1 /* 2 ****************************************************************************************** 3 * Copyright (C) 2009-2014, Google, Inc.; International Business Machines Corporation and * 4 * others. All Rights Reserved. * 5 ****************************************************************************************** 6 */ 7 8 package org.unicode.cldr.unittest; 9 10 import java.util.Set; 11 import java.util.TreeSet; 12 13 import com.ibm.icu.dev.test.TestFmwk; 14 import com.ibm.icu.util.LocaleMatcher; 15 import com.ibm.icu.util.LocaleMatcher.LanguageMatcherData; 16 import com.ibm.icu.util.LocalePriorityList; 17 import com.ibm.icu.util.ULocale; 18 19 /** 20 * Test the LocaleMatcher. 21 * 22 * @author markdavis 23 */ 24 @SuppressWarnings("deprecation") 25 public class LocaleMatcherTest extends TestFmwk { 26 27 private static final ULocale ZH_MO = new ULocale("zh_MO"); 28 private static final ULocale ZH_HK = new ULocale("zh_HK"); 29 static LanguageMatcherData LANGUAGE_MATCHER_DATA = LocaleMatcherShim.load(); 30 31 private LocaleMatcher newLocaleMatcher(LocalePriorityList build) { 32 return new LocaleMatcher(build, LANGUAGE_MATCHER_DATA); 33 } 34 35 private LocaleMatcher newLocaleMatcher(LocalePriorityList build, LanguageMatcherData data) { 36 return new LocaleMatcher(build, data == null ? LANGUAGE_MATCHER_DATA : data); 37 } 38 39 private LocaleMatcher newLocaleMatcher(LocalePriorityList lpl, LanguageMatcherData data, double d) { 40 return new LocaleMatcher(lpl, data == null ? LANGUAGE_MATCHER_DATA : data, d); 41 } 42 43 private LocaleMatcher newLocaleMatcher(String string) { 44 return new LocaleMatcher(LocalePriorityList.add(string).build(), LANGUAGE_MATCHER_DATA); 45 } 46 47 // public LocaleMatcher(LocalePriorityList languagePriorityList, 48 // LocaleMatcherData matcherData, double threshold) 49 50 public static void main(String[] args) throws Exception { 51 new LocaleMatcherTest().run(args); 52 } 53 54 public void testParentLocales() { 55 if (logKnownIssue("ICU-21241", "waiting on LocaleMatcherData update")) { 56 return; 57 } 58 59 assertCloser("es_AR", "es_419", "es_ES"); 60 assertCloser("es_AR", "es_419", "es"); 61 62 assertCloser("es_AR", "es_MX", "es"); 63 assertCloser("es_AR", "es_MX", "es"); 64 65 assertCloser("en_AU", "en_GB", "en_US"); 66 assertCloser("en_AU", "en_GB", "en"); 67 68 assertCloser("en_AU", "en_NZ", "en_US"); 69 assertCloser("en_AU", "en_NZ", "en"); 70 71 assertCloser("pt_AO", "pt_PT", "pt_BR"); 72 assertCloser("pt_AO", "pt_PT", "pt"); 73 74 assertCloser("zh_HK", "zh_MO", "zh_TW"); 75 assertCloser("zh_HK", "zh_MO", "zh_CN"); 76 assertCloser("zh_HK", "zh_MO", "zh"); 77 } 78 79 private void assertCloser(String a, String closer, String further) { 80 LocaleMatcher matcher = newLocaleMatcher(further + ", " + closer); 81 assertEquals("test " + a + " is closer to " + closer + " than to " + further, new ULocale(closer), matcher.getBestMatch(a)); 82 matcher = newLocaleMatcher(closer + ", " + further); 83 assertEquals("test " + a + " is closer to " + closer + " than to " + further, new ULocale(closer), matcher.getBestMatch(a)); 84 } 85 86 // public void testParentLocales() { 87 // // find all the regions that have a closer relation because of an explicit parent 88 // Set<String> explicitParents = new HashSet<>(INFO.getExplicitParents()); 89 // explicitParents.remove("root"); 90 // Set<String> otherParents = new HashSet<>(INFO.getExplicitParents()); 91 // for (String locale : explicitParents) { 92 // while (true) { 93 // locale = LocaleIDParser.getParent(locale); 94 // if (locale == null || locale.equals("root")) { 95 // break; 96 // } 97 // otherParents.add(locale); 98 // } 99 // } 100 // otherParents.remove("root"); 101 // 102 // for (String locale : CONFIG.getCldrFactory().getAvailable()) { 103 // String parentId = LocaleIDParser.getParent(locale); 104 // String parentIdSimple = LocaleIDParser.getSimpleParent(locale); 105 // if (!explicitParents.contains(parentId) && !otherParents.contains(parentIdSimple)) { 106 // continue; 107 // } 108 // System.out.println(locale + "\t" + CONFIG.getEnglish().getName(locale) + "\t" + parentId + "\t" + parentIdSimple); 109 // } 110 // } 111 112 public void testChinese() { 113 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 114 return; 115 } 116 LocaleMatcher matcher = newLocaleMatcher("zh_CN, zh_TW, iw"); 117 ULocale taiwanChinese = new ULocale("zh_TW"); 118 ULocale chinaChinese = new ULocale("zh_CN"); 119 assertEquals("zh_CN, zh_TW, iw;", taiwanChinese, matcher.getBestMatch("zh_Hant_TW")); 120 assertEquals("zh_CN, zh_TW, iw;", taiwanChinese, matcher.getBestMatch("zh_Hant")); 121 assertEquals("zh_CN, zh_TW, iw;", taiwanChinese, matcher.getBestMatch("zh_TW")); 122 assertEquals("zh_CN, zh_TW, iw;", chinaChinese, matcher.getBestMatch("zh_Hans_CN")); 123 assertEquals("zh_CN, zh_TW, iw;", chinaChinese, matcher.getBestMatch("zh_CN")); 124 assertEquals("zh_CN, zh_TW, iw;", chinaChinese, matcher.getBestMatch("zh")); 125 assertEquals("zh_CN, zh_TW, iw;", taiwanChinese, matcher.getBestMatch("zh_Hant_HK")); 126 } 127 128 public void testenGB() { 129 if (logKnownIssue("ICU-21241", "waiting on LocaleMatcherData update")) { 130 return; 131 } 132 133 final LocaleMatcher matcher = newLocaleMatcher("fr, en, en_GB, es_MX, es_419, es"); 134 assertEquals("en_GB", matcher.getBestMatch("en_NZ").toString()); 135 assertEquals("es", matcher.getBestMatch("es_ES").toString()); 136 assertEquals("es_419", matcher.getBestMatch("es_AR").toString()); 137 assertEquals("es_MX", matcher.getBestMatch("es_MX").toString()); 138 } 139 140 public void testFallbacks() { 141 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 142 return; 143 } 144 LocalePriorityList lpl = LocalePriorityList.add("en, hi").build(); 145 final LocaleMatcher matcher = newLocaleMatcher(lpl, null, 0.09); 146 assertEquals("hi", matcher.getBestMatch("sa").toString()); 147 } 148 149 public void testOverrideData() { 150 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 151 return; 152 } 153 double threshold = 0.05; 154 LanguageMatcherData localeMatcherData = new LanguageMatcherData() 155 .addDistance("br", "fr", 10, true) 156 .addDistance("es", "cy", 10, true); 157 logln(localeMatcherData.toString()); 158 159 final LocaleMatcher matcher = newLocaleMatcher( 160 LocalePriorityList 161 .add(ULocale.ENGLISH) 162 .add(ULocale.FRENCH) 163 .add(ULocale.UK) 164 .build(), 165 localeMatcherData, threshold); 166 logln(matcher.toString()); 167 168 assertEquals(ULocale.FRENCH, matcher.getBestMatch(new ULocale("br"))); 169 assertEquals(ULocale.ENGLISH, matcher.getBestMatch(new ULocale("es"))); // one 170 // way 171 } 172 173 public void testBasics() { 174 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 175 return; 176 } 177 final LocaleMatcher matcher = newLocaleMatcher(LocalePriorityList.add(ULocale.FRENCH).add(ULocale.UK) 178 .add(ULocale.ENGLISH).build()); 179 logln(matcher.toString()); 180 181 assertEquals(ULocale.UK, matcher.getBestMatch(ULocale.UK)); 182 assertEquals(ULocale.ENGLISH, matcher.getBestMatch(ULocale.US)); 183 assertEquals(ULocale.FRENCH, matcher.getBestMatch(ULocale.FRANCE)); 184 assertEquals(ULocale.FRENCH, matcher.getBestMatch(ULocale.JAPAN)); 185 } 186 187 public void testFallback() { 188 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 189 return; 190 } 191 // check that script fallbacks are handled right 192 final LocaleMatcher matcher = newLocaleMatcher("zh_CN, zh_TW, iw"); 193 assertEquals(new ULocale("zh_TW"), matcher.getBestMatch("zh_Hant")); 194 assertEquals(new ULocale("zh_CN"), matcher.getBestMatch("zh")); 195 assertEquals(new ULocale("zh_CN"), matcher.getBestMatch("zh_Hans_CN")); 196 assertEquals(new ULocale("zh_TW"), matcher.getBestMatch("zh_Hant_HK")); 197 assertEquals(new ULocale("he"), matcher.getBestMatch("iw_IT")); 198 } 199 200 public void testSpecials() { 201 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 202 return; 203 } 204 // check that nearby languages are handled 205 final LocaleMatcher matcher = newLocaleMatcher("en, fil, ro, nn"); 206 assertEquals(new ULocale("fil"), matcher.getBestMatch("tl")); 207 assertEquals(new ULocale("ro"), matcher.getBestMatch("mo")); 208 assertEquals(new ULocale("nn"), matcher.getBestMatch("nb")); 209 // make sure default works 210 assertEquals(new ULocale("en"), matcher.getBestMatch("ja")); 211 } 212 213 public void testRegionalSpecials() { 214 if (logKnownIssue("ICU-21241", "waiting on LocaleMatcherData update")) { 215 return; 216 } 217 218 // verify that en_AU is closer to en_GB than to en (which is en_US) 219 final LocaleMatcher matcher = newLocaleMatcher("en, en_GB, es, es_419"); 220 assertEquals("es_MX in {en, en_GB, es, es_419}", new ULocale("es_419"), matcher.getBestMatch("es_MX")); 221 assertEquals("en_AU in {en, en_GB, es, es_419}", new ULocale("en_GB"), matcher.getBestMatch("en_AU")); 222 assertEquals("es_ES in {en, en_GB, es, es_419}", new ULocale("es"), matcher.getBestMatch("es_ES")); 223 } 224 225 public void testHK() { 226 if (logKnownIssue("ICU-21241", "waiting on LocaleMatcherData update")) { 227 return; 228 } 229 230 // HK and MO are closer to each other for Hant than to TW 231 final LocaleMatcher matcher = newLocaleMatcher("zh, zh_TW, zh_MO"); 232 assertEquals("zh_HK in {zh, zh_TW, zh_MO}", ZH_MO, matcher.getBestMatch("zh_HK")); 233 final LocaleMatcher matcher2 = newLocaleMatcher("zh, zh_TW, zh_HK"); 234 assertEquals("zh_MO in {zh, zh_TW, zh_HK}", ZH_HK, matcher2.getBestMatch("zh_MO")); 235 } 236 237 public void TestLocaleMatcherCoverage() { 238 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 239 return; 240 } 241 // Add tests for better code coverage 242 LocaleMatcher matcher = newLocaleMatcher(LocalePriorityList.add(null, 0).build(), null); 243 logln(matcher.toString()); 244 245 LanguageMatcherData data = new LanguageMatcherData(); 246 247 LanguageMatcherData clone = data.cloneAsThawed(); 248 249 if (clone.equals(data)) { 250 errln("Error cloneAsThawed() is equal."); 251 } 252 253 if (data.isFrozen()) { 254 errln("Error LocaleMatcherData is frozen!"); 255 } 256 } 257 258 private void assertEquals(Object expected, Object string) { 259 assertEquals("", expected, string); 260 } 261 262 private void assertNull(Object bestMatch) { 263 assertNull("", bestMatch); 264 } 265 266 public void testEmpty() { 267 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 268 return; 269 } 270 final LocaleMatcher matcher = newLocaleMatcher(""); 271 assertNull(matcher.getBestMatch(ULocale.FRENCH)); 272 } 273 274 static final ULocale ENGLISH_CANADA = new ULocale("en_CA"); 275 276 public void testMatch_exact() { 277 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 278 return; 279 } 280 assertEquals(1.0, 281 LocaleMatcher.match(ENGLISH_CANADA, ENGLISH_CANADA)); 282 } 283 284 public void testMatch_none() { 285 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 286 return; 287 } 288 double match = LocaleMatcher.match( 289 new ULocale("ar_MK"), 290 ENGLISH_CANADA); 291 assertTrue("Actual < 0: " + match, 0 <= match); 292 assertTrue("Actual > 0.15 (~ language + script distance): " + match, 0.2 > match); 293 } 294 295 public void testMatch_matchOnMazimized() { 296 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 297 return; 298 } 299 ULocale undTw = new ULocale("und_TW"); 300 ULocale zhHant = new ULocale("zh_Hant"); 301 double matchZh = LocaleMatcher.match(undTw, new ULocale("zh")); 302 double matchZhHant = LocaleMatcher.match(undTw, zhHant); 303 assertTrue("und_TW should be closer to zh_Hant (" + matchZhHant + 304 ") than to zh (" + matchZh + ")", 305 matchZh < matchZhHant); 306 double matchEnHantTw = LocaleMatcher.match(new ULocale("en_Hant_TW"), 307 zhHant); 308 assertTrue("zh_Hant should be closer to und_TW (" + matchZhHant + 309 ") than to en_Hant_TW (" + matchEnHantTw + ")", 310 matchEnHantTw < matchZhHant); 311 assertTrue("zh should be closer to und_TW (" + matchZh + 312 ") than to en_Hant_TW (" + matchEnHantTw + ")", 313 matchEnHantTw < matchZh); 314 } 315 316 public void testMatchLegacyCode() { 317 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 318 return; 319 } 320 final LocaleMatcher matcher = newLocaleMatcher("fr, i_klingon, en_Latn_US"); 321 assertEquals("en_Latn_US", matcher.getBestMatch("en_GB_oed").toString()); 322 // assertEquals("tlh", matcher.getBestMatch("i_klingon").toString()); 323 } 324 325 public void testGetBestMatchForList_exactMatch() { 326 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 327 return; 328 } 329 final LocaleMatcher matcher = newLocaleMatcher("fr, en_GB, ja, es_ES, es_MX"); 330 assertEquals("ja", matcher.getBestMatch("ja, de").toString()); 331 } 332 333 public void testGetBestMatchForList_simpleVariantMatch() { 334 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 335 return; 336 } 337 final LocaleMatcher matcher = newLocaleMatcher("fr, en_GB, ja, es_ES, es_MX"); 338 // Intentionally avoiding a perfect_match or two candidates for variant 339 // matches. 340 assertEquals("en_GB", matcher.getBestMatch("de, en_US").toString()); 341 // Fall back. 342 assertEquals("fr", matcher.getBestMatch("de, zh").toString()); 343 } 344 345 public void testGetBestMatchForList_matchOnMaximized() { 346 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 347 return; 348 } 349 final LocaleMatcher matcher = newLocaleMatcher("en, ja"); 350 // final LocaleMatcher matcher = 351 // newLocaleMatcher("fr, en, ja, es_ES, es_MX"); 352 // Check that if the preference is maximized already, it works as well. 353 assertEquals("Match for ja_Jpan_JP (maximized already)", 354 "ja", matcher.getBestMatch("ja_Jpan_JP, en-AU").toString()); 355 if (true) 356 return; 357 // ja_JP matches ja on likely subtags, and it's listed first, thus it 358 // wins over 359 // thus it wins over the second preference en_GB. 360 assertEquals("Match for ja_JP, with likely region subtag", 361 "ja", matcher.getBestMatch("ja_JP, en_US").toString()); 362 // Check that if the preference is maximized already, it works as well. 363 assertEquals("Match for ja_Jpan_JP (maximized already)", 364 "ja", matcher.getBestMatch("ja_Jpan_JP, en_US").toString()); 365 } 366 367 public void testGetBestMatchForList_noMatchOnMaximized() { 368 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 369 return; 370 } 371 // Regression test for http://b/5714572 . 372 final LocaleMatcher matcher = newLocaleMatcher("en, de, fr, ja"); 373 // de maximizes to de_DE. Pick the exact match for the secondary 374 // language instead. 375 assertEquals("de", matcher.getBestMatch("de_CH, fr").toString()); 376 } 377 378 public void testBestMatchForTraditionalChinese() { 379 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 380 return; 381 } 382 // Scenario: An application that only supports Simplified Chinese (and 383 // some other languages), 384 // but does not support Traditional Chinese. zh_Hans_CN could be 385 // replaced with zh_CN, zh, or 386 // zh_Hans, it wouldn't make much of a difference. 387 final LocaleMatcher matcher = newLocaleMatcher("fr, zh_Hans_CN, en_US"); 388 389 // The script distance (simplified vs. traditional Han) is considered 390 // small enough 391 // to be an acceptable match. The regional difference is considered 392 // almost insignificant. 393 assertEquals("zh_Hans_CN", matcher.getBestMatch("zh_TW").toString()); 394 assertEquals("zh_Hans_CN", matcher.getBestMatch("zh_Hant").toString()); 395 396 // For geo_political reasons, you might want to avoid a zh_Hant -> 397 // zh_Hans match. 398 // In this case, if zh_TW, zh_HK or a tag starting with zh_Hant is 399 // requested, you can 400 // change your call to getBestMatch to include a 2nd language 401 // preference. 402 // "en" is a better match since its distance to "en_US" is closer than 403 // the distance 404 // from "zh_TW" to "zh_CN" (script distance). 405 assertEquals("en_US", matcher.getBestMatch("zh_TW, en").toString()); 406 assertEquals("en_US", matcher.getBestMatch("zh_Hant_CN, en").toString()); 407 assertEquals("zh_Hans_CN", matcher.getBestMatch("zh_Hans, en").toString()); 408 } 409 410 public void testUndefined() { 411 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 412 return; 413 } 414 // When the undefined language doesn't match anything in the list, 415 // getBestMatch returns 416 // the default, as usual. 417 LocaleMatcher matcher = newLocaleMatcher("it,fr"); 418 assertEquals("it", matcher.getBestMatch("und").toString()); 419 420 // When it *does* occur in the list, BestMatch returns it, as expected. 421 // Note that as of ICU 64, "und" is normalized to "" 422 matcher = newLocaleMatcher("it,und"); 423 assertEquals("", matcher.getBestMatch("und").toString()); 424 425 // The unusual part: 426 // max("und") = "en_Latn_US", and since matching is based on maximized 427 // tags, the undefined 428 // language would normally match English. But that would produce the 429 // counterintuitive results 430 // that getBestMatch("und", LocaleMatcher("it,en")) would be "en", and 431 // getBestMatch("en", LocaleMatcher("it,und")) would be "und". 432 // 433 // To avoid that, we change the matcher's definitions of max 434 // (AddLikelySubtagsWithDefaults) 435 // so that max("und")="und". That produces the following, more desirable 436 // results: 437 matcher = newLocaleMatcher("it,en"); 438 assertEquals("it", matcher.getBestMatch("und").toString()); 439 matcher = newLocaleMatcher("it,und"); 440 assertEquals("it", matcher.getBestMatch("en").toString()); 441 } 442 443 // public void testGetBestMatch_emptyList() { 444 // final LocaleMatcher matcher = newLocaleMatcher( 445 // new LocalePriorityList(new HashMap())); 446 // assertNull(matcher.getBestMatch(ULocale.ENGLISH)); 447 // } 448 449 public void testGetBestMatch_googlePseudoLocales() { 450 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 451 return; 452 } 453 // Google pseudo locales are primarily based on variant subtags. 454 // See http://sites/intl_eng/pseudo_locales. 455 // (See below for the region code based fall back options.) 456 final LocaleMatcher matcher = newLocaleMatcher( 457 "fr, pt"); 458 assertEquals("fr", matcher.getBestMatch("de").toString()); 459 assertEquals("fr", matcher.getBestMatch("en_US").toString()); 460 assertEquals("fr", matcher.getBestMatch("en").toString()); 461 assertEquals("pt", matcher.getBestMatch("pt_BR").toString()); 462 } 463 464 public void testGetBestMatch_regionDistance() { 465 if (logKnownIssue("ICU-21241", "waiting on LocaleMatcherData update")) { 466 return; 467 } 468 469 LocaleMatcher matcher = newLocaleMatcher("es_AR, es"); 470 assertEquals("es_AR", matcher.getBestMatch("es_MX").toString()); 471 472 matcher = newLocaleMatcher("fr, en, en_GB"); 473 assertEquals("en_GB", matcher.getBestMatch("en_CA").toString()); 474 475 matcher = newLocaleMatcher("de_AT, de_DE, de_CH"); 476 assertEquals("de_DE", matcher.getBestMatch("de").toString()); 477 478 showDistance(matcher, "en", "en_CA"); 479 showDistance(matcher, "en_CA", "en"); 480 showDistance(matcher, "en_US", "en_CA"); 481 showDistance(matcher, "en_CA", "en_US"); 482 showDistance(matcher, "en_GB", "en_CA"); 483 showDistance(matcher, "en_CA", "en_GB"); 484 showDistance(matcher, "en", "en_UM"); 485 showDistance(matcher, "en_UM", "en"); 486 } 487 488 private void showDistance(LocaleMatcher matcher, String desired, String supported) { 489 ULocale desired2 = new ULocale(desired); 490 ULocale supported2 = new ULocale(supported); 491 double distance = matcher.match(desired2, ULocale.addLikelySubtags(desired2), supported2, ULocale.addLikelySubtags(supported2)); 492 logln(desired + " to " + supported + " :\t" + distance); 493 } 494 495 /** 496 * If all the base languages are the same, then each sublocale matches 497 * itself most closely 498 */ 499 public void testExactMatches() { 500 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 501 return; 502 } 503 String lastBase = ""; 504 TreeSet<ULocale> sorted = new TreeSet<>(); 505 for (ULocale loc : ULocale.getAvailableLocales()) { 506 String language = loc.getLanguage(); 507 if (!lastBase.equals(language)) { 508 check(sorted); 509 sorted.clear(); 510 lastBase = language; 511 } 512 sorted.add(loc); 513 } 514 check(sorted); 515 } 516 517 private void check(Set<ULocale> sorted) { 518 if (sorted.isEmpty()) { 519 return; 520 } 521 check2(sorted); 522 ULocale first = sorted.iterator().next(); 523 ULocale max = ULocale.addLikelySubtags(first); 524 sorted.add(max); 525 check2(sorted); 526 } 527 528 /** 529 * @param sorted 530 */ 531 private void check2(Set<ULocale> sorted) { 532 // TODO Auto-generated method stub 533 logln("Checking: " + sorted); 534 LocaleMatcher matcher = newLocaleMatcher( 535 LocalePriorityList.add( 536 sorted.toArray(new ULocale[sorted.size()])) 537 .build()); 538 for (ULocale loc : sorted) { 539 String stringLoc = loc.toString(); 540 assertEquals(stringLoc, matcher.getBestMatch(stringLoc).toString()); 541 } 542 } 543 544 // public void testComputeDistance_monkeyTest() { 545 // RegionCode[] codes = RegionCode.values(); 546 // Random random = new Random(); 547 // for (int i = 0; i < 1000; ++i) { 548 // RegionCode x = codes[random.nextInt(codes.length)]; 549 // RegionCode y = codes[random.nextInt(codes.length)]; 550 // double d = LocaleMatcher.getRegionDistance(x, y, null, null); 551 // if (x == RegionCode.ZZ || y == RegionCode.ZZ) { 552 // assertEquals(LocaleMatcher.REGION_DISTANCE, d); 553 // } else if (x == y) { 554 // assertEquals(0.0, d); 555 // } else { 556 // assertTrue(d > 0); 557 // assertTrue(d <= LocaleMatcher.REGION_DISTANCE); 558 // } 559 // } 560 // } 561 562 public void testGetBestMatchForList_matchOnMaximized2() { 563 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 564 return; 565 } 566 final LocaleMatcher matcher = newLocaleMatcher("fr, en-GB, ja, es-ES, es-MX"); 567 // ja-JP matches ja on likely subtags, and it's listed first, thus it wins over 568 // thus it wins over the second preference en-GB. 569 assertEquals("Match for ja-JP, with likely region subtag", 570 "ja", matcher.getBestMatch("ja-JP, en-GB").toString()); 571 // Check that if the preference is maximized already, it works as well. 572 assertEquals("Match for ja-Jpan-JP (maximized already)", 573 "ja", matcher.getBestMatch("ja-Jpan-JP, en-GB").toString()); 574 } 575 576 public void testGetBestMatchForList_closeEnoughMatchOnMaximized() { 577 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 578 return; 579 } 580 final LocaleMatcher matcher = newLocaleMatcher("en-GB, en, de, fr, ja"); 581 assertEquals("de", matcher.getBestMatch("de-CH, fr").toString()); 582 assertEquals("en", matcher.getBestMatch("en-US, ar, nl, de, ja").toString()); 583 } 584 585 public void testGetBestMatchForPortuguese() { 586 if (logKnownIssue("ICU-21241", "waiting on LocaleMatcherData update")) { 587 return; 588 } 589 final LocaleMatcher withPTExplicit = newLocaleMatcher("pt_PT, pt_BR, es, es_419"); 590 final LocaleMatcher withPTImplicit = newLocaleMatcher("pt_PT, pt, es, es_419"); 591 // Could happen because "pt_BR" is a tier_1 language and "pt_PT" is tier_2. 592 593 final LocaleMatcher withoutPT = newLocaleMatcher("pt_BR, es, es_419"); 594 // European user who prefers Spanish over Brazillian Portuguese as a fallback. 595 596 assertEquals("pt_PT", withPTExplicit.getBestMatch("pt_PT, es, pt").toString()); 597 assertEquals("pt_PT", withPTImplicit.getBestMatch("pt_PT, es, pt").toString()); 598 assertEquals("es", withoutPT.getBestMatch("pt_PT, es, pt").toString()); 599 600 // Brazillian user who prefers South American Spanish over European Portuguese as a fallback. 601 // The asymmetry between this case and above is because it's "pt_PT" that's missing between the 602 // matchers as "pt_BR" is a much more common language. 603 assertEquals("pt_BR", withPTExplicit.getBestMatch("pt, es_419, pt_PT").toString()); 604 assertEquals("pt", withPTImplicit.getBestMatch("pt, es_419, pt_PT").toString()); 605 assertEquals("pt_BR", withoutPT.getBestMatch("pt, es_419, pt_PT").toString()); 606 607 // Code that adds the user's country can get "pt_US" for a user's language. 608 // That should fall back to "pt_BR". 609 assertEquals("pt_BR", withPTExplicit.getBestMatch("pt_US, pt_PT").toString()); 610 assertEquals("pt", withPTImplicit.getBestMatch("pt_US, pt_PT").toString()); 611 } 612 613 public void testVariantWithScriptMatch() { 614 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 615 return; 616 } 617 final LocaleMatcher matcher = newLocaleMatcher("fr, en, sv"); 618 assertEquals("en", matcher.getBestMatch("en-GB").toString()); 619 assertEquals("en", matcher.getBestMatch("en-GB, sv").toString()); 620 } 621 622 public void testVariantWithScriptMatch2() { 623 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 624 return; 625 } 626 final LocaleMatcher matcher = newLocaleMatcher("en, sv"); 627 assertEquals("en", matcher.getBestMatch("en-GB, sv").toString()); 628 } 629 630 public void testPerf() { 631 if (logKnownIssue("CLDR-14166", "Skip until CLDR updated for new ICU4J LocaleMatcher")) { 632 return; 633 } 634 final String desired = "sv, en"; 635 636 final LocaleMatcher matcherShort = newLocaleMatcher(desired); 637 final LocaleMatcher matcherLong = newLocaleMatcher( 638 "af, am, ar, az, be, bg, bn, bs, ca, cs, cy, cy, da, de, el, en, en-GB, es, es-419, et, eu, fa, fi, fil, fr, ga, gl, gu, hi, hr, hu, hy, id, is, it, iw, ja, ka, kk, km, kn, ko, ky, lo, lt, lv, mk, ml, mn, mr, ms, my, ne, nl, no, pa, pl, pt, pt-PT, ro, ru, si, sk, sl, sq, sr, sr-Latn, sv, sw, ta, te, th, tr, uk, ur, uz, vi, zh-CN, zh-TW, zu"); 639 final LocaleMatcher matcherVeryLong = newLocaleMatcher( 640 "af, af_NA, af_ZA, agq, agq_CM, ak, ak_GH, am, am_ET, ar, ar_001, ar_AE, ar_BH, ar_DJ, ar_DZ, ar_EG, ar_EH, ar_ER, ar_IL, ar_IQ, ar_JO, ar_KM, ar_KW, ar_LB, ar_LY, ar_MA, ar_MR, ar_OM, ar_PS, ar_QA, ar_SA, ar_SD, ar_SO, ar_SS, ar_SY, ar_TD, ar_TN, ar_YE, as, as_IN, asa, asa_TZ, ast, ast_ES, az, az_Cyrl, az_Cyrl_AZ, az_Latn, az_Latn_AZ, bas, bas_CM, be, be_BY, bem, bem_ZM, bez, bez_TZ, bg, bg_BG, bm, bm_ML, bn, bn_BD, bn_IN, bo, bo_CN, bo_IN, br, br_FR, brx, brx_IN, bs, bs_Cyrl, bs_Cyrl_BA, bs_Latn, bs_Latn_BA, ca, ca_AD, ca_ES, ca_ES_VALENCIA, ca_FR, ca_IT, ce, ce_RU, cgg, cgg_UG, chr, chr_US, ckb, ckb_IQ, ckb_IR, cs, cs_CZ, cu, cu_RU, cy, cy_GB, da, da_DK, da_GL, dav, dav_KE, de, de_AT, de_BE, de_CH, de_DE, de_LI, de_LU, dje, dje_NE, dsb, dsb_DE, dua, dua_CM, dyo, dyo_SN, dz, dz_BT, ebu, ebu_KE, ee, ee_GH, ee_TG, el, el_CY, el_GR, en, en_001, en_150, en_AG, en_AI, en_AS, en_AT, en_AU, en_BB, en_BE, en_BI, en_BM, en_BS, en_BW, en_BZ, en_CA, en_CC, en_CH, en_CK, en_CM, en_CX, en_CY, en_DE, en_DG, en_DK, en_DM, en_ER, en_FI, en_FJ, en_FK, en_FM, en_GB, en_GD, en_GG, en_GH, en_GI, en_GM, en_GU, en_GY, en_HK, en_IE, en_IL, en_IM, en_IN, en_IO, en_JE, en_JM, en_KE, en_KI, en_KN, en_KY, en_LC, en_LR, en_LS, en_MG, en_MH, en_MO, en_MP, en_MS, en_MT, en_MU, en_MW, en_MY, en_NA, en_NF, en_NG, en_NL, en_NR, en_NU, en_NZ, en_PG, en_PH, en_PK, en_PN, en_PR, en_PW, en_RW, en_SB, en_SC, en_SD, en_SE, en_SG, en_SH, en_SI, en_SL, en_SS, en_SX, en_SZ, en_TC, en_TK, en_TO, en_TT, en_TV, en_TZ, en_UG, en_UM, en_US, en_US_POSIX, en_VC, en_VG, en_VI, en_VU, en_WS, en_ZA, en_ZM, en_ZW, eo, eo_001, es, es_419, es_AR, es_BO, es_CL, es_CO, es_CR, es_CU, es_DO, es_EA, es_EC, es_ES, es_GQ, es_GT, es_HN, es_IC, es_MX, es_NI, es_PA, es_PE, es_PH, es_PR, es_PY, es_SV, es_US, es_UY, es_VE, et, et_EE, eu, eu_ES, ewo, ewo_CM, fa, fa_AF, fa_IR, ff, ff_CM, ff_GN, ff_MR, ff_SN, fi, fi_FI, fil, fil_PH, fo, fo_DK, fo_FO, fr, fr_BE, fr_BF, fr_BI, fr_BJ, fr_BL, fr_CA, fr_CD, fr_CF, fr_CG, fr_CH, fr_CI, fr_CM, fr_DJ, fr_DZ, fr_FR, fr_GA, fr_GF, fr_GN, fr_GP, fr_GQ, fr_HT, fr_KM, fr_LU, fr_MA, fr_MC, fr_MF, fr_MG, fr_ML, fr_MQ, fr_MR, fr_MU, fr_NC, fr_NE, fr_PF, fr_PM, fr_RE, fr_RW, fr_SC, fr_SN, fr_SY, fr_TD, fr_TG, fr_TN, fr_VU, fr_WF, fr_YT, fur, fur_IT, fy, fy_NL, ga, ga_IE, gd, gd_GB, gl, gl_ES, gsw, gsw_CH, gsw_FR, gsw_LI, gu, gu_IN, guz, guz_KE, gv, gv_IM, ha, ha_GH, ha_NE, ha_NG, haw, haw_US, he, he_IL, hi, hi_IN, hr, hr_BA, hr_HR, hsb, hsb_DE, hu, hu_HU, hy, hy_AM, id, id_ID, ig, ig_NG, ii, ii_CN, is, is_IS, it, it_CH, it_IT, it_SM, ja, ja_JP, jgo, jgo_CM, jmc, jmc_TZ, ka, ka_GE, kab, kab_DZ, kam, kam_KE, kde, kde_TZ, kea, kea_CV, khq, khq_ML, ki, ki_KE, kk, kk_KZ, kkj, kkj_CM, kl, kl_GL, kln, kln_KE, km, km_KH, kn, kn_IN, ko, ko_KP, ko_KR, kok, kok_IN, ks, ks_IN, ksb, ksb_TZ, ksf, ksf_CM, ksh, ksh_DE, kw, kw_GB, ky, ky_KG, lag, lag_TZ, lb, lb_LU, lg, lg_UG, lkt, lkt_US, ln, ln_AO, ln_CD, ln_CF, ln_CG, lo, lo_LA, lrc, lrc_IQ, lrc_IR, lt, lt_LT, lu, lu_CD, luo, luo_KE, luy, luy_KE, lv, lv_LV, mas, mas_KE, mas_TZ, mer, mer_KE, mfe, mfe_MU, mg, mg_MG, mgh, mgh_MZ, mgo, mgo_CM, mk, mk_MK, ml, ml_IN, mn, mn_MN, mr, mr_IN, ms, ms_BN, ms_MY, ms_SG, mt, mt_MT, mua, mua_CM, my, my_MM, mzn, mzn_IR, naq, naq_NA, nb, nb_NO, nb_SJ, nd, nd_ZW, ne, ne_IN, ne_NP, nl, nl_AW, nl_BE, nl_BQ, nl_CW, nl_NL, nl_SR, nl_SX, nmg, nmg_CM, nn, nn_NO, nnh, nnh_CM, nus, nus_SS, nyn, nyn_UG, om, om_ET, om_KE, or, or_IN, os, os_GE, os_RU, pa, pa_Arab, pa_Arab_PK, pa_Guru, pa_Guru_IN, pl, pl_PL, prg, prg_001, ps, ps_AF, pt, pt_AO, pt_BR, pt_CV, pt_GW, pt_MO, pt_MZ, pt_PT, pt_ST, pt_TL, qu, qu_BO, qu_EC, qu_PE, rm, rm_CH, rn, rn_BI, ro, ro_MD, ro_RO, rof, rof_TZ, root, ru, ru_BY, ru_KG, ru_KZ, ru_MD, ru_RU, ru_UA, rw, rw_RW, rwk, rwk_TZ, sah, sah_RU, saq, saq_KE, sbp, sbp_TZ, se, se_FI, se_NO, se_SE, seh, seh_MZ, ses, ses_ML, sg, sg_CF, shi, shi_Latn, shi_Latn_MA, shi_Tfng, shi_Tfng_MA, si, si_LK, sk, sk_SK, sl, sl_SI, smn, smn_FI, sn, sn_ZW, so, so_DJ, so_ET, so_KE, so_SO, sq, sq_AL, sq_MK, sq_XK, sr, sr_Cyrl, sr_Cyrl_BA, sr_Cyrl_ME, sr_Cyrl_RS, sr_Cyrl_XK, sr_Latn, sr_Latn_BA, sr_Latn_ME, sr_Latn_RS, sr_Latn_XK, sv, sv_AX, sv_FI, sv_SE, sw, sw_CD, sw_KE, sw_TZ, sw_UG, ta, ta_IN, ta_LK, ta_MY, ta_SG, te, te_IN, teo, teo_KE, teo_UG, th, th_TH, ti, ti_ER, ti_ET, tk, tk_TM, to, to_TO, tr, tr_CY, tr_TR, twq, twq_NE, tzm, tzm_MA, ug, ug_CN, uk, uk_UA, ur, ur_IN, ur_PK, uz, uz_Arab, uz_Arab_AF, uz_Cyrl, uz_Cyrl_UZ, uz_Latn, uz_Latn_UZ, vai, vai_Latn, vai_Latn_LR, vai_Vaii, vai_Vaii_LR, vi, vi_VN, vo, vo_001, vun, vun_TZ, wae, wae_CH, xog, xog_UG, yav, yav_CM, yi, yi_001, yo, yo_BJ, yo_NG, zgh, zgh_MA, zh, zh_Hans, zh_Hans_CN, zh_Hans_HK, zh_Hans_MO, zh_Hans_SG, zh_Hant, zh_Hant_HK, zh_Hant_MO, zh_Hant_TW, zu, zu_ZA"); 641 642 //LocaleMatcher.DEBUG = true; 643 ULocale expected = new ULocale("sv"); 644 assertEquals(expected, matcherShort.getBestMatch(desired)); 645 assertEquals(expected, matcherLong.getBestMatch(desired)); 646 assertEquals(expected, matcherVeryLong.getBestMatch(desired)); 647 //LocaleMatcher.DEBUG = false; 648 649 for (int i = 0; i < 2; ++i) { 650 int iterations = i == 0 ? 1000 : 100000; 651 boolean showMessage = i != 0; 652 long timeShort = timeLocaleMatcher("Duration (few supported):\t", desired, matcherShort, showMessage, iterations, 0); 653 timeLocaleMatcher("Duration (med. supported):\t", desired, matcherLong, showMessage, iterations, timeShort); 654 timeLocaleMatcher("Duration (many supported):\t", desired, matcherVeryLong, showMessage, iterations, timeShort); 655 } 656 } 657 658 private long timeLocaleMatcher(String title, String desired, LocaleMatcher matcher, 659 boolean showmessage, int iterations, long comparisonTime) { 660 long start = System.nanoTime(); 661 for (int i = iterations; i > 0; --i) { 662 matcher.getBestMatch(desired); 663 } 664 long delta = System.nanoTime() - start; 665 if (showmessage) logln(title + (delta / iterations) + " nanos, " 666 + (comparisonTime > 0 ? (delta * 100 / comparisonTime - 100) + "% longer" : "")); 667 return delta; 668 } 669 } 670