blob: abfa17f66db1cb38f3cf6c9997bdb0ab9fe9dddd [file] [log] [blame]
William Kurkianea869482019-04-09 15:16:11 -04001// Copyright 2013 The Go Authors. All rights reserved.
2// Use of this source code is governed by a BSD-style
3// license that can be found in the LICENSE file.
4
Abhilash S.L3b494632019-07-16 15:51:09 +05305//go:generate go run gen.go -output tables.go
William Kurkianea869482019-04-09 15:16:11 -04006
7package language
8
9// TODO: Remove above NOTE after:
10// - verifying that tables are dropped correctly (most notably matcher tables).
11
12import (
William Kurkianea869482019-04-09 15:16:11 -040013 "strings"
William Kurkianea869482019-04-09 15:16:11 -040014
Abhilash S.L3b494632019-07-16 15:51:09 +053015 "golang.org/x/text/internal/language"
16 "golang.org/x/text/internal/language/compact"
William Kurkianea869482019-04-09 15:16:11 -040017)
18
19// Tag represents a BCP 47 language tag. It is used to specify an instance of a
20// specific language or locale. All language tag values are guaranteed to be
21// well-formed.
Abhilash S.L3b494632019-07-16 15:51:09 +053022type Tag compact.Tag
William Kurkianea869482019-04-09 15:16:11 -040023
Abhilash S.L3b494632019-07-16 15:51:09 +053024func makeTag(t language.Tag) (tag Tag) {
25 return Tag(compact.Make(t))
William Kurkianea869482019-04-09 15:16:11 -040026}
27
Abhilash S.L3b494632019-07-16 15:51:09 +053028func (t *Tag) tag() language.Tag {
29 return (*compact.Tag)(t).Tag()
30}
31
32func (t *Tag) isCompact() bool {
33 return (*compact.Tag)(t).IsCompact()
34}
35
36// TODO: improve performance.
37func (t *Tag) lang() language.Language { return t.tag().LangID }
38func (t *Tag) region() language.Region { return t.tag().RegionID }
39func (t *Tag) script() language.Script { return t.tag().ScriptID }
40
William Kurkianea869482019-04-09 15:16:11 -040041// Make is a convenience wrapper for Parse that omits the error.
42// In case of an error, a sensible default is returned.
43func Make(s string) Tag {
44 return Default.Make(s)
45}
46
47// Make is a convenience wrapper for c.Parse that omits the error.
48// In case of an error, a sensible default is returned.
49func (c CanonType) Make(s string) Tag {
50 t, _ := c.Parse(s)
51 return t
52}
53
54// Raw returns the raw base language, script and region, without making an
55// attempt to infer their values.
56func (t Tag) Raw() (b Base, s Script, r Region) {
Abhilash S.L3b494632019-07-16 15:51:09 +053057 tt := t.tag()
58 return Base{tt.LangID}, Script{tt.ScriptID}, Region{tt.RegionID}
William Kurkianea869482019-04-09 15:16:11 -040059}
60
61// IsRoot returns true if t is equal to language "und".
62func (t Tag) IsRoot() bool {
Abhilash S.L3b494632019-07-16 15:51:09 +053063 return compact.Tag(t).IsRoot()
William Kurkianea869482019-04-09 15:16:11 -040064}
65
66// CanonType can be used to enable or disable various types of canonicalization.
67type CanonType int
68
69const (
70 // Replace deprecated base languages with their preferred replacements.
71 DeprecatedBase CanonType = 1 << iota
72 // Replace deprecated scripts with their preferred replacements.
73 DeprecatedScript
74 // Replace deprecated regions with their preferred replacements.
75 DeprecatedRegion
76 // Remove redundant scripts.
77 SuppressScript
78 // Normalize legacy encodings. This includes legacy languages defined in
79 // CLDR as well as bibliographic codes defined in ISO-639.
80 Legacy
81 // Map the dominant language of a macro language group to the macro language
82 // subtag. For example cmn -> zh.
83 Macro
84 // The CLDR flag should be used if full compatibility with CLDR is required.
85 // There are a few cases where language.Tag may differ from CLDR. To follow all
86 // of CLDR's suggestions, use All|CLDR.
87 CLDR
88
89 // Raw can be used to Compose or Parse without Canonicalization.
90 Raw CanonType = 0
91
92 // Replace all deprecated tags with their preferred replacements.
93 Deprecated = DeprecatedBase | DeprecatedScript | DeprecatedRegion
94
95 // All canonicalizations recommended by BCP 47.
96 BCP47 = Deprecated | SuppressScript
97
98 // All canonicalizations.
99 All = BCP47 | Legacy | Macro
100
101 // Default is the canonicalization used by Parse, Make and Compose. To
102 // preserve as much information as possible, canonicalizations that remove
103 // potentially valuable information are not included. The Matcher is
104 // designed to recognize similar tags that would be the same if
105 // they were canonicalized using All.
106 Default = Deprecated | Legacy
107
108 canonLang = DeprecatedBase | Legacy | Macro
109
110 // TODO: LikelyScript, LikelyRegion: suppress similar to ICU.
111)
112
113// canonicalize returns the canonicalized equivalent of the tag and
114// whether there was any change.
Abhilash S.L3b494632019-07-16 15:51:09 +0530115func canonicalize(c CanonType, t language.Tag) (language.Tag, bool) {
William Kurkianea869482019-04-09 15:16:11 -0400116 if c == Raw {
117 return t, false
118 }
119 changed := false
120 if c&SuppressScript != 0 {
Abhilash S.L3b494632019-07-16 15:51:09 +0530121 if t.LangID.SuppressScript() == t.ScriptID {
122 t.ScriptID = 0
William Kurkianea869482019-04-09 15:16:11 -0400123 changed = true
124 }
125 }
126 if c&canonLang != 0 {
127 for {
Abhilash S.L3b494632019-07-16 15:51:09 +0530128 if l, aliasType := t.LangID.Canonicalize(); l != t.LangID {
William Kurkianea869482019-04-09 15:16:11 -0400129 switch aliasType {
Abhilash S.L3b494632019-07-16 15:51:09 +0530130 case language.Legacy:
William Kurkianea869482019-04-09 15:16:11 -0400131 if c&Legacy != 0 {
Abhilash S.L3b494632019-07-16 15:51:09 +0530132 if t.LangID == _sh && t.ScriptID == 0 {
133 t.ScriptID = _Latn
William Kurkianea869482019-04-09 15:16:11 -0400134 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530135 t.LangID = l
William Kurkianea869482019-04-09 15:16:11 -0400136 changed = true
137 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530138 case language.Macro:
William Kurkianea869482019-04-09 15:16:11 -0400139 if c&Macro != 0 {
140 // We deviate here from CLDR. The mapping "nb" -> "no"
141 // qualifies as a typical Macro language mapping. However,
142 // for legacy reasons, CLDR maps "no", the macro language
143 // code for Norwegian, to the dominant variant "nb". This
144 // change is currently under consideration for CLDR as well.
Abhilash S.L3b494632019-07-16 15:51:09 +0530145 // See https://unicode.org/cldr/trac/ticket/2698 and also
146 // https://unicode.org/cldr/trac/ticket/1790 for some of the
William Kurkianea869482019-04-09 15:16:11 -0400147 // practical implications. TODO: this check could be removed
148 // if CLDR adopts this change.
Abhilash S.L3b494632019-07-16 15:51:09 +0530149 if c&CLDR == 0 || t.LangID != _nb {
William Kurkianea869482019-04-09 15:16:11 -0400150 changed = true
Abhilash S.L3b494632019-07-16 15:51:09 +0530151 t.LangID = l
William Kurkianea869482019-04-09 15:16:11 -0400152 }
153 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530154 case language.Deprecated:
William Kurkianea869482019-04-09 15:16:11 -0400155 if c&DeprecatedBase != 0 {
Abhilash S.L3b494632019-07-16 15:51:09 +0530156 if t.LangID == _mo && t.RegionID == 0 {
157 t.RegionID = _MD
William Kurkianea869482019-04-09 15:16:11 -0400158 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530159 t.LangID = l
William Kurkianea869482019-04-09 15:16:11 -0400160 changed = true
161 // Other canonicalization types may still apply.
162 continue
163 }
164 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530165 } else if c&Legacy != 0 && t.LangID == _no && c&CLDR != 0 {
166 t.LangID = _nb
William Kurkianea869482019-04-09 15:16:11 -0400167 changed = true
168 }
169 break
170 }
171 }
172 if c&DeprecatedScript != 0 {
Abhilash S.L3b494632019-07-16 15:51:09 +0530173 if t.ScriptID == _Qaai {
William Kurkianea869482019-04-09 15:16:11 -0400174 changed = true
Abhilash S.L3b494632019-07-16 15:51:09 +0530175 t.ScriptID = _Zinh
William Kurkianea869482019-04-09 15:16:11 -0400176 }
177 }
178 if c&DeprecatedRegion != 0 {
Abhilash S.L3b494632019-07-16 15:51:09 +0530179 if r := t.RegionID.Canonicalize(); r != t.RegionID {
William Kurkianea869482019-04-09 15:16:11 -0400180 changed = true
Abhilash S.L3b494632019-07-16 15:51:09 +0530181 t.RegionID = r
William Kurkianea869482019-04-09 15:16:11 -0400182 }
183 }
184 return t, changed
185}
186
187// Canonicalize returns the canonicalized equivalent of the tag.
188func (c CanonType) Canonicalize(t Tag) (Tag, error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530189 // First try fast path.
190 if t.isCompact() {
191 if _, changed := canonicalize(c, compact.Tag(t).Tag()); !changed {
192 return t, nil
193 }
194 }
195 // It is unlikely that one will canonicalize a tag after matching. So do
196 // a slow but simple approach here.
197 if tag, changed := canonicalize(c, t.tag()); changed {
198 tag.RemakeString()
199 return makeTag(tag), nil
William Kurkianea869482019-04-09 15:16:11 -0400200 }
201 return t, nil
Abhilash S.L3b494632019-07-16 15:51:09 +0530202
William Kurkianea869482019-04-09 15:16:11 -0400203}
204
205// Confidence indicates the level of certainty for a given return value.
206// For example, Serbian may be written in Cyrillic or Latin script.
207// The confidence level indicates whether a value was explicitly specified,
208// whether it is typically the only possible value, or whether there is
209// an ambiguity.
210type Confidence int
211
212const (
213 No Confidence = iota // full confidence that there was no match
214 Low // most likely value picked out of a set of alternatives
215 High // value is generally assumed to be the correct match
216 Exact // exact match or explicitly specified value
217)
218
219var confName = []string{"No", "Low", "High", "Exact"}
220
221func (c Confidence) String() string {
222 return confName[c]
223}
224
William Kurkianea869482019-04-09 15:16:11 -0400225// String returns the canonical string representation of the language tag.
226func (t Tag) String() string {
Abhilash S.L3b494632019-07-16 15:51:09 +0530227 return t.tag().String()
William Kurkianea869482019-04-09 15:16:11 -0400228}
229
230// MarshalText implements encoding.TextMarshaler.
231func (t Tag) MarshalText() (text []byte, err error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530232 return t.tag().MarshalText()
William Kurkianea869482019-04-09 15:16:11 -0400233}
234
235// UnmarshalText implements encoding.TextUnmarshaler.
236func (t *Tag) UnmarshalText(text []byte) error {
Abhilash S.L3b494632019-07-16 15:51:09 +0530237 var tag language.Tag
238 err := tag.UnmarshalText(text)
239 *t = makeTag(tag)
William Kurkianea869482019-04-09 15:16:11 -0400240 return err
241}
242
243// Base returns the base language of the language tag. If the base language is
244// unspecified, an attempt will be made to infer it from the context.
245// It uses a variant of CLDR's Add Likely Subtags algorithm. This is subject to change.
246func (t Tag) Base() (Base, Confidence) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530247 if b := t.lang(); b != 0 {
248 return Base{b}, Exact
William Kurkianea869482019-04-09 15:16:11 -0400249 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530250 tt := t.tag()
William Kurkianea869482019-04-09 15:16:11 -0400251 c := High
Abhilash S.L3b494632019-07-16 15:51:09 +0530252 if tt.ScriptID == 0 && !tt.RegionID.IsCountry() {
William Kurkianea869482019-04-09 15:16:11 -0400253 c = Low
254 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530255 if tag, err := tt.Maximize(); err == nil && tag.LangID != 0 {
256 return Base{tag.LangID}, c
William Kurkianea869482019-04-09 15:16:11 -0400257 }
258 return Base{0}, No
259}
260
261// Script infers the script for the language tag. If it was not explicitly given, it will infer
262// a most likely candidate.
263// If more than one script is commonly used for a language, the most likely one
264// is returned with a low confidence indication. For example, it returns (Cyrl, Low)
265// for Serbian.
266// If a script cannot be inferred (Zzzz, No) is returned. We do not use Zyyy (undetermined)
267// as one would suspect from the IANA registry for BCP 47. In a Unicode context Zyyy marks
268// common characters (like 1, 2, 3, '.', etc.) and is therefore more like multiple scripts.
Abhilash S.L3b494632019-07-16 15:51:09 +0530269// See https://www.unicode.org/reports/tr24/#Values for more details. Zzzz is also used for
William Kurkianea869482019-04-09 15:16:11 -0400270// unknown value in CLDR. (Zzzz, Exact) is returned if Zzzz was explicitly specified.
271// Note that an inferred script is never guaranteed to be the correct one. Latin is
272// almost exclusively used for Afrikaans, but Arabic has been used for some texts
273// in the past. Also, the script that is commonly used may change over time.
274// It uses a variant of CLDR's Add Likely Subtags algorithm. This is subject to change.
275func (t Tag) Script() (Script, Confidence) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530276 if scr := t.script(); scr != 0 {
277 return Script{scr}, Exact
William Kurkianea869482019-04-09 15:16:11 -0400278 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530279 tt := t.tag()
280 sc, c := language.Script(_Zzzz), No
281 if scr := tt.LangID.SuppressScript(); scr != 0 {
282 // Note: it is not always the case that a language with a suppress
283 // script value is only written in one script (e.g. kk, ms, pa).
284 if tt.RegionID == 0 {
285 return Script{scr}, High
William Kurkianea869482019-04-09 15:16:11 -0400286 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530287 sc, c = scr, High
William Kurkianea869482019-04-09 15:16:11 -0400288 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530289 if tag, err := tt.Maximize(); err == nil {
290 if tag.ScriptID != sc {
291 sc, c = tag.ScriptID, Low
William Kurkianea869482019-04-09 15:16:11 -0400292 }
293 } else {
Abhilash S.L3b494632019-07-16 15:51:09 +0530294 tt, _ = canonicalize(Deprecated|Macro, tt)
295 if tag, err := tt.Maximize(); err == nil && tag.ScriptID != sc {
296 sc, c = tag.ScriptID, Low
William Kurkianea869482019-04-09 15:16:11 -0400297 }
298 }
299 return Script{sc}, c
300}
301
302// Region returns the region for the language tag. If it was not explicitly given, it will
303// infer a most likely candidate from the context.
304// It uses a variant of CLDR's Add Likely Subtags algorithm. This is subject to change.
305func (t Tag) Region() (Region, Confidence) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530306 if r := t.region(); r != 0 {
307 return Region{r}, Exact
William Kurkianea869482019-04-09 15:16:11 -0400308 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530309 tt := t.tag()
310 if tt, err := tt.Maximize(); err == nil {
311 return Region{tt.RegionID}, Low // TODO: differentiate between high and low.
William Kurkianea869482019-04-09 15:16:11 -0400312 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530313 tt, _ = canonicalize(Deprecated|Macro, tt)
314 if tag, err := tt.Maximize(); err == nil {
315 return Region{tag.RegionID}, Low
William Kurkianea869482019-04-09 15:16:11 -0400316 }
317 return Region{_ZZ}, No // TODO: return world instead of undetermined?
318}
319
Abhilash S.L3b494632019-07-16 15:51:09 +0530320// Variants returns the variants specified explicitly for this language tag.
William Kurkianea869482019-04-09 15:16:11 -0400321// or nil if no variant was specified.
322func (t Tag) Variants() []Variant {
Abhilash S.L3b494632019-07-16 15:51:09 +0530323 if !compact.Tag(t).MayHaveVariants() {
324 return nil
325 }
William Kurkianea869482019-04-09 15:16:11 -0400326 v := []Variant{}
Abhilash S.L3b494632019-07-16 15:51:09 +0530327 x, str := "", t.tag().Variants()
328 for str != "" {
329 x, str = nextToken(str)
330 v = append(v, Variant{x})
William Kurkianea869482019-04-09 15:16:11 -0400331 }
332 return v
333}
334
335// Parent returns the CLDR parent of t. In CLDR, missing fields in data for a
336// specific language are substituted with fields from the parent language.
337// The parent for a language may change for newer versions of CLDR.
Abhilash S.L3b494632019-07-16 15:51:09 +0530338//
339// Parent returns a tag for a less specific language that is mutually
340// intelligible or Und if there is no such language. This may not be the same as
341// simply stripping the last BCP 47 subtag. For instance, the parent of "zh-TW"
342// is "zh-Hant", and the parent of "zh-Hant" is "und".
William Kurkianea869482019-04-09 15:16:11 -0400343func (t Tag) Parent() Tag {
Abhilash S.L3b494632019-07-16 15:51:09 +0530344 return Tag(compact.Tag(t).Parent())
William Kurkianea869482019-04-09 15:16:11 -0400345}
346
347// returns token t and the rest of the string.
348func nextToken(s string) (t, tail string) {
349 p := strings.Index(s[1:], "-")
350 if p == -1 {
351 return s[1:], ""
352 }
353 p++
354 return s[1:p], s[p:]
355}
356
357// Extension is a single BCP 47 extension.
358type Extension struct {
359 s string
360}
361
362// String returns the string representation of the extension, including the
363// type tag.
364func (e Extension) String() string {
365 return e.s
366}
367
368// ParseExtension parses s as an extension and returns it on success.
369func ParseExtension(s string) (e Extension, err error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530370 ext, err := language.ParseExtension(s)
371 return Extension{ext}, err
William Kurkianea869482019-04-09 15:16:11 -0400372}
373
374// Type returns the one-byte extension type of e. It returns 0 for the zero
375// exception.
376func (e Extension) Type() byte {
377 if e.s == "" {
378 return 0
379 }
380 return e.s[0]
381}
382
383// Tokens returns the list of tokens of e.
384func (e Extension) Tokens() []string {
385 return strings.Split(e.s, "-")
386}
387
388// Extension returns the extension of type x for tag t. It will return
389// false for ok if t does not have the requested extension. The returned
390// extension will be invalid in this case.
391func (t Tag) Extension(x byte) (ext Extension, ok bool) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530392 if !compact.Tag(t).MayHaveExtensions() {
393 return Extension{}, false
William Kurkianea869482019-04-09 15:16:11 -0400394 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530395 e, ok := t.tag().Extension(x)
396 return Extension{e}, ok
William Kurkianea869482019-04-09 15:16:11 -0400397}
398
399// Extensions returns all extensions of t.
400func (t Tag) Extensions() []Extension {
Abhilash S.L3b494632019-07-16 15:51:09 +0530401 if !compact.Tag(t).MayHaveExtensions() {
402 return nil
403 }
William Kurkianea869482019-04-09 15:16:11 -0400404 e := []Extension{}
Abhilash S.L3b494632019-07-16 15:51:09 +0530405 for _, ext := range t.tag().Extensions() {
William Kurkianea869482019-04-09 15:16:11 -0400406 e = append(e, Extension{ext})
407 }
408 return e
409}
410
411// TypeForKey returns the type associated with the given key, where key and type
412// are of the allowed values defined for the Unicode locale extension ('u') in
Abhilash S.L3b494632019-07-16 15:51:09 +0530413// https://www.unicode.org/reports/tr35/#Unicode_Language_and_Locale_Identifiers.
William Kurkianea869482019-04-09 15:16:11 -0400414// TypeForKey will traverse the inheritance chain to get the correct value.
415func (t Tag) TypeForKey(key string) string {
Abhilash S.L3b494632019-07-16 15:51:09 +0530416 if !compact.Tag(t).MayHaveExtensions() {
417 if key != "rg" && key != "va" {
418 return ""
419 }
William Kurkianea869482019-04-09 15:16:11 -0400420 }
Abhilash S.L3b494632019-07-16 15:51:09 +0530421 return t.tag().TypeForKey(key)
William Kurkianea869482019-04-09 15:16:11 -0400422}
423
William Kurkianea869482019-04-09 15:16:11 -0400424// SetTypeForKey returns a new Tag with the key set to type, where key and type
425// are of the allowed values defined for the Unicode locale extension ('u') in
Abhilash S.L3b494632019-07-16 15:51:09 +0530426// https://www.unicode.org/reports/tr35/#Unicode_Language_and_Locale_Identifiers.
William Kurkianea869482019-04-09 15:16:11 -0400427// An empty value removes an existing pair with the same key.
428func (t Tag) SetTypeForKey(key, value string) (Tag, error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530429 tt, err := t.tag().SetTypeForKey(key, value)
430 return makeTag(tt), err
William Kurkianea869482019-04-09 15:16:11 -0400431}
432
Abhilash S.L3b494632019-07-16 15:51:09 +0530433// NumCompactTags is the number of compact tags. The maximum tag is
434// NumCompactTags-1.
435const NumCompactTags = compact.NumCompactTags
William Kurkianea869482019-04-09 15:16:11 -0400436
437// CompactIndex returns an index, where 0 <= index < NumCompactTags, for tags
Abhilash S.L3b494632019-07-16 15:51:09 +0530438// for which data exists in the text repository.The index will change over time
439// and should not be stored in persistent storage. If t does not match a compact
440// index, exact will be false and the compact index will be returned for the
441// first match after repeatedly taking the Parent of t.
442func CompactIndex(t Tag) (index int, exact bool) {
443 id, exact := compact.LanguageID(compact.Tag(t))
444 return int(id), exact
William Kurkianea869482019-04-09 15:16:11 -0400445}
446
Abhilash S.L3b494632019-07-16 15:51:09 +0530447var root = language.Tag{}
448
William Kurkianea869482019-04-09 15:16:11 -0400449// Base is an ISO 639 language code, used for encoding the base language
450// of a language tag.
451type Base struct {
Abhilash S.L3b494632019-07-16 15:51:09 +0530452 langID language.Language
William Kurkianea869482019-04-09 15:16:11 -0400453}
454
455// ParseBase parses a 2- or 3-letter ISO 639 code.
456// It returns a ValueError if s is a well-formed but unknown language identifier
457// or another error if another error occurred.
458func ParseBase(s string) (Base, error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530459 l, err := language.ParseBase(s)
William Kurkianea869482019-04-09 15:16:11 -0400460 return Base{l}, err
461}
462
Abhilash S.L3b494632019-07-16 15:51:09 +0530463// String returns the BCP 47 representation of the base language.
464func (b Base) String() string {
465 return b.langID.String()
466}
467
468// ISO3 returns the ISO 639-3 language code.
469func (b Base) ISO3() string {
470 return b.langID.ISO3()
471}
472
473// IsPrivateUse reports whether this language code is reserved for private use.
474func (b Base) IsPrivateUse() bool {
475 return b.langID.IsPrivateUse()
476}
477
William Kurkianea869482019-04-09 15:16:11 -0400478// Script is a 4-letter ISO 15924 code for representing scripts.
479// It is idiomatically represented in title case.
480type Script struct {
Abhilash S.L3b494632019-07-16 15:51:09 +0530481 scriptID language.Script
William Kurkianea869482019-04-09 15:16:11 -0400482}
483
484// ParseScript parses a 4-letter ISO 15924 code.
485// It returns a ValueError if s is a well-formed but unknown script identifier
486// or another error if another error occurred.
487func ParseScript(s string) (Script, error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530488 sc, err := language.ParseScript(s)
William Kurkianea869482019-04-09 15:16:11 -0400489 return Script{sc}, err
490}
491
Abhilash S.L3b494632019-07-16 15:51:09 +0530492// String returns the script code in title case.
493// It returns "Zzzz" for an unspecified script.
494func (s Script) String() string {
495 return s.scriptID.String()
496}
497
498// IsPrivateUse reports whether this script code is reserved for private use.
499func (s Script) IsPrivateUse() bool {
500 return s.scriptID.IsPrivateUse()
501}
502
William Kurkianea869482019-04-09 15:16:11 -0400503// Region is an ISO 3166-1 or UN M.49 code for representing countries and regions.
504type Region struct {
Abhilash S.L3b494632019-07-16 15:51:09 +0530505 regionID language.Region
William Kurkianea869482019-04-09 15:16:11 -0400506}
507
508// EncodeM49 returns the Region for the given UN M.49 code.
509// It returns an error if r is not a valid code.
510func EncodeM49(r int) (Region, error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530511 rid, err := language.EncodeM49(r)
William Kurkianea869482019-04-09 15:16:11 -0400512 return Region{rid}, err
513}
514
515// ParseRegion parses a 2- or 3-letter ISO 3166-1 or a UN M.49 code.
516// It returns a ValueError if s is a well-formed but unknown region identifier
517// or another error if another error occurred.
518func ParseRegion(s string) (Region, error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530519 r, err := language.ParseRegion(s)
William Kurkianea869482019-04-09 15:16:11 -0400520 return Region{r}, err
521}
522
Abhilash S.L3b494632019-07-16 15:51:09 +0530523// String returns the BCP 47 representation for the region.
524// It returns "ZZ" for an unspecified region.
525func (r Region) String() string {
526 return r.regionID.String()
527}
528
529// ISO3 returns the 3-letter ISO code of r.
530// Note that not all regions have a 3-letter ISO code.
531// In such cases this method returns "ZZZ".
532func (r Region) ISO3() string {
533 return r.regionID.ISO3()
534}
535
536// M49 returns the UN M.49 encoding of r, or 0 if this encoding
537// is not defined for r.
538func (r Region) M49() int {
539 return r.regionID.M49()
540}
541
542// IsPrivateUse reports whether r has the ISO 3166 User-assigned status. This
543// may include private-use tags that are assigned by CLDR and used in this
544// implementation. So IsPrivateUse and IsCountry can be simultaneously true.
545func (r Region) IsPrivateUse() bool {
546 return r.regionID.IsPrivateUse()
547}
548
William Kurkianea869482019-04-09 15:16:11 -0400549// IsCountry returns whether this region is a country or autonomous area. This
550// includes non-standard definitions from CLDR.
551func (r Region) IsCountry() bool {
Abhilash S.L3b494632019-07-16 15:51:09 +0530552 return r.regionID.IsCountry()
William Kurkianea869482019-04-09 15:16:11 -0400553}
554
555// IsGroup returns whether this region defines a collection of regions. This
556// includes non-standard definitions from CLDR.
557func (r Region) IsGroup() bool {
Abhilash S.L3b494632019-07-16 15:51:09 +0530558 return r.regionID.IsGroup()
William Kurkianea869482019-04-09 15:16:11 -0400559}
560
561// Contains returns whether Region c is contained by Region r. It returns true
562// if c == r.
563func (r Region) Contains(c Region) bool {
Abhilash S.L3b494632019-07-16 15:51:09 +0530564 return r.regionID.Contains(c.regionID)
William Kurkianea869482019-04-09 15:16:11 -0400565}
566
William Kurkianea869482019-04-09 15:16:11 -0400567// TLD returns the country code top-level domain (ccTLD). UK is returned for GB.
568// In all other cases it returns either the region itself or an error.
569//
570// This method may return an error for a region for which there exists a
571// canonical form with a ccTLD. To get that ccTLD canonicalize r first. The
572// region will already be canonicalized it was obtained from a Tag that was
573// obtained using any of the default methods.
574func (r Region) TLD() (Region, error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530575 tld, err := r.regionID.TLD()
576 return Region{tld}, err
William Kurkianea869482019-04-09 15:16:11 -0400577}
578
579// Canonicalize returns the region or a possible replacement if the region is
580// deprecated. It will not return a replacement for deprecated regions that
581// are split into multiple regions.
582func (r Region) Canonicalize() Region {
Abhilash S.L3b494632019-07-16 15:51:09 +0530583 return Region{r.regionID.Canonicalize()}
William Kurkianea869482019-04-09 15:16:11 -0400584}
585
586// Variant represents a registered variant of a language as defined by BCP 47.
587type Variant struct {
588 variant string
589}
590
591// ParseVariant parses and returns a Variant. An error is returned if s is not
592// a valid variant.
593func ParseVariant(s string) (Variant, error) {
Abhilash S.L3b494632019-07-16 15:51:09 +0530594 v, err := language.ParseVariant(s)
595 return Variant{v.String()}, err
William Kurkianea869482019-04-09 15:16:11 -0400596}
597
598// String returns the string representation of the variant.
599func (v Variant) String() string {
600 return v.variant
601}