === 1) { |
|
this.buffer = jpxImage.tiles[0].items; |
|
} else { |
|
var data = new Uint8Array(width * height * componentsCount); |
|
|
|
for (var k = 0; k < tileCount; k++) { |
|
var tileComponents = jpxImage.tiles[k]; |
|
var tileWidth = tileComponents.width; |
|
var tileHeight = tileComponents.height; |
|
var tileLeft = tileComponents.left; |
|
var tileTop = tileComponents.top; |
|
|
|
var src = tileComponents.items; |
|
var srcPosition = 0; |
|
var dataPosition = (width * tileTop + tileLeft) * componentsCount; |
|
var imgRowSize = width * componentsCount; |
|
var tileRowSize = tileWidth * componentsCount; |
|
|
|
for (var j = 0; j < tileHeight; j++) { |
|
var rowBytes = src.subarray(srcPosition, srcPosition + tileRowSize); |
|
data.set(rowBytes, dataPosition); |
|
srcPosition += tileRowSize; |
|
dataPosition += imgRowSize; |
|
} |
|
} |
|
this.buffer = data; |
|
} |
|
this.bufferLength = this.buffer.length; |
|
this.eof = true; |
|
}; |
|
|
|
return JpxStream; |
|
})(); |
|
|
|
/** |
|
* For JBIG2's we use a library to decode these images and |
|
* the stream behaves like all the other DecodeStreams. |
|
*/ |
|
var Jbig2Stream = (function Jbig2StreamClosure() { |
|
function Jbig2Stream(stream, maybeLength, dict) { |
|
this.stream = stream; |
|
this.maybeLength = maybeLength; |
|
this.dict = dict; |
|
|
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
Jbig2Stream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
Object.defineProperty(Jbig2Stream.prototype, 'bytes', { |
|
get: function Jbig2Stream_bytes() { |
|
// If this.maybeLength is null, we'll get the entire stream. |
|
return shadow(this, 'bytes', this.stream.getBytes(this.maybeLength)); |
|
}, |
|
configurable: true |
|
}); |
|
|
|
Jbig2Stream.prototype.ensureBuffer = function Jbig2Stream_ensureBuffer(req) { |
|
if (this.bufferLength) { |
|
return; |
|
} |
|
|
|
var jbig2Image = new Jbig2Image(); |
|
|
|
var chunks = [], xref = this.dict.xref; |
|
var decodeParams = xref.fetchIfRef(this.dict.get('DecodeParms')); |
|
|
|
// According to the PDF specification, DecodeParms can be either |
|
// a dictionary, or an array whose elements are dictionaries. |
|
if (isArray(decodeParams)) { |
|
if (decodeParams.length > 1) { |
|
warn('JBIG2 - \'DecodeParms\' array with multiple elements ' + |
|
'not supported.'); |
|
} |
|
decodeParams = xref.fetchIfRef(decodeParams[0]); |
|
} |
|
if (decodeParams && decodeParams.has('JBIG2Globals')) { |
|
var globalsStream = decodeParams.get('JBIG2Globals'); |
|
var globals = globalsStream.getBytes(); |
|
chunks.push({data: globals, start: 0, end: globals.length}); |
|
} |
|
chunks.push({data: this.bytes, start: 0, end: this.bytes.length}); |
|
var data = jbig2Image.parseChunks(chunks); |
|
var dataLength = data.length; |
|
|
|
// JBIG2 had black as 1 and white as 0, inverting the colors |
|
for (var i = 0; i < dataLength; i++) { |
|
data[i] ^= 0xFF; |
|
} |
|
|
|
this.buffer = data; |
|
this.bufferLength = dataLength; |
|
this.eof = true; |
|
}; |
|
|
|
return Jbig2Stream; |
|
})(); |
|
|
|
var DecryptStream = (function DecryptStreamClosure() { |
|
function DecryptStream(str, maybeLength, decrypt) { |
|
this.str = str; |
|
this.dict = str.dict; |
|
this.decrypt = decrypt; |
|
this.nextChunk = null; |
|
this.initialized = false; |
|
|
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
var chunkSize = 512; |
|
|
|
DecryptStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
DecryptStream.prototype.readBlock = function DecryptStream_readBlock() { |
|
var chunk; |
|
if (this.initialized) { |
|
chunk = this.nextChunk; |
|
} else { |
|
chunk = this.str.getBytes(chunkSize); |
|
this.initialized = true; |
|
} |
|
if (!chunk || chunk.length === 0) { |
|
this.eof = true; |
|
return; |
|
} |
|
this.nextChunk = this.str.getBytes(chunkSize); |
|
var hasMoreData = this.nextChunk && this.nextChunk.length > 0; |
|
|
|
var decrypt = this.decrypt; |
|
chunk = decrypt(chunk, !hasMoreData); |
|
|
|
var bufferLength = this.bufferLength; |
|
var i, n = chunk.length; |
|
var buffer = this.ensureBuffer(bufferLength + n); |
|
for (i = 0; i < n; i++) { |
|
buffer[bufferLength++] = chunk[i]; |
|
} |
|
this.bufferLength = bufferLength; |
|
}; |
|
|
|
return DecryptStream; |
|
})(); |
|
|
|
var Ascii85Stream = (function Ascii85StreamClosure() { |
|
function Ascii85Stream(str, maybeLength) { |
|
this.str = str; |
|
this.dict = str.dict; |
|
this.input = new Uint8Array(5); |
|
|
|
// Most streams increase in size when decoded, but Ascii85 streams |
|
// typically shrink by ~20%. |
|
if (maybeLength) { |
|
maybeLength = 0.8 * maybeLength; |
|
} |
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
Ascii85Stream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
Ascii85Stream.prototype.readBlock = function Ascii85Stream_readBlock() { |
|
var TILDA_CHAR = 0x7E; // '~' |
|
var Z_LOWER_CHAR = 0x7A; // 'z' |
|
var EOF = -1; |
|
|
|
var str = this.str; |
|
|
|
var c = str.getByte(); |
|
while (Lexer.isSpace(c)) { |
|
c = str.getByte(); |
|
} |
|
|
|
if (c === EOF || c === TILDA_CHAR) { |
|
this.eof = true; |
|
return; |
|
} |
|
|
|
var bufferLength = this.bufferLength, buffer; |
|
var i; |
|
|
|
// special code for z |
|
if (c === Z_LOWER_CHAR) { |
|
buffer = this.ensureBuffer(bufferLength + 4); |
|
for (i = 0; i < 4; ++i) { |
|
buffer[bufferLength + i] = 0; |
|
} |
|
this.bufferLength += 4; |
|
} else { |
|
var input = this.input; |
|
input[0] = c; |
|
for (i = 1; i < 5; ++i) { |
|
c = str.getByte(); |
|
while (Lexer.isSpace(c)) { |
|
c = str.getByte(); |
|
} |
|
|
|
input[i] = c; |
|
|
|
if (c === EOF || c === TILDA_CHAR) { |
|
break; |
|
} |
|
} |
|
buffer = this.ensureBuffer(bufferLength + i - 1); |
|
this.bufferLength += i - 1; |
|
|
|
// partial ending; |
|
if (i < 5) { |
|
for (; i < 5; ++i) { |
|
input[i] = 0x21 + 84; |
|
} |
|
this.eof = true; |
|
} |
|
var t = 0; |
|
for (i = 0; i < 5; ++i) { |
|
t = t * 85 + (input[i] - 0x21); |
|
} |
|
|
|
for (i = 3; i >= 0; --i) { |
|
buffer[bufferLength + i] = t & 0xFF; |
|
t >>= 8; |
|
} |
|
} |
|
}; |
|
|
|
return Ascii85Stream; |
|
})(); |
|
|
|
var AsciiHexStream = (function AsciiHexStreamClosure() { |
|
function AsciiHexStream(str, maybeLength) { |
|
this.str = str; |
|
this.dict = str.dict; |
|
|
|
this.firstDigit = -1; |
|
|
|
// Most streams increase in size when decoded, but AsciiHex streams shrink |
|
// by 50%. |
|
if (maybeLength) { |
|
maybeLength = 0.5 * maybeLength; |
|
} |
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
AsciiHexStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
AsciiHexStream.prototype.readBlock = function AsciiHexStream_readBlock() { |
|
var UPSTREAM_BLOCK_SIZE = 8000; |
|
var bytes = this.str.getBytes(UPSTREAM_BLOCK_SIZE); |
|
if (!bytes.length) { |
|
this.eof = true; |
|
return; |
|
} |
|
|
|
var maxDecodeLength = (bytes.length + 1) >> 1; |
|
var buffer = this.ensureBuffer(this.bufferLength + maxDecodeLength); |
|
var bufferLength = this.bufferLength; |
|
|
|
var firstDigit = this.firstDigit; |
|
for (var i = 0, ii = bytes.length; i < ii; i++) { |
|
var ch = bytes[i], digit; |
|
if (ch >= 0x30 && ch <= 0x39) { // '0'-'9' |
|
digit = ch & 0x0F; |
|
} else if ((ch >= 0x41 && ch <= 0x46) || (ch >= 0x61 && ch <= 0x66)) { |
|
// 'A'-'Z', 'a'-'z' |
|
digit = (ch & 0x0F) + 9; |
|
} else if (ch === 0x3E) { // '>' |
|
this.eof = true; |
|
break; |
|
} else { // probably whitespace |
|
continue; // ignoring |
|
} |
|
if (firstDigit < 0) { |
|
firstDigit = digit; |
|
} else { |
|
buffer[bufferLength++] = (firstDigit << 4) | digit; |
|
firstDigit = -1; |
|
} |
|
} |
|
if (firstDigit >= 0 && this.eof) { |
|
// incomplete byte |
|
buffer[bufferLength++] = (firstDigit << 4); |
|
firstDigit = -1; |
|
} |
|
this.firstDigit = firstDigit; |
|
this.bufferLength = bufferLength; |
|
}; |
|
|
|
return AsciiHexStream; |
|
})(); |
|
|
|
var RunLengthStream = (function RunLengthStreamClosure() { |
|
function RunLengthStream(str, maybeLength) { |
|
this.str = str; |
|
this.dict = str.dict; |
|
|
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
RunLengthStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
RunLengthStream.prototype.readBlock = function RunLengthStream_readBlock() { |
|
// The repeatHeader has following format. The first byte defines type of run |
|
// and amount of bytes to repeat/copy: n = 0 through 127 - copy next n bytes |
|
// (in addition to the second byte from the header), n = 129 through 255 - |
|
// duplicate the second byte from the header (257 - n) times, n = 128 - end. |
|
var repeatHeader = this.str.getBytes(2); |
|
if (!repeatHeader || repeatHeader.length < 2 || repeatHeader[0] === 128) { |
|
this.eof = true; |
|
return; |
|
} |
|
|
|
var buffer; |
|
var bufferLength = this.bufferLength; |
|
var n = repeatHeader[0]; |
|
if (n < 128) { |
|
// copy n bytes |
|
buffer = this.ensureBuffer(bufferLength + n + 1); |
|
buffer[bufferLength++] = repeatHeader[1]; |
|
if (n > 0) { |
|
var source = this.str.getBytes(n); |
|
buffer.set(source, bufferLength); |
|
bufferLength += n; |
|
} |
|
} else { |
|
n = 257 - n; |
|
var b = repeatHeader[1]; |
|
buffer = this.ensureBuffer(bufferLength + n + 1); |
|
for (var i = 0; i < n; i++) { |
|
buffer[bufferLength++] = b; |
|
} |
|
} |
|
this.bufferLength = bufferLength; |
|
}; |
|
|
|
return RunLengthStream; |
|
})(); |
|
|
|
var CCITTFaxStream = (function CCITTFaxStreamClosure() { |
|
|
|
var ccittEOL = -2; |
|
var twoDimPass = 0; |
|
var twoDimHoriz = 1; |
|
var twoDimVert0 = 2; |
|
var twoDimVertR1 = 3; |
|
var twoDimVertL1 = 4; |
|
var twoDimVertR2 = 5; |
|
var twoDimVertL2 = 6; |
|
var twoDimVertR3 = 7; |
|
var twoDimVertL3 = 8; |
|
|
|
var twoDimTable = [ |
|
[-1, -1], [-1, -1], // 000000x |
|
[7, twoDimVertL3], // 0000010 |
|
[7, twoDimVertR3], // 0000011 |
|
[6, twoDimVertL2], [6, twoDimVertL2], // 000010x |
|
[6, twoDimVertR2], [6, twoDimVertR2], // 000011x |
|
[4, twoDimPass], [4, twoDimPass], // 0001xxx |
|
[4, twoDimPass], [4, twoDimPass], |
|
[4, twoDimPass], [4, twoDimPass], |
|
[4, twoDimPass], [4, twoDimPass], |
|
[3, twoDimHoriz], [3, twoDimHoriz], // 001xxxx |
|
[3, twoDimHoriz], [3, twoDimHoriz], |
|
[3, twoDimHoriz], [3, twoDimHoriz], |
|
[3, twoDimHoriz], [3, twoDimHoriz], |
|
[3, twoDimHoriz], [3, twoDimHoriz], |
|
[3, twoDimHoriz], [3, twoDimHoriz], |
|
[3, twoDimHoriz], [3, twoDimHoriz], |
|
[3, twoDimHoriz], [3, twoDimHoriz], |
|
[3, twoDimVertL1], [3, twoDimVertL1], // 010xxxx |
|
[3, twoDimVertL1], [3, twoDimVertL1], |
|
[3, twoDimVertL1], [3, twoDimVertL1], |
|
[3, twoDimVertL1], [3, twoDimVertL1], |
|
[3, twoDimVertL1], [3, twoDimVertL1], |
|
[3, twoDimVertL1], [3, twoDimVertL1], |
|
[3, twoDimVertL1], [3, twoDimVertL1], |
|
[3, twoDimVertL1], [3, twoDimVertL1], |
|
[3, twoDimVertR1], [3, twoDimVertR1], // 011xxxx |
|
[3, twoDimVertR1], [3, twoDimVertR1], |
|
[3, twoDimVertR1], [3, twoDimVertR1], |
|
[3, twoDimVertR1], [3, twoDimVertR1], |
|
[3, twoDimVertR1], [3, twoDimVertR1], |
|
[3, twoDimVertR1], [3, twoDimVertR1], |
|
[3, twoDimVertR1], [3, twoDimVertR1], |
|
[3, twoDimVertR1], [3, twoDimVertR1], |
|
[1, twoDimVert0], [1, twoDimVert0], // 1xxxxxx |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0], |
|
[1, twoDimVert0], [1, twoDimVert0] |
|
]; |
|
|
|
var whiteTable1 = [ |
|
[-1, -1], // 00000 |
|
[12, ccittEOL], // 00001 |
|
[-1, -1], [-1, -1], // 0001x |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 001xx |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 010xx |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 011xx |
|
[11, 1792], [11, 1792], // 1000x |
|
[12, 1984], // 10010 |
|
[12, 2048], // 10011 |
|
[12, 2112], // 10100 |
|
[12, 2176], // 10101 |
|
[12, 2240], // 10110 |
|
[12, 2304], // 10111 |
|
[11, 1856], [11, 1856], // 1100x |
|
[11, 1920], [11, 1920], // 1101x |
|
[12, 2368], // 11100 |
|
[12, 2432], // 11101 |
|
[12, 2496], // 11110 |
|
[12, 2560] // 11111 |
|
]; |
|
|
|
var whiteTable2 = [ |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 0000000xx |
|
[8, 29], [8, 29], // 00000010x |
|
[8, 30], [8, 30], // 00000011x |
|
[8, 45], [8, 45], // 00000100x |
|
[8, 46], [8, 46], // 00000101x |
|
[7, 22], [7, 22], [7, 22], [7, 22], // 0000011xx |
|
[7, 23], [7, 23], [7, 23], [7, 23], // 0000100xx |
|
[8, 47], [8, 47], // 00001010x |
|
[8, 48], [8, 48], // 00001011x |
|
[6, 13], [6, 13], [6, 13], [6, 13], // 000011xxx |
|
[6, 13], [6, 13], [6, 13], [6, 13], |
|
[7, 20], [7, 20], [7, 20], [7, 20], // 0001000xx |
|
[8, 33], [8, 33], // 00010010x |
|
[8, 34], [8, 34], // 00010011x |
|
[8, 35], [8, 35], // 00010100x |
|
[8, 36], [8, 36], // 00010101x |
|
[8, 37], [8, 37], // 00010110x |
|
[8, 38], [8, 38], // 00010111x |
|
[7, 19], [7, 19], [7, 19], [7, 19], // 0001100xx |
|
[8, 31], [8, 31], // 00011010x |
|
[8, 32], [8, 32], // 00011011x |
|
[6, 1], [6, 1], [6, 1], [6, 1], // 000111xxx |
|
[6, 1], [6, 1], [6, 1], [6, 1], |
|
[6, 12], [6, 12], [6, 12], [6, 12], // 001000xxx |
|
[6, 12], [6, 12], [6, 12], [6, 12], |
|
[8, 53], [8, 53], // 00100100x |
|
[8, 54], [8, 54], // 00100101x |
|
[7, 26], [7, 26], [7, 26], [7, 26], // 0010011xx |
|
[8, 39], [8, 39], // 00101000x |
|
[8, 40], [8, 40], // 00101001x |
|
[8, 41], [8, 41], // 00101010x |
|
[8, 42], [8, 42], // 00101011x |
|
[8, 43], [8, 43], // 00101100x |
|
[8, 44], [8, 44], // 00101101x |
|
[7, 21], [7, 21], [7, 21], [7, 21], // 0010111xx |
|
[7, 28], [7, 28], [7, 28], [7, 28], // 0011000xx |
|
[8, 61], [8, 61], // 00110010x |
|
[8, 62], [8, 62], // 00110011x |
|
[8, 63], [8, 63], // 00110100x |
|
[8, 0], [8, 0], // 00110101x |
|
[8, 320], [8, 320], // 00110110x |
|
[8, 384], [8, 384], // 00110111x |
|
[5, 10], [5, 10], [5, 10], [5, 10], // 00111xxxx |
|
[5, 10], [5, 10], [5, 10], [5, 10], |
|
[5, 10], [5, 10], [5, 10], [5, 10], |
|
[5, 10], [5, 10], [5, 10], [5, 10], |
|
[5, 11], [5, 11], [5, 11], [5, 11], // 01000xxxx |
|
[5, 11], [5, 11], [5, 11], [5, 11], |
|
[5, 11], [5, 11], [5, 11], [5, 11], |
|
[5, 11], [5, 11], [5, 11], [5, 11], |
|
[7, 27], [7, 27], [7, 27], [7, 27], // 0100100xx |
|
[8, 59], [8, 59], // 01001010x |
|
[8, 60], [8, 60], // 01001011x |
|
[9, 1472], // 010011000 |
|
[9, 1536], // 010011001 |
|
[9, 1600], // 010011010 |
|
[9, 1728], // 010011011 |
|
[7, 18], [7, 18], [7, 18], [7, 18], // 0100111xx |
|
[7, 24], [7, 24], [7, 24], [7, 24], // 0101000xx |
|
[8, 49], [8, 49], // 01010010x |
|
[8, 50], [8, 50], // 01010011x |
|
[8, 51], [8, 51], // 01010100x |
|
[8, 52], [8, 52], // 01010101x |
|
[7, 25], [7, 25], [7, 25], [7, 25], // 0101011xx |
|
[8, 55], [8, 55], // 01011000x |
|
[8, 56], [8, 56], // 01011001x |
|
[8, 57], [8, 57], // 01011010x |
|
[8, 58], [8, 58], // 01011011x |
|
[6, 192], [6, 192], [6, 192], [6, 192], // 010111xxx |
|
[6, 192], [6, 192], [6, 192], [6, 192], |
|
[6, 1664], [6, 1664], [6, 1664], [6, 1664], // 011000xxx |
|
[6, 1664], [6, 1664], [6, 1664], [6, 1664], |
|
[8, 448], [8, 448], // 01100100x |
|
[8, 512], [8, 512], // 01100101x |
|
[9, 704], // 011001100 |
|
[9, 768], // 011001101 |
|
[8, 640], [8, 640], // 01100111x |
|
[8, 576], [8, 576], // 01101000x |
|
[9, 832], // 011010010 |
|
[9, 896], // 011010011 |
|
[9, 960], // 011010100 |
|
[9, 1024], // 011010101 |
|
[9, 1088], // 011010110 |
|
[9, 1152], // 011010111 |
|
[9, 1216], // 011011000 |
|
[9, 1280], // 011011001 |
|
[9, 1344], // 011011010 |
|
[9, 1408], // 011011011 |
|
[7, 256], [7, 256], [7, 256], [7, 256], // 0110111xx |
|
[4, 2], [4, 2], [4, 2], [4, 2], // 0111xxxxx |
|
[4, 2], [4, 2], [4, 2], [4, 2], |
|
[4, 2], [4, 2], [4, 2], [4, 2], |
|
[4, 2], [4, 2], [4, 2], [4, 2], |
|
[4, 2], [4, 2], [4, 2], [4, 2], |
|
[4, 2], [4, 2], [4, 2], [4, 2], |
|
[4, 2], [4, 2], [4, 2], [4, 2], |
|
[4, 2], [4, 2], [4, 2], [4, 2], |
|
[4, 3], [4, 3], [4, 3], [4, 3], // 1000xxxxx |
|
[4, 3], [4, 3], [4, 3], [4, 3], |
|
[4, 3], [4, 3], [4, 3], [4, 3], |
|
[4, 3], [4, 3], [4, 3], [4, 3], |
|
[4, 3], [4, 3], [4, 3], [4, 3], |
|
[4, 3], [4, 3], [4, 3], [4, 3], |
|
[4, 3], [4, 3], [4, 3], [4, 3], |
|
[4, 3], [4, 3], [4, 3], [4, 3], |
|
[5, 128], [5, 128], [5, 128], [5, 128], // 10010xxxx |
|
[5, 128], [5, 128], [5, 128], [5, 128], |
|
[5, 128], [5, 128], [5, 128], [5, 128], |
|
[5, 128], [5, 128], [5, 128], [5, 128], |
|
[5, 8], [5, 8], [5, 8], [5, 8], // 10011xxxx |
|
[5, 8], [5, 8], [5, 8], [5, 8], |
|
[5, 8], [5, 8], [5, 8], [5, 8], |
|
[5, 8], [5, 8], [5, 8], [5, 8], |
|
[5, 9], [5, 9], [5, 9], [5, 9], // 10100xxxx |
|
[5, 9], [5, 9], [5, 9], [5, 9], |
|
[5, 9], [5, 9], [5, 9], [5, 9], |
|
[5, 9], [5, 9], [5, 9], [5, 9], |
|
[6, 16], [6, 16], [6, 16], [6, 16], // 101010xxx |
|
[6, 16], [6, 16], [6, 16], [6, 16], |
|
[6, 17], [6, 17], [6, 17], [6, 17], // 101011xxx |
|
[6, 17], [6, 17], [6, 17], [6, 17], |
|
[4, 4], [4, 4], [4, 4], [4, 4], // 1011xxxxx |
|
[4, 4], [4, 4], [4, 4], [4, 4], |
|
[4, 4], [4, 4], [4, 4], [4, 4], |
|
[4, 4], [4, 4], [4, 4], [4, 4], |
|
[4, 4], [4, 4], [4, 4], [4, 4], |
|
[4, 4], [4, 4], [4, 4], [4, 4], |
|
[4, 4], [4, 4], [4, 4], [4, 4], |
|
[4, 4], [4, 4], [4, 4], [4, 4], |
|
[4, 5], [4, 5], [4, 5], [4, 5], // 1100xxxxx |
|
[4, 5], [4, 5], [4, 5], [4, 5], |
|
[4, 5], [4, 5], [4, 5], [4, 5], |
|
[4, 5], [4, 5], [4, 5], [4, 5], |
|
[4, 5], [4, 5], [4, 5], [4, 5], |
|
[4, 5], [4, 5], [4, 5], [4, 5], |
|
[4, 5], [4, 5], [4, 5], [4, 5], |
|
[4, 5], [4, 5], [4, 5], [4, 5], |
|
[6, 14], [6, 14], [6, 14], [6, 14], // 110100xxx |
|
[6, 14], [6, 14], [6, 14], [6, 14], |
|
[6, 15], [6, 15], [6, 15], [6, 15], // 110101xxx |
|
[6, 15], [6, 15], [6, 15], [6, 15], |
|
[5, 64], [5, 64], [5, 64], [5, 64], // 11011xxxx |
|
[5, 64], [5, 64], [5, 64], [5, 64], |
|
[5, 64], [5, 64], [5, 64], [5, 64], |
|
[5, 64], [5, 64], [5, 64], [5, 64], |
|
[4, 6], [4, 6], [4, 6], [4, 6], // 1110xxxxx |
|
[4, 6], [4, 6], [4, 6], [4, 6], |
|
[4, 6], [4, 6], [4, 6], [4, 6], |
|
[4, 6], [4, 6], [4, 6], [4, 6], |
|
[4, 6], [4, 6], [4, 6], [4, 6], |
|
[4, 6], [4, 6], [4, 6], [4, 6], |
|
[4, 6], [4, 6], [4, 6], [4, 6], |
|
[4, 6], [4, 6], [4, 6], [4, 6], |
|
[4, 7], [4, 7], [4, 7], [4, 7], // 1111xxxxx |
|
[4, 7], [4, 7], [4, 7], [4, 7], |
|
[4, 7], [4, 7], [4, 7], [4, 7], |
|
[4, 7], [4, 7], [4, 7], [4, 7], |
|
[4, 7], [4, 7], [4, 7], [4, 7], |
|
[4, 7], [4, 7], [4, 7], [4, 7], |
|
[4, 7], [4, 7], [4, 7], [4, 7], |
|
[4, 7], [4, 7], [4, 7], [4, 7] |
|
]; |
|
|
|
var blackTable1 = [ |
|
[-1, -1], [-1, -1], // 000000000000x |
|
[12, ccittEOL], [12, ccittEOL], // 000000000001x |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 00000000001xx |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 00000000010xx |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 00000000011xx |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 00000000100xx |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 00000000101xx |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 00000000110xx |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 00000000111xx |
|
[11, 1792], [11, 1792], [11, 1792], [11, 1792], // 00000001000xx |
|
[12, 1984], [12, 1984], // 000000010010x |
|
[12, 2048], [12, 2048], // 000000010011x |
|
[12, 2112], [12, 2112], // 000000010100x |
|
[12, 2176], [12, 2176], // 000000010101x |
|
[12, 2240], [12, 2240], // 000000010110x |
|
[12, 2304], [12, 2304], // 000000010111x |
|
[11, 1856], [11, 1856], [11, 1856], [11, 1856], // 00000001100xx |
|
[11, 1920], [11, 1920], [11, 1920], [11, 1920], // 00000001101xx |
|
[12, 2368], [12, 2368], // 000000011100x |
|
[12, 2432], [12, 2432], // 000000011101x |
|
[12, 2496], [12, 2496], // 000000011110x |
|
[12, 2560], [12, 2560], // 000000011111x |
|
[10, 18], [10, 18], [10, 18], [10, 18], // 0000001000xxx |
|
[10, 18], [10, 18], [10, 18], [10, 18], |
|
[12, 52], [12, 52], // 000000100100x |
|
[13, 640], // 0000001001010 |
|
[13, 704], // 0000001001011 |
|
[13, 768], // 0000001001100 |
|
[13, 832], // 0000001001101 |
|
[12, 55], [12, 55], // 000000100111x |
|
[12, 56], [12, 56], // 000000101000x |
|
[13, 1280], // 0000001010010 |
|
[13, 1344], // 0000001010011 |
|
[13, 1408], // 0000001010100 |
|
[13, 1472], // 0000001010101 |
|
[12, 59], [12, 59], // 000000101011x |
|
[12, 60], [12, 60], // 000000101100x |
|
[13, 1536], // 0000001011010 |
|
[13, 1600], // 0000001011011 |
|
[11, 24], [11, 24], [11, 24], [11, 24], // 00000010111xx |
|
[11, 25], [11, 25], [11, 25], [11, 25], // 00000011000xx |
|
[13, 1664], // 0000001100100 |
|
[13, 1728], // 0000001100101 |
|
[12, 320], [12, 320], // 000000110011x |
|
[12, 384], [12, 384], // 000000110100x |
|
[12, 448], [12, 448], // 000000110101x |
|
[13, 512], // 0000001101100 |
|
[13, 576], // 0000001101101 |
|
[12, 53], [12, 53], // 000000110111x |
|
[12, 54], [12, 54], // 000000111000x |
|
[13, 896], // 0000001110010 |
|
[13, 960], // 0000001110011 |
|
[13, 1024], // 0000001110100 |
|
[13, 1088], // 0000001110101 |
|
[13, 1152], // 0000001110110 |
|
[13, 1216], // 0000001110111 |
|
[10, 64], [10, 64], [10, 64], [10, 64], // 0000001111xxx |
|
[10, 64], [10, 64], [10, 64], [10, 64] |
|
]; |
|
|
|
var blackTable2 = [ |
|
[8, 13], [8, 13], [8, 13], [8, 13], // 00000100xxxx |
|
[8, 13], [8, 13], [8, 13], [8, 13], |
|
[8, 13], [8, 13], [8, 13], [8, 13], |
|
[8, 13], [8, 13], [8, 13], [8, 13], |
|
[11, 23], [11, 23], // 00000101000x |
|
[12, 50], // 000001010010 |
|
[12, 51], // 000001010011 |
|
[12, 44], // 000001010100 |
|
[12, 45], // 000001010101 |
|
[12, 46], // 000001010110 |
|
[12, 47], // 000001010111 |
|
[12, 57], // 000001011000 |
|
[12, 58], // 000001011001 |
|
[12, 61], // 000001011010 |
|
[12, 256], // 000001011011 |
|
[10, 16], [10, 16], [10, 16], [10, 16], // 0000010111xx |
|
[10, 17], [10, 17], [10, 17], [10, 17], // 0000011000xx |
|
[12, 48], // 000001100100 |
|
[12, 49], // 000001100101 |
|
[12, 62], // 000001100110 |
|
[12, 63], // 000001100111 |
|
[12, 30], // 000001101000 |
|
[12, 31], // 000001101001 |
|
[12, 32], // 000001101010 |
|
[12, 33], // 000001101011 |
|
[12, 40], // 000001101100 |
|
[12, 41], // 000001101101 |
|
[11, 22], [11, 22], // 00000110111x |
|
[8, 14], [8, 14], [8, 14], [8, 14], // 00000111xxxx |
|
[8, 14], [8, 14], [8, 14], [8, 14], |
|
[8, 14], [8, 14], [8, 14], [8, 14], |
|
[8, 14], [8, 14], [8, 14], [8, 14], |
|
[7, 10], [7, 10], [7, 10], [7, 10], // 0000100xxxxx |
|
[7, 10], [7, 10], [7, 10], [7, 10], |
|
[7, 10], [7, 10], [7, 10], [7, 10], |
|
[7, 10], [7, 10], [7, 10], [7, 10], |
|
[7, 10], [7, 10], [7, 10], [7, 10], |
|
[7, 10], [7, 10], [7, 10], [7, 10], |
|
[7, 10], [7, 10], [7, 10], [7, 10], |
|
[7, 10], [7, 10], [7, 10], [7, 10], |
|
[7, 11], [7, 11], [7, 11], [7, 11], // 0000101xxxxx |
|
[7, 11], [7, 11], [7, 11], [7, 11], |
|
[7, 11], [7, 11], [7, 11], [7, 11], |
|
[7, 11], [7, 11], [7, 11], [7, 11], |
|
[7, 11], [7, 11], [7, 11], [7, 11], |
|
[7, 11], [7, 11], [7, 11], [7, 11], |
|
[7, 11], [7, 11], [7, 11], [7, 11], |
|
[7, 11], [7, 11], [7, 11], [7, 11], |
|
[9, 15], [9, 15], [9, 15], [9, 15], // 000011000xxx |
|
[9, 15], [9, 15], [9, 15], [9, 15], |
|
[12, 128], // 000011001000 |
|
[12, 192], // 000011001001 |
|
[12, 26], // 000011001010 |
|
[12, 27], // 000011001011 |
|
[12, 28], // 000011001100 |
|
[12, 29], // 000011001101 |
|
[11, 19], [11, 19], // 00001100111x |
|
[11, 20], [11, 20], // 00001101000x |
|
[12, 34], // 000011010010 |
|
[12, 35], // 000011010011 |
|
[12, 36], // 000011010100 |
|
[12, 37], // 000011010101 |
|
[12, 38], // 000011010110 |
|
[12, 39], // 000011010111 |
|
[11, 21], [11, 21], // 00001101100x |
|
[12, 42], // 000011011010 |
|
[12, 43], // 000011011011 |
|
[10, 0], [10, 0], [10, 0], [10, 0], // 0000110111xx |
|
[7, 12], [7, 12], [7, 12], [7, 12], // 0000111xxxxx |
|
[7, 12], [7, 12], [7, 12], [7, 12], |
|
[7, 12], [7, 12], [7, 12], [7, 12], |
|
[7, 12], [7, 12], [7, 12], [7, 12], |
|
[7, 12], [7, 12], [7, 12], [7, 12], |
|
[7, 12], [7, 12], [7, 12], [7, 12], |
|
[7, 12], [7, 12], [7, 12], [7, 12], |
|
[7, 12], [7, 12], [7, 12], [7, 12] |
|
]; |
|
|
|
var blackTable3 = [ |
|
[-1, -1], [-1, -1], [-1, -1], [-1, -1], // 0000xx |
|
[6, 9], // 000100 |
|
[6, 8], // 000101 |
|
[5, 7], [5, 7], // 00011x |
|
[4, 6], [4, 6], [4, 6], [4, 6], // 0010xx |
|
[4, 5], [4, 5], [4, 5], [4, 5], // 0011xx |
|
[3, 1], [3, 1], [3, 1], [3, 1], // 010xxx |
|
[3, 1], [3, 1], [3, 1], [3, 1], |
|
[3, 4], [3, 4], [3, 4], [3, 4], // 011xxx |
|
[3, 4], [3, 4], [3, 4], [3, 4], |
|
[2, 3], [2, 3], [2, 3], [2, 3], // 10xxxx |
|
[2, 3], [2, 3], [2, 3], [2, 3], |
|
[2, 3], [2, 3], [2, 3], [2, 3], |
|
[2, 3], [2, 3], [2, 3], [2, 3], |
|
[2, 2], [2, 2], [2, 2], [2, 2], // 11xxxx |
|
[2, 2], [2, 2], [2, 2], [2, 2], |
|
[2, 2], [2, 2], [2, 2], [2, 2], |
|
[2, 2], [2, 2], [2, 2], [2, 2] |
|
]; |
|
|
|
function CCITTFaxStream(str, maybeLength, params) { |
|
this.str = str; |
|
this.dict = str.dict; |
|
|
|
params = params || Dict.empty; |
|
|
|
this.encoding = params.get('K') || 0; |
|
this.eoline = params.get('EndOfLine') || false; |
|
this.byteAlign = params.get('EncodedByteAlign') || false; |
|
this.columns = params.get('Columns') || 1728; |
|
this.rows = params.get('Rows') || 0; |
|
var eoblock = params.get('EndOfBlock'); |
|
if (eoblock === null || eoblock === undefined) { |
|
eoblock = true; |
|
} |
|
this.eoblock = eoblock; |
|
this.black = params.get('BlackIs1') || false; |
|
|
|
this.codingLine = new Uint32Array(this.columns + 1); |
|
this.refLine = new Uint32Array(this.columns + 2); |
|
|
|
this.codingLine[0] = this.columns; |
|
this.codingPos = 0; |
|
|
|
this.row = 0; |
|
this.nextLine2D = this.encoding < 0; |
|
this.inputBits = 0; |
|
this.inputBuf = 0; |
|
this.outputBits = 0; |
|
|
|
var code1; |
|
while ((code1 = this.lookBits(12)) === 0) { |
|
this.eatBits(1); |
|
} |
|
if (code1 === 1) { |
|
this.eatBits(12); |
|
} |
|
if (this.encoding > 0) { |
|
this.nextLine2D = !this.lookBits(1); |
|
this.eatBits(1); |
|
} |
|
|
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
CCITTFaxStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
CCITTFaxStream.prototype.readBlock = function CCITTFaxStream_readBlock() { |
|
while (!this.eof) { |
|
var c = this.lookChar(); |
|
this.ensureBuffer(this.bufferLength + 1); |
|
this.buffer[this.bufferLength++] = c; |
|
} |
|
}; |
|
|
|
CCITTFaxStream.prototype.addPixels = |
|
function ccittFaxStreamAddPixels(a1, blackPixels) { |
|
var codingLine = this.codingLine; |
|
var codingPos = this.codingPos; |
|
|
|
if (a1 > codingLine[codingPos]) { |
|
if (a1 > this.columns) { |
|
info('row is wrong length'); |
|
this.err = true; |
|
a1 = this.columns; |
|
} |
|
if ((codingPos & 1) ^ blackPixels) { |
|
++codingPos; |
|
} |
|
|
|
codingLine[codingPos] = a1; |
|
} |
|
this.codingPos = codingPos; |
|
}; |
|
|
|
CCITTFaxStream.prototype.addPixelsNeg = |
|
function ccittFaxStreamAddPixelsNeg(a1, blackPixels) { |
|
var codingLine = this.codingLine; |
|
var codingPos = this.codingPos; |
|
|
|
if (a1 > codingLine[codingPos]) { |
|
if (a1 > this.columns) { |
|
info('row is wrong length'); |
|
this.err = true; |
|
a1 = this.columns; |
|
} |
|
if ((codingPos & 1) ^ blackPixels) { |
|
++codingPos; |
|
} |
|
|
|
codingLine[codingPos] = a1; |
|
} else if (a1 < codingLine[codingPos]) { |
|
if (a1 < 0) { |
|
info('invalid code'); |
|
this.err = true; |
|
a1 = 0; |
|
} |
|
while (codingPos > 0 && a1 < codingLine[codingPos - 1]) { |
|
--codingPos; |
|
} |
|
codingLine[codingPos] = a1; |
|
} |
|
|
|
this.codingPos = codingPos; |
|
}; |
|
|
|
CCITTFaxStream.prototype.lookChar = function CCITTFaxStream_lookChar() { |
|
var refLine = this.refLine; |
|
var codingLine = this.codingLine; |
|
var columns = this.columns; |
|
|
|
var refPos, blackPixels, bits, i; |
|
|
|
if (this.outputBits === 0) { |
|
if (this.eof) { |
|
return null; |
|
} |
|
this.err = false; |
|
|
|
var code1, code2, code3; |
|
if (this.nextLine2D) { |
|
for (i = 0; codingLine[i] < columns; ++i) { |
|
refLine[i] = codingLine[i]; |
|
} |
|
refLine[i++] = columns; |
|
refLine[i] = columns; |
|
codingLine[0] = 0; |
|
this.codingPos = 0; |
|
refPos = 0; |
|
blackPixels = 0; |
|
|
|
while (codingLine[this.codingPos] < columns) { |
|
code1 = this.getTwoDimCode(); |
|
switch (code1) { |
|
case twoDimPass: |
|
this.addPixels(refLine[refPos + 1], blackPixels); |
|
if (refLine[refPos + 1] < columns) { |
|
refPos += 2; |
|
} |
|
break; |
|
case twoDimHoriz: |
|
code1 = code2 = 0; |
|
if (blackPixels) { |
|
do { |
|
code1 += (code3 = this.getBlackCode()); |
|
} while (code3 >= 64); |
|
do { |
|
code2 += (code3 = this.getWhiteCode()); |
|
} while (code3 >= 64); |
|
} else { |
|
do { |
|
code1 += (code3 = this.getWhiteCode()); |
|
} while (code3 >= 64); |
|
do { |
|
code2 += (code3 = this.getBlackCode()); |
|
} while (code3 >= 64); |
|
} |
|
this.addPixels(codingLine[this.codingPos] + |
|
code1, blackPixels); |
|
if (codingLine[this.codingPos] < columns) { |
|
this.addPixels(codingLine[this.codingPos] + code2, |
|
blackPixels ^ 1); |
|
} |
|
while (refLine[refPos] <= codingLine[this.codingPos] && |
|
refLine[refPos] < columns) { |
|
refPos += 2; |
|
} |
|
break; |
|
case twoDimVertR3: |
|
this.addPixels(refLine[refPos] + 3, blackPixels); |
|
blackPixels ^= 1; |
|
if (codingLine[this.codingPos] < columns) { |
|
++refPos; |
|
while (refLine[refPos] <= codingLine[this.codingPos] && |
|
refLine[refPos] < columns) { |
|
refPos += 2; |
|
} |
|
} |
|
break; |
|
case twoDimVertR2: |
|
this.addPixels(refLine[refPos] + 2, blackPixels); |
|
blackPixels ^= 1; |
|
if (codingLine[this.codingPos] < columns) { |
|
++refPos; |
|
while (refLine[refPos] <= codingLine[this.codingPos] && |
|
refLine[refPos] < columns) { |
|
refPos += 2; |
|
} |
|
} |
|
break; |
|
case twoDimVertR1: |
|
this.addPixels(refLine[refPos] + 1, blackPixels); |
|
blackPixels ^= 1; |
|
if (codingLine[this.codingPos] < columns) { |
|
++refPos; |
|
while (refLine[refPos] <= codingLine[this.codingPos] && |
|
refLine[refPos] < columns) { |
|
refPos += 2; |
|
} |
|
} |
|
break; |
|
case twoDimVert0: |
|
this.addPixels(refLine[refPos], blackPixels); |
|
blackPixels ^= 1; |
|
if (codingLine[this.codingPos] < columns) { |
|
++refPos; |
|
while (refLine[refPos] <= codingLine[this.codingPos] && |
|
refLine[refPos] < columns) { |
|
refPos += 2; |
|
} |
|
} |
|
break; |
|
case twoDimVertL3: |
|
this.addPixelsNeg(refLine[refPos] - 3, blackPixels); |
|
blackPixels ^= 1; |
|
if (codingLine[this.codingPos] < columns) { |
|
if (refPos > 0) { |
|
--refPos; |
|
} else { |
|
++refPos; |
|
} |
|
while (refLine[refPos] <= codingLine[this.codingPos] && |
|
refLine[refPos] < columns) { |
|
refPos += 2; |
|
} |
|
} |
|
break; |
|
case twoDimVertL2: |
|
this.addPixelsNeg(refLine[refPos] - 2, blackPixels); |
|
blackPixels ^= 1; |
|
if (codingLine[this.codingPos] < columns) { |
|
if (refPos > 0) { |
|
--refPos; |
|
} else { |
|
++refPos; |
|
} |
|
while (refLine[refPos] <= codingLine[this.codingPos] && |
|
refLine[refPos] < columns) { |
|
refPos += 2; |
|
} |
|
} |
|
break; |
|
case twoDimVertL1: |
|
this.addPixelsNeg(refLine[refPos] - 1, blackPixels); |
|
blackPixels ^= 1; |
|
if (codingLine[this.codingPos] < columns) { |
|
if (refPos > 0) { |
|
--refPos; |
|
} else { |
|
++refPos; |
|
} |
|
while (refLine[refPos] <= codingLine[this.codingPos] && |
|
refLine[refPos] < columns) { |
|
refPos += 2; |
|
} |
|
} |
|
break; |
|
case EOF: |
|
this.addPixels(columns, 0); |
|
this.eof = true; |
|
break; |
|
default: |
|
info('bad 2d code'); |
|
this.addPixels(columns, 0); |
|
this.err = true; |
|
} |
|
} |
|
} else { |
|
codingLine[0] = 0; |
|
this.codingPos = 0; |
|
blackPixels = 0; |
|
while (codingLine[this.codingPos] < columns) { |
|
code1 = 0; |
|
if (blackPixels) { |
|
do { |
|
code1 += (code3 = this.getBlackCode()); |
|
} while (code3 >= 64); |
|
} else { |
|
do { |
|
code1 += (code3 = this.getWhiteCode()); |
|
} while (code3 >= 64); |
|
} |
|
this.addPixels(codingLine[this.codingPos] + code1, blackPixels); |
|
blackPixels ^= 1; |
|
} |
|
} |
|
|
|
var gotEOL = false; |
|
|
|
if (this.byteAlign) { |
|
this.inputBits &= ~7; |
|
} |
|
|
|
if (!this.eoblock && this.row === this.rows - 1) { |
|
this.eof = true; |
|
} else { |
|
code1 = this.lookBits(12); |
|
if (this.eoline) { |
|
while (code1 !== EOF && code1 !== 1) { |
|
this.eatBits(1); |
|
code1 = this.lookBits(12); |
|
} |
|
} else { |
|
while (code1 === 0) { |
|
this.eatBits(1); |
|
code1 = this.lookBits(12); |
|
} |
|
} |
|
if (code1 === 1) { |
|
this.eatBits(12); |
|
gotEOL = true; |
|
} else if (code1 === EOF) { |
|
this.eof = true; |
|
} |
|
} |
|
|
|
if (!this.eof && this.encoding > 0) { |
|
this.nextLine2D = !this.lookBits(1); |
|
this.eatBits(1); |
|
} |
|
|
|
if (this.eoblock && gotEOL && this.byteAlign) { |
|
code1 = this.lookBits(12); |
|
if (code1 === 1) { |
|
this.eatBits(12); |
|
if (this.encoding > 0) { |
|
this.lookBits(1); |
|
this.eatBits(1); |
|
} |
|
if (this.encoding >= 0) { |
|
for (i = 0; i < 4; ++i) { |
|
code1 = this.lookBits(12); |
|
if (code1 !== 1) { |
|
info('bad rtc code: ' + code1); |
|
} |
|
this.eatBits(12); |
|
if (this.encoding > 0) { |
|
this.lookBits(1); |
|
this.eatBits(1); |
|
} |
|
} |
|
} |
|
this.eof = true; |
|
} |
|
} else if (this.err && this.eoline) { |
|
while (true) { |
|
code1 = this.lookBits(13); |
|
if (code1 === EOF) { |
|
this.eof = true; |
|
return null; |
|
} |
|
if ((code1 >> 1) === 1) { |
|
break; |
|
} |
|
this.eatBits(1); |
|
} |
|
this.eatBits(12); |
|
if (this.encoding > 0) { |
|
this.eatBits(1); |
|
this.nextLine2D = !(code1 & 1); |
|
} |
|
} |
|
|
|
if (codingLine[0] > 0) { |
|
this.outputBits = codingLine[this.codingPos = 0]; |
|
} else { |
|
this.outputBits = codingLine[this.codingPos = 1]; |
|
} |
|
this.row++; |
|
} |
|
|
|
var c; |
|
if (this.outputBits >= 8) { |
|
c = (this.codingPos & 1) ? 0 : 0xFF; |
|
this.outputBits -= 8; |
|
if (this.outputBits === 0 && codingLine[this.codingPos] < columns) { |
|
this.codingPos++; |
|
this.outputBits = (codingLine[this.codingPos] - |
|
codingLine[this.codingPos - 1]); |
|
} |
|
} else { |
|
bits = 8; |
|
c = 0; |
|
do { |
|
if (this.outputBits > bits) { |
|
c <<= bits; |
|
if (!(this.codingPos & 1)) { |
|
c |= 0xFF >> (8 - bits); |
|
} |
|
this.outputBits -= bits; |
|
bits = 0; |
|
} else { |
|
c <<= this.outputBits; |
|
if (!(this.codingPos & 1)) { |
|
c |= 0xFF >> (8 - this.outputBits); |
|
} |
|
bits -= this.outputBits; |
|
this.outputBits = 0; |
|
if (codingLine[this.codingPos] < columns) { |
|
this.codingPos++; |
|
this.outputBits = (codingLine[this.codingPos] - |
|
codingLine[this.codingPos - 1]); |
|
} else if (bits > 0) { |
|
c <<= bits; |
|
bits = 0; |
|
} |
|
} |
|
} while (bits); |
|
} |
|
if (this.black) { |
|
c ^= 0xFF; |
|
} |
|
return c; |
|
}; |
|
|
|
// This functions returns the code found from the table. |
|
// The start and end parameters set the boundaries for searching the table. |
|
// The limit parameter is optional. Function returns an array with three |
|
// values. The first array element indicates whether a valid code is being |
|
// returned. The second array element is the actual code. The third array |
|
// element indicates whether EOF was reached. |
|
CCITTFaxStream.prototype.findTableCode = |
|
function ccittFaxStreamFindTableCode(start, end, table, limit) { |
|
|
|
var limitValue = limit || 0; |
|
for (var i = start; i <= end; ++i) { |
|
var code = this.lookBits(i); |
|
if (code === EOF) { |
|
return [true, 1, false]; |
|
} |
|
if (i < end) { |
|
code <<= end - i; |
|
} |
|
if (!limitValue || code >= limitValue) { |
|
var p = table[code - limitValue]; |
|
if (p[0] === i) { |
|
this.eatBits(i); |
|
return [true, p[1], true]; |
|
} |
|
} |
|
} |
|
return [false, 0, false]; |
|
}; |
|
|
|
CCITTFaxStream.prototype.getTwoDimCode = |
|
function ccittFaxStreamGetTwoDimCode() { |
|
|
|
var code = 0; |
|
var p; |
|
if (this.eoblock) { |
|
code = this.lookBits(7); |
|
p = twoDimTable[code]; |
|
if (p && p[0] > 0) { |
|
this.eatBits(p[0]); |
|
return p[1]; |
|
} |
|
} else { |
|
var result = this.findTableCode(1, 7, twoDimTable); |
|
if (result[0] && result[2]) { |
|
return result[1]; |
|
} |
|
} |
|
info('Bad two dim code'); |
|
return EOF; |
|
}; |
|
|
|
CCITTFaxStream.prototype.getWhiteCode = |
|
function ccittFaxStreamGetWhiteCode() { |
|
|
|
var code = 0; |
|
var p; |
|
if (this.eoblock) { |
|
code = this.lookBits(12); |
|
if (code === EOF) { |
|
return 1; |
|
} |
|
|
|
if ((code >> 5) === 0) { |
|
p = whiteTable1[code]; |
|
} else { |
|
p = whiteTable2[code >> 3]; |
|
} |
|
|
|
if (p[0] > 0) { |
|
this.eatBits(p[0]); |
|
return p[1]; |
|
} |
|
} else { |
|
var result = this.findTableCode(1, 9, whiteTable2); |
|
if (result[0]) { |
|
return result[1]; |
|
} |
|
|
|
result = this.findTableCode(11, 12, whiteTable1); |
|
if (result[0]) { |
|
return result[1]; |
|
} |
|
} |
|
info('bad white code'); |
|
this.eatBits(1); |
|
return 1; |
|
}; |
|
|
|
CCITTFaxStream.prototype.getBlackCode = |
|
function ccittFaxStreamGetBlackCode() { |
|
|
|
var code, p; |
|
if (this.eoblock) { |
|
code = this.lookBits(13); |
|
if (code === EOF) { |
|
return 1; |
|
} |
|
if ((code >> 7) === 0) { |
|
p = blackTable1[code]; |
|
} else if ((code >> 9) === 0 && (code >> 7) !== 0) { |
|
p = blackTable2[(code >> 1) - 64]; |
|
} else { |
|
p = blackTable3[code >> 7]; |
|
} |
|
|
|
if (p[0] > 0) { |
|
this.eatBits(p[0]); |
|
return p[1]; |
|
} |
|
} else { |
|
var result = this.findTableCode(2, 6, blackTable3); |
|
if (result[0]) { |
|
return result[1]; |
|
} |
|
|
|
result = this.findTableCode(7, 12, blackTable2, 64); |
|
if (result[0]) { |
|
return result[1]; |
|
} |
|
|
|
result = this.findTableCode(10, 13, blackTable1); |
|
if (result[0]) { |
|
return result[1]; |
|
} |
|
} |
|
info('bad black code'); |
|
this.eatBits(1); |
|
return 1; |
|
}; |
|
|
|
CCITTFaxStream.prototype.lookBits = function CCITTFaxStream_lookBits(n) { |
|
var c; |
|
while (this.inputBits < n) { |
|
if ((c = this.str.getByte()) === -1) { |
|
if (this.inputBits === 0) { |
|
return EOF; |
|
} |
|
return ((this.inputBuf << (n - this.inputBits)) & |
|
(0xFFFF >> (16 - n))); |
|
} |
|
this.inputBuf = (this.inputBuf << 8) + c; |
|
this.inputBits += 8; |
|
} |
|
return (this.inputBuf >> (this.inputBits - n)) & (0xFFFF >> (16 - n)); |
|
}; |
|
|
|
CCITTFaxStream.prototype.eatBits = function CCITTFaxStream_eatBits(n) { |
|
if ((this.inputBits -= n) < 0) { |
|
this.inputBits = 0; |
|
} |
|
}; |
|
|
|
return CCITTFaxStream; |
|
})(); |
|
|
|
var LZWStream = (function LZWStreamClosure() { |
|
function LZWStream(str, maybeLength, earlyChange) { |
|
this.str = str; |
|
this.dict = str.dict; |
|
this.cachedData = 0; |
|
this.bitsCached = 0; |
|
|
|
var maxLzwDictionarySize = 4096; |
|
var lzwState = { |
|
earlyChange: earlyChange, |
|
codeLength: 9, |
|
nextCode: 258, |
|
dictionaryValues: new Uint8Array(maxLzwDictionarySize), |
|
dictionaryLengths: new Uint16Array(maxLzwDictionarySize), |
|
dictionaryPrevCodes: new Uint16Array(maxLzwDictionarySize), |
|
currentSequence: new Uint8Array(maxLzwDictionarySize), |
|
currentSequenceLength: 0 |
|
}; |
|
for (var i = 0; i < 256; ++i) { |
|
lzwState.dictionaryValues[i] = i; |
|
lzwState.dictionaryLengths[i] = 1; |
|
} |
|
this.lzwState = lzwState; |
|
|
|
DecodeStream.call(this, maybeLength); |
|
} |
|
|
|
LZWStream.prototype = Object.create(DecodeStream.prototype); |
|
|
|
LZWStream.prototype.readBits = function LZWStream_readBits(n) { |
|
var bitsCached = this.bitsCached; |
|
var cachedData = this.cachedData; |
|
while (bitsCached < n) { |
|
var c = this.str.getByte(); |
|
if (c === -1) { |
|
this.eof = true; |
|
return null; |
|
} |
|
cachedData = (cachedData << 8) | c; |
|
bitsCached += 8; |
|
} |
|
this.bitsCached = (bitsCached -= n); |
|
this.cachedData = cachedData; |
|
this.lastCode = null; |
|
return (cachedData >>> bitsCached) & ((1 << n) - 1); |
|
}; |
|
|
|
LZWStream.prototype.readBlock = function LZWStream_readBlock() { |
|
var blockSize = 512; |
|
var estimatedDecodedSize = blockSize * 2, decodedSizeDelta = blockSize; |
|
var i, j, q; |
|
|
|
var lzwState = this.lzwState; |
|
if (!lzwState) { |
|
return; // eof was found |
|
} |
|
|
|
var earlyChange = lzwState.earlyChange; |
|
var nextCode = lzwState.nextCode; |
|
var dictionaryValues = lzwState.dictionaryValues; |
|
var dictionaryLengths = lzwState.dictionaryLengths; |
|
var dictionaryPrevCodes = lzwState.dictionaryPrevCodes; |
|
var codeLength = lzwState.codeLength; |
|
var prevCode = lzwState.prevCode; |
|
var currentSequence = lzwState.currentSequence; |
|
var currentSequenceLength = lzwState.currentSequenceLength; |
|
|
|
var decodedLength = 0; |
|
var currentBufferLength = this.bufferLength; |
|
var buffer = this.ensureBuffer(this.bufferLength + estimatedDecodedSize); |
|
|
|
for (i = 0; i < blockSize; i++) { |
|
var code = this.readBits(codeLength); |
|
var hasPrev = currentSequenceLength > 0; |
|
if (code < 256) { |
|
currentSequence[0] = code; |
|
currentSequenceLength = 1; |
|
} else if (code >= 258) { |
|
if (code < nextCode) { |
|
currentSequenceLength = dictionaryLengths[code]; |
|
for (j = currentSequenceLength - 1, q = code; j >= 0; j--) { |
|
currentSequence[j] = dictionaryValues[q]; |
|
q = dictionaryPrevCodes[q]; |
|
} |
|
} else { |
|
currentSequence[currentSequenceLength++] = currentSequence[0]; |
|
} |
|
} else if (code === 256) { |
|
codeLength = 9; |
|
nextCode = 258; |
|
currentSequenceLength = 0; |
|
continue; |
|
} else { |
|
this.eof = true; |
|
delete this.lzwState; |
|
break; |
|
} |
|
|
|
if (hasPrev) { |
|
dictionaryPrevCodes[nextCode] = prevCode; |
|
dictionaryLengths[nextCode] = dictionaryLengths[prevCode] + 1; |
|
dictionaryValues[nextCode] = currentSequence[0]; |
|
nextCode++; |
|
codeLength = (nextCode + earlyChange) & (nextCode + earlyChange - 1) ? |
|
codeLength : Math.min(Math.log(nextCode + earlyChange) / |
|
0.6931471805599453 + 1, 12) | 0; |
|
} |
|
prevCode = code; |
|
|
|
decodedLength += currentSequenceLength; |
|
if (estimatedDecodedSize < decodedLength) { |
|
do { |
|
estimatedDecodedSize += decodedSizeDelta; |
|
} while (estimatedDecodedSize < decodedLength); |
|
buffer = this.ensureBuffer(this.bufferLength + estimatedDecodedSize); |
|
} |
|
for (j = 0; j < currentSequenceLength; j++) { |
|
buffer[currentBufferLength++] = currentSequence[j]; |
|
} |
|
} |
|
lzwState.nextCode = nextCode; |
|
lzwState.codeLength = codeLength; |
|
lzwState.prevCode = prevCode; |
|
lzwState.currentSequenceLength = currentSequenceLength; |
|
|
|
this.bufferLength = currentBufferLength; |
|
}; |
|
|
|
return LZWStream; |
|
})(); |
|
|
|
var NullStream = (function NullStreamClosure() { |
|
function NullStream() { |
|
Stream.call(this, new Uint8Array(0)); |
|
} |
|
|
|
NullStream.prototype = Stream.prototype; |
|
|
|
return NullStream; |
|
})(); |
|
|
|
// TODO refactor to remove dependency on parser.js |
|
function _setCoreParser(coreParser_) { |
|
coreParser = coreParser_; |
|
EOF = coreParser_.EOF; |
|
Lexer = coreParser_.Lexer; |
|
} |
|
exports._setCoreParser = _setCoreParser; |
|
|
|
// TODO refactor to remove dependency on colorspace.js |
|
function _setCoreColorSpace(coreColorSpace_) { |
|
coreColorSpace = coreColorSpace_; |
|
ColorSpace = coreColorSpace_.ColorSpace; |
|
} |
|
exports._setCoreColorSpace = _setCoreColorSpace; |
|
|
|
exports.Ascii85Stream = Ascii85Stream; |
|
exports.AsciiHexStream = AsciiHexStream; |
|
exports.CCITTFaxStream = CCITTFaxStream; |
|
exports.DecryptStream = DecryptStream; |
|
exports.DecodeStream = DecodeStream; |
|
exports.FlateStream = FlateStream; |
|
exports.Jbig2Stream = Jbig2Stream; |
|
exports.JpegStream = JpegStream; |
|
exports.JpxStream = JpxStream; |
|
exports.NullStream = NullStream; |
|
exports.PredictorStream = PredictorStream; |
|
exports.RunLengthStream = RunLengthStream; |
|
exports.Stream = Stream; |
|
exports.StreamsSequenceStream = StreamsSequenceStream; |
|
exports.StringStream = StringStream; |
|
exports.LZWStream = LZWStream; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreCrypto = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreStream); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreStream) { |
|
|
|
var PasswordException = sharedUtil.PasswordException; |
|
var PasswordResponses = sharedUtil.PasswordResponses; |
|
var bytesToString = sharedUtil.bytesToString; |
|
var error = sharedUtil.error; |
|
var isInt = sharedUtil.isInt; |
|
var stringToBytes = sharedUtil.stringToBytes; |
|
var utf8StringToString = sharedUtil.utf8StringToString; |
|
var warn = sharedUtil.warn; |
|
var Name = corePrimitives.Name; |
|
var isName = corePrimitives.isName; |
|
var isDict = corePrimitives.isDict; |
|
var DecryptStream = coreStream.DecryptStream; |
|
|
|
var ARCFourCipher = (function ARCFourCipherClosure() { |
|
function ARCFourCipher(key) { |
|
this.a = 0; |
|
this.b = 0; |
|
var s = new Uint8Array(256); |
|
var i, j = 0, tmp, keyLength = key.length; |
|
for (i = 0; i < 256; ++i) { |
|
s[i] = i; |
|
} |
|
for (i = 0; i < 256; ++i) { |
|
tmp = s[i]; |
|
j = (j + tmp + key[i % keyLength]) & 0xFF; |
|
s[i] = s[j]; |
|
s[j] = tmp; |
|
} |
|
this.s = s; |
|
} |
|
|
|
ARCFourCipher.prototype = { |
|
encryptBlock: function ARCFourCipher_encryptBlock(data) { |
|
var i, n = data.length, tmp, tmp2; |
|
var a = this.a, b = this.b, s = this.s; |
|
var output = new Uint8Array(n); |
|
for (i = 0; i < n; ++i) { |
|
a = (a + 1) & 0xFF; |
|
tmp = s[a]; |
|
b = (b + tmp) & 0xFF; |
|
tmp2 = s[b]; |
|
s[a] = tmp2; |
|
s[b] = tmp; |
|
output[i] = data[i] ^ s[(tmp + tmp2) & 0xFF]; |
|
} |
|
this.a = a; |
|
this.b = b; |
|
return output; |
|
} |
|
}; |
|
ARCFourCipher.prototype.decryptBlock = ARCFourCipher.prototype.encryptBlock; |
|
|
|
return ARCFourCipher; |
|
})(); |
|
|
|
var calculateMD5 = (function calculateMD5Closure() { |
|
var r = new Uint8Array([ |
|
7, 12, 17, 22, 7, 12, 17, 22, 7, 12, 17, 22, 7, 12, 17, 22, |
|
5, 9, 14, 20, 5, 9, 14, 20, 5, 9, 14, 20, 5, 9, 14, 20, |
|
4, 11, 16, 23, 4, 11, 16, 23, 4, 11, 16, 23, 4, 11, 16, 23, |
|
6, 10, 15, 21, 6, 10, 15, 21, 6, 10, 15, 21, 6, 10, 15, 21]); |
|
|
|
var k = new Int32Array([ |
|
-680876936, -389564586, 606105819, -1044525330, -176418897, 1200080426, |
|
-1473231341, -45705983, 1770035416, -1958414417, -42063, -1990404162, |
|
1804603682, -40341101, -1502002290, 1236535329, -165796510, -1069501632, |
|
643717713, -373897302, -701558691, 38016083, -660478335, -405537848, |
|
568446438, -1019803690, -187363961, 1163531501, -1444681467, -51403784, |
|
1735328473, -1926607734, -378558, -2022574463, 1839030562, -35309556, |
|
-1530992060, 1272893353, -155497632, -1094730640, 681279174, -358537222, |
|
-722521979, 76029189, -640364487, -421815835, 530742520, -995338651, |
|
-198630844, 1126891415, -1416354905, -57434055, 1700485571, -1894986606, |
|
-1051523, -2054922799, 1873313359, -30611744, -1560198380, 1309151649, |
|
-145523070, -1120210379, 718787259, -343485551]); |
|
|
|
function hash(data, offset, length) { |
|
var h0 = 1732584193, h1 = -271733879, h2 = -1732584194, h3 = 271733878; |
|
// pre-processing |
|
var paddedLength = (length + 72) & ~63; // data + 9 extra bytes |
|
var padded = new Uint8Array(paddedLength); |
|
var i, j, n; |
|
for (i = 0; i < length; ++i) { |
|
padded[i] = data[offset++]; |
|
} |
|
padded[i++] = 0x80; |
|
n = paddedLength - 8; |
|
while (i < n) { |
|
padded[i++] = 0; |
|
} |
|
padded[i++] = (length << 3) & 0xFF; |
|
padded[i++] = (length >> 5) & 0xFF; |
|
padded[i++] = (length >> 13) & 0xFF; |
|
padded[i++] = (length >> 21) & 0xFF; |
|
padded[i++] = (length >>> 29) & 0xFF; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
var w = new Int32Array(16); |
|
for (i = 0; i < paddedLength;) { |
|
for (j = 0; j < 16; ++j, i += 4) { |
|
w[j] = (padded[i] | (padded[i + 1] << 8) | |
|
(padded[i + 2] << 16) | (padded[i + 3] << 24)); |
|
} |
|
var a = h0, b = h1, c = h2, d = h3, f, g; |
|
for (j = 0; j < 64; ++j) { |
|
if (j < 16) { |
|
f = (b & c) | ((~b) & d); |
|
g = j; |
|
} else if (j < 32) { |
|
f = (d & b) | ((~d) & c); |
|
g = (5 * j + 1) & 15; |
|
} else if (j < 48) { |
|
f = b ^ c ^ d; |
|
g = (3 * j + 5) & 15; |
|
} else { |
|
f = c ^ (b | (~d)); |
|
g = (7 * j) & 15; |
|
} |
|
var tmp = d, rotateArg = (a + f + k[j] + w[g]) | 0, rotate = r[j]; |
|
d = c; |
|
c = b; |
|
b = (b + ((rotateArg << rotate) | (rotateArg >>> (32 - rotate)))) | 0; |
|
a = tmp; |
|
} |
|
h0 = (h0 + a) | 0; |
|
h1 = (h1 + b) | 0; |
|
h2 = (h2 + c) | 0; |
|
h3 = (h3 + d) | 0; |
|
} |
|
return new Uint8Array([ |
|
h0 & 0xFF, (h0 >> 8) & 0xFF, (h0 >> 16) & 0xFF, (h0 >>> 24) & 0xFF, |
|
h1 & 0xFF, (h1 >> 8) & 0xFF, (h1 >> 16) & 0xFF, (h1 >>> 24) & 0xFF, |
|
h2 & 0xFF, (h2 >> 8) & 0xFF, (h2 >> 16) & 0xFF, (h2 >>> 24) & 0xFF, |
|
h3 & 0xFF, (h3 >> 8) & 0xFF, (h3 >> 16) & 0xFF, (h3 >>> 24) & 0xFF |
|
]); |
|
} |
|
|
|
return hash; |
|
})(); |
|
var Word64 = (function Word64Closure() { |
|
function Word64(highInteger, lowInteger) { |
|
this.high = highInteger | 0; |
|
this.low = lowInteger | 0; |
|
} |
|
Word64.prototype = { |
|
and: function Word64_and(word) { |
|
this.high &= word.high; |
|
this.low &= word.low; |
|
}, |
|
xor: function Word64_xor(word) { |
|
this.high ^= word.high; |
|
this.low ^= word.low; |
|
}, |
|
|
|
or: function Word64_or(word) { |
|
this.high |= word.high; |
|
this.low |= word.low; |
|
}, |
|
|
|
shiftRight: function Word64_shiftRight(places) { |
|
if (places >= 32) { |
|
this.low = (this.high >>> (places - 32)) | 0; |
|
this.high = 0; |
|
} else { |
|
this.low = (this.low >>> places) | (this.high << (32 - places)); |
|
this.high = (this.high >>> places) | 0; |
|
} |
|
}, |
|
|
|
shiftLeft: function Word64_shiftLeft(places) { |
|
if (places >= 32) { |
|
this.high = this.low << (places - 32); |
|
this.low = 0; |
|
} else { |
|
this.high = (this.high << places) | (this.low >>> (32 - places)); |
|
this.low = this.low << places; |
|
} |
|
}, |
|
|
|
rotateRight: function Word64_rotateRight(places) { |
|
var low, high; |
|
if (places & 32) { |
|
high = this.low; |
|
low = this.high; |
|
} else { |
|
low = this.low; |
|
high = this.high; |
|
} |
|
places &= 31; |
|
this.low = (low >>> places) | (high << (32 - places)); |
|
this.high = (high >>> places) | (low << (32 - places)); |
|
}, |
|
|
|
not: function Word64_not() { |
|
this.high = ~this.high; |
|
this.low = ~this.low; |
|
}, |
|
|
|
add: function Word64_add(word) { |
|
var lowAdd = (this.low >>> 0) + (word.low >>> 0); |
|
var highAdd = (this.high >>> 0) + (word.high >>> 0); |
|
if (lowAdd > 0xFFFFFFFF) { |
|
highAdd += 1; |
|
} |
|
this.low = lowAdd | 0; |
|
this.high = highAdd | 0; |
|
}, |
|
|
|
copyTo: function Word64_copyTo(bytes, offset) { |
|
bytes[offset] = (this.high >>> 24) & 0xFF; |
|
bytes[offset + 1] = (this.high >> 16) & 0xFF; |
|
bytes[offset + 2] = (this.high >> 8) & 0xFF; |
|
bytes[offset + 3] = this.high & 0xFF; |
|
bytes[offset + 4] = (this.low >>> 24) & 0xFF; |
|
bytes[offset + 5] = (this.low >> 16) & 0xFF; |
|
bytes[offset + 6] = (this.low >> 8) & 0xFF; |
|
bytes[offset + 7] = this.low & 0xFF; |
|
}, |
|
|
|
assign: function Word64_assign(word) { |
|
this.high = word.high; |
|
this.low = word.low; |
|
} |
|
}; |
|
return Word64; |
|
})(); |
|
|
|
var calculateSHA256 = (function calculateSHA256Closure() { |
|
function rotr(x, n) { |
|
return (x >>> n) | (x << 32 - n); |
|
} |
|
|
|
function ch(x, y, z) { |
|
return (x & y) ^ (~x & z); |
|
} |
|
|
|
function maj(x, y, z) { |
|
return (x & y) ^ (x & z) ^ (y & z); |
|
} |
|
|
|
function sigma(x) { |
|
return rotr(x, 2) ^ rotr(x, 13) ^ rotr(x, 22); |
|
} |
|
|
|
function sigmaPrime(x) { |
|
return rotr(x, 6) ^ rotr(x, 11) ^ rotr(x, 25); |
|
} |
|
|
|
function littleSigma(x) { |
|
return rotr(x, 7) ^ rotr(x, 18) ^ x >>> 3; |
|
} |
|
|
|
function littleSigmaPrime(x) { |
|
return rotr(x, 17) ^ rotr(x, 19) ^ x >>> 10; |
|
} |
|
|
|
var k = [0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5, |
|
0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5, |
|
0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3, |
|
0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174, |
|
0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc, |
|
0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da, |
|
0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7, |
|
0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967, |
|
0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13, |
|
0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85, |
|
0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3, |
|
0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070, |
|
0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5, |
|
0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3, |
|
0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208, |
|
0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2]; |
|
|
|
function hash(data, offset, length) { |
|
// initial hash values |
|
var h0 = 0x6a09e667, h1 = 0xbb67ae85, h2 = 0x3c6ef372, |
|
h3 = 0xa54ff53a, h4 = 0x510e527f, h5 = 0x9b05688c, |
|
h6 = 0x1f83d9ab, h7 = 0x5be0cd19; |
|
// pre-processing |
|
var paddedLength = Math.ceil((length + 9) / 64) * 64; |
|
var padded = new Uint8Array(paddedLength); |
|
var i, j, n; |
|
for (i = 0; i < length; ++i) { |
|
padded[i] = data[offset++]; |
|
} |
|
padded[i++] = 0x80; |
|
n = paddedLength - 8; |
|
while (i < n) { |
|
padded[i++] = 0; |
|
} |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = (length >>> 29) & 0xFF; |
|
padded[i++] = (length >> 21) & 0xFF; |
|
padded[i++] = (length >> 13) & 0xFF; |
|
padded[i++] = (length >> 5) & 0xFF; |
|
padded[i++] = (length << 3) & 0xFF; |
|
var w = new Uint32Array(64); |
|
// for each 512 bit block |
|
for (i = 0; i < paddedLength;) { |
|
for (j = 0; j < 16; ++j) { |
|
w[j] = (padded[i] << 24 | (padded[i + 1] << 16) | |
|
(padded[i + 2] << 8) | (padded[i + 3])); |
|
i += 4; |
|
} |
|
|
|
for (j = 16; j < 64; ++j) { |
|
w[j] = littleSigmaPrime(w[j - 2]) + w[j - 7] + |
|
littleSigma(w[j - 15]) + w[j - 16] | 0; |
|
} |
|
var a = h0, b = h1, c = h2, d = h3, e = h4, |
|
f = h5, g = h6, h = h7, t1, t2; |
|
for (j = 0; j < 64; ++j) { |
|
t1 = h + sigmaPrime(e) + ch(e, f, g) + k[j] + w[j]; |
|
t2 = sigma(a) + maj(a, b, c); |
|
h = g; |
|
g = f; |
|
f = e; |
|
e = (d + t1) | 0; |
|
d = c; |
|
c = b; |
|
b = a; |
|
a = (t1 + t2) | 0; |
|
} |
|
h0 = (h0 + a) | 0; |
|
h1 = (h1 + b) | 0; |
|
h2 = (h2 + c) | 0; |
|
h3 = (h3 + d) | 0; |
|
h4 = (h4 + e) | 0; |
|
h5 = (h5 + f) | 0; |
|
h6 = (h6 + g) | 0; |
|
h7 = (h7 + h) | 0; |
|
} |
|
return new Uint8Array([ |
|
(h0 >> 24) & 0xFF, (h0 >> 16) & 0xFF, (h0 >> 8) & 0xFF, (h0) & 0xFF, |
|
(h1 >> 24) & 0xFF, (h1 >> 16) & 0xFF, (h1 >> 8) & 0xFF, (h1) & 0xFF, |
|
(h2 >> 24) & 0xFF, (h2 >> 16) & 0xFF, (h2 >> 8) & 0xFF, (h2) & 0xFF, |
|
(h3 >> 24) & 0xFF, (h3 >> 16) & 0xFF, (h3 >> 8) & 0xFF, (h3) & 0xFF, |
|
(h4 >> 24) & 0xFF, (h4 >> 16) & 0xFF, (h4 >> 8) & 0xFF, (h4) & 0xFF, |
|
(h5 >> 24) & 0xFF, (h5 >> 16) & 0xFF, (h5 >> 8) & 0xFF, (h5) & 0xFF, |
|
(h6 >> 24) & 0xFF, (h6 >> 16) & 0xFF, (h6 >> 8) & 0xFF, (h6) & 0xFF, |
|
(h7 >> 24) & 0xFF, (h7 >> 16) & 0xFF, (h7 >> 8) & 0xFF, (h7) & 0xFF |
|
]); |
|
} |
|
|
|
return hash; |
|
})(); |
|
|
|
var calculateSHA512 = (function calculateSHA512Closure() { |
|
function ch(result, x, y, z, tmp) { |
|
result.assign(x); |
|
result.and(y); |
|
tmp.assign(x); |
|
tmp.not(); |
|
tmp.and(z); |
|
result.xor(tmp); |
|
} |
|
|
|
function maj(result, x, y, z, tmp) { |
|
result.assign(x); |
|
result.and(y); |
|
tmp.assign(x); |
|
tmp.and(z); |
|
result.xor(tmp); |
|
tmp.assign(y); |
|
tmp.and(z); |
|
result.xor(tmp); |
|
} |
|
|
|
function sigma(result, x, tmp) { |
|
result.assign(x); |
|
result.rotateRight(28); |
|
tmp.assign(x); |
|
tmp.rotateRight(34); |
|
result.xor(tmp); |
|
tmp.assign(x); |
|
tmp.rotateRight(39); |
|
result.xor(tmp); |
|
} |
|
|
|
function sigmaPrime(result, x, tmp) { |
|
result.assign(x); |
|
result.rotateRight(14); |
|
tmp.assign(x); |
|
tmp.rotateRight(18); |
|
result.xor(tmp); |
|
tmp.assign(x); |
|
tmp.rotateRight(41); |
|
result.xor(tmp); |
|
} |
|
|
|
function littleSigma(result, x, tmp) { |
|
result.assign(x); |
|
result.rotateRight(1); |
|
tmp.assign(x); |
|
tmp.rotateRight(8); |
|
result.xor(tmp); |
|
tmp.assign(x); |
|
tmp.shiftRight(7); |
|
result.xor(tmp); |
|
} |
|
|
|
function littleSigmaPrime(result, x, tmp) { |
|
result.assign(x); |
|
result.rotateRight(19); |
|
tmp.assign(x); |
|
tmp.rotateRight(61); |
|
result.xor(tmp); |
|
tmp.assign(x); |
|
tmp.shiftRight(6); |
|
result.xor(tmp); |
|
} |
|
|
|
var k = [ |
|
new Word64(0x428a2f98, 0xd728ae22), new Word64(0x71374491, 0x23ef65cd), |
|
new Word64(0xb5c0fbcf, 0xec4d3b2f), new Word64(0xe9b5dba5, 0x8189dbbc), |
|
new Word64(0x3956c25b, 0xf348b538), new Word64(0x59f111f1, 0xb605d019), |
|
new Word64(0x923f82a4, 0xaf194f9b), new Word64(0xab1c5ed5, 0xda6d8118), |
|
new Word64(0xd807aa98, 0xa3030242), new Word64(0x12835b01, 0x45706fbe), |
|
new Word64(0x243185be, 0x4ee4b28c), new Word64(0x550c7dc3, 0xd5ffb4e2), |
|
new Word64(0x72be5d74, 0xf27b896f), new Word64(0x80deb1fe, 0x3b1696b1), |
|
new Word64(0x9bdc06a7, 0x25c71235), new Word64(0xc19bf174, 0xcf692694), |
|
new Word64(0xe49b69c1, 0x9ef14ad2), new Word64(0xefbe4786, 0x384f25e3), |
|
new Word64(0x0fc19dc6, 0x8b8cd5b5), new Word64(0x240ca1cc, 0x77ac9c65), |
|
new Word64(0x2de92c6f, 0x592b0275), new Word64(0x4a7484aa, 0x6ea6e483), |
|
new Word64(0x5cb0a9dc, 0xbd41fbd4), new Word64(0x76f988da, 0x831153b5), |
|
new Word64(0x983e5152, 0xee66dfab), new Word64(0xa831c66d, 0x2db43210), |
|
new Word64(0xb00327c8, 0x98fb213f), new Word64(0xbf597fc7, 0xbeef0ee4), |
|
new Word64(0xc6e00bf3, 0x3da88fc2), new Word64(0xd5a79147, 0x930aa725), |
|
new Word64(0x06ca6351, 0xe003826f), new Word64(0x14292967, 0x0a0e6e70), |
|
new Word64(0x27b70a85, 0x46d22ffc), new Word64(0x2e1b2138, 0x5c26c926), |
|
new Word64(0x4d2c6dfc, 0x5ac42aed), new Word64(0x53380d13, 0x9d95b3df), |
|
new Word64(0x650a7354, 0x8baf63de), new Word64(0x766a0abb, 0x3c77b2a8), |
|
new Word64(0x81c2c92e, 0x47edaee6), new Word64(0x92722c85, 0x1482353b), |
|
new Word64(0xa2bfe8a1, 0x4cf10364), new Word64(0xa81a664b, 0xbc423001), |
|
new Word64(0xc24b8b70, 0xd0f89791), new Word64(0xc76c51a3, 0x0654be30), |
|
new Word64(0xd192e819, 0xd6ef5218), new Word64(0xd6990624, 0x5565a910), |
|
new Word64(0xf40e3585, 0x5771202a), new Word64(0x106aa070, 0x32bbd1b8), |
|
new Word64(0x19a4c116, 0xb8d2d0c8), new Word64(0x1e376c08, 0x5141ab53), |
|
new Word64(0x2748774c, 0xdf8eeb99), new Word64(0x34b0bcb5, 0xe19b48a8), |
|
new Word64(0x391c0cb3, 0xc5c95a63), new Word64(0x4ed8aa4a, 0xe3418acb), |
|
new Word64(0x5b9cca4f, 0x7763e373), new Word64(0x682e6ff3, 0xd6b2b8a3), |
|
new Word64(0x748f82ee, 0x5defb2fc), new Word64(0x78a5636f, 0x43172f60), |
|
new Word64(0x84c87814, 0xa1f0ab72), new Word64(0x8cc70208, 0x1a6439ec), |
|
new Word64(0x90befffa, 0x23631e28), new Word64(0xa4506ceb, 0xde82bde9), |
|
new Word64(0xbef9a3f7, 0xb2c67915), new Word64(0xc67178f2, 0xe372532b), |
|
new Word64(0xca273ece, 0xea26619c), new Word64(0xd186b8c7, 0x21c0c207), |
|
new Word64(0xeada7dd6, 0xcde0eb1e), new Word64(0xf57d4f7f, 0xee6ed178), |
|
new Word64(0x06f067aa, 0x72176fba), new Word64(0x0a637dc5, 0xa2c898a6), |
|
new Word64(0x113f9804, 0xbef90dae), new Word64(0x1b710b35, 0x131c471b), |
|
new Word64(0x28db77f5, 0x23047d84), new Word64(0x32caab7b, 0x40c72493), |
|
new Word64(0x3c9ebe0a, 0x15c9bebc), new Word64(0x431d67c4, 0x9c100d4c), |
|
new Word64(0x4cc5d4be, 0xcb3e42b6), new Word64(0x597f299c, 0xfc657e2a), |
|
new Word64(0x5fcb6fab, 0x3ad6faec), new Word64(0x6c44198c, 0x4a475817)]; |
|
|
|
function hash(data, offset, length, mode384) { |
|
mode384 = !!mode384; |
|
// initial hash values |
|
var h0, h1, h2, h3, h4, h5, h6, h7; |
|
if (!mode384) { |
|
h0 = new Word64(0x6a09e667, 0xf3bcc908); |
|
h1 = new Word64(0xbb67ae85, 0x84caa73b); |
|
h2 = new Word64(0x3c6ef372, 0xfe94f82b); |
|
h3 = new Word64(0xa54ff53a, 0x5f1d36f1); |
|
h4 = new Word64(0x510e527f, 0xade682d1); |
|
h5 = new Word64(0x9b05688c, 0x2b3e6c1f); |
|
h6 = new Word64(0x1f83d9ab, 0xfb41bd6b); |
|
h7 = new Word64(0x5be0cd19, 0x137e2179); |
|
} |
|
else { |
|
// SHA384 is exactly the same |
|
// except with different starting values and a trimmed result |
|
h0 = new Word64(0xcbbb9d5d, 0xc1059ed8); |
|
h1 = new Word64(0x629a292a, 0x367cd507); |
|
h2 = new Word64(0x9159015a, 0x3070dd17); |
|
h3 = new Word64(0x152fecd8, 0xf70e5939); |
|
h4 = new Word64(0x67332667, 0xffc00b31); |
|
h5 = new Word64(0x8eb44a87, 0x68581511); |
|
h6 = new Word64(0xdb0c2e0d, 0x64f98fa7); |
|
h7 = new Word64(0x47b5481d, 0xbefa4fa4); |
|
} |
|
|
|
// pre-processing |
|
var paddedLength = Math.ceil((length + 17) / 128) * 128; |
|
var padded = new Uint8Array(paddedLength); |
|
var i, j, n; |
|
for (i = 0; i < length; ++i) { |
|
padded[i] = data[offset++]; |
|
} |
|
padded[i++] = 0x80; |
|
n = paddedLength - 16; |
|
while (i < n) { |
|
padded[i++] = 0; |
|
} |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = 0; |
|
padded[i++] = (length >>> 29) & 0xFF; |
|
padded[i++] = (length >> 21) & 0xFF; |
|
padded[i++] = (length >> 13) & 0xFF; |
|
padded[i++] = (length >> 5) & 0xFF; |
|
padded[i++] = (length << 3) & 0xFF; |
|
|
|
var w = new Array(80); |
|
for (i = 0; i < 80; i++) { |
|
w[i] = new Word64(0, 0); |
|
} |
|
var a = new Word64(0, 0), b = new Word64(0, 0), c = new Word64(0, 0); |
|
var d = new Word64(0, 0), e = new Word64(0, 0), f = new Word64(0, 0); |
|
var g = new Word64(0, 0), h = new Word64(0, 0); |
|
var t1 = new Word64(0, 0), t2 = new Word64(0, 0); |
|
var tmp1 = new Word64(0, 0), tmp2 = new Word64(0, 0), tmp3; |
|
|
|
// for each 1024 bit block |
|
for (i = 0; i < paddedLength;) { |
|
for (j = 0; j < 16; ++j) { |
|
w[j].high = (padded[i] << 24) | (padded[i + 1] << 16) | |
|
(padded[i + 2] << 8) | (padded[i + 3]); |
|
w[j].low = (padded[i + 4]) << 24 | (padded[i + 5]) << 16 | |
|
(padded[i + 6]) << 8 | (padded[i + 7]); |
|
i += 8; |
|
} |
|
for (j = 16; j < 80; ++j) { |
|
tmp3 = w[j]; |
|
littleSigmaPrime(tmp3, w[j - 2], tmp2); |
|
tmp3.add(w[j - 7]); |
|
littleSigma(tmp1, w[j - 15], tmp2); |
|
tmp3.add(tmp1); |
|
tmp3.add(w[j - 16]); |
|
} |
|
|
|
a.assign(h0); b.assign(h1); c.assign(h2); d.assign(h3); |
|
e.assign(h4); f.assign(h5); g.assign(h6); h.assign(h7); |
|
for (j = 0; j < 80; ++j) { |
|
t1.assign(h); |
|
sigmaPrime(tmp1, e, tmp2); |
|
t1.add(tmp1); |
|
ch(tmp1, e, f, g, tmp2); |
|
t1.add(tmp1); |
|
t1.add(k[j]); |
|
t1.add(w[j]); |
|
|
|
sigma(t2, a, tmp2); |
|
maj(tmp1, a, b, c, tmp2); |
|
t2.add(tmp1); |
|
|
|
tmp3 = h; |
|
h = g; |
|
g = f; |
|
f = e; |
|
d.add(t1); |
|
e = d; |
|
d = c; |
|
c = b; |
|
b = a; |
|
tmp3.assign(t1); |
|
tmp3.add(t2); |
|
a = tmp3; |
|
} |
|
h0.add(a); |
|
h1.add(b); |
|
h2.add(c); |
|
h3.add(d); |
|
h4.add(e); |
|
h5.add(f); |
|
h6.add(g); |
|
h7.add(h); |
|
} |
|
|
|
var result; |
|
if (!mode384) { |
|
result = new Uint8Array(64); |
|
h0.copyTo(result,0); |
|
h1.copyTo(result,8); |
|
h2.copyTo(result,16); |
|
h3.copyTo(result,24); |
|
h4.copyTo(result,32); |
|
h5.copyTo(result,40); |
|
h6.copyTo(result,48); |
|
h7.copyTo(result,56); |
|
} |
|
else { |
|
result = new Uint8Array(48); |
|
h0.copyTo(result,0); |
|
h1.copyTo(result,8); |
|
h2.copyTo(result,16); |
|
h3.copyTo(result,24); |
|
h4.copyTo(result,32); |
|
h5.copyTo(result,40); |
|
} |
|
return result; |
|
} |
|
|
|
return hash; |
|
})(); |
|
var calculateSHA384 = (function calculateSHA384Closure() { |
|
function hash(data, offset, length) { |
|
return calculateSHA512(data, offset, length, true); |
|
} |
|
|
|
return hash; |
|
})(); |
|
var NullCipher = (function NullCipherClosure() { |
|
function NullCipher() { |
|
} |
|
|
|
NullCipher.prototype = { |
|
decryptBlock: function NullCipher_decryptBlock(data) { |
|
return data; |
|
} |
|
}; |
|
|
|
return NullCipher; |
|
})(); |
|
|
|
var AES128Cipher = (function AES128CipherClosure() { |
|
var rcon = new Uint8Array([ |
|
0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, |
|
0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, |
|
0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, |
|
0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, |
|
0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, |
|
0x1b, 0x36, 0x6c, 0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, |
|
0x97, 0x35, 0x6a, 0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, |
|
0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, |
|
0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, |
|
0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, 0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, |
|
0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, 0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, |
|
0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, |
|
0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, |
|
0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, 0xd8, 0xab, 0x4d, |
|
0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, 0xd4, 0xb3, 0x7d, |
|
0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, |
|
0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb, |
|
0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, |
|
0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, |
|
0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, |
|
0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, |
|
0x74, 0xe8, 0xcb, 0x8d]); |
|
|
|
var s = new Uint8Array([ |
|
0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, |
|
0xfe, 0xd7, 0xab, 0x76, 0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, |
|
0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0, 0xb7, 0xfd, 0x93, 0x26, |
|
0x36, 0x3f, 0xf7, 0xcc, 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15, |
|
0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a, 0x07, 0x12, 0x80, 0xe2, |
|
0xeb, 0x27, 0xb2, 0x75, 0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0, |
|
0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84, 0x53, 0xd1, 0x00, 0xed, |
|
0x20, 0xfc, 0xb1, 0x5b, 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf, |
|
0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85, 0x45, 0xf9, 0x02, 0x7f, |
|
0x50, 0x3c, 0x9f, 0xa8, 0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5, |
|
0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2, 0xcd, 0x0c, 0x13, 0xec, |
|
0x5f, 0x97, 0x44, 0x17, 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73, |
|
0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88, 0x46, 0xee, 0xb8, 0x14, |
|
0xde, 0x5e, 0x0b, 0xdb, 0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c, |
|
0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79, 0xe7, 0xc8, 0x37, 0x6d, |
|
0x8d, 0xd5, 0x4e, 0xa9, 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08, |
|
0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6, 0xe8, 0xdd, 0x74, 0x1f, |
|
0x4b, 0xbd, 0x8b, 0x8a, 0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e, |
|
0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e, 0xe1, 0xf8, 0x98, 0x11, |
|
0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf, |
|
0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, |
|
0xb0, 0x54, 0xbb, 0x16]); |
|
|
|
var inv_s = new Uint8Array([ |
|
0x52, 0x09, 0x6a, 0xd5, 0x30, 0x36, 0xa5, 0x38, 0xbf, 0x40, 0xa3, 0x9e, |
|
0x81, 0xf3, 0xd7, 0xfb, 0x7c, 0xe3, 0x39, 0x82, 0x9b, 0x2f, 0xff, 0x87, |
|
0x34, 0x8e, 0x43, 0x44, 0xc4, 0xde, 0xe9, 0xcb, 0x54, 0x7b, 0x94, 0x32, |
|
0xa6, 0xc2, 0x23, 0x3d, 0xee, 0x4c, 0x95, 0x0b, 0x42, 0xfa, 0xc3, 0x4e, |
|
0x08, 0x2e, 0xa1, 0x66, 0x28, 0xd9, 0x24, 0xb2, 0x76, 0x5b, 0xa2, 0x49, |
|
0x6d, 0x8b, 0xd1, 0x25, 0x72, 0xf8, 0xf6, 0x64, 0x86, 0x68, 0x98, 0x16, |
|
0xd4, 0xa4, 0x5c, 0xcc, 0x5d, 0x65, 0xb6, 0x92, 0x6c, 0x70, 0x48, 0x50, |
|
0xfd, 0xed, 0xb9, 0xda, 0x5e, 0x15, 0x46, 0x57, 0xa7, 0x8d, 0x9d, 0x84, |
|
0x90, 0xd8, 0xab, 0x00, 0x8c, 0xbc, 0xd3, 0x0a, 0xf7, 0xe4, 0x58, 0x05, |
|
0xb8, 0xb3, 0x45, 0x06, 0xd0, 0x2c, 0x1e, 0x8f, 0xca, 0x3f, 0x0f, 0x02, |
|
0xc1, 0xaf, 0xbd, 0x03, 0x01, 0x13, 0x8a, 0x6b, 0x3a, 0x91, 0x11, 0x41, |
|
0x4f, 0x67, 0xdc, 0xea, 0x97, 0xf2, 0xcf, 0xce, 0xf0, 0xb4, 0xe6, 0x73, |
|
0x96, 0xac, 0x74, 0x22, 0xe7, 0xad, 0x35, 0x85, 0xe2, 0xf9, 0x37, 0xe8, |
|
0x1c, 0x75, 0xdf, 0x6e, 0x47, 0xf1, 0x1a, 0x71, 0x1d, 0x29, 0xc5, 0x89, |
|
0x6f, 0xb7, 0x62, 0x0e, 0xaa, 0x18, 0xbe, 0x1b, 0xfc, 0x56, 0x3e, 0x4b, |
|
0xc6, 0xd2, 0x79, 0x20, 0x9a, 0xdb, 0xc0, 0xfe, 0x78, 0xcd, 0x5a, 0xf4, |
|
0x1f, 0xdd, 0xa8, 0x33, 0x88, 0x07, 0xc7, 0x31, 0xb1, 0x12, 0x10, 0x59, |
|
0x27, 0x80, 0xec, 0x5f, 0x60, 0x51, 0x7f, 0xa9, 0x19, 0xb5, 0x4a, 0x0d, |
|
0x2d, 0xe5, 0x7a, 0x9f, 0x93, 0xc9, 0x9c, 0xef, 0xa0, 0xe0, 0x3b, 0x4d, |
|
0xae, 0x2a, 0xf5, 0xb0, 0xc8, 0xeb, 0xbb, 0x3c, 0x83, 0x53, 0x99, 0x61, |
|
0x17, 0x2b, 0x04, 0x7e, 0xba, 0x77, 0xd6, 0x26, 0xe1, 0x69, 0x14, 0x63, |
|
0x55, 0x21, 0x0c, 0x7d]); |
|
var mixCol = new Uint8Array(256); |
|
for (var i = 0; i < 256; i++) { |
|
if (i < 128) { |
|
mixCol[i] = i << 1; |
|
} else { |
|
mixCol[i] = (i << 1) ^ 0x1b; |
|
} |
|
} |
|
var mix = new Uint32Array([ |
|
0x00000000, 0x0e090d0b, 0x1c121a16, 0x121b171d, 0x3824342c, 0x362d3927, |
|
0x24362e3a, 0x2a3f2331, 0x70486858, 0x7e416553, 0x6c5a724e, 0x62537f45, |
|
0x486c5c74, 0x4665517f, 0x547e4662, 0x5a774b69, 0xe090d0b0, 0xee99ddbb, |
|
0xfc82caa6, 0xf28bc7ad, 0xd8b4e49c, 0xd6bde997, 0xc4a6fe8a, 0xcaaff381, |
|
0x90d8b8e8, 0x9ed1b5e3, 0x8ccaa2fe, 0x82c3aff5, 0xa8fc8cc4, 0xa6f581cf, |
|
0xb4ee96d2, 0xbae79bd9, 0xdb3bbb7b, 0xd532b670, 0xc729a16d, 0xc920ac66, |
|
0xe31f8f57, 0xed16825c, 0xff0d9541, 0xf104984a, 0xab73d323, 0xa57ade28, |
|
0xb761c935, 0xb968c43e, 0x9357e70f, 0x9d5eea04, 0x8f45fd19, 0x814cf012, |
|
0x3bab6bcb, 0x35a266c0, 0x27b971dd, 0x29b07cd6, 0x038f5fe7, 0x0d8652ec, |
|
0x1f9d45f1, 0x119448fa, 0x4be30393, 0x45ea0e98, 0x57f11985, 0x59f8148e, |
|
0x73c737bf, 0x7dce3ab4, 0x6fd52da9, 0x61dc20a2, 0xad766df6, 0xa37f60fd, |
|
0xb16477e0, 0xbf6d7aeb, 0x955259da, 0x9b5b54d1, 0x894043cc, 0x87494ec7, |
|
0xdd3e05ae, 0xd33708a5, 0xc12c1fb8, 0xcf2512b3, 0xe51a3182, 0xeb133c89, |
|
0xf9082b94, 0xf701269f, 0x4de6bd46, 0x43efb04d, 0x51f4a750, 0x5ffdaa5b, |
|
0x75c2896a, 0x7bcb8461, 0x69d0937c, 0x67d99e77, 0x3daed51e, 0x33a7d815, |
|
0x21bccf08, 0x2fb5c203, 0x058ae132, 0x0b83ec39, 0x1998fb24, 0x1791f62f, |
|
0x764dd68d, 0x7844db86, 0x6a5fcc9b, 0x6456c190, 0x4e69e2a1, 0x4060efaa, |
|
0x527bf8b7, 0x5c72f5bc, 0x0605bed5, 0x080cb3de, 0x1a17a4c3, 0x141ea9c8, |
|
0x3e218af9, 0x302887f2, 0x223390ef, 0x2c3a9de4, 0x96dd063d, 0x98d40b36, |
|
0x8acf1c2b, 0x84c61120, 0xaef93211, 0xa0f03f1a, 0xb2eb2807, 0xbce2250c, |
|
0xe6956e65, 0xe89c636e, 0xfa877473, 0xf48e7978, 0xdeb15a49, 0xd0b85742, |
|
0xc2a3405f, 0xccaa4d54, 0x41ecdaf7, 0x4fe5d7fc, 0x5dfec0e1, 0x53f7cdea, |
|
0x79c8eedb, 0x77c1e3d0, 0x65daf4cd, 0x6bd3f9c6, 0x31a4b2af, 0x3fadbfa4, |
|
0x2db6a8b9, 0x23bfa5b2, 0x09808683, 0x07898b88, 0x15929c95, 0x1b9b919e, |
|
0xa17c0a47, 0xaf75074c, 0xbd6e1051, 0xb3671d5a, 0x99583e6b, 0x97513360, |
|
0x854a247d, 0x8b432976, 0xd134621f, 0xdf3d6f14, 0xcd267809, 0xc32f7502, |
|
0xe9105633, 0xe7195b38, 0xf5024c25, 0xfb0b412e, 0x9ad7618c, 0x94de6c87, |
|
0x86c57b9a, 0x88cc7691, 0xa2f355a0, 0xacfa58ab, 0xbee14fb6, 0xb0e842bd, |
|
0xea9f09d4, 0xe49604df, 0xf68d13c2, 0xf8841ec9, 0xd2bb3df8, 0xdcb230f3, |
|
0xcea927ee, 0xc0a02ae5, 0x7a47b13c, 0x744ebc37, 0x6655ab2a, 0x685ca621, |
|
0x42638510, 0x4c6a881b, 0x5e719f06, 0x5078920d, 0x0a0fd964, 0x0406d46f, |
|
0x161dc372, 0x1814ce79, 0x322bed48, 0x3c22e043, 0x2e39f75e, 0x2030fa55, |
|
0xec9ab701, 0xe293ba0a, 0xf088ad17, 0xfe81a01c, 0xd4be832d, 0xdab78e26, |
|
0xc8ac993b, 0xc6a59430, 0x9cd2df59, 0x92dbd252, 0x80c0c54f, 0x8ec9c844, |
|
0xa4f6eb75, 0xaaffe67e, 0xb8e4f163, 0xb6edfc68, 0x0c0a67b1, 0x02036aba, |
|
0x10187da7, 0x1e1170ac, 0x342e539d, 0x3a275e96, 0x283c498b, 0x26354480, |
|
0x7c420fe9, 0x724b02e2, 0x605015ff, 0x6e5918f4, 0x44663bc5, 0x4a6f36ce, |
|
0x587421d3, 0x567d2cd8, 0x37a10c7a, 0x39a80171, 0x2bb3166c, 0x25ba1b67, |
|
0x0f853856, 0x018c355d, 0x13972240, 0x1d9e2f4b, 0x47e96422, 0x49e06929, |
|
0x5bfb7e34, 0x55f2733f, 0x7fcd500e, 0x71c45d05, 0x63df4a18, 0x6dd64713, |
|
0xd731dcca, 0xd938d1c1, 0xcb23c6dc, 0xc52acbd7, 0xef15e8e6, 0xe11ce5ed, |
|
0xf307f2f0, 0xfd0efffb, 0xa779b492, 0xa970b999, 0xbb6bae84, 0xb562a38f, |
|
0x9f5d80be, 0x91548db5, 0x834f9aa8, 0x8d4697a3]); |
|
|
|
function expandKey128(cipherKey) { |
|
var b = 176, result = new Uint8Array(b); |
|
result.set(cipherKey); |
|
for (var j = 16, i = 1; j < b; ++i) { |
|
// RotWord |
|
var t1 = result[j - 3], t2 = result[j - 2], |
|
t3 = result[j - 1], t4 = result[j - 4]; |
|
// SubWord |
|
t1 = s[t1]; |
|
t2 = s[t2]; |
|
t3 = s[t3]; |
|
t4 = s[t4]; |
|
// Rcon |
|
t1 = t1 ^ rcon[i]; |
|
for (var n = 0; n < 4; ++n) { |
|
result[j] = (t1 ^= result[j - 16]); |
|
j++; |
|
result[j] = (t2 ^= result[j - 16]); |
|
j++; |
|
result[j] = (t3 ^= result[j - 16]); |
|
j++; |
|
result[j] = (t4 ^= result[j - 16]); |
|
j++; |
|
} |
|
} |
|
return result; |
|
} |
|
|
|
function decrypt128(input, key) { |
|
var state = new Uint8Array(16); |
|
state.set(input); |
|
var i, j, k; |
|
var t, u, v; |
|
// AddRoundKey |
|
for (j = 0, k = 160; j < 16; ++j, ++k) { |
|
state[j] ^= key[k]; |
|
} |
|
for (i = 9; i >= 1; --i) { |
|
// InvShiftRows |
|
t = state[13]; |
|
state[13] = state[9]; |
|
state[9] = state[5]; |
|
state[5] = state[1]; |
|
state[1] = t; |
|
t = state[14]; |
|
u = state[10]; |
|
state[14] = state[6]; |
|
state[10] = state[2]; |
|
state[6] = t; |
|
state[2] = u; |
|
t = state[15]; |
|
u = state[11]; |
|
v = state[7]; |
|
state[15] = state[3]; |
|
state[11] = t; |
|
state[7] = u; |
|
state[3] = v; |
|
// InvSubBytes |
|
for (j = 0; j < 16; ++j) { |
|
state[j] = inv_s[state[j]]; |
|
} |
|
// AddRoundKey |
|
for (j = 0, k = i * 16; j < 16; ++j, ++k) { |
|
state[j] ^= key[k]; |
|
} |
|
// InvMixColumns |
|
for (j = 0; j < 16; j += 4) { |
|
var s0 = mix[state[j]], s1 = mix[state[j + 1]], |
|
s2 = mix[state[j + 2]], s3 = mix[state[j + 3]]; |
|
t = (s0 ^ (s1 >>> 8) ^ (s1 << 24) ^ (s2 >>> 16) ^ (s2 << 16) ^ |
|
(s3 >>> 24) ^ (s3 << 8)); |
|
state[j] = (t >>> 24) & 0xFF; |
|
state[j + 1] = (t >> 16) & 0xFF; |
|
state[j + 2] = (t >> 8) & 0xFF; |
|
state[j + 3] = t & 0xFF; |
|
} |
|
} |
|
// InvShiftRows |
|
t = state[13]; |
|
state[13] = state[9]; |
|
state[9] = state[5]; |
|
state[5] = state[1]; |
|
state[1] = t; |
|
t = state[14]; |
|
u = state[10]; |
|
state[14] = state[6]; |
|
state[10] = state[2]; |
|
state[6] = t; |
|
state[2] = u; |
|
t = state[15]; |
|
u = state[11]; |
|
v = state[7]; |
|
state[15] = state[3]; |
|
state[11] = t; |
|
state[7] = u; |
|
state[3] = v; |
|
for (j = 0; j < 16; ++j) { |
|
// InvSubBytes |
|
state[j] = inv_s[state[j]]; |
|
// AddRoundKey |
|
state[j] ^= key[j]; |
|
} |
|
return state; |
|
} |
|
|
|
function encrypt128(input, key) { |
|
var t, u, v, k; |
|
var state = new Uint8Array(16); |
|
state.set(input); |
|
for (j = 0; j < 16; ++j) { |
|
// AddRoundKey |
|
state[j] ^= key[j]; |
|
} |
|
|
|
for (i = 1; i < 10; i++) { |
|
//SubBytes |
|
for (j = 0; j < 16; ++j) { |
|
state[j] = s[state[j]]; |
|
} |
|
//ShiftRows |
|
v = state[1]; |
|
state[1] = state[5]; |
|
state[5] = state[9]; |
|
state[9] = state[13]; |
|
state[13] = v; |
|
v = state[2]; |
|
u = state[6]; |
|
state[2] = state[10]; |
|
state[6] = state[14]; |
|
state[10] = v; |
|
state[14] = u; |
|
v = state[3]; |
|
u = state[7]; |
|
t = state[11]; |
|
state[3] = state[15]; |
|
state[7] = v; |
|
state[11] = u; |
|
state[15] = t; |
|
//MixColumns |
|
for (var j = 0; j < 16; j += 4) { |
|
var s0 = state[j + 0], s1 = state[j + 1]; |
|
var s2 = state[j + 2], s3 = state[j + 3]; |
|
t = s0 ^ s1 ^ s2 ^ s3; |
|
state[j + 0] ^= t ^ mixCol[s0 ^ s1]; |
|
state[j + 1] ^= t ^ mixCol[s1 ^ s2]; |
|
state[j + 2] ^= t ^ mixCol[s2 ^ s3]; |
|
state[j + 3] ^= t ^ mixCol[s3 ^ s0]; |
|
} |
|
//AddRoundKey |
|
for (j = 0, k = i * 16; j < 16; ++j, ++k) { |
|
state[j] ^= key[k]; |
|
} |
|
} |
|
|
|
//SubBytes |
|
for (j = 0; j < 16; ++j) { |
|
state[j] = s[state[j]]; |
|
} |
|
//ShiftRows |
|
v = state[1]; |
|
state[1] = state[5]; |
|
state[5] = state[9]; |
|
state[9] = state[13]; |
|
state[13] = v; |
|
v = state[2]; |
|
u = state[6]; |
|
state[2] = state[10]; |
|
state[6] = state[14]; |
|
state[10] = v; |
|
state[14] = u; |
|
v = state[3]; |
|
u = state[7]; |
|
t = state[11]; |
|
state[3] = state[15]; |
|
state[7] = v; |
|
state[11] = u; |
|
state[15] = t; |
|
//AddRoundKey |
|
for (j = 0, k = 160; j < 16; ++j, ++k) { |
|
state[j] ^= key[k]; |
|
} |
|
return state; |
|
} |
|
|
|
function AES128Cipher(key) { |
|
this.key = expandKey128(key); |
|
this.buffer = new Uint8Array(16); |
|
this.bufferPosition = 0; |
|
} |
|
|
|
function decryptBlock2(data, finalize) { |
|
var i, j, ii, sourceLength = data.length, |
|
buffer = this.buffer, bufferLength = this.bufferPosition, |
|
result = [], iv = this.iv; |
|
for (i = 0; i < sourceLength; ++i) { |
|
buffer[bufferLength] = data[i]; |
|
++bufferLength; |
|
if (bufferLength < 16) { |
|
continue; |
|
} |
|
// buffer is full, decrypting |
|
var plain = decrypt128(buffer, this.key); |
|
// xor-ing the IV vector to get plain text |
|
for (j = 0; j < 16; ++j) { |
|
plain[j] ^= iv[j]; |
|
} |
|
iv = buffer; |
|
result.push(plain); |
|
buffer = new Uint8Array(16); |
|
bufferLength = 0; |
|
} |
|
// saving incomplete buffer |
|
this.buffer = buffer; |
|
this.bufferLength = bufferLength; |
|
this.iv = iv; |
|
if (result.length === 0) { |
|
return new Uint8Array([]); |
|
} |
|
// combining plain text blocks into one |
|
var outputLength = 16 * result.length; |
|
if (finalize) { |
|
// undo a padding that is described in RFC 2898 |
|
var lastBlock = result[result.length - 1]; |
|
var psLen = lastBlock[15]; |
|
if (psLen <= 16) { |
|
for (i = 15, ii = 16 - psLen; i >= ii; --i) { |
|
if (lastBlock[i] !== psLen) { |
|
// Invalid padding, assume that the block has no padding. |
|
psLen = 0; |
|
break; |
|
} |
|
} |
|
outputLength -= psLen; |
|
result[result.length - 1] = lastBlock.subarray(0, 16 - psLen); |
|
} |
|
} |
|
var output = new Uint8Array(outputLength); |
|
for (i = 0, j = 0, ii = result.length; i < ii; ++i, j += 16) { |
|
output.set(result[i], j); |
|
} |
|
return output; |
|
} |
|
|
|
AES128Cipher.prototype = { |
|
decryptBlock: function AES128Cipher_decryptBlock(data, finalize) { |
|
var i, sourceLength = data.length; |
|
var buffer = this.buffer, bufferLength = this.bufferPosition; |
|
// waiting for IV values -- they are at the start of the stream |
|
for (i = 0; bufferLength < 16 && i < sourceLength; ++i, ++bufferLength) { |
|
buffer[bufferLength] = data[i]; |
|
} |
|
if (bufferLength < 16) { |
|
// need more data |
|
this.bufferLength = bufferLength; |
|
return new Uint8Array([]); |
|
} |
|
this.iv = buffer; |
|
this.buffer = new Uint8Array(16); |
|
this.bufferLength = 0; |
|
// starting decryption |
|
this.decryptBlock = decryptBlock2; |
|
return this.decryptBlock(data.subarray(16), finalize); |
|
}, |
|
encrypt: function AES128Cipher_encrypt(data, iv) { |
|
var i, j, ii, sourceLength = data.length, |
|
buffer = this.buffer, bufferLength = this.bufferPosition, |
|
result = []; |
|
if (!iv) { |
|
iv = new Uint8Array(16); |
|
} |
|
for (i = 0; i < sourceLength; ++i) { |
|
buffer[bufferLength] = data[i]; |
|
++bufferLength; |
|
if (bufferLength < 16) { |
|
continue; |
|
} |
|
for (j = 0; j < 16; ++j) { |
|
buffer[j] ^= iv[j]; |
|
} |
|
|
|
// buffer is full, encrypting |
|
var cipher = encrypt128(buffer, this.key); |
|
iv = cipher; |
|
result.push(cipher); |
|
buffer = new Uint8Array(16); |
|
bufferLength = 0; |
|
} |
|
// saving incomplete buffer |
|
this.buffer = buffer; |
|
this.bufferLength = bufferLength; |
|
this.iv = iv; |
|
if (result.length === 0) { |
|
return new Uint8Array([]); |
|
} |
|
// combining plain text blocks into one |
|
var outputLength = 16 * result.length; |
|
var output = new Uint8Array(outputLength); |
|
for (i = 0, j = 0, ii = result.length; i < ii; ++i, j += 16) { |
|
output.set(result[i], j); |
|
} |
|
return output; |
|
} |
|
}; |
|
|
|
return AES128Cipher; |
|
})(); |
|
|
|
var AES256Cipher = (function AES256CipherClosure() { |
|
var rcon = new Uint8Array([ |
|
0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, |
|
0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, |
|
0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, |
|
0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, |
|
0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, |
|
0x1b, 0x36, 0x6c, 0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, |
|
0x97, 0x35, 0x6a, 0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, |
|
0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, |
|
0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, |
|
0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, 0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, |
|
0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, 0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, |
|
0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, |
|
0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, |
|
0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, 0xd8, 0xab, 0x4d, |
|
0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, 0xd4, 0xb3, 0x7d, |
|
0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, |
|
0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb, |
|
0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, |
|
0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, |
|
0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, |
|
0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, |
|
0x74, 0xe8, 0xcb, 0x8d]); |
|
|
|
var s = new Uint8Array([ |
|
0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, |
|
0xfe, 0xd7, 0xab, 0x76, 0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, |
|
0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0, 0xb7, 0xfd, 0x93, 0x26, |
|
0x36, 0x3f, 0xf7, 0xcc, 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15, |
|
0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a, 0x07, 0x12, 0x80, 0xe2, |
|
0xeb, 0x27, 0xb2, 0x75, 0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0, |
|
0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84, 0x53, 0xd1, 0x00, 0xed, |
|
0x20, 0xfc, 0xb1, 0x5b, 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf, |
|
0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85, 0x45, 0xf9, 0x02, 0x7f, |
|
0x50, 0x3c, 0x9f, 0xa8, 0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5, |
|
0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2, 0xcd, 0x0c, 0x13, 0xec, |
|
0x5f, 0x97, 0x44, 0x17, 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73, |
|
0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88, 0x46, 0xee, 0xb8, 0x14, |
|
0xde, 0x5e, 0x0b, 0xdb, 0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c, |
|
0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79, 0xe7, 0xc8, 0x37, 0x6d, |
|
0x8d, 0xd5, 0x4e, 0xa9, 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08, |
|
0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6, 0xe8, 0xdd, 0x74, 0x1f, |
|
0x4b, 0xbd, 0x8b, 0x8a, 0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e, |
|
0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e, 0xe1, 0xf8, 0x98, 0x11, |
|
0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf, |
|
0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, |
|
0xb0, 0x54, 0xbb, 0x16]); |
|
|
|
var inv_s = new Uint8Array([ |
|
0x52, 0x09, 0x6a, 0xd5, 0x30, 0x36, 0xa5, 0x38, 0xbf, 0x40, 0xa3, 0x9e, |
|
0x81, 0xf3, 0xd7, 0xfb, 0x7c, 0xe3, 0x39, 0x82, 0x9b, 0x2f, 0xff, 0x87, |
|
0x34, 0x8e, 0x43, 0x44, 0xc4, 0xde, 0xe9, 0xcb, 0x54, 0x7b, 0x94, 0x32, |
|
0xa6, 0xc2, 0x23, 0x3d, 0xee, 0x4c, 0x95, 0x0b, 0x42, 0xfa, 0xc3, 0x4e, |
|
0x08, 0x2e, 0xa1, 0x66, 0x28, 0xd9, 0x24, 0xb2, 0x76, 0x5b, 0xa2, 0x49, |
|
0x6d, 0x8b, 0xd1, 0x25, 0x72, 0xf8, 0xf6, 0x64, 0x86, 0x68, 0x98, 0x16, |
|
0xd4, 0xa4, 0x5c, 0xcc, 0x5d, 0x65, 0xb6, 0x92, 0x6c, 0x70, 0x48, 0x50, |
|
0xfd, 0xed, 0xb9, 0xda, 0x5e, 0x15, 0x46, 0x57, 0xa7, 0x8d, 0x9d, 0x84, |
|
0x90, 0xd8, 0xab, 0x00, 0x8c, 0xbc, 0xd3, 0x0a, 0xf7, 0xe4, 0x58, 0x05, |
|
0xb8, 0xb3, 0x45, 0x06, 0xd0, 0x2c, 0x1e, 0x8f, 0xca, 0x3f, 0x0f, 0x02, |
|
0xc1, 0xaf, 0xbd, 0x03, 0x01, 0x13, 0x8a, 0x6b, 0x3a, 0x91, 0x11, 0x41, |
|
0x4f, 0x67, 0xdc, 0xea, 0x97, 0xf2, 0xcf, 0xce, 0xf0, 0xb4, 0xe6, 0x73, |
|
0x96, 0xac, 0x74, 0x22, 0xe7, 0xad, 0x35, 0x85, 0xe2, 0xf9, 0x37, 0xe8, |
|
0x1c, 0x75, 0xdf, 0x6e, 0x47, 0xf1, 0x1a, 0x71, 0x1d, 0x29, 0xc5, 0x89, |
|
0x6f, 0xb7, 0x62, 0x0e, 0xaa, 0x18, 0xbe, 0x1b, 0xfc, 0x56, 0x3e, 0x4b, |
|
0xc6, 0xd2, 0x79, 0x20, 0x9a, 0xdb, 0xc0, 0xfe, 0x78, 0xcd, 0x5a, 0xf4, |
|
0x1f, 0xdd, 0xa8, 0x33, 0x88, 0x07, 0xc7, 0x31, 0xb1, 0x12, 0x10, 0x59, |
|
0x27, 0x80, 0xec, 0x5f, 0x60, 0x51, 0x7f, 0xa9, 0x19, 0xb5, 0x4a, 0x0d, |
|
0x2d, 0xe5, 0x7a, 0x9f, 0x93, 0xc9, 0x9c, 0xef, 0xa0, 0xe0, 0x3b, 0x4d, |
|
0xae, 0x2a, 0xf5, 0xb0, 0xc8, 0xeb, 0xbb, 0x3c, 0x83, 0x53, 0x99, 0x61, |
|
0x17, 0x2b, 0x04, 0x7e, 0xba, 0x77, 0xd6, 0x26, 0xe1, 0x69, 0x14, 0x63, |
|
0x55, 0x21, 0x0c, 0x7d]); |
|
|
|
var mixCol = new Uint8Array(256); |
|
for (var i = 0; i < 256; i++) { |
|
if (i < 128) { |
|
mixCol[i] = i << 1; |
|
} else { |
|
mixCol[i] = (i << 1) ^ 0x1b; |
|
} |
|
} |
|
var mix = new Uint32Array([ |
|
0x00000000, 0x0e090d0b, 0x1c121a16, 0x121b171d, 0x3824342c, 0x362d3927, |
|
0x24362e3a, 0x2a3f2331, 0x70486858, 0x7e416553, 0x6c5a724e, 0x62537f45, |
|
0x486c5c74, 0x4665517f, 0x547e4662, 0x5a774b69, 0xe090d0b0, 0xee99ddbb, |
|
0xfc82caa6, 0xf28bc7ad, 0xd8b4e49c, 0xd6bde997, 0xc4a6fe8a, 0xcaaff381, |
|
0x90d8b8e8, 0x9ed1b5e3, 0x8ccaa2fe, 0x82c3aff5, 0xa8fc8cc4, 0xa6f581cf, |
|
0xb4ee96d2, 0xbae79bd9, 0xdb3bbb7b, 0xd532b670, 0xc729a16d, 0xc920ac66, |
|
0xe31f8f57, 0xed16825c, 0xff0d9541, 0xf104984a, 0xab73d323, 0xa57ade28, |
|
0xb761c935, 0xb968c43e, 0x9357e70f, 0x9d5eea04, 0x8f45fd19, 0x814cf012, |
|
0x3bab6bcb, 0x35a266c0, 0x27b971dd, 0x29b07cd6, 0x038f5fe7, 0x0d8652ec, |
|
0x1f9d45f1, 0x119448fa, 0x4be30393, 0x45ea0e98, 0x57f11985, 0x59f8148e, |
|
0x73c737bf, 0x7dce3ab4, 0x6fd52da9, 0x61dc20a2, 0xad766df6, 0xa37f60fd, |
|
0xb16477e0, 0xbf6d7aeb, 0x955259da, 0x9b5b54d1, 0x894043cc, 0x87494ec7, |
|
0xdd3e05ae, 0xd33708a5, 0xc12c1fb8, 0xcf2512b3, 0xe51a3182, 0xeb133c89, |
|
0xf9082b94, 0xf701269f, 0x4de6bd46, 0x43efb04d, 0x51f4a750, 0x5ffdaa5b, |
|
0x75c2896a, 0x7bcb8461, 0x69d0937c, 0x67d99e77, 0x3daed51e, 0x33a7d815, |
|
0x21bccf08, 0x2fb5c203, 0x058ae132, 0x0b83ec39, 0x1998fb24, 0x1791f62f, |
|
0x764dd68d, 0x7844db86, 0x6a5fcc9b, 0x6456c190, 0x4e69e2a1, 0x4060efaa, |
|
0x527bf8b7, 0x5c72f5bc, 0x0605bed5, 0x080cb3de, 0x1a17a4c3, 0x141ea9c8, |
|
0x3e218af9, 0x302887f2, 0x223390ef, 0x2c3a9de4, 0x96dd063d, 0x98d40b36, |
|
0x8acf1c2b, 0x84c61120, 0xaef93211, 0xa0f03f1a, 0xb2eb2807, 0xbce2250c, |
|
0xe6956e65, 0xe89c636e, 0xfa877473, 0xf48e7978, 0xdeb15a49, 0xd0b85742, |
|
0xc2a3405f, 0xccaa4d54, 0x41ecdaf7, 0x4fe5d7fc, 0x5dfec0e1, 0x53f7cdea, |
|
0x79c8eedb, 0x77c1e3d0, 0x65daf4cd, 0x6bd3f9c6, 0x31a4b2af, 0x3fadbfa4, |
|
0x2db6a8b9, 0x23bfa5b2, 0x09808683, 0x07898b88, 0x15929c95, 0x1b9b919e, |
|
0xa17c0a47, 0xaf75074c, 0xbd6e1051, 0xb3671d5a, 0x99583e6b, 0x97513360, |
|
0x854a247d, 0x8b432976, 0xd134621f, 0xdf3d6f14, 0xcd267809, 0xc32f7502, |
|
0xe9105633, 0xe7195b38, 0xf5024c25, 0xfb0b412e, 0x9ad7618c, 0x94de6c87, |
|
0x86c57b9a, 0x88cc7691, 0xa2f355a0, 0xacfa58ab, 0xbee14fb6, 0xb0e842bd, |
|
0xea9f09d4, 0xe49604df, 0xf68d13c2, 0xf8841ec9, 0xd2bb3df8, 0xdcb230f3, |
|
0xcea927ee, 0xc0a02ae5, 0x7a47b13c, 0x744ebc37, 0x6655ab2a, 0x685ca621, |
|
0x42638510, 0x4c6a881b, 0x5e719f06, 0x5078920d, 0x0a0fd964, 0x0406d46f, |
|
0x161dc372, 0x1814ce79, 0x322bed48, 0x3c22e043, 0x2e39f75e, 0x2030fa55, |
|
0xec9ab701, 0xe293ba0a, 0xf088ad17, 0xfe81a01c, 0xd4be832d, 0xdab78e26, |
|
0xc8ac993b, 0xc6a59430, 0x9cd2df59, 0x92dbd252, 0x80c0c54f, 0x8ec9c844, |
|
0xa4f6eb75, 0xaaffe67e, 0xb8e4f163, 0xb6edfc68, 0x0c0a67b1, 0x02036aba, |
|
0x10187da7, 0x1e1170ac, 0x342e539d, 0x3a275e96, 0x283c498b, 0x26354480, |
|
0x7c420fe9, 0x724b02e2, 0x605015ff, 0x6e5918f4, 0x44663bc5, 0x4a6f36ce, |
|
0x587421d3, 0x567d2cd8, 0x37a10c7a, 0x39a80171, 0x2bb3166c, 0x25ba1b67, |
|
0x0f853856, 0x018c355d, 0x13972240, 0x1d9e2f4b, 0x47e96422, 0x49e06929, |
|
0x5bfb7e34, 0x55f2733f, 0x7fcd500e, 0x71c45d05, 0x63df4a18, 0x6dd64713, |
|
0xd731dcca, 0xd938d1c1, 0xcb23c6dc, 0xc52acbd7, 0xef15e8e6, 0xe11ce5ed, |
|
0xf307f2f0, 0xfd0efffb, 0xa779b492, 0xa970b999, 0xbb6bae84, 0xb562a38f, |
|
0x9f5d80be, 0x91548db5, 0x834f9aa8, 0x8d4697a3]); |
|
|
|
function expandKey256(cipherKey) { |
|
var b = 240, result = new Uint8Array(b); |
|
var r = 1; |
|
|
|
result.set(cipherKey); |
|
for (var j = 32, i = 1; j < b; ++i) { |
|
if (j % 32 === 16) { |
|
t1 = s[t1]; |
|
t2 = s[t2]; |
|
t3 = s[t3]; |
|
t4 = s[t4]; |
|
} else if (j % 32 === 0) { |
|
// RotWord |
|
var t1 = result[j - 3], t2 = result[j - 2], |
|
t3 = result[j - 1], t4 = result[j - 4]; |
|
// SubWord |
|
t1 = s[t1]; |
|
t2 = s[t2]; |
|
t3 = s[t3]; |
|
t4 = s[t4]; |
|
// Rcon |
|
t1 = t1 ^ r; |
|
if ((r <<= 1) >= 256) { |
|
r = (r ^ 0x1b) & 0xFF; |
|
} |
|
} |
|
|
|
for (var n = 0; n < 4; ++n) { |
|
result[j] = (t1 ^= result[j - 32]); |
|
j++; |
|
result[j] = (t2 ^= result[j - 32]); |
|
j++; |
|
result[j] = (t3 ^= result[j - 32]); |
|
j++; |
|
result[j] = (t4 ^= result[j - 32]); |
|
j++; |
|
} |
|
} |
|
return result; |
|
} |
|
|
|
function decrypt256(input, key) { |
|
var state = new Uint8Array(16); |
|
state.set(input); |
|
var i, j, k; |
|
var t, u, v; |
|
// AddRoundKey |
|
for (j = 0, k = 224; j < 16; ++j, ++k) { |
|
state[j] ^= key[k]; |
|
} |
|
for (i = 13; i >= 1; --i) { |
|
// InvShiftRows |
|
t = state[13]; |
|
state[13] = state[9]; |
|
state[9] = state[5]; |
|
state[5] = state[1]; |
|
state[1] = t; |
|
t = state[14]; |
|
u = state[10]; |
|
state[14] = state[6]; |
|
state[10] = state[2]; |
|
state[6] = t; |
|
state[2] = u; |
|
t = state[15]; |
|
u = state[11]; |
|
v = state[7]; |
|
state[15] = state[3]; |
|
state[11] = t; |
|
state[7] = u; |
|
state[3] = v; |
|
// InvSubBytes |
|
for (j = 0; j < 16; ++j) { |
|
state[j] = inv_s[state[j]]; |
|
} |
|
// AddRoundKey |
|
for (j = 0, k = i * 16; j < 16; ++j, ++k) { |
|
state[j] ^= key[k]; |
|
} |
|
// InvMixColumns |
|
for (j = 0; j < 16; j += 4) { |
|
var s0 = mix[state[j]], s1 = mix[state[j + 1]], |
|
s2 = mix[state[j + 2]], s3 = mix[state[j + 3]]; |
|
t = (s0 ^ (s1 >>> 8) ^ (s1 << 24) ^ (s2 >>> 16) ^ (s2 << 16) ^ |
|
(s3 >>> 24) ^ (s3 << 8)); |
|
state[j] = (t >>> 24) & 0xFF; |
|
state[j + 1] = (t >> 16) & 0xFF; |
|
state[j + 2] = (t >> 8) & 0xFF; |
|
state[j + 3] = t & 0xFF; |
|
} |
|
} |
|
// InvShiftRows |
|
t = state[13]; |
|
state[13] = state[9]; |
|
state[9] = state[5]; |
|
state[5] = state[1]; |
|
state[1] = t; |
|
t = state[14]; |
|
u = state[10]; |
|
state[14] = state[6]; |
|
state[10] = state[2]; |
|
state[6] = t; |
|
state[2] = u; |
|
t = state[15]; |
|
u = state[11]; |
|
v = state[7]; |
|
state[15] = state[3]; |
|
state[11] = t; |
|
state[7] = u; |
|
state[3] = v; |
|
for (j = 0; j < 16; ++j) { |
|
// InvSubBytes |
|
state[j] = inv_s[state[j]]; |
|
// AddRoundKey |
|
state[j] ^= key[j]; |
|
} |
|
return state; |
|
} |
|
|
|
function encrypt256(input, key) { |
|
var t, u, v, k; |
|
var state = new Uint8Array(16); |
|
state.set(input); |
|
for (j = 0; j < 16; ++j) { |
|
// AddRoundKey |
|
state[j] ^= key[j]; |
|
} |
|
|
|
for (i = 1; i < 14; i++) { |
|
//SubBytes |
|
for (j = 0; j < 16; ++j) { |
|
state[j] = s[state[j]]; |
|
} |
|
//ShiftRows |
|
v = state[1]; |
|
state[1] = state[5]; |
|
state[5] = state[9]; |
|
state[9] = state[13]; |
|
state[13] = v; |
|
v = state[2]; |
|
u = state[6]; |
|
state[2] = state[10]; |
|
state[6] = state[14]; |
|
state[10] = v; |
|
state[14] = u; |
|
v = state[3]; |
|
u = state[7]; |
|
t = state[11]; |
|
state[3] = state[15]; |
|
state[7] = v; |
|
state[11] = u; |
|
state[15] = t; |
|
//MixColumns |
|
for (var j = 0; j < 16; j += 4) { |
|
var s0 = state[j + 0], s1 = state[j + 1]; |
|
var s2 = state[j + 2], s3 = state[j + 3]; |
|
t = s0 ^ s1 ^ s2 ^ s3; |
|
state[j + 0] ^= t ^ mixCol[s0 ^ s1]; |
|
state[j + 1] ^= t ^ mixCol[s1 ^ s2]; |
|
state[j + 2] ^= t ^ mixCol[s2 ^ s3]; |
|
state[j + 3] ^= t ^ mixCol[s3 ^ s0]; |
|
} |
|
//AddRoundKey |
|
for (j = 0, k = i * 16; j < 16; ++j, ++k) { |
|
state[j] ^= key[k]; |
|
} |
|
} |
|
|
|
//SubBytes |
|
for (j = 0; j < 16; ++j) { |
|
state[j] = s[state[j]]; |
|
} |
|
//ShiftRows |
|
v = state[1]; |
|
state[1] = state[5]; |
|
state[5] = state[9]; |
|
state[9] = state[13]; |
|
state[13] = v; |
|
v = state[2]; |
|
u = state[6]; |
|
state[2] = state[10]; |
|
state[6] = state[14]; |
|
state[10] = v; |
|
state[14] = u; |
|
v = state[3]; |
|
u = state[7]; |
|
t = state[11]; |
|
state[3] = state[15]; |
|
state[7] = v; |
|
state[11] = u; |
|
state[15] = t; |
|
//AddRoundKey |
|
for (j = 0, k = 224; j < 16; ++j, ++k) { |
|
state[j] ^= key[k]; |
|
} |
|
|
|
return state; |
|
|
|
} |
|
|
|
function AES256Cipher(key) { |
|
this.key = expandKey256(key); |
|
this.buffer = new Uint8Array(16); |
|
this.bufferPosition = 0; |
|
} |
|
|
|
function decryptBlock2(data, finalize) { |
|
var i, j, ii, sourceLength = data.length, |
|
buffer = this.buffer, bufferLength = this.bufferPosition, |
|
result = [], iv = this.iv; |
|
|
|
for (i = 0; i < sourceLength; ++i) { |
|
buffer[bufferLength] = data[i]; |
|
++bufferLength; |
|
if (bufferLength < 16) { |
|
continue; |
|
} |
|
// buffer is full, decrypting |
|
var plain = decrypt256(buffer, this.key); |
|
// xor-ing the IV vector to get plain text |
|
for (j = 0; j < 16; ++j) { |
|
plain[j] ^= iv[j]; |
|
} |
|
iv = buffer; |
|
result.push(plain); |
|
buffer = new Uint8Array(16); |
|
bufferLength = 0; |
|
} |
|
// saving incomplete buffer |
|
this.buffer = buffer; |
|
this.bufferLength = bufferLength; |
|
this.iv = iv; |
|
if (result.length === 0) { |
|
return new Uint8Array([]); |
|
} |
|
// combining plain text blocks into one |
|
var outputLength = 16 * result.length; |
|
if (finalize) { |
|
// undo a padding that is described in RFC 2898 |
|
var lastBlock = result[result.length - 1]; |
|
var psLen = lastBlock[15]; |
|
if (psLen <= 16) { |
|
for (i = 15, ii = 16 - psLen; i >= ii; --i) { |
|
if (lastBlock[i] !== psLen) { |
|
// Invalid padding, assume that the block has no padding. |
|
psLen = 0; |
|
break; |
|
} |
|
} |
|
outputLength -= psLen; |
|
result[result.length - 1] = lastBlock.subarray(0, 16 - psLen); |
|
} |
|
} |
|
var output = new Uint8Array(outputLength); |
|
for (i = 0, j = 0, ii = result.length; i < ii; ++i, j += 16) { |
|
output.set(result[i], j); |
|
} |
|
return output; |
|
|
|
} |
|
|
|
AES256Cipher.prototype = { |
|
decryptBlock: function AES256Cipher_decryptBlock(data, finalize, iv) { |
|
var i, sourceLength = data.length; |
|
var buffer = this.buffer, bufferLength = this.bufferPosition; |
|
// if not supplied an IV wait for IV values |
|
// they are at the start of the stream |
|
if (iv) { |
|
this.iv = iv; |
|
} else { |
|
for (i = 0; bufferLength < 16 && |
|
i < sourceLength; ++i, ++bufferLength) { |
|
buffer[bufferLength] = data[i]; |
|
} |
|
if (bufferLength < 16) { |
|
//need more data |
|
this.bufferLength = bufferLength; |
|
return new Uint8Array([]); |
|
} |
|
this.iv = buffer; |
|
data = data.subarray(16); |
|
} |
|
this.buffer = new Uint8Array(16); |
|
this.bufferLength = 0; |
|
// starting decryption |
|
this.decryptBlock = decryptBlock2; |
|
return this.decryptBlock(data, finalize); |
|
}, |
|
encrypt: function AES256Cipher_encrypt(data, iv) { |
|
var i, j, ii, sourceLength = data.length, |
|
buffer = this.buffer, bufferLength = this.bufferPosition, |
|
result = []; |
|
if (!iv) { |
|
iv = new Uint8Array(16); |
|
} |
|
for (i = 0; i < sourceLength; ++i) { |
|
buffer[bufferLength] = data[i]; |
|
++bufferLength; |
|
if (bufferLength < 16) { |
|
continue; |
|
} |
|
for (j = 0; j < 16; ++j) { |
|
buffer[j] ^= iv[j]; |
|
} |
|
|
|
// buffer is full, encrypting |
|
var cipher = encrypt256(buffer, this.key); |
|
this.iv = cipher; |
|
result.push(cipher); |
|
buffer = new Uint8Array(16); |
|
bufferLength = 0; |
|
} |
|
// saving incomplete buffer |
|
this.buffer = buffer; |
|
this.bufferLength = bufferLength; |
|
this.iv = iv; |
|
if (result.length === 0) { |
|
return new Uint8Array([]); |
|
} |
|
// combining plain text blocks into one |
|
var outputLength = 16 * result.length; |
|
var output = new Uint8Array(outputLength); |
|
for (i = 0, j = 0, ii = result.length; i < ii; ++i, j += 16) { |
|
output.set(result[i], j); |
|
} |
|
return output; |
|
} |
|
}; |
|
|
|
return AES256Cipher; |
|
})(); |
|
|
|
var PDF17 = (function PDF17Closure() { |
|
|
|
function compareByteArrays(array1, array2) { |
|
if (array1.length !== array2.length) { |
|
return false; |
|
} |
|
for (var i = 0; i < array1.length; i++) { |
|
if (array1[i] !== array2[i]) { |
|
return false; |
|
} |
|
} |
|
return true; |
|
} |
|
|
|
function PDF17() { |
|
} |
|
|
|
PDF17.prototype = { |
|
checkOwnerPassword: function PDF17_checkOwnerPassword(password, |
|
ownerValidationSalt, |
|
userBytes, |
|
ownerPassword) { |
|
var hashData = new Uint8Array(password.length + 56); |
|
hashData.set(password, 0); |
|
hashData.set(ownerValidationSalt, password.length); |
|
hashData.set(userBytes, password.length + ownerValidationSalt.length); |
|
var result = calculateSHA256(hashData, 0, hashData.length); |
|
return compareByteArrays(result, ownerPassword); |
|
}, |
|
checkUserPassword: function PDF17_checkUserPassword(password, |
|
userValidationSalt, |
|
userPassword) { |
|
var hashData = new Uint8Array(password.length + 8); |
|
hashData.set(password, 0); |
|
hashData.set(userValidationSalt, password.length); |
|
var result = calculateSHA256(hashData, 0, hashData.length); |
|
return compareByteArrays(result, userPassword); |
|
}, |
|
getOwnerKey: function PDF17_getOwnerKey(password, ownerKeySalt, userBytes, |
|
ownerEncryption) { |
|
var hashData = new Uint8Array(password.length + 56); |
|
hashData.set(password, 0); |
|
hashData.set(ownerKeySalt, password.length); |
|
hashData.set(userBytes, password.length + ownerKeySalt.length); |
|
var key = calculateSHA256(hashData, 0, hashData.length); |
|
var cipher = new AES256Cipher(key); |
|
return cipher.decryptBlock(ownerEncryption, |
|
false, |
|
new Uint8Array(16)); |
|
|
|
}, |
|
getUserKey: function PDF17_getUserKey(password, userKeySalt, |
|
userEncryption) { |
|
var hashData = new Uint8Array(password.length + 8); |
|
hashData.set(password, 0); |
|
hashData.set(userKeySalt, password.length); |
|
//key is the decryption key for the UE string |
|
var key = calculateSHA256(hashData, 0, hashData.length); |
|
var cipher = new AES256Cipher(key); |
|
return cipher.decryptBlock(userEncryption, |
|
false, |
|
new Uint8Array(16)); |
|
} |
|
}; |
|
return PDF17; |
|
})(); |
|
|
|
var PDF20 = (function PDF20Closure() { |
|
|
|
function concatArrays(array1, array2) { |
|
var t = new Uint8Array(array1.length + array2.length); |
|
t.set(array1, 0); |
|
t.set(array2, array1.length); |
|
return t; |
|
} |
|
|
|
function calculatePDF20Hash(password, input, userBytes) { |
|
//This refers to Algorithm 2.B as defined in ISO 32000-2 |
|
var k = calculateSHA256(input, 0, input.length).subarray(0, 32); |
|
var e = [0]; |
|
var i = 0; |
|
while (i < 64 || e[e.length - 1] > i - 32) { |
|
var arrayLength = password.length + k.length + userBytes.length; |
|
|
|
var k1 = new Uint8Array(arrayLength * 64); |
|
var array = concatArrays(password, k); |
|
array = concatArrays(array, userBytes); |
|
for (var j = 0, pos = 0; j < 64; j++, pos += arrayLength) { |
|
k1.set(array, pos); |
|
} |
|
//AES128 CBC NO PADDING with |
|
//first 16 bytes of k as the key and the second 16 as the iv. |
|
var cipher = new AES128Cipher(k.subarray(0, 16)); |
|
e = cipher.encrypt(k1, k.subarray(16, 32)); |
|
//Now we have to take the first 16 bytes of an unsigned |
|
//big endian integer... and compute the remainder |
|
//modulo 3.... That is a fairly large number and |
|
//JavaScript isn't going to handle that well... |
|
//So we're using a trick that allows us to perform |
|
//modulo math byte by byte |
|
var remainder = 0; |
|
for (var z = 0; z < 16; z++) { |
|
remainder *= (256 % 3); |
|
remainder %= 3; |
|
remainder += ((e[z] >>> 0) % 3); |
|
remainder %= 3; |
|
} |
|
if (remainder === 0) { |
|
k = calculateSHA256(e, 0, e.length); |
|
} |
|
else if (remainder === 1) { |
|
k = calculateSHA384(e, 0, e.length); |
|
} |
|
else if (remainder === 2) { |
|
k = calculateSHA512(e, 0, e.length); |
|
} |
|
i++; |
|
} |
|
return k.subarray(0, 32); |
|
} |
|
|
|
function PDF20() { |
|
} |
|
|
|
function compareByteArrays(array1, array2) { |
|
if (array1.length !== array2.length) { |
|
return false; |
|
} |
|
for (var i = 0; i < array1.length; i++) { |
|
if (array1[i] !== array2[i]) { |
|
return false; |
|
} |
|
} |
|
return true; |
|
} |
|
|
|
PDF20.prototype = { |
|
hash: function PDF20_hash(password, concatBytes, userBytes) { |
|
return calculatePDF20Hash(password, concatBytes, userBytes); |
|
}, |
|
checkOwnerPassword: function PDF20_checkOwnerPassword(password, |
|
ownerValidationSalt, |
|
userBytes, |
|
ownerPassword) { |
|
var hashData = new Uint8Array(password.length + 56); |
|
hashData.set(password, 0); |
|
hashData.set(ownerValidationSalt, password.length); |
|
hashData.set(userBytes, password.length + ownerValidationSalt.length); |
|
var result = calculatePDF20Hash(password, hashData, userBytes); |
|
return compareByteArrays(result, ownerPassword); |
|
}, |
|
checkUserPassword: function PDF20_checkUserPassword(password, |
|
userValidationSalt, |
|
userPassword) { |
|
var hashData = new Uint8Array(password.length + 8); |
|
hashData.set(password, 0); |
|
hashData.set(userValidationSalt, password.length); |
|
var result = calculatePDF20Hash(password, hashData, []); |
|
return compareByteArrays(result, userPassword); |
|
}, |
|
getOwnerKey: function PDF20_getOwnerKey(password, ownerKeySalt, userBytes, |
|
ownerEncryption) { |
|
var hashData = new Uint8Array(password.length + 56); |
|
hashData.set(password, 0); |
|
hashData.set(ownerKeySalt, password.length); |
|
hashData.set(userBytes, password.length + ownerKeySalt.length); |
|
var key = calculatePDF20Hash(password, hashData, userBytes); |
|
var cipher = new AES256Cipher(key); |
|
return cipher.decryptBlock(ownerEncryption, |
|
false, |
|
new Uint8Array(16)); |
|
|
|
}, |
|
getUserKey: function PDF20_getUserKey(password, userKeySalt, |
|
userEncryption) { |
|
var hashData = new Uint8Array(password.length + 8); |
|
hashData.set(password, 0); |
|
hashData.set(userKeySalt, password.length); |
|
//key is the decryption key for the UE string |
|
var key = calculatePDF20Hash(password, hashData, []); |
|
var cipher = new AES256Cipher(key); |
|
return cipher.decryptBlock(userEncryption, |
|
false, |
|
new Uint8Array(16)); |
|
} |
|
}; |
|
return PDF20; |
|
})(); |
|
|
|
var CipherTransform = (function CipherTransformClosure() { |
|
function CipherTransform(stringCipherConstructor, streamCipherConstructor) { |
|
this.stringCipherConstructor = stringCipherConstructor; |
|
this.streamCipherConstructor = streamCipherConstructor; |
|
} |
|
|
|
CipherTransform.prototype = { |
|
createStream: function CipherTransform_createStream(stream, length) { |
|
var cipher = new this.streamCipherConstructor(); |
|
return new DecryptStream(stream, length, |
|
function cipherTransformDecryptStream(data, finalize) { |
|
return cipher.decryptBlock(data, finalize); |
|
} |
|
); |
|
}, |
|
decryptString: function CipherTransform_decryptString(s) { |
|
var cipher = new this.stringCipherConstructor(); |
|
var data = stringToBytes(s); |
|
data = cipher.decryptBlock(data, true); |
|
return bytesToString(data); |
|
} |
|
}; |
|
return CipherTransform; |
|
})(); |
|
|
|
var CipherTransformFactory = (function CipherTransformFactoryClosure() { |
|
var defaultPasswordBytes = new Uint8Array([ |
|
0x28, 0xBF, 0x4E, 0x5E, 0x4E, 0x75, 0x8A, 0x41, |
|
0x64, 0x00, 0x4E, 0x56, 0xFF, 0xFA, 0x01, 0x08, |
|
0x2E, 0x2E, 0x00, 0xB6, 0xD0, 0x68, 0x3E, 0x80, |
|
0x2F, 0x0C, 0xA9, 0xFE, 0x64, 0x53, 0x69, 0x7A]); |
|
|
|
function createEncryptionKey20(revision, password, ownerPassword, |
|
ownerValidationSalt, ownerKeySalt, uBytes, |
|
userPassword, userValidationSalt, userKeySalt, |
|
ownerEncryption, userEncryption, perms) { |
|
if (password) { |
|
var passwordLength = Math.min(127, password.length); |
|
password = password.subarray(0, passwordLength); |
|
} else { |
|
password = []; |
|
} |
|
var pdfAlgorithm; |
|
if (revision === 6) { |
|
pdfAlgorithm = new PDF20(); |
|
} else { |
|
pdfAlgorithm = new PDF17(); |
|
} |
|
|
|
if (pdfAlgorithm) { |
|
if (pdfAlgorithm.checkUserPassword(password, userValidationSalt, |
|
userPassword)) { |
|
return pdfAlgorithm.getUserKey(password, userKeySalt, userEncryption); |
|
} else if (password.length && pdfAlgorithm.checkOwnerPassword(password, |
|
ownerValidationSalt, |
|
uBytes, |
|
ownerPassword)) { |
|
return pdfAlgorithm.getOwnerKey(password, ownerKeySalt, uBytes, |
|
ownerEncryption); |
|
} |
|
} |
|
|
|
return null; |
|
} |
|
|
|
function prepareKeyData(fileId, password, ownerPassword, userPassword, |
|
flags, revision, keyLength, encryptMetadata) { |
|
var hashDataSize = 40 + ownerPassword.length + fileId.length; |
|
var hashData = new Uint8Array(hashDataSize), i = 0, j, n; |
|
if (password) { |
|
n = Math.min(32, password.length); |
|
for (; i < n; ++i) { |
|
hashData[i] = password[i]; |
|
} |
|
} |
|
j = 0; |
|
while (i < 32) { |
|
hashData[i++] = defaultPasswordBytes[j++]; |
|
} |
|
// as now the padded password in the hashData[0..i] |
|
for (j = 0, n = ownerPassword.length; j < n; ++j) { |
|
hashData[i++] = ownerPassword[j]; |
|
} |
|
hashData[i++] = flags & 0xFF; |
|
hashData[i++] = (flags >> 8) & 0xFF; |
|
hashData[i++] = (flags >> 16) & 0xFF; |
|
hashData[i++] = (flags >>> 24) & 0xFF; |
|
for (j = 0, n = fileId.length; j < n; ++j) { |
|
hashData[i++] = fileId[j]; |
|
} |
|
if (revision >= 4 && !encryptMetadata) { |
|
hashData[i++] = 0xFF; |
|
hashData[i++] = 0xFF; |
|
hashData[i++] = 0xFF; |
|
hashData[i++] = 0xFF; |
|
} |
|
var hash = calculateMD5(hashData, 0, i); |
|
var keyLengthInBytes = keyLength >> 3; |
|
if (revision >= 3) { |
|
for (j = 0; j < 50; ++j) { |
|
hash = calculateMD5(hash, 0, keyLengthInBytes); |
|
} |
|
} |
|
var encryptionKey = hash.subarray(0, keyLengthInBytes); |
|
var cipher, checkData; |
|
|
|
if (revision >= 3) { |
|
for (i = 0; i < 32; ++i) { |
|
hashData[i] = defaultPasswordBytes[i]; |
|
} |
|
for (j = 0, n = fileId.length; j < n; ++j) { |
|
hashData[i++] = fileId[j]; |
|
} |
|
cipher = new ARCFourCipher(encryptionKey); |
|
checkData = cipher.encryptBlock(calculateMD5(hashData, 0, i)); |
|
n = encryptionKey.length; |
|
var derivedKey = new Uint8Array(n), k; |
|
for (j = 1; j <= 19; ++j) { |
|
for (k = 0; k < n; ++k) { |
|
derivedKey[k] = encryptionKey[k] ^ j; |
|
} |
|
cipher = new ARCFourCipher(derivedKey); |
|
checkData = cipher.encryptBlock(checkData); |
|
} |
|
for (j = 0, n = checkData.length; j < n; ++j) { |
|
if (userPassword[j] !== checkData[j]) { |
|
return null; |
|
} |
|
} |
|
} else { |
|
cipher = new ARCFourCipher(encryptionKey); |
|
checkData = cipher.encryptBlock(defaultPasswordBytes); |
|
for (j = 0, n = checkData.length; j < n; ++j) { |
|
if (userPassword[j] !== checkData[j]) { |
|
return null; |
|
} |
|
} |
|
} |
|
return encryptionKey; |
|
} |
|
|
|
function decodeUserPassword(password, ownerPassword, revision, keyLength) { |
|
var hashData = new Uint8Array(32), i = 0, j, n; |
|
n = Math.min(32, password.length); |
|
for (; i < n; ++i) { |
|
hashData[i] = password[i]; |
|
} |
|
j = 0; |
|
while (i < 32) { |
|
hashData[i++] = defaultPasswordBytes[j++]; |
|
} |
|
var hash = calculateMD5(hashData, 0, i); |
|
var keyLengthInBytes = keyLength >> 3; |
|
if (revision >= 3) { |
|
for (j = 0; j < 50; ++j) { |
|
hash = calculateMD5(hash, 0, hash.length); |
|
} |
|
} |
|
|
|
var cipher, userPassword; |
|
if (revision >= 3) { |
|
userPassword = ownerPassword; |
|
var derivedKey = new Uint8Array(keyLengthInBytes), k; |
|
for (j = 19; j >= 0; j--) { |
|
for (k = 0; k < keyLengthInBytes; ++k) { |
|
derivedKey[k] = hash[k] ^ j; |
|
} |
|
cipher = new ARCFourCipher(derivedKey); |
|
userPassword = cipher.encryptBlock(userPassword); |
|
} |
|
} else { |
|
cipher = new ARCFourCipher(hash.subarray(0, keyLengthInBytes)); |
|
userPassword = cipher.encryptBlock(ownerPassword); |
|
} |
|
return userPassword; |
|
} |
|
|
|
var identityName = Name.get('Identity'); |
|
|
|
function CipherTransformFactory(dict, fileId, password) { |
|
var filter = dict.get('Filter'); |
|
if (!isName(filter) || filter.name !== 'Standard') { |
|
error('unknown encryption method'); |
|
} |
|
this.dict = dict; |
|
var algorithm = dict.get('V'); |
|
if (!isInt(algorithm) || |
|
(algorithm !== 1 && algorithm !== 2 && algorithm !== 4 && |
|
algorithm !== 5)) { |
|
error('unsupported encryption algorithm'); |
|
} |
|
this.algorithm = algorithm; |
|
var keyLength = dict.get('Length'); |
|
if (!keyLength) { |
|
// Spec asks to rely on encryption dictionary's Length entry, however |
|
// some PDFs don't have it. Trying to recover. |
|
if (algorithm <= 3) { |
|
// For 1 and 2 it's fixed to 40-bit, for 3 40-bit is a minimal value. |
|
keyLength = 40; |
|
} else { |
|
// Trying to find default handler -- it usually has Length. |
|
var cfDict = dict.get('CF'); |
|
var streamCryptoName = dict.get('StmF'); |
|
if (isDict(cfDict) && isName(streamCryptoName)) { |
|
var handlerDict = cfDict.get(streamCryptoName.name); |
|
keyLength = (handlerDict && handlerDict.get('Length')) || 128; |
|
if (keyLength < 40) { |
|
// Sometimes it's incorrect value of bits, generators specify bytes. |
|
keyLength <<= 3; |
|
} |
|
} |
|
} |
|
} |
|
if (!isInt(keyLength) || |
|
keyLength < 40 || (keyLength % 8) !== 0) { |
|
error('invalid key length'); |
|
} |
|
|
|
// prepare keys |
|
var ownerPassword = stringToBytes(dict.get('O')).subarray(0, 32); |
|
var userPassword = stringToBytes(dict.get('U')).subarray(0, 32); |
|
var flags = dict.get('P'); |
|
var revision = dict.get('R'); |
|
// meaningful when V is 4 or 5 |
|
var encryptMetadata = ((algorithm === 4 || algorithm === 5) && |
|
dict.get('EncryptMetadata') !== false); |
|
this.encryptMetadata = encryptMetadata; |
|
|
|
var fileIdBytes = stringToBytes(fileId); |
|
var passwordBytes; |
|
if (password) { |
|
if (revision === 6) { |
|
try { |
|
password = utf8StringToString(password); |
|
} catch (ex) { |
|
warn('CipherTransformFactory: ' + |
|
'Unable to convert UTF8 encoded password.'); |
|
} |
|
} |
|
passwordBytes = stringToBytes(password); |
|
} |
|
|
|
var encryptionKey; |
|
if (algorithm !== 5) { |
|
encryptionKey = prepareKeyData(fileIdBytes, passwordBytes, |
|
ownerPassword, userPassword, flags, |
|
revision, keyLength, encryptMetadata); |
|
} |
|
else { |
|
var ownerValidationSalt = stringToBytes(dict.get('O')).subarray(32, 40); |
|
var ownerKeySalt = stringToBytes(dict.get('O')).subarray(40, 48); |
|
var uBytes = stringToBytes(dict.get('U')).subarray(0, 48); |
|
var userValidationSalt = stringToBytes(dict.get('U')).subarray(32, 40); |
|
var userKeySalt = stringToBytes(dict.get('U')).subarray(40, 48); |
|
var ownerEncryption = stringToBytes(dict.get('OE')); |
|
var userEncryption = stringToBytes(dict.get('UE')); |
|
var perms = stringToBytes(dict.get('Perms')); |
|
encryptionKey = |
|
createEncryptionKey20(revision, passwordBytes, |
|
ownerPassword, ownerValidationSalt, |
|
ownerKeySalt, uBytes, |
|
userPassword, userValidationSalt, |
|
userKeySalt, ownerEncryption, |
|
userEncryption, perms); |
|
} |
|
if (!encryptionKey && !password) { |
|
throw new PasswordException('No password given', |
|
PasswordResponses.NEED_PASSWORD); |
|
} else if (!encryptionKey && password) { |
|
// Attempting use the password as an owner password |
|
var decodedPassword = decodeUserPassword(passwordBytes, ownerPassword, |
|
revision, keyLength); |
|
encryptionKey = prepareKeyData(fileIdBytes, decodedPassword, |
|
ownerPassword, userPassword, flags, |
|
revision, keyLength, encryptMetadata); |
|
} |
|
|
|
if (!encryptionKey) { |
|
throw new PasswordException('Incorrect Password', |
|
PasswordResponses.INCORRECT_PASSWORD); |
|
} |
|
|
|
this.encryptionKey = encryptionKey; |
|
|
|
if (algorithm >= 4) { |
|
this.cf = dict.get('CF'); |
|
this.stmf = dict.get('StmF') || identityName; |
|
this.strf = dict.get('StrF') || identityName; |
|
this.eff = dict.get('EFF') || this.stmf; |
|
} |
|
} |
|
|
|
function buildObjectKey(num, gen, encryptionKey, isAes) { |
|
var key = new Uint8Array(encryptionKey.length + 9), i, n; |
|
for (i = 0, n = encryptionKey.length; i < n; ++i) { |
|
key[i] = encryptionKey[i]; |
|
} |
|
key[i++] = num & 0xFF; |
|
key[i++] = (num >> 8) & 0xFF; |
|
key[i++] = (num >> 16) & 0xFF; |
|
key[i++] = gen & 0xFF; |
|
key[i++] = (gen >> 8) & 0xFF; |
|
if (isAes) { |
|
key[i++] = 0x73; |
|
key[i++] = 0x41; |
|
key[i++] = 0x6C; |
|
key[i++] = 0x54; |
|
} |
|
var hash = calculateMD5(key, 0, i); |
|
return hash.subarray(0, Math.min(encryptionKey.length + 5, 16)); |
|
} |
|
|
|
function buildCipherConstructor(cf, name, num, gen, key) { |
|
var cryptFilter = cf.get(name.name); |
|
var cfm; |
|
if (cryptFilter !== null && cryptFilter !== undefined) { |
|
cfm = cryptFilter.get('CFM'); |
|
} |
|
if (!cfm || cfm.name === 'None') { |
|
return function cipherTransformFactoryBuildCipherConstructorNone() { |
|
return new NullCipher(); |
|
}; |
|
} |
|
if ('V2' === cfm.name) { |
|
return function cipherTransformFactoryBuildCipherConstructorV2() { |
|
return new ARCFourCipher(buildObjectKey(num, gen, key, false)); |
|
}; |
|
} |
|
if ('AESV2' === cfm.name) { |
|
return function cipherTransformFactoryBuildCipherConstructorAESV2() { |
|
return new AES128Cipher(buildObjectKey(num, gen, key, true)); |
|
}; |
|
} |
|
if ('AESV3' === cfm.name) { |
|
return function cipherTransformFactoryBuildCipherConstructorAESV3() { |
|
return new AES256Cipher(key); |
|
}; |
|
} |
|
error('Unknown crypto method'); |
|
} |
|
|
|
CipherTransformFactory.prototype = { |
|
createCipherTransform: |
|
function CipherTransformFactory_createCipherTransform(num, gen) { |
|
if (this.algorithm === 4 || this.algorithm === 5) { |
|
return new CipherTransform( |
|
buildCipherConstructor(this.cf, this.stmf, |
|
num, gen, this.encryptionKey), |
|
buildCipherConstructor(this.cf, this.strf, |
|
num, gen, this.encryptionKey)); |
|
} |
|
// algorithms 1 and 2 |
|
var key = buildObjectKey(num, gen, this.encryptionKey, false); |
|
var cipherConstructor = function buildCipherCipherConstructor() { |
|
return new ARCFourCipher(key); |
|
}; |
|
return new CipherTransform(cipherConstructor, cipherConstructor); |
|
} |
|
}; |
|
|
|
return CipherTransformFactory; |
|
})(); |
|
|
|
exports.AES128Cipher = AES128Cipher; |
|
exports.AES256Cipher = AES256Cipher; |
|
exports.ARCFourCipher = ARCFourCipher; |
|
exports.CipherTransformFactory = CipherTransformFactory; |
|
exports.PDF17 = PDF17; |
|
exports.PDF20 = PDF20; |
|
exports.calculateMD5 = calculateMD5; |
|
exports.calculateSHA256 = calculateSHA256; |
|
exports.calculateSHA384 = calculateSHA384; |
|
exports.calculateSHA512 = calculateSHA512; |
|
})); |
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreFontRenderer = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCoreStream, root.pdfjsCoreGlyphList); |
|
} |
|
}(this, function (exports, sharedUtil, coreStream, coreGlyphList) { |
|
|
|
var Util = sharedUtil.Util; |
|
var bytesToString = sharedUtil.bytesToString; |
|
var error = sharedUtil.error; |
|
var Stream = coreStream.Stream; |
|
var GlyphsUnicode = coreGlyphList.GlyphsUnicode; |
|
|
|
var coreFonts; // see _setCoreFonts below |
|
var CFFParser; // = coreFonts.CFFParser; |
|
var Encodings; // = coreFonts.Encodings; |
|
|
|
var FontRendererFactory = (function FontRendererFactoryClosure() { |
|
function getLong(data, offset) { |
|
return (data[offset] << 24) | (data[offset + 1] << 16) | |
|
(data[offset + 2] << 8) | data[offset + 3]; |
|
} |
|
|
|
function getUshort(data, offset) { |
|
return (data[offset] << 8) | data[offset + 1]; |
|
} |
|
|
|
function parseCmap(data, start, end) { |
|
var offset = (getUshort(data, start + 2) === 1 ? |
|
getLong(data, start + 8) : getLong(data, start + 16)); |
|
var format = getUshort(data, start + offset); |
|
var length, ranges, p, i; |
|
if (format === 4) { |
|
length = getUshort(data, start + offset + 2); |
|
var segCount = getUshort(data, start + offset + 6) >> 1; |
|
p = start + offset + 14; |
|
ranges = []; |
|
for (i = 0; i < segCount; i++, p += 2) { |
|
ranges[i] = {end: getUshort(data, p)}; |
|
} |
|
p += 2; |
|
for (i = 0; i < segCount; i++, p += 2) { |
|
ranges[i].start = getUshort(data, p); |
|
} |
|
for (i = 0; i < segCount; i++, p += 2) { |
|
ranges[i].idDelta = getUshort(data, p); |
|
} |
|
for (i = 0; i < segCount; i++, p += 2) { |
|
var idOffset = getUshort(data, p); |
|
if (idOffset === 0) { |
|
continue; |
|
} |
|
ranges[i].ids = []; |
|
for (var j = 0, jj = ranges[i].end - ranges[i].start + 1; j < jj; j++) { |
|
ranges[i].ids[j] = getUshort(data, p + idOffset); |
|
idOffset += 2; |
|
} |
|
} |
|
return ranges; |
|
} else if (format === 12) { |
|
length = getLong(data, start + offset + 4); |
|
var groups = getLong(data, start + offset + 12); |
|
p = start + offset + 16; |
|
ranges = []; |
|
for (i = 0; i < groups; i++) { |
|
ranges.push({ |
|
start: getLong(data, p), |
|
end: getLong(data, p + 4), |
|
idDelta: getLong(data, p + 8) - getLong(data, p) |
|
}); |
|
p += 12; |
|
} |
|
return ranges; |
|
} |
|
error('not supported cmap: ' + format); |
|
} |
|
|
|
function parseCff(data, start, end) { |
|
var properties = {}; |
|
var parser = new CFFParser(new Stream(data, start, end - start), |
|
properties); |
|
var cff = parser.parse(); |
|
return { |
|
glyphs: cff.charStrings.objects, |
|
subrs: (cff.topDict.privateDict && cff.topDict.privateDict.subrsIndex && |
|
cff.topDict.privateDict.subrsIndex.objects), |
|
gsubrs: cff.globalSubrIndex && cff.globalSubrIndex.objects |
|
}; |
|
} |
|
|
|
function parseGlyfTable(glyf, loca, isGlyphLocationsLong) { |
|
var itemSize, itemDecode; |
|
if (isGlyphLocationsLong) { |
|
itemSize = 4; |
|
itemDecode = function fontItemDecodeLong(data, offset) { |
|
return (data[offset] << 24) | (data[offset + 1] << 16) | |
|
(data[offset + 2] << 8) | data[offset + 3]; |
|
}; |
|
} else { |
|
itemSize = 2; |
|
itemDecode = function fontItemDecode(data, offset) { |
|
return (data[offset] << 9) | (data[offset + 1] << 1); |
|
}; |
|
} |
|
var glyphs = []; |
|
var startOffset = itemDecode(loca, 0); |
|
for (var j = itemSize; j < loca.length; j += itemSize) { |
|
var endOffset = itemDecode(loca, j); |
|
glyphs.push(glyf.subarray(startOffset, endOffset)); |
|
startOffset = endOffset; |
|
} |
|
return glyphs; |
|
} |
|
|
|
function lookupCmap(ranges, unicode) { |
|
var code = unicode.charCodeAt(0); |
|
var l = 0, r = ranges.length - 1; |
|
while (l < r) { |
|
var c = (l + r + 1) >> 1; |
|
if (code < ranges[c].start) { |
|
r = c - 1; |
|
} else { |
|
l = c; |
|
} |
|
} |
|
if (ranges[l].start <= code && code <= ranges[l].end) { |
|
return (ranges[l].idDelta + (ranges[l].ids ? |
|
ranges[l].ids[code - ranges[l].start] : code)) & 0xFFFF; |
|
} |
|
return 0; |
|
} |
|
|
|
function compileGlyf(code, cmds, font) { |
|
function moveTo(x, y) { |
|
cmds.push({cmd: 'moveTo', args: [x, y]}); |
|
} |
|
function lineTo(x, y) { |
|
cmds.push({cmd: 'lineTo', args: [x, y]}); |
|
} |
|
function quadraticCurveTo(xa, ya, x, y) { |
|
cmds.push({cmd: 'quadraticCurveTo', args: [xa, ya, x, y]}); |
|
} |
|
|
|
var i = 0; |
|
var numberOfContours = ((code[i] << 24) | (code[i + 1] << 16)) >> 16; |
|
var flags; |
|
var x = 0, y = 0; |
|
i += 10; |
|
if (numberOfContours < 0) { |
|
// composite glyph |
|
do { |
|
flags = (code[i] << 8) | code[i + 1]; |
|
var glyphIndex = (code[i + 2] << 8) | code[i + 3]; |
|
i += 4; |
|
var arg1, arg2; |
|
if ((flags & 0x01)) { |
|
arg1 = ((code[i] << 24) | (code[i + 1] << 16)) >> 16; |
|
arg2 = ((code[i + 2] << 24) | (code[i + 3] << 16)) >> 16; |
|
i += 4; |
|
} else { |
|
arg1 = code[i++]; arg2 = code[i++]; |
|
} |
|
if ((flags & 0x02)) { |
|
x = arg1; |
|
y = arg2; |
|
} else { |
|
x = 0; y = 0; // TODO "they are points" ? |
|
} |
|
var scaleX = 1, scaleY = 1, scale01 = 0, scale10 = 0; |
|
if ((flags & 0x08)) { |
|
scaleX = |
|
scaleY = ((code[i] << 24) | (code[i + 1] << 16)) / 1073741824; |
|
i += 2; |
|
} else if ((flags & 0x40)) { |
|
scaleX = ((code[i] << 24) | (code[i + 1] << 16)) / 1073741824; |
|
scaleY = ((code[i + 2] << 24) | (code[i + 3] << 16)) / 1073741824; |
|
i += 4; |
|
} else if ((flags & 0x80)) { |
|
scaleX = ((code[i] << 24) | (code[i + 1] << 16)) / 1073741824; |
|
scale01 = ((code[i + 2] << 24) | (code[i + 3] << 16)) / 1073741824; |
|
scale10 = ((code[i + 4] << 24) | (code[i + 5] << 16)) / 1073741824; |
|
scaleY = ((code[i + 6] << 24) | (code[i + 7] << 16)) / 1073741824; |
|
i += 8; |
|
} |
|
var subglyph = font.glyphs[glyphIndex]; |
|
if (subglyph) { |
|
cmds.push({cmd: 'save'}); |
|
cmds.push({cmd: 'transform', |
|
args: [scaleX, scale01, scale10, scaleY, x, y]}); |
|
compileGlyf(subglyph, cmds, font); |
|
cmds.push({cmd: 'restore'}); |
|
} |
|
} while ((flags & 0x20)); |
|
} else { |
|
// simple glyph |
|
var endPtsOfContours = []; |
|
var j, jj; |
|
for (j = 0; j < numberOfContours; j++) { |
|
endPtsOfContours.push((code[i] << 8) | code[i + 1]); |
|
i += 2; |
|
} |
|
var instructionLength = (code[i] << 8) | code[i + 1]; |
|
i += 2 + instructionLength; // skipping the instructions |
|
var numberOfPoints = endPtsOfContours[endPtsOfContours.length - 1] + 1; |
|
var points = []; |
|
while (points.length < numberOfPoints) { |
|
flags = code[i++]; |
|
var repeat = 1; |
|
if ((flags & 0x08)) { |
|
repeat += code[i++]; |
|
} |
|
while (repeat-- > 0) { |
|
points.push({flags: flags}); |
|
} |
|
} |
|
for (j = 0; j < numberOfPoints; j++) { |
|
switch (points[j].flags & 0x12) { |
|
case 0x00: |
|
x += ((code[i] << 24) | (code[i + 1] << 16)) >> 16; |
|
i += 2; |
|
break; |
|
case 0x02: |
|
x -= code[i++]; |
|
break; |
|
case 0x12: |
|
x += code[i++]; |
|
break; |
|
} |
|
points[j].x = x; |
|
} |
|
for (j = 0; j < numberOfPoints; j++) { |
|
switch (points[j].flags & 0x24) { |
|
case 0x00: |
|
y += ((code[i] << 24) | (code[i + 1] << 16)) >> 16; |
|
i += 2; |
|
break; |
|
case 0x04: |
|
y -= code[i++]; |
|
break; |
|
case 0x24: |
|
y += code[i++]; |
|
break; |
|
} |
|
points[j].y = y; |
|
} |
|
|
|
var startPoint = 0; |
|
for (i = 0; i < numberOfContours; i++) { |
|
var endPoint = endPtsOfContours[i]; |
|
// contours might have implicit points, which is located in the middle |
|
// between two neighboring off-curve points |
|
var contour = points.slice(startPoint, endPoint + 1); |
|
if ((contour[0].flags & 1)) { |
|
contour.push(contour[0]); // using start point at the contour end |
|
} else if ((contour[contour.length - 1].flags & 1)) { |
|
// first is off-curve point, trying to use one from the end |
|
contour.unshift(contour[contour.length - 1]); |
|
} else { |
|
// start and end are off-curve points, creating implicit one |
|
var p = { |
|
flags: 1, |
|
x: (contour[0].x + contour[contour.length - 1].x) / 2, |
|
y: (contour[0].y + contour[contour.length - 1].y) / 2 |
|
}; |
|
contour.unshift(p); |
|
contour.push(p); |
|
} |
|
moveTo(contour[0].x, contour[0].y); |
|
for (j = 1, jj = contour.length; j < jj; j++) { |
|
if ((contour[j].flags & 1)) { |
|
lineTo(contour[j].x, contour[j].y); |
|
} else if ((contour[j + 1].flags & 1)){ |
|
quadraticCurveTo(contour[j].x, contour[j].y, |
|
contour[j + 1].x, contour[j + 1].y); |
|
j++; |
|
} else { |
|
quadraticCurveTo(contour[j].x, contour[j].y, |
|
(contour[j].x + contour[j + 1].x) / 2, |
|
(contour[j].y + contour[j + 1].y) / 2); |
|
} |
|
} |
|
startPoint = endPoint + 1; |
|
} |
|
} |
|
} |
|
|
|
function compileCharString(code, cmds, font) { |
|
var stack = []; |
|
var x = 0, y = 0; |
|
var stems = 0; |
|
|
|
function moveTo(x, y) { |
|
cmds.push({cmd: 'moveTo', args: [x, y]}); |
|
} |
|
function lineTo(x, y) { |
|
cmds.push({cmd: 'lineTo', args: [x, y]}); |
|
} |
|
function bezierCurveTo(x1, y1, x2, y2, x, y) { |
|
cmds.push({cmd: 'bezierCurveTo', args: [x1, y1, x2, y2, x, y]}); |
|
} |
|
|
|
function parse(code) { |
|
var i = 0; |
|
while (i < code.length) { |
|
var stackClean = false; |
|
var v = code[i++]; |
|
var xa, xb, ya, yb, y1, y2, y3, n, subrCode; |
|
switch (v) { |
|
case 1: // hstem |
|
stems += stack.length >> 1; |
|
stackClean = true; |
|
break; |
|
case 3: // vstem |
|
stems += stack.length >> 1; |
|
stackClean = true; |
|
break; |
|
case 4: // vmoveto |
|
y += stack.pop(); |
|
moveTo(x, y); |
|
stackClean = true; |
|
break; |
|
case 5: // rlineto |
|
while (stack.length > 0) { |
|
x += stack.shift(); |
|
y += stack.shift(); |
|
lineTo(x, y); |
|
} |
|
break; |
|
case 6: // hlineto |
|
while (stack.length > 0) { |
|
x += stack.shift(); |
|
lineTo(x, y); |
|
if (stack.length === 0) { |
|
break; |
|
} |
|
y += stack.shift(); |
|
lineTo(x, y); |
|
} |
|
break; |
|
case 7: // vlineto |
|
while (stack.length > 0) { |
|
y += stack.shift(); |
|
lineTo(x, y); |
|
if (stack.length === 0) { |
|
break; |
|
} |
|
x += stack.shift(); |
|
lineTo(x, y); |
|
} |
|
break; |
|
case 8: // rrcurveto |
|
while (stack.length > 0) { |
|
xa = x + stack.shift(); ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb + stack.shift(); y = yb + stack.shift(); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
} |
|
break; |
|
case 10: // callsubr |
|
n = stack.pop() + font.subrsBias; |
|
subrCode = font.subrs[n]; |
|
if (subrCode) { |
|
parse(subrCode); |
|
} |
|
break; |
|
case 11: // return |
|
return; |
|
case 12: |
|
v = code[i++]; |
|
switch (v) { |
|
case 34: // flex |
|
xa = x + stack.shift(); |
|
xb = xa + stack.shift(); y1 = y + stack.shift(); |
|
x = xb + stack.shift(); |
|
bezierCurveTo(xa, y, xb, y1, x, y1); |
|
xa = x + stack.shift(); |
|
xb = xa + stack.shift(); |
|
x = xb + stack.shift(); |
|
bezierCurveTo(xa, y1, xb, y, x, y); |
|
break; |
|
case 35: // flex |
|
xa = x + stack.shift(); ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb + stack.shift(); y = yb + stack.shift(); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
xa = x + stack.shift(); ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb + stack.shift(); y = yb + stack.shift(); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
stack.pop(); // fd |
|
break; |
|
case 36: // hflex1 |
|
xa = x + stack.shift(); y1 = y + stack.shift(); |
|
xb = xa + stack.shift(); y2 = y1 + stack.shift(); |
|
x = xb + stack.shift(); |
|
bezierCurveTo(xa, y1, xb, y2, x, y2); |
|
xa = x + stack.shift(); |
|
xb = xa + stack.shift(); y3 = y2 + stack.shift(); |
|
x = xb + stack.shift(); |
|
bezierCurveTo(xa, y2, xb, y3, x, y); |
|
break; |
|
case 37: // flex1 |
|
var x0 = x, y0 = y; |
|
xa = x + stack.shift(); ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb + stack.shift(); y = yb + stack.shift(); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
xa = x + stack.shift(); ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb; y = yb; |
|
if (Math.abs(x - x0) > Math.abs(y - y0)) { |
|
x += stack.shift(); |
|
} else { |
|
y += stack.shift(); |
|
} |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
break; |
|
default: |
|
error('unknown operator: 12 ' + v); |
|
} |
|
break; |
|
case 14: // endchar |
|
if (stack.length >= 4) { |
|
var achar = stack.pop(); |
|
var bchar = stack.pop(); |
|
y = stack.pop(); |
|
x = stack.pop(); |
|
cmds.push({cmd: 'save'}); |
|
cmds.push({cmd: 'translate', args: [x, y]}); |
|
var gid = lookupCmap(font.cmap, String.fromCharCode( |
|
font.glyphNameMap[Encodings.StandardEncoding[achar]])); |
|
compileCharString(font.glyphs[gid], cmds, font); |
|
cmds.push({cmd: 'restore'}); |
|
|
|
gid = lookupCmap(font.cmap, String.fromCharCode( |
|
font.glyphNameMap[Encodings.StandardEncoding[bchar]])); |
|
compileCharString(font.glyphs[gid], cmds, font); |
|
} |
|
return; |
|
case 18: // hstemhm |
|
stems += stack.length >> 1; |
|
stackClean = true; |
|
break; |
|
case 19: // hintmask |
|
stems += stack.length >> 1; |
|
i += (stems + 7) >> 3; |
|
stackClean = true; |
|
break; |
|
case 20: // cntrmask |
|
stems += stack.length >> 1; |
|
i += (stems + 7) >> 3; |
|
stackClean = true; |
|
break; |
|
case 21: // rmoveto |
|
y += stack.pop(); |
|
x += stack.pop(); |
|
moveTo(x, y); |
|
stackClean = true; |
|
break; |
|
case 22: // hmoveto |
|
x += stack.pop(); |
|
moveTo(x, y); |
|
stackClean = true; |
|
break; |
|
case 23: // vstemhm |
|
stems += stack.length >> 1; |
|
stackClean = true; |
|
break; |
|
case 24: // rcurveline |
|
while (stack.length > 2) { |
|
xa = x + stack.shift(); ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb + stack.shift(); y = yb + stack.shift(); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
} |
|
x += stack.shift(); |
|
y += stack.shift(); |
|
lineTo(x, y); |
|
break; |
|
case 25: // rlinecurve |
|
while (stack.length > 6) { |
|
x += stack.shift(); |
|
y += stack.shift(); |
|
lineTo(x, y); |
|
} |
|
xa = x + stack.shift(); ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb + stack.shift(); y = yb + stack.shift(); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
break; |
|
case 26: // vvcurveto |
|
if (stack.length % 2) { |
|
x += stack.shift(); |
|
} |
|
while (stack.length > 0) { |
|
xa = x; ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb; y = yb + stack.shift(); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
} |
|
break; |
|
case 27: // hhcurveto |
|
if (stack.length % 2) { |
|
y += stack.shift(); |
|
} |
|
while (stack.length > 0) { |
|
xa = x + stack.shift(); ya = y; |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb + stack.shift(); y = yb; |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
} |
|
break; |
|
case 28: |
|
stack.push(((code[i] << 24) | (code[i + 1] << 16)) >> 16); |
|
i += 2; |
|
break; |
|
case 29: // callgsubr |
|
n = stack.pop() + font.gsubrsBias; |
|
subrCode = font.gsubrs[n]; |
|
if (subrCode) { |
|
parse(subrCode); |
|
} |
|
break; |
|
case 30: // vhcurveto |
|
while (stack.length > 0) { |
|
xa = x; ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb + stack.shift(); |
|
y = yb + (stack.length === 1 ? stack.shift() : 0); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
if (stack.length === 0) { |
|
break; |
|
} |
|
|
|
xa = x + stack.shift(); ya = y; |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
y = yb + stack.shift(); |
|
x = xb + (stack.length === 1 ? stack.shift() : 0); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
} |
|
break; |
|
case 31: // hvcurveto |
|
while (stack.length > 0) { |
|
xa = x + stack.shift(); ya = y; |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
y = yb + stack.shift(); |
|
x = xb + (stack.length === 1 ? stack.shift() : 0); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
if (stack.length === 0) { |
|
break; |
|
} |
|
|
|
xa = x; ya = y + stack.shift(); |
|
xb = xa + stack.shift(); yb = ya + stack.shift(); |
|
x = xb + stack.shift(); |
|
y = yb + (stack.length === 1 ? stack.shift() : 0); |
|
bezierCurveTo(xa, ya, xb, yb, x, y); |
|
} |
|
break; |
|
default: |
|
if (v < 32) { |
|
error('unknown operator: ' + v); |
|
} |
|
if (v < 247) { |
|
stack.push(v - 139); |
|
} else if (v < 251) { |
|
stack.push((v - 247) * 256 + code[i++] + 108); |
|
} else if (v < 255) { |
|
stack.push(-(v - 251) * 256 - code[i++] - 108); |
|
} else { |
|
stack.push(((code[i] << 24) | (code[i + 1] << 16) | |
|
(code[i + 2] << 8) | code[i + 3]) / 65536); |
|
i += 4; |
|
} |
|
break; |
|
} |
|
if (stackClean) { |
|
stack.length = 0; |
|
} |
|
} |
|
} |
|
parse(code); |
|
} |
|
|
|
var noop = ''; |
|
|
|
function CompiledFont(fontMatrix) { |
|
this.compiledGlyphs = {}; |
|
this.fontMatrix = fontMatrix; |
|
} |
|
CompiledFont.prototype = { |
|
getPathJs: function (unicode) { |
|
var gid = lookupCmap(this.cmap, unicode); |
|
var fn = this.compiledGlyphs[gid]; |
|
if (!fn) { |
|
this.compiledGlyphs[gid] = fn = this.compileGlyph(this.glyphs[gid]); |
|
} |
|
return fn; |
|
}, |
|
|
|
compileGlyph: function (code) { |
|
if (!code || code.length === 0 || code[0] === 14) { |
|
return noop; |
|
} |
|
|
|
var cmds = []; |
|
cmds.push({cmd: 'save'}); |
|
cmds.push({cmd: 'transform', args: this.fontMatrix.slice()}); |
|
cmds.push({cmd: 'scale', args: ['size', '-size']}); |
|
|
|
this.compileGlyphImpl(code, cmds); |
|
|
|
cmds.push({cmd: 'restore'}); |
|
|
|
return cmds; |
|
}, |
|
|
|
compileGlyphImpl: function () { |
|
error('Children classes should implement this.'); |
|
}, |
|
|
|
hasBuiltPath: function (unicode) { |
|
var gid = lookupCmap(this.cmap, unicode); |
|
return gid in this.compiledGlyphs; |
|
} |
|
}; |
|
|
|
function TrueTypeCompiled(glyphs, cmap, fontMatrix) { |
|
fontMatrix = fontMatrix || [0.000488, 0, 0, 0.000488, 0, 0]; |
|
CompiledFont.call(this, fontMatrix); |
|
|
|
this.glyphs = glyphs; |
|
this.cmap = cmap; |
|
|
|
this.compiledGlyphs = []; |
|
} |
|
|
|
Util.inherit(TrueTypeCompiled, CompiledFont, { |
|
compileGlyphImpl: function (code, cmds) { |
|
compileGlyf(code, cmds, this); |
|
} |
|
}); |
|
|
|
function Type2Compiled(cffInfo, cmap, fontMatrix, glyphNameMap) { |
|
fontMatrix = fontMatrix || [0.001, 0, 0, 0.001, 0, 0]; |
|
CompiledFont.call(this, fontMatrix); |
|
this.glyphs = cffInfo.glyphs; |
|
this.gsubrs = cffInfo.gsubrs || []; |
|
this.subrs = cffInfo.subrs || []; |
|
this.cmap = cmap; |
|
this.glyphNameMap = glyphNameMap || GlyphsUnicode; |
|
|
|
this.compiledGlyphs = []; |
|
this.gsubrsBias = (this.gsubrs.length < 1240 ? |
|
107 : (this.gsubrs.length < 33900 ? 1131 : 32768)); |
|
this.subrsBias = (this.subrs.length < 1240 ? |
|
107 : (this.subrs.length < 33900 ? 1131 : 32768)); |
|
} |
|
|
|
Util.inherit(Type2Compiled, CompiledFont, { |
|
compileGlyphImpl: function (code, cmds) { |
|
compileCharString(code, cmds, this); |
|
} |
|
}); |
|
|
|
|
|
return { |
|
create: function FontRendererFactory_create(font) { |
|
var data = new Uint8Array(font.data); |
|
var cmap, glyf, loca, cff, indexToLocFormat, unitsPerEm; |
|
var numTables = getUshort(data, 4); |
|
for (var i = 0, p = 12; i < numTables; i++, p += 16) { |
|
var tag = bytesToString(data.subarray(p, p + 4)); |
|
var offset = getLong(data, p + 8); |
|
var length = getLong(data, p + 12); |
|
switch (tag) { |
|
case 'cmap': |
|
cmap = parseCmap(data, offset, offset + length); |
|
break; |
|
case 'glyf': |
|
glyf = data.subarray(offset, offset + length); |
|
break; |
|
case 'loca': |
|
loca = data.subarray(offset, offset + length); |
|
break; |
|
case 'head': |
|
unitsPerEm = getUshort(data, offset + 18); |
|
indexToLocFormat = getUshort(data, offset + 50); |
|
break; |
|
case 'CFF ': |
|
cff = parseCff(data, offset, offset + length); |
|
break; |
|
} |
|
} |
|
|
|
if (glyf) { |
|
var fontMatrix = (!unitsPerEm ? font.fontMatrix : |
|
[1 / unitsPerEm, 0, 0, 1 / unitsPerEm, 0, 0]); |
|
return new TrueTypeCompiled( |
|
parseGlyfTable(glyf, loca, indexToLocFormat), cmap, fontMatrix); |
|
} else { |
|
return new Type2Compiled(cff, cmap, font.fontMatrix, font.glyphNameMap); |
|
} |
|
} |
|
}; |
|
})(); |
|
|
|
|
|
// TODO refactor to remove cyclic dependency on fonts.js |
|
function _setCoreFonts(coreFonts_) { |
|
coreFonts = coreFonts_; |
|
Encodings = coreFonts_.Encodings; |
|
CFFParser = coreFonts_.CFFParser; |
|
} |
|
exports._setCoreFonts = _setCoreFonts; |
|
|
|
exports.FontRendererFactory = FontRendererFactory; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreParser = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreStream); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreStream) { |
|
|
|
var MissingDataException = sharedUtil.MissingDataException; |
|
var StreamType = sharedUtil.StreamType; |
|
var assert = sharedUtil.assert; |
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isArray = sharedUtil.isArray; |
|
var isInt = sharedUtil.isInt; |
|
var isNum = sharedUtil.isNum; |
|
var isString = sharedUtil.isString; |
|
var warn = sharedUtil.warn; |
|
var Cmd = corePrimitives.Cmd; |
|
var Dict = corePrimitives.Dict; |
|
var Name = corePrimitives.Name; |
|
var Ref = corePrimitives.Ref; |
|
var isCmd = corePrimitives.isCmd; |
|
var isDict = corePrimitives.isDict; |
|
var isName = corePrimitives.isName; |
|
var Ascii85Stream = coreStream.Ascii85Stream; |
|
var AsciiHexStream = coreStream.AsciiHexStream; |
|
var CCITTFaxStream = coreStream.CCITTFaxStream; |
|
var FlateStream = coreStream.FlateStream; |
|
var Jbig2Stream = coreStream.Jbig2Stream; |
|
var JpegStream = coreStream.JpegStream; |
|
var JpxStream = coreStream.JpxStream; |
|
var LZWStream = coreStream.LZWStream; |
|
var NullStream = coreStream.NullStream; |
|
var PredictorStream = coreStream.PredictorStream; |
|
var RunLengthStream = coreStream.RunLengthStream; |
|
|
|
var EOF = {}; |
|
|
|
function isEOF(v) { |
|
return (v === EOF); |
|
} |
|
|
|
var MAX_LENGTH_TO_CACHE = 1000; |
|
|
|
var Parser = (function ParserClosure() { |
|
function Parser(lexer, allowStreams, xref) { |
|
this.lexer = lexer; |
|
this.allowStreams = allowStreams; |
|
this.xref = xref; |
|
this.imageCache = {}; |
|
this.refill(); |
|
} |
|
|
|
Parser.prototype = { |
|
refill: function Parser_refill() { |
|
this.buf1 = this.lexer.getObj(); |
|
this.buf2 = this.lexer.getObj(); |
|
}, |
|
shift: function Parser_shift() { |
|
if (isCmd(this.buf2, 'ID')) { |
|
this.buf1 = this.buf2; |
|
this.buf2 = null; |
|
} else { |
|
this.buf1 = this.buf2; |
|
this.buf2 = this.lexer.getObj(); |
|
} |
|
}, |
|
tryShift: function Parser_tryShift() { |
|
try { |
|
this.shift(); |
|
return true; |
|
} catch (e) { |
|
if (e instanceof MissingDataException) { |
|
throw e; |
|
} |
|
// Upon failure, the caller should reset this.lexer.pos to a known good |
|
// state and call this.shift() twice to reset the buffers. |
|
return false; |
|
} |
|
}, |
|
getObj: function Parser_getObj(cipherTransform) { |
|
var buf1 = this.buf1; |
|
this.shift(); |
|
|
|
if (buf1 instanceof Cmd) { |
|
switch (buf1.cmd) { |
|
case 'BI': // inline image |
|
return this.makeInlineImage(cipherTransform); |
|
case '[': // array |
|
var array = []; |
|
while (!isCmd(this.buf1, ']') && !isEOF(this.buf1)) { |
|
array.push(this.getObj(cipherTransform)); |
|
} |
|
if (isEOF(this.buf1)) { |
|
error('End of file inside array'); |
|
} |
|
this.shift(); |
|
return array; |
|
case '<<': // dictionary or stream |
|
var dict = new Dict(this.xref); |
|
while (!isCmd(this.buf1, '>>') && !isEOF(this.buf1)) { |
|
if (!isName(this.buf1)) { |
|
info('Malformed dictionary: key must be a name object'); |
|
this.shift(); |
|
continue; |
|
} |
|
|
|
var key = this.buf1.name; |
|
this.shift(); |
|
if (isEOF(this.buf1)) { |
|
break; |
|
} |
|
dict.set(key, this.getObj(cipherTransform)); |
|
} |
|
if (isEOF(this.buf1)) { |
|
error('End of file inside dictionary'); |
|
} |
|
|
|
// Stream objects are not allowed inside content streams or |
|
// object streams. |
|
if (isCmd(this.buf2, 'stream')) { |
|
return (this.allowStreams ? |
|
this.makeStream(dict, cipherTransform) : dict); |
|
} |
|
this.shift(); |
|
return dict; |
|
default: // simple object |
|
return buf1; |
|
} |
|
} |
|
|
|
if (isInt(buf1)) { // indirect reference or integer |
|
var num = buf1; |
|
if (isInt(this.buf1) && isCmd(this.buf2, 'R')) { |
|
var ref = new Ref(num, this.buf1); |
|
this.shift(); |
|
this.shift(); |
|
return ref; |
|
} |
|
return num; |
|
} |
|
|
|
if (isString(buf1)) { // string |
|
var str = buf1; |
|
if (cipherTransform) { |
|
str = cipherTransform.decryptString(str); |
|
} |
|
return str; |
|
} |
|
|
|
// simple object |
|
return buf1; |
|
}, |
|
/** |
|
* Find the end of the stream by searching for the /EI\s/. |
|
* @returns {number} The inline stream length. |
|
*/ |
|
findDefaultInlineStreamEnd: |
|
function Parser_findDefaultInlineStreamEnd(stream) { |
|
var E = 0x45, I = 0x49, SPACE = 0x20, LF = 0xA, CR = 0xD; |
|
var startPos = stream.pos, state = 0, ch, i, n, followingBytes; |
|
while ((ch = stream.getByte()) !== -1) { |
|
if (state === 0) { |
|
state = (ch === E) ? 1 : 0; |
|
} else if (state === 1) { |
|
state = (ch === I) ? 2 : 0; |
|
} else { |
|
assert(state === 2); |
|
if (ch === SPACE || ch === LF || ch === CR) { |
|
// Let's check the next five bytes are ASCII... just be sure. |
|
n = 5; |
|
followingBytes = stream.peekBytes(n); |
|
for (i = 0; i < n; i++) { |
|
ch = followingBytes[i]; |
|
if (ch !== LF && ch !== CR && (ch < SPACE || ch > 0x7F)) { |
|
// Not a LF, CR, SPACE or any visible ASCII character, i.e. |
|
// it's binary stuff. Resetting the state. |
|
state = 0; |
|
break; |
|
} |
|
} |
|
if (state === 2) { |
|
break; // Finished! |
|
} |
|
} else { |
|
state = 0; |
|
} |
|
} |
|
} |
|
return ((stream.pos - 4) - startPos); |
|
}, |
|
/** |
|
* Find the EOI (end-of-image) marker 0xFFD9 of the stream. |
|
* @returns {number} The inline stream length. |
|
*/ |
|
findDCTDecodeInlineStreamEnd: |
|
function Parser_findDCTDecodeInlineStreamEnd(stream) { |
|
var startPos = stream.pos, foundEOI = false, b, markerLength, length; |
|
while ((b = stream.getByte()) !== -1) { |
|
if (b !== 0xFF) { // Not a valid marker. |
|
continue; |
|
} |
|
switch (stream.getByte()) { |
|
case 0x00: // Byte stuffing. |
|
// 0xFF00 appears to be a very common byte sequence in JPEG images. |
|
break; |
|
|
|
case 0xFF: // Fill byte. |
|
// Avoid skipping a valid marker, resetting the stream position. |
|
stream.skip(-1); |
|
break; |
|
|
|
case 0xD9: // EOI |
|
foundEOI = true; |
|
break; |
|
|
|
case 0xC0: // SOF0 |
|
case 0xC1: // SOF1 |
|
case 0xC2: // SOF2 |
|
case 0xC3: // SOF3 |
|
|
|
case 0xC5: // SOF5 |
|
case 0xC6: // SOF6 |
|
case 0xC7: // SOF7 |
|
|
|
case 0xC9: // SOF9 |
|
case 0xCA: // SOF10 |
|
case 0xCB: // SOF11 |
|
|
|
case 0xCD: // SOF13 |
|
case 0xCE: // SOF14 |
|
case 0xCF: // SOF15 |
|
|
|
case 0xC4: // DHT |
|
case 0xCC: // DAC |
|
|
|
case 0xDA: // SOS |
|
case 0xDB: // DQT |
|
case 0xDC: // DNL |
|
case 0xDD: // DRI |
|
case 0xDE: // DHP |
|
case 0xDF: // EXP |
|
|
|
case 0xE0: // APP0 |
|
case 0xE1: // APP1 |
|
case 0xE2: // APP2 |
|
case 0xE3: // APP3 |
|
case 0xE4: // APP4 |
|
case 0xE5: // APP5 |
|
case 0xE6: // APP6 |
|
case 0xE7: // APP7 |
|
case 0xE8: // APP8 |
|
case 0xE9: // APP9 |
|
case 0xEA: // APP10 |
|
case 0xEB: // APP11 |
|
case 0xEC: // APP12 |
|
case 0xED: // APP13 |
|
case 0xEE: // APP14 |
|
case 0xEF: // APP15 |
|
|
|
case 0xFE: // COM |
|
// The marker should be followed by the length of the segment. |
|
markerLength = stream.getUint16(); |
|
if (markerLength > 2) { |
|
// |markerLength| contains the byte length of the marker segment, |
|
// including its own length (2 bytes) and excluding the marker. |
|
stream.skip(markerLength - 2); // Jump to the next marker. |
|
} else { |
|
// The marker length is invalid, resetting the stream position. |
|
stream.skip(-2); |
|
} |
|
break; |
|
} |
|
if (foundEOI) { |
|
break; |
|
} |
|
} |
|
length = stream.pos - startPos; |
|
if (b === -1) { |
|
warn('Inline DCTDecode image stream: ' + |
|
'EOI marker not found, searching for /EI/ instead.'); |
|
stream.skip(-length); // Reset the stream position. |
|
return this.findDefaultInlineStreamEnd(stream); |
|
} |
|
this.inlineStreamSkipEI(stream); |
|
return length; |
|
}, |
|
/** |
|
* Find the EOD (end-of-data) marker '~>' (i.e. TILDE + GT) of the stream. |
|
* @returns {number} The inline stream length. |
|
*/ |
|
findASCII85DecodeInlineStreamEnd: |
|
function Parser_findASCII85DecodeInlineStreamEnd(stream) { |
|
var TILDE = 0x7E, GT = 0x3E; |
|
var startPos = stream.pos, ch, length; |
|
while ((ch = stream.getByte()) !== -1) { |
|
if (ch === TILDE && stream.peekByte() === GT) { |
|
stream.skip(); |
|
break; |
|
} |
|
} |
|
length = stream.pos - startPos; |
|
if (ch === -1) { |
|
warn('Inline ASCII85Decode image stream: ' + |
|
'EOD marker not found, searching for /EI/ instead.'); |
|
stream.skip(-length); // Reset the stream position. |
|
return this.findDefaultInlineStreamEnd(stream); |
|
} |
|
this.inlineStreamSkipEI(stream); |
|
return length; |
|
}, |
|
/** |
|
* Find the EOD (end-of-data) marker '>' (i.e. GT) of the stream. |
|
* @returns {number} The inline stream length. |
|
*/ |
|
findASCIIHexDecodeInlineStreamEnd: |
|
function Parser_findASCIIHexDecodeInlineStreamEnd(stream) { |
|
var GT = 0x3E; |
|
var startPos = stream.pos, ch, length; |
|
while ((ch = stream.getByte()) !== -1) { |
|
if (ch === GT) { |
|
break; |
|
} |
|
} |
|
length = stream.pos - startPos; |
|
if (ch === -1) { |
|
warn('Inline ASCIIHexDecode image stream: ' + |
|
'EOD marker not found, searching for /EI/ instead.'); |
|
stream.skip(-length); // Reset the stream position. |
|
return this.findDefaultInlineStreamEnd(stream); |
|
} |
|
this.inlineStreamSkipEI(stream); |
|
return length; |
|
}, |
|
/** |
|
* Skip over the /EI/ for streams where we search for an EOD marker. |
|
*/ |
|
inlineStreamSkipEI: function Parser_inlineStreamSkipEI(stream) { |
|
var E = 0x45, I = 0x49; |
|
var state = 0, ch; |
|
while ((ch = stream.getByte()) !== -1) { |
|
if (state === 0) { |
|
state = (ch === E) ? 1 : 0; |
|
} else if (state === 1) { |
|
state = (ch === I) ? 2 : 0; |
|
} else if (state === 2) { |
|
break; |
|
} |
|
} |
|
}, |
|
makeInlineImage: function Parser_makeInlineImage(cipherTransform) { |
|
var lexer = this.lexer; |
|
var stream = lexer.stream; |
|
|
|
// Parse dictionary. |
|
var dict = new Dict(this.xref); |
|
while (!isCmd(this.buf1, 'ID') && !isEOF(this.buf1)) { |
|
if (!isName(this.buf1)) { |
|
error('Dictionary key must be a name object'); |
|
} |
|
var key = this.buf1.name; |
|
this.shift(); |
|
if (isEOF(this.buf1)) { |
|
break; |
|
} |
|
dict.set(key, this.getObj(cipherTransform)); |
|
} |
|
|
|
// Extract the name of the first (i.e. the current) image filter. |
|
var filter = dict.get('Filter', 'F'), filterName; |
|
if (isName(filter)) { |
|
filterName = filter.name; |
|
} else if (isArray(filter) && isName(filter[0])) { |
|
filterName = filter[0].name; |
|
} |
|
|
|
// Parse image stream. |
|
var startPos = stream.pos, length, i, ii; |
|
if (filterName === 'DCTDecode' || filterName === 'DCT') { |
|
length = this.findDCTDecodeInlineStreamEnd(stream); |
|
} else if (filterName === 'ASCII85Decide' || filterName === 'A85') { |
|
length = this.findASCII85DecodeInlineStreamEnd(stream); |
|
} else if (filterName === 'ASCIIHexDecode' || filterName === 'AHx') { |
|
length = this.findASCIIHexDecodeInlineStreamEnd(stream); |
|
} else { |
|
length = this.findDefaultInlineStreamEnd(stream); |
|
} |
|
var imageStream = stream.makeSubStream(startPos, length, dict); |
|
|
|
// Cache all images below the MAX_LENGTH_TO_CACHE threshold by their |
|
// adler32 checksum. |
|
var adler32; |
|
if (length < MAX_LENGTH_TO_CACHE) { |
|
var imageBytes = imageStream.getBytes(); |
|
imageStream.reset(); |
|
|
|
var a = 1; |
|
var b = 0; |
|
for (i = 0, ii = imageBytes.length; i < ii; ++i) { |
|
// No modulo required in the loop if imageBytes.length < 5552. |
|
a += imageBytes[i] & 0xff; |
|
b += a; |
|
} |
|
adler32 = ((b % 65521) << 16) | (a % 65521); |
|
|
|
if (this.imageCache.adler32 === adler32) { |
|
this.buf2 = Cmd.get('EI'); |
|
this.shift(); |
|
|
|
this.imageCache[adler32].reset(); |
|
return this.imageCache[adler32]; |
|
} |
|
} |
|
|
|
if (cipherTransform) { |
|
imageStream = cipherTransform.createStream(imageStream, length); |
|
} |
|
|
|
imageStream = this.filter(imageStream, dict, length); |
|
imageStream.dict = dict; |
|
if (adler32 !== undefined) { |
|
imageStream.cacheKey = 'inline_' + length + '_' + adler32; |
|
this.imageCache[adler32] = imageStream; |
|
} |
|
|
|
this.buf2 = Cmd.get('EI'); |
|
this.shift(); |
|
|
|
return imageStream; |
|
}, |
|
makeStream: function Parser_makeStream(dict, cipherTransform) { |
|
var lexer = this.lexer; |
|
var stream = lexer.stream; |
|
|
|
// get stream start position |
|
lexer.skipToNextLine(); |
|
var pos = stream.pos - 1; |
|
|
|
// get length |
|
var length = dict.get('Length'); |
|
if (!isInt(length)) { |
|
info('Bad ' + length + ' attribute in stream'); |
|
length = 0; |
|
} |
|
|
|
// skip over the stream data |
|
stream.pos = pos + length; |
|
lexer.nextChar(); |
|
|
|
// Shift '>>' and check whether the new object marks the end of the stream |
|
if (this.tryShift() && isCmd(this.buf2, 'endstream')) { |
|
this.shift(); // 'stream' |
|
} else { |
|
// bad stream length, scanning for endstream |
|
stream.pos = pos; |
|
var SCAN_BLOCK_SIZE = 2048; |
|
var ENDSTREAM_SIGNATURE_LENGTH = 9; |
|
var ENDSTREAM_SIGNATURE = [0x65, 0x6E, 0x64, 0x73, 0x74, 0x72, 0x65, |
|
0x61, 0x6D]; |
|
var skipped = 0, found = false, i, j; |
|
while (stream.pos < stream.end) { |
|
var scanBytes = stream.peekBytes(SCAN_BLOCK_SIZE); |
|
var scanLength = scanBytes.length - ENDSTREAM_SIGNATURE_LENGTH; |
|
if (scanLength <= 0) { |
|
break; |
|
} |
|
found = false; |
|
i = 0; |
|
while (i < scanLength) { |
|
j = 0; |
|
while (j < ENDSTREAM_SIGNATURE_LENGTH && |
|
scanBytes[i + j] === ENDSTREAM_SIGNATURE[j]) { |
|
j++; |
|
} |
|
if (j >= ENDSTREAM_SIGNATURE_LENGTH) { |
|
found = true; |
|
break; |
|
} |
|
i++; |
|
} |
|
if (found) { |
|
skipped += i; |
|
stream.pos += i; |
|
break; |
|
} |
|
skipped += scanLength; |
|
stream.pos += scanLength; |
|
} |
|
if (!found) { |
|
error('Missing endstream'); |
|
} |
|
length = skipped; |
|
|
|
lexer.nextChar(); |
|
this.shift(); |
|
this.shift(); |
|
} |
|
this.shift(); // 'endstream' |
|
|
|
stream = stream.makeSubStream(pos, length, dict); |
|
if (cipherTransform) { |
|
stream = cipherTransform.createStream(stream, length); |
|
} |
|
stream = this.filter(stream, dict, length); |
|
stream.dict = dict; |
|
return stream; |
|
}, |
|
filter: function Parser_filter(stream, dict, length) { |
|
var filter = dict.get('Filter', 'F'); |
|
var params = dict.get('DecodeParms', 'DP'); |
|
if (isName(filter)) { |
|
return this.makeFilter(stream, filter.name, length, params); |
|
} |
|
|
|
var maybeLength = length; |
|
if (isArray(filter)) { |
|
var filterArray = filter; |
|
var paramsArray = params; |
|
for (var i = 0, ii = filterArray.length; i < ii; ++i) { |
|
filter = filterArray[i]; |
|
if (!isName(filter)) { |
|
error('Bad filter name: ' + filter); |
|
} |
|
|
|
params = null; |
|
if (isArray(paramsArray) && (i in paramsArray)) { |
|
params = paramsArray[i]; |
|
} |
|
stream = this.makeFilter(stream, filter.name, maybeLength, params); |
|
// after the first stream the length variable is invalid |
|
maybeLength = null; |
|
} |
|
} |
|
return stream; |
|
}, |
|
makeFilter: function Parser_makeFilter(stream, name, maybeLength, params) { |
|
if (stream.dict.get('Length') === 0 && !maybeLength) { |
|
warn('Empty "' + name + '" stream.'); |
|
return new NullStream(stream); |
|
} |
|
try { |
|
if (params && this.xref) { |
|
params = this.xref.fetchIfRef(params); |
|
} |
|
var xrefStreamStats = this.xref.stats.streamTypes; |
|
if (name === 'FlateDecode' || name === 'Fl') { |
|
xrefStreamStats[StreamType.FLATE] = true; |
|
if (params) { |
|
return new PredictorStream(new FlateStream(stream, maybeLength), |
|
maybeLength, params); |
|
} |
|
return new FlateStream(stream, maybeLength); |
|
} |
|
if (name === 'LZWDecode' || name === 'LZW') { |
|
xrefStreamStats[StreamType.LZW] = true; |
|
var earlyChange = 1; |
|
if (params) { |
|
if (params.has('EarlyChange')) { |
|
earlyChange = params.get('EarlyChange'); |
|
} |
|
return new PredictorStream( |
|
new LZWStream(stream, maybeLength, earlyChange), |
|
maybeLength, params); |
|
} |
|
return new LZWStream(stream, maybeLength, earlyChange); |
|
} |
|
if (name === 'DCTDecode' || name === 'DCT') { |
|
xrefStreamStats[StreamType.DCT] = true; |
|
return new JpegStream(stream, maybeLength, stream.dict, this.xref); |
|
} |
|
if (name === 'JPXDecode' || name === 'JPX') { |
|
xrefStreamStats[StreamType.JPX] = true; |
|
return new JpxStream(stream, maybeLength, stream.dict); |
|
} |
|
if (name === 'ASCII85Decode' || name === 'A85') { |
|
xrefStreamStats[StreamType.A85] = true; |
|
return new Ascii85Stream(stream, maybeLength); |
|
} |
|
if (name === 'ASCIIHexDecode' || name === 'AHx') { |
|
xrefStreamStats[StreamType.AHX] = true; |
|
return new AsciiHexStream(stream, maybeLength); |
|
} |
|
if (name === 'CCITTFaxDecode' || name === 'CCF') { |
|
xrefStreamStats[StreamType.CCF] = true; |
|
return new CCITTFaxStream(stream, maybeLength, params); |
|
} |
|
if (name === 'RunLengthDecode' || name === 'RL') { |
|
xrefStreamStats[StreamType.RL] = true; |
|
return new RunLengthStream(stream, maybeLength); |
|
} |
|
if (name === 'JBIG2Decode') { |
|
xrefStreamStats[StreamType.JBIG] = true; |
|
return new Jbig2Stream(stream, maybeLength, stream.dict); |
|
} |
|
warn('filter "' + name + '" not supported yet'); |
|
return stream; |
|
} catch (ex) { |
|
if (ex instanceof MissingDataException) { |
|
throw ex; |
|
} |
|
warn('Invalid stream: \"' + ex + '\"'); |
|
return new NullStream(stream); |
|
} |
|
} |
|
}; |
|
|
|
return Parser; |
|
})(); |
|
|
|
var Lexer = (function LexerClosure() { |
|
function Lexer(stream, knownCommands) { |
|
this.stream = stream; |
|
this.nextChar(); |
|
|
|
// While lexing, we build up many strings one char at a time. Using += for |
|
// this can result in lots of garbage strings. It's better to build an |
|
// array of single-char strings and then join() them together at the end. |
|
// And reusing a single array (i.e. |this.strBuf|) over and over for this |
|
// purpose uses less memory than using a new array for each string. |
|
this.strBuf = []; |
|
|
|
// The PDFs might have "glued" commands with other commands, operands or |
|
// literals, e.g. "q1". The knownCommands is a dictionary of the valid |
|
// commands and their prefixes. The prefixes are built the following way: |
|
// if there a command that is a prefix of the other valid command or |
|
// literal (e.g. 'f' and 'false') the following prefixes must be included, |
|
// 'fa', 'fal', 'fals'. The prefixes are not needed, if the command has no |
|
// other commands or literals as a prefix. The knowCommands is optional. |
|
this.knownCommands = knownCommands; |
|
} |
|
|
|
Lexer.isSpace = function Lexer_isSpace(ch) { |
|
// Space is one of the following characters: SPACE, TAB, CR or LF. |
|
return (ch === 0x20 || ch === 0x09 || ch === 0x0D || ch === 0x0A); |
|
}; |
|
|
|
// A '1' in this array means the character is white space. A '1' or |
|
// '2' means the character ends a name or command. |
|
var specialChars = [ |
|
1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, // 0x |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 1x |
|
1, 0, 0, 0, 0, 2, 0, 0, 2, 2, 0, 0, 0, 0, 0, 2, // 2x |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 2, 0, // 3x |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 4x |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 2, 0, 0, // 5x |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 6x |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 2, 0, 0, // 7x |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 8x |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 9x |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // ax |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // bx |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // cx |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // dx |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // ex |
|
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 // fx |
|
]; |
|
|
|
function toHexDigit(ch) { |
|
if (ch >= 0x30 && ch <= 0x39) { // '0'-'9' |
|
return ch & 0x0F; |
|
} |
|
if ((ch >= 0x41 && ch <= 0x46) || (ch >= 0x61 && ch <= 0x66)) { |
|
// 'A'-'F', 'a'-'f' |
|
return (ch & 0x0F) + 9; |
|
} |
|
return -1; |
|
} |
|
|
|
Lexer.prototype = { |
|
nextChar: function Lexer_nextChar() { |
|
return (this.currentChar = this.stream.getByte()); |
|
}, |
|
peekChar: function Lexer_peekChar() { |
|
return this.stream.peekByte(); |
|
}, |
|
getNumber: function Lexer_getNumber() { |
|
var ch = this.currentChar; |
|
var eNotation = false; |
|
var divideBy = 0; // different from 0 if it's a floating point value |
|
var sign = 1; |
|
|
|
if (ch === 0x2D) { // '-' |
|
sign = -1; |
|
ch = this.nextChar(); |
|
|
|
if (ch === 0x2D) { // '-' |
|
// Ignore double negative (this is consistent with Adobe Reader). |
|
ch = this.nextChar(); |
|
} |
|
} else if (ch === 0x2B) { // '+' |
|
ch = this.nextChar(); |
|
} |
|
if (ch === 0x2E) { // '.' |
|
divideBy = 10; |
|
ch = this.nextChar(); |
|
} |
|
if (ch < 0x30 || ch > 0x39) { // '0' - '9' |
|
error('Invalid number: ' + String.fromCharCode(ch)); |
|
return 0; |
|
} |
|
|
|
var baseValue = ch - 0x30; // '0' |
|
var powerValue = 0; |
|
var powerValueSign = 1; |
|
|
|
while ((ch = this.nextChar()) >= 0) { |
|
if (0x30 <= ch && ch <= 0x39) { // '0' - '9' |
|
var currentDigit = ch - 0x30; // '0' |
|
if (eNotation) { // We are after an 'e' or 'E' |
|
powerValue = powerValue * 10 + currentDigit; |
|
} else { |
|
if (divideBy !== 0) { // We are after a point |
|
divideBy *= 10; |
|
} |
|
baseValue = baseValue * 10 + currentDigit; |
|
} |
|
} else if (ch === 0x2E) { // '.' |
|
if (divideBy === 0) { |
|
divideBy = 1; |
|
} else { |
|
// A number can have only one '.' |
|
break; |
|
} |
|
} else if (ch === 0x2D) { // '-' |
|
// ignore minus signs in the middle of numbers to match |
|
// Adobe's behavior |
|
warn('Badly formated number'); |
|
} else if (ch === 0x45 || ch === 0x65) { // 'E', 'e' |
|
// 'E' can be either a scientific notation or the beginning of a new |
|
// operator |
|
ch = this.peekChar(); |
|
if (ch === 0x2B || ch === 0x2D) { // '+', '-' |
|
powerValueSign = (ch === 0x2D) ? -1 : 1; |
|
this.nextChar(); // Consume the sign character |
|
} else if (ch < 0x30 || ch > 0x39) { // '0' - '9' |
|
// The 'E' must be the beginning of a new operator |
|
break; |
|
} |
|
eNotation = true; |
|
} else { |
|
// the last character doesn't belong to us |
|
break; |
|
} |
|
} |
|
|
|
if (divideBy !== 0) { |
|
baseValue /= divideBy; |
|
} |
|
if (eNotation) { |
|
baseValue *= Math.pow(10, powerValueSign * powerValue); |
|
} |
|
return sign * baseValue; |
|
}, |
|
getString: function Lexer_getString() { |
|
var numParen = 1; |
|
var done = false; |
|
var strBuf = this.strBuf; |
|
strBuf.length = 0; |
|
|
|
var ch = this.nextChar(); |
|
while (true) { |
|
var charBuffered = false; |
|
switch (ch | 0) { |
|
case -1: |
|
warn('Unterminated string'); |
|
done = true; |
|
break; |
|
case 0x28: // '(' |
|
++numParen; |
|
strBuf.push('('); |
|
break; |
|
case 0x29: // ')' |
|
if (--numParen === 0) { |
|
this.nextChar(); // consume strings ')' |
|
done = true; |
|
} else { |
|
strBuf.push(')'); |
|
} |
|
break; |
|
case 0x5C: // '\\' |
|
ch = this.nextChar(); |
|
switch (ch) { |
|
case -1: |
|
warn('Unterminated string'); |
|
done = true; |
|
break; |
|
case 0x6E: // 'n' |
|
strBuf.push('\n'); |
|
break; |
|
case 0x72: // 'r' |
|
strBuf.push('\r'); |
|
break; |
|
case 0x74: // 't' |
|
strBuf.push('\t'); |
|
break; |
|
case 0x62: // 'b' |
|
strBuf.push('\b'); |
|
break; |
|
case 0x66: // 'f' |
|
strBuf.push('\f'); |
|
break; |
|
case 0x5C: // '\' |
|
case 0x28: // '(' |
|
case 0x29: // ')' |
|
strBuf.push(String.fromCharCode(ch)); |
|
break; |
|
case 0x30: case 0x31: case 0x32: case 0x33: // '0'-'3' |
|
case 0x34: case 0x35: case 0x36: case 0x37: // '4'-'7' |
|
var x = ch & 0x0F; |
|
ch = this.nextChar(); |
|
charBuffered = true; |
|
if (ch >= 0x30 && ch <= 0x37) { // '0'-'7' |
|
x = (x << 3) + (ch & 0x0F); |
|
ch = this.nextChar(); |
|
if (ch >= 0x30 && ch <= 0x37) { // '0'-'7' |
|
charBuffered = false; |
|
x = (x << 3) + (ch & 0x0F); |
|
} |
|
} |
|
strBuf.push(String.fromCharCode(x)); |
|
break; |
|
case 0x0D: // CR |
|
if (this.peekChar() === 0x0A) { // LF |
|
this.nextChar(); |
|
} |
|
break; |
|
case 0x0A: // LF |
|
break; |
|
default: |
|
strBuf.push(String.fromCharCode(ch)); |
|
break; |
|
} |
|
break; |
|
default: |
|
strBuf.push(String.fromCharCode(ch)); |
|
break; |
|
} |
|
if (done) { |
|
break; |
|
} |
|
if (!charBuffered) { |
|
ch = this.nextChar(); |
|
} |
|
} |
|
return strBuf.join(''); |
|
}, |
|
getName: function Lexer_getName() { |
|
var ch, previousCh; |
|
var strBuf = this.strBuf; |
|
strBuf.length = 0; |
|
while ((ch = this.nextChar()) >= 0 && !specialChars[ch]) { |
|
if (ch === 0x23) { // '#' |
|
ch = this.nextChar(); |
|
if (specialChars[ch]) { |
|
warn('Lexer_getName: ' + |
|
'NUMBER SIGN (#) should be followed by a hexadecimal number.'); |
|
strBuf.push('#'); |
|
break; |
|
} |
|
var x = toHexDigit(ch); |
|
if (x !== -1) { |
|
previousCh = ch; |
|
ch = this.nextChar(); |
|
var x2 = toHexDigit(ch); |
|
if (x2 === -1) { |
|
warn('Lexer_getName: Illegal digit (' + |
|
String.fromCharCode(ch) +') in hexadecimal number.'); |
|
strBuf.push('#', String.fromCharCode(previousCh)); |
|
if (specialChars[ch]) { |
|
break; |
|
} |
|
strBuf.push(String.fromCharCode(ch)); |
|
continue; |
|
} |
|
strBuf.push(String.fromCharCode((x << 4) | x2)); |
|
} else { |
|
strBuf.push('#', String.fromCharCode(ch)); |
|
} |
|
} else { |
|
strBuf.push(String.fromCharCode(ch)); |
|
} |
|
} |
|
if (strBuf.length > 127) { |
|
warn('name token is longer than allowed by the spec: ' + strBuf.length); |
|
} |
|
return Name.get(strBuf.join('')); |
|
}, |
|
getHexString: function Lexer_getHexString() { |
|
var strBuf = this.strBuf; |
|
strBuf.length = 0; |
|
var ch = this.currentChar; |
|
var isFirstHex = true; |
|
var firstDigit; |
|
var secondDigit; |
|
while (true) { |
|
if (ch < 0) { |
|
warn('Unterminated hex string'); |
|
break; |
|
} else if (ch === 0x3E) { // '>' |
|
this.nextChar(); |
|
break; |
|
} else if (specialChars[ch] === 1) { |
|
ch = this.nextChar(); |
|
continue; |
|
} else { |
|
if (isFirstHex) { |
|
firstDigit = toHexDigit(ch); |
|
if (firstDigit === -1) { |
|
warn('Ignoring invalid character "' + ch + '" in hex string'); |
|
ch = this.nextChar(); |
|
continue; |
|
} |
|
} else { |
|
secondDigit = toHexDigit(ch); |
|
if (secondDigit === -1) { |
|
warn('Ignoring invalid character "' + ch + '" in hex string'); |
|
ch = this.nextChar(); |
|
continue; |
|
} |
|
strBuf.push(String.fromCharCode((firstDigit << 4) | secondDigit)); |
|
} |
|
isFirstHex = !isFirstHex; |
|
ch = this.nextChar(); |
|
} |
|
} |
|
return strBuf.join(''); |
|
}, |
|
getObj: function Lexer_getObj() { |
|
// skip whitespace and comments |
|
var comment = false; |
|
var ch = this.currentChar; |
|
while (true) { |
|
if (ch < 0) { |
|
return EOF; |
|
} |
|
if (comment) { |
|
if (ch === 0x0A || ch === 0x0D) { // LF, CR |
|
comment = false; |
|
} |
|
} else if (ch === 0x25) { // '%' |
|
comment = true; |
|
} else if (specialChars[ch] !== 1) { |
|
break; |
|
} |
|
ch = this.nextChar(); |
|
} |
|
|
|
// start reading token |
|
switch (ch | 0) { |
|
case 0x30: case 0x31: case 0x32: case 0x33: case 0x34: // '0'-'4' |
|
case 0x35: case 0x36: case 0x37: case 0x38: case 0x39: // '5'-'9' |
|
case 0x2B: case 0x2D: case 0x2E: // '+', '-', '.' |
|
return this.getNumber(); |
|
case 0x28: // '(' |
|
return this.getString(); |
|
case 0x2F: // '/' |
|
return this.getName(); |
|
// array punctuation |
|
case 0x5B: // '[' |
|
this.nextChar(); |
|
return Cmd.get('['); |
|
case 0x5D: // ']' |
|
this.nextChar(); |
|
return Cmd.get(']'); |
|
// hex string or dict punctuation |
|
case 0x3C: // '<' |
|
ch = this.nextChar(); |
|
if (ch === 0x3C) { |
|
// dict punctuation |
|
this.nextChar(); |
|
return Cmd.get('<<'); |
|
} |
|
return this.getHexString(); |
|
// dict punctuation |
|
case 0x3E: // '>' |
|
ch = this.nextChar(); |
|
if (ch === 0x3E) { |
|
this.nextChar(); |
|
return Cmd.get('>>'); |
|
} |
|
return Cmd.get('>'); |
|
case 0x7B: // '{' |
|
this.nextChar(); |
|
return Cmd.get('{'); |
|
case 0x7D: // '}' |
|
this.nextChar(); |
|
return Cmd.get('}'); |
|
case 0x29: // ')' |
|
error('Illegal character: ' + ch); |
|
break; |
|
} |
|
|
|
// command |
|
var str = String.fromCharCode(ch); |
|
var knownCommands = this.knownCommands; |
|
var knownCommandFound = knownCommands && knownCommands[str] !== undefined; |
|
while ((ch = this.nextChar()) >= 0 && !specialChars[ch]) { |
|
// stop if known command is found and next character does not make |
|
// the str a command |
|
var possibleCommand = str + String.fromCharCode(ch); |
|
if (knownCommandFound && knownCommands[possibleCommand] === undefined) { |
|
break; |
|
} |
|
if (str.length === 128) { |
|
error('Command token too long: ' + str.length); |
|
} |
|
str = possibleCommand; |
|
knownCommandFound = knownCommands && knownCommands[str] !== undefined; |
|
} |
|
if (str === 'true') { |
|
return true; |
|
} |
|
if (str === 'false') { |
|
return false; |
|
} |
|
if (str === 'null') { |
|
return null; |
|
} |
|
return Cmd.get(str); |
|
}, |
|
skipToNextLine: function Lexer_skipToNextLine() { |
|
var ch = this.currentChar; |
|
while (ch >= 0) { |
|
if (ch === 0x0D) { // CR |
|
ch = this.nextChar(); |
|
if (ch === 0x0A) { // LF |
|
this.nextChar(); |
|
} |
|
break; |
|
} else if (ch === 0x0A) { // LF |
|
this.nextChar(); |
|
break; |
|
} |
|
ch = this.nextChar(); |
|
} |
|
} |
|
}; |
|
|
|
return Lexer; |
|
})(); |
|
|
|
var Linearization = { |
|
create: function LinearizationCreate(stream) { |
|
function getInt(name, allowZeroValue) { |
|
var obj = linDict.get(name); |
|
if (isInt(obj) && (allowZeroValue ? obj >= 0 : obj > 0)) { |
|
return obj; |
|
} |
|
throw new Error('The "' + name + '" parameter in the linearization ' + |
|
'dictionary is invalid.'); |
|
} |
|
function getHints() { |
|
var hints = linDict.get('H'), hintsLength, item; |
|
if (isArray(hints) && |
|
((hintsLength = hints.length) === 2 || hintsLength === 4)) { |
|
for (var index = 0; index < hintsLength; index++) { |
|
if (!(isInt(item = hints[index]) && item > 0)) { |
|
throw new Error('Hint (' + index + |
|
') in the linearization dictionary is invalid.'); |
|
} |
|
} |
|
return hints; |
|
} |
|
throw new Error('Hint array in the linearization dictionary is invalid.'); |
|
} |
|
var parser = new Parser(new Lexer(stream), false, null); |
|
var obj1 = parser.getObj(); |
|
var obj2 = parser.getObj(); |
|
var obj3 = parser.getObj(); |
|
var linDict = parser.getObj(); |
|
var obj, length; |
|
if (!(isInt(obj1) && isInt(obj2) && isCmd(obj3, 'obj') && isDict(linDict) && |
|
isNum(obj = linDict.get('Linearized')) && obj > 0)) { |
|
return null; // No valid linearization dictionary found. |
|
} else if ((length = getInt('L')) !== stream.length) { |
|
throw new Error('The "L" parameter in the linearization dictionary ' + |
|
'does not equal the stream length.'); |
|
} |
|
return { |
|
length: length, |
|
hints: getHints(), |
|
objectNumberFirst: getInt('O'), |
|
endFirst: getInt('E'), |
|
numPages: getInt('N'), |
|
mainXRefEntriesOffset: getInt('T'), |
|
pageFirst: (linDict.has('P') ? getInt('P', true) : 0) |
|
}; |
|
} |
|
}; |
|
|
|
exports.EOF = EOF; |
|
exports.Lexer = Lexer; |
|
exports.Linearization = Linearization; |
|
exports.Parser = Parser; |
|
exports.isEOF = isEOF; |
|
|
|
// TODO refactor to remove dependency on stream.js |
|
coreStream._setCoreParser(exports); |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreCMap = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreStream, root.pdfjsCoreParser); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreStream, coreParser) { |
|
|
|
var Util = sharedUtil.Util; |
|
var assert = sharedUtil.assert; |
|
var error = sharedUtil.error; |
|
var isInt = sharedUtil.isInt; |
|
var isString = sharedUtil.isString; |
|
var warn = sharedUtil.warn; |
|
var isName = corePrimitives.isName; |
|
var isCmd = corePrimitives.isCmd; |
|
var isStream = corePrimitives.isStream; |
|
var StringStream = coreStream.StringStream; |
|
var Lexer = coreParser.Lexer; |
|
var isEOF = coreParser.isEOF; |
|
|
|
var BUILT_IN_CMAPS = [ |
|
// << Start unicode maps. |
|
'Adobe-GB1-UCS2', |
|
'Adobe-CNS1-UCS2', |
|
'Adobe-Japan1-UCS2', |
|
'Adobe-Korea1-UCS2', |
|
// >> End unicode maps. |
|
'78-EUC-H', |
|
'78-EUC-V', |
|
'78-H', |
|
'78-RKSJ-H', |
|
'78-RKSJ-V', |
|
'78-V', |
|
'78ms-RKSJ-H', |
|
'78ms-RKSJ-V', |
|
'83pv-RKSJ-H', |
|
'90ms-RKSJ-H', |
|
'90ms-RKSJ-V', |
|
'90msp-RKSJ-H', |
|
'90msp-RKSJ-V', |
|
'90pv-RKSJ-H', |
|
'90pv-RKSJ-V', |
|
'Add-H', |
|
'Add-RKSJ-H', |
|
'Add-RKSJ-V', |
|
'Add-V', |
|
'Adobe-CNS1-0', |
|
'Adobe-CNS1-1', |
|
'Adobe-CNS1-2', |
|
'Adobe-CNS1-3', |
|
'Adobe-CNS1-4', |
|
'Adobe-CNS1-5', |
|
'Adobe-CNS1-6', |
|
'Adobe-GB1-0', |
|
'Adobe-GB1-1', |
|
'Adobe-GB1-2', |
|
'Adobe-GB1-3', |
|
'Adobe-GB1-4', |
|
'Adobe-GB1-5', |
|
'Adobe-Japan1-0', |
|
'Adobe-Japan1-1', |
|
'Adobe-Japan1-2', |
|
'Adobe-Japan1-3', |
|
'Adobe-Japan1-4', |
|
'Adobe-Japan1-5', |
|
'Adobe-Japan1-6', |
|
'Adobe-Korea1-0', |
|
'Adobe-Korea1-1', |
|
'Adobe-Korea1-2', |
|
'B5-H', |
|
'B5-V', |
|
'B5pc-H', |
|
'B5pc-V', |
|
'CNS-EUC-H', |
|
'CNS-EUC-V', |
|
'CNS1-H', |
|
'CNS1-V', |
|
'CNS2-H', |
|
'CNS2-V', |
|
'ETHK-B5-H', |
|
'ETHK-B5-V', |
|
'ETen-B5-H', |
|
'ETen-B5-V', |
|
'ETenms-B5-H', |
|
'ETenms-B5-V', |
|
'EUC-H', |
|
'EUC-V', |
|
'Ext-H', |
|
'Ext-RKSJ-H', |
|
'Ext-RKSJ-V', |
|
'Ext-V', |
|
'GB-EUC-H', |
|
'GB-EUC-V', |
|
'GB-H', |
|
'GB-V', |
|
'GBK-EUC-H', |
|
'GBK-EUC-V', |
|
'GBK2K-H', |
|
'GBK2K-V', |
|
'GBKp-EUC-H', |
|
'GBKp-EUC-V', |
|
'GBT-EUC-H', |
|
'GBT-EUC-V', |
|
'GBT-H', |
|
'GBT-V', |
|
'GBTpc-EUC-H', |
|
'GBTpc-EUC-V', |
|
'GBpc-EUC-H', |
|
'GBpc-EUC-V', |
|
'H', |
|
'HKdla-B5-H', |
|
'HKdla-B5-V', |
|
'HKdlb-B5-H', |
|
'HKdlb-B5-V', |
|
'HKgccs-B5-H', |
|
'HKgccs-B5-V', |
|
'HKm314-B5-H', |
|
'HKm314-B5-V', |
|
'HKm471-B5-H', |
|
'HKm471-B5-V', |
|
'HKscs-B5-H', |
|
'HKscs-B5-V', |
|
'Hankaku', |
|
'Hiragana', |
|
'KSC-EUC-H', |
|
'KSC-EUC-V', |
|
'KSC-H', |
|
'KSC-Johab-H', |
|
'KSC-Johab-V', |
|
'KSC-V', |
|
'KSCms-UHC-H', |
|
'KSCms-UHC-HW-H', |
|
'KSCms-UHC-HW-V', |
|
'KSCms-UHC-V', |
|
'KSCpc-EUC-H', |
|
'KSCpc-EUC-V', |
|
'Katakana', |
|
'NWP-H', |
|
'NWP-V', |
|
'RKSJ-H', |
|
'RKSJ-V', |
|
'Roman', |
|
'UniCNS-UCS2-H', |
|
'UniCNS-UCS2-V', |
|
'UniCNS-UTF16-H', |
|
'UniCNS-UTF16-V', |
|
'UniCNS-UTF32-H', |
|
'UniCNS-UTF32-V', |
|
'UniCNS-UTF8-H', |
|
'UniCNS-UTF8-V', |
|
'UniGB-UCS2-H', |
|
'UniGB-UCS2-V', |
|
'UniGB-UTF16-H', |
|
'UniGB-UTF16-V', |
|
'UniGB-UTF32-H', |
|
'UniGB-UTF32-V', |
|
'UniGB-UTF8-H', |
|
'UniGB-UTF8-V', |
|
'UniJIS-UCS2-H', |
|
'UniJIS-UCS2-HW-H', |
|
'UniJIS-UCS2-HW-V', |
|
'UniJIS-UCS2-V', |
|
'UniJIS-UTF16-H', |
|
'UniJIS-UTF16-V', |
|
'UniJIS-UTF32-H', |
|
'UniJIS-UTF32-V', |
|
'UniJIS-UTF8-H', |
|
'UniJIS-UTF8-V', |
|
'UniJIS2004-UTF16-H', |
|
'UniJIS2004-UTF16-V', |
|
'UniJIS2004-UTF32-H', |
|
'UniJIS2004-UTF32-V', |
|
'UniJIS2004-UTF8-H', |
|
'UniJIS2004-UTF8-V', |
|
'UniJISPro-UCS2-HW-V', |
|
'UniJISPro-UCS2-V', |
|
'UniJISPro-UTF8-V', |
|
'UniJISX0213-UTF32-H', |
|
'UniJISX0213-UTF32-V', |
|
'UniJISX02132004-UTF32-H', |
|
'UniJISX02132004-UTF32-V', |
|
'UniKS-UCS2-H', |
|
'UniKS-UCS2-V', |
|
'UniKS-UTF16-H', |
|
'UniKS-UTF16-V', |
|
'UniKS-UTF32-H', |
|
'UniKS-UTF32-V', |
|
'UniKS-UTF8-H', |
|
'UniKS-UTF8-V', |
|
'V', |
|
'WP-Symbol']; |
|
|
|
// CMap, not to be confused with TrueType's cmap. |
|
var CMap = (function CMapClosure() { |
|
function CMap(builtInCMap) { |
|
// Codespace ranges are stored as follows: |
|
// [[1BytePairs], [2BytePairs], [3BytePairs], [4BytePairs]] |
|
// where nBytePairs are ranges e.g. [low1, high1, low2, high2, ...] |
|
this.codespaceRanges = [[], [], [], []]; |
|
this.numCodespaceRanges = 0; |
|
// Map entries have one of two forms. |
|
// - cid chars are 16-bit unsigned integers, stored as integers. |
|
// - bf chars are variable-length byte sequences, stored as strings, with |
|
// one byte per character. |
|
this._map = []; |
|
this.name = ''; |
|
this.vertical = false; |
|
this.useCMap = null; |
|
this.builtInCMap = builtInCMap; |
|
} |
|
CMap.prototype = { |
|
addCodespaceRange: function(n, low, high) { |
|
this.codespaceRanges[n - 1].push(low, high); |
|
this.numCodespaceRanges++; |
|
}, |
|
|
|
mapCidRange: function(low, high, dstLow) { |
|
while (low <= high) { |
|
this._map[low++] = dstLow++; |
|
} |
|
}, |
|
|
|
mapBfRange: function(low, high, dstLow) { |
|
var lastByte = dstLow.length - 1; |
|
while (low <= high) { |
|
this._map[low++] = dstLow; |
|
// Only the last byte has to be incremented. |
|
dstLow = dstLow.substr(0, lastByte) + |
|
String.fromCharCode(dstLow.charCodeAt(lastByte) + 1); |
|
} |
|
}, |
|
|
|
mapBfRangeToArray: function(low, high, array) { |
|
var i = 0, ii = array.length; |
|
while (low <= high && i < ii) { |
|
this._map[low] = array[i++]; |
|
++low; |
|
} |
|
}, |
|
|
|
// This is used for both bf and cid chars. |
|
mapOne: function(src, dst) { |
|
this._map[src] = dst; |
|
}, |
|
|
|
lookup: function(code) { |
|
return this._map[code]; |
|
}, |
|
|
|
contains: function(code) { |
|
return this._map[code] !== undefined; |
|
}, |
|
|
|
forEach: function(callback) { |
|
// Most maps have fewer than 65536 entries, and for those we use normal |
|
// array iteration. But really sparse tables are possible -- e.g. with |
|
// indices in the *billions*. For such tables we use for..in, which isn't |
|
// ideal because it stringifies the indices for all present elements, but |
|
// it does avoid iterating over every undefined entry. |
|
var map = this._map; |
|
var length = map.length; |
|
var i; |
|
if (length <= 0x10000) { |
|
for (i = 0; i < length; i++) { |
|
if (map[i] !== undefined) { |
|
callback(i, map[i]); |
|
} |
|
} |
|
} else { |
|
for (i in this._map) { |
|
callback(i, map[i]); |
|
} |
|
} |
|
}, |
|
|
|
charCodeOf: function(value) { |
|
return this._map.indexOf(value); |
|
}, |
|
|
|
getMap: function() { |
|
return this._map; |
|
}, |
|
|
|
readCharCode: function(str, offset, out) { |
|
var c = 0; |
|
var codespaceRanges = this.codespaceRanges; |
|
var codespaceRangesLen = this.codespaceRanges.length; |
|
// 9.7.6.2 CMap Mapping |
|
// The code length is at most 4. |
|
for (var n = 0; n < codespaceRangesLen; n++) { |
|
c = ((c << 8) | str.charCodeAt(offset + n)) >>> 0; |
|
// Check each codespace range to see if it falls within. |
|
var codespaceRange = codespaceRanges[n]; |
|
for (var k = 0, kk = codespaceRange.length; k < kk;) { |
|
var low = codespaceRange[k++]; |
|
var high = codespaceRange[k++]; |
|
if (c >= low && c <= high) { |
|
out.charcode = c; |
|
out.length = n + 1; |
|
return; |
|
} |
|
} |
|
} |
|
out.charcode = 0; |
|
out.length = 1; |
|
}, |
|
|
|
get length() { |
|
return this._map.length; |
|
}, |
|
|
|
get isIdentityCMap() { |
|
if (!(this.name === 'Identity-H' || this.name === 'Identity-V')) { |
|
return false; |
|
} |
|
if (this._map.length !== 0x10000) { |
|
return false; |
|
} |
|
for (var i = 0; i < 0x10000; i++) { |
|
if (this._map[i] !== i) { |
|
return false; |
|
} |
|
} |
|
return true; |
|
} |
|
}; |
|
return CMap; |
|
})(); |
|
|
|
// A special case of CMap, where the _map array implicitly has a length of |
|
// 65536 and each element is equal to its index. |
|
var IdentityCMap = (function IdentityCMapClosure() { |
|
function IdentityCMap(vertical, n) { |
|
CMap.call(this); |
|
this.vertical = vertical; |
|
this.addCodespaceRange(n, 0, 0xffff); |
|
} |
|
Util.inherit(IdentityCMap, CMap, {}); |
|
|
|
IdentityCMap.prototype = { |
|
addCodespaceRange: CMap.prototype.addCodespaceRange, |
|
|
|
mapCidRange: function(low, high, dstLow) { |
|
error('should not call mapCidRange'); |
|
}, |
|
|
|
mapBfRange: function(low, high, dstLow) { |
|
error('should not call mapBfRange'); |
|
}, |
|
|
|
mapBfRangeToArray: function(low, high, array) { |
|
error('should not call mapBfRangeToArray'); |
|
}, |
|
|
|
mapOne: function(src, dst) { |
|
error('should not call mapCidOne'); |
|
}, |
|
|
|
lookup: function(code) { |
|
return (isInt(code) && code <= 0xffff) ? code : undefined; |
|
}, |
|
|
|
contains: function(code) { |
|
return isInt(code) && code <= 0xffff; |
|
}, |
|
|
|
forEach: function(callback) { |
|
for (var i = 0; i <= 0xffff; i++) { |
|
callback(i, i); |
|
} |
|
}, |
|
|
|
charCodeOf: function(value) { |
|
return (isInt(value) && value <= 0xffff) ? value : -1; |
|
}, |
|
|
|
getMap: function() { |
|
// Sometimes identity maps must be instantiated, but it's rare. |
|
var map = new Array(0x10000); |
|
for (var i = 0; i <= 0xffff; i++) { |
|
map[i] = i; |
|
} |
|
return map; |
|
}, |
|
|
|
readCharCode: CMap.prototype.readCharCode, |
|
|
|
get length() { |
|
return 0x10000; |
|
}, |
|
|
|
get isIdentityCMap() { |
|
error('should not access .isIdentityCMap'); |
|
} |
|
}; |
|
|
|
return IdentityCMap; |
|
})(); |
|
|
|
var BinaryCMapReader = (function BinaryCMapReaderClosure() { |
|
function fetchBinaryData(url) { |
|
var nonBinaryRequest = PDFJS.disableWorker; |
|
var request = new XMLHttpRequest(); |
|
request.open('GET', url, false); |
|
if (!nonBinaryRequest) { |
|
try { |
|
request.responseType = 'arraybuffer'; |
|
nonBinaryRequest = request.responseType !== 'arraybuffer'; |
|
} catch (e) { |
|
nonBinaryRequest = true; |
|
} |
|
} |
|
if (nonBinaryRequest && request.overrideMimeType) { |
|
request.overrideMimeType('text/plain; charset=x-user-defined'); |
|
} |
|
request.send(null); |
|
if (nonBinaryRequest ? !request.responseText : !request.response) { |
|
error('Unable to get binary cMap at: ' + url); |
|
} |
|
if (nonBinaryRequest) { |
|
var data = Array.prototype.map.call(request.responseText, function (ch) { |
|
return ch.charCodeAt(0) & 255; |
|
}); |
|
return new Uint8Array(data); |
|
} |
|
return new Uint8Array(request.response); |
|
} |
|
|
|
function hexToInt(a, size) { |
|
var n = 0; |
|
for (var i = 0; i <= size; i++) { |
|
n = (n << 8) | a[i]; |
|
} |
|
return n >>> 0; |
|
} |
|
|
|
function hexToStr(a, size) { |
|
// This code is hot. Special-case some common values to avoid creating an |
|
// object with subarray(). |
|
if (size === 1) { |
|
return String.fromCharCode(a[0], a[1]); |
|
} |
|
if (size === 3) { |
|
return String.fromCharCode(a[0], a[1], a[2], a[3]); |
|
} |
|
return String.fromCharCode.apply(null, a.subarray(0, size + 1)); |
|
} |
|
|
|
function addHex(a, b, size) { |
|
var c = 0; |
|
for (var i = size; i >= 0; i--) { |
|
c += a[i] + b[i]; |
|
a[i] = c & 255; |
|
c >>= 8; |
|
} |
|
} |
|
|
|
function incHex(a, size) { |
|
var c = 1; |
|
for (var i = size; i >= 0 && c > 0; i--) { |
|
c += a[i]; |
|
a[i] = c & 255; |
|
c >>= 8; |
|
} |
|
} |
|
|
|
var MAX_NUM_SIZE = 16; |
|
var MAX_ENCODED_NUM_SIZE = 19; // ceil(MAX_NUM_SIZE * 7 / 8) |
|
|
|
function BinaryCMapStream(data) { |
|
this.buffer = data; |
|
this.pos = 0; |
|
this.end = data.length; |
|
this.tmpBuf = new Uint8Array(MAX_ENCODED_NUM_SIZE); |
|
} |
|
|
|
BinaryCMapStream.prototype = { |
|
readByte: function () { |
|
if (this.pos >= this.end) { |
|
return -1; |
|
} |
|
return this.buffer[this.pos++]; |
|
}, |
|
readNumber: function () { |
|
var n = 0; |
|
var last; |
|
do { |
|
var b = this.readByte(); |
|
if (b < 0) { |
|
error('unexpected EOF in bcmap'); |
|
} |
|
last = !(b & 0x80); |
|
n = (n << 7) | (b & 0x7F); |
|
} while (!last); |
|
return n; |
|
}, |
|
readSigned: function () { |
|
var n = this.readNumber(); |
|
return (n & 1) ? ~(n >>> 1) : n >>> 1; |
|
}, |
|
readHex: function (num, size) { |
|
num.set(this.buffer.subarray(this.pos, |
|
this.pos + size + 1)); |
|
this.pos += size + 1; |
|
}, |
|
readHexNumber: function (num, size) { |
|
var last; |
|
var stack = this.tmpBuf, sp = 0; |
|
do { |
|
var b = this.readByte(); |
|
if (b < 0) { |
|
error('unexpected EOF in bcmap'); |
|
} |
|
last = !(b & 0x80); |
|
stack[sp++] = b & 0x7F; |
|
} while (!last); |
|
var i = size, buffer = 0, bufferSize = 0; |
|
while (i >= 0) { |
|
while (bufferSize < 8 && stack.length > 0) { |
|
buffer = (stack[--sp] << bufferSize) | buffer; |
|
bufferSize += 7; |
|
} |
|
num[i] = buffer & 255; |
|
i--; |
|
buffer >>= 8; |
|
bufferSize -= 8; |
|
} |
|
}, |
|
readHexSigned: function (num, size) { |
|
this.readHexNumber(num, size); |
|
var sign = num[size] & 1 ? 255 : 0; |
|
var c = 0; |
|
for (var i = 0; i <= size; i++) { |
|
c = ((c & 1) << 8) | num[i]; |
|
num[i] = (c >> 1) ^ sign; |
|
} |
|
}, |
|
readString: function () { |
|
var len = this.readNumber(); |
|
var s = ''; |
|
for (var i = 0; i < len; i++) { |
|
s += String.fromCharCode(this.readNumber()); |
|
} |
|
return s; |
|
} |
|
}; |
|
|
|
function processBinaryCMap(url, cMap, extend) { |
|
var data = fetchBinaryData(url); |
|
var stream = new BinaryCMapStream(data); |
|
|
|
var header = stream.readByte(); |
|
cMap.vertical = !!(header & 1); |
|
|
|
var useCMap = null; |
|
var start = new Uint8Array(MAX_NUM_SIZE); |
|
var end = new Uint8Array(MAX_NUM_SIZE); |
|
var char = new Uint8Array(MAX_NUM_SIZE); |
|
var charCode = new Uint8Array(MAX_NUM_SIZE); |
|
var tmp = new Uint8Array(MAX_NUM_SIZE); |
|
var code; |
|
|
|
var b; |
|
while ((b = stream.readByte()) >= 0) { |
|
var type = b >> 5; |
|
if (type === 7) { // metadata, e.g. comment or usecmap |
|
switch (b & 0x1F) { |
|
case 0: |
|
stream.readString(); // skipping comment |
|
break; |
|
case 1: |
|
useCMap = stream.readString(); |
|
break; |
|
} |
|
continue; |
|
} |
|
var sequence = !!(b & 0x10); |
|
var dataSize = b & 15; |
|
|
|
assert(dataSize + 1 <= MAX_NUM_SIZE); |
|
|
|
var ucs2DataSize = 1; |
|
var subitemsCount = stream.readNumber(); |
|
var i; |
|
switch (type) { |
|
case 0: // codespacerange |
|
stream.readHex(start, dataSize); |
|
stream.readHexNumber(end, dataSize); |
|
addHex(end, start, dataSize); |
|
cMap.addCodespaceRange(dataSize + 1, hexToInt(start, dataSize), |
|
hexToInt(end, dataSize)); |
|
for (i = 1; i < subitemsCount; i++) { |
|
incHex(end, dataSize); |
|
stream.readHexNumber(start, dataSize); |
|
addHex(start, end, dataSize); |
|
stream.readHexNumber(end, dataSize); |
|
addHex(end, start, dataSize); |
|
cMap.addCodespaceRange(dataSize + 1, hexToInt(start, dataSize), |
|
hexToInt(end, dataSize)); |
|
} |
|
break; |
|
case 1: // notdefrange |
|
stream.readHex(start, dataSize); |
|
stream.readHexNumber(end, dataSize); |
|
addHex(end, start, dataSize); |
|
code = stream.readNumber(); |
|
// undefined range, skipping |
|
for (i = 1; i < subitemsCount; i++) { |
|
incHex(end, dataSize); |
|
stream.readHexNumber(start, dataSize); |
|
addHex(start, end, dataSize); |
|
stream.readHexNumber(end, dataSize); |
|
addHex(end, start, dataSize); |
|
code = stream.readNumber(); |
|
// nop |
|
} |
|
break; |
|
case 2: // cidchar |
|
stream.readHex(char, dataSize); |
|
code = stream.readNumber(); |
|
cMap.mapOne(hexToInt(char, dataSize), code); |
|
for (i = 1; i < subitemsCount; i++) { |
|
incHex(char, dataSize); |
|
if (!sequence) { |
|
stream.readHexNumber(tmp, dataSize); |
|
addHex(char, tmp, dataSize); |
|
} |
|
code = stream.readSigned() + (code + 1); |
|
cMap.mapOne(hexToInt(char, dataSize), code); |
|
} |
|
break; |
|
case 3: // cidrange |
|
stream.readHex(start, dataSize); |
|
stream.readHexNumber(end, dataSize); |
|
addHex(end, start, dataSize); |
|
code = stream.readNumber(); |
|
cMap.mapCidRange(hexToInt(start, dataSize), hexToInt(end, dataSize), |
|
code); |
|
for (i = 1; i < subitemsCount; i++) { |
|
incHex(end, dataSize); |
|
if (!sequence) { |
|
stream.readHexNumber(start, dataSize); |
|
addHex(start, end, dataSize); |
|
} else { |
|
start.set(end); |
|
} |
|
stream.readHexNumber(end, dataSize); |
|
addHex(end, start, dataSize); |
|
code = stream.readNumber(); |
|
cMap.mapCidRange(hexToInt(start, dataSize), hexToInt(end, dataSize), |
|
code); |
|
} |
|
break; |
|
case 4: // bfchar |
|
stream.readHex(char, ucs2DataSize); |
|
stream.readHex(charCode, dataSize); |
|
cMap.mapOne(hexToInt(char, ucs2DataSize), |
|
hexToStr(charCode, dataSize)); |
|
for (i = 1; i < subitemsCount; i++) { |
|
incHex(char, ucs2DataSize); |
|
if (!sequence) { |
|
stream.readHexNumber(tmp, ucs2DataSize); |
|
addHex(char, tmp, ucs2DataSize); |
|
} |
|
incHex(charCode, dataSize); |
|
stream.readHexSigned(tmp, dataSize); |
|
addHex(charCode, tmp, dataSize); |
|
cMap.mapOne(hexToInt(char, ucs2DataSize), |
|
hexToStr(charCode, dataSize)); |
|
} |
|
break; |
|
case 5: // bfrange |
|
stream.readHex(start, ucs2DataSize); |
|
stream.readHexNumber(end, ucs2DataSize); |
|
addHex(end, start, ucs2DataSize); |
|
stream.readHex(charCode, dataSize); |
|
cMap.mapBfRange(hexToInt(start, ucs2DataSize), |
|
hexToInt(end, ucs2DataSize), |
|
hexToStr(charCode, dataSize)); |
|
for (i = 1; i < subitemsCount; i++) { |
|
incHex(end, ucs2DataSize); |
|
if (!sequence) { |
|
stream.readHexNumber(start, ucs2DataSize); |
|
addHex(start, end, ucs2DataSize); |
|
} else { |
|
start.set(end); |
|
} |
|
stream.readHexNumber(end, ucs2DataSize); |
|
addHex(end, start, ucs2DataSize); |
|
stream.readHex(charCode, dataSize); |
|
cMap.mapBfRange(hexToInt(start, ucs2DataSize), |
|
hexToInt(end, ucs2DataSize), |
|
hexToStr(charCode, dataSize)); |
|
} |
|
break; |
|
default: |
|
error('Unknown type: ' + type); |
|
break; |
|
} |
|
} |
|
|
|
if (useCMap) { |
|
extend(useCMap); |
|
} |
|
return cMap; |
|
} |
|
|
|
function BinaryCMapReader() {} |
|
|
|
BinaryCMapReader.prototype = { |
|
read: processBinaryCMap |
|
}; |
|
|
|
return BinaryCMapReader; |
|
})(); |
|
|
|
var CMapFactory = (function CMapFactoryClosure() { |
|
function strToInt(str) { |
|
var a = 0; |
|
for (var i = 0; i < str.length; i++) { |
|
a = (a << 8) | str.charCodeAt(i); |
|
} |
|
return a >>> 0; |
|
} |
|
|
|
function expectString(obj) { |
|
if (!isString(obj)) { |
|
error('Malformed CMap: expected string.'); |
|
} |
|
} |
|
|
|
function expectInt(obj) { |
|
if (!isInt(obj)) { |
|
error('Malformed CMap: expected int.'); |
|
} |
|
} |
|
|
|
function parseBfChar(cMap, lexer) { |
|
while (true) { |
|
var obj = lexer.getObj(); |
|
if (isEOF(obj)) { |
|
break; |
|
} |
|
if (isCmd(obj, 'endbfchar')) { |
|
return; |
|
} |
|
expectString(obj); |
|
var src = strToInt(obj); |
|
obj = lexer.getObj(); |
|
// TODO are /dstName used? |
|
expectString(obj); |
|
var dst = obj; |
|
cMap.mapOne(src, dst); |
|
} |
|
} |
|
|
|
function parseBfRange(cMap, lexer) { |
|
while (true) { |
|
var obj = lexer.getObj(); |
|
if (isEOF(obj)) { |
|
break; |
|
} |
|
if (isCmd(obj, 'endbfrange')) { |
|
return; |
|
} |
|
expectString(obj); |
|
var low = strToInt(obj); |
|
obj = lexer.getObj(); |
|
expectString(obj); |
|
var high = strToInt(obj); |
|
obj = lexer.getObj(); |
|
if (isInt(obj) || isString(obj)) { |
|
var dstLow = isInt(obj) ? String.fromCharCode(obj) : obj; |
|
cMap.mapBfRange(low, high, dstLow); |
|
} else if (isCmd(obj, '[')) { |
|
obj = lexer.getObj(); |
|
var array = []; |
|
while (!isCmd(obj, ']') && !isEOF(obj)) { |
|
array.push(obj); |
|
obj = lexer.getObj(); |
|
} |
|
cMap.mapBfRangeToArray(low, high, array); |
|
} else { |
|
break; |
|
} |
|
} |
|
error('Invalid bf range.'); |
|
} |
|
|
|
function parseCidChar(cMap, lexer) { |
|
while (true) { |
|
var obj = lexer.getObj(); |
|
if (isEOF(obj)) { |
|
break; |
|
} |
|
if (isCmd(obj, 'endcidchar')) { |
|
return; |
|
} |
|
expectString(obj); |
|
var src = strToInt(obj); |
|
obj = lexer.getObj(); |
|
expectInt(obj); |
|
var dst = obj; |
|
cMap.mapOne(src, dst); |
|
} |
|
} |
|
|
|
function parseCidRange(cMap, lexer) { |
|
while (true) { |
|
var obj = lexer.getObj(); |
|
if (isEOF(obj)) { |
|
break; |
|
} |
|
if (isCmd(obj, 'endcidrange')) { |
|
return; |
|
} |
|
expectString(obj); |
|
var low = strToInt(obj); |
|
obj = lexer.getObj(); |
|
expectString(obj); |
|
var high = strToInt(obj); |
|
obj = lexer.getObj(); |
|
expectInt(obj); |
|
var dstLow = obj; |
|
cMap.mapCidRange(low, high, dstLow); |
|
} |
|
} |
|
|
|
function parseCodespaceRange(cMap, lexer) { |
|
while (true) { |
|
var obj = lexer.getObj(); |
|
if (isEOF(obj)) { |
|
break; |
|
} |
|
if (isCmd(obj, 'endcodespacerange')) { |
|
return; |
|
} |
|
if (!isString(obj)) { |
|
break; |
|
} |
|
var low = strToInt(obj); |
|
obj = lexer.getObj(); |
|
if (!isString(obj)) { |
|
break; |
|
} |
|
var high = strToInt(obj); |
|
cMap.addCodespaceRange(obj.length, low, high); |
|
} |
|
error('Invalid codespace range.'); |
|
} |
|
|
|
function parseWMode(cMap, lexer) { |
|
var obj = lexer.getObj(); |
|
if (isInt(obj)) { |
|
cMap.vertical = !!obj; |
|
} |
|
} |
|
|
|
function parseCMapName(cMap, lexer) { |
|
var obj = lexer.getObj(); |
|
if (isName(obj) && isString(obj.name)) { |
|
cMap.name = obj.name; |
|
} |
|
} |
|
|
|
function parseCMap(cMap, lexer, builtInCMapParams, useCMap) { |
|
var previous; |
|
var embededUseCMap; |
|
objLoop: while (true) { |
|
var obj = lexer.getObj(); |
|
if (isEOF(obj)) { |
|
break; |
|
} else if (isName(obj)) { |
|
if (obj.name === 'WMode') { |
|
parseWMode(cMap, lexer); |
|
} else if (obj.name === 'CMapName') { |
|
parseCMapName(cMap, lexer); |
|
} |
|
previous = obj; |
|
} else if (isCmd(obj)) { |
|
switch (obj.cmd) { |
|
case 'endcmap': |
|
break objLoop; |
|
case 'usecmap': |
|
if (isName(previous)) { |
|
embededUseCMap = previous.name; |
|
} |
|
break; |
|
case 'begincodespacerange': |
|
parseCodespaceRange(cMap, lexer); |
|
break; |
|
case 'beginbfchar': |
|
parseBfChar(cMap, lexer); |
|
break; |
|
case 'begincidchar': |
|
parseCidChar(cMap, lexer); |
|
break; |
|
case 'beginbfrange': |
|
parseBfRange(cMap, lexer); |
|
break; |
|
case 'begincidrange': |
|
parseCidRange(cMap, lexer); |
|
break; |
|
} |
|
} |
|
} |
|
|
|
if (!useCMap && embededUseCMap) { |
|
// Load the usecmap definition from the file only if there wasn't one |
|
// specified. |
|
useCMap = embededUseCMap; |
|
} |
|
if (useCMap) { |
|
extendCMap(cMap, builtInCMapParams, useCMap); |
|
} |
|
} |
|
|
|
function extendCMap(cMap, builtInCMapParams, useCMap) { |
|
cMap.useCMap = createBuiltInCMap(useCMap, builtInCMapParams); |
|
// If there aren't any code space ranges defined clone all the parent ones |
|
// into this cMap. |
|
if (cMap.numCodespaceRanges === 0) { |
|
var useCodespaceRanges = cMap.useCMap.codespaceRanges; |
|
for (var i = 0; i < useCodespaceRanges.length; i++) { |
|
cMap.codespaceRanges[i] = useCodespaceRanges[i].slice(); |
|
} |
|
cMap.numCodespaceRanges = cMap.useCMap.numCodespaceRanges; |
|
} |
|
// Merge the map into the current one, making sure not to override |
|
// any previously defined entries. |
|
cMap.useCMap.forEach(function(key, value) { |
|
if (!cMap.contains(key)) { |
|
cMap.mapOne(key, cMap.useCMap.lookup(key)); |
|
} |
|
}); |
|
} |
|
|
|
function parseBinaryCMap(name, builtInCMapParams) { |
|
var url = builtInCMapParams.url + name + '.bcmap'; |
|
var cMap = new CMap(true); |
|
new BinaryCMapReader().read(url, cMap, function (useCMap) { |
|
extendCMap(cMap, builtInCMapParams, useCMap); |
|
}); |
|
return cMap; |
|
} |
|
|
|
function createBuiltInCMap(name, builtInCMapParams) { |
|
if (name === 'Identity-H') { |
|
return new IdentityCMap(false, 2); |
|
} else if (name === 'Identity-V') { |
|
return new IdentityCMap(true, 2); |
|
} |
|
if (BUILT_IN_CMAPS.indexOf(name) === -1) { |
|
error('Unknown cMap name: ' + name); |
|
} |
|
assert(builtInCMapParams, 'built-in cMap parameters are not provided'); |
|
|
|
if (builtInCMapParams.packed) { |
|
return parseBinaryCMap(name, builtInCMapParams); |
|
} |
|
|
|
var request = new XMLHttpRequest(); |
|
var url = builtInCMapParams.url + name; |
|
request.open('GET', url, false); |
|
request.send(null); |
|
if (!request.responseText) { |
|
error('Unable to get cMap at: ' + url); |
|
} |
|
var cMap = new CMap(true); |
|
var lexer = new Lexer(new StringStream(request.responseText)); |
|
parseCMap(cMap, lexer, builtInCMapParams, null); |
|
return cMap; |
|
} |
|
|
|
return { |
|
create: function (encoding, builtInCMapParams, useCMap) { |
|
if (isName(encoding)) { |
|
return createBuiltInCMap(encoding.name, builtInCMapParams); |
|
} else if (isStream(encoding)) { |
|
var cMap = new CMap(); |
|
var lexer = new Lexer(encoding); |
|
try { |
|
parseCMap(cMap, lexer, builtInCMapParams, useCMap); |
|
} catch (e) { |
|
warn('Invalid CMap data. ' + e); |
|
} |
|
if (cMap.isIdentityCMap) { |
|
return createBuiltInCMap(cMap.name, builtInCMapParams); |
|
} |
|
return cMap; |
|
} |
|
error('Encoding required.'); |
|
} |
|
}; |
|
})(); |
|
|
|
exports.CMap = CMap; |
|
exports.CMapFactory = CMapFactory; |
|
exports.IdentityCMap = IdentityCMap; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreObj = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreCrypto, root.pdfjsCoreParser, |
|
root.pdfjsCoreChunkedStream); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreCrypto, coreParser, |
|
coreChunkedStream) { |
|
|
|
var InvalidPDFException = sharedUtil.InvalidPDFException; |
|
var MissingDataException = sharedUtil.MissingDataException; |
|
var XRefParseException = sharedUtil.XRefParseException; |
|
var assert = sharedUtil.assert; |
|
var bytesToString = sharedUtil.bytesToString; |
|
var createPromiseCapability = sharedUtil.createPromiseCapability; |
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isArray = sharedUtil.isArray; |
|
var isInt = sharedUtil.isInt; |
|
var isString = sharedUtil.isString; |
|
var shadow = sharedUtil.shadow; |
|
var stringToPDFString = sharedUtil.stringToPDFString; |
|
var stringToUTF8String = sharedUtil.stringToUTF8String; |
|
var warn = sharedUtil.warn; |
|
var isValidUrl = sharedUtil.isValidUrl; |
|
var Util = sharedUtil.Util; |
|
var Ref = corePrimitives.Ref; |
|
var RefSet = corePrimitives.RefSet; |
|
var RefSetCache = corePrimitives.RefSetCache; |
|
var isName = corePrimitives.isName; |
|
var isCmd = corePrimitives.isCmd; |
|
var isDict = corePrimitives.isDict; |
|
var isRef = corePrimitives.isRef; |
|
var isStream = corePrimitives.isStream; |
|
var CipherTransformFactory = coreCrypto.CipherTransformFactory; |
|
var Lexer = coreParser.Lexer; |
|
var Parser = coreParser.Parser; |
|
var ChunkedStream = coreChunkedStream.ChunkedStream; |
|
|
|
var Catalog = (function CatalogClosure() { |
|
function Catalog(pdfManager, xref, pageFactory) { |
|
this.pdfManager = pdfManager; |
|
this.xref = xref; |
|
this.catDict = xref.getCatalogObj(); |
|
this.fontCache = new RefSetCache(); |
|
assert(isDict(this.catDict), |
|
'catalog object is not a dictionary'); |
|
|
|
// TODO refactor to move getPage() to the PDFDocument. |
|
this.pageFactory = pageFactory; |
|
this.pagePromises = []; |
|
} |
|
|
|
Catalog.prototype = { |
|
get metadata() { |
|
var streamRef = this.catDict.getRaw('Metadata'); |
|
if (!isRef(streamRef)) { |
|
return shadow(this, 'metadata', null); |
|
} |
|
|
|
var encryptMetadata = (!this.xref.encrypt ? false : |
|
this.xref.encrypt.encryptMetadata); |
|
|
|
var stream = this.xref.fetch(streamRef, !encryptMetadata); |
|
var metadata; |
|
if (stream && isDict(stream.dict)) { |
|
var type = stream.dict.get('Type'); |
|
var subtype = stream.dict.get('Subtype'); |
|
|
|
if (isName(type) && isName(subtype) && |
|
type.name === 'Metadata' && subtype.name === 'XML') { |
|
// XXX: This should examine the charset the XML document defines, |
|
// however since there are currently no real means to decode |
|
// arbitrary charsets, let's just hope that the author of the PDF |
|
// was reasonable enough to stick with the XML default charset, |
|
// which is UTF-8. |
|
try { |
|
metadata = stringToUTF8String(bytesToString(stream.getBytes())); |
|
} catch (e) { |
|
info('Skipping invalid metadata.'); |
|
} |
|
} |
|
} |
|
|
|
return shadow(this, 'metadata', metadata); |
|
}, |
|
get toplevelPagesDict() { |
|
var pagesObj = this.catDict.get('Pages'); |
|
assert(isDict(pagesObj), 'invalid top-level pages dictionary'); |
|
// shadow the prototype getter |
|
return shadow(this, 'toplevelPagesDict', pagesObj); |
|
}, |
|
get documentOutline() { |
|
var obj = null; |
|
try { |
|
obj = this.readDocumentOutline(); |
|
} catch (ex) { |
|
if (ex instanceof MissingDataException) { |
|
throw ex; |
|
} |
|
warn('Unable to read document outline'); |
|
} |
|
return shadow(this, 'documentOutline', obj); |
|
}, |
|
readDocumentOutline: function Catalog_readDocumentOutline() { |
|
var xref = this.xref; |
|
var obj = this.catDict.get('Outlines'); |
|
var root = { items: [] }; |
|
if (isDict(obj)) { |
|
obj = obj.getRaw('First'); |
|
var processed = new RefSet(); |
|
if (isRef(obj)) { |
|
var queue = [{obj: obj, parent: root}]; |
|
// to avoid recursion keeping track of the items |
|
// in the processed dictionary |
|
processed.put(obj); |
|
while (queue.length > 0) { |
|
var i = queue.shift(); |
|
var outlineDict = xref.fetchIfRef(i.obj); |
|
if (outlineDict === null) { |
|
continue; |
|
} |
|
if (!outlineDict.has('Title')) { |
|
error('Invalid outline item'); |
|
} |
|
var actionDict = outlineDict.get('A'), dest = null, url = null; |
|
if (actionDict) { |
|
var destEntry = actionDict.get('D'); |
|
if (destEntry) { |
|
dest = destEntry; |
|
} else { |
|
var uriEntry = actionDict.get('URI'); |
|
if (isString(uriEntry) && isValidUrl(uriEntry, false)) { |
|
url = uriEntry; |
|
} |
|
} |
|
} else if (outlineDict.has('Dest')) { |
|
dest = outlineDict.getRaw('Dest'); |
|
if (isName(dest)) { |
|
dest = dest.name; |
|
} |
|
} |
|
var title = outlineDict.get('Title'); |
|
var outlineItem = { |
|
dest: dest, |
|
url: url, |
|
title: stringToPDFString(title), |
|
color: outlineDict.get('C') || [0, 0, 0], |
|
count: outlineDict.get('Count'), |
|
bold: !!(outlineDict.get('F') & 2), |
|
italic: !!(outlineDict.get('F') & 1), |
|
items: [] |
|
}; |
|
i.parent.items.push(outlineItem); |
|
obj = outlineDict.getRaw('First'); |
|
if (isRef(obj) && !processed.has(obj)) { |
|
queue.push({obj: obj, parent: outlineItem}); |
|
processed.put(obj); |
|
} |
|
obj = outlineDict.getRaw('Next'); |
|
if (isRef(obj) && !processed.has(obj)) { |
|
queue.push({obj: obj, parent: i.parent}); |
|
processed.put(obj); |
|
} |
|
} |
|
} |
|
} |
|
return (root.items.length > 0 ? root.items : null); |
|
}, |
|
get numPages() { |
|
var obj = this.toplevelPagesDict.get('Count'); |
|
assert( |
|
isInt(obj), |
|
'page count in top level pages object is not an integer' |
|
); |
|
// shadow the prototype getter |
|
return shadow(this, 'num', obj); |
|
}, |
|
get destinations() { |
|
function fetchDestination(dest) { |
|
return isDict(dest) ? dest.get('D') : dest; |
|
} |
|
|
|
var xref = this.xref; |
|
var dests = {}, nameTreeRef, nameDictionaryRef; |
|
var obj = this.catDict.get('Names'); |
|
if (obj && obj.has('Dests')) { |
|
nameTreeRef = obj.getRaw('Dests'); |
|
} else if (this.catDict.has('Dests')) { |
|
nameDictionaryRef = this.catDict.get('Dests'); |
|
} |
|
|
|
if (nameDictionaryRef) { |
|
// reading simple destination dictionary |
|
obj = nameDictionaryRef; |
|
obj.forEach(function catalogForEach(key, value) { |
|
if (!value) { |
|
return; |
|
} |
|
dests[key] = fetchDestination(value); |
|
}); |
|
} |
|
if (nameTreeRef) { |
|
var nameTree = new NameTree(nameTreeRef, xref); |
|
var names = nameTree.getAll(); |
|
for (var name in names) { |
|
if (!names.hasOwnProperty(name)) { |
|
continue; |
|
} |
|
dests[name] = fetchDestination(names[name]); |
|
} |
|
} |
|
return shadow(this, 'destinations', dests); |
|
}, |
|
getDestination: function Catalog_getDestination(destinationId) { |
|
function fetchDestination(dest) { |
|
return isDict(dest) ? dest.get('D') : dest; |
|
} |
|
|
|
var xref = this.xref; |
|
var dest = null, nameTreeRef, nameDictionaryRef; |
|
var obj = this.catDict.get('Names'); |
|
if (obj && obj.has('Dests')) { |
|
nameTreeRef = obj.getRaw('Dests'); |
|
} else if (this.catDict.has('Dests')) { |
|
nameDictionaryRef = this.catDict.get('Dests'); |
|
} |
|
|
|
if (nameDictionaryRef) { // Simple destination dictionary. |
|
var value = nameDictionaryRef.get(destinationId); |
|
if (value) { |
|
dest = fetchDestination(value); |
|
} |
|
} |
|
if (nameTreeRef) { |
|
var nameTree = new NameTree(nameTreeRef, xref); |
|
dest = fetchDestination(nameTree.get(destinationId)); |
|
} |
|
return dest; |
|
}, |
|
|
|
get pageLabels() { |
|
var obj = null; |
|
try { |
|
obj = this.readPageLabels(); |
|
} catch (ex) { |
|
if (ex instanceof MissingDataException) { |
|
throw ex; |
|
} |
|
warn('Unable to read page labels.'); |
|
} |
|
return shadow(this, 'pageLabels', obj); |
|
}, |
|
readPageLabels: function Catalog_readPageLabels() { |
|
var obj = this.catDict.getRaw('PageLabels'); |
|
if (!obj) { |
|
return null; |
|
} |
|
var pageLabels = new Array(this.numPages); |
|
var style = null; |
|
var prefix = ''; |
|
var start = 1; |
|
|
|
var numberTree = new NumberTree(obj, this.xref); |
|
var nums = numberTree.getAll(); |
|
var currentLabel = '', currentIndex = 1; |
|
|
|
for (var i = 0, ii = this.numPages; i < ii; i++) { |
|
if (nums.hasOwnProperty(i)) { |
|
var labelDict = nums[i]; |
|
assert(isDict(labelDict), 'The PageLabel is not a dictionary.'); |
|
|
|
var type = labelDict.get('Type'); |
|
assert(!type || (isName(type) && type.name === 'PageLabel'), |
|
'Invalid type in PageLabel dictionary.'); |
|
|
|
var s = labelDict.get('S'); |
|
assert(!s || isName(s), 'Invalid style in PageLabel dictionary.'); |
|
style = (s ? s.name : null); |
|
|
|
prefix = labelDict.get('P') || ''; |
|
assert(isString(prefix), 'Invalid prefix in PageLabel dictionary.'); |
|
|
|
start = labelDict.get('St') || 1; |
|
assert(isInt(start), 'Invalid start in PageLabel dictionary.'); |
|
currentIndex = start; |
|
} |
|
|
|
switch (style) { |
|
case 'D': |
|
currentLabel = currentIndex; |
|
break; |
|
case 'R': |
|
case 'r': |
|
currentLabel = Util.toRoman(currentIndex, style === 'r'); |
|
break; |
|
case 'A': |
|
case 'a': |
|
var LIMIT = 26; // Use only the characters A--Z, or a--z. |
|
var A_UPPER_CASE = 0x41, A_LOWER_CASE = 0x61; |
|
|
|
var baseCharCode = (style === 'a' ? A_LOWER_CASE : A_UPPER_CASE); |
|
var letterIndex = currentIndex - 1; |
|
var character = String.fromCharCode(baseCharCode + |
|
(letterIndex % LIMIT)); |
|
var charBuf = []; |
|
for (var j = 0, jj = (letterIndex / LIMIT) | 0; j <= jj; j++) { |
|
charBuf.push(character); |
|
} |
|
currentLabel = charBuf.join(''); |
|
break; |
|
default: |
|
assert(!style, |
|
'Invalid style "' + style + '" in PageLabel dictionary.'); |
|
} |
|
pageLabels[i] = prefix + currentLabel; |
|
|
|
currentLabel = ''; |
|
currentIndex++; |
|
} |
|
return pageLabels; |
|
}, |
|
|
|
get attachments() { |
|
var xref = this.xref; |
|
var attachments = null, nameTreeRef; |
|
var obj = this.catDict.get('Names'); |
|
if (obj) { |
|
nameTreeRef = obj.getRaw('EmbeddedFiles'); |
|
} |
|
|
|
if (nameTreeRef) { |
|
var nameTree = new NameTree(nameTreeRef, xref); |
|
var names = nameTree.getAll(); |
|
for (var name in names) { |
|
if (!names.hasOwnProperty(name)) { |
|
continue; |
|
} |
|
var fs = new FileSpec(names[name], xref); |
|
if (!attachments) { |
|
attachments = {}; |
|
} |
|
attachments[stringToPDFString(name)] = fs.serializable; |
|
} |
|
} |
|
return shadow(this, 'attachments', attachments); |
|
}, |
|
get javaScript() { |
|
var xref = this.xref; |
|
var obj = this.catDict.get('Names'); |
|
|
|
var javaScript = []; |
|
function appendIfJavaScriptDict(jsDict) { |
|
var type = jsDict.get('S'); |
|
if (!isName(type) || type.name !== 'JavaScript') { |
|
return; |
|
} |
|
var js = jsDict.get('JS'); |
|
if (isStream(js)) { |
|
js = bytesToString(js.getBytes()); |
|
} else if (!isString(js)) { |
|
return; |
|
} |
|
javaScript.push(stringToPDFString(js)); |
|
} |
|
if (obj && obj.has('JavaScript')) { |
|
var nameTree = new NameTree(obj.getRaw('JavaScript'), xref); |
|
var names = nameTree.getAll(); |
|
for (var name in names) { |
|
if (!names.hasOwnProperty(name)) { |
|
continue; |
|
} |
|
// We don't really use the JavaScript right now. This code is |
|
// defensive so we don't cause errors on document load. |
|
var jsDict = names[name]; |
|
if (isDict(jsDict)) { |
|
appendIfJavaScriptDict(jsDict); |
|
} |
|
} |
|
} |
|
|
|
// Append OpenAction actions to javaScript array |
|
var openactionDict = this.catDict.get('OpenAction'); |
|
if (isDict(openactionDict, 'Action')) { |
|
var actionType = openactionDict.get('S'); |
|
if (isName(actionType) && actionType.name === 'Named') { |
|
// The named Print action is not a part of the PDF 1.7 specification, |
|
// but is supported by many PDF readers/writers (including Adobe's). |
|
var action = openactionDict.get('N'); |
|
if (isName(action) && action.name === 'Print') { |
|
javaScript.push('print({});'); |
|
} |
|
} else { |
|
appendIfJavaScriptDict(openactionDict); |
|
} |
|
} |
|
|
|
return shadow(this, 'javaScript', javaScript); |
|
}, |
|
|
|
cleanup: function Catalog_cleanup() { |
|
var promises = []; |
|
this.fontCache.forEach(function (promise) { |
|
promises.push(promise); |
|
}); |
|
return Promise.all(promises).then(function (translatedFonts) { |
|
for (var i = 0, ii = translatedFonts.length; i < ii; i++) { |
|
var font = translatedFonts[i].dict; |
|
delete font.translated; |
|
} |
|
this.fontCache.clear(); |
|
}.bind(this)); |
|
}, |
|
|
|
getPage: function Catalog_getPage(pageIndex) { |
|
if (!(pageIndex in this.pagePromises)) { |
|
this.pagePromises[pageIndex] = this.getPageDict(pageIndex).then( |
|
function (a) { |
|
var dict = a[0]; |
|
var ref = a[1]; |
|
return this.pageFactory.createPage(pageIndex, dict, ref, |
|
this.fontCache); |
|
}.bind(this) |
|
); |
|
} |
|
return this.pagePromises[pageIndex]; |
|
}, |
|
|
|
getPageDict: function Catalog_getPageDict(pageIndex) { |
|
var capability = createPromiseCapability(); |
|
var nodesToVisit = [this.catDict.getRaw('Pages')]; |
|
var currentPageIndex = 0; |
|
var xref = this.xref; |
|
var checkAllKids = false; |
|
|
|
function next() { |
|
while (nodesToVisit.length) { |
|
var currentNode = nodesToVisit.pop(); |
|
|
|
if (isRef(currentNode)) { |
|
xref.fetchAsync(currentNode).then(function (obj) { |
|
if (isDict(obj, 'Page') || (isDict(obj) && !obj.has('Kids'))) { |
|
if (pageIndex === currentPageIndex) { |
|
capability.resolve([obj, currentNode]); |
|
} else { |
|
currentPageIndex++; |
|
next(); |
|
} |
|
return; |
|
} |
|
nodesToVisit.push(obj); |
|
next(); |
|
}, capability.reject); |
|
return; |
|
} |
|
|
|
// Must be a child page dictionary. |
|
assert( |
|
isDict(currentNode), |
|
'page dictionary kid reference points to wrong type of object' |
|
); |
|
var count = currentNode.get('Count'); |
|
// If the current node doesn't have any children, avoid getting stuck |
|
// in an empty node further down in the tree (see issue5644.pdf). |
|
if (count === 0) { |
|
checkAllKids = true; |
|
} |
|
// Skip nodes where the page can't be. |
|
if (currentPageIndex + count <= pageIndex) { |
|
currentPageIndex += count; |
|
continue; |
|
} |
|
|
|
var kids = currentNode.get('Kids'); |
|
assert(isArray(kids), 'page dictionary kids object is not an array'); |
|
if (!checkAllKids && count === kids.length) { |
|
// Nodes that don't have the page have been skipped and this is the |
|
// bottom of the tree which means the page requested must be a |
|
// descendant of this pages node. Ideally we would just resolve the |
|
// promise with the page ref here, but there is the case where more |
|
// pages nodes could link to single a page (see issue 3666 pdf). To |
|
// handle this push it back on the queue so if it is a pages node it |
|
// will be descended into. |
|
nodesToVisit = [kids[pageIndex - currentPageIndex]]; |
|
currentPageIndex = pageIndex; |
|
continue; |
|
} else { |
|
for (var last = kids.length - 1; last >= 0; last--) { |
|
nodesToVisit.push(kids[last]); |
|
} |
|
} |
|
} |
|
capability.reject('Page index ' + pageIndex + ' not found.'); |
|
} |
|
next(); |
|
return capability.promise; |
|
}, |
|
|
|
getPageIndex: function Catalog_getPageIndex(ref) { |
|
// The page tree nodes have the count of all the leaves below them. To get |
|
// how many pages are before we just have to walk up the tree and keep |
|
// adding the count of siblings to the left of the node. |
|
var xref = this.xref; |
|
function pagesBeforeRef(kidRef) { |
|
var total = 0; |
|
var parentRef; |
|
return xref.fetchAsync(kidRef).then(function (node) { |
|
if (!node) { |
|
return null; |
|
} |
|
parentRef = node.getRaw('Parent'); |
|
return node.getAsync('Parent'); |
|
}).then(function (parent) { |
|
if (!parent) { |
|
return null; |
|
} |
|
return parent.getAsync('Kids'); |
|
}).then(function (kids) { |
|
if (!kids) { |
|
return null; |
|
} |
|
var kidPromises = []; |
|
var found = false; |
|
for (var i = 0; i < kids.length; i++) { |
|
var kid = kids[i]; |
|
assert(isRef(kid), 'kids must be a ref'); |
|
if (kid.num === kidRef.num) { |
|
found = true; |
|
break; |
|
} |
|
kidPromises.push(xref.fetchAsync(kid).then(function (kid) { |
|
if (kid.has('Count')) { |
|
var count = kid.get('Count'); |
|
total += count; |
|
} else { // page leaf node |
|
total++; |
|
} |
|
})); |
|
} |
|
if (!found) { |
|
error('kid ref not found in parents kids'); |
|
} |
|
return Promise.all(kidPromises).then(function () { |
|
return [total, parentRef]; |
|
}); |
|
}); |
|
} |
|
|
|
var total = 0; |
|
function next(ref) { |
|
return pagesBeforeRef(ref).then(function (args) { |
|
if (!args) { |
|
return total; |
|
} |
|
var count = args[0]; |
|
var parentRef = args[1]; |
|
total += count; |
|
return next(parentRef); |
|
}); |
|
} |
|
|
|
return next(ref); |
|
} |
|
}; |
|
|
|
return Catalog; |
|
})(); |
|
|
|
var XRef = (function XRefClosure() { |
|
function XRef(stream, password) { |
|
this.stream = stream; |
|
this.entries = []; |
|
this.xrefstms = {}; |
|
// prepare the XRef cache |
|
this.cache = []; |
|
this.password = password; |
|
this.stats = { |
|
streamTypes: [], |
|
fontTypes: [] |
|
}; |
|
} |
|
|
|
XRef.prototype = { |
|
setStartXRef: function XRef_setStartXRef(startXRef) { |
|
// Store the starting positions of xref tables as we process them |
|
// so we can recover from missing data errors |
|
this.startXRefQueue = [startXRef]; |
|
}, |
|
|
|
parse: function XRef_parse(recoveryMode) { |
|
var trailerDict; |
|
if (!recoveryMode) { |
|
trailerDict = this.readXRef(); |
|
} else { |
|
warn('Indexing all PDF objects'); |
|
trailerDict = this.indexObjects(); |
|
} |
|
trailerDict.assignXref(this); |
|
this.trailer = trailerDict; |
|
var encrypt = trailerDict.get('Encrypt'); |
|
if (encrypt) { |
|
var ids = trailerDict.get('ID'); |
|
var fileId = (ids && ids.length) ? ids[0] : ''; |
|
this.encrypt = new CipherTransformFactory(encrypt, fileId, |
|
this.password); |
|
} |
|
|
|
// get the root dictionary (catalog) object |
|
if (!(this.root = trailerDict.get('Root'))) { |
|
error('Invalid root reference'); |
|
} |
|
}, |
|
|
|
processXRefTable: function XRef_processXRefTable(parser) { |
|
if (!('tableState' in this)) { |
|
// Stores state of the table as we process it so we can resume |
|
// from middle of table in case of missing data error |
|
this.tableState = { |
|
entryNum: 0, |
|
streamPos: parser.lexer.stream.pos, |
|
parserBuf1: parser.buf1, |
|
parserBuf2: parser.buf2 |
|
}; |
|
} |
|
|
|
var obj = this.readXRefTable(parser); |
|
|
|
// Sanity check |
|
if (!isCmd(obj, 'trailer')) { |
|
error('Invalid XRef table: could not find trailer dictionary'); |
|
} |
|
// Read trailer dictionary, e.g. |
|
// trailer |
|
// << /Size 22 |
|
// /Root 20R |
|
// /Info 10R |
|
// /ID [ <81b14aafa313db63dbd6f981e49f94f4> ] |
|
// >> |
|
// The parser goes through the entire stream << ... >> and provides |
|
// a getter interface for the key-value table |
|
var dict = parser.getObj(); |
|
|
|
// The pdflib PDF generator can generate a nested trailer dictionary |
|
if (!isDict(dict) && dict.dict) { |
|
dict = dict.dict; |
|
} |
|
if (!isDict(dict)) { |
|
error('Invalid XRef table: could not parse trailer dictionary'); |
|
} |
|
delete this.tableState; |
|
|
|
return dict; |
|
}, |
|
|
|
readXRefTable: function XRef_readXRefTable(parser) { |
|
// Example of cross-reference table: |
|
// xref |
|
// 0 1 <-- subsection header (first obj #, obj count) |
|
// 0000000000 65535 f <-- actual object (offset, generation #, f/n) |
|
// 23 2 <-- subsection header ... and so on ... |
|
// 0000025518 00002 n |
|
// 0000025635 00000 n |
|
// trailer |
|
// ... |
|
|
|
var stream = parser.lexer.stream; |
|
var tableState = this.tableState; |
|
stream.pos = tableState.streamPos; |
|
parser.buf1 = tableState.parserBuf1; |
|
parser.buf2 = tableState.parserBuf2; |
|
|
|
// Outer loop is over subsection headers |
|
var obj; |
|
|
|
while (true) { |
|
if (!('firstEntryNum' in tableState) || !('entryCount' in tableState)) { |
|
if (isCmd(obj = parser.getObj(), 'trailer')) { |
|
break; |
|
} |
|
tableState.firstEntryNum = obj; |
|
tableState.entryCount = parser.getObj(); |
|
} |
|
|
|
var first = tableState.firstEntryNum; |
|
var count = tableState.entryCount; |
|
if (!isInt(first) || !isInt(count)) { |
|
error('Invalid XRef table: wrong types in subsection header'); |
|
} |
|
// Inner loop is over objects themselves |
|
for (var i = tableState.entryNum; i < count; i++) { |
|
tableState.streamPos = stream.pos; |
|
tableState.entryNum = i; |
|
tableState.parserBuf1 = parser.buf1; |
|
tableState.parserBuf2 = parser.buf2; |
|
|
|
var entry = {}; |
|
entry.offset = parser.getObj(); |
|
entry.gen = parser.getObj(); |
|
var type = parser.getObj(); |
|
|
|
if (isCmd(type, 'f')) { |
|
entry.free = true; |
|
} else if (isCmd(type, 'n')) { |
|
entry.uncompressed = true; |
|
} |
|
|
|
// Validate entry obj |
|
if (!isInt(entry.offset) || !isInt(entry.gen) || |
|
!(entry.free || entry.uncompressed)) { |
|
error('Invalid entry in XRef subsection: ' + first + ', ' + count); |
|
} |
|
|
|
if (!this.entries[i + first]) { |
|
this.entries[i + first] = entry; |
|
} |
|
} |
|
|
|
tableState.entryNum = 0; |
|
tableState.streamPos = stream.pos; |
|
tableState.parserBuf1 = parser.buf1; |
|
tableState.parserBuf2 = parser.buf2; |
|
delete tableState.firstEntryNum; |
|
delete tableState.entryCount; |
|
} |
|
|
|
// Per issue 3248: hp scanners generate bad XRef |
|
if (first === 1 && this.entries[1] && this.entries[1].free) { |
|
// shifting the entries |
|
this.entries.shift(); |
|
} |
|
|
|
// Sanity check: as per spec, first object must be free |
|
if (this.entries[0] && !this.entries[0].free) { |
|
error('Invalid XRef table: unexpected first object'); |
|
} |
|
return obj; |
|
}, |
|
|
|
processXRefStream: function XRef_processXRefStream(stream) { |
|
if (!('streamState' in this)) { |
|
// Stores state of the stream as we process it so we can resume |
|
// from middle of stream in case of missing data error |
|
var streamParameters = stream.dict; |
|
var byteWidths = streamParameters.get('W'); |
|
var range = streamParameters.get('Index'); |
|
if (!range) { |
|
range = [0, streamParameters.get('Size')]; |
|
} |
|
|
|
this.streamState = { |
|
entryRanges: range, |
|
byteWidths: byteWidths, |
|
entryNum: 0, |
|
streamPos: stream.pos |
|
}; |
|
} |
|
this.readXRefStream(stream); |
|
delete this.streamState; |
|
|
|
return stream.dict; |
|
}, |
|
|
|
readXRefStream: function XRef_readXRefStream(stream) { |
|
var i, j; |
|
var streamState = this.streamState; |
|
stream.pos = streamState.streamPos; |
|
|
|
var byteWidths = streamState.byteWidths; |
|
var typeFieldWidth = byteWidths[0]; |
|
var offsetFieldWidth = byteWidths[1]; |
|
var generationFieldWidth = byteWidths[2]; |
|
|
|
var entryRanges = streamState.entryRanges; |
|
while (entryRanges.length > 0) { |
|
var first = entryRanges[0]; |
|
var n = entryRanges[1]; |
|
|
|
if (!isInt(first) || !isInt(n)) { |
|
error('Invalid XRef range fields: ' + first + ', ' + n); |
|
} |
|
if (!isInt(typeFieldWidth) || !isInt(offsetFieldWidth) || |
|
!isInt(generationFieldWidth)) { |
|
error('Invalid XRef entry fields length: ' + first + ', ' + n); |
|
} |
|
for (i = streamState.entryNum; i < n; ++i) { |
|
streamState.entryNum = i; |
|
streamState.streamPos = stream.pos; |
|
|
|
var type = 0, offset = 0, generation = 0; |
|
for (j = 0; j < typeFieldWidth; ++j) { |
|
type = (type << 8) | stream.getByte(); |
|
} |
|
// if type field is absent, its default value is 1 |
|
if (typeFieldWidth === 0) { |
|
type = 1; |
|
} |
|
for (j = 0; j < offsetFieldWidth; ++j) { |
|
offset = (offset << 8) | stream.getByte(); |
|
} |
|
for (j = 0; j < generationFieldWidth; ++j) { |
|
generation = (generation << 8) | stream.getByte(); |
|
} |
|
var entry = {}; |
|
entry.offset = offset; |
|
entry.gen = generation; |
|
switch (type) { |
|
case 0: |
|
entry.free = true; |
|
break; |
|
case 1: |
|
entry.uncompressed = true; |
|
break; |
|
case 2: |
|
break; |
|
default: |
|
error('Invalid XRef entry type: ' + type); |
|
} |
|
if (!this.entries[first + i]) { |
|
this.entries[first + i] = entry; |
|
} |
|
} |
|
|
|
streamState.entryNum = 0; |
|
streamState.streamPos = stream.pos; |
|
entryRanges.splice(0, 2); |
|
} |
|
}, |
|
|
|
indexObjects: function XRef_indexObjects() { |
|
// Simple scan through the PDF content to find objects, |
|
// trailers and XRef streams. |
|
var TAB = 0x9, LF = 0xA, CR = 0xD, SPACE = 0x20; |
|
var PERCENT = 0x25, LT = 0x3C; |
|
|
|
function readToken(data, offset) { |
|
var token = '', ch = data[offset]; |
|
while (ch !== LF && ch !== CR && ch !== LT) { |
|
if (++offset >= data.length) { |
|
break; |
|
} |
|
token += String.fromCharCode(ch); |
|
ch = data[offset]; |
|
} |
|
return token; |
|
} |
|
function skipUntil(data, offset, what) { |
|
var length = what.length, dataLength = data.length; |
|
var skipped = 0; |
|
// finding byte sequence |
|
while (offset < dataLength) { |
|
var i = 0; |
|
while (i < length && data[offset + i] === what[i]) { |
|
++i; |
|
} |
|
if (i >= length) { |
|
break; // sequence found |
|
} |
|
offset++; |
|
skipped++; |
|
} |
|
return skipped; |
|
} |
|
var objRegExp = /^(\d+)\s+(\d+)\s+obj\b/; |
|
var trailerBytes = new Uint8Array([116, 114, 97, 105, 108, 101, 114]); |
|
var startxrefBytes = new Uint8Array([115, 116, 97, 114, 116, 120, 114, |
|
101, 102]); |
|
var endobjBytes = new Uint8Array([101, 110, 100, 111, 98, 106]); |
|
var xrefBytes = new Uint8Array([47, 88, 82, 101, 102]); |
|
|
|
// Clear out any existing entries, since they may be bogus. |
|
this.entries.length = 0; |
|
|
|
var stream = this.stream; |
|
stream.pos = 0; |
|
var buffer = stream.getBytes(); |
|
var position = stream.start, length = buffer.length; |
|
var trailers = [], xrefStms = []; |
|
while (position < length) { |
|
var ch = buffer[position]; |
|
if (ch === TAB || ch === LF || ch === CR || ch === SPACE) { |
|
++position; |
|
continue; |
|
} |
|
if (ch === PERCENT) { // %-comment |
|
do { |
|
++position; |
|
if (position >= length) { |
|
break; |
|
} |
|
ch = buffer[position]; |
|
} while (ch !== LF && ch !== CR); |
|
continue; |
|
} |
|
var token = readToken(buffer, position); |
|
var m; |
|
if (token.indexOf('xref') === 0 && |
|
(token.length === 4 || /\s/.test(token[4]))) { |
|
position += skipUntil(buffer, position, trailerBytes); |
|
trailers.push(position); |
|
position += skipUntil(buffer, position, startxrefBytes); |
|
} else if ((m = objRegExp.exec(token))) { |
|
if (typeof this.entries[m[1]] === 'undefined') { |
|
this.entries[m[1]] = { |
|
offset: position - stream.start, |
|
gen: m[2] | 0, |
|
uncompressed: true |
|
}; |
|
} |
|
var contentLength = skipUntil(buffer, position, endobjBytes) + 7; |
|
var content = buffer.subarray(position, position + contentLength); |
|
|
|
// checking XRef stream suspect |
|
// (it shall have '/XRef' and next char is not a letter) |
|
var xrefTagOffset = skipUntil(content, 0, xrefBytes); |
|
if (xrefTagOffset < contentLength && |
|
content[xrefTagOffset + 5] < 64) { |
|
xrefStms.push(position - stream.start); |
|
this.xrefstms[position - stream.start] = 1; // Avoid recursion |
|
} |
|
|
|
position += contentLength; |
|
} else if (token.indexOf('trailer') === 0 && |
|
(token.length === 7 || /\s/.test(token[7]))) { |
|
trailers.push(position); |
|
position += skipUntil(buffer, position, startxrefBytes); |
|
} else { |
|
position += token.length + 1; |
|
} |
|
} |
|
// reading XRef streams |
|
var i, ii; |
|
for (i = 0, ii = xrefStms.length; i < ii; ++i) { |
|
this.startXRefQueue.push(xrefStms[i]); |
|
this.readXRef(/* recoveryMode */ true); |
|
} |
|
// finding main trailer |
|
var dict; |
|
for (i = 0, ii = trailers.length; i < ii; ++i) { |
|
stream.pos = trailers[i]; |
|
var parser = new Parser(new Lexer(stream), true, this); |
|
var obj = parser.getObj(); |
|
if (!isCmd(obj, 'trailer')) { |
|
continue; |
|
} |
|
// read the trailer dictionary |
|
if (!isDict(dict = parser.getObj())) { |
|
continue; |
|
} |
|
// taking the first one with 'ID' |
|
if (dict.has('ID')) { |
|
return dict; |
|
} |
|
} |
|
// no tailer with 'ID', taking last one (if exists) |
|
if (dict) { |
|
return dict; |
|
} |
|
// nothing helps |
|
// calling error() would reject worker with an UnknownErrorException. |
|
throw new InvalidPDFException('Invalid PDF structure'); |
|
}, |
|
|
|
readXRef: function XRef_readXRef(recoveryMode) { |
|
var stream = this.stream; |
|
|
|
try { |
|
while (this.startXRefQueue.length) { |
|
var startXRef = this.startXRefQueue[0]; |
|
|
|
stream.pos = startXRef + stream.start; |
|
|
|
var parser = new Parser(new Lexer(stream), true, this); |
|
var obj = parser.getObj(); |
|
var dict; |
|
|
|
// Get dictionary |
|
if (isCmd(obj, 'xref')) { |
|
// Parse end-of-file XRef |
|
dict = this.processXRefTable(parser); |
|
if (!this.topDict) { |
|
this.topDict = dict; |
|
} |
|
|
|
// Recursively get other XRefs 'XRefStm', if any |
|
obj = dict.get('XRefStm'); |
|
if (isInt(obj)) { |
|
var pos = obj; |
|
// ignore previously loaded xref streams |
|
// (possible infinite recursion) |
|
if (!(pos in this.xrefstms)) { |
|
this.xrefstms[pos] = 1; |
|
this.startXRefQueue.push(pos); |
|
} |
|
} |
|
} else if (isInt(obj)) { |
|
// Parse in-stream XRef |
|
if (!isInt(parser.getObj()) || |
|
!isCmd(parser.getObj(), 'obj') || |
|
!isStream(obj = parser.getObj())) { |
|
error('Invalid XRef stream'); |
|
} |
|
dict = this.processXRefStream(obj); |
|
if (!this.topDict) { |
|
this.topDict = dict; |
|
} |
|
if (!dict) { |
|
error('Failed to read XRef stream'); |
|
} |
|
} else { |
|
error('Invalid XRef stream header'); |
|
} |
|
|
|
// Recursively get previous dictionary, if any |
|
obj = dict.get('Prev'); |
|
if (isInt(obj)) { |
|
this.startXRefQueue.push(obj); |
|
} else if (isRef(obj)) { |
|
// The spec says Prev must not be a reference, i.e. "/Prev NNN" |
|
// This is a fallback for non-compliant PDFs, i.e. "/Prev NNN 0 R" |
|
this.startXRefQueue.push(obj.num); |
|
} |
|
|
|
this.startXRefQueue.shift(); |
|
} |
|
|
|
return this.topDict; |
|
} catch (e) { |
|
if (e instanceof MissingDataException) { |
|
throw e; |
|
} |
|
info('(while reading XRef): ' + e); |
|
} |
|
|
|
if (recoveryMode) { |
|
return; |
|
} |
|
throw new XRefParseException(); |
|
}, |
|
|
|
getEntry: function XRef_getEntry(i) { |
|
var xrefEntry = this.entries[i]; |
|
if (xrefEntry && !xrefEntry.free && xrefEntry.offset) { |
|
return xrefEntry; |
|
} |
|
return null; |
|
}, |
|
|
|
fetchIfRef: function XRef_fetchIfRef(obj) { |
|
if (!isRef(obj)) { |
|
return obj; |
|
} |
|
return this.fetch(obj); |
|
}, |
|
|
|
fetch: function XRef_fetch(ref, suppressEncryption) { |
|
assert(isRef(ref), 'ref object is not a reference'); |
|
var num = ref.num; |
|
if (num in this.cache) { |
|
var cacheEntry = this.cache[num]; |
|
return cacheEntry; |
|
} |
|
|
|
var xrefEntry = this.getEntry(num); |
|
|
|
// the referenced entry can be free |
|
if (xrefEntry === null) { |
|
return (this.cache[num] = null); |
|
} |
|
|
|
if (xrefEntry.uncompressed) { |
|
xrefEntry = this.fetchUncompressed(ref, xrefEntry, suppressEncryption); |
|
} else { |
|
xrefEntry = this.fetchCompressed(xrefEntry, suppressEncryption); |
|
} |
|
if (isDict(xrefEntry)){ |
|
xrefEntry.objId = ref.toString(); |
|
} else if (isStream(xrefEntry)) { |
|
xrefEntry.dict.objId = ref.toString(); |
|
} |
|
return xrefEntry; |
|
}, |
|
|
|
fetchUncompressed: function XRef_fetchUncompressed(ref, xrefEntry, |
|
suppressEncryption) { |
|
var gen = ref.gen; |
|
var num = ref.num; |
|
if (xrefEntry.gen !== gen) { |
|
error('inconsistent generation in XRef'); |
|
} |
|
var stream = this.stream.makeSubStream(xrefEntry.offset + |
|
this.stream.start); |
|
var parser = new Parser(new Lexer(stream), true, this); |
|
var obj1 = parser.getObj(); |
|
var obj2 = parser.getObj(); |
|
var obj3 = parser.getObj(); |
|
if (!isInt(obj1) || parseInt(obj1, 10) !== num || |
|
!isInt(obj2) || parseInt(obj2, 10) !== gen || |
|
!isCmd(obj3)) { |
|
error('bad XRef entry'); |
|
} |
|
if (!isCmd(obj3, 'obj')) { |
|
// some bad PDFs use "obj1234" and really mean 1234 |
|
if (obj3.cmd.indexOf('obj') === 0) { |
|
num = parseInt(obj3.cmd.substring(3), 10); |
|
if (!isNaN(num)) { |
|
return num; |
|
} |
|
} |
|
error('bad XRef entry'); |
|
} |
|
if (this.encrypt && !suppressEncryption) { |
|
xrefEntry = parser.getObj(this.encrypt.createCipherTransform(num, gen)); |
|
} else { |
|
xrefEntry = parser.getObj(); |
|
} |
|
if (!isStream(xrefEntry)) { |
|
this.cache[num] = xrefEntry; |
|
} |
|
return xrefEntry; |
|
}, |
|
|
|
fetchCompressed: function XRef_fetchCompressed(xrefEntry, |
|
suppressEncryption) { |
|
var tableOffset = xrefEntry.offset; |
|
var stream = this.fetch(new Ref(tableOffset, 0)); |
|
if (!isStream(stream)) { |
|
error('bad ObjStm stream'); |
|
} |
|
var first = stream.dict.get('First'); |
|
var n = stream.dict.get('N'); |
|
if (!isInt(first) || !isInt(n)) { |
|
error('invalid first and n parameters for ObjStm stream'); |
|
} |
|
var parser = new Parser(new Lexer(stream), false, this); |
|
parser.allowStreams = true; |
|
var i, entries = [], num, nums = []; |
|
// read the object numbers to populate cache |
|
for (i = 0; i < n; ++i) { |
|
num = parser.getObj(); |
|
if (!isInt(num)) { |
|
error('invalid object number in the ObjStm stream: ' + num); |
|
} |
|
nums.push(num); |
|
var offset = parser.getObj(); |
|
if (!isInt(offset)) { |
|
error('invalid object offset in the ObjStm stream: ' + offset); |
|
} |
|
} |
|
// read stream objects for cache |
|
for (i = 0; i < n; ++i) { |
|
entries.push(parser.getObj()); |
|
num = nums[i]; |
|
var entry = this.entries[num]; |
|
if (entry && entry.offset === tableOffset && entry.gen === i) { |
|
this.cache[num] = entries[i]; |
|
} |
|
} |
|
xrefEntry = entries[xrefEntry.gen]; |
|
if (xrefEntry === undefined) { |
|
error('bad XRef entry for compressed object'); |
|
} |
|
return xrefEntry; |
|
}, |
|
|
|
fetchIfRefAsync: function XRef_fetchIfRefAsync(obj) { |
|
if (!isRef(obj)) { |
|
return Promise.resolve(obj); |
|
} |
|
return this.fetchAsync(obj); |
|
}, |
|
|
|
fetchAsync: function XRef_fetchAsync(ref, suppressEncryption) { |
|
var streamManager = this.stream.manager; |
|
var xref = this; |
|
return new Promise(function tryFetch(resolve, reject) { |
|
try { |
|
resolve(xref.fetch(ref, suppressEncryption)); |
|
} catch (e) { |
|
if (e instanceof MissingDataException) { |
|
streamManager.requestRange(e.begin, e.end).then(function () { |
|
tryFetch(resolve, reject); |
|
}, reject); |
|
return; |
|
} |
|
reject(e); |
|
} |
|
}); |
|
}, |
|
|
|
getCatalogObj: function XRef_getCatalogObj() { |
|
return this.root; |
|
} |
|
}; |
|
|
|
return XRef; |
|
})(); |
|
|
|
/** |
|
* A NameTree/NumberTree is like a Dict but has some advantageous properties, |
|
* see the specification (7.9.6 and 7.9.7) for additional details. |
|
* TODO: implement all the Dict functions and make this more efficient. |
|
*/ |
|
var NameOrNumberTree = (function NameOrNumberTreeClosure() { |
|
function NameOrNumberTree(root, xref) { |
|
throw new Error('Cannot initialize NameOrNumberTree.'); |
|
} |
|
|
|
NameOrNumberTree.prototype = { |
|
getAll: function NameOrNumberTree_getAll() { |
|
var dict = {}; |
|
if (!this.root) { |
|
return dict; |
|
} |
|
var xref = this.xref; |
|
// Reading Name/Number tree. |
|
var processed = new RefSet(); |
|
processed.put(this.root); |
|
var queue = [this.root]; |
|
while (queue.length > 0) { |
|
var i, n; |
|
var obj = xref.fetchIfRef(queue.shift()); |
|
if (!isDict(obj)) { |
|
continue; |
|
} |
|
if (obj.has('Kids')) { |
|
var kids = obj.get('Kids'); |
|
for (i = 0, n = kids.length; i < n; i++) { |
|
var kid = kids[i]; |
|
assert(!processed.has(kid), |
|
'Duplicate entry in "' + this._type + '" tree.'); |
|
queue.push(kid); |
|
processed.put(kid); |
|
} |
|
continue; |
|
} |
|
var entries = obj.get(this._type); |
|
if (isArray(entries)) { |
|
for (i = 0, n = entries.length; i < n; i += 2) { |
|
dict[xref.fetchIfRef(entries[i])] = xref.fetchIfRef(entries[i + 1]); |
|
} |
|
} |
|
} |
|
return dict; |
|
}, |
|
|
|
get: function NameOrNumberTree_get(key) { |
|
if (!this.root) { |
|
return null; |
|
} |
|
|
|
var xref = this.xref; |
|
var kidsOrEntries = xref.fetchIfRef(this.root); |
|
var loopCount = 0; |
|
var MAX_LEVELS = 10; |
|
var l, r, m; |
|
|
|
// Perform a binary search to quickly find the entry that |
|
// contains the key we are looking for. |
|
while (kidsOrEntries.has('Kids')) { |
|
if (++loopCount > MAX_LEVELS) { |
|
warn('Search depth limit reached for "' + this._type + '" tree.'); |
|
return null; |
|
} |
|
|
|
var kids = kidsOrEntries.get('Kids'); |
|
if (!isArray(kids)) { |
|
return null; |
|
} |
|
|
|
l = 0; |
|
r = kids.length - 1; |
|
while (l <= r) { |
|
m = (l + r) >> 1; |
|
var kid = xref.fetchIfRef(kids[m]); |
|
var limits = kid.get('Limits'); |
|
|
|
if (key < xref.fetchIfRef(limits[0])) { |
|
r = m - 1; |
|
} else if (key > xref.fetchIfRef(limits[1])) { |
|
l = m + 1; |
|
} else { |
|
kidsOrEntries = xref.fetchIfRef(kids[m]); |
|
break; |
|
} |
|
} |
|
if (l > r) { |
|
return null; |
|
} |
|
} |
|
|
|
// If we get here, then we have found the right entry. Now go through the |
|
// entries in the dictionary until we find the key we're looking for. |
|
var entries = kidsOrEntries.get(this._type); |
|
if (isArray(entries)) { |
|
// Perform a binary search to reduce the lookup time. |
|
l = 0; |
|
r = entries.length - 2; |
|
while (l <= r) { |
|
// Check only even indices (0, 2, 4, ...) because the |
|
// odd indices contain the actual data. |
|
m = (l + r) & ~1; |
|
var currentKey = xref.fetchIfRef(entries[m]); |
|
if (key < currentKey) { |
|
r = m - 2; |
|
} else if (key > currentKey) { |
|
l = m + 2; |
|
} else { |
|
return xref.fetchIfRef(entries[m + 1]); |
|
} |
|
} |
|
} |
|
return null; |
|
} |
|
}; |
|
return NameOrNumberTree; |
|
})(); |
|
|
|
var NameTree = (function NameTreeClosure() { |
|
function NameTree(root, xref) { |
|
this.root = root; |
|
this.xref = xref; |
|
this._type = 'Names'; |
|
} |
|
|
|
Util.inherit(NameTree, NameOrNumberTree, {}); |
|
|
|
return NameTree; |
|
})(); |
|
|
|
var NumberTree = (function NumberTreeClosure() { |
|
function NumberTree(root, xref) { |
|
this.root = root; |
|
this.xref = xref; |
|
this._type = 'Nums'; |
|
} |
|
|
|
Util.inherit(NumberTree, NameOrNumberTree, {}); |
|
|
|
return NumberTree; |
|
})(); |
|
|
|
/** |
|
* "A PDF file can refer to the contents of another file by using a File |
|
* Specification (PDF 1.1)", see the spec (7.11) for more details. |
|
* NOTE: Only embedded files are supported (as part of the attachments support) |
|
* TODO: support the 'URL' file system (with caching if !/V), portable |
|
* collections attributes and related files (/RF) |
|
*/ |
|
var FileSpec = (function FileSpecClosure() { |
|
function FileSpec(root, xref) { |
|
if (!root || !isDict(root)) { |
|
return; |
|
} |
|
this.xref = xref; |
|
this.root = root; |
|
if (root.has('FS')) { |
|
this.fs = root.get('FS'); |
|
} |
|
this.description = root.has('Desc') ? |
|
stringToPDFString(root.get('Desc')) : |
|
''; |
|
if (root.has('RF')) { |
|
warn('Related file specifications are not supported'); |
|
} |
|
this.contentAvailable = true; |
|
if (!root.has('EF')) { |
|
this.contentAvailable = false; |
|
warn('Non-embedded file specifications are not supported'); |
|
} |
|
} |
|
|
|
function pickPlatformItem(dict) { |
|
// Look for the filename in this order: |
|
// UF, F, Unix, Mac, DOS |
|
if (dict.has('UF')) { |
|
return dict.get('UF'); |
|
} else if (dict.has('F')) { |
|
return dict.get('F'); |
|
} else if (dict.has('Unix')) { |
|
return dict.get('Unix'); |
|
} else if (dict.has('Mac')) { |
|
return dict.get('Mac'); |
|
} else if (dict.has('DOS')) { |
|
return dict.get('DOS'); |
|
} else { |
|
return null; |
|
} |
|
} |
|
|
|
FileSpec.prototype = { |
|
get filename() { |
|
if (!this._filename && this.root) { |
|
var filename = pickPlatformItem(this.root) || 'unnamed'; |
|
this._filename = stringToPDFString(filename). |
|
replace(/\\\\/g, '\\'). |
|
replace(/\\\//g, '/'). |
|
replace(/\\/g, '/'); |
|
} |
|
return this._filename; |
|
}, |
|
get content() { |
|
if (!this.contentAvailable) { |
|
return null; |
|
} |
|
if (!this.contentRef && this.root) { |
|
this.contentRef = pickPlatformItem(this.root.get('EF')); |
|
} |
|
var content = null; |
|
if (this.contentRef) { |
|
var xref = this.xref; |
|
var fileObj = xref.fetchIfRef(this.contentRef); |
|
if (fileObj && isStream(fileObj)) { |
|
content = fileObj.getBytes(); |
|
} else { |
|
warn('Embedded file specification points to non-existing/invalid ' + |
|
'content'); |
|
} |
|
} else { |
|
warn('Embedded file specification does not have a content'); |
|
} |
|
return content; |
|
}, |
|
get serializable() { |
|
return { |
|
filename: this.filename, |
|
content: this.content |
|
}; |
|
} |
|
}; |
|
return FileSpec; |
|
})(); |
|
|
|
/** |
|
* A helper for loading missing data in object graphs. It traverses the graph |
|
* depth first and queues up any objects that have missing data. Once it has |
|
* has traversed as many objects that are available it attempts to bundle the |
|
* missing data requests and then resume from the nodes that weren't ready. |
|
* |
|
* NOTE: It provides protection from circular references by keeping track of |
|
* of loaded references. However, you must be careful not to load any graphs |
|
* that have references to the catalog or other pages since that will cause the |
|
* entire PDF document object graph to be traversed. |
|
*/ |
|
var ObjectLoader = (function() { |
|
function mayHaveChildren(value) { |
|
return isRef(value) || isDict(value) || isArray(value) || isStream(value); |
|
} |
|
|
|
function addChildren(node, nodesToVisit) { |
|
var value; |
|
if (isDict(node) || isStream(node)) { |
|
var map; |
|
if (isDict(node)) { |
|
map = node.map; |
|
} else { |
|
map = node.dict.map; |
|
} |
|
for (var key in map) { |
|
value = map[key]; |
|
if (mayHaveChildren(value)) { |
|
nodesToVisit.push(value); |
|
} |
|
} |
|
} else if (isArray(node)) { |
|
for (var i = 0, ii = node.length; i < ii; i++) { |
|
value = node[i]; |
|
if (mayHaveChildren(value)) { |
|
nodesToVisit.push(value); |
|
} |
|
} |
|
} |
|
} |
|
|
|
function ObjectLoader(obj, keys, xref) { |
|
this.obj = obj; |
|
this.keys = keys; |
|
this.xref = xref; |
|
this.refSet = null; |
|
this.capability = null; |
|
} |
|
|
|
ObjectLoader.prototype = { |
|
load: function ObjectLoader_load() { |
|
var keys = this.keys; |
|
this.capability = createPromiseCapability(); |
|
// Don't walk the graph if all the data is already loaded. |
|
if (!(this.xref.stream instanceof ChunkedStream) || |
|
this.xref.stream.getMissingChunks().length === 0) { |
|
this.capability.resolve(); |
|
return this.capability.promise; |
|
} |
|
|
|
this.refSet = new RefSet(); |
|
// Setup the initial nodes to visit. |
|
var nodesToVisit = []; |
|
for (var i = 0; i < keys.length; i++) { |
|
nodesToVisit.push(this.obj[keys[i]]); |
|
} |
|
|
|
this._walk(nodesToVisit); |
|
return this.capability.promise; |
|
}, |
|
|
|
_walk: function ObjectLoader_walk(nodesToVisit) { |
|
var nodesToRevisit = []; |
|
var pendingRequests = []; |
|
// DFS walk of the object graph. |
|
while (nodesToVisit.length) { |
|
var currentNode = nodesToVisit.pop(); |
|
|
|
// Only references or chunked streams can cause missing data exceptions. |
|
if (isRef(currentNode)) { |
|
// Skip nodes that have already been visited. |
|
if (this.refSet.has(currentNode)) { |
|
continue; |
|
} |
|
try { |
|
var ref = currentNode; |
|
this.refSet.put(ref); |
|
currentNode = this.xref.fetch(currentNode); |
|
} catch (e) { |
|
if (!(e instanceof MissingDataException)) { |
|
throw e; |
|
} |
|
nodesToRevisit.push(currentNode); |
|
pendingRequests.push({ begin: e.begin, end: e.end }); |
|
} |
|
} |
|
if (currentNode && currentNode.getBaseStreams) { |
|
var baseStreams = currentNode.getBaseStreams(); |
|
var foundMissingData = false; |
|
for (var i = 0; i < baseStreams.length; i++) { |
|
var stream = baseStreams[i]; |
|
if (stream.getMissingChunks && stream.getMissingChunks().length) { |
|
foundMissingData = true; |
|
pendingRequests.push({ |
|
begin: stream.start, |
|
end: stream.end |
|
}); |
|
} |
|
} |
|
if (foundMissingData) { |
|
nodesToRevisit.push(currentNode); |
|
} |
|
} |
|
|
|
addChildren(currentNode, nodesToVisit); |
|
} |
|
|
|
if (pendingRequests.length) { |
|
this.xref.stream.manager.requestRanges(pendingRequests).then( |
|
function pendingRequestCallback() { |
|
nodesToVisit = nodesToRevisit; |
|
for (var i = 0; i < nodesToRevisit.length; i++) { |
|
var node = nodesToRevisit[i]; |
|
// Remove any reference nodes from the currrent refset so they |
|
// aren't skipped when we revist them. |
|
if (isRef(node)) { |
|
this.refSet.remove(node); |
|
} |
|
} |
|
this._walk(nodesToVisit); |
|
}.bind(this), this.capability.reject); |
|
return; |
|
} |
|
// Everything is loaded. |
|
this.refSet = null; |
|
this.capability.resolve(); |
|
} |
|
}; |
|
|
|
return ObjectLoader; |
|
})(); |
|
|
|
exports.Catalog = Catalog; |
|
exports.ObjectLoader = ObjectLoader; |
|
exports.XRef = XRef; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCorePsParser = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCoreParser); |
|
} |
|
}(this, function (exports, sharedUtil, coreParser) { |
|
|
|
var error = sharedUtil.error; |
|
var EOF = coreParser.EOF; |
|
var Lexer = coreParser.Lexer; |
|
|
|
var PostScriptParser = (function PostScriptParserClosure() { |
|
function PostScriptParser(lexer) { |
|
this.lexer = lexer; |
|
this.operators = []; |
|
this.token = null; |
|
this.prev = null; |
|
} |
|
PostScriptParser.prototype = { |
|
nextToken: function PostScriptParser_nextToken() { |
|
this.prev = this.token; |
|
this.token = this.lexer.getToken(); |
|
}, |
|
accept: function PostScriptParser_accept(type) { |
|
if (this.token.type === type) { |
|
this.nextToken(); |
|
return true; |
|
} |
|
return false; |
|
}, |
|
expect: function PostScriptParser_expect(type) { |
|
if (this.accept(type)) { |
|
return true; |
|
} |
|
error('Unexpected symbol: found ' + this.token.type + ' expected ' + |
|
type + '.'); |
|
}, |
|
parse: function PostScriptParser_parse() { |
|
this.nextToken(); |
|
this.expect(PostScriptTokenTypes.LBRACE); |
|
this.parseBlock(); |
|
this.expect(PostScriptTokenTypes.RBRACE); |
|
return this.operators; |
|
}, |
|
parseBlock: function PostScriptParser_parseBlock() { |
|
while (true) { |
|
if (this.accept(PostScriptTokenTypes.NUMBER)) { |
|
this.operators.push(this.prev.value); |
|
} else if (this.accept(PostScriptTokenTypes.OPERATOR)) { |
|
this.operators.push(this.prev.value); |
|
} else if (this.accept(PostScriptTokenTypes.LBRACE)) { |
|
this.parseCondition(); |
|
} else { |
|
return; |
|
} |
|
} |
|
}, |
|
parseCondition: function PostScriptParser_parseCondition() { |
|
// Add two place holders that will be updated later |
|
var conditionLocation = this.operators.length; |
|
this.operators.push(null, null); |
|
|
|
this.parseBlock(); |
|
this.expect(PostScriptTokenTypes.RBRACE); |
|
if (this.accept(PostScriptTokenTypes.IF)) { |
|
// The true block is right after the 'if' so it just falls through on |
|
// true else it jumps and skips the true block. |
|
this.operators[conditionLocation] = this.operators.length; |
|
this.operators[conditionLocation + 1] = 'jz'; |
|
} else if (this.accept(PostScriptTokenTypes.LBRACE)) { |
|
var jumpLocation = this.operators.length; |
|
this.operators.push(null, null); |
|
var endOfTrue = this.operators.length; |
|
this.parseBlock(); |
|
this.expect(PostScriptTokenTypes.RBRACE); |
|
this.expect(PostScriptTokenTypes.IFELSE); |
|
// The jump is added at the end of the true block to skip the false |
|
// block. |
|
this.operators[jumpLocation] = this.operators.length; |
|
this.operators[jumpLocation + 1] = 'j'; |
|
|
|
this.operators[conditionLocation] = endOfTrue; |
|
this.operators[conditionLocation + 1] = 'jz'; |
|
} else { |
|
error('PS Function: error parsing conditional.'); |
|
} |
|
} |
|
}; |
|
return PostScriptParser; |
|
})(); |
|
|
|
var PostScriptTokenTypes = { |
|
LBRACE: 0, |
|
RBRACE: 1, |
|
NUMBER: 2, |
|
OPERATOR: 3, |
|
IF: 4, |
|
IFELSE: 5 |
|
}; |
|
|
|
var PostScriptToken = (function PostScriptTokenClosure() { |
|
function PostScriptToken(type, value) { |
|
this.type = type; |
|
this.value = value; |
|
} |
|
|
|
var opCache = {}; |
|
|
|
PostScriptToken.getOperator = function PostScriptToken_getOperator(op) { |
|
var opValue = opCache[op]; |
|
if (opValue) { |
|
return opValue; |
|
} |
|
return opCache[op] = new PostScriptToken(PostScriptTokenTypes.OPERATOR, op); |
|
}; |
|
|
|
PostScriptToken.LBRACE = new PostScriptToken(PostScriptTokenTypes.LBRACE, |
|
'{'); |
|
PostScriptToken.RBRACE = new PostScriptToken(PostScriptTokenTypes.RBRACE, |
|
'}'); |
|
PostScriptToken.IF = new PostScriptToken(PostScriptTokenTypes.IF, 'IF'); |
|
PostScriptToken.IFELSE = new PostScriptToken(PostScriptTokenTypes.IFELSE, |
|
'IFELSE'); |
|
return PostScriptToken; |
|
})(); |
|
|
|
var PostScriptLexer = (function PostScriptLexerClosure() { |
|
function PostScriptLexer(stream) { |
|
this.stream = stream; |
|
this.nextChar(); |
|
|
|
this.strBuf = []; |
|
} |
|
PostScriptLexer.prototype = { |
|
nextChar: function PostScriptLexer_nextChar() { |
|
return (this.currentChar = this.stream.getByte()); |
|
}, |
|
getToken: function PostScriptLexer_getToken() { |
|
var comment = false; |
|
var ch = this.currentChar; |
|
|
|
// skip comments |
|
while (true) { |
|
if (ch < 0) { |
|
return EOF; |
|
} |
|
|
|
if (comment) { |
|
if (ch === 0x0A || ch === 0x0D) { |
|
comment = false; |
|
} |
|
} else if (ch === 0x25) { // '%' |
|
comment = true; |
|
} else if (!Lexer.isSpace(ch)) { |
|
break; |
|
} |
|
ch = this.nextChar(); |
|
} |
|
switch (ch | 0) { |
|
case 0x30: case 0x31: case 0x32: case 0x33: case 0x34: // '0'-'4' |
|
case 0x35: case 0x36: case 0x37: case 0x38: case 0x39: // '5'-'9' |
|
case 0x2B: case 0x2D: case 0x2E: // '+', '-', '.' |
|
return new PostScriptToken(PostScriptTokenTypes.NUMBER, |
|
this.getNumber()); |
|
case 0x7B: // '{' |
|
this.nextChar(); |
|
return PostScriptToken.LBRACE; |
|
case 0x7D: // '}' |
|
this.nextChar(); |
|
return PostScriptToken.RBRACE; |
|
} |
|
// operator |
|
var strBuf = this.strBuf; |
|
strBuf.length = 0; |
|
strBuf[0] = String.fromCharCode(ch); |
|
|
|
while ((ch = this.nextChar()) >= 0 && // and 'A'-'Z', 'a'-'z' |
|
((ch >= 0x41 && ch <= 0x5A) || (ch >= 0x61 && ch <= 0x7A))) { |
|
strBuf.push(String.fromCharCode(ch)); |
|
} |
|
var str = strBuf.join(''); |
|
switch (str.toLowerCase()) { |
|
case 'if': |
|
return PostScriptToken.IF; |
|
case 'ifelse': |
|
return PostScriptToken.IFELSE; |
|
default: |
|
return PostScriptToken.getOperator(str); |
|
} |
|
}, |
|
getNumber: function PostScriptLexer_getNumber() { |
|
var ch = this.currentChar; |
|
var strBuf = this.strBuf; |
|
strBuf.length = 0; |
|
strBuf[0] = String.fromCharCode(ch); |
|
|
|
while ((ch = this.nextChar()) >= 0) { |
|
if ((ch >= 0x30 && ch <= 0x39) || // '0'-'9' |
|
ch === 0x2D || ch === 0x2E) { // '-', '.' |
|
strBuf.push(String.fromCharCode(ch)); |
|
} else { |
|
break; |
|
} |
|
} |
|
var value = parseFloat(strBuf.join('')); |
|
if (isNaN(value)) { |
|
error('Invalid floating point number: ' + value); |
|
} |
|
return value; |
|
} |
|
}; |
|
return PostScriptLexer; |
|
})(); |
|
|
|
exports.PostScriptLexer = PostScriptLexer; |
|
exports.PostScriptParser = PostScriptParser; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreFonts = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreStream, root.pdfjsCoreParser, |
|
root.pdfjsCoreCMap, root.pdfjsCoreGlyphList, root.pdfjsCoreCharsets, |
|
root.pdfjsCoreFontRenderer); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreStream, coreParser, |
|
coreCMap, coreGlyphList, coreCharsets, coreFontRenderer) { |
|
|
|
var FONT_IDENTITY_MATRIX = sharedUtil.FONT_IDENTITY_MATRIX; |
|
var FontType = sharedUtil.FontType; |
|
var Util = sharedUtil.Util; |
|
var assert = sharedUtil.assert; |
|
var bytesToString = sharedUtil.bytesToString; |
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isArray = sharedUtil.isArray; |
|
var isInt = sharedUtil.isInt; |
|
var isNum = sharedUtil.isNum; |
|
var readUint32 = sharedUtil.readUint32; |
|
var shadow = sharedUtil.shadow; |
|
var stringToBytes = sharedUtil.stringToBytes; |
|
var string32 = sharedUtil.string32; |
|
var warn = sharedUtil.warn; |
|
var Name = corePrimitives.Name; |
|
var Stream = coreStream.Stream; |
|
var Lexer = coreParser.Lexer; |
|
var CMapFactory = coreCMap.CMapFactory; |
|
var IdentityCMap = coreCMap.IdentityCMap; |
|
var GlyphsUnicode = coreGlyphList.GlyphsUnicode; |
|
var DingbatsGlyphsUnicode = coreGlyphList.DingbatsGlyphsUnicode; |
|
var ISOAdobeCharset = coreCharsets.ISOAdobeCharset; |
|
var ExpertCharset = coreCharsets.ExpertCharset; |
|
var ExpertSubsetCharset = coreCharsets.ExpertSubsetCharset; |
|
var FontRendererFactory = coreFontRenderer.FontRendererFactory; |
|
|
|
// Unicode Private Use Area |
|
var PRIVATE_USE_OFFSET_START = 0xE000; |
|
var PRIVATE_USE_OFFSET_END = 0xF8FF; |
|
var SKIP_PRIVATE_USE_RANGE_F000_TO_F01F = false; |
|
|
|
// PDF Glyph Space Units are one Thousandth of a TextSpace Unit |
|
// except for Type 3 fonts |
|
var PDF_GLYPH_SPACE_UNITS = 1000; |
|
|
|
// Hinting is currently disabled due to unknown problems on windows |
|
// in tracemonkey and various other pdfs with type1 fonts. |
|
var HINTING_ENABLED = false; |
|
|
|
// Accented charactars are not displayed properly on windows, using this flag |
|
// to control analysis of seac charstrings. |
|
var SEAC_ANALYSIS_ENABLED = false; |
|
|
|
// Maximum subroutine call depth of type 2 chartrings. Matches OTS. |
|
var MAX_SUBR_NESTING = 10; |
|
|
|
var FontFlags = { |
|
FixedPitch: 1, |
|
Serif: 2, |
|
Symbolic: 4, |
|
Script: 8, |
|
Nonsymbolic: 32, |
|
Italic: 64, |
|
AllCap: 65536, |
|
SmallCap: 131072, |
|
ForceBold: 262144 |
|
}; |
|
|
|
var Encodings = { |
|
ExpertEncoding: ['', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'space', 'exclamsmall', 'Hungarumlautsmall', '', 'dollaroldstyle', |
|
'dollarsuperior', 'ampersandsmall', 'Acutesmall', 'parenleftsuperior', |
|
'parenrightsuperior', 'twodotenleader', 'onedotenleader', 'comma', |
|
'hyphen', 'period', 'fraction', 'zerooldstyle', 'oneoldstyle', |
|
'twooldstyle', 'threeoldstyle', 'fouroldstyle', 'fiveoldstyle', |
|
'sixoldstyle', 'sevenoldstyle', 'eightoldstyle', 'nineoldstyle', 'colon', |
|
'semicolon', 'commasuperior', 'threequartersemdash', 'periodsuperior', |
|
'questionsmall', '', 'asuperior', 'bsuperior', 'centsuperior', 'dsuperior', |
|
'esuperior', '', '', 'isuperior', '', '', 'lsuperior', 'msuperior', |
|
'nsuperior', 'osuperior', '', '', 'rsuperior', 'ssuperior', 'tsuperior', |
|
'', 'ff', 'fi', 'fl', 'ffi', 'ffl', 'parenleftinferior', '', |
|
'parenrightinferior', 'Circumflexsmall', 'hyphensuperior', 'Gravesmall', |
|
'Asmall', 'Bsmall', 'Csmall', 'Dsmall', 'Esmall', 'Fsmall', 'Gsmall', |
|
'Hsmall', 'Ismall', 'Jsmall', 'Ksmall', 'Lsmall', 'Msmall', 'Nsmall', |
|
'Osmall', 'Psmall', 'Qsmall', 'Rsmall', 'Ssmall', 'Tsmall', 'Usmall', |
|
'Vsmall', 'Wsmall', 'Xsmall', 'Ysmall', 'Zsmall', 'colonmonetary', |
|
'onefitted', 'rupiah', 'Tildesmall', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', 'exclamdownsmall', 'centoldstyle', 'Lslashsmall', |
|
'', '', 'Scaronsmall', 'Zcaronsmall', 'Dieresissmall', 'Brevesmall', |
|
'Caronsmall', '', 'Dotaccentsmall', '', '', 'Macronsmall', '', '', |
|
'figuredash', 'hypheninferior', '', '', 'Ogoneksmall', 'Ringsmall', |
|
'Cedillasmall', '', '', '', 'onequarter', 'onehalf', 'threequarters', |
|
'questiondownsmall', 'oneeighth', 'threeeighths', 'fiveeighths', |
|
'seveneighths', 'onethird', 'twothirds', '', '', 'zerosuperior', |
|
'onesuperior', 'twosuperior', 'threesuperior', 'foursuperior', |
|
'fivesuperior', 'sixsuperior', 'sevensuperior', 'eightsuperior', |
|
'ninesuperior', 'zeroinferior', 'oneinferior', 'twoinferior', |
|
'threeinferior', 'fourinferior', 'fiveinferior', 'sixinferior', |
|
'seveninferior', 'eightinferior', 'nineinferior', 'centinferior', |
|
'dollarinferior', 'periodinferior', 'commainferior', 'Agravesmall', |
|
'Aacutesmall', 'Acircumflexsmall', 'Atildesmall', 'Adieresissmall', |
|
'Aringsmall', 'AEsmall', 'Ccedillasmall', 'Egravesmall', 'Eacutesmall', |
|
'Ecircumflexsmall', 'Edieresissmall', 'Igravesmall', 'Iacutesmall', |
|
'Icircumflexsmall', 'Idieresissmall', 'Ethsmall', 'Ntildesmall', |
|
'Ogravesmall', 'Oacutesmall', 'Ocircumflexsmall', 'Otildesmall', |
|
'Odieresissmall', 'OEsmall', 'Oslashsmall', 'Ugravesmall', 'Uacutesmall', |
|
'Ucircumflexsmall', 'Udieresissmall', 'Yacutesmall', 'Thornsmall', |
|
'Ydieresissmall'], |
|
MacExpertEncoding: ['', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'space', 'exclamsmall', 'Hungarumlautsmall', 'centoldstyle', |
|
'dollaroldstyle', 'dollarsuperior', 'ampersandsmall', 'Acutesmall', |
|
'parenleftsuperior', 'parenrightsuperior', 'twodotenleader', |
|
'onedotenleader', 'comma', 'hyphen', 'period', 'fraction', 'zerooldstyle', |
|
'oneoldstyle', 'twooldstyle', 'threeoldstyle', 'fouroldstyle', |
|
'fiveoldstyle', 'sixoldstyle', 'sevenoldstyle', 'eightoldstyle', |
|
'nineoldstyle', 'colon', 'semicolon', '', 'threequartersemdash', '', |
|
'questionsmall', '', '', '', '', 'Ethsmall', '', '', 'onequarter', |
|
'onehalf', 'threequarters', 'oneeighth', 'threeeighths', 'fiveeighths', |
|
'seveneighths', 'onethird', 'twothirds', '', '', '', '', '', '', 'ff', |
|
'fi', 'fl', 'ffi', 'ffl', 'parenleftinferior', '', 'parenrightinferior', |
|
'Circumflexsmall', 'hypheninferior', 'Gravesmall', 'Asmall', 'Bsmall', |
|
'Csmall', 'Dsmall', 'Esmall', 'Fsmall', 'Gsmall', 'Hsmall', 'Ismall', |
|
'Jsmall', 'Ksmall', 'Lsmall', 'Msmall', 'Nsmall', 'Osmall', 'Psmall', |
|
'Qsmall', 'Rsmall', 'Ssmall', 'Tsmall', 'Usmall', 'Vsmall', 'Wsmall', |
|
'Xsmall', 'Ysmall', 'Zsmall', 'colonmonetary', 'onefitted', 'rupiah', |
|
'Tildesmall', '', '', 'asuperior', 'centsuperior', '', '', '', '', |
|
'Aacutesmall', 'Agravesmall', 'Acircumflexsmall', 'Adieresissmall', |
|
'Atildesmall', 'Aringsmall', 'Ccedillasmall', 'Eacutesmall', 'Egravesmall', |
|
'Ecircumflexsmall', 'Edieresissmall', 'Iacutesmall', 'Igravesmall', |
|
'Icircumflexsmall', 'Idieresissmall', 'Ntildesmall', 'Oacutesmall', |
|
'Ogravesmall', 'Ocircumflexsmall', 'Odieresissmall', 'Otildesmall', |
|
'Uacutesmall', 'Ugravesmall', 'Ucircumflexsmall', 'Udieresissmall', '', |
|
'eightsuperior', 'fourinferior', 'threeinferior', 'sixinferior', |
|
'eightinferior', 'seveninferior', 'Scaronsmall', '', 'centinferior', |
|
'twoinferior', '', 'Dieresissmall', '', 'Caronsmall', 'osuperior', |
|
'fiveinferior', '', 'commainferior', 'periodinferior', 'Yacutesmall', '', |
|
'dollarinferior', '', 'Thornsmall', '', 'nineinferior', 'zeroinferior', |
|
'Zcaronsmall', 'AEsmall', 'Oslashsmall', 'questiondownsmall', |
|
'oneinferior', 'Lslashsmall', '', '', '', '', '', '', 'Cedillasmall', '', |
|
'', '', '', '', 'OEsmall', 'figuredash', 'hyphensuperior', '', '', '', '', |
|
'exclamdownsmall', '', 'Ydieresissmall', '', 'onesuperior', 'twosuperior', |
|
'threesuperior', 'foursuperior', 'fivesuperior', 'sixsuperior', |
|
'sevensuperior', 'ninesuperior', 'zerosuperior', '', 'esuperior', |
|
'rsuperior', 'tsuperior', '', '', 'isuperior', 'ssuperior', 'dsuperior', |
|
'', '', '', '', '', 'lsuperior', 'Ogoneksmall', 'Brevesmall', |
|
'Macronsmall', 'bsuperior', 'nsuperior', 'msuperior', 'commasuperior', |
|
'periodsuperior', 'Dotaccentsmall', 'Ringsmall'], |
|
MacRomanEncoding: ['', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'space', 'exclam', 'quotedbl', 'numbersign', 'dollar', 'percent', |
|
'ampersand', 'quotesingle', 'parenleft', 'parenright', 'asterisk', 'plus', |
|
'comma', 'hyphen', 'period', 'slash', 'zero', 'one', 'two', 'three', |
|
'four', 'five', 'six', 'seven', 'eight', 'nine', 'colon', 'semicolon', |
|
'less', 'equal', 'greater', 'question', 'at', 'A', 'B', 'C', 'D', 'E', 'F', |
|
'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', |
|
'V', 'W', 'X', 'Y', 'Z', 'bracketleft', 'backslash', 'bracketright', |
|
'asciicircum', 'underscore', 'grave', 'a', 'b', 'c', 'd', 'e', 'f', 'g', |
|
'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', |
|
'w', 'x', 'y', 'z', 'braceleft', 'bar', 'braceright', 'asciitilde', '', |
|
'Adieresis', 'Aring', 'Ccedilla', 'Eacute', 'Ntilde', 'Odieresis', |
|
'Udieresis', 'aacute', 'agrave', 'acircumflex', 'adieresis', 'atilde', |
|
'aring', 'ccedilla', 'eacute', 'egrave', 'ecircumflex', 'edieresis', |
|
'iacute', 'igrave', 'icircumflex', 'idieresis', 'ntilde', 'oacute', |
|
'ograve', 'ocircumflex', 'odieresis', 'otilde', 'uacute', 'ugrave', |
|
'ucircumflex', 'udieresis', 'dagger', 'degree', 'cent', 'sterling', |
|
'section', 'bullet', 'paragraph', 'germandbls', 'registered', 'copyright', |
|
'trademark', 'acute', 'dieresis', 'notequal', 'AE', 'Oslash', 'infinity', |
|
'plusminus', 'lessequal', 'greaterequal', 'yen', 'mu', 'partialdiff', |
|
'summation', 'product', 'pi', 'integral', 'ordfeminine', 'ordmasculine', |
|
'Omega', 'ae', 'oslash', 'questiondown', 'exclamdown', 'logicalnot', |
|
'radical', 'florin', 'approxequal', 'Delta', 'guillemotleft', |
|
'guillemotright', 'ellipsis', 'space', 'Agrave', 'Atilde', 'Otilde', 'OE', |
|
'oe', 'endash', 'emdash', 'quotedblleft', 'quotedblright', 'quoteleft', |
|
'quoteright', 'divide', 'lozenge', 'ydieresis', 'Ydieresis', 'fraction', |
|
'currency', 'guilsinglleft', 'guilsinglright', 'fi', 'fl', 'daggerdbl', |
|
'periodcentered', 'quotesinglbase', 'quotedblbase', 'perthousand', |
|
'Acircumflex', 'Ecircumflex', 'Aacute', 'Edieresis', 'Egrave', 'Iacute', |
|
'Icircumflex', 'Idieresis', 'Igrave', 'Oacute', 'Ocircumflex', 'apple', |
|
'Ograve', 'Uacute', 'Ucircumflex', 'Ugrave', 'dotlessi', 'circumflex', |
|
'tilde', 'macron', 'breve', 'dotaccent', 'ring', 'cedilla', 'hungarumlaut', |
|
'ogonek', 'caron'], |
|
StandardEncoding: ['', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'space', 'exclam', 'quotedbl', 'numbersign', 'dollar', 'percent', |
|
'ampersand', 'quoteright', 'parenleft', 'parenright', 'asterisk', 'plus', |
|
'comma', 'hyphen', 'period', 'slash', 'zero', 'one', 'two', 'three', |
|
'four', 'five', 'six', 'seven', 'eight', 'nine', 'colon', 'semicolon', |
|
'less', 'equal', 'greater', 'question', 'at', 'A', 'B', 'C', 'D', 'E', 'F', |
|
'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', |
|
'V', 'W', 'X', 'Y', 'Z', 'bracketleft', 'backslash', 'bracketright', |
|
'asciicircum', 'underscore', 'quoteleft', 'a', 'b', 'c', 'd', 'e', 'f', |
|
'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', |
|
'v', 'w', 'x', 'y', 'z', 'braceleft', 'bar', 'braceright', 'asciitilde', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', 'exclamdown', |
|
'cent', 'sterling', 'fraction', 'yen', 'florin', 'section', 'currency', |
|
'quotesingle', 'quotedblleft', 'guillemotleft', 'guilsinglleft', |
|
'guilsinglright', 'fi', 'fl', '', 'endash', 'dagger', 'daggerdbl', |
|
'periodcentered', '', 'paragraph', 'bullet', 'quotesinglbase', |
|
'quotedblbase', 'quotedblright', 'guillemotright', 'ellipsis', |
|
'perthousand', '', 'questiondown', '', 'grave', 'acute', 'circumflex', |
|
'tilde', 'macron', 'breve', 'dotaccent', 'dieresis', '', 'ring', 'cedilla', |
|
'', 'hungarumlaut', 'ogonek', 'caron', 'emdash', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', 'AE', '', 'ordfeminine', '', '', |
|
'', '', 'Lslash', 'Oslash', 'OE', 'ordmasculine', '', '', '', '', '', 'ae', |
|
'', '', '', 'dotlessi', '', '', 'lslash', 'oslash', 'oe', 'germandbls'], |
|
WinAnsiEncoding: ['', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'space', 'exclam', 'quotedbl', 'numbersign', 'dollar', 'percent', |
|
'ampersand', 'quotesingle', 'parenleft', 'parenright', 'asterisk', 'plus', |
|
'comma', 'hyphen', 'period', 'slash', 'zero', 'one', 'two', 'three', |
|
'four', 'five', 'six', 'seven', 'eight', 'nine', 'colon', 'semicolon', |
|
'less', 'equal', 'greater', 'question', 'at', 'A', 'B', 'C', 'D', 'E', 'F', |
|
'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', |
|
'V', 'W', 'X', 'Y', 'Z', 'bracketleft', 'backslash', 'bracketright', |
|
'asciicircum', 'underscore', 'grave', 'a', 'b', 'c', 'd', 'e', 'f', 'g', |
|
'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', |
|
'w', 'x', 'y', 'z', 'braceleft', 'bar', 'braceright', 'asciitilde', |
|
'bullet', 'Euro', 'bullet', 'quotesinglbase', 'florin', 'quotedblbase', |
|
'ellipsis', 'dagger', 'daggerdbl', 'circumflex', 'perthousand', 'Scaron', |
|
'guilsinglleft', 'OE', 'bullet', 'Zcaron', 'bullet', 'bullet', 'quoteleft', |
|
'quoteright', 'quotedblleft', 'quotedblright', 'bullet', 'endash', |
|
'emdash', 'tilde', 'trademark', 'scaron', 'guilsinglright', 'oe', 'bullet', |
|
'zcaron', 'Ydieresis', 'space', 'exclamdown', 'cent', 'sterling', |
|
'currency', 'yen', 'brokenbar', 'section', 'dieresis', 'copyright', |
|
'ordfeminine', 'guillemotleft', 'logicalnot', 'hyphen', 'registered', |
|
'macron', 'degree', 'plusminus', 'twosuperior', 'threesuperior', 'acute', |
|
'mu', 'paragraph', 'periodcentered', 'cedilla', 'onesuperior', |
|
'ordmasculine', 'guillemotright', 'onequarter', 'onehalf', 'threequarters', |
|
'questiondown', 'Agrave', 'Aacute', 'Acircumflex', 'Atilde', 'Adieresis', |
|
'Aring', 'AE', 'Ccedilla', 'Egrave', 'Eacute', 'Ecircumflex', 'Edieresis', |
|
'Igrave', 'Iacute', 'Icircumflex', 'Idieresis', 'Eth', 'Ntilde', 'Ograve', |
|
'Oacute', 'Ocircumflex', 'Otilde', 'Odieresis', 'multiply', 'Oslash', |
|
'Ugrave', 'Uacute', 'Ucircumflex', 'Udieresis', 'Yacute', 'Thorn', |
|
'germandbls', 'agrave', 'aacute', 'acircumflex', 'atilde', 'adieresis', |
|
'aring', 'ae', 'ccedilla', 'egrave', 'eacute', 'ecircumflex', 'edieresis', |
|
'igrave', 'iacute', 'icircumflex', 'idieresis', 'eth', 'ntilde', 'ograve', |
|
'oacute', 'ocircumflex', 'otilde', 'odieresis', 'divide', 'oslash', |
|
'ugrave', 'uacute', 'ucircumflex', 'udieresis', 'yacute', 'thorn', |
|
'ydieresis'], |
|
SymbolSetEncoding: ['', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'space', 'exclam', 'universal', 'numbersign', 'existential', 'percent', |
|
'ampersand', 'suchthat', 'parenleft', 'parenright', 'asteriskmath', 'plus', |
|
'comma', 'minus', 'period', 'slash', 'zero', 'one', 'two', 'three', 'four', |
|
'five', 'six', 'seven', 'eight', 'nine', 'colon', 'semicolon', 'less', |
|
'equal', 'greater', 'question', 'congruent', 'Alpha', 'Beta', 'Chi', |
|
'Delta', 'Epsilon', 'Phi', 'Gamma', 'Eta', 'Iota', 'theta1', 'Kappa', |
|
'Lambda', 'Mu', 'Nu', 'Omicron', 'Pi', 'Theta', 'Rho', 'Sigma', 'Tau', |
|
'Upsilon', 'sigma1', 'Omega', 'Xi', 'Psi', 'Zeta', 'bracketleft', |
|
'therefore', 'bracketright', 'perpendicular', 'underscore', 'radicalex', |
|
'alpha', 'beta', 'chi', 'delta', 'epsilon', 'phi', 'gamma', 'eta', 'iota', |
|
'phi1', 'kappa', 'lambda', 'mu', 'nu', 'omicron', 'pi', 'theta', 'rho', |
|
'sigma', 'tau', 'upsilon', 'omega1', 'omega', 'xi', 'psi', 'zeta', |
|
'braceleft', 'bar', 'braceright', 'similar', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', 'Euro', 'Upsilon1', 'minute', 'lessequal', |
|
'fraction', 'infinity', 'florin', 'club', 'diamond', 'heart', 'spade', |
|
'arrowboth', 'arrowleft', 'arrowup', 'arrowright', 'arrowdown', 'degree', |
|
'plusminus', 'second', 'greaterequal', 'multiply', 'proportional', |
|
'partialdiff', 'bullet', 'divide', 'notequal', 'equivalence', |
|
'approxequal', 'ellipsis', 'arrowvertex', 'arrowhorizex', 'carriagereturn', |
|
'aleph', 'Ifraktur', 'Rfraktur', 'weierstrass', 'circlemultiply', |
|
'circleplus', 'emptyset', 'intersection', 'union', 'propersuperset', |
|
'reflexsuperset', 'notsubset', 'propersubset', 'reflexsubset', 'element', |
|
'notelement', 'angle', 'gradient', 'registerserif', 'copyrightserif', |
|
'trademarkserif', 'product', 'radical', 'dotmath', 'logicalnot', |
|
'logicaland', 'logicalor', 'arrowdblboth', 'arrowdblleft', 'arrowdblup', |
|
'arrowdblright', 'arrowdbldown', 'lozenge', 'angleleft', 'registersans', |
|
'copyrightsans', 'trademarksans', 'summation', 'parenlefttp', |
|
'parenleftex', 'parenleftbt', 'bracketlefttp', 'bracketleftex', |
|
'bracketleftbt', 'bracelefttp', 'braceleftmid', 'braceleftbt', 'braceex', |
|
'', 'angleright', 'integral', 'integraltp', 'integralex', 'integralbt', |
|
'parenrighttp', 'parenrightex', 'parenrightbt', 'bracketrighttp', |
|
'bracketrightex', 'bracketrightbt', 'bracerighttp', 'bracerightmid', |
|
'bracerightbt'], |
|
ZapfDingbatsEncoding: ['', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', '', |
|
'space', 'a1', 'a2', 'a202', 'a3', 'a4', 'a5', 'a119', 'a118', 'a117', |
|
'a11', 'a12', 'a13', 'a14', 'a15', 'a16', 'a105', 'a17', 'a18', 'a19', |
|
'a20', 'a21', 'a22', 'a23', 'a24', 'a25', 'a26', 'a27', 'a28', 'a6', 'a7', |
|
'a8', 'a9', 'a10', 'a29', 'a30', 'a31', 'a32', 'a33', 'a34', 'a35', 'a36', |
|
'a37', 'a38', 'a39', 'a40', 'a41', 'a42', 'a43', 'a44', 'a45', 'a46', |
|
'a47', 'a48', 'a49', 'a50', 'a51', 'a52', 'a53', 'a54', 'a55', 'a56', |
|
'a57', 'a58', 'a59', 'a60', 'a61', 'a62', 'a63', 'a64', 'a65', 'a66', |
|
'a67', 'a68', 'a69', 'a70', 'a71', 'a72', 'a73', 'a74', 'a203', 'a75', |
|
'a204', 'a76', 'a77', 'a78', 'a79', 'a81', 'a82', 'a83', 'a84', 'a97', |
|
'a98', 'a99', 'a100', '', 'a89', 'a90', 'a93', 'a94', 'a91', 'a92', 'a205', |
|
'a85', 'a206', 'a86', 'a87', 'a88', 'a95', 'a96', '', '', '', '', '', '', |
|
'', '', '', '', '', '', '', '', '', '', '', '', '', 'a101', 'a102', 'a103', |
|
'a104', 'a106', 'a107', 'a108', 'a112', 'a111', 'a110', 'a109', 'a120', |
|
'a121', 'a122', 'a123', 'a124', 'a125', 'a126', 'a127', 'a128', 'a129', |
|
'a130', 'a131', 'a132', 'a133', 'a134', 'a135', 'a136', 'a137', 'a138', |
|
'a139', 'a140', 'a141', 'a142', 'a143', 'a144', 'a145', 'a146', 'a147', |
|
'a148', 'a149', 'a150', 'a151', 'a152', 'a153', 'a154', 'a155', 'a156', |
|
'a157', 'a158', 'a159', 'a160', 'a161', 'a163', 'a164', 'a196', 'a165', |
|
'a192', 'a166', 'a167', 'a168', 'a169', 'a170', 'a171', 'a172', 'a173', |
|
'a162', 'a174', 'a175', 'a176', 'a177', 'a178', 'a179', 'a193', 'a180', |
|
'a199', 'a181', 'a200', 'a182', '', 'a201', 'a183', 'a184', 'a197', 'a185', |
|
'a194', 'a198', 'a186', 'a195', 'a187', 'a188', 'a189', 'a190', 'a191'] |
|
}; |
|
|
|
/** |
|
* Hold a map of decoded fonts and of the standard fourteen Type1 |
|
* fonts and their acronyms. |
|
*/ |
|
var stdFontMap = { |
|
'ArialNarrow': 'Helvetica', |
|
'ArialNarrow-Bold': 'Helvetica-Bold', |
|
'ArialNarrow-BoldItalic': 'Helvetica-BoldOblique', |
|
'ArialNarrow-Italic': 'Helvetica-Oblique', |
|
'ArialBlack': 'Helvetica', |
|
'ArialBlack-Bold': 'Helvetica-Bold', |
|
'ArialBlack-BoldItalic': 'Helvetica-BoldOblique', |
|
'ArialBlack-Italic': 'Helvetica-Oblique', |
|
'Arial': 'Helvetica', |
|
'Arial-Bold': 'Helvetica-Bold', |
|
'Arial-BoldItalic': 'Helvetica-BoldOblique', |
|
'Arial-Italic': 'Helvetica-Oblique', |
|
'Arial-BoldItalicMT': 'Helvetica-BoldOblique', |
|
'Arial-BoldMT': 'Helvetica-Bold', |
|
'Arial-ItalicMT': 'Helvetica-Oblique', |
|
'ArialMT': 'Helvetica', |
|
'Courier-Bold': 'Courier-Bold', |
|
'Courier-BoldItalic': 'Courier-BoldOblique', |
|
'Courier-Italic': 'Courier-Oblique', |
|
'CourierNew': 'Courier', |
|
'CourierNew-Bold': 'Courier-Bold', |
|
'CourierNew-BoldItalic': 'Courier-BoldOblique', |
|
'CourierNew-Italic': 'Courier-Oblique', |
|
'CourierNewPS-BoldItalicMT': 'Courier-BoldOblique', |
|
'CourierNewPS-BoldMT': 'Courier-Bold', |
|
'CourierNewPS-ItalicMT': 'Courier-Oblique', |
|
'CourierNewPSMT': 'Courier', |
|
'Helvetica': 'Helvetica', |
|
'Helvetica-Bold': 'Helvetica-Bold', |
|
'Helvetica-BoldItalic': 'Helvetica-BoldOblique', |
|
'Helvetica-BoldOblique': 'Helvetica-BoldOblique', |
|
'Helvetica-Italic': 'Helvetica-Oblique', |
|
'Helvetica-Oblique':'Helvetica-Oblique', |
|
'Symbol-Bold': 'Symbol', |
|
'Symbol-BoldItalic': 'Symbol', |
|
'Symbol-Italic': 'Symbol', |
|
'TimesNewRoman': 'Times-Roman', |
|
'TimesNewRoman-Bold': 'Times-Bold', |
|
'TimesNewRoman-BoldItalic': 'Times-BoldItalic', |
|
'TimesNewRoman-Italic': 'Times-Italic', |
|
'TimesNewRomanPS': 'Times-Roman', |
|
'TimesNewRomanPS-Bold': 'Times-Bold', |
|
'TimesNewRomanPS-BoldItalic': 'Times-BoldItalic', |
|
'TimesNewRomanPS-BoldItalicMT': 'Times-BoldItalic', |
|
'TimesNewRomanPS-BoldMT': 'Times-Bold', |
|
'TimesNewRomanPS-Italic': 'Times-Italic', |
|
'TimesNewRomanPS-ItalicMT': 'Times-Italic', |
|
'TimesNewRomanPSMT': 'Times-Roman', |
|
'TimesNewRomanPSMT-Bold': 'Times-Bold', |
|
'TimesNewRomanPSMT-BoldItalic': 'Times-BoldItalic', |
|
'TimesNewRomanPSMT-Italic': 'Times-Italic' |
|
}; |
|
|
|
/** |
|
* Holds the map of the non-standard fonts that might be included as a standard |
|
* fonts without glyph data. |
|
*/ |
|
var nonStdFontMap = { |
|
'CenturyGothic': 'Helvetica', |
|
'CenturyGothic-Bold': 'Helvetica-Bold', |
|
'CenturyGothic-BoldItalic': 'Helvetica-BoldOblique', |
|
'CenturyGothic-Italic': 'Helvetica-Oblique', |
|
'ComicSansMS': 'Comic Sans MS', |
|
'ComicSansMS-Bold': 'Comic Sans MS-Bold', |
|
'ComicSansMS-BoldItalic': 'Comic Sans MS-BoldItalic', |
|
'ComicSansMS-Italic': 'Comic Sans MS-Italic', |
|
'LucidaConsole': 'Courier', |
|
'LucidaConsole-Bold': 'Courier-Bold', |
|
'LucidaConsole-BoldItalic': 'Courier-BoldOblique', |
|
'LucidaConsole-Italic': 'Courier-Oblique', |
|
'MS-Gothic': 'MS Gothic', |
|
'MS-Gothic-Bold': 'MS Gothic-Bold', |
|
'MS-Gothic-BoldItalic': 'MS Gothic-BoldItalic', |
|
'MS-Gothic-Italic': 'MS Gothic-Italic', |
|
'MS-Mincho': 'MS Mincho', |
|
'MS-Mincho-Bold': 'MS Mincho-Bold', |
|
'MS-Mincho-BoldItalic': 'MS Mincho-BoldItalic', |
|
'MS-Mincho-Italic': 'MS Mincho-Italic', |
|
'MS-PGothic': 'MS PGothic', |
|
'MS-PGothic-Bold': 'MS PGothic-Bold', |
|
'MS-PGothic-BoldItalic': 'MS PGothic-BoldItalic', |
|
'MS-PGothic-Italic': 'MS PGothic-Italic', |
|
'MS-PMincho': 'MS PMincho', |
|
'MS-PMincho-Bold': 'MS PMincho-Bold', |
|
'MS-PMincho-BoldItalic': 'MS PMincho-BoldItalic', |
|
'MS-PMincho-Italic': 'MS PMincho-Italic', |
|
'Wingdings': 'ZapfDingbats' |
|
}; |
|
|
|
var serifFonts = { |
|
'Adobe Jenson': true, 'Adobe Text': true, 'Albertus': true, |
|
'Aldus': true, 'Alexandria': true, 'Algerian': true, |
|
'American Typewriter': true, 'Antiqua': true, 'Apex': true, |
|
'Arno': true, 'Aster': true, 'Aurora': true, |
|
'Baskerville': true, 'Bell': true, 'Bembo': true, |
|
'Bembo Schoolbook': true, 'Benguiat': true, 'Berkeley Old Style': true, |
|
'Bernhard Modern': true, 'Berthold City': true, 'Bodoni': true, |
|
'Bauer Bodoni': true, 'Book Antiqua': true, 'Bookman': true, |
|
'Bordeaux Roman': true, 'Californian FB': true, 'Calisto': true, |
|
'Calvert': true, 'Capitals': true, 'Cambria': true, |
|
'Cartier': true, 'Caslon': true, 'Catull': true, |
|
'Centaur': true, 'Century Old Style': true, 'Century Schoolbook': true, |
|
'Chaparral': true, 'Charis SIL': true, 'Cheltenham': true, |
|
'Cholla Slab': true, 'Clarendon': true, 'Clearface': true, |
|
'Cochin': true, 'Colonna': true, 'Computer Modern': true, |
|
'Concrete Roman': true, 'Constantia': true, 'Cooper Black': true, |
|
'Corona': true, 'Ecotype': true, 'Egyptienne': true, |
|
'Elephant': true, 'Excelsior': true, 'Fairfield': true, |
|
'FF Scala': true, 'Folkard': true, 'Footlight': true, |
|
'FreeSerif': true, 'Friz Quadrata': true, 'Garamond': true, |
|
'Gentium': true, 'Georgia': true, 'Gloucester': true, |
|
'Goudy Old Style': true, 'Goudy Schoolbook': true, 'Goudy Pro Font': true, |
|
'Granjon': true, 'Guardian Egyptian': true, 'Heather': true, |
|
'Hercules': true, 'High Tower Text': true, 'Hiroshige': true, |
|
'Hoefler Text': true, 'Humana Serif': true, 'Imprint': true, |
|
'Ionic No. 5': true, 'Janson': true, 'Joanna': true, |
|
'Korinna': true, 'Lexicon': true, 'Liberation Serif': true, |
|
'Linux Libertine': true, 'Literaturnaya': true, 'Lucida': true, |
|
'Lucida Bright': true, 'Melior': true, 'Memphis': true, |
|
'Miller': true, 'Minion': true, 'Modern': true, |
|
'Mona Lisa': true, 'Mrs Eaves': true, 'MS Serif': true, |
|
'Museo Slab': true, 'New York': true, 'Nimbus Roman': true, |
|
'NPS Rawlinson Roadway': true, 'Palatino': true, 'Perpetua': true, |
|
'Plantin': true, 'Plantin Schoolbook': true, 'Playbill': true, |
|
'Poor Richard': true, 'Rawlinson Roadway': true, 'Renault': true, |
|
'Requiem': true, 'Rockwell': true, 'Roman': true, |
|
'Rotis Serif': true, 'Sabon': true, 'Scala': true, |
|
'Seagull': true, 'Sistina': true, 'Souvenir': true, |
|
'STIX': true, 'Stone Informal': true, 'Stone Serif': true, |
|
'Sylfaen': true, 'Times': true, 'Trajan': true, |
|
'Trinité': true, 'Trump Mediaeval': true, 'Utopia': true, |
|
'Vale Type': true, 'Bitstream Vera': true, 'Vera Serif': true, |
|
'Versailles': true, 'Wanted': true, 'Weiss': true, |
|
'Wide Latin': true, 'Windsor': true, 'XITS': true |
|
}; |
|
|
|
var symbolsFonts = { |
|
'Dingbats': true, 'Symbol': true, 'ZapfDingbats': true |
|
}; |
|
|
|
// Glyph map for well-known standard fonts. Sometimes Ghostscript uses CID fonts |
|
// but does not embed the CID to GID mapping. The mapping is incomplete for all |
|
// glyphs, but common for some set of the standard fonts. |
|
var GlyphMapForStandardFonts = { |
|
'2': 10, '3': 32, '4': 33, '5': 34, '6': 35, '7': 36, '8': 37, '9': 38, |
|
'10': 39, '11': 40, '12': 41, '13': 42, '14': 43, '15': 44, '16': 45, |
|
'17': 46, '18': 47, '19': 48, '20': 49, '21': 50, '22': 51, '23': 52, |
|
'24': 53, '25': 54, '26': 55, '27': 56, '28': 57, '29': 58, '30': 894, |
|
'31': 60, '32': 61, '33': 62, '34': 63, '35': 64, '36': 65, '37': 66, |
|
'38': 67, '39': 68, '40': 69, '41': 70, '42': 71, '43': 72, '44': 73, |
|
'45': 74, '46': 75, '47': 76, '48': 77, '49': 78, '50': 79, '51': 80, |
|
'52': 81, '53': 82, '54': 83, '55': 84, '56': 85, '57': 86, '58': 87, |
|
'59': 88, '60': 89, '61': 90, '62': 91, '63': 92, '64': 93, '65': 94, |
|
'66': 95, '67': 96, '68': 97, '69': 98, '70': 99, '71': 100, '72': 101, |
|
'73': 102, '74': 103, '75': 104, '76': 105, '77': 106, '78': 107, '79': 108, |
|
'80': 109, '81': 110, '82': 111, '83': 112, '84': 113, '85': 114, '86': 115, |
|
'87': 116, '88': 117, '89': 118, '90': 119, '91': 120, '92': 121, '93': 122, |
|
'94': 123, '95': 124, '96': 125, '97': 126, '98': 196, '99': 197, '100': 199, |
|
'101': 201, '102': 209, '103': 214, '104': 220, '105': 225, '106': 224, |
|
'107': 226, '108': 228, '109': 227, '110': 229, '111': 231, '112': 233, |
|
'113': 232, '114': 234, '115': 235, '116': 237, '117': 236, '118': 238, |
|
'119': 239, '120': 241, '121': 243, '122': 242, '123': 244, '124': 246, |
|
'125': 245, '126': 250, '127': 249, '128': 251, '129': 252, '130': 8224, |
|
'131': 176, '132': 162, '133': 163, '134': 167, '135': 8226, '136': 182, |
|
'137': 223, '138': 174, '139': 169, '140': 8482, '141': 180, '142': 168, |
|
'143': 8800, '144': 198, '145': 216, '146': 8734, '147': 177, '148': 8804, |
|
'149': 8805, '150': 165, '151': 181, '152': 8706, '153': 8721, '154': 8719, |
|
'156': 8747, '157': 170, '158': 186, '159': 8486, '160': 230, '161': 248, |
|
'162': 191, '163': 161, '164': 172, '165': 8730, '166': 402, '167': 8776, |
|
'168': 8710, '169': 171, '170': 187, '171': 8230, '210': 218, '223': 711, |
|
'224': 321, '225': 322, '227': 353, '229': 382, '234': 253, '252': 263, |
|
'253': 268, '254': 269, '258': 258, '260': 260, '261': 261, '265': 280, |
|
'266': 281, '268': 283, '269': 313, '275': 323, '276': 324, '278': 328, |
|
'284': 345, '285': 346, '286': 347, '292': 367, '295': 377, '296': 378, |
|
'298': 380, '305': 963, |
|
'306': 964, '307': 966, '308': 8215, '309': 8252, '310': 8319, '311': 8359, |
|
'312': 8592, '313': 8593, '337': 9552, '493': 1039, '494': 1040, '705': 1524, |
|
'706': 8362, '710': 64288, '711': 64298, '759': 1617, '761': 1776, |
|
'763': 1778, '775': 1652, '777': 1764, '778': 1780, '779': 1781, '780': 1782, |
|
'782': 771, '783': 64726, '786': 8363, '788': 8532, '790': 768, '791': 769, |
|
'792': 768, '795': 803, '797': 64336, '798': 64337, '799': 64342, |
|
'800': 64343, '801': 64344, '802': 64345, '803': 64362, '804': 64363, |
|
'805': 64364, '2424': 7821, '2425': 7822, '2426': 7823, '2427': 7824, |
|
'2428': 7825, '2429': 7826, '2430': 7827, '2433': 7682, '2678': 8045, |
|
'2679': 8046, '2830': 1552, '2838': 686, '2840': 751, '2842': 753, |
|
'2843': 754, '2844': 755, '2846': 757, '2856': 767, '2857': 848, '2858': 849, |
|
'2862': 853, '2863': 854, '2864': 855, '2865': 861, '2866': 862, '2906': 7460, |
|
'2908': 7462, '2909': 7463, '2910': 7464, '2912': 7466, '2913': 7467, |
|
'2914': 7468, '2916': 7470, '2917': 7471, '2918': 7472, '2920': 7474, |
|
'2921': 7475, '2922': 7476, '2924': 7478, '2925': 7479, '2926': 7480, |
|
'2928': 7482, '2929': 7483, '2930': 7484, '2932': 7486, '2933': 7487, |
|
'2934': 7488, '2936': 7490, '2937': 7491, '2938': 7492, '2940': 7494, |
|
'2941': 7495, '2942': 7496, '2944': 7498, '2946': 7500, '2948': 7502, |
|
'2950': 7504, '2951': 7505, '2952': 7506, '2954': 7508, '2955': 7509, |
|
'2956': 7510, '2958': 7512, '2959': 7513, '2960': 7514, '2962': 7516, |
|
'2963': 7517, '2964': 7518, '2966': 7520, '2967': 7521, '2968': 7522, |
|
'2970': 7524, '2971': 7525, '2972': 7526, '2974': 7528, '2975': 7529, |
|
'2976': 7530, '2978': 1537, '2979': 1538, '2980': 1539, '2982': 1549, |
|
'2983': 1551, '2984': 1552, '2986': 1554, '2987': 1555, '2988': 1556, |
|
'2990': 1623, '2991': 1624, '2995': 1775, '2999': 1791, '3002': 64290, |
|
'3003': 64291, '3004': 64292, '3006': 64294, '3007': 64295, '3008': 64296, |
|
'3011': 1900, '3014': 8223, '3015': 8244, '3017': 7532, '3018': 7533, |
|
'3019': 7534, '3075': 7590, '3076': 7591, '3079': 7594, '3080': 7595, |
|
'3083': 7598, '3084': 7599, '3087': 7602, '3088': 7603, '3091': 7606, |
|
'3092': 7607, '3095': 7610, '3096': 7611, '3099': 7614, '3100': 7615, |
|
'3103': 7618, '3104': 7619, '3107': 8337, '3108': 8338, '3116': 1884, |
|
'3119': 1885, '3120': 1885, '3123': 1886, '3124': 1886, '3127': 1887, |
|
'3128': 1887, '3131': 1888, '3132': 1888, '3135': 1889, '3136': 1889, |
|
'3139': 1890, '3140': 1890, '3143': 1891, '3144': 1891, '3147': 1892, |
|
'3148': 1892, '3153': 580, '3154': 581, '3157': 584, '3158': 585, '3161': 588, |
|
'3162': 589, '3165': 891, '3166': 892, '3169': 1274, '3170': 1275, |
|
'3173': 1278, '3174': 1279, '3181': 7622, '3182': 7623, '3282': 11799, |
|
'3316': 578, '3379': 42785, '3393': 1159, '3416': 8377 |
|
}; |
|
|
|
// The glyph map for ArialBlack differs slightly from the glyph map used for |
|
// other well-known standard fonts. Hence we use this (incomplete) CID to GID |
|
// mapping to adjust the glyph map for non-embedded ArialBlack fonts. |
|
var SupplementalGlyphMapForArialBlack = { |
|
'227': 322, '264': 261, '291': 346, |
|
}; |
|
|
|
// Some characters, e.g. copyrightserif, are mapped to the private use area and |
|
// might not be displayed using standard fonts. Mapping/hacking well-known chars |
|
// to the similar equivalents in the normal characters range. |
|
var SpecialPUASymbols = { |
|
'63721': 0x00A9, // copyrightsans (0xF8E9) => copyright |
|
'63193': 0x00A9, // copyrightserif (0xF6D9) => copyright |
|
'63720': 0x00AE, // registersans (0xF8E8) => registered |
|
'63194': 0x00AE, // registerserif (0xF6DA) => registered |
|
'63722': 0x2122, // trademarksans (0xF8EA) => trademark |
|
'63195': 0x2122, // trademarkserif (0xF6DB) => trademark |
|
'63729': 0x23A7, // bracelefttp (0xF8F1) |
|
'63730': 0x23A8, // braceleftmid (0xF8F2) |
|
'63731': 0x23A9, // braceleftbt (0xF8F3) |
|
'63740': 0x23AB, // bracerighttp (0xF8FC) |
|
'63741': 0x23AC, // bracerightmid (0xF8FD) |
|
'63742': 0x23AD, // bracerightbt (0xF8FE) |
|
'63726': 0x23A1, // bracketlefttp (0xF8EE) |
|
'63727': 0x23A2, // bracketleftex (0xF8EF) |
|
'63728': 0x23A3, // bracketleftbt (0xF8F0) |
|
'63737': 0x23A4, // bracketrighttp (0xF8F9) |
|
'63738': 0x23A5, // bracketrightex (0xF8FA) |
|
'63739': 0x23A6, // bracketrightbt (0xF8FB) |
|
'63723': 0x239B, // parenlefttp (0xF8EB) |
|
'63724': 0x239C, // parenleftex (0xF8EC) |
|
'63725': 0x239D, // parenleftbt (0xF8ED) |
|
'63734': 0x239E, // parenrighttp (0xF8F6) |
|
'63735': 0x239F, // parenrightex (0xF8F7) |
|
'63736': 0x23A0, // parenrightbt (0xF8F8) |
|
}; |
|
function mapSpecialUnicodeValues(code) { |
|
if (code >= 0xFFF0 && code <= 0xFFFF) { // Specials unicode block. |
|
return 0; |
|
} else if (code >= 0xF600 && code <= 0xF8FF) { |
|
return (SpecialPUASymbols[code] || code); |
|
} |
|
return code; |
|
} |
|
|
|
var UnicodeRanges = [ |
|
{ 'begin': 0x0000, 'end': 0x007F }, // Basic Latin |
|
{ 'begin': 0x0080, 'end': 0x00FF }, // Latin-1 Supplement |
|
{ 'begin': 0x0100, 'end': 0x017F }, // Latin Extended-A |
|
{ 'begin': 0x0180, 'end': 0x024F }, // Latin Extended-B |
|
{ 'begin': 0x0250, 'end': 0x02AF }, // IPA Extensions |
|
{ 'begin': 0x02B0, 'end': 0x02FF }, // Spacing Modifier Letters |
|
{ 'begin': 0x0300, 'end': 0x036F }, // Combining Diacritical Marks |
|
{ 'begin': 0x0370, 'end': 0x03FF }, // Greek and Coptic |
|
{ 'begin': 0x2C80, 'end': 0x2CFF }, // Coptic |
|
{ 'begin': 0x0400, 'end': 0x04FF }, // Cyrillic |
|
{ 'begin': 0x0530, 'end': 0x058F }, // Armenian |
|
{ 'begin': 0x0590, 'end': 0x05FF }, // Hebrew |
|
{ 'begin': 0xA500, 'end': 0xA63F }, // Vai |
|
{ 'begin': 0x0600, 'end': 0x06FF }, // Arabic |
|
{ 'begin': 0x07C0, 'end': 0x07FF }, // NKo |
|
{ 'begin': 0x0900, 'end': 0x097F }, // Devanagari |
|
{ 'begin': 0x0980, 'end': 0x09FF }, // Bengali |
|
{ 'begin': 0x0A00, 'end': 0x0A7F }, // Gurmukhi |
|
{ 'begin': 0x0A80, 'end': 0x0AFF }, // Gujarati |
|
{ 'begin': 0x0B00, 'end': 0x0B7F }, // Oriya |
|
{ 'begin': 0x0B80, 'end': 0x0BFF }, // Tamil |
|
{ 'begin': 0x0C00, 'end': 0x0C7F }, // Telugu |
|
{ 'begin': 0x0C80, 'end': 0x0CFF }, // Kannada |
|
{ 'begin': 0x0D00, 'end': 0x0D7F }, // Malayalam |
|
{ 'begin': 0x0E00, 'end': 0x0E7F }, // Thai |
|
{ 'begin': 0x0E80, 'end': 0x0EFF }, // Lao |
|
{ 'begin': 0x10A0, 'end': 0x10FF }, // Georgian |
|
{ 'begin': 0x1B00, 'end': 0x1B7F }, // Balinese |
|
{ 'begin': 0x1100, 'end': 0x11FF }, // Hangul Jamo |
|
{ 'begin': 0x1E00, 'end': 0x1EFF }, // Latin Extended Additional |
|
{ 'begin': 0x1F00, 'end': 0x1FFF }, // Greek Extended |
|
{ 'begin': 0x2000, 'end': 0x206F }, // General Punctuation |
|
{ 'begin': 0x2070, 'end': 0x209F }, // Superscripts And Subscripts |
|
{ 'begin': 0x20A0, 'end': 0x20CF }, // Currency Symbol |
|
{ 'begin': 0x20D0, 'end': 0x20FF }, // Combining Diacritical Marks For Symbols |
|
{ 'begin': 0x2100, 'end': 0x214F }, // Letterlike Symbols |
|
{ 'begin': 0x2150, 'end': 0x218F }, // Number Forms |
|
{ 'begin': 0x2190, 'end': 0x21FF }, // Arrows |
|
{ 'begin': 0x2200, 'end': 0x22FF }, // Mathematical Operators |
|
{ 'begin': 0x2300, 'end': 0x23FF }, // Miscellaneous Technical |
|
{ 'begin': 0x2400, 'end': 0x243F }, // Control Pictures |
|
{ 'begin': 0x2440, 'end': 0x245F }, // Optical Character Recognition |
|
{ 'begin': 0x2460, 'end': 0x24FF }, // Enclosed Alphanumerics |
|
{ 'begin': 0x2500, 'end': 0x257F }, // Box Drawing |
|
{ 'begin': 0x2580, 'end': 0x259F }, // Block Elements |
|
{ 'begin': 0x25A0, 'end': 0x25FF }, // Geometric Shapes |
|
{ 'begin': 0x2600, 'end': 0x26FF }, // Miscellaneous Symbols |
|
{ 'begin': 0x2700, 'end': 0x27BF }, // Dingbats |
|
{ 'begin': 0x3000, 'end': 0x303F }, // CJK Symbols And Punctuation |
|
{ 'begin': 0x3040, 'end': 0x309F }, // Hiragana |
|
{ 'begin': 0x30A0, 'end': 0x30FF }, // Katakana |
|
{ 'begin': 0x3100, 'end': 0x312F }, // Bopomofo |
|
{ 'begin': 0x3130, 'end': 0x318F }, // Hangul Compatibility Jamo |
|
{ 'begin': 0xA840, 'end': 0xA87F }, // Phags-pa |
|
{ 'begin': 0x3200, 'end': 0x32FF }, // Enclosed CJK Letters And Months |
|
{ 'begin': 0x3300, 'end': 0x33FF }, // CJK Compatibility |
|
{ 'begin': 0xAC00, 'end': 0xD7AF }, // Hangul Syllables |
|
{ 'begin': 0xD800, 'end': 0xDFFF }, // Non-Plane 0 * |
|
{ 'begin': 0x10900, 'end': 0x1091F }, // Phoenicia |
|
{ 'begin': 0x4E00, 'end': 0x9FFF }, // CJK Unified Ideographs |
|
{ 'begin': 0xE000, 'end': 0xF8FF }, // Private Use Area (plane 0) |
|
{ 'begin': 0x31C0, 'end': 0x31EF }, // CJK Strokes |
|
{ 'begin': 0xFB00, 'end': 0xFB4F }, // Alphabetic Presentation Forms |
|
{ 'begin': 0xFB50, 'end': 0xFDFF }, // Arabic Presentation Forms-A |
|
{ 'begin': 0xFE20, 'end': 0xFE2F }, // Combining Half Marks |
|
{ 'begin': 0xFE10, 'end': 0xFE1F }, // Vertical Forms |
|
{ 'begin': 0xFE50, 'end': 0xFE6F }, // Small Form Variants |
|
{ 'begin': 0xFE70, 'end': 0xFEFF }, // Arabic Presentation Forms-B |
|
{ 'begin': 0xFF00, 'end': 0xFFEF }, // Halfwidth And Fullwidth Forms |
|
{ 'begin': 0xFFF0, 'end': 0xFFFF }, // Specials |
|
{ 'begin': 0x0F00, 'end': 0x0FFF }, // Tibetan |
|
{ 'begin': 0x0700, 'end': 0x074F }, // Syriac |
|
{ 'begin': 0x0780, 'end': 0x07BF }, // Thaana |
|
{ 'begin': 0x0D80, 'end': 0x0DFF }, // Sinhala |
|
{ 'begin': 0x1000, 'end': 0x109F }, // Myanmar |
|
{ 'begin': 0x1200, 'end': 0x137F }, // Ethiopic |
|
{ 'begin': 0x13A0, 'end': 0x13FF }, // Cherokee |
|
{ 'begin': 0x1400, 'end': 0x167F }, // Unified Canadian Aboriginal Syllabics |
|
{ 'begin': 0x1680, 'end': 0x169F }, // Ogham |
|
{ 'begin': 0x16A0, 'end': 0x16FF }, // Runic |
|
{ 'begin': 0x1780, 'end': 0x17FF }, // Khmer |
|
{ 'begin': 0x1800, 'end': 0x18AF }, // Mongolian |
|
{ 'begin': 0x2800, 'end': 0x28FF }, // Braille Patterns |
|
{ 'begin': 0xA000, 'end': 0xA48F }, // Yi Syllables |
|
{ 'begin': 0x1700, 'end': 0x171F }, // Tagalog |
|
{ 'begin': 0x10300, 'end': 0x1032F }, // Old Italic |
|
{ 'begin': 0x10330, 'end': 0x1034F }, // Gothic |
|
{ 'begin': 0x10400, 'end': 0x1044F }, // Deseret |
|
{ 'begin': 0x1D000, 'end': 0x1D0FF }, // Byzantine Musical Symbols |
|
{ 'begin': 0x1D400, 'end': 0x1D7FF }, // Mathematical Alphanumeric Symbols |
|
{ 'begin': 0xFF000, 'end': 0xFFFFD }, // Private Use (plane 15) |
|
{ 'begin': 0xFE00, 'end': 0xFE0F }, // Variation Selectors |
|
{ 'begin': 0xE0000, 'end': 0xE007F }, // Tags |
|
{ 'begin': 0x1900, 'end': 0x194F }, // Limbu |
|
{ 'begin': 0x1950, 'end': 0x197F }, // Tai Le |
|
{ 'begin': 0x1980, 'end': 0x19DF }, // New Tai Lue |
|
{ 'begin': 0x1A00, 'end': 0x1A1F }, // Buginese |
|
{ 'begin': 0x2C00, 'end': 0x2C5F }, // Glagolitic |
|
{ 'begin': 0x2D30, 'end': 0x2D7F }, // Tifinagh |
|
{ 'begin': 0x4DC0, 'end': 0x4DFF }, // Yijing Hexagram Symbols |
|
{ 'begin': 0xA800, 'end': 0xA82F }, // Syloti Nagri |
|
{ 'begin': 0x10000, 'end': 0x1007F }, // Linear B Syllabary |
|
{ 'begin': 0x10140, 'end': 0x1018F }, // Ancient Greek Numbers |
|
{ 'begin': 0x10380, 'end': 0x1039F }, // Ugaritic |
|
{ 'begin': 0x103A0, 'end': 0x103DF }, // Old Persian |
|
{ 'begin': 0x10450, 'end': 0x1047F }, // Shavian |
|
{ 'begin': 0x10480, 'end': 0x104AF }, // Osmanya |
|
{ 'begin': 0x10800, 'end': 0x1083F }, // Cypriot Syllabary |
|
{ 'begin': 0x10A00, 'end': 0x10A5F }, // Kharoshthi |
|
{ 'begin': 0x1D300, 'end': 0x1D35F }, // Tai Xuan Jing Symbols |
|
{ 'begin': 0x12000, 'end': 0x123FF }, // Cuneiform |
|
{ 'begin': 0x1D360, 'end': 0x1D37F }, // Counting Rod Numerals |
|
{ 'begin': 0x1B80, 'end': 0x1BBF }, // Sundanese |
|
{ 'begin': 0x1C00, 'end': 0x1C4F }, // Lepcha |
|
{ 'begin': 0x1C50, 'end': 0x1C7F }, // Ol Chiki |
|
{ 'begin': 0xA880, 'end': 0xA8DF }, // Saurashtra |
|
{ 'begin': 0xA900, 'end': 0xA92F }, // Kayah Li |
|
{ 'begin': 0xA930, 'end': 0xA95F }, // Rejang |
|
{ 'begin': 0xAA00, 'end': 0xAA5F }, // Cham |
|
{ 'begin': 0x10190, 'end': 0x101CF }, // Ancient Symbols |
|
{ 'begin': 0x101D0, 'end': 0x101FF }, // Phaistos Disc |
|
{ 'begin': 0x102A0, 'end': 0x102DF }, // Carian |
|
{ 'begin': 0x1F030, 'end': 0x1F09F } // Domino Tiles |
|
]; |
|
|
|
var MacStandardGlyphOrdering = [ |
|
'.notdef', '.null', 'nonmarkingreturn', 'space', 'exclam', 'quotedbl', |
|
'numbersign', 'dollar', 'percent', 'ampersand', 'quotesingle', 'parenleft', |
|
'parenright', 'asterisk', 'plus', 'comma', 'hyphen', 'period', 'slash', |
|
'zero', 'one', 'two', 'three', 'four', 'five', 'six', 'seven', 'eight', |
|
'nine', 'colon', 'semicolon', 'less', 'equal', 'greater', 'question', 'at', |
|
'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', |
|
'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 'bracketleft', |
|
'backslash', 'bracketright', 'asciicircum', 'underscore', 'grave', 'a', 'b', |
|
'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', |
|
'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', 'braceleft', 'bar', 'braceright', |
|
'asciitilde', 'Adieresis', 'Aring', 'Ccedilla', 'Eacute', 'Ntilde', |
|
'Odieresis', 'Udieresis', 'aacute', 'agrave', 'acircumflex', 'adieresis', |
|
'atilde', 'aring', 'ccedilla', 'eacute', 'egrave', 'ecircumflex', 'edieresis', |
|
'iacute', 'igrave', 'icircumflex', 'idieresis', 'ntilde', 'oacute', 'ograve', |
|
'ocircumflex', 'odieresis', 'otilde', 'uacute', 'ugrave', 'ucircumflex', |
|
'udieresis', 'dagger', 'degree', 'cent', 'sterling', 'section', 'bullet', |
|
'paragraph', 'germandbls', 'registered', 'copyright', 'trademark', 'acute', |
|
'dieresis', 'notequal', 'AE', 'Oslash', 'infinity', 'plusminus', 'lessequal', |
|
'greaterequal', 'yen', 'mu', 'partialdiff', 'summation', 'product', 'pi', |
|
'integral', 'ordfeminine', 'ordmasculine', 'Omega', 'ae', 'oslash', |
|
'questiondown', 'exclamdown', 'logicalnot', 'radical', 'florin', |
|
'approxequal', 'Delta', 'guillemotleft', 'guillemotright', 'ellipsis', |
|
'nonbreakingspace', 'Agrave', 'Atilde', 'Otilde', 'OE', 'oe', 'endash', |
|
'emdash', 'quotedblleft', 'quotedblright', 'quoteleft', 'quoteright', |
|
'divide', 'lozenge', 'ydieresis', 'Ydieresis', 'fraction', 'currency', |
|
'guilsinglleft', 'guilsinglright', 'fi', 'fl', 'daggerdbl', 'periodcentered', |
|
'quotesinglbase', 'quotedblbase', 'perthousand', 'Acircumflex', |
|
'Ecircumflex', 'Aacute', 'Edieresis', 'Egrave', 'Iacute', 'Icircumflex', |
|
'Idieresis', 'Igrave', 'Oacute', 'Ocircumflex', 'apple', 'Ograve', 'Uacute', |
|
'Ucircumflex', 'Ugrave', 'dotlessi', 'circumflex', 'tilde', 'macron', |
|
'breve', 'dotaccent', 'ring', 'cedilla', 'hungarumlaut', 'ogonek', 'caron', |
|
'Lslash', 'lslash', 'Scaron', 'scaron', 'Zcaron', 'zcaron', 'brokenbar', |
|
'Eth', 'eth', 'Yacute', 'yacute', 'Thorn', 'thorn', 'minus', 'multiply', |
|
'onesuperior', 'twosuperior', 'threesuperior', 'onehalf', 'onequarter', |
|
'threequarters', 'franc', 'Gbreve', 'gbreve', 'Idotaccent', 'Scedilla', |
|
'scedilla', 'Cacute', 'cacute', 'Ccaron', 'ccaron', 'dcroat']; |
|
|
|
function getUnicodeRangeFor(value) { |
|
for (var i = 0, ii = UnicodeRanges.length; i < ii; i++) { |
|
var range = UnicodeRanges[i]; |
|
if (value >= range.begin && value < range.end) { |
|
return i; |
|
} |
|
} |
|
return -1; |
|
} |
|
|
|
function isRTLRangeFor(value) { |
|
var range = UnicodeRanges[13]; |
|
if (value >= range.begin && value < range.end) { |
|
return true; |
|
} |
|
range = UnicodeRanges[11]; |
|
if (value >= range.begin && value < range.end) { |
|
return true; |
|
} |
|
return false; |
|
} |
|
|
|
// The normalization table is obtained by filtering the Unicode characters |
|
// database with <compat> entries. |
|
var NormalizedUnicodes = { |
|
'\u00A8': '\u0020\u0308', |
|
'\u00AF': '\u0020\u0304', |
|
'\u00B4': '\u0020\u0301', |
|
'\u00B5': '\u03BC', |
|
'\u00B8': '\u0020\u0327', |
|
'\u0132': '\u0049\u004A', |
|
'\u0133': '\u0069\u006A', |
|
'\u013F': '\u004C\u00B7', |
|
'\u0140': '\u006C\u00B7', |
|
'\u0149': '\u02BC\u006E', |
|
'\u017F': '\u0073', |
|
'\u01C4': '\u0044\u017D', |
|
'\u01C5': '\u0044\u017E', |
|
'\u01C6': '\u0064\u017E', |
|
'\u01C7': '\u004C\u004A', |
|
'\u01C8': '\u004C\u006A', |
|
'\u01C9': '\u006C\u006A', |
|
'\u01CA': '\u004E\u004A', |
|
'\u01CB': '\u004E\u006A', |
|
'\u01CC': '\u006E\u006A', |
|
'\u01F1': '\u0044\u005A', |
|
'\u01F2': '\u0044\u007A', |
|
'\u01F3': '\u0064\u007A', |
|
'\u02D8': '\u0020\u0306', |
|
'\u02D9': '\u0020\u0307', |
|
'\u02DA': '\u0020\u030A', |
|
'\u02DB': '\u0020\u0328', |
|
'\u02DC': '\u0020\u0303', |
|
'\u02DD': '\u0020\u030B', |
|
'\u037A': '\u0020\u0345', |
|
'\u0384': '\u0020\u0301', |
|
'\u03D0': '\u03B2', |
|
'\u03D1': '\u03B8', |
|
'\u03D2': '\u03A5', |
|
'\u03D5': '\u03C6', |
|
'\u03D6': '\u03C0', |
|
'\u03F0': '\u03BA', |
|
'\u03F1': '\u03C1', |
|
'\u03F2': '\u03C2', |
|
'\u03F4': '\u0398', |
|
'\u03F5': '\u03B5', |
|
'\u03F9': '\u03A3', |
|
'\u0587': '\u0565\u0582', |
|
'\u0675': '\u0627\u0674', |
|
'\u0676': '\u0648\u0674', |
|
'\u0677': '\u06C7\u0674', |
|
'\u0678': '\u064A\u0674', |
|
'\u0E33': '\u0E4D\u0E32', |
|
'\u0EB3': '\u0ECD\u0EB2', |
|
'\u0EDC': '\u0EAB\u0E99', |
|
'\u0EDD': '\u0EAB\u0EA1', |
|
'\u0F77': '\u0FB2\u0F81', |
|
'\u0F79': '\u0FB3\u0F81', |
|
'\u1E9A': '\u0061\u02BE', |
|
'\u1FBD': '\u0020\u0313', |
|
'\u1FBF': '\u0020\u0313', |
|
'\u1FC0': '\u0020\u0342', |
|
'\u1FFE': '\u0020\u0314', |
|
'\u2002': '\u0020', |
|
'\u2003': '\u0020', |
|
'\u2004': '\u0020', |
|
'\u2005': '\u0020', |
|
'\u2006': '\u0020', |
|
'\u2008': '\u0020', |
|
'\u2009': '\u0020', |
|
'\u200A': '\u0020', |
|
'\u2017': '\u0020\u0333', |
|
'\u2024': '\u002E', |
|
'\u2025': '\u002E\u002E', |
|
'\u2026': '\u002E\u002E\u002E', |
|
'\u2033': '\u2032\u2032', |
|
'\u2034': '\u2032\u2032\u2032', |
|
'\u2036': '\u2035\u2035', |
|
'\u2037': '\u2035\u2035\u2035', |
|
'\u203C': '\u0021\u0021', |
|
'\u203E': '\u0020\u0305', |
|
'\u2047': '\u003F\u003F', |
|
'\u2048': '\u003F\u0021', |
|
'\u2049': '\u0021\u003F', |
|
'\u2057': '\u2032\u2032\u2032\u2032', |
|
'\u205F': '\u0020', |
|
'\u20A8': '\u0052\u0073', |
|
'\u2100': '\u0061\u002F\u0063', |
|
'\u2101': '\u0061\u002F\u0073', |
|
'\u2103': '\u00B0\u0043', |
|
'\u2105': '\u0063\u002F\u006F', |
|
'\u2106': '\u0063\u002F\u0075', |
|
'\u2107': '\u0190', |
|
'\u2109': '\u00B0\u0046', |
|
'\u2116': '\u004E\u006F', |
|
'\u2121': '\u0054\u0045\u004C', |
|
'\u2135': '\u05D0', |
|
'\u2136': '\u05D1', |
|
'\u2137': '\u05D2', |
|
'\u2138': '\u05D3', |
|
'\u213B': '\u0046\u0041\u0058', |
|
'\u2160': '\u0049', |
|
'\u2161': '\u0049\u0049', |
|
'\u2162': '\u0049\u0049\u0049', |
|
'\u2163': '\u0049\u0056', |
|
'\u2164': '\u0056', |
|
'\u2165': '\u0056\u0049', |
|
'\u2166': '\u0056\u0049\u0049', |
|
'\u2167': '\u0056\u0049\u0049\u0049', |
|
'\u2168': '\u0049\u0058', |
|
'\u2169': '\u0058', |
|
'\u216A': '\u0058\u0049', |
|
'\u216B': '\u0058\u0049\u0049', |
|
'\u216C': '\u004C', |
|
'\u216D': '\u0043', |
|
'\u216E': '\u0044', |
|
'\u216F': '\u004D', |
|
'\u2170': '\u0069', |
|
'\u2171': '\u0069\u0069', |
|
'\u2172': '\u0069\u0069\u0069', |
|
'\u2173': '\u0069\u0076', |
|
'\u2174': '\u0076', |
|
'\u2175': '\u0076\u0069', |
|
'\u2176': '\u0076\u0069\u0069', |
|
'\u2177': '\u0076\u0069\u0069\u0069', |
|
'\u2178': '\u0069\u0078', |
|
'\u2179': '\u0078', |
|
'\u217A': '\u0078\u0069', |
|
'\u217B': '\u0078\u0069\u0069', |
|
'\u217C': '\u006C', |
|
'\u217D': '\u0063', |
|
'\u217E': '\u0064', |
|
'\u217F': '\u006D', |
|
'\u222C': '\u222B\u222B', |
|
'\u222D': '\u222B\u222B\u222B', |
|
'\u222F': '\u222E\u222E', |
|
'\u2230': '\u222E\u222E\u222E', |
|
'\u2474': '\u0028\u0031\u0029', |
|
'\u2475': '\u0028\u0032\u0029', |
|
'\u2476': '\u0028\u0033\u0029', |
|
'\u2477': '\u0028\u0034\u0029', |
|
'\u2478': '\u0028\u0035\u0029', |
|
'\u2479': '\u0028\u0036\u0029', |
|
'\u247A': '\u0028\u0037\u0029', |
|
'\u247B': '\u0028\u0038\u0029', |
|
'\u247C': '\u0028\u0039\u0029', |
|
'\u247D': '\u0028\u0031\u0030\u0029', |
|
'\u247E': '\u0028\u0031\u0031\u0029', |
|
'\u247F': '\u0028\u0031\u0032\u0029', |
|
'\u2480': '\u0028\u0031\u0033\u0029', |
|
'\u2481': '\u0028\u0031\u0034\u0029', |
|
'\u2482': '\u0028\u0031\u0035\u0029', |
|
'\u2483': '\u0028\u0031\u0036\u0029', |
|
'\u2484': '\u0028\u0031\u0037\u0029', |
|
'\u2485': '\u0028\u0031\u0038\u0029', |
|
'\u2486': '\u0028\u0031\u0039\u0029', |
|
'\u2487': '\u0028\u0032\u0030\u0029', |
|
'\u2488': '\u0031\u002E', |
|
'\u2489': '\u0032\u002E', |
|
'\u248A': '\u0033\u002E', |
|
'\u248B': '\u0034\u002E', |
|
'\u248C': '\u0035\u002E', |
|
'\u248D': '\u0036\u002E', |
|
'\u248E': '\u0037\u002E', |
|
'\u248F': '\u0038\u002E', |
|
'\u2490': '\u0039\u002E', |
|
'\u2491': '\u0031\u0030\u002E', |
|
'\u2492': '\u0031\u0031\u002E', |
|
'\u2493': '\u0031\u0032\u002E', |
|
'\u2494': '\u0031\u0033\u002E', |
|
'\u2495': '\u0031\u0034\u002E', |
|
'\u2496': '\u0031\u0035\u002E', |
|
'\u2497': '\u0031\u0036\u002E', |
|
'\u2498': '\u0031\u0037\u002E', |
|
'\u2499': '\u0031\u0038\u002E', |
|
'\u249A': '\u0031\u0039\u002E', |
|
'\u249B': '\u0032\u0030\u002E', |
|
'\u249C': '\u0028\u0061\u0029', |
|
'\u249D': '\u0028\u0062\u0029', |
|
'\u249E': '\u0028\u0063\u0029', |
|
'\u249F': '\u0028\u0064\u0029', |
|
'\u24A0': '\u0028\u0065\u0029', |
|
'\u24A1': '\u0028\u0066\u0029', |
|
'\u24A2': '\u0028\u0067\u0029', |
|
'\u24A3': '\u0028\u0068\u0029', |
|
'\u24A4': '\u0028\u0069\u0029', |
|
'\u24A5': '\u0028\u006A\u0029', |
|
'\u24A6': '\u0028\u006B\u0029', |
|
'\u24A7': '\u0028\u006C\u0029', |
|
'\u24A8': '\u0028\u006D\u0029', |
|
'\u24A9': '\u0028\u006E\u0029', |
|
'\u24AA': '\u0028\u006F\u0029', |
|
'\u24AB': '\u0028\u0070\u0029', |
|
'\u24AC': '\u0028\u0071\u0029', |
|
'\u24AD': '\u0028\u0072\u0029', |
|
'\u24AE': '\u0028\u0073\u0029', |
|
'\u24AF': '\u0028\u0074\u0029', |
|
'\u24B0': '\u0028\u0075\u0029', |
|
'\u24B1': '\u0028\u0076\u0029', |
|
'\u24B2': '\u0028\u0077\u0029', |
|
'\u24B3': '\u0028\u0078\u0029', |
|
'\u24B4': '\u0028\u0079\u0029', |
|
'\u24B5': '\u0028\u007A\u0029', |
|
'\u2A0C': '\u222B\u222B\u222B\u222B', |
|
'\u2A74': '\u003A\u003A\u003D', |
|
'\u2A75': '\u003D\u003D', |
|
'\u2A76': '\u003D\u003D\u003D', |
|
'\u2E9F': '\u6BCD', |
|
'\u2EF3': '\u9F9F', |
|
'\u2F00': '\u4E00', |
|
'\u2F01': '\u4E28', |
|
'\u2F02': '\u4E36', |
|
'\u2F03': '\u4E3F', |
|
'\u2F04': '\u4E59', |
|
'\u2F05': '\u4E85', |
|
'\u2F06': '\u4E8C', |
|
'\u2F07': '\u4EA0', |
|
'\u2F08': '\u4EBA', |
|
'\u2F09': '\u513F', |
|
'\u2F0A': '\u5165', |
|
'\u2F0B': '\u516B', |
|
'\u2F0C': '\u5182', |
|
'\u2F0D': '\u5196', |
|
'\u2F0E': '\u51AB', |
|
'\u2F0F': '\u51E0', |
|
'\u2F10': '\u51F5', |
|
'\u2F11': '\u5200', |
|
'\u2F12': '\u529B', |
|
'\u2F13': '\u52F9', |
|
'\u2F14': '\u5315', |
|
'\u2F15': '\u531A', |
|
'\u2F16': '\u5338', |
|
'\u2F17': '\u5341', |
|
'\u2F18': '\u535C', |
|
'\u2F19': '\u5369', |
|
'\u2F1A': '\u5382', |
|
'\u2F1B': '\u53B6', |
|
'\u2F1C': '\u53C8', |
|
'\u2F1D': '\u53E3', |
|
'\u2F1E': '\u56D7', |
|
'\u2F1F': '\u571F', |
|
'\u2F20': '\u58EB', |
|
'\u2F21': '\u5902', |
|
'\u2F22': '\u590A', |
|
'\u2F23': '\u5915', |
|
'\u2F24': '\u5927', |
|
'\u2F25': '\u5973', |
|
'\u2F26': '\u5B50', |
|
'\u2F27': '\u5B80', |
|
'\u2F28': '\u5BF8', |
|
'\u2F29': '\u5C0F', |
|
'\u2F2A': '\u5C22', |
|
'\u2F2B': '\u5C38', |
|
'\u2F2C': '\u5C6E', |
|
'\u2F2D': '\u5C71', |
|
'\u2F2E': '\u5DDB', |
|
'\u2F2F': '\u5DE5', |
|
'\u2F30': '\u5DF1', |
|
'\u2F31': '\u5DFE', |
|
'\u2F32': '\u5E72', |
|
'\u2F33': '\u5E7A', |
|
'\u2F34': '\u5E7F', |
|
'\u2F35': '\u5EF4', |
|
'\u2F36': '\u5EFE', |
|
'\u2F37': '\u5F0B', |
|
'\u2F38': '\u5F13', |
|
'\u2F39': '\u5F50', |
|
'\u2F3A': '\u5F61', |
|
'\u2F3B': '\u5F73', |
|
'\u2F3C': '\u5FC3', |
|
'\u2F3D': '\u6208', |
|
'\u2F3E': '\u6236', |
|
'\u2F3F': '\u624B', |
|
'\u2F40': '\u652F', |
|
'\u2F41': '\u6534', |
|
'\u2F42': '\u6587', |
|
'\u2F43': '\u6597', |
|
'\u2F44': '\u65A4', |
|
'\u2F45': '\u65B9', |
|
'\u2F46': '\u65E0', |
|
'\u2F47': '\u65E5', |
|
'\u2F48': '\u66F0', |
|
'\u2F49': '\u6708', |
|
'\u2F4A': '\u6728', |
|
'\u2F4B': '\u6B20', |
|
'\u2F4C': '\u6B62', |
|
'\u2F4D': '\u6B79', |
|
'\u2F4E': '\u6BB3', |
|
'\u2F4F': '\u6BCB', |
|
'\u2F50': '\u6BD4', |
|
'\u2F51': '\u6BDB', |
|
'\u2F52': '\u6C0F', |
|
'\u2F53': '\u6C14', |
|
'\u2F54': '\u6C34', |
|
'\u2F55': '\u706B', |
|
'\u2F56': '\u722A', |
|
'\u2F57': '\u7236', |
|
'\u2F58': '\u723B', |
|
'\u2F59': '\u723F', |
|
'\u2F5A': '\u7247', |
|
'\u2F5B': '\u7259', |
|
'\u2F5C': '\u725B', |
|
'\u2F5D': '\u72AC', |
|
'\u2F5E': '\u7384', |
|
'\u2F5F': '\u7389', |
|
'\u2F60': '\u74DC', |
|
'\u2F61': '\u74E6', |
|
'\u2F62': '\u7518', |
|
'\u2F63': '\u751F', |
|
'\u2F64': '\u7528', |
|
'\u2F65': '\u7530', |
|
'\u2F66': '\u758B', |
|
'\u2F67': '\u7592', |
|
'\u2F68': '\u7676', |
|
'\u2F69': '\u767D', |
|
'\u2F6A': '\u76AE', |
|
'\u2F6B': '\u76BF', |
|
'\u2F6C': '\u76EE', |
|
'\u2F6D': '\u77DB', |
|
'\u2F6E': '\u77E2', |
|
'\u2F6F': '\u77F3', |
|
'\u2F70': '\u793A', |
|
'\u2F71': '\u79B8', |
|
'\u2F72': '\u79BE', |
|
'\u2F73': '\u7A74', |
|
'\u2F74': '\u7ACB', |
|
'\u2F75': '\u7AF9', |
|
'\u2F76': '\u7C73', |
|
'\u2F77': '\u7CF8', |
|
'\u2F78': '\u7F36', |
|
'\u2F79': '\u7F51', |
|
'\u2F7A': '\u7F8A', |
|
'\u2F7B': '\u7FBD', |
|
'\u2F7C': '\u8001', |
|
'\u2F7D': '\u800C', |
|
'\u2F7E': '\u8012', |
|
'\u2F7F': '\u8033', |
|
'\u2F80': '\u807F', |
|
'\u2F81': '\u8089', |
|
'\u2F82': '\u81E3', |
|
'\u2F83': '\u81EA', |
|
'\u2F84': '\u81F3', |
|
'\u2F85': '\u81FC', |
|
'\u2F86': '\u820C', |
|
'\u2F87': '\u821B', |
|
'\u2F88': '\u821F', |
|
'\u2F89': '\u826E', |
|
'\u2F8A': '\u8272', |
|
'\u2F8B': '\u8278', |
|
'\u2F8C': '\u864D', |
|
'\u2F8D': '\u866B', |
|
'\u2F8E': '\u8840', |
|
'\u2F8F': '\u884C', |
|
'\u2F90': '\u8863', |
|
'\u2F91': '\u897E', |
|
'\u2F92': '\u898B', |
|
'\u2F93': '\u89D2', |
|
'\u2F94': '\u8A00', |
|
'\u2F95': '\u8C37', |
|
'\u2F96': '\u8C46', |
|
'\u2F97': '\u8C55', |
|
'\u2F98': '\u8C78', |
|
'\u2F99': '\u8C9D', |
|
'\u2F9A': '\u8D64', |
|
'\u2F9B': '\u8D70', |
|
'\u2F9C': '\u8DB3', |
|
'\u2F9D': '\u8EAB', |
|
'\u2F9E': '\u8ECA', |
|
'\u2F9F': '\u8F9B', |
|
'\u2FA0': '\u8FB0', |
|
'\u2FA1': '\u8FB5', |
|
'\u2FA2': '\u9091', |
|
'\u2FA3': '\u9149', |
|
'\u2FA4': '\u91C6', |
|
'\u2FA5': '\u91CC', |
|
'\u2FA6': '\u91D1', |
|
'\u2FA7': '\u9577', |
|
'\u2FA8': '\u9580', |
|
'\u2FA9': '\u961C', |
|
'\u2FAA': '\u96B6', |
|
'\u2FAB': '\u96B9', |
|
'\u2FAC': '\u96E8', |
|
'\u2FAD': '\u9751', |
|
'\u2FAE': '\u975E', |
|
'\u2FAF': '\u9762', |
|
'\u2FB0': '\u9769', |
|
'\u2FB1': '\u97CB', |
|
'\u2FB2': '\u97ED', |
|
'\u2FB3': '\u97F3', |
|
'\u2FB4': '\u9801', |
|
'\u2FB5': '\u98A8', |
|
'\u2FB6': '\u98DB', |
|
'\u2FB7': '\u98DF', |
|
'\u2FB8': '\u9996', |
|
'\u2FB9': '\u9999', |
|
'\u2FBA': '\u99AC', |
|
'\u2FBB': '\u9AA8', |
|
'\u2FBC': '\u9AD8', |
|
'\u2FBD': '\u9ADF', |
|
'\u2FBE': '\u9B25', |
|
'\u2FBF': '\u9B2F', |
|
'\u2FC0': '\u9B32', |
|
'\u2FC1': '\u9B3C', |
|
'\u2FC2': '\u9B5A', |
|
'\u2FC3': '\u9CE5', |
|
'\u2FC4': '\u9E75', |
|
'\u2FC5': '\u9E7F', |
|
'\u2FC6': '\u9EA5', |
|
'\u2FC7': '\u9EBB', |
|
'\u2FC8': '\u9EC3', |
|
'\u2FC9': '\u9ECD', |
|
'\u2FCA': '\u9ED1', |
|
'\u2FCB': '\u9EF9', |
|
'\u2FCC': '\u9EFD', |
|
'\u2FCD': '\u9F0E', |
|
'\u2FCE': '\u9F13', |
|
'\u2FCF': '\u9F20', |
|
'\u2FD0': '\u9F3B', |
|
'\u2FD1': '\u9F4A', |
|
'\u2FD2': '\u9F52', |
|
'\u2FD3': '\u9F8D', |
|
'\u2FD4': '\u9F9C', |
|
'\u2FD5': '\u9FA0', |
|
'\u3036': '\u3012', |
|
'\u3038': '\u5341', |
|
'\u3039': '\u5344', |
|
'\u303A': '\u5345', |
|
'\u309B': '\u0020\u3099', |
|
'\u309C': '\u0020\u309A', |
|
'\u3131': '\u1100', |
|
'\u3132': '\u1101', |
|
'\u3133': '\u11AA', |
|
'\u3134': '\u1102', |
|
'\u3135': '\u11AC', |
|
'\u3136': '\u11AD', |
|
'\u3137': '\u1103', |
|
'\u3138': '\u1104', |
|
'\u3139': '\u1105', |
|
'\u313A': '\u11B0', |
|
'\u313B': '\u11B1', |
|
'\u313C': '\u11B2', |
|
'\u313D': '\u11B3', |
|
'\u313E': '\u11B4', |
|
'\u313F': '\u11B5', |
|
'\u3140': '\u111A', |
|
'\u3141': '\u1106', |
|
'\u3142': '\u1107', |
|
'\u3143': '\u1108', |
|
'\u3144': '\u1121', |
|
'\u3145': '\u1109', |
|
'\u3146': '\u110A', |
|
'\u3147': '\u110B', |
|
'\u3148': '\u110C', |
|
'\u3149': '\u110D', |
|
'\u314A': '\u110E', |
|
'\u314B': '\u110F', |
|
'\u314C': '\u1110', |
|
'\u314D': '\u1111', |
|
'\u314E': '\u1112', |
|
'\u314F': '\u1161', |
|
'\u3150': '\u1162', |
|
'\u3151': '\u1163', |
|
'\u3152': '\u1164', |
|
'\u3153': '\u1165', |
|
'\u3154': '\u1166', |
|
'\u3155': '\u1167', |
|
'\u3156': '\u1168', |
|
'\u3157': '\u1169', |
|
'\u3158': '\u116A', |
|
'\u3159': '\u116B', |
|
'\u315A': '\u116C', |
|
'\u315B': '\u116D', |
|
'\u315C': '\u116E', |
|
'\u315D': '\u116F', |
|
'\u315E': '\u1170', |
|
'\u315F': '\u1171', |
|
'\u3160': '\u1172', |
|
'\u3161': '\u1173', |
|
'\u3162': '\u1174', |
|
'\u3163': '\u1175', |
|
'\u3164': '\u1160', |
|
'\u3165': '\u1114', |
|
'\u3166': '\u1115', |
|
'\u3167': '\u11C7', |
|
'\u3168': '\u11C8', |
|
'\u3169': '\u11CC', |
|
'\u316A': '\u11CE', |
|
'\u316B': '\u11D3', |
|
'\u316C': '\u11D7', |
|
'\u316D': '\u11D9', |
|
'\u316E': '\u111C', |
|
'\u316F': '\u11DD', |
|
'\u3170': '\u11DF', |
|
'\u3171': '\u111D', |
|
'\u3172': '\u111E', |
|
'\u3173': '\u1120', |
|
'\u3174': '\u1122', |
|
'\u3175': '\u1123', |
|
'\u3176': '\u1127', |
|
'\u3177': '\u1129', |
|
'\u3178': '\u112B', |
|
'\u3179': '\u112C', |
|
'\u317A': '\u112D', |
|
'\u317B': '\u112E', |
|
'\u317C': '\u112F', |
|
'\u317D': '\u1132', |
|
'\u317E': '\u1136', |
|
'\u317F': '\u1140', |
|
'\u3180': '\u1147', |
|
'\u3181': '\u114C', |
|
'\u3182': '\u11F1', |
|
'\u3183': '\u11F2', |
|
'\u3184': '\u1157', |
|
'\u3185': '\u1158', |
|
'\u3186': '\u1159', |
|
'\u3187': '\u1184', |
|
'\u3188': '\u1185', |
|
'\u3189': '\u1188', |
|
'\u318A': '\u1191', |
|
'\u318B': '\u1192', |
|
'\u318C': '\u1194', |
|
'\u318D': '\u119E', |
|
'\u318E': '\u11A1', |
|
'\u3200': '\u0028\u1100\u0029', |
|
'\u3201': '\u0028\u1102\u0029', |
|
'\u3202': '\u0028\u1103\u0029', |
|
'\u3203': '\u0028\u1105\u0029', |
|
'\u3204': '\u0028\u1106\u0029', |
|
'\u3205': '\u0028\u1107\u0029', |
|
'\u3206': '\u0028\u1109\u0029', |
|
'\u3207': '\u0028\u110B\u0029', |
|
'\u3208': '\u0028\u110C\u0029', |
|
'\u3209': '\u0028\u110E\u0029', |
|
'\u320A': '\u0028\u110F\u0029', |
|
'\u320B': '\u0028\u1110\u0029', |
|
'\u320C': '\u0028\u1111\u0029', |
|
'\u320D': '\u0028\u1112\u0029', |
|
'\u320E': '\u0028\u1100\u1161\u0029', |
|
'\u320F': '\u0028\u1102\u1161\u0029', |
|
'\u3210': '\u0028\u1103\u1161\u0029', |
|
'\u3211': '\u0028\u1105\u1161\u0029', |
|
'\u3212': '\u0028\u1106\u1161\u0029', |
|
'\u3213': '\u0028\u1107\u1161\u0029', |
|
'\u3214': '\u0028\u1109\u1161\u0029', |
|
'\u3215': '\u0028\u110B\u1161\u0029', |
|
'\u3216': '\u0028\u110C\u1161\u0029', |
|
'\u3217': '\u0028\u110E\u1161\u0029', |
|
'\u3218': '\u0028\u110F\u1161\u0029', |
|
'\u3219': '\u0028\u1110\u1161\u0029', |
|
'\u321A': '\u0028\u1111\u1161\u0029', |
|
'\u321B': '\u0028\u1112\u1161\u0029', |
|
'\u321C': '\u0028\u110C\u116E\u0029', |
|
'\u321D': '\u0028\u110B\u1169\u110C\u1165\u11AB\u0029', |
|
'\u321E': '\u0028\u110B\u1169\u1112\u116E\u0029', |
|
'\u3220': '\u0028\u4E00\u0029', |
|
'\u3221': '\u0028\u4E8C\u0029', |
|
'\u3222': '\u0028\u4E09\u0029', |
|
'\u3223': '\u0028\u56DB\u0029', |
|
'\u3224': '\u0028\u4E94\u0029', |
|
'\u3225': '\u0028\u516D\u0029', |
|
'\u3226': '\u0028\u4E03\u0029', |
|
'\u3227': '\u0028\u516B\u0029', |
|
'\u3228': '\u0028\u4E5D\u0029', |
|
'\u3229': '\u0028\u5341\u0029', |
|
'\u322A': '\u0028\u6708\u0029', |
|
'\u322B': '\u0028\u706B\u0029', |
|
'\u322C': '\u0028\u6C34\u0029', |
|
'\u322D': '\u0028\u6728\u0029', |
|
'\u322E': '\u0028\u91D1\u0029', |
|
'\u322F': '\u0028\u571F\u0029', |
|
'\u3230': '\u0028\u65E5\u0029', |
|
'\u3231': '\u0028\u682A\u0029', |
|
'\u3232': '\u0028\u6709\u0029', |
|
'\u3233': '\u0028\u793E\u0029', |
|
'\u3234': '\u0028\u540D\u0029', |
|
'\u3235': '\u0028\u7279\u0029', |
|
'\u3236': '\u0028\u8CA1\u0029', |
|
'\u3237': '\u0028\u795D\u0029', |
|
'\u3238': '\u0028\u52B4\u0029', |
|
'\u3239': '\u0028\u4EE3\u0029', |
|
'\u323A': '\u0028\u547C\u0029', |
|
'\u323B': '\u0028\u5B66\u0029', |
|
'\u323C': '\u0028\u76E3\u0029', |
|
'\u323D': '\u0028\u4F01\u0029', |
|
'\u323E': '\u0028\u8CC7\u0029', |
|
'\u323F': '\u0028\u5354\u0029', |
|
'\u3240': '\u0028\u796D\u0029', |
|
'\u3241': '\u0028\u4F11\u0029', |
|
'\u3242': '\u0028\u81EA\u0029', |
|
'\u3243': '\u0028\u81F3\u0029', |
|
'\u32C0': '\u0031\u6708', |
|
'\u32C1': '\u0032\u6708', |
|
'\u32C2': '\u0033\u6708', |
|
'\u32C3': '\u0034\u6708', |
|
'\u32C4': '\u0035\u6708', |
|
'\u32C5': '\u0036\u6708', |
|
'\u32C6': '\u0037\u6708', |
|
'\u32C7': '\u0038\u6708', |
|
'\u32C8': '\u0039\u6708', |
|
'\u32C9': '\u0031\u0030\u6708', |
|
'\u32CA': '\u0031\u0031\u6708', |
|
'\u32CB': '\u0031\u0032\u6708', |
|
'\u3358': '\u0030\u70B9', |
|
'\u3359': '\u0031\u70B9', |
|
'\u335A': '\u0032\u70B9', |
|
'\u335B': '\u0033\u70B9', |
|
'\u335C': '\u0034\u70B9', |
|
'\u335D': '\u0035\u70B9', |
|
'\u335E': '\u0036\u70B9', |
|
'\u335F': '\u0037\u70B9', |
|
'\u3360': '\u0038\u70B9', |
|
'\u3361': '\u0039\u70B9', |
|
'\u3362': '\u0031\u0030\u70B9', |
|
'\u3363': '\u0031\u0031\u70B9', |
|
'\u3364': '\u0031\u0032\u70B9', |
|
'\u3365': '\u0031\u0033\u70B9', |
|
'\u3366': '\u0031\u0034\u70B9', |
|
'\u3367': '\u0031\u0035\u70B9', |
|
'\u3368': '\u0031\u0036\u70B9', |
|
'\u3369': '\u0031\u0037\u70B9', |
|
'\u336A': '\u0031\u0038\u70B9', |
|
'\u336B': '\u0031\u0039\u70B9', |
|
'\u336C': '\u0032\u0030\u70B9', |
|
'\u336D': '\u0032\u0031\u70B9', |
|
'\u336E': '\u0032\u0032\u70B9', |
|
'\u336F': '\u0032\u0033\u70B9', |
|
'\u3370': '\u0032\u0034\u70B9', |
|
'\u33E0': '\u0031\u65E5', |
|
'\u33E1': '\u0032\u65E5', |
|
'\u33E2': '\u0033\u65E5', |
|
'\u33E3': '\u0034\u65E5', |
|
'\u33E4': '\u0035\u65E5', |
|
'\u33E5': '\u0036\u65E5', |
|
'\u33E6': '\u0037\u65E5', |
|
'\u33E7': '\u0038\u65E5', |
|
'\u33E8': '\u0039\u65E5', |
|
'\u33E9': '\u0031\u0030\u65E5', |
|
'\u33EA': '\u0031\u0031\u65E5', |
|
'\u33EB': '\u0031\u0032\u65E5', |
|
'\u33EC': '\u0031\u0033\u65E5', |
|
'\u33ED': '\u0031\u0034\u65E5', |
|
'\u33EE': '\u0031\u0035\u65E5', |
|
'\u33EF': '\u0031\u0036\u65E5', |
|
'\u33F0': '\u0031\u0037\u65E5', |
|
'\u33F1': '\u0031\u0038\u65E5', |
|
'\u33F2': '\u0031\u0039\u65E5', |
|
'\u33F3': '\u0032\u0030\u65E5', |
|
'\u33F4': '\u0032\u0031\u65E5', |
|
'\u33F5': '\u0032\u0032\u65E5', |
|
'\u33F6': '\u0032\u0033\u65E5', |
|
'\u33F7': '\u0032\u0034\u65E5', |
|
'\u33F8': '\u0032\u0035\u65E5', |
|
'\u33F9': '\u0032\u0036\u65E5', |
|
'\u33FA': '\u0032\u0037\u65E5', |
|
'\u33FB': '\u0032\u0038\u65E5', |
|
'\u33FC': '\u0032\u0039\u65E5', |
|
'\u33FD': '\u0033\u0030\u65E5', |
|
'\u33FE': '\u0033\u0031\u65E5', |
|
'\uFB00': '\u0066\u0066', |
|
'\uFB01': '\u0066\u0069', |
|
'\uFB02': '\u0066\u006C', |
|
'\uFB03': '\u0066\u0066\u0069', |
|
'\uFB04': '\u0066\u0066\u006C', |
|
'\uFB05': '\u017F\u0074', |
|
'\uFB06': '\u0073\u0074', |
|
'\uFB13': '\u0574\u0576', |
|
'\uFB14': '\u0574\u0565', |
|
'\uFB15': '\u0574\u056B', |
|
'\uFB16': '\u057E\u0576', |
|
'\uFB17': '\u0574\u056D', |
|
'\uFB4F': '\u05D0\u05DC', |
|
'\uFB50': '\u0671', |
|
'\uFB51': '\u0671', |
|
'\uFB52': '\u067B', |
|
'\uFB53': '\u067B', |
|
'\uFB54': '\u067B', |
|
'\uFB55': '\u067B', |
|
'\uFB56': '\u067E', |
|
'\uFB57': '\u067E', |
|
'\uFB58': '\u067E', |
|
'\uFB59': '\u067E', |
|
'\uFB5A': '\u0680', |
|
'\uFB5B': '\u0680', |
|
'\uFB5C': '\u0680', |
|
'\uFB5D': '\u0680', |
|
'\uFB5E': '\u067A', |
|
'\uFB5F': '\u067A', |
|
'\uFB60': '\u067A', |
|
'\uFB61': '\u067A', |
|
'\uFB62': '\u067F', |
|
'\uFB63': '\u067F', |
|
'\uFB64': '\u067F', |
|
'\uFB65': '\u067F', |
|
'\uFB66': '\u0679', |
|
'\uFB67': '\u0679', |
|
'\uFB68': '\u0679', |
|
'\uFB69': '\u0679', |
|
'\uFB6A': '\u06A4', |
|
'\uFB6B': '\u06A4', |
|
'\uFB6C': '\u06A4', |
|
'\uFB6D': '\u06A4', |
|
'\uFB6E': '\u06A6', |
|
'\uFB6F': '\u06A6', |
|
'\uFB70': '\u06A6', |
|
'\uFB71': '\u06A6', |
|
'\uFB72': '\u0684', |
|
'\uFB73': '\u0684', |
|
'\uFB74': '\u0684', |
|
'\uFB75': '\u0684', |
|
'\uFB76': '\u0683', |
|
'\uFB77': '\u0683', |
|
'\uFB78': '\u0683', |
|
'\uFB79': '\u0683', |
|
'\uFB7A': '\u0686', |
|
'\uFB7B': '\u0686', |
|
'\uFB7C': '\u0686', |
|
'\uFB7D': '\u0686', |
|
'\uFB7E': '\u0687', |
|
'\uFB7F': '\u0687', |
|
'\uFB80': '\u0687', |
|
'\uFB81': '\u0687', |
|
'\uFB82': '\u068D', |
|
'\uFB83': '\u068D', |
|
'\uFB84': '\u068C', |
|
'\uFB85': '\u068C', |
|
'\uFB86': '\u068E', |
|
'\uFB87': '\u068E', |
|
'\uFB88': '\u0688', |
|
'\uFB89': '\u0688', |
|
'\uFB8A': '\u0698', |
|
'\uFB8B': '\u0698', |
|
'\uFB8C': '\u0691', |
|
'\uFB8D': '\u0691', |
|
'\uFB8E': '\u06A9', |
|
'\uFB8F': '\u06A9', |
|
'\uFB90': '\u06A9', |
|
'\uFB91': '\u06A9', |
|
'\uFB92': '\u06AF', |
|
'\uFB93': '\u06AF', |
|
'\uFB94': '\u06AF', |
|
'\uFB95': '\u06AF', |
|
'\uFB96': '\u06B3', |
|
'\uFB97': '\u06B3', |
|
'\uFB98': '\u06B3', |
|
'\uFB99': '\u06B3', |
|
'\uFB9A': '\u06B1', |
|
'\uFB9B': '\u06B1', |
|
'\uFB9C': '\u06B1', |
|
'\uFB9D': '\u06B1', |
|
'\uFB9E': '\u06BA', |
|
'\uFB9F': '\u06BA', |
|
'\uFBA0': '\u06BB', |
|
'\uFBA1': '\u06BB', |
|
'\uFBA2': '\u06BB', |
|
'\uFBA3': '\u06BB', |
|
'\uFBA4': '\u06C0', |
|
'\uFBA5': '\u06C0', |
|
'\uFBA6': '\u06C1', |
|
'\uFBA7': '\u06C1', |
|
'\uFBA8': '\u06C1', |
|
'\uFBA9': '\u06C1', |
|
'\uFBAA': '\u06BE', |
|
'\uFBAB': '\u06BE', |
|
'\uFBAC': '\u06BE', |
|
'\uFBAD': '\u06BE', |
|
'\uFBAE': '\u06D2', |
|
'\uFBAF': '\u06D2', |
|
'\uFBB0': '\u06D3', |
|
'\uFBB1': '\u06D3', |
|
'\uFBD3': '\u06AD', |
|
'\uFBD4': '\u06AD', |
|
'\uFBD5': '\u06AD', |
|
'\uFBD6': '\u06AD', |
|
'\uFBD7': '\u06C7', |
|
'\uFBD8': '\u06C7', |
|
'\uFBD9': '\u06C6', |
|
'\uFBDA': '\u06C6', |
|
'\uFBDB': '\u06C8', |
|
'\uFBDC': '\u06C8', |
|
'\uFBDD': '\u0677', |
|
'\uFBDE': '\u06CB', |
|
'\uFBDF': '\u06CB', |
|
'\uFBE0': '\u06C5', |
|
'\uFBE1': '\u06C5', |
|
'\uFBE2': '\u06C9', |
|
'\uFBE3': '\u06C9', |
|
'\uFBE4': '\u06D0', |
|
'\uFBE5': '\u06D0', |
|
'\uFBE6': '\u06D0', |
|
'\uFBE7': '\u06D0', |
|
'\uFBE8': '\u0649', |
|
'\uFBE9': '\u0649', |
|
'\uFBEA': '\u0626\u0627', |
|
'\uFBEB': '\u0626\u0627', |
|
'\uFBEC': '\u0626\u06D5', |
|
'\uFBED': '\u0626\u06D5', |
|
'\uFBEE': '\u0626\u0648', |
|
'\uFBEF': '\u0626\u0648', |
|
'\uFBF0': '\u0626\u06C7', |
|
'\uFBF1': '\u0626\u06C7', |
|
'\uFBF2': '\u0626\u06C6', |
|
'\uFBF3': '\u0626\u06C6', |
|
'\uFBF4': '\u0626\u06C8', |
|
'\uFBF5': '\u0626\u06C8', |
|
'\uFBF6': '\u0626\u06D0', |
|
'\uFBF7': '\u0626\u06D0', |
|
'\uFBF8': '\u0626\u06D0', |
|
'\uFBF9': '\u0626\u0649', |
|
'\uFBFA': '\u0626\u0649', |
|
'\uFBFB': '\u0626\u0649', |
|
'\uFBFC': '\u06CC', |
|
'\uFBFD': '\u06CC', |
|
'\uFBFE': '\u06CC', |
|
'\uFBFF': '\u06CC', |
|
'\uFC00': '\u0626\u062C', |
|
'\uFC01': '\u0626\u062D', |
|
'\uFC02': '\u0626\u0645', |
|
'\uFC03': '\u0626\u0649', |
|
'\uFC04': '\u0626\u064A', |
|
'\uFC05': '\u0628\u062C', |
|
'\uFC06': '\u0628\u062D', |
|
'\uFC07': '\u0628\u062E', |
|
'\uFC08': '\u0628\u0645', |
|
'\uFC09': '\u0628\u0649', |
|
'\uFC0A': '\u0628\u064A', |
|
'\uFC0B': '\u062A\u062C', |
|
'\uFC0C': '\u062A\u062D', |
|
'\uFC0D': '\u062A\u062E', |
|
'\uFC0E': '\u062A\u0645', |
|
'\uFC0F': '\u062A\u0649', |
|
'\uFC10': '\u062A\u064A', |
|
'\uFC11': '\u062B\u062C', |
|
'\uFC12': '\u062B\u0645', |
|
'\uFC13': '\u062B\u0649', |
|
'\uFC14': '\u062B\u064A', |
|
'\uFC15': '\u062C\u062D', |
|
'\uFC16': '\u062C\u0645', |
|
'\uFC17': '\u062D\u062C', |
|
'\uFC18': '\u062D\u0645', |
|
'\uFC19': '\u062E\u062C', |
|
'\uFC1A': '\u062E\u062D', |
|
'\uFC1B': '\u062E\u0645', |
|
'\uFC1C': '\u0633\u062C', |
|
'\uFC1D': '\u0633\u062D', |
|
'\uFC1E': '\u0633\u062E', |
|
'\uFC1F': '\u0633\u0645', |
|
'\uFC20': '\u0635\u062D', |
|
'\uFC21': '\u0635\u0645', |
|
'\uFC22': '\u0636\u062C', |
|
'\uFC23': '\u0636\u062D', |
|
'\uFC24': '\u0636\u062E', |
|
'\uFC25': '\u0636\u0645', |
|
'\uFC26': '\u0637\u062D', |
|
'\uFC27': '\u0637\u0645', |
|
'\uFC28': '\u0638\u0645', |
|
'\uFC29': '\u0639\u062C', |
|
'\uFC2A': '\u0639\u0645', |
|
'\uFC2B': '\u063A\u062C', |
|
'\uFC2C': '\u063A\u0645', |
|
'\uFC2D': '\u0641\u062C', |
|
'\uFC2E': '\u0641\u062D', |
|
'\uFC2F': '\u0641\u062E', |
|
'\uFC30': '\u0641\u0645', |
|
'\uFC31': '\u0641\u0649', |
|
'\uFC32': '\u0641\u064A', |
|
'\uFC33': '\u0642\u062D', |
|
'\uFC34': '\u0642\u0645', |
|
'\uFC35': '\u0642\u0649', |
|
'\uFC36': '\u0642\u064A', |
|
'\uFC37': '\u0643\u0627', |
|
'\uFC38': '\u0643\u062C', |
|
'\uFC39': '\u0643\u062D', |
|
'\uFC3A': '\u0643\u062E', |
|
'\uFC3B': '\u0643\u0644', |
|
'\uFC3C': '\u0643\u0645', |
|
'\uFC3D': '\u0643\u0649', |
|
'\uFC3E': '\u0643\u064A', |
|
'\uFC3F': '\u0644\u062C', |
|
'\uFC40': '\u0644\u062D', |
|
'\uFC41': '\u0644\u062E', |
|
'\uFC42': '\u0644\u0645', |
|
'\uFC43': '\u0644\u0649', |
|
'\uFC44': '\u0644\u064A', |
|
'\uFC45': '\u0645\u062C', |
|
'\uFC46': '\u0645\u062D', |
|
'\uFC47': '\u0645\u062E', |
|
'\uFC48': '\u0645\u0645', |
|
'\uFC49': '\u0645\u0649', |
|
'\uFC4A': '\u0645\u064A', |
|
'\uFC4B': '\u0646\u062C', |
|
'\uFC4C': '\u0646\u062D', |
|
'\uFC4D': '\u0646\u062E', |
|
'\uFC4E': '\u0646\u0645', |
|
'\uFC4F': '\u0646\u0649', |
|
'\uFC50': '\u0646\u064A', |
|
'\uFC51': '\u0647\u062C', |
|
'\uFC52': '\u0647\u0645', |
|
'\uFC53': '\u0647\u0649', |
|
'\uFC54': '\u0647\u064A', |
|
'\uFC55': '\u064A\u062C', |
|
'\uFC56': '\u064A\u062D', |
|
'\uFC57': '\u064A\u062E', |
|
'\uFC58': '\u064A\u0645', |
|
'\uFC59': '\u064A\u0649', |
|
'\uFC5A': '\u064A\u064A', |
|
'\uFC5B': '\u0630\u0670', |
|
'\uFC5C': '\u0631\u0670', |
|
'\uFC5D': '\u0649\u0670', |
|
'\uFC5E': '\u0020\u064C\u0651', |
|
'\uFC5F': '\u0020\u064D\u0651', |
|
'\uFC60': '\u0020\u064E\u0651', |
|
'\uFC61': '\u0020\u064F\u0651', |
|
'\uFC62': '\u0020\u0650\u0651', |
|
'\uFC63': '\u0020\u0651\u0670', |
|
'\uFC64': '\u0626\u0631', |
|
'\uFC65': '\u0626\u0632', |
|
'\uFC66': '\u0626\u0645', |
|
'\uFC67': '\u0626\u0646', |
|
'\uFC68': '\u0626\u0649', |
|
'\uFC69': '\u0626\u064A', |
|
'\uFC6A': '\u0628\u0631', |
|
'\uFC6B': '\u0628\u0632', |
|
'\uFC6C': '\u0628\u0645', |
|
'\uFC6D': '\u0628\u0646', |
|
'\uFC6E': '\u0628\u0649', |
|
'\uFC6F': '\u0628\u064A', |
|
'\uFC70': '\u062A\u0631', |
|
'\uFC71': '\u062A\u0632', |
|
'\uFC72': '\u062A\u0645', |
|
'\uFC73': '\u062A\u0646', |
|
'\uFC74': '\u062A\u0649', |
|
'\uFC75': '\u062A\u064A', |
|
'\uFC76': '\u062B\u0631', |
|
'\uFC77': '\u062B\u0632', |
|
'\uFC78': '\u062B\u0645', |
|
'\uFC79': '\u062B\u0646', |
|
'\uFC7A': '\u062B\u0649', |
|
'\uFC7B': '\u062B\u064A', |
|
'\uFC7C': '\u0641\u0649', |
|
'\uFC7D': '\u0641\u064A', |
|
'\uFC7E': '\u0642\u0649', |
|
'\uFC7F': '\u0642\u064A', |
|
'\uFC80': '\u0643\u0627', |
|
'\uFC81': '\u0643\u0644', |
|
'\uFC82': '\u0643\u0645', |
|
'\uFC83': '\u0643\u0649', |
|
'\uFC84': '\u0643\u064A', |
|
'\uFC85': '\u0644\u0645', |
|
'\uFC86': '\u0644\u0649', |
|
'\uFC87': '\u0644\u064A', |
|
'\uFC88': '\u0645\u0627', |
|
'\uFC89': '\u0645\u0645', |
|
'\uFC8A': '\u0646\u0631', |
|
'\uFC8B': '\u0646\u0632', |
|
'\uFC8C': '\u0646\u0645', |
|
'\uFC8D': '\u0646\u0646', |
|
'\uFC8E': '\u0646\u0649', |
|
'\uFC8F': '\u0646\u064A', |
|
'\uFC90': '\u0649\u0670', |
|
'\uFC91': '\u064A\u0631', |
|
'\uFC92': '\u064A\u0632', |
|
'\uFC93': '\u064A\u0645', |
|
'\uFC94': '\u064A\u0646', |
|
'\uFC95': '\u064A\u0649', |
|
'\uFC96': '\u064A\u064A', |
|
'\uFC97': '\u0626\u062C', |
|
'\uFC98': '\u0626\u062D', |
|
'\uFC99': '\u0626\u062E', |
|
'\uFC9A': '\u0626\u0645', |
|
'\uFC9B': '\u0626\u0647', |
|
'\uFC9C': '\u0628\u062C', |
|
'\uFC9D': '\u0628\u062D', |
|
'\uFC9E': '\u0628\u062E', |
|
'\uFC9F': '\u0628\u0645', |
|
'\uFCA0': '\u0628\u0647', |
|
'\uFCA1': '\u062A\u062C', |
|
'\uFCA2': '\u062A\u062D', |
|
'\uFCA3': '\u062A\u062E', |
|
'\uFCA4': '\u062A\u0645', |
|
'\uFCA5': '\u062A\u0647', |
|
'\uFCA6': '\u062B\u0645', |
|
'\uFCA7': '\u062C\u062D', |
|
'\uFCA8': '\u062C\u0645', |
|
'\uFCA9': '\u062D\u062C', |
|
'\uFCAA': '\u062D\u0645', |
|
'\uFCAB': '\u062E\u062C', |
|
'\uFCAC': '\u062E\u0645', |
|
'\uFCAD': '\u0633\u062C', |
|
'\uFCAE': '\u0633\u062D', |
|
'\uFCAF': '\u0633\u062E', |
|
'\uFCB0': '\u0633\u0645', |
|
'\uFCB1': '\u0635\u062D', |
|
'\uFCB2': '\u0635\u062E', |
|
'\uFCB3': '\u0635\u0645', |
|
'\uFCB4': '\u0636\u062C', |
|
'\uFCB5': '\u0636\u062D', |
|
'\uFCB6': '\u0636\u062E', |
|
'\uFCB7': '\u0636\u0645', |
|
'\uFCB8': '\u0637\u062D', |
|
'\uFCB9': '\u0638\u0645', |
|
'\uFCBA': '\u0639\u062C', |
|
'\uFCBB': '\u0639\u0645', |
|
'\uFCBC': '\u063A\u062C', |
|
'\uFCBD': '\u063A\u0645', |
|
'\uFCBE': '\u0641\u062C', |
|
'\uFCBF': '\u0641\u062D', |
|
'\uFCC0': '\u0641\u062E', |
|
'\uFCC1': '\u0641\u0645', |
|
'\uFCC2': '\u0642\u062D', |
|
'\uFCC3': '\u0642\u0645', |
|
'\uFCC4': '\u0643\u062C', |
|
'\uFCC5': '\u0643\u062D', |
|
'\uFCC6': '\u0643\u062E', |
|
'\uFCC7': '\u0643\u0644', |
|
'\uFCC8': '\u0643\u0645', |
|
'\uFCC9': '\u0644\u062C', |
|
'\uFCCA': '\u0644\u062D', |
|
'\uFCCB': '\u0644\u062E', |
|
'\uFCCC': '\u0644\u0645', |
|
'\uFCCD': '\u0644\u0647', |
|
'\uFCCE': '\u0645\u062C', |
|
'\uFCCF': '\u0645\u062D', |
|
'\uFCD0': '\u0645\u062E', |
|
'\uFCD1': '\u0645\u0645', |
|
'\uFCD2': '\u0646\u062C', |
|
'\uFCD3': '\u0646\u062D', |
|
'\uFCD4': '\u0646\u062E', |
|
'\uFCD5': '\u0646\u0645', |
|
'\uFCD6': '\u0646\u0647', |
|
'\uFCD7': '\u0647\u062C', |
|
'\uFCD8': '\u0647\u0645', |
|
'\uFCD9': '\u0647\u0670', |
|
'\uFCDA': '\u064A\u062C', |
|
'\uFCDB': '\u064A\u062D', |
|
'\uFCDC': '\u064A\u062E', |
|
'\uFCDD': '\u064A\u0645', |
|
'\uFCDE': '\u064A\u0647', |
|
'\uFCDF': '\u0626\u0645', |
|
'\uFCE0': '\u0626\u0647', |
|
'\uFCE1': '\u0628\u0645', |
|
'\uFCE2': '\u0628\u0647', |
|
'\uFCE3': '\u062A\u0645', |
|
'\uFCE4': '\u062A\u0647', |
|
'\uFCE5': '\u062B\u0645', |
|
'\uFCE6': '\u062B\u0647', |
|
'\uFCE7': '\u0633\u0645', |
|
'\uFCE8': '\u0633\u0647', |
|
'\uFCE9': '\u0634\u0645', |
|
'\uFCEA': '\u0634\u0647', |
|
'\uFCEB': '\u0643\u0644', |
|
'\uFCEC': '\u0643\u0645', |
|
'\uFCED': '\u0644\u0645', |
|
'\uFCEE': '\u0646\u0645', |
|
'\uFCEF': '\u0646\u0647', |
|
'\uFCF0': '\u064A\u0645', |
|
'\uFCF1': '\u064A\u0647', |
|
'\uFCF2': '\u0640\u064E\u0651', |
|
'\uFCF3': '\u0640\u064F\u0651', |
|
'\uFCF4': '\u0640\u0650\u0651', |
|
'\uFCF5': '\u0637\u0649', |
|
'\uFCF6': '\u0637\u064A', |
|
'\uFCF7': '\u0639\u0649', |
|
'\uFCF8': '\u0639\u064A', |
|
'\uFCF9': '\u063A\u0649', |
|
'\uFCFA': '\u063A\u064A', |
|
'\uFCFB': '\u0633\u0649', |
|
'\uFCFC': '\u0633\u064A', |
|
'\uFCFD': '\u0634\u0649', |
|
'\uFCFE': '\u0634\u064A', |
|
'\uFCFF': '\u062D\u0649', |
|
'\uFD00': '\u062D\u064A', |
|
'\uFD01': '\u062C\u0649', |
|
'\uFD02': '\u062C\u064A', |
|
'\uFD03': '\u062E\u0649', |
|
'\uFD04': '\u062E\u064A', |
|
'\uFD05': '\u0635\u0649', |
|
'\uFD06': '\u0635\u064A', |
|
'\uFD07': '\u0636\u0649', |
|
'\uFD08': '\u0636\u064A', |
|
'\uFD09': '\u0634\u062C', |
|
'\uFD0A': '\u0634\u062D', |
|
'\uFD0B': '\u0634\u062E', |
|
'\uFD0C': '\u0634\u0645', |
|
'\uFD0D': '\u0634\u0631', |
|
'\uFD0E': '\u0633\u0631', |
|
'\uFD0F': '\u0635\u0631', |
|
'\uFD10': '\u0636\u0631', |
|
'\uFD11': '\u0637\u0649', |
|
'\uFD12': '\u0637\u064A', |
|
'\uFD13': '\u0639\u0649', |
|
'\uFD14': '\u0639\u064A', |
|
'\uFD15': '\u063A\u0649', |
|
'\uFD16': '\u063A\u064A', |
|
'\uFD17': '\u0633\u0649', |
|
'\uFD18': '\u0633\u064A', |
|
'\uFD19': '\u0634\u0649', |
|
'\uFD1A': '\u0634\u064A', |
|
'\uFD1B': '\u062D\u0649', |
|
'\uFD1C': '\u062D\u064A', |
|
'\uFD1D': '\u062C\u0649', |
|
'\uFD1E': '\u062C\u064A', |
|
'\uFD1F': '\u062E\u0649', |
|
'\uFD20': '\u062E\u064A', |
|
'\uFD21': '\u0635\u0649', |
|
'\uFD22': '\u0635\u064A', |
|
'\uFD23': '\u0636\u0649', |
|
'\uFD24': '\u0636\u064A', |
|
'\uFD25': '\u0634\u062C', |
|
'\uFD26': '\u0634\u062D', |
|
'\uFD27': '\u0634\u062E', |
|
'\uFD28': '\u0634\u0645', |
|
'\uFD29': '\u0634\u0631', |
|
'\uFD2A': '\u0633\u0631', |
|
'\uFD2B': '\u0635\u0631', |
|
'\uFD2C': '\u0636\u0631', |
|
'\uFD2D': '\u0634\u062C', |
|
'\uFD2E': '\u0634\u062D', |
|
'\uFD2F': '\u0634\u062E', |
|
'\uFD30': '\u0634\u0645', |
|
'\uFD31': '\u0633\u0647', |
|
'\uFD32': '\u0634\u0647', |
|
'\uFD33': '\u0637\u0645', |
|
'\uFD34': '\u0633\u062C', |
|
'\uFD35': '\u0633\u062D', |
|
'\uFD36': '\u0633\u062E', |
|
'\uFD37': '\u0634\u062C', |
|
'\uFD38': '\u0634\u062D', |
|
'\uFD39': '\u0634\u062E', |
|
'\uFD3A': '\u0637\u0645', |
|
'\uFD3B': '\u0638\u0645', |
|
'\uFD3C': '\u0627\u064B', |
|
'\uFD3D': '\u0627\u064B', |
|
'\uFD50': '\u062A\u062C\u0645', |
|
'\uFD51': '\u062A\u062D\u062C', |
|
'\uFD52': '\u062A\u062D\u062C', |
|
'\uFD53': '\u062A\u062D\u0645', |
|
'\uFD54': '\u062A\u062E\u0645', |
|
'\uFD55': '\u062A\u0645\u062C', |
|
'\uFD56': '\u062A\u0645\u062D', |
|
'\uFD57': '\u062A\u0645\u062E', |
|
'\uFD58': '\u062C\u0645\u062D', |
|
'\uFD59': '\u062C\u0645\u062D', |
|
'\uFD5A': '\u062D\u0645\u064A', |
|
'\uFD5B': '\u062D\u0645\u0649', |
|
'\uFD5C': '\u0633\u062D\u062C', |
|
'\uFD5D': '\u0633\u062C\u062D', |
|
'\uFD5E': '\u0633\u062C\u0649', |
|
'\uFD5F': '\u0633\u0645\u062D', |
|
'\uFD60': '\u0633\u0645\u062D', |
|
'\uFD61': '\u0633\u0645\u062C', |
|
'\uFD62': '\u0633\u0645\u0645', |
|
'\uFD63': '\u0633\u0645\u0645', |
|
'\uFD64': '\u0635\u062D\u062D', |
|
'\uFD65': '\u0635\u062D\u062D', |
|
'\uFD66': '\u0635\u0645\u0645', |
|
'\uFD67': '\u0634\u062D\u0645', |
|
'\uFD68': '\u0634\u062D\u0645', |
|
'\uFD69': '\u0634\u062C\u064A', |
|
'\uFD6A': '\u0634\u0645\u062E', |
|
'\uFD6B': '\u0634\u0645\u062E', |
|
'\uFD6C': '\u0634\u0645\u0645', |
|
'\uFD6D': '\u0634\u0645\u0645', |
|
'\uFD6E': '\u0636\u062D\u0649', |
|
'\uFD6F': '\u0636\u062E\u0645', |
|
'\uFD70': '\u0636\u062E\u0645', |
|
'\uFD71': '\u0637\u0645\u062D', |
|
'\uFD72': '\u0637\u0645\u062D', |
|
'\uFD73': '\u0637\u0645\u0645', |
|
'\uFD74': '\u0637\u0645\u064A', |
|
'\uFD75': '\u0639\u062C\u0645', |
|
'\uFD76': '\u0639\u0645\u0645', |
|
'\uFD77': '\u0639\u0645\u0645', |
|
'\uFD78': '\u0639\u0645\u0649', |
|
'\uFD79': '\u063A\u0645\u0645', |
|
'\uFD7A': '\u063A\u0645\u064A', |
|
'\uFD7B': '\u063A\u0645\u0649', |
|
'\uFD7C': '\u0641\u062E\u0645', |
|
'\uFD7D': '\u0641\u062E\u0645', |
|
'\uFD7E': '\u0642\u0645\u062D', |
|
'\uFD7F': '\u0642\u0645\u0645', |
|
'\uFD80': '\u0644\u062D\u0645', |
|
'\uFD81': '\u0644\u062D\u064A', |
|
'\uFD82': '\u0644\u062D\u0649', |
|
'\uFD83': '\u0644\u062C\u062C', |
|
'\uFD84': '\u0644\u062C\u062C', |
|
'\uFD85': '\u0644\u062E\u0645', |
|
'\uFD86': '\u0644\u062E\u0645', |
|
'\uFD87': '\u0644\u0645\u062D', |
|
'\uFD88': '\u0644\u0645\u062D', |
|
'\uFD89': '\u0645\u062D\u062C', |
|
'\uFD8A': '\u0645\u062D\u0645', |
|
'\uFD8B': '\u0645\u062D\u064A', |
|
'\uFD8C': '\u0645\u062C\u062D', |
|
'\uFD8D': '\u0645\u062C\u0645', |
|
'\uFD8E': '\u0645\u062E\u062C', |
|
'\uFD8F': '\u0645\u062E\u0645', |
|
'\uFD92': '\u0645\u062C\u062E', |
|
'\uFD93': '\u0647\u0645\u062C', |
|
'\uFD94': '\u0647\u0645\u0645', |
|
'\uFD95': '\u0646\u062D\u0645', |
|
'\uFD96': '\u0646\u062D\u0649', |
|
'\uFD97': '\u0646\u062C\u0645', |
|
'\uFD98': '\u0646\u062C\u0645', |
|
'\uFD99': '\u0646\u062C\u0649', |
|
'\uFD9A': '\u0646\u0645\u064A', |
|
'\uFD9B': '\u0646\u0645\u0649', |
|
'\uFD9C': '\u064A\u0645\u0645', |
|
'\uFD9D': '\u064A\u0645\u0645', |
|
'\uFD9E': '\u0628\u062E\u064A', |
|
'\uFD9F': '\u062A\u062C\u064A', |
|
'\uFDA0': '\u062A\u062C\u0649', |
|
'\uFDA1': '\u062A\u062E\u064A', |
|
'\uFDA2': '\u062A\u062E\u0649', |
|
'\uFDA3': '\u062A\u0645\u064A', |
|
'\uFDA4': '\u062A\u0645\u0649', |
|
'\uFDA5': '\u062C\u0645\u064A', |
|
'\uFDA6': '\u062C\u062D\u0649', |
|
'\uFDA7': '\u062C\u0645\u0649', |
|
'\uFDA8': '\u0633\u062E\u0649', |
|
'\uFDA9': '\u0635\u062D\u064A', |
|
'\uFDAA': '\u0634\u062D\u064A', |
|
'\uFDAB': '\u0636\u062D\u064A', |
|
'\uFDAC': '\u0644\u062C\u064A', |
|
'\uFDAD': '\u0644\u0645\u064A', |
|
'\uFDAE': '\u064A\u062D\u064A', |
|
'\uFDAF': '\u064A\u062C\u064A', |
|
'\uFDB0': '\u064A\u0645\u064A', |
|
'\uFDB1': '\u0645\u0645\u064A', |
|
'\uFDB2': '\u0642\u0645\u064A', |
|
'\uFDB3': '\u0646\u062D\u064A', |
|
'\uFDB4': '\u0642\u0645\u062D', |
|
'\uFDB5': '\u0644\u062D\u0645', |
|
'\uFDB6': '\u0639\u0645\u064A', |
|
'\uFDB7': '\u0643\u0645\u064A', |
|
'\uFDB8': '\u0646\u062C\u062D', |
|
'\uFDB9': '\u0645\u062E\u064A', |
|
'\uFDBA': '\u0644\u062C\u0645', |
|
'\uFDBB': '\u0643\u0645\u0645', |
|
'\uFDBC': '\u0644\u062C\u0645', |
|
'\uFDBD': '\u0646\u062C\u062D', |
|
'\uFDBE': '\u062C\u062D\u064A', |
|
'\uFDBF': '\u062D\u062C\u064A', |
|
'\uFDC0': '\u0645\u062C\u064A', |
|
'\uFDC1': '\u0641\u0645\u064A', |
|
'\uFDC2': '\u0628\u062D\u064A', |
|
'\uFDC3': '\u0643\u0645\u0645', |
|
'\uFDC4': '\u0639\u062C\u0645', |
|
'\uFDC5': '\u0635\u0645\u0645', |
|
'\uFDC6': '\u0633\u062E\u064A', |
|
'\uFDC7': '\u0646\u062C\u064A', |
|
'\uFE49': '\u203E', |
|
'\uFE4A': '\u203E', |
|
'\uFE4B': '\u203E', |
|
'\uFE4C': '\u203E', |
|
'\uFE4D': '\u005F', |
|
'\uFE4E': '\u005F', |
|
'\uFE4F': '\u005F', |
|
'\uFE80': '\u0621', |
|
'\uFE81': '\u0622', |
|
'\uFE82': '\u0622', |
|
'\uFE83': '\u0623', |
|
'\uFE84': '\u0623', |
|
'\uFE85': '\u0624', |
|
'\uFE86': '\u0624', |
|
'\uFE87': '\u0625', |
|
'\uFE88': '\u0625', |
|
'\uFE89': '\u0626', |
|
'\uFE8A': '\u0626', |
|
'\uFE8B': '\u0626', |
|
'\uFE8C': '\u0626', |
|
'\uFE8D': '\u0627', |
|
'\uFE8E': '\u0627', |
|
'\uFE8F': '\u0628', |
|
'\uFE90': '\u0628', |
|
'\uFE91': '\u0628', |
|
'\uFE92': '\u0628', |
|
'\uFE93': '\u0629', |
|
'\uFE94': '\u0629', |
|
'\uFE95': '\u062A', |
|
'\uFE96': '\u062A', |
|
'\uFE97': '\u062A', |
|
'\uFE98': '\u062A', |
|
'\uFE99': '\u062B', |
|
'\uFE9A': '\u062B', |
|
'\uFE9B': '\u062B', |
|
'\uFE9C': '\u062B', |
|
'\uFE9D': '\u062C', |
|
'\uFE9E': '\u062C', |
|
'\uFE9F': '\u062C', |
|
'\uFEA0': '\u062C', |
|
'\uFEA1': '\u062D', |
|
'\uFEA2': '\u062D', |
|
'\uFEA3': '\u062D', |
|
'\uFEA4': '\u062D', |
|
'\uFEA5': '\u062E', |
|
'\uFEA6': '\u062E', |
|
'\uFEA7': '\u062E', |
|
'\uFEA8': '\u062E', |
|
'\uFEA9': '\u062F', |
|
'\uFEAA': '\u062F', |
|
'\uFEAB': '\u0630', |
|
'\uFEAC': '\u0630', |
|
'\uFEAD': '\u0631', |
|
'\uFEAE': '\u0631', |
|
'\uFEAF': '\u0632', |
|
'\uFEB0': '\u0632', |
|
'\uFEB1': '\u0633', |
|
'\uFEB2': '\u0633', |
|
'\uFEB3': '\u0633', |
|
'\uFEB4': '\u0633', |
|
'\uFEB5': '\u0634', |
|
'\uFEB6': '\u0634', |
|
'\uFEB7': '\u0634', |
|
'\uFEB8': '\u0634', |
|
'\uFEB9': '\u0635', |
|
'\uFEBA': '\u0635', |
|
'\uFEBB': '\u0635', |
|
'\uFEBC': '\u0635', |
|
'\uFEBD': '\u0636', |
|
'\uFEBE': '\u0636', |
|
'\uFEBF': '\u0636', |
|
'\uFEC0': '\u0636', |
|
'\uFEC1': '\u0637', |
|
'\uFEC2': '\u0637', |
|
'\uFEC3': '\u0637', |
|
'\uFEC4': '\u0637', |
|
'\uFEC5': '\u0638', |
|
'\uFEC6': '\u0638', |
|
'\uFEC7': '\u0638', |
|
'\uFEC8': '\u0638', |
|
'\uFEC9': '\u0639', |
|
'\uFECA': '\u0639', |
|
'\uFECB': '\u0639', |
|
'\uFECC': '\u0639', |
|
'\uFECD': '\u063A', |
|
'\uFECE': '\u063A', |
|
'\uFECF': '\u063A', |
|
'\uFED0': '\u063A', |
|
'\uFED1': '\u0641', |
|
'\uFED2': '\u0641', |
|
'\uFED3': '\u0641', |
|
'\uFED4': '\u0641', |
|
'\uFED5': '\u0642', |
|
'\uFED6': '\u0642', |
|
'\uFED7': '\u0642', |
|
'\uFED8': '\u0642', |
|
'\uFED9': '\u0643', |
|
'\uFEDA': '\u0643', |
|
'\uFEDB': '\u0643', |
|
'\uFEDC': '\u0643', |
|
'\uFEDD': '\u0644', |
|
'\uFEDE': '\u0644', |
|
'\uFEDF': '\u0644', |
|
'\uFEE0': '\u0644', |
|
'\uFEE1': '\u0645', |
|
'\uFEE2': '\u0645', |
|
'\uFEE3': '\u0645', |
|
'\uFEE4': '\u0645', |
|
'\uFEE5': '\u0646', |
|
'\uFEE6': '\u0646', |
|
'\uFEE7': '\u0646', |
|
'\uFEE8': '\u0646', |
|
'\uFEE9': '\u0647', |
|
'\uFEEA': '\u0647', |
|
'\uFEEB': '\u0647', |
|
'\uFEEC': '\u0647', |
|
'\uFEED': '\u0648', |
|
'\uFEEE': '\u0648', |
|
'\uFEEF': '\u0649', |
|
'\uFEF0': '\u0649', |
|
'\uFEF1': '\u064A', |
|
'\uFEF2': '\u064A', |
|
'\uFEF3': '\u064A', |
|
'\uFEF4': '\u064A', |
|
'\uFEF5': '\u0644\u0622', |
|
'\uFEF6': '\u0644\u0622', |
|
'\uFEF7': '\u0644\u0623', |
|
'\uFEF8': '\u0644\u0623', |
|
'\uFEF9': '\u0644\u0625', |
|
'\uFEFA': '\u0644\u0625', |
|
'\uFEFB': '\u0644\u0627', |
|
'\uFEFC': '\u0644\u0627' |
|
}; |
|
|
|
function reverseIfRtl(chars) { |
|
var charsLength = chars.length; |
|
//reverse an arabic ligature |
|
if (charsLength <= 1 || !isRTLRangeFor(chars.charCodeAt(0))) { |
|
return chars; |
|
} |
|
var s = ''; |
|
for (var ii = charsLength - 1; ii >= 0; ii--) { |
|
s += chars[ii]; |
|
} |
|
return s; |
|
} |
|
|
|
function adjustWidths(properties) { |
|
if (!properties.fontMatrix) { |
|
return; |
|
} |
|
if (properties.fontMatrix[0] === FONT_IDENTITY_MATRIX[0]) { |
|
return; |
|
} |
|
// adjusting width to fontMatrix scale |
|
var scale = 0.001 / properties.fontMatrix[0]; |
|
var glyphsWidths = properties.widths; |
|
for (var glyph in glyphsWidths) { |
|
glyphsWidths[glyph] *= scale; |
|
} |
|
properties.defaultWidth *= scale; |
|
} |
|
|
|
function getFontType(type, subtype) { |
|
switch (type) { |
|
case 'Type1': |
|
return subtype === 'Type1C' ? FontType.TYPE1C : FontType.TYPE1; |
|
case 'CIDFontType0': |
|
return subtype === 'CIDFontType0C' ? FontType.CIDFONTTYPE0C : |
|
FontType.CIDFONTTYPE0; |
|
case 'OpenType': |
|
return FontType.OPENTYPE; |
|
case 'TrueType': |
|
return FontType.TRUETYPE; |
|
case 'CIDFontType2': |
|
return FontType.CIDFONTTYPE2; |
|
case 'MMType1': |
|
return FontType.MMTYPE1; |
|
case 'Type0': |
|
return FontType.TYPE0; |
|
default: |
|
return FontType.UNKNOWN; |
|
} |
|
} |
|
|
|
var Glyph = (function GlyphClosure() { |
|
function Glyph(fontChar, unicode, accent, width, vmetric, operatorListId, |
|
isSpace) { |
|
this.fontChar = fontChar; |
|
this.unicode = unicode; |
|
this.accent = accent; |
|
this.width = width; |
|
this.vmetric = vmetric; |
|
this.operatorListId = operatorListId; |
|
this.isSpace = isSpace; |
|
} |
|
|
|
Glyph.prototype.matchesForCache = function(fontChar, unicode, accent, width, |
|
vmetric, operatorListId, isSpace) { |
|
return this.fontChar === fontChar && |
|
this.unicode === unicode && |
|
this.accent === accent && |
|
this.width === width && |
|
this.vmetric === vmetric && |
|
this.operatorListId === operatorListId && |
|
this.isSpace === isSpace; |
|
}; |
|
|
|
return Glyph; |
|
})(); |
|
|
|
var ToUnicodeMap = (function ToUnicodeMapClosure() { |
|
function ToUnicodeMap(cmap) { |
|
// The elements of this._map can be integers or strings, depending on how |
|
// |cmap| was created. |
|
this._map = cmap; |
|
} |
|
|
|
ToUnicodeMap.prototype = { |
|
get length() { |
|
return this._map.length; |
|
}, |
|
|
|
forEach: function(callback) { |
|
for (var charCode in this._map) { |
|
callback(charCode, this._map[charCode].charCodeAt(0)); |
|
} |
|
}, |
|
|
|
has: function(i) { |
|
return this._map[i] !== undefined; |
|
}, |
|
|
|
get: function(i) { |
|
return this._map[i]; |
|
}, |
|
|
|
charCodeOf: function(v) { |
|
return this._map.indexOf(v); |
|
} |
|
}; |
|
|
|
return ToUnicodeMap; |
|
})(); |
|
|
|
var IdentityToUnicodeMap = (function IdentityToUnicodeMapClosure() { |
|
function IdentityToUnicodeMap(firstChar, lastChar) { |
|
this.firstChar = firstChar; |
|
this.lastChar = lastChar; |
|
} |
|
|
|
IdentityToUnicodeMap.prototype = { |
|
get length() { |
|
return (this.lastChar + 1) - this.firstChar; |
|
}, |
|
|
|
forEach: function (callback) { |
|
for (var i = this.firstChar, ii = this.lastChar; i <= ii; i++) { |
|
callback(i, i); |
|
} |
|
}, |
|
|
|
has: function (i) { |
|
return this.firstChar <= i && i <= this.lastChar; |
|
}, |
|
|
|
get: function (i) { |
|
if (this.firstChar <= i && i <= this.lastChar) { |
|
return String.fromCharCode(i); |
|
} |
|
return undefined; |
|
}, |
|
|
|
charCodeOf: function (v) { |
|
return (isInt(v) && v >= this.firstChar && v <= this.lastChar) ? v : -1; |
|
} |
|
}; |
|
|
|
return IdentityToUnicodeMap; |
|
})(); |
|
|
|
var OpenTypeFileBuilder = (function OpenTypeFileBuilderClosure() { |
|
function writeInt16(dest, offset, num) { |
|
dest[offset] = (num >> 8) & 0xFF; |
|
dest[offset + 1] = num & 0xFF; |
|
} |
|
|
|
function writeInt32(dest, offset, num) { |
|
dest[offset] = (num >> 24) & 0xFF; |
|
dest[offset + 1] = (num >> 16) & 0xFF; |
|
dest[offset + 2] = (num >> 8) & 0xFF; |
|
dest[offset + 3] = num & 0xFF; |
|
} |
|
|
|
function writeData(dest, offset, data) { |
|
var i, ii; |
|
if (data instanceof Uint8Array) { |
|
dest.set(data, offset); |
|
} else if (typeof data === 'string') { |
|
for (i = 0, ii = data.length; i < ii; i++) { |
|
dest[offset++] = data.charCodeAt(i) & 0xFF; |
|
} |
|
} else { |
|
// treating everything else as array |
|
for (i = 0, ii = data.length; i < ii; i++) { |
|
dest[offset++] = data[i] & 0xFF; |
|
} |
|
} |
|
} |
|
|
|
function OpenTypeFileBuilder(sfnt) { |
|
this.sfnt = sfnt; |
|
this.tables = Object.create(null); |
|
} |
|
|
|
OpenTypeFileBuilder.getSearchParams = |
|
function OpenTypeFileBuilder_getSearchParams(entriesCount, entrySize) { |
|
var maxPower2 = 1, log2 = 0; |
|
while ((maxPower2 ^ entriesCount) > maxPower2) { |
|
maxPower2 <<= 1; |
|
log2++; |
|
} |
|
var searchRange = maxPower2 * entrySize; |
|
return { |
|
range: searchRange, |
|
entry: log2, |
|
rangeShift: entrySize * entriesCount - searchRange |
|
}; |
|
}; |
|
|
|
var OTF_HEADER_SIZE = 12; |
|
var OTF_TABLE_ENTRY_SIZE = 16; |
|
|
|
OpenTypeFileBuilder.prototype = { |
|
toArray: function OpenTypeFileBuilder_toArray() { |
|
var sfnt = this.sfnt; |
|
|
|
// Tables needs to be written by ascendant alphabetic order |
|
var tables = this.tables; |
|
var tablesNames = Object.keys(tables); |
|
tablesNames.sort(); |
|
var numTables = tablesNames.length; |
|
|
|
var i, j, jj, table, tableName; |
|
// layout the tables data |
|
var offset = OTF_HEADER_SIZE + numTables * OTF_TABLE_ENTRY_SIZE; |
|
var tableOffsets = [offset]; |
|
for (i = 0; i < numTables; i++) { |
|
table = tables[tablesNames[i]]; |
|
var paddedLength = ((table.length + 3) & ~3) >>> 0; |
|
offset += paddedLength; |
|
tableOffsets.push(offset); |
|
} |
|
|
|
var file = new Uint8Array(offset); |
|
// write the table data first (mostly for checksum) |
|
for (i = 0; i < numTables; i++) { |
|
table = tables[tablesNames[i]]; |
|
writeData(file, tableOffsets[i], table); |
|
} |
|
|
|
// sfnt version (4 bytes) |
|
if (sfnt === 'true') { |
|
// Windows hates the Mac TrueType sfnt version number |
|
sfnt = string32(0x00010000); |
|
} |
|
file[0] = sfnt.charCodeAt(0) & 0xFF; |
|
file[1] = sfnt.charCodeAt(1) & 0xFF; |
|
file[2] = sfnt.charCodeAt(2) & 0xFF; |
|
file[3] = sfnt.charCodeAt(3) & 0xFF; |
|
|
|
// numTables (2 bytes) |
|
writeInt16(file, 4, numTables); |
|
|
|
var searchParams = OpenTypeFileBuilder.getSearchParams(numTables, 16); |
|
|
|
// searchRange (2 bytes) |
|
writeInt16(file, 6, searchParams.range); |
|
// entrySelector (2 bytes) |
|
writeInt16(file, 8, searchParams.entry); |
|
// rangeShift (2 bytes) |
|
writeInt16(file, 10, searchParams.rangeShift); |
|
|
|
offset = OTF_HEADER_SIZE; |
|
// writing table entries |
|
for (i = 0; i < numTables; i++) { |
|
tableName = tablesNames[i]; |
|
file[offset] = tableName.charCodeAt(0) & 0xFF; |
|
file[offset + 1] = tableName.charCodeAt(1) & 0xFF; |
|
file[offset + 2] = tableName.charCodeAt(2) & 0xFF; |
|
file[offset + 3] = tableName.charCodeAt(3) & 0xFF; |
|
|
|
// checksum |
|
var checksum = 0; |
|
for (j = tableOffsets[i], jj = tableOffsets[i + 1]; j < jj; j += 4) { |
|
var quad = (file[j] << 24) + (file[j + 1] << 16) + |
|
(file[j + 2] << 8) + file[j + 3]; |
|
checksum = (checksum + quad) | 0; |
|
} |
|
writeInt32(file, offset + 4, checksum); |
|
|
|
// offset |
|
writeInt32(file, offset + 8, tableOffsets[i]); |
|
// length |
|
writeInt32(file, offset + 12, tables[tableName].length); |
|
|
|
offset += OTF_TABLE_ENTRY_SIZE; |
|
} |
|
return file; |
|
}, |
|
|
|
addTable: function OpenTypeFileBuilder_addTable(tag, data) { |
|
if (tag in this.tables) { |
|
throw new Error('Table ' + tag + ' already exists'); |
|
} |
|
this.tables[tag] = data; |
|
} |
|
}; |
|
|
|
return OpenTypeFileBuilder; |
|
})(); |
|
|
|
// Problematic Unicode characters in the fonts that needs to be moved to avoid |
|
// issues when they are painted on the canvas, e.g. complex-script shaping or |
|
// control/whitespace characters. The ranges are listed in pairs: the first item |
|
// is a code of the first problematic code, the second one is the next |
|
// non-problematic code. The ranges must be in sorted order. |
|
var ProblematicCharRanges = new Int32Array([ |
|
// Control characters. |
|
0x0000, 0x0020, |
|
0x007F, 0x00A1, |
|
0x00AD, 0x00AE, |
|
// Chars that is used in complex-script shaping. |
|
0x0600, 0x0780, |
|
0x08A0, 0x10A0, |
|
0x1780, 0x1800, |
|
// General punctuation chars. |
|
0x2000, 0x2010, |
|
0x2011, 0x2012, |
|
0x2028, 0x2030, |
|
0x205F, 0x2070, |
|
0x25CC, 0x25CD, |
|
// Chars that is used in complex-script shaping. |
|
0xAA60, 0xAA80, |
|
// Specials Unicode block. |
|
0xFFF0, 0x10000 |
|
]); |
|
|
|
/** |
|
* 'Font' is the class the outside world should use, it encapsulate all the font |
|
* decoding logics whatever type it is (assuming the font type is supported). |
|
* |
|
* For example to read a Type1 font and to attach it to the document: |
|
* var type1Font = new Font("MyFontName", binaryFile, propertiesObject); |
|
* type1Font.bind(); |
|
*/ |
|
var Font = (function FontClosure() { |
|
function Font(name, file, properties) { |
|
var charCode, glyphName, fontChar; |
|
|
|
this.name = name; |
|
this.loadedName = properties.loadedName; |
|
this.isType3Font = properties.isType3Font; |
|
this.sizes = []; |
|
|
|
this.glyphCache = {}; |
|
|
|
var names = name.split('+'); |
|
names = names.length > 1 ? names[1] : names[0]; |
|
names = names.split(/[-,_]/g)[0]; |
|
this.isSerifFont = !!(properties.flags & FontFlags.Serif); |
|
this.isSymbolicFont = !!(properties.flags & FontFlags.Symbolic); |
|
this.isMonospace = !!(properties.flags & FontFlags.FixedPitch); |
|
|
|
var type = properties.type; |
|
var subtype = properties.subtype; |
|
this.type = type; |
|
|
|
this.fallbackName = (this.isMonospace ? 'monospace' : |
|
(this.isSerifFont ? 'serif' : 'sans-serif')); |
|
|
|
this.differences = properties.differences; |
|
this.widths = properties.widths; |
|
this.defaultWidth = properties.defaultWidth; |
|
this.composite = properties.composite; |
|
this.wideChars = properties.wideChars; |
|
this.cMap = properties.cMap; |
|
this.ascent = properties.ascent / PDF_GLYPH_SPACE_UNITS; |
|
this.descent = properties.descent / PDF_GLYPH_SPACE_UNITS; |
|
this.fontMatrix = properties.fontMatrix; |
|
this.bbox = properties.bbox; |
|
|
|
this.toUnicode = properties.toUnicode = this.buildToUnicode(properties); |
|
|
|
this.toFontChar = []; |
|
|
|
if (properties.type === 'Type3') { |
|
for (charCode = 0; charCode < 256; charCode++) { |
|
this.toFontChar[charCode] = (this.differences[charCode] || |
|
properties.defaultEncoding[charCode]); |
|
} |
|
this.fontType = FontType.TYPE3; |
|
return; |
|
} |
|
|
|
this.cidEncoding = properties.cidEncoding; |
|
this.vertical = properties.vertical; |
|
if (this.vertical) { |
|
this.vmetrics = properties.vmetrics; |
|
this.defaultVMetrics = properties.defaultVMetrics; |
|
} |
|
|
|
if (!file || file.isEmpty) { |
|
if (file) { |
|
// Some bad PDF generators will include empty font files, |
|
// attempting to recover by assuming that no file exists. |
|
warn('Font file is empty in "' + name + '" (' + this.loadedName + ')'); |
|
} |
|
|
|
this.missingFile = true; |
|
// The file data is not specified. Trying to fix the font name |
|
// to be used with the canvas.font. |
|
var fontName = name.replace(/[,_]/g, '-'); |
|
var isStandardFont = !!stdFontMap[fontName] || |
|
!!(nonStdFontMap[fontName] && stdFontMap[nonStdFontMap[fontName]]); |
|
fontName = stdFontMap[fontName] || nonStdFontMap[fontName] || fontName; |
|
|
|
this.bold = (fontName.search(/bold/gi) !== -1); |
|
this.italic = ((fontName.search(/oblique/gi) !== -1) || |
|
(fontName.search(/italic/gi) !== -1)); |
|
|
|
// Use 'name' instead of 'fontName' here because the original |
|
// name ArialBlack for example will be replaced by Helvetica. |
|
this.black = (name.search(/Black/g) !== -1); |
|
|
|
// if at least one width is present, remeasure all chars when exists |
|
this.remeasure = Object.keys(this.widths).length > 0; |
|
if (isStandardFont && type === 'CIDFontType2' && |
|
properties.cidEncoding.indexOf('Identity-') === 0) { |
|
// Standard fonts might be embedded as CID font without glyph mapping. |
|
// Building one based on GlyphMapForStandardFonts. |
|
var map = []; |
|
for (charCode in GlyphMapForStandardFonts) { |
|
map[+charCode] = GlyphMapForStandardFonts[charCode]; |
|
} |
|
if (/ArialBlack/i.test(name)) { |
|
for (charCode in SupplementalGlyphMapForArialBlack) { |
|
map[+charCode] = SupplementalGlyphMapForArialBlack[charCode]; |
|
} |
|
} |
|
var isIdentityUnicode = this.toUnicode instanceof IdentityToUnicodeMap; |
|
if (!isIdentityUnicode) { |
|
this.toUnicode.forEach(function(charCode, unicodeCharCode) { |
|
map[+charCode] = unicodeCharCode; |
|
}); |
|
} |
|
this.toFontChar = map; |
|
this.toUnicode = new ToUnicodeMap(map); |
|
} else if (/Symbol/i.test(fontName)) { |
|
var symbols = Encodings.SymbolSetEncoding; |
|
for (charCode in symbols) { |
|
fontChar = GlyphsUnicode[symbols[charCode]]; |
|
if (!fontChar) { |
|
continue; |
|
} |
|
this.toFontChar[charCode] = fontChar; |
|
} |
|
for (charCode in properties.differences) { |
|
fontChar = GlyphsUnicode[properties.differences[charCode]]; |
|
if (!fontChar) { |
|
continue; |
|
} |
|
this.toFontChar[charCode] = fontChar; |
|
} |
|
} else if (/Dingbats/i.test(fontName)) { |
|
if (/Wingdings/i.test(name)) { |
|
warn('Wingdings font without embedded font file, ' + |
|
'falling back to the ZapfDingbats encoding.'); |
|
} |
|
var dingbats = Encodings.ZapfDingbatsEncoding; |
|
for (charCode in dingbats) { |
|
fontChar = DingbatsGlyphsUnicode[dingbats[charCode]]; |
|
if (!fontChar) { |
|
continue; |
|
} |
|
this.toFontChar[charCode] = fontChar; |
|
} |
|
for (charCode in properties.differences) { |
|
fontChar = DingbatsGlyphsUnicode[properties.differences[charCode]]; |
|
if (!fontChar) { |
|
continue; |
|
} |
|
this.toFontChar[charCode] = fontChar; |
|
} |
|
} else if (isStandardFont) { |
|
this.toFontChar = []; |
|
for (charCode in properties.defaultEncoding) { |
|
glyphName = (properties.differences[charCode] || |
|
properties.defaultEncoding[charCode]); |
|
this.toFontChar[charCode] = GlyphsUnicode[glyphName]; |
|
} |
|
} else { |
|
var unicodeCharCode, notCidFont = (type.indexOf('CIDFontType') === -1); |
|
this.toUnicode.forEach(function(charCode, unicodeCharCode) { |
|
if (notCidFont) { |
|
glyphName = (properties.differences[charCode] || |
|
properties.defaultEncoding[charCode]); |
|
unicodeCharCode = (GlyphsUnicode[glyphName] || unicodeCharCode); |
|
} |
|
this.toFontChar[charCode] = unicodeCharCode; |
|
}.bind(this)); |
|
} |
|
this.loadedName = fontName.split('-')[0]; |
|
this.loading = false; |
|
this.fontType = getFontType(type, subtype); |
|
return; |
|
} |
|
|
|
// Some fonts might use wrong font types for Type1C or CIDFontType0C |
|
if (subtype === 'Type1C' && (type !== 'Type1' && type !== 'MMType1')) { |
|
// Some TrueType fonts by mistake claim Type1C |
|
if (isTrueTypeFile(file)) { |
|
subtype = 'TrueType'; |
|
} else { |
|
type = 'Type1'; |
|
} |
|
} |
|
if (subtype === 'CIDFontType0C' && type !== 'CIDFontType0') { |
|
type = 'CIDFontType0'; |
|
} |
|
if (subtype === 'OpenType') { |
|
type = 'OpenType'; |
|
} |
|
// Some CIDFontType0C fonts by mistake claim CIDFontType0. |
|
if (type === 'CIDFontType0') { |
|
if (isType1File(file)) { |
|
subtype = 'CIDFontType0'; |
|
} else if (isOpenTypeFile(file)) { |
|
// Sometimes the type/subtype can be a complete lie (see issue6782.pdf). |
|
type = subtype = 'OpenType'; |
|
} else { |
|
subtype = 'CIDFontType0C'; |
|
} |
|
} |
|
|
|
var data; |
|
switch (type) { |
|
case 'MMType1': |
|
info('MMType1 font (' + name + '), falling back to Type1.'); |
|
/* falls through */ |
|
case 'Type1': |
|
case 'CIDFontType0': |
|
this.mimetype = 'font/opentype'; |
|
|
|
var cff = (subtype === 'Type1C' || subtype === 'CIDFontType0C') ? |
|
new CFFFont(file, properties) : new Type1Font(name, file, properties); |
|
|
|
adjustWidths(properties); |
|
|
|
// Wrap the CFF data inside an OTF font file |
|
data = this.convert(name, cff, properties); |
|
break; |
|
|
|
case 'OpenType': |
|
case 'TrueType': |
|
case 'CIDFontType2': |
|
this.mimetype = 'font/opentype'; |
|
|
|
// Repair the TrueType file. It is can be damaged in the point of |
|
// view of the sanitizer |
|
data = this.checkAndRepair(name, file, properties); |
|
if (this.isOpenType) { |
|
adjustWidths(properties); |
|
|
|
type = 'OpenType'; |
|
} |
|
break; |
|
|
|
default: |
|
error('Font ' + type + ' is not supported'); |
|
break; |
|
} |
|
|
|
this.data = data; |
|
this.fontType = getFontType(type, subtype); |
|
|
|
// Transfer some properties again that could change during font conversion |
|
this.fontMatrix = properties.fontMatrix; |
|
this.widths = properties.widths; |
|
this.defaultWidth = properties.defaultWidth; |
|
this.encoding = properties.baseEncoding; |
|
this.seacMap = properties.seacMap; |
|
|
|
this.loading = true; |
|
} |
|
|
|
Font.getFontID = (function () { |
|
var ID = 1; |
|
return function Font_getFontID() { |
|
return String(ID++); |
|
}; |
|
})(); |
|
|
|
function int16(b0, b1) { |
|
return (b0 << 8) + b1; |
|
} |
|
|
|
function int32(b0, b1, b2, b3) { |
|
return (b0 << 24) + (b1 << 16) + (b2 << 8) + b3; |
|
} |
|
|
|
function string16(value) { |
|
return String.fromCharCode((value >> 8) & 0xff, value & 0xff); |
|
} |
|
|
|
function safeString16(value) { |
|
// clamp value to the 16-bit int range |
|
value = (value > 0x7FFF ? 0x7FFF : (value < -0x8000 ? -0x8000 : value)); |
|
return String.fromCharCode((value >> 8) & 0xff, value & 0xff); |
|
} |
|
|
|
function isTrueTypeFile(file) { |
|
var header = file.peekBytes(4); |
|
return readUint32(header, 0) === 0x00010000; |
|
} |
|
|
|
function isOpenTypeFile(file) { |
|
var header = file.peekBytes(4); |
|
return bytesToString(header) === 'OTTO'; |
|
} |
|
|
|
function isType1File(file) { |
|
var header = file.peekBytes(2); |
|
// All Type1 font programs must begin with the comment '%!' (0x25 + 0x21). |
|
if (header[0] === 0x25 && header[1] === 0x21) { |
|
return true; |
|
} |
|
// ... obviously some fonts violate that part of the specification, |
|
// please refer to the comment in |Type1Font| below. |
|
if (header[0] === 0x80 && header[1] === 0x01) { // pfb file header. |
|
return true; |
|
} |
|
return false; |
|
} |
|
|
|
/** |
|
* Helper function for |adjustMapping|. |
|
* @return {boolean} |
|
*/ |
|
function isProblematicUnicodeLocation(code) { |
|
// Using binary search to find a range start. |
|
var i = 0, j = ProblematicCharRanges.length - 1; |
|
while (i < j) { |
|
var c = (i + j + 1) >> 1; |
|
if (code < ProblematicCharRanges[c]) { |
|
j = c - 1; |
|
} else { |
|
i = c; |
|
} |
|
} |
|
// Even index means code in problematic range. |
|
return !(i & 1); |
|
} |
|
|
|
/** |
|
* Rebuilds the char code to glyph ID map by trying to replace the char codes |
|
* with their unicode value. It also moves char codes that are in known |
|
* problematic locations. |
|
* @return {Object} Two properties: |
|
* 'toFontChar' - maps original char codes(the value that will be read |
|
* from commands such as show text) to the char codes that will be used in the |
|
* font that we build |
|
* 'charCodeToGlyphId' - maps the new font char codes to glyph ids |
|
*/ |
|
function adjustMapping(charCodeToGlyphId, properties) { |
|
var toUnicode = properties.toUnicode; |
|
var isSymbolic = !!(properties.flags & FontFlags.Symbolic); |
|
var isIdentityUnicode = |
|
properties.toUnicode instanceof IdentityToUnicodeMap; |
|
var newMap = Object.create(null); |
|
var toFontChar = []; |
|
var usedFontCharCodes = []; |
|
var nextAvailableFontCharCode = PRIVATE_USE_OFFSET_START; |
|
for (var originalCharCode in charCodeToGlyphId) { |
|
originalCharCode |= 0; |
|
var glyphId = charCodeToGlyphId[originalCharCode]; |
|
var fontCharCode = originalCharCode; |
|
// First try to map the value to a unicode position if a non identity map |
|
// was created. |
|
if (!isIdentityUnicode && toUnicode.has(originalCharCode)) { |
|
var unicode = toUnicode.get(fontCharCode); |
|
// TODO: Try to map ligatures to the correct spot. |
|
if (unicode.length === 1) { |
|
fontCharCode = unicode.charCodeAt(0); |
|
} |
|
} |
|
// Try to move control characters, special characters and already mapped |
|
// characters to the private use area since they will not be drawn by |
|
// canvas if left in their current position. Also, move characters if the |
|
// font was symbolic and there is only an identity unicode map since the |
|
// characters probably aren't in the correct position (fixes an issue |
|
// with firefox and thuluthfont). |
|
if ((usedFontCharCodes[fontCharCode] !== undefined || |
|
isProblematicUnicodeLocation(fontCharCode) || |
|
(isSymbolic && isIdentityUnicode)) && |
|
nextAvailableFontCharCode <= PRIVATE_USE_OFFSET_END) { // Room left. |
|
// Loop to try and find a free spot in the private use area. |
|
do { |
|
fontCharCode = nextAvailableFontCharCode++; |
|
|
|
if (SKIP_PRIVATE_USE_RANGE_F000_TO_F01F && fontCharCode === 0xF000) { |
|
fontCharCode = 0xF020; |
|
nextAvailableFontCharCode = fontCharCode + 1; |
|
} |
|
|
|
} while (usedFontCharCodes[fontCharCode] !== undefined && |
|
nextAvailableFontCharCode <= PRIVATE_USE_OFFSET_END); |
|
} |
|
|
|
newMap[fontCharCode] = glyphId; |
|
toFontChar[originalCharCode] = fontCharCode; |
|
usedFontCharCodes[fontCharCode] = true; |
|
} |
|
return { |
|
toFontChar: toFontChar, |
|
charCodeToGlyphId: newMap, |
|
nextAvailableFontCharCode: nextAvailableFontCharCode |
|
}; |
|
} |
|
|
|
function getRanges(glyphs, numGlyphs) { |
|
// Array.sort() sorts by characters, not numerically, so convert to an |
|
// array of characters. |
|
var codes = []; |
|
for (var charCode in glyphs) { |
|
// Remove an invalid glyph ID mappings to make OTS happy. |
|
if (glyphs[charCode] >= numGlyphs) { |
|
continue; |
|
} |
|
codes.push({ fontCharCode: charCode | 0, glyphId: glyphs[charCode] }); |
|
} |
|
codes.sort(function fontGetRangesSort(a, b) { |
|
return a.fontCharCode - b.fontCharCode; |
|
}); |
|
|
|
// Split the sorted codes into ranges. |
|
var ranges = []; |
|
var length = codes.length; |
|
for (var n = 0; n < length; ) { |
|
var start = codes[n].fontCharCode; |
|
var codeIndices = [codes[n].glyphId]; |
|
++n; |
|
var end = start; |
|
while (n < length && end + 1 === codes[n].fontCharCode) { |
|
codeIndices.push(codes[n].glyphId); |
|
++end; |
|
++n; |
|
if (end === 0xFFFF) { |
|
break; |
|
} |
|
} |
|
ranges.push([start, end, codeIndices]); |
|
} |
|
|
|
return ranges; |
|
} |
|
|
|
function createCmapTable(glyphs, numGlyphs) { |
|
var ranges = getRanges(glyphs, numGlyphs); |
|
var numTables = ranges[ranges.length - 1][1] > 0xFFFF ? 2 : 1; |
|
var cmap = '\x00\x00' + // version |
|
string16(numTables) + // numTables |
|
'\x00\x03' + // platformID |
|
'\x00\x01' + // encodingID |
|
string32(4 + numTables * 8); // start of the table record |
|
|
|
var i, ii, j, jj; |
|
for (i = ranges.length - 1; i >= 0; --i) { |
|
if (ranges[i][0] <= 0xFFFF) { break; } |
|
} |
|
var bmpLength = i + 1; |
|
|
|
if (ranges[i][0] < 0xFFFF && ranges[i][1] === 0xFFFF) { |
|
ranges[i][1] = 0xFFFE; |
|
} |
|
var trailingRangesCount = ranges[i][1] < 0xFFFF ? 1 : 0; |
|
var segCount = bmpLength + trailingRangesCount; |
|
var searchParams = OpenTypeFileBuilder.getSearchParams(segCount, 2); |
|
|
|
// Fill up the 4 parallel arrays describing the segments. |
|
var startCount = ''; |
|
var endCount = ''; |
|
var idDeltas = ''; |
|
var idRangeOffsets = ''; |
|
var glyphsIds = ''; |
|
var bias = 0; |
|
|
|
var range, start, end, codes; |
|
for (i = 0, ii = bmpLength; i < ii; i++) { |
|
range = ranges[i]; |
|
start = range[0]; |
|
end = range[1]; |
|
startCount += string16(start); |
|
endCount += string16(end); |
|
codes = range[2]; |
|
var contiguous = true; |
|
for (j = 1, jj = codes.length; j < jj; ++j) { |
|
if (codes[j] !== codes[j - 1] + 1) { |
|
contiguous = false; |
|
break; |
|
} |
|
} |
|
if (!contiguous) { |
|
var offset = (segCount - i) * 2 + bias * 2; |
|
bias += (end - start + 1); |
|
|
|
idDeltas += string16(0); |
|
idRangeOffsets += string16(offset); |
|
|
|
for (j = 0, jj = codes.length; j < jj; ++j) { |
|
glyphsIds += string16(codes[j]); |
|
} |
|
} else { |
|
var startCode = codes[0]; |
|
|
|
idDeltas += string16((startCode - start) & 0xFFFF); |
|
idRangeOffsets += string16(0); |
|
} |
|
} |
|
|
|
if (trailingRangesCount > 0) { |
|
endCount += '\xFF\xFF'; |
|
startCount += '\xFF\xFF'; |
|
idDeltas += '\x00\x01'; |
|
idRangeOffsets += '\x00\x00'; |
|
} |
|
|
|
var format314 = '\x00\x00' + // language |
|
string16(2 * segCount) + |
|
string16(searchParams.range) + |
|
string16(searchParams.entry) + |
|
string16(searchParams.rangeShift) + |
|
endCount + '\x00\x00' + startCount + |
|
idDeltas + idRangeOffsets + glyphsIds; |
|
|
|
var format31012 = ''; |
|
var header31012 = ''; |
|
if (numTables > 1) { |
|
cmap += '\x00\x03' + // platformID |
|
'\x00\x0A' + // encodingID |
|
string32(4 + numTables * 8 + |
|
4 + format314.length); // start of the table record |
|
format31012 = ''; |
|
for (i = 0, ii = ranges.length; i < ii; i++) { |
|
range = ranges[i]; |
|
start = range[0]; |
|
codes = range[2]; |
|
var code = codes[0]; |
|
for (j = 1, jj = codes.length; j < jj; ++j) { |
|
if (codes[j] !== codes[j - 1] + 1) { |
|
end = range[0] + j - 1; |
|
format31012 += string32(start) + // startCharCode |
|
string32(end) + // endCharCode |
|
string32(code); // startGlyphID |
|
start = end + 1; |
|
code = codes[j]; |
|
} |
|
} |
|
format31012 += string32(start) + // startCharCode |
|
string32(range[1]) + // endCharCode |
|
string32(code); // startGlyphID |
|
} |
|
header31012 = '\x00\x0C' + // format |
|
'\x00\x00' + // reserved |
|
string32(format31012.length + 16) + // length |
|
'\x00\x00\x00\x00' + // language |
|
string32(format31012.length / 12); // nGroups |
|
} |
|
|
|
return cmap + '\x00\x04' + // format |
|
string16(format314.length + 4) + // length |
|
format314 + header31012 + format31012; |
|
} |
|
|
|
function validateOS2Table(os2) { |
|
var stream = new Stream(os2.data); |
|
var version = stream.getUint16(); |
|
// TODO verify all OS/2 tables fields, but currently we validate only those |
|
// that give us issues |
|
stream.getBytes(60); // skipping type, misc sizes, panose, unicode ranges |
|
var selection = stream.getUint16(); |
|
if (version < 4 && (selection & 0x0300)) { |
|
return false; |
|
} |
|
var firstChar = stream.getUint16(); |
|
var lastChar = stream.getUint16(); |
|
if (firstChar > lastChar) { |
|
return false; |
|
} |
|
stream.getBytes(6); // skipping sTypoAscender/Descender/LineGap |
|
var usWinAscent = stream.getUint16(); |
|
if (usWinAscent === 0) { // makes font unreadable by windows |
|
return false; |
|
} |
|
|
|
// OS/2 appears to be valid, resetting some fields |
|
os2.data[8] = os2.data[9] = 0; // IE rejects fonts if fsType != 0 |
|
return true; |
|
} |
|
|
|
function createOS2Table(properties, charstrings, override) { |
|
override = override || { |
|
unitsPerEm: 0, |
|
yMax: 0, |
|
yMin: 0, |
|
ascent: 0, |
|
descent: 0 |
|
}; |
|
|
|
var ulUnicodeRange1 = 0; |
|
var ulUnicodeRange2 = 0; |
|
var ulUnicodeRange3 = 0; |
|
var ulUnicodeRange4 = 0; |
|
|
|
var firstCharIndex = null; |
|
var lastCharIndex = 0; |
|
|
|
if (charstrings) { |
|
for (var code in charstrings) { |
|
code |= 0; |
|
if (firstCharIndex > code || !firstCharIndex) { |
|
firstCharIndex = code; |
|
} |
|
if (lastCharIndex < code) { |
|
lastCharIndex = code; |
|
} |
|
|
|
var position = getUnicodeRangeFor(code); |
|
if (position < 32) { |
|
ulUnicodeRange1 |= 1 << position; |
|
} else if (position < 64) { |
|
ulUnicodeRange2 |= 1 << position - 32; |
|
} else if (position < 96) { |
|
ulUnicodeRange3 |= 1 << position - 64; |
|
} else if (position < 123) { |
|
ulUnicodeRange4 |= 1 << position - 96; |
|
} else { |
|
error('Unicode ranges Bits > 123 are reserved for internal usage'); |
|
} |
|
} |
|
} else { |
|
// TODO |
|
firstCharIndex = 0; |
|
lastCharIndex = 255; |
|
} |
|
|
|
var bbox = properties.bbox || [0, 0, 0, 0]; |
|
var unitsPerEm = (override.unitsPerEm || |
|
1 / (properties.fontMatrix || FONT_IDENTITY_MATRIX)[0]); |
|
|
|
// if the font units differ to the PDF glyph space units |
|
// then scale up the values |
|
var scale = (properties.ascentScaled ? 1.0 : |
|
unitsPerEm / PDF_GLYPH_SPACE_UNITS); |
|
|
|
var typoAscent = (override.ascent || |
|
Math.round(scale * (properties.ascent || bbox[3]))); |
|
var typoDescent = (override.descent || |
|
Math.round(scale * (properties.descent || bbox[1]))); |
|
if (typoDescent > 0 && properties.descent > 0 && bbox[1] < 0) { |
|
typoDescent = -typoDescent; // fixing incorrect descent |
|
} |
|
var winAscent = override.yMax || typoAscent; |
|
var winDescent = -override.yMin || -typoDescent; |
|
|
|
return '\x00\x03' + // version |
|
'\x02\x24' + // xAvgCharWidth |
|
'\x01\xF4' + // usWeightClass |
|
'\x00\x05' + // usWidthClass |
|
'\x00\x00' + // fstype (0 to let the font loads via font-face on IE) |
|
'\x02\x8A' + // ySubscriptXSize |
|
'\x02\xBB' + // ySubscriptYSize |
|
'\x00\x00' + // ySubscriptXOffset |
|
'\x00\x8C' + // ySubscriptYOffset |
|
'\x02\x8A' + // ySuperScriptXSize |
|
'\x02\xBB' + // ySuperScriptYSize |
|
'\x00\x00' + // ySuperScriptXOffset |
|
'\x01\xDF' + // ySuperScriptYOffset |
|
'\x00\x31' + // yStrikeOutSize |
|
'\x01\x02' + // yStrikeOutPosition |
|
'\x00\x00' + // sFamilyClass |
|
'\x00\x00\x06' + |
|
String.fromCharCode(properties.fixedPitch ? 0x09 : 0x00) + |
|
'\x00\x00\x00\x00\x00\x00' + // Panose |
|
string32(ulUnicodeRange1) + // ulUnicodeRange1 (Bits 0-31) |
|
string32(ulUnicodeRange2) + // ulUnicodeRange2 (Bits 32-63) |
|
string32(ulUnicodeRange3) + // ulUnicodeRange3 (Bits 64-95) |
|
string32(ulUnicodeRange4) + // ulUnicodeRange4 (Bits 96-127) |
|
'\x2A\x32\x31\x2A' + // achVendID |
|
string16(properties.italicAngle ? 1 : 0) + // fsSelection |
|
string16(firstCharIndex || |
|
properties.firstChar) + // usFirstCharIndex |
|
string16(lastCharIndex || properties.lastChar) + // usLastCharIndex |
|
string16(typoAscent) + // sTypoAscender |
|
string16(typoDescent) + // sTypoDescender |
|
'\x00\x64' + // sTypoLineGap (7%-10% of the unitsPerEM value) |
|
string16(winAscent) + // usWinAscent |
|
string16(winDescent) + // usWinDescent |
|
'\x00\x00\x00\x00' + // ulCodePageRange1 (Bits 0-31) |
|
'\x00\x00\x00\x00' + // ulCodePageRange2 (Bits 32-63) |
|
string16(properties.xHeight) + // sxHeight |
|
string16(properties.capHeight) + // sCapHeight |
|
string16(0) + // usDefaultChar |
|
string16(firstCharIndex || properties.firstChar) + // usBreakChar |
|
'\x00\x03'; // usMaxContext |
|
} |
|
|
|
function createPostTable(properties) { |
|
var angle = Math.floor(properties.italicAngle * (Math.pow(2, 16))); |
|
return ('\x00\x03\x00\x00' + // Version number |
|
string32(angle) + // italicAngle |
|
'\x00\x00' + // underlinePosition |
|
'\x00\x00' + // underlineThickness |
|
string32(properties.fixedPitch) + // isFixedPitch |
|
'\x00\x00\x00\x00' + // minMemType42 |
|
'\x00\x00\x00\x00' + // maxMemType42 |
|
'\x00\x00\x00\x00' + // minMemType1 |
|
'\x00\x00\x00\x00'); // maxMemType1 |
|
} |
|
|
|
function createNameTable(name, proto) { |
|
if (!proto) { |
|
proto = [[], []]; // no strings and unicode strings |
|
} |
|
|
|
var strings = [ |
|
proto[0][0] || 'Original licence', // 0.Copyright |
|
proto[0][1] || name, // 1.Font family |
|
proto[0][2] || 'Unknown', // 2.Font subfamily (font weight) |
|
proto[0][3] || 'uniqueID', // 3.Unique ID |
|
proto[0][4] || name, // 4.Full font name |
|
proto[0][5] || 'Version 0.11', // 5.Version |
|
proto[0][6] || '', // 6.Postscript name |
|
proto[0][7] || 'Unknown', // 7.Trademark |
|
proto[0][8] || 'Unknown', // 8.Manufacturer |
|
proto[0][9] || 'Unknown' // 9.Designer |
|
]; |
|
|
|
// Mac want 1-byte per character strings while Windows want |
|
// 2-bytes per character, so duplicate the names table |
|
var stringsUnicode = []; |
|
var i, ii, j, jj, str; |
|
for (i = 0, ii = strings.length; i < ii; i++) { |
|
str = proto[1][i] || strings[i]; |
|
|
|
var strBufUnicode = []; |
|
for (j = 0, jj = str.length; j < jj; j++) { |
|
strBufUnicode.push(string16(str.charCodeAt(j))); |
|
} |
|
stringsUnicode.push(strBufUnicode.join('')); |
|
} |
|
|
|
var names = [strings, stringsUnicode]; |
|
var platforms = ['\x00\x01', '\x00\x03']; |
|
var encodings = ['\x00\x00', '\x00\x01']; |
|
var languages = ['\x00\x00', '\x04\x09']; |
|
|
|
var namesRecordCount = strings.length * platforms.length; |
|
var nameTable = |
|
'\x00\x00' + // format |
|
string16(namesRecordCount) + // Number of names Record |
|
string16(namesRecordCount * 12 + 6); // Storage |
|
|
|
// Build the name records field |
|
var strOffset = 0; |
|
for (i = 0, ii = platforms.length; i < ii; i++) { |
|
var strs = names[i]; |
|
for (j = 0, jj = strs.length; j < jj; j++) { |
|
str = strs[j]; |
|
var nameRecord = |
|
platforms[i] + // platform ID |
|
encodings[i] + // encoding ID |
|
languages[i] + // language ID |
|
string16(j) + // name ID |
|
string16(str.length) + |
|
string16(strOffset); |
|
nameTable += nameRecord; |
|
strOffset += str.length; |
|
} |
|
} |
|
|
|
nameTable += strings.join('') + stringsUnicode.join(''); |
|
return nameTable; |
|
} |
|
|
|
Font.prototype = { |
|
name: null, |
|
font: null, |
|
mimetype: null, |
|
encoding: null, |
|
get renderer() { |
|
var renderer = FontRendererFactory.create(this); |
|
return shadow(this, 'renderer', renderer); |
|
}, |
|
|
|
exportData: function Font_exportData() { |
|
var data = {}; |
|
for (var i in this) { |
|
if (this.hasOwnProperty(i)) { |
|
data[i] = this[i]; |
|
} |
|
} |
|
return data; |
|
}, |
|
|
|
checkAndRepair: function Font_checkAndRepair(name, font, properties) { |
|
function readTableEntry(file) { |
|
var tag = bytesToString(file.getBytes(4)); |
|
|
|
var checksum = file.getInt32(); |
|
var offset = file.getInt32() >>> 0; |
|
var length = file.getInt32() >>> 0; |
|
|
|
// Read the table associated data |
|
var previousPosition = file.pos; |
|
file.pos = file.start ? file.start : 0; |
|
file.skip(offset); |
|
var data = file.getBytes(length); |
|
file.pos = previousPosition; |
|
|
|
if (tag === 'head') { |
|
// clearing checksum adjustment |
|
data[8] = data[9] = data[10] = data[11] = 0; |
|
data[17] |= 0x20; //Set font optimized for cleartype flag |
|
} |
|
|
|
return { |
|
tag: tag, |
|
checksum: checksum, |
|
length: length, |
|
offset: offset, |
|
data: data |
|
}; |
|
} |
|
|
|
function readOpenTypeHeader(ttf) { |
|
return { |
|
version: bytesToString(ttf.getBytes(4)), |
|
numTables: ttf.getUint16(), |
|
searchRange: ttf.getUint16(), |
|
entrySelector: ttf.getUint16(), |
|
rangeShift: ttf.getUint16() |
|
}; |
|
} |
|
|
|
/** |
|
* Read the appropriate subtable from the cmap according to 9.6.6.4 from |
|
* PDF spec |
|
*/ |
|
function readCmapTable(cmap, font, isSymbolicFont, hasEncoding) { |
|
if (!cmap) { |
|
warn('No cmap table available.'); |
|
return { |
|
platformId: -1, |
|
encodingId: -1, |
|
mappings: [], |
|
hasShortCmap: false |
|
}; |
|
} |
|
var segment; |
|
var start = (font.start ? font.start : 0) + cmap.offset; |
|
font.pos = start; |
|
|
|
var version = font.getUint16(); |
|
var numTables = font.getUint16(); |
|
|
|
var potentialTable; |
|
var canBreak = false; |
|
// There's an order of preference in terms of which cmap subtable to |
|
// use: |
|
// - non-symbolic fonts the preference is a 3,1 table then a 1,0 table |
|
// - symbolic fonts the preference is a 3,0 table then a 1,0 table |
|
// The following takes advantage of the fact that the tables are sorted |
|
// to work. |
|
for (var i = 0; i < numTables; i++) { |
|
var platformId = font.getUint16(); |
|
var encodingId = font.getUint16(); |
|
var offset = font.getInt32() >>> 0; |
|
var useTable = false; |
|
|
|
if (platformId === 0 && encodingId === 0) { |
|
useTable = true; |
|
// Continue the loop since there still may be a higher priority |
|
// table. |
|
} else if (platformId === 1 && encodingId === 0) { |
|
useTable = true; |
|
// Continue the loop since there still may be a higher priority |
|
// table. |
|
} else if (platformId === 3 && encodingId === 1 && |
|
((!isSymbolicFont && hasEncoding) || !potentialTable)) { |
|
useTable = true; |
|
if (!isSymbolicFont) { |
|
canBreak = true; |
|
} |
|
} else if (isSymbolicFont && platformId === 3 && encodingId === 0) { |
|
useTable = true; |
|
canBreak = true; |
|
} |
|
|
|
if (useTable) { |
|
potentialTable = { |
|
platformId: platformId, |
|
encodingId: encodingId, |
|
offset: offset |
|
}; |
|
} |
|
if (canBreak) { |
|
break; |
|
} |
|
} |
|
|
|
if (potentialTable) { |
|
font.pos = start + potentialTable.offset; |
|
} |
|
if (!potentialTable || font.peekByte() === -1) { |
|
warn('Could not find a preferred cmap table.'); |
|
return { |
|
platformId: -1, |
|
encodingId: -1, |
|
mappings: [], |
|
hasShortCmap: false |
|
}; |
|
} |
|
|
|
var format = font.getUint16(); |
|
var length = font.getUint16(); |
|
var language = font.getUint16(); |
|
|
|
var hasShortCmap = false; |
|
var mappings = []; |
|
var j, glyphId; |
|
|
|
// TODO(mack): refactor this cmap subtable reading logic out |
|
if (format === 0) { |
|
for (j = 0; j < 256; j++) { |
|
var index = font.getByte(); |
|
if (!index) { |
|
continue; |
|
} |
|
mappings.push({ |
|
charCode: j, |
|
glyphId: index |
|
}); |
|
} |
|
hasShortCmap = true; |
|
} else if (format === 4) { |
|
// re-creating the table in format 4 since the encoding |
|
// might be changed |
|
var segCount = (font.getUint16() >> 1); |
|
font.getBytes(6); // skipping range fields |
|
var segIndex, segments = []; |
|
for (segIndex = 0; segIndex < segCount; segIndex++) { |
|
segments.push({ end: font.getUint16() }); |
|
} |
|
font.getUint16(); |
|
for (segIndex = 0; segIndex < segCount; segIndex++) { |
|
segments[segIndex].start = font.getUint16(); |
|
} |
|
|
|
for (segIndex = 0; segIndex < segCount; segIndex++) { |
|
segments[segIndex].delta = font.getUint16(); |
|
} |
|
|
|
var offsetsCount = 0; |
|
for (segIndex = 0; segIndex < segCount; segIndex++) { |
|
segment = segments[segIndex]; |
|
var rangeOffset = font.getUint16(); |
|
if (!rangeOffset) { |
|
segment.offsetIndex = -1; |
|
continue; |
|
} |
|
|
|
var offsetIndex = (rangeOffset >> 1) - (segCount - segIndex); |
|
segment.offsetIndex = offsetIndex; |
|
offsetsCount = Math.max(offsetsCount, offsetIndex + |
|
segment.end - segment.start + 1); |
|
} |
|
|
|
var offsets = []; |
|
for (j = 0; j < offsetsCount; j++) { |
|
offsets.push(font.getUint16()); |
|
} |
|
|
|
for (segIndex = 0; segIndex < segCount; segIndex++) { |
|
segment = segments[segIndex]; |
|
start = segment.start; |
|
var end = segment.end; |
|
var delta = segment.delta; |
|
offsetIndex = segment.offsetIndex; |
|
|
|
for (j = start; j <= end; j++) { |
|
if (j === 0xFFFF) { |
|
continue; |
|
} |
|
|
|
glyphId = (offsetIndex < 0 ? |
|
j : offsets[offsetIndex + j - start]); |
|
glyphId = (glyphId + delta) & 0xFFFF; |
|
if (glyphId === 0) { |
|
continue; |
|
} |
|
mappings.push({ |
|
charCode: j, |
|
glyphId: glyphId |
|
}); |
|
} |
|
} |
|
} else if (format === 6) { |
|
// Format 6 is a 2-bytes dense mapping, which means the font data |
|
// lives glue together even if they are pretty far in the unicode |
|
// table. (This looks weird, so I can have missed something), this |
|
// works on Linux but seems to fails on Mac so let's rewrite the |
|
// cmap table to a 3-1-4 style |
|
var firstCode = font.getUint16(); |
|
var entryCount = font.getUint16(); |
|
|
|
for (j = 0; j < entryCount; j++) { |
|
glyphId = font.getUint16(); |
|
var charCode = firstCode + j; |
|
|
|
mappings.push({ |
|
charCode: charCode, |
|
glyphId: glyphId |
|
}); |
|
} |
|
} else { |
|
warn('cmap table has unsupported format: ' + format); |
|
return { |
|
platformId: -1, |
|
encodingId: -1, |
|
mappings: [], |
|
hasShortCmap: false |
|
}; |
|
} |
|
|
|
// removing duplicate entries |
|
mappings.sort(function (a, b) { |
|
return a.charCode - b.charCode; |
|
}); |
|
for (i = 1; i < mappings.length; i++) { |
|
if (mappings[i - 1].charCode === mappings[i].charCode) { |
|
mappings.splice(i, 1); |
|
i--; |
|
} |
|
} |
|
|
|
return { |
|
platformId: potentialTable.platformId, |
|
encodingId: potentialTable.encodingId, |
|
mappings: mappings, |
|
hasShortCmap: hasShortCmap |
|
}; |
|
} |
|
|
|
function sanitizeMetrics(font, header, metrics, numGlyphs) { |
|
if (!header) { |
|
if (metrics) { |
|
metrics.data = null; |
|
} |
|
return; |
|
} |
|
|
|
font.pos = (font.start ? font.start : 0) + header.offset; |
|
font.pos += header.length - 2; |
|
var numOfMetrics = font.getUint16(); |
|
|
|
if (numOfMetrics > numGlyphs) { |
|
info('The numOfMetrics (' + numOfMetrics + ') should not be ' + |
|
'greater than the numGlyphs (' + numGlyphs + ')'); |
|
// Reduce numOfMetrics if it is greater than numGlyphs |
|
numOfMetrics = numGlyphs; |
|
header.data[34] = (numOfMetrics & 0xff00) >> 8; |
|
header.data[35] = numOfMetrics & 0x00ff; |
|
} |
|
|
|
var numOfSidebearings = numGlyphs - numOfMetrics; |
|
var numMissing = numOfSidebearings - |
|
((metrics.length - numOfMetrics * 4) >> 1); |
|
|
|
if (numMissing > 0) { |
|
// For each missing glyph, we set both the width and lsb to 0 (zero). |
|
// Since we need to add two properties for each glyph, this explains |
|
// the use of |numMissing * 2| when initializing the typed array. |
|
var entries = new Uint8Array(metrics.length + numMissing * 2); |
|
entries.set(metrics.data); |
|
metrics.data = entries; |
|
} |
|
} |
|
|
|
function sanitizeGlyph(source, sourceStart, sourceEnd, dest, destStart, |
|
hintsValid) { |
|
if (sourceEnd - sourceStart <= 12) { |
|
// glyph with data less than 12 is invalid one |
|
return 0; |
|
} |
|
var glyf = source.subarray(sourceStart, sourceEnd); |
|
var contoursCount = (glyf[0] << 8) | glyf[1]; |
|
if (contoursCount & 0x8000) { |
|
// complex glyph, writing as is |
|
dest.set(glyf, destStart); |
|
return glyf.length; |
|
} |
|
|
|
var i, j = 10, flagsCount = 0; |
|
for (i = 0; i < contoursCount; i++) { |
|
var endPoint = (glyf[j] << 8) | glyf[j + 1]; |
|
flagsCount = endPoint + 1; |
|
j += 2; |
|
} |
|
// skipping instructions |
|
var instructionsStart = j; |
|
var instructionsLength = (glyf[j] << 8) | glyf[j + 1]; |
|
j += 2 + instructionsLength; |
|
var instructionsEnd = j; |
|
// validating flags |
|
var coordinatesLength = 0; |
|
for (i = 0; i < flagsCount; i++) { |
|
var flag = glyf[j++]; |
|
if (flag & 0xC0) { |
|
// reserved flags must be zero, cleaning up |
|
glyf[j - 1] = flag & 0x3F; |
|
} |
|
var xyLength = ((flag & 2) ? 1 : (flag & 16) ? 0 : 2) + |
|
((flag & 4) ? 1 : (flag & 32) ? 0 : 2); |
|
coordinatesLength += xyLength; |
|
if (flag & 8) { |
|
var repeat = glyf[j++]; |
|
i += repeat; |
|
coordinatesLength += repeat * xyLength; |
|
} |
|
} |
|
// glyph without coordinates will be rejected |
|
if (coordinatesLength === 0) { |
|
return 0; |
|
} |
|
var glyphDataLength = j + coordinatesLength; |
|
if (glyphDataLength > glyf.length) { |
|
// not enough data for coordinates |
|
return 0; |
|
} |
|
if (!hintsValid && instructionsLength > 0) { |
|
dest.set(glyf.subarray(0, instructionsStart), destStart); |
|
dest.set([0, 0], destStart + instructionsStart); |
|
dest.set(glyf.subarray(instructionsEnd, glyphDataLength), |
|
destStart + instructionsStart + 2); |
|
glyphDataLength -= instructionsLength; |
|
if (glyf.length - glyphDataLength > 3) { |
|
glyphDataLength = (glyphDataLength + 3) & ~3; |
|
} |
|
return glyphDataLength; |
|
} |
|
if (glyf.length - glyphDataLength > 3) { |
|
// truncating and aligning to 4 bytes the long glyph data |
|
glyphDataLength = (glyphDataLength + 3) & ~3; |
|
dest.set(glyf.subarray(0, glyphDataLength), destStart); |
|
return glyphDataLength; |
|
} |
|
// glyph data is fine |
|
dest.set(glyf, destStart); |
|
return glyf.length; |
|
} |
|
|
|
function sanitizeHead(head, numGlyphs, locaLength) { |
|
var data = head.data; |
|
|
|
// Validate version: |
|
// Should always be 0x00010000 |
|
var version = int32(data[0], data[1], data[2], data[3]); |
|
if (version >> 16 !== 1) { |
|
info('Attempting to fix invalid version in head table: ' + version); |
|
data[0] = 0; |
|
data[1] = 1; |
|
data[2] = 0; |
|
data[3] = 0; |
|
} |
|
|
|
var indexToLocFormat = int16(data[50], data[51]); |
|
if (indexToLocFormat < 0 || indexToLocFormat > 1) { |
|
info('Attempting to fix invalid indexToLocFormat in head table: ' + |
|
indexToLocFormat); |
|
|
|
// The value of indexToLocFormat should be 0 if the loca table |
|
// consists of short offsets, and should be 1 if the loca table |
|
// consists of long offsets. |
|
// |
|
// The number of entries in the loca table should be numGlyphs + 1. |
|
// |
|
// Using this information, we can work backwards to deduce if the |
|
// size of each offset in the loca table, and thus figure out the |
|
// appropriate value for indexToLocFormat. |
|
|
|
var numGlyphsPlusOne = numGlyphs + 1; |
|
if (locaLength === numGlyphsPlusOne << 1) { |
|
// 0x0000 indicates the loca table consists of short offsets |
|
data[50] = 0; |
|
data[51] = 0; |
|
} else if (locaLength === numGlyphsPlusOne << 2) { |
|
// 0x0001 indicates the loca table consists of long offsets |
|
data[50] = 0; |
|
data[51] = 1; |
|
} else { |
|
warn('Could not fix indexToLocFormat: ' + indexToLocFormat); |
|
} |
|
} |
|
} |
|
|
|
function sanitizeGlyphLocations(loca, glyf, numGlyphs, |
|
isGlyphLocationsLong, hintsValid, |
|
dupFirstEntry) { |
|
var itemSize, itemDecode, itemEncode; |
|
if (isGlyphLocationsLong) { |
|
itemSize = 4; |
|
itemDecode = function fontItemDecodeLong(data, offset) { |
|
return (data[offset] << 24) | (data[offset + 1] << 16) | |
|
(data[offset + 2] << 8) | data[offset + 3]; |
|
}; |
|
itemEncode = function fontItemEncodeLong(data, offset, value) { |
|
data[offset] = (value >>> 24) & 0xFF; |
|
data[offset + 1] = (value >> 16) & 0xFF; |
|
data[offset + 2] = (value >> 8) & 0xFF; |
|
data[offset + 3] = value & 0xFF; |
|
}; |
|
} else { |
|
itemSize = 2; |
|
itemDecode = function fontItemDecode(data, offset) { |
|
return (data[offset] << 9) | (data[offset + 1] << 1); |
|
}; |
|
itemEncode = function fontItemEncode(data, offset, value) { |
|
data[offset] = (value >> 9) & 0xFF; |
|
data[offset + 1] = (value >> 1) & 0xFF; |
|
}; |
|
} |
|
var locaData = loca.data; |
|
var locaDataSize = itemSize * (1 + numGlyphs); |
|
// is loca.data too short or long? |
|
if (locaData.length !== locaDataSize) { |
|
locaData = new Uint8Array(locaDataSize); |
|
locaData.set(loca.data.subarray(0, locaDataSize)); |
|
loca.data = locaData; |
|
} |
|
// removing the invalid glyphs |
|
var oldGlyfData = glyf.data; |
|
var oldGlyfDataLength = oldGlyfData.length; |
|
var newGlyfData = new Uint8Array(oldGlyfDataLength); |
|
var startOffset = itemDecode(locaData, 0); |
|
var writeOffset = 0; |
|
var missingGlyphData = {}; |
|
itemEncode(locaData, 0, writeOffset); |
|
var i, j; |
|
for (i = 0, j = itemSize; i < numGlyphs; i++, j += itemSize) { |
|
var endOffset = itemDecode(locaData, j); |
|
if (endOffset > oldGlyfDataLength && |
|
((oldGlyfDataLength + 3) & ~3) === endOffset) { |
|
// Aspose breaks fonts by aligning the glyphs to the qword, but not |
|
// the glyf table size, which makes last glyph out of range. |
|
endOffset = oldGlyfDataLength; |
|
} |
|
if (endOffset > oldGlyfDataLength) { |
|
// glyph end offset points outside glyf data, rejecting the glyph |
|
itemEncode(locaData, j, writeOffset); |
|
startOffset = endOffset; |
|
continue; |
|
} |
|
|
|
if (startOffset === endOffset) { |
|
missingGlyphData[i] = true; |
|
} |
|
|
|
var newLength = sanitizeGlyph(oldGlyfData, startOffset, endOffset, |
|
newGlyfData, writeOffset, hintsValid); |
|
writeOffset += newLength; |
|
itemEncode(locaData, j, writeOffset); |
|
startOffset = endOffset; |
|
} |
|
|
|
if (writeOffset === 0) { |
|
// glyf table cannot be empty -- redoing the glyf and loca tables |
|
// to have single glyph with one point |
|
var simpleGlyph = new Uint8Array( |
|
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 49, 0]); |
|
for (i = 0, j = itemSize; i < numGlyphs; i++, j += itemSize) { |
|
itemEncode(locaData, j, simpleGlyph.length); |
|
} |
|
glyf.data = simpleGlyph; |
|
return missingGlyphData; |
|
} |
|
|
|
if (dupFirstEntry) { |
|
var firstEntryLength = itemDecode(locaData, itemSize); |
|
if (newGlyfData.length > firstEntryLength + writeOffset) { |
|
glyf.data = newGlyfData.subarray(0, firstEntryLength + writeOffset); |
|
} else { |
|
glyf.data = new Uint8Array(firstEntryLength + writeOffset); |
|
glyf.data.set(newGlyfData.subarray(0, writeOffset)); |
|
} |
|
glyf.data.set(newGlyfData.subarray(0, firstEntryLength), writeOffset); |
|
itemEncode(loca.data, locaData.length - itemSize, |
|
writeOffset + firstEntryLength); |
|
} else { |
|
glyf.data = newGlyfData.subarray(0, writeOffset); |
|
} |
|
return missingGlyphData; |
|
} |
|
|
|
function readPostScriptTable(post, properties, maxpNumGlyphs) { |
|
var start = (font.start ? font.start : 0) + post.offset; |
|
font.pos = start; |
|
|
|
var length = post.length, end = start + length; |
|
var version = font.getInt32(); |
|
// skip rest to the tables |
|
font.getBytes(28); |
|
|
|
var glyphNames; |
|
var valid = true; |
|
var i; |
|
|
|
switch (version) { |
|
case 0x00010000: |
|
glyphNames = MacStandardGlyphOrdering; |
|
break; |
|
case 0x00020000: |
|
var numGlyphs = font.getUint16(); |
|
if (numGlyphs !== maxpNumGlyphs) { |
|
valid = false; |
|
break; |
|
} |
|
var glyphNameIndexes = []; |
|
for (i = 0; i < numGlyphs; ++i) { |
|
var index = font.getUint16(); |
|
if (index >= 32768) { |
|
valid = false; |
|
break; |
|
} |
|
glyphNameIndexes.push(index); |
|
} |
|
if (!valid) { |
|
break; |
|
} |
|
var customNames = []; |
|
var strBuf = []; |
|
while (font.pos < end) { |
|
var stringLength = font.getByte(); |
|
strBuf.length = stringLength; |
|
for (i = 0; i < stringLength; ++i) { |
|
strBuf[i] = String.fromCharCode(font.getByte()); |
|
} |
|
customNames.push(strBuf.join('')); |
|
} |
|
glyphNames = []; |
|
for (i = 0; i < numGlyphs; ++i) { |
|
var j = glyphNameIndexes[i]; |
|
if (j < 258) { |
|
glyphNames.push(MacStandardGlyphOrdering[j]); |
|
continue; |
|
} |
|
glyphNames.push(customNames[j - 258]); |
|
} |
|
break; |
|
case 0x00030000: |
|
break; |
|
default: |
|
warn('Unknown/unsupported post table version ' + version); |
|
valid = false; |
|
if (properties.defaultEncoding) { |
|
glyphNames = properties.defaultEncoding; |
|
} |
|
break; |
|
} |
|
properties.glyphNames = glyphNames; |
|
return valid; |
|
} |
|
|
|
function readNameTable(nameTable) { |
|
var start = (font.start ? font.start : 0) + nameTable.offset; |
|
font.pos = start; |
|
|
|
var names = [[], []]; |
|
var length = nameTable.length, end = start + length; |
|
var format = font.getUint16(); |
|
var FORMAT_0_HEADER_LENGTH = 6; |
|
if (format !== 0 || length < FORMAT_0_HEADER_LENGTH) { |
|
// unsupported name table format or table "too" small |
|
return names; |
|
} |
|
var numRecords = font.getUint16(); |
|
var stringsStart = font.getUint16(); |
|
var records = []; |
|
var NAME_RECORD_LENGTH = 12; |
|
var i, ii; |
|
|
|
for (i = 0; i < numRecords && |
|
font.pos + NAME_RECORD_LENGTH <= end; i++) { |
|
var r = { |
|
platform: font.getUint16(), |
|
encoding: font.getUint16(), |
|
language: font.getUint16(), |
|
name: font.getUint16(), |
|
length: font.getUint16(), |
|
offset: font.getUint16() |
|
}; |
|
// using only Macintosh and Windows platform/encoding names |
|
if ((r.platform === 1 && r.encoding === 0 && r.language === 0) || |
|
(r.platform === 3 && r.encoding === 1 && r.language === 0x409)) { |
|
records.push(r); |
|
} |
|
} |
|
for (i = 0, ii = records.length; i < ii; i++) { |
|
var record = records[i]; |
|
var pos = start + stringsStart + record.offset; |
|
if (pos + record.length > end) { |
|
continue; // outside of name table, ignoring |
|
} |
|
font.pos = pos; |
|
var nameIndex = record.name; |
|
if (record.encoding) { |
|
// unicode |
|
var str = ''; |
|
for (var j = 0, jj = record.length; j < jj; j += 2) { |
|
str += String.fromCharCode(font.getUint16()); |
|
} |
|
names[1][nameIndex] = str; |
|
} else { |
|
names[0][nameIndex] = bytesToString(font.getBytes(record.length)); |
|
} |
|
} |
|
return names; |
|
} |
|
|
|
var TTOpsStackDeltas = [ |
|
0, 0, 0, 0, 0, 0, 0, 0, -2, -2, -2, -2, 0, 0, -2, -5, |
|
-1, -1, -1, -1, -1, -1, -1, -1, 0, 0, -1, 0, -1, -1, -1, -1, |
|
1, -1, -999, 0, 1, 0, -1, -2, 0, -1, -2, -1, -1, 0, -1, -1, |
|
0, 0, -999, -999, -1, -1, -1, -1, -2, -999, -2, -2, -999, 0, -2, -2, |
|
0, 0, -2, 0, -2, 0, 0, 0, -2, -1, -1, 1, 1, 0, 0, -1, |
|
-1, -1, -1, -1, -1, -1, 0, 0, -1, 0, -1, -1, 0, -999, -1, -1, |
|
-1, -1, -1, -1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, |
|
-2, -999, -999, -999, -999, -999, -1, -1, -2, -2, 0, 0, 0, 0, -1, -1, |
|
-999, -2, -2, 0, 0, -1, -2, -2, 0, 0, 0, -1, -1, -1, -2]; |
|
// 0xC0-DF == -1 and 0xE0-FF == -2 |
|
|
|
function sanitizeTTProgram(table, ttContext) { |
|
var data = table.data; |
|
var i = 0, j, n, b, funcId, pc, lastEndf = 0, lastDeff = 0; |
|
var stack = []; |
|
var callstack = []; |
|
var functionsCalled = []; |
|
var tooComplexToFollowFunctions = |
|
ttContext.tooComplexToFollowFunctions; |
|
var inFDEF = false, ifLevel = 0, inELSE = 0; |
|
for (var ii = data.length; i < ii;) { |
|
var op = data[i++]; |
|
// The TrueType instruction set docs can be found at |
|
// https://developer.apple.com/fonts/TTRefMan/RM05/Chap5.html |
|
if (op === 0x40) { // NPUSHB - pushes n bytes |
|
n = data[i++]; |
|
if (inFDEF || inELSE) { |
|
i += n; |
|
} else { |
|
for (j = 0; j < n; j++) { |
|
stack.push(data[i++]); |
|
} |
|
} |
|
} else if (op === 0x41) { // NPUSHW - pushes n words |
|
n = data[i++]; |
|
if (inFDEF || inELSE) { |
|
i += n * 2; |
|
} else { |
|
for (j = 0; j < n; j++) { |
|
b = data[i++]; |
|
stack.push((b << 8) | data[i++]); |
|
} |
|
} |
|
} else if ((op & 0xF8) === 0xB0) { // PUSHB - pushes bytes |
|
n = op - 0xB0 + 1; |
|
if (inFDEF || inELSE) { |
|
i += n; |
|
} else { |
|
for (j = 0; j < n; j++) { |
|
stack.push(data[i++]); |
|
} |
|
} |
|
} else if ((op & 0xF8) === 0xB8) { // PUSHW - pushes words |
|
n = op - 0xB8 + 1; |
|
if (inFDEF || inELSE) { |
|
i += n * 2; |
|
} else { |
|
for (j = 0; j < n; j++) { |
|
b = data[i++]; |
|
stack.push((b << 8) | data[i++]); |
|
} |
|
} |
|
} else if (op === 0x2B && !tooComplexToFollowFunctions) { // CALL |
|
if (!inFDEF && !inELSE) { |
|
// collecting inforamtion about which functions are used |
|
funcId = stack[stack.length - 1]; |
|
ttContext.functionsUsed[funcId] = true; |
|
if (funcId in ttContext.functionsStackDeltas) { |
|
stack.length += ttContext.functionsStackDeltas[funcId]; |
|
} else if (funcId in ttContext.functionsDefined && |
|
functionsCalled.indexOf(funcId) < 0) { |
|
callstack.push({data: data, i: i, stackTop: stack.length - 1}); |
|
functionsCalled.push(funcId); |
|
pc = ttContext.functionsDefined[funcId]; |
|
if (!pc) { |
|
warn('TT: CALL non-existent function'); |
|
ttContext.hintsValid = false; |
|
return; |
|
} |
|
data = pc.data; |
|
i = pc.i; |
|
} |
|
} |
|
} else if (op === 0x2C && !tooComplexToFollowFunctions) { // FDEF |
|
if (inFDEF || inELSE) { |
|
warn('TT: nested FDEFs not allowed'); |
|
tooComplexToFollowFunctions = true; |
|
} |
|
inFDEF = true; |
|
// collecting inforamtion about which functions are defined |
|
lastDeff = i; |
|
funcId = stack.pop(); |
|
ttContext.functionsDefined[funcId] = {data: data, i: i}; |
|
} else if (op === 0x2D) { // ENDF - end of function |
|
if (inFDEF) { |
|
inFDEF = false; |
|
lastEndf = i; |
|
} else { |
|
pc = callstack.pop(); |
|
if (!pc) { |
|
warn('TT: ENDF bad stack'); |
|
ttContext.hintsValid = false; |
|
return; |
|
} |
|
funcId = functionsCalled.pop(); |
|
data = pc.data; |
|
i = pc.i; |
|
ttContext.functionsStackDeltas[funcId] = |
|
stack.length - pc.stackTop; |
|
} |
|
} else if (op === 0x89) { // IDEF - instruction definition |
|
if (inFDEF || inELSE) { |
|
warn('TT: nested IDEFs not allowed'); |
|
tooComplexToFollowFunctions = true; |
|
} |
|
inFDEF = true; |
|
// recording it as a function to track ENDF |
|
lastDeff = i; |
|
} else if (op === 0x58) { // IF |
|
++ifLevel; |
|
} else if (op === 0x1B) { // ELSE |
|
inELSE = ifLevel; |
|
} else if (op === 0x59) { // EIF |
|
if (inELSE === ifLevel) { |
|
inELSE = 0; |
|
} |
|
--ifLevel; |
|
} else if (op === 0x1C) { // JMPR |
|
if (!inFDEF && !inELSE) { |
|
var offset = stack[stack.length - 1]; |
|
// only jumping forward to prevent infinite loop |
|
if (offset > 0) { |
|
i += offset - 1; |
|
} |
|
} |
|
} |
|
// Adjusting stack not extactly, but just enough to get function id |
|
if (!inFDEF && !inELSE) { |
|
var stackDelta = op <= 0x8E ? TTOpsStackDeltas[op] : |
|
op >= 0xC0 && op <= 0xDF ? -1 : op >= 0xE0 ? -2 : 0; |
|
if (op >= 0x71 && op <= 0x75) { |
|
n = stack.pop(); |
|
if (n === n) { |
|
stackDelta = -n * 2; |
|
} |
|
} |
|
while (stackDelta < 0 && stack.length > 0) { |
|
stack.pop(); |
|
stackDelta++; |
|
} |
|
while (stackDelta > 0) { |
|
stack.push(NaN); // pushing any number into stack |
|
stackDelta--; |
|
} |
|
} |
|
} |
|
ttContext.tooComplexToFollowFunctions = tooComplexToFollowFunctions; |
|
var content = [data]; |
|
if (i > data.length) { |
|
content.push(new Uint8Array(i - data.length)); |
|
} |
|
if (lastDeff > lastEndf) { |
|
warn('TT: complementing a missing function tail'); |
|
// new function definition started, but not finished |
|
// complete function by [CLEAR, ENDF] |
|
content.push(new Uint8Array([0x22, 0x2D])); |
|
} |
|
foldTTTable(table, content); |
|
} |
|
|
|
function checkInvalidFunctions(ttContext, maxFunctionDefs) { |
|
if (ttContext.tooComplexToFollowFunctions) { |
|
return; |
|
} |
|
if (ttContext.functionsDefined.length > maxFunctionDefs) { |
|
warn('TT: more functions defined than expected'); |
|
ttContext.hintsValid = false; |
|
return; |
|
} |
|
for (var j = 0, jj = ttContext.functionsUsed.length; j < jj; j++) { |
|
if (j > maxFunctionDefs) { |
|
warn('TT: invalid function id: ' + j); |
|
ttContext.hintsValid = false; |
|
return; |
|
} |
|
if (ttContext.functionsUsed[j] && !ttContext.functionsDefined[j]) { |
|
warn('TT: undefined function: ' + j); |
|
ttContext.hintsValid = false; |
|
return; |
|
} |
|
} |
|
} |
|
|
|
function foldTTTable(table, content) { |
|
if (content.length > 1) { |
|
// concatenating the content items |
|
var newLength = 0; |
|
var j, jj; |
|
for (j = 0, jj = content.length; j < jj; j++) { |
|
newLength += content[j].length; |
|
} |
|
newLength = (newLength + 3) & ~3; |
|
var result = new Uint8Array(newLength); |
|
var pos = 0; |
|
for (j = 0, jj = content.length; j < jj; j++) { |
|
result.set(content[j], pos); |
|
pos += content[j].length; |
|
} |
|
table.data = result; |
|
table.length = newLength; |
|
} |
|
} |
|
|
|
function sanitizeTTPrograms(fpgm, prep, cvt) { |
|
var ttContext = { |
|
functionsDefined: [], |
|
functionsUsed: [], |
|
functionsStackDeltas: [], |
|
tooComplexToFollowFunctions: false, |
|
hintsValid: true |
|
}; |
|
if (fpgm) { |
|
sanitizeTTProgram(fpgm, ttContext); |
|
} |
|
if (prep) { |
|
sanitizeTTProgram(prep, ttContext); |
|
} |
|
if (fpgm) { |
|
checkInvalidFunctions(ttContext, maxFunctionDefs); |
|
} |
|
if (cvt && (cvt.length & 1)) { |
|
var cvtData = new Uint8Array(cvt.length + 1); |
|
cvtData.set(cvt.data); |
|
cvt.data = cvtData; |
|
} |
|
return ttContext.hintsValid; |
|
} |
|
|
|
// The following steps modify the original font data, making copy |
|
font = new Stream(new Uint8Array(font.getBytes())); |
|
|
|
var VALID_TABLES = ['OS/2', 'cmap', 'head', 'hhea', 'hmtx', 'maxp', |
|
'name', 'post', 'loca', 'glyf', 'fpgm', 'prep', 'cvt ', 'CFF ']; |
|
|
|
var header = readOpenTypeHeader(font); |
|
var numTables = header.numTables; |
|
var cff, cffFile; |
|
|
|
var tables = { 'OS/2': null, cmap: null, head: null, hhea: null, |
|
hmtx: null, maxp: null, name: null, post: null }; |
|
var table; |
|
for (var i = 0; i < numTables; i++) { |
|
table = readTableEntry(font); |
|
if (VALID_TABLES.indexOf(table.tag) < 0) { |
|
continue; // skipping table if it's not a required or optional table |
|
} |
|
if (table.length === 0) { |
|
continue; // skipping empty tables |
|
} |
|
tables[table.tag] = table; |
|
} |
|
|
|
var isTrueType = !tables['CFF ']; |
|
if (!isTrueType) { |
|
// OpenType font |
|
if ((header.version === 'OTTO' && properties.type !== 'CIDFontType2') || |
|
!tables.head || !tables.hhea || !tables.maxp || !tables.post) { |
|
// no major tables: throwing everything at CFFFont |
|
cffFile = new Stream(tables['CFF '].data); |
|
cff = new CFFFont(cffFile, properties); |
|
|
|
adjustWidths(properties); |
|
|
|
return this.convert(name, cff, properties); |
|
} |
|
|
|
delete tables.glyf; |
|
delete tables.loca; |
|
delete tables.fpgm; |
|
delete tables.prep; |
|
delete tables['cvt ']; |
|
this.isOpenType = true; |
|
} else { |
|
if (!tables.loca) { |
|
error('Required "loca" table is not found'); |
|
} |
|
if (!tables.glyf) { |
|
warn('Required "glyf" table is not found -- trying to recover.'); |
|
// Note: We use `sanitizeGlyphLocations` to add dummy glyf data below. |
|
tables.glyf = { |
|
tag: 'glyf', |
|
data: new Uint8Array(0), |
|
}; |
|
} |
|
this.isOpenType = false; |
|
} |
|
|
|
if (!tables.maxp) { |
|
error('Required "maxp" table is not found'); |
|
} |
|
|
|
font.pos = (font.start || 0) + tables.maxp.offset; |
|
var version = font.getInt32(); |
|
var numGlyphs = font.getUint16(); |
|
var maxFunctionDefs = 0; |
|
if (version >= 0x00010000 && tables.maxp.length >= 22) { |
|
// maxZones can be invalid |
|
font.pos += 8; |
|
var maxZones = font.getUint16(); |
|
if (maxZones > 2) { // reset to 2 if font has invalid maxZones |
|
tables.maxp.data[14] = 0; |
|
tables.maxp.data[15] = 2; |
|
} |
|
font.pos += 4; |
|
maxFunctionDefs = font.getUint16(); |
|
} |
|
|
|
var dupFirstEntry = false; |
|
if (properties.type === 'CIDFontType2' && properties.toUnicode && |
|
properties.toUnicode.get(0) > '\u0000') { |
|
// oracle's defect (see 3427), duplicating first entry |
|
dupFirstEntry = true; |
|
numGlyphs++; |
|
tables.maxp.data[4] = numGlyphs >> 8; |
|
tables.maxp.data[5] = numGlyphs & 255; |
|
} |
|
|
|
var hintsValid = sanitizeTTPrograms(tables.fpgm, tables.prep, |
|
tables['cvt '], maxFunctionDefs); |
|
if (!hintsValid) { |
|
delete tables.fpgm; |
|
delete tables.prep; |
|
delete tables['cvt ']; |
|
} |
|
|
|
// Ensure the hmtx table contains the advance width and |
|
// sidebearings information for numGlyphs in the maxp table |
|
sanitizeMetrics(font, tables.hhea, tables.hmtx, numGlyphs); |
|
|
|
if (!tables.head) { |
|
error('Required "head" table is not found'); |
|
} |
|
|
|
sanitizeHead(tables.head, numGlyphs, isTrueType ? tables.loca.length : 0); |
|
|
|
var missingGlyphs = {}; |
|
if (isTrueType) { |
|
var isGlyphLocationsLong = int16(tables.head.data[50], |
|
tables.head.data[51]); |
|
missingGlyphs = sanitizeGlyphLocations(tables.loca, tables.glyf, |
|
numGlyphs, isGlyphLocationsLong, |
|
hintsValid, dupFirstEntry); |
|
} |
|
|
|
if (!tables.hhea) { |
|
error('Required "hhea" table is not found'); |
|
} |
|
|
|
// Sanitizer reduces the glyph advanceWidth to the maxAdvanceWidth |
|
// Sometimes it's 0. That needs to be fixed |
|
if (tables.hhea.data[10] === 0 && tables.hhea.data[11] === 0) { |
|
tables.hhea.data[10] = 0xFF; |
|
tables.hhea.data[11] = 0xFF; |
|
} |
|
|
|
// Extract some more font properties from the OpenType head and |
|
// hhea tables; yMin and descent value are always negative. |
|
var metricsOverride = { |
|
unitsPerEm: int16(tables.head.data[18], tables.head.data[19]), |
|
yMax: int16(tables.head.data[42], tables.head.data[43]), |
|
yMin: int16(tables.head.data[38], tables.head.data[39]) - 0x10000, |
|
ascent: int16(tables.hhea.data[4], tables.hhea.data[5]), |
|
descent: int16(tables.hhea.data[6], tables.hhea.data[7]) - 0x10000 |
|
}; |
|
|
|
// PDF FontDescriptor metrics lie -- using data from actual font. |
|
this.ascent = metricsOverride.ascent / metricsOverride.unitsPerEm; |
|
this.descent = metricsOverride.descent / metricsOverride.unitsPerEm; |
|
|
|
// The 'post' table has glyphs names. |
|
if (tables.post) { |
|
var valid = readPostScriptTable(tables.post, properties, numGlyphs); |
|
if (!valid) { |
|
tables.post = null; |
|
} |
|
} |
|
|
|
var charCodeToGlyphId = [], charCode; |
|
var toUnicode = properties.toUnicode, widths = properties.widths; |
|
var skipToUnicode = (toUnicode instanceof IdentityToUnicodeMap || |
|
toUnicode.length === 0x10000); |
|
|
|
// Helper function to try to skip mapping of empty glyphs. |
|
// Note: In some cases, just relying on the glyph data doesn't work, |
|
// hence we also use a few heuristics to fix various PDF files. |
|
function hasGlyph(glyphId, charCode, widthCode) { |
|
if (!missingGlyphs[glyphId]) { |
|
return true; |
|
} |
|
if (!skipToUnicode && charCode >= 0 && toUnicode.has(charCode)) { |
|
return true; |
|
} |
|
if (widths && widthCode >= 0 && isNum(widths[widthCode])) { |
|
return true; |
|
} |
|
return false; |
|
} |
|
|
|
if (properties.type === 'CIDFontType2') { |
|
var cidToGidMap = properties.cidToGidMap || []; |
|
var isCidToGidMapEmpty = cidToGidMap.length === 0; |
|
|
|
properties.cMap.forEach(function(charCode, cid) { |
|
assert(cid <= 0xffff, 'Max size of CID is 65,535'); |
|
var glyphId = -1; |
|
if (isCidToGidMapEmpty) { |
|
glyphId = charCode; |
|
} else if (cidToGidMap[cid] !== undefined) { |
|
glyphId = cidToGidMap[cid]; |
|
} |
|
|
|
if (glyphId >= 0 && glyphId < numGlyphs && |
|
hasGlyph(glyphId, charCode, cid)) { |
|
charCodeToGlyphId[charCode] = glyphId; |
|
} |
|
}); |
|
if (dupFirstEntry) { |
|
charCodeToGlyphId[0] = numGlyphs - 1; |
|
} |
|
} else { |
|
// Most of the following logic in this code branch is based on the |
|
// 9.6.6.4 of the PDF spec. |
|
var hasEncoding = |
|
properties.differences.length > 0 || !!properties.baseEncodingName; |
|
var cmapTable = |
|
readCmapTable(tables.cmap, font, this.isSymbolicFont, hasEncoding); |
|
var cmapPlatformId = cmapTable.platformId; |
|
var cmapEncodingId = cmapTable.encodingId; |
|
var cmapMappings = cmapTable.mappings; |
|
var cmapMappingsLength = cmapMappings.length; |
|
|
|
// The spec seems to imply that if the font is symbolic the encoding |
|
// should be ignored, this doesn't appear to work for 'preistabelle.pdf' |
|
// where the the font is symbolic and it has an encoding. |
|
if (hasEncoding && |
|
(cmapPlatformId === 3 && cmapEncodingId === 1 || |
|
cmapPlatformId === 1 && cmapEncodingId === 0) || |
|
(cmapPlatformId === -1 && cmapEncodingId === -1 && // Temporary hack |
|
!!Encodings[properties.baseEncodingName])) { // Temporary hack |
|
// When no preferred cmap table was found and |baseEncodingName| is |
|
// one of the predefined encodings, we seem to obtain a better |
|
// |charCodeToGlyphId| map from the code below (fixes bug 1057544). |
|
// TODO: Note that this is a hack which should be removed as soon as |
|
// we have proper support for more exotic cmap tables. |
|
|
|
var baseEncoding = []; |
|
if (properties.baseEncodingName === 'MacRomanEncoding' || |
|
properties.baseEncodingName === 'WinAnsiEncoding') { |
|
baseEncoding = Encodings[properties.baseEncodingName]; |
|
} |
|
for (charCode = 0; charCode < 256; charCode++) { |
|
var glyphName; |
|
if (this.differences && charCode in this.differences) { |
|
glyphName = this.differences[charCode]; |
|
} else if (charCode in baseEncoding && |
|
baseEncoding[charCode] !== '') { |
|
glyphName = baseEncoding[charCode]; |
|
} else { |
|
glyphName = Encodings.StandardEncoding[charCode]; |
|
} |
|
if (!glyphName) { |
|
continue; |
|
} |
|
var unicodeOrCharCode, isUnicode = false; |
|
if (cmapPlatformId === 3 && cmapEncodingId === 1) { |
|
unicodeOrCharCode = GlyphsUnicode[glyphName]; |
|
isUnicode = true; |
|
} else if (cmapPlatformId === 1 && cmapEncodingId === 0) { |
|
// TODO: the encoding needs to be updated with mac os table. |
|
unicodeOrCharCode = Encodings.MacRomanEncoding.indexOf(glyphName); |
|
} |
|
|
|
var found = false; |
|
for (i = 0; i < cmapMappingsLength; ++i) { |
|
if (cmapMappings[i].charCode !== unicodeOrCharCode) { |
|
continue; |
|
} |
|
var code = isUnicode ? charCode : unicodeOrCharCode; |
|
if (hasGlyph(cmapMappings[i].glyphId, code, -1)) { |
|
charCodeToGlyphId[charCode] = cmapMappings[i].glyphId; |
|
found = true; |
|
break; |
|
} |
|
} |
|
if (!found && properties.glyphNames) { |
|
// Try to map using the post table. |
|
var glyphId = properties.glyphNames.indexOf(glyphName); |
|
if (glyphId > 0 && hasGlyph(glyphId, -1, -1)) { |
|
charCodeToGlyphId[charCode] = glyphId; |
|
found = true; |
|
} |
|
} |
|
if (!found) { |
|
charCodeToGlyphId[charCode] = 0; // notdef |
|
} |
|
} |
|
} else if (cmapPlatformId === 0 && cmapEncodingId === 0) { |
|
// Default Unicode semantics, use the charcodes as is. |
|
for (i = 0; i < cmapMappingsLength; ++i) { |
|
charCodeToGlyphId[cmapMappings[i].charCode] = |
|
cmapMappings[i].glyphId; |
|
} |
|
} else { |
|
// For (3, 0) cmap tables: |
|
// The charcode key being stored in charCodeToGlyphId is the lower |
|
// byte of the two-byte charcodes of the cmap table since according to |
|
// the spec: 'each byte from the string shall be prepended with the |
|
// high byte of the range [of charcodes in the cmap table], to form |
|
// a two-byte character, which shall be used to select the |
|
// associated glyph description from the subtable'. |
|
// |
|
// For (1, 0) cmap tables: |
|
// 'single bytes from the string shall be used to look up the |
|
// associated glyph descriptions from the subtable'. This means |
|
// charcodes in the cmap will be single bytes, so no-op since |
|
// glyph.charCode & 0xFF === glyph.charCode |
|
for (i = 0; i < cmapMappingsLength; ++i) { |
|
charCode = cmapMappings[i].charCode & 0xFF; |
|
charCodeToGlyphId[charCode] = cmapMappings[i].glyphId; |
|
} |
|
} |
|
} |
|
|
|
if (charCodeToGlyphId.length === 0) { |
|
// defines at least one glyph |
|
charCodeToGlyphId[0] = 0; |
|
} |
|
|
|
// Converting glyphs and ids into font's cmap table |
|
var newMapping = adjustMapping(charCodeToGlyphId, properties); |
|
this.toFontChar = newMapping.toFontChar; |
|
tables.cmap = { |
|
tag: 'cmap', |
|
data: createCmapTable(newMapping.charCodeToGlyphId, numGlyphs) |
|
}; |
|
|
|
if (!tables['OS/2'] || !validateOS2Table(tables['OS/2'])) { |
|
tables['OS/2'] = { |
|
tag: 'OS/2', |
|
data: createOS2Table(properties, newMapping.charCodeToGlyphId, |
|
metricsOverride) |
|
}; |
|
} |
|
|
|
// Rewrite the 'post' table if needed |
|
if (!tables.post) { |
|
tables.post = { |
|
tag: 'post', |
|
data: createPostTable(properties) |
|
}; |
|
} |
|
|
|
if (!isTrueType) { |
|
try { |
|
// Trying to repair CFF file |
|
cffFile = new Stream(tables['CFF '].data); |
|
var parser = new CFFParser(cffFile, properties); |
|
cff = parser.parse(); |
|
var compiler = new CFFCompiler(cff); |
|
tables['CFF '].data = compiler.compile(); |
|
} catch (e) { |
|
warn('Failed to compile font ' + properties.loadedName); |
|
} |
|
} |
|
|
|
// Re-creating 'name' table |
|
if (!tables.name) { |
|
tables.name = { |
|
tag: 'name', |
|
data: createNameTable(this.name) |
|
}; |
|
} else { |
|
// ... using existing 'name' table as prototype |
|
var namePrototype = readNameTable(tables.name); |
|
tables.name.data = createNameTable(name, namePrototype); |
|
} |
|
|
|
var builder = new OpenTypeFileBuilder(header.version); |
|
for (var tableTag in tables) { |
|
builder.addTable(tableTag, tables[tableTag].data); |
|
} |
|
return builder.toArray(); |
|
}, |
|
|
|
convert: function Font_convert(fontName, font, properties) { |
|
// TODO: Check the charstring widths to determine this. |
|
properties.fixedPitch = false; |
|
|
|
var mapping = font.getGlyphMapping(properties); |
|
var newMapping = adjustMapping(mapping, properties); |
|
this.toFontChar = newMapping.toFontChar; |
|
var numGlyphs = font.numGlyphs; |
|
|
|
function getCharCodes(charCodeToGlyphId, glyphId) { |
|
var charCodes = null; |
|
for (var charCode in charCodeToGlyphId) { |
|
if (glyphId === charCodeToGlyphId[charCode]) { |
|
if (!charCodes) { |
|
charCodes = []; |
|
} |
|
charCodes.push(charCode | 0); |
|
} |
|
} |
|
return charCodes; |
|
} |
|
|
|
function createCharCode(charCodeToGlyphId, glyphId) { |
|
for (var charCode in charCodeToGlyphId) { |
|
if (glyphId === charCodeToGlyphId[charCode]) { |
|
return charCode | 0; |
|
} |
|
} |
|
newMapping.charCodeToGlyphId[newMapping.nextAvailableFontCharCode] = |
|
glyphId; |
|
return newMapping.nextAvailableFontCharCode++; |
|
} |
|
|
|
var seacs = font.seacs; |
|
if (SEAC_ANALYSIS_ENABLED && seacs && seacs.length) { |
|
var matrix = properties.fontMatrix || FONT_IDENTITY_MATRIX; |
|
var charset = font.getCharset(); |
|
var seacMap = Object.create(null); |
|
for (var glyphId in seacs) { |
|
glyphId |= 0; |
|
var seac = seacs[glyphId]; |
|
var baseGlyphName = Encodings.StandardEncoding[seac[2]]; |
|
var accentGlyphName = Encodings.StandardEncoding[seac[3]]; |
|
var baseGlyphId = charset.indexOf(baseGlyphName); |
|
var accentGlyphId = charset.indexOf(accentGlyphName); |
|
if (baseGlyphId < 0 || accentGlyphId < 0) { |
|
continue; |
|
} |
|
var accentOffset = { |
|
x: seac[0] * matrix[0] + seac[1] * matrix[2] + matrix[4], |
|
y: seac[0] * matrix[1] + seac[1] * matrix[3] + matrix[5] |
|
}; |
|
|
|
var charCodes = getCharCodes(mapping, glyphId); |
|
if (!charCodes) { |
|
// There's no point in mapping it if the char code was never mapped |
|
// to begin with. |
|
continue; |
|
} |
|
for (var i = 0, ii = charCodes.length; i < ii; i++) { |
|
var charCode = charCodes[i]; |
|
// Find a fontCharCode that maps to the base and accent glyphs. |
|
// If one doesn't exists, create it. |
|
var charCodeToGlyphId = newMapping.charCodeToGlyphId; |
|
var baseFontCharCode = createCharCode(charCodeToGlyphId, |
|
baseGlyphId); |
|
var accentFontCharCode = createCharCode(charCodeToGlyphId, |
|
accentGlyphId); |
|
seacMap[charCode] = { |
|
baseFontCharCode: baseFontCharCode, |
|
accentFontCharCode: accentFontCharCode, |
|
accentOffset: accentOffset |
|
}; |
|
} |
|
} |
|
properties.seacMap = seacMap; |
|
} |
|
|
|
var unitsPerEm = 1 / (properties.fontMatrix || FONT_IDENTITY_MATRIX)[0]; |
|
|
|
var builder = new OpenTypeFileBuilder('\x4F\x54\x54\x4F'); |
|
// PostScript Font Program |
|
builder.addTable('CFF ', font.data); |
|
// OS/2 and Windows Specific metrics |
|
builder.addTable('OS/2', createOS2Table(properties, |
|
newMapping.charCodeToGlyphId)); |
|
// Character to glyphs mapping |
|
builder.addTable('cmap', createCmapTable(newMapping.charCodeToGlyphId, |
|
numGlyphs)); |
|
// Font header |
|
builder.addTable('head', |
|
'\x00\x01\x00\x00' + // Version number |
|
'\x00\x00\x10\x00' + // fontRevision |
|
'\x00\x00\x00\x00' + // checksumAdjustement |
|
'\x5F\x0F\x3C\xF5' + // magicNumber |
|
'\x00\x00' + // Flags |
|
safeString16(unitsPerEm) + // unitsPerEM |
|
'\x00\x00\x00\x00\x9e\x0b\x7e\x27' + // creation date |
|
'\x00\x00\x00\x00\x9e\x0b\x7e\x27' + // modifification date |
|
'\x00\x00' + // xMin |
|
safeString16(properties.descent) + // yMin |
|
'\x0F\xFF' + // xMax |
|
safeString16(properties.ascent) + // yMax |
|
string16(properties.italicAngle ? 2 : 0) + // macStyle |
|
'\x00\x11' + // lowestRecPPEM |
|
'\x00\x00' + // fontDirectionHint |
|
'\x00\x00' + // indexToLocFormat |
|
'\x00\x00'); // glyphDataFormat |
|
|
|
// Horizontal header |
|
builder.addTable('hhea', |
|
'\x00\x01\x00\x00' + // Version number |
|
safeString16(properties.ascent) + // Typographic Ascent |
|
safeString16(properties.descent) + // Typographic Descent |
|
'\x00\x00' + // Line Gap |
|
'\xFF\xFF' + // advanceWidthMax |
|
'\x00\x00' + // minLeftSidebearing |
|
'\x00\x00' + // minRightSidebearing |
|
'\x00\x00' + // xMaxExtent |
|
safeString16(properties.capHeight) + // caretSlopeRise |
|
safeString16(Math.tan(properties.italicAngle) * |
|
properties.xHeight) + // caretSlopeRun |
|
'\x00\x00' + // caretOffset |
|
'\x00\x00' + // -reserved- |
|
'\x00\x00' + // -reserved- |
|
'\x00\x00' + // -reserved- |
|
'\x00\x00' + // -reserved- |
|
'\x00\x00' + // metricDataFormat |
|
string16(numGlyphs)); // Number of HMetrics |
|
|
|
// Horizontal metrics |
|
builder.addTable('hmtx', (function fontFieldsHmtx() { |
|
var charstrings = font.charstrings; |
|
var cffWidths = font.cff ? font.cff.widths : null; |
|
var hmtx = '\x00\x00\x00\x00'; // Fake .notdef |
|
for (var i = 1, ii = numGlyphs; i < ii; i++) { |
|
var width = 0; |
|
if (charstrings) { |
|
var charstring = charstrings[i - 1]; |
|
width = 'width' in charstring ? charstring.width : 0; |
|
} else if (cffWidths) { |
|
width = Math.ceil(cffWidths[i] || 0); |
|
} |
|
hmtx += string16(width) + string16(0); |
|
} |
|
return hmtx; |
|
})()); |
|
|
|
// Maximum profile |
|
builder.addTable('maxp', |
|
'\x00\x00\x50\x00' + // Version number |
|
string16(numGlyphs)); // Num of glyphs |
|
|
|
// Naming tables |
|
builder.addTable('name', createNameTable(fontName)); |
|
|
|
// PostScript informations |
|
builder.addTable('post', createPostTable(properties)); |
|
|
|
return builder.toArray(); |
|
}, |
|
|
|
/** |
|
* Builds a char code to unicode map based on section 9.10 of the spec. |
|
* @param {Object} properties Font properties object. |
|
* @return {Object} A ToUnicodeMap object. |
|
*/ |
|
buildToUnicode: function Font_buildToUnicode(properties) { |
|
// Section 9.10.2 Mapping Character Codes to Unicode Values |
|
if (properties.toUnicode && properties.toUnicode.length !== 0) { |
|
return properties.toUnicode; |
|
} |
|
// According to the spec if the font is a simple font we should only map |
|
// to unicode if the base encoding is MacRoman, MacExpert, or WinAnsi or |
|
// the differences array only contains adobe standard or symbol set names, |
|
// in pratice it seems better to always try to create a toUnicode |
|
// map based of the default encoding. |
|
var toUnicode, charcode; |
|
if (!properties.composite /* is simple font */) { |
|
toUnicode = []; |
|
var encoding = properties.defaultEncoding.slice(); |
|
var baseEncodingName = properties.baseEncodingName; |
|
// Merge in the differences array. |
|
var differences = properties.differences; |
|
for (charcode in differences) { |
|
encoding[charcode] = differences[charcode]; |
|
} |
|
for (charcode in encoding) { |
|
// a) Map the character code to a character name. |
|
var glyphName = encoding[charcode]; |
|
// b) Look up the character name in the Adobe Glyph List (see the |
|
// Bibliography) to obtain the corresponding Unicode value. |
|
if (glyphName === '') { |
|
continue; |
|
} else if (GlyphsUnicode[glyphName] === undefined) { |
|
// (undocumented) c) Few heuristics to recognize unknown glyphs |
|
// NOTE: Adobe Reader does not do this step, but OSX Preview does |
|
var code = 0; |
|
switch (glyphName[0]) { |
|
case 'G': // Gxx glyph |
|
if (glyphName.length === 3) { |
|
code = parseInt(glyphName.substr(1), 16); |
|
} |
|
break; |
|
case 'g': // g00xx glyph |
|
if (glyphName.length === 5) { |
|
code = parseInt(glyphName.substr(1), 16); |
|
} |
|
break; |
|
case 'C': // Cddd glyph |
|
case 'c': // cddd glyph |
|
if (glyphName.length >= 3) { |
|
code = +glyphName.substr(1); |
|
} |
|
break; |
|
} |
|
if (code) { |
|
// If |baseEncodingName| is one the predefined encodings, |
|
// and |code| equals |charcode|, using the glyph defined in the |
|
// baseEncoding seems to yield a better |toUnicode| mapping |
|
// (fixes issue 5070). |
|
if (baseEncodingName && code === +charcode) { |
|
var baseEncoding = Encodings[baseEncodingName]; |
|
if (baseEncoding && (glyphName = baseEncoding[charcode])) { |
|
toUnicode[charcode] = |
|
String.fromCharCode(GlyphsUnicode[glyphName]); |
|
continue; |
|
} |
|
} |
|
toUnicode[charcode] = String.fromCharCode(code); |
|
} |
|
continue; |
|
} |
|
toUnicode[charcode] = String.fromCharCode(GlyphsUnicode[glyphName]); |
|
} |
|
return new ToUnicodeMap(toUnicode); |
|
} |
|
// If the font is a composite font that uses one of the predefined CMaps |
|
// listed in Table 118 (except Identity–H and Identity–V) or whose |
|
// descendant CIDFont uses the Adobe-GB1, Adobe-CNS1, Adobe-Japan1, or |
|
// Adobe-Korea1 character collection: |
|
if (properties.composite && ( |
|
(properties.cMap.builtInCMap && |
|
!(properties.cMap instanceof IdentityCMap)) || |
|
(properties.cidSystemInfo.registry === 'Adobe' && |
|
(properties.cidSystemInfo.ordering === 'GB1' || |
|
properties.cidSystemInfo.ordering === 'CNS1' || |
|
properties.cidSystemInfo.ordering === 'Japan1' || |
|
properties.cidSystemInfo.ordering === 'Korea1')))) { |
|
// Then: |
|
// a) Map the character code to a character identifier (CID) according |
|
// to the font’s CMap. |
|
// b) Obtain the registry and ordering of the character collection used |
|
// by the font’s CMap (for example, Adobe and Japan1) from its |
|
// CIDSystemInfo dictionary. |
|
var registry = properties.cidSystemInfo.registry; |
|
var ordering = properties.cidSystemInfo.ordering; |
|
// c) Construct a second CMap name by concatenating the registry and |
|
// ordering obtained in step (b) in the format registry–ordering–UCS2 |
|
// (for example, Adobe–Japan1–UCS2). |
|
var ucs2CMapName = new Name(registry + '-' + ordering + '-UCS2'); |
|
// d) Obtain the CMap with the name constructed in step (c) (available |
|
// from the ASN Web site; see the Bibliography). |
|
var ucs2CMap = CMapFactory.create(ucs2CMapName, |
|
{ url: PDFJS.cMapUrl, packed: PDFJS.cMapPacked }, null); |
|
var cMap = properties.cMap; |
|
toUnicode = []; |
|
cMap.forEach(function(charcode, cid) { |
|
assert(cid <= 0xffff, 'Max size of CID is 65,535'); |
|
// e) Map the CID obtained in step (a) according to the CMap obtained |
|
// in step (d), producing a Unicode value. |
|
var ucs2 = ucs2CMap.lookup(cid); |
|
if (ucs2) { |
|
toUnicode[charcode] = |
|
String.fromCharCode((ucs2.charCodeAt(0) << 8) + |
|
ucs2.charCodeAt(1)); |
|
} |
|
}); |
|
return new ToUnicodeMap(toUnicode); |
|
} |
|
|
|
// The viewer's choice, just use an identity map. |
|
return new IdentityToUnicodeMap(properties.firstChar, |
|
properties.lastChar); |
|
}, |
|
|
|
get spaceWidth() { |
|
if ('_shadowWidth' in this) { |
|
return this._shadowWidth; |
|
} |
|
|
|
// trying to estimate space character width |
|
var possibleSpaceReplacements = ['space', 'minus', 'one', 'i']; |
|
var width; |
|
for (var i = 0, ii = possibleSpaceReplacements.length; i < ii; i++) { |
|
var glyphName = possibleSpaceReplacements[i]; |
|
// if possible, getting width by glyph name |
|
if (glyphName in this.widths) { |
|
width = this.widths[glyphName]; |
|
break; |
|
} |
|
var glyphUnicode = GlyphsUnicode[glyphName]; |
|
// finding the charcode via unicodeToCID map |
|
var charcode = 0; |
|
if (this.composite) { |
|
if (this.cMap.contains(glyphUnicode)) { |
|
charcode = this.cMap.lookup(glyphUnicode); |
|
} |
|
} |
|
// ... via toUnicode map |
|
if (!charcode && this.toUnicode) { |
|
charcode = this.toUnicode.charCodeOf(glyphUnicode); |
|
} |
|
// setting it to unicode if negative or undefined |
|
if (charcode <= 0) { |
|
charcode = glyphUnicode; |
|
} |
|
// trying to get width via charcode |
|
width = this.widths[charcode]; |
|
if (width) { |
|
break; // the non-zero width found |
|
} |
|
} |
|
width = width || this.defaultWidth; |
|
// Do not shadow the property here. See discussion: |
|
// https://github.com/mozilla/pdf.js/pull/2127#discussion_r1662280 |
|
this._shadowWidth = width; |
|
return width; |
|
}, |
|
|
|
charToGlyph: function Font_charToGlyph(charcode, isSpace) { |
|
var fontCharCode, width, operatorListId; |
|
|
|
var widthCode = charcode; |
|
if (this.cMap && this.cMap.contains(charcode)) { |
|
widthCode = this.cMap.lookup(charcode); |
|
} |
|
width = this.widths[widthCode]; |
|
width = isNum(width) ? width : this.defaultWidth; |
|
var vmetric = this.vmetrics && this.vmetrics[widthCode]; |
|
|
|
var unicode = this.toUnicode.get(charcode) || charcode; |
|
if (typeof unicode === 'number') { |
|
unicode = String.fromCharCode(unicode); |
|
} |
|
|
|
// First try the toFontChar map, if it's not there then try falling |
|
// back to the char code. |
|
fontCharCode = this.toFontChar[charcode] || charcode; |
|
if (this.missingFile) { |
|
fontCharCode = mapSpecialUnicodeValues(fontCharCode); |
|
} |
|
|
|
if (this.isType3Font) { |
|
// Font char code in this case is actually a glyph name. |
|
operatorListId = fontCharCode; |
|
} |
|
|
|
var accent = null; |
|
if (this.seacMap && this.seacMap[charcode]) { |
|
var seac = this.seacMap[charcode]; |
|
fontCharCode = seac.baseFontCharCode; |
|
accent = { |
|
fontChar: String.fromCharCode(seac.accentFontCharCode), |
|
offset: seac.accentOffset |
|
}; |
|
} |
|
|
|
var fontChar = String.fromCharCode(fontCharCode); |
|
|
|
var glyph = this.glyphCache[charcode]; |
|
if (!glyph || |
|
!glyph.matchesForCache(fontChar, unicode, accent, width, vmetric, |
|
operatorListId, isSpace)) { |
|
glyph = new Glyph(fontChar, unicode, accent, width, vmetric, |
|
operatorListId, isSpace); |
|
this.glyphCache[charcode] = glyph; |
|
} |
|
return glyph; |
|
}, |
|
|
|
charsToGlyphs: function Font_charsToGlyphs(chars) { |
|
var charsCache = this.charsCache; |
|
var glyphs, glyph, charcode; |
|
|
|
// if we translated this string before, just grab it from the cache |
|
if (charsCache) { |
|
glyphs = charsCache[chars]; |
|
if (glyphs) { |
|
return glyphs; |
|
} |
|
} |
|
|
|
// lazily create the translation cache |
|
if (!charsCache) { |
|
charsCache = this.charsCache = Object.create(null); |
|
} |
|
|
|
glyphs = []; |
|
var charsCacheKey = chars; |
|
var i = 0, ii; |
|
|
|
if (this.cMap) { |
|
// composite fonts have multi-byte strings convert the string from |
|
// single-byte to multi-byte |
|
var c = {}; |
|
while (i < chars.length) { |
|
this.cMap.readCharCode(chars, i, c); |
|
charcode = c.charcode; |
|
var length = c.length; |
|
i += length; |
|
// Space is char with code 0x20 and length 1 in multiple-byte codes. |
|
var isSpace = length === 1 && chars.charCodeAt(i - 1) === 0x20; |
|
glyph = this.charToGlyph(charcode, isSpace); |
|
glyphs.push(glyph); |
|
} |
|
} else { |
|
for (i = 0, ii = chars.length; i < ii; ++i) { |
|
charcode = chars.charCodeAt(i); |
|
glyph = this.charToGlyph(charcode, charcode === 0x20); |
|
glyphs.push(glyph); |
|
} |
|
} |
|
|
|
// Enter the translated string into the cache |
|
return (charsCache[charsCacheKey] = glyphs); |
|
} |
|
}; |
|
|
|
return Font; |
|
})(); |
|
|
|
var ErrorFont = (function ErrorFontClosure() { |
|
function ErrorFont(error) { |
|
this.error = error; |
|
this.loadedName = 'g_font_error'; |
|
this.loading = false; |
|
} |
|
|
|
ErrorFont.prototype = { |
|
charsToGlyphs: function ErrorFont_charsToGlyphs() { |
|
return []; |
|
}, |
|
exportData: function ErrorFont_exportData() { |
|
return {error: this.error}; |
|
} |
|
}; |
|
|
|
return ErrorFont; |
|
})(); |
|
|
|
/** |
|
* Shared logic for building a char code to glyph id mapping for Type1 and |
|
* simple CFF fonts. See section 9.6.6.2 of the spec. |
|
* @param {Object} properties Font properties object. |
|
* @param {Object} builtInEncoding The encoding contained within the actual font |
|
* data. |
|
* @param {Array} Array of glyph names where the index is the glyph ID. |
|
* @returns {Object} A char code to glyph ID map. |
|
*/ |
|
function type1FontGlyphMapping(properties, builtInEncoding, glyphNames) { |
|
var charCodeToGlyphId = Object.create(null); |
|
var glyphId, charCode, baseEncoding; |
|
|
|
if (properties.baseEncodingName) { |
|
// If a valid base encoding name was used, the mapping is initialized with |
|
// that. |
|
baseEncoding = Encodings[properties.baseEncodingName]; |
|
for (charCode = 0; charCode < baseEncoding.length; charCode++) { |
|
glyphId = glyphNames.indexOf(baseEncoding[charCode]); |
|
if (glyphId >= 0) { |
|
charCodeToGlyphId[charCode] = glyphId; |
|
} else { |
|
charCodeToGlyphId[charCode] = 0; // notdef |
|
} |
|
} |
|
} else if (!!(properties.flags & FontFlags.Symbolic)) { |
|
// For a symbolic font the encoding should be the fonts built-in |
|
// encoding. |
|
for (charCode in builtInEncoding) { |
|
charCodeToGlyphId[charCode] = builtInEncoding[charCode]; |
|
} |
|
} else { |
|
// For non-symbolic fonts that don't have a base encoding the standard |
|
// encoding should be used. |
|
baseEncoding = Encodings.StandardEncoding; |
|
for (charCode = 0; charCode < baseEncoding.length; charCode++) { |
|
glyphId = glyphNames.indexOf(baseEncoding[charCode]); |
|
if (glyphId >= 0) { |
|
charCodeToGlyphId[charCode] = glyphId; |
|
} else { |
|
charCodeToGlyphId[charCode] = 0; // notdef |
|
} |
|
} |
|
} |
|
|
|
// Lastly, merge in the differences. |
|
var differences = properties.differences; |
|
if (differences) { |
|
for (charCode in differences) { |
|
var glyphName = differences[charCode]; |
|
glyphId = glyphNames.indexOf(glyphName); |
|
if (glyphId >= 0) { |
|
charCodeToGlyphId[charCode] = glyphId; |
|
} else { |
|
charCodeToGlyphId[charCode] = 0; // notdef |
|
} |
|
} |
|
} |
|
return charCodeToGlyphId; |
|
} |
|
|
|
/* |
|
* CharStrings are encoded following the the CharString Encoding sequence |
|
* describe in Chapter 6 of the "Adobe Type1 Font Format" specification. |
|
* The value in a byte indicates a command, a number, or subsequent bytes |
|
* that are to be interpreted in a special way. |
|
* |
|
* CharString Number Encoding: |
|
* A CharString byte containing the values from 32 through 255 inclusive |
|
* indicate an integer. These values are decoded in four ranges. |
|
* |
|
* 1. A CharString byte containing a value, v, between 32 and 246 inclusive, |
|
* indicate the integer v - 139. Thus, the integer values from -107 through |
|
* 107 inclusive may be encoded in single byte. |
|
* |
|
* 2. A CharString byte containing a value, v, between 247 and 250 inclusive, |
|
* indicates an integer involving the next byte, w, according to the formula: |
|
* [(v - 247) x 256] + w + 108 |
|
* |
|
* 3. A CharString byte containing a value, v, between 251 and 254 inclusive, |
|
* indicates an integer involving the next byte, w, according to the formula: |
|
* -[(v - 251) * 256] - w - 108 |
|
* |
|
* 4. A CharString containing the value 255 indicates that the next 4 bytes |
|
* are a two complement signed integer. The first of these bytes contains the |
|
* highest order bits, the second byte contains the next higher order bits |
|
* and the fourth byte contain the lowest order bits. |
|
* |
|
* |
|
* CharString Command Encoding: |
|
* CharStrings commands are encoded in 1 or 2 bytes. |
|
* |
|
* Single byte commands are encoded in 1 byte that contains a value between |
|
* 0 and 31 inclusive. |
|
* If a command byte contains the value 12, then the value in the next byte |
|
* indicates a command. This "escape" mechanism allows many extra commands |
|
* to be encoded and this encoding technique helps to minimize the length of |
|
* the charStrings. |
|
*/ |
|
var Type1CharString = (function Type1CharStringClosure() { |
|
var COMMAND_MAP = { |
|
'hstem': [1], |
|
'vstem': [3], |
|
'vmoveto': [4], |
|
'rlineto': [5], |
|
'hlineto': [6], |
|
'vlineto': [7], |
|
'rrcurveto': [8], |
|
'callsubr': [10], |
|
'flex': [12, 35], |
|
'drop' : [12, 18], |
|
'endchar': [14], |
|
'rmoveto': [21], |
|
'hmoveto': [22], |
|
'vhcurveto': [30], |
|
'hvcurveto': [31] |
|
}; |
|
|
|
function Type1CharString() { |
|
this.width = 0; |
|
this.lsb = 0; |
|
this.flexing = false; |
|
this.output = []; |
|
this.stack = []; |
|
} |
|
|
|
Type1CharString.prototype = { |
|
convert: function Type1CharString_convert(encoded, subrs) { |
|
var count = encoded.length; |
|
var error = false; |
|
var wx, sbx, subrNumber; |
|
for (var i = 0; i < count; i++) { |
|
var value = encoded[i]; |
|
if (value < 32) { |
|
if (value === 12) { |
|
value = (value << 8) + encoded[++i]; |
|
} |
|
switch (value) { |
|
case 1: // hstem |
|
if (!HINTING_ENABLED) { |
|
this.stack = []; |
|
break; |
|
} |
|
error = this.executeCommand(2, COMMAND_MAP.hstem); |
|
break; |
|
case 3: // vstem |
|
if (!HINTING_ENABLED) { |
|
this.stack = []; |
|
break; |
|
} |
|
error = this.executeCommand(2, COMMAND_MAP.vstem); |
|
break; |
|
case 4: // vmoveto |
|
if (this.flexing) { |
|
if (this.stack.length < 1) { |
|
error = true; |
|
break; |
|
} |
|
// Add the dx for flex and but also swap the values so they are |
|
// the right order. |
|
var dy = this.stack.pop(); |
|
this.stack.push(0, dy); |
|
break; |
|
} |
|
error = this.executeCommand(1, COMMAND_MAP.vmoveto); |
|
break; |
|
case 5: // rlineto |
|
error = this.executeCommand(2, COMMAND_MAP.rlineto); |
|
break; |
|
case 6: // hlineto |
|
error = this.executeCommand(1, COMMAND_MAP.hlineto); |
|
break; |
|
case 7: // vlineto |
|
error = this.executeCommand(1, COMMAND_MAP.vlineto); |
|
break; |
|
case 8: // rrcurveto |
|
error = this.executeCommand(6, COMMAND_MAP.rrcurveto); |
|
break; |
|
case 9: // closepath |
|
// closepath is a Type1 command that does not take argument and is |
|
// useless in Type2 and it can simply be ignored. |
|
this.stack = []; |
|
break; |
|
case 10: // callsubr |
|
if (this.stack.length < 1) { |
|
error = true; |
|
break; |
|
} |
|
subrNumber = this.stack.pop(); |
|
error = this.convert(subrs[subrNumber], subrs); |
|
break; |
|
case 11: // return |
|
return error; |
|
case 13: // hsbw |
|
if (this.stack.length < 2) { |
|
error = true; |
|
break; |
|
} |
|
// To convert to type2 we have to move the width value to the |
|
// first part of the charstring and then use hmoveto with lsb. |
|
wx = this.stack.pop(); |
|
sbx = this.stack.pop(); |
|
this.lsb = sbx; |
|
this.width = wx; |
|
this.stack.push(wx, sbx); |
|
error = this.executeCommand(2, COMMAND_MAP.hmoveto); |
|
break; |
|
case 14: // endchar |
|
this.output.push(COMMAND_MAP.endchar[0]); |
|
break; |
|
case 21: // rmoveto |
|
if (this.flexing) { |
|
break; |
|
} |
|
error = this.executeCommand(2, COMMAND_MAP.rmoveto); |
|
break; |
|
case 22: // hmoveto |
|
if (this.flexing) { |
|
// Add the dy for flex. |
|
this.stack.push(0); |
|
break; |
|
} |
|
error = this.executeCommand(1, COMMAND_MAP.hmoveto); |
|
break; |
|
case 30: // vhcurveto |
|
error = this.executeCommand(4, COMMAND_MAP.vhcurveto); |
|
break; |
|
case 31: // hvcurveto |
|
error = this.executeCommand(4, COMMAND_MAP.hvcurveto); |
|
break; |
|
case (12 << 8) + 0: // dotsection |
|
// dotsection is a Type1 command to specify some hinting feature |
|
// for dots that do not take a parameter and it can safely be |
|
// ignored for Type2. |
|
this.stack = []; |
|
break; |
|
case (12 << 8) + 1: // vstem3 |
|
if (!HINTING_ENABLED) { |
|
this.stack = []; |
|
break; |
|
} |
|
// [vh]stem3 are Type1 only and Type2 supports [vh]stem with |
|
// multiple parameters, so instead of returning [vh]stem3 take a |
|
// shortcut and return [vhstem] instead. |
|
error = this.executeCommand(2, COMMAND_MAP.vstem); |
|
break; |
|
case (12 << 8) + 2: // hstem3 |
|
if (!HINTING_ENABLED) { |
|
this.stack = []; |
|
break; |
|
} |
|
// See vstem3. |
|
error = this.executeCommand(2, COMMAND_MAP.hstem); |
|
break; |
|
case (12 << 8) + 6: // seac |
|
// seac is like type 2's special endchar but it doesn't use the |
|
// first argument asb, so remove it. |
|
if (SEAC_ANALYSIS_ENABLED) { |
|
this.seac = this.stack.splice(-4, 4); |
|
error = this.executeCommand(0, COMMAND_MAP.endchar); |
|
} else { |
|
error = this.executeCommand(4, COMMAND_MAP.endchar); |
|
} |
|
break; |
|
case (12 << 8) + 7: // sbw |
|
if (this.stack.length < 4) { |
|
error = true; |
|
break; |
|
} |
|
// To convert to type2 we have to move the width value to the |
|
// first part of the charstring and then use rmoveto with |
|
// (dx, dy). The height argument will not be used for vmtx and |
|
// vhea tables reconstruction -- ignoring it. |
|
var wy = this.stack.pop(); |
|
wx = this.stack.pop(); |
|
var sby = this.stack.pop(); |
|
sbx = this.stack.pop(); |
|
this.lsb = sbx; |
|
this.width = wx; |
|
this.stack.push(wx, sbx, sby); |
|
error = this.executeCommand(3, COMMAND_MAP.rmoveto); |
|
break; |
|
case (12 << 8) + 12: // div |
|
if (this.stack.length < 2) { |
|
error = true; |
|
break; |
|
} |
|
var num2 = this.stack.pop(); |
|
var num1 = this.stack.pop(); |
|
this.stack.push(num1 / num2); |
|
break; |
|
case (12 << 8) + 16: // callothersubr |
|
if (this.stack.length < 2) { |
|
error = true; |
|
break; |
|
} |
|
subrNumber = this.stack.pop(); |
|
var numArgs = this.stack.pop(); |
|
if (subrNumber === 0 && numArgs === 3) { |
|
var flexArgs = this.stack.splice(this.stack.length - 17, 17); |
|
this.stack.push( |
|
flexArgs[2] + flexArgs[0], // bcp1x + rpx |
|
flexArgs[3] + flexArgs[1], // bcp1y + rpy |
|
flexArgs[4], // bcp2x |
|
flexArgs[5], // bcp2y |
|
flexArgs[6], // p2x |
|
flexArgs[7], // p2y |
|
flexArgs[8], // bcp3x |
|
flexArgs[9], // bcp3y |
|
flexArgs[10], // bcp4x |
|
flexArgs[11], // bcp4y |
|
flexArgs[12], // p3x |
|
flexArgs[13], // p3y |
|
flexArgs[14] // flexDepth |
|
// 15 = finalx unused by flex |
|
// 16 = finaly unused by flex |
|
); |
|
error = this.executeCommand(13, COMMAND_MAP.flex, true); |
|
this.flexing = false; |
|
this.stack.push(flexArgs[15], flexArgs[16]); |
|
} else if (subrNumber === 1 && numArgs === 0) { |
|
this.flexing = true; |
|
} |
|
break; |
|
case (12 << 8) + 17: // pop |
|
// Ignore this since it is only used with othersubr. |
|
break; |
|
case (12 << 8) + 33: // setcurrentpoint |
|
// Ignore for now. |
|
this.stack = []; |
|
break; |
|
default: |
|
warn('Unknown type 1 charstring command of "' + value + '"'); |
|
break; |
|
} |
|
if (error) { |
|
break; |
|
} |
|
continue; |
|
} else if (value <= 246) { |
|
value = value - 139; |
|
} else if (value <= 250) { |
|
value = ((value - 247) * 256) + encoded[++i] + 108; |
|
} else if (value <= 254) { |
|
value = -((value - 251) * 256) - encoded[++i] - 108; |
|
} else { |
|
value = (encoded[++i] & 0xff) << 24 | (encoded[++i] & 0xff) << 16 | |
|
(encoded[++i] & 0xff) << 8 | (encoded[++i] & 0xff) << 0; |
|
} |
|
this.stack.push(value); |
|
} |
|
return error; |
|
}, |
|
|
|
executeCommand: function(howManyArgs, command, keepStack) { |
|
var stackLength = this.stack.length; |
|
if (howManyArgs > stackLength) { |
|
return true; |
|
} |
|
var start = stackLength - howManyArgs; |
|
for (var i = start; i < stackLength; i++) { |
|
var value = this.stack[i]; |
|
if (value === (value | 0)) { // int |
|
this.output.push(28, (value >> 8) & 0xff, value & 0xff); |
|
} else { // fixed point |
|
value = (65536 * value) | 0; |
|
this.output.push(255, |
|
(value >> 24) & 0xFF, |
|
(value >> 16) & 0xFF, |
|
(value >> 8) & 0xFF, |
|
value & 0xFF); |
|
} |
|
} |
|
this.output.push.apply(this.output, command); |
|
if (keepStack) { |
|
this.stack.splice(start, howManyArgs); |
|
} else { |
|
this.stack.length = 0; |
|
} |
|
return false; |
|
} |
|
}; |
|
|
|
return Type1CharString; |
|
})(); |
|
|
|
/* |
|
* Type1Parser encapsulate the needed code for parsing a Type1 font |
|
* program. Some of its logic depends on the Type2 charstrings |
|
* structure. |
|
* Note: this doesn't really parse the font since that would require evaluation |
|
* of PostScript, but it is possible in most cases to extract what we need |
|
* without a full parse. |
|
*/ |
|
var Type1Parser = (function Type1ParserClosure() { |
|
/* |
|
* Decrypt a Sequence of Ciphertext Bytes to Produce the Original Sequence |
|
* of Plaintext Bytes. The function took a key as a parameter which can be |
|
* for decrypting the eexec block of for decoding charStrings. |
|
*/ |
|
var EEXEC_ENCRYPT_KEY = 55665; |
|
var CHAR_STRS_ENCRYPT_KEY = 4330; |
|
|
|
function isHexDigit(code) { |
|
return code >= 48 && code <= 57 || // '0'-'9' |
|
code >= 65 && code <= 70 || // 'A'-'F' |
|
code >= 97 && code <= 102; // 'a'-'f' |
|
} |
|
|
|
function decrypt(data, key, discardNumber) { |
|
var r = key | 0, c1 = 52845, c2 = 22719; |
|
var count = data.length; |
|
var decrypted = new Uint8Array(count); |
|
for (var i = 0; i < count; i++) { |
|
var value = data[i]; |
|
decrypted[i] = value ^ (r >> 8); |
|
r = ((value + r) * c1 + c2) & ((1 << 16) - 1); |
|
} |
|
return Array.prototype.slice.call(decrypted, discardNumber); |
|
} |
|
|
|
function decryptAscii(data, key, discardNumber) { |
|
var r = key | 0, c1 = 52845, c2 = 22719; |
|
var count = data.length, maybeLength = count >>> 1; |
|
var decrypted = new Uint8Array(maybeLength); |
|
var i, j; |
|
for (i = 0, j = 0; i < count; i++) { |
|
var digit1 = data[i]; |
|
if (!isHexDigit(digit1)) { |
|
continue; |
|
} |
|
i++; |
|
var digit2; |
|
while (i < count && !isHexDigit(digit2 = data[i])) { |
|
i++; |
|
} |
|
if (i < count) { |
|
var value = parseInt(String.fromCharCode(digit1, digit2), 16); |
|
decrypted[j++] = value ^ (r >> 8); |
|
r = ((value + r) * c1 + c2) & ((1 << 16) - 1); |
|
} |
|
} |
|
return Array.prototype.slice.call(decrypted, discardNumber, j); |
|
} |
|
|
|
function isSpecial(c) { |
|
return c === 0x2F || // '/' |
|
c === 0x5B || c === 0x5D || // '[', ']' |
|
c === 0x7B || c === 0x7D || // '{', '}' |
|
c === 0x28 || c === 0x29; // '(', ')' |
|
} |
|
|
|
function Type1Parser(stream, encrypted) { |
|
if (encrypted) { |
|
var data = stream.getBytes(); |
|
var isBinary = !(isHexDigit(data[0]) && isHexDigit(data[1]) && |
|
isHexDigit(data[2]) && isHexDigit(data[3])); |
|
stream = new Stream(isBinary ? decrypt(data, EEXEC_ENCRYPT_KEY, 4) : |
|
decryptAscii(data, EEXEC_ENCRYPT_KEY, 4)); |
|
} |
|
this.stream = stream; |
|
this.nextChar(); |
|
} |
|
|
|
Type1Parser.prototype = { |
|
readNumberArray: function Type1Parser_readNumberArray() { |
|
this.getToken(); // read '[' or '{' (arrays can start with either) |
|
var array = []; |
|
while (true) { |
|
var token = this.getToken(); |
|
if (token === null || token === ']' || token === '}') { |
|
break; |
|
} |
|
array.push(parseFloat(token || 0)); |
|
} |
|
return array; |
|
}, |
|
|
|
readNumber: function Type1Parser_readNumber() { |
|
var token = this.getToken(); |
|
return parseFloat(token || 0); |
|
}, |
|
|
|
readInt: function Type1Parser_readInt() { |
|
// Use '| 0' to prevent setting a double into length such as the double |
|
// does not flow into the loop variable. |
|
var token = this.getToken(); |
|
return parseInt(token || 0, 10) | 0; |
|
}, |
|
|
|
readBoolean: function Type1Parser_readBoolean() { |
|
var token = this.getToken(); |
|
|
|
// Use 1 and 0 since that's what type2 charstrings use. |
|
return token === 'true' ? 1 : 0; |
|
}, |
|
|
|
nextChar : function Type1_nextChar() { |
|
return (this.currentChar = this.stream.getByte()); |
|
}, |
|
|
|
getToken: function Type1Parser_getToken() { |
|
// Eat whitespace and comments. |
|
var comment = false; |
|
var ch = this.currentChar; |
|
while (true) { |
|
if (ch === -1) { |
|
return null; |
|
} |
|
|
|
if (comment) { |
|
if (ch === 0x0A || ch === 0x0D) { |
|
comment = false; |
|
} |
|
} else if (ch === 0x25) { // '%' |
|
comment = true; |
|
} else if (!Lexer.isSpace(ch)) { |
|
break; |
|
} |
|
ch = this.nextChar(); |
|
} |
|
if (isSpecial(ch)) { |
|
this.nextChar(); |
|
return String.fromCharCode(ch); |
|
} |
|
var token = ''; |
|
do { |
|
token += String.fromCharCode(ch); |
|
ch = this.nextChar(); |
|
} while (ch >= 0 && !Lexer.isSpace(ch) && !isSpecial(ch)); |
|
return token; |
|
}, |
|
|
|
/* |
|
* Returns an object containing a Subrs array and a CharStrings |
|
* array extracted from and eexec encrypted block of data |
|
*/ |
|
extractFontProgram: function Type1Parser_extractFontProgram() { |
|
var stream = this.stream; |
|
|
|
var subrs = [], charstrings = []; |
|
var program = { |
|
subrs: [], |
|
charstrings: [], |
|
properties: { |
|
'privateData': { |
|
'lenIV': 4 |
|
} |
|
} |
|
}; |
|
var token, length, data, lenIV, encoded; |
|
while ((token = this.getToken()) !== null) { |
|
if (token !== '/') { |
|
continue; |
|
} |
|
token = this.getToken(); |
|
switch (token) { |
|
case 'CharStrings': |
|
// The number immediately following CharStrings must be greater or |
|
// equal to the number of CharStrings. |
|
this.getToken(); |
|
this.getToken(); // read in 'dict' |
|
this.getToken(); // read in 'dup' |
|
this.getToken(); // read in 'begin' |
|
while(true) { |
|
token = this.getToken(); |
|
if (token === null || token === 'end') { |
|
break; |
|
} |
|
|
|
if (token !== '/') { |
|
continue; |
|
} |
|
var glyph = this.getToken(); |
|
length = this.readInt(); |
|
this.getToken(); // read in 'RD' or '-|' |
|
data = stream.makeSubStream(stream.pos, length); |
|
lenIV = program.properties.privateData['lenIV']; |
|
encoded = decrypt(data.getBytes(), CHAR_STRS_ENCRYPT_KEY, lenIV); |
|
// Skip past the required space and binary data. |
|
stream.skip(length); |
|
this.nextChar(); |
|
token = this.getToken(); // read in 'ND' or '|-' |
|
if (token === 'noaccess') { |
|
this.getToken(); // read in 'def' |
|
} |
|
charstrings.push({ |
|
glyph: glyph, |
|
encoded: encoded |
|
}); |
|
} |
|
break; |
|
case 'Subrs': |
|
var num = this.readInt(); |
|
this.getToken(); // read in 'array' |
|
while ((token = this.getToken()) === 'dup') { |
|
var index = this.readInt(); |
|
length = this.readInt(); |
|
this.getToken(); // read in 'RD' or '-|' |
|
data = stream.makeSubStream(stream.pos, length); |
|
lenIV = program.properties.privateData['lenIV']; |
|
encoded = decrypt(data.getBytes(), CHAR_STRS_ENCRYPT_KEY, lenIV); |
|
// Skip past the required space and binary data. |
|
stream.skip(length); |
|
this.nextChar(); |
|
token = this.getToken(); // read in 'NP' or '|' |
|
if (token === 'noaccess') { |
|
this.getToken(); // read in 'put' |
|
} |
|
subrs[index] = encoded; |
|
} |
|
break; |
|
case 'BlueValues': |
|
case 'OtherBlues': |
|
case 'FamilyBlues': |
|
case 'FamilyOtherBlues': |
|
var blueArray = this.readNumberArray(); |
|
// *Blue* values may contain invalid data: disables reading of |
|
// those values when hinting is disabled. |
|
if (blueArray.length > 0 && (blueArray.length % 2) === 0 && |
|
HINTING_ENABLED) { |
|
program.properties.privateData[token] = blueArray; |
|
} |
|
break; |
|
case 'StemSnapH': |
|
case 'StemSnapV': |
|
program.properties.privateData[token] = this.readNumberArray(); |
|
break; |
|
case 'StdHW': |
|
case 'StdVW': |
|
program.properties.privateData[token] = |
|
this.readNumberArray()[0]; |
|
break; |
|
case 'BlueShift': |
|
case 'lenIV': |
|
case 'BlueFuzz': |
|
case 'BlueScale': |
|
case 'LanguageGroup': |
|
case 'ExpansionFactor': |
|
program.properties.privateData[token] = this.readNumber(); |
|
break; |
|
case 'ForceBold': |
|
program.properties.privateData[token] = this.readBoolean(); |
|
break; |
|
} |
|
} |
|
|
|
for (var i = 0; i < charstrings.length; i++) { |
|
glyph = charstrings[i].glyph; |
|
encoded = charstrings[i].encoded; |
|
var charString = new Type1CharString(); |
|
var error = charString.convert(encoded, subrs); |
|
var output = charString.output; |
|
if (error) { |
|
// It seems when FreeType encounters an error while evaluating a glyph |
|
// that it completely ignores the glyph so we'll mimic that behaviour |
|
// here and put an endchar to make the validator happy. |
|
output = [14]; |
|
} |
|
program.charstrings.push({ |
|
glyphName: glyph, |
|
charstring: output, |
|
width: charString.width, |
|
lsb: charString.lsb, |
|
seac: charString.seac |
|
}); |
|
} |
|
|
|
return program; |
|
}, |
|
|
|
extractFontHeader: function Type1Parser_extractFontHeader(properties) { |
|
var token; |
|
while ((token = this.getToken()) !== null) { |
|
if (token !== '/') { |
|
continue; |
|
} |
|
token = this.getToken(); |
|
switch (token) { |
|
case 'FontMatrix': |
|
var matrix = this.readNumberArray(); |
|
properties.fontMatrix = matrix; |
|
break; |
|
case 'Encoding': |
|
var encodingArg = this.getToken(); |
|
var encoding; |
|
if (!/^\d+$/.test(encodingArg)) { |
|
// encoding name is specified |
|
encoding = Encodings[encodingArg]; |
|
} else { |
|
encoding = []; |
|
var size = parseInt(encodingArg, 10) | 0; |
|
this.getToken(); // read in 'array' |
|
|
|
for (var j = 0; j < size; j++) { |
|
token = this.getToken(); |
|
// skipping till first dup or def (e.g. ignoring for statement) |
|
while (token !== 'dup' && token !== 'def') { |
|
token = this.getToken(); |
|
if (token === null) { |
|
return; // invalid header |
|
} |
|
} |
|
if (token === 'def') { |
|
break; // read all array data |
|
} |
|
var index = this.readInt(); |
|
this.getToken(); // read in '/' |
|
var glyph = this.getToken(); |
|
encoding[index] = glyph; |
|
this.getToken(); // read the in 'put' |
|
} |
|
} |
|
properties.builtInEncoding = encoding; |
|
break; |
|
case 'FontBBox': |
|
var fontBBox = this.readNumberArray(); |
|
// adjusting ascent/descent |
|
properties.ascent = fontBBox[3]; |
|
properties.descent = fontBBox[1]; |
|
properties.ascentScaled = true; |
|
break; |
|
} |
|
} |
|
} |
|
}; |
|
|
|
return Type1Parser; |
|
})(); |
|
|
|
/** |
|
* The CFF class takes a Type1 file and wrap it into a |
|
* 'Compact Font Format' which itself embed Type2 charstrings. |
|
*/ |
|
var CFFStandardStrings = [ |
|
'.notdef', 'space', 'exclam', 'quotedbl', 'numbersign', 'dollar', 'percent', |
|
'ampersand', 'quoteright', 'parenleft', 'parenright', 'asterisk', 'plus', |
|
'comma', 'hyphen', 'period', 'slash', 'zero', 'one', 'two', 'three', 'four', |
|
'five', 'six', 'seven', 'eight', 'nine', 'colon', 'semicolon', 'less', |
|
'equal', 'greater', 'question', 'at', 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', |
|
'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', |
|
'X', 'Y', 'Z', 'bracketleft', 'backslash', 'bracketright', 'asciicircum', |
|
'underscore', 'quoteleft', 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', |
|
'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', |
|
'z', 'braceleft', 'bar', 'braceright', 'asciitilde', 'exclamdown', 'cent', |
|
'sterling', 'fraction', 'yen', 'florin', 'section', 'currency', |
|
'quotesingle', 'quotedblleft', 'guillemotleft', 'guilsinglleft', |
|
'guilsinglright', 'fi', 'fl', 'endash', 'dagger', 'daggerdbl', |
|
'periodcentered', 'paragraph', 'bullet', 'quotesinglbase', 'quotedblbase', |
|
'quotedblright', 'guillemotright', 'ellipsis', 'perthousand', 'questiondown', |
|
'grave', 'acute', 'circumflex', 'tilde', 'macron', 'breve', 'dotaccent', |
|
'dieresis', 'ring', 'cedilla', 'hungarumlaut', 'ogonek', 'caron', 'emdash', |
|
'AE', 'ordfeminine', 'Lslash', 'Oslash', 'OE', 'ordmasculine', 'ae', |
|
'dotlessi', 'lslash', 'oslash', 'oe', 'germandbls', 'onesuperior', |
|
'logicalnot', 'mu', 'trademark', 'Eth', 'onehalf', 'plusminus', 'Thorn', |
|
'onequarter', 'divide', 'brokenbar', 'degree', 'thorn', 'threequarters', |
|
'twosuperior', 'registered', 'minus', 'eth', 'multiply', 'threesuperior', |
|
'copyright', 'Aacute', 'Acircumflex', 'Adieresis', 'Agrave', 'Aring', |
|
'Atilde', 'Ccedilla', 'Eacute', 'Ecircumflex', 'Edieresis', 'Egrave', |
|
'Iacute', 'Icircumflex', 'Idieresis', 'Igrave', 'Ntilde', 'Oacute', |
|
'Ocircumflex', 'Odieresis', 'Ograve', 'Otilde', 'Scaron', 'Uacute', |
|
'Ucircumflex', 'Udieresis', 'Ugrave', 'Yacute', 'Ydieresis', 'Zcaron', |
|
'aacute', 'acircumflex', 'adieresis', 'agrave', 'aring', 'atilde', |
|
'ccedilla', 'eacute', 'ecircumflex', 'edieresis', 'egrave', 'iacute', |
|
'icircumflex', 'idieresis', 'igrave', 'ntilde', 'oacute', 'ocircumflex', |
|
'odieresis', 'ograve', 'otilde', 'scaron', 'uacute', 'ucircumflex', |
|
'udieresis', 'ugrave', 'yacute', 'ydieresis', 'zcaron', 'exclamsmall', |
|
'Hungarumlautsmall', 'dollaroldstyle', 'dollarsuperior', 'ampersandsmall', |
|
'Acutesmall', 'parenleftsuperior', 'parenrightsuperior', 'twodotenleader', |
|
'onedotenleader', 'zerooldstyle', 'oneoldstyle', 'twooldstyle', |
|
'threeoldstyle', 'fouroldstyle', 'fiveoldstyle', 'sixoldstyle', |
|
'sevenoldstyle', 'eightoldstyle', 'nineoldstyle', 'commasuperior', |
|
'threequartersemdash', 'periodsuperior', 'questionsmall', 'asuperior', |
|
'bsuperior', 'centsuperior', 'dsuperior', 'esuperior', 'isuperior', |
|
'lsuperior', 'msuperior', 'nsuperior', 'osuperior', 'rsuperior', 'ssuperior', |
|
'tsuperior', 'ff', 'ffi', 'ffl', 'parenleftinferior', 'parenrightinferior', |
|
'Circumflexsmall', 'hyphensuperior', 'Gravesmall', 'Asmall', 'Bsmall', |
|
'Csmall', 'Dsmall', 'Esmall', 'Fsmall', 'Gsmall', 'Hsmall', 'Ismall', |
|
'Jsmall', 'Ksmall', 'Lsmall', 'Msmall', 'Nsmall', 'Osmall', 'Psmall', |
|
'Qsmall', 'Rsmall', 'Ssmall', 'Tsmall', 'Usmall', 'Vsmall', 'Wsmall', |
|
'Xsmall', 'Ysmall', 'Zsmall', 'colonmonetary', 'onefitted', 'rupiah', |
|
'Tildesmall', 'exclamdownsmall', 'centoldstyle', 'Lslashsmall', |
|
'Scaronsmall', 'Zcaronsmall', 'Dieresissmall', 'Brevesmall', 'Caronsmall', |
|
'Dotaccentsmall', 'Macronsmall', 'figuredash', 'hypheninferior', |
|
'Ogoneksmall', 'Ringsmall', 'Cedillasmall', 'questiondownsmall', 'oneeighth', |
|
'threeeighths', 'fiveeighths', 'seveneighths', 'onethird', 'twothirds', |
|
'zerosuperior', 'foursuperior', 'fivesuperior', 'sixsuperior', |
|
'sevensuperior', 'eightsuperior', 'ninesuperior', 'zeroinferior', |
|
'oneinferior', 'twoinferior', 'threeinferior', 'fourinferior', |
|
'fiveinferior', 'sixinferior', 'seveninferior', 'eightinferior', |
|
'nineinferior', 'centinferior', 'dollarinferior', 'periodinferior', |
|
'commainferior', 'Agravesmall', 'Aacutesmall', 'Acircumflexsmall', |
|
'Atildesmall', 'Adieresissmall', 'Aringsmall', 'AEsmall', 'Ccedillasmall', |
|
'Egravesmall', 'Eacutesmall', 'Ecircumflexsmall', 'Edieresissmall', |
|
'Igravesmall', 'Iacutesmall', 'Icircumflexsmall', 'Idieresissmall', |
|
'Ethsmall', 'Ntildesmall', 'Ogravesmall', 'Oacutesmall', 'Ocircumflexsmall', |
|
'Otildesmall', 'Odieresissmall', 'OEsmall', 'Oslashsmall', 'Ugravesmall', |
|
'Uacutesmall', 'Ucircumflexsmall', 'Udieresissmall', 'Yacutesmall', |
|
'Thornsmall', 'Ydieresissmall', '001.000', '001.001', '001.002', '001.003', |
|
'Black', 'Bold', 'Book', 'Light', 'Medium', 'Regular', 'Roman', 'Semibold' |
|
]; |
|
|
|
// Type1Font is also a CIDFontType0. |
|
var Type1Font = function Type1Font(name, file, properties) { |
|
// Some bad generators embed pfb file as is, we have to strip 6-byte headers. |
|
// Also, length1 and length2 might be off by 6 bytes as well. |
|
// http://www.math.ubc.ca/~cass/piscript/type1.pdf |
|
var PFB_HEADER_SIZE = 6; |
|
var headerBlockLength = properties.length1; |
|
var eexecBlockLength = properties.length2; |
|
var pfbHeader = file.peekBytes(PFB_HEADER_SIZE); |
|
var pfbHeaderPresent = pfbHeader[0] === 0x80 && pfbHeader[1] === 0x01; |
|
if (pfbHeaderPresent) { |
|
file.skip(PFB_HEADER_SIZE); |
|
headerBlockLength = (pfbHeader[5] << 24) | (pfbHeader[4] << 16) | |
|
(pfbHeader[3] << 8) | pfbHeader[2]; |
|
} |
|
|
|
// Get the data block containing glyphs and subrs informations |
|
var headerBlock = new Stream(file.getBytes(headerBlockLength)); |
|
var headerBlockParser = new Type1Parser(headerBlock); |
|
headerBlockParser.extractFontHeader(properties); |
|
|
|
if (pfbHeaderPresent) { |
|
pfbHeader = file.getBytes(PFB_HEADER_SIZE); |
|
eexecBlockLength = (pfbHeader[5] << 24) | (pfbHeader[4] << 16) | |
|
(pfbHeader[3] << 8) | pfbHeader[2]; |
|
} |
|
|
|
// Decrypt the data blocks and retrieve it's content |
|
var eexecBlock = new Stream(file.getBytes(eexecBlockLength)); |
|
var eexecBlockParser = new Type1Parser(eexecBlock, true); |
|
var data = eexecBlockParser.extractFontProgram(); |
|
for (var info in data.properties) { |
|
properties[info] = data.properties[info]; |
|
} |
|
|
|
var charstrings = data.charstrings; |
|
var type2Charstrings = this.getType2Charstrings(charstrings); |
|
var subrs = this.getType2Subrs(data.subrs); |
|
|
|
this.charstrings = charstrings; |
|
this.data = this.wrap(name, type2Charstrings, this.charstrings, |
|
subrs, properties); |
|
this.seacs = this.getSeacs(data.charstrings); |
|
}; |
|
|
|
Type1Font.prototype = { |
|
get numGlyphs() { |
|
return this.charstrings.length + 1; |
|
}, |
|
|
|
getCharset: function Type1Font_getCharset() { |
|
var charset = ['.notdef']; |
|
var charstrings = this.charstrings; |
|
for (var glyphId = 0; glyphId < charstrings.length; glyphId++) { |
|
charset.push(charstrings[glyphId].glyphName); |
|
} |
|
return charset; |
|
}, |
|
|
|
getGlyphMapping: function Type1Font_getGlyphMapping(properties) { |
|
var charstrings = this.charstrings; |
|
var glyphNames = ['.notdef'], glyphId; |
|
for (glyphId = 0; glyphId < charstrings.length; glyphId++) { |
|
glyphNames.push(charstrings[glyphId].glyphName); |
|
} |
|
var encoding = properties.builtInEncoding; |
|
if (encoding) { |
|
var builtInEncoding = {}; |
|
for (var charCode in encoding) { |
|
glyphId = glyphNames.indexOf(encoding[charCode]); |
|
if (glyphId >= 0) { |
|
builtInEncoding[charCode] = glyphId; |
|
} |
|
} |
|
} |
|
|
|
return type1FontGlyphMapping(properties, builtInEncoding, glyphNames); |
|
}, |
|
|
|
getSeacs: function Type1Font_getSeacs(charstrings) { |
|
var i, ii; |
|
var seacMap = []; |
|
for (i = 0, ii = charstrings.length; i < ii; i++) { |
|
var charstring = charstrings[i]; |
|
if (charstring.seac) { |
|
// Offset by 1 for .notdef |
|
seacMap[i + 1] = charstring.seac; |
|
} |
|
} |
|
return seacMap; |
|
}, |
|
|
|
getType2Charstrings: function Type1Font_getType2Charstrings( |
|
type1Charstrings) { |
|
var type2Charstrings = []; |
|
for (var i = 0, ii = type1Charstrings.length; i < ii; i++) { |
|
type2Charstrings.push(type1Charstrings[i].charstring); |
|
} |
|
return type2Charstrings; |
|
}, |
|
|
|
getType2Subrs: function Type1Font_getType2Subrs(type1Subrs) { |
|
var bias = 0; |
|
var count = type1Subrs.length; |
|
if (count < 1133) { |
|
bias = 107; |
|
} else if (count < 33769) { |
|
bias = 1131; |
|
} else { |
|
bias = 32768; |
|
} |
|
|
|
// Add a bunch of empty subrs to deal with the Type2 bias |
|
var type2Subrs = []; |
|
var i; |
|
for (i = 0; i < bias; i++) { |
|
type2Subrs.push([0x0B]); |
|
} |
|
|
|
for (i = 0; i < count; i++) { |
|
type2Subrs.push(type1Subrs[i]); |
|
} |
|
|
|
return type2Subrs; |
|
}, |
|
|
|
wrap: function Type1Font_wrap(name, glyphs, charstrings, subrs, properties) { |
|
var cff = new CFF(); |
|
cff.header = new CFFHeader(1, 0, 4, 4); |
|
|
|
cff.names = [name]; |
|
|
|
var topDict = new CFFTopDict(); |
|
// CFF strings IDs 0...390 are predefined names, so refering |
|
// to entries in our own String INDEX starts at SID 391. |
|
topDict.setByName('version', 391); |
|
topDict.setByName('Notice', 392); |
|
topDict.setByName('FullName', 393); |
|
topDict.setByName('FamilyName', 394); |
|
topDict.setByName('Weight', 395); |
|
topDict.setByName('Encoding', null); // placeholder |
|
topDict.setByName('FontMatrix', properties.fontMatrix); |
|
topDict.setByName('FontBBox', properties.bbox); |
|
topDict.setByName('charset', null); // placeholder |
|
topDict.setByName('CharStrings', null); // placeholder |
|
topDict.setByName('Private', null); // placeholder |
|
cff.topDict = topDict; |
|
|
|
var strings = new CFFStrings(); |
|
strings.add('Version 0.11'); // Version |
|
strings.add('See original notice'); // Notice |
|
strings.add(name); // FullName |
|
strings.add(name); // FamilyName |
|
strings.add('Medium'); // Weight |
|
cff.strings = strings; |
|
|
|
cff.globalSubrIndex = new CFFIndex(); |
|
|
|
var count = glyphs.length; |
|
var charsetArray = [0]; |
|
var i, ii; |
|
for (i = 0; i < count; i++) { |
|
var index = CFFStandardStrings.indexOf(charstrings[i].glyphName); |
|
// TODO: Insert the string and correctly map it. Previously it was |
|
// thought mapping names that aren't in the standard strings to .notdef |
|
// was fine, however in issue818 when mapping them all to .notdef the |
|
// adieresis glyph no longer worked. |
|
if (index === -1) { |
|
index = 0; |
|
} |
|
charsetArray.push((index >> 8) & 0xff, index & 0xff); |
|
} |
|
cff.charset = new CFFCharset(false, 0, [], charsetArray); |
|
|
|
var charStringsIndex = new CFFIndex(); |
|
charStringsIndex.add([0x8B, 0x0E]); // .notdef |
|
for (i = 0; i < count; i++) { |
|
charStringsIndex.add(glyphs[i]); |
|
} |
|
cff.charStrings = charStringsIndex; |
|
|
|
var privateDict = new CFFPrivateDict(); |
|
privateDict.setByName('Subrs', null); // placeholder |
|
var fields = [ |
|
'BlueValues', |
|
'OtherBlues', |
|
'FamilyBlues', |
|
'FamilyOtherBlues', |
|
'StemSnapH', |
|
'StemSnapV', |
|
'BlueShift', |
|
'BlueFuzz', |
|
'BlueScale', |
|
'LanguageGroup', |
|
'ExpansionFactor', |
|
'ForceBold', |
|
'StdHW', |
|
'StdVW' |
|
]; |
|
for (i = 0, ii = fields.length; i < ii; i++) { |
|
var field = fields[i]; |
|
if (!properties.privateData.hasOwnProperty(field)) { |
|
continue; |
|
} |
|
var value = properties.privateData[field]; |
|
if (isArray(value)) { |
|
// All of the private dictionary array data in CFF must be stored as |
|
// "delta-encoded" numbers. |
|
for (var j = value.length - 1; j > 0; j--) { |
|
value[j] -= value[j - 1]; // ... difference from previous value |
|
} |
|
} |
|
privateDict.setByName(field, value); |
|
} |
|
cff.topDict.privateDict = privateDict; |
|
|
|
var subrIndex = new CFFIndex(); |
|
for (i = 0, ii = subrs.length; i < ii; i++) { |
|
subrIndex.add(subrs[i]); |
|
} |
|
privateDict.subrsIndex = subrIndex; |
|
|
|
var compiler = new CFFCompiler(cff); |
|
return compiler.compile(); |
|
} |
|
}; |
|
|
|
var CFFFont = (function CFFFontClosure() { |
|
function CFFFont(file, properties) { |
|
this.properties = properties; |
|
|
|
var parser = new CFFParser(file, properties); |
|
this.cff = parser.parse(); |
|
var compiler = new CFFCompiler(this.cff); |
|
this.seacs = this.cff.seacs; |
|
try { |
|
this.data = compiler.compile(); |
|
} catch (e) { |
|
warn('Failed to compile font ' + properties.loadedName); |
|
// There may have just been an issue with the compiler, set the data |
|
// anyway and hope the font loaded. |
|
this.data = file; |
|
} |
|
} |
|
|
|
CFFFont.prototype = { |
|
get numGlyphs() { |
|
return this.cff.charStrings.count; |
|
}, |
|
getCharset: function CFFFont_getCharset() { |
|
return this.cff.charset.charset; |
|
}, |
|
getGlyphMapping: function CFFFont_getGlyphMapping() { |
|
var cff = this.cff; |
|
var properties = this.properties; |
|
var charsets = cff.charset.charset; |
|
var charCodeToGlyphId; |
|
var glyphId; |
|
|
|
if (properties.composite) { |
|
charCodeToGlyphId = Object.create(null); |
|
if (cff.isCIDFont) { |
|
// If the font is actually a CID font then we should use the charset |
|
// to map CIDs to GIDs. |
|
for (glyphId = 0; glyphId < charsets.length; glyphId++) { |
|
var cid = charsets[glyphId]; |
|
var charCode = properties.cMap.charCodeOf(cid); |
|
charCodeToGlyphId[charCode] = glyphId; |
|
} |
|
} else { |
|
// If it is NOT actually a CID font then CIDs should be mapped |
|
// directly to GIDs. |
|
for (glyphId = 0; glyphId < cff.charStrings.count; glyphId++) { |
|
charCodeToGlyphId[glyphId] = glyphId; |
|
} |
|
} |
|
return charCodeToGlyphId; |
|
} |
|
|
|
var encoding = cff.encoding ? cff.encoding.encoding : null; |
|
charCodeToGlyphId = type1FontGlyphMapping(properties, encoding, charsets); |
|
return charCodeToGlyphId; |
|
} |
|
}; |
|
|
|
return CFFFont; |
|
})(); |
|
|
|
var CFFParser = (function CFFParserClosure() { |
|
var CharstringValidationData = [ |
|
null, |
|
{ id: 'hstem', min: 2, stackClearing: true, stem: true }, |
|
null, |
|
{ id: 'vstem', min: 2, stackClearing: true, stem: true }, |
|
{ id: 'vmoveto', min: 1, stackClearing: true }, |
|
{ id: 'rlineto', min: 2, resetStack: true }, |
|
{ id: 'hlineto', min: 1, resetStack: true }, |
|
{ id: 'vlineto', min: 1, resetStack: true }, |
|
{ id: 'rrcurveto', min: 6, resetStack: true }, |
|
null, |
|
{ id: 'callsubr', min: 1, undefStack: true }, |
|
{ id: 'return', min: 0, undefStack: true }, |
|
null, // 12 |
|
null, |
|
{ id: 'endchar', min: 0, stackClearing: true }, |
|
null, |
|
null, |
|
null, |
|
{ id: 'hstemhm', min: 2, stackClearing: true, stem: true }, |
|
{ id: 'hintmask', min: 0, stackClearing: true }, |
|
{ id: 'cntrmask', min: 0, stackClearing: true }, |
|
{ id: 'rmoveto', min: 2, stackClearing: true }, |
|
{ id: 'hmoveto', min: 1, stackClearing: true }, |
|
{ id: 'vstemhm', min: 2, stackClearing: true, stem: true }, |
|
{ id: 'rcurveline', min: 8, resetStack: true }, |
|
{ id: 'rlinecurve', min: 8, resetStack: true }, |
|
{ id: 'vvcurveto', min: 4, resetStack: true }, |
|
{ id: 'hhcurveto', min: 4, resetStack: true }, |
|
null, // shortint |
|
{ id: 'callgsubr', min: 1, undefStack: true }, |
|
{ id: 'vhcurveto', min: 4, resetStack: true }, |
|
{ id: 'hvcurveto', min: 4, resetStack: true } |
|
]; |
|
var CharstringValidationData12 = [ |
|
null, |
|
null, |
|
null, |
|
{ id: 'and', min: 2, stackDelta: -1 }, |
|
{ id: 'or', min: 2, stackDelta: -1 }, |
|
{ id: 'not', min: 1, stackDelta: 0 }, |
|
null, |
|
null, |
|
null, |
|
{ id: 'abs', min: 1, stackDelta: 0 }, |
|
{ id: 'add', min: 2, stackDelta: -1, |
|
stackFn: function stack_div(stack, index) { |
|
stack[index - 2] = stack[index - 2] + stack[index - 1]; |
|
} |
|
}, |
|
{ id: 'sub', min: 2, stackDelta: -1, |
|
stackFn: function stack_div(stack, index) { |
|
stack[index - 2] = stack[index - 2] - stack[index - 1]; |
|
} |
|
}, |
|
{ id: 'div', min: 2, stackDelta: -1, |
|
stackFn: function stack_div(stack, index) { |
|
stack[index - 2] = stack[index - 2] / stack[index - 1]; |
|
} |
|
}, |
|
null, |
|
{ id: 'neg', min: 1, stackDelta: 0, |
|
stackFn: function stack_div(stack, index) { |
|
stack[index - 1] = -stack[index - 1]; |
|
} |
|
}, |
|
{ id: 'eq', min: 2, stackDelta: -1 }, |
|
null, |
|
null, |
|
{ id: 'drop', min: 1, stackDelta: -1 }, |
|
null, |
|
{ id: 'put', min: 2, stackDelta: -2 }, |
|
{ id: 'get', min: 1, stackDelta: 0 }, |
|
{ id: 'ifelse', min: 4, stackDelta: -3 }, |
|
{ id: 'random', min: 0, stackDelta: 1 }, |
|
{ id: 'mul', min: 2, stackDelta: -1, |
|
stackFn: function stack_div(stack, index) { |
|
stack[index - 2] = stack[index - 2] * stack[index - 1]; |
|
} |
|
}, |
|
null, |
|
{ id: 'sqrt', min: 1, stackDelta: 0 }, |
|
{ id: 'dup', min: 1, stackDelta: 1 }, |
|
{ id: 'exch', min: 2, stackDelta: 0 }, |
|
{ id: 'index', min: 2, stackDelta: 0 }, |
|
{ id: 'roll', min: 3, stackDelta: -2 }, |
|
null, |
|
null, |
|
null, |
|
{ id: 'hflex', min: 7, resetStack: true }, |
|
{ id: 'flex', min: 13, resetStack: true }, |
|
{ id: 'hflex1', min: 9, resetStack: true }, |
|
{ id: 'flex1', min: 11, resetStack: true } |
|
]; |
|
|
|
function CFFParser(file, properties) { |
|
this.bytes = file.getBytes(); |
|
this.properties = properties; |
|
} |
|
CFFParser.prototype = { |
|
parse: function CFFParser_parse() { |
|
var properties = this.properties; |
|
var cff = new CFF(); |
|
this.cff = cff; |
|
|
|
// The first five sections must be in order, all the others are reached |
|
// via offsets contained in one of the below. |
|
var header = this.parseHeader(); |
|
var nameIndex = this.parseIndex(header.endPos); |
|
var topDictIndex = this.parseIndex(nameIndex.endPos); |
|
var stringIndex = this.parseIndex(topDictIndex.endPos); |
|
var globalSubrIndex = this.parseIndex(stringIndex.endPos); |
|
|
|
var topDictParsed = this.parseDict(topDictIndex.obj.get(0)); |
|
var topDict = this.createDict(CFFTopDict, topDictParsed, cff.strings); |
|
|
|
cff.header = header.obj; |
|
cff.names = this.parseNameIndex(nameIndex.obj); |
|
cff.strings = this.parseStringIndex(stringIndex.obj); |
|
cff.topDict = topDict; |
|
cff.globalSubrIndex = globalSubrIndex.obj; |
|
|
|
this.parsePrivateDict(cff.topDict); |
|
|
|
cff.isCIDFont = topDict.hasName('ROS'); |
|
|
|
var charStringOffset = topDict.getByName('CharStrings'); |
|
var charStringIndex = this.parseIndex(charStringOffset).obj; |
|
|
|
var fontMatrix = topDict.getByName('FontMatrix'); |
|
if (fontMatrix) { |
|
properties.fontMatrix = fontMatrix; |
|
} |
|
|
|
var fontBBox = topDict.getByName('FontBBox'); |
|
if (fontBBox) { |
|
// adjusting ascent/descent |
|
properties.ascent = fontBBox[3]; |
|
properties.descent = fontBBox[1]; |
|
properties.ascentScaled = true; |
|
} |
|
|
|
var charset, encoding; |
|
if (cff.isCIDFont) { |
|
var fdArrayIndex = this.parseIndex(topDict.getByName('FDArray')).obj; |
|
for (var i = 0, ii = fdArrayIndex.count; i < ii; ++i) { |
|
var dictRaw = fdArrayIndex.get(i); |
|
var fontDict = this.createDict(CFFTopDict, this.parseDict(dictRaw), |
|
cff.strings); |
|
this.parsePrivateDict(fontDict); |
|
cff.fdArray.push(fontDict); |
|
} |
|
// cid fonts don't have an encoding |
|
encoding = null; |
|
charset = this.parseCharsets(topDict.getByName('charset'), |
|
charStringIndex.count, cff.strings, true); |
|
cff.fdSelect = this.parseFDSelect(topDict.getByName('FDSelect'), |
|
charStringIndex.count); |
|
} else { |
|
charset = this.parseCharsets(topDict.getByName('charset'), |
|
charStringIndex.count, cff.strings, false); |
|
encoding = this.parseEncoding(topDict.getByName('Encoding'), |
|
properties, |
|
cff.strings, charset.charset); |
|
} |
|
|
|
cff.charset = charset; |
|
cff.encoding = encoding; |
|
|
|
var charStringsAndSeacs = this.parseCharStrings( |
|
charStringIndex, |
|
topDict.privateDict.subrsIndex, |
|
globalSubrIndex.obj, |
|
cff.fdSelect, |
|
cff.fdArray); |
|
cff.charStrings = charStringsAndSeacs.charStrings; |
|
cff.seacs = charStringsAndSeacs.seacs; |
|
cff.widths = charStringsAndSeacs.widths; |
|
|
|
return cff; |
|
}, |
|
parseHeader: function CFFParser_parseHeader() { |
|
var bytes = this.bytes; |
|
var bytesLength = bytes.length; |
|
var offset = 0; |
|
|
|
// Prevent an infinite loop, by checking that the offset is within the |
|
// bounds of the bytes array. Necessary in empty, or invalid, font files. |
|
while (offset < bytesLength && bytes[offset] !== 1) { |
|
++offset; |
|
} |
|
if (offset >= bytesLength) { |
|
error('Invalid CFF header'); |
|
} else if (offset !== 0) { |
|
info('cff data is shifted'); |
|
bytes = bytes.subarray(offset); |
|
this.bytes = bytes; |
|
} |
|
var major = bytes[0]; |
|
var minor = bytes[1]; |
|
var hdrSize = bytes[2]; |
|
var offSize = bytes[3]; |
|
var header = new CFFHeader(major, minor, hdrSize, offSize); |
|
return { obj: header, endPos: hdrSize }; |
|
}, |
|
parseDict: function CFFParser_parseDict(dict) { |
|
var pos = 0; |
|
|
|
function parseOperand() { |
|
var value = dict[pos++]; |
|
if (value === 30) { |
|
return parseFloatOperand(pos); |
|
} else if (value === 28) { |
|
value = dict[pos++]; |
|
value = ((value << 24) | (dict[pos++] << 16)) >> 16; |
|
return value; |
|
} else if (value === 29) { |
|
value = dict[pos++]; |
|
value = (value << 8) | dict[pos++]; |
|
value = (value << 8) | dict[pos++]; |
|
value = (value << 8) | dict[pos++]; |
|
return value; |
|
} else if (value >= 32 && value <= 246) { |
|
return value - 139; |
|
} else if (value >= 247 && value <= 250) { |
|
return ((value - 247) * 256) + dict[pos++] + 108; |
|
} else if (value >= 251 && value <= 254) { |
|
return -((value - 251) * 256) - dict[pos++] - 108; |
|
} else { |
|
error('255 is not a valid DICT command'); |
|
} |
|
return -1; |
|
} |
|
|
|
function parseFloatOperand() { |
|
var str = ''; |
|
var eof = 15; |
|
var lookup = ['0', '1', '2', '3', '4', '5', '6', '7', '8', |
|
'9', '.', 'E', 'E-', null, '-']; |
|
var length = dict.length; |
|
while (pos < length) { |
|
var b = dict[pos++]; |
|
var b1 = b >> 4; |
|
var b2 = b & 15; |
|
|
|
if (b1 === eof) { |
|
break; |
|
} |
|
str += lookup[b1]; |
|
|
|
if (b2 === eof) { |
|
break; |
|
} |
|
str += lookup[b2]; |
|
} |
|
return parseFloat(str); |
|
} |
|
|
|
var operands = []; |
|
var entries = []; |
|
|
|
pos = 0; |
|
var end = dict.length; |
|
while (pos < end) { |
|
var b = dict[pos]; |
|
if (b <= 21) { |
|
if (b === 12) { |
|
b = (b << 8) | dict[++pos]; |
|
} |
|
entries.push([b, operands]); |
|
operands = []; |
|
++pos; |
|
} else { |
|
operands.push(parseOperand()); |
|
} |
|
} |
|
return entries; |
|
}, |
|
parseIndex: function CFFParser_parseIndex(pos) { |
|
var cffIndex = new CFFIndex(); |
|
var bytes = this.bytes; |
|
var count = (bytes[pos++] << 8) | bytes[pos++]; |
|
var offsets = []; |
|
var end = pos; |
|
var i, ii; |
|
|
|
if (count !== 0) { |
|
var offsetSize = bytes[pos++]; |
|
// add 1 for offset to determine size of last object |
|
var startPos = pos + ((count + 1) * offsetSize) - 1; |
|
|
|
for (i = 0, ii = count + 1; i < ii; ++i) { |
|
var offset = 0; |
|
for (var j = 0; j < offsetSize; ++j) { |
|
offset <<= 8; |
|
offset += bytes[pos++]; |
|
} |
|
offsets.push(startPos + offset); |
|
} |
|
end = offsets[count]; |
|
} |
|
for (i = 0, ii = offsets.length - 1; i < ii; ++i) { |
|
var offsetStart = offsets[i]; |
|
var offsetEnd = offsets[i + 1]; |
|
cffIndex.add(bytes.subarray(offsetStart, offsetEnd)); |
|
} |
|
return {obj: cffIndex, endPos: end}; |
|
}, |
|
parseNameIndex: function CFFParser_parseNameIndex(index) { |
|
var names = []; |
|
for (var i = 0, ii = index.count; i < ii; ++i) { |
|
var name = index.get(i); |
|
// OTS doesn't allow names to be over 127 characters. |
|
var length = Math.min(name.length, 127); |
|
var data = []; |
|
// OTS also only permits certain characters in the name. |
|
for (var j = 0; j < length; ++j) { |
|
var c = name[j]; |
|
if (j === 0 && c === 0) { |
|
data[j] = c; |
|
continue; |
|
} |
|
if ((c < 33 || c > 126) || c === 91 /* [ */ || c === 93 /* ] */ || |
|
c === 40 /* ( */ || c === 41 /* ) */ || c === 123 /* { */ || |
|
c === 125 /* } */ || c === 60 /* < */ || c === 62 /* > */ || |
|
c === 47 /* / */ || c === 37 /* % */ || c === 35 /* # */) { |
|
data[j] = 95; |
|
continue; |
|
} |
|
data[j] = c; |
|
} |
|
names.push(bytesToString(data)); |
|
} |
|
return names; |
|
}, |
|
parseStringIndex: function CFFParser_parseStringIndex(index) { |
|
var strings = new CFFStrings(); |
|
for (var i = 0, ii = index.count; i < ii; ++i) { |
|
var data = index.get(i); |
|
strings.add(bytesToString(data)); |
|
} |
|
return strings; |
|
}, |
|
createDict: function CFFParser_createDict(Type, dict, strings) { |
|
var cffDict = new Type(strings); |
|
for (var i = 0, ii = dict.length; i < ii; ++i) { |
|
var pair = dict[i]; |
|
var key = pair[0]; |
|
var value = pair[1]; |
|
cffDict.setByKey(key, value); |
|
} |
|
return cffDict; |
|
}, |
|
parseCharString: function CFFParser_parseCharString(state, data, |
|
localSubrIndex, |
|
globalSubrIndex) { |
|
if (state.callDepth > MAX_SUBR_NESTING) { |
|
return false; |
|
} |
|
var stackSize = state.stackSize; |
|
var stack = state.stack; |
|
|
|
var length = data.length; |
|
|
|
for (var j = 0; j < length;) { |
|
var value = data[j++]; |
|
var validationCommand = null; |
|
if (value === 12) { |
|
var q = data[j++]; |
|
if (q === 0) { |
|
// The CFF specification state that the 'dotsection' command |
|
// (12, 0) is deprecated and treated as a no-op, but all Type2 |
|
// charstrings processors should support them. Unfortunately |
|
// the font sanitizer don't. As a workaround the sequence (12, 0) |
|
// is replaced by a useless (0, hmoveto). |
|
data[j - 2] = 139; |
|
data[j - 1] = 22; |
|
stackSize = 0; |
|
} else { |
|
validationCommand = CharstringValidationData12[q]; |
|
} |
|
} else if (value === 28) { // number (16 bit) |
|
stack[stackSize] = ((data[j] << 24) | (data[j + 1] << 16)) >> 16; |
|
j += 2; |
|
stackSize++; |
|
} else if (value === 14) { |
|
if (stackSize >= 4) { |
|
stackSize -= 4; |
|
if (SEAC_ANALYSIS_ENABLED) { |
|
state.seac = stack.slice(stackSize, stackSize + 4); |
|
return false; |
|
} |
|
} |
|
validationCommand = CharstringValidationData[value]; |
|
} else if (value >= 32 && value <= 246) { // number |
|
stack[stackSize] = value - 139; |
|
stackSize++; |
|
} else if (value >= 247 && value <= 254) { // number (+1 bytes) |
|
stack[stackSize] = (value < 251 ? |
|
((value - 247) << 8) + data[j] + 108 : |
|
-((value - 251) << 8) - data[j] - 108); |
|
j++; |
|
stackSize++; |
|
} else if (value === 255) { // number (32 bit) |
|
stack[stackSize] = ((data[j] << 24) | (data[j + 1] << 16) | |
|
(data[j + 2] << 8) | data[j + 3]) / 65536; |
|
j += 4; |
|
stackSize++; |
|
} else if (value === 19 || value === 20) { |
|
state.hints += stackSize >> 1; |
|
// skipping right amount of hints flag data |
|
j += (state.hints + 7) >> 3; |
|
stackSize %= 2; |
|
validationCommand = CharstringValidationData[value]; |
|
} else if (value === 10 || value === 29) { |
|
var subrsIndex; |
|
if (value === 10) { |
|
subrsIndex = localSubrIndex; |
|
} else { |
|
subrsIndex = globalSubrIndex; |
|
} |
|
if (!subrsIndex) { |
|
validationCommand = CharstringValidationData[value]; |
|
warn('Missing subrsIndex for ' + validationCommand.id); |
|
return false; |
|
} |
|
var bias = 32768; |
|
if (subrsIndex.count < 1240) { |
|
bias = 107; |
|
} else if (subrsIndex.count < 33900) { |
|
bias = 1131; |
|
} |
|
var subrNumber = stack[--stackSize] + bias; |
|
if (subrNumber < 0 || subrNumber >= subrsIndex.count) { |
|
validationCommand = CharstringValidationData[value]; |
|
warn('Out of bounds subrIndex for ' + validationCommand.id); |
|
return false; |
|
} |
|
state.stackSize = stackSize; |
|
state.callDepth++; |
|
var valid = this.parseCharString(state, subrsIndex.get(subrNumber), |
|
localSubrIndex, globalSubrIndex); |
|
if (!valid) { |
|
return false; |
|
} |
|
state.callDepth--; |
|
stackSize = state.stackSize; |
|
continue; |
|
} else if (value === 11) { |
|
state.stackSize = stackSize; |
|
return true; |
|
} else { |
|
validationCommand = CharstringValidationData[value]; |
|
} |
|
if (validationCommand) { |
|
if (validationCommand.stem) { |
|
state.hints += stackSize >> 1; |
|
} |
|
if ('min' in validationCommand) { |
|
if (!state.undefStack && stackSize < validationCommand.min) { |
|
warn('Not enough parameters for ' + validationCommand.id + |
|
'; actual: ' + stackSize + |
|
', expected: ' + validationCommand.min); |
|
return false; |
|
} |
|
} |
|
if (state.firstStackClearing && validationCommand.stackClearing) { |
|
state.firstStackClearing = false; |
|
// the optional character width can be found before the first |
|
// stack-clearing command arguments |
|
stackSize -= validationCommand.min; |
|
if (stackSize >= 2 && validationCommand.stem) { |
|
// there are even amount of arguments for stem commands |
|
stackSize %= 2; |
|
} else if (stackSize > 1) { |
|
warn('Found too many parameters for stack-clearing command'); |
|
} |
|
if (stackSize > 0 && stack[stackSize - 1] >= 0) { |
|
state.width = stack[stackSize - 1]; |
|
} |
|
} |
|
if ('stackDelta' in validationCommand) { |
|
if ('stackFn' in validationCommand) { |
|
validationCommand.stackFn(stack, stackSize); |
|
} |
|
stackSize += validationCommand.stackDelta; |
|
} else if (validationCommand.stackClearing) { |
|
stackSize = 0; |
|
} else if (validationCommand.resetStack) { |
|
stackSize = 0; |
|
state.undefStack = false; |
|
} else if (validationCommand.undefStack) { |
|
stackSize = 0; |
|
state.undefStack = true; |
|
state.firstStackClearing = false; |
|
} |
|
} |
|
} |
|
state.stackSize = stackSize; |
|
return true; |
|
}, |
|
parseCharStrings: function CFFParser_parseCharStrings(charStrings, |
|
localSubrIndex, |
|
globalSubrIndex, |
|
fdSelect, |
|
fdArray) { |
|
var seacs = []; |
|
var widths = []; |
|
var count = charStrings.count; |
|
for (var i = 0; i < count; i++) { |
|
var charstring = charStrings.get(i); |
|
var state = { |
|
callDepth: 0, |
|
stackSize: 0, |
|
stack: [], |
|
undefStack: true, |
|
hints: 0, |
|
firstStackClearing: true, |
|
seac: null, |
|
width: null |
|
}; |
|
var valid = true; |
|
var localSubrToUse = null; |
|
if (fdSelect && fdArray.length) { |
|
var fdIndex = fdSelect.getFDIndex(i); |
|
if (fdIndex === -1) { |
|
warn('Glyph index is not in fd select.'); |
|
valid = false; |
|
} |
|
if (fdIndex >= fdArray.length) { |
|
warn('Invalid fd index for glyph index.'); |
|
valid = false; |
|
} |
|
if (valid) { |
|
localSubrToUse = fdArray[fdIndex].privateDict.subrsIndex; |
|
} |
|
} else if (localSubrIndex) { |
|
localSubrToUse = localSubrIndex; |
|
} |
|
if (valid) { |
|
valid = this.parseCharString(state, charstring, localSubrToUse, |
|
globalSubrIndex); |
|
} |
|
if (state.width !== null) { |
|
widths[i] = state.width; |
|
} |
|
if (state.seac !== null) { |
|
seacs[i] = state.seac; |
|
} |
|
if (!valid) { |
|
// resetting invalid charstring to single 'endchar' |
|
charStrings.set(i, new Uint8Array([14])); |
|
} |
|
} |
|
return { charStrings: charStrings, seacs: seacs, widths: widths }; |
|
}, |
|
emptyPrivateDictionary: |
|
function CFFParser_emptyPrivateDictionary(parentDict) { |
|
var privateDict = this.createDict(CFFPrivateDict, [], |
|
parentDict.strings); |
|
parentDict.setByKey(18, [0, 0]); |
|
parentDict.privateDict = privateDict; |
|
}, |
|
parsePrivateDict: function CFFParser_parsePrivateDict(parentDict) { |
|
// no private dict, do nothing |
|
if (!parentDict.hasName('Private')) { |
|
this.emptyPrivateDictionary(parentDict); |
|
return; |
|
} |
|
var privateOffset = parentDict.getByName('Private'); |
|
// make sure the params are formatted correctly |
|
if (!isArray(privateOffset) || privateOffset.length !== 2) { |
|
parentDict.removeByName('Private'); |
|
return; |
|
} |
|
var size = privateOffset[0]; |
|
var offset = privateOffset[1]; |
|
// remove empty dicts or ones that refer to invalid location |
|
if (size === 0 || offset >= this.bytes.length) { |
|
this.emptyPrivateDictionary(parentDict); |
|
return; |
|
} |
|
|
|
var privateDictEnd = offset + size; |
|
var dictData = this.bytes.subarray(offset, privateDictEnd); |
|
var dict = this.parseDict(dictData); |
|
var privateDict = this.createDict(CFFPrivateDict, dict, |
|
parentDict.strings); |
|
parentDict.privateDict = privateDict; |
|
|
|
// Parse the Subrs index also since it's relative to the private dict. |
|
if (!privateDict.getByName('Subrs')) { |
|
return; |
|
} |
|
var subrsOffset = privateDict.getByName('Subrs'); |
|
var relativeOffset = offset + subrsOffset; |
|
// Validate the offset. |
|
if (subrsOffset === 0 || relativeOffset >= this.bytes.length) { |
|
this.emptyPrivateDictionary(parentDict); |
|
return; |
|
} |
|
var subrsIndex = this.parseIndex(relativeOffset); |
|
privateDict.subrsIndex = subrsIndex.obj; |
|
}, |
|
parseCharsets: function CFFParser_parseCharsets(pos, length, strings, cid) { |
|
if (pos === 0) { |
|
return new CFFCharset(true, CFFCharsetPredefinedTypes.ISO_ADOBE, |
|
ISOAdobeCharset); |
|
} else if (pos === 1) { |
|
return new CFFCharset(true, CFFCharsetPredefinedTypes.EXPERT, |
|
ExpertCharset); |
|
} else if (pos === 2) { |
|
return new CFFCharset(true, CFFCharsetPredefinedTypes.EXPERT_SUBSET, |
|
ExpertSubsetCharset); |
|
} |
|
|
|
var bytes = this.bytes; |
|
var start = pos; |
|
var format = bytes[pos++]; |
|
var charset = ['.notdef']; |
|
var id, count, i; |
|
|
|
// subtract 1 for the .notdef glyph |
|
length -= 1; |
|
|
|
switch (format) { |
|
case 0: |
|
for (i = 0; i < length; i++) { |
|
id = (bytes[pos++] << 8) | bytes[pos++]; |
|
charset.push(cid ? id : strings.get(id)); |
|
} |
|
break; |
|
case 1: |
|
while (charset.length <= length) { |
|
id = (bytes[pos++] << 8) | bytes[pos++]; |
|
count = bytes[pos++]; |
|
for (i = 0; i <= count; i++) { |
|
charset.push(cid ? id++ : strings.get(id++)); |
|
} |
|
} |
|
break; |
|
case 2: |
|
while (charset.length <= length) { |
|
id = (bytes[pos++] << 8) | bytes[pos++]; |
|
count = (bytes[pos++] << 8) | bytes[pos++]; |
|
for (i = 0; i <= count; i++) { |
|
charset.push(cid ? id++ : strings.get(id++)); |
|
} |
|
} |
|
break; |
|
default: |
|
error('Unknown charset format'); |
|
} |
|
// Raw won't be needed if we actually compile the charset. |
|
var end = pos; |
|
var raw = bytes.subarray(start, end); |
|
|
|
return new CFFCharset(false, format, charset, raw); |
|
}, |
|
parseEncoding: function CFFParser_parseEncoding(pos, |
|
properties, |
|
strings, |
|
charset) { |
|
var encoding = {}; |
|
var bytes = this.bytes; |
|
var predefined = false; |
|
var hasSupplement = false; |
|
var format, i, ii; |
|
var raw = null; |
|
|
|
function readSupplement() { |
|
var supplementsCount = bytes[pos++]; |
|
for (i = 0; i < supplementsCount; i++) { |
|
var code = bytes[pos++]; |
|
var sid = (bytes[pos++] << 8) + (bytes[pos++] & 0xff); |
|
encoding[code] = charset.indexOf(strings.get(sid)); |
|
} |
|
} |
|
|
|
if (pos === 0 || pos === 1) { |
|
predefined = true; |
|
format = pos; |
|
var baseEncoding = pos ? Encodings.ExpertEncoding : |
|
Encodings.StandardEncoding; |
|
for (i = 0, ii = charset.length; i < ii; i++) { |
|
var index = baseEncoding.indexOf(charset[i]); |
|
if (index !== -1) { |
|
encoding[index] = i; |
|
} |
|
} |
|
} else { |
|
var dataStart = pos; |
|
format = bytes[pos++]; |
|
switch (format & 0x7f) { |
|
case 0: |
|
var glyphsCount = bytes[pos++]; |
|
for (i = 1; i <= glyphsCount; i++) { |
|
encoding[bytes[pos++]] = i; |
|
} |
|
break; |
|
|
|
case 1: |
|
var rangesCount = bytes[pos++]; |
|
var gid = 1; |
|
for (i = 0; i < rangesCount; i++) { |
|
var start = bytes[pos++]; |
|
var left = bytes[pos++]; |
|
for (var j = start; j <= start + left; j++) { |
|
encoding[j] = gid++; |
|
} |
|
} |
|
break; |
|
|
|
default: |
|
error('Unknow encoding format: ' + format + ' in CFF'); |
|
break; |
|
} |
|
var dataEnd = pos; |
|
if (format & 0x80) { |
|
// The font sanitizer does not support CFF encoding with a |
|
// supplement, since the encoding is not really used to map |
|
// between gid to glyph, let's overwrite what is declared in |
|
// the top dictionary to let the sanitizer think the font use |
|
// StandardEncoding, that's a lie but that's ok. |
|
bytes[dataStart] &= 0x7f; |
|
readSupplement(); |
|
hasSupplement = true; |
|
} |
|
raw = bytes.subarray(dataStart, dataEnd); |
|
} |
|
format = format & 0x7f; |
|
return new CFFEncoding(predefined, format, encoding, raw); |
|
}, |
|
parseFDSelect: function CFFParser_parseFDSelect(pos, length) { |
|
var start = pos; |
|
var bytes = this.bytes; |
|
var format = bytes[pos++]; |
|
var fdSelect = []; |
|
var i; |
|
|
|
switch (format) { |
|
case 0: |
|
for (i = 0; i < length; ++i) { |
|
var id = bytes[pos++]; |
|
fdSelect.push(id); |
|
} |
|
break; |
|
case 3: |
|
var rangesCount = (bytes[pos++] << 8) | bytes[pos++]; |
|
for (i = 0; i < rangesCount; ++i) { |
|
var first = (bytes[pos++] << 8) | bytes[pos++]; |
|
var fdIndex = bytes[pos++]; |
|
var next = (bytes[pos] << 8) | bytes[pos + 1]; |
|
for (var j = first; j < next; ++j) { |
|
fdSelect.push(fdIndex); |
|
} |
|
} |
|
// Advance past the sentinel(next). |
|
pos += 2; |
|
break; |
|
default: |
|
error('Unknown fdselect format ' + format); |
|
break; |
|
} |
|
var end = pos; |
|
return new CFFFDSelect(fdSelect, bytes.subarray(start, end)); |
|
} |
|
}; |
|
return CFFParser; |
|
})(); |
|
|
|
// Compact Font Format |
|
var CFF = (function CFFClosure() { |
|
function CFF() { |
|
this.header = null; |
|
this.names = []; |
|
this.topDict = null; |
|
this.strings = new CFFStrings(); |
|
this.globalSubrIndex = null; |
|
|
|
// The following could really be per font, but since we only have one font |
|
// store them here. |
|
this.encoding = null; |
|
this.charset = null; |
|
this.charStrings = null; |
|
this.fdArray = []; |
|
this.fdSelect = null; |
|
|
|
this.isCIDFont = false; |
|
} |
|
return CFF; |
|
})(); |
|
|
|
var CFFHeader = (function CFFHeaderClosure() { |
|
function CFFHeader(major, minor, hdrSize, offSize) { |
|
this.major = major; |
|
this.minor = minor; |
|
this.hdrSize = hdrSize; |
|
this.offSize = offSize; |
|
} |
|
return CFFHeader; |
|
})(); |
|
|
|
var CFFStrings = (function CFFStringsClosure() { |
|
function CFFStrings() { |
|
this.strings = []; |
|
} |
|
CFFStrings.prototype = { |
|
get: function CFFStrings_get(index) { |
|
if (index >= 0 && index <= 390) { |
|
return CFFStandardStrings[index]; |
|
} |
|
if (index - 391 <= this.strings.length) { |
|
return this.strings[index - 391]; |
|
} |
|
return CFFStandardStrings[0]; |
|
}, |
|
add: function CFFStrings_add(value) { |
|
this.strings.push(value); |
|
}, |
|
get count() { |
|
return this.strings.length; |
|
} |
|
}; |
|
return CFFStrings; |
|
})(); |
|
|
|
var CFFIndex = (function CFFIndexClosure() { |
|
function CFFIndex() { |
|
this.objects = []; |
|
this.length = 0; |
|
} |
|
CFFIndex.prototype = { |
|
add: function CFFIndex_add(data) { |
|
this.length += data.length; |
|
this.objects.push(data); |
|
}, |
|
set: function CFFIndex_set(index, data) { |
|
this.length += data.length - this.objects[index].length; |
|
this.objects[index] = data; |
|
}, |
|
get: function CFFIndex_get(index) { |
|
return this.objects[index]; |
|
}, |
|
get count() { |
|
return this.objects.length; |
|
} |
|
}; |
|
return CFFIndex; |
|
})(); |
|
|
|
var CFFDict = (function CFFDictClosure() { |
|
function CFFDict(tables, strings) { |
|
this.keyToNameMap = tables.keyToNameMap; |
|
this.nameToKeyMap = tables.nameToKeyMap; |
|
this.defaults = tables.defaults; |
|
this.types = tables.types; |
|
this.opcodes = tables.opcodes; |
|
this.order = tables.order; |
|
this.strings = strings; |
|
this.values = {}; |
|
} |
|
CFFDict.prototype = { |
|
// value should always be an array |
|
setByKey: function CFFDict_setByKey(key, value) { |
|
if (!(key in this.keyToNameMap)) { |
|
return false; |
|
} |
|
// ignore empty values |
|
if (value.length === 0) { |
|
return true; |
|
} |
|
var type = this.types[key]; |
|
// remove the array wrapping these types of values |
|
if (type === 'num' || type === 'sid' || type === 'offset') { |
|
value = value[0]; |
|
} |
|
this.values[key] = value; |
|
return true; |
|
}, |
|
setByName: function CFFDict_setByName(name, value) { |
|
if (!(name in this.nameToKeyMap)) { |
|
error('Invalid dictionary name "' + name + '"'); |
|
} |
|
this.values[this.nameToKeyMap[name]] = value; |
|
}, |
|
hasName: function CFFDict_hasName(name) { |
|
return this.nameToKeyMap[name] in this.values; |
|
}, |
|
getByName: function CFFDict_getByName(name) { |
|
if (!(name in this.nameToKeyMap)) { |
|
error('Invalid dictionary name "' + name + '"'); |
|
} |
|
var key = this.nameToKeyMap[name]; |
|
if (!(key in this.values)) { |
|
return this.defaults[key]; |
|
} |
|
return this.values[key]; |
|
}, |
|
removeByName: function CFFDict_removeByName(name) { |
|
delete this.values[this.nameToKeyMap[name]]; |
|
} |
|
}; |
|
CFFDict.createTables = function CFFDict_createTables(layout) { |
|
var tables = { |
|
keyToNameMap: {}, |
|
nameToKeyMap: {}, |
|
defaults: {}, |
|
types: {}, |
|
opcodes: {}, |
|
order: [] |
|
}; |
|
for (var i = 0, ii = layout.length; i < ii; ++i) { |
|
var entry = layout[i]; |
|
var key = isArray(entry[0]) ? (entry[0][0] << 8) + entry[0][1] : entry[0]; |
|
tables.keyToNameMap[key] = entry[1]; |
|
tables.nameToKeyMap[entry[1]] = key; |
|
tables.types[key] = entry[2]; |
|
tables.defaults[key] = entry[3]; |
|
tables.opcodes[key] = isArray(entry[0]) ? entry[0] : [entry[0]]; |
|
tables.order.push(key); |
|
} |
|
return tables; |
|
}; |
|
return CFFDict; |
|
})(); |
|
|
|
var CFFTopDict = (function CFFTopDictClosure() { |
|
var layout = [ |
|
[[12, 30], 'ROS', ['sid', 'sid', 'num'], null], |
|
[[12, 20], 'SyntheticBase', 'num', null], |
|
[0, 'version', 'sid', null], |
|
[1, 'Notice', 'sid', null], |
|
[[12, 0], 'Copyright', 'sid', null], |
|
[2, 'FullName', 'sid', null], |
|
[3, 'FamilyName', 'sid', null], |
|
[4, 'Weight', 'sid', null], |
|
[[12, 1], 'isFixedPitch', 'num', 0], |
|
[[12, 2], 'ItalicAngle', 'num', 0], |
|
[[12, 3], 'UnderlinePosition', 'num', -100], |
|
[[12, 4], 'UnderlineThickness', 'num', 50], |
|
[[12, 5], 'PaintType', 'num', 0], |
|
[[12, 6], 'CharstringType', 'num', 2], |
|
[[12, 7], 'FontMatrix', ['num', 'num', 'num', 'num', 'num', 'num'], |
|
[0.001, 0, 0, 0.001, 0, 0]], |
|
[13, 'UniqueID', 'num', null], |
|
[5, 'FontBBox', ['num', 'num', 'num', 'num'], [0, 0, 0, 0]], |
|
[[12, 8], 'StrokeWidth', 'num', 0], |
|
[14, 'XUID', 'array', null], |
|
[15, 'charset', 'offset', 0], |
|
[16, 'Encoding', 'offset', 0], |
|
[17, 'CharStrings', 'offset', 0], |
|
[18, 'Private', ['offset', 'offset'], null], |
|
[[12, 21], 'PostScript', 'sid', null], |
|
[[12, 22], 'BaseFontName', 'sid', null], |
|
[[12, 23], 'BaseFontBlend', 'delta', null], |
|
[[12, 31], 'CIDFontVersion', 'num', 0], |
|
[[12, 32], 'CIDFontRevision', 'num', 0], |
|
[[12, 33], 'CIDFontType', 'num', 0], |
|
[[12, 34], 'CIDCount', 'num', 8720], |
|
[[12, 35], 'UIDBase', 'num', null], |
|
// XXX: CID Fonts on DirectWrite 6.1 only seem to work if FDSelect comes |
|
// before FDArray. |
|
[[12, 37], 'FDSelect', 'offset', null], |
|
[[12, 36], 'FDArray', 'offset', null], |
|
[[12, 38], 'FontName', 'sid', null] |
|
]; |
|
var tables = null; |
|
function CFFTopDict(strings) { |
|
if (tables === null) { |
|
tables = CFFDict.createTables(layout); |
|
} |
|
CFFDict.call(this, tables, strings); |
|
this.privateDict = null; |
|
} |
|
CFFTopDict.prototype = Object.create(CFFDict.prototype); |
|
return CFFTopDict; |
|
})(); |
|
|
|
var CFFPrivateDict = (function CFFPrivateDictClosure() { |
|
var layout = [ |
|
[6, 'BlueValues', 'delta', null], |
|
[7, 'OtherBlues', 'delta', null], |
|
[8, 'FamilyBlues', 'delta', null], |
|
[9, 'FamilyOtherBlues', 'delta', null], |
|
[[12, 9], 'BlueScale', 'num', 0.039625], |
|
[[12, 10], 'BlueShift', 'num', 7], |
|
[[12, 11], 'BlueFuzz', 'num', 1], |
|
[10, 'StdHW', 'num', null], |
|
[11, 'StdVW', 'num', null], |
|
[[12, 12], 'StemSnapH', 'delta', null], |
|
[[12, 13], 'StemSnapV', 'delta', null], |
|
[[12, 14], 'ForceBold', 'num', 0], |
|
[[12, 17], 'LanguageGroup', 'num', 0], |
|
[[12, 18], 'ExpansionFactor', 'num', 0.06], |
|
[[12, 19], 'initialRandomSeed', 'num', 0], |
|
[20, 'defaultWidthX', 'num', 0], |
|
[21, 'nominalWidthX', 'num', 0], |
|
[19, 'Subrs', 'offset', null] |
|
]; |
|
var tables = null; |
|
function CFFPrivateDict(strings) { |
|
if (tables === null) { |
|
tables = CFFDict.createTables(layout); |
|
} |
|
CFFDict.call(this, tables, strings); |
|
this.subrsIndex = null; |
|
} |
|
CFFPrivateDict.prototype = Object.create(CFFDict.prototype); |
|
return CFFPrivateDict; |
|
})(); |
|
|
|
var CFFCharsetPredefinedTypes = { |
|
ISO_ADOBE: 0, |
|
EXPERT: 1, |
|
EXPERT_SUBSET: 2 |
|
}; |
|
var CFFCharset = (function CFFCharsetClosure() { |
|
function CFFCharset(predefined, format, charset, raw) { |
|
this.predefined = predefined; |
|
this.format = format; |
|
this.charset = charset; |
|
this.raw = raw; |
|
} |
|
return CFFCharset; |
|
})(); |
|
|
|
var CFFEncoding = (function CFFEncodingClosure() { |
|
function CFFEncoding(predefined, format, encoding, raw) { |
|
this.predefined = predefined; |
|
this.format = format; |
|
this.encoding = encoding; |
|
this.raw = raw; |
|
} |
|
return CFFEncoding; |
|
})(); |
|
|
|
var CFFFDSelect = (function CFFFDSelectClosure() { |
|
function CFFFDSelect(fdSelect, raw) { |
|
this.fdSelect = fdSelect; |
|
this.raw = raw; |
|
} |
|
CFFFDSelect.prototype = { |
|
getFDIndex: function CFFFDSelect_get(glyphIndex) { |
|
if (glyphIndex < 0 || glyphIndex >= this.fdSelect.length) { |
|
return -1; |
|
} |
|
return this.fdSelect[glyphIndex]; |
|
} |
|
}; |
|
return CFFFDSelect; |
|
})(); |
|
|
|
// Helper class to keep track of where an offset is within the data and helps |
|
// filling in that offset once it's known. |
|
var CFFOffsetTracker = (function CFFOffsetTrackerClosure() { |
|
function CFFOffsetTracker() { |
|
this.offsets = {}; |
|
} |
|
CFFOffsetTracker.prototype = { |
|
isTracking: function CFFOffsetTracker_isTracking(key) { |
|
return key in this.offsets; |
|
}, |
|
track: function CFFOffsetTracker_track(key, location) { |
|
if (key in this.offsets) { |
|
error('Already tracking location of ' + key); |
|
} |
|
this.offsets[key] = location; |
|
}, |
|
offset: function CFFOffsetTracker_offset(value) { |
|
for (var key in this.offsets) { |
|
this.offsets[key] += value; |
|
} |
|
}, |
|
setEntryLocation: function CFFOffsetTracker_setEntryLocation(key, |
|
values, |
|
output) { |
|
if (!(key in this.offsets)) { |
|
error('Not tracking location of ' + key); |
|
} |
|
var data = output.data; |
|
var dataOffset = this.offsets[key]; |
|
var size = 5; |
|
for (var i = 0, ii = values.length; i < ii; ++i) { |
|
var offset0 = i * size + dataOffset; |
|
var offset1 = offset0 + 1; |
|
var offset2 = offset0 + 2; |
|
var offset3 = offset0 + 3; |
|
var offset4 = offset0 + 4; |
|
// It's easy to screw up offsets so perform this sanity check. |
|
if (data[offset0] !== 0x1d || data[offset1] !== 0 || |
|
data[offset2] !== 0 || data[offset3] !== 0 || data[offset4] !== 0) { |
|
error('writing to an offset that is not empty'); |
|
} |
|
var value = values[i]; |
|
data[offset0] = 0x1d; |
|
data[offset1] = (value >> 24) & 0xFF; |
|
data[offset2] = (value >> 16) & 0xFF; |
|
data[offset3] = (value >> 8) & 0xFF; |
|
data[offset4] = value & 0xFF; |
|
} |
|
} |
|
}; |
|
return CFFOffsetTracker; |
|
})(); |
|
|
|
// Takes a CFF and converts it to the binary representation. |
|
var CFFCompiler = (function CFFCompilerClosure() { |
|
function CFFCompiler(cff) { |
|
this.cff = cff; |
|
} |
|
CFFCompiler.prototype = { |
|
compile: function CFFCompiler_compile() { |
|
var cff = this.cff; |
|
var output = { |
|
data: [], |
|
length: 0, |
|
add: function CFFCompiler_add(data) { |
|
this.data = this.data.concat(data); |
|
this.length = this.data.length; |
|
} |
|
}; |
|
|
|
// Compile the five entries that must be in order. |
|
var header = this.compileHeader(cff.header); |
|
output.add(header); |
|
|
|
var nameIndex = this.compileNameIndex(cff.names); |
|
output.add(nameIndex); |
|
|
|
if (cff.isCIDFont) { |
|
// The spec is unclear on how font matrices should relate to each other |
|
// when there is one in the main top dict and the sub top dicts. |
|
// Windows handles this differently than linux and osx so we have to |
|
// normalize to work on all. |
|
// Rules based off of some mailing list discussions: |
|
// - If main font has a matrix and subfont doesn't, use the main matrix. |
|
// - If no main font matrix and there is a subfont matrix, use the |
|
// subfont matrix. |
|
// - If both have matrices, concat together. |
|
// - If neither have matrices, use default. |
|
// To make this work on all platforms we move the top matrix into each |
|
// sub top dict and concat if necessary. |
|
if (cff.topDict.hasName('FontMatrix')) { |
|
var base = cff.topDict.getByName('FontMatrix'); |
|
cff.topDict.removeByName('FontMatrix'); |
|
for (var i = 0, ii = cff.fdArray.length; i < ii; i++) { |
|
var subDict = cff.fdArray[i]; |
|
var matrix = base.slice(0); |
|
if (subDict.hasName('FontMatrix')) { |
|
matrix = Util.transform(matrix, subDict.getByName('FontMatrix')); |
|
} |
|
subDict.setByName('FontMatrix', matrix); |
|
} |
|
} |
|
} |
|
|
|
var compiled = this.compileTopDicts([cff.topDict], |
|
output.length, |
|
cff.isCIDFont); |
|
output.add(compiled.output); |
|
var topDictTracker = compiled.trackers[0]; |
|
|
|
var stringIndex = this.compileStringIndex(cff.strings.strings); |
|
output.add(stringIndex); |
|
|
|
var globalSubrIndex = this.compileIndex(cff.globalSubrIndex); |
|
output.add(globalSubrIndex); |
|
|
|
// Now start on the other entries that have no specfic order. |
|
if (cff.encoding && cff.topDict.hasName('Encoding')) { |
|
if (cff.encoding.predefined) { |
|
topDictTracker.setEntryLocation('Encoding', [cff.encoding.format], |
|
output); |
|
} else { |
|
var encoding = this.compileEncoding(cff.encoding); |
|
topDictTracker.setEntryLocation('Encoding', [output.length], output); |
|
output.add(encoding); |
|
} |
|
} |
|
|
|
if (cff.charset && cff.topDict.hasName('charset')) { |
|
if (cff.charset.predefined) { |
|
topDictTracker.setEntryLocation('charset', [cff.charset.format], |
|
output); |
|
} else { |
|
var charset = this.compileCharset(cff.charset); |
|
topDictTracker.setEntryLocation('charset', [output.length], output); |
|
output.add(charset); |
|
} |
|
} |
|
|
|
var charStrings = this.compileCharStrings(cff.charStrings); |
|
topDictTracker.setEntryLocation('CharStrings', [output.length], output); |
|
output.add(charStrings); |
|
|
|
if (cff.isCIDFont) { |
|
// For some reason FDSelect must be in front of FDArray on windows. OSX |
|
// and linux don't seem to care. |
|
topDictTracker.setEntryLocation('FDSelect', [output.length], output); |
|
var fdSelect = this.compileFDSelect(cff.fdSelect.raw); |
|
output.add(fdSelect); |
|
// It is unclear if the sub font dictionary can have CID related |
|
// dictionary keys, but the sanitizer doesn't like them so remove them. |
|
compiled = this.compileTopDicts(cff.fdArray, output.length, true); |
|
topDictTracker.setEntryLocation('FDArray', [output.length], output); |
|
output.add(compiled.output); |
|
var fontDictTrackers = compiled.trackers; |
|
|
|
this.compilePrivateDicts(cff.fdArray, fontDictTrackers, output); |
|
} |
|
|
|
this.compilePrivateDicts([cff.topDict], [topDictTracker], output); |
|
|
|
// If the font data ends with INDEX whose object data is zero-length, |
|
// the sanitizer will bail out. Add a dummy byte to avoid that. |
|
output.add([0]); |
|
|
|
return output.data; |
|
}, |
|
encodeNumber: function CFFCompiler_encodeNumber(value) { |
|
if (parseFloat(value) === parseInt(value, 10) && !isNaN(value)) { // isInt |
|
return this.encodeInteger(value); |
|
} else { |
|
return this.encodeFloat(value); |
|
} |
|
}, |
|
encodeFloat: function CFFCompiler_encodeFloat(num) { |
|
var value = num.toString(); |
|
|
|
// rounding inaccurate doubles |
|
var m = /\.(\d*?)(?:9{5,20}|0{5,20})\d{0,2}(?:e(.+)|$)/.exec(value); |
|
if (m) { |
|
var epsilon = parseFloat('1e' + ((m[2] ? +m[2] : 0) + m[1].length)); |
|
value = (Math.round(num * epsilon) / epsilon).toString(); |
|
} |
|
|
|
var nibbles = ''; |
|
var i, ii; |
|
for (i = 0, ii = value.length; i < ii; ++i) { |
|
var a = value[i]; |
|
if (a === 'e') { |
|
nibbles += value[++i] === '-' ? 'c' : 'b'; |
|
} else if (a === '.') { |
|
nibbles += 'a'; |
|
} else if (a === '-') { |
|
nibbles += 'e'; |
|
} else { |
|
nibbles += a; |
|
} |
|
} |
|
nibbles += (nibbles.length & 1) ? 'f' : 'ff'; |
|
var out = [30]; |
|
for (i = 0, ii = nibbles.length; i < ii; i += 2) { |
|
out.push(parseInt(nibbles.substr(i, 2), 16)); |
|
} |
|
return out; |
|
}, |
|
encodeInteger: function CFFCompiler_encodeInteger(value) { |
|
var code; |
|
if (value >= -107 && value <= 107) { |
|
code = [value + 139]; |
|
} else if (value >= 108 && value <= 1131) { |
|
value = [value - 108]; |
|
code = [(value >> 8) + 247, value & 0xFF]; |
|
} else if (value >= -1131 && value <= -108) { |
|
value = -value - 108; |
|
code = [(value >> 8) + 251, value & 0xFF]; |
|
} else if (value >= -32768 && value <= 32767) { |
|
code = [0x1c, (value >> 8) & 0xFF, value & 0xFF]; |
|
} else { |
|
code = [0x1d, |
|
(value >> 24) & 0xFF, |
|
(value >> 16) & 0xFF, |
|
(value >> 8) & 0xFF, |
|
value & 0xFF]; |
|
} |
|
return code; |
|
}, |
|
compileHeader: function CFFCompiler_compileHeader(header) { |
|
return [ |
|
header.major, |
|
header.minor, |
|
header.hdrSize, |
|
header.offSize |
|
]; |
|
}, |
|
compileNameIndex: function CFFCompiler_compileNameIndex(names) { |
|
var nameIndex = new CFFIndex(); |
|
for (var i = 0, ii = names.length; i < ii; ++i) { |
|
nameIndex.add(stringToBytes(names[i])); |
|
} |
|
return this.compileIndex(nameIndex); |
|
}, |
|
compileTopDicts: function CFFCompiler_compileTopDicts(dicts, |
|
length, |
|
removeCidKeys) { |
|
var fontDictTrackers = []; |
|
var fdArrayIndex = new CFFIndex(); |
|
for (var i = 0, ii = dicts.length; i < ii; ++i) { |
|
var fontDict = dicts[i]; |
|
if (removeCidKeys) { |
|
fontDict.removeByName('CIDFontVersion'); |
|
fontDict.removeByName('CIDFontRevision'); |
|
fontDict.removeByName('CIDFontType'); |
|
fontDict.removeByName('CIDCount'); |
|
fontDict.removeByName('UIDBase'); |
|
} |
|
var fontDictTracker = new CFFOffsetTracker(); |
|
var fontDictData = this.compileDict(fontDict, fontDictTracker); |
|
fontDictTrackers.push(fontDictTracker); |
|
fdArrayIndex.add(fontDictData); |
|
fontDictTracker.offset(length); |
|
} |
|
fdArrayIndex = this.compileIndex(fdArrayIndex, fontDictTrackers); |
|
return { |
|
trackers: fontDictTrackers, |
|
output: fdArrayIndex |
|
}; |
|
}, |
|
compilePrivateDicts: function CFFCompiler_compilePrivateDicts(dicts, |
|
trackers, |
|
output) { |
|
for (var i = 0, ii = dicts.length; i < ii; ++i) { |
|
var fontDict = dicts[i]; |
|
assert(fontDict.privateDict && fontDict.hasName('Private'), |
|
'There must be an private dictionary.'); |
|
var privateDict = fontDict.privateDict; |
|
var privateDictTracker = new CFFOffsetTracker(); |
|
var privateDictData = this.compileDict(privateDict, privateDictTracker); |
|
|
|
var outputLength = output.length; |
|
privateDictTracker.offset(outputLength); |
|
if (!privateDictData.length) { |
|
// The private dictionary was empty, set the output length to zero to |
|
// ensure the offset length isn't out of bounds in the eyes of the |
|
// sanitizer. |
|
outputLength = 0; |
|
} |
|
|
|
trackers[i].setEntryLocation('Private', |
|
[privateDictData.length, outputLength], |
|
output); |
|
output.add(privateDictData); |
|
|
|
if (privateDict.subrsIndex && privateDict.hasName('Subrs')) { |
|
var subrs = this.compileIndex(privateDict.subrsIndex); |
|
privateDictTracker.setEntryLocation('Subrs', [privateDictData.length], |
|
output); |
|
output.add(subrs); |
|
} |
|
} |
|
}, |
|
compileDict: function CFFCompiler_compileDict(dict, offsetTracker) { |
|
var out = []; |
|
// The dictionary keys must be in a certain order. |
|
var order = dict.order; |
|
for (var i = 0; i < order.length; ++i) { |
|
var key = order[i]; |
|
if (!(key in dict.values)) { |
|
continue; |
|
} |
|
var values = dict.values[key]; |
|
var types = dict.types[key]; |
|
if (!isArray(types)) { |
|
types = [types]; |
|
} |
|
if (!isArray(values)) { |
|
values = [values]; |
|
} |
|
|
|
// Remove any empty dict values. |
|
if (values.length === 0) { |
|
continue; |
|
} |
|
|
|
for (var j = 0, jj = types.length; j < jj; ++j) { |
|
var type = types[j]; |
|
var value = values[j]; |
|
switch (type) { |
|
case 'num': |
|
case 'sid': |
|
out = out.concat(this.encodeNumber(value)); |
|
break; |
|
case 'offset': |
|
// For offsets we just insert a 32bit integer so we don't have to |
|
// deal with figuring out the length of the offset when it gets |
|
// replaced later on by the compiler. |
|
var name = dict.keyToNameMap[key]; |
|
// Some offsets have the offset and the length, so just record the |
|
// position of the first one. |
|
if (!offsetTracker.isTracking(name)) { |
|
offsetTracker.track(name, out.length); |
|
} |
|
out = out.concat([0x1d, 0, 0, 0, 0]); |
|
break; |
|
case 'array': |
|
case 'delta': |
|
out = out.concat(this.encodeNumber(value)); |
|
for (var k = 1, kk = values.length; k < kk; ++k) { |
|
out = out.concat(this.encodeNumber(values[k])); |
|
} |
|
break; |
|
default: |
|
error('Unknown data type of ' + type); |
|
break; |
|
} |
|
} |
|
out = out.concat(dict.opcodes[key]); |
|
} |
|
return out; |
|
}, |
|
compileStringIndex: function CFFCompiler_compileStringIndex(strings) { |
|
var stringIndex = new CFFIndex(); |
|
for (var i = 0, ii = strings.length; i < ii; ++i) { |
|
stringIndex.add(stringToBytes(strings[i])); |
|
} |
|
return this.compileIndex(stringIndex); |
|
}, |
|
compileGlobalSubrIndex: function CFFCompiler_compileGlobalSubrIndex() { |
|
var globalSubrIndex = this.cff.globalSubrIndex; |
|
this.out.writeByteArray(this.compileIndex(globalSubrIndex)); |
|
}, |
|
compileCharStrings: function CFFCompiler_compileCharStrings(charStrings) { |
|
return this.compileIndex(charStrings); |
|
}, |
|
compileCharset: function CFFCompiler_compileCharset(charset) { |
|
return this.compileTypedArray(charset.raw); |
|
}, |
|
compileEncoding: function CFFCompiler_compileEncoding(encoding) { |
|
return this.compileTypedArray(encoding.raw); |
|
}, |
|
compileFDSelect: function CFFCompiler_compileFDSelect(fdSelect) { |
|
return this.compileTypedArray(fdSelect); |
|
}, |
|
compileTypedArray: function CFFCompiler_compileTypedArray(data) { |
|
var out = []; |
|
for (var i = 0, ii = data.length; i < ii; ++i) { |
|
out[i] = data[i]; |
|
} |
|
return out; |
|
}, |
|
compileIndex: function CFFCompiler_compileIndex(index, trackers) { |
|
trackers = trackers || []; |
|
var objects = index.objects; |
|
// First 2 bytes contains the number of objects contained into this index |
|
var count = objects.length; |
|
|
|
// If there is no object, just create an index. This technically |
|
// should just be [0, 0] but OTS has an issue with that. |
|
if (count === 0) { |
|
return [0, 0, 0]; |
|
} |
|
|
|
var data = [(count >> 8) & 0xFF, count & 0xff]; |
|
|
|
var lastOffset = 1, i; |
|
for (i = 0; i < count; ++i) { |
|
lastOffset += objects[i].length; |
|
} |
|
|
|
var offsetSize; |
|
if (lastOffset < 0x100) { |
|
offsetSize = 1; |
|
} else if (lastOffset < 0x10000) { |
|
offsetSize = 2; |
|
} else if (lastOffset < 0x1000000) { |
|
offsetSize = 3; |
|
} else { |
|
offsetSize = 4; |
|
} |
|
|
|
// Next byte contains the offset size use to reference object in the file |
|
data.push(offsetSize); |
|
|
|
// Add another offset after this one because we need a new offset |
|
var relativeOffset = 1; |
|
for (i = 0; i < count + 1; i++) { |
|
if (offsetSize === 1) { |
|
data.push(relativeOffset & 0xFF); |
|
} else if (offsetSize === 2) { |
|
data.push((relativeOffset >> 8) & 0xFF, |
|
relativeOffset & 0xFF); |
|
} else if (offsetSize === 3) { |
|
data.push((relativeOffset >> 16) & 0xFF, |
|
(relativeOffset >> 8) & 0xFF, |
|
relativeOffset & 0xFF); |
|
} else { |
|
data.push((relativeOffset >>> 24) & 0xFF, |
|
(relativeOffset >> 16) & 0xFF, |
|
(relativeOffset >> 8) & 0xFF, |
|
relativeOffset & 0xFF); |
|
} |
|
|
|
if (objects[i]) { |
|
relativeOffset += objects[i].length; |
|
} |
|
} |
|
|
|
for (i = 0; i < count; i++) { |
|
// Notify the tracker where the object will be offset in the data. |
|
if (trackers[i]) { |
|
trackers[i].offset(data.length); |
|
} |
|
for (var j = 0, jj = objects[i].length; j < jj; j++) { |
|
data.push(objects[i][j]); |
|
} |
|
} |
|
return data; |
|
} |
|
}; |
|
return CFFCompiler; |
|
})(); |
|
|
|
function _enableSeacAnalysis(enabled) { |
|
exports.SEAC_ANALYSIS_ENABLED = SEAC_ANALYSIS_ENABLED = enabled; |
|
} |
|
|
|
// Workaround for seac on Windows. |
|
(function checkSeacSupport() { |
|
if (typeof navigator !== 'undefined' && /Windows/.test(navigator.userAgent)) { |
|
SEAC_ANALYSIS_ENABLED = true; |
|
} |
|
})(); |
|
|
|
// Workaround for Private Use Area characters in Chrome on Windows |
|
// http://code.google.com/p/chromium/issues/detail?id=122465 |
|
// https://github.com/mozilla/pdf.js/issues/1689 |
|
(function checkChromeWindows() { |
|
if (typeof navigator !== 'undefined' && |
|
/Windows.*Chrome/.test(navigator.userAgent)) { |
|
SKIP_PRIVATE_USE_RANGE_F000_TO_F01F = true; |
|
} |
|
})(); |
|
|
|
exports.SEAC_ANALYSIS_ENABLED = SEAC_ANALYSIS_ENABLED; |
|
exports.CFFCompiler = CFFCompiler; |
|
exports.CFFIndex = CFFIndex; |
|
exports.CFFParser = CFFParser; |
|
exports.CFFStrings = CFFStrings; |
|
exports.Encodings = Encodings; |
|
exports.ErrorFont = ErrorFont; |
|
exports.FontFlags = FontFlags; |
|
exports.Font = Font; |
|
exports.IdentityToUnicodeMap = IdentityToUnicodeMap; |
|
exports.NormalizedUnicodes = NormalizedUnicodes; |
|
exports.ToUnicodeMap = ToUnicodeMap; |
|
exports.Type1Parser = Type1Parser; |
|
exports.getFontType = getFontType; |
|
exports.reverseIfRtl = reverseIfRtl; |
|
exports.serifFonts = serifFonts; |
|
exports.symbolsFonts = symbolsFonts; |
|
exports.stdFontMap = stdFontMap; |
|
exports._enableSeacAnalysis = _enableSeacAnalysis; |
|
|
|
// TODO refactor to remove cyclic dependency on font_renderer.js |
|
coreFontRenderer._setCoreFonts(exports); |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreFunction = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCorePsParser); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, corePsParser) { |
|
|
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isArray = sharedUtil.isArray; |
|
var isBool = sharedUtil.isBool; |
|
var isDict = corePrimitives.isDict; |
|
var isStream = corePrimitives.isStream; |
|
var PostScriptLexer = corePsParser.PostScriptLexer; |
|
var PostScriptParser = corePsParser.PostScriptParser; |
|
|
|
var PDFFunction = (function PDFFunctionClosure() { |
|
var CONSTRUCT_SAMPLED = 0; |
|
var CONSTRUCT_INTERPOLATED = 2; |
|
var CONSTRUCT_STICHED = 3; |
|
var CONSTRUCT_POSTSCRIPT = 4; |
|
|
|
return { |
|
getSampleArray: function PDFFunction_getSampleArray(size, outputSize, bps, |
|
str) { |
|
var i, ii; |
|
var length = 1; |
|
for (i = 0, ii = size.length; i < ii; i++) { |
|
length *= size[i]; |
|
} |
|
length *= outputSize; |
|
|
|
var array = new Array(length); |
|
var codeSize = 0; |
|
var codeBuf = 0; |
|
// 32 is a valid bps so shifting won't work |
|
var sampleMul = 1.0 / (Math.pow(2.0, bps) - 1); |
|
|
|
var strBytes = str.getBytes((length * bps + 7) / 8); |
|
var strIdx = 0; |
|
for (i = 0; i < length; i++) { |
|
while (codeSize < bps) { |
|
codeBuf <<= 8; |
|
codeBuf |= strBytes[strIdx++]; |
|
codeSize += 8; |
|
} |
|
codeSize -= bps; |
|
array[i] = (codeBuf >> codeSize) * sampleMul; |
|
codeBuf &= (1 << codeSize) - 1; |
|
} |
|
return array; |
|
}, |
|
|
|
getIR: function PDFFunction_getIR(xref, fn) { |
|
var dict = fn.dict; |
|
if (!dict) { |
|
dict = fn; |
|
} |
|
|
|
var types = [this.constructSampled, |
|
null, |
|
this.constructInterpolated, |
|
this.constructStiched, |
|
this.constructPostScript]; |
|
|
|
var typeNum = dict.get('FunctionType'); |
|
var typeFn = types[typeNum]; |
|
if (!typeFn) { |
|
error('Unknown type of function'); |
|
} |
|
|
|
return typeFn.call(this, fn, dict, xref); |
|
}, |
|
|
|
fromIR: function PDFFunction_fromIR(IR) { |
|
var type = IR[0]; |
|
switch (type) { |
|
case CONSTRUCT_SAMPLED: |
|
return this.constructSampledFromIR(IR); |
|
case CONSTRUCT_INTERPOLATED: |
|
return this.constructInterpolatedFromIR(IR); |
|
case CONSTRUCT_STICHED: |
|
return this.constructStichedFromIR(IR); |
|
//case CONSTRUCT_POSTSCRIPT: |
|
default: |
|
return this.constructPostScriptFromIR(IR); |
|
} |
|
}, |
|
|
|
parse: function PDFFunction_parse(xref, fn) { |
|
var IR = this.getIR(xref, fn); |
|
return this.fromIR(IR); |
|
}, |
|
|
|
parseArray: function PDFFunction_parseArray(xref, fnObj) { |
|
if (!isArray(fnObj)) { |
|
// not an array -- parsing as regular function |
|
return this.parse(xref, fnObj); |
|
} |
|
|
|
var fnArray = []; |
|
for (var j = 0, jj = fnObj.length; j < jj; j++) { |
|
var obj = xref.fetchIfRef(fnObj[j]); |
|
fnArray.push(PDFFunction.parse(xref, obj)); |
|
} |
|
return function (src, srcOffset, dest, destOffset) { |
|
for (var i = 0, ii = fnArray.length; i < ii; i++) { |
|
fnArray[i](src, srcOffset, dest, destOffset + i); |
|
} |
|
}; |
|
}, |
|
|
|
constructSampled: function PDFFunction_constructSampled(str, dict) { |
|
function toMultiArray(arr) { |
|
var inputLength = arr.length; |
|
var out = []; |
|
var index = 0; |
|
for (var i = 0; i < inputLength; i += 2) { |
|
out[index] = [arr[i], arr[i + 1]]; |
|
++index; |
|
} |
|
return out; |
|
} |
|
var domain = dict.get('Domain'); |
|
var range = dict.get('Range'); |
|
|
|
if (!domain || !range) { |
|
error('No domain or range'); |
|
} |
|
|
|
var inputSize = domain.length / 2; |
|
var outputSize = range.length / 2; |
|
|
|
domain = toMultiArray(domain); |
|
range = toMultiArray(range); |
|
|
|
var size = dict.get('Size'); |
|
var bps = dict.get('BitsPerSample'); |
|
var order = dict.get('Order') || 1; |
|
if (order !== 1) { |
|
// No description how cubic spline interpolation works in PDF32000:2008 |
|
// As in poppler, ignoring order, linear interpolation may work as good |
|
info('No support for cubic spline interpolation: ' + order); |
|
} |
|
|
|
var encode = dict.get('Encode'); |
|
if (!encode) { |
|
encode = []; |
|
for (var i = 0; i < inputSize; ++i) { |
|
encode.push(0); |
|
encode.push(size[i] - 1); |
|
} |
|
} |
|
encode = toMultiArray(encode); |
|
|
|
var decode = dict.get('Decode'); |
|
if (!decode) { |
|
decode = range; |
|
} else { |
|
decode = toMultiArray(decode); |
|
} |
|
|
|
var samples = this.getSampleArray(size, outputSize, bps, str); |
|
|
|
return [ |
|
CONSTRUCT_SAMPLED, inputSize, domain, encode, decode, samples, size, |
|
outputSize, Math.pow(2, bps) - 1, range |
|
]; |
|
}, |
|
|
|
constructSampledFromIR: function PDFFunction_constructSampledFromIR(IR) { |
|
// See chapter 3, page 109 of the PDF reference |
|
function interpolate(x, xmin, xmax, ymin, ymax) { |
|
return ymin + ((x - xmin) * ((ymax - ymin) / (xmax - xmin))); |
|
} |
|
|
|
return function constructSampledFromIRResult(src, srcOffset, |
|
dest, destOffset) { |
|
// See chapter 3, page 110 of the PDF reference. |
|
var m = IR[1]; |
|
var domain = IR[2]; |
|
var encode = IR[3]; |
|
var decode = IR[4]; |
|
var samples = IR[5]; |
|
var size = IR[6]; |
|
var n = IR[7]; |
|
//var mask = IR[8]; |
|
var range = IR[9]; |
|
|
|
// Building the cube vertices: its part and sample index |
|
// http://rjwagner49.com/Mathematics/Interpolation.pdf |
|
var cubeVertices = 1 << m; |
|
var cubeN = new Float64Array(cubeVertices); |
|
var cubeVertex = new Uint32Array(cubeVertices); |
|
var i, j; |
|
for (j = 0; j < cubeVertices; j++) { |
|
cubeN[j] = 1; |
|
} |
|
|
|
var k = n, pos = 1; |
|
// Map x_i to y_j for 0 <= i < m using the sampled function. |
|
for (i = 0; i < m; ++i) { |
|
// x_i' = min(max(x_i, Domain_2i), Domain_2i+1) |
|
var domain_2i = domain[i][0]; |
|
var domain_2i_1 = domain[i][1]; |
|
var xi = Math.min(Math.max(src[srcOffset +i], domain_2i), |
|
domain_2i_1); |
|
|
|
// e_i = Interpolate(x_i', Domain_2i, Domain_2i+1, |
|
// Encode_2i, Encode_2i+1) |
|
var e = interpolate(xi, domain_2i, domain_2i_1, |
|
encode[i][0], encode[i][1]); |
|
|
|
// e_i' = min(max(e_i, 0), Size_i - 1) |
|
var size_i = size[i]; |
|
e = Math.min(Math.max(e, 0), size_i - 1); |
|
|
|
// Adjusting the cube: N and vertex sample index |
|
var e0 = e < size_i - 1 ? Math.floor(e) : e - 1; // e1 = e0 + 1; |
|
var n0 = e0 + 1 - e; // (e1 - e) / (e1 - e0); |
|
var n1 = e - e0; // (e - e0) / (e1 - e0); |
|
var offset0 = e0 * k; |
|
var offset1 = offset0 + k; // e1 * k |
|
for (j = 0; j < cubeVertices; j++) { |
|
if (j & pos) { |
|
cubeN[j] *= n1; |
|
cubeVertex[j] += offset1; |
|
} else { |
|
cubeN[j] *= n0; |
|
cubeVertex[j] += offset0; |
|
} |
|
} |
|
|
|
k *= size_i; |
|
pos <<= 1; |
|
} |
|
|
|
for (j = 0; j < n; ++j) { |
|
// Sum all cube vertices' samples portions |
|
var rj = 0; |
|
for (i = 0; i < cubeVertices; i++) { |
|
rj += samples[cubeVertex[i] + j] * cubeN[i]; |
|
} |
|
|
|
// r_j' = Interpolate(r_j, 0, 2^BitsPerSample - 1, |
|
// Decode_2j, Decode_2j+1) |
|
rj = interpolate(rj, 0, 1, decode[j][0], decode[j][1]); |
|
|
|
// y_j = min(max(r_j, range_2j), range_2j+1) |
|
dest[destOffset + j] = Math.min(Math.max(rj, range[j][0]), |
|
range[j][1]); |
|
} |
|
}; |
|
}, |
|
|
|
constructInterpolated: function PDFFunction_constructInterpolated(str, |
|
dict) { |
|
var c0 = dict.get('C0') || [0]; |
|
var c1 = dict.get('C1') || [1]; |
|
var n = dict.get('N'); |
|
|
|
if (!isArray(c0) || !isArray(c1)) { |
|
error('Illegal dictionary for interpolated function'); |
|
} |
|
|
|
var length = c0.length; |
|
var diff = []; |
|
for (var i = 0; i < length; ++i) { |
|
diff.push(c1[i] - c0[i]); |
|
} |
|
|
|
return [CONSTRUCT_INTERPOLATED, c0, diff, n]; |
|
}, |
|
|
|
constructInterpolatedFromIR: |
|
function PDFFunction_constructInterpolatedFromIR(IR) { |
|
var c0 = IR[1]; |
|
var diff = IR[2]; |
|
var n = IR[3]; |
|
|
|
var length = diff.length; |
|
|
|
return function constructInterpolatedFromIRResult(src, srcOffset, |
|
dest, destOffset) { |
|
var x = n === 1 ? src[srcOffset] : Math.pow(src[srcOffset], n); |
|
|
|
for (var j = 0; j < length; ++j) { |
|
dest[destOffset + j] = c0[j] + (x * diff[j]); |
|
} |
|
}; |
|
}, |
|
|
|
constructStiched: function PDFFunction_constructStiched(fn, dict, xref) { |
|
var domain = dict.get('Domain'); |
|
|
|
if (!domain) { |
|
error('No domain'); |
|
} |
|
|
|
var inputSize = domain.length / 2; |
|
if (inputSize !== 1) { |
|
error('Bad domain for stiched function'); |
|
} |
|
|
|
var fnRefs = dict.get('Functions'); |
|
var fns = []; |
|
for (var i = 0, ii = fnRefs.length; i < ii; ++i) { |
|
fns.push(PDFFunction.getIR(xref, xref.fetchIfRef(fnRefs[i]))); |
|
} |
|
|
|
var bounds = dict.get('Bounds'); |
|
var encode = dict.get('Encode'); |
|
|
|
return [CONSTRUCT_STICHED, domain, bounds, encode, fns]; |
|
}, |
|
|
|
constructStichedFromIR: function PDFFunction_constructStichedFromIR(IR) { |
|
var domain = IR[1]; |
|
var bounds = IR[2]; |
|
var encode = IR[3]; |
|
var fnsIR = IR[4]; |
|
var fns = []; |
|
var tmpBuf = new Float32Array(1); |
|
|
|
for (var i = 0, ii = fnsIR.length; i < ii; i++) { |
|
fns.push(PDFFunction.fromIR(fnsIR[i])); |
|
} |
|
|
|
return function constructStichedFromIRResult(src, srcOffset, |
|
dest, destOffset) { |
|
var clip = function constructStichedFromIRClip(v, min, max) { |
|
if (v > max) { |
|
v = max; |
|
} else if (v < min) { |
|
v = min; |
|
} |
|
return v; |
|
}; |
|
|
|
// clip to domain |
|
var v = clip(src[srcOffset], domain[0], domain[1]); |
|
// calulate which bound the value is in |
|
for (var i = 0, ii = bounds.length; i < ii; ++i) { |
|
if (v < bounds[i]) { |
|
break; |
|
} |
|
} |
|
|
|
// encode value into domain of function |
|
var dmin = domain[0]; |
|
if (i > 0) { |
|
dmin = bounds[i - 1]; |
|
} |
|
var dmax = domain[1]; |
|
if (i < bounds.length) { |
|
dmax = bounds[i]; |
|
} |
|
|
|
var rmin = encode[2 * i]; |
|
var rmax = encode[2 * i + 1]; |
|
|
|
// Prevent the value from becoming NaN as a result |
|
// of division by zero (fixes issue6113.pdf). |
|
tmpBuf[0] = dmin === dmax ? rmin : |
|
rmin + (v - dmin) * (rmax - rmin) / (dmax - dmin); |
|
|
|
// call the appropriate function |
|
fns[i](tmpBuf, 0, dest, destOffset); |
|
}; |
|
}, |
|
|
|
constructPostScript: function PDFFunction_constructPostScript(fn, dict, |
|
xref) { |
|
var domain = dict.get('Domain'); |
|
var range = dict.get('Range'); |
|
|
|
if (!domain) { |
|
error('No domain.'); |
|
} |
|
|
|
if (!range) { |
|
error('No range.'); |
|
} |
|
|
|
var lexer = new PostScriptLexer(fn); |
|
var parser = new PostScriptParser(lexer); |
|
var code = parser.parse(); |
|
|
|
return [CONSTRUCT_POSTSCRIPT, domain, range, code]; |
|
}, |
|
|
|
constructPostScriptFromIR: function PDFFunction_constructPostScriptFromIR( |
|
IR) { |
|
var domain = IR[1]; |
|
var range = IR[2]; |
|
var code = IR[3]; |
|
|
|
var compiled = (new PostScriptCompiler()).compile(code, domain, range); |
|
if (compiled) { |
|
// Compiled function consists of simple expressions such as addition, |
|
// subtraction, Math.max, and also contains 'var' and 'return' |
|
// statements. See the generation in the PostScriptCompiler below. |
|
/*jshint -W054 */ |
|
return new Function('src', 'srcOffset', 'dest', 'destOffset', compiled); |
|
} |
|
|
|
info('Unable to compile PS function'); |
|
|
|
var numOutputs = range.length >> 1; |
|
var numInputs = domain.length >> 1; |
|
var evaluator = new PostScriptEvaluator(code); |
|
// Cache the values for a big speed up, the cache size is limited though |
|
// since the number of possible values can be huge from a PS function. |
|
var cache = {}; |
|
// The MAX_CACHE_SIZE is set to ~4x the maximum number of distinct values |
|
// seen in our tests. |
|
var MAX_CACHE_SIZE = 2048 * 4; |
|
var cache_available = MAX_CACHE_SIZE; |
|
var tmpBuf = new Float32Array(numInputs); |
|
|
|
return function constructPostScriptFromIRResult(src, srcOffset, |
|
dest, destOffset) { |
|
var i, value; |
|
var key = ''; |
|
var input = tmpBuf; |
|
for (i = 0; i < numInputs; i++) { |
|
value = src[srcOffset + i]; |
|
input[i] = value; |
|
key += value + '_'; |
|
} |
|
|
|
var cachedValue = cache[key]; |
|
if (cachedValue !== undefined) { |
|
dest.set(cachedValue, destOffset); |
|
return; |
|
} |
|
|
|
var output = new Float32Array(numOutputs); |
|
var stack = evaluator.execute(input); |
|
var stackIndex = stack.length - numOutputs; |
|
for (i = 0; i < numOutputs; i++) { |
|
value = stack[stackIndex + i]; |
|
var bound = range[i * 2]; |
|
if (value < bound) { |
|
value = bound; |
|
} else { |
|
bound = range[i * 2 +1]; |
|
if (value > bound) { |
|
value = bound; |
|
} |
|
} |
|
output[i] = value; |
|
} |
|
if (cache_available > 0) { |
|
cache_available--; |
|
cache[key] = output; |
|
} |
|
dest.set(output, destOffset); |
|
}; |
|
} |
|
}; |
|
})(); |
|
|
|
function isPDFFunction(v) { |
|
var fnDict; |
|
if (typeof v !== 'object') { |
|
return false; |
|
} else if (isDict(v)) { |
|
fnDict = v; |
|
} else if (isStream(v)) { |
|
fnDict = v.dict; |
|
} else { |
|
return false; |
|
} |
|
return fnDict.has('FunctionType'); |
|
} |
|
|
|
var PostScriptStack = (function PostScriptStackClosure() { |
|
var MAX_STACK_SIZE = 100; |
|
function PostScriptStack(initialStack) { |
|
this.stack = !initialStack ? [] : |
|
Array.prototype.slice.call(initialStack, 0); |
|
} |
|
|
|
PostScriptStack.prototype = { |
|
push: function PostScriptStack_push(value) { |
|
if (this.stack.length >= MAX_STACK_SIZE) { |
|
error('PostScript function stack overflow.'); |
|
} |
|
this.stack.push(value); |
|
}, |
|
pop: function PostScriptStack_pop() { |
|
if (this.stack.length <= 0) { |
|
error('PostScript function stack underflow.'); |
|
} |
|
return this.stack.pop(); |
|
}, |
|
copy: function PostScriptStack_copy(n) { |
|
if (this.stack.length + n >= MAX_STACK_SIZE) { |
|
error('PostScript function stack overflow.'); |
|
} |
|
var stack = this.stack; |
|
for (var i = stack.length - n, j = n - 1; j >= 0; j--, i++) { |
|
stack.push(stack[i]); |
|
} |
|
}, |
|
index: function PostScriptStack_index(n) { |
|
this.push(this.stack[this.stack.length - n - 1]); |
|
}, |
|
// rotate the last n stack elements p times |
|
roll: function PostScriptStack_roll(n, p) { |
|
var stack = this.stack; |
|
var l = stack.length - n; |
|
var r = stack.length - 1, c = l + (p - Math.floor(p / n) * n), i, j, t; |
|
for (i = l, j = r; i < j; i++, j--) { |
|
t = stack[i]; stack[i] = stack[j]; stack[j] = t; |
|
} |
|
for (i = l, j = c - 1; i < j; i++, j--) { |
|
t = stack[i]; stack[i] = stack[j]; stack[j] = t; |
|
} |
|
for (i = c, j = r; i < j; i++, j--) { |
|
t = stack[i]; stack[i] = stack[j]; stack[j] = t; |
|
} |
|
} |
|
}; |
|
return PostScriptStack; |
|
})(); |
|
var PostScriptEvaluator = (function PostScriptEvaluatorClosure() { |
|
function PostScriptEvaluator(operators) { |
|
this.operators = operators; |
|
} |
|
PostScriptEvaluator.prototype = { |
|
execute: function PostScriptEvaluator_execute(initialStack) { |
|
var stack = new PostScriptStack(initialStack); |
|
var counter = 0; |
|
var operators = this.operators; |
|
var length = operators.length; |
|
var operator, a, b; |
|
while (counter < length) { |
|
operator = operators[counter++]; |
|
if (typeof operator === 'number') { |
|
// Operator is really an operand and should be pushed to the stack. |
|
stack.push(operator); |
|
continue; |
|
} |
|
switch (operator) { |
|
// non standard ps operators |
|
case 'jz': // jump if false |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
if (!a) { |
|
counter = b; |
|
} |
|
break; |
|
case 'j': // jump |
|
a = stack.pop(); |
|
counter = a; |
|
break; |
|
|
|
// all ps operators in alphabetical order (excluding if/ifelse) |
|
case 'abs': |
|
a = stack.pop(); |
|
stack.push(Math.abs(a)); |
|
break; |
|
case 'add': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a + b); |
|
break; |
|
case 'and': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
if (isBool(a) && isBool(b)) { |
|
stack.push(a && b); |
|
} else { |
|
stack.push(a & b); |
|
} |
|
break; |
|
case 'atan': |
|
a = stack.pop(); |
|
stack.push(Math.atan(a)); |
|
break; |
|
case 'bitshift': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
if (a > 0) { |
|
stack.push(a << b); |
|
} else { |
|
stack.push(a >> b); |
|
} |
|
break; |
|
case 'ceiling': |
|
a = stack.pop(); |
|
stack.push(Math.ceil(a)); |
|
break; |
|
case 'copy': |
|
a = stack.pop(); |
|
stack.copy(a); |
|
break; |
|
case 'cos': |
|
a = stack.pop(); |
|
stack.push(Math.cos(a)); |
|
break; |
|
case 'cvi': |
|
a = stack.pop() | 0; |
|
stack.push(a); |
|
break; |
|
case 'cvr': |
|
// noop |
|
break; |
|
case 'div': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a / b); |
|
break; |
|
case 'dup': |
|
stack.copy(1); |
|
break; |
|
case 'eq': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a === b); |
|
break; |
|
case 'exch': |
|
stack.roll(2, 1); |
|
break; |
|
case 'exp': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(Math.pow(a, b)); |
|
break; |
|
case 'false': |
|
stack.push(false); |
|
break; |
|
case 'floor': |
|
a = stack.pop(); |
|
stack.push(Math.floor(a)); |
|
break; |
|
case 'ge': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a >= b); |
|
break; |
|
case 'gt': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a > b); |
|
break; |
|
case 'idiv': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push((a / b) | 0); |
|
break; |
|
case 'index': |
|
a = stack.pop(); |
|
stack.index(a); |
|
break; |
|
case 'le': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a <= b); |
|
break; |
|
case 'ln': |
|
a = stack.pop(); |
|
stack.push(Math.log(a)); |
|
break; |
|
case 'log': |
|
a = stack.pop(); |
|
stack.push(Math.log(a) / Math.LN10); |
|
break; |
|
case 'lt': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a < b); |
|
break; |
|
case 'mod': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a % b); |
|
break; |
|
case 'mul': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a * b); |
|
break; |
|
case 'ne': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a !== b); |
|
break; |
|
case 'neg': |
|
a = stack.pop(); |
|
stack.push(-a); |
|
break; |
|
case 'not': |
|
a = stack.pop(); |
|
if (isBool(a)) { |
|
stack.push(!a); |
|
} else { |
|
stack.push(~a); |
|
} |
|
break; |
|
case 'or': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
if (isBool(a) && isBool(b)) { |
|
stack.push(a || b); |
|
} else { |
|
stack.push(a | b); |
|
} |
|
break; |
|
case 'pop': |
|
stack.pop(); |
|
break; |
|
case 'roll': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.roll(a, b); |
|
break; |
|
case 'round': |
|
a = stack.pop(); |
|
stack.push(Math.round(a)); |
|
break; |
|
case 'sin': |
|
a = stack.pop(); |
|
stack.push(Math.sin(a)); |
|
break; |
|
case 'sqrt': |
|
a = stack.pop(); |
|
stack.push(Math.sqrt(a)); |
|
break; |
|
case 'sub': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
stack.push(a - b); |
|
break; |
|
case 'true': |
|
stack.push(true); |
|
break; |
|
case 'truncate': |
|
a = stack.pop(); |
|
a = a < 0 ? Math.ceil(a) : Math.floor(a); |
|
stack.push(a); |
|
break; |
|
case 'xor': |
|
b = stack.pop(); |
|
a = stack.pop(); |
|
if (isBool(a) && isBool(b)) { |
|
stack.push(a !== b); |
|
} else { |
|
stack.push(a ^ b); |
|
} |
|
break; |
|
default: |
|
error('Unknown operator ' + operator); |
|
break; |
|
} |
|
} |
|
return stack.stack; |
|
} |
|
}; |
|
return PostScriptEvaluator; |
|
})(); |
|
|
|
// Most of the PDFs functions consist of simple operations such as: |
|
// roll, exch, sub, cvr, pop, index, dup, mul, if, gt, add. |
|
// |
|
// We can compile most of such programs, and at the same moment, we can |
|
// optimize some expressions using basic math properties. Keeping track of |
|
// min/max values will allow us to avoid extra Math.min/Math.max calls. |
|
var PostScriptCompiler = (function PostScriptCompilerClosure() { |
|
function AstNode(type) { |
|
this.type = type; |
|
} |
|
AstNode.prototype.visit = function (visitor) { |
|
throw new Error('abstract method'); |
|
}; |
|
|
|
function AstArgument(index, min, max) { |
|
AstNode.call(this, 'args'); |
|
this.index = index; |
|
this.min = min; |
|
this.max = max; |
|
} |
|
AstArgument.prototype = Object.create(AstNode.prototype); |
|
AstArgument.prototype.visit = function (visitor) { |
|
visitor.visitArgument(this); |
|
}; |
|
|
|
function AstLiteral(number) { |
|
AstNode.call(this, 'literal'); |
|
this.number = number; |
|
this.min = number; |
|
this.max = number; |
|
} |
|
AstLiteral.prototype = Object.create(AstNode.prototype); |
|
AstLiteral.prototype.visit = function (visitor) { |
|
visitor.visitLiteral(this); |
|
}; |
|
|
|
function AstBinaryOperation(op, arg1, arg2, min, max) { |
|
AstNode.call(this, 'binary'); |
|
this.op = op; |
|
this.arg1 = arg1; |
|
this.arg2 = arg2; |
|
this.min = min; |
|
this.max = max; |
|
} |
|
AstBinaryOperation.prototype = Object.create(AstNode.prototype); |
|
AstBinaryOperation.prototype.visit = function (visitor) { |
|
visitor.visitBinaryOperation(this); |
|
}; |
|
|
|
function AstMin(arg, max) { |
|
AstNode.call(this, 'max'); |
|
this.arg = arg; |
|
this.min = arg.min; |
|
this.max = max; |
|
} |
|
AstMin.prototype = Object.create(AstNode.prototype); |
|
AstMin.prototype.visit = function (visitor) { |
|
visitor.visitMin(this); |
|
}; |
|
|
|
function AstVariable(index, min, max) { |
|
AstNode.call(this, 'var'); |
|
this.index = index; |
|
this.min = min; |
|
this.max = max; |
|
} |
|
AstVariable.prototype = Object.create(AstNode.prototype); |
|
AstVariable.prototype.visit = function (visitor) { |
|
visitor.visitVariable(this); |
|
}; |
|
|
|
function AstVariableDefinition(variable, arg) { |
|
AstNode.call(this, 'definition'); |
|
this.variable = variable; |
|
this.arg = arg; |
|
} |
|
AstVariableDefinition.prototype = Object.create(AstNode.prototype); |
|
AstVariableDefinition.prototype.visit = function (visitor) { |
|
visitor.visitVariableDefinition(this); |
|
}; |
|
|
|
function ExpressionBuilderVisitor() { |
|
this.parts = []; |
|
} |
|
ExpressionBuilderVisitor.prototype = { |
|
visitArgument: function (arg) { |
|
this.parts.push('Math.max(', arg.min, ', Math.min(', |
|
arg.max, ', src[srcOffset + ', arg.index, ']))'); |
|
}, |
|
visitVariable: function (variable) { |
|
this.parts.push('v', variable.index); |
|
}, |
|
visitLiteral: function (literal) { |
|
this.parts.push(literal.number); |
|
}, |
|
visitBinaryOperation: function (operation) { |
|
this.parts.push('('); |
|
operation.arg1.visit(this); |
|
this.parts.push(' ', operation.op, ' '); |
|
operation.arg2.visit(this); |
|
this.parts.push(')'); |
|
}, |
|
visitVariableDefinition: function (definition) { |
|
this.parts.push('var '); |
|
definition.variable.visit(this); |
|
this.parts.push(' = '); |
|
definition.arg.visit(this); |
|
this.parts.push(';'); |
|
}, |
|
visitMin: function (max) { |
|
this.parts.push('Math.min('); |
|
max.arg.visit(this); |
|
this.parts.push(', ', max.max, ')'); |
|
}, |
|
toString: function () { |
|
return this.parts.join(''); |
|
} |
|
}; |
|
|
|
function buildAddOperation(num1, num2) { |
|
if (num2.type === 'literal' && num2.number === 0) { |
|
// optimization: second operand is 0 |
|
return num1; |
|
} |
|
if (num1.type === 'literal' && num1.number === 0) { |
|
// optimization: first operand is 0 |
|
return num2; |
|
} |
|
if (num2.type === 'literal' && num1.type === 'literal') { |
|
// optimization: operands operand are literals |
|
return new AstLiteral(num1.number + num2.number); |
|
} |
|
return new AstBinaryOperation('+', num1, num2, |
|
num1.min + num2.min, num1.max + num2.max); |
|
} |
|
|
|
function buildMulOperation(num1, num2) { |
|
if (num2.type === 'literal') { |
|
// optimization: second operands is a literal... |
|
if (num2.number === 0) { |
|
return new AstLiteral(0); // and it's 0 |
|
} else if (num2.number === 1) { |
|
return num1; // and it's 1 |
|
} else if (num1.type === 'literal') { |
|
// ... and first operands is a literal too |
|
return new AstLiteral(num1.number * num2.number); |
|
} |
|
} |
|
if (num1.type === 'literal') { |
|
// optimization: first operands is a literal... |
|
if (num1.number === 0) { |
|
return new AstLiteral(0); // and it's 0 |
|
} else if (num1.number === 1) { |
|
return num2; // and it's 1 |
|
} |
|
} |
|
var min = Math.min(num1.min * num2.min, num1.min * num2.max, |
|
num1.max * num2.min, num1.max * num2.max); |
|
var max = Math.max(num1.min * num2.min, num1.min * num2.max, |
|
num1.max * num2.min, num1.max * num2.max); |
|
return new AstBinaryOperation('*', num1, num2, min, max); |
|
} |
|
|
|
function buildSubOperation(num1, num2) { |
|
if (num2.type === 'literal') { |
|
// optimization: second operands is a literal... |
|
if (num2.number === 0) { |
|
return num1; // ... and it's 0 |
|
} else if (num1.type === 'literal') { |
|
// ... and first operands is a literal too |
|
return new AstLiteral(num1.number - num2.number); |
|
} |
|
} |
|
if (num2.type === 'binary' && num2.op === '-' && |
|
num1.type === 'literal' && num1.number === 1 && |
|
num2.arg1.type === 'literal' && num2.arg1.number === 1) { |
|
// optimization for case: 1 - (1 - x) |
|
return num2.arg2; |
|
} |
|
return new AstBinaryOperation('-', num1, num2, |
|
num1.min - num2.max, num1.max - num2.min); |
|
} |
|
|
|
function buildMinOperation(num1, max) { |
|
if (num1.min >= max) { |
|
// optimization: num1 min value is not less than required max |
|
return new AstLiteral(max); // just returning max |
|
} else if (num1.max <= max) { |
|
// optimization: num1 max value is not greater than required max |
|
return num1; // just returning an argument |
|
} |
|
return new AstMin(num1, max); |
|
} |
|
|
|
function PostScriptCompiler() {} |
|
PostScriptCompiler.prototype = { |
|
compile: function PostScriptCompiler_compile(code, domain, range) { |
|
var stack = []; |
|
var i, ii; |
|
var instructions = []; |
|
var inputSize = domain.length >> 1, outputSize = range.length >> 1; |
|
var lastRegister = 0; |
|
var n, j; |
|
var num1, num2, ast1, ast2, tmpVar, item; |
|
for (i = 0; i < inputSize; i++) { |
|
stack.push(new AstArgument(i, domain[i * 2], domain[i * 2 + 1])); |
|
} |
|
|
|
for (i = 0, ii = code.length; i < ii; i++) { |
|
item = code[i]; |
|
if (typeof item === 'number') { |
|
stack.push(new AstLiteral(item)); |
|
continue; |
|
} |
|
|
|
switch (item) { |
|
case 'add': |
|
if (stack.length < 2) { |
|
return null; |
|
} |
|
num2 = stack.pop(); |
|
num1 = stack.pop(); |
|
stack.push(buildAddOperation(num1, num2)); |
|
break; |
|
case 'cvr': |
|
if (stack.length < 1) { |
|
return null; |
|
} |
|
break; |
|
case 'mul': |
|
if (stack.length < 2) { |
|
return null; |
|
} |
|
num2 = stack.pop(); |
|
num1 = stack.pop(); |
|
stack.push(buildMulOperation(num1, num2)); |
|
break; |
|
case 'sub': |
|
if (stack.length < 2) { |
|
return null; |
|
} |
|
num2 = stack.pop(); |
|
num1 = stack.pop(); |
|
stack.push(buildSubOperation(num1, num2)); |
|
break; |
|
case 'exch': |
|
if (stack.length < 2) { |
|
return null; |
|
} |
|
ast1 = stack.pop(); ast2 = stack.pop(); |
|
stack.push(ast1, ast2); |
|
break; |
|
case 'pop': |
|
if (stack.length < 1) { |
|
return null; |
|
} |
|
stack.pop(); |
|
break; |
|
case 'index': |
|
if (stack.length < 1) { |
|
return null; |
|
} |
|
num1 = stack.pop(); |
|
if (num1.type !== 'literal') { |
|
return null; |
|
} |
|
n = num1.number; |
|
if (n < 0 || (n|0) !== n || stack.length < n) { |
|
return null; |
|
} |
|
ast1 = stack[stack.length - n - 1]; |
|
if (ast1.type === 'literal' || ast1.type === 'var') { |
|
stack.push(ast1); |
|
break; |
|
} |
|
tmpVar = new AstVariable(lastRegister++, ast1.min, ast1.max); |
|
stack[stack.length - n - 1] = tmpVar; |
|
stack.push(tmpVar); |
|
instructions.push(new AstVariableDefinition(tmpVar, ast1)); |
|
break; |
|
case 'dup': |
|
if (stack.length < 1) { |
|
return null; |
|
} |
|
if (typeof code[i + 1] === 'number' && code[i + 2] === 'gt' && |
|
code[i + 3] === i + 7 && code[i + 4] === 'jz' && |
|
code[i + 5] === 'pop' && code[i + 6] === code[i + 1]) { |
|
// special case of the commands sequence for the min operation |
|
num1 = stack.pop(); |
|
stack.push(buildMinOperation(num1, code[i + 1])); |
|
i += 6; |
|
break; |
|
} |
|
ast1 = stack[stack.length - 1]; |
|
if (ast1.type === 'literal' || ast1.type === 'var') { |
|
// we don't have to save into intermediate variable a literal or |
|
// variable. |
|
stack.push(ast1); |
|
break; |
|
} |
|
tmpVar = new AstVariable(lastRegister++, ast1.min, ast1.max); |
|
stack[stack.length - 1] = tmpVar; |
|
stack.push(tmpVar); |
|
instructions.push(new AstVariableDefinition(tmpVar, ast1)); |
|
break; |
|
case 'roll': |
|
if (stack.length < 2) { |
|
return null; |
|
} |
|
num2 = stack.pop(); |
|
num1 = stack.pop(); |
|
if (num2.type !== 'literal' || num1.type !== 'literal') { |
|
// both roll operands must be numbers |
|
return null; |
|
} |
|
j = num2.number; |
|
n = num1.number; |
|
if (n <= 0 || (n|0) !== n || (j|0) !== j || stack.length < n) { |
|
// ... and integers |
|
return null; |
|
} |
|
j = ((j % n) + n) % n; |
|
if (j === 0) { |
|
break; // just skipping -- there are nothing to rotate |
|
} |
|
Array.prototype.push.apply(stack, |
|
stack.splice(stack.length - n, n - j)); |
|
break; |
|
default: |
|
return null; // unsupported operator |
|
} |
|
} |
|
|
|
if (stack.length !== outputSize) { |
|
return null; |
|
} |
|
|
|
var result = []; |
|
instructions.forEach(function (instruction) { |
|
var statementBuilder = new ExpressionBuilderVisitor(); |
|
instruction.visit(statementBuilder); |
|
result.push(statementBuilder.toString()); |
|
}); |
|
stack.forEach(function (expr, i) { |
|
var statementBuilder = new ExpressionBuilderVisitor(); |
|
expr.visit(statementBuilder); |
|
var min = range[i * 2], max = range[i * 2 + 1]; |
|
var out = [statementBuilder.toString()]; |
|
if (min > expr.min) { |
|
out.unshift('Math.max(', min, ', '); |
|
out.push(')'); |
|
} |
|
if (max < expr.max) { |
|
out.unshift('Math.min(', max, ', '); |
|
out.push(')'); |
|
} |
|
out.unshift('dest[destOffset + ', i, '] = '); |
|
out.push(';'); |
|
result.push(out.join('')); |
|
}); |
|
return result.join('\n'); |
|
} |
|
}; |
|
|
|
return PostScriptCompiler; |
|
})(); |
|
|
|
exports.isPDFFunction = isPDFFunction; |
|
exports.PDFFunction = PDFFunction; |
|
exports.PostScriptEvaluator = PostScriptEvaluator; |
|
exports.PostScriptCompiler = PostScriptCompiler; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreColorSpace = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreFunction, root.pdfjsCoreStream); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreFunction, |
|
coreStream) { |
|
|
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isArray = sharedUtil.isArray; |
|
var isString = sharedUtil.isString; |
|
var shadow = sharedUtil.shadow; |
|
var warn = sharedUtil.warn; |
|
var isDict = corePrimitives.isDict; |
|
var isName = corePrimitives.isName; |
|
var isStream = corePrimitives.isStream; |
|
var PDFFunction = coreFunction.PDFFunction; |
|
|
|
var coreImage; // see _setCoreImage below |
|
var PDFImage; // = coreImage.PDFImage; |
|
|
|
var ColorSpace = (function ColorSpaceClosure() { |
|
// Constructor should define this.numComps, this.defaultColor, this.name |
|
function ColorSpace() { |
|
error('should not call ColorSpace constructor'); |
|
} |
|
|
|
ColorSpace.prototype = { |
|
/** |
|
* Converts the color value to the RGB color. The color components are |
|
* located in the src array starting from the srcOffset. Returns the array |
|
* of the rgb components, each value ranging from [0,255]. |
|
*/ |
|
getRgb: function ColorSpace_getRgb(src, srcOffset) { |
|
var rgb = new Uint8Array(3); |
|
this.getRgbItem(src, srcOffset, rgb, 0); |
|
return rgb; |
|
}, |
|
/** |
|
* Converts the color value to the RGB color, similar to the getRgb method. |
|
* The result placed into the dest array starting from the destOffset. |
|
*/ |
|
getRgbItem: function ColorSpace_getRgbItem(src, srcOffset, |
|
dest, destOffset) { |
|
error('Should not call ColorSpace.getRgbItem'); |
|
}, |
|
/** |
|
* Converts the specified number of the color values to the RGB colors. |
|
* The colors are located in the src array starting from the srcOffset. |
|
* The result is placed into the dest array starting from the destOffset. |
|
* The src array items shall be in [0,2^bits) range, the dest array items |
|
* will be in [0,255] range. alpha01 indicates how many alpha components |
|
* there are in the dest array; it will be either 0 (RGB array) or 1 (RGBA |
|
* array). |
|
*/ |
|
getRgbBuffer: function ColorSpace_getRgbBuffer(src, srcOffset, count, |
|
dest, destOffset, bits, |
|
alpha01) { |
|
error('Should not call ColorSpace.getRgbBuffer'); |
|
}, |
|
/** |
|
* Determines the number of bytes required to store the result of the |
|
* conversion done by the getRgbBuffer method. As in getRgbBuffer, |
|
* |alpha01| is either 0 (RGB output) or 1 (RGBA output). |
|
*/ |
|
getOutputLength: function ColorSpace_getOutputLength(inputLength, |
|
alpha01) { |
|
error('Should not call ColorSpace.getOutputLength'); |
|
}, |
|
/** |
|
* Returns true if source data will be equal the result/output data. |
|
*/ |
|
isPassthrough: function ColorSpace_isPassthrough(bits) { |
|
return false; |
|
}, |
|
/** |
|
* Fills in the RGB colors in the destination buffer. alpha01 indicates |
|
* how many alpha components there are in the dest array; it will be either |
|
* 0 (RGB array) or 1 (RGBA array). |
|
*/ |
|
fillRgb: function ColorSpace_fillRgb(dest, originalWidth, |
|
originalHeight, width, height, |
|
actualHeight, bpc, comps, alpha01) { |
|
var count = originalWidth * originalHeight; |
|
var rgbBuf = null; |
|
var numComponentColors = 1 << bpc; |
|
var needsResizing = originalHeight !== height || originalWidth !== width; |
|
var i, ii; |
|
|
|
if (this.isPassthrough(bpc)) { |
|
rgbBuf = comps; |
|
} else if (this.numComps === 1 && count > numComponentColors && |
|
this.name !== 'DeviceGray' && this.name !== 'DeviceRGB') { |
|
// Optimization: create a color map when there is just one component and |
|
// we are converting more colors than the size of the color map. We |
|
// don't build the map if the colorspace is gray or rgb since those |
|
// methods are faster than building a map. This mainly offers big speed |
|
// ups for indexed and alternate colorspaces. |
|
// |
|
// TODO it may be worth while to cache the color map. While running |
|
// testing I never hit a cache so I will leave that out for now (perhaps |
|
// we are reparsing colorspaces too much?). |
|
var allColors = bpc <= 8 ? new Uint8Array(numComponentColors) : |
|
new Uint16Array(numComponentColors); |
|
var key; |
|
for (i = 0; i < numComponentColors; i++) { |
|
allColors[i] = i; |
|
} |
|
var colorMap = new Uint8Array(numComponentColors * 3); |
|
this.getRgbBuffer(allColors, 0, numComponentColors, colorMap, 0, bpc, |
|
/* alpha01 = */ 0); |
|
|
|
var destPos, rgbPos; |
|
if (!needsResizing) { |
|
// Fill in the RGB values directly into |dest|. |
|
destPos = 0; |
|
for (i = 0; i < count; ++i) { |
|
key = comps[i] * 3; |
|
dest[destPos++] = colorMap[key]; |
|
dest[destPos++] = colorMap[key + 1]; |
|
dest[destPos++] = colorMap[key + 2]; |
|
destPos += alpha01; |
|
} |
|
} else { |
|
rgbBuf = new Uint8Array(count * 3); |
|
rgbPos = 0; |
|
for (i = 0; i < count; ++i) { |
|
key = comps[i] * 3; |
|
rgbBuf[rgbPos++] = colorMap[key]; |
|
rgbBuf[rgbPos++] = colorMap[key + 1]; |
|
rgbBuf[rgbPos++] = colorMap[key + 2]; |
|
} |
|
} |
|
} else { |
|
if (!needsResizing) { |
|
// Fill in the RGB values directly into |dest|. |
|
this.getRgbBuffer(comps, 0, width * actualHeight, dest, 0, bpc, |
|
alpha01); |
|
} else { |
|
rgbBuf = new Uint8Array(count * 3); |
|
this.getRgbBuffer(comps, 0, count, rgbBuf, 0, bpc, |
|
/* alpha01 = */ 0); |
|
} |
|
} |
|
|
|
if (rgbBuf) { |
|
if (needsResizing) { |
|
PDFImage.resize(rgbBuf, bpc, 3, originalWidth, originalHeight, width, |
|
height, dest, alpha01); |
|
} else { |
|
rgbPos = 0; |
|
destPos = 0; |
|
for (i = 0, ii = width * actualHeight; i < ii; i++) { |
|
dest[destPos++] = rgbBuf[rgbPos++]; |
|
dest[destPos++] = rgbBuf[rgbPos++]; |
|
dest[destPos++] = rgbBuf[rgbPos++]; |
|
destPos += alpha01; |
|
} |
|
} |
|
} |
|
}, |
|
/** |
|
* True if the colorspace has components in the default range of [0, 1]. |
|
* This should be true for all colorspaces except for lab color spaces |
|
* which are [0,100], [-128, 127], [-128, 127]. |
|
*/ |
|
usesZeroToOneRange: true |
|
}; |
|
|
|
ColorSpace.parse = function ColorSpace_parse(cs, xref, res) { |
|
var IR = ColorSpace.parseToIR(cs, xref, res); |
|
if (IR instanceof AlternateCS) { |
|
return IR; |
|
} |
|
return ColorSpace.fromIR(IR); |
|
}; |
|
|
|
ColorSpace.fromIR = function ColorSpace_fromIR(IR) { |
|
var name = isArray(IR) ? IR[0] : IR; |
|
var whitePoint, blackPoint, gamma; |
|
|
|
switch (name) { |
|
case 'DeviceGrayCS': |
|
return this.singletons.gray; |
|
case 'DeviceRgbCS': |
|
return this.singletons.rgb; |
|
case 'DeviceCmykCS': |
|
return this.singletons.cmyk; |
|
case 'CalGrayCS': |
|
whitePoint = IR[1].WhitePoint; |
|
blackPoint = IR[1].BlackPoint; |
|
gamma = IR[1].Gamma; |
|
return new CalGrayCS(whitePoint, blackPoint, gamma); |
|
case 'CalRGBCS': |
|
whitePoint = IR[1].WhitePoint; |
|
blackPoint = IR[1].BlackPoint; |
|
gamma = IR[1].Gamma; |
|
var matrix = IR[1].Matrix; |
|
return new CalRGBCS(whitePoint, blackPoint, gamma, matrix); |
|
case 'PatternCS': |
|
var basePatternCS = IR[1]; |
|
if (basePatternCS) { |
|
basePatternCS = ColorSpace.fromIR(basePatternCS); |
|
} |
|
return new PatternCS(basePatternCS); |
|
case 'IndexedCS': |
|
var baseIndexedCS = IR[1]; |
|
var hiVal = IR[2]; |
|
var lookup = IR[3]; |
|
return new IndexedCS(ColorSpace.fromIR(baseIndexedCS), hiVal, lookup); |
|
case 'AlternateCS': |
|
var numComps = IR[1]; |
|
var alt = IR[2]; |
|
var tintFnIR = IR[3]; |
|
|
|
return new AlternateCS(numComps, ColorSpace.fromIR(alt), |
|
PDFFunction.fromIR(tintFnIR)); |
|
case 'LabCS': |
|
whitePoint = IR[1].WhitePoint; |
|
blackPoint = IR[1].BlackPoint; |
|
var range = IR[1].Range; |
|
return new LabCS(whitePoint, blackPoint, range); |
|
default: |
|
error('Unknown name ' + name); |
|
} |
|
return null; |
|
}; |
|
|
|
ColorSpace.parseToIR = function ColorSpace_parseToIR(cs, xref, res) { |
|
if (isName(cs)) { |
|
var colorSpaces = res.get('ColorSpace'); |
|
if (isDict(colorSpaces)) { |
|
var refcs = colorSpaces.get(cs.name); |
|
if (refcs) { |
|
cs = refcs; |
|
} |
|
} |
|
} |
|
|
|
cs = xref.fetchIfRef(cs); |
|
var mode; |
|
|
|
if (isName(cs)) { |
|
mode = cs.name; |
|
this.mode = mode; |
|
|
|
switch (mode) { |
|
case 'DeviceGray': |
|
case 'G': |
|
return 'DeviceGrayCS'; |
|
case 'DeviceRGB': |
|
case 'RGB': |
|
return 'DeviceRgbCS'; |
|
case 'DeviceCMYK': |
|
case 'CMYK': |
|
return 'DeviceCmykCS'; |
|
case 'Pattern': |
|
return ['PatternCS', null]; |
|
default: |
|
error('unrecognized colorspace ' + mode); |
|
} |
|
} else if (isArray(cs)) { |
|
mode = xref.fetchIfRef(cs[0]).name; |
|
this.mode = mode; |
|
var numComps, params, alt; |
|
|
|
switch (mode) { |
|
case 'DeviceGray': |
|
case 'G': |
|
return 'DeviceGrayCS'; |
|
case 'DeviceRGB': |
|
case 'RGB': |
|
return 'DeviceRgbCS'; |
|
case 'DeviceCMYK': |
|
case 'CMYK': |
|
return 'DeviceCmykCS'; |
|
case 'CalGray': |
|
params = xref.fetchIfRef(cs[1]).getAll(); |
|
return ['CalGrayCS', params]; |
|
case 'CalRGB': |
|
params = xref.fetchIfRef(cs[1]).getAll(); |
|
return ['CalRGBCS', params]; |
|
case 'ICCBased': |
|
var stream = xref.fetchIfRef(cs[1]); |
|
var dict = stream.dict; |
|
numComps = dict.get('N'); |
|
alt = dict.get('Alternate'); |
|
if (alt) { |
|
var altIR = ColorSpace.parseToIR(alt, xref, res); |
|
// Parse the /Alternate CS to ensure that the number of components |
|
// are correct, and also (indirectly) that it is not a PatternCS. |
|
var altCS = ColorSpace.fromIR(altIR); |
|
if (altCS.numComps === numComps) { |
|
return altIR; |
|
} |
|
warn('ICCBased color space: Ignoring incorrect /Alternate entry.'); |
|
} |
|
if (numComps === 1) { |
|
return 'DeviceGrayCS'; |
|
} else if (numComps === 3) { |
|
return 'DeviceRgbCS'; |
|
} else if (numComps === 4) { |
|
return 'DeviceCmykCS'; |
|
} |
|
break; |
|
case 'Pattern': |
|
var basePatternCS = cs[1] || null; |
|
if (basePatternCS) { |
|
basePatternCS = ColorSpace.parseToIR(basePatternCS, xref, res); |
|
} |
|
return ['PatternCS', basePatternCS]; |
|
case 'Indexed': |
|
case 'I': |
|
var baseIndexedCS = ColorSpace.parseToIR(cs[1], xref, res); |
|
var hiVal = xref.fetchIfRef(cs[2]) + 1; |
|
var lookup = xref.fetchIfRef(cs[3]); |
|
if (isStream(lookup)) { |
|
lookup = lookup.getBytes(); |
|
} |
|
return ['IndexedCS', baseIndexedCS, hiVal, lookup]; |
|
case 'Separation': |
|
case 'DeviceN': |
|
var name = xref.fetchIfRef(cs[1]); |
|
numComps = 1; |
|
if (isName(name)) { |
|
numComps = 1; |
|
} else if (isArray(name)) { |
|
numComps = name.length; |
|
} |
|
alt = ColorSpace.parseToIR(cs[2], xref, res); |
|
var tintFnIR = PDFFunction.getIR(xref, xref.fetchIfRef(cs[3])); |
|
return ['AlternateCS', numComps, alt, tintFnIR]; |
|
case 'Lab': |
|
params = xref.fetchIfRef(cs[1]).getAll(); |
|
return ['LabCS', params]; |
|
default: |
|
error('unimplemented color space object "' + mode + '"'); |
|
} |
|
} else { |
|
error('unrecognized color space object: "' + cs + '"'); |
|
} |
|
return null; |
|
}; |
|
/** |
|
* Checks if a decode map matches the default decode map for a color space. |
|
* This handles the general decode maps where there are two values per |
|
* component. e.g. [0, 1, 0, 1, 0, 1] for a RGB color. |
|
* This does not handle Lab, Indexed, or Pattern decode maps since they are |
|
* slightly different. |
|
* @param {Array} decode Decode map (usually from an image). |
|
* @param {Number} n Number of components the color space has. |
|
*/ |
|
ColorSpace.isDefaultDecode = function ColorSpace_isDefaultDecode(decode, n) { |
|
if (!isArray(decode)) { |
|
return true; |
|
} |
|
|
|
if (n * 2 !== decode.length) { |
|
warn('The decode map is not the correct length'); |
|
return true; |
|
} |
|
for (var i = 0, ii = decode.length; i < ii; i += 2) { |
|
if (decode[i] !== 0 || decode[i + 1] !== 1) { |
|
return false; |
|
} |
|
} |
|
return true; |
|
}; |
|
|
|
ColorSpace.singletons = { |
|
get gray() { |
|
return shadow(this, 'gray', new DeviceGrayCS()); |
|
}, |
|
get rgb() { |
|
return shadow(this, 'rgb', new DeviceRgbCS()); |
|
}, |
|
get cmyk() { |
|
return shadow(this, 'cmyk', new DeviceCmykCS()); |
|
} |
|
}; |
|
|
|
return ColorSpace; |
|
})(); |
|
|
|
/** |
|
* Alternate color space handles both Separation and DeviceN color spaces. A |
|
* Separation color space is actually just a DeviceN with one color component. |
|
* Both color spaces use a tinting function to convert colors to a base color |
|
* space. |
|
*/ |
|
var AlternateCS = (function AlternateCSClosure() { |
|
function AlternateCS(numComps, base, tintFn) { |
|
this.name = 'Alternate'; |
|
this.numComps = numComps; |
|
this.defaultColor = new Float32Array(numComps); |
|
for (var i = 0; i < numComps; ++i) { |
|
this.defaultColor[i] = 1; |
|
} |
|
this.base = base; |
|
this.tintFn = tintFn; |
|
this.tmpBuf = new Float32Array(base.numComps); |
|
} |
|
|
|
AlternateCS.prototype = { |
|
getRgb: ColorSpace.prototype.getRgb, |
|
getRgbItem: function AlternateCS_getRgbItem(src, srcOffset, |
|
dest, destOffset) { |
|
var tmpBuf = this.tmpBuf; |
|
this.tintFn(src, srcOffset, tmpBuf, 0); |
|
this.base.getRgbItem(tmpBuf, 0, dest, destOffset); |
|
}, |
|
getRgbBuffer: function AlternateCS_getRgbBuffer(src, srcOffset, count, |
|
dest, destOffset, bits, |
|
alpha01) { |
|
var tintFn = this.tintFn; |
|
var base = this.base; |
|
var scale = 1 / ((1 << bits) - 1); |
|
var baseNumComps = base.numComps; |
|
var usesZeroToOneRange = base.usesZeroToOneRange; |
|
var isPassthrough = (base.isPassthrough(8) || !usesZeroToOneRange) && |
|
alpha01 === 0; |
|
var pos = isPassthrough ? destOffset : 0; |
|
var baseBuf = isPassthrough ? dest : new Uint8Array(baseNumComps * count); |
|
var numComps = this.numComps; |
|
|
|
var scaled = new Float32Array(numComps); |
|
var tinted = new Float32Array(baseNumComps); |
|
var i, j; |
|
if (usesZeroToOneRange) { |
|
for (i = 0; i < count; i++) { |
|
for (j = 0; j < numComps; j++) { |
|
scaled[j] = src[srcOffset++] * scale; |
|
} |
|
tintFn(scaled, 0, tinted, 0); |
|
for (j = 0; j < baseNumComps; j++) { |
|
baseBuf[pos++] = tinted[j] * 255; |
|
} |
|
} |
|
} else { |
|
for (i = 0; i < count; i++) { |
|
for (j = 0; j < numComps; j++) { |
|
scaled[j] = src[srcOffset++] * scale; |
|
} |
|
tintFn(scaled, 0, tinted, 0); |
|
base.getRgbItem(tinted, 0, baseBuf, pos); |
|
pos += baseNumComps; |
|
} |
|
} |
|
if (!isPassthrough) { |
|
base.getRgbBuffer(baseBuf, 0, count, dest, destOffset, 8, alpha01); |
|
} |
|
}, |
|
getOutputLength: function AlternateCS_getOutputLength(inputLength, |
|
alpha01) { |
|
return this.base.getOutputLength(inputLength * |
|
this.base.numComps / this.numComps, |
|
alpha01); |
|
}, |
|
isPassthrough: ColorSpace.prototype.isPassthrough, |
|
fillRgb: ColorSpace.prototype.fillRgb, |
|
isDefaultDecode: function AlternateCS_isDefaultDecode(decodeMap) { |
|
return ColorSpace.isDefaultDecode(decodeMap, this.numComps); |
|
}, |
|
usesZeroToOneRange: true |
|
}; |
|
|
|
return AlternateCS; |
|
})(); |
|
|
|
var PatternCS = (function PatternCSClosure() { |
|
function PatternCS(baseCS) { |
|
this.name = 'Pattern'; |
|
this.base = baseCS; |
|
} |
|
PatternCS.prototype = {}; |
|
|
|
return PatternCS; |
|
})(); |
|
|
|
var IndexedCS = (function IndexedCSClosure() { |
|
function IndexedCS(base, highVal, lookup) { |
|
this.name = 'Indexed'; |
|
this.numComps = 1; |
|
this.defaultColor = new Uint8Array([0]); |
|
this.base = base; |
|
this.highVal = highVal; |
|
|
|
var baseNumComps = base.numComps; |
|
var length = baseNumComps * highVal; |
|
var lookupArray; |
|
|
|
if (isStream(lookup)) { |
|
lookupArray = new Uint8Array(length); |
|
var bytes = lookup.getBytes(length); |
|
lookupArray.set(bytes); |
|
} else if (isString(lookup)) { |
|
lookupArray = new Uint8Array(length); |
|
for (var i = 0; i < length; ++i) { |
|
lookupArray[i] = lookup.charCodeAt(i); |
|
} |
|
} else if (lookup instanceof Uint8Array || lookup instanceof Array) { |
|
lookupArray = lookup; |
|
} else { |
|
error('Unrecognized lookup table: ' + lookup); |
|
} |
|
this.lookup = lookupArray; |
|
} |
|
|
|
IndexedCS.prototype = { |
|
getRgb: ColorSpace.prototype.getRgb, |
|
getRgbItem: function IndexedCS_getRgbItem(src, srcOffset, |
|
dest, destOffset) { |
|
var numComps = this.base.numComps; |
|
var start = src[srcOffset] * numComps; |
|
this.base.getRgbItem(this.lookup, start, dest, destOffset); |
|
}, |
|
getRgbBuffer: function IndexedCS_getRgbBuffer(src, srcOffset, count, |
|
dest, destOffset, bits, |
|
alpha01) { |
|
var base = this.base; |
|
var numComps = base.numComps; |
|
var outputDelta = base.getOutputLength(numComps, alpha01); |
|
var lookup = this.lookup; |
|
|
|
for (var i = 0; i < count; ++i) { |
|
var lookupPos = src[srcOffset++] * numComps; |
|
base.getRgbBuffer(lookup, lookupPos, 1, dest, destOffset, 8, alpha01); |
|
destOffset += outputDelta; |
|
} |
|
}, |
|
getOutputLength: function IndexedCS_getOutputLength(inputLength, alpha01) { |
|
return this.base.getOutputLength(inputLength * this.base.numComps, |
|
alpha01); |
|
}, |
|
isPassthrough: ColorSpace.prototype.isPassthrough, |
|
fillRgb: ColorSpace.prototype.fillRgb, |
|
isDefaultDecode: function IndexedCS_isDefaultDecode(decodeMap) { |
|
// indexed color maps shouldn't be changed |
|
return true; |
|
}, |
|
usesZeroToOneRange: true |
|
}; |
|
return IndexedCS; |
|
})(); |
|
|
|
var DeviceGrayCS = (function DeviceGrayCSClosure() { |
|
function DeviceGrayCS() { |
|
this.name = 'DeviceGray'; |
|
this.numComps = 1; |
|
this.defaultColor = new Float32Array([0]); |
|
} |
|
|
|
DeviceGrayCS.prototype = { |
|
getRgb: ColorSpace.prototype.getRgb, |
|
getRgbItem: function DeviceGrayCS_getRgbItem(src, srcOffset, |
|
dest, destOffset) { |
|
var c = (src[srcOffset] * 255) | 0; |
|
c = c < 0 ? 0 : c > 255 ? 255 : c; |
|
dest[destOffset] = dest[destOffset + 1] = dest[destOffset + 2] = c; |
|
}, |
|
getRgbBuffer: function DeviceGrayCS_getRgbBuffer(src, srcOffset, count, |
|
dest, destOffset, bits, |
|
alpha01) { |
|
var scale = 255 / ((1 << bits) - 1); |
|
var j = srcOffset, q = destOffset; |
|
for (var i = 0; i < count; ++i) { |
|
var c = (scale * src[j++]) | 0; |
|
dest[q++] = c; |
|
dest[q++] = c; |
|
dest[q++] = c; |
|
q += alpha01; |
|
} |
|
}, |
|
getOutputLength: function DeviceGrayCS_getOutputLength(inputLength, |
|
alpha01) { |
|
return inputLength * (3 + alpha01); |
|
}, |
|
isPassthrough: ColorSpace.prototype.isPassthrough, |
|
fillRgb: ColorSpace.prototype.fillRgb, |
|
isDefaultDecode: function DeviceGrayCS_isDefaultDecode(decodeMap) { |
|
return ColorSpace.isDefaultDecode(decodeMap, this.numComps); |
|
}, |
|
usesZeroToOneRange: true |
|
}; |
|
return DeviceGrayCS; |
|
})(); |
|
|
|
var DeviceRgbCS = (function DeviceRgbCSClosure() { |
|
function DeviceRgbCS() { |
|
this.name = 'DeviceRGB'; |
|
this.numComps = 3; |
|
this.defaultColor = new Float32Array([0, 0, 0]); |
|
} |
|
DeviceRgbCS.prototype = { |
|
getRgb: ColorSpace.prototype.getRgb, |
|
getRgbItem: function DeviceRgbCS_getRgbItem(src, srcOffset, |
|
dest, destOffset) { |
|
var r = (src[srcOffset] * 255) | 0; |
|
var g = (src[srcOffset + 1] * 255) | 0; |
|
var b = (src[srcOffset + 2] * 255) | 0; |
|
dest[destOffset] = r < 0 ? 0 : r > 255 ? 255 : r; |
|
dest[destOffset + 1] = g < 0 ? 0 : g > 255 ? 255 : g; |
|
dest[destOffset + 2] = b < 0 ? 0 : b > 255 ? 255 : b; |
|
}, |
|
getRgbBuffer: function DeviceRgbCS_getRgbBuffer(src, srcOffset, count, |
|
dest, destOffset, bits, |
|
alpha01) { |
|
if (bits === 8 && alpha01 === 0) { |
|
dest.set(src.subarray(srcOffset, srcOffset + count * 3), destOffset); |
|
return; |
|
} |
|
var scale = 255 / ((1 << bits) - 1); |
|
var j = srcOffset, q = destOffset; |
|
for (var i = 0; i < count; ++i) { |
|
dest[q++] = (scale * src[j++]) | 0; |
|
dest[q++] = (scale * src[j++]) | 0; |
|
dest[q++] = (scale * src[j++]) | 0; |
|
q += alpha01; |
|
} |
|
}, |
|
getOutputLength: function DeviceRgbCS_getOutputLength(inputLength, |
|
alpha01) { |
|
return (inputLength * (3 + alpha01) / 3) | 0; |
|
}, |
|
isPassthrough: function DeviceRgbCS_isPassthrough(bits) { |
|
return bits === 8; |
|
}, |
|
fillRgb: ColorSpace.prototype.fillRgb, |
|
isDefaultDecode: function DeviceRgbCS_isDefaultDecode(decodeMap) { |
|
return ColorSpace.isDefaultDecode(decodeMap, this.numComps); |
|
}, |
|
usesZeroToOneRange: true |
|
}; |
|
return DeviceRgbCS; |
|
})(); |
|
|
|
var DeviceCmykCS = (function DeviceCmykCSClosure() { |
|
// The coefficients below was found using numerical analysis: the method of |
|
// steepest descent for the sum((f_i - color_value_i)^2) for r/g/b colors, |
|
// where color_value is the tabular value from the table of sampled RGB colors |
|
// from CMYK US Web Coated (SWOP) colorspace, and f_i is the corresponding |
|
// CMYK color conversion using the estimation below: |
|
// f(A, B,.. N) = Acc+Bcm+Ccy+Dck+c+Fmm+Gmy+Hmk+Im+Jyy+Kyk+Ly+Mkk+Nk+255 |
|
function convertToRgb(src, srcOffset, srcScale, dest, destOffset) { |
|
var c = src[srcOffset + 0] * srcScale; |
|
var m = src[srcOffset + 1] * srcScale; |
|
var y = src[srcOffset + 2] * srcScale; |
|
var k = src[srcOffset + 3] * srcScale; |
|
|
|
var r = |
|
(c * (-4.387332384609988 * c + 54.48615194189176 * m + |
|
18.82290502165302 * y + 212.25662451639585 * k + |
|
-285.2331026137004) + |
|
m * (1.7149763477362134 * m - 5.6096736904047315 * y + |
|
-17.873870861415444 * k - 5.497006427196366) + |
|
y * (-2.5217340131683033 * y - 21.248923337353073 * k + |
|
17.5119270841813) + |
|
k * (-21.86122147463605 * k - 189.48180835922747) + 255) | 0; |
|
var g = |
|
(c * (8.841041422036149 * c + 60.118027045597366 * m + |
|
6.871425592049007 * y + 31.159100130055922 * k + |
|
-79.2970844816548) + |
|
m * (-15.310361306967817 * m + 17.575251261109482 * y + |
|
131.35250912493976 * k - 190.9453302588951) + |
|
y * (4.444339102852739 * y + 9.8632861493405 * k - 24.86741582555878) + |
|
k * (-20.737325471181034 * k - 187.80453709719578) + 255) | 0; |
|
var b = |
|
(c * (0.8842522430003296 * c + 8.078677503112928 * m + |
|
30.89978309703729 * y - 0.23883238689178934 * k + |
|
-14.183576799673286) + |
|
m * (10.49593273432072 * m + 63.02378494754052 * y + |
|
50.606957656360734 * k - 112.23884253719248) + |
|
y * (0.03296041114873217 * y + 115.60384449646641 * k + |
|
-193.58209356861505) + |
|
k * (-22.33816807309886 * k - 180.12613974708367) + 255) | 0; |
|
|
|
dest[destOffset] = r > 255 ? 255 : r < 0 ? 0 : r; |
|
dest[destOffset + 1] = g > 255 ? 255 : g < 0 ? 0 : g; |
|
dest[destOffset + 2] = b > 255 ? 255 : b < 0 ? 0 : b; |
|
} |
|
|
|
function DeviceCmykCS() { |
|
this.name = 'DeviceCMYK'; |
|
this.numComps = 4; |
|
this.defaultColor = new Float32Array([0, 0, 0, 1]); |
|
} |
|
DeviceCmykCS.prototype = { |
|
getRgb: ColorSpace.prototype.getRgb, |
|
getRgbItem: function DeviceCmykCS_getRgbItem(src, srcOffset, |
|
dest, destOffset) { |
|
convertToRgb(src, srcOffset, 1, dest, destOffset); |
|
}, |
|
getRgbBuffer: function DeviceCmykCS_getRgbBuffer(src, srcOffset, count, |
|
dest, destOffset, bits, |
|
alpha01) { |
|
var scale = 1 / ((1 << bits) - 1); |
|
for (var i = 0; i < count; i++) { |
|
convertToRgb(src, srcOffset, scale, dest, destOffset); |
|
srcOffset += 4; |
|
destOffset += 3 + alpha01; |
|
} |
|
}, |
|
getOutputLength: function DeviceCmykCS_getOutputLength(inputLength, |
|
alpha01) { |
|
return (inputLength / 4 * (3 + alpha01)) | 0; |
|
}, |
|
isPassthrough: ColorSpace.prototype.isPassthrough, |
|
fillRgb: ColorSpace.prototype.fillRgb, |
|
isDefaultDecode: function DeviceCmykCS_isDefaultDecode(decodeMap) { |
|
return ColorSpace.isDefaultDecode(decodeMap, this.numComps); |
|
}, |
|
usesZeroToOneRange: true |
|
}; |
|
|
|
return DeviceCmykCS; |
|
})(); |
|
|
|
// |
|
// CalGrayCS: Based on "PDF Reference, Sixth Ed", p.245 |
|
// |
|
var CalGrayCS = (function CalGrayCSClosure() { |
|
function CalGrayCS(whitePoint, blackPoint, gamma) { |
|
this.name = 'CalGray'; |
|
this.numComps = 1; |
|
this.defaultColor = new Float32Array([0]); |
|
|
|
if (!whitePoint) { |
|
error('WhitePoint missing - required for color space CalGray'); |
|
} |
|
blackPoint = blackPoint || [0, 0, 0]; |
|
gamma = gamma || 1; |
|
|
|
// Translate arguments to spec variables. |
|
this.XW = whitePoint[0]; |
|
this.YW = whitePoint[1]; |
|
this.ZW = whitePoint[2]; |
|
|
|
this.XB = blackPoint[0]; |
|
this.YB = blackPoint[1]; |
|
this.ZB = blackPoint[2]; |
|
|
|
this.G = gamma; |
|
|
|
// Validate variables as per spec. |
|
if (this.XW < 0 || this.ZW < 0 || this.YW !== 1) { |
|
error('Invalid WhitePoint components for ' + this.name + |
|
', no fallback available'); |
|
} |
|
|
|
if (this.XB < 0 || this.YB < 0 || this.ZB < 0) { |
|
info('Invalid BlackPoint for ' + this.name + ', falling back to default'); |
|
this.XB = this.YB = this.ZB = 0; |
|
} |
|
|
|
if (this.XB !== 0 || this.YB !== 0 || this.ZB !== 0) { |
|
warn(this.name + ', BlackPoint: XB: ' + this.XB + ', YB: ' + this.YB + |
|
', ZB: ' + this.ZB + ', only default values are supported.'); |
|
} |
|
|
|
if (this.G < 1) { |
|
info('Invalid Gamma: ' + this.G + ' for ' + this.name + |
|
', falling back to default'); |
|
this.G = 1; |
|
} |
|
} |
|
|
|
function convertToRgb(cs, src, srcOffset, dest, destOffset, scale) { |
|
// A represents a gray component of a calibrated gray space. |
|
// A <---> AG in the spec |
|
var A = src[srcOffset] * scale; |
|
var AG = Math.pow(A, cs.G); |
|
|
|
// Computes L as per spec. ( = cs.YW * AG ) |
|
// Except if other than default BlackPoint values are used. |
|
var L = cs.YW * AG; |
|
// http://www.poynton.com/notes/colour_and_gamma/ColorFAQ.html, Ch 4. |
|
// Convert values to rgb range [0, 255]. |
|
var val = Math.max(295.8 * Math.pow(L, 0.333333333333333333) - 40.8, 0) | 0; |
|
dest[destOffset] = val; |
|
dest[destOffset + 1] = val; |
|
dest[destOffset + 2] = val; |
|
} |
|
|
|
CalGrayCS.prototype = { |
|
getRgb: ColorSpace.prototype.getRgb, |
|
getRgbItem: function CalGrayCS_getRgbItem(src, srcOffset, |
|
dest, destOffset) { |
|
convertToRgb(this, src, srcOffset, dest, destOffset, 1); |
|
}, |
|
getRgbBuffer: function CalGrayCS_getRgbBuffer(src, srcOffset, count, |
|
dest, destOffset, bits, |
|
alpha01) { |
|
var scale = 1 / ((1 << bits) - 1); |
|
|
|
for (var i = 0; i < count; ++i) { |
|
convertToRgb(this, src, srcOffset, dest, destOffset, scale); |
|
srcOffset += 1; |
|
destOffset += 3 + alpha01; |
|
} |
|
}, |
|
getOutputLength: function CalGrayCS_getOutputLength(inputLength, alpha01) { |
|
return inputLength * (3 + alpha01); |
|
}, |
|
isPassthrough: ColorSpace.prototype.isPassthrough, |
|
fillRgb: ColorSpace.prototype.fillRgb, |
|
isDefaultDecode: function CalGrayCS_isDefaultDecode(decodeMap) { |
|
return ColorSpace.isDefaultDecode(decodeMap, this.numComps); |
|
}, |
|
usesZeroToOneRange: true |
|
}; |
|
return CalGrayCS; |
|
})(); |
|
|
|
// |
|
// CalRGBCS: Based on "PDF Reference, Sixth Ed", p.247 |
|
// |
|
var CalRGBCS = (function CalRGBCSClosure() { |
|
|
|
// See http://www.brucelindbloom.com/index.html?Eqn_ChromAdapt.html for these |
|
// matrices. |
|
var BRADFORD_SCALE_MATRIX = new Float32Array([ |
|
0.8951, 0.2664, -0.1614, |
|
-0.7502, 1.7135, 0.0367, |
|
0.0389, -0.0685, 1.0296]); |
|
|
|
var BRADFORD_SCALE_INVERSE_MATRIX = new Float32Array([ |
|
0.9869929, -0.1470543, 0.1599627, |
|
0.4323053, 0.5183603, 0.0492912, |
|
-0.0085287, 0.0400428, 0.9684867]); |
|
|
|
// See http://www.brucelindbloom.com/index.html?Eqn_RGB_XYZ_Matrix.html. |
|
var SRGB_D65_XYZ_TO_RGB_MATRIX = new Float32Array([ |
|
3.2404542, -1.5371385, -0.4985314, |
|
-0.9692660, 1.8760108, 0.0415560, |
|
0.0556434, -0.2040259, 1.0572252]); |
|
|
|
var FLAT_WHITEPOINT_MATRIX = new Float32Array([1, 1, 1]); |
|
|
|
var tempNormalizeMatrix = new Float32Array(3); |
|
var tempConvertMatrix1 = new Float32Array(3); |
|
var tempConvertMatrix2 = new Float32Array(3); |
|
|
|
var DECODE_L_CONSTANT = Math.pow(((8 + 16) / 116), 3) / 8.0; |
|
|
|
function CalRGBCS(whitePoint, blackPoint, gamma, matrix) { |
|
this.name = 'CalRGB'; |
|
this.numComps = 3; |
|
this.defaultColor = new Float32Array(3); |
|
|
|
if (!whitePoint) { |
|
error('WhitePoint missing - required for color space CalRGB'); |
|
} |
|
blackPoint = blackPoint || new Float32Array(3); |
|
gamma = gamma || new Float32Array([1, 1, 1]); |
|
matrix = matrix || new Float32Array([1, 0, 0, 0, 1, 0, 0, 0, 1]); |
|
|
|
// Translate arguments to spec variables. |
|
var XW = whitePoint[0]; |
|
var YW = whitePoint[1]; |
|
var ZW = whitePoint[2]; |
|
this.whitePoint = whitePoint; |
|
|
|
var XB = blackPoint[0]; |
|
var YB = blackPoint[1]; |
|
var ZB = blackPoint[2]; |
|
this.blackPoint = blackPoint; |
|
|
|
this.GR = gamma[0]; |
|
this.GG = gamma[1]; |
|
this.GB = gamma[2]; |
|
|
|
this.MXA = matrix[0]; |
|
this.MYA = matrix[1]; |
|
this.MZA = matrix[2]; |
|
this.MXB = matrix[3]; |
|
this.MYB = matrix[4]; |
|
this.MZB = matrix[5]; |
|
this.MXC = matrix[6]; |
|
this.MYC = matrix[7]; |
|
this.MZC = matrix[8]; |
|
|
|
// Validate variables as per spec. |
|
if (XW < 0 || ZW < 0 || YW !== 1) { |
|
error('Invalid WhitePoint components for ' + this.name + |
|
', no fallback available'); |
|
} |
|
|
|
if (XB < 0 || YB < 0 || ZB < 0) { |
|
info('Invalid BlackPoint for ' + this.name + ' [' + XB + ', ' + YB + |
|
', ' + ZB + '], falling back to default'); |
|
this.blackPoint = new Float32Array(3); |
|
} |
|
|
|
if (this.GR < 0 || this.GG < 0 || this.GB < 0) { |
|
info('Invalid Gamma [' + this.GR + ', ' + this.GG + ', ' + this.GB + |
|
'] for ' + this.name + ', falling back to default'); |
|
this.GR = this.GG = this.GB = 1; |
|
} |
|
|
|
if (this.MXA < 0 || this.MYA < 0 || this.MZA < 0 || |
|
this.MXB < 0 || this.MYB < 0 || this.MZB < 0 || |
|
this.MXC < 0 || this.MYC < 0 || this.MZC < 0) { |
|
info('Invalid Matrix for ' + this.name + ' [' + |
|
this.MXA + ', ' + this.MYA + ', ' + this.MZA + |
|
this.MXB + ', ' + this.MYB + ', ' + this.MZB + |
|
this.MXC + ', ' + this.MYC + ', ' + this.MZC + |
|
'], falling back to default'); |
|
this.MXA = this.MYB = this.MZC = 1; |
|
this.MXB = this.MYA = this.MZA = this.MXC = this.MYC = this.MZB = 0; |
|
} |
|
} |
|
|
|
function matrixProduct(a, b, result) { |
|
result[0] = a[0] * b[0] + a[1] * b[1] + a[2] * b[2]; |
|
result[1] = a[3] * b[0] + a[4] * b[1] + a[5] * b[2]; |
|
result[2] = a[6] * b[0] + a[7] * b[1] + a[8] * b[2]; |
|
} |
|
|
|
function convertToFlat(sourceWhitePoint, LMS, result) { |
|
result[0] = LMS[0] * 1 / sourceWhitePoint[0]; |
|
result[1] = LMS[1] * 1 / sourceWhitePoint[1]; |
|
result[2] = LMS[2] * 1 / sourceWhitePoint[2]; |
|
} |
|
|
|
function convertToD65(sourceWhitePoint, LMS, result) { |
|
var D65X = 0.95047; |
|
var D65Y = 1; |
|
var D65Z = 1.08883; |
|
|
|
result[0] = LMS[0] * D65X / sourceWhitePoint[0]; |
|
result[1] = LMS[1] * D65Y / sourceWhitePoint[1]; |
|
result[2] = LMS[2] * D65Z / sourceWhitePoint[2]; |
|
} |
|
|
|
function sRGBTransferFunction(color) { |
|
// See http://en.wikipedia.org/wiki/SRGB. |
|
if (color <= 0.0031308){ |
|
return adjustToRange(0, 1, 12.92 * color); |
|
} |
|
|
|
return adjustToRange(0, 1, (1 + 0.055) * Math.pow(color, 1 / 2.4) - 0.055); |
|
} |
|
|
|
function adjustToRange(min, max, value) { |
|
return Math.max(min, Math.min(max, value)); |
|
} |
|
|
|
function decodeL(L) { |
|
if (L < 0) { |
|
return -decodeL(-L); |
|
} |
|
|
|
if (L > 8.0) { |
|
return Math.pow(((L + 16) / 116), 3); |
|
} |
|
|
|
return L * DECODE_L_CONSTANT; |
|
} |
|
|
|
function compensateBlackPoint(sourceBlackPoint, XYZ_Flat, result) { |
|
|
|
// In case the blackPoint is already the default blackPoint then there is |
|
// no need to do compensation. |
|
if (sourceBlackPoint[0] === 0 && |
|
sourceBlackPoint[1] === 0 && |
|
sourceBlackPoint[2] === 0) { |
|
result[0] = XYZ_Flat[0]; |
|
result[1] = XYZ_Flat[1]; |
|
result[2] = XYZ_Flat[2]; |
|
return; |
|
} |
|
|
|
// For the blackPoint calculation details, please see |
|
// http://www.adobe.com/content/dam/Adobe/en/devnet/photoshop/sdk/ |
|
// AdobeBPC.pdf. |
|
// The destination blackPoint is the default blackPoint [0, 0, 0]. |
|
var zeroDecodeL = decodeL(0); |
|
|
|
var X_DST = zeroDecodeL; |
|
var X_SRC = decodeL(sourceBlackPoint[0]); |
|
|
|
var Y_DST = zeroDecodeL; |
|
var Y_SRC = decodeL(sourceBlackPoint[1]); |
|
|
|
var Z_DST = zeroDecodeL; |
|
var Z_SRC = decodeL(sourceBlackPoint[2]); |
|
|
|
var X_Scale = (1 - X_DST) / (1 - X_SRC); |
|
var X_Offset = 1 - X_Scale; |
|
|
|
var Y_Scale = (1 - Y_DST) / (1 - Y_SRC); |
|
var Y_Offset = 1 - Y_Scale; |
|
|
|
var Z_Scale = (1 - Z_DST) / (1 - Z_SRC); |
|
var Z_Offset = 1 - Z_Scale; |
|
|
|
result[0] = XYZ_Flat[0] * X_Scale + X_Offset; |
|
result[1] = XYZ_Flat[1] * Y_Scale + Y_Offset; |
|
result[2] = XYZ_Flat[2] * Z_Scale + Z_Offset; |
|
} |
|
|
|
function normalizeWhitePointToFlat(sourceWhitePoint, XYZ_In, result) { |
|
|
|
// In case the whitePoint is already flat then there is no need to do |
|
// normalization. |
|
if (sourceWhitePoint[0] === 1 && sourceWhitePoint[2] === 1) { |
|
result[0] = XYZ_In[0]; |
|
result[1] = XYZ_In[1]; |
|
result[2] = XYZ_In[2]; |
|
return; |
|
} |
|
|
|
var LMS = result; |
|
matrixProduct(BRADFORD_SCALE_MATRIX, XYZ_In, LMS); |
|
|
|
var LMS_Flat = tempNormalizeMatrix; |
|
convertToFlat(sourceWhitePoint, LMS, LMS_Flat); |
|
|
|
matrixProduct(BRADFORD_SCALE_INVERSE_MATRIX, LMS_Flat, result); |
|
} |
|
|
|
function normalizeWhitePointToD65(sourceWhitePoint, XYZ_In, result) { |
|
|
|
var LMS = result; |
|
matrixProduct(BRADFORD_SCALE_MATRIX, XYZ_In, LMS); |
|
|
|
var LMS_D65 = tempNormalizeMatrix; |
|
convertToD65(sourceWhitePoint, LMS, LMS_D65); |
|
|
|
matrixProduct(BRADFORD_SCALE_INVERSE_MATRIX, LMS_D65, result); |
|
} |
|
|
|
function convertToRgb(cs, src, srcOffset, dest, destOffset, scale) { |
|
// A, B and C represent a red, green and blue components of a calibrated |
|
// rgb space. |
|
var A = adjustToRange(0, 1, src[srcOffset] * scale); |
|
var B = adjustToRange(0, 1, src[srcOffset + 1] * scale); |
|
var C = adjustToRange(0, 1, src[srcOffset + 2] * scale); |
|
|
|
// A <---> AGR in the spec |
|
// B <---> BGG in the spec |
|
// C <---> CGB in the spec |
|
var AGR = Math.pow(A, cs.GR); |
|
var BGG = Math.pow(B, cs.GG); |
|
var CGB = Math.pow(C, cs.GB); |
|
|
|
// Computes intermediate variables L, M, N as per spec. |
|
// To decode X, Y, Z values map L, M, N directly to them. |
|
var X = cs.MXA * AGR + cs.MXB * BGG + cs.MXC * CGB; |
|
var Y = cs.MYA * AGR + cs.MYB * BGG + cs.MYC * CGB; |
|
var Z = cs.MZA * AGR + cs.MZB * BGG + cs.MZC * CGB; |
|
|
|
// The following calculations are based on this document: |
|
// http://www.adobe.com/content/dam/Adobe/en/devnet/photoshop/sdk/ |
|
// AdobeBPC.pdf. |
|
var XYZ = tempConvertMatrix1; |
|
XYZ[0] = X; |
|
XYZ[1] = Y; |
|
XYZ[2] = Z; |
|
var XYZ_Flat = tempConvertMatrix2; |
|
|
|
normalizeWhitePointToFlat(cs.whitePoint, XYZ, XYZ_Flat); |
|
|
|
var XYZ_Black = tempConvertMatrix1; |
|
compensateBlackPoint(cs.blackPoint, XYZ_Flat, XYZ_Black); |
|
|
|
var XYZ_D65 = tempConvertMatrix2; |
|
normalizeWhitePointToD65(FLAT_WHITEPOINT_MATRIX, XYZ_Black, XYZ_D65); |
|
|
|
var SRGB = tempConvertMatrix1; |
|
matrixProduct(SRGB_D65_XYZ_TO_RGB_MATRIX, XYZ_D65, SRGB); |
|
|
|
var sR = sRGBTransferFunction(SRGB[0]); |
|
var sG = sRGBTransferFunction(SRGB[1]); |
|
var sB = sRGBTransferFunction(SRGB[2]); |
|
|
|
// Convert the values to rgb range [0, 255]. |
|
dest[destOffset] = Math.round(sR * 255); |
|
dest[destOffset + 1] = Math.round(sG * 255); |
|
dest[destOffset + 2] = Math.round(sB * 255); |
|
} |
|
|
|
CalRGBCS.prototype = { |
|
getRgb: function CalRGBCS_getRgb(src, srcOffset) { |
|
var rgb = new Uint8Array(3); |
|
this.getRgbItem(src, srcOffset, rgb, 0); |
|
return rgb; |
|
}, |
|
getRgbItem: function CalRGBCS_getRgbItem(src, srcOffset, |
|
dest, destOffset) { |
|
convertToRgb(this, src, srcOffset, dest, destOffset, 1); |
|
}, |
|
getRgbBuffer: function CalRGBCS_getRgbBuffer(src, srcOffset, count, |
|
dest, destOffset, bits, |
|
alpha01) { |
|
var scale = 1 / ((1 << bits) - 1); |
|
|
|
for (var i = 0; i < count; ++i) { |
|
convertToRgb(this, src, srcOffset, dest, destOffset, scale); |
|
srcOffset += 3; |
|
destOffset += 3 + alpha01; |
|
} |
|
}, |
|
getOutputLength: function CalRGBCS_getOutputLength(inputLength, alpha01) { |
|
return (inputLength * (3 + alpha01) / 3) | 0; |
|
}, |
|
isPassthrough: ColorSpace.prototype.isPassthrough, |
|
fillRgb: ColorSpace.prototype.fillRgb, |
|
isDefaultDecode: function CalRGBCS_isDefaultDecode(decodeMap) { |
|
return ColorSpace.isDefaultDecode(decodeMap, this.numComps); |
|
}, |
|
usesZeroToOneRange: true |
|
}; |
|
return CalRGBCS; |
|
})(); |
|
|
|
// |
|
// LabCS: Based on "PDF Reference, Sixth Ed", p.250 |
|
// |
|
var LabCS = (function LabCSClosure() { |
|
function LabCS(whitePoint, blackPoint, range) { |
|
this.name = 'Lab'; |
|
this.numComps = 3; |
|
this.defaultColor = new Float32Array([0, 0, 0]); |
|
|
|
if (!whitePoint) { |
|
error('WhitePoint missing - required for color space Lab'); |
|
} |
|
blackPoint = blackPoint || [0, 0, 0]; |
|
range = range || [-100, 100, -100, 100]; |
|
|
|
// Translate args to spec variables |
|
this.XW = whitePoint[0]; |
|
this.YW = whitePoint[1]; |
|
this.ZW = whitePoint[2]; |
|
this.amin = range[0]; |
|
this.amax = range[1]; |
|
this.bmin = range[2]; |
|
this.bmax = range[3]; |
|
|
|
// These are here just for completeness - the spec doesn't offer any |
|
// formulas that use BlackPoint in Lab |
|
this.XB = blackPoint[0]; |
|
this.YB = blackPoint[1]; |
|
this.ZB = blackPoint[2]; |
|
|
|
// Validate vars as per spec |
|
if (this.XW < 0 || this.ZW < 0 || this.YW !== 1) { |
|
error('Invalid WhitePoint components, no fallback available'); |
|
} |
|
|
|
if (this.XB < 0 || this.YB < 0 || this.ZB < 0) { |
|
info('Invalid BlackPoint, falling back to default'); |
|
this.XB = this.YB = this.ZB = 0; |
|
} |
|
|
|
if (this.amin > this.amax || this.bmin > this.bmax) { |
|
info('Invalid Range, falling back to defaults'); |
|
this.amin = -100; |
|
this.amax = 100; |
|
this.bmin = -100; |
|
this.bmax = 100; |
|
} |
|
} |
|
|
|
// Function g(x) from spec |
|
function fn_g(x) { |
|
if (x >= 6 / 29) { |
|
return x * x * x; |
|
} else { |
|
return (108 / 841) * (x - 4 / 29); |
|
} |
|
} |
|
|
|
function decode(value, high1, low2, high2) { |
|
return low2 + (value) * (high2 - low2) / (high1); |
|
} |
|
|
|
// If decoding is needed maxVal should be 2^bits per component - 1. |
|
function convertToRgb(cs, src, srcOffset, maxVal, dest, destOffset) { |
|
// XXX: Lab input is in the range of [0, 100], [amin, amax], [bmin, bmax] |
|
// not the usual [0, 1]. If a command like setFillColor is used the src |
|
// values will already be within the correct range. However, if we are |
|
// converting an image we have to map the values to the correct range given |
|
// above. |
|
// Ls,as,bs <---> L*,a*,b* in the spec |
|
var Ls = src[srcOffset]; |
|
var as = src[srcOffset + 1]; |
|
var bs = src[srcOffset + 2]; |
|
if (maxVal !== false) { |
|
Ls = decode(Ls, maxVal, 0, 100); |
|
as = decode(as, maxVal, cs.amin, cs.amax); |
|
bs = decode(bs, maxVal, cs.bmin, cs.bmax); |
|
} |
|
|
|
// Adjust limits of 'as' and 'bs' |
|
as = as > cs.amax ? cs.amax : as < cs.amin ? cs.amin : as; |
|
bs = bs > cs.bmax ? cs.bmax : bs < cs.bmin ? cs.bmin : bs; |
|
|
|
// Computes intermediate variables X,Y,Z as per spec |
|
var M = (Ls + 16) / 116; |
|
var L = M + (as / 500); |
|
var N = M - (bs / 200); |
|
|
|
var X = cs.XW * fn_g(L); |
|
var Y = cs.YW * fn_g(M); |
|
var Z = cs.ZW * fn_g(N); |
|
|
|
var r, g, b; |
|
// Using different conversions for D50 and D65 white points, |
|
// per http://www.color.org/srgb.pdf |
|
if (cs.ZW < 1) { |
|
// Assuming D50 (X=0.9642, Y=1.00, Z=0.8249) |
|
r = X * 3.1339 + Y * -1.6170 + Z * -0.4906; |
|
g = X * -0.9785 + Y * 1.9160 + Z * 0.0333; |
|
b = X * 0.0720 + Y * -0.2290 + Z * 1.4057; |
|
} else { |
|
// Assuming D65 (X=0.9505, Y=1.00, Z=1.0888) |
|
r = X * 3.2406 + Y * -1.5372 + Z * -0.4986; |
|
g = X * -0.9689 + Y * 1.8758 + Z * 0.0415; |
|
b = X * 0.0557 + Y * -0.2040 + Z * 1.0570; |
|
} |
|
// clamp color values to [0,1] range then convert to [0,255] range. |
|
dest[destOffset] = r <= 0 ? 0 : r >= 1 ? 255 : Math.sqrt(r) * 255 | 0; |
|
dest[destOffset + 1] = g <= 0 ? 0 : g >= 1 ? 255 : Math.sqrt(g) * 255 | 0; |
|
dest[destOffset + 2] = b <= 0 ? 0 : b >= 1 ? 255 : Math.sqrt(b) * 255 | 0; |
|
} |
|
|
|
LabCS.prototype = { |
|
getRgb: ColorSpace.prototype.getRgb, |
|
getRgbItem: function LabCS_getRgbItem(src, srcOffset, dest, destOffset) { |
|
convertToRgb(this, src, srcOffset, false, dest, destOffset); |
|
}, |
|
getRgbBuffer: function LabCS_getRgbBuffer(src, srcOffset, count, |
|
dest, destOffset, bits, |
|
alpha01) { |
|
var maxVal = (1 << bits) - 1; |
|
for (var i = 0; i < count; i++) { |
|
convertToRgb(this, src, srcOffset, maxVal, dest, destOffset); |
|
srcOffset += 3; |
|
destOffset += 3 + alpha01; |
|
} |
|
}, |
|
getOutputLength: function LabCS_getOutputLength(inputLength, alpha01) { |
|
return (inputLength * (3 + alpha01) / 3) | 0; |
|
}, |
|
isPassthrough: ColorSpace.prototype.isPassthrough, |
|
fillRgb: ColorSpace.prototype.fillRgb, |
|
isDefaultDecode: function LabCS_isDefaultDecode(decodeMap) { |
|
// XXX: Decoding is handled with the lab conversion because of the strange |
|
// ranges that are used. |
|
return true; |
|
}, |
|
usesZeroToOneRange: false |
|
}; |
|
return LabCS; |
|
})(); |
|
|
|
// TODO refactor to remove dependency on image.js |
|
function _setCoreImage(coreImage_) { |
|
coreImage = coreImage_; |
|
PDFImage = coreImage_.PDFImage; |
|
} |
|
exports._setCoreImage = _setCoreImage; |
|
|
|
exports.ColorSpace = ColorSpace; |
|
|
|
// TODO refactor to remove dependency on colorspace.js |
|
coreStream._setCoreColorSpace(exports); |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreImage = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreColorSpace, root.pdfjsCoreStream, |
|
root.pdfjsCoreJpx); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreColorSpace, |
|
coreStream, coreJpx) { |
|
|
|
var ImageKind = sharedUtil.ImageKind; |
|
var assert = sharedUtil.assert; |
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isArray = sharedUtil.isArray; |
|
var warn = sharedUtil.warn; |
|
var Name = corePrimitives.Name; |
|
var isStream = corePrimitives.isStream; |
|
var ColorSpace = coreColorSpace.ColorSpace; |
|
var DecodeStream = coreStream.DecodeStream; |
|
var Stream = coreStream.Stream; |
|
var JpegStream = coreStream.JpegStream; |
|
var JpxImage = coreJpx.JpxImage; |
|
|
|
var PDFImage = (function PDFImageClosure() { |
|
/** |
|
* Decode the image in the main thread if it supported. Resovles the promise |
|
* when the image data is ready. |
|
*/ |
|
function handleImageData(handler, xref, res, image) { |
|
if (image instanceof JpegStream && image.isNativelyDecodable(xref, res)) { |
|
// For natively supported jpegs send them to the main thread for decoding. |
|
var dict = image.dict; |
|
var colorSpace = dict.get('ColorSpace', 'CS'); |
|
colorSpace = ColorSpace.parse(colorSpace, xref, res); |
|
var numComps = colorSpace.numComps; |
|
var decodePromise = handler.sendWithPromise('JpegDecode', |
|
[image.getIR(), numComps]); |
|
return decodePromise.then(function (message) { |
|
var data = message.data; |
|
return new Stream(data, 0, data.length, image.dict); |
|
}); |
|
} else { |
|
return Promise.resolve(image); |
|
} |
|
} |
|
|
|
/** |
|
* Decode and clamp a value. The formula is different from the spec because we |
|
* don't decode to float range [0,1], we decode it in the [0,max] range. |
|
*/ |
|
function decodeAndClamp(value, addend, coefficient, max) { |
|
value = addend + value * coefficient; |
|
// Clamp the value to the range |
|
return (value < 0 ? 0 : (value > max ? max : value)); |
|
} |
|
|
|
function PDFImage(xref, res, image, inline, smask, mask, isMask) { |
|
this.image = image; |
|
var dict = image.dict; |
|
if (dict.has('Filter')) { |
|
var filter = dict.get('Filter').name; |
|
if (filter === 'JPXDecode') { |
|
var jpxImage = new JpxImage(); |
|
jpxImage.parseImageProperties(image.stream); |
|
image.stream.reset(); |
|
image.bitsPerComponent = jpxImage.bitsPerComponent; |
|
image.numComps = jpxImage.componentsCount; |
|
} else if (filter === 'JBIG2Decode') { |
|
image.bitsPerComponent = 1; |
|
image.numComps = 1; |
|
} |
|
} |
|
// TODO cache rendered images? |
|
|
|
this.width = dict.get('Width', 'W'); |
|
this.height = dict.get('Height', 'H'); |
|
|
|
if (this.width < 1 || this.height < 1) { |
|
error('Invalid image width: ' + this.width + ' or height: ' + |
|
this.height); |
|
} |
|
|
|
this.interpolate = dict.get('Interpolate', 'I') || false; |
|
this.imageMask = dict.get('ImageMask', 'IM') || false; |
|
this.matte = dict.get('Matte') || false; |
|
|
|
var bitsPerComponent = image.bitsPerComponent; |
|
if (!bitsPerComponent) { |
|
bitsPerComponent = dict.get('BitsPerComponent', 'BPC'); |
|
if (!bitsPerComponent) { |
|
if (this.imageMask) { |
|
bitsPerComponent = 1; |
|
} else { |
|
error('Bits per component missing in image: ' + this.imageMask); |
|
} |
|
} |
|
} |
|
this.bpc = bitsPerComponent; |
|
|
|
if (!this.imageMask) { |
|
var colorSpace = dict.get('ColorSpace', 'CS'); |
|
if (!colorSpace) { |
|
info('JPX images (which do not require color spaces)'); |
|
switch (image.numComps) { |
|
case 1: |
|
colorSpace = Name.get('DeviceGray'); |
|
break; |
|
case 3: |
|
colorSpace = Name.get('DeviceRGB'); |
|
break; |
|
case 4: |
|
colorSpace = Name.get('DeviceCMYK'); |
|
break; |
|
default: |
|
error('JPX images with ' + this.numComps + |
|
' color components not supported.'); |
|
} |
|
} |
|
this.colorSpace = ColorSpace.parse(colorSpace, xref, res); |
|
this.numComps = this.colorSpace.numComps; |
|
} |
|
|
|
this.decode = dict.get('Decode', 'D'); |
|
this.needsDecode = false; |
|
if (this.decode && |
|
((this.colorSpace && !this.colorSpace.isDefaultDecode(this.decode)) || |
|
(isMask && !ColorSpace.isDefaultDecode(this.decode, 1)))) { |
|
this.needsDecode = true; |
|
// Do some preprocessing to avoid more math. |
|
var max = (1 << bitsPerComponent) - 1; |
|
this.decodeCoefficients = []; |
|
this.decodeAddends = []; |
|
for (var i = 0, j = 0; i < this.decode.length; i += 2, ++j) { |
|
var dmin = this.decode[i]; |
|
var dmax = this.decode[i + 1]; |
|
this.decodeCoefficients[j] = dmax - dmin; |
|
this.decodeAddends[j] = max * dmin; |
|
} |
|
} |
|
|
|
if (smask) { |
|
this.smask = new PDFImage(xref, res, smask, false); |
|
} else if (mask) { |
|
if (isStream(mask)) { |
|
var maskDict = mask.dict, imageMask = maskDict.get('ImageMask', 'IM'); |
|
if (!imageMask) { |
|
warn('Ignoring /Mask in image without /ImageMask.'); |
|
} else { |
|
this.mask = new PDFImage(xref, res, mask, false, null, null, true); |
|
} |
|
} else { |
|
// Color key mask (just an array). |
|
this.mask = mask; |
|
} |
|
} |
|
} |
|
/** |
|
* Handles processing of image data and returns the Promise that is resolved |
|
* with a PDFImage when the image is ready to be used. |
|
*/ |
|
PDFImage.buildImage = function PDFImage_buildImage(handler, xref, |
|
res, image, inline) { |
|
var imagePromise = handleImageData(handler, xref, res, image); |
|
var smaskPromise; |
|
var maskPromise; |
|
|
|
var smask = image.dict.get('SMask'); |
|
var mask = image.dict.get('Mask'); |
|
|
|
if (smask) { |
|
smaskPromise = handleImageData(handler, xref, res, smask); |
|
maskPromise = Promise.resolve(null); |
|
} else { |
|
smaskPromise = Promise.resolve(null); |
|
if (mask) { |
|
if (isStream(mask)) { |
|
maskPromise = handleImageData(handler, xref, res, mask); |
|
} else if (isArray(mask)) { |
|
maskPromise = Promise.resolve(mask); |
|
} else { |
|
warn('Unsupported mask format.'); |
|
maskPromise = Promise.resolve(null); |
|
} |
|
} else { |
|
maskPromise = Promise.resolve(null); |
|
} |
|
} |
|
return Promise.all([imagePromise, smaskPromise, maskPromise]).then( |
|
function(results) { |
|
var imageData = results[0]; |
|
var smaskData = results[1]; |
|
var maskData = results[2]; |
|
return new PDFImage(xref, res, imageData, inline, smaskData, maskData); |
|
}); |
|
}; |
|
|
|
/** |
|
* Resize an image using the nearest neighbor algorithm. Currently only |
|
* supports one and three component images. |
|
* @param {TypedArray} pixels The original image with one component. |
|
* @param {Number} bpc Number of bits per component. |
|
* @param {Number} components Number of color components, 1 or 3 is supported. |
|
* @param {Number} w1 Original width. |
|
* @param {Number} h1 Original height. |
|
* @param {Number} w2 New width. |
|
* @param {Number} h2 New height. |
|
* @param {TypedArray} dest (Optional) The destination buffer. |
|
* @param {Number} alpha01 (Optional) Size reserved for the alpha channel. |
|
* @return {TypedArray} Resized image data. |
|
*/ |
|
PDFImage.resize = function PDFImage_resize(pixels, bpc, components, |
|
w1, h1, w2, h2, dest, alpha01) { |
|
|
|
if (components !== 1 && components !== 3) { |
|
error('Unsupported component count for resizing.'); |
|
} |
|
|
|
var length = w2 * h2 * components; |
|
var temp = dest ? dest : (bpc <= 8 ? new Uint8Array(length) : |
|
(bpc <= 16 ? new Uint16Array(length) : new Uint32Array(length))); |
|
var xRatio = w1 / w2; |
|
var yRatio = h1 / h2; |
|
var i, j, py, newIndex = 0, oldIndex; |
|
var xScaled = new Uint16Array(w2); |
|
var w1Scanline = w1 * components; |
|
if (alpha01 !== 1) { |
|
alpha01 = 0; |
|
} |
|
|
|
for (j = 0; j < w2; j++) { |
|
xScaled[j] = Math.floor(j * xRatio) * components; |
|
} |
|
|
|
if (components === 1) { |
|
for (i = 0; i < h2; i++) { |
|
py = Math.floor(i * yRatio) * w1Scanline; |
|
for (j = 0; j < w2; j++) { |
|
oldIndex = py + xScaled[j]; |
|
temp[newIndex++] = pixels[oldIndex]; |
|
} |
|
} |
|
} else if (components === 3) { |
|
for (i = 0; i < h2; i++) { |
|
py = Math.floor(i * yRatio) * w1Scanline; |
|
for (j = 0; j < w2; j++) { |
|
oldIndex = py + xScaled[j]; |
|
temp[newIndex++] = pixels[oldIndex++]; |
|
temp[newIndex++] = pixels[oldIndex++]; |
|
temp[newIndex++] = pixels[oldIndex++]; |
|
newIndex += alpha01; |
|
} |
|
} |
|
} |
|
return temp; |
|
}; |
|
|
|
PDFImage.createMask = |
|
function PDFImage_createMask(imgArray, width, height, |
|
imageIsFromDecodeStream, inverseDecode) { |
|
|
|
// |imgArray| might not contain full data for every pixel of the mask, so |
|
// we need to distinguish between |computedLength| and |actualLength|. |
|
// In particular, if inverseDecode is true, then the array we return must |
|
// have a length of |computedLength|. |
|
|
|
var computedLength = ((width + 7) >> 3) * height; |
|
var actualLength = imgArray.byteLength; |
|
var haveFullData = computedLength === actualLength; |
|
var data, i; |
|
|
|
if (imageIsFromDecodeStream && (!inverseDecode || haveFullData)) { |
|
// imgArray came from a DecodeStream and its data is in an appropriate |
|
// form, so we can just transfer it. |
|
data = imgArray; |
|
} else if (!inverseDecode) { |
|
data = new Uint8Array(actualLength); |
|
data.set(imgArray); |
|
} else { |
|
data = new Uint8Array(computedLength); |
|
data.set(imgArray); |
|
for (i = actualLength; i < computedLength; i++) { |
|
data[i] = 0xff; |
|
} |
|
} |
|
|
|
// If necessary, invert the original mask data (but not any extra we might |
|
// have added above). It's safe to modify the array -- whether it's the |
|
// original or a copy, we're about to transfer it anyway, so nothing else |
|
// in this thread can be relying on its contents. |
|
if (inverseDecode) { |
|
for (i = 0; i < actualLength; i++) { |
|
data[i] = ~data[i]; |
|
} |
|
} |
|
|
|
return {data: data, width: width, height: height}; |
|
}; |
|
|
|
PDFImage.prototype = { |
|
get drawWidth() { |
|
return Math.max(this.width, |
|
this.smask && this.smask.width || 0, |
|
this.mask && this.mask.width || 0); |
|
}, |
|
|
|
get drawHeight() { |
|
return Math.max(this.height, |
|
this.smask && this.smask.height || 0, |
|
this.mask && this.mask.height || 0); |
|
}, |
|
|
|
decodeBuffer: function PDFImage_decodeBuffer(buffer) { |
|
var bpc = this.bpc; |
|
var numComps = this.numComps; |
|
|
|
var decodeAddends = this.decodeAddends; |
|
var decodeCoefficients = this.decodeCoefficients; |
|
var max = (1 << bpc) - 1; |
|
var i, ii; |
|
|
|
if (bpc === 1) { |
|
// If the buffer needed decode that means it just needs to be inverted. |
|
for (i = 0, ii = buffer.length; i < ii; i++) { |
|
buffer[i] = +!(buffer[i]); |
|
} |
|
return; |
|
} |
|
var index = 0; |
|
for (i = 0, ii = this.width * this.height; i < ii; i++) { |
|
for (var j = 0; j < numComps; j++) { |
|
buffer[index] = decodeAndClamp(buffer[index], decodeAddends[j], |
|
decodeCoefficients[j], max); |
|
index++; |
|
} |
|
} |
|
}, |
|
|
|
getComponents: function PDFImage_getComponents(buffer) { |
|
var bpc = this.bpc; |
|
|
|
// This image doesn't require any extra work. |
|
if (bpc === 8) { |
|
return buffer; |
|
} |
|
|
|
var width = this.width; |
|
var height = this.height; |
|
var numComps = this.numComps; |
|
|
|
var length = width * height * numComps; |
|
var bufferPos = 0; |
|
var output = (bpc <= 8 ? new Uint8Array(length) : |
|
(bpc <= 16 ? new Uint16Array(length) : new Uint32Array(length))); |
|
var rowComps = width * numComps; |
|
|
|
var max = (1 << bpc) - 1; |
|
var i = 0, ii, buf; |
|
|
|
if (bpc === 1) { |
|
// Optimization for reading 1 bpc images. |
|
var mask, loop1End, loop2End; |
|
for (var j = 0; j < height; j++) { |
|
loop1End = i + (rowComps & ~7); |
|
loop2End = i + rowComps; |
|
|
|
// unroll loop for all full bytes |
|
while (i < loop1End) { |
|
buf = buffer[bufferPos++]; |
|
output[i] = (buf >> 7) & 1; |
|
output[i + 1] = (buf >> 6) & 1; |
|
output[i + 2] = (buf >> 5) & 1; |
|
output[i + 3] = (buf >> 4) & 1; |
|
output[i + 4] = (buf >> 3) & 1; |
|
output[i + 5] = (buf >> 2) & 1; |
|
output[i + 6] = (buf >> 1) & 1; |
|
output[i + 7] = buf & 1; |
|
i += 8; |
|
} |
|
|
|
// handle remaing bits |
|
if (i < loop2End) { |
|
buf = buffer[bufferPos++]; |
|
mask = 128; |
|
while (i < loop2End) { |
|
output[i++] = +!!(buf & mask); |
|
mask >>= 1; |
|
} |
|
} |
|
} |
|
} else { |
|
// The general case that handles all other bpc values. |
|
var bits = 0; |
|
buf = 0; |
|
for (i = 0, ii = length; i < ii; ++i) { |
|
if (i % rowComps === 0) { |
|
buf = 0; |
|
bits = 0; |
|
} |
|
|
|
while (bits < bpc) { |
|
buf = (buf << 8) | buffer[bufferPos++]; |
|
bits += 8; |
|
} |
|
|
|
var remainingBits = bits - bpc; |
|
var value = buf >> remainingBits; |
|
output[i] = (value < 0 ? 0 : (value > max ? max : value)); |
|
buf = buf & ((1 << remainingBits) - 1); |
|
bits = remainingBits; |
|
} |
|
} |
|
return output; |
|
}, |
|
|
|
fillOpacity: function PDFImage_fillOpacity(rgbaBuf, width, height, |
|
actualHeight, image) { |
|
var smask = this.smask; |
|
var mask = this.mask; |
|
var alphaBuf, sw, sh, i, ii, j; |
|
|
|
if (smask) { |
|
sw = smask.width; |
|
sh = smask.height; |
|
alphaBuf = new Uint8Array(sw * sh); |
|
smask.fillGrayBuffer(alphaBuf); |
|
if (sw !== width || sh !== height) { |
|
alphaBuf = PDFImage.resize(alphaBuf, smask.bpc, 1, sw, sh, width, |
|
height); |
|
} |
|
} else if (mask) { |
|
if (mask instanceof PDFImage) { |
|
sw = mask.width; |
|
sh = mask.height; |
|
alphaBuf = new Uint8Array(sw * sh); |
|
mask.numComps = 1; |
|
mask.fillGrayBuffer(alphaBuf); |
|
|
|
// Need to invert values in rgbaBuf |
|
for (i = 0, ii = sw * sh; i < ii; ++i) { |
|
alphaBuf[i] = 255 - alphaBuf[i]; |
|
} |
|
|
|
if (sw !== width || sh !== height) { |
|
alphaBuf = PDFImage.resize(alphaBuf, mask.bpc, 1, sw, sh, width, |
|
height); |
|
} |
|
} else if (isArray(mask)) { |
|
// Color key mask: if any of the compontents are outside the range |
|
// then they should be painted. |
|
alphaBuf = new Uint8Array(width * height); |
|
var numComps = this.numComps; |
|
for (i = 0, ii = width * height; i < ii; ++i) { |
|
var opacity = 0; |
|
var imageOffset = i * numComps; |
|
for (j = 0; j < numComps; ++j) { |
|
var color = image[imageOffset + j]; |
|
var maskOffset = j * 2; |
|
if (color < mask[maskOffset] || color > mask[maskOffset + 1]) { |
|
opacity = 255; |
|
break; |
|
} |
|
} |
|
alphaBuf[i] = opacity; |
|
} |
|
} else { |
|
error('Unknown mask format.'); |
|
} |
|
} |
|
|
|
if (alphaBuf) { |
|
for (i = 0, j = 3, ii = width * actualHeight; i < ii; ++i, j += 4) { |
|
rgbaBuf[j] = alphaBuf[i]; |
|
} |
|
} else { |
|
// No mask. |
|
for (i = 0, j = 3, ii = width * actualHeight; i < ii; ++i, j += 4) { |
|
rgbaBuf[j] = 255; |
|
} |
|
} |
|
}, |
|
|
|
undoPreblend: function PDFImage_undoPreblend(buffer, width, height) { |
|
var matte = this.smask && this.smask.matte; |
|
if (!matte) { |
|
return; |
|
} |
|
var matteRgb = this.colorSpace.getRgb(matte, 0); |
|
var matteR = matteRgb[0]; |
|
var matteG = matteRgb[1]; |
|
var matteB = matteRgb[2]; |
|
var length = width * height * 4; |
|
var r, g, b; |
|
for (var i = 0; i < length; i += 4) { |
|
var alpha = buffer[i + 3]; |
|
if (alpha === 0) { |
|
// according formula we have to get Infinity in all components |
|
// making it white (typical paper color) should be okay |
|
buffer[i] = 255; |
|
buffer[i + 1] = 255; |
|
buffer[i + 2] = 255; |
|
continue; |
|
} |
|
var k = 255 / alpha; |
|
r = (buffer[i] - matteR) * k + matteR; |
|
g = (buffer[i + 1] - matteG) * k + matteG; |
|
b = (buffer[i + 2] - matteB) * k + matteB; |
|
buffer[i] = r <= 0 ? 0 : r >= 255 ? 255 : r | 0; |
|
buffer[i + 1] = g <= 0 ? 0 : g >= 255 ? 255 : g | 0; |
|
buffer[i + 2] = b <= 0 ? 0 : b >= 255 ? 255 : b | 0; |
|
} |
|
}, |
|
|
|
createImageData: function PDFImage_createImageData(forceRGBA) { |
|
var drawWidth = this.drawWidth; |
|
var drawHeight = this.drawHeight; |
|
var imgData = { // other fields are filled in below |
|
width: drawWidth, |
|
height: drawHeight |
|
}; |
|
|
|
var numComps = this.numComps; |
|
var originalWidth = this.width; |
|
var originalHeight = this.height; |
|
var bpc = this.bpc; |
|
|
|
// Rows start at byte boundary. |
|
var rowBytes = (originalWidth * numComps * bpc + 7) >> 3; |
|
var imgArray; |
|
|
|
if (!forceRGBA) { |
|
// If it is a 1-bit-per-pixel grayscale (i.e. black-and-white) image |
|
// without any complications, we pass a same-sized copy to the main |
|
// thread rather than expanding by 32x to RGBA form. This saves *lots* |
|
// of memory for many scanned documents. It's also much faster. |
|
// |
|
// Similarly, if it is a 24-bit-per pixel RGB image without any |
|
// complications, we avoid expanding by 1.333x to RGBA form. |
|
var kind; |
|
if (this.colorSpace.name === 'DeviceGray' && bpc === 1) { |
|
kind = ImageKind.GRAYSCALE_1BPP; |
|
} else if (this.colorSpace.name === 'DeviceRGB' && bpc === 8 && |
|
!this.needsDecode) { |
|
kind = ImageKind.RGB_24BPP; |
|
} |
|
if (kind && !this.smask && !this.mask && |
|
drawWidth === originalWidth && drawHeight === originalHeight) { |
|
imgData.kind = kind; |
|
|
|
imgArray = this.getImageBytes(originalHeight * rowBytes); |
|
// If imgArray came from a DecodeStream, we're safe to transfer it |
|
// (and thus neuter it) because it will constitute the entire |
|
// DecodeStream's data. But if it came from a Stream, we need to |
|
// copy it because it'll only be a portion of the Stream's data, and |
|
// the rest will be read later on. |
|
if (this.image instanceof DecodeStream) { |
|
imgData.data = imgArray; |
|
} else { |
|
var newArray = new Uint8Array(imgArray.length); |
|
newArray.set(imgArray); |
|
imgData.data = newArray; |
|
} |
|
if (this.needsDecode) { |
|
// Invert the buffer (which must be grayscale if we reached here). |
|
assert(kind === ImageKind.GRAYSCALE_1BPP); |
|
var buffer = imgData.data; |
|
for (var i = 0, ii = buffer.length; i < ii; i++) { |
|
buffer[i] ^= 0xff; |
|
} |
|
} |
|
return imgData; |
|
} |
|
if (this.image instanceof JpegStream && !this.smask && !this.mask && |
|
(this.colorSpace.name === 'DeviceGray' || |
|
this.colorSpace.name === 'DeviceRGB' || |
|
this.colorSpace.name === 'DeviceCMYK')) { |
|
imgData.kind = ImageKind.RGB_24BPP; |
|
imgData.data = this.getImageBytes(originalHeight * rowBytes, |
|
drawWidth, drawHeight, true); |
|
return imgData; |
|
} |
|
} |
|
|
|
imgArray = this.getImageBytes(originalHeight * rowBytes); |
|
// imgArray can be incomplete (e.g. after CCITT fax encoding). |
|
var actualHeight = 0 | (imgArray.length / rowBytes * |
|
drawHeight / originalHeight); |
|
|
|
var comps = this.getComponents(imgArray); |
|
|
|
// If opacity data is present, use RGBA_32BPP form. Otherwise, use the |
|
// more compact RGB_24BPP form if allowable. |
|
var alpha01, maybeUndoPreblend; |
|
if (!forceRGBA && !this.smask && !this.mask) { |
|
imgData.kind = ImageKind.RGB_24BPP; |
|
imgData.data = new Uint8Array(drawWidth * drawHeight * 3); |
|
alpha01 = 0; |
|
maybeUndoPreblend = false; |
|
} else { |
|
imgData.kind = ImageKind.RGBA_32BPP; |
|
imgData.data = new Uint8Array(drawWidth * drawHeight * 4); |
|
alpha01 = 1; |
|
maybeUndoPreblend = true; |
|
|
|
// Color key masking (opacity) must be performed before decoding. |
|
this.fillOpacity(imgData.data, drawWidth, drawHeight, actualHeight, |
|
comps); |
|
} |
|
|
|
if (this.needsDecode) { |
|
this.decodeBuffer(comps); |
|
} |
|
this.colorSpace.fillRgb(imgData.data, originalWidth, originalHeight, |
|
drawWidth, drawHeight, actualHeight, bpc, comps, |
|
alpha01); |
|
if (maybeUndoPreblend) { |
|
this.undoPreblend(imgData.data, drawWidth, actualHeight); |
|
} |
|
|
|
return imgData; |
|
}, |
|
|
|
fillGrayBuffer: function PDFImage_fillGrayBuffer(buffer) { |
|
var numComps = this.numComps; |
|
if (numComps !== 1) { |
|
error('Reading gray scale from a color image: ' + numComps); |
|
} |
|
|
|
var width = this.width; |
|
var height = this.height; |
|
var bpc = this.bpc; |
|
|
|
// rows start at byte boundary |
|
var rowBytes = (width * numComps * bpc + 7) >> 3; |
|
var imgArray = this.getImageBytes(height * rowBytes); |
|
|
|
var comps = this.getComponents(imgArray); |
|
var i, length; |
|
|
|
if (bpc === 1) { |
|
// inline decoding (= inversion) for 1 bpc images |
|
length = width * height; |
|
if (this.needsDecode) { |
|
// invert and scale to {0, 255} |
|
for (i = 0; i < length; ++i) { |
|
buffer[i] = (comps[i] - 1) & 255; |
|
} |
|
} else { |
|
// scale to {0, 255} |
|
for (i = 0; i < length; ++i) { |
|
buffer[i] = (-comps[i]) & 255; |
|
} |
|
} |
|
return; |
|
} |
|
|
|
if (this.needsDecode) { |
|
this.decodeBuffer(comps); |
|
} |
|
length = width * height; |
|
// we aren't using a colorspace so we need to scale the value |
|
var scale = 255 / ((1 << bpc) - 1); |
|
for (i = 0; i < length; ++i) { |
|
buffer[i] = (scale * comps[i]) | 0; |
|
} |
|
}, |
|
|
|
getImageBytes: function PDFImage_getImageBytes(length, |
|
drawWidth, drawHeight, |
|
forceRGB) { |
|
this.image.reset(); |
|
this.image.drawWidth = drawWidth || this.width; |
|
this.image.drawHeight = drawHeight || this.height; |
|
this.image.forceRGB = !!forceRGB; |
|
return this.image.getBytes(length); |
|
} |
|
}; |
|
return PDFImage; |
|
})(); |
|
|
|
exports.PDFImage = PDFImage; |
|
|
|
// TODO refactor to remove dependency on colorspace.js |
|
coreColorSpace._setCoreImage(exports); |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCorePattern = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreFunction, |
|
root.pdfjsCoreColorSpace); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreFunction, |
|
coreColorSpace) { |
|
|
|
var UNSUPPORTED_FEATURES = sharedUtil.UNSUPPORTED_FEATURES; |
|
var MissingDataException = sharedUtil.MissingDataException; |
|
var Util = sharedUtil.Util; |
|
var assert = sharedUtil.assert; |
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var warn = sharedUtil.warn; |
|
var isStream = corePrimitives.isStream; |
|
var PDFFunction = coreFunction.PDFFunction; |
|
var ColorSpace = coreColorSpace.ColorSpace; |
|
|
|
var ShadingType = { |
|
FUNCTION_BASED: 1, |
|
AXIAL: 2, |
|
RADIAL: 3, |
|
FREE_FORM_MESH: 4, |
|
LATTICE_FORM_MESH: 5, |
|
COONS_PATCH_MESH: 6, |
|
TENSOR_PATCH_MESH: 7 |
|
}; |
|
|
|
var Pattern = (function PatternClosure() { |
|
// Constructor should define this.getPattern |
|
function Pattern() { |
|
error('should not call Pattern constructor'); |
|
} |
|
|
|
Pattern.prototype = { |
|
// Input: current Canvas context |
|
// Output: the appropriate fillStyle or strokeStyle |
|
getPattern: function Pattern_getPattern(ctx) { |
|
error('Should not call Pattern.getStyle: ' + ctx); |
|
} |
|
}; |
|
|
|
Pattern.parseShading = function Pattern_parseShading(shading, matrix, xref, |
|
res, handler) { |
|
|
|
var dict = isStream(shading) ? shading.dict : shading; |
|
var type = dict.get('ShadingType'); |
|
|
|
try { |
|
switch (type) { |
|
case ShadingType.AXIAL: |
|
case ShadingType.RADIAL: |
|
// Both radial and axial shadings are handled by RadialAxial shading. |
|
return new Shadings.RadialAxial(dict, matrix, xref, res); |
|
case ShadingType.FREE_FORM_MESH: |
|
case ShadingType.LATTICE_FORM_MESH: |
|
case ShadingType.COONS_PATCH_MESH: |
|
case ShadingType.TENSOR_PATCH_MESH: |
|
return new Shadings.Mesh(shading, matrix, xref, res); |
|
default: |
|
throw new Error('Unsupported ShadingType: ' + type); |
|
} |
|
} catch (ex) { |
|
if (ex instanceof MissingDataException) { |
|
throw ex; |
|
} |
|
handler.send('UnsupportedFeature', |
|
{featureId: UNSUPPORTED_FEATURES.shadingPattern}); |
|
warn(ex); |
|
return new Shadings.Dummy(); |
|
} |
|
}; |
|
return Pattern; |
|
})(); |
|
|
|
var Shadings = {}; |
|
|
|
// A small number to offset the first/last color stops so we can insert ones to |
|
// support extend. Number.MIN_VALUE is too small and breaks the extend. |
|
Shadings.SMALL_NUMBER = 1e-6; |
|
|
|
// Radial and axial shading have very similar implementations |
|
// If needed, the implementations can be broken into two classes |
|
Shadings.RadialAxial = (function RadialAxialClosure() { |
|
function RadialAxial(dict, matrix, xref, res) { |
|
this.matrix = matrix; |
|
this.coordsArr = dict.get('Coords'); |
|
this.shadingType = dict.get('ShadingType'); |
|
this.type = 'Pattern'; |
|
var cs = dict.get('ColorSpace', 'CS'); |
|
cs = ColorSpace.parse(cs, xref, res); |
|
this.cs = cs; |
|
|
|
var t0 = 0.0, t1 = 1.0; |
|
if (dict.has('Domain')) { |
|
var domainArr = dict.get('Domain'); |
|
t0 = domainArr[0]; |
|
t1 = domainArr[1]; |
|
} |
|
|
|
var extendStart = false, extendEnd = false; |
|
if (dict.has('Extend')) { |
|
var extendArr = dict.get('Extend'); |
|
extendStart = extendArr[0]; |
|
extendEnd = extendArr[1]; |
|
} |
|
|
|
if (this.shadingType === ShadingType.RADIAL && |
|
(!extendStart || !extendEnd)) { |
|
// Radial gradient only currently works if either circle is fully within |
|
// the other circle. |
|
var x1 = this.coordsArr[0]; |
|
var y1 = this.coordsArr[1]; |
|
var r1 = this.coordsArr[2]; |
|
var x2 = this.coordsArr[3]; |
|
var y2 = this.coordsArr[4]; |
|
var r2 = this.coordsArr[5]; |
|
var distance = Math.sqrt((x1 - x2) * (x1 - x2) + (y1 - y2) * (y1 - y2)); |
|
if (r1 <= r2 + distance && |
|
r2 <= r1 + distance) { |
|
warn('Unsupported radial gradient.'); |
|
} |
|
} |
|
|
|
this.extendStart = extendStart; |
|
this.extendEnd = extendEnd; |
|
|
|
var fnObj = dict.get('Function'); |
|
var fn = PDFFunction.parseArray(xref, fnObj); |
|
|
|
// 10 samples seems good enough for now, but probably won't work |
|
// if there are sharp color changes. Ideally, we would implement |
|
// the spec faithfully and add lossless optimizations. |
|
var diff = t1 - t0; |
|
var step = diff / 10; |
|
|
|
var colorStops = this.colorStops = []; |
|
|
|
// Protect against bad domains so we don't end up in an infinte loop below. |
|
if (t0 >= t1 || step <= 0) { |
|
// Acrobat doesn't seem to handle these cases so we'll ignore for |
|
// now. |
|
info('Bad shading domain.'); |
|
return; |
|
} |
|
|
|
var color = new Float32Array(cs.numComps), ratio = new Float32Array(1); |
|
var rgbColor; |
|
for (var i = t0; i <= t1; i += step) { |
|
ratio[0] = i; |
|
fn(ratio, 0, color, 0); |
|
rgbColor = cs.getRgb(color, 0); |
|
var cssColor = Util.makeCssRgb(rgbColor[0], rgbColor[1], rgbColor[2]); |
|
colorStops.push([(i - t0) / diff, cssColor]); |
|
} |
|
|
|
var background = 'transparent'; |
|
if (dict.has('Background')) { |
|
rgbColor = cs.getRgb(dict.get('Background'), 0); |
|
background = Util.makeCssRgb(rgbColor[0], rgbColor[1], rgbColor[2]); |
|
} |
|
|
|
if (!extendStart) { |
|
// Insert a color stop at the front and offset the first real color stop |
|
// so it doesn't conflict with the one we insert. |
|
colorStops.unshift([0, background]); |
|
colorStops[1][0] += Shadings.SMALL_NUMBER; |
|
} |
|
if (!extendEnd) { |
|
// Same idea as above in extendStart but for the end. |
|
colorStops[colorStops.length - 1][0] -= Shadings.SMALL_NUMBER; |
|
colorStops.push([1, background]); |
|
} |
|
|
|
this.colorStops = colorStops; |
|
} |
|
|
|
RadialAxial.prototype = { |
|
getIR: function RadialAxial_getIR() { |
|
var coordsArr = this.coordsArr; |
|
var shadingType = this.shadingType; |
|
var type, p0, p1, r0, r1; |
|
if (shadingType === ShadingType.AXIAL) { |
|
p0 = [coordsArr[0], coordsArr[1]]; |
|
p1 = [coordsArr[2], coordsArr[3]]; |
|
r0 = null; |
|
r1 = null; |
|
type = 'axial'; |
|
} else if (shadingType === ShadingType.RADIAL) { |
|
p0 = [coordsArr[0], coordsArr[1]]; |
|
p1 = [coordsArr[3], coordsArr[4]]; |
|
r0 = coordsArr[2]; |
|
r1 = coordsArr[5]; |
|
type = 'radial'; |
|
} else { |
|
error('getPattern type unknown: ' + shadingType); |
|
} |
|
|
|
var matrix = this.matrix; |
|
if (matrix) { |
|
p0 = Util.applyTransform(p0, matrix); |
|
p1 = Util.applyTransform(p1, matrix); |
|
if (shadingType === ShadingType.RADIAL) { |
|
var scale = Util.singularValueDecompose2dScale(matrix); |
|
r0 *= scale[0]; |
|
r1 *= scale[1]; |
|
} |
|
} |
|
|
|
return ['RadialAxial', type, this.colorStops, p0, p1, r0, r1]; |
|
} |
|
}; |
|
|
|
return RadialAxial; |
|
})(); |
|
|
|
// All mesh shading. For now, they will be presented as set of the triangles |
|
// to be drawn on the canvas and rgb color for each vertex. |
|
Shadings.Mesh = (function MeshClosure() { |
|
function MeshStreamReader(stream, context) { |
|
this.stream = stream; |
|
this.context = context; |
|
this.buffer = 0; |
|
this.bufferLength = 0; |
|
|
|
var numComps = context.numComps; |
|
this.tmpCompsBuf = new Float32Array(numComps); |
|
var csNumComps = context.colorSpace.numComps; |
|
this.tmpCsCompsBuf = context.colorFn ? new Float32Array(csNumComps) : |
|
this.tmpCompsBuf; |
|
} |
|
MeshStreamReader.prototype = { |
|
get hasData() { |
|
if (this.stream.end) { |
|
return this.stream.pos < this.stream.end; |
|
} |
|
if (this.bufferLength > 0) { |
|
return true; |
|
} |
|
var nextByte = this.stream.getByte(); |
|
if (nextByte < 0) { |
|
return false; |
|
} |
|
this.buffer = nextByte; |
|
this.bufferLength = 8; |
|
return true; |
|
}, |
|
readBits: function MeshStreamReader_readBits(n) { |
|
var buffer = this.buffer; |
|
var bufferLength = this.bufferLength; |
|
if (n === 32) { |
|
if (bufferLength === 0) { |
|
return ((this.stream.getByte() << 24) | |
|
(this.stream.getByte() << 16) | (this.stream.getByte() << 8) | |
|
this.stream.getByte()) >>> 0; |
|
} |
|
buffer = (buffer << 24) | (this.stream.getByte() << 16) | |
|
(this.stream.getByte() << 8) | this.stream.getByte(); |
|
var nextByte = this.stream.getByte(); |
|
this.buffer = nextByte & ((1 << bufferLength) - 1); |
|
return ((buffer << (8 - bufferLength)) | |
|
((nextByte & 0xFF) >> bufferLength)) >>> 0; |
|
} |
|
if (n === 8 && bufferLength === 0) { |
|
return this.stream.getByte(); |
|
} |
|
while (bufferLength < n) { |
|
buffer = (buffer << 8) | this.stream.getByte(); |
|
bufferLength += 8; |
|
} |
|
bufferLength -= n; |
|
this.bufferLength = bufferLength; |
|
this.buffer = buffer & ((1 << bufferLength) - 1); |
|
return buffer >> bufferLength; |
|
}, |
|
align: function MeshStreamReader_align() { |
|
this.buffer = 0; |
|
this.bufferLength = 0; |
|
}, |
|
readFlag: function MeshStreamReader_readFlag() { |
|
return this.readBits(this.context.bitsPerFlag); |
|
}, |
|
readCoordinate: function MeshStreamReader_readCoordinate() { |
|
var bitsPerCoordinate = this.context.bitsPerCoordinate; |
|
var xi = this.readBits(bitsPerCoordinate); |
|
var yi = this.readBits(bitsPerCoordinate); |
|
var decode = this.context.decode; |
|
var scale = bitsPerCoordinate < 32 ? 1 / ((1 << bitsPerCoordinate) - 1) : |
|
2.3283064365386963e-10; // 2 ^ -32 |
|
return [ |
|
xi * scale * (decode[1] - decode[0]) + decode[0], |
|
yi * scale * (decode[3] - decode[2]) + decode[2] |
|
]; |
|
}, |
|
readComponents: function MeshStreamReader_readComponents() { |
|
var numComps = this.context.numComps; |
|
var bitsPerComponent = this.context.bitsPerComponent; |
|
var scale = bitsPerComponent < 32 ? 1 / ((1 << bitsPerComponent) - 1) : |
|
2.3283064365386963e-10; // 2 ^ -32 |
|
var decode = this.context.decode; |
|
var components = this.tmpCompsBuf; |
|
for (var i = 0, j = 4; i < numComps; i++, j += 2) { |
|
var ci = this.readBits(bitsPerComponent); |
|
components[i] = ci * scale * (decode[j + 1] - decode[j]) + decode[j]; |
|
} |
|
var color = this.tmpCsCompsBuf; |
|
if (this.context.colorFn) { |
|
this.context.colorFn(components, 0, color, 0); |
|
} |
|
return this.context.colorSpace.getRgb(color, 0); |
|
} |
|
}; |
|
|
|
function decodeType4Shading(mesh, reader) { |
|
var coords = mesh.coords; |
|
var colors = mesh.colors; |
|
var operators = []; |
|
var ps = []; // not maintaining cs since that will match ps |
|
var verticesLeft = 0; // assuming we have all data to start a new triangle |
|
while (reader.hasData) { |
|
var f = reader.readFlag(); |
|
var coord = reader.readCoordinate(); |
|
var color = reader.readComponents(); |
|
if (verticesLeft === 0) { // ignoring flags if we started a triangle |
|
assert(0 <= f && f <= 2, 'Unknown type4 flag'); |
|
switch (f) { |
|
case 0: |
|
verticesLeft = 3; |
|
break; |
|
case 1: |
|
ps.push(ps[ps.length - 2], ps[ps.length - 1]); |
|
verticesLeft = 1; |
|
break; |
|
case 2: |
|
ps.push(ps[ps.length - 3], ps[ps.length - 1]); |
|
verticesLeft = 1; |
|
break; |
|
} |
|
operators.push(f); |
|
} |
|
ps.push(coords.length); |
|
coords.push(coord); |
|
colors.push(color); |
|
verticesLeft--; |
|
|
|
reader.align(); |
|
} |
|
mesh.figures.push({ |
|
type: 'triangles', |
|
coords: new Int32Array(ps), |
|
colors: new Int32Array(ps), |
|
}); |
|
} |
|
|
|
function decodeType5Shading(mesh, reader, verticesPerRow) { |
|
var coords = mesh.coords; |
|
var colors = mesh.colors; |
|
var ps = []; // not maintaining cs since that will match ps |
|
while (reader.hasData) { |
|
var coord = reader.readCoordinate(); |
|
var color = reader.readComponents(); |
|
ps.push(coords.length); |
|
coords.push(coord); |
|
colors.push(color); |
|
} |
|
mesh.figures.push({ |
|
type: 'lattice', |
|
coords: new Int32Array(ps), |
|
colors: new Int32Array(ps), |
|
verticesPerRow: verticesPerRow |
|
}); |
|
} |
|
|
|
var MIN_SPLIT_PATCH_CHUNKS_AMOUNT = 3; |
|
var MAX_SPLIT_PATCH_CHUNKS_AMOUNT = 20; |
|
|
|
var TRIANGLE_DENSITY = 20; // count of triangles per entire mesh bounds |
|
|
|
var getB = (function getBClosure() { |
|
function buildB(count) { |
|
var lut = []; |
|
for (var i = 0; i <= count; i++) { |
|
var t = i / count, t_ = 1 - t; |
|
lut.push(new Float32Array([t_ * t_ * t_, 3 * t * t_ * t_, |
|
3 * t * t * t_, t * t * t])); |
|
} |
|
return lut; |
|
} |
|
var cache = []; |
|
return function getB(count) { |
|
if (!cache[count]) { |
|
cache[count] = buildB(count); |
|
} |
|
return cache[count]; |
|
}; |
|
})(); |
|
|
|
function buildFigureFromPatch(mesh, index) { |
|
var figure = mesh.figures[index]; |
|
assert(figure.type === 'patch', 'Unexpected patch mesh figure'); |
|
|
|
var coords = mesh.coords, colors = mesh.colors; |
|
var pi = figure.coords; |
|
var ci = figure.colors; |
|
|
|
var figureMinX = Math.min(coords[pi[0]][0], coords[pi[3]][0], |
|
coords[pi[12]][0], coords[pi[15]][0]); |
|
var figureMinY = Math.min(coords[pi[0]][1], coords[pi[3]][1], |
|
coords[pi[12]][1], coords[pi[15]][1]); |
|
var figureMaxX = Math.max(coords[pi[0]][0], coords[pi[3]][0], |
|
coords[pi[12]][0], coords[pi[15]][0]); |
|
var figureMaxY = Math.max(coords[pi[0]][1], coords[pi[3]][1], |
|
coords[pi[12]][1], coords[pi[15]][1]); |
|
var splitXBy = Math.ceil((figureMaxX - figureMinX) * TRIANGLE_DENSITY / |
|
(mesh.bounds[2] - mesh.bounds[0])); |
|
splitXBy = Math.max(MIN_SPLIT_PATCH_CHUNKS_AMOUNT, |
|
Math.min(MAX_SPLIT_PATCH_CHUNKS_AMOUNT, splitXBy)); |
|
var splitYBy = Math.ceil((figureMaxY - figureMinY) * TRIANGLE_DENSITY / |
|
(mesh.bounds[3] - mesh.bounds[1])); |
|
splitYBy = Math.max(MIN_SPLIT_PATCH_CHUNKS_AMOUNT, |
|
Math.min(MAX_SPLIT_PATCH_CHUNKS_AMOUNT, splitYBy)); |
|
|
|
var verticesPerRow = splitXBy + 1; |
|
var figureCoords = new Int32Array((splitYBy + 1) * verticesPerRow); |
|
var figureColors = new Int32Array((splitYBy + 1) * verticesPerRow); |
|
var k = 0; |
|
var cl = new Uint8Array(3), cr = new Uint8Array(3); |
|
var c0 = colors[ci[0]], c1 = colors[ci[1]], |
|
c2 = colors[ci[2]], c3 = colors[ci[3]]; |
|
var bRow = getB(splitYBy), bCol = getB(splitXBy); |
|
for (var row = 0; row <= splitYBy; row++) { |
|
cl[0] = ((c0[0] * (splitYBy - row) + c2[0] * row) / splitYBy) | 0; |
|
cl[1] = ((c0[1] * (splitYBy - row) + c2[1] * row) / splitYBy) | 0; |
|
cl[2] = ((c0[2] * (splitYBy - row) + c2[2] * row) / splitYBy) | 0; |
|
|
|
cr[0] = ((c1[0] * (splitYBy - row) + c3[0] * row) / splitYBy) | 0; |
|
cr[1] = ((c1[1] * (splitYBy - row) + c3[1] * row) / splitYBy) | 0; |
|
cr[2] = ((c1[2] * (splitYBy - row) + c3[2] * row) / splitYBy) | 0; |
|
|
|
for (var col = 0; col <= splitXBy; col++, k++) { |
|
if ((row === 0 || row === splitYBy) && |
|
(col === 0 || col === splitXBy)) { |
|
continue; |
|
} |
|
var x = 0, y = 0; |
|
var q = 0; |
|
for (var i = 0; i <= 3; i++) { |
|
for (var j = 0; j <= 3; j++, q++) { |
|
var m = bRow[row][i] * bCol[col][j]; |
|
x += coords[pi[q]][0] * m; |
|
y += coords[pi[q]][1] * m; |
|
} |
|
} |
|
figureCoords[k] = coords.length; |
|
coords.push([x, y]); |
|
figureColors[k] = colors.length; |
|
var newColor = new Uint8Array(3); |
|
newColor[0] = ((cl[0] * (splitXBy - col) + cr[0] * col) / splitXBy) | 0; |
|
newColor[1] = ((cl[1] * (splitXBy - col) + cr[1] * col) / splitXBy) | 0; |
|
newColor[2] = ((cl[2] * (splitXBy - col) + cr[2] * col) / splitXBy) | 0; |
|
colors.push(newColor); |
|
} |
|
} |
|
figureCoords[0] = pi[0]; |
|
figureColors[0] = ci[0]; |
|
figureCoords[splitXBy] = pi[3]; |
|
figureColors[splitXBy] = ci[1]; |
|
figureCoords[verticesPerRow * splitYBy] = pi[12]; |
|
figureColors[verticesPerRow * splitYBy] = ci[2]; |
|
figureCoords[verticesPerRow * splitYBy + splitXBy] = pi[15]; |
|
figureColors[verticesPerRow * splitYBy + splitXBy] = ci[3]; |
|
|
|
mesh.figures[index] = { |
|
type: 'lattice', |
|
coords: figureCoords, |
|
colors: figureColors, |
|
verticesPerRow: verticesPerRow |
|
}; |
|
} |
|
|
|
function decodeType6Shading(mesh, reader) { |
|
// A special case of Type 7. The p11, p12, p21, p22 automatically filled |
|
var coords = mesh.coords; |
|
var colors = mesh.colors; |
|
var ps = new Int32Array(16); // p00, p10, ..., p30, p01, ..., p33 |
|
var cs = new Int32Array(4); // c00, c30, c03, c33 |
|
while (reader.hasData) { |
|
var f = reader.readFlag(); |
|
assert(0 <= f && f <= 3, 'Unknown type6 flag'); |
|
var i, ii; |
|
var pi = coords.length; |
|
for (i = 0, ii = (f !== 0 ? 8 : 12); i < ii; i++) { |
|
coords.push(reader.readCoordinate()); |
|
} |
|
var ci = colors.length; |
|
for (i = 0, ii = (f !== 0 ? 2 : 4); i < ii; i++) { |
|
colors.push(reader.readComponents()); |
|
} |
|
var tmp1, tmp2, tmp3, tmp4; |
|
switch (f) { |
|
case 0: |
|
ps[12] = pi + 3; ps[13] = pi + 4; ps[14] = pi + 5; ps[15] = pi + 6; |
|
ps[ 8] = pi + 2; /* values for 5, 6, 9, 10 are */ ps[11] = pi + 7; |
|
ps[ 4] = pi + 1; /* calculated below */ ps[ 7] = pi + 8; |
|
ps[ 0] = pi; ps[ 1] = pi + 11; ps[ 2] = pi + 10; ps[ 3] = pi + 9; |
|
cs[2] = ci + 1; cs[3] = ci + 2; |
|
cs[0] = ci; cs[1] = ci + 3; |
|
break; |
|
case 1: |
|
tmp1 = ps[12]; tmp2 = ps[13]; tmp3 = ps[14]; tmp4 = ps[15]; |
|
ps[12] = tmp4; ps[13] = pi + 0; ps[14] = pi + 1; ps[15] = pi + 2; |
|
ps[ 8] = tmp3; /* values for 5, 6, 9, 10 are */ ps[11] = pi + 3; |
|
ps[ 4] = tmp2; /* calculated below */ ps[ 7] = pi + 4; |
|
ps[ 0] = tmp1; ps[ 1] = pi + 7; ps[ 2] = pi + 6; ps[ 3] = pi + 5; |
|
tmp1 = cs[2]; tmp2 = cs[3]; |
|
cs[2] = tmp2; cs[3] = ci; |
|
cs[0] = tmp1; cs[1] = ci + 1; |
|
break; |
|
case 2: |
|
tmp1 = ps[15]; |
|
tmp2 = ps[11]; |
|
ps[12] = ps[3]; ps[13] = pi + 0; ps[14] = pi + 1; ps[15] = pi + 2; |
|
ps[ 8] = ps[7]; /* values for 5, 6, 9, 10 are */ ps[11] = pi + 3; |
|
ps[ 4] = tmp2; /* calculated below */ ps[ 7] = pi + 4; |
|
ps[ 0] = tmp1; ps[ 1] = pi + 7; ps[ 2] = pi + 6; ps[ 3] = pi + 5; |
|
tmp1 = cs[3]; |
|
cs[2] = cs[1]; cs[3] = ci; |
|
cs[0] = tmp1; cs[1] = ci + 1; |
|
break; |
|
case 3: |
|
ps[12] = ps[0]; ps[13] = pi + 0; ps[14] = pi + 1; ps[15] = pi + 2; |
|
ps[ 8] = ps[1]; /* values for 5, 6, 9, 10 are */ ps[11] = pi + 3; |
|
ps[ 4] = ps[2]; /* calculated below */ ps[ 7] = pi + 4; |
|
ps[ 0] = ps[3]; ps[ 1] = pi + 7; ps[ 2] = pi + 6; ps[ 3] = pi + 5; |
|
cs[2] = cs[0]; cs[3] = ci; |
|
cs[0] = cs[1]; cs[1] = ci + 1; |
|
break; |
|
} |
|
// set p11, p12, p21, p22 |
|
ps[5] = coords.length; |
|
coords.push([ |
|
(-4 * coords[ps[0]][0] - coords[ps[15]][0] + |
|
6 * (coords[ps[4]][0] + coords[ps[1]][0]) - |
|
2 * (coords[ps[12]][0] + coords[ps[3]][0]) + |
|
3 * (coords[ps[13]][0] + coords[ps[7]][0])) / 9, |
|
(-4 * coords[ps[0]][1] - coords[ps[15]][1] + |
|
6 * (coords[ps[4]][1] + coords[ps[1]][1]) - |
|
2 * (coords[ps[12]][1] + coords[ps[3]][1]) + |
|
3 * (coords[ps[13]][1] + coords[ps[7]][1])) / 9 |
|
]); |
|
ps[6] = coords.length; |
|
coords.push([ |
|
(-4 * coords[ps[3]][0] - coords[ps[12]][0] + |
|
6 * (coords[ps[2]][0] + coords[ps[7]][0]) - |
|
2 * (coords[ps[0]][0] + coords[ps[15]][0]) + |
|
3 * (coords[ps[4]][0] + coords[ps[14]][0])) / 9, |
|
(-4 * coords[ps[3]][1] - coords[ps[12]][1] + |
|
6 * (coords[ps[2]][1] + coords[ps[7]][1]) - |
|
2 * (coords[ps[0]][1] + coords[ps[15]][1]) + |
|
3 * (coords[ps[4]][1] + coords[ps[14]][1])) / 9 |
|
]); |
|
ps[9] = coords.length; |
|
coords.push([ |
|
(-4 * coords[ps[12]][0] - coords[ps[3]][0] + |
|
6 * (coords[ps[8]][0] + coords[ps[13]][0]) - |
|
2 * (coords[ps[0]][0] + coords[ps[15]][0]) + |
|
3 * (coords[ps[11]][0] + coords[ps[1]][0])) / 9, |
|
(-4 * coords[ps[12]][1] - coords[ps[3]][1] + |
|
6 * (coords[ps[8]][1] + coords[ps[13]][1]) - |
|
2 * (coords[ps[0]][1] + coords[ps[15]][1]) + |
|
3 * (coords[ps[11]][1] + coords[ps[1]][1])) / 9 |
|
]); |
|
ps[10] = coords.length; |
|
coords.push([ |
|
(-4 * coords[ps[15]][0] - coords[ps[0]][0] + |
|
6 * (coords[ps[11]][0] + coords[ps[14]][0]) - |
|
2 * (coords[ps[12]][0] + coords[ps[3]][0]) + |
|
3 * (coords[ps[2]][0] + coords[ps[8]][0])) / 9, |
|
(-4 * coords[ps[15]][1] - coords[ps[0]][1] + |
|
6 * (coords[ps[11]][1] + coords[ps[14]][1]) - |
|
2 * (coords[ps[12]][1] + coords[ps[3]][1]) + |
|
3 * (coords[ps[2]][1] + coords[ps[8]][1])) / 9 |
|
]); |
|
mesh.figures.push({ |
|
type: 'patch', |
|
coords: new Int32Array(ps), // making copies of ps and cs |
|
colors: new Int32Array(cs) |
|
}); |
|
} |
|
} |
|
|
|
function decodeType7Shading(mesh, reader) { |
|
var coords = mesh.coords; |
|
var colors = mesh.colors; |
|
var ps = new Int32Array(16); // p00, p10, ..., p30, p01, ..., p33 |
|
var cs = new Int32Array(4); // c00, c30, c03, c33 |
|
while (reader.hasData) { |
|
var f = reader.readFlag(); |
|
assert(0 <= f && f <= 3, 'Unknown type7 flag'); |
|
var i, ii; |
|
var pi = coords.length; |
|
for (i = 0, ii = (f !== 0 ? 12 : 16); i < ii; i++) { |
|
coords.push(reader.readCoordinate()); |
|
} |
|
var ci = colors.length; |
|
for (i = 0, ii = (f !== 0 ? 2 : 4); i < ii; i++) { |
|
colors.push(reader.readComponents()); |
|
} |
|
var tmp1, tmp2, tmp3, tmp4; |
|
switch (f) { |
|
case 0: |
|
ps[12] = pi + 3; ps[13] = pi + 4; ps[14] = pi + 5; ps[15] = pi + 6; |
|
ps[ 8] = pi + 2; ps[ 9] = pi + 13; ps[10] = pi + 14; ps[11] = pi + 7; |
|
ps[ 4] = pi + 1; ps[ 5] = pi + 12; ps[ 6] = pi + 15; ps[ 7] = pi + 8; |
|
ps[ 0] = pi; ps[ 1] = pi + 11; ps[ 2] = pi + 10; ps[ 3] = pi + 9; |
|
cs[2] = ci + 1; cs[3] = ci + 2; |
|
cs[0] = ci; cs[1] = ci + 3; |
|
break; |
|
case 1: |
|
tmp1 = ps[12]; tmp2 = ps[13]; tmp3 = ps[14]; tmp4 = ps[15]; |
|
ps[12] = tmp4; ps[13] = pi + 0; ps[14] = pi + 1; ps[15] = pi + 2; |
|
ps[ 8] = tmp3; ps[ 9] = pi + 9; ps[10] = pi + 10; ps[11] = pi + 3; |
|
ps[ 4] = tmp2; ps[ 5] = pi + 8; ps[ 6] = pi + 11; ps[ 7] = pi + 4; |
|
ps[ 0] = tmp1; ps[ 1] = pi + 7; ps[ 2] = pi + 6; ps[ 3] = pi + 5; |
|
tmp1 = cs[2]; tmp2 = cs[3]; |
|
cs[2] = tmp2; cs[3] = ci; |
|
cs[0] = tmp1; cs[1] = ci + 1; |
|
break; |
|
case 2: |
|
tmp1 = ps[15]; |
|
tmp2 = ps[11]; |
|
ps[12] = ps[3]; ps[13] = pi + 0; ps[14] = pi + 1; ps[15] = pi + 2; |
|
ps[ 8] = ps[7]; ps[ 9] = pi + 9; ps[10] = pi + 10; ps[11] = pi + 3; |
|
ps[ 4] = tmp2; ps[ 5] = pi + 8; ps[ 6] = pi + 11; ps[ 7] = pi + 4; |
|
ps[ 0] = tmp1; ps[ 1] = pi + 7; ps[ 2] = pi + 6; ps[ 3] = pi + 5; |
|
tmp1 = cs[3]; |
|
cs[2] = cs[1]; cs[3] = ci; |
|
cs[0] = tmp1; cs[1] = ci + 1; |
|
break; |
|
case 3: |
|
ps[12] = ps[0]; ps[13] = pi + 0; ps[14] = pi + 1; ps[15] = pi + 2; |
|
ps[ 8] = ps[1]; ps[ 9] = pi + 9; ps[10] = pi + 10; ps[11] = pi + 3; |
|
ps[ 4] = ps[2]; ps[ 5] = pi + 8; ps[ 6] = pi + 11; ps[ 7] = pi + 4; |
|
ps[ 0] = ps[3]; ps[ 1] = pi + 7; ps[ 2] = pi + 6; ps[ 3] = pi + 5; |
|
cs[2] = cs[0]; cs[3] = ci; |
|
cs[0] = cs[1]; cs[1] = ci + 1; |
|
break; |
|
} |
|
mesh.figures.push({ |
|
type: 'patch', |
|
coords: new Int32Array(ps), // making copies of ps and cs |
|
colors: new Int32Array(cs) |
|
}); |
|
} |
|
} |
|
|
|
function updateBounds(mesh) { |
|
var minX = mesh.coords[0][0], minY = mesh.coords[0][1], |
|
maxX = minX, maxY = minY; |
|
for (var i = 1, ii = mesh.coords.length; i < ii; i++) { |
|
var x = mesh.coords[i][0], y = mesh.coords[i][1]; |
|
minX = minX > x ? x : minX; |
|
minY = minY > y ? y : minY; |
|
maxX = maxX < x ? x : maxX; |
|
maxY = maxY < y ? y : maxY; |
|
} |
|
mesh.bounds = [minX, minY, maxX, maxY]; |
|
} |
|
|
|
function packData(mesh) { |
|
var i, ii, j, jj; |
|
|
|
var coords = mesh.coords; |
|
var coordsPacked = new Float32Array(coords.length * 2); |
|
for (i = 0, j = 0, ii = coords.length; i < ii; i++) { |
|
var xy = coords[i]; |
|
coordsPacked[j++] = xy[0]; |
|
coordsPacked[j++] = xy[1]; |
|
} |
|
mesh.coords = coordsPacked; |
|
|
|
var colors = mesh.colors; |
|
var colorsPacked = new Uint8Array(colors.length * 3); |
|
for (i = 0, j = 0, ii = colors.length; i < ii; i++) { |
|
var c = colors[i]; |
|
colorsPacked[j++] = c[0]; |
|
colorsPacked[j++] = c[1]; |
|
colorsPacked[j++] = c[2]; |
|
} |
|
mesh.colors = colorsPacked; |
|
|
|
var figures = mesh.figures; |
|
for (i = 0, ii = figures.length; i < ii; i++) { |
|
var figure = figures[i], ps = figure.coords, cs = figure.colors; |
|
for (j = 0, jj = ps.length; j < jj; j++) { |
|
ps[j] *= 2; |
|
cs[j] *= 3; |
|
} |
|
} |
|
} |
|
|
|
function Mesh(stream, matrix, xref, res) { |
|
assert(isStream(stream), 'Mesh data is not a stream'); |
|
var dict = stream.dict; |
|
this.matrix = matrix; |
|
this.shadingType = dict.get('ShadingType'); |
|
this.type = 'Pattern'; |
|
this.bbox = dict.get('BBox'); |
|
var cs = dict.get('ColorSpace', 'CS'); |
|
cs = ColorSpace.parse(cs, xref, res); |
|
this.cs = cs; |
|
this.background = dict.has('Background') ? |
|
cs.getRgb(dict.get('Background'), 0) : null; |
|
|
|
var fnObj = dict.get('Function'); |
|
var fn = fnObj ? PDFFunction.parseArray(xref, fnObj) : null; |
|
|
|
this.coords = []; |
|
this.colors = []; |
|
this.figures = []; |
|
|
|
var decodeContext = { |
|
bitsPerCoordinate: dict.get('BitsPerCoordinate'), |
|
bitsPerComponent: dict.get('BitsPerComponent'), |
|
bitsPerFlag: dict.get('BitsPerFlag'), |
|
decode: dict.get('Decode'), |
|
colorFn: fn, |
|
colorSpace: cs, |
|
numComps: fn ? 1 : cs.numComps |
|
}; |
|
var reader = new MeshStreamReader(stream, decodeContext); |
|
|
|
var patchMesh = false; |
|
switch (this.shadingType) { |
|
case ShadingType.FREE_FORM_MESH: |
|
decodeType4Shading(this, reader); |
|
break; |
|
case ShadingType.LATTICE_FORM_MESH: |
|
var verticesPerRow = dict.get('VerticesPerRow') | 0; |
|
assert(verticesPerRow >= 2, 'Invalid VerticesPerRow'); |
|
decodeType5Shading(this, reader, verticesPerRow); |
|
break; |
|
case ShadingType.COONS_PATCH_MESH: |
|
decodeType6Shading(this, reader); |
|
patchMesh = true; |
|
break; |
|
case ShadingType.TENSOR_PATCH_MESH: |
|
decodeType7Shading(this, reader); |
|
patchMesh = true; |
|
break; |
|
default: |
|
error('Unsupported mesh type.'); |
|
break; |
|
} |
|
|
|
if (patchMesh) { |
|
// dirty bounds calculation for determining, how dense shall be triangles |
|
updateBounds(this); |
|
for (var i = 0, ii = this.figures.length; i < ii; i++) { |
|
buildFigureFromPatch(this, i); |
|
} |
|
} |
|
// calculate bounds |
|
updateBounds(this); |
|
|
|
packData(this); |
|
} |
|
|
|
Mesh.prototype = { |
|
getIR: function Mesh_getIR() { |
|
return ['Mesh', this.shadingType, this.coords, this.colors, this.figures, |
|
this.bounds, this.matrix, this.bbox, this.background]; |
|
} |
|
}; |
|
|
|
return Mesh; |
|
})(); |
|
|
|
Shadings.Dummy = (function DummyClosure() { |
|
function Dummy() { |
|
this.type = 'Pattern'; |
|
} |
|
|
|
Dummy.prototype = { |
|
getIR: function Dummy_getIR() { |
|
return ['Dummy']; |
|
} |
|
}; |
|
return Dummy; |
|
})(); |
|
|
|
function getTilingPatternIR(operatorList, dict, args) { |
|
var matrix = dict.get('Matrix'); |
|
var bbox = dict.get('BBox'); |
|
var xstep = dict.get('XStep'); |
|
var ystep = dict.get('YStep'); |
|
var paintType = dict.get('PaintType'); |
|
var tilingType = dict.get('TilingType'); |
|
|
|
return [ |
|
'TilingPattern', args, operatorList, matrix, bbox, xstep, ystep, |
|
paintType, tilingType |
|
]; |
|
} |
|
|
|
exports.Pattern = Pattern; |
|
exports.getTilingPatternIR = getTilingPatternIR; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreEvaluator = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreStream, root.pdfjsCoreParser, |
|
root.pdfjsCoreImage, root.pdfjsCoreColorSpace, root.pdfjsCoreMurmurHash3, |
|
root.pdfjsCoreFonts, root.pdfjsCoreFunction, root.pdfjsCorePattern, |
|
root.pdfjsCoreCMap, root.pdfjsCoreMetrics, root.pdfjsCoreBidi); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreStream, coreParser, |
|
coreImage, coreColorSpace, coreMurmurHash3, coreFonts, |
|
coreFunction, corePattern, coreCMap, coreMetrics, coreBidi) { |
|
|
|
var FONT_IDENTITY_MATRIX = sharedUtil.FONT_IDENTITY_MATRIX; |
|
var IDENTITY_MATRIX = sharedUtil.IDENTITY_MATRIX; |
|
var UNSUPPORTED_FEATURES = sharedUtil.UNSUPPORTED_FEATURES; |
|
var ImageKind = sharedUtil.ImageKind; |
|
var OPS = sharedUtil.OPS; |
|
var TextRenderingMode = sharedUtil.TextRenderingMode; |
|
var Util = sharedUtil.Util; |
|
var assert = sharedUtil.assert; |
|
var createPromiseCapability = sharedUtil.createPromiseCapability; |
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isArray = sharedUtil.isArray; |
|
var isNum = sharedUtil.isNum; |
|
var isString = sharedUtil.isString; |
|
var warn = sharedUtil.warn; |
|
var Dict = corePrimitives.Dict; |
|
var Name = corePrimitives.Name; |
|
var isCmd = corePrimitives.isCmd; |
|
var isDict = corePrimitives.isDict; |
|
var isName = corePrimitives.isName; |
|
var isRef = corePrimitives.isRef; |
|
var isStream = corePrimitives.isStream; |
|
var DecodeStream = coreStream.DecodeStream; |
|
var JpegStream = coreStream.JpegStream; |
|
var Lexer = coreParser.Lexer; |
|
var Parser = coreParser.Parser; |
|
var isEOF = coreParser.isEOF; |
|
var PDFImage = coreImage.PDFImage; |
|
var ColorSpace = coreColorSpace.ColorSpace; |
|
var MurmurHash3_64 = coreMurmurHash3.MurmurHash3_64; |
|
var Encodings = coreFonts.Encodings; |
|
var ErrorFont = coreFonts.ErrorFont; |
|
var FontFlags = coreFonts.FontFlags; |
|
var Font = coreFonts.Font; |
|
var IdentityToUnicodeMap = coreFonts.IdentityToUnicodeMap; |
|
var NormalizedUnicodes = coreFonts.NormalizedUnicodes; |
|
var ToUnicodeMap = coreFonts.ToUnicodeMap; |
|
var getFontType = coreFonts.getFontType; |
|
var reverseIfRtl = coreFonts.reverseIfRtl; |
|
var serifFonts = coreFonts.serifFonts; |
|
var symbolsFonts = coreFonts.symbolsFonts; |
|
var stdFontMap = coreFonts.stdFontMap; |
|
var isPDFFunction = coreFunction.isPDFFunction; |
|
var PDFFunction = coreFunction.PDFFunction; |
|
var Pattern = corePattern.Pattern; |
|
var getTilingPatternIR = corePattern.getTilingPatternIR; |
|
var CMapFactory = coreCMap.CMapFactory; |
|
var IdentityCMap = coreCMap.IdentityCMap; |
|
var Metrics = coreMetrics.Metrics; |
|
var bidi = coreBidi.bidi; |
|
|
|
var PartialEvaluator = (function PartialEvaluatorClosure() { |
|
function PartialEvaluator(pdfManager, xref, handler, pageIndex, |
|
uniquePrefix, idCounters, fontCache) { |
|
this.pdfManager = pdfManager; |
|
this.xref = xref; |
|
this.handler = handler; |
|
this.pageIndex = pageIndex; |
|
this.uniquePrefix = uniquePrefix; |
|
this.idCounters = idCounters; |
|
this.fontCache = fontCache; |
|
} |
|
|
|
// Trying to minimize Date.now() usage and check every 100 time |
|
var TIME_SLOT_DURATION_MS = 20; |
|
var CHECK_TIME_EVERY = 100; |
|
function TimeSlotManager() { |
|
this.reset(); |
|
} |
|
TimeSlotManager.prototype = { |
|
check: function TimeSlotManager_check() { |
|
if (++this.checked < CHECK_TIME_EVERY) { |
|
return false; |
|
} |
|
this.checked = 0; |
|
return this.endTime <= Date.now(); |
|
}, |
|
reset: function TimeSlotManager_reset() { |
|
this.endTime = Date.now() + TIME_SLOT_DURATION_MS; |
|
this.checked = 0; |
|
} |
|
}; |
|
|
|
var deferred = Promise.resolve(); |
|
|
|
var TILING_PATTERN = 1, SHADING_PATTERN = 2; |
|
|
|
PartialEvaluator.prototype = { |
|
hasBlendModes: function PartialEvaluator_hasBlendModes(resources) { |
|
if (!isDict(resources)) { |
|
return false; |
|
} |
|
|
|
var processed = Object.create(null); |
|
if (resources.objId) { |
|
processed[resources.objId] = true; |
|
} |
|
|
|
var nodes = [resources]; |
|
while (nodes.length) { |
|
var key; |
|
var node = nodes.shift(); |
|
// First check the current resources for blend modes. |
|
var graphicStates = node.get('ExtGState'); |
|
if (isDict(graphicStates)) { |
|
graphicStates = graphicStates.getAll(); |
|
for (key in graphicStates) { |
|
var graphicState = graphicStates[key]; |
|
var bm = graphicState['BM']; |
|
if (isName(bm) && bm.name !== 'Normal') { |
|
return true; |
|
} |
|
} |
|
} |
|
// Descend into the XObjects to look for more resources and blend modes. |
|
var xObjects = node.get('XObject'); |
|
if (!isDict(xObjects)) { |
|
continue; |
|
} |
|
xObjects = xObjects.getAll(); |
|
for (key in xObjects) { |
|
var xObject = xObjects[key]; |
|
if (!isStream(xObject)) { |
|
continue; |
|
} |
|
if (xObject.dict.objId) { |
|
if (processed[xObject.dict.objId]) { |
|
// stream has objId and is processed already |
|
continue; |
|
} |
|
processed[xObject.dict.objId] = true; |
|
} |
|
var xResources = xObject.dict.get('Resources'); |
|
// Checking objId to detect an infinite loop. |
|
if (isDict(xResources) && |
|
(!xResources.objId || !processed[xResources.objId])) { |
|
nodes.push(xResources); |
|
if (xResources.objId) { |
|
processed[xResources.objId] = true; |
|
} |
|
} |
|
} |
|
} |
|
return false; |
|
}, |
|
|
|
buildFormXObject: function PartialEvaluator_buildFormXObject(resources, |
|
xobj, smask, |
|
operatorList, |
|
task, |
|
initialState) { |
|
var matrix = xobj.dict.getArray('Matrix'); |
|
var bbox = xobj.dict.getArray('BBox'); |
|
var group = xobj.dict.get('Group'); |
|
if (group) { |
|
var groupOptions = { |
|
matrix: matrix, |
|
bbox: bbox, |
|
smask: smask, |
|
isolated: false, |
|
knockout: false |
|
}; |
|
|
|
var groupSubtype = group.get('S'); |
|
var colorSpace; |
|
if (isName(groupSubtype) && groupSubtype.name === 'Transparency') { |
|
groupOptions.isolated = (group.get('I') || false); |
|
groupOptions.knockout = (group.get('K') || false); |
|
colorSpace = (group.has('CS') ? |
|
ColorSpace.parse(group.get('CS'), this.xref, resources) : null); |
|
} |
|
|
|
if (smask && smask.backdrop) { |
|
colorSpace = colorSpace || ColorSpace.singletons.rgb; |
|
smask.backdrop = colorSpace.getRgb(smask.backdrop, 0); |
|
} |
|
|
|
operatorList.addOp(OPS.beginGroup, [groupOptions]); |
|
} |
|
|
|
operatorList.addOp(OPS.paintFormXObjectBegin, [matrix, bbox]); |
|
|
|
return this.getOperatorList(xobj, task, |
|
(xobj.dict.get('Resources') || resources), operatorList, initialState). |
|
then(function () { |
|
operatorList.addOp(OPS.paintFormXObjectEnd, []); |
|
|
|
if (group) { |
|
operatorList.addOp(OPS.endGroup, [groupOptions]); |
|
} |
|
}); |
|
}, |
|
|
|
buildPaintImageXObject: |
|
function PartialEvaluator_buildPaintImageXObject(resources, image, |
|
inline, operatorList, |
|
cacheKey, imageCache) { |
|
var self = this; |
|
var dict = image.dict; |
|
var w = dict.get('Width', 'W'); |
|
var h = dict.get('Height', 'H'); |
|
|
|
if (!(w && isNum(w)) || !(h && isNum(h))) { |
|
warn('Image dimensions are missing, or not numbers.'); |
|
return; |
|
} |
|
if (PDFJS.maxImageSize !== -1 && w * h > PDFJS.maxImageSize) { |
|
warn('Image exceeded maximum allowed size and was removed.'); |
|
return; |
|
} |
|
|
|
var imageMask = (dict.get('ImageMask', 'IM') || false); |
|
var imgData, args; |
|
if (imageMask) { |
|
// This depends on a tmpCanvas being filled with the |
|
// current fillStyle, such that processing the pixel |
|
// data can't be done here. Instead of creating a |
|
// complete PDFImage, only read the information needed |
|
// for later. |
|
|
|
var width = dict.get('Width', 'W'); |
|
var height = dict.get('Height', 'H'); |
|
var bitStrideLength = (width + 7) >> 3; |
|
var imgArray = image.getBytes(bitStrideLength * height); |
|
var decode = dict.get('Decode', 'D'); |
|
var inverseDecode = (!!decode && decode[0] > 0); |
|
|
|
imgData = PDFImage.createMask(imgArray, width, height, |
|
image instanceof DecodeStream, |
|
inverseDecode); |
|
imgData.cached = true; |
|
args = [imgData]; |
|
operatorList.addOp(OPS.paintImageMaskXObject, args); |
|
if (cacheKey) { |
|
imageCache[cacheKey] = { |
|
fn: OPS.paintImageMaskXObject, |
|
args: args |
|
}; |
|
} |
|
return; |
|
} |
|
|
|
var softMask = (dict.get('SMask', 'SM') || false); |
|
var mask = (dict.get('Mask') || false); |
|
|
|
var SMALL_IMAGE_DIMENSIONS = 200; |
|
// Inlining small images into the queue as RGB data |
|
if (inline && !softMask && !mask && !(image instanceof JpegStream) && |
|
(w + h) < SMALL_IMAGE_DIMENSIONS) { |
|
var imageObj = new PDFImage(this.xref, resources, image, |
|
inline, null, null); |
|
// We force the use of RGBA_32BPP images here, because we can't handle |
|
// any other kind. |
|
imgData = imageObj.createImageData(/* forceRGBA = */ true); |
|
operatorList.addOp(OPS.paintInlineImageXObject, [imgData]); |
|
return; |
|
} |
|
|
|
// If there is no imageMask, create the PDFImage and a lot |
|
// of image processing can be done here. |
|
var uniquePrefix = (this.uniquePrefix || ''); |
|
var objId = 'img_' + uniquePrefix + (++this.idCounters.obj); |
|
operatorList.addDependency(objId); |
|
args = [objId, w, h]; |
|
|
|
if (!softMask && !mask && image instanceof JpegStream && |
|
image.isNativelySupported(this.xref, resources)) { |
|
// These JPEGs don't need any more processing so we can just send it. |
|
operatorList.addOp(OPS.paintJpegXObject, args); |
|
this.handler.send('obj', |
|
[objId, this.pageIndex, 'JpegStream', image.getIR()]); |
|
return; |
|
} |
|
|
|
PDFImage.buildImage(self.handler, self.xref, resources, image, inline). |
|
then(function(imageObj) { |
|
var imgData = imageObj.createImageData(/* forceRGBA = */ false); |
|
self.handler.send('obj', [objId, self.pageIndex, 'Image', imgData], |
|
[imgData.data.buffer]); |
|
}).then(undefined, function (reason) { |
|
warn('Unable to decode image: ' + reason); |
|
self.handler.send('obj', [objId, self.pageIndex, 'Image', null]); |
|
}); |
|
|
|
operatorList.addOp(OPS.paintImageXObject, args); |
|
if (cacheKey) { |
|
imageCache[cacheKey] = { |
|
fn: OPS.paintImageXObject, |
|
args: args |
|
}; |
|
} |
|
}, |
|
|
|
handleSMask: function PartialEvaluator_handleSmask(smask, resources, |
|
operatorList, task, |
|
stateManager) { |
|
var smaskContent = smask.get('G'); |
|
var smaskOptions = { |
|
subtype: smask.get('S').name, |
|
backdrop: smask.get('BC') |
|
}; |
|
|
|
// The SMask might have a alpha/luminosity value transfer function -- |
|
// we will build a map of integer values in range 0..255 to be fast. |
|
var transferObj = smask.get('TR'); |
|
if (isPDFFunction(transferObj)) { |
|
var transferFn = PDFFunction.parse(this.xref, transferObj); |
|
var transferMap = new Uint8Array(256); |
|
var tmp = new Float32Array(1); |
|
for (var i = 0; i < 255; i++) { |
|
tmp[0] = i / 255; |
|
transferFn(tmp, 0, tmp, 0); |
|
transferMap[i] = (tmp[0] * 255) | 0; |
|
} |
|
smaskOptions.transferMap = transferMap; |
|
} |
|
|
|
return this.buildFormXObject(resources, smaskContent, smaskOptions, |
|
operatorList, task, stateManager.state.clone()); |
|
}, |
|
|
|
handleTilingType: |
|
function PartialEvaluator_handleTilingType(fn, args, resources, |
|
pattern, patternDict, |
|
operatorList, task) { |
|
// Create an IR of the pattern code. |
|
var tilingOpList = new OperatorList(); |
|
// Merge the available resources, to prevent issues when the patternDict |
|
// is missing some /Resources entries (fixes issue6541.pdf). |
|
var resourcesArray = [patternDict.get('Resources'), resources]; |
|
var patternResources = Dict.merge(this.xref, resourcesArray); |
|
|
|
return this.getOperatorList(pattern, task, patternResources, |
|
tilingOpList).then(function () { |
|
// Add the dependencies to the parent operator list so they are |
|
// resolved before sub operator list is executed synchronously. |
|
operatorList.addDependencies(tilingOpList.dependencies); |
|
operatorList.addOp(fn, getTilingPatternIR({ |
|
fnArray: tilingOpList.fnArray, |
|
argsArray: tilingOpList.argsArray |
|
}, patternDict, args)); |
|
}); |
|
}, |
|
|
|
handleSetFont: |
|
function PartialEvaluator_handleSetFont(resources, fontArgs, fontRef, |
|
operatorList, task, state) { |
|
// TODO(mack): Not needed? |
|
var fontName; |
|
if (fontArgs) { |
|
fontArgs = fontArgs.slice(); |
|
fontName = fontArgs[0].name; |
|
} |
|
|
|
var self = this; |
|
return this.loadFont(fontName, fontRef, this.xref, resources).then( |
|
function (translated) { |
|
if (!translated.font.isType3Font) { |
|
return translated; |
|
} |
|
return translated.loadType3Data(self, resources, operatorList, task). |
|
then(function () { |
|
return translated; |
|
}, function (reason) { |
|
// Error in the font data -- sending unsupported feature notification. |
|
self.handler.send('UnsupportedFeature', |
|
{featureId: UNSUPPORTED_FEATURES.font}); |
|
return new TranslatedFont('g_font_error', |
|
new ErrorFont('Type3 font load error: ' + reason), translated.font); |
|
}); |
|
}).then(function (translated) { |
|
state.font = translated.font; |
|
translated.send(self.handler); |
|
return translated.loadedName; |
|
}); |
|
}, |
|
|
|
handleText: function PartialEvaluator_handleText(chars, state) { |
|
var font = state.font; |
|
var glyphs = font.charsToGlyphs(chars); |
|
var isAddToPathSet = !!(state.textRenderingMode & |
|
TextRenderingMode.ADD_TO_PATH_FLAG); |
|
if (font.data && (isAddToPathSet || PDFJS.disableFontFace)) { |
|
var buildPath = function (fontChar) { |
|
if (!font.renderer.hasBuiltPath(fontChar)) { |
|
var path = font.renderer.getPathJs(fontChar); |
|
this.handler.send('commonobj', [ |
|
font.loadedName + '_path_' + fontChar, |
|
'FontPath', |
|
path |
|
]); |
|
} |
|
}.bind(this); |
|
|
|
for (var i = 0, ii = glyphs.length; i < ii; i++) { |
|
var glyph = glyphs[i]; |
|
buildPath(glyph.fontChar); |
|
|
|
// If the glyph has an accent we need to build a path for its |
|
// fontChar too, otherwise CanvasGraphics_paintChar will fail. |
|
var accent = glyph.accent; |
|
if (accent && accent.fontChar) { |
|
buildPath(accent.fontChar); |
|
} |
|
} |
|
} |
|
|
|
return glyphs; |
|
}, |
|
|
|
setGState: function PartialEvaluator_setGState(resources, gState, |
|
operatorList, task, |
|
xref, stateManager) { |
|
// This array holds the converted/processed state data. |
|
var gStateObj = []; |
|
var gStateMap = gState.map; |
|
var self = this; |
|
var promise = Promise.resolve(); |
|
for (var key in gStateMap) { |
|
var value = gStateMap[key]; |
|
switch (key) { |
|
case 'Type': |
|
break; |
|
case 'LW': |
|
case 'LC': |
|
case 'LJ': |
|
case 'ML': |
|
case 'D': |
|
case 'RI': |
|
case 'FL': |
|
case 'CA': |
|
case 'ca': |
|
gStateObj.push([key, value]); |
|
break; |
|
case 'Font': |
|
promise = promise.then(function () { |
|
return self.handleSetFont(resources, null, value[0], operatorList, |
|
task, stateManager.state). |
|
then(function (loadedName) { |
|
operatorList.addDependency(loadedName); |
|
gStateObj.push([key, [loadedName, value[1]]]); |
|
}); |
|
}); |
|
break; |
|
case 'BM': |
|
gStateObj.push([key, value]); |
|
break; |
|
case 'SMask': |
|
if (isName(value) && value.name === 'None') { |
|
gStateObj.push([key, false]); |
|
break; |
|
} |
|
var dict = xref.fetchIfRef(value); |
|
if (isDict(dict)) { |
|
promise = promise.then(function () { |
|
return self.handleSMask(dict, resources, operatorList, |
|
task, stateManager); |
|
}); |
|
gStateObj.push([key, true]); |
|
} else { |
|
warn('Unsupported SMask type'); |
|
} |
|
|
|
break; |
|
// Only generate info log messages for the following since |
|
// they are unlikely to have a big impact on the rendering. |
|
case 'OP': |
|
case 'op': |
|
case 'OPM': |
|
case 'BG': |
|
case 'BG2': |
|
case 'UCR': |
|
case 'UCR2': |
|
case 'TR': |
|
case 'TR2': |
|
case 'HT': |
|
case 'SM': |
|
case 'SA': |
|
case 'AIS': |
|
case 'TK': |
|
// TODO implement these operators. |
|
info('graphic state operator ' + key); |
|
break; |
|
default: |
|
info('Unknown graphic state operator ' + key); |
|
break; |
|
} |
|
} |
|
return promise.then(function () { |
|
if (gStateObj.length >= 0) { |
|
operatorList.addOp(OPS.setGState, [gStateObj]); |
|
} |
|
}); |
|
}, |
|
|
|
loadFont: function PartialEvaluator_loadFont(fontName, font, xref, |
|
resources) { |
|
|
|
function errorFont() { |
|
return Promise.resolve(new TranslatedFont('g_font_error', |
|
new ErrorFont('Font ' + fontName + ' is not available'), font)); |
|
} |
|
var fontRef; |
|
if (font) { // Loading by ref. |
|
assert(isRef(font)); |
|
fontRef = font; |
|
} else { // Loading by name. |
|
var fontRes = resources.get('Font'); |
|
if (fontRes) { |
|
fontRef = fontRes.getRaw(fontName); |
|
} else { |
|
warn('fontRes not available'); |
|
return errorFont(); |
|
} |
|
} |
|
if (!fontRef) { |
|
warn('fontRef not available'); |
|
return errorFont(); |
|
} |
|
|
|
if (this.fontCache.has(fontRef)) { |
|
return this.fontCache.get(fontRef); |
|
} |
|
|
|
font = xref.fetchIfRef(fontRef); |
|
if (!isDict(font)) { |
|
return errorFont(); |
|
} |
|
|
|
// We are holding font.translated references just for fontRef that are not |
|
// dictionaries (Dict). See explanation below. |
|
if (font.translated) { |
|
return font.translated; |
|
} |
|
|
|
var fontCapability = createPromiseCapability(); |
|
|
|
var preEvaluatedFont = this.preEvaluateFont(font, xref); |
|
var descriptor = preEvaluatedFont.descriptor; |
|
var fontID = fontRef.num + '_' + fontRef.gen; |
|
if (isDict(descriptor)) { |
|
if (!descriptor.fontAliases) { |
|
descriptor.fontAliases = Object.create(null); |
|
} |
|
|
|
var fontAliases = descriptor.fontAliases; |
|
var hash = preEvaluatedFont.hash; |
|
if (fontAliases[hash]) { |
|
var aliasFontRef = fontAliases[hash].aliasRef; |
|
if (aliasFontRef && this.fontCache.has(aliasFontRef)) { |
|
this.fontCache.putAlias(fontRef, aliasFontRef); |
|
return this.fontCache.get(fontRef); |
|
} |
|
} |
|
|
|
if (!fontAliases[hash]) { |
|
fontAliases[hash] = { |
|
fontID: Font.getFontID() |
|
}; |
|
} |
|
|
|
fontAliases[hash].aliasRef = fontRef; |
|
fontID = fontAliases[hash].fontID; |
|
} |
|
|
|
// Workaround for bad PDF generators that don't reference fonts |
|
// properly, i.e. by not using an object identifier. |
|
// Check if the fontRef is a Dict (as opposed to a standard object), |
|
// in which case we don't cache the font and instead reference it by |
|
// fontName in font.loadedName below. |
|
var fontRefIsDict = isDict(fontRef); |
|
if (!fontRefIsDict) { |
|
this.fontCache.put(fontRef, fontCapability.promise); |
|
} |
|
|
|
// Keep track of each font we translated so the caller can |
|
// load them asynchronously before calling display on a page. |
|
font.loadedName = 'g_' + this.pdfManager.docId + '_f' + (fontRefIsDict ? |
|
fontName.replace(/\W/g, '') : fontID); |
|
|
|
font.translated = fontCapability.promise; |
|
|
|
// TODO move promises into translate font |
|
var translatedPromise; |
|
try { |
|
translatedPromise = Promise.resolve( |
|
this.translateFont(preEvaluatedFont, xref)); |
|
} catch (e) { |
|
translatedPromise = Promise.reject(e); |
|
} |
|
|
|
var self = this; |
|
translatedPromise.then(function (translatedFont) { |
|
if (translatedFont.fontType !== undefined) { |
|
var xrefFontStats = xref.stats.fontTypes; |
|
xrefFontStats[translatedFont.fontType] = true; |
|
} |
|
|
|
fontCapability.resolve(new TranslatedFont(font.loadedName, |
|
translatedFont, font)); |
|
}, function (reason) { |
|
// TODO fontCapability.reject? |
|
// Error in the font data -- sending unsupported feature notification. |
|
self.handler.send('UnsupportedFeature', |
|
{featureId: UNSUPPORTED_FEATURES.font}); |
|
|
|
try { |
|
// error, but it's still nice to have font type reported |
|
var descriptor = preEvaluatedFont.descriptor; |
|
var fontFile3 = descriptor && descriptor.get('FontFile3'); |
|
var subtype = fontFile3 && fontFile3.get('Subtype'); |
|
var fontType = getFontType(preEvaluatedFont.type, |
|
subtype && subtype.name); |
|
var xrefFontStats = xref.stats.fontTypes; |
|
xrefFontStats[fontType] = true; |
|
} catch (ex) { } |
|
|
|
fontCapability.resolve(new TranslatedFont(font.loadedName, |
|
new ErrorFont(reason instanceof Error ? reason.message : reason), |
|
font)); |
|
}); |
|
return fontCapability.promise; |
|
}, |
|
|
|
buildPath: function PartialEvaluator_buildPath(operatorList, fn, args) { |
|
var lastIndex = operatorList.length - 1; |
|
if (!args) { |
|
args = []; |
|
} |
|
if (lastIndex < 0 || |
|
operatorList.fnArray[lastIndex] !== OPS.constructPath) { |
|
operatorList.addOp(OPS.constructPath, [[fn], args]); |
|
} else { |
|
var opArgs = operatorList.argsArray[lastIndex]; |
|
opArgs[0].push(fn); |
|
Array.prototype.push.apply(opArgs[1], args); |
|
} |
|
}, |
|
|
|
handleColorN: function PartialEvaluator_handleColorN(operatorList, fn, args, |
|
cs, patterns, resources, task, xref) { |
|
// compile tiling patterns |
|
var patternName = args[args.length - 1]; |
|
// SCN/scn applies patterns along with normal colors |
|
var pattern; |
|
if (isName(patternName) && |
|
(pattern = patterns.get(patternName.name))) { |
|
var dict = (isStream(pattern) ? pattern.dict : pattern); |
|
var typeNum = dict.get('PatternType'); |
|
|
|
if (typeNum === TILING_PATTERN) { |
|
var color = cs.base ? cs.base.getRgb(args, 0) : null; |
|
return this.handleTilingType(fn, color, resources, pattern, |
|
dict, operatorList, task); |
|
} else if (typeNum === SHADING_PATTERN) { |
|
var shading = dict.get('Shading'); |
|
var matrix = dict.get('Matrix'); |
|
pattern = Pattern.parseShading(shading, matrix, xref, resources, |
|
this.handler); |
|
operatorList.addOp(fn, pattern.getIR()); |
|
return Promise.resolve(); |
|
} else { |
|
return Promise.reject('Unknown PatternType: ' + typeNum); |
|
} |
|
} |
|
// TODO shall we fail here? |
|
operatorList.addOp(fn, args); |
|
return Promise.resolve(); |
|
}, |
|
|
|
getOperatorList: function PartialEvaluator_getOperatorList(stream, |
|
task, |
|
resources, |
|
operatorList, |
|
initialState) { |
|
|
|
var self = this; |
|
var xref = this.xref; |
|
var imageCache = {}; |
|
|
|
assert(operatorList); |
|
|
|
resources = (resources || Dict.empty); |
|
var xobjs = (resources.get('XObject') || Dict.empty); |
|
var patterns = (resources.get('Pattern') || Dict.empty); |
|
var stateManager = new StateManager(initialState || new EvalState()); |
|
var preprocessor = new EvaluatorPreprocessor(stream, xref, stateManager); |
|
var timeSlotManager = new TimeSlotManager(); |
|
|
|
return new Promise(function next(resolve, reject) { |
|
task.ensureNotTerminated(); |
|
timeSlotManager.reset(); |
|
var stop, operation = {}, i, ii, cs; |
|
while (!(stop = timeSlotManager.check())) { |
|
// The arguments parsed by read() are used beyond this loop, so we |
|
// cannot reuse the same array on each iteration. Therefore we pass |
|
// in |null| as the initial value (see the comment on |
|
// EvaluatorPreprocessor_read() for why). |
|
operation.args = null; |
|
if (!(preprocessor.read(operation))) { |
|
break; |
|
} |
|
var args = operation.args; |
|
var fn = operation.fn; |
|
|
|
switch (fn | 0) { |
|
case OPS.paintXObject: |
|
if (args[0].code) { |
|
break; |
|
} |
|
// eagerly compile XForm objects |
|
var name = args[0].name; |
|
if (!name) { |
|
warn('XObject must be referred to by name.'); |
|
continue; |
|
} |
|
if (imageCache[name] !== undefined) { |
|
operatorList.addOp(imageCache[name].fn, imageCache[name].args); |
|
args = null; |
|
continue; |
|
} |
|
|
|
var xobj = xobjs.get(name); |
|
if (xobj) { |
|
assert(isStream(xobj), 'XObject should be a stream'); |
|
|
|
var type = xobj.dict.get('Subtype'); |
|
assert(isName(type), |
|
'XObject should have a Name subtype'); |
|
|
|
if (type.name === 'Form') { |
|
stateManager.save(); |
|
return self.buildFormXObject(resources, xobj, null, |
|
operatorList, task, |
|
stateManager.state.clone()). |
|
then(function () { |
|
stateManager.restore(); |
|
next(resolve, reject); |
|
}, reject); |
|
} else if (type.name === 'Image') { |
|
self.buildPaintImageXObject(resources, xobj, false, |
|
operatorList, name, imageCache); |
|
args = null; |
|
continue; |
|
} else if (type.name === 'PS') { |
|
// PostScript XObjects are unused when viewing documents. |
|
// See section 4.7.1 of Adobe's PDF reference. |
|
info('Ignored XObject subtype PS'); |
|
continue; |
|
} else { |
|
error('Unhandled XObject subtype ' + type.name); |
|
} |
|
} |
|
break; |
|
case OPS.setFont: |
|
var fontSize = args[1]; |
|
// eagerly collect all fonts |
|
return self.handleSetFont(resources, args, null, operatorList, |
|
task, stateManager.state). |
|
then(function (loadedName) { |
|
operatorList.addDependency(loadedName); |
|
operatorList.addOp(OPS.setFont, [loadedName, fontSize]); |
|
next(resolve, reject); |
|
}, reject); |
|
case OPS.endInlineImage: |
|
var cacheKey = args[0].cacheKey; |
|
if (cacheKey) { |
|
var cacheEntry = imageCache[cacheKey]; |
|
if (cacheEntry !== undefined) { |
|
operatorList.addOp(cacheEntry.fn, cacheEntry.args); |
|
args = null; |
|
continue; |
|
} |
|
} |
|
self.buildPaintImageXObject(resources, args[0], true, |
|
operatorList, cacheKey, imageCache); |
|
args = null; |
|
continue; |
|
case OPS.showText: |
|
args[0] = self.handleText(args[0], stateManager.state); |
|
break; |
|
case OPS.showSpacedText: |
|
var arr = args[0]; |
|
var combinedGlyphs = []; |
|
var arrLength = arr.length; |
|
var state = stateManager.state; |
|
for (i = 0; i < arrLength; ++i) { |
|
var arrItem = arr[i]; |
|
if (isString(arrItem)) { |
|
Array.prototype.push.apply(combinedGlyphs, |
|
self.handleText(arrItem, state)); |
|
} else if (isNum(arrItem)) { |
|
combinedGlyphs.push(arrItem); |
|
} |
|
} |
|
args[0] = combinedGlyphs; |
|
fn = OPS.showText; |
|
break; |
|
case OPS.nextLineShowText: |
|
operatorList.addOp(OPS.nextLine); |
|
args[0] = self.handleText(args[0], stateManager.state); |
|
fn = OPS.showText; |
|
break; |
|
case OPS.nextLineSetSpacingShowText: |
|
operatorList.addOp(OPS.nextLine); |
|
operatorList.addOp(OPS.setWordSpacing, [args.shift()]); |
|
operatorList.addOp(OPS.setCharSpacing, [args.shift()]); |
|
args[0] = self.handleText(args[0], stateManager.state); |
|
fn = OPS.showText; |
|
break; |
|
case OPS.setTextRenderingMode: |
|
stateManager.state.textRenderingMode = args[0]; |
|
break; |
|
|
|
case OPS.setFillColorSpace: |
|
stateManager.state.fillColorSpace = |
|
ColorSpace.parse(args[0], xref, resources); |
|
continue; |
|
case OPS.setStrokeColorSpace: |
|
stateManager.state.strokeColorSpace = |
|
ColorSpace.parse(args[0], xref, resources); |
|
continue; |
|
case OPS.setFillColor: |
|
cs = stateManager.state.fillColorSpace; |
|
args = cs.getRgb(args, 0); |
|
fn = OPS.setFillRGBColor; |
|
break; |
|
case OPS.setStrokeColor: |
|
cs = stateManager.state.strokeColorSpace; |
|
args = cs.getRgb(args, 0); |
|
fn = OPS.setStrokeRGBColor; |
|
break; |
|
case OPS.setFillGray: |
|
stateManager.state.fillColorSpace = ColorSpace.singletons.gray; |
|
args = ColorSpace.singletons.gray.getRgb(args, 0); |
|
fn = OPS.setFillRGBColor; |
|
break; |
|
case OPS.setStrokeGray: |
|
stateManager.state.strokeColorSpace = ColorSpace.singletons.gray; |
|
args = ColorSpace.singletons.gray.getRgb(args, 0); |
|
fn = OPS.setStrokeRGBColor; |
|
break; |
|
case OPS.setFillCMYKColor: |
|
stateManager.state.fillColorSpace = ColorSpace.singletons.cmyk; |
|
args = ColorSpace.singletons.cmyk.getRgb(args, 0); |
|
fn = OPS.setFillRGBColor; |
|
break; |
|
case OPS.setStrokeCMYKColor: |
|
stateManager.state.strokeColorSpace = ColorSpace.singletons.cmyk; |
|
args = ColorSpace.singletons.cmyk.getRgb(args, 0); |
|
fn = OPS.setStrokeRGBColor; |
|
break; |
|
case OPS.setFillRGBColor: |
|
stateManager.state.fillColorSpace = ColorSpace.singletons.rgb; |
|
args = ColorSpace.singletons.rgb.getRgb(args, 0); |
|
break; |
|
case OPS.setStrokeRGBColor: |
|
stateManager.state.strokeColorSpace = ColorSpace.singletons.rgb; |
|
args = ColorSpace.singletons.rgb.getRgb(args, 0); |
|
break; |
|
case OPS.setFillColorN: |
|
cs = stateManager.state.fillColorSpace; |
|
if (cs.name === 'Pattern') { |
|
return self.handleColorN(operatorList, OPS.setFillColorN, |
|
args, cs, patterns, resources, task, xref).then(function() { |
|
next(resolve, reject); |
|
}, reject); |
|
} |
|
args = cs.getRgb(args, 0); |
|
fn = OPS.setFillRGBColor; |
|
break; |
|
case OPS.setStrokeColorN: |
|
cs = stateManager.state.strokeColorSpace; |
|
if (cs.name === 'Pattern') { |
|
return self.handleColorN(operatorList, OPS.setStrokeColorN, |
|
args, cs, patterns, resources, task, xref).then(function() { |
|
next(resolve, reject); |
|
}, reject); |
|
} |
|
args = cs.getRgb(args, 0); |
|
fn = OPS.setStrokeRGBColor; |
|
break; |
|
|
|
case OPS.shadingFill: |
|
var shadingRes = resources.get('Shading'); |
|
if (!shadingRes) { |
|
error('No shading resource found'); |
|
} |
|
|
|
var shading = shadingRes.get(args[0].name); |
|
if (!shading) { |
|
error('No shading object found'); |
|
} |
|
|
|
var shadingFill = Pattern.parseShading(shading, null, xref, |
|
resources, self.handler); |
|
var patternIR = shadingFill.getIR(); |
|
args = [patternIR]; |
|
fn = OPS.shadingFill; |
|
break; |
|
case OPS.setGState: |
|
var dictName = args[0]; |
|
var extGState = resources.get('ExtGState'); |
|
|
|
if (!isDict(extGState) || !extGState.has(dictName.name)) { |
|
break; |
|
} |
|
|
|
var gState = extGState.get(dictName.name); |
|
return self.setGState(resources, gState, operatorList, task, |
|
xref, stateManager).then(function() { |
|
next(resolve, reject); |
|
}, reject); |
|
case OPS.moveTo: |
|
case OPS.lineTo: |
|
case OPS.curveTo: |
|
case OPS.curveTo2: |
|
case OPS.curveTo3: |
|
case OPS.closePath: |
|
self.buildPath(operatorList, fn, args); |
|
continue; |
|
case OPS.rectangle: |
|
self.buildPath(operatorList, fn, args); |
|
continue; |
|
case OPS.markPoint: |
|
case OPS.markPointProps: |
|
case OPS.beginMarkedContent: |
|
case OPS.beginMarkedContentProps: |
|
case OPS.endMarkedContent: |
|
case OPS.beginCompat: |
|
case OPS.endCompat: |
|
// Ignore operators where the corresponding handlers are known to |
|
// be no-op in CanvasGraphics (display/canvas.js). This prevents |
|
// serialization errors and is also a bit more efficient. |
|
// We could also try to serialize all objects in a general way, |
|
// e.g. as done in https://github.com/mozilla/pdf.js/pull/6266, |
|
// but doing so is meaningless without knowing the semantics. |
|
continue; |
|
default: |
|
// Note: Let's hope that the ignored operator does not have any |
|
// non-serializable arguments, otherwise postMessage will throw |
|
// "An object could not be cloned.". |
|
} |
|
operatorList.addOp(fn, args); |
|
} |
|
if (stop) { |
|
deferred.then(function () { |
|
next(resolve, reject); |
|
}, reject); |
|
return; |
|
} |
|
// Some PDFs don't close all restores inside object/form. |
|
// Closing those for them. |
|
for (i = 0, ii = preprocessor.savedStatesDepth; i < ii; i++) { |
|
operatorList.addOp(OPS.restore, []); |
|
} |
|
resolve(); |
|
}); |
|
}, |
|
|
|
getTextContent: |
|
function PartialEvaluator_getTextContent(stream, task, resources, |
|
stateManager, |
|
normalizeWhitespace) { |
|
|
|
stateManager = (stateManager || new StateManager(new TextState())); |
|
|
|
var WhitespaceRegexp = /\s/g; |
|
|
|
var textContent = { |
|
items: [], |
|
styles: Object.create(null) |
|
}; |
|
var textContentItem = { |
|
initialized: false, |
|
str: [], |
|
width: 0, |
|
height: 0, |
|
vertical: false, |
|
lastAdvanceWidth: 0, |
|
lastAdvanceHeight: 0, |
|
textAdvanceScale: 0, |
|
spaceWidth: 0, |
|
fakeSpaceMin: Infinity, |
|
fakeMultiSpaceMin: Infinity, |
|
fakeMultiSpaceMax: -0, |
|
textRunBreakAllowed: false, |
|
transform: null, |
|
fontName: null |
|
}; |
|
var SPACE_FACTOR = 0.3; |
|
var MULTI_SPACE_FACTOR = 1.5; |
|
var MULTI_SPACE_FACTOR_MAX = 4; |
|
|
|
var self = this; |
|
var xref = this.xref; |
|
|
|
resources = (xref.fetchIfRef(resources) || Dict.empty); |
|
|
|
// The xobj is parsed iff it's needed, e.g. if there is a `DO` cmd. |
|
var xobjs = null; |
|
var xobjsCache = {}; |
|
|
|
var preprocessor = new EvaluatorPreprocessor(stream, xref, stateManager); |
|
|
|
var textState; |
|
|
|
function ensureTextContentItem() { |
|
if (textContentItem.initialized) { |
|
return textContentItem; |
|
} |
|
var font = textState.font; |
|
if (!(font.loadedName in textContent.styles)) { |
|
textContent.styles[font.loadedName] = { |
|
fontFamily: font.fallbackName, |
|
ascent: font.ascent, |
|
descent: font.descent, |
|
vertical: font.vertical |
|
}; |
|
} |
|
textContentItem.fontName = font.loadedName; |
|
|
|
// 9.4.4 Text Space Details |
|
var tsm = [textState.fontSize * textState.textHScale, 0, |
|
0, textState.fontSize, |
|
0, textState.textRise]; |
|
|
|
if (font.isType3Font && |
|
textState.fontMatrix !== FONT_IDENTITY_MATRIX && |
|
textState.fontSize === 1) { |
|
var glyphHeight = font.bbox[3] - font.bbox[1]; |
|
if (glyphHeight > 0) { |
|
glyphHeight = glyphHeight * textState.fontMatrix[3]; |
|
tsm[3] *= glyphHeight; |
|
} |
|
} |
|
|
|
var trm = Util.transform(textState.ctm, |
|
Util.transform(textState.textMatrix, tsm)); |
|
textContentItem.transform = trm; |
|
if (!font.vertical) { |
|
textContentItem.width = 0; |
|
textContentItem.height = Math.sqrt(trm[2] * trm[2] + trm[3] * trm[3]); |
|
textContentItem.vertical = false; |
|
} else { |
|
textContentItem.width = Math.sqrt(trm[0] * trm[0] + trm[1] * trm[1]); |
|
textContentItem.height = 0; |
|
textContentItem.vertical = true; |
|
} |
|
|
|
var a = textState.textLineMatrix[0]; |
|
var b = textState.textLineMatrix[1]; |
|
var scaleLineX = Math.sqrt(a * a + b * b); |
|
a = textState.ctm[0]; |
|
b = textState.ctm[1]; |
|
var scaleCtmX = Math.sqrt(a * a + b * b); |
|
textContentItem.textAdvanceScale = scaleCtmX * scaleLineX; |
|
textContentItem.lastAdvanceWidth = 0; |
|
textContentItem.lastAdvanceHeight = 0; |
|
|
|
var spaceWidth = font.spaceWidth / 1000 * textState.fontSize; |
|
if (spaceWidth) { |
|
textContentItem.spaceWidth = spaceWidth; |
|
textContentItem.fakeSpaceMin = spaceWidth * SPACE_FACTOR; |
|
textContentItem.fakeMultiSpaceMin = spaceWidth * MULTI_SPACE_FACTOR; |
|
textContentItem.fakeMultiSpaceMax = |
|
spaceWidth * MULTI_SPACE_FACTOR_MAX; |
|
// It's okay for monospace fonts to fake as much space as needed. |
|
textContentItem.textRunBreakAllowed = !font.isMonospace; |
|
} else { |
|
textContentItem.spaceWidth = 0; |
|
textContentItem.fakeSpaceMin = Infinity; |
|
textContentItem.fakeMultiSpaceMin = Infinity; |
|
textContentItem.fakeMultiSpaceMax = 0; |
|
textContentItem.textRunBreakAllowed = false; |
|
} |
|
|
|
|
|
textContentItem.initialized = true; |
|
return textContentItem; |
|
} |
|
|
|
function replaceWhitespace(str) { |
|
// Replaces all whitespaces with standard spaces (0x20), to avoid |
|
// alignment issues between the textLayer and the canvas if the text |
|
// contains e.g. tabs (fixes issue6612.pdf). |
|
var i = 0, ii = str.length, code; |
|
while (i < ii && (code = str.charCodeAt(i)) >= 0x20 && code <= 0x7F) { |
|
i++; |
|
} |
|
return (i < ii ? str.replace(WhitespaceRegexp, ' ') : str); |
|
} |
|
|
|
function runBidiTransform(textChunk) { |
|
var str = textChunk.str.join(''); |
|
var bidiResult = PDFJS.bidi(str, -1, textChunk.vertical); |
|
return { |
|
str: (normalizeWhitespace ? replaceWhitespace(bidiResult.str) : |
|
bidiResult.str), |
|
dir: bidiResult.dir, |
|
width: textChunk.width, |
|
height: textChunk.height, |
|
transform: textChunk.transform, |
|
fontName: textChunk.fontName |
|
}; |
|
} |
|
|
|
function handleSetFont(fontName, fontRef) { |
|
return self.loadFont(fontName, fontRef, xref, resources). |
|
then(function (translated) { |
|
textState.font = translated.font; |
|
textState.fontMatrix = translated.font.fontMatrix || |
|
FONT_IDENTITY_MATRIX; |
|
}); |
|
} |
|
|
|
function buildTextContentItem(chars) { |
|
var font = textState.font; |
|
var textChunk = ensureTextContentItem(); |
|
var width = 0; |
|
var height = 0; |
|
var glyphs = font.charsToGlyphs(chars); |
|
var defaultVMetrics = font.defaultVMetrics; |
|
for (var i = 0; i < glyphs.length; i++) { |
|
var glyph = glyphs[i]; |
|
var vMetricX = null; |
|
var vMetricY = null; |
|
var glyphWidth = null; |
|
if (font.vertical) { |
|
if (glyph.vmetric) { |
|
glyphWidth = glyph.vmetric[0]; |
|
vMetricX = glyph.vmetric[1]; |
|
vMetricY = glyph.vmetric[2]; |
|
} else { |
|
glyphWidth = glyph.width; |
|
vMetricX = glyph.width * 0.5; |
|
vMetricY = defaultVMetrics[2]; |
|
} |
|
} else { |
|
glyphWidth = glyph.width; |
|
} |
|
|
|
var glyphUnicode = glyph.unicode; |
|
if (NormalizedUnicodes[glyphUnicode] !== undefined) { |
|
glyphUnicode = NormalizedUnicodes[glyphUnicode]; |
|
} |
|
glyphUnicode = reverseIfRtl(glyphUnicode); |
|
|
|
// The following will calculate the x and y of the individual glyphs. |
|
// if (font.vertical) { |
|
// tsm[4] -= vMetricX * Math.abs(textState.fontSize) * |
|
// textState.fontMatrix[0]; |
|
// tsm[5] -= vMetricY * textState.fontSize * |
|
// textState.fontMatrix[0]; |
|
// } |
|
// var trm = Util.transform(textState.textMatrix, tsm); |
|
// var pt = Util.applyTransform([trm[4], trm[5]], textState.ctm); |
|
// var x = pt[0]; |
|
// var y = pt[1]; |
|
|
|
var charSpacing = textState.charSpacing; |
|
if (glyph.isSpace) { |
|
var wordSpacing = textState.wordSpacing; |
|
charSpacing += wordSpacing; |
|
if (wordSpacing > 0) { |
|
addFakeSpaces(wordSpacing, textChunk.str); |
|
} |
|
} |
|
|
|
var tx = 0; |
|
var ty = 0; |
|
if (!font.vertical) { |
|
var w0 = glyphWidth * textState.fontMatrix[0]; |
|
tx = (w0 * textState.fontSize + charSpacing) * |
|
textState.textHScale; |
|
width += tx; |
|
} else { |
|
var w1 = glyphWidth * textState.fontMatrix[0]; |
|
ty = w1 * textState.fontSize + charSpacing; |
|
height += ty; |
|
} |
|
textState.translateTextMatrix(tx, ty); |
|
|
|
textChunk.str.push(glyphUnicode); |
|
} |
|
|
|
if (!font.vertical) { |
|
textChunk.lastAdvanceWidth = width; |
|
textChunk.width += width * textChunk.textAdvanceScale; |
|
} else { |
|
textChunk.lastAdvanceHeight = height; |
|
textChunk.height += Math.abs(height * textChunk.textAdvanceScale); |
|
} |
|
|
|
return textChunk; |
|
} |
|
|
|
function addFakeSpaces(width, strBuf) { |
|
if (width < textContentItem.fakeSpaceMin) { |
|
return; |
|
} |
|
if (width < textContentItem.fakeMultiSpaceMin) { |
|
strBuf.push(' '); |
|
return; |
|
} |
|
var fakeSpaces = Math.round(width / textContentItem.spaceWidth); |
|
while (fakeSpaces-- > 0) { |
|
strBuf.push(' '); |
|
} |
|
} |
|
|
|
function flushTextContentItem() { |
|
if (!textContentItem.initialized) { |
|
return; |
|
} |
|
textContent.items.push(runBidiTransform(textContentItem)); |
|
|
|
textContentItem.initialized = false; |
|
textContentItem.str.length = 0; |
|
} |
|
|
|
var timeSlotManager = new TimeSlotManager(); |
|
|
|
return new Promise(function next(resolve, reject) { |
|
task.ensureNotTerminated(); |
|
timeSlotManager.reset(); |
|
var stop, operation = {}, args = []; |
|
while (!(stop = timeSlotManager.check())) { |
|
// The arguments parsed by read() are not used beyond this loop, so |
|
// we can reuse the same array on every iteration, thus avoiding |
|
// unnecessary allocations. |
|
args.length = 0; |
|
operation.args = args; |
|
if (!(preprocessor.read(operation))) { |
|
break; |
|
} |
|
textState = stateManager.state; |
|
var fn = operation.fn; |
|
args = operation.args; |
|
var advance; |
|
|
|
switch (fn | 0) { |
|
case OPS.setFont: |
|
flushTextContentItem(); |
|
textState.fontSize = args[1]; |
|
return handleSetFont(args[0].name).then(function() { |
|
next(resolve, reject); |
|
}, reject); |
|
case OPS.setTextRise: |
|
flushTextContentItem(); |
|
textState.textRise = args[0]; |
|
break; |
|
case OPS.setHScale: |
|
flushTextContentItem(); |
|
textState.textHScale = args[0] / 100; |
|
break; |
|
case OPS.setLeading: |
|
flushTextContentItem(); |
|
textState.leading = args[0]; |
|
break; |
|
case OPS.moveText: |
|
// Optimization to treat same line movement as advance |
|
var isSameTextLine = !textState.font ? false : |
|
((textState.font.vertical ? args[0] : args[1]) === 0); |
|
advance = args[0] - args[1]; |
|
if (isSameTextLine && textContentItem.initialized && |
|
advance > 0 && |
|
advance <= textContentItem.fakeMultiSpaceMax) { |
|
textState.translateTextLineMatrix(args[0], args[1]); |
|
textContentItem.width += |
|
(args[0] - textContentItem.lastAdvanceWidth); |
|
textContentItem.height += |
|
(args[1] - textContentItem.lastAdvanceHeight); |
|
var diff = (args[0] - textContentItem.lastAdvanceWidth) - |
|
(args[1] - textContentItem.lastAdvanceHeight); |
|
addFakeSpaces(diff, textContentItem.str); |
|
break; |
|
} |
|
|
|
flushTextContentItem(); |
|
textState.translateTextLineMatrix(args[0], args[1]); |
|
textState.textMatrix = textState.textLineMatrix.slice(); |
|
break; |
|
case OPS.setLeadingMoveText: |
|
flushTextContentItem(); |
|
textState.leading = -args[1]; |
|
textState.translateTextLineMatrix(args[0], args[1]); |
|
textState.textMatrix = textState.textLineMatrix.slice(); |
|
break; |
|
case OPS.nextLine: |
|
flushTextContentItem(); |
|
textState.carriageReturn(); |
|
break; |
|
case OPS.setTextMatrix: |
|
flushTextContentItem(); |
|
textState.setTextMatrix(args[0], args[1], args[2], args[3], |
|
args[4], args[5]); |
|
textState.setTextLineMatrix(args[0], args[1], args[2], args[3], |
|
args[4], args[5]); |
|
break; |
|
case OPS.setCharSpacing: |
|
textState.charSpacing = args[0]; |
|
break; |
|
case OPS.setWordSpacing: |
|
textState.wordSpacing = args[0]; |
|
break; |
|
case OPS.beginText: |
|
flushTextContentItem(); |
|
textState.textMatrix = IDENTITY_MATRIX.slice(); |
|
textState.textLineMatrix = IDENTITY_MATRIX.slice(); |
|
break; |
|
case OPS.showSpacedText: |
|
var items = args[0]; |
|
var offset; |
|
for (var j = 0, jj = items.length; j < jj; j++) { |
|
if (typeof items[j] === 'string') { |
|
buildTextContentItem(items[j]); |
|
} else { |
|
ensureTextContentItem(); |
|
|
|
// PDF Specification 5.3.2 states: |
|
// The number is expressed in thousandths of a unit of text |
|
// space. |
|
// This amount is subtracted from the current horizontal or |
|
// vertical coordinate, depending on the writing mode. |
|
// In the default coordinate system, a positive adjustment |
|
// has the effect of moving the next glyph painted either to |
|
// the left or down by the given amount. |
|
advance = items[j] * textState.fontSize / 1000; |
|
var breakTextRun = false; |
|
if (textState.font.vertical) { |
|
offset = advance * |
|
(textState.textHScale * textState.textMatrix[2] + |
|
textState.textMatrix[3]); |
|
textState.translateTextMatrix(0, advance); |
|
breakTextRun = textContentItem.textRunBreakAllowed && |
|
advance > textContentItem.fakeMultiSpaceMax; |
|
if (!breakTextRun) { |
|
// Value needs to be added to height to paint down. |
|
textContentItem.height += offset; |
|
} |
|
} else { |
|
advance = -advance; |
|
offset = advance * ( |
|
textState.textHScale * textState.textMatrix[0] + |
|
textState.textMatrix[1]); |
|
textState.translateTextMatrix(advance, 0); |
|
breakTextRun = textContentItem.textRunBreakAllowed && |
|
advance > textContentItem.fakeMultiSpaceMax; |
|
if (!breakTextRun) { |
|
// Value needs to be subtracted from width to paint left. |
|
textContentItem.width += offset; |
|
} |
|
} |
|
if (breakTextRun) { |
|
flushTextContentItem(); |
|
} else if (advance > 0) { |
|
addFakeSpaces(advance, textContentItem.str); |
|
} |
|
} |
|
} |
|
break; |
|
case OPS.showText: |
|
buildTextContentItem(args[0]); |
|
break; |
|
case OPS.nextLineShowText: |
|
flushTextContentItem(); |
|
textState.carriageReturn(); |
|
buildTextContentItem(args[0]); |
|
break; |
|
case OPS.nextLineSetSpacingShowText: |
|
flushTextContentItem(); |
|
textState.wordSpacing = args[0]; |
|
textState.charSpacing = args[1]; |
|
textState.carriageReturn(); |
|
buildTextContentItem(args[2]); |
|
break; |
|
case OPS.paintXObject: |
|
flushTextContentItem(); |
|
if (args[0].code) { |
|
break; |
|
} |
|
|
|
if (!xobjs) { |
|
xobjs = (resources.get('XObject') || Dict.empty); |
|
} |
|
|
|
var name = args[0].name; |
|
if (xobjsCache.key === name) { |
|
if (xobjsCache.texts) { |
|
Util.appendToArray(textContent.items, xobjsCache.texts.items); |
|
Util.extendObj(textContent.styles, xobjsCache.texts.styles); |
|
} |
|
break; |
|
} |
|
|
|
var xobj = xobjs.get(name); |
|
if (!xobj) { |
|
break; |
|
} |
|
assert(isStream(xobj), 'XObject should be a stream'); |
|
|
|
var type = xobj.dict.get('Subtype'); |
|
assert(isName(type), |
|
'XObject should have a Name subtype'); |
|
|
|
if ('Form' !== type.name) { |
|
xobjsCache.key = name; |
|
xobjsCache.texts = null; |
|
break; |
|
} |
|
|
|
stateManager.save(); |
|
var matrix = xobj.dict.get('Matrix'); |
|
if (isArray(matrix) && matrix.length === 6) { |
|
stateManager.transform(matrix); |
|
} |
|
|
|
return self.getTextContent(xobj, task, |
|
xobj.dict.get('Resources') || resources, stateManager, |
|
normalizeWhitespace).then(function (formTextContent) { |
|
Util.appendToArray(textContent.items, formTextContent.items); |
|
Util.extendObj(textContent.styles, formTextContent.styles); |
|
stateManager.restore(); |
|
|
|
xobjsCache.key = name; |
|
xobjsCache.texts = formTextContent; |
|
|
|
next(resolve, reject); |
|
}, reject); |
|
case OPS.setGState: |
|
flushTextContentItem(); |
|
var dictName = args[0]; |
|
var extGState = resources.get('ExtGState'); |
|
|
|
if (!isDict(extGState) || !extGState.has(dictName.name)) { |
|
break; |
|
} |
|
|
|
var gsStateMap = extGState.get(dictName.name); |
|
var gsStateFont = null; |
|
for (var key in gsStateMap) { |
|
if (key === 'Font') { |
|
assert(!gsStateFont); |
|
gsStateFont = gsStateMap[key]; |
|
} |
|
} |
|
if (gsStateFont) { |
|
textState.fontSize = gsStateFont[1]; |
|
return handleSetFont(gsStateFont[0]).then(function() { |
|
next(resolve, reject); |
|
}, reject); |
|
} |
|
break; |
|
} // switch |
|
} // while |
|
if (stop) { |
|
deferred.then(function () { |
|
next(resolve, reject); |
|
}, reject); |
|
return; |
|
} |
|
flushTextContentItem(); |
|
resolve(textContent); |
|
}); |
|
}, |
|
|
|
extractDataStructures: function |
|
partialEvaluatorExtractDataStructures(dict, baseDict, |
|
xref, properties) { |
|
// 9.10.2 |
|
var toUnicode = (dict.get('ToUnicode') || baseDict.get('ToUnicode')); |
|
if (toUnicode) { |
|
properties.toUnicode = this.readToUnicode(toUnicode); |
|
} |
|
if (properties.composite) { |
|
// CIDSystemInfo helps to match CID to glyphs |
|
var cidSystemInfo = dict.get('CIDSystemInfo'); |
|
if (isDict(cidSystemInfo)) { |
|
properties.cidSystemInfo = { |
|
registry: cidSystemInfo.get('Registry'), |
|
ordering: cidSystemInfo.get('Ordering'), |
|
supplement: cidSystemInfo.get('Supplement') |
|
}; |
|
} |
|
|
|
var cidToGidMap = dict.get('CIDToGIDMap'); |
|
if (isStream(cidToGidMap)) { |
|
properties.cidToGidMap = this.readCidToGidMap(cidToGidMap); |
|
} |
|
} |
|
|
|
// Based on 9.6.6 of the spec the encoding can come from multiple places |
|
// and depends on the font type. The base encoding and differences are |
|
// read here, but the encoding that is actually used is chosen during |
|
// glyph mapping in the font. |
|
// TODO: Loading the built in encoding in the font would allow the |
|
// differences to be merged in here not require us to hold on to it. |
|
var differences = []; |
|
var baseEncodingName = null; |
|
var encoding; |
|
if (dict.has('Encoding')) { |
|
encoding = dict.get('Encoding'); |
|
if (isDict(encoding)) { |
|
baseEncodingName = encoding.get('BaseEncoding'); |
|
baseEncodingName = (isName(baseEncodingName) ? |
|
baseEncodingName.name : null); |
|
// Load the differences between the base and original |
|
if (encoding.has('Differences')) { |
|
var diffEncoding = encoding.get('Differences'); |
|
var index = 0; |
|
for (var j = 0, jj = diffEncoding.length; j < jj; j++) { |
|
var data = diffEncoding[j]; |
|
if (isNum(data)) { |
|
index = data; |
|
} else if (isName(data)) { |
|
differences[index++] = data.name; |
|
} else if (isRef(data)) { |
|
diffEncoding[j--] = xref.fetch(data); |
|
continue; |
|
} else { |
|
error('Invalid entry in \'Differences\' array: ' + data); |
|
} |
|
} |
|
} |
|
} else if (isName(encoding)) { |
|
baseEncodingName = encoding.name; |
|
} else { |
|
error('Encoding is not a Name nor a Dict'); |
|
} |
|
// According to table 114 if the encoding is a named encoding it must be |
|
// one of these predefined encodings. |
|
if ((baseEncodingName !== 'MacRomanEncoding' && |
|
baseEncodingName !== 'MacExpertEncoding' && |
|
baseEncodingName !== 'WinAnsiEncoding')) { |
|
baseEncodingName = null; |
|
} |
|
} |
|
|
|
if (baseEncodingName) { |
|
properties.defaultEncoding = Encodings[baseEncodingName].slice(); |
|
} else { |
|
encoding = (properties.type === 'TrueType' ? |
|
Encodings.WinAnsiEncoding : Encodings.StandardEncoding); |
|
// The Symbolic attribute can be misused for regular fonts |
|
// Heuristic: we have to check if the font is a standard one also |
|
if (!!(properties.flags & FontFlags.Symbolic)) { |
|
encoding = Encodings.MacRomanEncoding; |
|
if (!properties.file) { |
|
if (/Symbol/i.test(properties.name)) { |
|
encoding = Encodings.SymbolSetEncoding; |
|
} else if (/Dingbats/i.test(properties.name)) { |
|
encoding = Encodings.ZapfDingbatsEncoding; |
|
} |
|
} |
|
} |
|
properties.defaultEncoding = encoding; |
|
} |
|
|
|
properties.differences = differences; |
|
properties.baseEncodingName = baseEncodingName; |
|
properties.dict = dict; |
|
}, |
|
|
|
readToUnicode: function PartialEvaluator_readToUnicode(toUnicode) { |
|
var cmap, cmapObj = toUnicode; |
|
if (isName(cmapObj)) { |
|
cmap = CMapFactory.create(cmapObj, |
|
{ url: PDFJS.cMapUrl, packed: PDFJS.cMapPacked }, null); |
|
if (cmap instanceof IdentityCMap) { |
|
return new IdentityToUnicodeMap(0, 0xFFFF); |
|
} |
|
return new ToUnicodeMap(cmap.getMap()); |
|
} else if (isStream(cmapObj)) { |
|
cmap = CMapFactory.create(cmapObj, |
|
{ url: PDFJS.cMapUrl, packed: PDFJS.cMapPacked }, null); |
|
if (cmap instanceof IdentityCMap) { |
|
return new IdentityToUnicodeMap(0, 0xFFFF); |
|
} |
|
var map = new Array(cmap.length); |
|
// Convert UTF-16BE |
|
// NOTE: cmap can be a sparse array, so use forEach instead of for(;;) |
|
// to iterate over all keys. |
|
cmap.forEach(function(charCode, token) { |
|
var str = []; |
|
for (var k = 0; k < token.length; k += 2) { |
|
var w1 = (token.charCodeAt(k) << 8) | token.charCodeAt(k + 1); |
|
if ((w1 & 0xF800) !== 0xD800) { // w1 < 0xD800 || w1 > 0xDFFF |
|
str.push(w1); |
|
continue; |
|
} |
|
k += 2; |
|
var w2 = (token.charCodeAt(k) << 8) | token.charCodeAt(k + 1); |
|
str.push(((w1 & 0x3ff) << 10) + (w2 & 0x3ff) + 0x10000); |
|
} |
|
map[charCode] = String.fromCharCode.apply(String, str); |
|
}); |
|
return new ToUnicodeMap(map); |
|
} |
|
return null; |
|
}, |
|
|
|
readCidToGidMap: function PartialEvaluator_readCidToGidMap(cidToGidStream) { |
|
// Extract the encoding from the CIDToGIDMap |
|
var glyphsData = cidToGidStream.getBytes(); |
|
|
|
// Set encoding 0 to later verify the font has an encoding |
|
var result = []; |
|
for (var j = 0, jj = glyphsData.length; j < jj; j++) { |
|
var glyphID = (glyphsData[j++] << 8) | glyphsData[j]; |
|
if (glyphID === 0) { |
|
continue; |
|
} |
|
var code = j >> 1; |
|
result[code] = glyphID; |
|
} |
|
return result; |
|
}, |
|
|
|
extractWidths: function PartialEvaluator_extractWidths(dict, xref, |
|
descriptor, |
|
properties) { |
|
var glyphsWidths = []; |
|
var defaultWidth = 0; |
|
var glyphsVMetrics = []; |
|
var defaultVMetrics; |
|
var i, ii, j, jj, start, code, widths; |
|
if (properties.composite) { |
|
defaultWidth = dict.get('DW') || 1000; |
|
|
|
widths = dict.get('W'); |
|
if (widths) { |
|
for (i = 0, ii = widths.length; i < ii; i++) { |
|
start = widths[i++]; |
|
code = xref.fetchIfRef(widths[i]); |
|
if (isArray(code)) { |
|
for (j = 0, jj = code.length; j < jj; j++) { |
|
glyphsWidths[start++] = code[j]; |
|
} |
|
} else { |
|
var width = widths[++i]; |
|
for (j = start; j <= code; j++) { |
|
glyphsWidths[j] = width; |
|
} |
|
} |
|
} |
|
} |
|
|
|
if (properties.vertical) { |
|
var vmetrics = (dict.get('DW2') || [880, -1000]); |
|
defaultVMetrics = [vmetrics[1], defaultWidth * 0.5, vmetrics[0]]; |
|
vmetrics = dict.get('W2'); |
|
if (vmetrics) { |
|
for (i = 0, ii = vmetrics.length; i < ii; i++) { |
|
start = vmetrics[i++]; |
|
code = xref.fetchIfRef(vmetrics[i]); |
|
if (isArray(code)) { |
|
for (j = 0, jj = code.length; j < jj; j++) { |
|
glyphsVMetrics[start++] = [code[j++], code[j++], code[j]]; |
|
} |
|
} else { |
|
var vmetric = [vmetrics[++i], vmetrics[++i], vmetrics[++i]]; |
|
for (j = start; j <= code; j++) { |
|
glyphsVMetrics[j] = vmetric; |
|
} |
|
} |
|
} |
|
} |
|
} |
|
} else { |
|
var firstChar = properties.firstChar; |
|
widths = dict.get('Widths'); |
|
if (widths) { |
|
j = firstChar; |
|
for (i = 0, ii = widths.length; i < ii; i++) { |
|
glyphsWidths[j++] = widths[i]; |
|
} |
|
defaultWidth = (parseFloat(descriptor.get('MissingWidth')) || 0); |
|
} else { |
|
// Trying get the BaseFont metrics (see comment above). |
|
var baseFontName = dict.get('BaseFont'); |
|
if (isName(baseFontName)) { |
|
var metrics = this.getBaseFontMetrics(baseFontName.name); |
|
|
|
glyphsWidths = this.buildCharCodeToWidth(metrics.widths, |
|
properties); |
|
defaultWidth = metrics.defaultWidth; |
|
} |
|
} |
|
} |
|
|
|
// Heuristic: detection of monospace font by checking all non-zero widths |
|
var isMonospace = true; |
|
var firstWidth = defaultWidth; |
|
for (var glyph in glyphsWidths) { |
|
var glyphWidth = glyphsWidths[glyph]; |
|
if (!glyphWidth) { |
|
continue; |
|
} |
|
if (!firstWidth) { |
|
firstWidth = glyphWidth; |
|
continue; |
|
} |
|
if (firstWidth !== glyphWidth) { |
|
isMonospace = false; |
|
break; |
|
} |
|
} |
|
if (isMonospace) { |
|
properties.flags |= FontFlags.FixedPitch; |
|
} |
|
|
|
properties.defaultWidth = defaultWidth; |
|
properties.widths = glyphsWidths; |
|
properties.defaultVMetrics = defaultVMetrics; |
|
properties.vmetrics = glyphsVMetrics; |
|
}, |
|
|
|
isSerifFont: function PartialEvaluator_isSerifFont(baseFontName) { |
|
// Simulating descriptor flags attribute |
|
var fontNameWoStyle = baseFontName.split('-')[0]; |
|
return (fontNameWoStyle in serifFonts) || |
|
(fontNameWoStyle.search(/serif/gi) !== -1); |
|
}, |
|
|
|
getBaseFontMetrics: function PartialEvaluator_getBaseFontMetrics(name) { |
|
var defaultWidth = 0; |
|
var widths = []; |
|
var monospace = false; |
|
var lookupName = (stdFontMap[name] || name); |
|
|
|
if (!(lookupName in Metrics)) { |
|
// Use default fonts for looking up font metrics if the passed |
|
// font is not a base font |
|
if (this.isSerifFont(name)) { |
|
lookupName = 'Times-Roman'; |
|
} else { |
|
lookupName = 'Helvetica'; |
|
} |
|
} |
|
var glyphWidths = Metrics[lookupName]; |
|
|
|
if (isNum(glyphWidths)) { |
|
defaultWidth = glyphWidths; |
|
monospace = true; |
|
} else { |
|
widths = glyphWidths; |
|
} |
|
|
|
return { |
|
defaultWidth: defaultWidth, |
|
monospace: monospace, |
|
widths: widths |
|
}; |
|
}, |
|
|
|
buildCharCodeToWidth: |
|
function PartialEvaluator_bulildCharCodeToWidth(widthsByGlyphName, |
|
properties) { |
|
var widths = Object.create(null); |
|
var differences = properties.differences; |
|
var encoding = properties.defaultEncoding; |
|
for (var charCode = 0; charCode < 256; charCode++) { |
|
if (charCode in differences && |
|
widthsByGlyphName[differences[charCode]]) { |
|
widths[charCode] = widthsByGlyphName[differences[charCode]]; |
|
continue; |
|
} |
|
if (charCode in encoding && widthsByGlyphName[encoding[charCode]]) { |
|
widths[charCode] = widthsByGlyphName[encoding[charCode]]; |
|
continue; |
|
} |
|
} |
|
return widths; |
|
}, |
|
|
|
preEvaluateFont: function PartialEvaluator_preEvaluateFont(dict, xref) { |
|
var baseDict = dict; |
|
var type = dict.get('Subtype'); |
|
assert(isName(type), 'invalid font Subtype'); |
|
|
|
var composite = false; |
|
var uint8array; |
|
if (type.name === 'Type0') { |
|
// If font is a composite |
|
// - get the descendant font |
|
// - set the type according to the descendant font |
|
// - get the FontDescriptor from the descendant font |
|
var df = dict.get('DescendantFonts'); |
|
if (!df) { |
|
error('Descendant fonts are not specified'); |
|
} |
|
dict = (isArray(df) ? xref.fetchIfRef(df[0]) : df); |
|
|
|
type = dict.get('Subtype'); |
|
assert(isName(type), 'invalid font Subtype'); |
|
composite = true; |
|
} |
|
|
|
var descriptor = dict.get('FontDescriptor'); |
|
if (descriptor) { |
|
var hash = new MurmurHash3_64(); |
|
var encoding = baseDict.getRaw('Encoding'); |
|
if (isName(encoding)) { |
|
hash.update(encoding.name); |
|
} else if (isRef(encoding)) { |
|
hash.update(encoding.num + '_' + encoding.gen); |
|
} else if (isDict(encoding)) { |
|
var keys = encoding.getKeys(); |
|
for (var i = 0, ii = keys.length; i < ii; i++) { |
|
var entry = encoding.getRaw(keys[i]); |
|
if (isName(entry)) { |
|
hash.update(entry.name); |
|
} else if (isRef(entry)) { |
|
hash.update(entry.num + '_' + entry.gen); |
|
} else if (isArray(entry)) { // 'Differences' entry. |
|
// Ideally we should check the contents of the array, but to avoid |
|
// parsing it here and then again in |extractDataStructures|, |
|
// we only use the array length for now (fixes bug1157493.pdf). |
|
hash.update(entry.length.toString()); |
|
} |
|
} |
|
} |
|
|
|
var toUnicode = dict.get('ToUnicode') || baseDict.get('ToUnicode'); |
|
if (isStream(toUnicode)) { |
|
var stream = toUnicode.str || toUnicode; |
|
uint8array = stream.buffer ? |
|
new Uint8Array(stream.buffer.buffer, 0, stream.bufferLength) : |
|
new Uint8Array(stream.bytes.buffer, |
|
stream.start, stream.end - stream.start); |
|
hash.update(uint8array); |
|
|
|
} else if (isName(toUnicode)) { |
|
hash.update(toUnicode.name); |
|
} |
|
|
|
var widths = dict.get('Widths') || baseDict.get('Widths'); |
|
if (widths) { |
|
uint8array = new Uint8Array(new Uint32Array(widths).buffer); |
|
hash.update(uint8array); |
|
} |
|
} |
|
|
|
return { |
|
descriptor: descriptor, |
|
dict: dict, |
|
baseDict: baseDict, |
|
composite: composite, |
|
type: type.name, |
|
hash: hash ? hash.hexdigest() : '' |
|
}; |
|
}, |
|
|
|
translateFont: function PartialEvaluator_translateFont(preEvaluatedFont, |
|
xref) { |
|
var baseDict = preEvaluatedFont.baseDict; |
|
var dict = preEvaluatedFont.dict; |
|
var composite = preEvaluatedFont.composite; |
|
var descriptor = preEvaluatedFont.descriptor; |
|
var type = preEvaluatedFont.type; |
|
var maxCharIndex = (composite ? 0xFFFF : 0xFF); |
|
var properties; |
|
|
|
if (!descriptor) { |
|
if (type === 'Type3') { |
|
// FontDescriptor is only required for Type3 fonts when the document |
|
// is a tagged pdf. Create a barbebones one to get by. |
|
descriptor = new Dict(null); |
|
descriptor.set('FontName', Name.get(type)); |
|
descriptor.set('FontBBox', dict.get('FontBBox')); |
|
} else { |
|
// Before PDF 1.5 if the font was one of the base 14 fonts, having a |
|
// FontDescriptor was not required. |
|
// This case is here for compatibility. |
|
var baseFontName = dict.get('BaseFont'); |
|
if (!isName(baseFontName)) { |
|
error('Base font is not specified'); |
|
} |
|
|
|
// Using base font name as a font name. |
|
baseFontName = baseFontName.name.replace(/[,_]/g, '-'); |
|
var metrics = this.getBaseFontMetrics(baseFontName); |
|
|
|
// Simulating descriptor flags attribute |
|
var fontNameWoStyle = baseFontName.split('-')[0]; |
|
var flags = |
|
(this.isSerifFont(fontNameWoStyle) ? FontFlags.Serif : 0) | |
|
(metrics.monospace ? FontFlags.FixedPitch : 0) | |
|
(symbolsFonts[fontNameWoStyle] ? FontFlags.Symbolic : |
|
FontFlags.Nonsymbolic); |
|
|
|
properties = { |
|
type: type, |
|
name: baseFontName, |
|
widths: metrics.widths, |
|
defaultWidth: metrics.defaultWidth, |
|
flags: flags, |
|
firstChar: 0, |
|
lastChar: maxCharIndex |
|
}; |
|
this.extractDataStructures(dict, dict, xref, properties); |
|
properties.widths = this.buildCharCodeToWidth(metrics.widths, |
|
properties); |
|
return new Font(baseFontName, null, properties); |
|
} |
|
} |
|
|
|
// According to the spec if 'FontDescriptor' is declared, 'FirstChar', |
|
// 'LastChar' and 'Widths' should exist too, but some PDF encoders seem |
|
// to ignore this rule when a variant of a standart font is used. |
|
// TODO Fill the width array depending on which of the base font this is |
|
// a variant. |
|
var firstChar = (dict.get('FirstChar') || 0); |
|
var lastChar = (dict.get('LastChar') || maxCharIndex); |
|
|
|
var fontName = descriptor.get('FontName'); |
|
var baseFont = dict.get('BaseFont'); |
|
// Some bad PDFs have a string as the font name. |
|
if (isString(fontName)) { |
|
fontName = Name.get(fontName); |
|
} |
|
if (isString(baseFont)) { |
|
baseFont = Name.get(baseFont); |
|
} |
|
|
|
if (type !== 'Type3') { |
|
var fontNameStr = fontName && fontName.name; |
|
var baseFontStr = baseFont && baseFont.name; |
|
if (fontNameStr !== baseFontStr) { |
|
info('The FontDescriptor\'s FontName is "' + fontNameStr + |
|
'" but should be the same as the Font\'s BaseFont "' + |
|
baseFontStr + '"'); |
|
// Workaround for cases where e.g. fontNameStr = 'Arial' and |
|
// baseFontStr = 'Arial,Bold' (needed when no font file is embedded). |
|
if (fontNameStr && baseFontStr && |
|
baseFontStr.indexOf(fontNameStr) === 0) { |
|
fontName = baseFont; |
|
} |
|
} |
|
} |
|
fontName = (fontName || baseFont); |
|
|
|
assert(isName(fontName), 'invalid font name'); |
|
|
|
var fontFile = descriptor.get('FontFile', 'FontFile2', 'FontFile3'); |
|
if (fontFile) { |
|
if (fontFile.dict) { |
|
var subtype = fontFile.dict.get('Subtype'); |
|
if (subtype) { |
|
subtype = subtype.name; |
|
} |
|
var length1 = fontFile.dict.get('Length1'); |
|
var length2 = fontFile.dict.get('Length2'); |
|
} |
|
} |
|
|
|
properties = { |
|
type: type, |
|
name: fontName.name, |
|
subtype: subtype, |
|
file: fontFile, |
|
length1: length1, |
|
length2: length2, |
|
loadedName: baseDict.loadedName, |
|
composite: composite, |
|
wideChars: composite, |
|
fixedPitch: false, |
|
fontMatrix: (dict.get('FontMatrix') || FONT_IDENTITY_MATRIX), |
|
firstChar: firstChar || 0, |
|
lastChar: (lastChar || maxCharIndex), |
|
bbox: descriptor.get('FontBBox'), |
|
ascent: descriptor.get('Ascent'), |
|
descent: descriptor.get('Descent'), |
|
xHeight: descriptor.get('XHeight'), |
|
capHeight: descriptor.get('CapHeight'), |
|
flags: descriptor.get('Flags'), |
|
italicAngle: descriptor.get('ItalicAngle'), |
|
coded: false |
|
}; |
|
|
|
if (composite) { |
|
var cidEncoding = baseDict.get('Encoding'); |
|
if (isName(cidEncoding)) { |
|
properties.cidEncoding = cidEncoding.name; |
|
} |
|
properties.cMap = CMapFactory.create(cidEncoding, |
|
{ url: PDFJS.cMapUrl, packed: PDFJS.cMapPacked }, null); |
|
properties.vertical = properties.cMap.vertical; |
|
} |
|
this.extractDataStructures(dict, baseDict, xref, properties); |
|
this.extractWidths(dict, xref, descriptor, properties); |
|
|
|
if (type === 'Type3') { |
|
properties.isType3Font = true; |
|
} |
|
|
|
return new Font(fontName.name, fontFile, properties); |
|
} |
|
}; |
|
|
|
return PartialEvaluator; |
|
})(); |
|
|
|
var TranslatedFont = (function TranslatedFontClosure() { |
|
function TranslatedFont(loadedName, font, dict) { |
|
this.loadedName = loadedName; |
|
this.font = font; |
|
this.dict = dict; |
|
this.type3Loaded = null; |
|
this.sent = false; |
|
} |
|
TranslatedFont.prototype = { |
|
send: function (handler) { |
|
if (this.sent) { |
|
return; |
|
} |
|
var fontData = this.font.exportData(); |
|
handler.send('commonobj', [ |
|
this.loadedName, |
|
'Font', |
|
fontData |
|
]); |
|
this.sent = true; |
|
}, |
|
loadType3Data: function (evaluator, resources, parentOperatorList, task) { |
|
assert(this.font.isType3Font); |
|
|
|
if (this.type3Loaded) { |
|
return this.type3Loaded; |
|
} |
|
|
|
var translatedFont = this.font; |
|
var loadCharProcsPromise = Promise.resolve(); |
|
var charProcs = this.dict.get('CharProcs').getAll(); |
|
var fontResources = this.dict.get('Resources') || resources; |
|
var charProcKeys = Object.keys(charProcs); |
|
var charProcOperatorList = {}; |
|
for (var i = 0, n = charProcKeys.length; i < n; ++i) { |
|
loadCharProcsPromise = loadCharProcsPromise.then(function (key) { |
|
var glyphStream = charProcs[key]; |
|
var operatorList = new OperatorList(); |
|
return evaluator.getOperatorList(glyphStream, task, fontResources, |
|
operatorList).then(function () { |
|
charProcOperatorList[key] = operatorList.getIR(); |
|
|
|
// Add the dependencies to the parent operator list so they are |
|
// resolved before sub operator list is executed synchronously. |
|
parentOperatorList.addDependencies(operatorList.dependencies); |
|
}, function (reason) { |
|
warn('Type3 font resource \"' + key + '\" is not available'); |
|
var operatorList = new OperatorList(); |
|
charProcOperatorList[key] = operatorList.getIR(); |
|
}); |
|
}.bind(this, charProcKeys[i])); |
|
} |
|
this.type3Loaded = loadCharProcsPromise.then(function () { |
|
translatedFont.charProcOperatorList = charProcOperatorList; |
|
}); |
|
return this.type3Loaded; |
|
} |
|
}; |
|
return TranslatedFont; |
|
})(); |
|
|
|
var OperatorList = (function OperatorListClosure() { |
|
var CHUNK_SIZE = 1000; |
|
var CHUNK_SIZE_ABOUT = CHUNK_SIZE - 5; // close to chunk size |
|
|
|
function getTransfers(queue) { |
|
var transfers = []; |
|
var fnArray = queue.fnArray, argsArray = queue.argsArray; |
|
for (var i = 0, ii = queue.length; i < ii; i++) { |
|
switch (fnArray[i]) { |
|
case OPS.paintInlineImageXObject: |
|
case OPS.paintInlineImageXObjectGroup: |
|
case OPS.paintImageMaskXObject: |
|
var arg = argsArray[i][0]; // first param in imgData |
|
if (!arg.cached) { |
|
transfers.push(arg.data.buffer); |
|
} |
|
break; |
|
} |
|
} |
|
return transfers; |
|
} |
|
|
|
function OperatorList(intent, messageHandler, pageIndex) { |
|
this.messageHandler = messageHandler; |
|
this.fnArray = []; |
|
this.argsArray = []; |
|
this.dependencies = {}; |
|
this._totalLength = 0; |
|
this.pageIndex = pageIndex; |
|
this.intent = intent; |
|
} |
|
|
|
OperatorList.prototype = { |
|
get length() { |
|
return this.argsArray.length; |
|
}, |
|
|
|
/** |
|
* @returns {number} The total length of the entire operator list, |
|
* since `this.length === 0` after flushing. |
|
*/ |
|
get totalLength() { |
|
return (this._totalLength + this.length); |
|
}, |
|
|
|
addOp: function(fn, args) { |
|
this.fnArray.push(fn); |
|
this.argsArray.push(args); |
|
if (this.messageHandler) { |
|
if (this.fnArray.length >= CHUNK_SIZE) { |
|
this.flush(); |
|
} else if (this.fnArray.length >= CHUNK_SIZE_ABOUT && |
|
(fn === OPS.restore || fn === OPS.endText)) { |
|
// heuristic to flush on boundary of restore or endText |
|
this.flush(); |
|
} |
|
} |
|
}, |
|
|
|
addDependency: function(dependency) { |
|
if (dependency in this.dependencies) { |
|
return; |
|
} |
|
this.dependencies[dependency] = true; |
|
this.addOp(OPS.dependency, [dependency]); |
|
}, |
|
|
|
addDependencies: function(dependencies) { |
|
for (var key in dependencies) { |
|
this.addDependency(key); |
|
} |
|
}, |
|
|
|
addOpList: function(opList) { |
|
Util.extendObj(this.dependencies, opList.dependencies); |
|
for (var i = 0, ii = opList.length; i < ii; i++) { |
|
this.addOp(opList.fnArray[i], opList.argsArray[i]); |
|
} |
|
}, |
|
|
|
getIR: function() { |
|
return { |
|
fnArray: this.fnArray, |
|
argsArray: this.argsArray, |
|
length: this.length |
|
}; |
|
}, |
|
|
|
flush: function(lastChunk) { |
|
if (this.intent !== 'oplist') { |
|
new QueueOptimizer().optimize(this); |
|
} |
|
var transfers = getTransfers(this); |
|
var length = this.length; |
|
this._totalLength += length; |
|
|
|
this.messageHandler.send('RenderPageChunk', { |
|
operatorList: { |
|
fnArray: this.fnArray, |
|
argsArray: this.argsArray, |
|
lastChunk: lastChunk, |
|
length: length |
|
}, |
|
pageIndex: this.pageIndex, |
|
intent: this.intent |
|
}, transfers); |
|
this.dependencies = {}; |
|
this.fnArray.length = 0; |
|
this.argsArray.length = 0; |
|
} |
|
}; |
|
|
|
return OperatorList; |
|
})(); |
|
|
|
var StateManager = (function StateManagerClosure() { |
|
function StateManager(initialState) { |
|
this.state = initialState; |
|
this.stateStack = []; |
|
} |
|
StateManager.prototype = { |
|
save: function () { |
|
var old = this.state; |
|
this.stateStack.push(this.state); |
|
this.state = old.clone(); |
|
}, |
|
restore: function () { |
|
var prev = this.stateStack.pop(); |
|
if (prev) { |
|
this.state = prev; |
|
} |
|
}, |
|
transform: function (args) { |
|
this.state.ctm = Util.transform(this.state.ctm, args); |
|
} |
|
}; |
|
return StateManager; |
|
})(); |
|
|
|
var TextState = (function TextStateClosure() { |
|
function TextState() { |
|
this.ctm = new Float32Array(IDENTITY_MATRIX); |
|
this.fontSize = 0; |
|
this.font = null; |
|
this.fontMatrix = FONT_IDENTITY_MATRIX; |
|
this.textMatrix = IDENTITY_MATRIX.slice(); |
|
this.textLineMatrix = IDENTITY_MATRIX.slice(); |
|
this.charSpacing = 0; |
|
this.wordSpacing = 0; |
|
this.leading = 0; |
|
this.textHScale = 1; |
|
this.textRise = 0; |
|
} |
|
|
|
TextState.prototype = { |
|
setTextMatrix: function TextState_setTextMatrix(a, b, c, d, e, f) { |
|
var m = this.textMatrix; |
|
m[0] = a; m[1] = b; m[2] = c; m[3] = d; m[4] = e; m[5] = f; |
|
}, |
|
setTextLineMatrix: function TextState_setTextMatrix(a, b, c, d, e, f) { |
|
var m = this.textLineMatrix; |
|
m[0] = a; m[1] = b; m[2] = c; m[3] = d; m[4] = e; m[5] = f; |
|
}, |
|
translateTextMatrix: function TextState_translateTextMatrix(x, y) { |
|
var m = this.textMatrix; |
|
m[4] = m[0] * x + m[2] * y + m[4]; |
|
m[5] = m[1] * x + m[3] * y + m[5]; |
|
}, |
|
translateTextLineMatrix: function TextState_translateTextMatrix(x, y) { |
|
var m = this.textLineMatrix; |
|
m[4] = m[0] * x + m[2] * y + m[4]; |
|
m[5] = m[1] * x + m[3] * y + m[5]; |
|
}, |
|
calcRenderMatrix: function TextState_calcRendeMatrix(ctm) { |
|
// 9.4.4 Text Space Details |
|
var tsm = [this.fontSize * this.textHScale, 0, |
|
0, this.fontSize, |
|
0, this.textRise]; |
|
return Util.transform(ctm, Util.transform(this.textMatrix, tsm)); |
|
}, |
|
carriageReturn: function TextState_carriageReturn() { |
|
this.translateTextLineMatrix(0, -this.leading); |
|
this.textMatrix = this.textLineMatrix.slice(); |
|
}, |
|
clone: function TextState_clone() { |
|
var clone = Object.create(this); |
|
clone.textMatrix = this.textMatrix.slice(); |
|
clone.textLineMatrix = this.textLineMatrix.slice(); |
|
clone.fontMatrix = this.fontMatrix.slice(); |
|
return clone; |
|
} |
|
}; |
|
return TextState; |
|
})(); |
|
|
|
var EvalState = (function EvalStateClosure() { |
|
function EvalState() { |
|
this.ctm = new Float32Array(IDENTITY_MATRIX); |
|
this.font = null; |
|
this.textRenderingMode = TextRenderingMode.FILL; |
|
this.fillColorSpace = ColorSpace.singletons.gray; |
|
this.strokeColorSpace = ColorSpace.singletons.gray; |
|
} |
|
EvalState.prototype = { |
|
clone: function CanvasExtraState_clone() { |
|
return Object.create(this); |
|
}, |
|
}; |
|
return EvalState; |
|
})(); |
|
|
|
var EvaluatorPreprocessor = (function EvaluatorPreprocessorClosure() { |
|
// Specifies properties for each command |
|
// |
|
// If variableArgs === true: [0, `numArgs`] expected |
|
// If variableArgs === false: exactly `numArgs` expected |
|
var OP_MAP = { |
|
// Graphic state |
|
w: { id: OPS.setLineWidth, numArgs: 1, variableArgs: false }, |
|
J: { id: OPS.setLineCap, numArgs: 1, variableArgs: false }, |
|
j: { id: OPS.setLineJoin, numArgs: 1, variableArgs: false }, |
|
M: { id: OPS.setMiterLimit, numArgs: 1, variableArgs: false }, |
|
d: { id: OPS.setDash, numArgs: 2, variableArgs: false }, |
|
ri: { id: OPS.setRenderingIntent, numArgs: 1, variableArgs: false }, |
|
i: { id: OPS.setFlatness, numArgs: 1, variableArgs: false }, |
|
gs: { id: OPS.setGState, numArgs: 1, variableArgs: false }, |
|
q: { id: OPS.save, numArgs: 0, variableArgs: false }, |
|
Q: { id: OPS.restore, numArgs: 0, variableArgs: false }, |
|
cm: { id: OPS.transform, numArgs: 6, variableArgs: false }, |
|
|
|
// Path |
|
m: { id: OPS.moveTo, numArgs: 2, variableArgs: false }, |
|
l: { id: OPS.lineTo, numArgs: 2, variableArgs: false }, |
|
c: { id: OPS.curveTo, numArgs: 6, variableArgs: false }, |
|
v: { id: OPS.curveTo2, numArgs: 4, variableArgs: false }, |
|
y: { id: OPS.curveTo3, numArgs: 4, variableArgs: false }, |
|
h: { id: OPS.closePath, numArgs: 0, variableArgs: false }, |
|
re: { id: OPS.rectangle, numArgs: 4, variableArgs: false }, |
|
S: { id: OPS.stroke, numArgs: 0, variableArgs: false }, |
|
s: { id: OPS.closeStroke, numArgs: 0, variableArgs: false }, |
|
f: { id: OPS.fill, numArgs: 0, variableArgs: false }, |
|
F: { id: OPS.fill, numArgs: 0, variableArgs: false }, |
|
'f*': { id: OPS.eoFill, numArgs: 0, variableArgs: false }, |
|
B: { id: OPS.fillStroke, numArgs: 0, variableArgs: false }, |
|
'B*': { id: OPS.eoFillStroke, numArgs: 0, variableArgs: false }, |
|
b: { id: OPS.closeFillStroke, numArgs: 0, variableArgs: false }, |
|
'b*': { id: OPS.closeEOFillStroke, numArgs: 0, variableArgs: false }, |
|
n: { id: OPS.endPath, numArgs: 0, variableArgs: false }, |
|
|
|
// Clipping |
|
W: { id: OPS.clip, numArgs: 0, variableArgs: false }, |
|
'W*': { id: OPS.eoClip, numArgs: 0, variableArgs: false }, |
|
|
|
// Text |
|
BT: { id: OPS.beginText, numArgs: 0, variableArgs: false }, |
|
ET: { id: OPS.endText, numArgs: 0, variableArgs: false }, |
|
Tc: { id: OPS.setCharSpacing, numArgs: 1, variableArgs: false }, |
|
Tw: { id: OPS.setWordSpacing, numArgs: 1, variableArgs: false }, |
|
Tz: { id: OPS.setHScale, numArgs: 1, variableArgs: false }, |
|
TL: { id: OPS.setLeading, numArgs: 1, variableArgs: false }, |
|
Tf: { id: OPS.setFont, numArgs: 2, variableArgs: false }, |
|
Tr: { id: OPS.setTextRenderingMode, numArgs: 1, variableArgs: false }, |
|
Ts: { id: OPS.setTextRise, numArgs: 1, variableArgs: false }, |
|
Td: { id: OPS.moveText, numArgs: 2, variableArgs: false }, |
|
TD: { id: OPS.setLeadingMoveText, numArgs: 2, variableArgs: false }, |
|
Tm: { id: OPS.setTextMatrix, numArgs: 6, variableArgs: false }, |
|
'T*': { id: OPS.nextLine, numArgs: 0, variableArgs: false }, |
|
Tj: { id: OPS.showText, numArgs: 1, variableArgs: false }, |
|
TJ: { id: OPS.showSpacedText, numArgs: 1, variableArgs: false }, |
|
'\'': { id: OPS.nextLineShowText, numArgs: 1, variableArgs: false }, |
|
'"': { id: OPS.nextLineSetSpacingShowText, numArgs: 3, |
|
variableArgs: false }, |
|
|
|
// Type3 fonts |
|
d0: { id: OPS.setCharWidth, numArgs: 2, variableArgs: false }, |
|
d1: { id: OPS.setCharWidthAndBounds, numArgs: 6, variableArgs: false }, |
|
|
|
// Color |
|
CS: { id: OPS.setStrokeColorSpace, numArgs: 1, variableArgs: false }, |
|
cs: { id: OPS.setFillColorSpace, numArgs: 1, variableArgs: false }, |
|
SC: { id: OPS.setStrokeColor, numArgs: 4, variableArgs: true }, |
|
SCN: { id: OPS.setStrokeColorN, numArgs: 33, variableArgs: true }, |
|
sc: { id: OPS.setFillColor, numArgs: 4, variableArgs: true }, |
|
scn: { id: OPS.setFillColorN, numArgs: 33, variableArgs: true }, |
|
G: { id: OPS.setStrokeGray, numArgs: 1, variableArgs: false }, |
|
g: { id: OPS.setFillGray, numArgs: 1, variableArgs: false }, |
|
RG: { id: OPS.setStrokeRGBColor, numArgs: 3, variableArgs: false }, |
|
rg: { id: OPS.setFillRGBColor, numArgs: 3, variableArgs: false }, |
|
K: { id: OPS.setStrokeCMYKColor, numArgs: 4, variableArgs: false }, |
|
k: { id: OPS.setFillCMYKColor, numArgs: 4, variableArgs: false }, |
|
|
|
// Shading |
|
sh: { id: OPS.shadingFill, numArgs: 1, variableArgs: false }, |
|
|
|
// Images |
|
BI: { id: OPS.beginInlineImage, numArgs: 0, variableArgs: false }, |
|
ID: { id: OPS.beginImageData, numArgs: 0, variableArgs: false }, |
|
EI: { id: OPS.endInlineImage, numArgs: 1, variableArgs: false }, |
|
|
|
// XObjects |
|
Do: { id: OPS.paintXObject, numArgs: 1, variableArgs: false }, |
|
MP: { id: OPS.markPoint, numArgs: 1, variableArgs: false }, |
|
DP: { id: OPS.markPointProps, numArgs: 2, variableArgs: false }, |
|
BMC: { id: OPS.beginMarkedContent, numArgs: 1, variableArgs: false }, |
|
BDC: { id: OPS.beginMarkedContentProps, numArgs: 2, |
|
variableArgs: false }, |
|
EMC: { id: OPS.endMarkedContent, numArgs: 0, variableArgs: false }, |
|
|
|
// Compatibility |
|
BX: { id: OPS.beginCompat, numArgs: 0, variableArgs: false }, |
|
EX: { id: OPS.endCompat, numArgs: 0, variableArgs: false }, |
|
|
|
// (reserved partial commands for the lexer) |
|
BM: null, |
|
BD: null, |
|
'true': null, |
|
fa: null, |
|
fal: null, |
|
fals: null, |
|
'false': null, |
|
nu: null, |
|
nul: null, |
|
'null': null |
|
}; |
|
|
|
function EvaluatorPreprocessor(stream, xref, stateManager) { |
|
// TODO(mduan): pass array of knownCommands rather than OP_MAP |
|
// dictionary |
|
this.parser = new Parser(new Lexer(stream, OP_MAP), false, xref); |
|
this.stateManager = stateManager; |
|
this.nonProcessedArgs = []; |
|
} |
|
|
|
EvaluatorPreprocessor.prototype = { |
|
get savedStatesDepth() { |
|
return this.stateManager.stateStack.length; |
|
}, |
|
|
|
// |operation| is an object with two fields: |
|
// |
|
// - |fn| is an out param. |
|
// |
|
// - |args| is an inout param. On entry, it should have one of two values. |
|
// |
|
// - An empty array. This indicates that the caller is providing the |
|
// array in which the args will be stored in. The caller should use |
|
// this value if it can reuse a single array for each call to read(). |
|
// |
|
// - |null|. This indicates that the caller needs this function to create |
|
// the array in which any args are stored in. If there are zero args, |
|
// this function will leave |operation.args| as |null| (thus avoiding |
|
// allocations that would occur if we used an empty array to represent |
|
// zero arguments). Otherwise, it will replace |null| with a new array |
|
// containing the arguments. The caller should use this value if it |
|
// cannot reuse an array for each call to read(). |
|
// |
|
// These two modes are present because this function is very hot and so |
|
// avoiding allocations where possible is worthwhile. |
|
// |
|
read: function EvaluatorPreprocessor_read(operation) { |
|
var args = operation.args; |
|
while (true) { |
|
var obj = this.parser.getObj(); |
|
if (isCmd(obj)) { |
|
var cmd = obj.cmd; |
|
// Check that the command is valid |
|
var opSpec = OP_MAP[cmd]; |
|
if (!opSpec) { |
|
warn('Unknown command "' + cmd + '"'); |
|
continue; |
|
} |
|
|
|
var fn = opSpec.id; |
|
var numArgs = opSpec.numArgs; |
|
var argsLength = args !== null ? args.length : 0; |
|
|
|
if (!opSpec.variableArgs) { |
|
// Postscript commands can be nested, e.g. /F2 /GS2 gs 5.711 Tf |
|
if (argsLength !== numArgs) { |
|
var nonProcessedArgs = this.nonProcessedArgs; |
|
while (argsLength > numArgs) { |
|
nonProcessedArgs.push(args.shift()); |
|
argsLength--; |
|
} |
|
while (argsLength < numArgs && nonProcessedArgs.length !== 0) { |
|
if (!args) { |
|
args = []; |
|
} |
|
args.unshift(nonProcessedArgs.pop()); |
|
argsLength++; |
|
} |
|
} |
|
|
|
if (argsLength < numArgs) { |
|
// If we receive too few args, it's not possible to possible |
|
// to execute the command, so skip the command |
|
info('Command ' + fn + ': because expected ' + |
|
numArgs + ' args, but received ' + argsLength + |
|
' args; skipping'); |
|
args = null; |
|
continue; |
|
} |
|
} else if (argsLength > numArgs) { |
|
info('Command ' + fn + ': expected [0,' + numArgs + |
|
'] args, but received ' + argsLength + ' args'); |
|
} |
|
|
|
// TODO figure out how to type-check vararg functions |
|
this.preprocessCommand(fn, args); |
|
|
|
operation.fn = fn; |
|
operation.args = args; |
|
return true; |
|
} else { |
|
if (isEOF(obj)) { |
|
return false; // no more commands |
|
} |
|
// argument |
|
if (obj !== null) { |
|
if (!args) { |
|
args = []; |
|
} |
|
args.push((obj instanceof Dict ? obj.getAll() : obj)); |
|
assert(args.length <= 33, 'Too many arguments'); |
|
} |
|
} |
|
} |
|
}, |
|
|
|
preprocessCommand: |
|
function EvaluatorPreprocessor_preprocessCommand(fn, args) { |
|
switch (fn | 0) { |
|
case OPS.save: |
|
this.stateManager.save(); |
|
break; |
|
case OPS.restore: |
|
this.stateManager.restore(); |
|
break; |
|
case OPS.transform: |
|
this.stateManager.transform(args); |
|
break; |
|
} |
|
} |
|
}; |
|
return EvaluatorPreprocessor; |
|
})(); |
|
|
|
var QueueOptimizer = (function QueueOptimizerClosure() { |
|
function addState(parentState, pattern, fn) { |
|
var state = parentState; |
|
for (var i = 0, ii = pattern.length - 1; i < ii; i++) { |
|
var item = pattern[i]; |
|
state = (state[item] || (state[item] = [])); |
|
} |
|
state[pattern[pattern.length - 1]] = fn; |
|
} |
|
|
|
function handlePaintSolidColorImageMask(iFirstSave, count, fnArray, |
|
argsArray) { |
|
// Handles special case of mainly LaTeX documents which use image masks to |
|
// draw lines with the current fill style. |
|
// 'count' groups of (save, transform, paintImageMaskXObject, restore)+ |
|
// have been found at iFirstSave. |
|
var iFirstPIMXO = iFirstSave + 2; |
|
for (var i = 0; i < count; i++) { |
|
var arg = argsArray[iFirstPIMXO + 4 * i]; |
|
var imageMask = arg.length === 1 && arg[0]; |
|
if (imageMask && imageMask.width === 1 && imageMask.height === 1 && |
|
(!imageMask.data.length || |
|
(imageMask.data.length === 1 && imageMask.data[0] === 0))) { |
|
fnArray[iFirstPIMXO + 4 * i] = OPS.paintSolidColorImageMask; |
|
continue; |
|
} |
|
break; |
|
} |
|
return count - i; |
|
} |
|
|
|
var InitialState = []; |
|
|
|
// This replaces (save, transform, paintInlineImageXObject, restore)+ |
|
// sequences with one |paintInlineImageXObjectGroup| operation. |
|
addState(InitialState, |
|
[OPS.save, OPS.transform, OPS.paintInlineImageXObject, OPS.restore], |
|
function foundInlineImageGroup(context) { |
|
var MIN_IMAGES_IN_INLINE_IMAGES_BLOCK = 10; |
|
var MAX_IMAGES_IN_INLINE_IMAGES_BLOCK = 200; |
|
var MAX_WIDTH = 1000; |
|
var IMAGE_PADDING = 1; |
|
|
|
var fnArray = context.fnArray, argsArray = context.argsArray; |
|
var curr = context.iCurr; |
|
var iFirstSave = curr - 3; |
|
var iFirstTransform = curr - 2; |
|
var iFirstPIIXO = curr - 1; |
|
|
|
// Look for the quartets. |
|
var i = iFirstSave + 4; |
|
var ii = fnArray.length; |
|
while (i + 3 < ii) { |
|
if (fnArray[i] !== OPS.save || |
|
fnArray[i + 1] !== OPS.transform || |
|
fnArray[i + 2] !== OPS.paintInlineImageXObject || |
|
fnArray[i + 3] !== OPS.restore) { |
|
break; // ops don't match |
|
} |
|
i += 4; |
|
} |
|
|
|
// At this point, i is the index of the first op past the last valid |
|
// quartet. |
|
var count = Math.min((i - iFirstSave) / 4, |
|
MAX_IMAGES_IN_INLINE_IMAGES_BLOCK); |
|
if (count < MIN_IMAGES_IN_INLINE_IMAGES_BLOCK) { |
|
return i; |
|
} |
|
|
|
// assuming that heights of those image is too small (~1 pixel) |
|
// packing as much as possible by lines |
|
var maxX = 0; |
|
var map = [], maxLineHeight = 0; |
|
var currentX = IMAGE_PADDING, currentY = IMAGE_PADDING; |
|
var q; |
|
for (q = 0; q < count; q++) { |
|
var transform = argsArray[iFirstTransform + (q << 2)]; |
|
var img = argsArray[iFirstPIIXO + (q << 2)][0]; |
|
if (currentX + img.width > MAX_WIDTH) { |
|
// starting new line |
|
maxX = Math.max(maxX, currentX); |
|
currentY += maxLineHeight + 2 * IMAGE_PADDING; |
|
currentX = 0; |
|
maxLineHeight = 0; |
|
} |
|
map.push({ |
|
transform: transform, |
|
x: currentX, y: currentY, |
|
w: img.width, h: img.height |
|
}); |
|
currentX += img.width + 2 * IMAGE_PADDING; |
|
maxLineHeight = Math.max(maxLineHeight, img.height); |
|
} |
|
var imgWidth = Math.max(maxX, currentX) + IMAGE_PADDING; |
|
var imgHeight = currentY + maxLineHeight + IMAGE_PADDING; |
|
var imgData = new Uint8Array(imgWidth * imgHeight * 4); |
|
var imgRowSize = imgWidth << 2; |
|
for (q = 0; q < count; q++) { |
|
var data = argsArray[iFirstPIIXO + (q << 2)][0].data; |
|
// Copy image by lines and extends pixels into padding. |
|
var rowSize = map[q].w << 2; |
|
var dataOffset = 0; |
|
var offset = (map[q].x + map[q].y * imgWidth) << 2; |
|
imgData.set(data.subarray(0, rowSize), offset - imgRowSize); |
|
for (var k = 0, kk = map[q].h; k < kk; k++) { |
|
imgData.set(data.subarray(dataOffset, dataOffset + rowSize), offset); |
|
dataOffset += rowSize; |
|
offset += imgRowSize; |
|
} |
|
imgData.set(data.subarray(dataOffset - rowSize, dataOffset), offset); |
|
while (offset >= 0) { |
|
data[offset - 4] = data[offset]; |
|
data[offset - 3] = data[offset + 1]; |
|
data[offset - 2] = data[offset + 2]; |
|
data[offset - 1] = data[offset + 3]; |
|
data[offset + rowSize] = data[offset + rowSize - 4]; |
|
data[offset + rowSize + 1] = data[offset + rowSize - 3]; |
|
data[offset + rowSize + 2] = data[offset + rowSize - 2]; |
|
data[offset + rowSize + 3] = data[offset + rowSize - 1]; |
|
offset -= imgRowSize; |
|
} |
|
} |
|
|
|
// Replace queue items. |
|
fnArray.splice(iFirstSave, count * 4, OPS.paintInlineImageXObjectGroup); |
|
argsArray.splice(iFirstSave, count * 4, |
|
[{ width: imgWidth, height: imgHeight, kind: ImageKind.RGBA_32BPP, |
|
data: imgData }, map]); |
|
|
|
return iFirstSave + 1; |
|
}); |
|
|
|
// This replaces (save, transform, paintImageMaskXObject, restore)+ |
|
// sequences with one |paintImageMaskXObjectGroup| or one |
|
// |paintImageMaskXObjectRepeat| operation. |
|
addState(InitialState, |
|
[OPS.save, OPS.transform, OPS.paintImageMaskXObject, OPS.restore], |
|
function foundImageMaskGroup(context) { |
|
var MIN_IMAGES_IN_MASKS_BLOCK = 10; |
|
var MAX_IMAGES_IN_MASKS_BLOCK = 100; |
|
var MAX_SAME_IMAGES_IN_MASKS_BLOCK = 1000; |
|
|
|
var fnArray = context.fnArray, argsArray = context.argsArray; |
|
var curr = context.iCurr; |
|
var iFirstSave = curr - 3; |
|
var iFirstTransform = curr - 2; |
|
var iFirstPIMXO = curr - 1; |
|
|
|
// Look for the quartets. |
|
var i = iFirstSave + 4; |
|
var ii = fnArray.length; |
|
while (i + 3 < ii) { |
|
if (fnArray[i] !== OPS.save || |
|
fnArray[i + 1] !== OPS.transform || |
|
fnArray[i + 2] !== OPS.paintImageMaskXObject || |
|
fnArray[i + 3] !== OPS.restore) { |
|
break; // ops don't match |
|
} |
|
i += 4; |
|
} |
|
|
|
// At this point, i is the index of the first op past the last valid |
|
// quartet. |
|
var count = (i - iFirstSave) / 4; |
|
count = handlePaintSolidColorImageMask(iFirstSave, count, fnArray, |
|
argsArray); |
|
if (count < MIN_IMAGES_IN_MASKS_BLOCK) { |
|
return i; |
|
} |
|
|
|
var q; |
|
var isSameImage = false; |
|
var iTransform, transformArgs; |
|
var firstPIMXOArg0 = argsArray[iFirstPIMXO][0]; |
|
if (argsArray[iFirstTransform][1] === 0 && |
|
argsArray[iFirstTransform][2] === 0) { |
|
isSameImage = true; |
|
var firstTransformArg0 = argsArray[iFirstTransform][0]; |
|
var firstTransformArg3 = argsArray[iFirstTransform][3]; |
|
iTransform = iFirstTransform + 4; |
|
var iPIMXO = iFirstPIMXO + 4; |
|
for (q = 1; q < count; q++, iTransform += 4, iPIMXO += 4) { |
|
transformArgs = argsArray[iTransform]; |
|
if (argsArray[iPIMXO][0] !== firstPIMXOArg0 || |
|
transformArgs[0] !== firstTransformArg0 || |
|
transformArgs[1] !== 0 || |
|
transformArgs[2] !== 0 || |
|
transformArgs[3] !== firstTransformArg3) { |
|
if (q < MIN_IMAGES_IN_MASKS_BLOCK) { |
|
isSameImage = false; |
|
} else { |
|
count = q; |
|
} |
|
break; // different image or transform |
|
} |
|
} |
|
} |
|
|
|
if (isSameImage) { |
|
count = Math.min(count, MAX_SAME_IMAGES_IN_MASKS_BLOCK); |
|
var positions = new Float32Array(count * 2); |
|
iTransform = iFirstTransform; |
|
for (q = 0; q < count; q++, iTransform += 4) { |
|
transformArgs = argsArray[iTransform]; |
|
positions[(q << 1)] = transformArgs[4]; |
|
positions[(q << 1) + 1] = transformArgs[5]; |
|
} |
|
|
|
// Replace queue items. |
|
fnArray.splice(iFirstSave, count * 4, OPS.paintImageMaskXObjectRepeat); |
|
argsArray.splice(iFirstSave, count * 4, |
|
[firstPIMXOArg0, firstTransformArg0, firstTransformArg3, positions]); |
|
} else { |
|
count = Math.min(count, MAX_IMAGES_IN_MASKS_BLOCK); |
|
var images = []; |
|
for (q = 0; q < count; q++) { |
|
transformArgs = argsArray[iFirstTransform + (q << 2)]; |
|
var maskParams = argsArray[iFirstPIMXO + (q << 2)][0]; |
|
images.push({ data: maskParams.data, width: maskParams.width, |
|
height: maskParams.height, |
|
transform: transformArgs }); |
|
} |
|
|
|
// Replace queue items. |
|
fnArray.splice(iFirstSave, count * 4, OPS.paintImageMaskXObjectGroup); |
|
argsArray.splice(iFirstSave, count * 4, [images]); |
|
} |
|
|
|
return iFirstSave + 1; |
|
}); |
|
|
|
// This replaces (save, transform, paintImageXObject, restore)+ sequences |
|
// with one paintImageXObjectRepeat operation, if the |transform| and |
|
// |paintImageXObjectRepeat| ops are appropriate. |
|
addState(InitialState, |
|
[OPS.save, OPS.transform, OPS.paintImageXObject, OPS.restore], |
|
function (context) { |
|
var MIN_IMAGES_IN_BLOCK = 3; |
|
var MAX_IMAGES_IN_BLOCK = 1000; |
|
|
|
var fnArray = context.fnArray, argsArray = context.argsArray; |
|
var curr = context.iCurr; |
|
var iFirstSave = curr - 3; |
|
var iFirstTransform = curr - 2; |
|
var iFirstPIXO = curr - 1; |
|
var iFirstRestore = curr; |
|
|
|
if (argsArray[iFirstTransform][1] !== 0 || |
|
argsArray[iFirstTransform][2] !== 0) { |
|
return iFirstRestore + 1; // transform has the wrong form |
|
} |
|
|
|
// Look for the quartets. |
|
var firstPIXOArg0 = argsArray[iFirstPIXO][0]; |
|
var firstTransformArg0 = argsArray[iFirstTransform][0]; |
|
var firstTransformArg3 = argsArray[iFirstTransform][3]; |
|
var i = iFirstSave + 4; |
|
var ii = fnArray.length; |
|
while (i + 3 < ii) { |
|
if (fnArray[i] !== OPS.save || |
|
fnArray[i + 1] !== OPS.transform || |
|
fnArray[i + 2] !== OPS.paintImageXObject || |
|
fnArray[i + 3] !== OPS.restore) { |
|
break; // ops don't match |
|
} |
|
if (argsArray[i + 1][0] !== firstTransformArg0 || |
|
argsArray[i + 1][1] !== 0 || |
|
argsArray[i + 1][2] !== 0 || |
|
argsArray[i + 1][3] !== firstTransformArg3) { |
|
break; // transforms don't match |
|
} |
|
if (argsArray[i + 2][0] !== firstPIXOArg0) { |
|
break; // images don't match |
|
} |
|
i += 4; |
|
} |
|
|
|
// At this point, i is the index of the first op past the last valid |
|
// quartet. |
|
var count = Math.min((i - iFirstSave) / 4, MAX_IMAGES_IN_BLOCK); |
|
if (count < MIN_IMAGES_IN_BLOCK) { |
|
return i; |
|
} |
|
|
|
// Extract the (x,y) positions from all of the matching transforms. |
|
var positions = new Float32Array(count * 2); |
|
var iTransform = iFirstTransform; |
|
for (var q = 0; q < count; q++, iTransform += 4) { |
|
var transformArgs = argsArray[iTransform]; |
|
positions[(q << 1)] = transformArgs[4]; |
|
positions[(q << 1) + 1] = transformArgs[5]; |
|
} |
|
|
|
// Replace queue items. |
|
var args = [firstPIXOArg0, firstTransformArg0, firstTransformArg3, |
|
positions]; |
|
fnArray.splice(iFirstSave, count * 4, OPS.paintImageXObjectRepeat); |
|
argsArray.splice(iFirstSave, count * 4, args); |
|
|
|
return iFirstSave + 1; |
|
}); |
|
|
|
// This replaces (beginText, setFont, setTextMatrix, showText, endText)+ |
|
// sequences with (beginText, setFont, (setTextMatrix, showText)+, endText)+ |
|
// sequences, if the font for each one is the same. |
|
addState(InitialState, |
|
[OPS.beginText, OPS.setFont, OPS.setTextMatrix, OPS.showText, OPS.endText], |
|
function (context) { |
|
var MIN_CHARS_IN_BLOCK = 3; |
|
var MAX_CHARS_IN_BLOCK = 1000; |
|
|
|
var fnArray = context.fnArray, argsArray = context.argsArray; |
|
var curr = context.iCurr; |
|
var iFirstBeginText = curr - 4; |
|
var iFirstSetFont = curr - 3; |
|
var iFirstSetTextMatrix = curr - 2; |
|
var iFirstShowText = curr - 1; |
|
var iFirstEndText = curr; |
|
|
|
// Look for the quintets. |
|
var firstSetFontArg0 = argsArray[iFirstSetFont][0]; |
|
var firstSetFontArg1 = argsArray[iFirstSetFont][1]; |
|
var i = iFirstBeginText + 5; |
|
var ii = fnArray.length; |
|
while (i + 4 < ii) { |
|
if (fnArray[i] !== OPS.beginText || |
|
fnArray[i + 1] !== OPS.setFont || |
|
fnArray[i + 2] !== OPS.setTextMatrix || |
|
fnArray[i + 3] !== OPS.showText || |
|
fnArray[i + 4] !== OPS.endText) { |
|
break; // ops don't match |
|
} |
|
if (argsArray[i + 1][0] !== firstSetFontArg0 || |
|
argsArray[i + 1][1] !== firstSetFontArg1) { |
|
break; // fonts don't match |
|
} |
|
i += 5; |
|
} |
|
|
|
// At this point, i is the index of the first op past the last valid |
|
// quintet. |
|
var count = Math.min(((i - iFirstBeginText) / 5), MAX_CHARS_IN_BLOCK); |
|
if (count < MIN_CHARS_IN_BLOCK) { |
|
return i; |
|
} |
|
|
|
// If the preceding quintet is (<something>, setFont, setTextMatrix, |
|
// showText, endText), include that as well. (E.g. <something> might be |
|
// |dependency|.) |
|
var iFirst = iFirstBeginText; |
|
if (iFirstBeginText >= 4 && |
|
fnArray[iFirstBeginText - 4] === fnArray[iFirstSetFont] && |
|
fnArray[iFirstBeginText - 3] === fnArray[iFirstSetTextMatrix] && |
|
fnArray[iFirstBeginText - 2] === fnArray[iFirstShowText] && |
|
fnArray[iFirstBeginText - 1] === fnArray[iFirstEndText] && |
|
argsArray[iFirstBeginText - 4][0] === firstSetFontArg0 && |
|
argsArray[iFirstBeginText - 4][1] === firstSetFontArg1) { |
|
count++; |
|
iFirst -= 5; |
|
} |
|
|
|
// Remove (endText, beginText, setFont) trios. |
|
var iEndText = iFirst + 4; |
|
for (var q = 1; q < count; q++) { |
|
fnArray.splice(iEndText, 3); |
|
argsArray.splice(iEndText, 3); |
|
iEndText += 2; |
|
} |
|
|
|
return iEndText + 1; |
|
}); |
|
|
|
function QueueOptimizer() {} |
|
|
|
QueueOptimizer.prototype = { |
|
optimize: function QueueOptimizer_optimize(queue) { |
|
var fnArray = queue.fnArray, argsArray = queue.argsArray; |
|
var context = { |
|
iCurr: 0, |
|
fnArray: fnArray, |
|
argsArray: argsArray |
|
}; |
|
var state; |
|
var i = 0, ii = fnArray.length; |
|
while (i < ii) { |
|
state = (state || InitialState)[fnArray[i]]; |
|
if (typeof state === 'function') { // we found some handler |
|
context.iCurr = i; |
|
// state() returns the index of the first non-matching op (if we |
|
// didn't match) or the first op past the modified ops (if we did |
|
// match and replace). |
|
i = state(context); |
|
state = undefined; // reset the state machine |
|
ii = context.fnArray.length; |
|
} else { |
|
i++; |
|
} |
|
} |
|
} |
|
}; |
|
return QueueOptimizer; |
|
})(); |
|
|
|
exports.OperatorList = OperatorList; |
|
exports.PartialEvaluator = PartialEvaluator; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreAnnotation = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreStream, root.pdfjsCoreColorSpace, |
|
root.pdfjsCoreObj, root.pdfjsCoreEvaluator); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreStream, |
|
coreColorSpace, coreObj, coreEvaluator) { |
|
|
|
var AnnotationBorderStyleType = sharedUtil.AnnotationBorderStyleType; |
|
var AnnotationFlag = sharedUtil.AnnotationFlag; |
|
var AnnotationType = sharedUtil.AnnotationType; |
|
var OPS = sharedUtil.OPS; |
|
var Util = sharedUtil.Util; |
|
var isArray = sharedUtil.isArray; |
|
var isInt = sharedUtil.isInt; |
|
var isValidUrl = sharedUtil.isValidUrl; |
|
var stringToBytes = sharedUtil.stringToBytes; |
|
var stringToPDFString = sharedUtil.stringToPDFString; |
|
var stringToUTF8String = sharedUtil.stringToUTF8String; |
|
var warn = sharedUtil.warn; |
|
var Dict = corePrimitives.Dict; |
|
var isDict = corePrimitives.isDict; |
|
var isName = corePrimitives.isName; |
|
var Stream = coreStream.Stream; |
|
var ColorSpace = coreColorSpace.ColorSpace; |
|
var ObjectLoader = coreObj.ObjectLoader; |
|
var OperatorList = coreEvaluator.OperatorList; |
|
|
|
/** |
|
* @class |
|
* @alias AnnotationFactory |
|
*/ |
|
function AnnotationFactory() {} |
|
AnnotationFactory.prototype = /** @lends AnnotationFactory.prototype */ { |
|
/** |
|
* @param {XRef} xref |
|
* @param {Object} ref |
|
* @returns {Annotation} |
|
*/ |
|
create: function AnnotationFactory_create(xref, ref) { |
|
var dict = xref.fetchIfRef(ref); |
|
if (!isDict(dict)) { |
|
return; |
|
} |
|
|
|
// Determine the annotation's subtype. |
|
var subtype = dict.get('Subtype'); |
|
subtype = isName(subtype) ? subtype.name : ''; |
|
|
|
// Return the right annotation object based on the subtype and field type. |
|
var parameters = { |
|
dict: dict, |
|
ref: ref |
|
}; |
|
|
|
switch (subtype) { |
|
case 'Link': |
|
return new LinkAnnotation(parameters); |
|
|
|
case 'Text': |
|
return new TextAnnotation(parameters); |
|
|
|
case 'Widget': |
|
var fieldType = Util.getInheritableProperty(dict, 'FT'); |
|
if (isName(fieldType) && fieldType.name === 'Tx') { |
|
return new TextWidgetAnnotation(parameters); |
|
} |
|
return new WidgetAnnotation(parameters); |
|
|
|
case 'Popup': |
|
return new PopupAnnotation(parameters); |
|
|
|
case 'Highlight': |
|
return new HighlightAnnotation(parameters); |
|
|
|
case 'Underline': |
|
return new UnderlineAnnotation(parameters); |
|
|
|
case 'Squiggly': |
|
return new SquigglyAnnotation(parameters); |
|
|
|
case 'StrikeOut': |
|
return new StrikeOutAnnotation(parameters); |
|
|
|
default: |
|
warn('Unimplemented annotation type "' + subtype + '", ' + |
|
'falling back to base annotation'); |
|
return new Annotation(parameters); |
|
} |
|
} |
|
}; |
|
|
|
var Annotation = (function AnnotationClosure() { |
|
// 12.5.5: Algorithm: Appearance streams |
|
function getTransformMatrix(rect, bbox, matrix) { |
|
var bounds = Util.getAxialAlignedBoundingBox(bbox, matrix); |
|
var minX = bounds[0]; |
|
var minY = bounds[1]; |
|
var maxX = bounds[2]; |
|
var maxY = bounds[3]; |
|
|
|
if (minX === maxX || minY === maxY) { |
|
// From real-life file, bbox was [0, 0, 0, 0]. In this case, |
|
// just apply the transform for rect |
|
return [1, 0, 0, 1, rect[0], rect[1]]; |
|
} |
|
|
|
var xRatio = (rect[2] - rect[0]) / (maxX - minX); |
|
var yRatio = (rect[3] - rect[1]) / (maxY - minY); |
|
return [ |
|
xRatio, |
|
0, |
|
0, |
|
yRatio, |
|
rect[0] - minX * xRatio, |
|
rect[1] - minY * yRatio |
|
]; |
|
} |
|
|
|
function getDefaultAppearance(dict) { |
|
var appearanceState = dict.get('AP'); |
|
if (!isDict(appearanceState)) { |
|
return; |
|
} |
|
|
|
var appearance; |
|
var appearances = appearanceState.get('N'); |
|
if (isDict(appearances)) { |
|
var as = dict.get('AS'); |
|
if (as && appearances.has(as.name)) { |
|
appearance = appearances.get(as.name); |
|
} |
|
} else { |
|
appearance = appearances; |
|
} |
|
return appearance; |
|
} |
|
|
|
function Annotation(params) { |
|
var dict = params.dict; |
|
|
|
this.setFlags(dict.get('F')); |
|
this.setRectangle(dict.get('Rect')); |
|
this.setColor(dict.get('C')); |
|
this.setBorderStyle(dict); |
|
this.appearance = getDefaultAppearance(dict); |
|
|
|
// Expose public properties using a data object. |
|
this.data = {}; |
|
this.data.id = params.ref.toString(); |
|
this.data.subtype = dict.get('Subtype').name; |
|
this.data.annotationFlags = this.flags; |
|
this.data.rect = this.rectangle; |
|
this.data.color = this.color; |
|
this.data.borderStyle = this.borderStyle; |
|
this.data.hasAppearance = !!this.appearance; |
|
} |
|
|
|
Annotation.prototype = { |
|
/** |
|
* @return {boolean} |
|
*/ |
|
get viewable() { |
|
if (this.flags) { |
|
return !this.hasFlag(AnnotationFlag.INVISIBLE) && |
|
!this.hasFlag(AnnotationFlag.HIDDEN) && |
|
!this.hasFlag(AnnotationFlag.NOVIEW); |
|
} |
|
return true; |
|
}, |
|
|
|
/** |
|
* @return {boolean} |
|
*/ |
|
get printable() { |
|
if (this.flags) { |
|
return this.hasFlag(AnnotationFlag.PRINT) && |
|
!this.hasFlag(AnnotationFlag.INVISIBLE) && |
|
!this.hasFlag(AnnotationFlag.HIDDEN); |
|
} |
|
return false; |
|
}, |
|
|
|
/** |
|
* Set the flags. |
|
* |
|
* @public |
|
* @memberof Annotation |
|
* @param {number} flags - Unsigned 32-bit integer specifying annotation |
|
* characteristics |
|
* @see {@link shared/util.js} |
|
*/ |
|
setFlags: function Annotation_setFlags(flags) { |
|
if (isInt(flags)) { |
|
this.flags = flags; |
|
} else { |
|
this.flags = 0; |
|
} |
|
}, |
|
|
|
/** |
|
* Check if a provided flag is set. |
|
* |
|
* @public |
|
* @memberof Annotation |
|
* @param {number} flag - Hexadecimal representation for an annotation |
|
* characteristic |
|
* @return {boolean} |
|
* @see {@link shared/util.js} |
|
*/ |
|
hasFlag: function Annotation_hasFlag(flag) { |
|
if (this.flags) { |
|
return (this.flags & flag) > 0; |
|
} |
|
return false; |
|
}, |
|
|
|
/** |
|
* Set the rectangle. |
|
* |
|
* @public |
|
* @memberof Annotation |
|
* @param {Array} rectangle - The rectangle array with exactly four entries |
|
*/ |
|
setRectangle: function Annotation_setRectangle(rectangle) { |
|
if (isArray(rectangle) && rectangle.length === 4) { |
|
this.rectangle = Util.normalizeRect(rectangle); |
|
} else { |
|
this.rectangle = [0, 0, 0, 0]; |
|
} |
|
}, |
|
|
|
/** |
|
* Set the color and take care of color space conversion. |
|
* |
|
* @public |
|
* @memberof Annotation |
|
* @param {Array} color - The color array containing either 0 |
|
* (transparent), 1 (grayscale), 3 (RGB) or |
|
* 4 (CMYK) elements |
|
*/ |
|
setColor: function Annotation_setColor(color) { |
|
var rgbColor = new Uint8Array(3); // Black in RGB color space (default) |
|
if (!isArray(color)) { |
|
this.color = rgbColor; |
|
return; |
|
} |
|
|
|
switch (color.length) { |
|
case 0: // Transparent, which we indicate with a null value |
|
this.color = null; |
|
break; |
|
|
|
case 1: // Convert grayscale to RGB |
|
ColorSpace.singletons.gray.getRgbItem(color, 0, rgbColor, 0); |
|
this.color = rgbColor; |
|
break; |
|
|
|
case 3: // Convert RGB percentages to RGB |
|
ColorSpace.singletons.rgb.getRgbItem(color, 0, rgbColor, 0); |
|
this.color = rgbColor; |
|
break; |
|
|
|
case 4: // Convert CMYK to RGB |
|
ColorSpace.singletons.cmyk.getRgbItem(color, 0, rgbColor, 0); |
|
this.color = rgbColor; |
|
break; |
|
|
|
default: |
|
this.color = rgbColor; |
|
break; |
|
} |
|
}, |
|
|
|
/** |
|
* Set the border style (as AnnotationBorderStyle object). |
|
* |
|
* @public |
|
* @memberof Annotation |
|
* @param {Dict} borderStyle - The border style dictionary |
|
*/ |
|
setBorderStyle: function Annotation_setBorderStyle(borderStyle) { |
|
this.borderStyle = new AnnotationBorderStyle(); |
|
if (!isDict(borderStyle)) { |
|
return; |
|
} |
|
if (borderStyle.has('BS')) { |
|
var dict = borderStyle.get('BS'); |
|
var dictType; |
|
|
|
if (!dict.has('Type') || (isName(dictType = dict.get('Type')) && |
|
dictType.name === 'Border')) { |
|
this.borderStyle.setWidth(dict.get('W')); |
|
this.borderStyle.setStyle(dict.get('S')); |
|
this.borderStyle.setDashArray(dict.get('D')); |
|
} |
|
} else if (borderStyle.has('Border')) { |
|
var array = borderStyle.get('Border'); |
|
if (isArray(array) && array.length >= 3) { |
|
this.borderStyle.setHorizontalCornerRadius(array[0]); |
|
this.borderStyle.setVerticalCornerRadius(array[1]); |
|
this.borderStyle.setWidth(array[2]); |
|
|
|
if (array.length === 4) { // Dash array available |
|
this.borderStyle.setDashArray(array[3]); |
|
} |
|
} |
|
} else { |
|
// There are no border entries in the dictionary. According to the |
|
// specification, we should draw a solid border of width 1 in that |
|
// case, but Adobe Reader did not implement that part of the |
|
// specification and instead draws no border at all, so we do the same. |
|
// See also https://github.com/mozilla/pdf.js/issues/6179. |
|
this.borderStyle.setWidth(0); |
|
} |
|
}, |
|
|
|
loadResources: function Annotation_loadResources(keys) { |
|
return new Promise(function (resolve, reject) { |
|
this.appearance.dict.getAsync('Resources').then(function (resources) { |
|
if (!resources) { |
|
resolve(); |
|
return; |
|
} |
|
var objectLoader = new ObjectLoader(resources.map, |
|
keys, |
|
resources.xref); |
|
objectLoader.load().then(function() { |
|
resolve(resources); |
|
}, reject); |
|
}, reject); |
|
}.bind(this)); |
|
}, |
|
|
|
getOperatorList: function Annotation_getOperatorList(evaluator, task) { |
|
if (!this.appearance) { |
|
return Promise.resolve(new OperatorList()); |
|
} |
|
|
|
var data = this.data; |
|
var appearanceDict = this.appearance.dict; |
|
var resourcesPromise = this.loadResources([ |
|
'ExtGState', |
|
'ColorSpace', |
|
'Pattern', |
|
'Shading', |
|
'XObject', |
|
'Font' |
|
// ProcSet |
|
// Properties |
|
]); |
|
var bbox = appearanceDict.get('BBox') || [0, 0, 1, 1]; |
|
var matrix = appearanceDict.get('Matrix') || [1, 0, 0, 1, 0 ,0]; |
|
var transform = getTransformMatrix(data.rect, bbox, matrix); |
|
var self = this; |
|
|
|
return resourcesPromise.then(function(resources) { |
|
var opList = new OperatorList(); |
|
opList.addOp(OPS.beginAnnotation, [data.rect, transform, matrix]); |
|
return evaluator.getOperatorList(self.appearance, task, |
|
resources, opList). |
|
then(function () { |
|
opList.addOp(OPS.endAnnotation, []); |
|
self.appearance.reset(); |
|
return opList; |
|
}); |
|
}); |
|
} |
|
}; |
|
|
|
Annotation.appendToOperatorList = function Annotation_appendToOperatorList( |
|
annotations, opList, partialEvaluator, task, intent) { |
|
var annotationPromises = []; |
|
for (var i = 0, n = annotations.length; i < n; ++i) { |
|
if ((intent === 'display' && annotations[i].viewable) || |
|
(intent === 'print' && annotations[i].printable)) { |
|
annotationPromises.push( |
|
annotations[i].getOperatorList(partialEvaluator, task)); |
|
} |
|
} |
|
return Promise.all(annotationPromises).then(function(operatorLists) { |
|
opList.addOp(OPS.beginAnnotations, []); |
|
for (var i = 0, n = operatorLists.length; i < n; ++i) { |
|
opList.addOpList(operatorLists[i]); |
|
} |
|
opList.addOp(OPS.endAnnotations, []); |
|
}); |
|
}; |
|
|
|
return Annotation; |
|
})(); |
|
|
|
/** |
|
* Contains all data regarding an annotation's border style. |
|
* |
|
* @class |
|
*/ |
|
var AnnotationBorderStyle = (function AnnotationBorderStyleClosure() { |
|
/** |
|
* @constructor |
|
* @private |
|
*/ |
|
function AnnotationBorderStyle() { |
|
this.width = 1; |
|
this.style = AnnotationBorderStyleType.SOLID; |
|
this.dashArray = [3]; |
|
this.horizontalCornerRadius = 0; |
|
this.verticalCornerRadius = 0; |
|
} |
|
|
|
AnnotationBorderStyle.prototype = { |
|
/** |
|
* Set the width. |
|
* |
|
* @public |
|
* @memberof AnnotationBorderStyle |
|
* @param {integer} width - The width |
|
*/ |
|
setWidth: function AnnotationBorderStyle_setWidth(width) { |
|
if (width === (width | 0)) { |
|
this.width = width; |
|
} |
|
}, |
|
|
|
/** |
|
* Set the style. |
|
* |
|
* @public |
|
* @memberof AnnotationBorderStyle |
|
* @param {Object} style - The style object |
|
* @see {@link shared/util.js} |
|
*/ |
|
setStyle: function AnnotationBorderStyle_setStyle(style) { |
|
if (!style) { |
|
return; |
|
} |
|
switch (style.name) { |
|
case 'S': |
|
this.style = AnnotationBorderStyleType.SOLID; |
|
break; |
|
|
|
case 'D': |
|
this.style = AnnotationBorderStyleType.DASHED; |
|
break; |
|
|
|
case 'B': |
|
this.style = AnnotationBorderStyleType.BEVELED; |
|
break; |
|
|
|
case 'I': |
|
this.style = AnnotationBorderStyleType.INSET; |
|
break; |
|
|
|
case 'U': |
|
this.style = AnnotationBorderStyleType.UNDERLINE; |
|
break; |
|
|
|
default: |
|
break; |
|
} |
|
}, |
|
|
|
/** |
|
* Set the dash array. |
|
* |
|
* @public |
|
* @memberof AnnotationBorderStyle |
|
* @param {Array} dashArray - The dash array with at least one element |
|
*/ |
|
setDashArray: function AnnotationBorderStyle_setDashArray(dashArray) { |
|
// We validate the dash array, but we do not use it because CSS does not |
|
// allow us to change spacing of dashes. For more information, visit |
|
// http://www.w3.org/TR/css3-background/#the-border-style. |
|
if (isArray(dashArray) && dashArray.length > 0) { |
|
// According to the PDF specification: the elements in a dashArray |
|
// shall be numbers that are nonnegative and not all equal to zero. |
|
var isValid = true; |
|
var allZeros = true; |
|
for (var i = 0, len = dashArray.length; i < len; i++) { |
|
var element = dashArray[i]; |
|
var validNumber = (+element >= 0); |
|
if (!validNumber) { |
|
isValid = false; |
|
break; |
|
} else if (element > 0) { |
|
allZeros = false; |
|
} |
|
} |
|
if (isValid && !allZeros) { |
|
this.dashArray = dashArray; |
|
} else { |
|
this.width = 0; // Adobe behavior when the array is invalid. |
|
} |
|
} else if (dashArray) { |
|
this.width = 0; // Adobe behavior when the array is invalid. |
|
} |
|
}, |
|
|
|
/** |
|
* Set the horizontal corner radius (from a Border dictionary). |
|
* |
|
* @public |
|
* @memberof AnnotationBorderStyle |
|
* @param {integer} radius - The horizontal corner radius |
|
*/ |
|
setHorizontalCornerRadius: |
|
function AnnotationBorderStyle_setHorizontalCornerRadius(radius) { |
|
if (radius === (radius | 0)) { |
|
this.horizontalCornerRadius = radius; |
|
} |
|
}, |
|
|
|
/** |
|
* Set the vertical corner radius (from a Border dictionary). |
|
* |
|
* @public |
|
* @memberof AnnotationBorderStyle |
|
* @param {integer} radius - The vertical corner radius |
|
*/ |
|
setVerticalCornerRadius: |
|
function AnnotationBorderStyle_setVerticalCornerRadius(radius) { |
|
if (radius === (radius | 0)) { |
|
this.verticalCornerRadius = radius; |
|
} |
|
} |
|
}; |
|
|
|
return AnnotationBorderStyle; |
|
})(); |
|
|
|
var WidgetAnnotation = (function WidgetAnnotationClosure() { |
|
function WidgetAnnotation(params) { |
|
Annotation.call(this, params); |
|
|
|
var dict = params.dict; |
|
var data = this.data; |
|
|
|
data.annotationType = AnnotationType.WIDGET; |
|
data.fieldValue = stringToPDFString( |
|
Util.getInheritableProperty(dict, 'V') || ''); |
|
data.alternativeText = stringToPDFString(dict.get('TU') || ''); |
|
data.defaultAppearance = Util.getInheritableProperty(dict, 'DA') || ''; |
|
var fieldType = Util.getInheritableProperty(dict, 'FT'); |
|
data.fieldType = isName(fieldType) ? fieldType.name : ''; |
|
data.fieldFlags = Util.getInheritableProperty(dict, 'Ff') || 0; |
|
this.fieldResources = Util.getInheritableProperty(dict, 'DR') || Dict.empty; |
|
|
|
// Hide unsupported Widget signatures. |
|
if (data.fieldType === 'Sig') { |
|
warn('unimplemented annotation type: Widget signature'); |
|
this.setFlags(AnnotationFlag.HIDDEN); |
|
} |
|
|
|
// Building the full field name by collecting the field and |
|
// its ancestors 'T' data and joining them using '.'. |
|
var fieldName = []; |
|
var namedItem = dict; |
|
var ref = params.ref; |
|
while (namedItem) { |
|
var parent = namedItem.get('Parent'); |
|
var parentRef = namedItem.getRaw('Parent'); |
|
var name = namedItem.get('T'); |
|
if (name) { |
|
fieldName.unshift(stringToPDFString(name)); |
|
} else if (parent && ref) { |
|
// The field name is absent, that means more than one field |
|
// with the same name may exist. Replacing the empty name |
|
// with the '`' plus index in the parent's 'Kids' array. |
|
// This is not in the PDF spec but necessary to id the |
|
// the input controls. |
|
var kids = parent.get('Kids'); |
|
var j, jj; |
|
for (j = 0, jj = kids.length; j < jj; j++) { |
|
var kidRef = kids[j]; |
|
if (kidRef.num === ref.num && kidRef.gen === ref.gen) { |
|
break; |
|
} |
|
} |
|
fieldName.unshift('`' + j); |
|
} |
|
namedItem = parent; |
|
ref = parentRef; |
|
} |
|
data.fullName = fieldName.join('.'); |
|
} |
|
|
|
Util.inherit(WidgetAnnotation, Annotation, {}); |
|
|
|
return WidgetAnnotation; |
|
})(); |
|
|
|
var TextWidgetAnnotation = (function TextWidgetAnnotationClosure() { |
|
function TextWidgetAnnotation(params) { |
|
WidgetAnnotation.call(this, params); |
|
|
|
this.data.textAlignment = Util.getInheritableProperty(params.dict, 'Q'); |
|
this.data.hasHtml = !this.data.hasAppearance && !!this.data.fieldValue; |
|
} |
|
|
|
Util.inherit(TextWidgetAnnotation, WidgetAnnotation, { |
|
getOperatorList: function TextWidgetAnnotation_getOperatorList(evaluator, |
|
task) { |
|
if (this.appearance) { |
|
return Annotation.prototype.getOperatorList.call(this, evaluator, task); |
|
} |
|
|
|
var opList = new OperatorList(); |
|
var data = this.data; |
|
|
|
// Even if there is an appearance stream, ignore it. This is the |
|
// behaviour used by Adobe Reader. |
|
if (!data.defaultAppearance) { |
|
return Promise.resolve(opList); |
|
} |
|
|
|
var stream = new Stream(stringToBytes(data.defaultAppearance)); |
|
return evaluator.getOperatorList(stream, task, |
|
this.fieldResources, opList). |
|
then(function () { |
|
return opList; |
|
}); |
|
} |
|
}); |
|
|
|
return TextWidgetAnnotation; |
|
})(); |
|
|
|
var TextAnnotation = (function TextAnnotationClosure() { |
|
var DEFAULT_ICON_SIZE = 22; // px |
|
|
|
function TextAnnotation(parameters) { |
|
Annotation.call(this, parameters); |
|
|
|
this.data.annotationType = AnnotationType.TEXT; |
|
this.data.hasHtml = true; |
|
|
|
var dict = parameters.dict; |
|
if (this.data.hasAppearance) { |
|
this.data.name = 'NoIcon'; |
|
} else { |
|
this.data.rect[1] = this.data.rect[3] - DEFAULT_ICON_SIZE; |
|
this.data.rect[2] = this.data.rect[0] + DEFAULT_ICON_SIZE; |
|
this.data.name = dict.has('Name') ? dict.get('Name').name : 'Note'; |
|
} |
|
|
|
if (!dict.has('C')) { |
|
// Fall back to the default background color. |
|
this.data.color = null; |
|
} |
|
|
|
this.data.hasPopup = dict.has('Popup'); |
|
if (!this.data.hasPopup) { |
|
// There is no associated Popup annotation, so the Text annotation |
|
// must create its own popup. |
|
this.data.title = stringToPDFString(dict.get('T') || ''); |
|
this.data.contents = stringToPDFString(dict.get('Contents') || ''); |
|
this.data.hasHtml = (this.data.title || this.data.contents); |
|
} |
|
} |
|
|
|
Util.inherit(TextAnnotation, Annotation, {}); |
|
|
|
return TextAnnotation; |
|
})(); |
|
|
|
var LinkAnnotation = (function LinkAnnotationClosure() { |
|
function LinkAnnotation(params) { |
|
Annotation.call(this, params); |
|
|
|
var dict = params.dict; |
|
var data = this.data; |
|
data.annotationType = AnnotationType.LINK; |
|
data.hasHtml = true; |
|
|
|
var action = dict.get('A'); |
|
if (action && isDict(action)) { |
|
var linkType = action.get('S').name; |
|
if (linkType === 'URI') { |
|
var url = action.get('URI'); |
|
if (isName(url)) { |
|
// Some bad PDFs do not put parentheses around relative URLs. |
|
url = '/' + url.name; |
|
} else if (url) { |
|
url = addDefaultProtocolToUrl(url); |
|
} |
|
// TODO: pdf spec mentions urls can be relative to a Base |
|
// entry in the dictionary. |
|
if (!isValidUrl(url, false)) { |
|
url = ''; |
|
} |
|
// According to ISO 32000-1:2008, section 12.6.4.7, |
|
// URI should to be encoded in 7-bit ASCII. |
|
// Some bad PDFs may have URIs in UTF-8 encoding, see Bugzilla 1122280. |
|
try { |
|
data.url = stringToUTF8String(url); |
|
} catch (e) { |
|
// Fall back to a simple copy. |
|
data.url = url; |
|
} |
|
} else if (linkType === 'GoTo') { |
|
data.dest = action.get('D'); |
|
} else if (linkType === 'GoToR') { |
|
var urlDict = action.get('F'); |
|
if (isDict(urlDict)) { |
|
// We assume that the 'url' is a Filspec dictionary |
|
// and fetch the url without checking any further |
|
url = urlDict.get('F') || ''; |
|
} |
|
|
|
// TODO: pdf reference says that GoToR |
|
// can also have 'NewWindow' attribute |
|
if (!isValidUrl(url, false)) { |
|
url = ''; |
|
} |
|
data.url = url; |
|
data.dest = action.get('D'); |
|
} else if (linkType === 'Named') { |
|
data.action = action.get('N').name; |
|
} else { |
|
warn('unrecognized link type: ' + linkType); |
|
} |
|
} else if (dict.has('Dest')) { |
|
// simple destination link |
|
var dest = dict.get('Dest'); |
|
data.dest = isName(dest) ? dest.name : dest; |
|
} |
|
} |
|
|
|
// Lets URLs beginning with 'www.' default to using the 'http://' protocol. |
|
function addDefaultProtocolToUrl(url) { |
|
if (url && url.indexOf('www.') === 0) { |
|
return ('http://' + url); |
|
} |
|
return url; |
|
} |
|
|
|
Util.inherit(LinkAnnotation, Annotation, {}); |
|
|
|
return LinkAnnotation; |
|
})(); |
|
|
|
var PopupAnnotation = (function PopupAnnotationClosure() { |
|
function PopupAnnotation(parameters) { |
|
Annotation.call(this, parameters); |
|
|
|
this.data.annotationType = AnnotationType.POPUP; |
|
|
|
var dict = parameters.dict; |
|
var parentItem = dict.get('Parent'); |
|
if (!parentItem) { |
|
warn('Popup annotation has a missing or invalid parent annotation.'); |
|
return; |
|
} |
|
|
|
this.data.parentId = dict.getRaw('Parent').toString(); |
|
this.data.title = stringToPDFString(parentItem.get('T') || ''); |
|
this.data.contents = stringToPDFString(parentItem.get('Contents') || ''); |
|
|
|
if (!parentItem.has('C')) { |
|
// Fall back to the default background color. |
|
this.data.color = null; |
|
} else { |
|
this.setColor(parentItem.get('C')); |
|
this.data.color = this.color; |
|
} |
|
|
|
this.data.hasHtml = (this.data.title || this.data.contents); |
|
} |
|
|
|
Util.inherit(PopupAnnotation, Annotation, {}); |
|
|
|
return PopupAnnotation; |
|
})(); |
|
|
|
var HighlightAnnotation = (function HighlightAnnotationClosure() { |
|
function HighlightAnnotation(parameters) { |
|
Annotation.call(this, parameters); |
|
|
|
this.data.annotationType = AnnotationType.HIGHLIGHT; |
|
this.data.hasHtml = true; |
|
|
|
// PDF viewers completely ignore any border styles. |
|
this.data.borderStyle.setWidth(0); |
|
} |
|
|
|
Util.inherit(HighlightAnnotation, Annotation, {}); |
|
|
|
return HighlightAnnotation; |
|
})(); |
|
|
|
var UnderlineAnnotation = (function UnderlineAnnotationClosure() { |
|
function UnderlineAnnotation(parameters) { |
|
Annotation.call(this, parameters); |
|
|
|
this.data.annotationType = AnnotationType.UNDERLINE; |
|
this.data.hasHtml = true; |
|
|
|
// PDF viewers completely ignore any border styles. |
|
this.data.borderStyle.setWidth(0); |
|
} |
|
|
|
Util.inherit(UnderlineAnnotation, Annotation, {}); |
|
|
|
return UnderlineAnnotation; |
|
})(); |
|
|
|
var SquigglyAnnotation = (function SquigglyAnnotationClosure() { |
|
function SquigglyAnnotation(parameters) { |
|
Annotation.call(this, parameters); |
|
|
|
this.data.annotationType = AnnotationType.SQUIGGLY; |
|
this.data.hasHtml = true; |
|
|
|
// PDF viewers completely ignore any border styles. |
|
this.data.borderStyle.setWidth(0); |
|
} |
|
|
|
Util.inherit(SquigglyAnnotation, Annotation, {}); |
|
|
|
return SquigglyAnnotation; |
|
})(); |
|
|
|
var StrikeOutAnnotation = (function StrikeOutAnnotationClosure() { |
|
function StrikeOutAnnotation(parameters) { |
|
Annotation.call(this, parameters); |
|
|
|
this.data.annotationType = AnnotationType.STRIKEOUT; |
|
this.data.hasHtml = true; |
|
|
|
// PDF viewers completely ignore any border styles. |
|
this.data.borderStyle.setWidth(0); |
|
} |
|
|
|
Util.inherit(StrikeOutAnnotation, Annotation, {}); |
|
|
|
return StrikeOutAnnotation; |
|
})(); |
|
|
|
exports.Annotation = Annotation; |
|
exports.AnnotationBorderStyle = AnnotationBorderStyle; |
|
exports.AnnotationFactory = AnnotationFactory; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreDocument = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCoreStream, |
|
root.pdfjsCoreObj, root.pdfjsCoreParser, root.pdfjsCoreCrypto, |
|
root.pdfjsCoreEvaluator, root.pdfjsCoreAnnotation); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, coreStream, coreObj, |
|
coreParser, coreCrypto, coreEvaluator, coreAnnotation) { |
|
|
|
var MissingDataException = sharedUtil.MissingDataException; |
|
var Util = sharedUtil.Util; |
|
var assert = sharedUtil.assert; |
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isArray = sharedUtil.isArray; |
|
var isArrayBuffer = sharedUtil.isArrayBuffer; |
|
var isString = sharedUtil.isString; |
|
var shadow = sharedUtil.shadow; |
|
var stringToBytes = sharedUtil.stringToBytes; |
|
var stringToPDFString = sharedUtil.stringToPDFString; |
|
var warn = sharedUtil.warn; |
|
var Dict = corePrimitives.Dict; |
|
var isDict = corePrimitives.isDict; |
|
var isName = corePrimitives.isName; |
|
var isStream = corePrimitives.isStream; |
|
var NullStream = coreStream.NullStream; |
|
var Stream = coreStream.Stream; |
|
var StreamsSequenceStream = coreStream.StreamsSequenceStream; |
|
var Catalog = coreObj.Catalog; |
|
var ObjectLoader = coreObj.ObjectLoader; |
|
var XRef = coreObj.XRef; |
|
var Lexer = coreParser.Lexer; |
|
var Linearization = coreParser.Linearization; |
|
var calculateMD5 = coreCrypto.calculateMD5; |
|
var OperatorList = coreEvaluator.OperatorList; |
|
var PartialEvaluator = coreEvaluator.PartialEvaluator; |
|
var Annotation = coreAnnotation.Annotation; |
|
var AnnotationFactory = coreAnnotation.AnnotationFactory; |
|
|
|
var Page = (function PageClosure() { |
|
|
|
var LETTER_SIZE_MEDIABOX = [0, 0, 612, 792]; |
|
|
|
function Page(pdfManager, xref, pageIndex, pageDict, ref, fontCache) { |
|
this.pdfManager = pdfManager; |
|
this.pageIndex = pageIndex; |
|
this.pageDict = pageDict; |
|
this.xref = xref; |
|
this.ref = ref; |
|
this.fontCache = fontCache; |
|
this.idCounters = { |
|
obj: 0 |
|
}; |
|
this.resourcesPromise = null; |
|
} |
|
|
|
Page.prototype = { |
|
getPageProp: function Page_getPageProp(key) { |
|
return this.pageDict.get(key); |
|
}, |
|
|
|
getInheritedPageProp: function Page_getInheritedPageProp(key) { |
|
var dict = this.pageDict, valueArray = null, loopCount = 0; |
|
var MAX_LOOP_COUNT = 100; |
|
// Always walk up the entire parent chain, to be able to find |
|
// e.g. \Resources placed on multiple levels of the tree. |
|
while (dict) { |
|
var value = dict.get(key); |
|
if (value) { |
|
if (!valueArray) { |
|
valueArray = []; |
|
} |
|
valueArray.push(value); |
|
} |
|
if (++loopCount > MAX_LOOP_COUNT) { |
|
warn('Page_getInheritedPageProp: maximum loop count exceeded.'); |
|
break; |
|
} |
|
dict = dict.get('Parent'); |
|
} |
|
if (!valueArray) { |
|
return Dict.empty; |
|
} |
|
if (valueArray.length === 1 || !isDict(valueArray[0]) || |
|
loopCount > MAX_LOOP_COUNT) { |
|
return valueArray[0]; |
|
} |
|
return Dict.merge(this.xref, valueArray); |
|
}, |
|
|
|
get content() { |
|
return this.getPageProp('Contents'); |
|
}, |
|
|
|
get resources() { |
|
// For robustness: The spec states that a \Resources entry has to be |
|
// present, but can be empty. Some document omit it still, in this case |
|
// we return an empty dictionary. |
|
return shadow(this, 'resources', this.getInheritedPageProp('Resources')); |
|
}, |
|
|
|
get mediaBox() { |
|
var obj = this.getInheritedPageProp('MediaBox'); |
|
// Reset invalid media box to letter size. |
|
if (!isArray(obj) || obj.length !== 4) { |
|
obj = LETTER_SIZE_MEDIABOX; |
|
} |
|
return shadow(this, 'mediaBox', obj); |
|
}, |
|
|
|
get view() { |
|
var mediaBox = this.mediaBox; |
|
var cropBox = this.getInheritedPageProp('CropBox'); |
|
if (!isArray(cropBox) || cropBox.length !== 4) { |
|
return shadow(this, 'view', mediaBox); |
|
} |
|
|
|
// From the spec, 6th ed., p.963: |
|
// "The crop, bleed, trim, and art boxes should not ordinarily |
|
// extend beyond the boundaries of the media box. If they do, they are |
|
// effectively reduced to their intersection with the media box." |
|
cropBox = Util.intersect(cropBox, mediaBox); |
|
if (!cropBox) { |
|
return shadow(this, 'view', mediaBox); |
|
} |
|
return shadow(this, 'view', cropBox); |
|
}, |
|
|
|
get rotate() { |
|
var rotate = this.getInheritedPageProp('Rotate') || 0; |
|
// Normalize rotation so it's a multiple of 90 and between 0 and 270 |
|
if (rotate % 90 !== 0) { |
|
rotate = 0; |
|
} else if (rotate >= 360) { |
|
rotate = rotate % 360; |
|
} else if (rotate < 0) { |
|
// The spec doesn't cover negatives, assume its counterclockwise |
|
// rotation. The following is the other implementation of modulo. |
|
rotate = ((rotate % 360) + 360) % 360; |
|
} |
|
return shadow(this, 'rotate', rotate); |
|
}, |
|
|
|
getContentStream: function Page_getContentStream() { |
|
var content = this.content; |
|
var stream; |
|
if (isArray(content)) { |
|
// fetching items |
|
var xref = this.xref; |
|
var i, n = content.length; |
|
var streams = []; |
|
for (i = 0; i < n; ++i) { |
|
streams.push(xref.fetchIfRef(content[i])); |
|
} |
|
stream = new StreamsSequenceStream(streams); |
|
} else if (isStream(content)) { |
|
stream = content; |
|
} else { |
|
// replacing non-existent page content with empty one |
|
stream = new NullStream(); |
|
} |
|
return stream; |
|
}, |
|
|
|
loadResources: function Page_loadResources(keys) { |
|
if (!this.resourcesPromise) { |
|
// TODO: add async getInheritedPageProp and remove this. |
|
this.resourcesPromise = this.pdfManager.ensure(this, 'resources'); |
|
} |
|
return this.resourcesPromise.then(function resourceSuccess() { |
|
var objectLoader = new ObjectLoader(this.resources.map, |
|
keys, |
|
this.xref); |
|
return objectLoader.load(); |
|
}.bind(this)); |
|
}, |
|
|
|
getOperatorList: function Page_getOperatorList(handler, task, intent) { |
|
var self = this; |
|
|
|
var pdfManager = this.pdfManager; |
|
var contentStreamPromise = pdfManager.ensure(this, 'getContentStream', |
|
[]); |
|
var resourcesPromise = this.loadResources([ |
|
'ExtGState', |
|
'ColorSpace', |
|
'Pattern', |
|
'Shading', |
|
'XObject', |
|
'Font' |
|
// ProcSet |
|
// Properties |
|
]); |
|
|
|
var partialEvaluator = new PartialEvaluator(pdfManager, this.xref, |
|
handler, this.pageIndex, |
|
'p' + this.pageIndex + '_', |
|
this.idCounters, |
|
this.fontCache); |
|
|
|
var dataPromises = Promise.all([contentStreamPromise, resourcesPromise]); |
|
var pageListPromise = dataPromises.then(function(data) { |
|
var contentStream = data[0]; |
|
var opList = new OperatorList(intent, handler, self.pageIndex); |
|
|
|
handler.send('StartRenderPage', { |
|
transparency: partialEvaluator.hasBlendModes(self.resources), |
|
pageIndex: self.pageIndex, |
|
intent: intent |
|
}); |
|
return partialEvaluator.getOperatorList(contentStream, task, |
|
self.resources, opList).then(function () { |
|
return opList; |
|
}); |
|
}); |
|
|
|
var annotationsPromise = pdfManager.ensure(this, 'annotations'); |
|
return Promise.all([pageListPromise, annotationsPromise]).then( |
|
function(datas) { |
|
var pageOpList = datas[0]; |
|
var annotations = datas[1]; |
|
|
|
if (annotations.length === 0) { |
|
pageOpList.flush(true); |
|
return pageOpList; |
|
} |
|
|
|
var annotationsReadyPromise = Annotation.appendToOperatorList( |
|
annotations, pageOpList, partialEvaluator, task, intent); |
|
return annotationsReadyPromise.then(function () { |
|
pageOpList.flush(true); |
|
return pageOpList; |
|
}); |
|
}); |
|
}, |
|
|
|
extractTextContent: function Page_extractTextContent(task, |
|
normalizeWhitespace) { |
|
var handler = { |
|
on: function nullHandlerOn() {}, |
|
send: function nullHandlerSend() {} |
|
}; |
|
|
|
var self = this; |
|
|
|
var pdfManager = this.pdfManager; |
|
var contentStreamPromise = pdfManager.ensure(this, 'getContentStream', |
|
[]); |
|
|
|
var resourcesPromise = this.loadResources([ |
|
'ExtGState', |
|
'XObject', |
|
'Font' |
|
]); |
|
|
|
var dataPromises = Promise.all([contentStreamPromise, |
|
resourcesPromise]); |
|
return dataPromises.then(function(data) { |
|
var contentStream = data[0]; |
|
var partialEvaluator = new PartialEvaluator(pdfManager, self.xref, |
|
handler, self.pageIndex, |
|
'p' + self.pageIndex + '_', |
|
self.idCounters, |
|
self.fontCache); |
|
|
|
return partialEvaluator.getTextContent(contentStream, |
|
task, |
|
self.resources, |
|
/* stateManager = */ null, |
|
normalizeWhitespace); |
|
}); |
|
}, |
|
|
|
getAnnotationsData: function Page_getAnnotationsData(intent) { |
|
var annotations = this.annotations; |
|
var annotationsData = []; |
|
for (var i = 0, n = annotations.length; i < n; ++i) { |
|
if (intent) { |
|
if (!(intent === 'display' && annotations[i].viewable) && |
|
!(intent === 'print' && annotations[i].printable)) { |
|
continue; |
|
} |
|
} |
|
annotationsData.push(annotations[i].data); |
|
} |
|
return annotationsData; |
|
}, |
|
|
|
get annotations() { |
|
var annotations = []; |
|
var annotationRefs = this.getInheritedPageProp('Annots') || []; |
|
var annotationFactory = new AnnotationFactory(); |
|
for (var i = 0, n = annotationRefs.length; i < n; ++i) { |
|
var annotationRef = annotationRefs[i]; |
|
var annotation = annotationFactory.create(this.xref, annotationRef); |
|
if (annotation) { |
|
annotations.push(annotation); |
|
} |
|
} |
|
return shadow(this, 'annotations', annotations); |
|
} |
|
}; |
|
|
|
return Page; |
|
})(); |
|
|
|
/** |
|
* The `PDFDocument` holds all the data of the PDF file. Compared to the |
|
* `PDFDoc`, this one doesn't have any job management code. |
|
* Right now there exists one PDFDocument on the main thread + one object |
|
* for each worker. If there is no worker support enabled, there are two |
|
* `PDFDocument` objects on the main thread created. |
|
*/ |
|
var PDFDocument = (function PDFDocumentClosure() { |
|
var FINGERPRINT_FIRST_BYTES = 1024; |
|
var EMPTY_FINGERPRINT = '\x00\x00\x00\x00\x00\x00\x00' + |
|
'\x00\x00\x00\x00\x00\x00\x00\x00\x00'; |
|
|
|
function PDFDocument(pdfManager, arg, password) { |
|
if (isStream(arg)) { |
|
init.call(this, pdfManager, arg, password); |
|
} else if (isArrayBuffer(arg)) { |
|
init.call(this, pdfManager, new Stream(arg), password); |
|
} else { |
|
error('PDFDocument: Unknown argument type'); |
|
} |
|
} |
|
|
|
function init(pdfManager, stream, password) { |
|
assert(stream.length > 0, 'stream must have data'); |
|
this.pdfManager = pdfManager; |
|
this.stream = stream; |
|
var xref = new XRef(this.stream, password, pdfManager); |
|
this.xref = xref; |
|
} |
|
|
|
function find(stream, needle, limit, backwards) { |
|
var pos = stream.pos; |
|
var end = stream.end; |
|
var strBuf = []; |
|
if (pos + limit > end) { |
|
limit = end - pos; |
|
} |
|
for (var n = 0; n < limit; ++n) { |
|
strBuf.push(String.fromCharCode(stream.getByte())); |
|
} |
|
var str = strBuf.join(''); |
|
stream.pos = pos; |
|
var index = backwards ? str.lastIndexOf(needle) : str.indexOf(needle); |
|
if (index === -1) { |
|
return false; /* not found */ |
|
} |
|
stream.pos += index; |
|
return true; /* found */ |
|
} |
|
|
|
var DocumentInfoValidators = { |
|
get entries() { |
|
// Lazily build this since all the validation functions below are not |
|
// defined until after this file loads. |
|
return shadow(this, 'entries', { |
|
Title: isString, |
|
Author: isString, |
|
Subject: isString, |
|
Keywords: isString, |
|
Creator: isString, |
|
Producer: isString, |
|
CreationDate: isString, |
|
ModDate: isString, |
|
Trapped: isName |
|
}); |
|
} |
|
}; |
|
|
|
PDFDocument.prototype = { |
|
parse: function PDFDocument_parse(recoveryMode) { |
|
this.setup(recoveryMode); |
|
var version = this.catalog.catDict.get('Version'); |
|
if (isName(version)) { |
|
this.pdfFormatVersion = version.name; |
|
} |
|
try { |
|
// checking if AcroForm is present |
|
this.acroForm = this.catalog.catDict.get('AcroForm'); |
|
if (this.acroForm) { |
|
this.xfa = this.acroForm.get('XFA'); |
|
var fields = this.acroForm.get('Fields'); |
|
if ((!fields || !isArray(fields) || fields.length === 0) && |
|
!this.xfa) { |
|
// no fields and no XFA -- not a form (?) |
|
this.acroForm = null; |
|
} |
|
} |
|
} catch (ex) { |
|
info('Something wrong with AcroForm entry'); |
|
this.acroForm = null; |
|
} |
|
}, |
|
|
|
get linearization() { |
|
var linearization = null; |
|
if (this.stream.length) { |
|
try { |
|
linearization = Linearization.create(this.stream); |
|
} catch (err) { |
|
if (err instanceof MissingDataException) { |
|
throw err; |
|
} |
|
info(err); |
|
} |
|
} |
|
// shadow the prototype getter with a data property |
|
return shadow(this, 'linearization', linearization); |
|
}, |
|
get startXRef() { |
|
var stream = this.stream; |
|
var startXRef = 0; |
|
var linearization = this.linearization; |
|
if (linearization) { |
|
// Find end of first obj. |
|
stream.reset(); |
|
if (find(stream, 'endobj', 1024)) { |
|
startXRef = stream.pos + 6; |
|
} |
|
} else { |
|
// Find startxref by jumping backward from the end of the file. |
|
var step = 1024; |
|
var found = false, pos = stream.end; |
|
while (!found && pos > 0) { |
|
pos -= step - 'startxref'.length; |
|
if (pos < 0) { |
|
pos = 0; |
|
} |
|
stream.pos = pos; |
|
found = find(stream, 'startxref', step, true); |
|
} |
|
if (found) { |
|
stream.skip(9); |
|
var ch; |
|
do { |
|
ch = stream.getByte(); |
|
} while (Lexer.isSpace(ch)); |
|
var str = ''; |
|
while (ch >= 0x20 && ch <= 0x39) { // < '9' |
|
str += String.fromCharCode(ch); |
|
ch = stream.getByte(); |
|
} |
|
startXRef = parseInt(str, 10); |
|
if (isNaN(startXRef)) { |
|
startXRef = 0; |
|
} |
|
} |
|
} |
|
// shadow the prototype getter with a data property |
|
return shadow(this, 'startXRef', startXRef); |
|
}, |
|
get mainXRefEntriesOffset() { |
|
var mainXRefEntriesOffset = 0; |
|
var linearization = this.linearization; |
|
if (linearization) { |
|
mainXRefEntriesOffset = linearization.mainXRefEntriesOffset; |
|
} |
|
// shadow the prototype getter with a data property |
|
return shadow(this, 'mainXRefEntriesOffset', mainXRefEntriesOffset); |
|
}, |
|
// Find the header, remove leading garbage and setup the stream |
|
// starting from the header. |
|
checkHeader: function PDFDocument_checkHeader() { |
|
var stream = this.stream; |
|
stream.reset(); |
|
if (find(stream, '%PDF-', 1024)) { |
|
// Found the header, trim off any garbage before it. |
|
stream.moveStart(); |
|
// Reading file format version |
|
var MAX_VERSION_LENGTH = 12; |
|
var version = '', ch; |
|
while ((ch = stream.getByte()) > 0x20) { // SPACE |
|
if (version.length >= MAX_VERSION_LENGTH) { |
|
break; |
|
} |
|
version += String.fromCharCode(ch); |
|
} |
|
if (!this.pdfFormatVersion) { |
|
// removing "%PDF-"-prefix |
|
this.pdfFormatVersion = version.substring(5); |
|
} |
|
return; |
|
} |
|
// May not be a PDF file, continue anyway. |
|
}, |
|
parseStartXRef: function PDFDocument_parseStartXRef() { |
|
var startXRef = this.startXRef; |
|
this.xref.setStartXRef(startXRef); |
|
}, |
|
setup: function PDFDocument_setup(recoveryMode) { |
|
this.xref.parse(recoveryMode); |
|
var self = this; |
|
var pageFactory = { |
|
createPage: function (pageIndex, dict, ref, fontCache) { |
|
return new Page(self.pdfManager, self.xref, pageIndex, dict, ref, |
|
fontCache); |
|
} |
|
}; |
|
this.catalog = new Catalog(this.pdfManager, this.xref, pageFactory); |
|
}, |
|
get numPages() { |
|
var linearization = this.linearization; |
|
var num = linearization ? linearization.numPages : this.catalog.numPages; |
|
// shadow the prototype getter |
|
return shadow(this, 'numPages', num); |
|
}, |
|
get documentInfo() { |
|
var docInfo = { |
|
PDFFormatVersion: this.pdfFormatVersion, |
|
IsAcroFormPresent: !!this.acroForm, |
|
IsXFAPresent: !!this.xfa |
|
}; |
|
var infoDict; |
|
try { |
|
infoDict = this.xref.trailer.get('Info'); |
|
} catch (err) { |
|
info('The document information dictionary is invalid.'); |
|
} |
|
if (infoDict) { |
|
var validEntries = DocumentInfoValidators.entries; |
|
// Only fill the document info with valid entries from the spec. |
|
for (var key in validEntries) { |
|
if (infoDict.has(key)) { |
|
var value = infoDict.get(key); |
|
// Make sure the value conforms to the spec. |
|
if (validEntries[key](value)) { |
|
docInfo[key] = (typeof value !== 'string' ? |
|
value : stringToPDFString(value)); |
|
} else { |
|
info('Bad value in document info for "' + key + '"'); |
|
} |
|
} |
|
} |
|
} |
|
return shadow(this, 'documentInfo', docInfo); |
|
}, |
|
get fingerprint() { |
|
var xref = this.xref, hash, fileID = ''; |
|
var idArray = xref.trailer.get('ID'); |
|
|
|
if (idArray && isArray(idArray) && idArray[0] && isString(idArray[0]) && |
|
idArray[0] !== EMPTY_FINGERPRINT) { |
|
hash = stringToBytes(idArray[0]); |
|
} else { |
|
if (this.stream.ensureRange) { |
|
this.stream.ensureRange(0, |
|
Math.min(FINGERPRINT_FIRST_BYTES, this.stream.end)); |
|
} |
|
hash = calculateMD5(this.stream.bytes.subarray(0, |
|
FINGERPRINT_FIRST_BYTES), 0, FINGERPRINT_FIRST_BYTES); |
|
} |
|
|
|
for (var i = 0, n = hash.length; i < n; i++) { |
|
var hex = hash[i].toString(16); |
|
fileID += hex.length === 1 ? '0' + hex : hex; |
|
} |
|
|
|
return shadow(this, 'fingerprint', fileID); |
|
}, |
|
|
|
getPage: function PDFDocument_getPage(pageIndex) { |
|
return this.catalog.getPage(pageIndex); |
|
}, |
|
|
|
cleanup: function PDFDocument_cleanup() { |
|
return this.catalog.cleanup(); |
|
} |
|
}; |
|
|
|
return PDFDocument; |
|
})(); |
|
|
|
exports.Page = Page; |
|
exports.PDFDocument = PDFDocument; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCorePdfManager = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCoreStream, root.pdfjsCoreChunkedStream, |
|
root.pdfjsCoreDocument); |
|
} |
|
}(this, function (exports, sharedUtil, coreStream, coreChunkedStream, |
|
coreDocument) { |
|
|
|
var NotImplementedException = sharedUtil.NotImplementedException; |
|
var MissingDataException = sharedUtil.MissingDataException; |
|
var createPromiseCapability = sharedUtil.createPromiseCapability; |
|
var Util = sharedUtil.Util; |
|
var Stream = coreStream.Stream; |
|
var ChunkedStreamManager = coreChunkedStream.ChunkedStreamManager; |
|
var PDFDocument = coreDocument.PDFDocument; |
|
|
|
var BasePdfManager = (function BasePdfManagerClosure() { |
|
function BasePdfManager() { |
|
throw new Error('Cannot initialize BaseManagerManager'); |
|
} |
|
|
|
BasePdfManager.prototype = { |
|
get docId() { |
|
return this._docId; |
|
}, |
|
|
|
onLoadedStream: function BasePdfManager_onLoadedStream() { |
|
throw new NotImplementedException(); |
|
}, |
|
|
|
ensureDoc: function BasePdfManager_ensureDoc(prop, args) { |
|
return this.ensure(this.pdfDocument, prop, args); |
|
}, |
|
|
|
ensureXRef: function BasePdfManager_ensureXRef(prop, args) { |
|
return this.ensure(this.pdfDocument.xref, prop, args); |
|
}, |
|
|
|
ensureCatalog: function BasePdfManager_ensureCatalog(prop, args) { |
|
return this.ensure(this.pdfDocument.catalog, prop, args); |
|
}, |
|
|
|
getPage: function BasePdfManager_getPage(pageIndex) { |
|
return this.pdfDocument.getPage(pageIndex); |
|
}, |
|
|
|
cleanup: function BasePdfManager_cleanup() { |
|
return this.pdfDocument.cleanup(); |
|
}, |
|
|
|
ensure: function BasePdfManager_ensure(obj, prop, args) { |
|
return new NotImplementedException(); |
|
}, |
|
|
|
requestRange: function BasePdfManager_requestRange(begin, end) { |
|
return new NotImplementedException(); |
|
}, |
|
|
|
requestLoadedStream: function BasePdfManager_requestLoadedStream() { |
|
return new NotImplementedException(); |
|
}, |
|
|
|
sendProgressiveData: function BasePdfManager_sendProgressiveData(chunk) { |
|
return new NotImplementedException(); |
|
}, |
|
|
|
updatePassword: function BasePdfManager_updatePassword(password) { |
|
this.pdfDocument.xref.password = this.password = password; |
|
if (this._passwordChangedCapability) { |
|
this._passwordChangedCapability.resolve(); |
|
} |
|
}, |
|
|
|
passwordChanged: function BasePdfManager_passwordChanged() { |
|
this._passwordChangedCapability = createPromiseCapability(); |
|
return this._passwordChangedCapability.promise; |
|
}, |
|
|
|
terminate: function BasePdfManager_terminate() { |
|
return new NotImplementedException(); |
|
} |
|
}; |
|
|
|
return BasePdfManager; |
|
})(); |
|
|
|
var LocalPdfManager = (function LocalPdfManagerClosure() { |
|
function LocalPdfManager(docId, data, password) { |
|
this._docId = docId; |
|
var stream = new Stream(data); |
|
this.pdfDocument = new PDFDocument(this, stream, password); |
|
this._loadedStreamCapability = createPromiseCapability(); |
|
this._loadedStreamCapability.resolve(stream); |
|
} |
|
|
|
Util.inherit(LocalPdfManager, BasePdfManager, { |
|
ensure: function LocalPdfManager_ensure(obj, prop, args) { |
|
return new Promise(function (resolve, reject) { |
|
try { |
|
var value = obj[prop]; |
|
var result; |
|
if (typeof value === 'function') { |
|
result = value.apply(obj, args); |
|
} else { |
|
result = value; |
|
} |
|
resolve(result); |
|
} catch (e) { |
|
reject(e); |
|
} |
|
}); |
|
}, |
|
|
|
requestRange: function LocalPdfManager_requestRange(begin, end) { |
|
return Promise.resolve(); |
|
}, |
|
|
|
requestLoadedStream: function LocalPdfManager_requestLoadedStream() { |
|
return; |
|
}, |
|
|
|
onLoadedStream: function LocalPdfManager_onLoadedStream() { |
|
return this._loadedStreamCapability.promise; |
|
}, |
|
|
|
terminate: function LocalPdfManager_terminate() { |
|
return; |
|
} |
|
}); |
|
|
|
return LocalPdfManager; |
|
})(); |
|
|
|
var NetworkPdfManager = (function NetworkPdfManagerClosure() { |
|
function NetworkPdfManager(docId, args, msgHandler) { |
|
this._docId = docId; |
|
this.msgHandler = msgHandler; |
|
|
|
var params = { |
|
msgHandler: msgHandler, |
|
httpHeaders: args.httpHeaders, |
|
withCredentials: args.withCredentials, |
|
chunkedViewerLoading: args.chunkedViewerLoading, |
|
disableAutoFetch: args.disableAutoFetch, |
|
initialData: args.initialData |
|
}; |
|
this.streamManager = new ChunkedStreamManager(args.length, |
|
args.rangeChunkSize, |
|
args.url, params); |
|
this.pdfDocument = new PDFDocument(this, this.streamManager.getStream(), |
|
args.password); |
|
} |
|
|
|
Util.inherit(NetworkPdfManager, BasePdfManager, { |
|
ensure: function NetworkPdfManager_ensure(obj, prop, args) { |
|
var pdfManager = this; |
|
|
|
return new Promise(function (resolve, reject) { |
|
function ensureHelper() { |
|
try { |
|
var result; |
|
var value = obj[prop]; |
|
if (typeof value === 'function') { |
|
result = value.apply(obj, args); |
|
} else { |
|
result = value; |
|
} |
|
resolve(result); |
|
} catch(e) { |
|
if (!(e instanceof MissingDataException)) { |
|
reject(e); |
|
return; |
|
} |
|
pdfManager.streamManager.requestRange(e.begin, e.end). |
|
then(ensureHelper, reject); |
|
} |
|
} |
|
|
|
ensureHelper(); |
|
}); |
|
}, |
|
|
|
requestRange: function NetworkPdfManager_requestRange(begin, end) { |
|
return this.streamManager.requestRange(begin, end); |
|
}, |
|
|
|
requestLoadedStream: function NetworkPdfManager_requestLoadedStream() { |
|
this.streamManager.requestAllChunks(); |
|
}, |
|
|
|
sendProgressiveData: |
|
function NetworkPdfManager_sendProgressiveData(chunk) { |
|
this.streamManager.onReceiveData({ chunk: chunk }); |
|
}, |
|
|
|
onLoadedStream: function NetworkPdfManager_onLoadedStream() { |
|
return this.streamManager.onLoadedStream(); |
|
}, |
|
|
|
terminate: function NetworkPdfManager_terminate() { |
|
this.streamManager.abort(); |
|
} |
|
}); |
|
|
|
return NetworkPdfManager; |
|
})(); |
|
|
|
exports.LocalPdfManager = LocalPdfManager; |
|
exports.NetworkPdfManager = NetworkPdfManager; |
|
})); |
|
|
|
|
|
(function (root, factory) { |
|
{ |
|
factory((root.pdfjsCoreWorker = {}), root.pdfjsSharedUtil, |
|
root.pdfjsCorePrimitives, root.pdfjsCorePdfManager, |
|
root.pdfjsSharedGlobal); |
|
} |
|
}(this, function (exports, sharedUtil, corePrimitives, corePdfManager, |
|
sharedGlobal) { |
|
|
|
var UNSUPPORTED_FEATURES = sharedUtil.UNSUPPORTED_FEATURES; |
|
var InvalidPDFException = sharedUtil.InvalidPDFException; |
|
var MessageHandler = sharedUtil.MessageHandler; |
|
var MissingPDFException = sharedUtil.MissingPDFException; |
|
var UnexpectedResponseException = sharedUtil.UnexpectedResponseException; |
|
var PasswordException = sharedUtil.PasswordException; |
|
var PasswordResponses = sharedUtil.PasswordResponses; |
|
var UnknownErrorException = sharedUtil.UnknownErrorException; |
|
var XRefParseException = sharedUtil.XRefParseException; |
|
var createPromiseCapability = sharedUtil.createPromiseCapability; |
|
var error = sharedUtil.error; |
|
var info = sharedUtil.info; |
|
var isInt = sharedUtil.isInt; |
|
var warn = sharedUtil.warn; |
|
var Ref = corePrimitives.Ref; |
|
var LocalPdfManager = corePdfManager.LocalPdfManager; |
|
var NetworkPdfManager = corePdfManager.NetworkPdfManager; |
|
var globalScope = sharedGlobal.globalScope; |
|
var PDFJS = sharedGlobal.PDFJS; |
|
|
|
var WorkerTask = (function WorkerTaskClosure() { |
|
function WorkerTask(name) { |
|
this.name = name; |
|
this.terminated = false; |
|
this._capability = createPromiseCapability(); |
|
} |
|
|
|
WorkerTask.prototype = { |
|
get finished() { |
|
return this._capability.promise; |
|
}, |
|
|
|
finish: function () { |
|
this._capability.resolve(); |
|
}, |
|
|
|
terminate: function () { |
|
this.terminated = true; |
|
}, |
|
|
|
ensureNotTerminated: function () { |
|
if (this.terminated) { |
|
throw new Error('Worker task was terminated'); |
|
} |
|
} |
|
}; |
|
|
|
return WorkerTask; |
|
})(); |
|
|
|
var WorkerMessageHandler = PDFJS.WorkerMessageHandler = { |
|
setup: function wphSetup(handler, port) { |
|
var testMessageProcessed = false; |
|
handler.on('test', function wphSetupTest(data) { |
|
if (testMessageProcessed) { |
|
return; // we already processed 'test' message once |
|
} |
|
testMessageProcessed = true; |
|
|
|
// check if Uint8Array can be sent to worker |
|
if (!(data instanceof Uint8Array)) { |
|
handler.send('test', 'main', false); |
|
return; |
|
} |
|
// making sure postMessage transfers are working |
|
var supportTransfers = data[0] === 255; |
|
handler.postMessageTransfers = supportTransfers; |
|
// check if the response property is supported by xhr |
|
var xhr = new XMLHttpRequest(); |
|
var responseExists = 'response' in xhr; |
|
// check if the property is actually implemented |
|
try { |
|
var dummy = xhr.responseType; |
|
} catch (e) { |
|
responseExists = false; |
|
} |
|
if (!responseExists) { |
|
handler.send('test', false); |
|
return; |
|
} |
|
handler.send('test', { |
|
supportTypedArray: true, |
|
supportTransfers: supportTransfers |
|
}); |
|
}); |
|
|
|
handler.on('GetDocRequest', function wphSetupDoc(data) { |
|
return WorkerMessageHandler.createDocumentHandler(data, port); |
|
}); |
|
}, |
|
createDocumentHandler: function wphCreateDocumentHandler(docParams, port) { |
|
// This context is actually holds references on pdfManager and handler, |
|
// until the latter is destroyed. |
|
var pdfManager; |
|
var terminated = false; |
|
var cancelXHRs = null; |
|
var WorkerTasks = []; |
|
|
|
var docId = docParams.docId; |
|
var workerHandlerName = docParams.docId + '_worker'; |
|
var handler = new MessageHandler(workerHandlerName, docId, port); |
|
|
|
function ensureNotTerminated() { |
|
if (terminated) { |
|
throw new Error('Worker was terminated'); |
|
} |
|
} |
|
|
|
function startWorkerTask(task) { |
|
WorkerTasks.push(task); |
|
} |
|
|
|
function finishWorkerTask(task) { |
|
task.finish(); |
|
var i = WorkerTasks.indexOf(task); |
|
WorkerTasks.splice(i, 1); |
|
} |
|
|
|
function loadDocument(recoveryMode) { |
|
var loadDocumentCapability = createPromiseCapability(); |
|
|
|
var parseSuccess = function parseSuccess() { |
|
var numPagesPromise = pdfManager.ensureDoc('numPages'); |
|
var fingerprintPromise = pdfManager.ensureDoc('fingerprint'); |
|
var encryptedPromise = pdfManager.ensureXRef('encrypt'); |
|
Promise.all([numPagesPromise, fingerprintPromise, |
|
encryptedPromise]).then(function onDocReady(results) { |
|
var doc = { |
|
numPages: results[0], |
|
fingerprint: results[1], |
|
encrypted: !!results[2], |
|
}; |
|
loadDocumentCapability.resolve(doc); |
|
}, |
|
parseFailure); |
|
}; |
|
|
|
var parseFailure = function parseFailure(e) { |
|
loadDocumentCapability.reject(e); |
|
}; |
|
|
|
pdfManager.ensureDoc('checkHeader', []).then(function() { |
|
pdfManager.ensureDoc('parseStartXRef', []).then(function() { |
|
pdfManager.ensureDoc('parse', [recoveryMode]).then( |
|
parseSuccess, parseFailure); |
|
}, parseFailure); |
|
}, parseFailure); |
|
|
|
return loadDocumentCapability.promise; |
|
} |
|
|
|
function getPdfManager(data) { |
|
var pdfManagerCapability = createPromiseCapability(); |
|
var pdfManager; |
|
|
|
var source = data.source; |
|
var disableRange = data.disableRange; |
|
if (source.data) { |
|
try { |
|
pdfManager = new LocalPdfManager(docId, source.data, source.password); |
|
pdfManagerCapability.resolve(pdfManager); |
|
} catch (ex) { |
|
pdfManagerCapability.reject(ex); |
|
} |
|
return pdfManagerCapability.promise; |
|
} else if (source.chunkedViewerLoading) { |
|
try { |
|
pdfManager = new NetworkPdfManager(docId, source, handler); |
|
pdfManagerCapability.resolve(pdfManager); |
|
} catch (ex) { |
|
pdfManagerCapability.reject(ex); |
|
} |
|
return pdfManagerCapability.promise; |
|
} |
|
|
|
var networkManager = new NetworkManager(source.url, { |
|
httpHeaders: source.httpHeaders, |
|
withCredentials: source.withCredentials |
|
}); |
|
var cachedChunks = []; |
|
var fullRequestXhrId = networkManager.requestFull({ |
|
onHeadersReceived: function onHeadersReceived() { |
|
if (disableRange) { |
|
return; |
|
} |
|
|
|
var fullRequestXhr = networkManager.getRequestXhr(fullRequestXhrId); |
|
if (fullRequestXhr.getResponseHeader('Accept-Ranges') !== 'bytes') { |
|
return; |
|
} |
|
|
|
var contentEncoding = |
|
fullRequestXhr.getResponseHeader('Content-Encoding') || 'identity'; |
|
if (contentEncoding !== 'identity') { |
|
return; |
|
} |
|
|
|
var length = fullRequestXhr.getResponseHeader('Content-Length'); |
|
length = parseInt(length, 10); |
|
if (!isInt(length)) { |
|
return; |
|
} |
|
source.length = length; |
|
if (length <= 2 * source.rangeChunkSize) { |
|
// The file size is smaller than the size of two chunks, so it does |
|
// not make any sense to abort the request and retry with a range |
|
// request. |
|
return; |
|
} |
|
|
|
if (networkManager.isStreamingRequest(fullRequestXhrId)) { |
|
// We can continue fetching when progressive loading is enabled, |
|
// and we don't need the autoFetch feature. |
|
source.disableAutoFetch = true; |
|
} else { |
|
// NOTE: by cancelling the full request, and then issuing range |
|
// requests, there will be an issue for sites where you can only |
|
// request the pdf once. However, if this is the case, then the |
|
// server should not be returning that it can support range |
|
// requests. |
|
networkManager.abortRequest(fullRequestXhrId); |
|
} |
|
|
|
try { |
|
pdfManager = new NetworkPdfManager(docId, source, handler); |
|
pdfManagerCapability.resolve(pdfManager); |
|
} catch (ex) { |
|
pdfManagerCapability.reject(ex); |
|
} |
|
cancelXHRs = null; |
|
}, |
|
|
|
onProgressiveData: source.disableStream ? null : |
|
function onProgressiveData(chunk) { |
|
if (!pdfManager) { |
|
cachedChunks.push(chunk); |
|
return; |
|
} |
|
pdfManager.sendProgressiveData(chunk); |
|
}, |
|
|
|
onDone: function onDone(args) { |
|
if (pdfManager) { |
|
return; // already processed |
|
} |
|
|
|
var pdfFile; |
|
if (args === null) { |
|
// TODO add some streaming manager, e.g. for unknown length files. |
|
// The data was returned in the onProgressiveData, combining... |
|
var pdfFileLength = 0, pos = 0; |
|
cachedChunks.forEach(function (chunk) { |
|
pdfFileLength += chunk.byteLength; |
|
}); |
|
if (source.length && pdfFileLength !== source.length) { |
|
warn('reported HTTP length is different from actual'); |
|
} |
|
var pdfFileArray = new Uint8Array(pdfFileLength); |
|
cachedChunks.forEach(function (chunk) { |
|
pdfFileArray.set(new Uint8Array(chunk), pos); |
|
pos += chunk.byteLength; |
|
}); |
|
pdfFile = pdfFileArray.buffer; |
|
} else { |
|
pdfFile = args.chunk; |
|
} |
|
|
|
// the data is array, instantiating directly from it |
|
try { |
|
pdfManager = new LocalPdfManager(docId, pdfFile, source.password); |
|
pdfManagerCapability.resolve(pdfManager); |
|
} catch (ex) { |
|
pdfManagerCapability.reject(ex); |
|
} |
|
cancelXHRs = null; |
|
}, |
|
|
|
onError: function onError(status) { |
|
var exception; |
|
if (status === 404 || status === 0 && /^file:/.test(source.url)) { |
|
exception = new MissingPDFException('Missing PDF "' + |
|
source.url + '".'); |
|
handler.send('MissingPDF', exception); |
|
} else { |
|
exception = new UnexpectedResponseException( |
|
'Unexpected server response (' + status + |
|
') while retrieving PDF "' + source.url + '".', status); |
|
handler.send('UnexpectedResponse', exception); |
|
} |
|
cancelXHRs = null; |
|
}, |
|
|
|
onProgress: function onProgress(evt) { |
|
handler.send('DocProgress', { |
|
loaded: evt.loaded, |
|
total: evt.lengthComputable ? evt.total : source.length |
|
}); |
|
} |
|
}); |
|
|
|
cancelXHRs = function () { |
|
networkManager.abortRequest(fullRequestXhrId); |
|
}; |
|
|
|
return pdfManagerCapability.promise; |
|
} |
|
|
|
var setupDoc = function(data) { |
|
var onSuccess = function(doc) { |
|
ensureNotTerminated(); |
|
handler.send('GetDoc', { pdfInfo: doc }); |
|
}; |
|
|
|
var onFailure = function(e) { |
|
if (e instanceof PasswordException) { |
|
if (e.code === PasswordResponses.NEED_PASSWORD) { |
|
handler.send('NeedPassword', e); |
|
} else if (e.code === PasswordResponses.INCORRECT_PASSWORD) { |
|
handler.send('IncorrectPassword', e); |
|
} |
|
} else if (e instanceof InvalidPDFException) { |
|
handler.send('InvalidPDF', e); |
|
} else if (e instanceof MissingPDFException) { |
|
handler.send('MissingPDF', e); |
|
} else if (e instanceof UnexpectedResponseException) { |
|
handler.send('UnexpectedResponse', e); |
|
} else { |
|
handler.send('UnknownError', |
|
new UnknownErrorException(e.message, e.toString())); |
|
} |
|
}; |
|
|
|
ensureNotTerminated(); |
|
|
|
PDFJS.maxImageSize = data.maxImageSize === undefined ? |
|
-1 : data.maxImageSize; |
|
PDFJS.disableFontFace = data.disableFontFace; |
|
PDFJS.disableCreateObjectURL = data.disableCreateObjectURL; |
|
PDFJS.verbosity = data.verbosity; |
|
PDFJS.cMapUrl = data.cMapUrl === undefined ? |
|
null : data.cMapUrl; |
|
PDFJS.cMapPacked = data.cMapPacked === true; |
|
|
|
getPdfManager(data).then(function (newPdfManager) { |
|
if (terminated) { |
|
// We were in a process of setting up the manager, but it got |
|
// terminated in the middle. |
|
newPdfManager.terminate(); |
|
throw new Error('Worker was terminated'); |
|
} |
|
|
|
pdfManager = newPdfManager; |
|
handler.send('PDFManagerReady', null); |
|
pdfManager.onLoadedStream().then(function(stream) { |
|
handler.send('DataLoaded', { length: stream.bytes.byteLength }); |
|
}); |
|
}).then(function pdfManagerReady() { |
|
ensureNotTerminated(); |
|
|
|
loadDocument(false).then(onSuccess, function loadFailure(ex) { |
|
ensureNotTerminated(); |
|
|
|
// Try again with recoveryMode == true |
|
if (!(ex instanceof XRefParseException)) { |
|
if (ex instanceof PasswordException) { |
|
// after password exception prepare to receive a new password |
|
// to repeat loading |
|
pdfManager.passwordChanged().then(pdfManagerReady); |
|
} |
|
|
|
onFailure(ex); |
|
return; |
|
} |
|
|
|
pdfManager.requestLoadedStream(); |
|
pdfManager.onLoadedStream().then(function() { |
|
ensureNotTerminated(); |
|
|
|
loadDocument(true).then(onSuccess, onFailure); |
|
}); |
|
}, onFailure); |
|
}, onFailure); |
|
}; |
|
|
|
handler.on('GetPage', function wphSetupGetPage(data) { |
|
return pdfManager.getPage(data.pageIndex).then(function(page) { |
|
var rotatePromise = pdfManager.ensure(page, 'rotate'); |
|
var refPromise = pdfManager.ensure(page, 'ref'); |
|
var viewPromise = pdfManager.ensure(page, 'view'); |
|
|
|
return Promise.all([rotatePromise, refPromise, viewPromise]).then( |
|
function(results) { |
|
return { |
|
rotate: results[0], |
|
ref: results[1], |
|
view: results[2] |
|
}; |
|
}); |
|
}); |
|
}); |
|
|
|
handler.on('GetPageIndex', function wphSetupGetPageIndex(data) { |
|
var ref = new Ref(data.ref.num, data.ref.gen); |
|
var catalog = pdfManager.pdfDocument.catalog; |
|
return catalog.getPageIndex(ref); |
|
}); |
|
|
|
handler.on('GetDestinations', |
|
function wphSetupGetDestinations(data) { |
|
return pdfManager.ensureCatalog('destinations'); |
|
} |
|
); |
|
|
|
handler.on('GetDestination', |
|
function wphSetupGetDestination(data) { |
|
return pdfManager.ensureCatalog('getDestination', [data.id]); |
|
} |
|
); |
|
|
|
handler.on('GetPageLabels', |
|
function wphSetupGetPageLabels(data) { |
|
return pdfManager.ensureCatalog('pageLabels'); |
|
} |
|
); |
|
|
|
handler.on('GetAttachments', |
|
function wphSetupGetAttachments(data) { |
|
return pdfManager.ensureCatalog('attachments'); |
|
} |
|
); |
|
|
|
handler.on('GetJavaScript', |
|
function wphSetupGetJavaScript(data) { |
|
return pdfManager.ensureCatalog('javaScript'); |
|
} |
|
); |
|
|
|
handler.on('GetOutline', |
|
function wphSetupGetOutline(data) { |
|
return pdfManager.ensureCatalog('documentOutline'); |
|
} |
|
); |
|
|
|
handler.on('GetMetadata', |
|
function wphSetupGetMetadata(data) { |
|
return Promise.all([pdfManager.ensureDoc('documentInfo'), |
|
pdfManager.ensureCatalog('metadata')]); |
|
} |
|
); |
|
|
|
handler.on('GetData', function wphSetupGetData(data) { |
|
pdfManager.requestLoadedStream(); |
|
return pdfManager.onLoadedStream().then(function(stream) { |
|
return stream.bytes; |
|
}); |
|
}); |
|
|
|
handler.on('GetStats', |
|
function wphSetupGetStats(data) { |
|
return pdfManager.pdfDocument.xref.stats; |
|
} |
|
); |
|
|
|
handler.on('UpdatePassword', function wphSetupUpdatePassword(data) { |
|
pdfManager.updatePassword(data); |
|
}); |
|
|
|
handler.on('GetAnnotations', function wphSetupGetAnnotations(data) { |
|
return pdfManager.getPage(data.pageIndex).then(function(page) { |
|
return pdfManager.ensure(page, 'getAnnotationsData', [data.intent]); |
|
}); |
|
}); |
|
|
|
handler.on('RenderPageRequest', function wphSetupRenderPage(data) { |
|
var pageIndex = data.pageIndex; |
|
pdfManager.getPage(pageIndex).then(function(page) { |
|
var task = new WorkerTask('RenderPageRequest: page ' + pageIndex); |
|
startWorkerTask(task); |
|
|
|
var pageNum = pageIndex + 1; |
|
var start = Date.now(); |
|
// Pre compile the pdf page and fetch the fonts/images. |
|
page.getOperatorList(handler, task, data.intent).then( |
|
function(operatorList) { |
|
finishWorkerTask(task); |
|
|
|
info('page=' + pageNum + ' - getOperatorList: time=' + |
|
(Date.now() - start) + 'ms, len=' + operatorList.totalLength); |
|
}, function(e) { |
|
finishWorkerTask(task); |
|
if (task.terminated) { |
|
return; // ignoring errors from the terminated thread |
|
} |
|
|
|
// For compatibility with older behavior, generating unknown |
|
// unsupported feature notification on errors. |
|
handler.send('UnsupportedFeature', |
|
{featureId: UNSUPPORTED_FEATURES.unknown}); |
|
|
|
var minimumStackMessage = |
|
'worker.js: while trying to getPage() and getOperatorList()'; |
|
|
|
var wrappedException; |
|
|
|
// Turn the error into an obj that can be serialized |
|
if (typeof e === 'string') { |
|
wrappedException = { |
|
message: e, |
|
stack: minimumStackMessage |
|
}; |
|
} else if (typeof e === 'object') { |
|
wrappedException = { |
|
message: e.message || e.toString(), |
|
stack: e.stack || minimumStackMessage |
|
}; |
|
} else { |
|
wrappedException = { |
|
message: 'Unknown exception type: ' + (typeof e), |
|
stack: minimumStackMessage |
|
}; |
|
} |
|
|
|
handler.send('PageError', { |
|
pageNum: pageNum, |
|
error: wrappedException, |
|
intent: data.intent |
|
}); |
|
}); |
|
}); |
|
}, this); |
|
|
|
handler.on('GetTextContent', function wphExtractText(data) { |
|
var pageIndex = data.pageIndex; |
|
var normalizeWhitespace = data.normalizeWhitespace; |
|
return pdfManager.getPage(pageIndex).then(function(page) { |
|
var task = new WorkerTask('GetTextContent: page ' + pageIndex); |
|
startWorkerTask(task); |
|
var pageNum = pageIndex + 1; |
|
var start = Date.now(); |
|
return page.extractTextContent(task, normalizeWhitespace).then( |
|
function(textContent) { |
|
finishWorkerTask(task); |
|
info('text indexing: page=' + pageNum + ' - time=' + |
|
(Date.now() - start) + 'ms'); |
|
return textContent; |
|
}, function (reason) { |
|
finishWorkerTask(task); |
|
if (task.terminated) { |
|
return; // ignoring errors from the terminated thread |
|
} |
|
throw reason; |
|
}); |
|
}); |
|
}); |
|
|
|
handler.on('Cleanup', function wphCleanup(data) { |
|
return pdfManager.cleanup(); |
|
}); |
|
|
|
handler.on('Terminate', function wphTerminate(data) { |
|
terminated = true; |
|
if (pdfManager) { |
|
pdfManager.terminate(); |
|
pdfManager = null; |
|
} |
|
if (cancelXHRs) { |
|
cancelXHRs(); |
|
} |
|
|
|
var waitOn = []; |
|
WorkerTasks.forEach(function (task) { |
|
waitOn.push(task.finished); |
|
task.terminate(); |
|
}); |
|
|
|
return Promise.all(waitOn).then(function () { |
|
// Notice that even if we destroying handler, resolved response promise |
|
// must be sent back. |
|
handler.destroy(); |
|
handler = null; |
|
}); |
|
}); |
|
|
|
handler.on('Ready', function wphReady(data) { |
|
setupDoc(docParams); |
|
docParams = null; // we don't need docParams anymore -- saving memory. |
|
}); |
|
return workerHandlerName; |
|
} |
|
}; |
|
|
|
function initializeWorker() { |
|
if (!('console' in globalScope)) { |
|
var consoleTimer = {}; |
|
|
|
var workerConsole = { |
|
log: function log() { |
|
var args = Array.prototype.slice.call(arguments); |
|
globalScope.postMessage({ |
|
targetName: 'main', |
|
action: 'console_log', |
|
data: args |
|
}); |
|
}, |
|
|
|
error: function error() { |
|
var args = Array.prototype.slice.call(arguments); |
|
globalScope.postMessage({ |
|
targetName: 'main', |
|
action: 'console_error', |
|
data: args |
|
}); |
|
throw 'pdf.js execution error'; |
|
}, |
|
|
|
time: function time(name) { |
|
consoleTimer[name] = Date.now(); |
|
}, |
|
|
|
timeEnd: function timeEnd(name) { |
|
var time = consoleTimer[name]; |
|
if (!time) { |
|
error('Unknown timer name ' + name); |
|
} |
|
this.log('Timer:', name, Date.now() - time); |
|
} |
|
}; |
|
|
|
globalScope.console = workerConsole; |
|
} |
|
|
|
var handler = new MessageHandler('worker', 'main', self); |
|
WorkerMessageHandler.setup(handler, self); |
|
handler.send('ready', null); |
|
} |
|
|
|
// Worker thread (and not node.js)? |
|
if (typeof window === 'undefined' && |
|
!(typeof module !== 'undefined' && module.require)) { |
|
initializeWorker(); |
|
} |
|
|
|
exports.WorkerTask = WorkerTask; |
|
exports.WorkerMessageHandler = WorkerMessageHandler; |
|
})); |
|
|
|
|
|
}).call(pdfjsLibs); |
|
|
|
exports.PDFJS = pdfjsLibs.pdfjsSharedGlobal.PDFJS; |
|
|
|
})); |
|
|
|
|