diff options
| author | Thomas Voss <mail@thomasvoss.com> | 2024-08-07 00:21:12 +0200 |
|---|---|---|
| committer | Thomas Voss <mail@thomasvoss.com> | 2024-08-07 00:21:12 +0200 |
| commit | 351c15d28e0444fd8a78c510a0c4d62ed433c758 (patch) | |
| tree | b97aae6ec45c1b341075da147fb9e333246c19f7 /vendor/golang.org/x/text/language | |
Genesis commit
Diffstat (limited to 'vendor/golang.org/x/text/language')
| -rw-r--r-- | vendor/golang.org/x/text/language/coverage.go | 187 | ||||
| -rw-r--r-- | vendor/golang.org/x/text/language/doc.go | 98 | ||||
| -rw-r--r-- | vendor/golang.org/x/text/language/language.go | 605 | ||||
| -rw-r--r-- | vendor/golang.org/x/text/language/match.go | 735 | ||||
| -rw-r--r-- | vendor/golang.org/x/text/language/parse.go | 256 | ||||
| -rw-r--r-- | vendor/golang.org/x/text/language/tables.go | 298 | ||||
| -rw-r--r-- | vendor/golang.org/x/text/language/tags.go | 145 |
7 files changed, 2324 insertions, 0 deletions
diff --git a/vendor/golang.org/x/text/language/coverage.go b/vendor/golang.org/x/text/language/coverage.go new file mode 100644 index 0000000..a24fd1a --- /dev/null +++ b/vendor/golang.org/x/text/language/coverage.go @@ -0,0 +1,187 @@ +// 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 language + +import ( + "fmt" + "sort" + + "golang.org/x/text/internal/language" +) + +// The Coverage interface is used to define the level of coverage of an +// internationalization service. Note that not all types are supported by all +// services. As lists may be generated on the fly, it is recommended that users +// of a Coverage cache the results. +type Coverage interface { + // Tags returns the list of supported tags. + Tags() []Tag + + // BaseLanguages returns the list of supported base languages. + BaseLanguages() []Base + + // Scripts returns the list of supported scripts. + Scripts() []Script + + // Regions returns the list of supported regions. + Regions() []Region +} + +var ( + // Supported defines a Coverage that lists all supported subtags. Tags + // always returns nil. + Supported Coverage = allSubtags{} +) + +// TODO: +// - Support Variants, numbering systems. +// - CLDR coverage levels. +// - Set of common tags defined in this package. + +type allSubtags struct{} + +// Regions returns the list of supported regions. As all regions are in a +// consecutive range, it simply returns a slice of numbers in increasing order. +// The "undefined" region is not returned. +func (s allSubtags) Regions() []Region { + reg := make([]Region, language.NumRegions) + for i := range reg { + reg[i] = Region{language.Region(i + 1)} + } + return reg +} + +// Scripts returns the list of supported scripts. As all scripts are in a +// consecutive range, it simply returns a slice of numbers in increasing order. +// The "undefined" script is not returned. +func (s allSubtags) Scripts() []Script { + scr := make([]Script, language.NumScripts) + for i := range scr { + scr[i] = Script{language.Script(i + 1)} + } + return scr +} + +// BaseLanguages returns the list of all supported base languages. It generates +// the list by traversing the internal structures. +func (s allSubtags) BaseLanguages() []Base { + bs := language.BaseLanguages() + base := make([]Base, len(bs)) + for i, b := range bs { + base[i] = Base{b} + } + return base +} + +// Tags always returns nil. +func (s allSubtags) Tags() []Tag { + return nil +} + +// coverage is used by NewCoverage which is used as a convenient way for +// creating Coverage implementations for partially defined data. Very often a +// package will only need to define a subset of slices. coverage provides a +// convenient way to do this. Moreover, packages using NewCoverage, instead of +// their own implementation, will not break if later new slice types are added. +type coverage struct { + tags func() []Tag + bases func() []Base + scripts func() []Script + regions func() []Region +} + +func (s *coverage) Tags() []Tag { + if s.tags == nil { + return nil + } + return s.tags() +} + +// bases implements sort.Interface and is used to sort base languages. +type bases []Base + +func (b bases) Len() int { + return len(b) +} + +func (b bases) Swap(i, j int) { + b[i], b[j] = b[j], b[i] +} + +func (b bases) Less(i, j int) bool { + return b[i].langID < b[j].langID +} + +// BaseLanguages returns the result from calling s.bases if it is specified or +// otherwise derives the set of supported base languages from tags. +func (s *coverage) BaseLanguages() []Base { + if s.bases == nil { + tags := s.Tags() + if len(tags) == 0 { + return nil + } + a := make([]Base, len(tags)) + for i, t := range tags { + a[i] = Base{language.Language(t.lang())} + } + sort.Sort(bases(a)) + k := 0 + for i := 1; i < len(a); i++ { + if a[k] != a[i] { + k++ + a[k] = a[i] + } + } + return a[:k+1] + } + return s.bases() +} + +func (s *coverage) Scripts() []Script { + if s.scripts == nil { + return nil + } + return s.scripts() +} + +func (s *coverage) Regions() []Region { + if s.regions == nil { + return nil + } + return s.regions() +} + +// NewCoverage returns a Coverage for the given lists. It is typically used by +// packages providing internationalization services to define their level of +// coverage. A list may be of type []T or func() []T, where T is either Tag, +// Base, Script or Region. The returned Coverage derives the value for Bases +// from Tags if no func or slice for []Base is specified. For other unspecified +// types the returned Coverage will return nil for the respective methods. +func NewCoverage(list ...interface{}) Coverage { + s := &coverage{} + for _, x := range list { + switch v := x.(type) { + case func() []Base: + s.bases = v + case func() []Script: + s.scripts = v + case func() []Region: + s.regions = v + case func() []Tag: + s.tags = v + case []Base: + s.bases = func() []Base { return v } + case []Script: + s.scripts = func() []Script { return v } + case []Region: + s.regions = func() []Region { return v } + case []Tag: + s.tags = func() []Tag { return v } + default: + panic(fmt.Sprintf("language: unsupported set type %T", v)) + } + } + return s +} diff --git a/vendor/golang.org/x/text/language/doc.go b/vendor/golang.org/x/text/language/doc.go new file mode 100644 index 0000000..212b77c --- /dev/null +++ b/vendor/golang.org/x/text/language/doc.go @@ -0,0 +1,98 @@ +// Copyright 2017 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 language implements BCP 47 language tags and related functionality. +// +// The most important function of package language is to match a list of +// user-preferred languages to a list of supported languages. +// It alleviates the developer of dealing with the complexity of this process +// and provides the user with the best experience +// (see https://blog.golang.org/matchlang). +// +// # Matching preferred against supported languages +// +// A Matcher for an application that supports English, Australian English, +// Danish, and standard Mandarin can be created as follows: +// +// var matcher = language.NewMatcher([]language.Tag{ +// language.English, // The first language is used as fallback. +// language.MustParse("en-AU"), +// language.Danish, +// language.Chinese, +// }) +// +// This list of supported languages is typically implied by the languages for +// which there exists translations of the user interface. +// +// User-preferred languages usually come as a comma-separated list of BCP 47 +// language tags. +// The MatchString finds best matches for such strings: +// +// handler(w http.ResponseWriter, r *http.Request) { +// lang, _ := r.Cookie("lang") +// accept := r.Header.Get("Accept-Language") +// tag, _ := language.MatchStrings(matcher, lang.String(), accept) +// +// // tag should now be used for the initialization of any +// // locale-specific service. +// } +// +// The Matcher's Match method can be used to match Tags directly. +// +// Matchers are aware of the intricacies of equivalence between languages, such +// as deprecated subtags, legacy tags, macro languages, mutual +// intelligibility between scripts and languages, and transparently passing +// BCP 47 user configuration. +// For instance, it will know that a reader of Bokmål Danish can read Norwegian +// and will know that Cantonese ("yue") is a good match for "zh-HK". +// +// # Using match results +// +// To guarantee a consistent user experience to the user it is important to +// use the same language tag for the selection of any locale-specific services. +// For example, it is utterly confusing to substitute spelled-out numbers +// or dates in one language in text of another language. +// More subtly confusing is using the wrong sorting order or casing +// algorithm for a certain language. +// +// All the packages in x/text that provide locale-specific services +// (e.g. collate, cases) should be initialized with the tag that was +// obtained at the start of an interaction with the user. +// +// Note that Tag that is returned by Match and MatchString may differ from any +// of the supported languages, as it may contain carried over settings from +// the user tags. +// This may be inconvenient when your application has some additional +// locale-specific data for your supported languages. +// Match and MatchString both return the index of the matched supported tag +// to simplify associating such data with the matched tag. +// +// # Canonicalization +// +// If one uses the Matcher to compare languages one does not need to +// worry about canonicalization. +// +// The meaning of a Tag varies per application. The language package +// therefore delays canonicalization and preserves information as much +// as possible. The Matcher, however, will always take into account that +// two different tags may represent the same language. +// +// By default, only legacy and deprecated tags are converted into their +// canonical equivalent. All other information is preserved. This approach makes +// the confidence scores more accurate and allows matchers to distinguish +// between variants that are otherwise lost. +// +// As a consequence, two tags that should be treated as identical according to +// BCP 47 or CLDR, like "en-Latn" and "en", will be represented differently. The +// Matcher handles such distinctions, though, and is aware of the +// equivalence relations. The CanonType type can be used to alter the +// canonicalization form. +// +// # References +// +// BCP 47 - Tags for Identifying Languages http://tools.ietf.org/html/bcp47 +package language // import "golang.org/x/text/language" + +// TODO: explanation on how to match languages for your own locale-specific +// service. diff --git a/vendor/golang.org/x/text/language/language.go b/vendor/golang.org/x/text/language/language.go new file mode 100644 index 0000000..4d9c661 --- /dev/null +++ b/vendor/golang.org/x/text/language/language.go @@ -0,0 +1,605 @@ +// 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. + +//go:generate go run gen.go -output tables.go + +package language + +// TODO: Remove above NOTE after: +// - verifying that tables are dropped correctly (most notably matcher tables). + +import ( + "strings" + + "golang.org/x/text/internal/language" + "golang.org/x/text/internal/language/compact" +) + +// Tag represents a BCP 47 language tag. It is used to specify an instance of a +// specific language or locale. All language tag values are guaranteed to be +// well-formed. +type Tag compact.Tag + +func makeTag(t language.Tag) (tag Tag) { + return Tag(compact.Make(t)) +} + +func (t *Tag) tag() language.Tag { + return (*compact.Tag)(t).Tag() +} + +func (t *Tag) isCompact() bool { + return (*compact.Tag)(t).IsCompact() +} + +// TODO: improve performance. +func (t *Tag) lang() language.Language { return t.tag().LangID } +func (t *Tag) region() language.Region { return t.tag().RegionID } +func (t *Tag) script() language.Script { return t.tag().ScriptID } + +// Make is a convenience wrapper for Parse that omits the error. +// In case of an error, a sensible default is returned. +func Make(s string) Tag { + return Default.Make(s) +} + +// Make is a convenience wrapper for c.Parse that omits the error. +// In case of an error, a sensible default is returned. +func (c CanonType) Make(s string) Tag { + t, _ := c.Parse(s) + return t +} + +// Raw returns the raw base language, script and region, without making an +// attempt to infer their values. +func (t Tag) Raw() (b Base, s Script, r Region) { + tt := t.tag() + return Base{tt.LangID}, Script{tt.ScriptID}, Region{tt.RegionID} +} + +// IsRoot returns true if t is equal to language "und". +func (t Tag) IsRoot() bool { + return compact.Tag(t).IsRoot() +} + +// CanonType can be used to enable or disable various types of canonicalization. +type CanonType int + +const ( + // Replace deprecated base languages with their preferred replacements. + DeprecatedBase CanonType = 1 << iota + // Replace deprecated scripts with their preferred replacements. + DeprecatedScript + // Replace deprecated regions with their preferred replacements. + DeprecatedRegion + // Remove redundant scripts. + SuppressScript + // Normalize legacy encodings. This includes legacy languages defined in + // CLDR as well as bibliographic codes defined in ISO-639. + Legacy + // Map the dominant language of a macro language group to the macro language + // subtag. For example cmn -> zh. + Macro + // The CLDR flag should be used if full compatibility with CLDR is required. + // There are a few cases where language.Tag may differ from CLDR. To follow all + // of CLDR's suggestions, use All|CLDR. + CLDR + + // Raw can be used to Compose or Parse without Canonicalization. + Raw CanonType = 0 + + // Replace all deprecated tags with their preferred replacements. + Deprecated = DeprecatedBase | DeprecatedScript | DeprecatedRegion + + // All canonicalizations recommended by BCP 47. + BCP47 = Deprecated | SuppressScript + + // All canonicalizations. + All = BCP47 | Legacy | Macro + + // Default is the canonicalization used by Parse, Make and Compose. To + // preserve as much information as possible, canonicalizations that remove + // potentially valuable information are not included. The Matcher is + // designed to recognize similar tags that would be the same if + // they were canonicalized using All. + Default = Deprecated | Legacy + + canonLang = DeprecatedBase | Legacy | Macro + + // TODO: LikelyScript, LikelyRegion: suppress similar to ICU. +) + +// canonicalize returns the canonicalized equivalent of the tag and +// whether there was any change. +func canonicalize(c CanonType, t language.Tag) (language.Tag, bool) { + if c == Raw { + return t, false + } + changed := false + if c&SuppressScript != 0 { + if t.LangID.SuppressScript() == t.ScriptID { + t.ScriptID = 0 + changed = true + } + } + if c&canonLang != 0 { + for { + if l, aliasType := t.LangID.Canonicalize(); l != t.LangID { + switch aliasType { + case language.Legacy: + if c&Legacy != 0 { + if t.LangID == _sh && t.ScriptID == 0 { + t.ScriptID = _Latn + } + t.LangID = l + changed = true + } + case language.Macro: + if c&Macro != 0 { + // We deviate here from CLDR. The mapping "nb" -> "no" + // qualifies as a typical Macro language mapping. However, + // for legacy reasons, CLDR maps "no", the macro language + // code for Norwegian, to the dominant variant "nb". This + // change is currently under consideration for CLDR as well. + // See https://unicode.org/cldr/trac/ticket/2698 and also + // https://unicode.org/cldr/trac/ticket/1790 for some of the + // practical implications. TODO: this check could be removed + // if CLDR adopts this change. + if c&CLDR == 0 || t.LangID != _nb { + changed = true + t.LangID = l + } + } + case language.Deprecated: + if c&DeprecatedBase != 0 { + if t.LangID == _mo && t.RegionID == 0 { + t.RegionID = _MD + } + t.LangID = l + changed = true + // Other canonicalization types may still apply. + continue + } + } + } else if c&Legacy != 0 && t.LangID == _no && c&CLDR != 0 { + t.LangID = _nb + changed = true + } + break + } + } + if c&DeprecatedScript != 0 { + if t.ScriptID == _Qaai { + changed = true + t.ScriptID = _Zinh + } + } + if c&DeprecatedRegion != 0 { + if r := t.RegionID.Canonicalize(); r != t.RegionID { + changed = true + t.RegionID = r + } + } + return t, changed +} + +// Canonicalize returns the canonicalized equivalent of the tag. +func (c CanonType) Canonicalize(t Tag) (Tag, error) { + // First try fast path. + if t.isCompact() { + if _, changed := canonicalize(c, compact.Tag(t).Tag()); !changed { + return t, nil + } + } + // It is unlikely that one will canonicalize a tag after matching. So do + // a slow but simple approach here. + if tag, changed := canonicalize(c, t.tag()); changed { + tag.RemakeString() + return makeTag(tag), nil + } + return t, nil + +} + +// Confidence indicates the level of certainty for a given return value. +// For example, Serbian may be written in Cyrillic or Latin script. +// The confidence level indicates whether a value was explicitly specified, +// whether it is typically the only possible value, or whether there is +// an ambiguity. +type Confidence int + +const ( + No Confidence = iota // full confidence that there was no match + Low // most likely value picked out of a set of alternatives + High // value is generally assumed to be the correct match + Exact // exact match or explicitly specified value +) + +var confName = []string{"No", "Low", "High", "Exact"} + +func (c Confidence) String() string { + return confName[c] +} + +// String returns the canonical string representation of the language tag. +func (t Tag) String() string { + return t.tag().String() +} + +// MarshalText implements encoding.TextMarshaler. +func (t Tag) MarshalText() (text []byte, err error) { + return t.tag().MarshalText() +} + +// UnmarshalText implements encoding.TextUnmarshaler. +func (t *Tag) UnmarshalText(text []byte) error { + var tag language.Tag + err := tag.UnmarshalText(text) + *t = makeTag(tag) + return err +} + +// Base returns the base language of the language tag. If the base language is +// unspecified, an attempt will be made to infer it from the context. +// It uses a variant of CLDR's Add Likely Subtags algorithm. This is subject to change. +func (t Tag) Base() (Base, Confidence) { + if b := t.lang(); b != 0 { + return Base{b}, Exact + } + tt := t.tag() + c := High + if tt.ScriptID == 0 && !tt.RegionID.IsCountry() { + c = Low + } + if tag, err := tt.Maximize(); err == nil && tag.LangID != 0 { + return Base{tag.LangID}, c + } + return Base{0}, No +} + +// Script infers the script for the language tag. If it was not explicitly given, it will infer +// a most likely candidate. +// If more than one script is commonly used for a language, the most likely one +// is returned with a low confidence indication. For example, it returns (Cyrl, Low) +// for Serbian. +// If a script cannot be inferred (Zzzz, No) is returned. We do not use Zyyy (undetermined) +// as one would suspect from the IANA registry for BCP 47. In a Unicode context Zyyy marks +// common characters (like 1, 2, 3, '.', etc.) and is therefore more like multiple scripts. +// See https://www.unicode.org/reports/tr24/#Values for more details. Zzzz is also used for +// unknown value in CLDR. (Zzzz, Exact) is returned if Zzzz was explicitly specified. +// Note that an inferred script is never guaranteed to be the correct one. Latin is +// almost exclusively used for Afrikaans, but Arabic has been used for some texts +// in the past. Also, the script that is commonly used may change over time. +// It uses a variant of CLDR's Add Likely Subtags algorithm. This is subject to change. +func (t Tag) Script() (Script, Confidence) { + if scr := t.script(); scr != 0 { + return Script{scr}, Exact + } + tt := t.tag() + sc, c := language.Script(_Zzzz), No + if scr := tt.LangID.SuppressScript(); scr != 0 { + // Note: it is not always the case that a language with a suppress + // script value is only written in one script (e.g. kk, ms, pa). + if tt.RegionID == 0 { + return Script{scr}, High + } + sc, c = scr, High + } + if tag, err := tt.Maximize(); err == nil { + if tag.ScriptID != sc { + sc, c = tag.ScriptID, Low + } + } else { + tt, _ = canonicalize(Deprecated|Macro, tt) + if tag, err := tt.Maximize(); err == nil && tag.ScriptID != sc { + sc, c = tag.ScriptID, Low + } + } + return Script{sc}, c +} + +// Region returns the region for the language tag. If it was not explicitly given, it will +// infer a most likely candidate from the context. +// It uses a variant of CLDR's Add Likely Subtags algorithm. This is subject to change. +func (t Tag) Region() (Region, Confidence) { + if r := t.region(); r != 0 { + return Region{r}, Exact + } + tt := t.tag() + if tt, err := tt.Maximize(); err == nil { + return Region{tt.RegionID}, Low // TODO: differentiate between high and low. + } + tt, _ = canonicalize(Deprecated|Macro, tt) + if tag, err := tt.Maximize(); err == nil { + return Region{tag.RegionID}, Low + } + return Region{_ZZ}, No // TODO: return world instead of undetermined? +} + +// Variants returns the variants specified explicitly for this language tag. +// or nil if no variant was specified. +func (t Tag) Variants() []Variant { + if !compact.Tag(t).MayHaveVariants() { + return nil + } + v := []Variant{} + x, str := "", t.tag().Variants() + for str != "" { + x, str = nextToken(str) + v = append(v, Variant{x}) + } + return v +} + +// Parent returns the CLDR parent of t. In CLDR, missing fields in data for a +// specific language are substituted with fields from the parent language. +// The parent for a language may change for newer versions of CLDR. +// +// Parent returns a tag for a less specific language that is mutually +// intelligible or Und if there is no such language. This may not be the same as +// simply stripping the last BCP 47 subtag. For instance, the parent of "zh-TW" +// is "zh-Hant", and the parent of "zh-Hant" is "und". +func (t Tag) Parent() Tag { + return Tag(compact.Tag(t).Parent()) +} + +// nextToken returns token t and the rest of the string. +func nextToken(s string) (t, tail string) { + p := strings.Index(s[1:], "-") + if p == -1 { + return s[1:], "" + } + p++ + return s[1:p], s[p:] +} + +// Extension is a single BCP 47 extension. +type Extension struct { + s string +} + +// String returns the string representation of the extension, including the +// type tag. +func (e Extension) String() string { + return e.s +} + +// ParseExtension parses s as an extension and returns it on success. +func ParseExtension(s string) (e Extension, err error) { + ext, err := language.ParseExtension(s) + return Extension{ext}, err +} + +// Type returns the one-byte extension type of e. It returns 0 for the zero +// exception. +func (e Extension) Type() byte { + if e.s == "" { + return 0 + } + return e.s[0] +} + +// Tokens returns the list of tokens of e. +func (e Extension) Tokens() []string { + return strings.Split(e.s, "-") +} + +// Extension returns the extension of type x for tag t. It will return +// false for ok if t does not have the requested extension. The returned +// extension will be invalid in this case. +func (t Tag) Extension(x byte) (ext Extension, ok bool) { + if !compact.Tag(t).MayHaveExtensions() { + return Extension{}, false + } + e, ok := t.tag().Extension(x) + return Extension{e}, ok +} + +// Extensions returns all extensions of t. +func (t Tag) Extensions() []Extension { + if !compact.Tag(t).MayHaveExtensions() { + return nil + } + e := []Extension{} + for _, ext := range t.tag().Extensions() { + e = append(e, Extension{ext}) + } + return e +} + +// TypeForKey returns the type associated with the given key, where key and type +// are of the allowed values defined for the Unicode locale extension ('u') in +// https://www.unicode.org/reports/tr35/#Unicode_Language_and_Locale_Identifiers. +// TypeForKey will traverse the inheritance chain to get the correct value. +// +// If there are multiple types associated with a key, only the first will be +// returned. If there is no type associated with a key, it returns the empty +// string. +func (t Tag) TypeForKey(key string) string { + if !compact.Tag(t).MayHaveExtensions() { + if key != "rg" && key != "va" { + return "" + } + } + return t.tag().TypeForKey(key) +} + +// SetTypeForKey returns a new Tag with the key set to type, where key and type +// are of the allowed values defined for the Unicode locale extension ('u') in +// https://www.unicode.org/reports/tr35/#Unicode_Language_and_Locale_Identifiers. +// An empty value removes an existing pair with the same key. +func (t Tag) SetTypeForKey(key, value string) (Tag, error) { + tt, err := t.tag().SetTypeForKey(key, value) + return makeTag(tt), err +} + +// NumCompactTags is the number of compact tags. The maximum tag is +// NumCompactTags-1. +const NumCompactTags = compact.NumCompactTags + +// CompactIndex returns an index, where 0 <= index < NumCompactTags, for tags +// for which data exists in the text repository.The index will change over time +// and should not be stored in persistent storage. If t does not match a compact +// index, exact will be false and the compact index will be returned for the +// first match after repeatedly taking the Parent of t. +func CompactIndex(t Tag) (index int, exact bool) { + id, exact := compact.LanguageID(compact.Tag(t)) + return int(id), exact +} + +var root = language.Tag{} + +// Base is an ISO 639 language code, used for encoding the base language +// of a language tag. +type Base struct { + langID language.Language +} + +// ParseBase parses a 2- or 3-letter ISO 639 code. +// It returns a ValueError if s is a well-formed but unknown language identifier +// or another error if another error occurred. +func ParseBase(s string) (Base, error) { + l, err := language.ParseBase(s) + return Base{l}, err +} + +// String returns the BCP 47 representation of the base language. +func (b Base) String() string { + return b.langID.String() +} + +// ISO3 returns the ISO 639-3 language code. +func (b Base) ISO3() string { + return b.langID.ISO3() +} + +// IsPrivateUse reports whether this language code is reserved for private use. +func (b Base) IsPrivateUse() bool { + return b.langID.IsPrivateUse() +} + +// Script is a 4-letter ISO 15924 code for representing scripts. +// It is idiomatically represented in title case. +type Script struct { + scriptID language.Script +} + +// ParseScript parses a 4-letter ISO 15924 code. +// It returns a ValueError if s is a well-formed but unknown script identifier +// or another error if another error occurred. +func ParseScript(s string) (Script, error) { + sc, err := language.ParseScript(s) + return Script{sc}, err +} + +// String returns the script code in title case. +// It returns "Zzzz" for an unspecified script. +func (s Script) String() string { + return s.scriptID.String() +} + +// IsPrivateUse reports whether this script code is reserved for private use. +func (s Script) IsPrivateUse() bool { + return s.scriptID.IsPrivateUse() +} + +// Region is an ISO 3166-1 or UN M.49 code for representing countries and regions. +type Region struct { + regionID language.Region +} + +// EncodeM49 returns the Region for the given UN M.49 code. +// It returns an error if r is not a valid code. +func EncodeM49(r int) (Region, error) { + rid, err := language.EncodeM49(r) + return Region{rid}, err +} + +// ParseRegion parses a 2- or 3-letter ISO 3166-1 or a UN M.49 code. +// It returns a ValueError if s is a well-formed but unknown region identifier +// or another error if another error occurred. +func ParseRegion(s string) (Region, error) { + r, err := language.ParseRegion(s) + return Region{r}, err +} + +// String returns the BCP 47 representation for the region. +// It returns "ZZ" for an unspecified region. +func (r Region) String() string { + return r.regionID.String() +} + +// ISO3 returns the 3-letter ISO code of r. +// Note that not all regions have a 3-letter ISO code. +// In such cases this method returns "ZZZ". +func (r Region) ISO3() string { + return r.regionID.ISO3() +} + +// M49 returns the UN M.49 encoding of r, or 0 if this encoding +// is not defined for r. +func (r Region) M49() int { + return r.regionID.M49() +} + +// IsPrivateUse reports whether r has the ISO 3166 User-assigned status. This +// may include private-use tags that are assigned by CLDR and used in this +// implementation. So IsPrivateUse and IsCountry can be simultaneously true. +func (r Region) IsPrivateUse() bool { + return r.regionID.IsPrivateUse() +} + +// IsCountry returns whether this region is a country or autonomous area. This +// includes non-standard definitions from CLDR. +func (r Region) IsCountry() bool { + return r.regionID.IsCountry() +} + +// IsGroup returns whether this region defines a collection of regions. This +// includes non-standard definitions from CLDR. +func (r Region) IsGroup() bool { + return r.regionID.IsGroup() +} + +// Contains returns whether Region c is contained by Region r. It returns true +// if c == r. +func (r Region) Contains(c Region) bool { + return r.regionID.Contains(c.regionID) +} + +// TLD returns the country code top-level domain (ccTLD). UK is returned for GB. +// In all other cases it returns either the region itself or an error. +// +// This method may return an error for a region for which there exists a +// canonical form with a ccTLD. To get that ccTLD canonicalize r first. The +// region will already be canonicalized it was obtained from a Tag that was +// obtained using any of the default methods. +func (r Region) TLD() (Region, error) { + tld, err := r.regionID.TLD() + return Region{tld}, err +} + +// Canonicalize returns the region or a possible replacement if the region is +// deprecated. It will not return a replacement for deprecated regions that +// are split into multiple regions. +func (r Region) Canonicalize() Region { + return Region{r.regionID.Canonicalize()} +} + +// Variant represents a registered variant of a language as defined by BCP 47. +type Variant struct { + variant string +} + +// ParseVariant parses and returns a Variant. An error is returned if s is not +// a valid variant. +func ParseVariant(s string) (Variant, error) { + v, err := language.ParseVariant(s) + return Variant{v.String()}, err +} + +// String returns the string representation of the variant. +func (v Variant) String() string { + return v.variant +} diff --git a/vendor/golang.org/x/text/language/match.go b/vendor/golang.org/x/text/language/match.go new file mode 100644 index 0000000..1153baf --- /dev/null +++ b/vendor/golang.org/x/text/language/match.go @@ -0,0 +1,735 @@ +// 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 language + +import ( + "errors" + "strings" + + "golang.org/x/text/internal/language" +) + +// A MatchOption configures a Matcher. +type MatchOption func(*matcher) + +// PreferSameScript will, in the absence of a match, result in the first +// preferred tag with the same script as a supported tag to match this supported +// tag. The default is currently true, but this may change in the future. +func PreferSameScript(preferSame bool) MatchOption { + return func(m *matcher) { m.preferSameScript = preferSame } +} + +// TODO(v1.0.0): consider making Matcher a concrete type, instead of interface. +// There doesn't seem to be too much need for multiple types. +// Making it a concrete type allows MatchStrings to be a method, which will +// improve its discoverability. + +// MatchStrings parses and matches the given strings until one of them matches +// the language in the Matcher. A string may be an Accept-Language header as +// handled by ParseAcceptLanguage. The default language is returned if no +// other language matched. +func MatchStrings(m Matcher, lang ...string) (tag Tag, index int) { + for _, accept := range lang { + desired, _, err := ParseAcceptLanguage(accept) + if err != nil { + continue + } + if tag, index, conf := m.Match(desired...); conf != No { + return tag, index + } + } + tag, index, _ = m.Match() + return +} + +// Matcher is the interface that wraps the Match method. +// +// Match returns the best match for any of the given tags, along with +// a unique index associated with the returned tag and a confidence +// score. +type Matcher interface { + Match(t ...Tag) (tag Tag, index int, c Confidence) +} + +// Comprehends reports the confidence score for a speaker of a given language +// to being able to comprehend the written form of an alternative language. +func Comprehends(speaker, alternative Tag) Confidence { + _, _, c := NewMatcher([]Tag{alternative}).Match(speaker) + return c +} + +// NewMatcher returns a Matcher that matches an ordered list of preferred tags +// against a list of supported tags based on written intelligibility, closeness +// of dialect, equivalence of subtags and various other rules. It is initialized +// with the list of supported tags. The first element is used as the default +// value in case no match is found. +// +// Its Match method matches the first of the given Tags to reach a certain +// confidence threshold. The tags passed to Match should therefore be specified +// in order of preference. Extensions are ignored for matching. +// +// The index returned by the Match method corresponds to the index of the +// matched tag in t, but is augmented with the Unicode extension ('u')of the +// corresponding preferred tag. This allows user locale options to be passed +// transparently. +func NewMatcher(t []Tag, options ...MatchOption) Matcher { + return newMatcher(t, options) +} + +func (m *matcher) Match(want ...Tag) (t Tag, index int, c Confidence) { + var tt language.Tag + match, w, c := m.getBest(want...) + if match != nil { + tt, index = match.tag, match.index + } else { + // TODO: this should be an option + tt = m.default_.tag + if m.preferSameScript { + outer: + for _, w := range want { + script, _ := w.Script() + if script.scriptID == 0 { + // Don't do anything if there is no script, such as with + // private subtags. + continue + } + for i, h := range m.supported { + if script.scriptID == h.maxScript { + tt, index = h.tag, i + break outer + } + } + } + } + // TODO: select first language tag based on script. + } + if w.RegionID != tt.RegionID && w.RegionID != 0 { + if w.RegionID != 0 && tt.RegionID != 0 && tt.RegionID.Contains(w.RegionID) { + tt.RegionID = w.RegionID + tt.RemakeString() + } else if r := w.RegionID.String(); len(r) == 2 { + // TODO: also filter macro and deprecated. + tt, _ = tt.SetTypeForKey("rg", strings.ToLower(r)+"zzzz") + } + } + // Copy options from the user-provided tag into the result tag. This is hard + // to do after the fact, so we do it here. + // TODO: add in alternative variants to -u-va-. + // TODO: add preferred region to -u-rg-. + if e := w.Extensions(); len(e) > 0 { + b := language.Builder{} + b.SetTag(tt) + for _, e := range e { + b.AddExt(e) + } + tt = b.Make() + } + return makeTag(tt), index, c +} + +// ErrMissingLikelyTagsData indicates no information was available +// to compute likely values of missing tags. +var ErrMissingLikelyTagsData = errors.New("missing likely tags data") + +// func (t *Tag) setTagsFrom(id Tag) { +// t.LangID = id.LangID +// t.ScriptID = id.ScriptID +// t.RegionID = id.RegionID +// } + +// Tag Matching +// CLDR defines an algorithm for finding the best match between two sets of language +// tags. The basic algorithm defines how to score a possible match and then find +// the match with the best score +// (see https://www.unicode.org/reports/tr35/#LanguageMatching). +// Using scoring has several disadvantages. The scoring obfuscates the importance of +// the various factors considered, making the algorithm harder to understand. Using +// scoring also requires the full score to be computed for each pair of tags. +// +// We will use a different algorithm which aims to have the following properties: +// - clarity on the precedence of the various selection factors, and +// - improved performance by allowing early termination of a comparison. +// +// Matching algorithm (overview) +// Input: +// - supported: a set of supported tags +// - default: the default tag to return in case there is no match +// - desired: list of desired tags, ordered by preference, starting with +// the most-preferred. +// +// Algorithm: +// 1) Set the best match to the lowest confidence level +// 2) For each tag in "desired": +// a) For each tag in "supported": +// 1) compute the match between the two tags. +// 2) if the match is better than the previous best match, replace it +// with the new match. (see next section) +// b) if the current best match is Exact and pin is true the result will be +// frozen to the language found thusfar, although better matches may +// still be found for the same language. +// 3) If the best match so far is below a certain threshold, return "default". +// +// Ranking: +// We use two phases to determine whether one pair of tags are a better match +// than another pair of tags. First, we determine a rough confidence level. If the +// levels are different, the one with the highest confidence wins. +// Second, if the rough confidence levels are identical, we use a set of tie-breaker +// rules. +// +// The confidence level of matching a pair of tags is determined by finding the +// lowest confidence level of any matches of the corresponding subtags (the +// result is deemed as good as its weakest link). +// We define the following levels: +// Exact - An exact match of a subtag, before adding likely subtags. +// MaxExact - An exact match of a subtag, after adding likely subtags. +// [See Note 2]. +// High - High level of mutual intelligibility between different subtag +// variants. +// Low - Low level of mutual intelligibility between different subtag +// variants. +// No - No mutual intelligibility. +// +// The following levels can occur for each type of subtag: +// Base: Exact, MaxExact, High, Low, No +// Script: Exact, MaxExact [see Note 3], Low, No +// Region: Exact, MaxExact, High +// Variant: Exact, High +// Private: Exact, No +// +// Any result with a confidence level of Low or higher is deemed a possible match. +// Once a desired tag matches any of the supported tags with a level of MaxExact +// or higher, the next desired tag is not considered (see Step 2.b). +// Note that CLDR provides languageMatching data that defines close equivalence +// classes for base languages, scripts and regions. +// +// Tie-breaking +// If we get the same confidence level for two matches, we apply a sequence of +// tie-breaking rules. The first that succeeds defines the result. The rules are +// applied in the following order. +// 1) Original language was defined and was identical. +// 2) Original region was defined and was identical. +// 3) Distance between two maximized regions was the smallest. +// 4) Original script was defined and was identical. +// 5) Distance from want tag to have tag using the parent relation [see Note 5.] +// If there is still no winner after these rules are applied, the first match +// found wins. +// +// Notes: +// [2] In practice, as matching of Exact is done in a separate phase from +// matching the other levels, we reuse the Exact level to mean MaxExact in +// the second phase. As a consequence, we only need the levels defined by +// the Confidence type. The MaxExact confidence level is mapped to High in +// the public API. +// [3] We do not differentiate between maximized script values that were derived +// from suppressScript versus most likely tag data. We determined that in +// ranking the two, one ranks just after the other. Moreover, the two cannot +// occur concurrently. As a consequence, they are identical for practical +// purposes. +// [4] In case of deprecated, macro-equivalents and legacy mappings, we assign +// the MaxExact level to allow iw vs he to still be a closer match than +// en-AU vs en-US, for example. +// [5] In CLDR a locale inherits fields that are unspecified for this locale +// from its parent. Therefore, if a locale is a parent of another locale, +// it is a strong measure for closeness, especially when no other tie +// breaker rule applies. One could also argue it is inconsistent, for +// example, when pt-AO matches pt (which CLDR equates with pt-BR), even +// though its parent is pt-PT according to the inheritance rules. +// +// Implementation Details: +// There are several performance considerations worth pointing out. Most notably, +// we preprocess as much as possible (within reason) at the time of creation of a +// matcher. This includes: +// - creating a per-language map, which includes data for the raw base language +// and its canonicalized variant (if applicable), +// - expanding entries for the equivalence classes defined in CLDR's +// languageMatch data. +// The per-language map ensures that typically only a very small number of tags +// need to be considered. The pre-expansion of canonicalized subtags and +// equivalence classes reduces the amount of map lookups that need to be done at +// runtime. + +// matcher keeps a set of supported language tags, indexed by language. +type matcher struct { + default_ *haveTag + supported []*haveTag + index map[language.Language]*matchHeader + passSettings bool + preferSameScript bool +} + +// matchHeader has the lists of tags for exact matches and matches based on +// maximized and canonicalized tags for a given language. +type matchHeader struct { + haveTags []*haveTag + original bool +} + +// haveTag holds a supported Tag and its maximized script and region. The maximized +// or canonicalized language is not stored as it is not needed during matching. +type haveTag struct { + tag language.Tag + + // index of this tag in the original list of supported tags. + index int + + // conf is the maximum confidence that can result from matching this haveTag. + // When conf < Exact this means it was inserted after applying a CLDR equivalence rule. + conf Confidence + + // Maximized region and script. + maxRegion language.Region + maxScript language.Script + + // altScript may be checked as an alternative match to maxScript. If altScript + // matches, the confidence level for this match is Low. Theoretically there + // could be multiple alternative scripts. This does not occur in practice. + altScript language.Script + + // nextMax is the index of the next haveTag with the same maximized tags. + nextMax uint16 +} + +func makeHaveTag(tag language.Tag, index int) (haveTag, language.Language) { + max := tag + if tag.LangID != 0 || tag.RegionID != 0 || tag.ScriptID != 0 { + max, _ = canonicalize(All, max) + max, _ = max.Maximize() + max.RemakeString() + } + return haveTag{tag, index, Exact, max.RegionID, max.ScriptID, altScript(max.LangID, max.ScriptID), 0}, max.LangID +} + +// altScript returns an alternative script that may match the given script with +// a low confidence. At the moment, the langMatch data allows for at most one +// script to map to another and we rely on this to keep the code simple. +func altScript(l language.Language, s language.Script) language.Script { + for _, alt := range matchScript { + // TODO: also match cases where language is not the same. + if (language.Language(alt.wantLang) == l || language.Language(alt.haveLang) == l) && + language.Script(alt.haveScript) == s { + return language.Script(alt.wantScript) + } + } + return 0 +} + +// addIfNew adds a haveTag to the list of tags only if it is a unique tag. +// Tags that have the same maximized values are linked by index. +func (h *matchHeader) addIfNew(n haveTag, exact bool) { + h.original = h.original || exact + // Don't add new exact matches. + for _, v := range h.haveTags { + if equalsRest(v.tag, n.tag) { + return + } + } + // Allow duplicate maximized tags, but create a linked list to allow quickly + // comparing the equivalents and bail out. + for i, v := range h.haveTags { + if v.maxScript == n.maxScript && + v.maxRegion == n.maxRegion && + v.tag.VariantOrPrivateUseTags() == n.tag.VariantOrPrivateUseTags() { + for h.haveTags[i].nextMax != 0 { + i = int(h.haveTags[i].nextMax) + } + h.haveTags[i].nextMax = uint16(len(h.haveTags)) + break + } + } + h.haveTags = append(h.haveTags, &n) +} + +// header returns the matchHeader for the given language. It creates one if +// it doesn't already exist. +func (m *matcher) header(l language.Language) *matchHeader { + if h := m.index[l]; h != nil { + return h + } + h := &matchHeader{} + m.index[l] = h + return h +} + +func toConf(d uint8) Confidence { + if d <= 10 { + return High + } + if d < 30 { + return Low + } + return No +} + +// newMatcher builds an index for the given supported tags and returns it as +// a matcher. It also expands the index by considering various equivalence classes +// for a given tag. +func newMatcher(supported []Tag, options []MatchOption) *matcher { + m := &matcher{ + index: make(map[language.Language]*matchHeader), + preferSameScript: true, + } + for _, o := range options { + o(m) + } + if len(supported) == 0 { + m.default_ = &haveTag{} + return m + } + // Add supported languages to the index. Add exact matches first to give + // them precedence. + for i, tag := range supported { + tt := tag.tag() + pair, _ := makeHaveTag(tt, i) + m.header(tt.LangID).addIfNew(pair, true) + m.supported = append(m.supported, &pair) + } + m.default_ = m.header(supported[0].lang()).haveTags[0] + // Keep these in two different loops to support the case that two equivalent + // languages are distinguished, such as iw and he. + for i, tag := range supported { + tt := tag.tag() + pair, max := makeHaveTag(tt, i) + if max != tt.LangID { + m.header(max).addIfNew(pair, true) + } + } + + // update is used to add indexes in the map for equivalent languages. + // update will only add entries to original indexes, thus not computing any + // transitive relations. + update := func(want, have uint16, conf Confidence) { + if hh := m.index[language.Language(have)]; hh != nil { + if !hh.original { + return + } + hw := m.header(language.Language(want)) + for _, ht := range hh.haveTags { + v := *ht + if conf < v.conf { + v.conf = conf + } + v.nextMax = 0 // this value needs to be recomputed + if v.altScript != 0 { + v.altScript = altScript(language.Language(want), v.maxScript) + } + hw.addIfNew(v, conf == Exact && hh.original) + } + } + } + + // Add entries for languages with mutual intelligibility as defined by CLDR's + // languageMatch data. + for _, ml := range matchLang { + update(ml.want, ml.have, toConf(ml.distance)) + if !ml.oneway { + update(ml.have, ml.want, toConf(ml.distance)) + } + } + + // Add entries for possible canonicalizations. This is an optimization to + // ensure that only one map lookup needs to be done at runtime per desired tag. + // First we match deprecated equivalents. If they are perfect equivalents + // (their canonicalization simply substitutes a different language code, but + // nothing else), the match confidence is Exact, otherwise it is High. + for i, lm := range language.AliasMap { + // If deprecated codes match and there is no fiddling with the script + // or region, we consider it an exact match. + conf := Exact + if language.AliasTypes[i] != language.Macro { + if !isExactEquivalent(language.Language(lm.From)) { + conf = High + } + update(lm.To, lm.From, conf) + } + update(lm.From, lm.To, conf) + } + return m +} + +// getBest gets the best matching tag in m for any of the given tags, taking into +// account the order of preference of the given tags. +func (m *matcher) getBest(want ...Tag) (got *haveTag, orig language.Tag, c Confidence) { + best := bestMatch{} + for i, ww := range want { + w := ww.tag() + var max language.Tag + // Check for exact match first. + h := m.index[w.LangID] + if w.LangID != 0 { + if h == nil { + continue + } + // Base language is defined. + max, _ = canonicalize(Legacy|Deprecated|Macro, w) + // A region that is added through canonicalization is stronger than + // a maximized region: set it in the original (e.g. mo -> ro-MD). + if w.RegionID != max.RegionID { + w.RegionID = max.RegionID + } + // TODO: should we do the same for scripts? + // See test case: en, sr, nl ; sh ; sr + max, _ = max.Maximize() + } else { + // Base language is not defined. + if h != nil { + for i := range h.haveTags { + have := h.haveTags[i] + if equalsRest(have.tag, w) { + return have, w, Exact + } + } + } + if w.ScriptID == 0 && w.RegionID == 0 { + // We skip all tags matching und for approximate matching, including + // private tags. + continue + } + max, _ = w.Maximize() + if h = m.index[max.LangID]; h == nil { + continue + } + } + pin := true + for _, t := range want[i+1:] { + if w.LangID == t.lang() { + pin = false + break + } + } + // Check for match based on maximized tag. + for i := range h.haveTags { + have := h.haveTags[i] + best.update(have, w, max.ScriptID, max.RegionID, pin) + if best.conf == Exact { + for have.nextMax != 0 { + have = h.haveTags[have.nextMax] + best.update(have, w, max.ScriptID, max.RegionID, pin) + } + return best.have, best.want, best.conf + } + } + } + if best.conf <= No { + if len(want) != 0 { + return nil, want[0].tag(), No + } + return nil, language.Tag{}, No + } + return best.have, best.want, best.conf +} + +// bestMatch accumulates the best match so far. +type bestMatch struct { + have *haveTag + want language.Tag + conf Confidence + pinnedRegion language.Region + pinLanguage bool + sameRegionGroup bool + // Cached results from applying tie-breaking rules. + origLang bool + origReg bool + paradigmReg bool + regGroupDist uint8 + origScript bool +} + +// update updates the existing best match if the new pair is considered to be a +// better match. To determine if the given pair is a better match, it first +// computes the rough confidence level. If this surpasses the current match, it +// will replace it and update the tie-breaker rule cache. If there is a tie, it +// proceeds with applying a series of tie-breaker rules. If there is no +// conclusive winner after applying the tie-breaker rules, it leaves the current +// match as the preferred match. +// +// If pin is true and have and tag are a strong match, it will henceforth only +// consider matches for this language. This corresponds to the idea that most +// users have a strong preference for the first defined language. A user can +// still prefer a second language over a dialect of the preferred language by +// explicitly specifying dialects, e.g. "en, nl, en-GB". In this case pin should +// be false. +func (m *bestMatch) update(have *haveTag, tag language.Tag, maxScript language.Script, maxRegion language.Region, pin bool) { + // Bail if the maximum attainable confidence is below that of the current best match. + c := have.conf + if c < m.conf { + return + } + // Don't change the language once we already have found an exact match. + if m.pinLanguage && tag.LangID != m.want.LangID { + return + } + // Pin the region group if we are comparing tags for the same language. + if tag.LangID == m.want.LangID && m.sameRegionGroup { + _, sameGroup := regionGroupDist(m.pinnedRegion, have.maxRegion, have.maxScript, m.want.LangID) + if !sameGroup { + return + } + } + if c == Exact && have.maxScript == maxScript { + // If there is another language and then another entry of this language, + // don't pin anything, otherwise pin the language. + m.pinLanguage = pin + } + if equalsRest(have.tag, tag) { + } else if have.maxScript != maxScript { + // There is usually very little comprehension between different scripts. + // In a few cases there may still be Low comprehension. This possibility + // is pre-computed and stored in have.altScript. + if Low < m.conf || have.altScript != maxScript { + return + } + c = Low + } else if have.maxRegion != maxRegion { + if High < c { + // There is usually a small difference between languages across regions. + c = High + } + } + + // We store the results of the computations of the tie-breaker rules along + // with the best match. There is no need to do the checks once we determine + // we have a winner, but we do still need to do the tie-breaker computations. + // We use "beaten" to keep track if we still need to do the checks. + beaten := false // true if the new pair defeats the current one. + if c != m.conf { + if c < m.conf { + return + } + beaten = true + } + + // Tie-breaker rules: + // We prefer if the pre-maximized language was specified and identical. + origLang := have.tag.LangID == tag.LangID && tag.LangID != 0 + if !beaten && m.origLang != origLang { + if m.origLang { + return + } + beaten = true + } + + // We prefer if the pre-maximized region was specified and identical. + origReg := have.tag.RegionID == tag.RegionID && tag.RegionID != 0 + if !beaten && m.origReg != origReg { + if m.origReg { + return + } + beaten = true + } + + regGroupDist, sameGroup := regionGroupDist(have.maxRegion, maxRegion, maxScript, tag.LangID) + if !beaten && m.regGroupDist != regGroupDist { + if regGroupDist > m.regGroupDist { + return + } + beaten = true + } + + paradigmReg := isParadigmLocale(tag.LangID, have.maxRegion) + if !beaten && m.paradigmReg != paradigmReg { + if !paradigmReg { + return + } + beaten = true + } + + // Next we prefer if the pre-maximized script was specified and identical. + origScript := have.tag.ScriptID == tag.ScriptID && tag.ScriptID != 0 + if !beaten && m.origScript != origScript { + if m.origScript { + return + } + beaten = true + } + + // Update m to the newly found best match. + if beaten { + m.have = have + m.want = tag + m.conf = c + m.pinnedRegion = maxRegion + m.sameRegionGroup = sameGroup + m.origLang = origLang + m.origReg = origReg + m.paradigmReg = paradigmReg + m.origScript = origScript + m.regGroupDist = regGroupDist + } +} + +func isParadigmLocale(lang language.Language, r language.Region) bool { + for _, e := range paradigmLocales { + if language.Language(e[0]) == lang && (r == language.Region(e[1]) || r == language.Region(e[2])) { + return true + } + } + return false +} + +// regionGroupDist computes the distance between two regions based on their +// CLDR grouping. +func regionGroupDist(a, b language.Region, script language.Script, lang language.Language) (dist uint8, same bool) { + const defaultDistance = 4 + + aGroup := uint(regionToGroups[a]) << 1 + bGroup := uint(regionToGroups[b]) << 1 + for _, ri := range matchRegion { + if language.Language(ri.lang) == lang && (ri.script == 0 || language.Script(ri.script) == script) { + group := uint(1 << (ri.group &^ 0x80)) + if 0x80&ri.group == 0 { + if aGroup&bGroup&group != 0 { // Both regions are in the group. + return ri.distance, ri.distance == defaultDistance + } + } else { + if (aGroup|bGroup)&group == 0 { // Both regions are not in the group. + return ri.distance, ri.distance == defaultDistance + } + } + } + } + return defaultDistance, true +} + +// equalsRest compares everything except the language. +func equalsRest(a, b language.Tag) bool { + // TODO: don't include extensions in this comparison. To do this efficiently, + // though, we should handle private tags separately. + return a.ScriptID == b.ScriptID && a.RegionID == b.RegionID && a.VariantOrPrivateUseTags() == b.VariantOrPrivateUseTags() +} + +// isExactEquivalent returns true if canonicalizing the language will not alter +// the script or region of a tag. +func isExactEquivalent(l language.Language) bool { + for _, o := range notEquivalent { + if o == l { + return false + } + } + return true +} + +var notEquivalent []language.Language + +func init() { + // Create a list of all languages for which canonicalization may alter the + // script or region. + for _, lm := range language.AliasMap { + tag := language.Tag{LangID: language.Language(lm.From)} + if tag, _ = canonicalize(All, tag); tag.ScriptID != 0 || tag.RegionID != 0 { + notEquivalent = append(notEquivalent, language.Language(lm.From)) + } + } + // Maximize undefined regions of paradigm locales. + for i, v := range paradigmLocales { + t := language.Tag{LangID: language.Language(v[0])} + max, _ := t.Maximize() + if v[1] == 0 { + paradigmLocales[i][1] = uint16(max.RegionID) + } + if v[2] == 0 { + paradigmLocales[i][2] = uint16(max.RegionID) + } + } +} diff --git a/vendor/golang.org/x/text/language/parse.go b/vendor/golang.org/x/text/language/parse.go new file mode 100644 index 0000000..4d57222 --- /dev/null +++ b/vendor/golang.org/x/text/language/parse.go @@ -0,0 +1,256 @@ +// 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 language + +import ( + "errors" + "sort" + "strconv" + "strings" + + "golang.org/x/text/internal/language" +) + +// ValueError is returned by any of the parsing functions when the +// input is well-formed but the respective subtag is not recognized +// as a valid value. +type ValueError interface { + error + + // Subtag returns the subtag for which the error occurred. + Subtag() string +} + +// Parse parses the given BCP 47 string and returns a valid Tag. If parsing +// failed it returns an error and any part of the tag that could be parsed. +// If parsing succeeded but an unknown value was found, it returns +// ValueError. The Tag returned in this case is just stripped of the unknown +// value. All other values are preserved. It accepts tags in the BCP 47 format +// and extensions to this standard defined in +// https://www.unicode.org/reports/tr35/#Unicode_Language_and_Locale_Identifiers. +// The resulting tag is canonicalized using the default canonicalization type. +func Parse(s string) (t Tag, err error) { + return Default.Parse(s) +} + +// Parse parses the given BCP 47 string and returns a valid Tag. If parsing +// failed it returns an error and any part of the tag that could be parsed. +// If parsing succeeded but an unknown value was found, it returns +// ValueError. The Tag returned in this case is just stripped of the unknown +// value. All other values are preserved. It accepts tags in the BCP 47 format +// and extensions to this standard defined in +// https://www.unicode.org/reports/tr35/#Unicode_Language_and_Locale_Identifiers. +// The resulting tag is canonicalized using the canonicalization type c. +func (c CanonType) Parse(s string) (t Tag, err error) { + defer func() { + if recover() != nil { + t = Tag{} + err = language.ErrSyntax + } + }() + + tt, err := language.Parse(s) + if err != nil { + return makeTag(tt), err + } + tt, changed := canonicalize(c, tt) + if changed { + tt.RemakeString() + } + return makeTag(tt), err +} + +// Compose creates a Tag from individual parts, which may be of type Tag, Base, +// Script, Region, Variant, []Variant, Extension, []Extension or error. If a +// Base, Script or Region or slice of type Variant or Extension is passed more +// than once, the latter will overwrite the former. Variants and Extensions are +// accumulated, but if two extensions of the same type are passed, the latter +// will replace the former. For -u extensions, though, the key-type pairs are +// added, where later values overwrite older ones. A Tag overwrites all former +// values and typically only makes sense as the first argument. The resulting +// tag is returned after canonicalizing using the Default CanonType. If one or +// more errors are encountered, one of the errors is returned. +func Compose(part ...interface{}) (t Tag, err error) { + return Default.Compose(part...) +} + +// Compose creates a Tag from individual parts, which may be of type Tag, Base, +// Script, Region, Variant, []Variant, Extension, []Extension or error. If a +// Base, Script or Region or slice of type Variant or Extension is passed more +// than once, the latter will overwrite the former. Variants and Extensions are +// accumulated, but if two extensions of the same type are passed, the latter +// will replace the former. For -u extensions, though, the key-type pairs are +// added, where later values overwrite older ones. A Tag overwrites all former +// values and typically only makes sense as the first argument. The resulting +// tag is returned after canonicalizing using CanonType c. If one or more errors +// are encountered, one of the errors is returned. +func (c CanonType) Compose(part ...interface{}) (t Tag, err error) { + defer func() { + if recover() != nil { + t = Tag{} + err = language.ErrSyntax + } + }() + + var b language.Builder + if err = update(&b, part...); err != nil { + return und, err + } + b.Tag, _ = canonicalize(c, b.Tag) + return makeTag(b.Make()), err +} + +var errInvalidArgument = errors.New("invalid Extension or Variant") + +func update(b *language.Builder, part ...interface{}) (err error) { + for _, x := range part { + switch v := x.(type) { + case Tag: + b.SetTag(v.tag()) + case Base: + b.Tag.LangID = v.langID + case Script: + b.Tag.ScriptID = v.scriptID + case Region: + b.Tag.RegionID = v.regionID + case Variant: + if v.variant == "" { + err = errInvalidArgument + break + } + b.AddVariant(v.variant) + case Extension: + if v.s == "" { + err = errInvalidArgument + break + } + b.SetExt(v.s) + case []Variant: + b.ClearVariants() + for _, v := range v { + b.AddVariant(v.variant) + } + case []Extension: + b.ClearExtensions() + for _, e := range v { + b.SetExt(e.s) + } + // TODO: support parsing of raw strings based on morphology or just extensions? + case error: + if v != nil { + err = v + } + } + } + return +} + +var errInvalidWeight = errors.New("ParseAcceptLanguage: invalid weight") +var errTagListTooLarge = errors.New("tag list exceeds max length") + +// ParseAcceptLanguage parses the contents of an Accept-Language header as +// defined in http://www.ietf.org/rfc/rfc2616.txt and returns a list of Tags and +// a list of corresponding quality weights. It is more permissive than RFC 2616 +// and may return non-nil slices even if the input is not valid. +// The Tags will be sorted by highest weight first and then by first occurrence. +// Tags with a weight of zero will be dropped. An error will be returned if the +// input could not be parsed. +func ParseAcceptLanguage(s string) (tag []Tag, q []float32, err error) { + defer func() { + if recover() != nil { + tag = nil + q = nil + err = language.ErrSyntax + } + }() + + if strings.Count(s, "-") > 1000 { + return nil, nil, errTagListTooLarge + } + + var entry string + for s != "" { + if entry, s = split(s, ','); entry == "" { + continue + } + + entry, weight := split(entry, ';') + + // Scan the language. + t, err := Parse(entry) + if err != nil { + id, ok := acceptFallback[entry] + if !ok { + return nil, nil, err + } + t = makeTag(language.Tag{LangID: id}) + } + + // Scan the optional weight. + w := 1.0 + if weight != "" { + weight = consume(weight, 'q') + weight = consume(weight, '=') + // consume returns the empty string when a token could not be + // consumed, resulting in an error for ParseFloat. + if w, err = strconv.ParseFloat(weight, 32); err != nil { + return nil, nil, errInvalidWeight + } + // Drop tags with a quality weight of 0. + if w <= 0 { + continue + } + } + + tag = append(tag, t) + q = append(q, float32(w)) + } + sort.Stable(&tagSort{tag, q}) + return tag, q, nil +} + +// consume removes a leading token c from s and returns the result or the empty +// string if there is no such token. +func consume(s string, c byte) string { + if s == "" || s[0] != c { + return "" + } + return strings.TrimSpace(s[1:]) +} + +func split(s string, c byte) (head, tail string) { + if i := strings.IndexByte(s, c); i >= 0 { + return strings.TrimSpace(s[:i]), strings.TrimSpace(s[i+1:]) + } + return strings.TrimSpace(s), "" +} + +// Add hack mapping to deal with a small number of cases that occur +// in Accept-Language (with reasonable frequency). +var acceptFallback = map[string]language.Language{ + "english": _en, + "deutsch": _de, + "italian": _it, + "french": _fr, + "*": _mul, // defined in the spec to match all languages. +} + +type tagSort struct { + tag []Tag + q []float32 +} + +func (s *tagSort) Len() int { + return len(s.q) +} + +func (s *tagSort) Less(i, j int) bool { + return s.q[i] > s.q[j] +} + +func (s *tagSort) Swap(i, j int) { + s.tag[i], s.tag[j] = s.tag[j], s.tag[i] + s.q[i], s.q[j] = s.q[j], s.q[i] +} diff --git a/vendor/golang.org/x/text/language/tables.go b/vendor/golang.org/x/text/language/tables.go new file mode 100644 index 0000000..a6573dc --- /dev/null +++ b/vendor/golang.org/x/text/language/tables.go @@ -0,0 +1,298 @@ +// Code generated by running "go generate" in golang.org/x/text. DO NOT EDIT. + +package language + +// CLDRVersion is the CLDR version from which the tables in this package are derived. +const CLDRVersion = "32" + +const ( + _de = 269 + _en = 313 + _fr = 350 + _it = 505 + _mo = 784 + _no = 879 + _nb = 839 + _pt = 960 + _sh = 1031 + _mul = 806 + _und = 0 +) +const ( + _001 = 1 + _419 = 31 + _BR = 65 + _CA = 73 + _ES = 111 + _GB = 124 + _MD = 189 + _PT = 239 + _UK = 307 + _US = 310 + _ZZ = 358 + _XA = 324 + _XC = 326 + _XK = 334 +) +const ( + _Latn = 91 + _Hani = 57 + _Hans = 59 + _Hant = 60 + _Qaaa = 149 + _Qaai = 157 + _Qabx = 198 + _Zinh = 255 + _Zyyy = 260 + _Zzzz = 261 +) + +var regionToGroups = []uint8{ // 359 elements + // Entry 0 - 3F + 0x00, 0x00, 0x00, 0x04, 0x04, 0x00, 0x00, 0x04, + 0x00, 0x00, 0x00, 0x00, 0x04, 0x04, 0x04, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x04, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x04, 0x04, 0x00, + 0x00, 0x04, 0x00, 0x00, 0x04, 0x01, 0x00, 0x00, + 0x04, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x04, 0x04, 0x00, 0x04, + // Entry 40 - 7F + 0x04, 0x04, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x04, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x04, 0x00, 0x00, 0x04, 0x00, 0x00, 0x04, + 0x00, 0x00, 0x04, 0x00, 0x04, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x04, 0x04, 0x00, + 0x08, 0x00, 0x04, 0x00, 0x00, 0x08, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x04, + // Entry 80 - BF + 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x04, 0x00, + 0x00, 0x00, 0x04, 0x01, 0x00, 0x04, 0x02, 0x00, + 0x04, 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x04, 0x00, + 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x08, 0x08, 0x00, 0x00, 0x00, 0x04, + // Entry C0 - FF + 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x02, + 0x01, 0x04, 0x08, 0x04, 0x00, 0x00, 0x00, 0x00, + 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x04, 0x00, 0x04, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x04, 0x00, 0x05, 0x00, 0x00, + 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + // Entry 100 - 13F + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x04, + 0x00, 0x00, 0x00, 0x04, 0x04, 0x00, 0x00, 0x00, + 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x08, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x05, 0x04, + 0x00, 0x00, 0x04, 0x00, 0x04, 0x04, 0x05, 0x00, + // Entry 140 - 17F + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, +} // Size: 383 bytes + +var paradigmLocales = [][3]uint16{ // 3 elements + 0: [3]uint16{0x139, 0x0, 0x7c}, + 1: [3]uint16{0x13e, 0x0, 0x1f}, + 2: [3]uint16{0x3c0, 0x41, 0xef}, +} // Size: 42 bytes + +type mutualIntelligibility struct { + want uint16 + have uint16 + distance uint8 + oneway bool +} +type scriptIntelligibility struct { + wantLang uint16 + haveLang uint16 + wantScript uint8 + haveScript uint8 + distance uint8 +} +type regionIntelligibility struct { + lang uint16 + script uint8 + group uint8 + distance uint8 +} + +// matchLang holds pairs of langIDs of base languages that are typically +// mutually intelligible. Each pair is associated with a confidence and +// whether the intelligibility goes one or both ways. +var matchLang = []mutualIntelligibility{ // 113 elements + 0: {want: 0x1d1, have: 0xb7, distance: 0x4, oneway: false}, + 1: {want: 0x407, have: 0xb7, distance: 0x4, oneway: false}, + 2: {want: 0x407, have: 0x1d1, distance: 0x4, oneway: false}, + 3: {want: 0x407, have: 0x432, distance: 0x4, oneway: false}, + 4: {want: 0x43a, have: 0x1, distance: 0x4, oneway: false}, + 5: {want: 0x1a3, have: 0x10d, distance: 0x4, oneway: true}, + 6: {want: 0x295, have: 0x10d, distance: 0x4, oneway: true}, + 7: {want: 0x101, have: 0x36f, distance: 0x8, oneway: false}, + 8: {want: 0x101, have: 0x347, distance: 0x8, oneway: false}, + 9: {want: 0x5, have: 0x3e2, distance: 0xa, oneway: true}, + 10: {want: 0xd, have: 0x139, distance: 0xa, oneway: true}, + 11: {want: 0x16, have: 0x367, distance: 0xa, oneway: true}, + 12: {want: 0x21, have: 0x139, distance: 0xa, oneway: true}, + 13: {want: 0x56, have: 0x13e, distance: 0xa, oneway: true}, + 14: {want: 0x58, have: 0x3e2, distance: 0xa, oneway: true}, + 15: {want: 0x71, have: 0x3e2, distance: 0xa, oneway: true}, + 16: {want: 0x75, have: 0x139, distance: 0xa, oneway: true}, + 17: {want: 0x82, have: 0x1be, distance: 0xa, oneway: true}, + 18: {want: 0xa5, have: 0x139, distance: 0xa, oneway: true}, + 19: {want: 0xb2, have: 0x15e, distance: 0xa, oneway: true}, + 20: {want: 0xdd, have: 0x153, distance: 0xa, oneway: true}, + 21: {want: 0xe5, have: 0x139, distance: 0xa, oneway: true}, + 22: {want: 0xe9, have: 0x3a, distance: 0xa, oneway: true}, + 23: {want: 0xf0, have: 0x15e, distance: 0xa, oneway: true}, + 24: {want: 0xf9, have: 0x15e, distance: 0xa, oneway: true}, + 25: {want: 0x100, have: 0x139, distance: 0xa, oneway: true}, + 26: {want: 0x130, have: 0x139, distance: 0xa, oneway: true}, + 27: {want: 0x13c, have: 0x139, distance: 0xa, oneway: true}, + 28: {want: 0x140, have: 0x151, distance: 0xa, oneway: true}, + 29: {want: 0x145, have: 0x13e, distance: 0xa, oneway: true}, + 30: {want: 0x158, have: 0x101, distance: 0xa, oneway: true}, + 31: {want: 0x16d, have: 0x367, distance: 0xa, oneway: true}, + 32: {want: 0x16e, have: 0x139, distance: 0xa, oneway: true}, + 33: {want: 0x16f, have: 0x139, distance: 0xa, oneway: true}, + 34: {want: 0x17e, have: 0x139, distance: 0xa, oneway: true}, + 35: {want: 0x190, have: 0x13e, distance: 0xa, oneway: true}, + 36: {want: 0x194, have: 0x13e, distance: 0xa, oneway: true}, + 37: {want: 0x1a4, have: 0x1be, distance: 0xa, oneway: true}, + 38: {want: 0x1b4, have: 0x139, distance: 0xa, oneway: true}, + 39: {want: 0x1b8, have: 0x139, distance: 0xa, oneway: true}, + 40: {want: 0x1d4, have: 0x15e, distance: 0xa, oneway: true}, + 41: {want: 0x1d7, have: 0x3e2, distance: 0xa, oneway: true}, + 42: {want: 0x1d9, have: 0x139, distance: 0xa, oneway: true}, + 43: {want: 0x1e7, have: 0x139, distance: 0xa, oneway: true}, + 44: {want: 0x1f8, have: 0x139, distance: 0xa, oneway: true}, + 45: {want: 0x20e, have: 0x1e1, distance: 0xa, oneway: true}, + 46: {want: 0x210, have: 0x139, distance: 0xa, oneway: true}, + 47: {want: 0x22d, have: 0x15e, distance: 0xa, oneway: true}, + 48: {want: 0x242, have: 0x3e2, distance: 0xa, oneway: true}, + 49: {want: 0x24a, have: 0x139, distance: 0xa, oneway: true}, + 50: {want: 0x251, have: 0x139, distance: 0xa, oneway: true}, + 51: {want: 0x265, have: 0x139, distance: 0xa, oneway: true}, + 52: {want: 0x274, have: 0x48a, distance: 0xa, oneway: true}, + 53: {want: 0x28a, have: 0x3e2, distance: 0xa, oneway: true}, + 54: {want: 0x28e, have: 0x1f9, distance: 0xa, oneway: true}, + 55: {want: 0x2a3, have: 0x139, distance: 0xa, oneway: true}, + 56: {want: 0x2b5, have: 0x15e, distance: 0xa, oneway: true}, + 57: {want: 0x2b8, have: 0x139, distance: 0xa, oneway: true}, + 58: {want: 0x2be, have: 0x139, distance: 0xa, oneway: true}, + 59: {want: 0x2c3, have: 0x15e, distance: 0xa, oneway: true}, + 60: {want: 0x2ed, have: 0x139, distance: 0xa, oneway: true}, + 61: {want: 0x2f1, have: 0x15e, distance: 0xa, oneway: true}, + 62: {want: 0x2fa, have: 0x139, distance: 0xa, oneway: true}, + 63: {want: 0x2ff, have: 0x7e, distance: 0xa, oneway: true}, + 64: {want: 0x304, have: 0x139, distance: 0xa, oneway: true}, + 65: {want: 0x30b, have: 0x3e2, distance: 0xa, oneway: true}, + 66: {want: 0x31b, have: 0x1be, distance: 0xa, oneway: true}, + 67: {want: 0x31f, have: 0x1e1, distance: 0xa, oneway: true}, + 68: {want: 0x320, have: 0x139, distance: 0xa, oneway: true}, + 69: {want: 0x331, have: 0x139, distance: 0xa, oneway: true}, + 70: {want: 0x351, have: 0x139, distance: 0xa, oneway: true}, + 71: {want: 0x36a, have: 0x347, distance: 0xa, oneway: false}, + 72: {want: 0x36a, have: 0x36f, distance: 0xa, oneway: true}, + 73: {want: 0x37a, have: 0x139, distance: 0xa, oneway: true}, + 74: {want: 0x387, have: 0x139, distance: 0xa, oneway: true}, + 75: {want: 0x389, have: 0x139, distance: 0xa, oneway: true}, + 76: {want: 0x38b, have: 0x15e, distance: 0xa, oneway: true}, + 77: {want: 0x390, have: 0x139, distance: 0xa, oneway: true}, + 78: {want: 0x395, have: 0x139, distance: 0xa, oneway: true}, + 79: {want: 0x39d, have: 0x139, distance: 0xa, oneway: true}, + 80: {want: 0x3a5, have: 0x139, distance: 0xa, oneway: true}, + 81: {want: 0x3be, have: 0x139, distance: 0xa, oneway: true}, + 82: {want: 0x3c4, have: 0x13e, distance: 0xa, oneway: true}, + 83: {want: 0x3d4, have: 0x10d, distance: 0xa, oneway: true}, + 84: {want: 0x3d9, have: 0x139, distance: 0xa, oneway: true}, + 85: {want: 0x3e5, have: 0x15e, distance: 0xa, oneway: true}, + 86: {want: 0x3e9, have: 0x1be, distance: 0xa, oneway: true}, + 87: {want: 0x3fa, have: 0x139, distance: 0xa, oneway: true}, + 88: {want: 0x40c, have: 0x139, distance: 0xa, oneway: true}, + 89: {want: 0x423, have: 0x139, distance: 0xa, oneway: true}, + 90: {want: 0x429, have: 0x139, distance: 0xa, oneway: true}, + 91: {want: 0x431, have: 0x139, distance: 0xa, oneway: true}, + 92: {want: 0x43b, have: 0x139, distance: 0xa, oneway: true}, + 93: {want: 0x43e, have: 0x1e1, distance: 0xa, oneway: true}, + 94: {want: 0x445, have: 0x139, distance: 0xa, oneway: true}, + 95: {want: 0x450, have: 0x139, distance: 0xa, oneway: true}, + 96: {want: 0x461, have: 0x139, distance: 0xa, oneway: true}, + 97: {want: 0x467, have: 0x3e2, distance: 0xa, oneway: true}, + 98: {want: 0x46f, have: 0x139, distance: 0xa, oneway: true}, + 99: {want: 0x476, have: 0x3e2, distance: 0xa, oneway: true}, + 100: {want: 0x3883, have: 0x139, distance: 0xa, oneway: true}, + 101: {want: 0x480, have: 0x139, distance: 0xa, oneway: true}, + 102: {want: 0x482, have: 0x139, distance: 0xa, oneway: true}, + 103: {want: 0x494, have: 0x3e2, distance: 0xa, oneway: true}, + 104: {want: 0x49d, have: 0x139, distance: 0xa, oneway: true}, + 105: {want: 0x4ac, have: 0x529, distance: 0xa, oneway: true}, + 106: {want: 0x4b4, have: 0x139, distance: 0xa, oneway: true}, + 107: {want: 0x4bc, have: 0x3e2, distance: 0xa, oneway: true}, + 108: {want: 0x4e5, have: 0x15e, distance: 0xa, oneway: true}, + 109: {want: 0x4f2, have: 0x139, distance: 0xa, oneway: true}, + 110: {want: 0x512, have: 0x139, distance: 0xa, oneway: true}, + 111: {want: 0x518, have: 0x139, distance: 0xa, oneway: true}, + 112: {want: 0x52f, have: 0x139, distance: 0xa, oneway: true}, +} // Size: 702 bytes + +// matchScript holds pairs of scriptIDs where readers of one script +// can typically also read the other. Each is associated with a confidence. +var matchScript = []scriptIntelligibility{ // 26 elements + 0: {wantLang: 0x432, haveLang: 0x432, wantScript: 0x5b, haveScript: 0x20, distance: 0x5}, + 1: {wantLang: 0x432, haveLang: 0x432, wantScript: 0x20, haveScript: 0x5b, distance: 0x5}, + 2: {wantLang: 0x58, haveLang: 0x3e2, wantScript: 0x5b, haveScript: 0x20, distance: 0xa}, + 3: {wantLang: 0xa5, haveLang: 0x139, wantScript: 0xe, haveScript: 0x5b, distance: 0xa}, + 4: {wantLang: 0x1d7, haveLang: 0x3e2, wantScript: 0x8, haveScript: 0x20, distance: 0xa}, + 5: {wantLang: 0x210, haveLang: 0x139, wantScript: 0x2e, haveScript: 0x5b, distance: 0xa}, + 6: {wantLang: 0x24a, haveLang: 0x139, wantScript: 0x4f, haveScript: 0x5b, distance: 0xa}, + 7: {wantLang: 0x251, haveLang: 0x139, wantScript: 0x53, haveScript: 0x5b, distance: 0xa}, + 8: {wantLang: 0x2b8, haveLang: 0x139, wantScript: 0x58, haveScript: 0x5b, distance: 0xa}, + 9: {wantLang: 0x304, haveLang: 0x139, wantScript: 0x6f, haveScript: 0x5b, distance: 0xa}, + 10: {wantLang: 0x331, haveLang: 0x139, wantScript: 0x76, haveScript: 0x5b, distance: 0xa}, + 11: {wantLang: 0x351, haveLang: 0x139, wantScript: 0x22, haveScript: 0x5b, distance: 0xa}, + 12: {wantLang: 0x395, haveLang: 0x139, wantScript: 0x83, haveScript: 0x5b, distance: 0xa}, + 13: {wantLang: 0x39d, haveLang: 0x139, wantScript: 0x36, haveScript: 0x5b, distance: 0xa}, + 14: {wantLang: 0x3be, haveLang: 0x139, wantScript: 0x5, haveScript: 0x5b, distance: 0xa}, + 15: {wantLang: 0x3fa, haveLang: 0x139, wantScript: 0x5, haveScript: 0x5b, distance: 0xa}, + 16: {wantLang: 0x40c, haveLang: 0x139, wantScript: 0xd6, haveScript: 0x5b, distance: 0xa}, + 17: {wantLang: 0x450, haveLang: 0x139, wantScript: 0xe6, haveScript: 0x5b, distance: 0xa}, + 18: {wantLang: 0x461, haveLang: 0x139, wantScript: 0xe9, haveScript: 0x5b, distance: 0xa}, + 19: {wantLang: 0x46f, haveLang: 0x139, wantScript: 0x2c, haveScript: 0x5b, distance: 0xa}, + 20: {wantLang: 0x476, haveLang: 0x3e2, wantScript: 0x5b, haveScript: 0x20, distance: 0xa}, + 21: {wantLang: 0x4b4, haveLang: 0x139, wantScript: 0x5, haveScript: 0x5b, distance: 0xa}, + 22: {wantLang: 0x4bc, haveLang: 0x3e2, wantScript: 0x5b, haveScript: 0x20, distance: 0xa}, + 23: {wantLang: 0x512, haveLang: 0x139, wantScript: 0x3e, haveScript: 0x5b, distance: 0xa}, + 24: {wantLang: 0x529, haveLang: 0x529, wantScript: 0x3b, haveScript: 0x3c, distance: 0xf}, + 25: {wantLang: 0x529, haveLang: 0x529, wantScript: 0x3c, haveScript: 0x3b, distance: 0x13}, +} // Size: 232 bytes + +var matchRegion = []regionIntelligibility{ // 15 elements + 0: {lang: 0x3a, script: 0x0, group: 0x4, distance: 0x4}, + 1: {lang: 0x3a, script: 0x0, group: 0x84, distance: 0x4}, + 2: {lang: 0x139, script: 0x0, group: 0x1, distance: 0x4}, + 3: {lang: 0x139, script: 0x0, group: 0x81, distance: 0x4}, + 4: {lang: 0x13e, script: 0x0, group: 0x3, distance: 0x4}, + 5: {lang: 0x13e, script: 0x0, group: 0x83, distance: 0x4}, + 6: {lang: 0x3c0, script: 0x0, group: 0x3, distance: 0x4}, + 7: {lang: 0x3c0, script: 0x0, group: 0x83, distance: 0x4}, + 8: {lang: 0x529, script: 0x3c, group: 0x2, distance: 0x4}, + 9: {lang: 0x529, script: 0x3c, group: 0x82, distance: 0x4}, + 10: {lang: 0x3a, script: 0x0, group: 0x80, distance: 0x5}, + 11: {lang: 0x139, script: 0x0, group: 0x80, distance: 0x5}, + 12: {lang: 0x13e, script: 0x0, group: 0x80, distance: 0x5}, + 13: {lang: 0x3c0, script: 0x0, group: 0x80, distance: 0x5}, + 14: {lang: 0x529, script: 0x3c, group: 0x80, distance: 0x5}, +} // Size: 114 bytes + +// Total table size 1473 bytes (1KiB); checksum: 7BB90B5C diff --git a/vendor/golang.org/x/text/language/tags.go b/vendor/golang.org/x/text/language/tags.go new file mode 100644 index 0000000..42ea792 --- /dev/null +++ b/vendor/golang.org/x/text/language/tags.go @@ -0,0 +1,145 @@ +// 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 language + +import "golang.org/x/text/internal/language/compact" + +// TODO: Various sets of commonly use tags and regions. + +// MustParse is like Parse, but panics if the given BCP 47 tag cannot be parsed. +// It simplifies safe initialization of Tag values. +func MustParse(s string) Tag { + t, err := Parse(s) + if err != nil { + panic(err) + } + return t +} + +// MustParse is like Parse, but panics if the given BCP 47 tag cannot be parsed. +// It simplifies safe initialization of Tag values. +func (c CanonType) MustParse(s string) Tag { + t, err := c.Parse(s) + if err != nil { + panic(err) + } + return t +} + +// MustParseBase is like ParseBase, but panics if the given base cannot be parsed. +// It simplifies safe initialization of Base values. +func MustParseBase(s string) Base { + b, err := ParseBase(s) + if err != nil { + panic(err) + } + return b +} + +// MustParseScript is like ParseScript, but panics if the given script cannot be +// parsed. It simplifies safe initialization of Script values. +func MustParseScript(s string) Script { + scr, err := ParseScript(s) + if err != nil { + panic(err) + } + return scr +} + +// MustParseRegion is like ParseRegion, but panics if the given region cannot be +// parsed. It simplifies safe initialization of Region values. +func MustParseRegion(s string) Region { + r, err := ParseRegion(s) + if err != nil { + panic(err) + } + return r +} + +var ( + und = Tag{} + + Und Tag = Tag{} + + Afrikaans Tag = Tag(compact.Afrikaans) + Amharic Tag = Tag(compact.Amharic) + Arabic Tag = Tag(compact.Arabic) + ModernStandardArabic Tag = Tag(compact.ModernStandardArabic) + Azerbaijani Tag = Tag(compact.Azerbaijani) + Bulgarian Tag = Tag(compact.Bulgarian) + Bengali Tag = Tag(compact.Bengali) + Catalan Tag = Tag(compact.Catalan) + Czech Tag = Tag(compact.Czech) + Danish Tag = Tag(compact.Danish) + German Tag = Tag(compact.German) + Greek Tag = Tag(compact.Greek) + English Tag = Tag(compact.English) + AmericanEnglish Tag = Tag(compact.AmericanEnglish) + BritishEnglish Tag = Tag(compact.BritishEnglish) + Spanish Tag = Tag(compact.Spanish) + EuropeanSpanish Tag = Tag(compact.EuropeanSpanish) + LatinAmericanSpanish Tag = Tag(compact.LatinAmericanSpanish) + Estonian Tag = Tag(compact.Estonian) + Persian Tag = Tag(compact.Persian) + Finnish Tag = Tag(compact.Finnish) + Filipino Tag = Tag(compact.Filipino) + French Tag = Tag(compact.French) + CanadianFrench Tag = Tag(compact.CanadianFrench) + Gujarati Tag = Tag(compact.Gujarati) + Hebrew Tag = Tag(compact.Hebrew) + Hindi Tag = Tag(compact.Hindi) + Croatian Tag = Tag(compact.Croatian) + Hungarian Tag = Tag(compact.Hungarian) + Armenian Tag = Tag(compact.Armenian) + Indonesian Tag = Tag(compact.Indonesian) + Icelandic Tag = Tag(compact.Icelandic) + Italian Tag = Tag(compact.Italian) + Japanese Tag = Tag(compact.Japanese) + Georgian Tag = Tag(compact.Georgian) + Kazakh Tag = Tag(compact.Kazakh) + Khmer Tag = Tag(compact.Khmer) + Kannada Tag = Tag(compact.Kannada) + Korean Tag = Tag(compact.Korean) + Kirghiz Tag = Tag(compact.Kirghiz) + Lao Tag = Tag(compact.Lao) + Lithuanian Tag = Tag(compact.Lithuanian) + Latvian Tag = Tag(compact.Latvian) + Macedonian Tag = Tag(compact.Macedonian) + Malayalam Tag = Tag(compact.Malayalam) + Mongolian Tag = Tag(compact.Mongolian) + Marathi Tag = Tag(compact.Marathi) + Malay Tag = Tag(compact.Malay) + Burmese Tag = Tag(compact.Burmese) + Nepali Tag = Tag(compact.Nepali) + Dutch Tag = Tag(compact.Dutch) + Norwegian Tag = Tag(compact.Norwegian) + Punjabi Tag = Tag(compact.Punjabi) + Polish Tag = Tag(compact.Polish) + Portuguese Tag = Tag(compact.Portuguese) + BrazilianPortuguese Tag = Tag(compact.BrazilianPortuguese) + EuropeanPortuguese Tag = Tag(compact.EuropeanPortuguese) + Romanian Tag = Tag(compact.Romanian) + Russian Tag = Tag(compact.Russian) + Sinhala Tag = Tag(compact.Sinhala) + Slovak Tag = Tag(compact.Slovak) + Slovenian Tag = Tag(compact.Slovenian) + Albanian Tag = Tag(compact.Albanian) + Serbian Tag = Tag(compact.Serbian) + SerbianLatin Tag = Tag(compact.SerbianLatin) + Swedish Tag = Tag(compact.Swedish) + Swahili Tag = Tag(compact.Swahili) + Tamil Tag = Tag(compact.Tamil) + Telugu Tag = Tag(compact.Telugu) + Thai Tag = Tag(compact.Thai) + Turkish Tag = Tag(compact.Turkish) + Ukrainian Tag = Tag(compact.Ukrainian) + Urdu Tag = Tag(compact.Urdu) + Uzbek Tag = Tag(compact.Uzbek) + Vietnamese Tag = Tag(compact.Vietnamese) + Chinese Tag = Tag(compact.Chinese) + SimplifiedChinese Tag = Tag(compact.SimplifiedChinese) + TraditionalChinese Tag = Tag(compact.TraditionalChinese) + Zulu Tag = Tag(compact.Zulu) +) |