|
|
|
package log
|
|
|
|
|
|
|
|
import (
|
|
|
|
"bytes"
|
|
|
|
"fmt"
|
|
|
|
"log/slog"
|
|
|
|
"math/big"
|
|
|
|
"reflect"
|
|
|
|
"strconv"
|
|
|
|
"time"
|
|
|
|
"unicode/utf8"
|
|
|
|
|
|
|
|
"github.com/holiman/uint256"
|
|
|
|
)
|
|
|
|
|
|
|
|
const (
|
|
|
|
timeFormat = "2006-01-02T15:04:05-0700"
|
|
|
|
floatFormat = 'f'
|
|
|
|
termMsgJust = 40
|
|
|
|
termCtxMaxPadding = 40
|
|
|
|
)
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
// 40 spaces
|
|
|
|
var spaces = []byte(" ")
|
|
|
|
|
|
|
|
// TerminalStringer is an analogous interface to the stdlib stringer, allowing
|
|
|
|
// own types to have custom shortened serialization formats when printed to the
|
|
|
|
// screen.
|
|
|
|
type TerminalStringer interface {
|
|
|
|
TerminalString() string
|
|
|
|
}
|
|
|
|
|
log: remove lazy, remove unused interfaces, unexport methods (#28622)
This change
- Removes interface `log.Format`,
- Removes method `log.FormatFunc`,
- unexports `TerminalHandler.TerminalFormat` formatting methods (renamed to `TerminalHandler.format`)
- removes the notion of `log.Lazy` values
The lazy handler was useful in the old log package, since it
could defer the evaluation of costly attributes until later in the
log pipeline: thus, if the logging was done at 'Trace', we could
skip evaluation if logging only was set to 'Info'.
With the move to slog, this way of deferring evaluation is no longer
needed, since slog introduced 'Enabled': the caller can thus do
the evaluate-or-not decision at the callsite, which is much more
straight-forward than dealing with lazy reflect-based evaluation.
Also, lazy evaluation would not work with 'native' slog, as in, these
two statements would be evaluated differently:
```golang
log.Info("foo", "my lazy", lazyObj)
slog.Info("foo", "my lazy", lazyObj)
```
11 months ago
|
|
|
func (h *TerminalHandler) format(buf []byte, r slog.Record, usecolor bool) []byte {
|
|
|
|
msg := escapeMessage(r.Message)
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
var color = ""
|
|
|
|
if usecolor {
|
|
|
|
switch r.Level {
|
|
|
|
case LevelCrit:
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
color = "\x1b[35m"
|
|
|
|
case slog.LevelError:
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
color = "\x1b[31m"
|
|
|
|
case slog.LevelWarn:
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
color = "\x1b[33m"
|
|
|
|
case slog.LevelInfo:
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
color = "\x1b[32m"
|
|
|
|
case slog.LevelDebug:
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
color = "\x1b[36m"
|
|
|
|
case LevelTrace:
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
color = "\x1b[34m"
|
|
|
|
}
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
if buf == nil {
|
|
|
|
buf = make([]byte, 0, 30+termMsgJust)
|
|
|
|
}
|
|
|
|
b := bytes.NewBuffer(buf)
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
if color != "" { // Start color
|
|
|
|
b.WriteString(color)
|
|
|
|
b.WriteString(LevelAlignedString(r.Level))
|
|
|
|
b.WriteString("\x1b[0m")
|
|
|
|
} else {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
b.WriteString(LevelAlignedString(r.Level))
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
b.WriteString("[")
|
|
|
|
writeTimeTermFormat(b, r.Time)
|
|
|
|
b.WriteString("] ")
|
|
|
|
b.WriteString(msg)
|
|
|
|
|
|
|
|
// try to justify the log output for short messages
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
//length := utf8.RuneCountInString(msg)
|
|
|
|
length := len(msg)
|
|
|
|
if (r.NumAttrs()+len(h.attrs)) > 0 && length < termMsgJust {
|
|
|
|
b.Write(spaces[:termMsgJust-length])
|
|
|
|
}
|
log: remove lazy, remove unused interfaces, unexport methods (#28622)
This change
- Removes interface `log.Format`,
- Removes method `log.FormatFunc`,
- unexports `TerminalHandler.TerminalFormat` formatting methods (renamed to `TerminalHandler.format`)
- removes the notion of `log.Lazy` values
The lazy handler was useful in the old log package, since it
could defer the evaluation of costly attributes until later in the
log pipeline: thus, if the logging was done at 'Trace', we could
skip evaluation if logging only was set to 'Info'.
With the move to slog, this way of deferring evaluation is no longer
needed, since slog introduced 'Enabled': the caller can thus do
the evaluate-or-not decision at the callsite, which is much more
straight-forward than dealing with lazy reflect-based evaluation.
Also, lazy evaluation would not work with 'native' slog, as in, these
two statements would be evaluated differently:
```golang
log.Info("foo", "my lazy", lazyObj)
slog.Info("foo", "my lazy", lazyObj)
```
11 months ago
|
|
|
// print the attributes
|
|
|
|
h.formatAttributes(b, r, color)
|
|
|
|
|
|
|
|
return b.Bytes()
|
|
|
|
}
|
|
|
|
|
log: remove lazy, remove unused interfaces, unexport methods (#28622)
This change
- Removes interface `log.Format`,
- Removes method `log.FormatFunc`,
- unexports `TerminalHandler.TerminalFormat` formatting methods (renamed to `TerminalHandler.format`)
- removes the notion of `log.Lazy` values
The lazy handler was useful in the old log package, since it
could defer the evaluation of costly attributes until later in the
log pipeline: thus, if the logging was done at 'Trace', we could
skip evaluation if logging only was set to 'Info'.
With the move to slog, this way of deferring evaluation is no longer
needed, since slog introduced 'Enabled': the caller can thus do
the evaluate-or-not decision at the callsite, which is much more
straight-forward than dealing with lazy reflect-based evaluation.
Also, lazy evaluation would not work with 'native' slog, as in, these
two statements would be evaluated differently:
```golang
log.Info("foo", "my lazy", lazyObj)
slog.Info("foo", "my lazy", lazyObj)
```
11 months ago
|
|
|
func (h *TerminalHandler) formatAttributes(buf *bytes.Buffer, r slog.Record, color string) {
|
|
|
|
writeAttr := func(attr slog.Attr, last bool) {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
buf.WriteByte(' ')
|
|
|
|
|
|
|
|
if color != "" {
|
|
|
|
buf.WriteString(color)
|
|
|
|
buf.Write(appendEscapeString(buf.AvailableBuffer(), attr.Key))
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
buf.WriteString("\x1b[0m=")
|
|
|
|
} else {
|
|
|
|
buf.Write(appendEscapeString(buf.AvailableBuffer(), attr.Key))
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
buf.WriteByte('=')
|
|
|
|
}
|
|
|
|
val := FormatSlogValue(attr.Value, buf.AvailableBuffer())
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
padding := h.fieldPadding[attr.Key]
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
length := utf8.RuneCount(val)
|
|
|
|
if padding < length && length <= termCtxMaxPadding {
|
|
|
|
padding = length
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
h.fieldPadding[attr.Key] = padding
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
buf.Write(val)
|
|
|
|
if !last && padding > length {
|
|
|
|
buf.Write(spaces[:padding-length])
|
|
|
|
}
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
var n = 0
|
|
|
|
var nAttrs = len(h.attrs) + r.NumAttrs()
|
|
|
|
for _, attr := range h.attrs {
|
|
|
|
writeAttr(attr, n == nAttrs-1)
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
n++
|
|
|
|
}
|
|
|
|
r.Attrs(func(attr slog.Attr) bool {
|
|
|
|
writeAttr(attr, n == nAttrs-1)
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
n++
|
|
|
|
return true
|
|
|
|
})
|
|
|
|
buf.WriteByte('\n')
|
|
|
|
}
|
|
|
|
|
log: remove lazy, remove unused interfaces, unexport methods (#28622)
This change
- Removes interface `log.Format`,
- Removes method `log.FormatFunc`,
- unexports `TerminalHandler.TerminalFormat` formatting methods (renamed to `TerminalHandler.format`)
- removes the notion of `log.Lazy` values
The lazy handler was useful in the old log package, since it
could defer the evaluation of costly attributes until later in the
log pipeline: thus, if the logging was done at 'Trace', we could
skip evaluation if logging only was set to 'Info'.
With the move to slog, this way of deferring evaluation is no longer
needed, since slog introduced 'Enabled': the caller can thus do
the evaluate-or-not decision at the callsite, which is much more
straight-forward than dealing with lazy reflect-based evaluation.
Also, lazy evaluation would not work with 'native' slog, as in, these
two statements would be evaluated differently:
```golang
log.Info("foo", "my lazy", lazyObj)
slog.Info("foo", "my lazy", lazyObj)
```
11 months ago
|
|
|
// FormatSlogValue formats a slog.Value for serialization to terminal.
|
|
|
|
func FormatSlogValue(v slog.Value, tmp []byte) (result []byte) {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
var value any
|
|
|
|
defer func() {
|
|
|
|
if err := recover(); err != nil {
|
|
|
|
if v := reflect.ValueOf(value); v.Kind() == reflect.Ptr && v.IsNil() {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
result = []byte("<nil>")
|
|
|
|
} else {
|
|
|
|
panic(err)
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}()
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
switch v.Kind() {
|
|
|
|
case slog.KindString:
|
|
|
|
return appendEscapeString(tmp, v.String())
|
log: remove lazy, remove unused interfaces, unexport methods (#28622)
This change
- Removes interface `log.Format`,
- Removes method `log.FormatFunc`,
- unexports `TerminalHandler.TerminalFormat` formatting methods (renamed to `TerminalHandler.format`)
- removes the notion of `log.Lazy` values
The lazy handler was useful in the old log package, since it
could defer the evaluation of costly attributes until later in the
log pipeline: thus, if the logging was done at 'Trace', we could
skip evaluation if logging only was set to 'Info'.
With the move to slog, this way of deferring evaluation is no longer
needed, since slog introduced 'Enabled': the caller can thus do
the evaluate-or-not decision at the callsite, which is much more
straight-forward than dealing with lazy reflect-based evaluation.
Also, lazy evaluation would not work with 'native' slog, as in, these
two statements would be evaluated differently:
```golang
log.Info("foo", "my lazy", lazyObj)
slog.Info("foo", "my lazy", lazyObj)
```
11 months ago
|
|
|
case slog.KindInt64: // All int-types (int8, int16 etc) wind up here
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendInt64(tmp, v.Int64())
|
log: remove lazy, remove unused interfaces, unexport methods (#28622)
This change
- Removes interface `log.Format`,
- Removes method `log.FormatFunc`,
- unexports `TerminalHandler.TerminalFormat` formatting methods (renamed to `TerminalHandler.format`)
- removes the notion of `log.Lazy` values
The lazy handler was useful in the old log package, since it
could defer the evaluation of costly attributes until later in the
log pipeline: thus, if the logging was done at 'Trace', we could
skip evaluation if logging only was set to 'Info'.
With the move to slog, this way of deferring evaluation is no longer
needed, since slog introduced 'Enabled': the caller can thus do
the evaluate-or-not decision at the callsite, which is much more
straight-forward than dealing with lazy reflect-based evaluation.
Also, lazy evaluation would not work with 'native' slog, as in, these
two statements would be evaluated differently:
```golang
log.Info("foo", "my lazy", lazyObj)
slog.Info("foo", "my lazy", lazyObj)
```
11 months ago
|
|
|
case slog.KindUint64: // All uint-types (uint8, uint16 etc) wind up here
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendUint64(tmp, v.Uint64(), false)
|
|
|
|
case slog.KindFloat64:
|
|
|
|
return strconv.AppendFloat(tmp, v.Float64(), floatFormat, 3, 64)
|
|
|
|
case slog.KindBool:
|
|
|
|
return strconv.AppendBool(tmp, v.Bool())
|
|
|
|
case slog.KindDuration:
|
|
|
|
value = v.Duration()
|
|
|
|
case slog.KindTime:
|
|
|
|
// Performance optimization: No need for escaping since the provided
|
|
|
|
// timeFormat doesn't have any escape characters, and escaping is
|
|
|
|
// expensive.
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return v.Time().AppendFormat(tmp, timeFormat)
|
|
|
|
default:
|
|
|
|
value = v.Any()
|
|
|
|
}
|
|
|
|
if value == nil {
|
|
|
|
return []byte("<nil>")
|
|
|
|
}
|
|
|
|
switch v := value.(type) {
|
log: remove lazy, remove unused interfaces, unexport methods (#28622)
This change
- Removes interface `log.Format`,
- Removes method `log.FormatFunc`,
- unexports `TerminalHandler.TerminalFormat` formatting methods (renamed to `TerminalHandler.format`)
- removes the notion of `log.Lazy` values
The lazy handler was useful in the old log package, since it
could defer the evaluation of costly attributes until later in the
log pipeline: thus, if the logging was done at 'Trace', we could
skip evaluation if logging only was set to 'Info'.
With the move to slog, this way of deferring evaluation is no longer
needed, since slog introduced 'Enabled': the caller can thus do
the evaluate-or-not decision at the callsite, which is much more
straight-forward than dealing with lazy reflect-based evaluation.
Also, lazy evaluation would not work with 'native' slog, as in, these
two statements would be evaluated differently:
```golang
log.Info("foo", "my lazy", lazyObj)
slog.Info("foo", "my lazy", lazyObj)
```
11 months ago
|
|
|
case *big.Int: // Need to be before fmt.Stringer-clause
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendBigInt(tmp, v)
|
log: remove lazy, remove unused interfaces, unexport methods (#28622)
This change
- Removes interface `log.Format`,
- Removes method `log.FormatFunc`,
- unexports `TerminalHandler.TerminalFormat` formatting methods (renamed to `TerminalHandler.format`)
- removes the notion of `log.Lazy` values
The lazy handler was useful in the old log package, since it
could defer the evaluation of costly attributes until later in the
log pipeline: thus, if the logging was done at 'Trace', we could
skip evaluation if logging only was set to 'Info'.
With the move to slog, this way of deferring evaluation is no longer
needed, since slog introduced 'Enabled': the caller can thus do
the evaluate-or-not decision at the callsite, which is much more
straight-forward than dealing with lazy reflect-based evaluation.
Also, lazy evaluation would not work with 'native' slog, as in, these
two statements would be evaluated differently:
```golang
log.Info("foo", "my lazy", lazyObj)
slog.Info("foo", "my lazy", lazyObj)
```
11 months ago
|
|
|
case *uint256.Int: // Need to be before fmt.Stringer-clause
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendU256(tmp, v)
|
|
|
|
case error:
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendEscapeString(tmp, v.Error())
|
|
|
|
case TerminalStringer:
|
log: remove lazy, remove unused interfaces, unexport methods (#28622)
This change
- Removes interface `log.Format`,
- Removes method `log.FormatFunc`,
- unexports `TerminalHandler.TerminalFormat` formatting methods (renamed to `TerminalHandler.format`)
- removes the notion of `log.Lazy` values
The lazy handler was useful in the old log package, since it
could defer the evaluation of costly attributes until later in the
log pipeline: thus, if the logging was done at 'Trace', we could
skip evaluation if logging only was set to 'Info'.
With the move to slog, this way of deferring evaluation is no longer
needed, since slog introduced 'Enabled': the caller can thus do
the evaluate-or-not decision at the callsite, which is much more
straight-forward than dealing with lazy reflect-based evaluation.
Also, lazy evaluation would not work with 'native' slog, as in, these
two statements would be evaluated differently:
```golang
log.Info("foo", "my lazy", lazyObj)
slog.Info("foo", "my lazy", lazyObj)
```
11 months ago
|
|
|
return appendEscapeString(tmp, v.TerminalString())
|
|
|
|
case fmt.Stringer:
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendEscapeString(tmp, v.String())
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
|
|
|
|
// We can use the 'tmp' as a scratch-buffer, to first format the
|
|
|
|
// value, and in a second step do escaping.
|
|
|
|
internal := fmt.Appendf(tmp, "%+v", value)
|
|
|
|
return appendEscapeString(tmp, string(internal))
|
|
|
|
}
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
// appendInt64 formats n with thousand separators and writes into buffer dst.
|
|
|
|
func appendInt64(dst []byte, n int64) []byte {
|
|
|
|
if n < 0 {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendUint64(dst, uint64(-n), true)
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendUint64(dst, uint64(n), false)
|
|
|
|
}
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
// appendUint64 formats n with thousand separators and writes into buffer dst.
|
|
|
|
func appendUint64(dst []byte, n uint64, neg bool) []byte {
|
|
|
|
// Small numbers are fine as is
|
|
|
|
if n < 100000 {
|
|
|
|
if neg {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return strconv.AppendInt(dst, -int64(n), 10)
|
|
|
|
} else {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return strconv.AppendInt(dst, int64(n), 10)
|
|
|
|
}
|
|
|
|
}
|
|
|
|
// Large numbers should be split
|
|
|
|
const maxLength = 26
|
|
|
|
|
|
|
|
var (
|
|
|
|
out = make([]byte, maxLength)
|
|
|
|
i = maxLength - 1
|
|
|
|
comma = 0
|
|
|
|
)
|
|
|
|
for ; n > 0; i-- {
|
|
|
|
if comma == 3 {
|
|
|
|
comma = 0
|
|
|
|
out[i] = ','
|
|
|
|
} else {
|
|
|
|
comma++
|
|
|
|
out[i] = '0' + byte(n%10)
|
|
|
|
n /= 10
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if neg {
|
|
|
|
out[i] = '-'
|
|
|
|
i--
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return append(dst, out[i+1:]...)
|
|
|
|
}
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
// FormatLogfmtUint64 formats n with thousand separators.
|
|
|
|
func FormatLogfmtUint64(n uint64) string {
|
|
|
|
return string(appendUint64(nil, n, false))
|
|
|
|
}
|
|
|
|
|
|
|
|
// appendBigInt formats n with thousand separators and writes to dst.
|
|
|
|
func appendBigInt(dst []byte, n *big.Int) []byte {
|
|
|
|
if n.IsUint64() {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendUint64(dst, n.Uint64(), false)
|
|
|
|
}
|
|
|
|
if n.IsInt64() {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendInt64(dst, n.Int64())
|
|
|
|
}
|
|
|
|
|
|
|
|
var (
|
|
|
|
text = n.String()
|
|
|
|
buf = make([]byte, len(text)+len(text)/3)
|
|
|
|
comma = 0
|
|
|
|
i = len(buf) - 1
|
|
|
|
)
|
|
|
|
for j := len(text) - 1; j >= 0; j, i = j-1, i-1 {
|
|
|
|
c := text[j]
|
|
|
|
|
|
|
|
switch {
|
|
|
|
case c == '-':
|
|
|
|
buf[i] = c
|
|
|
|
case comma == 3:
|
|
|
|
buf[i] = ','
|
|
|
|
i--
|
|
|
|
comma = 0
|
|
|
|
fallthrough
|
|
|
|
default:
|
|
|
|
buf[i] = c
|
|
|
|
comma++
|
|
|
|
}
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return append(dst, buf[i+1:]...)
|
|
|
|
}
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
// appendU256 formats n with thousand separators.
|
|
|
|
func appendU256(dst []byte, n *uint256.Int) []byte {
|
|
|
|
if n.IsUint64() {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
return appendUint64(dst, n.Uint64(), false)
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
res := []byte(n.PrettyDec(','))
|
|
|
|
return append(dst, res...)
|
|
|
|
}
|
|
|
|
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
// appendEscapeString writes the string s to the given writer, with
|
|
|
|
// escaping/quoting if needed.
|
|
|
|
func appendEscapeString(dst []byte, s string) []byte {
|
|
|
|
needsQuoting := false
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
needsEscaping := false
|
|
|
|
for _, r := range s {
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
// If it contains spaces or equal-sign, we need to quote it.
|
|
|
|
if r == ' ' || r == '=' {
|
|
|
|
needsQuoting = true
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
continue
|
|
|
|
}
|
|
|
|
// We need to escape it, if it contains
|
|
|
|
// - character " (0x22) and lower (except space)
|
|
|
|
// - characters above ~ (0x7E), plus equal-sign
|
|
|
|
if r <= '"' || r > '~' {
|
|
|
|
needsEscaping = true
|
|
|
|
break
|
|
|
|
}
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
if needsEscaping {
|
|
|
|
return strconv.AppendQuote(dst, s)
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
// No escaping needed, but we might have to place within quote-marks, in case
|
|
|
|
// it contained a space
|
|
|
|
if needsQuoting {
|
|
|
|
dst = append(dst, '"')
|
|
|
|
dst = append(dst, []byte(s)...)
|
|
|
|
return append(dst, '"')
|
|
|
|
}
|
|
|
|
return append(dst, []byte(s)...)
|
|
|
|
}
|
|
|
|
|
|
|
|
// escapeMessage checks if the provided string needs escaping/quoting, similarly
|
|
|
|
// to escapeString. The difference is that this method is more lenient: it allows
|
|
|
|
// for spaces and linebreaks to occur without needing quoting.
|
|
|
|
func escapeMessage(s string) string {
|
|
|
|
needsQuoting := false
|
|
|
|
for _, r := range s {
|
|
|
|
// Allow CR/LF/TAB. This is to make multi-line messages work.
|
|
|
|
if r == '\r' || r == '\n' || r == '\t' {
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
// We quote everything below <space> (0x20) and above~ (0x7E),
|
|
|
|
// plus equal-sign
|
|
|
|
if r < ' ' || r > '~' || r == '=' {
|
|
|
|
needsQuoting = true
|
|
|
|
break
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if !needsQuoting {
|
|
|
|
return s
|
|
|
|
}
|
|
|
|
return strconv.Quote(s)
|
|
|
|
}
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
|
|
|
|
// writeTimeTermFormat writes on the format "01-02|15:04:05.000"
|
|
|
|
func writeTimeTermFormat(buf *bytes.Buffer, t time.Time) {
|
|
|
|
_, month, day := t.Date()
|
|
|
|
writePosIntWidth(buf, int(month), 2)
|
|
|
|
buf.WriteByte('-')
|
|
|
|
writePosIntWidth(buf, day, 2)
|
|
|
|
buf.WriteByte('|')
|
|
|
|
hour, min, sec := t.Clock()
|
|
|
|
writePosIntWidth(buf, hour, 2)
|
|
|
|
buf.WriteByte(':')
|
|
|
|
writePosIntWidth(buf, min, 2)
|
|
|
|
buf.WriteByte(':')
|
|
|
|
writePosIntWidth(buf, sec, 2)
|
|
|
|
ns := t.Nanosecond()
|
|
|
|
buf.WriteByte('.')
|
|
|
|
writePosIntWidth(buf, ns/1e6, 3)
|
|
|
|
}
|
|
|
|
|
|
|
|
// writePosIntWidth writes non-negative integer i to the buffer, padded on the left
|
|
|
|
// by zeroes to the given width. Use a width of 0 to omit padding.
|
|
|
|
// Adapted from pkg.go.dev/log/slog/internal/buffer
|
slog: faster and less memory-consumption (#28621)
These changes improves the performance of the non-coloured terminal formatting, _quite a lot_.
```
name old time/op new time/op delta
TerminalHandler-8 10.2µs ±15% 5.4µs ± 9% -47.02% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
TerminalHandler-8 2.17kB ± 0% 0.40kB ± 0% -81.46% (p=0.008 n=5+5)
name old allocs/op new allocs/op delta
TerminalHandler-8 33.0 ± 0% 5.0 ± 0% -84.85% (p=0.008 n=5+5)
```
I tried to _somewhat_ organize the commits, but the it might still be a bit chaotic. Some core insights:
- The function `terminalHandler.Handl` uses a mutex, and writes all output immediately to 'upstream'. Thus, it can reuse a scratch-buffer every time.
- This buffer can be propagated internally, making all the internal formatters either write directly to it,
- OR, make use of the `tmp := buf.AvailableBuffer()` in some cases, where a byte buffer "extra capacity" can be temporarily used.
- The `slog` package uses `Attr` by value. It makes sense to minimize operating on them, since iterating / collecting into a new slice, iterating again etc causes copy-on-heap. Better to operate on them only once.
- If we want to do padding, it's better to copy from a constant `space`-buffer than to invoke `bytes.Repeat` every single time.
11 months ago
|
|
|
func writePosIntWidth(b *bytes.Buffer, i, width int) {
|
|
|
|
// Cheap integer to fixed-width decimal ASCII.
|
|
|
|
// Copied from log/log.go.
|
|
|
|
if i < 0 {
|
|
|
|
panic("negative int")
|
|
|
|
}
|
|
|
|
// Assemble decimal in reverse order.
|
|
|
|
var bb [20]byte
|
|
|
|
bp := len(bb) - 1
|
|
|
|
for i >= 10 || width > 1 {
|
|
|
|
width--
|
|
|
|
q := i / 10
|
|
|
|
bb[bp] = byte('0' + i - q*10)
|
|
|
|
bp--
|
|
|
|
i = q
|
|
|
|
}
|
|
|
|
// i < 10
|
|
|
|
bb[bp] = byte('0' + i)
|
|
|
|
b.Write(bb[bp:])
|
|
|
|
}
|