beenull/moby
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

daemon/config: support alternate (common) unicode encodings using a BOM

Browse source 
This is a pragmatic but impure choice, in order to better support the
default tools available on Windows Server, and reduce user confusion due
to otherwise inscrutable-to-the-uninitiated errors like the following:

> invalid character 'þ' looking for beginning of value
> invalid character 'ÿ' looking for beginning of value

While meaningful to those who are familiar with and are equipped to
diagnose encoding issues, these characters will be hidden when the file
is edited with a BOM-aware text editor, and further confuse the user.

Signed-off-by: Bjorn Neergaard <bneergaard@mirantis.com>
This commit is contained in:
Bjorn Neergaard 2023-01-10 13:35:46 -07:00
parent 1dcf7d5b03
commit d42495033e
No known key found for this signature in database
13 changed files with 3417 additions and 10 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

37
daemon/config/config.go
View file

@ -11,6 +11,9 @@ import (
"strings" "strings"
"sync" "sync"
"golang.org/x/text/encoding/unicode"
"golang.org/x/text/transform"
"github.com/containerd/containerd/runtime/v2/shim" "github.com/containerd/containerd/runtime/v2/shim"
"github.com/docker/docker/opts" "github.com/docker/docker/opts"
"github.com/docker/docker/pkg/authorization" "github.com/docker/docker/pkg/authorization"
@ -410,10 +413,36 @@ func getConflictFreeConfiguration(configFile string, flags *pflag.FlagSet) (*Con
return nil, err return nil, err
} }
// Strip the UTF-8 BOM if present ([RFC 8259] allows JSON implementations to optionally strip the BOM for // Decode the contents of the JSON file using a [byte order mark] if present, instead of assuming UTF-8 without BOM.
// interoperability; do so here as Notepad on older versions of Windows Server insists on a BOM). // The BOM, if present, will be used to determine the encoding. If no BOM is present, we will assume the default
// [RFC 8259]: https://tools.ietf.org/html/rfc8259#section-8.1 // and preferred encoding for JSON as defined by [RFC 8259], UTF-8 without BOM.
b = bytes.TrimPrefix(b, []byte("\xef\xbb\xbf")) //
// While JSON is normatively UTF-8 with no BOM, there are a couple of reasons to decode here:
// * UTF-8 with BOM is something that new implementations should avoid producing; however, [RFC 8259 Section 8.1]
// allows implementations to ignore the UTF-8 BOM when present for interoperability. Older versions of Notepad,
// the only text editor available out of the box on Windows Server, writes UTF-8 with a BOM by default.
// * The default encoding for [Windows PowerShell] is UTF-16 LE with BOM. While encodings in PowerShell can be a
// bit idiosyncratic, BOMs are still generally written. There is no support for selecting UTF-8 without a BOM as
// the encoding in Windows PowerShell, though some Cmdlets only write UTF-8 with no BOM. PowerShell Core
// introduces `utf8NoBOM` and makes it the default, but PowerShell Core is unlikely to be the implementation for
// a majority of Windows Server + PowerShell users.
// * While [RFC 8259 Section 8.1] asserts that software that is not part of a closed ecosystem or that crosses a
// network boundary should only support UTF-8, and should never write a BOM, it does acknowledge older versions
// of the standard, such as [RFC 7159 Section 8.1]. In the interest of pragmatism and easing pain for Windows
// users, we consider Windows tools such as Windows PowerShell and Notepad part of our ecosystem, and support
// the two most common encodings: UTF-16 LE with BOM, and UTF-8 with BOM, in addition to the standard UTF-8
// without BOM.
//
// [byte order mark]: https://www.unicode.org/faq/utf_bom.html#BOM
// [RFC 8259]: https://www.rfc-editor.org/rfc/rfc8259
// [RFC 8259 Section 8.1]: https://www.rfc-editor.org/rfc/rfc8259#section-8.1
// [RFC 7159 Section 8.1]: https://www.rfc-editor.org/rfc/rfc7159#section-8.1
// [Windows PowerShell]: https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.core/about/about_character_encoding?view=powershell-5.1
decoder := unicode.BOMOverride(unicode.UTF8.NewDecoder())
b, _, err = transform.Bytes(decoder, b)
if err != nil {
return nil, err
}
// Trim whitespace so that an empty config can be detected for an early return. // Trim whitespace so that an empty config can be detected for an early return.
b = bytes.TrimSpace(b) b = bytes.TrimSpace(b)

39
daemon/config/config_test.go
View file

@ -13,6 +13,8 @@ import (
"github.com/google/go-cmp/cmp/cmpopts" "github.com/google/go-cmp/cmp/cmpopts"
"github.com/imdario/mergo" "github.com/imdario/mergo"
"github.com/spf13/pflag" "github.com/spf13/pflag"
"golang.org/x/text/encoding"
"golang.org/x/text/encoding/unicode"
"gotest.tools/v3/assert" "gotest.tools/v3/assert"
is "gotest.tools/v3/assert/cmp" is "gotest.tools/v3/assert/cmp"
"gotest.tools/v3/skip" "gotest.tools/v3/skip"
@ -38,12 +40,41 @@ func TestDaemonBrokenConfiguration(t *testing.T) {
assert.ErrorContains(t, err, `invalid character ' ' in literal true`) assert.ErrorContains(t, err, `invalid character ' ' in literal true`)
} }
// TestDaemonConfigurationWithBOM ensures that the UTF-8 byte order mark is ignored when reading the configuration file. // TestDaemonConfigurationUnicodeVariations feeds various variations of Unicode into the JSON parser, ensuring that we
func TestDaemonConfigurationWithBOM(t *testing.T) { // respect a BOM and otherwise default to UTF-8.
configFile := makeConfigFile(t, "\xef\xbb\xbf{\"debug\": true}") func TestDaemonConfigurationUnicodeVariations(t *testing.T) {
jsonData := `{"debug": true}`
_, err := MergeDaemonConfigurations(&Config{}, nil, configFile) testCases := []struct {
name string
encoding encoding.Encoding
}{
{
name: "UTF-8",
encoding: unicode.UTF8,
},
{
name: "UTF-8 (with BOM)",
encoding: unicode.UTF8BOM,
},
{
name: "UTF-16 (BE with BOM)",
encoding: unicode.UTF16(unicode.BigEndian, unicode.UseBOM),
},
{
name: "UTF-16 (LE with BOM)",
encoding: unicode.UTF16(unicode.LittleEndian, unicode.UseBOM),
},
}
for _, tc := range testCases {
t.Run(tc.name, func(t *testing.T) {
encodedJson, err := tc.encoding.NewEncoder().String(jsonData)
assert.NilError(t, err) assert.NilError(t, err)
configFile := makeConfigFile(t, encodedJson)
_, err = MergeDaemonConfigurations(&Config{}, nil, configFile)
assert.NilError(t, err)
})
}
} }
func TestFindConfigurationConflicts(t *testing.T) { func TestFindConfigurationConflicts(t *testing.T) {

2
vendor.mod
View file

@ -90,6 +90,7 @@ require (
golang.org/x/net v0.4.0 golang.org/x/net v0.4.0
golang.org/x/sync v0.1.0 golang.org/x/sync v0.1.0
golang.org/x/sys v0.3.0 golang.org/x/sys v0.3.0
golang.org/x/text v0.5.0
golang.org/x/time v0.1.0 golang.org/x/time v0.1.0
google.golang.org/genproto v0.0.0-20220706185917-7780775163c4 google.golang.org/genproto v0.0.0-20220706185917-7780775163c4
google.golang.org/grpc v1.50.1 google.golang.org/grpc v1.50.1
@ -170,7 +171,6 @@ require (
go.uber.org/zap v1.21.0 // indirect go.uber.org/zap v1.21.0 // indirect
golang.org/x/crypto v0.2.0 // indirect golang.org/x/crypto v0.2.0 // indirect
golang.org/x/oauth2 v0.1.0 // indirect golang.org/x/oauth2 v0.1.0 // indirect
golang.org/x/text v0.5.0 // indirect
google.golang.org/api v0.93.0 // indirect google.golang.org/api v0.93.0 // indirect
google.golang.org/appengine v1.6.7 // indirect google.golang.org/appengine v1.6.7 // indirect
google.golang.org/protobuf v1.28.1 // indirect google.golang.org/protobuf v1.28.1 // indirect

335
vendor/golang.org/x/text/encoding/encoding.go generated vendored Normal file
View file

@ -0,0 +1,335 @@
// Copyright 2013 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package encoding defines an interface for character encodings, such as Shift
// JIS and Windows 1252, that can convert to and from UTF-8.
//
// Encoding implementations are provided in other packages, such as
// golang.org/x/text/encoding/charmap and
// golang.org/x/text/encoding/japanese.
package encoding // import "golang.org/x/text/encoding"
import (
"errors"
"io"
"strconv"
"unicode/utf8"
"golang.org/x/text/encoding/internal/identifier"
"golang.org/x/text/transform"
)
// TODO:
// - There seems to be some inconsistency in when decoders return errors
// and when not. Also documentation seems to suggest they shouldn't return
// errors at all (except for UTF-16).
// - Encoders seem to rely on or at least benefit from the input being in NFC
// normal form. Perhaps add an example how users could prepare their output.
// Encoding is a character set encoding that can be transformed to and from
// UTF-8.
type Encoding interface {
// NewDecoder returns a Decoder.
NewDecoder() *Decoder
// NewEncoder returns an Encoder.
NewEncoder() *Encoder
}
// A Decoder converts bytes to UTF-8. It implements transform.Transformer.
//
// Transforming source bytes that are not of that encoding will not result in an
// error per se. Each byte that cannot be transcoded will be represented in the
// output by the UTF-8 encoding of '\uFFFD', the replacement rune.
type Decoder struct {
transform.Transformer
// This forces external creators of Decoders to use names in struct
// initializers, allowing for future extendibility without having to break
// code.
_ struct{}
}
// Bytes converts the given encoded bytes to UTF-8. It returns the converted
// bytes or nil, err if any error occurred.
func (d *Decoder) Bytes(b []byte) ([]byte, error) {
b, _, err := transform.Bytes(d, b)
if err != nil {
return nil, err
}
return b, nil
}
// String converts the given encoded string to UTF-8. It returns the converted
// string or "", err if any error occurred.
func (d *Decoder) String(s string) (string, error) {
s, _, err := transform.String(d, s)
if err != nil {
return "", err
}
return s, nil
}
// Reader wraps another Reader to decode its bytes.
//
// The Decoder may not be used for any other operation as long as the returned
// Reader is in use.
func (d *Decoder) Reader(r io.Reader) io.Reader {
return transform.NewReader(r, d)
}
// An Encoder converts bytes from UTF-8. It implements transform.Transformer.
//
// Each rune that cannot be transcoded will result in an error. In this case,
// the transform will consume all source byte up to, not including the offending
// rune. Transforming source bytes that are not valid UTF-8 will be replaced by
// `\uFFFD`. To return early with an error instead, use transform.Chain to
// preprocess the data with a UTF8Validator.
type Encoder struct {
transform.Transformer
// This forces external creators of Encoders to use names in struct
// initializers, allowing for future extendibility without having to break
// code.
_ struct{}
}
// Bytes converts bytes from UTF-8. It returns the converted bytes or nil, err if
// any error occurred.
func (e *Encoder) Bytes(b []byte) ([]byte, error) {
b, _, err := transform.Bytes(e, b)
if err != nil {
return nil, err
}
return b, nil
}
// String converts a string from UTF-8. It returns the converted string or
// "", err if any error occurred.
func (e *Encoder) String(s string) (string, error) {
s, _, err := transform.String(e, s)
if err != nil {
return "", err
}
return s, nil
}
// Writer wraps another Writer to encode its UTF-8 output.
//
// The Encoder may not be used for any other operation as long as the returned
// Writer is in use.
func (e *Encoder) Writer(w io.Writer) io.Writer {
return transform.NewWriter(w, e)
}
// ASCIISub is the ASCII substitute character, as recommended by
// https://unicode.org/reports/tr36/#Text_Comparison
const ASCIISub = '\x1a'
// Nop is the nop encoding. Its transformed bytes are the same as the source
// bytes; it does not replace invalid UTF-8 sequences.
var Nop Encoding = nop{}
type nop struct{}
func (nop) NewDecoder() *Decoder {
return &Decoder{Transformer: transform.Nop}
}
func (nop) NewEncoder() *Encoder {
return &Encoder{Transformer: transform.Nop}
}
// Replacement is the replacement encoding. Decoding from the replacement
// encoding yields a single '\uFFFD' replacement rune. Encoding from UTF-8 to
// the replacement encoding yields the same as the source bytes except that
// invalid UTF-8 is converted to '\uFFFD'.
//
// It is defined at http://encoding.spec.whatwg.org/#replacement
var Replacement Encoding = replacement{}
type replacement struct{}
func (replacement) NewDecoder() *Decoder {
return &Decoder{Transformer: replacementDecoder{}}
}
func (replacement) NewEncoder() *Encoder {
return &Encoder{Transformer: replacementEncoder{}}
}
func (replacement) ID() (mib identifier.MIB, other string) {
return identifier.Replacement, ""
}
type replacementDecoder struct{ transform.NopResetter }
func (replacementDecoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
if len(dst) < 3 {
return 0, 0, transform.ErrShortDst
}
if atEOF {
const fffd = "\ufffd"
dst[0] = fffd[0]
dst[1] = fffd[1]
dst[2] = fffd[2]
nDst = 3
}
return nDst, len(src), nil
}
type replacementEncoder struct{ transform.NopResetter }
func (replacementEncoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
r, size := rune(0), 0
for ; nSrc < len(src); nSrc += size {
r = rune(src[nSrc])
// Decode a 1-byte rune.
if r < utf8.RuneSelf {
size = 1
} else {
// Decode a multi-byte rune.
r, size = utf8.DecodeRune(src[nSrc:])
if size == 1 {
// All valid runes of size 1 (those below utf8.RuneSelf) were
// handled above. We have invalid UTF-8 or we haven't seen the
// full character yet.
if !atEOF && !utf8.FullRune(src[nSrc:]) {
err = transform.ErrShortSrc
break
}
r = '\ufffd'
}
}
if nDst+utf8.RuneLen(r) > len(dst) {
err = transform.ErrShortDst
break
}
nDst += utf8.EncodeRune(dst[nDst:], r)
}
return nDst, nSrc, err
}
// HTMLEscapeUnsupported wraps encoders to replace source runes outside the
// repertoire of the destination encoding with HTML escape sequences.
//
// This wrapper exists to comply to URL and HTML forms requiring a
// non-terminating legacy encoder. The produced sequences may lead to data
// loss as they are indistinguishable from legitimate input. To avoid this
// issue, use UTF-8 encodings whenever possible.
func HTMLEscapeUnsupported(e *Encoder) *Encoder {
return &Encoder{Transformer: &errorHandler{e, errorToHTML}}
}
// ReplaceUnsupported wraps encoders to replace source runes outside the
// repertoire of the destination encoding with an encoding-specific
// replacement.
//
// This wrapper is only provided for backwards compatibility and legacy
// handling. Its use is strongly discouraged. Use UTF-8 whenever possible.
func ReplaceUnsupported(e *Encoder) *Encoder {
return &Encoder{Transformer: &errorHandler{e, errorToReplacement}}
}
type errorHandler struct {
*Encoder
handler func(dst []byte, r rune, err repertoireError) (n int, ok bool)
}
// TODO: consider making this error public in some form.
type repertoireError interface {
Replacement() byte
}
func (h errorHandler) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
nDst, nSrc, err = h.Transformer.Transform(dst, src, atEOF)
for err != nil {
rerr, ok := err.(repertoireError)
if !ok {
return nDst, nSrc, err
}
r, sz := utf8.DecodeRune(src[nSrc:])
n, ok := h.handler(dst[nDst:], r, rerr)
if !ok {
return nDst, nSrc, transform.ErrShortDst
}
err = nil
nDst += n
if nSrc += sz; nSrc < len(src) {
var dn, sn int
dn, sn, err = h.Transformer.Transform(dst[nDst:], src[nSrc:], atEOF)
nDst += dn
nSrc += sn
}
}
return nDst, nSrc, err
}
func errorToHTML(dst []byte, r rune, err repertoireError) (n int, ok bool) {
buf := [8]byte{}
b := strconv.AppendUint(buf[:0], uint64(r), 10)
if n = len(b) + len("&#;"); n >= len(dst) {
return 0, false
}
dst[0] = '&'
dst[1] = '#'
dst[copy(dst[2:], b)+2] = ';'
return n, true
}
func errorToReplacement(dst []byte, r rune, err repertoireError) (n int, ok bool) {
if len(dst) == 0 {
return 0, false
}
dst[0] = err.Replacement()
return 1, true
}
// ErrInvalidUTF8 means that a transformer encountered invalid UTF-8.
var ErrInvalidUTF8 = errors.New("encoding: invalid UTF-8")
// UTF8Validator is a transformer that returns ErrInvalidUTF8 on the first
// input byte that is not valid UTF-8.
var UTF8Validator transform.Transformer = utf8Validator{}
type utf8Validator struct{ transform.NopResetter }
func (utf8Validator) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
n := len(src)
if n > len(dst) {
n = len(dst)
}
for i := 0; i < n; {
if c := src[i]; c < utf8.RuneSelf {
dst[i] = c
i++
continue
}
_, size := utf8.DecodeRune(src[i:])
if size == 1 {
// All valid runes of size 1 (those below utf8.RuneSelf) were
// handled above. We have invalid UTF-8 or we haven't seen the
// full character yet.
err = ErrInvalidUTF8
if !atEOF && !utf8.FullRune(src[i:]) {
err = transform.ErrShortSrc
}
return i, i, err
}
if i+size > len(dst) {
return i, i, transform.ErrShortDst
}
for ; size > 0; size-- {
dst[i] = src[i]
i++
}
}
if len(src) > len(dst) {
err = transform.ErrShortDst
}
return n, n, err
}

81
vendor/golang.org/x/text/encoding/internal/identifier/identifier.go generated vendored Normal file
View file

@ -0,0 +1,81 @@
// Copyright 2015 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:generate go run gen.go
// Package identifier defines the contract between implementations of Encoding
// and Index by defining identifiers that uniquely identify standardized coded
// character sets (CCS) and character encoding schemes (CES), which we will
// together refer to as encodings, for which Encoding implementations provide
// converters to and from UTF-8. This package is typically only of concern to
// implementers of Indexes and Encodings.
//
// One part of the identifier is the MIB code, which is defined by IANA and
// uniquely identifies a CCS or CES. Each code is associated with data that
// references authorities, official documentation as well as aliases and MIME
// names.
//
// Not all CESs are covered by the IANA registry. The "other" string that is
// returned by ID can be used to identify other character sets or versions of
// existing ones.
//
// It is recommended that each package that provides a set of Encodings provide
// the All and Common variables to reference all supported encodings and
// commonly used subset. This allows Index implementations to include all
// available encodings without explicitly referencing or knowing about them.
package identifier
// Note: this package is internal, but could be made public if there is a need
// for writing third-party Indexes and Encodings.
// References:
// - http://source.icu-project.org/repos/icu/icu/trunk/source/data/mappings/convrtrs.txt
// - http://www.iana.org/assignments/character-sets/character-sets.xhtml
// - http://www.iana.org/assignments/ianacharset-mib/ianacharset-mib
// - http://www.ietf.org/rfc/rfc2978.txt
// - https://www.unicode.org/reports/tr22/
// - http://www.w3.org/TR/encoding/
// - https://encoding.spec.whatwg.org/
// - https://encoding.spec.whatwg.org/encodings.json
// - https://tools.ietf.org/html/rfc6657#section-5
// Interface can be implemented by Encodings to define the CCS or CES for which
// it implements conversions.
type Interface interface {
// ID returns an encoding identifier. Exactly one of the mib and other
// values should be non-zero.
//
// In the usual case it is only necessary to indicate the MIB code. The
// other string can be used to specify encodings for which there is no MIB,
// such as "x-mac-dingbat".
//
// The other string may only contain the characters a-z, A-Z, 0-9, - and _.
ID() (mib MIB, other string)
// NOTE: the restrictions on the encoding are to allow extending the syntax
// with additional information such as versions, vendors and other variants.
}
// A MIB identifies an encoding. It is derived from the IANA MIB codes and adds
// some identifiers for some encodings that are not covered by the IANA
// standard.
//
// See http://www.iana.org/assignments/ianacharset-mib.
type MIB uint16
// These additional MIB types are not defined in IANA. They are added because
// they are common and defined within the text repo.
const (
// Unofficial marks the start of encodings not registered by IANA.
Unofficial MIB = 10000 + iota
// Replacement is the WhatWG replacement encoding.
Replacement
// XUserDefined is the code for x-user-defined.
XUserDefined
// MacintoshCyrillic is the code for x-mac-cyrillic.
MacintoshCyrillic
)

1627
vendor/golang.org/x/text/encoding/internal/identifier/mib.go generated vendored Normal file
View file

File diff suppressed because it is too large Load diff

75
vendor/golang.org/x/text/encoding/internal/internal.go generated vendored Normal file
View file

@ -0,0 +1,75 @@
// Copyright 2015 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package internal contains code that is shared among encoding implementations.
package internal
import (
"golang.org/x/text/encoding"
"golang.org/x/text/encoding/internal/identifier"
"golang.org/x/text/transform"
)
// Encoding is an implementation of the Encoding interface that adds the String
// and ID methods to an existing encoding.
type Encoding struct {
encoding.Encoding
Name string
MIB identifier.MIB
}
// _ verifies that Encoding implements identifier.Interface.
var _ identifier.Interface = (*Encoding)(nil)
func (e *Encoding) String() string {
return e.Name
}
func (e *Encoding) ID() (mib identifier.MIB, other string) {
return e.MIB, ""
}
// SimpleEncoding is an Encoding that combines two Transformers.
type SimpleEncoding struct {
Decoder transform.Transformer
Encoder transform.Transformer
}
func (e *SimpleEncoding) NewDecoder() *encoding.Decoder {
return &encoding.Decoder{Transformer: e.Decoder}
}
func (e *SimpleEncoding) NewEncoder() *encoding.Encoder {
return &encoding.Encoder{Transformer: e.Encoder}
}
// FuncEncoding is an Encoding that combines two functions returning a new
// Transformer.
type FuncEncoding struct {
Decoder func() transform.Transformer
Encoder func() transform.Transformer
}
func (e FuncEncoding) NewDecoder() *encoding.Decoder {
return &encoding.Decoder{Transformer: e.Decoder()}
}
func (e FuncEncoding) NewEncoder() *encoding.Encoder {
return &encoding.Encoder{Transformer: e.Encoder()}
}
// A RepertoireError indicates a rune is not in the repertoire of a destination
// encoding. It is associated with an encoding-specific suggested replacement
// byte.
type RepertoireError byte
// Error implements the error interrface.
func (r RepertoireError) Error() string {
return "encoding: rune not supported by encoding."
}
// Replacement returns the replacement string associated with this error.
func (r RepertoireError) Replacement() byte { return byte(r) }
var ErrASCIIReplacement = RepertoireError(encoding.ASCIISub)

82
vendor/golang.org/x/text/encoding/unicode/override.go generated vendored Normal file
View file

@ -0,0 +1,82 @@
// Copyright 2015 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package unicode
import (
"golang.org/x/text/transform"
)
// BOMOverride returns a new decoder transformer that is identical to fallback,
// except that the presence of a Byte Order Mark at the start of the input
// causes it to switch to the corresponding Unicode decoding. It will only
// consider BOMs for UTF-8, UTF-16BE, and UTF-16LE.
//
// This differs from using ExpectBOM by allowing a BOM to switch to UTF-8, not
// just UTF-16 variants, and allowing falling back to any encoding scheme.
//
// This technique is recommended by the W3C for use in HTML 5: "For
// compatibility with deployed content, the byte order mark (also known as BOM)
// is considered more authoritative than anything else."
// http://www.w3.org/TR/encoding/#specification-hooks
//
// Using BOMOverride is mostly intended for use cases where the first characters
// of a fallback encoding are known to not be a BOM, for example, for valid HTML
// and most encodings.
func BOMOverride(fallback transform.Transformer) transform.Transformer {
// TODO: possibly allow a variadic argument of unicode encodings to allow
// specifying details of which fallbacks are supported as well as
// specifying the details of the implementations. This would also allow for
// support for UTF-32, which should not be supported by default.
return &bomOverride{fallback: fallback}
}
type bomOverride struct {
fallback transform.Transformer
current transform.Transformer
}
func (d *bomOverride) Reset() {
d.current = nil
d.fallback.Reset()
}
var (
// TODO: we could use decode functions here, instead of allocating a new
// decoder on every NewDecoder as IgnoreBOM decoders can be stateless.
utf16le = UTF16(LittleEndian, IgnoreBOM)
utf16be = UTF16(BigEndian, IgnoreBOM)
)
const utf8BOM = "\ufeff"
func (d *bomOverride) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
if d.current != nil {
return d.current.Transform(dst, src, atEOF)
}
if len(src) < 3 && !atEOF {
return 0, 0, transform.ErrShortSrc
}
d.current = d.fallback
bomSize := 0
if len(src) >= 2 {
if src[0] == 0xFF && src[1] == 0xFE {
d.current = utf16le.NewDecoder()
bomSize = 2
} else if src[0] == 0xFE && src[1] == 0xFF {
d.current = utf16be.NewDecoder()
bomSize = 2
} else if len(src) >= 3 &&
src[0] == utf8BOM[0] &&
src[1] == utf8BOM[1] &&
src[2] == utf8BOM[2] {
d.current = transform.Nop
bomSize = 3
}
}
if bomSize < len(src) {
nDst, nSrc, err = d.current.Transform(dst, src[bomSize:], atEOF)
}
return nDst, nSrc + bomSize, err
}

512
vendor/golang.org/x/text/encoding/unicode/unicode.go generated vendored Normal file
View file

@ -0,0 +1,512 @@
// Copyright 2013 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package unicode provides Unicode encodings such as UTF-16.
package unicode // import "golang.org/x/text/encoding/unicode"
import (
"bytes"
"errors"
"unicode/utf16"
"unicode/utf8"
"golang.org/x/text/encoding"
"golang.org/x/text/encoding/internal"
"golang.org/x/text/encoding/internal/identifier"
"golang.org/x/text/internal/utf8internal"
"golang.org/x/text/runes"
"golang.org/x/text/transform"
)
// TODO: I think the Transformers really should return errors on unmatched
// surrogate pairs and odd numbers of bytes. This is not required by RFC 2781,
// which leaves it open, but is suggested by WhatWG. It will allow for all error
// modes as defined by WhatWG: fatal, HTML and Replacement. This would require
// the introduction of some kind of error type for conveying the erroneous code
// point.
// UTF8 is the UTF-8 encoding. It neither removes nor adds byte order marks.
var UTF8 encoding.Encoding = utf8enc
// UTF8BOM is an UTF-8 encoding where the decoder strips a leading byte order
// mark while the encoder adds one.
//
// Some editors add a byte order mark as a signature to UTF-8 files. Although
// the byte order mark is not useful for detecting byte order in UTF-8, it is
// sometimes used as a convention to mark UTF-8-encoded files. This relies on
// the observation that the UTF-8 byte order mark is either an illegal or at
// least very unlikely sequence in any other character encoding.
var UTF8BOM encoding.Encoding = utf8bomEncoding{}
type utf8bomEncoding struct{}
func (utf8bomEncoding) String() string {
return "UTF-8-BOM"
}
func (utf8bomEncoding) ID() (identifier.MIB, string) {
return identifier.Unofficial, "x-utf8bom"
}
func (utf8bomEncoding) NewEncoder() *encoding.Encoder {
return &encoding.Encoder{
Transformer: &utf8bomEncoder{t: runes.ReplaceIllFormed()},
}
}
func (utf8bomEncoding) NewDecoder() *encoding.Decoder {
return &encoding.Decoder{Transformer: &utf8bomDecoder{}}
}
var utf8enc = &internal.Encoding{
&internal.SimpleEncoding{utf8Decoder{}, runes.ReplaceIllFormed()},
"UTF-8",
identifier.UTF8,
}
type utf8bomDecoder struct {
checked bool
}
func (t *utf8bomDecoder) Reset() {
t.checked = false
}
func (t *utf8bomDecoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
if !t.checked {
if !atEOF && len(src) < len(utf8BOM) {
if len(src) == 0 {
return 0, 0, nil
}
return 0, 0, transform.ErrShortSrc
}
if bytes.HasPrefix(src, []byte(utf8BOM)) {
nSrc += len(utf8BOM)
src = src[len(utf8BOM):]
}
t.checked = true
}
nDst, n, err := utf8Decoder.Transform(utf8Decoder{}, dst[nDst:], src, atEOF)
nSrc += n
return nDst, nSrc, err
}
type utf8bomEncoder struct {
written bool
t transform.Transformer
}
func (t *utf8bomEncoder) Reset() {
t.written = false
t.t.Reset()
}
func (t *utf8bomEncoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
if !t.written {
if len(dst) < len(utf8BOM) {
return nDst, 0, transform.ErrShortDst
}
nDst = copy(dst, utf8BOM)
t.written = true
}
n, nSrc, err := utf8Decoder.Transform(utf8Decoder{}, dst[nDst:], src, atEOF)
nDst += n
return nDst, nSrc, err
}
type utf8Decoder struct{ transform.NopResetter }
func (utf8Decoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
var pSrc int // point from which to start copy in src
var accept utf8internal.AcceptRange
// The decoder can only make the input larger, not smaller.
n := len(src)
if len(dst) < n {
err = transform.ErrShortDst
n = len(dst)
atEOF = false
}
for nSrc < n {
c := src[nSrc]
if c < utf8.RuneSelf {
nSrc++
continue
}
first := utf8internal.First[c]
size := int(first & utf8internal.SizeMask)
if first == utf8internal.FirstInvalid {
goto handleInvalid // invalid starter byte
}
accept = utf8internal.AcceptRanges[first>>utf8internal.AcceptShift]
if nSrc+size > n {
if !atEOF {
// We may stop earlier than necessary here if the short sequence
// has invalid bytes. Not checking for this simplifies the code
// and may avoid duplicate computations in certain conditions.
if err == nil {
err = transform.ErrShortSrc
}
break
}
// Determine the maximal subpart of an ill-formed subsequence.
switch {
case nSrc+1 >= n || src[nSrc+1] < accept.Lo || accept.Hi < src[nSrc+1]:
size = 1
case nSrc+2 >= n || src[nSrc+2] < utf8internal.LoCB || utf8internal.HiCB < src[nSrc+2]:
size = 2
default:
size = 3 // As we are short, the maximum is 3.
}
goto handleInvalid
}
if c = src[nSrc+1]; c < accept.Lo || accept.Hi < c {
size = 1
goto handleInvalid // invalid continuation byte
} else if size == 2 {
} else if c = src[nSrc+2]; c < utf8internal.LoCB || utf8internal.HiCB < c {
size = 2
goto handleInvalid // invalid continuation byte
} else if size == 3 {
} else if c = src[nSrc+3]; c < utf8internal.LoCB || utf8internal.HiCB < c {
size = 3
goto handleInvalid // invalid continuation byte
}
nSrc += size
continue
handleInvalid:
// Copy the scanned input so far.
nDst += copy(dst[nDst:], src[pSrc:nSrc])
// Append RuneError to the destination.
const runeError = "\ufffd"
if nDst+len(runeError) > len(dst) {
return nDst, nSrc, transform.ErrShortDst
}
nDst += copy(dst[nDst:], runeError)
// Skip the maximal subpart of an ill-formed subsequence according to
// the W3C standard way instead of the Go way. This Transform is
// probably the only place in the text repo where it is warranted.
nSrc += size
pSrc = nSrc
// Recompute the maximum source length.
if sz := len(dst) - nDst; sz < len(src)-nSrc {
err = transform.ErrShortDst
n = nSrc + sz
atEOF = false
}
}
return nDst + copy(dst[nDst:], src[pSrc:nSrc]), nSrc, err
}
// UTF16 returns a UTF-16 Encoding for the given default endianness and byte
// order mark (BOM) policy.
//
// When decoding from UTF-16 to UTF-8, if the BOMPolicy is IgnoreBOM then
// neither BOMs U+FEFF nor noncharacters U+FFFE in the input stream will affect
// the endianness used for decoding, and will instead be output as their
// standard UTF-8 encodings: "\xef\xbb\xbf" and "\xef\xbf\xbe". If the BOMPolicy
// is UseBOM or ExpectBOM a staring BOM is not written to the UTF-8 output.
// Instead, it overrides the default endianness e for the remainder of the
// transformation. Any subsequent BOMs U+FEFF or noncharacters U+FFFE will not
// affect the endianness used, and will instead be output as their standard
// UTF-8 encodings. For UseBOM, if there is no starting BOM, it will proceed
// with the default Endianness. For ExpectBOM, in that case, the transformation
// will return early with an ErrMissingBOM error.
//
// When encoding from UTF-8 to UTF-16, a BOM will be inserted at the start of
// the output if the BOMPolicy is UseBOM or ExpectBOM. Otherwise, a BOM will not
// be inserted. The UTF-8 input does not need to contain a BOM.
//
// There is no concept of a 'native' endianness. If the UTF-16 data is produced
// and consumed in a greater context that implies a certain endianness, use
// IgnoreBOM. Otherwise, use ExpectBOM and always produce and consume a BOM.
//
// In the language of https://www.unicode.org/faq/utf_bom.html#bom10, IgnoreBOM
// corresponds to "Where the precise type of the data stream is known... the
// BOM should not be used" and ExpectBOM corresponds to "A particular
// protocol... may require use of the BOM".
func UTF16(e Endianness, b BOMPolicy) encoding.Encoding {
return utf16Encoding{config{e, b}, mibValue[e][b&bomMask]}
}
// mibValue maps Endianness and BOMPolicy settings to MIB constants. Note that
// some configurations map to the same MIB identifier. RFC 2781 has requirements
// and recommendations. Some of the "configurations" are merely recommendations,
// so multiple configurations could match.
var mibValue = map[Endianness][numBOMValues]identifier.MIB{
BigEndian: [numBOMValues]identifier.MIB{
IgnoreBOM: identifier.UTF16BE,
UseBOM: identifier.UTF16, // BigEnding default is preferred by RFC 2781.
// TODO: acceptBOM | strictBOM would map to UTF16BE as well.
},
LittleEndian: [numBOMValues]identifier.MIB{
IgnoreBOM: identifier.UTF16LE,
UseBOM: identifier.UTF16, // LittleEndian default is allowed and preferred on Windows.
// TODO: acceptBOM | strictBOM would map to UTF16LE as well.
},
// ExpectBOM is not widely used and has no valid MIB identifier.
}
// All lists a configuration for each IANA-defined UTF-16 variant.
var All = []encoding.Encoding{
UTF8,
UTF16(BigEndian, UseBOM),
UTF16(BigEndian, IgnoreBOM),
UTF16(LittleEndian, IgnoreBOM),
}
// BOMPolicy is a UTF-16 encoding's byte order mark policy.
type BOMPolicy uint8
const (
writeBOM BOMPolicy = 0x01
acceptBOM BOMPolicy = 0x02
requireBOM BOMPolicy = 0x04
bomMask BOMPolicy = 0x07
// HACK: numBOMValues == 8 triggers a bug in the 1.4 compiler (cannot have a
// map of an array of length 8 of a type that is also used as a key or value
// in another map). See golang.org/issue/11354.
// TODO: consider changing this value back to 8 if the use of 1.4.* has
// been minimized.
numBOMValues = 8 + 1
// IgnoreBOM means to ignore any byte order marks.
IgnoreBOM BOMPolicy = 0
// Common and RFC 2781-compliant interpretation for UTF-16BE/LE.
// UseBOM means that the UTF-16 form may start with a byte order mark, which
// will be used to override the default encoding.
UseBOM BOMPolicy = writeBOM | acceptBOM
// Common and RFC 2781-compliant interpretation for UTF-16.
// ExpectBOM means that the UTF-16 form must start with a byte order mark,
// which will be used to override the default encoding.
ExpectBOM BOMPolicy = writeBOM | acceptBOM | requireBOM
// Used in Java as Unicode (not to be confused with Java's UTF-16) and
// ICU's UTF-16,version=1. Not compliant with RFC 2781.
// TODO (maybe): strictBOM: BOM must match Endianness. This would allow:
// - UTF-16(B|L)E,version=1: writeBOM | acceptBOM | requireBOM | strictBOM
// (UnicodeBig and UnicodeLittle in Java)
// - RFC 2781-compliant, but less common interpretation for UTF-16(B|L)E:
// acceptBOM | strictBOM (e.g. assigned to CheckBOM).
// This addition would be consistent with supporting ExpectBOM.
)
// Endianness is a UTF-16 encoding's default endianness.
type Endianness bool
const (
// BigEndian is UTF-16BE.
BigEndian Endianness = false
// LittleEndian is UTF-16LE.
LittleEndian Endianness = true
)
// ErrMissingBOM means that decoding UTF-16 input with ExpectBOM did not find a
// starting byte order mark.
var ErrMissingBOM = errors.New("encoding: missing byte order mark")
type utf16Encoding struct {
config
mib identifier.MIB
}
type config struct {
endianness Endianness
bomPolicy BOMPolicy
}
func (u utf16Encoding) NewDecoder() *encoding.Decoder {
return &encoding.Decoder{Transformer: &utf16Decoder{
initial: u.config,
current: u.config,
}}
}
func (u utf16Encoding) NewEncoder() *encoding.Encoder {
return &encoding.Encoder{Transformer: &utf16Encoder{
endianness: u.endianness,
initialBOMPolicy: u.bomPolicy,
currentBOMPolicy: u.bomPolicy,
}}
}
func (u utf16Encoding) ID() (mib identifier.MIB, other string) {
return u.mib, ""
}
func (u utf16Encoding) String() string {
e, b := "B", ""
if u.endianness == LittleEndian {
e = "L"
}
switch u.bomPolicy {
case ExpectBOM:
b = "Expect"
case UseBOM:
b = "Use"
case IgnoreBOM:
b = "Ignore"
}
return "UTF-16" + e + "E (" + b + " BOM)"
}
type utf16Decoder struct {
initial config
current config
}
func (u *utf16Decoder) Reset() {
u.current = u.initial
}
func (u *utf16Decoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
if len(src) < 2 && atEOF && u.current.bomPolicy&requireBOM != 0 {
return 0, 0, ErrMissingBOM
}
if len(src) == 0 {
return 0, 0, nil
}
if len(src) >= 2 && u.current.bomPolicy&acceptBOM != 0 {
switch {
case src[0] == 0xfe && src[1] == 0xff:
u.current.endianness = BigEndian
nSrc = 2
case src[0] == 0xff && src[1] == 0xfe:
u.current.endianness = LittleEndian
nSrc = 2
default:
if u.current.bomPolicy&requireBOM != 0 {
return 0, 0, ErrMissingBOM
}
}
u.current.bomPolicy = IgnoreBOM
}
var r rune
var dSize, sSize int
for nSrc < len(src) {
if nSrc+1 < len(src) {
x := uint16(src[nSrc+0])<<8 | uint16(src[nSrc+1])
if u.current.endianness == LittleEndian {
x = x>>8 | x<<8
}
r, sSize = rune(x), 2
if utf16.IsSurrogate(r) {
if nSrc+3 < len(src) {
x = uint16(src[nSrc+2])<<8 | uint16(src[nSrc+3])
if u.current.endianness == LittleEndian {
x = x>>8 | x<<8
}
// Save for next iteration if it is not a high surrogate.
if isHighSurrogate(rune(x)) {
r, sSize = utf16.DecodeRune(r, rune(x)), 4
}
} else if !atEOF {
err = transform.ErrShortSrc
break
}
}
if dSize = utf8.RuneLen(r); dSize < 0 {
r, dSize = utf8.RuneError, 3
}
} else if atEOF {
// Single trailing byte.
r, dSize, sSize = utf8.RuneError, 3, 1
} else {
err = transform.ErrShortSrc
break
}
if nDst+dSize > len(dst) {
err = transform.ErrShortDst
break
}
nDst += utf8.EncodeRune(dst[nDst:], r)
nSrc += sSize
}
return nDst, nSrc, err
}
func isHighSurrogate(r rune) bool {
return 0xDC00 <= r && r <= 0xDFFF
}
type utf16Encoder struct {
endianness Endianness
initialBOMPolicy BOMPolicy
currentBOMPolicy BOMPolicy
}
func (u *utf16Encoder) Reset() {
u.currentBOMPolicy = u.initialBOMPolicy
}
func (u *utf16Encoder) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
if u.currentBOMPolicy&writeBOM != 0 {
if len(dst) < 2 {
return 0, 0, transform.ErrShortDst
}
dst[0], dst[1] = 0xfe, 0xff
u.currentBOMPolicy = IgnoreBOM
nDst = 2
}
r, size := rune(0), 0
for nSrc < len(src) {
r = rune(src[nSrc])
// Decode a 1-byte rune.
if r < utf8.RuneSelf {
size = 1
} else {
// Decode a multi-byte rune.
r, size = utf8.DecodeRune(src[nSrc:])
if size == 1 {
// All valid runes of size 1 (those below utf8.RuneSelf) were
// handled above. We have invalid UTF-8 or we haven't seen the
// full character yet.
if !atEOF && !utf8.FullRune(src[nSrc:]) {
err = transform.ErrShortSrc
break
}
}
}
if r <= 0xffff {
if nDst+2 > len(dst) {
err = transform.ErrShortDst
break
}
dst[nDst+0] = uint8(r >> 8)
dst[nDst+1] = uint8(r)
nDst += 2
} else {
if nDst+4 > len(dst) {
err = transform.ErrShortDst
break
}
r1, r2 := utf16.EncodeRune(r)
dst[nDst+0] = uint8(r1 >> 8)
dst[nDst+1] = uint8(r1)
dst[nDst+2] = uint8(r2 >> 8)
dst[nDst+3] = uint8(r2)
nDst += 4
}
nSrc += size
}
if u.endianness == LittleEndian {
for i := 0; i < nDst; i += 2 {
dst[i], dst[i+1] = dst[i+1], dst[i]
}
}
return nDst, nSrc, err
}

87
vendor/golang.org/x/text/internal/utf8internal/utf8internal.go generated vendored Normal file
View file

@ -0,0 +1,87 @@
// Copyright 2015 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package utf8internal contains low-level utf8-related constants, tables, etc.
// that are used internally by the text package.
package utf8internal
// The default lowest and highest continuation byte.
const (
LoCB = 0x80 // 1000 0000
HiCB = 0xBF // 1011 1111
)
// Constants related to getting information of first bytes of UTF-8 sequences.
const (
// ASCII identifies a UTF-8 byte as ASCII.
ASCII = as
// FirstInvalid indicates a byte is invalid as a first byte of a UTF-8
// sequence.
FirstInvalid = xx
// SizeMask is a mask for the size bits. Use use x&SizeMask to get the size.
SizeMask = 7
// AcceptShift is the right-shift count for the first byte info byte to get
// the index into the AcceptRanges table. See AcceptRanges.
AcceptShift = 4
// The names of these constants are chosen to give nice alignment in the
// table below. The first nibble is an index into acceptRanges or F for
// special one-byte cases. The second nibble is the Rune length or the
// Status for the special one-byte case.
xx = 0xF1 // invalid: size 1
as = 0xF0 // ASCII: size 1
s1 = 0x02 // accept 0, size 2
s2 = 0x13 // accept 1, size 3
s3 = 0x03 // accept 0, size 3
s4 = 0x23 // accept 2, size 3
s5 = 0x34 // accept 3, size 4
s6 = 0x04 // accept 0, size 4
s7 = 0x44 // accept 4, size 4
)
// First is information about the first byte in a UTF-8 sequence.
var First = [256]uint8{
// 1 2 3 4 5 6 7 8 9 A B C D E F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x00-0x0F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x10-0x1F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x20-0x2F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x30-0x3F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x40-0x4F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x50-0x5F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x60-0x6F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x70-0x7F
// 1 2 3 4 5 6 7 8 9 A B C D E F
xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0x80-0x8F
xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0x90-0x9F
xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0xA0-0xAF
xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0xB0-0xBF
xx, xx, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, // 0xC0-0xCF
s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, // 0xD0-0xDF
s2, s3, s3, s3, s3, s3, s3, s3, s3, s3, s3, s3, s3, s4, s3, s3, // 0xE0-0xEF
s5, s6, s6, s6, s7, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0xF0-0xFF
}
// AcceptRange gives the range of valid values for the second byte in a UTF-8
// sequence for any value for First that is not ASCII or FirstInvalid.
type AcceptRange struct {
Lo uint8 // lowest value for second byte.
Hi uint8 // highest value for second byte.
}
// AcceptRanges is a slice of AcceptRange values. For a given byte sequence b
//
// AcceptRanges[First[b[0]]>>AcceptShift]
//
// will give the value of AcceptRange for the multi-byte UTF-8 sequence starting
// at b[0].
var AcceptRanges = [...]AcceptRange{
0: {LoCB, HiCB},
1: {0xA0, HiCB},
2: {LoCB, 0x9F},
3: {0x90, HiCB},
4: {LoCB, 0x8F},
}

187
vendor/golang.org/x/text/runes/cond.go generated vendored Normal file
View file

@ -0,0 +1,187 @@
// Copyright 2015 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package runes
import (
"unicode/utf8"
"golang.org/x/text/transform"
)
// Note: below we pass invalid UTF-8 to the tIn and tNotIn transformers as is.
// This is done for various reasons:
// - To retain the semantics of the Nop transformer: if input is passed to a Nop
// one would expect it to be unchanged.
// - It would be very expensive to pass a converted RuneError to a transformer:
// a transformer might need more source bytes after RuneError, meaning that
// the only way to pass it safely is to create a new buffer and manage the
// intermingling of RuneErrors and normal input.
// - Many transformers leave ill-formed UTF-8 as is, so this is not
// inconsistent. Generally ill-formed UTF-8 is only replaced if it is a
// logical consequence of the operation (as for Map) or if it otherwise would
// pose security concerns (as for Remove).
// - An alternative would be to return an error on ill-formed UTF-8, but this
// would be inconsistent with other operations.
// If returns a transformer that applies tIn to consecutive runes for which
// s.Contains(r) and tNotIn to consecutive runes for which !s.Contains(r). Reset
// is called on tIn and tNotIn at the start of each run. A Nop transformer will
// substitute a nil value passed to tIn or tNotIn. Invalid UTF-8 is translated
// to RuneError to determine which transformer to apply, but is passed as is to
// the respective transformer.
func If(s Set, tIn, tNotIn transform.Transformer) Transformer {
if tIn == nil && tNotIn == nil {
return Transformer{transform.Nop}
}
if tIn == nil {
tIn = transform.Nop
}
if tNotIn == nil {
tNotIn = transform.Nop
}
sIn, ok := tIn.(transform.SpanningTransformer)
if !ok {
sIn = dummySpan{tIn}
}
sNotIn, ok := tNotIn.(transform.SpanningTransformer)
if !ok {
sNotIn = dummySpan{tNotIn}
}
a := &cond{
tIn: sIn,
tNotIn: sNotIn,
f: s.Contains,
}
a.Reset()
return Transformer{a}
}
type dummySpan struct{ transform.Transformer }
func (d dummySpan) Span(src []byte, atEOF bool) (n int, err error) {
return 0, transform.ErrEndOfSpan
}
type cond struct {
tIn, tNotIn transform.SpanningTransformer
f func(rune) bool
check func(rune) bool // current check to perform
t transform.SpanningTransformer // current transformer to use
}
// Reset implements transform.Transformer.
func (t *cond) Reset() {
t.check = t.is
t.t = t.tIn
t.t.Reset() // notIn will be reset on first usage.
}
func (t *cond) is(r rune) bool {
if t.f(r) {
return true
}
t.check = t.isNot
t.t = t.tNotIn
t.tNotIn.Reset()
return false
}
func (t *cond) isNot(r rune) bool {
if !t.f(r) {
return true
}
t.check = t.is
t.t = t.tIn
t.tIn.Reset()
return false
}
// This implementation of Span doesn't help all too much, but it needs to be
// there to satisfy this package's Transformer interface.
// TODO: there are certainly room for improvements, though. For example, if
// t.t == transform.Nop (which will a common occurrence) it will save a bundle
// to special-case that loop.
func (t *cond) Span(src []byte, atEOF bool) (n int, err error) {
p := 0
for n < len(src) && err == nil {
// Don't process too much at a time as the Spanner that will be
// called on this block may terminate early.
const maxChunk = 4096
max := len(src)
if v := n + maxChunk; v < max {
max = v
}
atEnd := false
size := 0
current := t.t
for ; p < max; p += size {
r := rune(src[p])
if r < utf8.RuneSelf {
size = 1
} else if r, size = utf8.DecodeRune(src[p:]); size == 1 {
if !atEOF && !utf8.FullRune(src[p:]) {
err = transform.ErrShortSrc
break
}
}
if !t.check(r) {
// The next rune will be the start of a new run.
atEnd = true
break
}
}
n2, err2 := current.Span(src[n:p], atEnd || (atEOF && p == len(src)))
n += n2
if err2 != nil {
return n, err2
}
// At this point either err != nil or t.check will pass for the rune at p.
p = n + size
}
return n, err
}
func (t *cond) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
p := 0
for nSrc < len(src) && err == nil {
// Don't process too much at a time, as the work might be wasted if the
// destination buffer isn't large enough to hold the result or a
// transform returns an error early.
const maxChunk = 4096
max := len(src)
if n := nSrc + maxChunk; n < len(src) {
max = n
}
atEnd := false
size := 0
current := t.t
for ; p < max; p += size {
r := rune(src[p])
if r < utf8.RuneSelf {
size = 1
} else if r, size = utf8.DecodeRune(src[p:]); size == 1 {
if !atEOF && !utf8.FullRune(src[p:]) {
err = transform.ErrShortSrc
break
}
}
if !t.check(r) {
// The next rune will be the start of a new run.
atEnd = true
break
}
}
nDst2, nSrc2, err2 := current.Transform(dst[nDst:], src[nSrc:p], atEnd || (atEOF && p == len(src)))
nDst += nDst2
nSrc += nSrc2
if err2 != nil {
return nDst, nSrc, err2
}
// At this point either err != nil or t.check will pass for the rune at p.
p = nSrc + size
}
return nDst, nSrc, err
}

355
vendor/golang.org/x/text/runes/runes.go generated vendored Normal file
View file

@ -0,0 +1,355 @@
// Copyright 2014 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package runes provide transforms for UTF-8 encoded text.
package runes // import "golang.org/x/text/runes"
import (
"unicode"
"unicode/utf8"
"golang.org/x/text/transform"
)
// A Set is a collection of runes.
type Set interface {
// Contains returns true if r is contained in the set.
Contains(r rune) bool
}
type setFunc func(rune) bool
func (s setFunc) Contains(r rune) bool {
return s(r)
}
// Note: using funcs here instead of wrapping types result in cleaner
// documentation and a smaller API.
// In creates a Set with a Contains method that returns true for all runes in
// the given RangeTable.
func In(rt *unicode.RangeTable) Set {
return setFunc(func(r rune) bool { return unicode.Is(rt, r) })
}
// NotIn creates a Set with a Contains method that returns true for all runes not
// in the given RangeTable.
func NotIn(rt *unicode.RangeTable) Set {
return setFunc(func(r rune) bool { return !unicode.Is(rt, r) })
}
// Predicate creates a Set with a Contains method that returns f(r).
func Predicate(f func(rune) bool) Set {
return setFunc(f)
}
// Transformer implements the transform.Transformer interface.
type Transformer struct {
t transform.SpanningTransformer
}
func (t Transformer) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
return t.t.Transform(dst, src, atEOF)
}
func (t Transformer) Span(b []byte, atEOF bool) (n int, err error) {
return t.t.Span(b, atEOF)
}
func (t Transformer) Reset() { t.t.Reset() }
// Bytes returns a new byte slice with the result of converting b using t. It
// calls Reset on t. It returns nil if any error was found. This can only happen
// if an error-producing Transformer is passed to If.
func (t Transformer) Bytes(b []byte) []byte {
b, _, err := transform.Bytes(t, b)
if err != nil {
return nil
}
return b
}
// String returns a string with the result of converting s using t. It calls
// Reset on t. It returns the empty string if any error was found. This can only
// happen if an error-producing Transformer is passed to If.
func (t Transformer) String(s string) string {
s, _, err := transform.String(t, s)
if err != nil {
return ""
}
return s
}
// TODO:
// - Copy: copying strings and bytes in whole-rune units.
// - Validation (maybe)
// - Well-formed-ness (maybe)
const runeErrorString = string(utf8.RuneError)
// Remove returns a Transformer that removes runes r for which s.Contains(r).
// Illegal input bytes are replaced by RuneError before being passed to f.
func Remove(s Set) Transformer {
if f, ok := s.(setFunc); ok {
// This little trick cuts the running time of BenchmarkRemove for sets
// created by Predicate roughly in half.
// TODO: special-case RangeTables as well.
return Transformer{remove(f)}
}
return Transformer{remove(s.Contains)}
}
// TODO: remove transform.RemoveFunc.
type remove func(r rune) bool
func (remove) Reset() {}
// Span implements transform.Spanner.
func (t remove) Span(src []byte, atEOF bool) (n int, err error) {
for r, size := rune(0), 0; n < len(src); {
if r = rune(src[n]); r < utf8.RuneSelf {
size = 1
} else if r, size = utf8.DecodeRune(src[n:]); size == 1 {
// Invalid rune.
if !atEOF && !utf8.FullRune(src[n:]) {
err = transform.ErrShortSrc
} else {
err = transform.ErrEndOfSpan
}
break
}
if t(r) {
err = transform.ErrEndOfSpan
break
}
n += size
}
return
}
// Transform implements transform.Transformer.
func (t remove) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
for r, size := rune(0), 0; nSrc < len(src); {
if r = rune(src[nSrc]); r < utf8.RuneSelf {
size = 1
} else if r, size = utf8.DecodeRune(src[nSrc:]); size == 1 {
// Invalid rune.
if !atEOF && !utf8.FullRune(src[nSrc:]) {
err = transform.ErrShortSrc
break
}
// We replace illegal bytes with RuneError. Not doing so might
// otherwise turn a sequence of invalid UTF-8 into valid UTF-8.
// The resulting byte sequence may subsequently contain runes
// for which t(r) is true that were passed unnoticed.
if !t(utf8.RuneError) {
if nDst+3 > len(dst) {
err = transform.ErrShortDst
break
}
dst[nDst+0] = runeErrorString[0]
dst[nDst+1] = runeErrorString[1]
dst[nDst+2] = runeErrorString[2]
nDst += 3
}
nSrc++
continue
}
if t(r) {
nSrc += size
continue
}
if nDst+size > len(dst) {
err = transform.ErrShortDst
break
}
for i := 0; i < size; i++ {
dst[nDst] = src[nSrc]
nDst++
nSrc++
}
}
return
}
// Map returns a Transformer that maps the runes in the input using the given
// mapping. Illegal bytes in the input are converted to utf8.RuneError before
// being passed to the mapping func.
func Map(mapping func(rune) rune) Transformer {
return Transformer{mapper(mapping)}
}
type mapper func(rune) rune
func (mapper) Reset() {}
// Span implements transform.Spanner.
func (t mapper) Span(src []byte, atEOF bool) (n int, err error) {
for r, size := rune(0), 0; n < len(src); n += size {
if r = rune(src[n]); r < utf8.RuneSelf {
size = 1
} else if r, size = utf8.DecodeRune(src[n:]); size == 1 {
// Invalid rune.
if !atEOF && !utf8.FullRune(src[n:]) {
err = transform.ErrShortSrc
} else {
err = transform.ErrEndOfSpan
}
break
}
if t(r) != r {
err = transform.ErrEndOfSpan
break
}
}
return n, err
}
// Transform implements transform.Transformer.
func (t mapper) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
var replacement rune
var b [utf8.UTFMax]byte
for r, size := rune(0), 0; nSrc < len(src); {
if r = rune(src[nSrc]); r < utf8.RuneSelf {
if replacement = t(r); replacement < utf8.RuneSelf {
if nDst == len(dst) {
err = transform.ErrShortDst
break
}
dst[nDst] = byte(replacement)
nDst++
nSrc++
continue
}
size = 1
} else if r, size = utf8.DecodeRune(src[nSrc:]); size == 1 {
// Invalid rune.
if !atEOF && !utf8.FullRune(src[nSrc:]) {
err = transform.ErrShortSrc
break
}
if replacement = t(utf8.RuneError); replacement == utf8.RuneError {
if nDst+3 > len(dst) {
err = transform.ErrShortDst
break
}
dst[nDst+0] = runeErrorString[0]
dst[nDst+1] = runeErrorString[1]
dst[nDst+2] = runeErrorString[2]
nDst += 3
nSrc++
continue
}
} else if replacement = t(r); replacement == r {
if nDst+size > len(dst) {
err = transform.ErrShortDst
break
}
for i := 0; i < size; i++ {
dst[nDst] = src[nSrc]
nDst++
nSrc++
}
continue
}
n := utf8.EncodeRune(b[:], replacement)
if nDst+n > len(dst) {
err = transform.ErrShortDst
break
}
for i := 0; i < n; i++ {
dst[nDst] = b[i]
nDst++
}
nSrc += size
}
return
}
// ReplaceIllFormed returns a transformer that replaces all input bytes that are
// not part of a well-formed UTF-8 code sequence with utf8.RuneError.
func ReplaceIllFormed() Transformer {
return Transformer{&replaceIllFormed{}}
}
type replaceIllFormed struct{ transform.NopResetter }
func (t replaceIllFormed) Span(src []byte, atEOF bool) (n int, err error) {
for n < len(src) {
// ASCII fast path.
if src[n] < utf8.RuneSelf {
n++
continue
}
r, size := utf8.DecodeRune(src[n:])
// Look for a valid non-ASCII rune.
if r != utf8.RuneError || size != 1 {
n += size
continue
}
// Look for short source data.
if !atEOF && !utf8.FullRune(src[n:]) {
err = transform.ErrShortSrc
break
}
// We have an invalid rune.
err = transform.ErrEndOfSpan
break
}
return n, err
}
func (t replaceIllFormed) Transform(dst, src []byte, atEOF bool) (nDst, nSrc int, err error) {
for nSrc < len(src) {
// ASCII fast path.
if r := src[nSrc]; r < utf8.RuneSelf {
if nDst == len(dst) {
err = transform.ErrShortDst
break
}
dst[nDst] = r
nDst++
nSrc++
continue
}
// Look for a valid non-ASCII rune.
if _, size := utf8.DecodeRune(src[nSrc:]); size != 1 {
if size != copy(dst[nDst:], src[nSrc:nSrc+size]) {
err = transform.ErrShortDst
break
}
nDst += size
nSrc += size
continue
}
// Look for short source data.
if !atEOF && !utf8.FullRune(src[nSrc:]) {
err = transform.ErrShortSrc
break
}
// We have an invalid rune.
if nDst+3 > len(dst) {
err = transform.ErrShortDst
break
}
dst[nDst+0] = runeErrorString[0]
dst[nDst+1] = runeErrorString[1]
dst[nDst+2] = runeErrorString[2]
nDst += 3
nSrc++
}
return nDst, nSrc, err
}

6
vendor/modules.txt vendored
View file

@ -1011,6 +1011,12 @@ golang.org/x/sys/windows/svc/eventlog
golang.org/x/sys/windows/svc/mgr golang.org/x/sys/windows/svc/mgr
# golang.org/x/text v0.5.0 # golang.org/x/text v0.5.0
## explicit; go 1.17 ## explicit; go 1.17
golang.org/x/text/encoding
golang.org/x/text/encoding/internal
golang.org/x/text/encoding/internal/identifier
golang.org/x/text/encoding/unicode
golang.org/x/text/internal/utf8internal
golang.org/x/text/runes
golang.org/x/text/secure/bidirule golang.org/x/text/secure/bidirule
golang.org/x/text/transform golang.org/x/text/transform
golang.org/x/text/unicode/bidi golang.org/x/text/unicode/bidi