1 /*
2 * Copyright (c) 1996, 2014, Oracle and/or its affiliates. All rights reserved.
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 *
5 * This code is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU General Public License version 2 only, as
7 * published by the Free Software Foundation. Oracle designates this
8 * particular file as subject to the "Classpath" exception as provided
9 * by Oracle in the LICENSE file that accompanied this code.
10 *
11 * This code is distributed in the hope that it will be useful, but WITHOUT
12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 * version 2 for more details (a copy is included in the LICENSE file that
15 * accompanied this code).
16 *
17 * You should have received a copy of the GNU General Public License version
18 * 2 along with this work; if not, write to the Free Software Foundation,
19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
20 *
21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
22 * or visit www.oracle.com if you need additional information or have any
23 * questions.
24 */
25
26 /*
27 * (C) Copyright Taligent, Inc. 1996, 1997 - All Rights Reserved
28 * (C) Copyright IBM Corp. 1996 - 1998 - All Rights Reserved
29 *
30 * The original version of this source code and documentation
31 * is copyrighted and owned by Taligent, Inc., a wholly-owned
32 * subsidiary of IBM. These materials are provided under terms
33 * of a License Agreement between Taligent and Sun. This technology
34 * is protected by multiple US and International patents.
35 *
36 * This notice and attribution to Taligent may not be removed.
37 * Taligent is a registered trademark of Taligent, Inc.
38 *
39 */
40
41 package java.util;
42
43 import java.io.IOException;
44 import java.io.ObjectInputStream;
45 import java.io.ObjectOutputStream;
46 import java.io.ObjectStreamField;
47 import java.io.Serializable;
48 import java.security.AccessController;
49 import java.text.MessageFormat;
50 import java.util.spi.LocaleNameProvider;
51
52 import sun.security.action.GetPropertyAction;
53 import sun.util.locale.BaseLocale;
54 import sun.util.locale.InternalLocaleBuilder;
55 import sun.util.locale.LanguageTag;
56 import sun.util.locale.LocaleExtensions;
57 import sun.util.locale.LocaleMatcher;
58 import sun.util.locale.LocaleObjectCache;
59 import sun.util.locale.LocaleSyntaxException;
60 import sun.util.locale.LocaleUtils;
61 import sun.util.locale.ParseStatus;
62 import sun.util.locale.provider.LocaleProviderAdapter;
63 import sun.util.locale.provider.LocaleResources;
64 import sun.util.locale.provider.LocaleServiceProviderPool;
65
66 /**
67 * A <code>Locale</code> object represents a specific geographical, political,
68 * or cultural region. An operation that requires a <code>Locale</code> to perform
69 * its task is called <em>locale-sensitive</em> and uses the <code>Locale</code>
70 * to tailor information for the user. For example, displaying a number
71 * is a locale-sensitive operation— the number should be formatted
72 * according to the customs and conventions of the user's native country,
73 * region, or culture.
74 *
75 * <p> The {@code Locale} class implements IETF BCP 47 which is composed of
76 * <a href="http://tools.ietf.org/html/rfc4647">RFC 4647 "Matching of Language
77 * Tags"</a> and <a href="http://tools.ietf.org/html/rfc5646">RFC 5646 "Tags
78 * for Identifying Languages"</a> with support for the LDML (UTS#35, "Unicode
79 * Locale Data Markup Language") BCP 47-compatible extensions for locale data
80 * exchange.
81 *
82 * <p> A <code>Locale</code> object logically consists of the fields
83 * described below.
84 *
85 * <dl>
86 * <dt><a name="def_language"><b>language</b></a></dt>
87 *
88 * <dd>ISO 639 alpha-2 or alpha-3 language code, or registered
89 * language subtags up to 8 alpha letters (for future enhancements).
90 * When a language has both an alpha-2 code and an alpha-3 code, the
91 * alpha-2 code must be used. You can find a full list of valid
92 * language codes in the IANA Language Subtag Registry (search for
93 * "Type: language"). The language field is case insensitive, but
94 * <code>Locale</code> always canonicalizes to lower case.</dd>
95 *
96 * <dd>Well-formed language values have the form
97 * <code>[a-zA-Z]{2,8}</code>. Note that this is not the full
98 * BCP47 language production, since it excludes extlang. They are
99 * not needed since modern three-letter language codes replace
100 * them.</dd>
101 *
102 * <dd>Example: "en" (English), "ja" (Japanese), "kok" (Konkani)</dd>
103 *
104 * <dt><a name="def_script"><b>script</b></a></dt>
105 *
106 * <dd>ISO 15924 alpha-4 script code. You can find a full list of
107 * valid script codes in the IANA Language Subtag Registry (search
108 * for "Type: script"). The script field is case insensitive, but
109 * <code>Locale</code> always canonicalizes to title case (the first
110 * letter is upper case and the rest of the letters are lower
111 * case).</dd>
112 *
113 * <dd>Well-formed script values have the form
114 * <code>[a-zA-Z]{4}</code></dd>
115 *
116 * <dd>Example: "Latn" (Latin), "Cyrl" (Cyrillic)</dd>
117 *
118 * <dt><a name="def_region"><b>country (region)</b></a></dt>
119 *
120 * <dd>ISO 3166 alpha-2 country code or UN M.49 numeric-3 area code.
121 * You can find a full list of valid country and region codes in the
122 * IANA Language Subtag Registry (search for "Type: region"). The
123 * country (region) field is case insensitive, but
124 * <code>Locale</code> always canonicalizes to upper case.</dd>
125 *
126 * <dd>Well-formed country/region values have
127 * the form <code>[a-zA-Z]{2} | [0-9]{3}</code></dd>
128 *
129 * <dd>Example: "US" (United States), "FR" (France), "029"
130 * (Caribbean)</dd>
131 *
132 * <dt><a name="def_variant"><b>variant</b></a></dt>
133 *
134 * <dd>Any arbitrary value used to indicate a variation of a
135 * <code>Locale</code>. Where there are two or more variant values
136 * each indicating its own semantics, these values should be ordered
137 * by importance, with most important first, separated by
138 * underscore('_'). The variant field is case sensitive.</dd>
139 *
140 * <dd>Note: IETF BCP 47 places syntactic restrictions on variant
141 * subtags. Also BCP 47 subtags are strictly used to indicate
142 * additional variations that define a language or its dialects that
143 * are not covered by any combinations of language, script and
144 * region subtags. You can find a full list of valid variant codes
145 * in the IANA Language Subtag Registry (search for "Type: variant").
146 *
147 * <p>However, the variant field in <code>Locale</code> has
148 * historically been used for any kind of variation, not just
149 * language variations. For example, some supported variants
150 * available in Java SE Runtime Environments indicate alternative
151 * cultural behaviors such as calendar type or number script. In
152 * BCP 47 this kind of information, which does not identify the
153 * language, is supported by extension subtags or private use
154 * subtags.</dd>
155 *
156 * <dd>Well-formed variant values have the form <code>SUBTAG
157 * (('_'|'-') SUBTAG)*</code> where <code>SUBTAG =
158 * [0-9][0-9a-zA-Z]{3} | [0-9a-zA-Z]{5,8}</code>. (Note: BCP 47 only
159 * uses hyphen ('-') as a delimiter, this is more lenient).</dd>
160 *
161 * <dd>Example: "polyton" (Polytonic Greek), "POSIX"</dd>
162 *
163 * <dt><a name="def_extensions"><b>extensions</b></a></dt>
164 *
165 * <dd>A map from single character keys to string values, indicating
166 * extensions apart from language identification. The extensions in
167 * <code>Locale</code> implement the semantics and syntax of BCP 47
168 * extension subtags and private use subtags. The extensions are
169 * case insensitive, but <code>Locale</code> canonicalizes all
170 * extension keys and values to lower case. Note that extensions
171 * cannot have empty values.</dd>
172 *
173 * <dd>Well-formed keys are single characters from the set
174 * <code>[0-9a-zA-Z]</code>. Well-formed values have the form
175 * <code>SUBTAG ('-' SUBTAG)*</code> where for the key 'x'
176 * <code>SUBTAG = [0-9a-zA-Z]{1,8}</code> and for other keys
177 * <code>SUBTAG = [0-9a-zA-Z]{2,8}</code> (that is, 'x' allows
178 * single-character subtags).</dd>
179 *
180 * <dd>Example: key="u"/value="ca-japanese" (Japanese Calendar),
181 * key="x"/value="java-1-7"</dd>
182 * </dl>
183 *
184 * <b>Note:</b> Although BCP 47 requires field values to be registered
185 * in the IANA Language Subtag Registry, the <code>Locale</code> class
186 * does not provide any validation features. The <code>Builder</code>
187 * only checks if an individual field satisfies the syntactic
188 * requirement (is well-formed), but does not validate the value
189 * itself. See {@link Builder} for details.
190 *
191 * <h3><a name="def_locale_extension">Unicode locale/language extension</a></h3>
192 *
193 * <p>UTS#35, "Unicode Locale Data Markup Language" defines optional
194 * attributes and keywords to override or refine the default behavior
195 * associated with a locale. A keyword is represented by a pair of
196 * key and type. For example, "nu-thai" indicates that Thai local
197 * digits (value:"thai") should be used for formatting numbers
198 * (key:"nu").
199 *
200 * <p>The keywords are mapped to a BCP 47 extension value using the
201 * extension key 'u' ({@link #UNICODE_LOCALE_EXTENSION}). The above
202 * example, "nu-thai", becomes the extension "u-nu-thai".code
203 *
204 * <p>Thus, when a <code>Locale</code> object contains Unicode locale
205 * attributes and keywords,
206 * <code>getExtension(UNICODE_LOCALE_EXTENSION)</code> will return a
207 * String representing this information, for example, "nu-thai". The
208 * <code>Locale</code> class also provides {@link
209 * #getUnicodeLocaleAttributes}, {@link #getUnicodeLocaleKeys}, and
210 * {@link #getUnicodeLocaleType} which allow you to access Unicode
211 * locale attributes and key/type pairs directly. When represented as
212 * a string, the Unicode Locale Extension lists attributes
213 * alphabetically, followed by key/type sequences with keys listed
214 * alphabetically (the order of subtags comprising a key's type is
215 * fixed when the type is defined)
216 *
217 * <p>A well-formed locale key has the form
218 * <code>[0-9a-zA-Z]{2}</code>. A well-formed locale type has the
219 * form <code>"" | [0-9a-zA-Z]{3,8} ('-' [0-9a-zA-Z]{3,8})*</code> (it
220 * can be empty, or a series of subtags 3-8 alphanums in length). A
221 * well-formed locale attribute has the form
222 * <code>[0-9a-zA-Z]{3,8}</code> (it is a single subtag with the same
223 * form as a locale type subtag).
224 *
225 * <p>The Unicode locale extension specifies optional behavior in
226 * locale-sensitive services. Although the LDML specification defines
227 * various keys and values, actual locale-sensitive service
228 * implementations in a Java Runtime Environment might not support any
229 * particular Unicode locale attributes or key/type pairs.
230 *
231 * <h4>Creating a Locale</h4>
232 *
233 * <p>There are several different ways to create a <code>Locale</code>
234 * object.
235 *
236 * <h5>Builder</h5>
237 *
238 * <p>Using {@link Builder} you can construct a <code>Locale</code> object
239 * that conforms to BCP 47 syntax.
240 *
241 * <h5>Constructors</h5>
242 *
243 * <p>The <code>Locale</code> class provides three constructors:
244 * <blockquote>
245 * <pre>
246 * {@link #Locale(String language)}
247 * {@link #Locale(String language, String country)}
248 * {@link #Locale(String language, String country, String variant)}
249 * </pre>
250 * </blockquote>
251 * These constructors allow you to create a <code>Locale</code> object
252 * with language, country and variant, but you cannot specify
253 * script or extensions.
254 *
255 * <h5>Factory Methods</h5>
256 *
257 * <p>The method {@link #forLanguageTag} creates a <code>Locale</code>
258 * object for a well-formed BCP 47 language tag.
259 *
260 * <h5>Locale Constants</h5>
261 *
262 * <p>The <code>Locale</code> class provides a number of convenient constants
263 * that you can use to create <code>Locale</code> objects for commonly used
264 * locales. For example, the following creates a <code>Locale</code> object
265 * for the United States:
266 * <blockquote>
267 * <pre>
268 * Locale.US
269 * </pre>
270 * </blockquote>
271 *
272 * <h4><a name="LocaleMatching">Locale Matching</a></h4>
273 *
274 * <p>If an application or a system is internationalized and provides localized
275 * resources for multiple locales, it sometimes needs to find one or more
276 * locales (or language tags) which meet each user's specific preferences. Note
277 * that a term "language tag" is used interchangeably with "locale" in this
278 * locale matching documentation.
279 *
280 * <p>In order to do matching a user's preferred locales to a set of language
281 * tags, <a href="http://tools.ietf.org/html/rfc4647">RFC 4647 Matching of
282 * Language Tags</a> defines two mechanisms: filtering and lookup.
283 * <em>Filtering</em> is used to get all matching locales, whereas
284 * <em>lookup</em> is to choose the best matching locale.
285 * Matching is done case-insensitively. These matching mechanisms are described
286 * in the following sections.
287 *
288 * <p>A user's preference is called a <em>Language Priority List</em> and is
289 * expressed as a list of language ranges. There are syntactically two types of
290 * language ranges: basic and extended. See
291 * {@link Locale.LanguageRange Locale.LanguageRange} for details.
292 *
293 * <h5>Filtering</h5>
294 *
295 * <p>The filtering operation returns all matching language tags. It is defined
296 * in RFC 4647 as follows:
297 * "In filtering, each language range represents the least specific language
298 * tag (that is, the language tag with fewest number of subtags) that is an
299 * acceptable match. All of the language tags in the matching set of tags will
300 * have an equal or greater number of subtags than the language range. Every
301 * non-wildcard subtag in the language range will appear in every one of the
302 * matching language tags."
303 *
304 * <p>There are two types of filtering: filtering for basic language ranges
305 * (called "basic filtering") and filtering for extended language ranges
306 * (called "extended filtering"). They may return different results by what
307 * kind of language ranges are included in the given Language Priority List.
308 * {@link Locale.FilteringMode} is a parameter to specify how filtering should
309 * be done.
310 *
311 * <h5>Lookup</h5>
312 *
313 * <p>The lookup operation returns the best matching language tags. It is
314 * defined in RFC 4647 as follows:
315 * "By contrast with filtering, each language range represents the most
316 * specific tag that is an acceptable match. The first matching tag found,
317 * according to the user's priority, is considered the closest match and is the
318 * item returned."
319 *
320 * <p>For example, if a Language Priority List consists of two language ranges,
321 * {@code "zh-Hant-TW"} and {@code "en-US"}, in prioritized order, lookup
322 * method progressively searches the language tags below in order to find the
323 * best matching language tag.
324 * <blockquote>
325 * <pre>
326 * 1. zh-Hant-TW
327 * 2. zh-Hant
328 * 3. zh
329 * 4. en-US
330 * 5. en
331 * </pre>
332 * </blockquote>
333 * If there is a language tag which matches completely to a language range
334 * above, the language tag is returned.
335 *
336 * <p>{@code "*"} is the special language range, and it is ignored in lookup.
337 *
338 * <p>If multiple language tags match as a result of the subtag {@code '*'}
339 * included in a language range, the first matching language tag returned by
340 * an {@link Iterator} over a {@link Collection} of language tags is treated as
341 * the best matching one.
342 *
343 * <h4>Use of Locale</h4>
344 *
345 * <p>Once you've created a <code>Locale</code> you can query it for information
346 * about itself. Use <code>getCountry</code> to get the country (or region)
347 * code and <code>getLanguage</code> to get the language code.
348 * You can use <code>getDisplayCountry</code> to get the
349 * name of the country suitable for displaying to the user. Similarly,
350 * you can use <code>getDisplayLanguage</code> to get the name of
351 * the language suitable for displaying to the user. Interestingly,
352 * the <code>getDisplayXXX</code> methods are themselves locale-sensitive
353 * and have two versions: one that uses the default
354 * {@link Locale.Category#DISPLAY DISPLAY} locale and one
355 * that uses the locale specified as an argument.
356 *
357 * <p>The Java Platform provides a number of classes that perform locale-sensitive
358 * operations. For example, the <code>NumberFormat</code> class formats
359 * numbers, currency, and percentages in a locale-sensitive manner. Classes
360 * such as <code>NumberFormat</code> have several convenience methods
361 * for creating a default object of that type. For example, the
362 * <code>NumberFormat</code> class provides these three convenience methods
363 * for creating a default <code>NumberFormat</code> object:
364 * <blockquote>
365 * <pre>
366 * NumberFormat.getInstance()
367 * NumberFormat.getCurrencyInstance()
368 * NumberFormat.getPercentInstance()
369 * </pre>
370 * </blockquote>
371 * Each of these methods has two variants; one with an explicit locale
372 * and one without; the latter uses the default
373 * {@link Locale.Category#FORMAT FORMAT} locale:
374 * <blockquote>
375 * <pre>
376 * NumberFormat.getInstance(myLocale)
377 * NumberFormat.getCurrencyInstance(myLocale)
378 * NumberFormat.getPercentInstance(myLocale)
379 * </pre>
380 * </blockquote>
381 * A <code>Locale</code> is the mechanism for identifying the kind of object
382 * (<code>NumberFormat</code>) that you would like to get. The locale is
383 * <STRONG>just</STRONG> a mechanism for identifying objects,
384 * <STRONG>not</STRONG> a container for the objects themselves.
385 *
386 * <h4>Compatibility</h4>
387 *
388 * <p>In order to maintain compatibility with existing usage, Locale's
389 * constructors retain their behavior prior to the Java Runtime
390 * Environment version 1.7. The same is largely true for the
391 * <code>toString</code> method. Thus Locale objects can continue to
392 * be used as they were. In particular, clients who parse the output
393 * of toString into language, country, and variant fields can continue
394 * to do so (although this is strongly discouraged), although the
395 * variant field will have additional information in it if script or
396 * extensions are present.
397 *
398 * <p>In addition, BCP 47 imposes syntax restrictions that are not
399 * imposed by Locale's constructors. This means that conversions
400 * between some Locales and BCP 47 language tags cannot be made without
401 * losing information. Thus <code>toLanguageTag</code> cannot
402 * represent the state of locales whose language, country, or variant
403 * do not conform to BCP 47.
404 *
405 * <p>Because of these issues, it is recommended that clients migrate
406 * away from constructing non-conforming locales and use the
407 * <code>forLanguageTag</code> and <code>Locale.Builder</code> APIs instead.
408 * Clients desiring a string representation of the complete locale can
409 * then always rely on <code>toLanguageTag</code> for this purpose.
410 *
411 * <h5><a name="special_cases_constructor">Special cases</a></h5>
412 *
413 * <p>For compatibility reasons, two
414 * non-conforming locales are treated as special cases. These are
415 * <b>{@code ja_JP_JP}</b> and <b>{@code th_TH_TH}</b>. These are ill-formed
416 * in BCP 47 since the variants are too short. To ease migration to BCP 47,
417 * these are treated specially during construction. These two cases (and only
418 * these) cause a constructor to generate an extension, all other values behave
419 * exactly as they did prior to Java 7.
420 *
421 * <p>Java has used {@code ja_JP_JP} to represent Japanese as used in
422 * Japan together with the Japanese Imperial calendar. This is now
423 * representable using a Unicode locale extension, by specifying the
424 * Unicode locale key {@code ca} (for "calendar") and type
425 * {@code japanese}. When the Locale constructor is called with the
426 * arguments "ja", "JP", "JP", the extension "u-ca-japanese" is
427 * automatically added.
428 *
429 * <p>Java has used {@code th_TH_TH} to represent Thai as used in
430 * Thailand together with Thai digits. This is also now representable using
431 * a Unicode locale extension, by specifying the Unicode locale key
432 * {@code nu} (for "number") and value {@code thai}. When the Locale
433 * constructor is called with the arguments "th", "TH", "TH", the
434 * extension "u-nu-thai" is automatically added.
435 *
436 * <h5>Serialization</h5>
437 *
438 * <p>During serialization, writeObject writes all fields to the output
439 * stream, including extensions.
440 *
441 * <p>During deserialization, readResolve adds extensions as described
442 * in <a href="#special_cases_constructor">Special Cases</a>, only
443 * for the two cases th_TH_TH and ja_JP_JP.
444 *
445 * <h5>Legacy language codes</h5>
446 *
447 * <p>Locale's constructor has always converted three language codes to
448 * their earlier, obsoleted forms: {@code he} maps to {@code iw},
449 * {@code yi} maps to {@code ji}, and {@code id} maps to
450 * {@code in}. This continues to be the case, in order to not break
451 * backwards compatibility.
452 *
453 * <p>The APIs added in 1.7 map between the old and new language codes,
454 * maintaining the old codes internal to Locale (so that
455 * <code>getLanguage</code> and <code>toString</code> reflect the old
456 * code), but using the new codes in the BCP 47 language tag APIs (so
457 * that <code>toLanguageTag</code> reflects the new one). This
458 * preserves the equivalence between Locales no matter which code or
459 * API is used to construct them. Java's default resource bundle
460 * lookup mechanism also implements this mapping, so that resources
461 * can be named using either convention, see {@link ResourceBundle.Control}.
462 *
463 * <h5>Three-letter language/country(region) codes</h5>
464 *
465 * <p>The Locale constructors have always specified that the language
466 * and the country param be two characters in length, although in
467 * practice they have accepted any length. The specification has now
468 * been relaxed to allow language codes of two to eight characters and
469 * country (region) codes of two to three characters, and in
470 * particular, three-letter language codes and three-digit region
471 * codes as specified in the IANA Language Subtag Registry. For
472 * compatibility, the implementation still does not impose a length
473 * constraint.
474 *
475 * @see Builder
476 * @see ResourceBundle
477 * @see java.text.Format
478 * @see java.text.NumberFormat
479 * @see java.text.Collator
480 * @author Mark Davis
481 * @since 1.1
482 */
483 public final class Locale implements Cloneable, Serializable {
484
485 private static final Cache LOCALECACHE = new Cache();
486
487 /** Useful constant for language.
488 */
489 public static final Locale ENGLISH = createConstant("en", "");
490
491 /** Useful constant for language.
492 */
493 public static final Locale FRENCH = createConstant("fr", "");
494
495 /** Useful constant for language.
496 */
497 public static final Locale GERMAN = createConstant("de", "");
498
499 /** Useful constant for language.
500 */
501 public static final Locale ITALIAN = createConstant("it", "");
502
503 /** Useful constant for language.
504 */
505 public static final Locale JAPANESE = createConstant("ja", "");
506
507 /** Useful constant for language.
508 */
509 public static final Locale KOREAN = createConstant("ko", "");
510
511 /** Useful constant for language.
512 */
513 public static final Locale CHINESE = createConstant("zh", "");
514
515 /** Useful constant for language.
516 */
517 public static final Locale SIMPLIFIED_CHINESE = createConstant("zh", "CN");
518
519 /** Useful constant for language.
520 */
521 public static final Locale TRADITIONAL_CHINESE = createConstant("zh", "TW");
522
523 /** Useful constant for country.
524 */
525 public static final Locale FRANCE = createConstant("fr", "FR");
526
527 /** Useful constant for country.
528 */
529 public static final Locale GERMANY = createConstant("de", "DE");
530
531 /** Useful constant for country.
532 */
533 public static final Locale ITALY = createConstant("it", "IT");
534
535 /** Useful constant for country.
536 */
537 public static final Locale JAPAN = createConstant("ja", "JP");
538
539 /** Useful constant for country.
540 */
541 public static final Locale KOREA = createConstant("ko", "KR");
542
543 /** Useful constant for country.
544 */
545 public static final Locale CHINA = SIMPLIFIED_CHINESE;
546
547 /** Useful constant for country.
548 */
549 public static final Locale PRC = SIMPLIFIED_CHINESE;
550
551 /** Useful constant for country.
552 */
553 public static final Locale TAIWAN = TRADITIONAL_CHINESE;
554
555 /** Useful constant for country.
556 */
557 public static final Locale UK = createConstant("en", "GB");
558
559 /** Useful constant for country.
560 */
561 public static final Locale US = createConstant("en", "US");
562
563 /** Useful constant for country.
564 */
565 public static final Locale CANADA = createConstant("en", "CA");
566
567 /** Useful constant for country.
568 */
569 public static final Locale CANADA_FRENCH = createConstant("fr", "CA");
570
571 /**
572 * Useful constant for the root locale. The root locale is the locale whose
573 * language, country, and variant are empty ("") strings. This is regarded
574 * as the base locale of all locales, and is used as the language/country
575 * neutral locale for the locale sensitive operations.
576 *
577 * @since 1.6
578 */
579 public static final Locale ROOT = createConstant("", "");
580
581 /**
582 * The key for the private use extension ('x').
583 *
584 * @see #getExtension(char)
585 * @see Builder#setExtension(char, String)
586 * @since 1.7
587 */
588 public static final char PRIVATE_USE_EXTENSION = 'x';
589
590 /**
591 * The key for Unicode locale extension ('u').
592 *
593 * @see #getExtension(char)
594 * @see Builder#setExtension(char, String)
595 * @since 1.7
596 */
597 public static final char UNICODE_LOCALE_EXTENSION = 'u';
598
599 /** serialization ID
600 */
601 static final long serialVersionUID = 9149081749638150636L;
602
603 /**
604 * Display types for retrieving localized names from the name providers.
605 */
606 private static final int DISPLAY_LANGUAGE = 0;
607 private static final int DISPLAY_COUNTRY = 1;
608 private static final int DISPLAY_VARIANT = 2;
609 private static final int DISPLAY_SCRIPT = 3;
610
611 /**
612 * Private constructor used by getInstance method
613 */
614 private Locale(BaseLocale baseLocale, LocaleExtensions extensions) {
615 this.baseLocale = baseLocale;
616 this.localeExtensions = extensions;
617 }
618
619 /**
620 * Construct a locale from language, country and variant.
621 * This constructor normalizes the language value to lowercase and
622 * the country value to uppercase.
623 * <p>
624 * <b>Note:</b>
625 * <ul>
626 * <li>ISO 639 is not a stable standard; some of the language codes it defines
627 * (specifically "iw", "ji", and "in") have changed. This constructor accepts both the
628 * old codes ("iw", "ji", and "in") and the new codes ("he", "yi", and "id"), but all other
629 * API on Locale will return only the OLD codes.
630 * <li>For backward compatibility reasons, this constructor does not make
631 * any syntactic checks on the input.
632 * <li>The two cases ("ja", "JP", "JP") and ("th", "TH", "TH") are handled specially,
633 * see <a href="#special_cases_constructor">Special Cases</a> for more information.
634 * </ul>
635 *
636 * @param language An ISO 639 alpha-2 or alpha-3 language code, or a language subtag
637 * up to 8 characters in length. See the <code>Locale</code> class description about
638 * valid language values.
639 * @param country An ISO 3166 alpha-2 country code or a UN M.49 numeric-3 area code.
640 * See the <code>Locale</code> class description about valid country values.
641 * @param variant Any arbitrary value used to indicate a variation of a <code>Locale</code>.
642 * See the <code>Locale</code> class description for the details.
643 * @exception NullPointerException thrown if any argument is null.
644 */
645 public Locale(String language, String country, String variant) {
646 if (language== null || country == null || variant == null) {
647 throw new NullPointerException();
648 }
649 baseLocale = BaseLocale.getInstance(convertOldISOCodes(language), "", country, variant);
650 localeExtensions = getCompatibilityExtensions(language, "", country, variant);
651 }
652
653 /**
654 * Construct a locale from language and country.
655 * This constructor normalizes the language value to lowercase and
656 * the country value to uppercase.
657 * <p>
658 * <b>Note:</b>
659 * <ul>
660 * <li>ISO 639 is not a stable standard; some of the language codes it defines
661 * (specifically "iw", "ji", and "in") have changed. This constructor accepts both the
662 * old codes ("iw", "ji", and "in") and the new codes ("he", "yi", and "id"), but all other
663 * API on Locale will return only the OLD codes.
664 * <li>For backward compatibility reasons, this constructor does not make
665 * any syntactic checks on the input.
666 * </ul>
667 *
668 * @param language An ISO 639 alpha-2 or alpha-3 language code, or a language subtag
669 * up to 8 characters in length. See the <code>Locale</code> class description about
670 * valid language values.
671 * @param country An ISO 3166 alpha-2 country code or a UN M.49 numeric-3 area code.
672 * See the <code>Locale</code> class description about valid country values.
673 * @exception NullPointerException thrown if either argument is null.
674 */
675 public Locale(String language, String country) {
676 this(language, country, "");
677 }
678
679 /**
680 * Construct a locale from a language code.
681 * This constructor normalizes the language value to lowercase.
682 * <p>
683 * <b>Note:</b>
684 * <ul>
685 * <li>ISO 639 is not a stable standard; some of the language codes it defines
686 * (specifically "iw", "ji", and "in") have changed. This constructor accepts both the
687 * old codes ("iw", "ji", and "in") and the new codes ("he", "yi", and "id"), but all other
688 * API on Locale will return only the OLD codes.
689 * <li>For backward compatibility reasons, this constructor does not make
690 * any syntactic checks on the input.
691 * </ul>
692 *
693 * @param language An ISO 639 alpha-2 or alpha-3 language code, or a language subtag
694 * up to 8 characters in length. See the <code>Locale</code> class description about
695 * valid language values.
696 * @exception NullPointerException thrown if argument is null.
697 * @since 1.4
698 */
699 public Locale(String language) {
700 this(language, "", "");
701 }
702
703 /**
704 * This method must be called only for creating the Locale.*
705 * constants due to making shortcuts.
706 */
707 private static Locale createConstant(String lang, String country) {
708 BaseLocale base = BaseLocale.createInstance(lang, country);
709 return getInstance(base, null);
710 }
711
712 /**
713 * Returns a <code>Locale</code> constructed from the given
714 * <code>language</code>, <code>country</code> and
715 * <code>variant</code>. If the same <code>Locale</code> instance
716 * is available in the cache, then that instance is
717 * returned. Otherwise, a new <code>Locale</code> instance is
718 * created and cached.
719 *
720 * @param language lowercase 2 to 8 language code.
721 * @param country uppercase two-letter ISO-3166 code and numeric-3 UN M.49 area code.
722 * @param variant vendor and browser specific code. See class description.
723 * @return the <code>Locale</code> instance requested
724 * @exception NullPointerException if any argument is null.
725 */
726 static Locale getInstance(String language, String country, String variant) {
727 return getInstance(language, "", country, variant, null);
728 }
729
730 static Locale getInstance(String language, String script, String country,
731 String variant, LocaleExtensions extensions) {
732 if (language== null || script == null || country == null || variant == null) {
733 throw new NullPointerException();
734 }
735
736 if (extensions == null) {
737 extensions = getCompatibilityExtensions(language, script, country, variant);
738 }
739
740 BaseLocale baseloc = BaseLocale.getInstance(language, script, country, variant);
741 return getInstance(baseloc, extensions);
742 }
743
744 static Locale getInstance(BaseLocale baseloc, LocaleExtensions extensions) {
745 LocaleKey key = new LocaleKey(baseloc, extensions);
746 return LOCALECACHE.get(key);
747 }
748
749 private static class Cache extends LocaleObjectCache<LocaleKey, Locale> {
750 private Cache() {
751 }
752
753 @Override
754 protected Locale createObject(LocaleKey key) {
755 return new Locale(key.base, key.exts);
756 }
757 }
758
759 private static final class LocaleKey {
760 private final BaseLocale base;
761 private final LocaleExtensions exts;
762 private final int hash;
763
764 private LocaleKey(BaseLocale baseLocale, LocaleExtensions extensions) {
765 base = baseLocale;
766 exts = extensions;
767
768 // Calculate the hash value here because it's always used.
769 int h = base.hashCode();
770 if (exts != null) {
771 h ^= exts.hashCode();
772 }
773 hash = h;
774 }
775
776 @Override
777 public boolean equals(Object obj) {
778 if (this == obj) {
779 return true;
780 }
781 if (!(obj instanceof LocaleKey)) {
782 return false;
783 }
784 LocaleKey other = (LocaleKey)obj;
785 if (hash != other.hash || !base.equals(other.base)) {
786 return false;
787 }
788 if (exts == null) {
789 return other.exts == null;
790 }
791 return exts.equals(other.exts);
792 }
793
794 @Override
795 public int hashCode() {
796 return hash;
797 }
798 }
799
800 /**
801 * Gets the current value of the default locale for this instance
802 * of the Java Virtual Machine.
803 * <p>
804 * The Java Virtual Machine sets the default locale during startup
805 * based on the host environment. It is used by many locale-sensitive
806 * methods if no locale is explicitly specified.
807 * It can be changed using the
808 * {@link #setDefault(java.util.Locale) setDefault} method.
809 *
810 * @return the default locale for this instance of the Java Virtual Machine
811 */
812 public static Locale getDefault() {
813 // do not synchronize this method - see 4071298
814 return defaultLocale;
815 }
816
817 /**
818 * Gets the current value of the default locale for the specified Category
819 * for this instance of the Java Virtual Machine.
820 * <p>
821 * The Java Virtual Machine sets the default locale during startup based
822 * on the host environment. It is used by many locale-sensitive methods
823 * if no locale is explicitly specified. It can be changed using the
824 * setDefault(Locale.Category, Locale) method.
825 *
826 * @param category - the specified category to get the default locale
827 * @throws NullPointerException if category is null
828 * @return the default locale for the specified Category for this instance
829 * of the Java Virtual Machine
830 * @see #setDefault(Locale.Category, Locale)
831 * @since 1.7
832 */
833 public static Locale getDefault(Locale.Category category) {
834 // do not synchronize this method - see 4071298
835 switch (category) {
836 case DISPLAY:
837 if (defaultDisplayLocale == null) {
838 synchronized(Locale.class) {
839 if (defaultDisplayLocale == null) {
840 defaultDisplayLocale = initDefault(category);
841 }
842 }
843 }
844 return defaultDisplayLocale;
845 case FORMAT:
846 if (defaultFormatLocale == null) {
847 synchronized(Locale.class) {
848 if (defaultFormatLocale == null) {
849 defaultFormatLocale = initDefault(category);
850 }
851 }
852 }
853 return defaultFormatLocale;
854 default:
855 assert false: "Unknown Category";
856 }
857 return getDefault();
858 }
859
860 private static Locale initDefault() {
861 String language, region, script, country, variant;
862 language = AccessController.doPrivileged(
863 new GetPropertyAction("user.language", "en"));
864 // for compatibility, check for old user.region property
865 region = AccessController.doPrivileged(
866 new GetPropertyAction("user.region"));
867 if (region != null) {
868 // region can be of form country, country_variant, or _variant
869 int i = region.indexOf('_');
870 if (i >= 0) {
871 country = region.substring(0, i);
872 variant = region.substring(i + 1);
873 } else {
874 country = region;
875 variant = "";
876 }
877 script = "";
878 } else {
879 script = AccessController.doPrivileged(
880 new GetPropertyAction("user.script", ""));
881 country = AccessController.doPrivileged(
882 new GetPropertyAction("user.country", ""));
883 variant = AccessController.doPrivileged(
884 new GetPropertyAction("user.variant", ""));
885 }
886
887 return getInstance(language, script, country, variant, null);
888 }
889
890 private static Locale initDefault(Locale.Category category) {
891 return getInstance(
892 AccessController.doPrivileged(
893 new GetPropertyAction(category.languageKey, defaultLocale.getLanguage())),
894 AccessController.doPrivileged(
895 new GetPropertyAction(category.scriptKey, defaultLocale.getScript())),
896 AccessController.doPrivileged(
897 new GetPropertyAction(category.countryKey, defaultLocale.getCountry())),
898 AccessController.doPrivileged(
899 new GetPropertyAction(category.variantKey, defaultLocale.getVariant())),
900 null);
901 }
902
903 /**
904 * Sets the default locale for this instance of the Java Virtual Machine.
905 * This does not affect the host locale.
906 * <p>
907 * If there is a security manager, its <code>checkPermission</code>
908 * method is called with a <code>PropertyPermission("user.language", "write")</code>
909 * permission before the default locale is changed.
910 * <p>
911 * The Java Virtual Machine sets the default locale during startup
912 * based on the host environment. It is used by many locale-sensitive
913 * methods if no locale is explicitly specified.
914 * <p>
915 * Since changing the default locale may affect many different areas
916 * of functionality, this method should only be used if the caller
917 * is prepared to reinitialize locale-sensitive code running
918 * within the same Java Virtual Machine.
919 * <p>
920 * By setting the default locale with this method, all of the default
921 * locales for each Category are also set to the specified default locale.
922 *
923 * @throws SecurityException
924 * if a security manager exists and its
925 * <code>checkPermission</code> method doesn't allow the operation.
926 * @throws NullPointerException if <code>newLocale</code> is null
927 * @param newLocale the new default locale
928 * @see SecurityManager#checkPermission
929 * @see java.util.PropertyPermission
930 */
931 public static synchronized void setDefault(Locale newLocale) {
932 setDefault(Category.DISPLAY, newLocale);
933 setDefault(Category.FORMAT, newLocale);
934 defaultLocale = newLocale;
935 }
936
937 /**
938 * Sets the default locale for the specified Category for this instance
939 * of the Java Virtual Machine. This does not affect the host locale.
940 * <p>
941 * If there is a security manager, its checkPermission method is called
942 * with a PropertyPermission("user.language", "write") permission before
943 * the default locale is changed.
944 * <p>
945 * The Java Virtual Machine sets the default locale during startup based
946 * on the host environment. It is used by many locale-sensitive methods
947 * if no locale is explicitly specified.
948 * <p>
949 * Since changing the default locale may affect many different areas of
950 * functionality, this method should only be used if the caller is
951 * prepared to reinitialize locale-sensitive code running within the
952 * same Java Virtual Machine.
953 *
954 * @param category - the specified category to set the default locale
955 * @param newLocale - the new default locale
956 * @throws SecurityException if a security manager exists and its
957 * checkPermission method doesn't allow the operation.
958 * @throws NullPointerException if category and/or newLocale is null
959 * @see SecurityManager#checkPermission(java.security.Permission)
960 * @see PropertyPermission
961 * @see #getDefault(Locale.Category)
962 * @since 1.7
963 */
964 public static synchronized void setDefault(Locale.Category category,
965 Locale newLocale) {
966 if (category == null)
967 throw new NullPointerException("Category cannot be NULL");
968 if (newLocale == null)
969 throw new NullPointerException("Can't set default locale to NULL");
970
971 SecurityManager sm = System.getSecurityManager();
972 if (sm != null) sm.checkPermission(new PropertyPermission
973 ("user.language", "write"));
974 switch (category) {
975 case DISPLAY:
976 defaultDisplayLocale = newLocale;
977 break;
978 case FORMAT:
979 defaultFormatLocale = newLocale;
980 break;
981 default:
982 assert false: "Unknown Category";
983 }
984 }
985
986 /**
987 * Returns an array of all installed locales.
988 * The returned array represents the union of locales supported
989 * by the Java runtime environment and by installed
990 * {@link java.util.spi.LocaleServiceProvider LocaleServiceProvider}
991 * implementations. It must contain at least a <code>Locale</code>
992 * instance equal to {@link java.util.Locale#US Locale.US}.
993 *
994 * @return An array of installed locales.
995 */
996 public static Locale[] getAvailableLocales() {
997 return LocaleServiceProviderPool.getAllAvailableLocales();
998 }
999
1000 /**
1001 * Returns a list of all 2-letter country codes defined in ISO 3166.
1002 * Can be used to create Locales.
1003 * <p>
1004 * <b>Note:</b> The <code>Locale</code> class also supports other codes for
1005 * country (region), such as 3-letter numeric UN M.49 area codes.
1006 * Therefore, the list returned by this method does not contain ALL valid
1007 * codes that can be used to create Locales.
1008 *
1009 * @return An array of ISO 3166 two-letter country codes.
1010 */
1011 public static String[] getISOCountries() {
1012 if (isoCountries == null) {
1013 isoCountries = getISO2Table(LocaleISOData.isoCountryTable);
1014 }
1015 String[] result = new String[isoCountries.length];
1016 System.arraycopy(isoCountries, 0, result, 0, isoCountries.length);
1017 return result;
1018 }
1019
1020 /**
1021 * Returns a list of all 2-letter language codes defined in ISO 639.
1022 * Can be used to create Locales.
1023 * <p>
1024 * <b>Note:</b>
1025 * <ul>
1026 * <li>ISO 639 is not a stable standard— some languages' codes have changed.
1027 * The list this function returns includes both the new and the old codes for the
1028 * languages whose codes have changed.
1029 * <li>The <code>Locale</code> class also supports language codes up to
1030 * 8 characters in length. Therefore, the list returned by this method does
1031 * not contain ALL valid codes that can be used to create Locales.
1032 * </ul>
1033 *
1034 * @return Am array of ISO 639 two-letter language codes.
1035 */
1036 public static String[] getISOLanguages() {
1037 if (isoLanguages == null) {
1038 isoLanguages = getISO2Table(LocaleISOData.isoLanguageTable);
1039 }
1040 String[] result = new String[isoLanguages.length];
1041 System.arraycopy(isoLanguages, 0, result, 0, isoLanguages.length);
1042 return result;
1043 }
1044
1045 private static String[] getISO2Table(String table) {
1046 int len = table.length() / 5;
1047 String[] isoTable = new String[len];
1048 for (int i = 0, j = 0; i < len; i++, j += 5) {
1049 isoTable[i] = table.substring(j, j + 2);
1050 }
1051 return isoTable;
1052 }
1053
1054 /**
1055 * Returns the language code of this Locale.
1056 *
1057 * <p><b>Note:</b> ISO 639 is not a stable standard— some languages' codes have changed.
1058 * Locale's constructor recognizes both the new and the old codes for the languages
1059 * whose codes have changed, but this function always returns the old code. If you
1060 * want to check for a specific language whose code has changed, don't do
1061 * <pre>
1062 * if (locale.getLanguage().equals("he")) // BAD!
1063 * ...
1064 * </pre>
1065 * Instead, do
1066 * <pre>
1067 * if (locale.getLanguage().equals(new Locale("he").getLanguage()))
1068 * ...
1069 * </pre>
1070 * @return The language code, or the empty string if none is defined.
1071 * @see #getDisplayLanguage
1072 */
1073 public String getLanguage() {
1074 return baseLocale.getLanguage();
1075 }
1076
1077 /**
1078 * Returns the script for this locale, which should
1079 * either be the empty string or an ISO 15924 4-letter script
1080 * code. The first letter is uppercase and the rest are
1081 * lowercase, for example, 'Latn', 'Cyrl'.
1082 *
1083 * @return The script code, or the empty string if none is defined.
1084 * @see #getDisplayScript
1085 * @since 1.7
1086 */
1087 public String getScript() {
1088 return baseLocale.getScript();
1089 }
1090
1091 /**
1092 * Returns the country/region code for this locale, which should
1093 * either be the empty string, an uppercase ISO 3166 2-letter code,
1094 * or a UN M.49 3-digit code.
1095 *
1096 * @return The country/region code, or the empty string if none is defined.
1097 * @see #getDisplayCountry
1098 */
1099 public String getCountry() {
1100 return baseLocale.getRegion();
1101 }
1102
1103 /**
1104 * Returns the variant code for this locale.
1105 *
1106 * @return The variant code, or the empty string if none is defined.
1107 * @see #getDisplayVariant
1108 */
1109 public String getVariant() {
1110 return baseLocale.getVariant();
1111 }
1112
1113 /**
1114 * Returns {@code true} if this {@code Locale} has any <a href="#def_extensions">
1115 * extensions</a>.
1116 *
1117 * @return {@code true} if this {@code Locale} has any extensions
1118 * @since 1.8
1119 */
1120 public boolean hasExtensions() {
1121 return localeExtensions != null;
1122 }
1123
1124 /**
1125 * Returns a copy of this {@code Locale} with no <a href="#def_extensions">
1126 * extensions</a>. If this {@code Locale} has no extensions, this {@code Locale}
1127 * is returned.
1128 *
1129 * @return a copy of this {@code Locale} with no extensions, or {@code this}
1130 * if {@code this} has no extensions
1131 * @since 1.8
1132 */
1133 public Locale stripExtensions() {
1134 return hasExtensions() ? Locale.getInstance(baseLocale, null) : this;
1135 }
1136
1137 /**
1138 * Returns the extension (or private use) value associated with
1139 * the specified key, or null if there is no extension
1140 * associated with the key. To be well-formed, the key must be one
1141 * of <code>[0-9A-Za-z]</code>. Keys are case-insensitive, so
1142 * for example 'z' and 'Z' represent the same extension.
1143 *
1144 * @param key the extension key
1145 * @return The extension, or null if this locale defines no
1146 * extension for the specified key.
1147 * @throws IllegalArgumentException if key is not well-formed
1148 * @see #PRIVATE_USE_EXTENSION
1149 * @see #UNICODE_LOCALE_EXTENSION
1150 * @since 1.7
1151 */
1152 public String getExtension(char key) {
1153 if (!LocaleExtensions.isValidKey(key)) {
1154 throw new IllegalArgumentException("Ill-formed extension key: " + key);
1155 }
1156 return hasExtensions() ? localeExtensions.getExtensionValue(key) : null;
1157 }
1158
1159 /**
1160 * Returns the set of extension keys associated with this locale, or the
1161 * empty set if it has no extensions. The returned set is unmodifiable.
1162 * The keys will all be lower-case.
1163 *
1164 * @return The set of extension keys, or the empty set if this locale has
1165 * no extensions.
1166 * @since 1.7
1167 */
1168 public Set<Character> getExtensionKeys() {
1169 if (!hasExtensions()) {
1170 return Collections.emptySet();
1171 }
1172 return localeExtensions.getKeys();
1173 }
1174
1175 /**
1176 * Returns the set of unicode locale attributes associated with
1177 * this locale, or the empty set if it has no attributes. The
1178 * returned set is unmodifiable.
1179 *
1180 * @return The set of attributes.
1181 * @since 1.7
1182 */
1183 public Set<String> getUnicodeLocaleAttributes() {
1184 if (!hasExtensions()) {
1185 return Collections.emptySet();
1186 }
1187 return localeExtensions.getUnicodeLocaleAttributes();
1188 }
1189
1190 /**
1191 * Returns the Unicode locale type associated with the specified Unicode locale key
1192 * for this locale. Returns the empty string for keys that are defined with no type.
1193 * Returns null if the key is not defined. Keys are case-insensitive. The key must
1194 * be two alphanumeric characters ([0-9a-zA-Z]), or an IllegalArgumentException is
1195 * thrown.
1196 *
1197 * @param key the Unicode locale key
1198 * @return The Unicode locale type associated with the key, or null if the
1199 * locale does not define the key.
1200 * @throws IllegalArgumentException if the key is not well-formed
1201 * @throws NullPointerException if <code>key</code> is null
1202 * @since 1.7
1203 */
1204 public String getUnicodeLocaleType(String key) {
1205 if (!isUnicodeExtensionKey(key)) {
1206 throw new IllegalArgumentException("Ill-formed Unicode locale key: " + key);
1207 }
1208 return hasExtensions() ? localeExtensions.getUnicodeLocaleType(key) : null;
1209 }
1210
1211 /**
1212 * Returns the set of Unicode locale keys defined by this locale, or the empty set if
1213 * this locale has none. The returned set is immutable. Keys are all lower case.
1214 *
1215 * @return The set of Unicode locale keys, or the empty set if this locale has
1216 * no Unicode locale keywords.
1217 * @since 1.7
1218 */
1219 public Set<String> getUnicodeLocaleKeys() {
1220 if (localeExtensions == null) {
1221 return Collections.emptySet();
1222 }
1223 return localeExtensions.getUnicodeLocaleKeys();
1224 }
1225
1226 /**
1227 * Package locale method returning the Locale's BaseLocale,
1228 * used by ResourceBundle
1229 * @return base locale of this Locale
1230 */
1231 BaseLocale getBaseLocale() {
1232 return baseLocale;
1233 }
1234
1235 /**
1236 * Package private method returning the Locale's LocaleExtensions,
1237 * used by ResourceBundle.
1238 * @return locale extensions of this Locale,
1239 * or {@code null} if no extensions are defined
1240 */
1241 LocaleExtensions getLocaleExtensions() {
1242 return localeExtensions;
1243 }
1244
1245 /**
1246 * Returns a string representation of this <code>Locale</code>
1247 * object, consisting of language, country, variant, script,
1248 * and extensions as below:
1249 * <blockquote>
1250 * language + "_" + country + "_" + (variant + "_#" | "#") + script + "_" + extensions
1251 * </blockquote>
1252 *
1253 * Language is always lower case, country is always upper case, script is always title
1254 * case, and extensions are always lower case. Extensions and private use subtags
1255 * will be in canonical order as explained in {@link #toLanguageTag}.
1256 *
1257 * <p>When the locale has neither script nor extensions, the result is the same as in
1258 * Java 6 and prior.
1259 *
1260 * <p>If both the language and country fields are missing, this function will return
1261 * the empty string, even if the variant, script, or extensions field is present (you
1262 * can't have a locale with just a variant, the variant must accompany a well-formed
1263 * language or country code).
1264 *
1265 * <p>If script or extensions are present and variant is missing, no underscore is
1266 * added before the "#".
1267 *
1268 * <p>This behavior is designed to support debugging and to be compatible with
1269 * previous uses of <code>toString</code> that expected language, country, and variant
1270 * fields only. To represent a Locale as a String for interchange purposes, use
1271 * {@link #toLanguageTag}.
1272 *
1273 * <p>Examples: <ul>
1274 * <li>{@code en}</li>
1275 * <li>{@code de_DE}</li>
1276 * <li>{@code _GB}</li>
1277 * <li>{@code en_US_WIN}</li>
1278 * <li>{@code de__POSIX}</li>
1279 * <li>{@code zh_CN_#Hans}</li>
1280 * <li>{@code zh_TW_#Hant_x-java}</li>
1281 * <li>{@code th_TH_TH_#u-nu-thai}</li></ul>
1282 *
1283 * @return A string representation of the Locale, for debugging.
1284 * @see #getDisplayName
1285 * @see #toLanguageTag
1286 */
1287 @Override
1288 public final String toString() {
1289 boolean l = (baseLocale.getLanguage().length() != 0);
1290 boolean s = (baseLocale.getScript().length() != 0);
1291 boolean r = (baseLocale.getRegion().length() != 0);
1292 boolean v = (baseLocale.getVariant().length() != 0);
1293 boolean e = (localeExtensions != null && localeExtensions.getID().length() != 0);
1294
1295 StringBuilder result = new StringBuilder(baseLocale.getLanguage());
1296 if (r || (l && (v || s || e))) {
1297 result.append('_')
1298 .append(baseLocale.getRegion()); // This may just append '_'
1299 }
1300 if (v && (l || r)) {
1301 result.append('_')
1302 .append(baseLocale.getVariant());
1303 }
1304
1305 if (s && (l || r)) {
1306 result.append("_#")
1307 .append(baseLocale.getScript());
1308 }
1309
1310 if (e && (l || r)) {
1311 result.append('_');
1312 if (!s) {
1313 result.append('#');
1314 }
1315 result.append(localeExtensions.getID());
1316 }
1317
1318 return result.toString();
1319 }
1320
1321 /**
1322 * Returns a well-formed IETF BCP 47 language tag representing
1323 * this locale.
1324 *
1325 * <p>If this <code>Locale</code> has a language, country, or
1326 * variant that does not satisfy the IETF BCP 47 language tag
1327 * syntax requirements, this method handles these fields as
1328 * described below:
1329 *
1330 * <p><b>Language:</b> If language is empty, or not <a
1331 * href="#def_language" >well-formed</a> (for example "a" or
1332 * "e2"), it will be emitted as "und" (Undetermined).
1333 *
1334 * <p><b>Country:</b> If country is not <a
1335 * href="#def_region">well-formed</a> (for example "12" or "USA"),
1336 * it will be omitted.
1337 *
1338 * <p><b>Variant:</b> If variant <b>is</b> <a
1339 * href="#def_variant">well-formed</a>, each sub-segment
1340 * (delimited by '-' or '_') is emitted as a subtag. Otherwise:
1341 * <ul>
1342 *
1343 * <li>if all sub-segments match <code>[0-9a-zA-Z]{1,8}</code>
1344 * (for example "WIN" or "Oracle_JDK_Standard_Edition"), the first
1345 * ill-formed sub-segment and all following will be appended to
1346 * the private use subtag. The first appended subtag will be
1347 * "lvariant", followed by the sub-segments in order, separated by
1348 * hyphen. For example, "x-lvariant-WIN",
1349 * "Oracle-x-lvariant-JDK-Standard-Edition".
1350 *
1351 * <li>if any sub-segment does not match
1352 * <code>[0-9a-zA-Z]{1,8}</code>, the variant will be truncated
1353 * and the problematic sub-segment and all following sub-segments
1354 * will be omitted. If the remainder is non-empty, it will be
1355 * emitted as a private use subtag as above (even if the remainder
1356 * turns out to be well-formed). For example,
1357 * "Solaris_isjustthecoolestthing" is emitted as
1358 * "x-lvariant-Solaris", not as "solaris".</li></ul>
1359 *
1360 * <p><b>Special Conversions:</b> Java supports some old locale
1361 * representations, including deprecated ISO language codes,
1362 * for compatibility. This method performs the following
1363 * conversions:
1364 * <ul>
1365 *
1366 * <li>Deprecated ISO language codes "iw", "ji", and "in" are
1367 * converted to "he", "yi", and "id", respectively.
1368 *
1369 * <li>A locale with language "no", country "NO", and variant
1370 * "NY", representing Norwegian Nynorsk (Norway), is converted
1371 * to a language tag "nn-NO".</li></ul>
1372 *
1373 * <p><b>Note:</b> Although the language tag created by this
1374 * method is well-formed (satisfies the syntax requirements
1375 * defined by the IETF BCP 47 specification), it is not
1376 * necessarily a valid BCP 47 language tag. For example,
1377 * <pre>
1378 * new Locale("xx", "YY").toLanguageTag();</pre>
1379 *
1380 * will return "xx-YY", but the language subtag "xx" and the
1381 * region subtag "YY" are invalid because they are not registered
1382 * in the IANA Language Subtag Registry.
1383 *
1384 * @return a BCP47 language tag representing the locale
1385 * @see #forLanguageTag(String)
1386 * @since 1.7
1387 */
1388 public String toLanguageTag() {
1389 if (languageTag != null) {
1390 return languageTag;
1391 }
1392
1393 LanguageTag tag = LanguageTag.parseLocale(baseLocale, localeExtensions);
1394 StringBuilder buf = new StringBuilder();
1395
1396 String subtag = tag.getLanguage();
1397 if (subtag.length() > 0) {
1398 buf.append(LanguageTag.canonicalizeLanguage(subtag));
1399 }
1400
1401 subtag = tag.getScript();
1402 if (subtag.length() > 0) {
1403 buf.append(LanguageTag.SEP);
1404 buf.append(LanguageTag.canonicalizeScript(subtag));
1405 }
1406
1407 subtag = tag.getRegion();
1408 if (subtag.length() > 0) {
1409 buf.append(LanguageTag.SEP);
1410 buf.append(LanguageTag.canonicalizeRegion(subtag));
1411 }
1412
1413 List<String>subtags = tag.getVariants();
1414 for (String s : subtags) {
1415 buf.append(LanguageTag.SEP);
1416 // preserve casing
1417 buf.append(s);
1418 }
1419
1420 subtags = tag.getExtensions();
1421 for (String s : subtags) {
1422 buf.append(LanguageTag.SEP);
1423 buf.append(LanguageTag.canonicalizeExtension(s));
1424 }
1425
1426 subtag = tag.getPrivateuse();
1427 if (subtag.length() > 0) {
1428 if (buf.length() > 0) {
1429 buf.append(LanguageTag.SEP);
1430 }
1431 buf.append(LanguageTag.PRIVATEUSE).append(LanguageTag.SEP);
1432 // preserve casing
1433 buf.append(subtag);
1434 }
1435
1436 String langTag = buf.toString();
1437 synchronized (this) {
1438 if (languageTag == null) {
1439 languageTag = langTag;
1440 }
1441 }
1442 return languageTag;
1443 }
1444
1445 /**
1446 * Returns a locale for the specified IETF BCP 47 language tag string.
1447 *
1448 * <p>If the specified language tag contains any ill-formed subtags,
1449 * the first such subtag and all following subtags are ignored. Compare
1450 * to {@link Locale.Builder#setLanguageTag} which throws an exception
1451 * in this case.
1452 *
1453 * <p>The following <b>conversions</b> are performed:<ul>
1454 *
1455 * <li>The language code "und" is mapped to language "".
1456 *
1457 * <li>The language codes "he", "yi", and "id" are mapped to "iw",
1458 * "ji", and "in" respectively. (This is the same canonicalization
1459 * that's done in Locale's constructors.)
1460 *
1461 * <li>The portion of a private use subtag prefixed by "lvariant",
1462 * if any, is removed and appended to the variant field in the
1463 * result locale (without case normalization). If it is then
1464 * empty, the private use subtag is discarded:
1465 *
1466 * <pre>
1467 * Locale loc;
1468 * loc = Locale.forLanguageTag("en-US-x-lvariant-POSIX");
1469 * loc.getVariant(); // returns "POSIX"
1470 * loc.getExtension('x'); // returns null
1471 *
1472 * loc = Locale.forLanguageTag("de-POSIX-x-URP-lvariant-Abc-Def");
1473 * loc.getVariant(); // returns "POSIX_Abc_Def"
1474 * loc.getExtension('x'); // returns "urp"
1475 * </pre>
1476 *
1477 * <li>When the languageTag argument contains an extlang subtag,
1478 * the first such subtag is used as the language, and the primary
1479 * language subtag and other extlang subtags are ignored:
1480 *
1481 * <pre>
1482 * Locale.forLanguageTag("ar-aao").getLanguage(); // returns "aao"
1483 * Locale.forLanguageTag("en-abc-def-us").toString(); // returns "abc_US"
1484 * </pre>
1485 *
1486 * <li>Case is normalized except for variant tags, which are left
1487 * unchanged. Language is normalized to lower case, script to
1488 * title case, country to upper case, and extensions to lower
1489 * case.
1490 *
1491 * <li>If, after processing, the locale would exactly match either
1492 * ja_JP_JP or th_TH_TH with no extensions, the appropriate
1493 * extensions are added as though the constructor had been called:
1494 *
1495 * <pre>
1496 * Locale.forLanguageTag("ja-JP-x-lvariant-JP").toLanguageTag();
1497 * // returns "ja-JP-u-ca-japanese-x-lvariant-JP"
1498 * Locale.forLanguageTag("th-TH-x-lvariant-TH").toLanguageTag();
1499 * // returns "th-TH-u-nu-thai-x-lvariant-TH"
1500 * </pre></ul>
1501 *
1502 * <p>This implements the 'Language-Tag' production of BCP47, and
1503 * so supports grandfathered (regular and irregular) as well as
1504 * private use language tags. Stand alone private use tags are
1505 * represented as empty language and extension 'x-whatever',
1506 * and grandfathered tags are converted to their canonical replacements
1507 * where they exist.
1508 *
1509 * <p>Grandfathered tags with canonical replacements are as follows:
1510 *
1511 * <table summary="Grandfathered tags with canonical replacements">
1512 * <tbody align="center">
1513 * <tr><th>grandfathered tag</th><th> </th><th>modern replacement</th></tr>
1514 * <tr><td>art-lojban</td><td> </td><td>jbo</td></tr>
1515 * <tr><td>i-ami</td><td> </td><td>ami</td></tr>
1516 * <tr><td>i-bnn</td><td> </td><td>bnn</td></tr>
1517 * <tr><td>i-hak</td><td> </td><td>hak</td></tr>
1518 * <tr><td>i-klingon</td><td> </td><td>tlh</td></tr>
1519 * <tr><td>i-lux</td><td> </td><td>lb</td></tr>
1520 * <tr><td>i-navajo</td><td> </td><td>nv</td></tr>
1521 * <tr><td>i-pwn</td><td> </td><td>pwn</td></tr>
1522 * <tr><td>i-tao</td><td> </td><td>tao</td></tr>
1523 * <tr><td>i-tay</td><td> </td><td>tay</td></tr>
1524 * <tr><td>i-tsu</td><td> </td><td>tsu</td></tr>
1525 * <tr><td>no-bok</td><td> </td><td>nb</td></tr>
1526 * <tr><td>no-nyn</td><td> </td><td>nn</td></tr>
1527 * <tr><td>sgn-BE-FR</td><td> </td><td>sfb</td></tr>
1528 * <tr><td>sgn-BE-NL</td><td> </td><td>vgt</td></tr>
1529 * <tr><td>sgn-CH-DE</td><td> </td><td>sgg</td></tr>
1530 * <tr><td>zh-guoyu</td><td> </td><td>cmn</td></tr>
1531 * <tr><td>zh-hakka</td><td> </td><td>hak</td></tr>
1532 * <tr><td>zh-min-nan</td><td> </td><td>nan</td></tr>
1533 * <tr><td>zh-xiang</td><td> </td><td>hsn</td></tr>
1534 * </tbody>
1535 * </table>
1536 *
1537 * <p>Grandfathered tags with no modern replacement will be
1538 * converted as follows:
1539 *
1540 * <table summary="Grandfathered tags with no modern replacement">
1541 * <tbody align="center">
1542 * <tr><th>grandfathered tag</th><th> </th><th>converts to</th></tr>
1543 * <tr><td>cel-gaulish</td><td> </td><td>xtg-x-cel-gaulish</td></tr>
1544 * <tr><td>en-GB-oed</td><td> </td><td>en-GB-x-oed</td></tr>
1545 * <tr><td>i-default</td><td> </td><td>en-x-i-default</td></tr>
1546 * <tr><td>i-enochian</td><td> </td><td>und-x-i-enochian</td></tr>
1547 * <tr><td>i-mingo</td><td> </td><td>see-x-i-mingo</td></tr>
1548 * <tr><td>zh-min</td><td> </td><td>nan-x-zh-min</td></tr>
1549 * </tbody>
1550 * </table>
1551 *
1552 * <p>For a list of all grandfathered tags, see the
1553 * IANA Language Subtag Registry (search for "Type: grandfathered").
1554 *
1555 * <p><b>Note</b>: there is no guarantee that <code>toLanguageTag</code>
1556 * and <code>forLanguageTag</code> will round-trip.
1557 *
1558 * @param languageTag the language tag
1559 * @return The locale that best represents the language tag.
1560 * @throws NullPointerException if <code>languageTag</code> is <code>null</code>
1561 * @see #toLanguageTag()
1562 * @see java.util.Locale.Builder#setLanguageTag(String)
1563 * @since 1.7
1564 */
1565 public static Locale forLanguageTag(String languageTag) {
1566 LanguageTag tag = LanguageTag.parse(languageTag, null);
1567 InternalLocaleBuilder bldr = new InternalLocaleBuilder();
1568 bldr.setLanguageTag(tag);
1569 BaseLocale base = bldr.getBaseLocale();
1570 LocaleExtensions exts = bldr.getLocaleExtensions();
1571 if (exts == null && base.getVariant().length() > 0) {
1572 exts = getCompatibilityExtensions(base.getLanguage(), base.getScript(),
1573 base.getRegion(), base.getVariant());
1574 }
1575 return getInstance(base, exts);
1576 }
1577
1578 /**
1579 * Returns a three-letter abbreviation of this locale's language.
1580 * If the language matches an ISO 639-1 two-letter code, the
1581 * corresponding ISO 639-2/T three-letter lowercase code is
1582 * returned. The ISO 639-2 language codes can be found on-line,
1583 * see "Codes for the Representation of Names of Languages Part 2:
1584 * Alpha-3 Code". If the locale specifies a three-letter
1585 * language, the language is returned as is. If the locale does
1586 * not specify a language the empty string is returned.
1587 *
1588 * @return A three-letter abbreviation of this locale's language.
1589 * @exception MissingResourceException Throws MissingResourceException if
1590 * three-letter language abbreviation is not available for this locale.
1591 */
1592 public String getISO3Language() throws MissingResourceException {
1593 String lang = baseLocale.getLanguage();
1594 if (lang.length() == 3) {
1595 return lang;
1596 }
1597
1598 String language3 = getISO3Code(lang, LocaleISOData.isoLanguageTable);
1599 if (language3 == null) {
1600 throw new MissingResourceException("Couldn't find 3-letter language code for "
1601 + lang, "FormatData_" + toString(), "ShortLanguage");
1602 }
1603 return language3;
1604 }
1605
1606 /**
1607 * Returns a three-letter abbreviation for this locale's country.
1608 * If the country matches an ISO 3166-1 alpha-2 code, the
1609 * corresponding ISO 3166-1 alpha-3 uppercase code is returned.
1610 * If the locale doesn't specify a country, this will be the empty
1611 * string.
1612 *
1613 * <p>The ISO 3166-1 codes can be found on-line.
1614 *
1615 * @return A three-letter abbreviation of this locale's country.
1616 * @exception MissingResourceException Throws MissingResourceException if the
1617 * three-letter country abbreviation is not available for this locale.
1618 */
1619 public String getISO3Country() throws MissingResourceException {
1620 String country3 = getISO3Code(baseLocale.getRegion(), LocaleISOData.isoCountryTable);
1621 if (country3 == null) {
1622 throw new MissingResourceException("Couldn't find 3-letter country code for "
1623 + baseLocale.getRegion(), "FormatData_" + toString(), "ShortCountry");
1624 }
1625 return country3;
1626 }
1627
1628 private static String getISO3Code(String iso2Code, String table) {
1629 int codeLength = iso2Code.length();
1630 if (codeLength == 0) {
1631 return "";
1632 }
1633
1634 int tableLength = table.length();
1635 int index = tableLength;
1636 if (codeLength == 2) {
1637 char c1 = iso2Code.charAt(0);
1638 char c2 = iso2Code.charAt(1);
1639 for (index = 0; index < tableLength; index += 5) {
1640 if (table.charAt(index) == c1
1641 && table.charAt(index + 1) == c2) {
1642 break;
1643 }
1644 }
1645 }
1646 return index < tableLength ? table.substring(index + 2, index + 5) : null;
1647 }
1648
1649 /**
1650 * Returns a name for the locale's language that is appropriate for display to the
1651 * user.
1652 * If possible, the name returned will be localized for the default
1653 * {@link Locale.Category#DISPLAY DISPLAY} locale.
1654 * For example, if the locale is fr_FR and the default
1655 * {@link Locale.Category#DISPLAY DISPLAY} locale
1656 * is en_US, getDisplayLanguage() will return "French"; if the locale is en_US and
1657 * the default {@link Locale.Category#DISPLAY DISPLAY} locale is fr_FR,
1658 * getDisplayLanguage() will return "anglais".
1659 * If the name returned cannot be localized for the default
1660 * {@link Locale.Category#DISPLAY DISPLAY} locale,
1661 * (say, we don't have a Japanese name for Croatian),
1662 * this function falls back on the English name, and uses the ISO code as a last-resort
1663 * value. If the locale doesn't specify a language, this function returns the empty string.
1664 *
1665 * @return The name of the display language.
1666 */
1667 public final String getDisplayLanguage() {
1668 return getDisplayLanguage(getDefault(Category.DISPLAY));
1669 }
1670
1671 /**
1672 * Returns a name for the locale's language that is appropriate for display to the
1673 * user.
1674 * If possible, the name returned will be localized according to inLocale.
1675 * For example, if the locale is fr_FR and inLocale
1676 * is en_US, getDisplayLanguage() will return "French"; if the locale is en_US and
1677 * inLocale is fr_FR, getDisplayLanguage() will return "anglais".
1678 * If the name returned cannot be localized according to inLocale,
1679 * (say, we don't have a Japanese name for Croatian),
1680 * this function falls back on the English name, and finally
1681 * on the ISO code as a last-resort value. If the locale doesn't specify a language,
1682 * this function returns the empty string.
1683 *
1684 * @param inLocale The locale for which to retrieve the display language.
1685 * @return The name of the display language appropriate to the given locale.
1686 * @exception NullPointerException if <code>inLocale</code> is <code>null</code>
1687 */
1688 public String getDisplayLanguage(Locale inLocale) {
1689 return getDisplayString(baseLocale.getLanguage(), inLocale, DISPLAY_LANGUAGE);
1690 }
1691
1692 /**
1693 * Returns a name for the locale's script that is appropriate for display to
1694 * the user. If possible, the name will be localized for the default
1695 * {@link Locale.Category#DISPLAY DISPLAY} locale. Returns
1696 * the empty string if this locale doesn't specify a script code.
1697 *
1698 * @return the display name of the script code for the current default
1699 * {@link Locale.Category#DISPLAY DISPLAY} locale
1700 * @since 1.7
1701 */
1702 public String getDisplayScript() {
1703 return getDisplayScript(getDefault(Category.DISPLAY));
1704 }
1705
1706 /**
1707 * Returns a name for the locale's script that is appropriate
1708 * for display to the user. If possible, the name will be
1709 * localized for the given locale. Returns the empty string if
1710 * this locale doesn't specify a script code.
1711 *
1712 * @param inLocale The locale for which to retrieve the display script.
1713 * @return the display name of the script code for the current default
1714 * {@link Locale.Category#DISPLAY DISPLAY} locale
1715 * @throws NullPointerException if <code>inLocale</code> is <code>null</code>
1716 * @since 1.7
1717 */
1718 public String getDisplayScript(Locale inLocale) {
1719 return getDisplayString(baseLocale.getScript(), inLocale, DISPLAY_SCRIPT);
1720 }
1721
1722 /**
1723 * Returns a name for the locale's country that is appropriate for display to the
1724 * user.
1725 * If possible, the name returned will be localized for the default
1726 * {@link Locale.Category#DISPLAY DISPLAY} locale.
1727 * For example, if the locale is fr_FR and the default
1728 * {@link Locale.Category#DISPLAY DISPLAY} locale
1729 * is en_US, getDisplayCountry() will return "France"; if the locale is en_US and
1730 * the default {@link Locale.Category#DISPLAY DISPLAY} locale is fr_FR,
1731 * getDisplayCountry() will return "Etats-Unis".
1732 * If the name returned cannot be localized for the default
1733 * {@link Locale.Category#DISPLAY DISPLAY} locale,
1734 * (say, we don't have a Japanese name for Croatia),
1735 * this function falls back on the English name, and uses the ISO code as a last-resort
1736 * value. If the locale doesn't specify a country, this function returns the empty string.
1737 *
1738 * @return The name of the country appropriate to the locale.
1739 */
1740 public final String getDisplayCountry() {
1741 return getDisplayCountry(getDefault(Category.DISPLAY));
1742 }
1743
1744 /**
1745 * Returns a name for the locale's country that is appropriate for display to the
1746 * user.
1747 * If possible, the name returned will be localized according to inLocale.
1748 * For example, if the locale is fr_FR and inLocale
1749 * is en_US, getDisplayCountry() will return "France"; if the locale is en_US and
1750 * inLocale is fr_FR, getDisplayCountry() will return "Etats-Unis".
1751 * If the name returned cannot be localized according to inLocale.
1752 * (say, we don't have a Japanese name for Croatia),
1753 * this function falls back on the English name, and finally
1754 * on the ISO code as a last-resort value. If the locale doesn't specify a country,
1755 * this function returns the empty string.
1756 *
1757 * @param inLocale The locale for which to retrieve the display country.
1758 * @return The name of the country appropriate to the given locale.
1759 * @exception NullPointerException if <code>inLocale</code> is <code>null</code>
1760 */
1761 public String getDisplayCountry(Locale inLocale) {
1762 return getDisplayString(baseLocale.getRegion(), inLocale, DISPLAY_COUNTRY);
1763 }
1764
1765 private String getDisplayString(String code, Locale inLocale, int type) {
1766 if (code.length() == 0) {
1767 return "";
1768 }
1769
1770 if (inLocale == null) {
1771 throw new NullPointerException();
1772 }
1773
1774 LocaleServiceProviderPool pool =
1775 LocaleServiceProviderPool.getPool(LocaleNameProvider.class);
1776 String key = (type == DISPLAY_VARIANT ? "%%"+code : code);
1777 String result = pool.getLocalizedObject(
1778 LocaleNameGetter.INSTANCE,
1779 inLocale, key, type, code);
1780 if (result != null) {
1781 return result;
1782 }
1783
1784 return code;
1785 }
1786
1787 /**
1788 * Returns a name for the locale's variant code that is appropriate for display to the
1789 * user. If possible, the name will be localized for the default
1790 * {@link Locale.Category#DISPLAY DISPLAY} locale. If the locale
1791 * doesn't specify a variant code, this function returns the empty string.
1792 *
1793 * @return The name of the display variant code appropriate to the locale.
1794 */
1795 public final String getDisplayVariant() {
1796 return getDisplayVariant(getDefault(Category.DISPLAY));
1797 }
1798
1799 /**
1800 * Returns a name for the locale's variant code that is appropriate for display to the
1801 * user. If possible, the name will be localized for inLocale. If the locale
1802 * doesn't specify a variant code, this function returns the empty string.
1803 *
1804 * @param inLocale The locale for which to retrieve the display variant code.
1805 * @return The name of the display variant code appropriate to the given locale.
1806 * @exception NullPointerException if <code>inLocale</code> is <code>null</code>
1807 */
1808 public String getDisplayVariant(Locale inLocale) {
1809 if (baseLocale.getVariant().length() == 0)
1810 return "";
1811
1812 LocaleResources lr = LocaleProviderAdapter.forJRE().getLocaleResources(inLocale);
1813
1814 String names[] = getDisplayVariantArray(inLocale);
1815
1816 // Get the localized patterns for formatting a list, and use
1817 // them to format the list.
1818 return formatList(names,
1819 lr.getLocaleName("ListPattern"),
1820 lr.getLocaleName("ListCompositionPattern"));
1821 }
1822
1823 /**
1824 * Returns a name for the locale that is appropriate for display to the
1825 * user. This will be the values returned by getDisplayLanguage(),
1826 * getDisplayScript(), getDisplayCountry(), and getDisplayVariant() assembled
1827 * into a single string. The the non-empty values are used in order,
1828 * with the second and subsequent names in parentheses. For example:
1829 * <blockquote>
1830 * language (script, country, variant)<br>
1831 * language (country)<br>
1832 * language (variant)<br>
1833 * script (country)<br>
1834 * country<br>
1835 * </blockquote>
1836 * depending on which fields are specified in the locale. If the
1837 * language, script, country, and variant fields are all empty,
1838 * this function returns the empty string.
1839 *
1840 * @return The name of the locale appropriate to display.
1841 */
1842 public final String getDisplayName() {
1843 return getDisplayName(getDefault(Category.DISPLAY));
1844 }
1845
1846 /**
1847 * Returns a name for the locale that is appropriate for display
1848 * to the user. This will be the values returned by
1849 * getDisplayLanguage(), getDisplayScript(),getDisplayCountry(),
1850 * and getDisplayVariant() assembled into a single string.
1851 * The non-empty values are used in order,
1852 * with the second and subsequent names in parentheses. For example:
1853 * <blockquote>
1854 * language (script, country, variant)<br>
1855 * language (country)<br>
1856 * language (variant)<br>
1857 * script (country)<br>
1858 * country<br>
1859 * </blockquote>
1860 * depending on which fields are specified in the locale. If the
1861 * language, script, country, and variant fields are all empty,
1862 * this function returns the empty string.
1863 *
1864 * @param inLocale The locale for which to retrieve the display name.
1865 * @return The name of the locale appropriate to display.
1866 * @throws NullPointerException if <code>inLocale</code> is <code>null</code>
1867 */
1868 public String getDisplayName(Locale inLocale) {
1869 LocaleResources lr = LocaleProviderAdapter.forJRE().getLocaleResources(inLocale);
1870
1871 String languageName = getDisplayLanguage(inLocale);
1872 String scriptName = getDisplayScript(inLocale);
1873 String countryName = getDisplayCountry(inLocale);
1874 String[] variantNames = getDisplayVariantArray(inLocale);
1875
1876 // Get the localized patterns for formatting a display name.
1877 String displayNamePattern = lr.getLocaleName("DisplayNamePattern");
1878 String listPattern = lr.getLocaleName("ListPattern");
1879 String listCompositionPattern = lr.getLocaleName("ListCompositionPattern");
1880
1881 // The display name consists of a main name, followed by qualifiers.
1882 // Typically, the format is "MainName (Qualifier, Qualifier)" but this
1883 // depends on what pattern is stored in the display locale.
1884 String mainName = null;
1885 String[] qualifierNames = null;
1886
1887 // The main name is the language, or if there is no language, the script,
1888 // then if no script, the country. If there is no language/script/country
1889 // (an anomalous situation) then the display name is simply the variant's
1890 // display name.
1891 if (languageName.length() == 0 && scriptName.length() == 0 && countryName.length() == 0) {
1892 if (variantNames.length == 0) {
1893 return "";
1894 } else {
1895 return formatList(variantNames, listPattern, listCompositionPattern);
1896 }
1897 }
1898 ArrayList<String> names = new ArrayList<>(4);
1899 if (languageName.length() != 0) {
1900 names.add(languageName);
1901 }
1902 if (scriptName.length() != 0) {
1903 names.add(scriptName);
1904 }
1905 if (countryName.length() != 0) {
1906 names.add(countryName);
1907 }
1908 if (variantNames.length != 0) {
1909 names.addAll(Arrays.asList(variantNames));
1910 }
1911
1912 // The first one in the main name
1913 mainName = names.get(0);
1914
1915 // Others are qualifiers
1916 int numNames = names.size();
1917 qualifierNames = (numNames > 1) ?
1918 names.subList(1, numNames).toArray(new String[numNames - 1]) : new String[0];
1919
1920 // Create an array whose first element is the number of remaining
1921 // elements. This serves as a selector into a ChoiceFormat pattern from
1922 // the resource. The second and third elements are the main name and
1923 // the qualifier; if there are no qualifiers, the third element is
1924 // unused by the format pattern.
1925 Object[] displayNames = {
1926 qualifierNames.length != 0 ? 2 : 1,
1927 mainName,
1928 // We could also just call formatList() and have it handle the empty
1929 // list case, but this is more efficient, and we want it to be
1930 // efficient since all the language-only locales will not have any
1931 // qualifiers.
1932 qualifierNames.length != 0 ? formatList(qualifierNames, listPattern, listCompositionPattern) : null
1933 };
1934
1935 if (displayNamePattern != null) {
1936 return new MessageFormat(displayNamePattern).format(displayNames);
1937 }
1938 else {
1939 // If we cannot get the message format pattern, then we use a simple
1940 // hard-coded pattern. This should not occur in practice unless the
1941 // installation is missing some core files (FormatData etc.).
1942 StringBuilder result = new StringBuilder();
1943 result.append((String)displayNames[1]);
1944 if (displayNames.length > 2) {
1945 result.append(" (");
1946 result.append((String)displayNames[2]);
1947 result.append(')');
1948 }
1949 return result.toString();
1950 }
1951 }
1952
1953 /**
1954 * Overrides Cloneable.
1955 */
1956 @Override
1957 public Object clone()
1958 {
1959 try {
1960 Locale that = (Locale)super.clone();
1961 return that;
1962 } catch (CloneNotSupportedException e) {
1963 throw new InternalError(e);
1964 }
1965 }
1966
1967 /**
1968 * Override hashCode.
1969 * Since Locales are often used in hashtables, caches the value
1970 * for speed.
1971 */
1972 @Override
1973 public int hashCode() {
1974 int hc = hashCodeValue;
1975 if (hc == 0) {
1976 hc = baseLocale.hashCode();
1977 if (localeExtensions != null) {
1978 hc ^= localeExtensions.hashCode();
1979 }
1980 hashCodeValue = hc;
1981 }
1982 return hc;
1983 }
1984
1985 // Overrides
1986
1987 /**
1988 * Returns true if this Locale is equal to another object. A Locale is
1989 * deemed equal to another Locale with identical language, script, country,
1990 * variant and extensions, and unequal to all other objects.
1991 *
1992 * @return true if this Locale is equal to the specified object.
1993 */
1994 @Override
1995 public boolean equals(Object obj) {
1996 if (this == obj) // quick check
1997 return true;
1998 if (!(obj instanceof Locale))
1999 return false;
2000 BaseLocale otherBase = ((Locale)obj).baseLocale;
2001 if (!baseLocale.equals(otherBase)) {
2002 return false;
2003 }
2004 if (localeExtensions == null) {
2005 return ((Locale)obj).localeExtensions == null;
2006 }
2007 return localeExtensions.equals(((Locale)obj).localeExtensions);
2008 }
2009
2010 // ================= privates =====================================
2011
2012 private transient BaseLocale baseLocale;
2013 private transient LocaleExtensions localeExtensions;
2014
2015 /**
2016 * Calculated hashcode
2017 */
2018 private transient volatile int hashCodeValue;
2019
2020 private static volatile Locale defaultLocale = initDefault();
2021 private static volatile Locale defaultDisplayLocale;
2022 private static volatile Locale defaultFormatLocale;
2023
2024 private transient volatile String languageTag;
2025
2026 /**
2027 * Return an array of the display names of the variant.
2028 * @param bundle the ResourceBundle to use to get the display names
2029 * @return an array of display names, possible of zero length.
2030 */
2031 private String[] getDisplayVariantArray(Locale inLocale) {
2032 // Split the variant name into tokens separated by '_'.
2033 StringTokenizer tokenizer = new StringTokenizer(baseLocale.getVariant(), "_");
2034 String[] names = new String[tokenizer.countTokens()];
2035
2036 // For each variant token, lookup the display name. If
2037 // not found, use the variant name itself.
2038 for (int i=0; i<names.length; ++i) {
2039 names[i] = getDisplayString(tokenizer.nextToken(),
2040 inLocale, DISPLAY_VARIANT);
2041 }
2042
2043 return names;
2044 }
2045
2046 /**
2047 * Format a list using given pattern strings.
2048 * If either of the patterns is null, then a the list is
2049 * formatted by concatenation with the delimiter ','.
2050 * @param stringList the list of strings to be formatted.
2051 * @param listPattern should create a MessageFormat taking 0-3 arguments
2052 * and formatting them into a list.
2053 * @param listCompositionPattern should take 2 arguments
2054 * and is used by composeList.
2055 * @return a string representing the list.
2056 */
2057 private static String formatList(String[] stringList, String listPattern, String listCompositionPattern) {
2058 // If we have no list patterns, compose the list in a simple,
2059 // non-localized way.
2060 if (listPattern == null || listCompositionPattern == null) {
2061 StringJoiner sj = new StringJoiner(",");
2062 for (int i = 0; i < stringList.length; ++i) {
2063 sj.add(stringList[i]);
2064 }
2065 return sj.toString();
2066 }
2067
2068 // Compose the list down to three elements if necessary
2069 if (stringList.length > 3) {
2070 MessageFormat format = new MessageFormat(listCompositionPattern);
2071 stringList = composeList(format, stringList);
2072 }
2073
2074 // Rebuild the argument list with the list length as the first element
2075 Object[] args = new Object[stringList.length + 1];
2076 System.arraycopy(stringList, 0, args, 1, stringList.length);
2077 args[0] = stringList.length;
2078
2079 // Format it using the pattern in the resource
2080 MessageFormat format = new MessageFormat(listPattern);
2081 return format.format(args);
2082 }
2083
2084 /**
2085 * Given a list of strings, return a list shortened to three elements.
2086 * Shorten it by applying the given format to the first two elements
2087 * recursively.
2088 * @param format a format which takes two arguments
2089 * @param list a list of strings
2090 * @return if the list is three elements or shorter, the same list;
2091 * otherwise, a new list of three elements.
2092 */
2093 private static String[] composeList(MessageFormat format, String[] list) {
2094 if (list.length <= 3) return list;
2095
2096 // Use the given format to compose the first two elements into one
2097 String[] listItems = { list[0], list[1] };
2098 String newItem = format.format(listItems);
2099
2100 // Form a new list one element shorter
2101 String[] newList = new String[list.length-1];
2102 System.arraycopy(list, 2, newList, 1, newList.length-1);
2103 newList[0] = newItem;
2104
2105 // Recurse
2106 return composeList(format, newList);
2107 }
2108
2109 // Duplicate of sun.util.locale.UnicodeLocaleExtension.isKey in order to
2110 // avoid its class loading.
2111 private static boolean isUnicodeExtensionKey(String s) {
2112 // 2alphanum
2113 return (s.length() == 2) && LocaleUtils.isAlphaNumericString(s);
2114 }
2115
2116 /**
2117 * @serialField language String
2118 * language subtag in lower case. (See <a href="java/util/Locale.html#getLanguage()">getLanguage()</a>)
2119 * @serialField country String
2120 * country subtag in upper case. (See <a href="java/util/Locale.html#getCountry()">getCountry()</a>)
2121 * @serialField variant String
2122 * variant subtags separated by LOWLINE characters. (See <a href="java/util/Locale.html#getVariant()">getVariant()</a>)
2123 * @serialField hashcode int
2124 * deprecated, for forward compatibility only
2125 * @serialField script String
2126 * script subtag in title case (See <a href="java/util/Locale.html#getScript()">getScript()</a>)
2127 * @serialField extensions String
2128 * canonical representation of extensions, that is,
2129 * BCP47 extensions in alphabetical order followed by
2130 * BCP47 private use subtags, all in lower case letters
2131 * separated by HYPHEN-MINUS characters.
2132 * (See <a href="java/util/Locale.html#getExtensionKeys()">getExtensionKeys()</a>,
2133 * <a href="java/util/Locale.html#getExtension(char)">getExtension(char)</a>)
2134 */
2135 private static final ObjectStreamField[] serialPersistentFields = {
2136 new ObjectStreamField("language", String.class),
2137 new ObjectStreamField("country", String.class),
2138 new ObjectStreamField("variant", String.class),
2139 new ObjectStreamField("hashcode", int.class),
2140 new ObjectStreamField("script", String.class),
2141 new ObjectStreamField("extensions", String.class),
2142 };
2143
2144 /**
2145 * Serializes this <code>Locale</code> to the specified <code>ObjectOutputStream</code>.
2146 * @param out the <code>ObjectOutputStream</code> to write
2147 * @throws IOException
2148 * @since 1.7
2149 */
2150 private void writeObject(ObjectOutputStream out) throws IOException {
2151 ObjectOutputStream.PutField fields = out.putFields();
2152 fields.put("language", baseLocale.getLanguage());
2153 fields.put("script", baseLocale.getScript());
2154 fields.put("country", baseLocale.getRegion());
2155 fields.put("variant", baseLocale.getVariant());
2156 fields.put("extensions", localeExtensions == null ? "" : localeExtensions.getID());
2157 fields.put("hashcode", -1); // place holder just for backward support
2158 out.writeFields();
2159 }
2160
2161 /**
2162 * Deserializes this <code>Locale</code>.
2163 * @param in the <code>ObjectInputStream</code> to read
2164 * @throws IOException
2165 * @throws ClassNotFoundException
2166 * @throws IllformedLocaleException
2167 * @since 1.7
2168 */
2169 private void readObject(ObjectInputStream in) throws IOException, ClassNotFoundException {
2170 ObjectInputStream.GetField fields = in.readFields();
2171 String language = (String)fields.get("language", "");
2172 String script = (String)fields.get("script", "");
2173 String country = (String)fields.get("country", "");
2174 String variant = (String)fields.get("variant", "");
2175 String extStr = (String)fields.get("extensions", "");
2176 baseLocale = BaseLocale.getInstance(convertOldISOCodes(language), script, country, variant);
2177 if (extStr.length() > 0) {
2178 try {
2179 InternalLocaleBuilder bldr = new InternalLocaleBuilder();
2180 bldr.setExtensions(extStr);
2181 localeExtensions = bldr.getLocaleExtensions();
2182 } catch (LocaleSyntaxException e) {
2183 throw new IllformedLocaleException(e.getMessage());
2184 }
2185 } else {
2186 localeExtensions = null;
2187 }
2188 }
2189
2190 /**
2191 * Returns a cached <code>Locale</code> instance equivalent to
2192 * the deserialized <code>Locale</code>. When serialized
2193 * language, country and variant fields read from the object data stream
2194 * are exactly "ja", "JP", "JP" or "th", "TH", "TH" and script/extensions
2195 * fields are empty, this method supplies <code>UNICODE_LOCALE_EXTENSION</code>
2196 * "ca"/"japanese" (calendar type is "japanese") or "nu"/"thai" (number script
2197 * type is "thai"). See <a href="Locale.html#special_cases_constructor">Special Cases</a>
2198 * for more information.
2199 *
2200 * @return an instance of <code>Locale</code> equivalent to
2201 * the deserialized <code>Locale</code>.
2202 * @throws java.io.ObjectStreamException
2203 */
2204 private Object readResolve() throws java.io.ObjectStreamException {
2205 return getInstance(baseLocale.getLanguage(), baseLocale.getScript(),
2206 baseLocale.getRegion(), baseLocale.getVariant(), localeExtensions);
2207 }
2208
2209 private static volatile String[] isoLanguages;
2210
2211 private static volatile String[] isoCountries;
2212
2213 private static String convertOldISOCodes(String language) {
2214 // we accept both the old and the new ISO codes for the languages whose ISO
2215 // codes have changed, but we always store the OLD code, for backward compatibility
2216 language = LocaleUtils.toLowerString(language).intern();
2217 if (language == "he") {
2218 return "iw";
2219 } else if (language == "yi") {
2220 return "ji";
2221 } else if (language == "id") {
2222 return "in";
2223 } else {
2224 return language;
2225 }
2226 }
2227
2228 private static LocaleExtensions getCompatibilityExtensions(String language,
2229 String script,
2230 String country,
2231 String variant) {
2232 LocaleExtensions extensions = null;
2233 // Special cases for backward compatibility support
2234 if (LocaleUtils.caseIgnoreMatch(language, "ja")
2235 && script.length() == 0
2236 && LocaleUtils.caseIgnoreMatch(country, "jp")
2237 && "JP".equals(variant)) {
2238 // ja_JP_JP -> u-ca-japanese (calendar = japanese)
2239 extensions = LocaleExtensions.CALENDAR_JAPANESE;
2240 } else if (LocaleUtils.caseIgnoreMatch(language, "th")
2241 && script.length() == 0
2242 && LocaleUtils.caseIgnoreMatch(country, "th")
2243 && "TH".equals(variant)) {
2244 // th_TH_TH -> u-nu-thai (numbersystem = thai)
2245 extensions = LocaleExtensions.NUMBER_THAI;
2246 }
2247 return extensions;
2248 }
2249
2250 /**
2251 * Obtains a localized locale names from a LocaleNameProvider
2252 * implementation.
2253 */
2254 private static class LocaleNameGetter
2255 implements LocaleServiceProviderPool.LocalizedObjectGetter<LocaleNameProvider, String> {
2256 private static final LocaleNameGetter INSTANCE = new LocaleNameGetter();
2257
2258 @Override
2259 public String getObject(LocaleNameProvider localeNameProvider,
2260 Locale locale,
2261 String key,
2262 Object... params) {
2263 assert params.length == 2;
2264 int type = (Integer)params[0];
2265 String code = (String)params[1];
2266
2267 switch(type) {
2268 case DISPLAY_LANGUAGE:
2269 return localeNameProvider.getDisplayLanguage(code, locale);
2270 case DISPLAY_COUNTRY:
2271 return localeNameProvider.getDisplayCountry(code, locale);
2272 case DISPLAY_VARIANT:
2273 return localeNameProvider.getDisplayVariant(code, locale);
2274 case DISPLAY_SCRIPT:
2275 return localeNameProvider.getDisplayScript(code, locale);
2276 default:
2277 assert false; // shouldn't happen
2278 }
2279
2280 return null;
2281 }
2282 }
2283
2284 /**
2285 * Enum for locale categories. These locale categories are used to get/set
2286 * the default locale for the specific functionality represented by the
2287 * category.
2288 *
2289 * @see #getDefault(Locale.Category)
2290 * @see #setDefault(Locale.Category, Locale)
2291 * @since 1.7
2292 */
2293 public enum Category {
2294
2295 /**
2296 * Category used to represent the default locale for
2297 * displaying user interfaces.
2298 */
2299 DISPLAY("user.language.display",
2300 "user.script.display",
2301 "user.country.display",
2302 "user.variant.display"),
2303
2304 /**
2305 * Category used to represent the default locale for
2306 * formatting dates, numbers, and/or currencies.
2307 */
2308 FORMAT("user.language.format",
2309 "user.script.format",
2310 "user.country.format",
2311 "user.variant.format");
2312
2313 Category(String languageKey, String scriptKey, String countryKey, String variantKey) {
2314 this.languageKey = languageKey;
2315 this.scriptKey = scriptKey;
2316 this.countryKey = countryKey;
2317 this.variantKey = variantKey;
2318 }
2319
2320 final String languageKey;
2321 final String scriptKey;
2322 final String countryKey;
2323 final String variantKey;
2324 }
2325
2326 /**
2327 * <code>Builder</code> is used to build instances of <code>Locale</code>
2328 * from values configured by the setters. Unlike the <code>Locale</code>
2329 * constructors, the <code>Builder</code> checks if a value configured by a
2330 * setter satisfies the syntax requirements defined by the <code>Locale</code>
2331 * class. A <code>Locale</code> object created by a <code>Builder</code> is
2332 * well-formed and can be transformed to a well-formed IETF BCP 47 language tag
2333 * without losing information.
2334 *
2335 * <p><b>Note:</b> The <code>Locale</code> class does not provide any
2336 * syntactic restrictions on variant, while BCP 47 requires each variant
2337 * subtag to be 5 to 8 alphanumerics or a single numeric followed by 3
2338 * alphanumerics. The method <code>setVariant</code> throws
2339 * <code>IllformedLocaleException</code> for a variant that does not satisfy
2340 * this restriction. If it is necessary to support such a variant, use a
2341 * Locale constructor. However, keep in mind that a <code>Locale</code>
2342 * object created this way might lose the variant information when
2343 * transformed to a BCP 47 language tag.
2344 *
2345 * <p>The following example shows how to create a <code>Locale</code> object
2346 * with the <code>Builder</code>.
2347 * <blockquote>
2348 * <pre>
2349 * Locale aLocale = new Builder().setLanguage("sr").setScript("Latn").setRegion("RS").build();
2350 * </pre>
2351 * </blockquote>
2352 *
2353 * <p>Builders can be reused; <code>clear()</code> resets all
2354 * fields to their default values.
2355 *
2356 * @see Locale#forLanguageTag
2357 * @since 1.7
2358 */
2359 public static final class Builder {
2360 private final InternalLocaleBuilder localeBuilder;
2361
2362 /**
2363 * Constructs an empty Builder. The default value of all
2364 * fields, extensions, and private use information is the
2365 * empty string.
2366 */
2367 public Builder() {
2368 localeBuilder = new InternalLocaleBuilder();
2369 }
2370
2371 /**
2372 * Resets the <code>Builder</code> to match the provided
2373 * <code>locale</code>. Existing state is discarded.
2374 *
2375 * <p>All fields of the locale must be well-formed, see {@link Locale}.
2376 *
2377 * <p>Locales with any ill-formed fields cause
2378 * <code>IllformedLocaleException</code> to be thrown, except for the
2379 * following three cases which are accepted for compatibility
2380 * reasons:<ul>
2381 * <li>Locale("ja", "JP", "JP") is treated as "ja-JP-u-ca-japanese"
2382 * <li>Locale("th", "TH", "TH") is treated as "th-TH-u-nu-thai"
2383 * <li>Locale("no", "NO", "NY") is treated as "nn-NO"</ul>
2384 *
2385 * @param locale the locale
2386 * @return This builder.
2387 * @throws IllformedLocaleException if <code>locale</code> has
2388 * any ill-formed fields.
2389 * @throws NullPointerException if <code>locale</code> is null.
2390 */
2391 public Builder setLocale(Locale locale) {
2392 try {
2393 localeBuilder.setLocale(locale.baseLocale, locale.localeExtensions);
2394 } catch (LocaleSyntaxException e) {
2395 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex());
2396 }
2397 return this;
2398 }
2399
2400 /**
2401 * Resets the Builder to match the provided IETF BCP 47
2402 * language tag. Discards the existing state. Null and the
2403 * empty string cause the builder to be reset, like {@link
2404 * #clear}. Grandfathered tags (see {@link
2405 * Locale#forLanguageTag}) are converted to their canonical
2406 * form before being processed. Otherwise, the language tag
2407 * must be well-formed (see {@link Locale}) or an exception is
2408 * thrown (unlike <code>Locale.forLanguageTag</code>, which
2409 * just discards ill-formed and following portions of the
2410 * tag).
2411 *
2412 * @param languageTag the language tag
2413 * @return This builder.
2414 * @throws IllformedLocaleException if <code>languageTag</code> is ill-formed
2415 * @see Locale#forLanguageTag(String)
2416 */
2417 public Builder setLanguageTag(String languageTag) {
2418 ParseStatus sts = new ParseStatus();
2419 LanguageTag tag = LanguageTag.parse(languageTag, sts);
2420 if (sts.isError()) {
2421 throw new IllformedLocaleException(sts.getErrorMessage(), sts.getErrorIndex());
2422 }
2423 localeBuilder.setLanguageTag(tag);
2424 return this;
2425 }
2426
2427 /**
2428 * Sets the language. If <code>language</code> is the empty string or
2429 * null, the language in this <code>Builder</code> is removed. Otherwise,
2430 * the language must be <a href="./Locale.html#def_language">well-formed</a>
2431 * or an exception is thrown.
2432 *
2433 * <p>The typical language value is a two or three-letter language
2434 * code as defined in ISO639.
2435 *
2436 * @param language the language
2437 * @return This builder.
2438 * @throws IllformedLocaleException if <code>language</code> is ill-formed
2439 */
2440 public Builder setLanguage(String language) {
2441 try {
2442 localeBuilder.setLanguage(language);
2443 } catch (LocaleSyntaxException e) {
2444 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex());
2445 }
2446 return this;
2447 }
2448
2449 /**
2450 * Sets the script. If <code>script</code> is null or the empty string,
2451 * the script in this <code>Builder</code> is removed.
2452 * Otherwise, the script must be <a href="./Locale.html#def_script">well-formed</a> or an
2453 * exception is thrown.
2454 *
2455 * <p>The typical script value is a four-letter script code as defined by ISO 15924.
2456 *
2457 * @param script the script
2458 * @return This builder.
2459 * @throws IllformedLocaleException if <code>script</code> is ill-formed
2460 */
2461 public Builder setScript(String script) {
2462 try {
2463 localeBuilder.setScript(script);
2464 } catch (LocaleSyntaxException e) {
2465 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex());
2466 }
2467 return this;
2468 }
2469
2470 /**
2471 * Sets the region. If region is null or the empty string, the region
2472 * in this <code>Builder</code> is removed. Otherwise,
2473 * the region must be <a href="./Locale.html#def_region">well-formed</a> or an
2474 * exception is thrown.
2475 *
2476 * <p>The typical region value is a two-letter ISO 3166 code or a
2477 * three-digit UN M.49 area code.
2478 *
2479 * <p>The country value in the <code>Locale</code> created by the
2480 * <code>Builder</code> is always normalized to upper case.
2481 *
2482 * @param region the region
2483 * @return This builder.
2484 * @throws IllformedLocaleException if <code>region</code> is ill-formed
2485 */
2486 public Builder setRegion(String region) {
2487 try {
2488 localeBuilder.setRegion(region);
2489 } catch (LocaleSyntaxException e) {
2490 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex());
2491 }
2492 return this;
2493 }
2494
2495 /**
2496 * Sets the variant. If variant is null or the empty string, the
2497 * variant in this <code>Builder</code> is removed. Otherwise, it
2498 * must consist of one or more <a href="./Locale.html#def_variant">well-formed</a>
2499 * subtags, or an exception is thrown.
2500 *
2501 * <p><b>Note:</b> This method checks if <code>variant</code>
2502 * satisfies the IETF BCP 47 variant subtag's syntax requirements,
2503 * and normalizes the value to lowercase letters. However,
2504 * the <code>Locale</code> class does not impose any syntactic
2505 * restriction on variant, and the variant value in
2506 * <code>Locale</code> is case sensitive. To set such a variant,
2507 * use a Locale constructor.
2508 *
2509 * @param variant the variant
2510 * @return This builder.
2511 * @throws IllformedLocaleException if <code>variant</code> is ill-formed
2512 */
2513 public Builder setVariant(String variant) {
2514 try {
2515 localeBuilder.setVariant(variant);
2516 } catch (LocaleSyntaxException e) {
2517 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex());
2518 }
2519 return this;
2520 }
2521
2522 /**
2523 * Sets the extension for the given key. If the value is null or the
2524 * empty string, the extension is removed. Otherwise, the extension
2525 * must be <a href="./Locale.html#def_extensions">well-formed</a> or an exception
2526 * is thrown.
2527 *
2528 * <p><b>Note:</b> The key {@link Locale#UNICODE_LOCALE_EXTENSION
2529 * UNICODE_LOCALE_EXTENSION} ('u') is used for the Unicode locale extension.
2530 * Setting a value for this key replaces any existing Unicode locale key/type
2531 * pairs with those defined in the extension.
2532 *
2533 * <p><b>Note:</b> The key {@link Locale#PRIVATE_USE_EXTENSION
2534 * PRIVATE_USE_EXTENSION} ('x') is used for the private use code. To be
2535 * well-formed, the value for this key needs only to have subtags of one to
2536 * eight alphanumeric characters, not two to eight as in the general case.
2537 *
2538 * @param key the extension key
2539 * @param value the extension value
2540 * @return This builder.
2541 * @throws IllformedLocaleException if <code>key</code> is illegal
2542 * or <code>value</code> is ill-formed
2543 * @see #setUnicodeLocaleKeyword(String, String)
2544 */
2545 public Builder setExtension(char key, String value) {
2546 try {
2547 localeBuilder.setExtension(key, value);
2548 } catch (LocaleSyntaxException e) {
2549 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex());
2550 }
2551 return this;
2552 }
2553
2554 /**
2555 * Sets the Unicode locale keyword type for the given key. If the type
2556 * is null, the Unicode keyword is removed. Otherwise, the key must be
2557 * non-null and both key and type must be <a
2558 * href="./Locale.html#def_locale_extension">well-formed</a> or an exception
2559 * is thrown.
2560 *
2561 * <p>Keys and types are converted to lower case.
2562 *
2563 * <p><b>Note</b>:Setting the 'u' extension via {@link #setExtension}
2564 * replaces all Unicode locale keywords with those defined in the
2565 * extension.
2566 *
2567 * @param key the Unicode locale key
2568 * @param type the Unicode locale type
2569 * @return This builder.
2570 * @throws IllformedLocaleException if <code>key</code> or <code>type</code>
2571 * is ill-formed
2572 * @throws NullPointerException if <code>key</code> is null
2573 * @see #setExtension(char, String)
2574 */
2575 public Builder setUnicodeLocaleKeyword(String key, String type) {
2576 try {
2577 localeBuilder.setUnicodeLocaleKeyword(key, type);
2578 } catch (LocaleSyntaxException e) {
2579 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex());
2580 }
2581 return this;
2582 }
2583
2584 /**
2585 * Adds a unicode locale attribute, if not already present, otherwise
2586 * has no effect. The attribute must not be null and must be <a
2587 * href="./Locale.html#def_locale_extension">well-formed</a> or an exception
2588 * is thrown.
2589 *
2590 * @param attribute the attribute
2591 * @return This builder.
2592 * @throws NullPointerException if <code>attribute</code> is null
2593 * @throws IllformedLocaleException if <code>attribute</code> is ill-formed
2594 * @see #setExtension(char, String)
2595 */
2596 public Builder addUnicodeLocaleAttribute(String attribute) {
2597 try {
2598 localeBuilder.addUnicodeLocaleAttribute(attribute);
2599 } catch (LocaleSyntaxException e) {
2600 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex());
2601 }
2602 return this;
2603 }
2604
2605 /**
2606 * Removes a unicode locale attribute, if present, otherwise has no
2607 * effect. The attribute must not be null and must be <a
2608 * href="./Locale.html#def_locale_extension">well-formed</a> or an exception
2609 * is thrown.
2610 *
2611 * <p>Attribute comparison for removal is case-insensitive.
2612 *
2613 * @param attribute the attribute
2614 * @return This builder.
2615 * @throws NullPointerException if <code>attribute</code> is null
2616 * @throws IllformedLocaleException if <code>attribute</code> is ill-formed
2617 * @see #setExtension(char, String)
2618 */
2619 public Builder removeUnicodeLocaleAttribute(String attribute) {
2620 try {
2621 localeBuilder.removeUnicodeLocaleAttribute(attribute);
2622 } catch (LocaleSyntaxException e) {
2623 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex());
2624 }
2625 return this;
2626 }
2627
2628 /**
2629 * Resets the builder to its initial, empty state.
2630 *
2631 * @return This builder.
2632 */
2633 public Builder clear() {
2634 localeBuilder.clear();
2635 return this;
2636 }
2637
2638 /**
2639 * Resets the extensions to their initial, empty state.
2640 * Language, script, region and variant are unchanged.
2641 *
2642 * @return This builder.
2643 * @see #setExtension(char, String)
2644 */
2645 public Builder clearExtensions() {
2646 localeBuilder.clearExtensions();
2647 return this;
2648 }
2649
2650 /**
2651 * Returns an instance of <code>Locale</code> created from the fields set
2652 * on this builder.
2653 *
2654 * <p>This applies the conversions listed in {@link Locale#forLanguageTag}
2655 * when constructing a Locale. (Grandfathered tags are handled in
2656 * {@link #setLanguageTag}.)
2657 *
2658 * @return A Locale.
2659 */
2660 public Locale build() {
2661 BaseLocale baseloc = localeBuilder.getBaseLocale();
2662 LocaleExtensions extensions = localeBuilder.getLocaleExtensions();
2663 if (extensions == null && baseloc.getVariant().length() > 0) {
2664 extensions = getCompatibilityExtensions(baseloc.getLanguage(), baseloc.getScript(),
2665 baseloc.getRegion(), baseloc.getVariant());
2666 }
2667 return Locale.getInstance(baseloc, extensions);
2668 }
2669 }
2670
2671 /**
2672 * This enum provides constants to select a filtering mode for locale
2673 * matching. Refer to <a href="http://tools.ietf.org/html/rfc4647">RFC 4647
2674 * Matching of Language Tags</a> for details.
2675 *
2676 * <p>As an example, think of two Language Priority Lists each of which
2677 * includes only one language range and a set of following language tags:
2678 *
2679 * <pre>
2680 * de (German)
2681 * de-DE (German, Germany)
2682 * de-Deva (German, in Devanagari script)
2683 * de-Deva-DE (German, in Devanagari script, Germany)
2684 * de-DE-1996 (German, Germany, orthography of 1996)
2685 * de-Latn-DE (German, in Latin script, Germany)
2686 * de-Latn-DE-1996 (German, in Latin script, Germany, orthography of 1996)
2687 * </pre>
2688 *
2689 * The filtering method will behave as follows:
2690 *
2691 * <table cellpadding=2 summary="Filtering method behavior">
2692 * <tr>
2693 * <th>Filtering Mode</th>
2694 * <th>Language Priority List: {@code "de-DE"}</th>
2695 * <th>Language Priority List: {@code "de-*-DE"}</th>
2696 * </tr>
2697 * <tr>
2698 * <td valign=top>
2699 * {@link FilteringMode#AUTOSELECT_FILTERING AUTOSELECT_FILTERING}
2700 * </td>
2701 * <td valign=top>
2702 * Performs <em>basic</em> filtering and returns {@code "de-DE"} and
2703 * {@code "de-DE-1996"}.
2704 * </td>
2705 * <td valign=top>
2706 * Performs <em>extended</em> filtering and returns {@code "de-DE"},
2707 * {@code "de-Deva-DE"}, {@code "de-DE-1996"}, {@code "de-Latn-DE"}, and
2708 * {@code "de-Latn-DE-1996"}.
2709 * </td>
2710 * </tr>
2711 * <tr>
2712 * <td valign=top>
2713 * {@link FilteringMode#EXTENDED_FILTERING EXTENDED_FILTERING}
2714 * </td>
2715 * <td valign=top>
2716 * Performs <em>extended</em> filtering and returns {@code "de-DE"},
2717 * {@code "de-Deva-DE"}, {@code "de-DE-1996"}, {@code "de-Latn-DE"}, and
2718 * {@code "de-Latn-DE-1996"}.
2719 * </td>
2720 * <td valign=top>Same as above.</td>
2721 * </tr>
2722 * <tr>
2723 * <td valign=top>
2724 * {@link FilteringMode#IGNORE_EXTENDED_RANGES IGNORE_EXTENDED_RANGES}
2725 * </td>
2726 * <td valign=top>
2727 * Performs <em>basic</em> filtering and returns {@code "de-DE"} and
2728 * {@code "de-DE-1996"}.
2729 * </td>
2730 * <td valign=top>
2731 * Performs <em>basic</em> filtering and returns {@code null} because
2732 * nothing matches.
2733 * </td>
2734 * </tr>
2735 * <tr>
2736 * <td valign=top>
2737 * {@link FilteringMode#MAP_EXTENDED_RANGES MAP_EXTENDED_RANGES}
2738 * </td>
2739 * <td valign=top>Same as above.</td>
2740 * <td valign=top>
2741 * Performs <em>basic</em> filtering and returns {@code "de-DE"} and
2742 * {@code "de-DE-1996"} because {@code "de-*-DE"} is mapped to
2743 * {@code "de-DE"}.
2744 * </td>
2745 * </tr>
2746 * <tr>
2747 * <td valign=top>
2748 * {@link FilteringMode#REJECT_EXTENDED_RANGES REJECT_EXTENDED_RANGES}
2749 * </td>
2750 * <td valign=top>Same as above.</td>
2751 * <td valign=top>
2752 * Throws {@link IllegalArgumentException} because {@code "de-*-DE"} is
2753 * not a valid basic language range.
2754 * </td>
2755 * </tr>
2756 * </table>
2757 *
2758 * @see #filter(List, Collection, FilteringMode)
2759 * @see #filterTags(List, Collection, FilteringMode)
2760 *
2761 * @since 1.8
2762 */
2763 public static enum FilteringMode {
2764 /**
2765 * Specifies automatic filtering mode based on the given Language
2766 * Priority List consisting of language ranges. If all of the ranges
2767 * are basic, basic filtering is selected. Otherwise, extended
2768 * filtering is selected.
2769 */
2770 AUTOSELECT_FILTERING,
2771
2772 /**
2773 * Specifies extended filtering.
2774 */
2775 EXTENDED_FILTERING,
2776
2777 /**
2778 * Specifies basic filtering: Note that any extended language ranges
2779 * included in the given Language Priority List are ignored.
2780 */
2781 IGNORE_EXTENDED_RANGES,
2782
2783 /**
2784 * Specifies basic filtering: If any extended language ranges are
2785 * included in the given Language Priority List, they are mapped to the
2786 * basic language range. Specifically, a language range starting with a
2787 * subtag {@code "*"} is treated as a language range {@code "*"}. For
2788 * example, {@code "*-US"} is treated as {@code "*"}. If {@code "*"} is
2789 * not the first subtag, {@code "*"} and extra {@code "-"} are removed.
2790 * For example, {@code "ja-*-JP"} is mapped to {@code "ja-JP"}.
2791 */
2792 MAP_EXTENDED_RANGES,
2793
2794 /**
2795 * Specifies basic filtering: If any extended language ranges are
2796 * included in the given Language Priority List, the list is rejected
2797 * and the filtering method throws {@link IllegalArgumentException}.
2798 */
2799 REJECT_EXTENDED_RANGES
2800 };
2801
2802 /**
2803 * This class expresses a <em>Language Range</em> defined in
2804 * <a href="http://tools.ietf.org/html/rfc4647">RFC 4647 Matching of
2805 * Language Tags</a>. A language range is an identifier which is used to
2806 * select language tag(s) meeting specific requirements by using the
2807 * mechanisms described in <a href="Locale.html#LocaleMatching">Locale
2808 * Matching</a>. A list which represents a user's preferences and consists
2809 * of language ranges is called a <em>Language Priority List</em>.
2810 *
2811 * <p>There are two types of language ranges: basic and extended. In RFC
2812 * 4647, the syntax of language ranges is expressed in
2813 * <a href="http://tools.ietf.org/html/rfc4234">ABNF</a> as follows:
2814 * <blockquote>
2815 * <pre>
2816 * basic-language-range = (1*8ALPHA *("-" 1*8alphanum)) / "*"
2817 * extended-language-range = (1*8ALPHA / "*")
2818 * *("-" (1*8alphanum / "*"))
2819 * alphanum = ALPHA / DIGIT
2820 * </pre>
2821 * </blockquote>
2822 * For example, {@code "en"} (English), {@code "ja-JP"} (Japanese, Japan),
2823 * {@code "*"} (special language range which matches any language tag) are
2824 * basic language ranges, whereas {@code "*-CH"} (any languages,
2825 * Switzerland), {@code "es-*"} (Spanish, any regions), and
2826 * {@code "zh-Hant-*"} (Traditional Chinese, any regions) are extended
2827 * language ranges.
2828 *
2829 * @see #filter
2830 * @see #filterTags
2831 * @see #lookup
2832 * @see #lookupTag
2833 *
2834 * @since 1.8
2835 */
2836 public static final class LanguageRange {
2837
2838 /**
2839 * A constant holding the maximum value of weight, 1.0, which indicates
2840 * that the language range is a good fit for the user.
2841 */
2842 public static final double MAX_WEIGHT = 1.0;
2843
2844 /**
2845 * A constant holding the minimum value of weight, 0.0, which indicates
2846 * that the language range is not a good fit for the user.
2847 */
2848 public static final double MIN_WEIGHT = 0.0;
2849
2850 private final String range;
2851 private final double weight;
2852
2853 private volatile int hash;
2854
2855 /**
2856 * Constructs a {@code LanguageRange} using the given {@code range}.
2857 * Note that no validation is done against the IANA Language Subtag
2858 * Registry at time of construction.
2859 *
2860 * <p>This is equivalent to {@code LanguageRange(range, MAX_WEIGHT)}.
2861 *
2862 * @param range a language range
2863 * @throws NullPointerException if the given {@code range} is
2864 * {@code null}
2865 */
2866 public LanguageRange(String range) {
2867 this(range, MAX_WEIGHT);
2868 }
2869
2870 /**
2871 * Constructs a {@code LanguageRange} using the given {@code range} and
2872 * {@code weight}. Note that no validation is done against the IANA
2873 * Language Subtag Registry at time of construction.
2874 *
2875 * @param range a language range
2876 * @param weight a weight value between {@code MIN_WEIGHT} and
2877 * {@code MAX_WEIGHT}
2878 * @throws NullPointerException if the given {@code range} is
2879 * {@code null}
2880 * @throws IllegalArgumentException if the given {@code weight} is less
2881 * than {@code MIN_WEIGHT} or greater than {@code MAX_WEIGHT}
2882 */
2883 public LanguageRange(String range, double weight) {
2884 if (range == null) {
2885 throw new NullPointerException();
2886 }
2887 if (weight < MIN_WEIGHT || weight > MAX_WEIGHT) {
2888 throw new IllegalArgumentException("weight=" + weight);
2889 }
2890
2891 range = range.toLowerCase();
2892
2893 // Do syntax check.
2894 boolean isIllFormed = false;
2895 String[] subtags = range.split("-");
2896 if (isSubtagIllFormed(subtags[0], true)
2897 || range.endsWith("-")) {
2898 isIllFormed = true;
2899 } else {
2900 for (int i = 1; i < subtags.length; i++) {
2901 if (isSubtagIllFormed(subtags[i], false)) {
2902 isIllFormed = true;
2903 break;
2904 }
2905 }
2906 }
2907 if (isIllFormed) {
2908 throw new IllegalArgumentException("range=" + range);
2909 }
2910
2911 this.range = range;
2912 this.weight = weight;
2913 }
2914
2915 private static boolean isSubtagIllFormed(String subtag,
2916 boolean isFirstSubtag) {
2917 if (subtag.equals("") || subtag.length() > 8) {
2918 return true;
2919 } else if (subtag.equals("*")) {
2920 return false;
2921 }
2922 char[] charArray = subtag.toCharArray();
2923 if (isFirstSubtag) { // ALPHA
2924 for (char c : charArray) {
2925 if (c < 'a' || c > 'z') {
2926 return true;
2927 }
2928 }
2929 } else { // ALPHA / DIGIT
2930 for (char c : charArray) {
2931 if (c < '0' || (c > '9' && c < 'a') || c > 'z') {
2932 return true;
2933 }
2934 }
2935 }
2936 return false;
2937 }
2938
2939 /**
2940 * Returns the language range of this {@code LanguageRange}.
2941 *
2942 * @return the language range.
2943 */
2944 public String getRange() {
2945 return range;
2946 }
2947
2948 /**
2949 * Returns the weight of this {@code LanguageRange}.
2950 *
2951 * @return the weight value.
2952 */
2953 public double getWeight() {
2954 return weight;
2955 }
2956
2957 /**
2958 * Parses the given {@code ranges} to generate a Language Priority List.
2959 *
2960 * <p>This method performs a syntactic check for each language range in
2961 * the given {@code ranges} but doesn't do validation using the IANA
2962 * Language Subtag Registry.
2963 *
2964 * <p>The {@code ranges} to be given can take one of the following
2965 * forms:
2966 *
2967 * <pre>
2968 * "Accept-Language: ja,en;q=0.4" (weighted list with Accept-Language prefix)
2969 * "ja,en;q=0.4" (weighted list)
2970 * "ja,en" (prioritized list)
2971 * </pre>
2972 *
2973 * In a weighted list, each language range is given a weight value.
2974 * The weight value is identical to the "quality value" in
2975 * <a href="http://tools.ietf.org/html/rfc2616">RFC 2616</a>, and it
2976 * expresses how much the user prefers the language. A weight value is
2977 * specified after a corresponding language range followed by
2978 * {@code ";q="}, and the default weight value is {@code MAX_WEIGHT}
2979 * when it is omitted.
2980 *
2981 * <p>Unlike a weighted list, language ranges in a prioritized list
2982 * are sorted in the descending order based on its priority. The first
2983 * language range has the highest priority and meets the user's
2984 * preference most.
2985 *
2986 * <p>In either case, language ranges are sorted in descending order in
2987 * the Language Priority List based on priority or weight. If a
2988 * language range appears in the given {@code ranges} more than once,
2989 * only the first one is included on the Language Priority List.
2990 *
2991 * <p>The returned list consists of language ranges from the given
2992 * {@code ranges} and their equivalents found in the IANA Language
2993 * Subtag Registry. For example, if the given {@code ranges} is
2994 * {@code "Accept-Language: iw,en-us;q=0.7,en;q=0.3"}, the elements in
2995 * the list to be returned are:
2996 *
2997 * <pre>
2998 * <b>Range</b> <b>Weight</b>
2999 * "iw" (older tag for Hebrew) 1.0
3000 * "he" (new preferred code for Hebrew) 1.0
3001 * "en-us" (English, United States) 0.7
3002 * "en" (English) 0.3
3003 * </pre>
3004 *
3005 * Two language ranges, {@code "iw"} and {@code "he"}, have the same
3006 * highest priority in the list. By adding {@code "he"} to the user's
3007 * Language Priority List, locale-matching method can find Hebrew as a
3008 * matching locale (or language tag) even if the application or system
3009 * offers only {@code "he"} as a supported locale (or language tag).
3010 *
3011 * @param ranges a list of comma-separated language ranges or a list of
3012 * language ranges in the form of the "Accept-Language" header
3013 * defined in <a href="http://tools.ietf.org/html/rfc2616">RFC
3014 * 2616</a>
3015 * @return a Language Priority List consisting of language ranges
3016 * included in the given {@code ranges} and their equivalent
3017 * language ranges if available. The list is modifiable.
3018 * @throws NullPointerException if {@code ranges} is null
3019 * @throws IllegalArgumentException if a language range or a weight
3020 * found in the given {@code ranges} is ill-formed
3021 */
3022 public static List<LanguageRange> parse(String ranges) {
3023 return LocaleMatcher.parse(ranges);
3024 }
3025
3026 /**
3027 * Parses the given {@code ranges} to generate a Language Priority
3028 * List, and then customizes the list using the given {@code map}.
3029 * This method is equivalent to
3030 * {@code mapEquivalents(parse(ranges), map)}.
3031 *
3032 * @param ranges a list of comma-separated language ranges or a list
3033 * of language ranges in the form of the "Accept-Language" header
3034 * defined in <a href="http://tools.ietf.org/html/rfc2616">RFC
3035 * 2616</a>
3036 * @param map a map containing information to customize language ranges
3037 * @return a Language Priority List with customization. The list is
3038 * modifiable.
3039 * @throws NullPointerException if {@code ranges} is null
3040 * @throws IllegalArgumentException if a language range or a weight
3041 * found in the given {@code ranges} is ill-formed
3042 * @see #parse(String)
3043 * @see #mapEquivalents
3044 */
3045 public static List<LanguageRange> parse(String ranges,
3046 Map<String, List<String>> map) {
3047 return mapEquivalents(parse(ranges), map);
3048 }
3049
3050 /**
3051 * Generates a new customized Language Priority List using the given
3052 * {@code priorityList} and {@code map}. If the given {@code map} is
3053 * empty, this method returns a copy of the given {@code priorityList}.
3054 *
3055 * <p>In the map, a key represents a language range whereas a value is
3056 * a list of equivalents of it. {@code '*'} cannot be used in the map.
3057 * Each equivalent language range has the same weight value as its
3058 * original language range.
3059 *
3060 * <pre>
3061 * An example of map:
3062 * <b>Key</b> <b>Value</b>
3063 * "zh" (Chinese) "zh",
3064 * "zh-Hans"(Simplified Chinese)
3065 * "zh-HK" (Chinese, Hong Kong) "zh-HK"
3066 * "zh-TW" (Chinese, Taiwan) "zh-TW"
3067 * </pre>
3068 *
3069 * The customization is performed after modification using the IANA
3070 * Language Subtag Registry.
3071 *
3072 * <p>For example, if a user's Language Priority List consists of five
3073 * language ranges ({@code "zh"}, {@code "zh-CN"}, {@code "en"},
3074 * {@code "zh-TW"}, and {@code "zh-HK"}), the newly generated Language
3075 * Priority List which is customized using the above map example will
3076 * consists of {@code "zh"}, {@code "zh-Hans"}, {@code "zh-CN"},
3077 * {@code "zh-Hans-CN"}, {@code "en"}, {@code "zh-TW"}, and
3078 * {@code "zh-HK"}.
3079 *
3080 * <p>{@code "zh-HK"} and {@code "zh-TW"} aren't converted to
3081 * {@code "zh-Hans-HK"} nor {@code "zh-Hans-TW"} even if they are
3082 * included in the Language Priority List. In this example, mapping
3083 * is used to clearly distinguish Simplified Chinese and Traditional
3084 * Chinese.
3085 *
3086 * <p>If the {@code "zh"}-to-{@code "zh"} mapping isn't included in the
3087 * map, a simple replacement will be performed and the customized list
3088 * won't include {@code "zh"} and {@code "zh-CN"}.
3089 *
3090 * @param priorityList user's Language Priority List
3091 * @param map a map containing information to customize language ranges
3092 * @return a new Language Priority List with customization. The list is
3093 * modifiable.
3094 * @throws NullPointerException if {@code priorityList} is {@code null}
3095 * @see #parse(String, Map)
3096 */
3097 public static List<LanguageRange> mapEquivalents(
3098 List<LanguageRange>priorityList,
3099 Map<String, List<String>> map) {
3100 return LocaleMatcher.mapEquivalents(priorityList, map);
3101 }
3102
3103 /**
3104 * Returns a hash code value for the object.
3105 *
3106 * @return a hash code value for this object.
3107 */
3108 @Override
3109 public int hashCode() {
3110 int h = hash;
3111 if (h == 0) {
3112 h = 17;
3113 h = 37*h + range.hashCode();
3114 long bitsWeight = Double.doubleToLongBits(weight);
3115 h = 37*h + (int)(bitsWeight ^ (bitsWeight >>> 32));
3116 if (h != 0) {
3117 hash = h;
3118 }
3119 }
3120 return h;
3121 }
3122
3123 /**
3124 * Compares this object to the specified object. The result is true if
3125 * and only if the argument is not {@code null} and is a
3126 * {@code LanguageRange} object that contains the same {@code range}
3127 * and {@code weight} values as this object.
3128 *
3129 * @param obj the object to compare with
3130 * @return {@code true} if this object's {@code range} and
3131 * {@code weight} are the same as the {@code obj}'s; {@code false}
3132 * otherwise.
3133 */
3134 @Override
3135 public boolean equals(Object obj) {
3136 if (this == obj) {
3137 return true;
3138 }
3139 if (!(obj instanceof LanguageRange)) {
3140 return false;
3141 }
3142 LanguageRange other = (LanguageRange)obj;
3143 return hash == other.hash
3144 && range.equals(other.range)
3145 && weight == other.weight;
3146 }
3147
3148 /**
3149 * Returns an informative string representation of this {@code LanguageRange}
3150 * object, consisting of language range and weight if the range is
3151 * weighted and the weight is less than the max weight.
3152 *
3153 * @return a string representation of this {@code LanguageRange} object.
3154 */
3155 @Override
3156 public String toString() {
3157 return (weight == MAX_WEIGHT) ? range : range + ";q=" + weight;
3158 }
3159 }
3160
3161 /**
3162 * Returns a list of matching {@code Locale} instances using the filtering
3163 * mechanism defined in RFC 4647.
3164 *
3165 * @param priorityList user's Language Priority List in which each language
3166 * tag is sorted in descending order based on priority or weight
3167 * @param locales {@code Locale} instances used for matching
3168 * @param mode filtering mode
3169 * @return a list of {@code Locale} instances for matching language tags
3170 * sorted in descending order based on priority or weight, or an empty
3171 * list if nothing matches. The list is modifiable.
3172 * @throws NullPointerException if {@code priorityList} or {@code locales}
3173 * is {@code null}
3174 * @throws IllegalArgumentException if one or more extended language ranges
3175 * are included in the given list when
3176 * {@link FilteringMode#REJECT_EXTENDED_RANGES} is specified
3177 *
3178 * @since 1.8
3179 */
3180 public static List<Locale> filter(List<LanguageRange> priorityList,
3181 Collection<Locale> locales,
3182 FilteringMode mode) {
3183 return LocaleMatcher.filter(priorityList, locales, mode);
3184 }
3185
3186 /**
3187 * Returns a list of matching {@code Locale} instances using the filtering
3188 * mechanism defined in RFC 4647. This is equivalent to
3189 * {@link #filter(List, Collection, FilteringMode)} when {@code mode} is
3190 * {@link FilteringMode#AUTOSELECT_FILTERING}.
3191 *
3192 * @param priorityList user's Language Priority List in which each language
3193 * tag is sorted in descending order based on priority or weight
3194 * @param locales {@code Locale} instances used for matching
3195 * @return a list of {@code Locale} instances for matching language tags
3196 * sorted in descending order based on priority or weight, or an empty
3197 * list if nothing matches. The list is modifiable.
3198 * @throws NullPointerException if {@code priorityList} or {@code locales}
3199 * is {@code null}
3200 *
3201 * @since 1.8
3202 */
3203 public static List<Locale> filter(List<LanguageRange> priorityList,
3204 Collection<Locale> locales) {
3205 return filter(priorityList, locales, FilteringMode.AUTOSELECT_FILTERING);
3206 }
3207
3208 /**
3209 * Returns a list of matching languages tags using the basic filtering
3210 * mechanism defined in RFC 4647.
3211 *
3212 * @param priorityList user's Language Priority List in which each language
3213 * tag is sorted in descending order based on priority or weight
3214 * @param tags language tags
3215 * @param mode filtering mode
3216 * @return a list of matching language tags sorted in descending order
3217 * based on priority or weight, or an empty list if nothing matches.
3218 * The list is modifiable.
3219 * @throws NullPointerException if {@code priorityList} or {@code tags} is
3220 * {@code null}
3221 * @throws IllegalArgumentException if one or more extended language ranges
3222 * are included in the given list when
3223 * {@link FilteringMode#REJECT_EXTENDED_RANGES} is specified
3224 *
3225 * @since 1.8
3226 */
3227 public static List<String> filterTags(List<LanguageRange> priorityList,
3228 Collection<String> tags,
3229 FilteringMode mode) {
3230 return LocaleMatcher.filterTags(priorityList, tags, mode);
3231 }
3232
3233 /**
3234 * Returns a list of matching languages tags using the basic filtering
3235 * mechanism defined in RFC 4647. This is equivalent to
3236 * {@link #filterTags(List, Collection, FilteringMode)} when {@code mode}
3237 * is {@link FilteringMode#AUTOSELECT_FILTERING}.
3238 *
3239 * @param priorityList user's Language Priority List in which each language
3240 * tag is sorted in descending order based on priority or weight
3241 * @param tags language tags
3242 * @return a list of matching language tags sorted in descending order
3243 * based on priority or weight, or an empty list if nothing matches.
3244 * The list is modifiable.
3245 * @throws NullPointerException if {@code priorityList} or {@code tags} is
3246 * {@code null}
3247 *
3248 * @since 1.8
3249 */
3250 public static List<String> filterTags(List<LanguageRange> priorityList,
3251 Collection<String> tags) {
3252 return filterTags(priorityList, tags, FilteringMode.AUTOSELECT_FILTERING);
3253 }
3254
3255 /**
3256 * Returns a {@code Locale} instance for the best-matching language
3257 * tag using the lookup mechanism defined in RFC 4647.
3258 *
3259 * @param priorityList user's Language Priority List in which each language
3260 * tag is sorted in descending order based on priority or weight
3261 * @param locales {@code Locale} instances used for matching
3262 * @return the best matching <code>Locale</code> instance chosen based on
3263 * priority or weight, or {@code null} if nothing matches.
3264 * @throws NullPointerException if {@code priorityList} or {@code tags} is
3265 * {@code null}
3266 *
3267 * @since 1.8
3268 */
3269 public static Locale lookup(List<LanguageRange> priorityList,
3270 Collection<Locale> locales) {
3271 return LocaleMatcher.lookup(priorityList, locales);
3272 }
3273
3274 /**
3275 * Returns the best-matching language tag using the lookup mechanism
3276 * defined in RFC 4647.
3277 *
3278 * @param priorityList user's Language Priority List in which each language
3279 * tag is sorted in descending order based on priority or weight
3280 * @param tags language tangs used for matching
3281 * @return the best matching language tag chosen based on priority or
3282 * weight, or {@code null} if nothing matches.
3283 * @throws NullPointerException if {@code priorityList} or {@code tags} is
3284 * {@code null}
3285 *
3286 * @since 1.8
3287 */
3288 public static String lookupTag(List<LanguageRange> priorityList,
3289 Collection<String> tags) {
3290 return LocaleMatcher.lookupTag(priorityList, tags);
3291 }
3292
3293 }