Locale.java revision 6d65625ab80fefa3a01b4daa89f598109647e295
1/* 2 * Copyright (C) 2014 The Android Open Source Project 3 * Copyright (c) 1996, 2014, Oracle and/or its affiliates. All rights reserved. 4 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. 5 * 6 * This code is free software; you can redistribute it and/or modify it 7 * under the terms of the GNU General Public License version 2 only, as 8 * published by the Free Software Foundation. Oracle designates this 9 * particular file as subject to the "Classpath" exception as provided 10 * by Oracle in the LICENSE file that accompanied this code. 11 * 12 * This code is distributed in the hope that it will be useful, but WITHOUT 13 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or 14 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License 15 * version 2 for more details (a copy is included in the LICENSE file that 16 * accompanied this code). 17 * 18 * You should have received a copy of the GNU General Public License version 19 * 2 along with this work; if not, write to the Free Software Foundation, 20 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. 21 * 22 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA 23 * or visit www.oracle.com if you need additional information or have any 24 * questions. 25 */ 26 27/* 28 * (C) Copyright Taligent, Inc. 1996, 1997 - All Rights Reserved 29 * (C) Copyright IBM Corp. 1996 - 1998 - All Rights Reserved 30 * 31 * The original version of this source code and documentation 32 * is copyrighted and owned by Taligent, Inc., a wholly-owned 33 * subsidiary of IBM. These materials are provided under terms 34 * of a License Agreement between Taligent and Sun. This technology 35 * is protected by multiple US and International patents. 36 * 37 * This notice and attribution to Taligent may not be removed. 38 * Taligent is a registered trademark of Taligent, Inc. 39 * 40 */ 41 42package java.util; 43 44import java.io.IOException; 45import java.io.ObjectInputStream; 46import java.io.ObjectOutputStream; 47import java.io.ObjectStreamField; 48import java.io.Serializable; 49import java.text.MessageFormat; 50import libcore.icu.ICU; 51 52import sun.util.locale.BaseLocale; 53import sun.util.locale.InternalLocaleBuilder; 54import sun.util.locale.LanguageTag; 55import sun.util.locale.LocaleExtensions; 56import sun.util.locale.LocaleMatcher; 57import sun.util.locale.LocaleObjectCache; 58import sun.util.locale.LocaleSyntaxException; 59import sun.util.locale.LocaleUtils; 60import sun.util.locale.ParseStatus; 61 62/** 63 * A <code>Locale</code> object represents a specific geographical, political, 64 * or cultural region. An operation that requires a <code>Locale</code> to perform 65 * its task is called <em>locale-sensitive</em> and uses the <code>Locale</code> 66 * to tailor information for the user. For example, displaying a number 67 * is a locale-sensitive operation— the number should be formatted 68 * according to the customs and conventions of the user's native country, 69 * region, or culture. 70 * 71 * <p> The {@code Locale} class implements IETF BCP 47 which is composed of 72 * <a href="http://tools.ietf.org/html/rfc4647">RFC 4647 "Matching of Language 73 * Tags"</a> and <a href="http://tools.ietf.org/html/rfc5646">RFC 5646 "Tags 74 * for Identifying Languages"</a> with support for the LDML (UTS#35, "Unicode 75 * Locale Data Markup Language") BCP 47-compatible extensions for locale data 76 * exchange. 77 * 78 * <p> A <code>Locale</code> object logically consists of the fields 79 * described below. 80 * 81 * <dl> 82 * <dt><a name="def_language"><b>language</b></a></dt> 83 * 84 * <dd>ISO 639 alpha-2 or alpha-3 language code, or registered 85 * language subtags up to 8 alpha letters (for future enhancements). 86 * When a language has both an alpha-2 code and an alpha-3 code, the 87 * alpha-2 code must be used. You can find a full list of valid 88 * language codes in the IANA Language Subtag Registry (search for 89 * "Type: language"). The language field is case insensitive, but 90 * <code>Locale</code> always canonicalizes to lower case.</dd> 91 * 92 * <dd>Well-formed language values have the form 93 * <code>[a-zA-Z]{2,8}</code>. Note that this is not the the full 94 * BCP47 language production, since it excludes extlang. They are 95 * not needed since modern three-letter language codes replace 96 * them.</dd> 97 * 98 * <dd>Example: "en" (English), "ja" (Japanese), "kok" (Konkani)</dd> 99 * 100 * <dt><a name="def_script"><b>script</b></a></dt> 101 * 102 * <dd>ISO 15924 alpha-4 script code. You can find a full list of 103 * valid script codes in the IANA Language Subtag Registry (search 104 * for "Type: script"). The script field is case insensitive, but 105 * <code>Locale</code> always canonicalizes to title case (the first 106 * letter is upper case and the rest of the letters are lower 107 * case).</dd> 108 * 109 * <dd>Well-formed script values have the form 110 * <code>[a-zA-Z]{4}</code></dd> 111 * 112 * <dd>Example: "Latn" (Latin), "Cyrl" (Cyrillic)</dd> 113 * 114 * <dt><a name="def_region"><b>country (region)</b></a></dt> 115 * 116 * <dd>ISO 3166 alpha-2 country code or UN M.49 numeric-3 area code. 117 * You can find a full list of valid country and region codes in the 118 * IANA Language Subtag Registry (search for "Type: region"). The 119 * country (region) field is case insensitive, but 120 * <code>Locale</code> always canonicalizes to upper case.</dd> 121 * 122 * <dd>Well-formed country/region values have 123 * the form <code>[a-zA-Z]{2} | [0-9]{3}</code></dd> 124 * 125 * <dd>Example: "US" (United States), "FR" (France), "029" 126 * (Caribbean)</dd> 127 * 128 * <dt><a name="def_variant"><b>variant</b></a></dt> 129 * 130 * <dd>Any arbitrary value used to indicate a variation of a 131 * <code>Locale</code>. Where there are two or more variant values 132 * each indicating its own semantics, these values should be ordered 133 * by importance, with most important first, separated by 134 * underscore('_'). The variant field is case sensitive.</dd> 135 * 136 * <dd>Note: IETF BCP 47 places syntactic restrictions on variant 137 * subtags. Also BCP 47 subtags are strictly used to indicate 138 * additional variations that define a language or its dialects that 139 * are not covered by any combinations of language, script and 140 * region subtags. You can find a full list of valid variant codes 141 * in the IANA Language Subtag Registry (search for "Type: variant"). 142 * 143 * <p>However, the variant field in <code>Locale</code> has 144 * historically been used for any kind of variation, not just 145 * language variations. For example, some supported variants 146 * available in Java SE Runtime Environments indicate alternative 147 * cultural behaviors such as calendar type or number script. In 148 * BCP 47 this kind of information, which does not identify the 149 * language, is supported by extension subtags or private use 150 * subtags.</dd> 151 * 152 * <dd>Well-formed variant values have the form <code>SUBTAG 153 * (('_'|'-') SUBTAG)*</code> where <code>SUBTAG = 154 * [0-9][0-9a-zA-Z]{3} | [0-9a-zA-Z]{5,8}</code>. (Note: BCP 47 only 155 * uses hyphen ('-') as a delimiter, this is more lenient).</dd> 156 * 157 * <dd>Example: "polyton" (Polytonic Greek), "POSIX"</dd> 158 * 159 * <dt><a name="def_extensions"><b>extensions</b></a></dt> 160 * 161 * <dd>A map from single character keys to string values, indicating 162 * extensions apart from language identification. The extensions in 163 * <code>Locale</code> implement the semantics and syntax of BCP 47 164 * extension subtags and private use subtags. The extensions are 165 * case insensitive, but <code>Locale</code> canonicalizes all 166 * extension keys and values to lower case. Note that extensions 167 * cannot have empty values.</dd> 168 * 169 * <dd>Well-formed keys are single characters from the set 170 * <code>[0-9a-zA-Z]</code>. Well-formed values have the form 171 * <code>SUBTAG ('-' SUBTAG)*</code> where for the key 'x' 172 * <code>SUBTAG = [0-9a-zA-Z]{1,8}</code> and for other keys 173 * <code>SUBTAG = [0-9a-zA-Z]{2,8}</code> (that is, 'x' allows 174 * single-character subtags).</dd> 175 * 176 * <dd>Example: key="u"/value="ca-japanese" (Japanese Calendar), 177 * key="x"/value="java-1-7"</dd> 178 * </dl> 179 * 180 * <b>Note:</b> Although BCP 47 requires field values to be registered 181 * in the IANA Language Subtag Registry, the <code>Locale</code> class 182 * does not provide any validation features. The <code>Builder</code> 183 * only checks if an individual field satisfies the syntactic 184 * requirement (is well-formed), but does not validate the value 185 * itself. See {@link Builder} for details. 186 * 187 * <h3><a name="def_locale_extension">Unicode locale/language extension</a></h3> 188 * 189 * <p>UTS#35, "Unicode Locale Data Markup Language" defines optional 190 * attributes and keywords to override or refine the default behavior 191 * associated with a locale. A keyword is represented by a pair of 192 * key and type. For example, "nu-thai" indicates that Thai local 193 * digits (value:"thai") should be used for formatting numbers 194 * (key:"nu"). 195 * 196 * <p>The keywords are mapped to a BCP 47 extension value using the 197 * extension key 'u' ({@link #UNICODE_LOCALE_EXTENSION}). The above 198 * example, "nu-thai", becomes the extension "u-nu-thai".code 199 * 200 * <p>Thus, when a <code>Locale</code> object contains Unicode locale 201 * attributes and keywords, 202 * <code>getExtension(UNICODE_LOCALE_EXTENSION)</code> will return a 203 * String representing this information, for example, "nu-thai". The 204 * <code>Locale</code> class also provides {@link 205 * #getUnicodeLocaleAttributes}, {@link #getUnicodeLocaleKeys}, and 206 * {@link #getUnicodeLocaleType} which allow you to access Unicode 207 * locale attributes and key/type pairs directly. When represented as 208 * a string, the Unicode Locale Extension lists attributes 209 * alphabetically, followed by key/type sequences with keys listed 210 * alphabetically (the order of subtags comprising a key's type is 211 * fixed when the type is defined) 212 * 213 * <p>A well-formed locale key has the form 214 * <code>[0-9a-zA-Z]{2}</code>. A well-formed locale type has the 215 * form <code>"" | [0-9a-zA-Z]{3,8} ('-' [0-9a-zA-Z]{3,8})*</code> (it 216 * can be empty, or a series of subtags 3-8 alphanums in length). A 217 * well-formed locale attribute has the form 218 * <code>[0-9a-zA-Z]{3,8}</code> (it is a single subtag with the same 219 * form as a locale type subtag). 220 * 221 * <p>The Unicode locale extension specifies optional behavior in 222 * locale-sensitive services. Although the LDML specification defines 223 * various keys and values, actual locale-sensitive service 224 * implementations in a Java Runtime Environment might not support any 225 * particular Unicode locale attributes or key/type pairs. 226 * 227 * <h4>Creating a Locale</h4> 228 * 229 * <p>There are several different ways to create a <code>Locale</code> 230 * object. 231 * 232 * <h5>Builder</h5> 233 * 234 * <p>Using {@link Builder} you can construct a <code>Locale</code> object 235 * that conforms to BCP 47 syntax. 236 * 237 * <h5>Constructors</h5> 238 * 239 * <p>The <code>Locale</code> class provides three constructors: 240 * <blockquote> 241 * <pre> 242 * {@link #Locale(String language)} 243 * {@link #Locale(String language, String country)} 244 * {@link #Locale(String language, String country, String variant)} 245 * </pre> 246 * </blockquote> 247 * These constructors allow you to create a <code>Locale</code> object 248 * with language, country and variant, but you cannot specify 249 * script or extensions. 250 * 251 * <h5>Factory Methods</h5> 252 * 253 * <p>The method {@link #forLanguageTag} creates a <code>Locale</code> 254 * object for a well-formed BCP 47 language tag. 255 * 256 * <h5>Locale Constants</h5> 257 * 258 * <p>The <code>Locale</code> class provides a number of convenient constants 259 * that you can use to create <code>Locale</code> objects for commonly used 260 * locales. For example, the following creates a <code>Locale</code> object 261 * for the United States: 262 * <blockquote> 263 * <pre> 264 * Locale.US 265 * </pre> 266 * </blockquote> 267 * 268 * <h4><a name="LocaleMatching">Locale Matching</a></h4> 269 * 270 * <p>If an application or a system is internationalized and provides localized 271 * resources for multiple locales, it sometimes needs to find one or more 272 * locales (or language tags) which meet each user's specific preferences. Note 273 * that a term "language tag" is used interchangeably with "locale" in this 274 * locale matching documentation. 275 * 276 * <p>In order to do matching a user's preferred locales to a set of language 277 * tags, <a href="http://tools.ietf.org/html/rfc4647">RFC 4647 Matching of 278 * Language Tags</a> defines two mechanisms: filtering and lookup. 279 * <em>Filtering</em> is used to get all matching locales, whereas 280 * <em>lookup</em> is to choose the best matching locale. 281 * Matching is done case-insensitively. These matching mechanisms are described 282 * in the following sections. 283 * 284 * <p>A user's preference is called a <em>Language Priority List</em> and is 285 * expressed as a list of language ranges. There are syntactically two types of 286 * language ranges: basic and extended. See 287 * {@link Locale.LanguageRange Locale.LanguageRange} for details. 288 * 289 * <h5>Filtering</h5> 290 * 291 * <p>The filtering operation returns all matching language tags. It is defined 292 * in RFC 4647 as follows: 293 * "In filtering, each language range represents the least specific language 294 * tag (that is, the language tag with fewest number of subtags) that is an 295 * acceptable match. All of the language tags in the matching set of tags will 296 * have an equal or greater number of subtags than the language range. Every 297 * non-wildcard subtag in the language range will appear in every one of the 298 * matching language tags." 299 * 300 * <p>There are two types of filtering: filtering for basic language ranges 301 * (called "basic filtering") and filtering for extended language ranges 302 * (called "extended filtering"). They may return different results by what 303 * kind of language ranges are included in the given Language Priority List. 304 * {@link Locale.FilteringMode} is a parameter to specify how filtering should 305 * be done. 306 * 307 * <h5>Lookup</h5> 308 * 309 * <p>The lookup operation returns the best matching language tags. It is 310 * defined in RFC 4647 as follows: 311 * "By contrast with filtering, each language range represents the most 312 * specific tag that is an acceptable match. The first matching tag found, 313 * according to the user's priority, is considered the closest match and is the 314 * item returned." 315 * 316 * <p>For example, if a Language Priority List consists of two language ranges, 317 * {@code "zh-Hant-TW"} and {@code "en-US"}, in prioritized order, lookup 318 * method progressively searches the language tags below in order to find the 319 * best matching language tag. 320 * <blockquote> 321 * <pre> 322 * 1. zh-Hant-TW 323 * 2. zh-Hant 324 * 3. zh 325 * 4. en-US 326 * 5. en 327 * </pre> 328 * </blockquote> 329 * If there is a language tag which matches completely to a language range 330 * above, the language tag is returned. 331 * 332 * <p>{@code "*"} is the special language range, and it is ignored in lookup. 333 * 334 * <p>If multiple language tags match as a result of the subtag {@code '*'} 335 * included in a language range, the first matching language tag returned by 336 * an {@link Iterator} over a {@link Collection} of language tags is treated as 337 * the best matching one. 338 * 339 * <h4>Use of Locale</h4> 340 * 341 * <p>Once you've created a <code>Locale</code> you can query it for information 342 * about itself. Use <code>getCountry</code> to get the country (or region) 343 * code and <code>getLanguage</code> to get the language code. 344 * You can use <code>getDisplayCountry</code> to get the 345 * name of the country suitable for displaying to the user. Similarly, 346 * you can use <code>getDisplayLanguage</code> to get the name of 347 * the language suitable for displaying to the user. Interestingly, 348 * the <code>getDisplayXXX</code> methods are themselves locale-sensitive 349 * and have two versions: one that uses the default 350 * {@link Locale.Category#DISPLAY DISPLAY} locale and one 351 * that uses the locale specified as an argument. 352 * 353 * <p>The Java Platform provides a number of classes that perform locale-sensitive 354 * operations. For example, the <code>NumberFormat</code> class formats 355 * numbers, currency, and percentages in a locale-sensitive manner. Classes 356 * such as <code>NumberFormat</code> have several convenience methods 357 * for creating a default object of that type. For example, the 358 * <code>NumberFormat</code> class provides these three convenience methods 359 * for creating a default <code>NumberFormat</code> object: 360 * <blockquote> 361 * <pre> 362 * NumberFormat.getInstance() 363 * NumberFormat.getCurrencyInstance() 364 * NumberFormat.getPercentInstance() 365 * </pre> 366 * </blockquote> 367 * Each of these methods has two variants; one with an explicit locale 368 * and one without; the latter uses the default 369 * {@link Locale.Category#FORMAT FORMAT} locale: 370 * <blockquote> 371 * <pre> 372 * NumberFormat.getInstance(myLocale) 373 * NumberFormat.getCurrencyInstance(myLocale) 374 * NumberFormat.getPercentInstance(myLocale) 375 * </pre> 376 * </blockquote> 377 * A <code>Locale</code> is the mechanism for identifying the kind of object 378 * (<code>NumberFormat</code>) that you would like to get. The locale is 379 * <STRONG>just</STRONG> a mechanism for identifying objects, 380 * <STRONG>not</STRONG> a container for the objects themselves. 381 * 382 * <h4>Compatibility</h4> 383 * 384 * <p>In order to maintain compatibility with existing usage, Locale's 385 * constructors retain their behavior prior to the Java Runtime 386 * Environment version 1.7. The same is largely true for the 387 * <code>toString</code> method. Thus Locale objects can continue to 388 * be used as they were. In particular, clients who parse the output 389 * of toString into language, country, and variant fields can continue 390 * to do so (although this is strongly discouraged), although the 391 * variant field will have additional information in it if script or 392 * extensions are present. 393 * 394 * <p>In addition, BCP 47 imposes syntax restrictions that are not 395 * imposed by Locale's constructors. This means that conversions 396 * between some Locales and BCP 47 language tags cannot be made without 397 * losing information. Thus <code>toLanguageTag</code> cannot 398 * represent the state of locales whose language, country, or variant 399 * do not conform to BCP 47. 400 * 401 * <p>Because of these issues, it is recommended that clients migrate 402 * away from constructing non-conforming locales and use the 403 * <code>forLanguageTag</code> and <code>Locale.Builder</code> APIs instead. 404 * Clients desiring a string representation of the complete locale can 405 * then always rely on <code>toLanguageTag</code> for this purpose. 406 * 407 * <h5><a name="special_cases_constructor">Special cases</a></h5> 408 * 409 * <p>For compatibility reasons, two 410 * non-conforming locales are treated as special cases. These are 411 * <b><tt>ja_JP_JP</tt></b> and <b><tt>th_TH_TH</tt></b>. These are ill-formed 412 * in BCP 47 since the variants are too short. To ease migration to BCP 47, 413 * these are treated specially during construction. These two cases (and only 414 * these) cause a constructor to generate an extension, all other values behave 415 * exactly as they did prior to Java 7. 416 * 417 * <p>Java has used <tt>ja_JP_JP</tt> to represent Japanese as used in 418 * Japan together with the Japanese Imperial calendar. This is now 419 * representable using a Unicode locale extension, by specifying the 420 * Unicode locale key <tt>ca</tt> (for "calendar") and type 421 * <tt>japanese</tt>. When the Locale constructor is called with the 422 * arguments "ja", "JP", "JP", the extension "u-ca-japanese" is 423 * automatically added. 424 * 425 * <p>Java has used <tt>th_TH_TH</tt> to represent Thai as used in 426 * Thailand together with Thai digits. This is also now representable using 427 * a Unicode locale extension, by specifying the Unicode locale key 428 * <tt>nu</tt> (for "number") and value <tt>thai</tt>. When the Locale 429 * constructor is called with the arguments "th", "TH", "TH", the 430 * extension "u-nu-thai" is automatically added. 431 * 432 * <h5>Serialization</h5> 433 * 434 * <p>During serialization, writeObject writes all fields to the output 435 * stream, including extensions. 436 * 437 * <p>During deserialization, readResolve adds extensions as described 438 * in <a href="#special_cases_constructor">Special Cases</a>, only 439 * for the two cases th_TH_TH and ja_JP_JP. 440 * 441 * <h5>Legacy language codes</h5> 442 * 443 * <p>Locale's constructor has always converted three language codes to 444 * their earlier, obsoleted forms: <tt>he</tt> maps to <tt>iw</tt>, 445 * <tt>yi</tt> maps to <tt>ji</tt>, and <tt>id</tt> maps to 446 * <tt>in</tt>. This continues to be the case, in order to not break 447 * backwards compatibility. 448 * 449 * <p>The APIs added in 1.7 map between the old and new language codes, 450 * maintaining the old codes internal to Locale (so that 451 * <code>getLanguage</code> and <code>toString</code> reflect the old 452 * code), but using the new codes in the BCP 47 language tag APIs (so 453 * that <code>toLanguageTag</code> reflects the new one). This 454 * preserves the equivalence between Locales no matter which code or 455 * API is used to construct them. Java's default resource bundle 456 * lookup mechanism also implements this mapping, so that resources 457 * can be named using either convention, see {@link ResourceBundle.Control}. 458 * 459 * <h5>Three-letter language/country(region) codes</h5> 460 * 461 * <p>The Locale constructors have always specified that the language 462 * and the country param be two characters in length, although in 463 * practice they have accepted any length. The specification has now 464 * been relaxed to allow language codes of two to eight characters and 465 * country (region) codes of two to three characters, and in 466 * particular, three-letter language codes and three-digit region 467 * codes as specified in the IANA Language Subtag Registry. For 468 * compatibility, the implementation still does not impose a length 469 * constraint. 470 * 471 * <a name="locale_data"></a><h4>Locale data</h4> 472 * <p>Note that locale data comes solely from ICU. User-supplied locale service providers (using 473 * the {@code java.text.spi} or {@code java.util.spi} mechanisms) are not supported. 474 * 475 * <p>Here are the versions of ICU (and the corresponding CLDR and Unicode versions) used in 476 * various Android releases: 477 * <table BORDER="1" WIDTH="100%" CELLPADDING="3" CELLSPACING="0" SUMMARY=""> 478 * <tr><td>Android 1.5 (Cupcake)/Android 1.6 (Donut)/Android 2.0 (Eclair)</td> 479 * <td>ICU 3.8</td> 480 * <td><a href="http://cldr.unicode.org/index/downloads/cldr-1-5">CLDR 1.5</a></td> 481 * <td><a href="http://www.unicode.org/versions/Unicode5.0.0/">Unicode 5.0</a></td></tr> 482 * <tr><td>Android 2.2 (Froyo)</td> 483 * <td>ICU 4.2</td> 484 * <td><a href="http://cldr.unicode.org/index/downloads/cldr-1-7">CLDR 1.7</a></td> 485 * <td><a href="http://www.unicode.org/versions/Unicode5.1.0/">Unicode 5.1</a></td></tr> 486 * <tr><td>Android 2.3 (Gingerbread)/Android 3.0 (Honeycomb)</td> 487 * <td>ICU 4.4</td> 488 * <td><a href="http://cldr.unicode.org/index/downloads/cldr-1-8">CLDR 1.8</a></td> 489 * <td><a href="http://www.unicode.org/versions/Unicode5.2.0/">Unicode 5.2</a></td></tr> 490 * <tr><td>Android 4.0 (Ice Cream Sandwich)</td> 491 * <td><a href="http://site.icu-project.org/download/46">ICU 4.6</a></td> 492 * <td><a href="http://cldr.unicode.org/index/downloads/cldr-1-9">CLDR 1.9</a></td> 493 * <td><a href="http://www.unicode.org/versions/Unicode6.0.0/">Unicode 6.0</a></td></tr> 494 * <tr><td>Android 4.1 (Jelly Bean)</td> 495 * <td><a href="http://site.icu-project.org/download/48">ICU 4.8</a></td> 496 * <td><a href="http://cldr.unicode.org/index/downloads/cldr-2-0">CLDR 2.0</a></td> 497 * <td><a href="http://www.unicode.org/versions/Unicode6.0.0/">Unicode 6.0</a></td></tr> 498 * <tr><td>Android 4.3 (Jelly Bean MR2)</td> 499 * <td><a href="http://site.icu-project.org/download/50">ICU 50</a></td> 500 * <td><a href="http://cldr.unicode.org/index/downloads/cldr-22-1">CLDR 22.1</a></td> 501 * <td><a href="http://www.unicode.org/versions/Unicode6.2.0/">Unicode 6.2</a></td></tr> 502 * <tr><td>Android 4.4 (KitKat)</td> 503 * <td><a href="http://site.icu-project.org/download/51">ICU 51</a></td> 504 * <td><a href="http://cldr.unicode.org/index/downloads/cldr-23">CLDR 23</a></td> 505 * <td><a href="http://www.unicode.org/versions/Unicode6.2.0/">Unicode 6.2</a></td></tr> 506 * <tr><td>Android 5.0 (Lollipop)</td> 507 * <td><a href="http://site.icu-project.org/download/53">ICU 53</a></td> 508 * <td><a href="http://cldr.unicode.org/index/downloads/cldr-25">CLDR 25</a></td> 509 * <td><a href="http://www.unicode.org/versions/Unicode6.3.0/">Unicode 6.3</a></td></tr> 510 * <tr><td>Android 6.0 (Marshmallow)</td> 511 * <td><a href="http://site.icu-project.org/download/55">ICU 55.1</a></td> 512 * <td><a href="http://cldr.unicode.org/index/downloads/cldr-27">CLDR 27.0.1</a></td> 513 * <td><a href="http://www.unicode.org/versions/Unicode7.0.0/">Unicode 7.0</a></td></tr> 514 * </table> 515 * 516 * <a name="default_locale"></a><h4>Be wary of the default locale</h3> 517 * <p>Note that there are many convenience methods that automatically use the default locale, but 518 * using them may lead to subtle bugs. 519 * 520 * <p>The default locale is appropriate for tasks that involve presenting data to the user. In 521 * this case, you want to use the user's date/time formats, number 522 * formats, rules for conversion to lowercase, and so on. In this case, it's safe to use the 523 * convenience methods. 524 * 525 * <p>The default locale is <i>not</i> appropriate for machine-readable output. The best choice 526 * there is usually {@code Locale.US} – this locale is guaranteed to be available on all 527 * devices, and the fact that it has no surprising special cases and is frequently used (especially 528 * for computer-computer communication) means that it tends to be the most efficient choice too. 529 * 530 * <p>A common mistake is to implicitly use the default locale when producing output meant to be 531 * machine-readable. This tends to work on the developer's test devices (especially because so many 532 * developers use en_US), but fails when run on a device whose user is in a more complex locale. 533 * 534 * <p>For example, if you're formatting integers some locales will use non-ASCII decimal 535 * digits. As another example, if you're formatting floating-point numbers some locales will use 536 * {@code ','} as the decimal point and {@code '.'} for digit grouping. That's correct for 537 * human-readable output, but likely to cause problems if presented to another 538 * computer ({@link Double#parseDouble} can't parse such a number, for example). 539 * You should also be wary of the {@link String#toLowerCase} and 540 * {@link String#toUpperCase} overloads that don't take a {@code Locale}: in Turkey, for example, 541 * the characters {@code 'i'} and {@code 'I'} won't be converted to {@code 'I'} and {@code 'i'}. 542 * This is the correct behavior for Turkish text (such as user input), but inappropriate for, say, 543 * HTTP headers. 544 * 545 * @see Builder 546 * @see ResourceBundle 547 * @see java.text.Format 548 * @see java.text.NumberFormat 549 * @see java.text.Collator 550 * @author Mark Davis 551 * @since 1.1 552 */ 553public final class Locale implements Cloneable, Serializable { 554 555 static private final Cache LOCALECACHE = new Cache(); 556 557 /** Useful constant for language. 558 */ 559 static public final Locale ENGLISH = createConstant("en", ""); 560 561 /** Useful constant for language. 562 */ 563 static public final Locale FRENCH = createConstant("fr", ""); 564 565 /** Useful constant for language. 566 */ 567 static public final Locale GERMAN = createConstant("de", ""); 568 569 /** Useful constant for language. 570 */ 571 static public final Locale ITALIAN = createConstant("it", ""); 572 573 /** Useful constant for language. 574 */ 575 static public final Locale JAPANESE = createConstant("ja", ""); 576 577 /** Useful constant for language. 578 */ 579 static public final Locale KOREAN = createConstant("ko", ""); 580 581 /** Useful constant for language. 582 */ 583 static public final Locale CHINESE = createConstant("zh", ""); 584 585 /** Useful constant for language. 586 */ 587 static public final Locale SIMPLIFIED_CHINESE = createConstant("zh", "CN"); 588 589 /** Useful constant for language. 590 */ 591 static public final Locale TRADITIONAL_CHINESE = createConstant("zh", "TW"); 592 593 /** Useful constant for country. 594 */ 595 static public final Locale FRANCE = createConstant("fr", "FR"); 596 597 /** Useful constant for country. 598 */ 599 static public final Locale GERMANY = createConstant("de", "DE"); 600 601 /** Useful constant for country. 602 */ 603 static public final Locale ITALY = createConstant("it", "IT"); 604 605 /** Useful constant for country. 606 */ 607 static public final Locale JAPAN = createConstant("ja", "JP"); 608 609 /** Useful constant for country. 610 */ 611 static public final Locale KOREA = createConstant("ko", "KR"); 612 613 /** Useful constant for country. 614 */ 615 static public final Locale CHINA = SIMPLIFIED_CHINESE; 616 617 /** Useful constant for country. 618 */ 619 static public final Locale PRC = SIMPLIFIED_CHINESE; 620 621 /** Useful constant for country. 622 */ 623 static public final Locale TAIWAN = TRADITIONAL_CHINESE; 624 625 /** Useful constant for country. 626 */ 627 static public final Locale UK = createConstant("en", "GB"); 628 629 /** Useful constant for country. 630 */ 631 static public final Locale US = createConstant("en", "US"); 632 633 /** Useful constant for country. 634 */ 635 static public final Locale CANADA = createConstant("en", "CA"); 636 637 /** Useful constant for country. 638 */ 639 static public final Locale CANADA_FRENCH = createConstant("fr", "CA"); 640 641 /** 642 * ISO 639-3 generic code for undetermined languages. 643 */ 644 private static final String UNDETERMINED_LANGUAGE = "und"; 645 646 /** 647 * Useful constant for the root locale. The root locale is the locale whose 648 * language, country, and variant are empty ("") strings. This is regarded 649 * as the base locale of all locales, and is used as the language/country 650 * neutral locale for the locale sensitive operations. 651 * 652 * @since 1.6 653 */ 654 static public final Locale ROOT = createConstant("", ""); 655 656 /** 657 * The key for the private use extension ('x'). 658 * 659 * @see #getExtension(char) 660 * @see Builder#setExtension(char, String) 661 * @since 1.7 662 */ 663 static public final char PRIVATE_USE_EXTENSION = 'x'; 664 665 /** 666 * The key for Unicode locale extension ('u'). 667 * 668 * @see #getExtension(char) 669 * @see Builder#setExtension(char, String) 670 * @since 1.7 671 */ 672 static public final char UNICODE_LOCALE_EXTENSION = 'u'; 673 674 /** serialization ID 675 */ 676 static final long serialVersionUID = 9149081749638150636L; 677 678 /** 679 * Display types for retrieving localized names from the name providers. 680 */ 681 private static final int DISPLAY_LANGUAGE = 0; 682 private static final int DISPLAY_COUNTRY = 1; 683 private static final int DISPLAY_VARIANT = 2; 684 private static final int DISPLAY_SCRIPT = 3; 685 686 /** 687 * Private constructor used by getInstance method 688 */ 689 private Locale(BaseLocale baseLocale, LocaleExtensions extensions) { 690 this.baseLocale = baseLocale; 691 this.localeExtensions = extensions; 692 } 693 694 /** 695 * Construct a locale from language, country and variant. 696 * This constructor normalizes the language value to lowercase and 697 * the country value to uppercase. 698 * <p> 699 * <b>Note:</b> 700 * <ul> 701 * <li>ISO 639 is not a stable standard; some of the language codes it defines 702 * (specifically "iw", "ji", and "in") have changed. This constructor accepts both the 703 * old codes ("iw", "ji", and "in") and the new codes ("he", "yi", and "id"), but all other 704 * API on Locale will return only the OLD codes. 705 * <li>For backward compatibility reasons, this constructor does not make 706 * any syntactic checks on the input. 707 * <li>The two cases ("ja", "JP", "JP") and ("th", "TH", "TH") are handled specially, 708 * see <a href="#special_cases_constructor">Special Cases</a> for more information. 709 * </ul> 710 * 711 * @param language An ISO 639 alpha-2 or alpha-3 language code, or a language subtag 712 * up to 8 characters in length. See the <code>Locale</code> class description about 713 * valid language values. 714 * @param country An ISO 3166 alpha-2 country code or a UN M.49 numeric-3 area code. 715 * See the <code>Locale</code> class description about valid country values. 716 * @param variant Any arbitrary value used to indicate a variation of a <code>Locale</code>. 717 * See the <code>Locale</code> class description for the details. 718 * @exception NullPointerException thrown if any argument is null. 719 */ 720 public Locale(String language, String country, String variant) { 721 if (language== null || country == null || variant == null) { 722 throw new NullPointerException(); 723 } 724 baseLocale = BaseLocale.getInstance(convertOldISOCodes(language), "", country, variant); 725 localeExtensions = getCompatibilityExtensions(language, "", country, variant); 726 } 727 728 /** 729 * Construct a locale from language and country. 730 * This constructor normalizes the language value to lowercase and 731 * the country value to uppercase. 732 * <p> 733 * <b>Note:</b> 734 * <ul> 735 * <li>ISO 639 is not a stable standard; some of the language codes it defines 736 * (specifically "iw", "ji", and "in") have changed. This constructor accepts both the 737 * old codes ("iw", "ji", and "in") and the new codes ("he", "yi", and "id"), but all other 738 * API on Locale will return only the OLD codes. 739 * <li>For backward compatibility reasons, this constructor does not make 740 * any syntactic checks on the input. 741 * </ul> 742 * 743 * @param language An ISO 639 alpha-2 or alpha-3 language code, or a language subtag 744 * up to 8 characters in length. See the <code>Locale</code> class description about 745 * valid language values. 746 * @param country An ISO 3166 alpha-2 country code or a UN M.49 numeric-3 area code. 747 * See the <code>Locale</code> class description about valid country values. 748 * @exception NullPointerException thrown if either argument is null. 749 */ 750 public Locale(String language, String country) { 751 this(language, country, ""); 752 } 753 754 /** 755 * Construct a locale from a language code. 756 * This constructor normalizes the language value to lowercase. 757 * <p> 758 * <b>Note:</b> 759 * <ul> 760 * <li>ISO 639 is not a stable standard; some of the language codes it defines 761 * (specifically "iw", "ji", and "in") have changed. This constructor accepts both the 762 * old codes ("iw", "ji", and "in") and the new codes ("he", "yi", and "id"), but all other 763 * API on Locale will return only the OLD codes. 764 * <li>For backward compatibility reasons, this constructor does not make 765 * any syntactic checks on the input. 766 * </ul> 767 * 768 * @param language An ISO 639 alpha-2 or alpha-3 language code, or a language subtag 769 * up to 8 characters in length. See the <code>Locale</code> class description about 770 * valid language values. 771 * @exception NullPointerException thrown if argument is null. 772 * @since 1.4 773 */ 774 public Locale(String language) { 775 this(language, "", ""); 776 } 777 778 /** 779 * This method must be called only for creating the Locale.* 780 * constants due to making shortcuts. 781 */ 782 private static Locale createConstant(String lang, String country) { 783 BaseLocale base = BaseLocale.createInstance(lang, country); 784 return getInstance(base, null); 785 } 786 787 /** 788 * Returns a <code>Locale</code> constructed from the given 789 * <code>language</code>, <code>country</code> and 790 * <code>variant</code>. If the same <code>Locale</code> instance 791 * is available in the cache, then that instance is 792 * returned. Otherwise, a new <code>Locale</code> instance is 793 * created and cached. 794 * 795 * @param language lowercase 2 to 8 language code. 796 * @param country uppercase two-letter ISO-3166 code and numric-3 UN M.49 area code. 797 * @param variant vendor and browser specific code. See class description. 798 * @return the <code>Locale</code> instance requested 799 * @exception NullPointerException if any argument is null. 800 */ 801 static Locale getInstance(String language, String country, String variant) { 802 return getInstance(language, "", country, variant, null); 803 } 804 805 static Locale getInstance(String language, String script, String country, 806 String variant, LocaleExtensions extensions) { 807 if (language== null || script == null || country == null || variant == null) { 808 throw new NullPointerException(); 809 } 810 811 if (extensions == null) { 812 extensions = getCompatibilityExtensions(language, script, country, variant); 813 } 814 815 BaseLocale baseloc = BaseLocale.getInstance(language, script, country, variant); 816 return getInstance(baseloc, extensions); 817 } 818 819 static Locale getInstance(BaseLocale baseloc, LocaleExtensions extensions) { 820 LocaleKey key = new LocaleKey(baseloc, extensions); 821 return LOCALECACHE.get(key); 822 } 823 824 private static class Cache extends LocaleObjectCache<LocaleKey, Locale> { 825 private Cache() { 826 } 827 828 @Override 829 protected Locale createObject(LocaleKey key) { 830 return new Locale(key.base, key.exts); 831 } 832 } 833 834 private static final class LocaleKey { 835 private final BaseLocale base; 836 private final LocaleExtensions exts; 837 private final int hash; 838 839 private LocaleKey(BaseLocale baseLocale, LocaleExtensions extensions) { 840 base = baseLocale; 841 exts = extensions; 842 843 // Calculate the hash value here because it's always used. 844 int h = base.hashCode(); 845 if (exts != null) { 846 h ^= exts.hashCode(); 847 } 848 hash = h; 849 } 850 851 @Override 852 public boolean equals(Object obj) { 853 if (this == obj) { 854 return true; 855 } 856 if (!(obj instanceof LocaleKey)) { 857 return false; 858 } 859 LocaleKey other = (LocaleKey)obj; 860 if (hash != other.hash || !base.equals(other.base)) { 861 return false; 862 } 863 if (exts == null) { 864 return other.exts == null; 865 } 866 return exts.equals(other.exts); 867 } 868 869 @Override 870 public int hashCode() { 871 return hash; 872 } 873 } 874 875 /** 876 * Gets the current value of the default locale for this instance 877 * of the Java Virtual Machine. 878 * <p> 879 * The Java Virtual Machine sets the default locale during startup 880 * based on the host environment. It is used by many locale-sensitive 881 * methods if no locale is explicitly specified. 882 * It can be changed using the 883 * {@link #setDefault(java.util.Locale) setDefault} method. 884 * 885 * @return the default locale for this instance of the Java Virtual Machine 886 */ 887 public static Locale getDefault() { 888 // do not synchronize this method - see 4071298 889 // Android-changed: Add NoImagePreloadHolder to allow compile-time initialization. 890 return NoImagePreloadHolder.defaultLocale; 891 } 892 893 /** 894 * Gets the current value of the default locale for the specified Category 895 * for this instance of the Java Virtual Machine. 896 * <p> 897 * The Java Virtual Machine sets the default locale during startup based 898 * on the host environment. It is used by many locale-sensitive methods 899 * if no locale is explicitly specified. It can be changed using the 900 * setDefault(Locale.Category, Locale) method. 901 * 902 * @param category - the specified category to get the default locale 903 * @throws NullPointerException - if category is null 904 * @return the default locale for the specified Category for this instance 905 * of the Java Virtual Machine 906 * @see #setDefault(Locale.Category, Locale) 907 * @since 1.7 908 */ 909 public static Locale getDefault(Locale.Category category) { 910 // do not synchronize this method - see 4071298 911 switch (category) { 912 case DISPLAY: 913 if (defaultDisplayLocale == null) { 914 synchronized(Locale.class) { 915 if (defaultDisplayLocale == null) { 916 defaultDisplayLocale = initDefault(category); 917 } 918 } 919 } 920 return defaultDisplayLocale; 921 case FORMAT: 922 if (defaultFormatLocale == null) { 923 synchronized(Locale.class) { 924 if (defaultFormatLocale == null) { 925 defaultFormatLocale = initDefault(category); 926 } 927 } 928 } 929 return defaultFormatLocale; 930 default: 931 assert false: "Unknown Category"; 932 } 933 return getDefault(); 934 } 935 936 // Android-changed BEGIN: 937 // 1.) In initDefault(), user.locale gets priority 938 // 2.) In both initDefault methods, use System.getProperty() instead 939 // of legacy AccessController / GetPropertyAction security code. 940 /** 941 * @hide visible for testing. 942 */ 943 public static Locale initDefault() { 944 // user.locale gets priority 945 final String languageTag = System.getProperty("user.locale", ""); 946 if (!languageTag.isEmpty()) { 947 return Locale.forLanguageTag(languageTag); 948 } 949 950 // user.locale is empty 951 String language, region, script, country, variant; 952 language = System.getProperty("user.language", "en"); 953 // for compatibility, check for old user.region property 954 region = System.getProperty("user.region"); 955 if (region != null) { 956 // region can be of form country, country_variant, or _variant 957 int i = region.indexOf('_'); 958 if (i >= 0) { 959 country = region.substring(0, i); 960 variant = region.substring(i + 1); 961 } else { 962 country = region; 963 variant = ""; 964 } 965 script = ""; 966 } else { 967 script = System.getProperty("user.script", ""); 968 country = System.getProperty("user.country", ""); 969 variant = System.getProperty("user.variant", ""); 970 } 971 972 return getInstance(language, script, country, variant, null); 973 } 974 975 private static Locale initDefault(Locale.Category category) { 976 // Android-changed: Add NoImagePreloadHolder to allow compile-time initialization. 977 final Locale defaultLocale = NoImagePreloadHolder.defaultLocale; 978 return getInstance( 979 System.getProperty(category.languageKey, defaultLocale.getLanguage()), 980 System.getProperty(category.scriptKey, defaultLocale.getScript()), 981 System.getProperty(category.countryKey, defaultLocale.getCountry()), 982 System.getProperty(category.variantKey, defaultLocale.getVariant()), 983 null); 984 } 985 // Android-changed END 986 987 /** 988 * Sets the default locale for this instance of the Java Virtual Machine. 989 * This does not affect the host locale. 990 * <p> 991 * If there is a security manager, its <code>checkPermission</code> 992 * method is called with a <code>PropertyPermission("user.language", "write")</code> 993 * permission before the default locale is changed. 994 * <p> 995 * The Java Virtual Machine sets the default locale during startup 996 * based on the host environment. It is used by many locale-sensitive 997 * methods if no locale is explicitly specified. 998 * <p> 999 * Since changing the default locale may affect many different areas 1000 * of functionality, this method should only be used if the caller 1001 * is prepared to reinitialize locale-sensitive code running 1002 * within the same Java Virtual Machine. 1003 * <p> 1004 * By setting the default locale with this method, all of the default 1005 * locales for each Category are also set to the specified default locale. 1006 * 1007 * @throws SecurityException 1008 * if a security manager exists and its 1009 * <code>checkPermission</code> method doesn't allow the operation. 1010 * @throws NullPointerException if <code>newLocale</code> is null 1011 * @param newLocale the new default locale 1012 * @see SecurityManager#checkPermission 1013 * @see java.util.PropertyPermission 1014 */ 1015 public static synchronized void setDefault(Locale newLocale) { 1016 setDefault(Category.DISPLAY, newLocale); 1017 setDefault(Category.FORMAT, newLocale); 1018 // Android-changed: Add NoImagePreloadHolder to allow compile-time initialization. 1019 NoImagePreloadHolder.defaultLocale = newLocale; 1020 // Android-added: Keep ICU state in sync with java.util. 1021 ICU.setDefaultLocale(newLocale.toLanguageTag()); 1022 } 1023 1024 /** 1025 * Sets the default locale for the specified Category for this instance 1026 * of the Java Virtual Machine. This does not affect the host locale. 1027 * <p> 1028 * If there is a security manager, its checkPermission method is called 1029 * with a PropertyPermission("user.language", "write") permission before 1030 * the default locale is changed. 1031 * <p> 1032 * The Java Virtual Machine sets the default locale during startup based 1033 * on the host environment. It is used by many locale-sensitive methods 1034 * if no locale is explicitly specified. 1035 * <p> 1036 * Since changing the default locale may affect many different areas of 1037 * functionality, this method should only be used if the caller is 1038 * prepared to reinitialize locale-sensitive code running within the 1039 * same Java Virtual Machine. 1040 * <p> 1041 * 1042 * @param category - the specified category to set the default locale 1043 * @param newLocale - the new default locale 1044 * @throws SecurityException - if a security manager exists and its 1045 * checkPermission method doesn't allow the operation. 1046 * @throws NullPointerException - if category and/or newLocale is null 1047 * @see SecurityManager#checkPermission(java.security.Permission) 1048 * @see PropertyPermission 1049 * @see #getDefault(Locale.Category) 1050 * @since 1.7 1051 */ 1052 public static synchronized void setDefault(Locale.Category category, 1053 Locale newLocale) { 1054 if (category == null) 1055 throw new NullPointerException("Category cannot be NULL"); 1056 if (newLocale == null) 1057 throw new NullPointerException("Can't set default locale to NULL"); 1058 1059 SecurityManager sm = System.getSecurityManager(); 1060 if (sm != null) sm.checkPermission(new PropertyPermission 1061 ("user.language", "write")); 1062 switch (category) { 1063 case DISPLAY: 1064 defaultDisplayLocale = newLocale; 1065 break; 1066 case FORMAT: 1067 defaultFormatLocale = newLocale; 1068 break; 1069 default: 1070 assert false: "Unknown Category"; 1071 } 1072 } 1073 1074 // Android-changed: Removed references to LocaleServiceProvider. 1075 /** 1076 * Returns an array of all installed locales. 1077 * 1078 * @return An array of installed locales. 1079 */ 1080 public static Locale[] getAvailableLocales() { 1081 // Android-changed: Removed used of LocaleServiceProviderPool. Switched to use ICU. 1082 return ICU.getAvailableLocales(); 1083 } 1084 1085 /** 1086 * Returns a list of all 2-letter country codes defined in ISO 3166. 1087 * Can be used to create Locales. 1088 * <p> 1089 * <b>Note:</b> The <code>Locale</code> class also supports other codes for 1090 * country (region), such as 3-letter numeric UN M.49 area codes. 1091 * Therefore, the list returned by this method does not contain ALL valid 1092 * codes that can be used to create Locales. 1093 * 1094 * @return An array of ISO 3166 two-letter country codes. 1095 */ 1096 public static String[] getISOCountries() { 1097 // Android-changed: Switched to use ICU. 1098 return ICU.getISOCountries(); 1099 } 1100 1101 /** 1102 * Returns a list of all 2-letter language codes defined in ISO 639. 1103 * Can be used to create Locales. 1104 * <p> 1105 * <b>Note:</b> 1106 * <ul> 1107 * <li>ISO 639 is not a stable standard— some languages' codes have changed. 1108 * The list this function returns includes both the new and the old codes for the 1109 * languages whose codes have changed. 1110 * <li>The <code>Locale</code> class also supports language codes up to 1111 * 8 characters in length. Therefore, the list returned by this method does 1112 * not contain ALL valid codes that can be used to create Locales. 1113 * </ul> 1114 * 1115 * @return Am array of ISO 639 two-letter language codes. 1116 */ 1117 public static String[] getISOLanguages() { 1118 // Android-changed: Switched to use ICU. 1119 return ICU.getISOLanguages(); 1120 } 1121 1122 /** 1123 * Returns the language code of this Locale. 1124 * 1125 * <p><b>Note:</b> ISO 639 is not a stable standard— some languages' codes have changed. 1126 * Locale's constructor recognizes both the new and the old codes for the languages 1127 * whose codes have changed, but this function always returns the old code. If you 1128 * want to check for a specific language whose code has changed, don't do 1129 * <pre> 1130 * if (locale.getLanguage().equals("he")) // BAD! 1131 * ... 1132 * </pre> 1133 * Instead, do 1134 * <pre> 1135 * if (locale.getLanguage().equals(new Locale("he").getLanguage())) 1136 * ... 1137 * </pre> 1138 * @return The language code, or the empty string if none is defined. 1139 * @see #getDisplayLanguage 1140 */ 1141 public String getLanguage() { 1142 return baseLocale.getLanguage(); 1143 } 1144 1145 /** 1146 * Returns the script for this locale, which should 1147 * either be the empty string or an ISO 15924 4-letter script 1148 * code. The first letter is uppercase and the rest are 1149 * lowercase, for example, 'Latn', 'Cyrl'. 1150 * 1151 * @return The script code, or the empty string if none is defined. 1152 * @see #getDisplayScript 1153 * @since 1.7 1154 */ 1155 public String getScript() { 1156 return baseLocale.getScript(); 1157 } 1158 1159 /** 1160 * Returns the country/region code for this locale, which should 1161 * either be the empty string, an uppercase ISO 3166 2-letter code, 1162 * or a UN M.49 3-digit code. 1163 * 1164 * @return The country/region code, or the empty string if none is defined. 1165 * @see #getDisplayCountry 1166 */ 1167 public String getCountry() { 1168 return baseLocale.getRegion(); 1169 } 1170 1171 /** 1172 * Returns the variant code for this locale. 1173 * 1174 * @return The variant code, or the empty string if none is defined. 1175 * @see #getDisplayVariant 1176 */ 1177 public String getVariant() { 1178 return baseLocale.getVariant(); 1179 } 1180 1181 /** 1182 * Returns {@code true} if this {@code Locale} has any <a href="#def_extensions"> 1183 * extensions</a>. 1184 * 1185 * @return {@code true} if this {@code Locale} has any extensions 1186 * @since 1.8 1187 */ 1188 public boolean hasExtensions() { 1189 return localeExtensions != null; 1190 } 1191 1192 /** 1193 * Returns a copy of this {@code Locale} with no <a href="#def_extensions"> 1194 * extensions</a>. If this {@code Locale} has no extensions, this {@code Locale} 1195 * is returned. 1196 * 1197 * @return a copy of this {@code Locale} with no extensions, or {@code this} 1198 * if {@code this} has no extensions 1199 * @since 1.8 1200 */ 1201 public Locale stripExtensions() { 1202 return hasExtensions() ? Locale.getInstance(baseLocale, null) : this; 1203 } 1204 1205 /** 1206 * Returns the extension (or private use) value associated with 1207 * the specified key, or null if there is no extension 1208 * associated with the key. To be well-formed, the key must be one 1209 * of <code>[0-9A-Za-z]</code>. Keys are case-insensitive, so 1210 * for example 'z' and 'Z' represent the same extension. 1211 * 1212 * @param key the extension key 1213 * @return The extension, or null if this locale defines no 1214 * extension for the specified key. 1215 * @throws IllegalArgumentException if key is not well-formed 1216 * @see #PRIVATE_USE_EXTENSION 1217 * @see #UNICODE_LOCALE_EXTENSION 1218 * @since 1.7 1219 */ 1220 public String getExtension(char key) { 1221 if (!LocaleExtensions.isValidKey(key)) { 1222 throw new IllegalArgumentException("Ill-formed extension key: " + key); 1223 } 1224 return hasExtensions() ? localeExtensions.getExtensionValue(key) : null; 1225 } 1226 1227 /** 1228 * Returns the set of extension keys associated with this locale, or the 1229 * empty set if it has no extensions. The returned set is unmodifiable. 1230 * The keys will all be lower-case. 1231 * 1232 * @return The set of extension keys, or the empty set if this locale has 1233 * no extensions. 1234 * @since 1.7 1235 */ 1236 public Set<Character> getExtensionKeys() { 1237 if (!hasExtensions()) { 1238 return Collections.emptySet(); 1239 } 1240 return localeExtensions.getKeys(); 1241 } 1242 1243 /** 1244 * Returns the set of unicode locale attributes associated with 1245 * this locale, or the empty set if it has no attributes. The 1246 * returned set is unmodifiable. 1247 * 1248 * @return The set of attributes. 1249 * @since 1.7 1250 */ 1251 public Set<String> getUnicodeLocaleAttributes() { 1252 if (!hasExtensions()) { 1253 return Collections.emptySet(); 1254 } 1255 return localeExtensions.getUnicodeLocaleAttributes(); 1256 } 1257 1258 /** 1259 * Returns the Unicode locale type associated with the specified Unicode locale key 1260 * for this locale. Returns the empty string for keys that are defined with no type. 1261 * Returns null if the key is not defined. Keys are case-insensitive. The key must 1262 * be two alphanumeric characters ([0-9a-zA-Z]), or an IllegalArgumentException is 1263 * thrown. 1264 * 1265 * @param key the Unicode locale key 1266 * @return The Unicode locale type associated with the key, or null if the 1267 * locale does not define the key. 1268 * @throws IllegalArgumentException if the key is not well-formed 1269 * @throws NullPointerException if <code>key</code> is null 1270 * @since 1.7 1271 */ 1272 public String getUnicodeLocaleType(String key) { 1273 if (!isUnicodeExtensionKey(key)) { 1274 throw new IllegalArgumentException("Ill-formed Unicode locale key: " + key); 1275 } 1276 return hasExtensions() ? localeExtensions.getUnicodeLocaleType(key) : null; 1277 } 1278 1279 /** 1280 * Returns the set of Unicode locale keys defined by this locale, or the empty set if 1281 * this locale has none. The returned set is immutable. Keys are all lower case. 1282 * 1283 * @return The set of Unicode locale keys, or the empty set if this locale has 1284 * no Unicode locale keywords. 1285 * @since 1.7 1286 */ 1287 public Set<String> getUnicodeLocaleKeys() { 1288 if (localeExtensions == null) { 1289 return Collections.emptySet(); 1290 } 1291 return localeExtensions.getUnicodeLocaleKeys(); 1292 } 1293 1294 /** 1295 * Package locale method returning the Locale's BaseLocale, 1296 * used by ResourceBundle 1297 * @return base locale of this Locale 1298 */ 1299 BaseLocale getBaseLocale() { 1300 return baseLocale; 1301 } 1302 1303 /** 1304 * Package private method returning the Locale's LocaleExtensions, 1305 * used by ResourceBundle. 1306 * @return locale exnteions of this Locale, 1307 * or {@code null} if no extensions are defined 1308 */ 1309 LocaleExtensions getLocaleExtensions() { 1310 return localeExtensions; 1311 } 1312 1313 /** 1314 * Returns a string representation of this <code>Locale</code> 1315 * object, consisting of language, country, variant, script, 1316 * and extensions as below: 1317 * <blockquote> 1318 * language + "_" + country + "_" + (variant + "_#" | "#") + script + "-" + extensions 1319 * </blockquote> 1320 * 1321 * Language is always lower case, country is always upper case, script is always title 1322 * case, and extensions are always lower case. Extensions and private use subtags 1323 * will be in canonical order as explained in {@link #toLanguageTag}. 1324 * 1325 * <p>When the locale has neither script nor extensions, the result is the same as in 1326 * Java 6 and prior. 1327 * 1328 * <p>If both the language and country fields are missing, this function will return 1329 * the empty string, even if the variant, script, or extensions field is present (you 1330 * can't have a locale with just a variant, the variant must accompany a well-formed 1331 * language or country code). 1332 * 1333 * <p>If script or extensions are present and variant is missing, no underscore is 1334 * added before the "#". 1335 * 1336 * <p>This behavior is designed to support debugging and to be compatible with 1337 * previous uses of <code>toString</code> that expected language, country, and variant 1338 * fields only. To represent a Locale as a String for interchange purposes, use 1339 * {@link #toLanguageTag}. 1340 * 1341 * <p>Examples: <ul> 1342 * <li><tt>en</tt></li> 1343 * <li><tt>de_DE</tt></li> 1344 * <li><tt>_GB</tt></li> 1345 * <li><tt>en_US_WIN</tt></li> 1346 * <li><tt>de__POSIX</tt></li> 1347 * <li><tt>zh_CN_#Hans</tt></li> 1348 * <li><tt>zh_TW_#Hant-x-java</tt></li> 1349 * <li><tt>th_TH_TH_#u-nu-thai</tt></li></ul> 1350 * 1351 * @return A string representation of the Locale, for debugging. 1352 * @see #getDisplayName 1353 * @see #toLanguageTag 1354 */ 1355 @Override 1356 public final String toString() { 1357 boolean l = (baseLocale.getLanguage().length() != 0); 1358 boolean s = (baseLocale.getScript().length() != 0); 1359 boolean r = (baseLocale.getRegion().length() != 0); 1360 boolean v = (baseLocale.getVariant().length() != 0); 1361 boolean e = (localeExtensions != null && localeExtensions.getID().length() != 0); 1362 1363 StringBuilder result = new StringBuilder(baseLocale.getLanguage()); 1364 if (r || (l && (v || s || e))) { 1365 result.append('_') 1366 .append(baseLocale.getRegion()); // This may just append '_' 1367 } 1368 if (v && (l || r)) { 1369 result.append('_') 1370 .append(baseLocale.getVariant()); 1371 } 1372 1373 if (s && (l || r)) { 1374 result.append("_#") 1375 .append(baseLocale.getScript()); 1376 } 1377 1378 if (e && (l || r)) { 1379 result.append('_'); 1380 if (!s) { 1381 result.append('#'); 1382 } 1383 result.append(localeExtensions.getID()); 1384 } 1385 1386 return result.toString(); 1387 } 1388 1389 /** 1390 * Returns a well-formed IETF BCP 47 language tag representing 1391 * this locale. 1392 * 1393 * <p>If this <code>Locale</code> has a language, country, or 1394 * variant that does not satisfy the IETF BCP 47 language tag 1395 * syntax requirements, this method handles these fields as 1396 * described below: 1397 * 1398 * <p><b>Language:</b> If language is empty, or not <a 1399 * href="#def_language" >well-formed</a> (for example "a" or 1400 * "e2"), it will be emitted as "und" (Undetermined). 1401 * 1402 * <p><b>Country:</b> If country is not <a 1403 * href="#def_region">well-formed</a> (for example "12" or "USA"), 1404 * it will be omitted. 1405 * 1406 * <p><b>Variant:</b> If variant <b>is</b> <a 1407 * href="#def_variant">well-formed</a>, each sub-segment 1408 * (delimited by '-' or '_') is emitted as a subtag. Otherwise: 1409 * <ul> 1410 * 1411 * <li>if all sub-segments match <code>[0-9a-zA-Z]{1,8}</code> 1412 * (for example "WIN" or "Oracle_JDK_Standard_Edition"), the first 1413 * ill-formed sub-segment and all following will be appended to 1414 * the private use subtag. The first appended subtag will be 1415 * "lvariant", followed by the sub-segments in order, separated by 1416 * hyphen. For example, "x-lvariant-WIN", 1417 * "Oracle-x-lvariant-JDK-Standard-Edition". 1418 * 1419 * <li>if any sub-segment does not match 1420 * <code>[0-9a-zA-Z]{1,8}</code>, the variant will be truncated 1421 * and the problematic sub-segment and all following sub-segments 1422 * will be omitted. If the remainder is non-empty, it will be 1423 * emitted as a private use subtag as above (even if the remainder 1424 * turns out to be well-formed). For example, 1425 * "Solaris_isjustthecoolestthing" is emitted as 1426 * "x-lvariant-Solaris", not as "solaris".</li></ul> 1427 * 1428 * <p><b>Special Conversions:</b> Java supports some old locale 1429 * representations, including deprecated ISO language codes, 1430 * for compatibility. This method performs the following 1431 * conversions: 1432 * <ul> 1433 * 1434 * <li>Deprecated ISO language codes "iw", "ji", and "in" are 1435 * converted to "he", "yi", and "id", respectively. 1436 * 1437 * <li>A locale with language "no", country "NO", and variant 1438 * "NY", representing Norwegian Nynorsk (Norway), is converted 1439 * to a language tag "nn-NO".</li></ul> 1440 * 1441 * <p><b>Note:</b> Although the language tag created by this 1442 * method is well-formed (satisfies the syntax requirements 1443 * defined by the IETF BCP 47 specification), it is not 1444 * necessarily a valid BCP 47 language tag. For example, 1445 * <pre> 1446 * new Locale("xx", "YY").toLanguageTag();</pre> 1447 * 1448 * will return "xx-YY", but the language subtag "xx" and the 1449 * region subtag "YY" are invalid because they are not registered 1450 * in the IANA Language Subtag Registry. 1451 * 1452 * @return a BCP47 language tag representing the locale 1453 * @see #forLanguageTag(String) 1454 * @since 1.7 1455 */ 1456 public String toLanguageTag() { 1457 if (languageTag != null) { 1458 return languageTag; 1459 } 1460 1461 LanguageTag tag = LanguageTag.parseLocale(baseLocale, localeExtensions); 1462 StringBuilder buf = new StringBuilder(); 1463 1464 String subtag = tag.getLanguage(); 1465 if (subtag.length() > 0) { 1466 buf.append(LanguageTag.canonicalizeLanguage(subtag)); 1467 } 1468 1469 subtag = tag.getScript(); 1470 if (subtag.length() > 0) { 1471 buf.append(LanguageTag.SEP); 1472 buf.append(LanguageTag.canonicalizeScript(subtag)); 1473 } 1474 1475 subtag = tag.getRegion(); 1476 if (subtag.length() > 0) { 1477 buf.append(LanguageTag.SEP); 1478 buf.append(LanguageTag.canonicalizeRegion(subtag)); 1479 } 1480 1481 List<String>subtags = tag.getVariants(); 1482 for (String s : subtags) { 1483 buf.append(LanguageTag.SEP); 1484 // preserve casing 1485 buf.append(s); 1486 } 1487 1488 subtags = tag.getExtensions(); 1489 for (String s : subtags) { 1490 buf.append(LanguageTag.SEP); 1491 buf.append(LanguageTag.canonicalizeExtension(s)); 1492 } 1493 1494 subtag = tag.getPrivateuse(); 1495 if (subtag.length() > 0) { 1496 if (buf.length() > 0) { 1497 buf.append(LanguageTag.SEP); 1498 } 1499 buf.append(LanguageTag.PRIVATEUSE).append(LanguageTag.SEP); 1500 // preserve casing 1501 buf.append(subtag); 1502 } 1503 1504 String langTag = buf.toString(); 1505 synchronized (this) { 1506 if (languageTag == null) { 1507 languageTag = langTag; 1508 } 1509 } 1510 return languageTag; 1511 } 1512 1513 /** 1514 * Returns a locale for the specified IETF BCP 47 language tag string. 1515 * 1516 * <p>If the specified language tag contains any ill-formed subtags, 1517 * the first such subtag and all following subtags are ignored. Compare 1518 * to {@link Locale.Builder#setLanguageTag} which throws an exception 1519 * in this case. 1520 * 1521 * <p>The following <b>conversions</b> are performed:<ul> 1522 * 1523 * <li>The language code "und" is mapped to language "". 1524 * 1525 * <li>The language codes "he", "yi", and "id" are mapped to "iw", 1526 * "ji", and "in" respectively. (This is the same canonicalization 1527 * that's done in Locale's constructors.) 1528 * 1529 * <li>The portion of a private use subtag prefixed by "lvariant", 1530 * if any, is removed and appended to the variant field in the 1531 * result locale (without case normalization). If it is then 1532 * empty, the private use subtag is discarded: 1533 * 1534 * <pre> 1535 * Locale loc; 1536 * loc = Locale.forLanguageTag("en-US-x-lvariant-POSIX"); 1537 * loc.getVariant(); // returns "POSIX" 1538 * loc.getExtension('x'); // returns null 1539 * 1540 * loc = Locale.forLanguageTag("de-POSIX-x-URP-lvariant-Abc-Def"); 1541 * loc.getVariant(); // returns "POSIX_Abc_Def" 1542 * loc.getExtension('x'); // returns "urp" 1543 * </pre> 1544 * 1545 * <li>When the languageTag argument contains an extlang subtag, 1546 * the first such subtag is used as the language, and the primary 1547 * language subtag and other extlang subtags are ignored: 1548 * 1549 * <pre> 1550 * Locale.forLanguageTag("ar-aao").getLanguage(); // returns "aao" 1551 * Locale.forLanguageTag("en-abc-def-us").toString(); // returns "abc_US" 1552 * </pre> 1553 * 1554 * <li>Case is normalized except for variant tags, which are left 1555 * unchanged. Language is normalized to lower case, script to 1556 * title case, country to upper case, and extensions to lower 1557 * case. 1558 * 1559 * <li>If, after processing, the locale would exactly match either 1560 * ja_JP_JP or th_TH_TH with no extensions, the appropriate 1561 * extensions are added as though the constructor had been called: 1562 * 1563 * <pre> 1564 * Locale.forLanguageTag("ja-JP-x-lvariant-JP").toLanguageTag(); 1565 * // returns "ja-JP-u-ca-japanese-x-lvariant-JP" 1566 * Locale.forLanguageTag("th-TH-x-lvariant-TH").toLanguageTag(); 1567 * // returns "th-TH-u-nu-thai-x-lvariant-TH" 1568 * </pre></ul> 1569 * 1570 * <p>This implements the 'Language-Tag' production of BCP47, and 1571 * so supports grandfathered (regular and irregular) as well as 1572 * private use language tags. Stand alone private use tags are 1573 * represented as empty language and extension 'x-whatever', 1574 * and grandfathered tags are converted to their canonical replacements 1575 * where they exist. 1576 * 1577 * <p>Grandfathered tags with canonical replacements are as follows: 1578 * 1579 * <table summary="Grandfathered tags with canonical replacements"> 1580 * <tbody align="center"> 1581 * <tr><th>grandfathered tag</th><th> </th><th>modern replacement</th></tr> 1582 * <tr><td>art-lojban</td><td> </td><td>jbo</td></tr> 1583 * <tr><td>i-ami</td><td> </td><td>ami</td></tr> 1584 * <tr><td>i-bnn</td><td> </td><td>bnn</td></tr> 1585 * <tr><td>i-hak</td><td> </td><td>hak</td></tr> 1586 * <tr><td>i-klingon</td><td> </td><td>tlh</td></tr> 1587 * <tr><td>i-lux</td><td> </td><td>lb</td></tr> 1588 * <tr><td>i-navajo</td><td> </td><td>nv</td></tr> 1589 * <tr><td>i-pwn</td><td> </td><td>pwn</td></tr> 1590 * <tr><td>i-tao</td><td> </td><td>tao</td></tr> 1591 * <tr><td>i-tay</td><td> </td><td>tay</td></tr> 1592 * <tr><td>i-tsu</td><td> </td><td>tsu</td></tr> 1593 * <tr><td>no-bok</td><td> </td><td>nb</td></tr> 1594 * <tr><td>no-nyn</td><td> </td><td>nn</td></tr> 1595 * <tr><td>sgn-BE-FR</td><td> </td><td>sfb</td></tr> 1596 * <tr><td>sgn-BE-NL</td><td> </td><td>vgt</td></tr> 1597 * <tr><td>sgn-CH-DE</td><td> </td><td>sgg</td></tr> 1598 * <tr><td>zh-guoyu</td><td> </td><td>cmn</td></tr> 1599 * <tr><td>zh-hakka</td><td> </td><td>hak</td></tr> 1600 * <tr><td>zh-min-nan</td><td> </td><td>nan</td></tr> 1601 * <tr><td>zh-xiang</td><td> </td><td>hsn</td></tr> 1602 * </tbody> 1603 * </table> 1604 * 1605 * <p>Grandfathered tags with no modern replacement will be 1606 * converted as follows: 1607 * 1608 * <table summary="Grandfathered tags with no modern replacement"> 1609 * <tbody align="center"> 1610 * <tr><th>grandfathered tag</th><th> </th><th>converts to</th></tr> 1611 * <tr><td>cel-gaulish</td><td> </td><td>xtg-x-cel-gaulish</td></tr> 1612 * <tr><td>en-GB-oed</td><td> </td><td>en-GB-x-oed</td></tr> 1613 * <tr><td>i-default</td><td> </td><td>en-x-i-default</td></tr> 1614 * <tr><td>i-enochian</td><td> </td><td>und-x-i-enochian</td></tr> 1615 * <tr><td>i-mingo</td><td> </td><td>see-x-i-mingo</td></tr> 1616 * <tr><td>zh-min</td><td> </td><td>nan-x-zh-min</td></tr> 1617 * </tbody> 1618 * </table> 1619 * 1620 * <p>For a list of all grandfathered tags, see the 1621 * IANA Language Subtag Registry (search for "Type: grandfathered"). 1622 * 1623 * <p><b>Note</b>: there is no guarantee that <code>toLanguageTag</code> 1624 * and <code>forLanguageTag</code> will round-trip. 1625 * 1626 * @param languageTag the language tag 1627 * @return The locale that best represents the language tag. 1628 * @throws NullPointerException if <code>languageTag</code> is <code>null</code> 1629 * @see #toLanguageTag() 1630 * @see java.util.Locale.Builder#setLanguageTag(String) 1631 * @since 1.7 1632 */ 1633 public static Locale forLanguageTag(String languageTag) { 1634 LanguageTag tag = LanguageTag.parse(languageTag, null); 1635 InternalLocaleBuilder bldr = new InternalLocaleBuilder(); 1636 bldr.setLanguageTag(tag); 1637 BaseLocale base = bldr.getBaseLocale(); 1638 LocaleExtensions exts = bldr.getLocaleExtensions(); 1639 if (exts == null && base.getVariant().length() > 0) { 1640 exts = getCompatibilityExtensions(base.getLanguage(), base.getScript(), 1641 base.getRegion(), base.getVariant()); 1642 } 1643 return getInstance(base, exts); 1644 } 1645 1646 /** 1647 * Returns a three-letter abbreviation of this locale's language. 1648 * If the language matches an ISO 639-1 two-letter code, the 1649 * corresponding ISO 639-2/T three-letter lowercase code is 1650 * returned. The ISO 639-2 language codes can be found on-line, 1651 * see "Codes for the Representation of Names of Languages Part 2: 1652 * Alpha-3 Code". If the locale specifies a three-letter 1653 * language, the language is returned as is. If the locale does 1654 * not specify a language the empty string is returned. 1655 * 1656 * @return A three-letter abbreviation of this locale's language. 1657 * @exception MissingResourceException Throws MissingResourceException if 1658 * three-letter language abbreviation is not available for this locale. 1659 */ 1660 public String getISO3Language() throws MissingResourceException { 1661 String lang = baseLocale.getLanguage(); 1662 if (lang.length() == 3) { 1663 return lang; 1664 } 1665 // Android-added BEGIN 1666 // return "" for empty languages for the sake of backwards compatibility. 1667 else if (lang.isEmpty()) { 1668 return ""; 1669 } 1670 // Android-added END 1671 1672 // Android-changed BEGIN: Use ICU. 1673 // String language3 = getISO3Code(lang, LocaleISOData.isoLanguageTable); 1674 // if (language3 == null) { 1675 String language3 = ICU.getISO3Language(lang); 1676 if (!lang.isEmpty() && language3.isEmpty()) { 1677 // Android-changed END 1678 throw new MissingResourceException("Couldn't find 3-letter language code for " 1679 + lang, "FormatData_" + toString(), "ShortLanguage"); 1680 } 1681 return language3; 1682 } 1683 1684 /** 1685 * Returns a three-letter abbreviation for this locale's country. 1686 * If the country matches an ISO 3166-1 alpha-2 code, the 1687 * corresponding ISO 3166-1 alpha-3 uppercase code is returned. 1688 * If the locale doesn't specify a country, this will be the empty 1689 * string. 1690 * 1691 * <p>The ISO 3166-1 codes can be found on-line. 1692 * 1693 * @return A three-letter abbreviation of this locale's country. 1694 * @exception MissingResourceException Throws MissingResourceException if the 1695 * three-letter country abbreviation is not available for this locale. 1696 */ 1697 public String getISO3Country() throws MissingResourceException { 1698 // Android-changed BEGIN: Use ICU. Also return "" for missing regions. 1699 final String region = baseLocale.getRegion(); 1700 // Note that this will return an UN.M49 region code 1701 if (region.length() == 3) { 1702 return baseLocale.getRegion(); 1703 } else if (region.isEmpty()) { 1704 return ""; 1705 } 1706 1707 // Prefix "en-" because ICU doesn't really care about what the language is. 1708 String country3 = ICU.getISO3Country("en-" + region); 1709 if (!region.isEmpty() && country3.isEmpty()) { 1710 throw new MissingResourceException("Couldn't find 3-letter country code for " 1711 + baseLocale.getRegion(), "FormatData_" + toString(), "ShortCountry"); 1712 } 1713 // Android-changed END 1714 return country3; 1715 } 1716 1717 /** 1718 * Returns a name for the locale's language that is appropriate for display to the 1719 * user. 1720 * If possible, the name returned will be localized for the default 1721 * {@link Locale.Category#DISPLAY DISPLAY} locale. 1722 * For example, if the locale is fr_FR and the default 1723 * {@link Locale.Category#DISPLAY DISPLAY} locale 1724 * is en_US, getDisplayLanguage() will return "French"; if the locale is en_US and 1725 * the default {@link Locale.Category#DISPLAY DISPLAY} locale is fr_FR, 1726 * getDisplayLanguage() will return "anglais". 1727 * If the name returned cannot be localized for the default 1728 * {@link Locale.Category#DISPLAY DISPLAY} locale, 1729 * (say, we don't have a Japanese name for Croatian), 1730 * this function falls back on the English name, and uses the ISO code as a last-resort 1731 * value. If the locale doesn't specify a language, this function returns the empty string. 1732 * 1733 * @return The name of the display language. 1734 */ 1735 public final String getDisplayLanguage() { 1736 return getDisplayLanguage(getDefault(Category.DISPLAY)); 1737 } 1738 1739 // Android-changed BEGIN: Use ICU; documentation; backwards compatibility hacks; 1740 // added private helper methods. 1741 /** 1742 * Returns the name of this locale's language, localized to {@code locale}. 1743 * If the language name is unknown, the language code is returned. 1744 */ 1745 public String getDisplayLanguage(Locale locale) { 1746 String languageCode = baseLocale.getLanguage(); 1747 if (languageCode.isEmpty()) { 1748 return ""; 1749 } 1750 1751 // Hacks for backward compatibility. 1752 // 1753 // Our language tag will contain "und" if the languageCode is invalid 1754 // or missing. ICU will then return "langue indéterminée" or the equivalent 1755 // display language for the indeterminate language code. 1756 // 1757 // Sigh... ugh... and what not. 1758 final String normalizedLanguage = normalizeAndValidateLanguage( 1759 languageCode, false /* strict */); 1760 if (UNDETERMINED_LANGUAGE.equals(normalizedLanguage)) { 1761 return languageCode; 1762 } 1763 1764 // TODO: We need a new hack or a complete fix for http://b/8049507 --- We would 1765 // cover the frameworks' tracks when they were using "tl" instead of "fil". 1766 String result = ICU.getDisplayLanguage(this, locale); 1767 if (result == null) { // TODO: do we need to do this, or does ICU do it for us? 1768 result = ICU.getDisplayLanguage(this, Locale.getDefault()); 1769 } 1770 return result; 1771 } 1772 1773 private static String normalizeAndValidateLanguage(String language, boolean strict) { 1774 if (language == null || language.isEmpty()) { 1775 return ""; 1776 } 1777 1778 final String lowercaseLanguage = language.toLowerCase(Locale.ROOT); 1779 if (!isValidBcp47Alpha(lowercaseLanguage, 2, 3)) { 1780 if (strict) { 1781 throw new IllformedLocaleException("Invalid language: " + language); 1782 } else { 1783 return UNDETERMINED_LANGUAGE; 1784 } 1785 } 1786 1787 return lowercaseLanguage; 1788 } 1789 1790 /* 1791 * Checks whether a given string is an ASCII alphanumeric string. 1792 */ 1793 private static boolean isAsciiAlphaNum(String string) { 1794 for (int i = 0; i < string.length(); i++) { 1795 final char character = string.charAt(i); 1796 if (!(character >= 'a' && character <= 'z' || 1797 character >= 'A' && character <= 'Z' || 1798 character >= '0' && character <= '9')) { 1799 return false; 1800 } 1801 } 1802 1803 return true; 1804 } 1805 // Android-changed END 1806 1807 /** 1808 * Returns a name for the the locale's script that is appropriate for display to 1809 * the user. If possible, the name will be localized for the default 1810 * {@link Locale.Category#DISPLAY DISPLAY} locale. Returns 1811 * the empty string if this locale doesn't specify a script code. 1812 * 1813 * @return the display name of the script code for the current default 1814 * {@link Locale.Category#DISPLAY DISPLAY} locale 1815 * @since 1.7 1816 */ 1817 public String getDisplayScript() { 1818 return getDisplayScript(getDefault(Category.DISPLAY)); 1819 } 1820 1821 /** 1822 * Returns a name for the locale's script that is appropriate 1823 * for display to the user. If possible, the name will be 1824 * localized for the given locale. Returns the empty string if 1825 * this locale doesn't specify a script code. 1826 * 1827 * @param inLocale The locale for which to retrieve the display script. 1828 * @return the display name of the script code for the current default 1829 * {@link Locale.Category#DISPLAY DISPLAY} locale 1830 * @throws NullPointerException if <code>inLocale</code> is <code>null</code> 1831 * @since 1.7 1832 */ 1833 public String getDisplayScript(Locale inLocale) { 1834 // Android-changed BEGIN: Use ICU. 1835 String scriptCode = baseLocale.getScript(); 1836 if (scriptCode.isEmpty()) { 1837 return ""; 1838 } 1839 1840 String result = ICU.getDisplayScript(this, inLocale); 1841 if (result == null) { // TODO: do we need to do this, or does ICU do it for us? 1842 result = ICU.getDisplayScript(this, Locale.getDefault(Category.DISPLAY)); 1843 } 1844 1845 return result; 1846 // Android-changed END 1847 } 1848 1849 /** 1850 * Returns a name for the locale's country that is appropriate for display to the 1851 * user. 1852 * If possible, the name returned will be localized for the default 1853 * {@link Locale.Category#DISPLAY DISPLAY} locale. 1854 * For example, if the locale is fr_FR and the default 1855 * {@link Locale.Category#DISPLAY DISPLAY} locale 1856 * is en_US, getDisplayCountry() will return "France"; if the locale is en_US and 1857 * the default {@link Locale.Category#DISPLAY DISPLAY} locale is fr_FR, 1858 * getDisplayCountry() will return "Etats-Unis". 1859 * If the name returned cannot be localized for the default 1860 * {@link Locale.Category#DISPLAY DISPLAY} locale, 1861 * (say, we don't have a Japanese name for Croatia), 1862 * this function falls back on the English name, and uses the ISO code as a last-resort 1863 * value. If the locale doesn't specify a country, this function returns the empty string. 1864 * 1865 * @return The name of the country appropriate to the locale. 1866 */ 1867 public final String getDisplayCountry() { 1868 return getDisplayCountry(getDefault(Category.DISPLAY)); 1869 } 1870 1871 // Android-changed BEGIN: Use ICU; documentation; added private helper methods. 1872 /** 1873 * Returns the name of this locale's country, localized to {@code locale}. 1874 * Returns the empty string if this locale does not correspond to a specific 1875 * country. 1876 */ 1877 public String getDisplayCountry(Locale locale) { 1878 String countryCode = baseLocale.getRegion(); 1879 if (countryCode.isEmpty()) { 1880 return ""; 1881 } 1882 1883 final String normalizedRegion = normalizeAndValidateRegion( 1884 countryCode, false /* strict */); 1885 if (normalizedRegion.isEmpty()) { 1886 return countryCode; 1887 } 1888 1889 String result = ICU.getDisplayCountry(this, locale); 1890 if (result == null) { // TODO: do we need to do this, or does ICU do it for us? 1891 result = ICU.getDisplayCountry(this, Locale.getDefault()); 1892 } 1893 return result; 1894 } 1895 1896 private static String normalizeAndValidateRegion(String region, boolean strict) { 1897 if (region == null || region.isEmpty()) { 1898 return ""; 1899 } 1900 1901 final String uppercaseRegion = region.toUpperCase(Locale.ROOT); 1902 if (!isValidBcp47Alpha(uppercaseRegion, 2, 2) && 1903 !isUnM49AreaCode(uppercaseRegion)) { 1904 if (strict) { 1905 throw new IllformedLocaleException("Invalid region: " + region); 1906 } else { 1907 return ""; 1908 } 1909 } 1910 1911 return uppercaseRegion; 1912 } 1913 1914 private static boolean isValidBcp47Alpha(String string, int lowerBound, int upperBound) { 1915 final int length = string.length(); 1916 if (length < lowerBound || length > upperBound) { 1917 return false; 1918 } 1919 1920 for (int i = 0; i < length; ++i) { 1921 final char character = string.charAt(i); 1922 if (!(character >= 'a' && character <= 'z' || 1923 character >= 'A' && character <= 'Z')) { 1924 return false; 1925 } 1926 } 1927 1928 return true; 1929 } 1930 1931 /** 1932 * A UN M.49 is a 3 digit numeric code. 1933 */ 1934 private static boolean isUnM49AreaCode(String code) { 1935 if (code.length() != 3) { 1936 return false; 1937 } 1938 1939 for (int i = 0; i < 3; ++i) { 1940 final char character = code.charAt(i); 1941 if (!(character >= '0' && character <= '9')) { 1942 return false; 1943 } 1944 } 1945 1946 return true; 1947 } 1948 // Android-changed END 1949 1950 /** 1951 * Returns a name for the locale's variant code that is appropriate for display to the 1952 * user. If possible, the name will be localized for the default 1953 * {@link Locale.Category#DISPLAY DISPLAY} locale. If the locale 1954 * doesn't specify a variant code, this function returns the empty string. 1955 * 1956 * @return The name of the display variant code appropriate to the locale. 1957 */ 1958 public final String getDisplayVariant() { 1959 return getDisplayVariant(getDefault(Category.DISPLAY)); 1960 } 1961 1962 /** 1963 * Returns a name for the locale's variant code that is appropriate for display to the 1964 * user. If possible, the name will be localized for inLocale. If the locale 1965 * doesn't specify a variant code, this function returns the empty string. 1966 * 1967 * @param inLocale The locale for which to retrieve the display variant code. 1968 * @return The name of the display variant code appropriate to the given locale. 1969 * @exception NullPointerException if <code>inLocale</code> is <code>null</code> 1970 */ 1971 // Android-changed BEGIN: Use ICU; added private helper methods. 1972 public String getDisplayVariant(Locale inLocale) { 1973 String variantCode = baseLocale.getVariant(); 1974 if (variantCode.isEmpty()) { 1975 return ""; 1976 } 1977 1978 try { 1979 normalizeAndValidateVariant(variantCode); 1980 } catch (IllformedLocaleException ilfe) { 1981 return variantCode; 1982 } 1983 1984 String result = ICU.getDisplayVariant(this, inLocale); 1985 if (result == null) { // TODO: do we need to do this, or does ICU do it for us? 1986 result = ICU.getDisplayVariant(this, Locale.getDefault()); 1987 } 1988 1989 // The "old style" locale constructors allow us to pass in variants that aren't 1990 // valid BCP-47 variant subtags. When that happens, toLanguageTag will not emit 1991 // them. Note that we know variantCode.length() > 0 due to the isEmpty check at 1992 // the beginning of this function. 1993 if (result.isEmpty()) { 1994 return variantCode; 1995 } 1996 return result; 1997 } 1998 1999 private static String normalizeAndValidateVariant(String variant) { 2000 if (variant == null || variant.isEmpty()) { 2001 return ""; 2002 } 2003 2004 // Note that unlike extensions, we canonicalize to lower case alphabets 2005 // and underscores instead of hyphens. 2006 final String normalizedVariant = variant.replace('-', '_'); 2007 String[] subTags = normalizedVariant.split("_"); 2008 2009 for (String subTag : subTags) { 2010 if (!isValidVariantSubtag(subTag)) { 2011 throw new IllformedLocaleException("Invalid variant: " + variant); 2012 } 2013 } 2014 2015 return normalizedVariant; 2016 } 2017 2018 private static boolean isValidVariantSubtag(String subTag) { 2019 // The BCP-47 spec states that : 2020 // - Subtags can be between [5, 8] alphanumeric chars in length. 2021 // - Subtags that start with a number are allowed to be 4 chars in length. 2022 if (subTag.length() >= 5 && subTag.length() <= 8) { 2023 if (isAsciiAlphaNum(subTag)) { 2024 return true; 2025 } 2026 } else if (subTag.length() == 4) { 2027 final char firstChar = subTag.charAt(0); 2028 if ((firstChar >= '0' && firstChar <= '9') && isAsciiAlphaNum(subTag)) { 2029 return true; 2030 } 2031 } 2032 2033 return false; 2034 } 2035 // Android-changed END 2036 2037 /** 2038 * Returns a name for the locale that is appropriate for display to the 2039 * user. This will be the values returned by getDisplayLanguage(), 2040 * getDisplayScript(), getDisplayCountry(), and getDisplayVariant() assembled 2041 * into a single string. The the non-empty values are used in order, 2042 * with the second and subsequent names in parentheses. For example: 2043 * <blockquote> 2044 * language (script, country, variant)<br> 2045 * language (country)<br> 2046 * language (variant)<br> 2047 * script (country)<br> 2048 * country<br> 2049 * </blockquote> 2050 * depending on which fields are specified in the locale. If the 2051 * language, script, country, and variant fields are all empty, 2052 * this function returns the empty string. 2053 * 2054 * @return The name of the locale appropriate to display. 2055 */ 2056 public final String getDisplayName() { 2057 return getDisplayName(getDefault(Category.DISPLAY)); 2058 } 2059 2060 // Android-changed BEGIN: Use ICU. 2061 /** 2062 * Returns this locale's language name, country name, and variant, localized 2063 * to {@code locale}. The exact output form depends on whether this locale 2064 * corresponds to a specific language, script, country and variant. 2065 * 2066 * <p>For example: 2067 * <ul> 2068 * <li>{@code new Locale("en").getDisplayName(Locale.US)} -> {@code English} 2069 * <li>{@code new Locale("en", "US").getDisplayName(Locale.US)} -> {@code English (United States)} 2070 * <li>{@code new Locale("en", "US", "POSIX").getDisplayName(Locale.US)} -> {@code English (United States,Computer)} 2071 * <li>{@code Locale.fromLanguageTag("zh-Hant-CN").getDisplayName(Locale.US)} -> {@code Chinese (Traditional Han,China)} 2072 * <li>{@code new Locale("en").getDisplayName(Locale.FRANCE)} -> {@code anglais} 2073 * <li>{@code new Locale("en", "US").getDisplayName(Locale.FRANCE)} -> {@code anglais (États-Unis)} 2074 * <li>{@code new Locale("en", "US", "POSIX").getDisplayName(Locale.FRANCE)} -> {@code anglais (États-Unis,informatique)}. 2075 * </ul> 2076 */ 2077 public String getDisplayName(Locale locale) { 2078 int count = 0; 2079 StringBuilder buffer = new StringBuilder(); 2080 String languageCode = baseLocale.getLanguage(); 2081 if (!languageCode.isEmpty()) { 2082 String displayLanguage = getDisplayLanguage(locale); 2083 buffer.append(displayLanguage.isEmpty() ? languageCode : displayLanguage); 2084 ++count; 2085 } 2086 String scriptCode = baseLocale.getScript(); 2087 if (!scriptCode.isEmpty()) { 2088 if (count == 1) { 2089 buffer.append(" ("); 2090 } 2091 String displayScript = getDisplayScript(locale); 2092 buffer.append(displayScript.isEmpty() ? scriptCode : displayScript); 2093 ++count; 2094 } 2095 String countryCode = baseLocale.getRegion(); 2096 if (!countryCode.isEmpty()) { 2097 if (count == 1) { 2098 buffer.append(" ("); 2099 } else if (count == 2) { 2100 buffer.append(","); 2101 } 2102 String displayCountry = getDisplayCountry(locale); 2103 buffer.append(displayCountry.isEmpty() ? countryCode : displayCountry); 2104 ++count; 2105 } 2106 String variantCode = baseLocale.getVariant(); 2107 if (!variantCode.isEmpty()) { 2108 if (count == 1) { 2109 buffer.append(" ("); 2110 } else if (count == 2 || count == 3) { 2111 buffer.append(","); 2112 } 2113 String displayVariant = getDisplayVariant(locale); 2114 buffer.append(displayVariant.isEmpty() ? variantCode : displayVariant); 2115 ++count; 2116 } 2117 if (count > 1) { 2118 buffer.append(")"); 2119 } 2120 return buffer.toString(); 2121 } 2122 // Android-changed END 2123 2124 /** 2125 * Overrides Cloneable. 2126 */ 2127 @Override 2128 public Object clone() 2129 { 2130 try { 2131 Locale that = (Locale)super.clone(); 2132 return that; 2133 } catch (CloneNotSupportedException e) { 2134 throw new InternalError(e); 2135 } 2136 } 2137 2138 /** 2139 * Override hashCode. 2140 * Since Locales are often used in hashtables, caches the value 2141 * for speed. 2142 */ 2143 @Override 2144 public int hashCode() { 2145 int hc = hashCodeValue; 2146 if (hc == 0) { 2147 hc = baseLocale.hashCode(); 2148 if (localeExtensions != null) { 2149 hc ^= localeExtensions.hashCode(); 2150 } 2151 hashCodeValue = hc; 2152 } 2153 return hc; 2154 } 2155 2156 // Overrides 2157 2158 /** 2159 * Returns true if this Locale is equal to another object. A Locale is 2160 * deemed equal to another Locale with identical language, script, country, 2161 * variant and extensions, and unequal to all other objects. 2162 * 2163 * @return true if this Locale is equal to the specified object. 2164 */ 2165 @Override 2166 public boolean equals(Object obj) { 2167 if (this == obj) // quick check 2168 return true; 2169 if (!(obj instanceof Locale)) 2170 return false; 2171 BaseLocale otherBase = ((Locale)obj).baseLocale; 2172 if (!baseLocale.equals(otherBase)) { 2173 return false; 2174 } 2175 if (localeExtensions == null) { 2176 return ((Locale)obj).localeExtensions == null; 2177 } 2178 return localeExtensions.equals(((Locale)obj).localeExtensions); 2179 } 2180 2181 // ================= privates ===================================== 2182 2183 private transient BaseLocale baseLocale; 2184 private transient LocaleExtensions localeExtensions; 2185 2186 /** 2187 * Calculated hashcode 2188 */ 2189 private transient volatile int hashCodeValue = 0; 2190 2191 // Android-changed: Add NoImagePreloadHolder to allow compile-time initialization. 2192 private static class NoImagePreloadHolder { 2193 public volatile static Locale defaultLocale = initDefault(); 2194 } 2195 private volatile static Locale defaultDisplayLocale = null; 2196 private volatile static Locale defaultFormatLocale = null; 2197 2198 private transient volatile String languageTag; 2199 2200 /** 2201 * Format a list using given pattern strings. 2202 * If either of the patterns is null, then a the list is 2203 * formatted by concatenation with the delimiter ','. 2204 * @param stringList the list of strings to be formatted. 2205 * @param listPattern should create a MessageFormat taking 0-3 arguments 2206 * and formatting them into a list. 2207 * @param listCompositionPattern should take 2 arguments 2208 * and is used by composeList. 2209 * @return a string representing the list. 2210 */ 2211 private static String formatList(String[] stringList, String listPattern, String listCompositionPattern) { 2212 // If we have no list patterns, compose the list in a simple, 2213 // non-localized way. 2214 if (listPattern == null || listCompositionPattern == null) { 2215 StringBuilder result = new StringBuilder(); 2216 for (int i = 0; i < stringList.length; ++i) { 2217 if (i > 0) { 2218 result.append(','); 2219 } 2220 result.append(stringList[i]); 2221 } 2222 return result.toString(); 2223 } 2224 2225 // Compose the list down to three elements if necessary 2226 if (stringList.length > 3) { 2227 MessageFormat format = new MessageFormat(listCompositionPattern); 2228 stringList = composeList(format, stringList); 2229 } 2230 2231 // Rebuild the argument list with the list length as the first element 2232 Object[] args = new Object[stringList.length + 1]; 2233 System.arraycopy(stringList, 0, args, 1, stringList.length); 2234 args[0] = new Integer(stringList.length); 2235 2236 // Format it using the pattern in the resource 2237 MessageFormat format = new MessageFormat(listPattern); 2238 return format.format(args); 2239 } 2240 2241 /** 2242 * Given a list of strings, return a list shortened to three elements. 2243 * Shorten it by applying the given format to the first two elements 2244 * recursively. 2245 * @param format a format which takes two arguments 2246 * @param list a list of strings 2247 * @return if the list is three elements or shorter, the same list; 2248 * otherwise, a new list of three elements. 2249 */ 2250 private static String[] composeList(MessageFormat format, String[] list) { 2251 if (list.length <= 3) return list; 2252 2253 // Use the given format to compose the first two elements into one 2254 String[] listItems = { list[0], list[1] }; 2255 String newItem = format.format(listItems); 2256 2257 // Form a new list one element shorter 2258 String[] newList = new String[list.length-1]; 2259 System.arraycopy(list, 2, newList, 1, newList.length-1); 2260 newList[0] = newItem; 2261 2262 // Recurse 2263 return composeList(format, newList); 2264 } 2265 2266 // Duplicate of sun.util.locale.UnicodeLocaleExtension.isKey in order to 2267 // avoid its class loading. 2268 private static boolean isUnicodeExtensionKey(String s) { 2269 // 2alphanum 2270 return (s.length() == 2) && LocaleUtils.isAlphaNumericString(s); 2271 } 2272 2273 /** 2274 * @serialField language String 2275 * language subtag in lower case. (See <a href="java/util/Locale.html#getLanguage()">getLanguage()</a>) 2276 * @serialField country String 2277 * country subtag in upper case. (See <a href="java/util/Locale.html#getCountry()">getCountry()</a>) 2278 * @serialField variant String 2279 * variant subtags separated by LOWLINE characters. (See <a href="java/util/Locale.html#getVariant()">getVariant()</a>) 2280 * @serialField hashcode int 2281 * deprecated, for forward compatibility only 2282 * @serialField script String 2283 * script subtag in title case (See <a href="java/util/Locale.html#getScript()">getScript()</a>) 2284 * @serialField extensions String 2285 * canonical representation of extensions, that is, 2286 * BCP47 extensions in alphabetical order followed by 2287 * BCP47 private use subtags, all in lower case letters 2288 * separated by HYPHEN-MINUS characters. 2289 * (See <a href="java/util/Locale.html#getExtensionKeys()">getExtensionKeys()</a>, 2290 * <a href="java/util/Locale.html#getExtension(char)">getExtension(char)</a>) 2291 */ 2292 private static final ObjectStreamField[] serialPersistentFields = { 2293 new ObjectStreamField("language", String.class), 2294 new ObjectStreamField("country", String.class), 2295 new ObjectStreamField("variant", String.class), 2296 new ObjectStreamField("hashcode", int.class), 2297 new ObjectStreamField("script", String.class), 2298 new ObjectStreamField("extensions", String.class), 2299 }; 2300 2301 /** 2302 * Serializes this <code>Locale</code> to the specified <code>ObjectOutputStream</code>. 2303 * @param out the <code>ObjectOutputStream</code> to write 2304 * @throws IOException 2305 * @since 1.7 2306 */ 2307 private void writeObject(ObjectOutputStream out) throws IOException { 2308 ObjectOutputStream.PutField fields = out.putFields(); 2309 fields.put("language", baseLocale.getLanguage()); 2310 fields.put("script", baseLocale.getScript()); 2311 fields.put("country", baseLocale.getRegion()); 2312 fields.put("variant", baseLocale.getVariant()); 2313 fields.put("extensions", localeExtensions == null ? "" : localeExtensions.getID()); 2314 fields.put("hashcode", -1); // place holder just for backward support 2315 out.writeFields(); 2316 } 2317 2318 /** 2319 * Deserializes this <code>Locale</code>. 2320 * @param in the <code>ObjectInputStream</code> to read 2321 * @throws IOException 2322 * @throws ClassNotFoundException 2323 * @throws IllformedLocaleException 2324 * @since 1.7 2325 */ 2326 private void readObject(ObjectInputStream in) throws IOException, ClassNotFoundException { 2327 ObjectInputStream.GetField fields = in.readFields(); 2328 String language = (String)fields.get("language", ""); 2329 String script = (String)fields.get("script", ""); 2330 String country = (String)fields.get("country", ""); 2331 String variant = (String)fields.get("variant", ""); 2332 String extStr = (String)fields.get("extensions", ""); 2333 baseLocale = BaseLocale.getInstance(convertOldISOCodes(language), script, country, variant); 2334 // Android-changed: Handle null for backwards compatible deserialization. http://b/26387905 2335 // was: if (extStr.length() > 0) { 2336 if (extStr != null && extStr.length() > 0) { 2337 try { 2338 InternalLocaleBuilder bldr = new InternalLocaleBuilder(); 2339 bldr.setExtensions(extStr); 2340 localeExtensions = bldr.getLocaleExtensions(); 2341 } catch (LocaleSyntaxException e) { 2342 throw new IllformedLocaleException(e.getMessage()); 2343 } 2344 } else { 2345 localeExtensions = null; 2346 } 2347 } 2348 2349 /** 2350 * Returns a cached <code>Locale</code> instance equivalent to 2351 * the deserialized <code>Locale</code>. When serialized 2352 * language, country and variant fields read from the object data stream 2353 * are exactly "ja", "JP", "JP" or "th", "TH", "TH" and script/extensions 2354 * fields are empty, this method supplies <code>UNICODE_LOCALE_EXTENSION</code> 2355 * "ca"/"japanese" (calendar type is "japanese") or "nu"/"thai" (number script 2356 * type is "thai"). See <a href="Locale.html#special_cases_constructor">Special Cases</a> 2357 * for more information. 2358 * 2359 * @return an instance of <code>Locale</code> equivalent to 2360 * the deserialized <code>Locale</code>. 2361 * @throws java.io.ObjectStreamException 2362 */ 2363 private Object readResolve() throws java.io.ObjectStreamException { 2364 return getInstance(baseLocale.getLanguage(), baseLocale.getScript(), 2365 baseLocale.getRegion(), baseLocale.getVariant(), localeExtensions); 2366 } 2367 2368 private static volatile String[] isoLanguages = null; 2369 2370 private static volatile String[] isoCountries = null; 2371 2372 private static String convertOldISOCodes(String language) { 2373 // we accept both the old and the new ISO codes for the languages whose ISO 2374 // codes have changed, but we always store the OLD code, for backward compatibility 2375 language = LocaleUtils.toLowerString(language).intern(); 2376 if (language == "he") { 2377 return "iw"; 2378 } else if (language == "yi") { 2379 return "ji"; 2380 } else if (language == "id") { 2381 return "in"; 2382 } else { 2383 return language; 2384 } 2385 } 2386 2387 private static LocaleExtensions getCompatibilityExtensions(String language, 2388 String script, 2389 String country, 2390 String variant) { 2391 LocaleExtensions extensions = null; 2392 // Special cases for backward compatibility support 2393 if (LocaleUtils.caseIgnoreMatch(language, "ja") 2394 && script.length() == 0 2395 && LocaleUtils.caseIgnoreMatch(country, "jp") 2396 && "JP".equals(variant)) { 2397 // ja_JP_JP -> u-ca-japanese (calendar = japanese) 2398 extensions = LocaleExtensions.CALENDAR_JAPANESE; 2399 } else if (LocaleUtils.caseIgnoreMatch(language, "th") 2400 && script.length() == 0 2401 && LocaleUtils.caseIgnoreMatch(country, "th") 2402 && "TH".equals(variant)) { 2403 // th_TH_TH -> u-nu-thai (numbersystem = thai) 2404 extensions = LocaleExtensions.NUMBER_THAI; 2405 } 2406 return extensions; 2407 } 2408 2409 // Android-removed: Drop nested private class LocaleNameGetter. 2410 // Android-added BEGIN: Add adjustLanguageCode(); for internal use only. 2411 /** @hide for internal use only. */ 2412 public static String adjustLanguageCode(String languageCode) { 2413 String adjusted = languageCode.toLowerCase(Locale.US); 2414 // Map new language codes to the obsolete language 2415 // codes so the correct resource bundles will be used. 2416 if (languageCode.equals("he")) { 2417 adjusted = "iw"; 2418 } else if (languageCode.equals("id")) { 2419 adjusted = "in"; 2420 } else if (languageCode.equals("yi")) { 2421 adjusted = "ji"; 2422 } 2423 2424 return adjusted; 2425 } 2426 // Android-added END 2427 2428 /** 2429 * Enum for locale categories. These locale categories are used to get/set 2430 * the default locale for the specific functionality represented by the 2431 * category. 2432 * 2433 * @see #getDefault(Locale.Category) 2434 * @see #setDefault(Locale.Category, Locale) 2435 * @since 1.7 2436 */ 2437 public enum Category { 2438 2439 /** 2440 * Category used to represent the default locale for 2441 * displaying user interfaces. 2442 */ 2443 DISPLAY("user.language.display", 2444 "user.script.display", 2445 "user.country.display", 2446 "user.variant.display"), 2447 2448 /** 2449 * Category used to represent the default locale for 2450 * formatting dates, numbers, and/or currencies. 2451 */ 2452 FORMAT("user.language.format", 2453 "user.script.format", 2454 "user.country.format", 2455 "user.variant.format"); 2456 2457 Category(String languageKey, String scriptKey, String countryKey, String variantKey) { 2458 this.languageKey = languageKey; 2459 this.scriptKey = scriptKey; 2460 this.countryKey = countryKey; 2461 this.variantKey = variantKey; 2462 } 2463 2464 final String languageKey; 2465 final String scriptKey; 2466 final String countryKey; 2467 final String variantKey; 2468 } 2469 2470 /** 2471 * <code>Builder</code> is used to build instances of <code>Locale</code> 2472 * from values configured by the setters. Unlike the <code>Locale</code> 2473 * constructors, the <code>Builder</code> checks if a value configured by a 2474 * setter satisfies the syntax requirements defined by the <code>Locale</code> 2475 * class. A <code>Locale</code> object created by a <code>Builder</code> is 2476 * well-formed and can be transformed to a well-formed IETF BCP 47 language tag 2477 * without losing information. 2478 * 2479 * <p><b>Note:</b> The <code>Locale</code> class does not provide any 2480 * syntactic restrictions on variant, while BCP 47 requires each variant 2481 * subtag to be 5 to 8 alphanumerics or a single numeric followed by 3 2482 * alphanumerics. The method <code>setVariant</code> throws 2483 * <code>IllformedLocaleException</code> for a variant that does not satisfy 2484 * this restriction. If it is necessary to support such a variant, use a 2485 * Locale constructor. However, keep in mind that a <code>Locale</code> 2486 * object created this way might lose the variant information when 2487 * transformed to a BCP 47 language tag. 2488 * 2489 * <p>The following example shows how to create a <code>Locale</code> object 2490 * with the <code>Builder</code>. 2491 * <blockquote> 2492 * <pre> 2493 * Locale aLocale = new Builder().setLanguage("sr").setScript("Latn").setRegion("RS").build(); 2494 * </pre> 2495 * </blockquote> 2496 * 2497 * <p>Builders can be reused; <code>clear()</code> resets all 2498 * fields to their default values. 2499 * 2500 * @see Locale#forLanguageTag 2501 * @since 1.7 2502 */ 2503 public static final class Builder { 2504 private final InternalLocaleBuilder localeBuilder; 2505 2506 /** 2507 * Constructs an empty Builder. The default value of all 2508 * fields, extensions, and private use information is the 2509 * empty string. 2510 */ 2511 public Builder() { 2512 localeBuilder = new InternalLocaleBuilder(); 2513 } 2514 2515 /** 2516 * Resets the <code>Builder</code> to match the provided 2517 * <code>locale</code>. Existing state is discarded. 2518 * 2519 * <p>All fields of the locale must be well-formed, see {@link Locale}. 2520 * 2521 * <p>Locales with any ill-formed fields cause 2522 * <code>IllformedLocaleException</code> to be thrown, except for the 2523 * following three cases which are accepted for compatibility 2524 * reasons:<ul> 2525 * <li>Locale("ja", "JP", "JP") is treated as "ja-JP-u-ca-japanese" 2526 * <li>Locale("th", "TH", "TH") is treated as "th-TH-u-nu-thai" 2527 * <li>Locale("no", "NO", "NY") is treated as "nn-NO"</ul> 2528 * 2529 * @param locale the locale 2530 * @return This builder. 2531 * @throws IllformedLocaleException if <code>locale</code> has 2532 * any ill-formed fields. 2533 * @throws NullPointerException if <code>locale</code> is null. 2534 */ 2535 public Builder setLocale(Locale locale) { 2536 try { 2537 localeBuilder.setLocale(locale.baseLocale, locale.localeExtensions); 2538 } catch (LocaleSyntaxException e) { 2539 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex()); 2540 } 2541 return this; 2542 } 2543 2544 /** 2545 * Resets the Builder to match the provided IETF BCP 47 2546 * language tag. Discards the existing state. Null and the 2547 * empty string cause the builder to be reset, like {@link 2548 * #clear}. Grandfathered tags (see {@link 2549 * Locale#forLanguageTag}) are converted to their canonical 2550 * form before being processed. Otherwise, the language tag 2551 * must be well-formed (see {@link Locale}) or an exception is 2552 * thrown (unlike <code>Locale.forLanguageTag</code>, which 2553 * just discards ill-formed and following portions of the 2554 * tag). 2555 * 2556 * @param languageTag the language tag 2557 * @return This builder. 2558 * @throws IllformedLocaleException if <code>languageTag</code> is ill-formed 2559 * @see Locale#forLanguageTag(String) 2560 */ 2561 public Builder setLanguageTag(String languageTag) { 2562 ParseStatus sts = new ParseStatus(); 2563 LanguageTag tag = LanguageTag.parse(languageTag, sts); 2564 if (sts.isError()) { 2565 throw new IllformedLocaleException(sts.getErrorMessage(), sts.getErrorIndex()); 2566 } 2567 localeBuilder.setLanguageTag(tag); 2568 return this; 2569 } 2570 2571 /** 2572 * Sets the language. If <code>language</code> is the empty string or 2573 * null, the language in this <code>Builder</code> is removed. Otherwise, 2574 * the language must be <a href="./Locale.html#def_language">well-formed</a> 2575 * or an exception is thrown. 2576 * 2577 * <p>The typical language value is a two or three-letter language 2578 * code as defined in ISO639. 2579 * 2580 * @param language the language 2581 * @return This builder. 2582 * @throws IllformedLocaleException if <code>language</code> is ill-formed 2583 */ 2584 public Builder setLanguage(String language) { 2585 try { 2586 localeBuilder.setLanguage(language); 2587 } catch (LocaleSyntaxException e) { 2588 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex()); 2589 } 2590 return this; 2591 } 2592 2593 /** 2594 * Sets the script. If <code>script</code> is null or the empty string, 2595 * the script in this <code>Builder</code> is removed. 2596 * Otherwise, the script must be <a href="./Locale.html#def_script">well-formed</a> or an 2597 * exception is thrown. 2598 * 2599 * <p>The typical script value is a four-letter script code as defined by ISO 15924. 2600 * 2601 * @param script the script 2602 * @return This builder. 2603 * @throws IllformedLocaleException if <code>script</code> is ill-formed 2604 */ 2605 public Builder setScript(String script) { 2606 try { 2607 localeBuilder.setScript(script); 2608 } catch (LocaleSyntaxException e) { 2609 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex()); 2610 } 2611 return this; 2612 } 2613 2614 /** 2615 * Sets the region. If region is null or the empty string, the region 2616 * in this <code>Builder</code> is removed. Otherwise, 2617 * the region must be <a href="./Locale.html#def_region">well-formed</a> or an 2618 * exception is thrown. 2619 * 2620 * <p>The typical region value is a two-letter ISO 3166 code or a 2621 * three-digit UN M.49 area code. 2622 * 2623 * <p>The country value in the <code>Locale</code> created by the 2624 * <code>Builder</code> is always normalized to upper case. 2625 * 2626 * @param region the region 2627 * @return This builder. 2628 * @throws IllformedLocaleException if <code>region</code> is ill-formed 2629 */ 2630 public Builder setRegion(String region) { 2631 try { 2632 localeBuilder.setRegion(region); 2633 } catch (LocaleSyntaxException e) { 2634 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex()); 2635 } 2636 return this; 2637 } 2638 2639 /** 2640 * Sets the variant. If variant is null or the empty string, the 2641 * variant in this <code>Builder</code> is removed. Otherwise, it 2642 * must consist of one or more <a href="./Locale.html#def_variant">well-formed</a> 2643 * subtags, or an exception is thrown. 2644 * 2645 * <p><b>Note:</b> This method checks if <code>variant</code> 2646 * satisfies the IETF BCP 47 variant subtag's syntax requirements, 2647 * and normalizes the value to lowercase letters. However, 2648 * the <code>Locale</code> class does not impose any syntactic 2649 * restriction on variant, and the variant value in 2650 * <code>Locale</code> is case sensitive. To set such a variant, 2651 * use a Locale constructor. 2652 * 2653 * @param variant the variant 2654 * @return This builder. 2655 * @throws IllformedLocaleException if <code>variant</code> is ill-formed 2656 */ 2657 public Builder setVariant(String variant) { 2658 try { 2659 localeBuilder.setVariant(variant); 2660 } catch (LocaleSyntaxException e) { 2661 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex()); 2662 } 2663 return this; 2664 } 2665 2666 /** 2667 * Sets the extension for the given key. If the value is null or the 2668 * empty string, the extension is removed. Otherwise, the extension 2669 * must be <a href="./Locale.html#def_extensions">well-formed</a> or an exception 2670 * is thrown. 2671 * 2672 * <p><b>Note:</b> The key {@link Locale#UNICODE_LOCALE_EXTENSION 2673 * UNICODE_LOCALE_EXTENSION} ('u') is used for the Unicode locale extension. 2674 * Setting a value for this key replaces any existing Unicode locale key/type 2675 * pairs with those defined in the extension. 2676 * 2677 * <p><b>Note:</b> The key {@link Locale#PRIVATE_USE_EXTENSION 2678 * PRIVATE_USE_EXTENSION} ('x') is used for the private use code. To be 2679 * well-formed, the value for this key needs only to have subtags of one to 2680 * eight alphanumeric characters, not two to eight as in the general case. 2681 * 2682 * @param key the extension key 2683 * @param value the extension value 2684 * @return This builder. 2685 * @throws IllformedLocaleException if <code>key</code> is illegal 2686 * or <code>value</code> is ill-formed 2687 * @see #setUnicodeLocaleKeyword(String, String) 2688 */ 2689 public Builder setExtension(char key, String value) { 2690 try { 2691 localeBuilder.setExtension(key, value); 2692 } catch (LocaleSyntaxException e) { 2693 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex()); 2694 } 2695 return this; 2696 } 2697 2698 /** 2699 * Sets the Unicode locale keyword type for the given key. If the type 2700 * is null, the Unicode keyword is removed. Otherwise, the key must be 2701 * non-null and both key and type must be <a 2702 * href="./Locale.html#def_locale_extension">well-formed</a> or an exception 2703 * is thrown. 2704 * 2705 * <p>Keys and types are converted to lower case. 2706 * 2707 * <p><b>Note</b>:Setting the 'u' extension via {@link #setExtension} 2708 * replaces all Unicode locale keywords with those defined in the 2709 * extension. 2710 * 2711 * @param key the Unicode locale key 2712 * @param type the Unicode locale type 2713 * @return This builder. 2714 * @throws IllformedLocaleException if <code>key</code> or <code>type</code> 2715 * is ill-formed 2716 * @throws NullPointerException if <code>key</code> is null 2717 * @see #setExtension(char, String) 2718 */ 2719 public Builder setUnicodeLocaleKeyword(String key, String type) { 2720 try { 2721 localeBuilder.setUnicodeLocaleKeyword(key, type); 2722 } catch (LocaleSyntaxException e) { 2723 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex()); 2724 } 2725 return this; 2726 } 2727 2728 /** 2729 * Adds a unicode locale attribute, if not already present, otherwise 2730 * has no effect. The attribute must not be null and must be <a 2731 * href="./Locale.html#def_locale_extension">well-formed</a> or an exception 2732 * is thrown. 2733 * 2734 * @param attribute the attribute 2735 * @return This builder. 2736 * @throws NullPointerException if <code>attribute</code> is null 2737 * @throws IllformedLocaleException if <code>attribute</code> is ill-formed 2738 * @see #setExtension(char, String) 2739 */ 2740 public Builder addUnicodeLocaleAttribute(String attribute) { 2741 try { 2742 localeBuilder.addUnicodeLocaleAttribute(attribute); 2743 } catch (LocaleSyntaxException e) { 2744 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex()); 2745 } 2746 return this; 2747 } 2748 2749 /** 2750 * Removes a unicode locale attribute, if present, otherwise has no 2751 * effect. The attribute must not be null and must be <a 2752 * href="./Locale.html#def_locale_extension">well-formed</a> or an exception 2753 * is thrown. 2754 * 2755 * <p>Attribute comparision for removal is case-insensitive. 2756 * 2757 * @param attribute the attribute 2758 * @return This builder. 2759 * @throws NullPointerException if <code>attribute</code> is null 2760 * @throws IllformedLocaleException if <code>attribute</code> is ill-formed 2761 * @see #setExtension(char, String) 2762 */ 2763 public Builder removeUnicodeLocaleAttribute(String attribute) { 2764 // Android-added BEGIN 2765 if (attribute == null) { 2766 throw new NullPointerException("attribute == null"); 2767 } 2768 // Android-added END 2769 2770 try { 2771 localeBuilder.removeUnicodeLocaleAttribute(attribute); 2772 } catch (LocaleSyntaxException e) { 2773 throw new IllformedLocaleException(e.getMessage(), e.getErrorIndex()); 2774 } 2775 return this; 2776 } 2777 2778 /** 2779 * Resets the builder to its initial, empty state. 2780 * 2781 * @return This builder. 2782 */ 2783 public Builder clear() { 2784 localeBuilder.clear(); 2785 return this; 2786 } 2787 2788 /** 2789 * Resets the extensions to their initial, empty state. 2790 * Language, script, region and variant are unchanged. 2791 * 2792 * @return This builder. 2793 * @see #setExtension(char, String) 2794 */ 2795 public Builder clearExtensions() { 2796 localeBuilder.clearExtensions(); 2797 return this; 2798 } 2799 2800 /** 2801 * Returns an instance of <code>Locale</code> created from the fields set 2802 * on this builder. 2803 * 2804 * <p>This applies the conversions listed in {@link Locale#forLanguageTag} 2805 * when constructing a Locale. (Grandfathered tags are handled in 2806 * {@link #setLanguageTag}.) 2807 * 2808 * @return A Locale. 2809 */ 2810 public Locale build() { 2811 BaseLocale baseloc = localeBuilder.getBaseLocale(); 2812 LocaleExtensions extensions = localeBuilder.getLocaleExtensions(); 2813 if (extensions == null && baseloc.getVariant().length() > 0) { 2814 extensions = getCompatibilityExtensions(baseloc.getLanguage(), baseloc.getScript(), 2815 baseloc.getRegion(), baseloc.getVariant()); 2816 } 2817 return Locale.getInstance(baseloc, extensions); 2818 } 2819 } 2820 2821 /** 2822 * This enum provides constants to select a filtering mode for locale 2823 * matching. Refer to <a href="http://tools.ietf.org/html/rfc4647">RFC 4647 2824 * Matching of Language Tags</a> for details. 2825 * 2826 * <p>As an example, think of two Language Priority Lists each of which 2827 * includes only one language range and a set of following language tags: 2828 * 2829 * <pre> 2830 * de (German) 2831 * de-DE (German, Germany) 2832 * de-Deva (German, in Devanagari script) 2833 * de-Deva-DE (German, in Devanagari script, Germany) 2834 * de-DE-1996 (German, Germany, orthography of 1996) 2835 * de-Latn-DE (German, in Latin script, Germany) 2836 * de-Latn-DE-1996 (German, in Latin script, Germany, orthography of 1996) 2837 * </pre> 2838 * 2839 * The filtering method will behave as follows: 2840 * 2841 * <table cellpadding=2 summary="Filtering method behavior"> 2842 * <tr> 2843 * <th>Filtering Mode</th> 2844 * <th>Language Priority List: {@code "de-DE"}</th> 2845 * <th>Language Priority List: {@code "de-*-DE"}</th> 2846 * </tr> 2847 * <tr> 2848 * <td valign=top> 2849 * {@link FilteringMode#AUTOSELECT_FILTERING AUTOSELECT_FILTERING} 2850 * </td> 2851 * <td valign=top> 2852 * Performs <em>basic</em> filtering and returns {@code "de-DE"} and 2853 * {@code "de-DE-1996"}. 2854 * </td> 2855 * <td valign=top> 2856 * Performs <em>extended</em> filtering and returns {@code "de-DE"}, 2857 * {@code "de-Deva-DE"}, {@code "de-DE-1996"}, {@code "de-Latn-DE"}, and 2858 * {@code "de-Latn-DE-1996"}. 2859 * </td> 2860 * </tr> 2861 * <tr> 2862 * <td valign=top> 2863 * {@link FilteringMode#EXTENDED_FILTERING EXTENDED_FILTERING} 2864 * </td> 2865 * <td valign=top> 2866 * Performs <em>extended</em> filtering and returns {@code "de-DE"}, 2867 * {@code "de-Deva-DE"}, {@code "de-DE-1996"}, {@code "de-Latn-DE"}, and 2868 * {@code "de-Latn-DE-1996"}. 2869 * </td> 2870 * <td valign=top>Same as above.</td> 2871 * </tr> 2872 * <tr> 2873 * <td valign=top> 2874 * {@link FilteringMode#IGNORE_EXTENDED_RANGES IGNORE_EXTENDED_RANGES} 2875 * </td> 2876 * <td valign=top> 2877 * Performs <em>basic</em> filtering and returns {@code "de-DE"} and 2878 * {@code "de-DE-1996"}. 2879 * </td> 2880 * <td valign=top> 2881 * Performs <em>basic</em> filtering and returns {@code null} because 2882 * nothing matches. 2883 * </td> 2884 * </tr> 2885 * <tr> 2886 * <td valign=top> 2887 * {@link FilteringMode#MAP_EXTENDED_RANGES MAP_EXTENDED_RANGES} 2888 * </td> 2889 * <td valign=top>Same as above.</td> 2890 * <td valign=top> 2891 * Performs <em>basic</em> filtering and returns {@code "de-DE"} and 2892 * {@code "de-DE-1996"} because {@code "de-*-DE"} is mapped to 2893 * {@code "de-DE"}. 2894 * </td> 2895 * </tr> 2896 * <tr> 2897 * <td valign=top> 2898 * {@link FilteringMode#REJECT_EXTENDED_RANGES REJECT_EXTENDED_RANGES} 2899 * </td> 2900 * <td valign=top>Same as above.</td> 2901 * <td valign=top> 2902 * Throws {@link IllegalArgumentException} because {@code "de-*-DE"} is 2903 * not a valid basic language range. 2904 * </td> 2905 * </tr> 2906 * </table> 2907 * 2908 * @see #filter(List, Collection, FilteringMode) 2909 * @see #filterTags(List, Collection, FilteringMode) 2910 * 2911 * @since 1.8 2912 */ 2913 public static enum FilteringMode { 2914 /** 2915 * Specifies automatic filtering mode based on the given Language 2916 * Priority List consisting of language ranges. If all of the ranges 2917 * are basic, basic filtering is selected. Otherwise, extended 2918 * filtering is selected. 2919 */ 2920 AUTOSELECT_FILTERING, 2921 2922 /** 2923 * Specifies extended filtering. 2924 */ 2925 EXTENDED_FILTERING, 2926 2927 /** 2928 * Specifies basic filtering: Note that any extended language ranges 2929 * included in the given Language Priority List are ignored. 2930 */ 2931 IGNORE_EXTENDED_RANGES, 2932 2933 /** 2934 * Specifies basic filtering: If any extended language ranges are 2935 * included in the given Language Priority List, they are mapped to the 2936 * basic language range. Specifically, a language range starting with a 2937 * subtag {@code "*"} is treated as a language range {@code "*"}. For 2938 * example, {@code "*-US"} is treated as {@code "*"}. If {@code "*"} is 2939 * not the first subtag, {@code "*"} and extra {@code "-"} are removed. 2940 * For example, {@code "ja-*-JP"} is mapped to {@code "ja-JP"}. 2941 */ 2942 MAP_EXTENDED_RANGES, 2943 2944 /** 2945 * Specifies basic filtering: If any extended language ranges are 2946 * included in the given Language Priority List, the list is rejected 2947 * and the filtering method throws {@link IllegalArgumentException}. 2948 */ 2949 REJECT_EXTENDED_RANGES 2950 }; 2951 2952 /** 2953 * This class expresses a <em>Language Range</em> defined in 2954 * <a href="http://tools.ietf.org/html/rfc4647">RFC 4647 Matching of 2955 * Language Tags</a>. A language range is an identifier which is used to 2956 * select language tag(s) meeting specific requirements by using the 2957 * mechanisms described in <a href="Locale.html#LocaleMatching">Locale 2958 * Matching</a>. A list which represents a user's preferences and consists 2959 * of language ranges is called a <em>Language Priority List</em>. 2960 * 2961 * <p>There are two types of language ranges: basic and extended. In RFC 2962 * 4647, the syntax of language ranges is expressed in 2963 * <a href="http://tools.ietf.org/html/rfc4234">ABNF</a> as follows: 2964 * <blockquote> 2965 * <pre> 2966 * basic-language-range = (1*8ALPHA *("-" 1*8alphanum)) / "*" 2967 * extended-language-range = (1*8ALPHA / "*") 2968 * *("-" (1*8alphanum / "*")) 2969 * alphanum = ALPHA / DIGIT 2970 * </pre> 2971 * </blockquote> 2972 * For example, {@code "en"} (English), {@code "ja-JP"} (Japanese, Japan), 2973 * {@code "*"} (special language range which matches any language tag) are 2974 * basic language ranges, whereas {@code "*-CH"} (any languages, 2975 * Switzerland), {@code "es-*"} (Spanish, any regions), and 2976 * {@code "zh-Hant-*"} (Traditional Chinese, any regions) are extended 2977 * language ranges. 2978 * 2979 * @see #filter 2980 * @see #filterTags 2981 * @see #lookup 2982 * @see #lookupTag 2983 * 2984 * @since 1.8 2985 */ 2986 public static final class LanguageRange { 2987 2988 /** 2989 * A constant holding the maximum value of weight, 1.0, which indicates 2990 * that the language range is a good fit for the user. 2991 */ 2992 public static final double MAX_WEIGHT = 1.0; 2993 2994 /** 2995 * A constant holding the minimum value of weight, 0.0, which indicates 2996 * that the language range is not a good fit for the user. 2997 */ 2998 public static final double MIN_WEIGHT = 0.0; 2999 3000 private final String range; 3001 private final double weight; 3002 3003 private volatile int hash = 0; 3004 3005 /** 3006 * Constructs a {@code LanguageRange} using the given {@code range}. 3007 * Note that no validation is done against the IANA Language Subtag 3008 * Registry at time of construction. 3009 * 3010 * <p>This is equivalent to {@code LanguageRange(range, MAX_WEIGHT)}. 3011 * 3012 * @param range a language range 3013 * @throws NullPointerException if the given {@code range} is 3014 * {@code null} 3015 */ 3016 public LanguageRange(String range) { 3017 this(range, MAX_WEIGHT); 3018 } 3019 3020 /** 3021 * Constructs a {@code LanguageRange} using the given {@code range} and 3022 * {@code weight}. Note that no validation is done against the IANA 3023 * Language Subtag Registry at time of construction. 3024 * 3025 * @param range a language range 3026 * @param weight a weight value between {@code MIN_WEIGHT} and 3027 * {@code MAX_WEIGHT} 3028 * @throws NullPointerException if the given {@code range} is 3029 * {@code null} 3030 * @throws IllegalArgumentException if the given {@code weight} is less 3031 * than {@code MIN_WEIGHT} or greater than {@code MAX_WEIGHT} 3032 */ 3033 public LanguageRange(String range, double weight) { 3034 if (range == null) { 3035 throw new NullPointerException(); 3036 } 3037 if (weight < MIN_WEIGHT || weight > MAX_WEIGHT) { 3038 throw new IllegalArgumentException("weight=" + weight); 3039 } 3040 3041 range = range.toLowerCase(); 3042 3043 // Do syntax check. 3044 boolean isIllFormed = false; 3045 String[] subtags = range.split("-"); 3046 if (isSubtagIllFormed(subtags[0], true) 3047 || range.endsWith("-")) { 3048 isIllFormed = true; 3049 } else { 3050 for (int i = 1; i < subtags.length; i++) { 3051 if (isSubtagIllFormed(subtags[i], false)) { 3052 isIllFormed = true; 3053 break; 3054 } 3055 } 3056 } 3057 if (isIllFormed) { 3058 throw new IllegalArgumentException("range=" + range); 3059 } 3060 3061 this.range = range; 3062 this.weight = weight; 3063 } 3064 3065 private static boolean isSubtagIllFormed(String subtag, 3066 boolean isFirstSubtag) { 3067 if (subtag.equals("") || subtag.length() > 8) { 3068 return true; 3069 } else if (subtag.equals("*")) { 3070 return false; 3071 } 3072 char[] charArray = subtag.toCharArray(); 3073 if (isFirstSubtag) { // ALPHA 3074 for (char c : charArray) { 3075 if (c < 'a' || c > 'z') { 3076 return true; 3077 } 3078 } 3079 } else { // ALPHA / DIGIT 3080 for (char c : charArray) { 3081 if (c < '0' || (c > '9' && c < 'a') || c > 'z') { 3082 return true; 3083 } 3084 } 3085 } 3086 return false; 3087 } 3088 3089 /** 3090 * Returns the language range of this {@code LanguageRange}. 3091 * 3092 * @return the language range. 3093 */ 3094 public String getRange() { 3095 return range; 3096 } 3097 3098 /** 3099 * Returns the weight of this {@code LanguageRange}. 3100 * 3101 * @return the weight value. 3102 */ 3103 public double getWeight() { 3104 return weight; 3105 } 3106 3107 /** 3108 * Parses the given {@code ranges} to generate a Language Priority List. 3109 * 3110 * <p>This method performs a syntactic check for each language range in 3111 * the given {@code ranges} but doesn't do validation using the IANA 3112 * Language Subtag Registry. 3113 * 3114 * <p>The {@code ranges} to be given can take one of the following 3115 * forms: 3116 * 3117 * <pre> 3118 * "Accept-Language: ja,en;q=0.4" (weighted list with Accept-Language prefix) 3119 * "ja,en;q=0.4" (weighted list) 3120 * "ja,en" (prioritized list) 3121 * </pre> 3122 * 3123 * In a weighted list, each language range is given a weight value. 3124 * The weight value is identical to the "quality value" in 3125 * <a href="http://tools.ietf.org/html/rfc2616">RFC 2616</a>, and it 3126 * expresses how much the user prefers the language. A weight value is 3127 * specified after a corresponding language range followed by 3128 * {@code ";q="}, and the default weight value is {@code MAX_WEIGHT} 3129 * when it is omitted. 3130 * 3131 * <p>Unlike a weighted list, language ranges in a prioritized list 3132 * are sorted in the descending order based on its priority. The first 3133 * language range has the highest priority and meets the user's 3134 * preference most. 3135 * 3136 * <p>In either case, language ranges are sorted in descending order in 3137 * the Language Priority List based on priority or weight. If a 3138 * language range appears in the given {@code ranges} more than once, 3139 * only the first one is included on the Language Priority List. 3140 * 3141 * <p>The returned list consists of language ranges from the given 3142 * {@code ranges} and their equivalents found in the IANA Language 3143 * Subtag Registry. For example, if the given {@code ranges} is 3144 * {@code "Accept-Language: iw,en-us;q=0.7,en;q=0.3"}, the elements in 3145 * the list to be returned are: 3146 * 3147 * <pre> 3148 * <b>Range</b> <b>Weight</b> 3149 * "iw" (older tag for Hebrew) 1.0 3150 * "he" (new preferred code for Hebrew) 1.0 3151 * "en-us" (English, United States) 0.7 3152 * "en" (English) 0.3 3153 * </pre> 3154 * 3155 * Two language ranges, {@code "iw"} and {@code "he"}, have the same 3156 * highest priority in the list. By adding {@code "he"} to the user's 3157 * Language Priority List, locale-matching method can find Hebrew as a 3158 * matching locale (or language tag) even if the application or system 3159 * offers only {@code "he"} as a supported locale (or language tag). 3160 * 3161 * @param ranges a list of comma-separated language ranges or a list of 3162 * language ranges in the form of the "Accept-Language" header 3163 * defined in <a href="http://tools.ietf.org/html/rfc2616">RFC 3164 * 2616</a> 3165 * @return a Language Priority List consisting of language ranges 3166 * included in the given {@code ranges} and their equivalent 3167 * language ranges if available. The list is modifiable. 3168 * @throws NullPointerException if {@code ranges} is null 3169 * @throws IllegalArgumentException if a language range or a weight 3170 * found in the given {@code ranges} is ill-formed 3171 */ 3172 public static List<LanguageRange> parse(String ranges) { 3173 return LocaleMatcher.parse(ranges); 3174 } 3175 3176 /** 3177 * Parses the given {@code ranges} to generate a Language Priority 3178 * List, and then customizes the list using the given {@code map}. 3179 * This method is equivalent to 3180 * {@code mapEquivalents(parse(ranges), map)}. 3181 * 3182 * @param ranges a list of comma-separated language ranges or a list 3183 * of language ranges in the form of the "Accept-Language" header 3184 * defined in <a href="http://tools.ietf.org/html/rfc2616">RFC 3185 * 2616</a> 3186 * @param map a map containing information to customize language ranges 3187 * @return a Language Priority List with customization. The list is 3188 * modifiable. 3189 * @throws NullPointerException if {@code ranges} is null 3190 * @throws IllegalArgumentException if a language range or a weight 3191 * found in the given {@code ranges} is ill-formed 3192 * @see #parse(String) 3193 * @see #mapEquivalents 3194 */ 3195 public static List<LanguageRange> parse(String ranges, 3196 Map<String, List<String>> map) { 3197 return mapEquivalents(parse(ranges), map); 3198 } 3199 3200 /** 3201 * Generates a new customized Language Priority List using the given 3202 * {@code priorityList} and {@code map}. If the given {@code map} is 3203 * empty, this method returns a copy of the given {@code priorityList}. 3204 * 3205 * <p>In the map, a key represents a language range whereas a value is 3206 * a list of equivalents of it. {@code '*'} cannot be used in the map. 3207 * Each equivalent language range has the same weight value as its 3208 * original language range. 3209 * 3210 * <pre> 3211 * An example of map: 3212 * <b>Key</b> <b>Value</b> 3213 * "zh" (Chinese) "zh", 3214 * "zh-Hans"(Simplified Chinese) 3215 * "zh-HK" (Chinese, Hong Kong) "zh-HK" 3216 * "zh-TW" (Chinese, Taiwan) "zh-TW" 3217 * </pre> 3218 * 3219 * The customization is performed after modification using the IANA 3220 * Language Subtag Registry. 3221 * 3222 * <p>For example, if a user's Language Priority List consists of five 3223 * language ranges ({@code "zh"}, {@code "zh-CN"}, {@code "en"}, 3224 * {@code "zh-TW"}, and {@code "zh-HK"}), the newly generated Language 3225 * Priority List which is customized using the above map example will 3226 * consists of {@code "zh"}, {@code "zh-Hans"}, {@code "zh-CN"}, 3227 * {@code "zh-Hans-CN"}, {@code "en"}, {@code "zh-TW"}, and 3228 * {@code "zh-HK"}. 3229 * 3230 * <p>{@code "zh-HK"} and {@code "zh-TW"} aren't converted to 3231 * {@code "zh-Hans-HK"} nor {@code "zh-Hans-TW"} even if they are 3232 * included in the Language Priority List. In this example, mapping 3233 * is used to clearly distinguish Simplified Chinese and Traditional 3234 * Chinese. 3235 * 3236 * <p>If the {@code "zh"}-to-{@code "zh"} mapping isn't included in the 3237 * map, a simple replacement will be performed and the customized list 3238 * won't include {@code "zh"} and {@code "zh-CN"}. 3239 * 3240 * @param priorityList user's Language Priority List 3241 * @param map a map containing information to customize language ranges 3242 * @return a new Language Priority List with customization. The list is 3243 * modifiable. 3244 * @throws NullPointerException if {@code priorityList} is {@code null} 3245 * @see #parse(String, Map) 3246 */ 3247 public static List<LanguageRange> mapEquivalents( 3248 List<LanguageRange>priorityList, 3249 Map<String, List<String>> map) { 3250 return LocaleMatcher.mapEquivalents(priorityList, map); 3251 } 3252 3253 /** 3254 * Returns a hash code value for the object. 3255 * 3256 * @return a hash code value for this object. 3257 */ 3258 @Override 3259 public int hashCode() { 3260 if (hash == 0) { 3261 int result = 17; 3262 result = 37*result + range.hashCode(); 3263 long bitsWeight = Double.doubleToLongBits(weight); 3264 result = 37*result + (int)(bitsWeight ^ (bitsWeight >>> 32)); 3265 hash = result; 3266 } 3267 return hash; 3268 } 3269 3270 /** 3271 * Compares this object to the specified object. The result is true if 3272 * and only if the argument is not {@code null} and is a 3273 * {@code LanguageRange} object that contains the same {@code range} 3274 * and {@code weight} values as this object. 3275 * 3276 * @param obj the object to compare with 3277 * @return {@code true} if this object's {@code range} and 3278 * {@code weight} are the same as the {@code obj}'s; {@code false} 3279 * otherwise. 3280 */ 3281 @Override 3282 public boolean equals(Object obj) { 3283 if (this == obj) { 3284 return true; 3285 } 3286 if (!(obj instanceof LanguageRange)) { 3287 return false; 3288 } 3289 LanguageRange other = (LanguageRange)obj; 3290 return hash == other.hash 3291 && range.equals(other.range) 3292 && weight == other.weight; 3293 } 3294 } 3295 3296 /** 3297 * Returns a list of matching {@code Locale} instances using the filtering 3298 * mechanism defined in RFC 4647. 3299 * 3300 * @param priorityList user's Language Priority List in which each language 3301 * tag is sorted in descending order based on priority or weight 3302 * @param locales {@code Locale} instances used for matching 3303 * @param mode filtering mode 3304 * @return a list of {@code Locale} instances for matching language tags 3305 * sorted in descending order based on priority or weight, or an empty 3306 * list if nothing matches. The list is modifiable. 3307 * @throws NullPointerException if {@code priorityList} or {@code locales} 3308 * is {@code null} 3309 * @throws IllegalArgumentException if one or more extended language ranges 3310 * are included in the given list when 3311 * {@link FilteringMode#REJECT_EXTENDED_RANGES} is specified 3312 * 3313 * @since 1.8 3314 */ 3315 public static List<Locale> filter(List<LanguageRange> priorityList, 3316 Collection<Locale> locales, 3317 FilteringMode mode) { 3318 return LocaleMatcher.filter(priorityList, locales, mode); 3319 } 3320 3321 /** 3322 * Returns a list of matching {@code Locale} instances using the filtering 3323 * mechanism defined in RFC 4647. This is equivalent to 3324 * {@link #filter(List, Collection, FilteringMode)} when {@code mode} is 3325 * {@link FilteringMode#AUTOSELECT_FILTERING}. 3326 * 3327 * @param priorityList user's Language Priority List in which each language 3328 * tag is sorted in descending order based on priority or weight 3329 * @param locales {@code Locale} instances used for matching 3330 * @return a list of {@code Locale} instances for matching language tags 3331 * sorted in descending order based on priority or weight, or an empty 3332 * list if nothing matches. The list is modifiable. 3333 * @throws NullPointerException if {@code priorityList} or {@code locales} 3334 * is {@code null} 3335 * 3336 * @since 1.8 3337 */ 3338 public static List<Locale> filter(List<LanguageRange> priorityList, 3339 Collection<Locale> locales) { 3340 return filter(priorityList, locales, FilteringMode.AUTOSELECT_FILTERING); 3341 } 3342 3343 /** 3344 * Returns a list of matching languages tags using the basic filtering 3345 * mechanism defined in RFC 4647. 3346 * 3347 * @param priorityList user's Language Priority List in which each language 3348 * tag is sorted in descending order based on priority or weight 3349 * @param tags language tags 3350 * @param mode filtering mode 3351 * @return a list of matching language tags sorted in descending order 3352 * based on priority or weight, or an empty list if nothing matches. 3353 * The list is modifiable. 3354 * @throws NullPointerException if {@code priorityList} or {@code tags} is 3355 * {@code null} 3356 * @throws IllegalArgumentException if one or more extended language ranges 3357 * are included in the given list when 3358 * {@link FilteringMode#REJECT_EXTENDED_RANGES} is specified 3359 * 3360 * @since 1.8 3361 */ 3362 public static List<String> filterTags(List<LanguageRange> priorityList, 3363 Collection<String> tags, 3364 FilteringMode mode) { 3365 return LocaleMatcher.filterTags(priorityList, tags, mode); 3366 } 3367 3368 /** 3369 * Returns a list of matching languages tags using the basic filtering 3370 * mechanism defined in RFC 4647. This is equivalent to 3371 * {@link #filterTags(List, Collection, FilteringMode)} when {@code mode} 3372 * is {@link FilteringMode#AUTOSELECT_FILTERING}. 3373 * 3374 * @param priorityList user's Language Priority List in which each language 3375 * tag is sorted in descending order based on priority or weight 3376 * @param tags language tags 3377 * @return a list of matching language tags sorted in descending order 3378 * based on priority or weight, or an empty list if nothing matches. 3379 * The list is modifiable. 3380 * @throws NullPointerException if {@code priorityList} or {@code tags} is 3381 * {@code null} 3382 * 3383 * @since 1.8 3384 */ 3385 public static List<String> filterTags(List<LanguageRange> priorityList, 3386 Collection<String> tags) { 3387 return filterTags(priorityList, tags, FilteringMode.AUTOSELECT_FILTERING); 3388 } 3389 3390 /** 3391 * Returns a {@code Locale} instance for the best-matching language 3392 * tag using the lookup mechanism defined in RFC 4647. 3393 * 3394 * @param priorityList user's Language Priority List in which each language 3395 * tag is sorted in descending order based on priority or weight 3396 * @param locales {@code Locale} instances used for matching 3397 * @return the best matching <code>Locale</code> instance chosen based on 3398 * priority or weight, or {@code null} if nothing matches. 3399 * @throws NullPointerException if {@code priorityList} or {@code tags} is 3400 * {@code null} 3401 * 3402 * @since 1.8 3403 */ 3404 public static Locale lookup(List<LanguageRange> priorityList, 3405 Collection<Locale> locales) { 3406 return LocaleMatcher.lookup(priorityList, locales); 3407 } 3408 3409 /** 3410 * Returns the best-matching language tag using the lookup mechanism 3411 * defined in RFC 4647. 3412 * 3413 * @param priorityList user's Language Priority List in which each language 3414 * tag is sorted in descending order based on priority or weight 3415 * @param tags language tangs used for matching 3416 * @return the best matching language tag chosen based on priority or 3417 * weight, or {@code null} if nothing matches. 3418 * @throws NullPointerException if {@code priorityList} or {@code tags} is 3419 * {@code null} 3420 * 3421 * @since 1.8 3422 */ 3423 public static String lookupTag(List<LanguageRange> priorityList, 3424 Collection<String> tags) { 3425 return LocaleMatcher.lookupTag(priorityList, tags); 3426 } 3427 3428} 3429