| File: | root/firefox-clang/intl/icu/source/common/locid.cpp |
| Warning: | line 1905, column 9 Value stored to 'separator' is never read |
Press '?' to see keyboard shortcuts
Keyboard shortcuts:
| 1 | // © 2016 and later: Unicode, Inc. and others. |
| 2 | // License & terms of use: http://www.unicode.org/copyright.html |
| 3 | /* |
| 4 | ********************************************************************** |
| 5 | * Copyright (C) 1997-2016, International Business Machines |
| 6 | * Corporation and others. All Rights Reserved. |
| 7 | ********************************************************************** |
| 8 | * |
| 9 | * File locid.cpp |
| 10 | * |
| 11 | * Created by: Richard Gillam |
| 12 | * |
| 13 | * Modification History: |
| 14 | * |
| 15 | * Date Name Description |
| 16 | * 02/11/97 aliu Changed gLocPath to fgDataDirectory and added |
| 17 | * methods to get and set it. |
| 18 | * 04/02/97 aliu Made operator!= inline; fixed return value |
| 19 | * of getName(). |
| 20 | * 04/15/97 aliu Cleanup for AIX/Win32. |
| 21 | * 04/24/97 aliu Numerous changes per code review. |
| 22 | * 08/18/98 stephen Changed getDisplayName() |
| 23 | * Added SIMPLIFIED_CHINESE, TRADITIONAL_CHINESE |
| 24 | * Added getISOCountries(), getISOLanguages(), |
| 25 | * getLanguagesForCountry() |
| 26 | * 03/16/99 bertrand rehaul. |
| 27 | * 07/21/99 stephen Added U_CFUNC setDefault |
| 28 | * 11/09/99 weiv Added const char * getName() const; |
| 29 | * 04/12/00 srl removing unicodestring api's and cached hash code |
| 30 | * 08/10/01 grhoten Change the static Locales to accessor functions |
| 31 | ****************************************************************************** |
| 32 | */ |
| 33 | |
| 34 | #include <optional> |
| 35 | #include <string_view> |
| 36 | #include <utility> |
| 37 | |
| 38 | #include "unicode/bytestream.h" |
| 39 | #include "unicode/locid.h" |
| 40 | #include "unicode/localebuilder.h" |
| 41 | #include "unicode/strenum.h" |
| 42 | #include "unicode/stringpiece.h" |
| 43 | #include "unicode/uloc.h" |
| 44 | #include "unicode/ures.h" |
| 45 | |
| 46 | #include "bytesinkutil.h" |
| 47 | #include "charstr.h" |
| 48 | #include "charstrmap.h" |
| 49 | #include "cmemory.h" |
| 50 | #include "cstring.h" |
| 51 | #include "mutex.h" |
| 52 | #include "putilimp.h" |
| 53 | #include "uassert.h" |
| 54 | #include "ucln_cmn.h" |
| 55 | #include "uhash.h" |
| 56 | #include "ulocimp.h" |
| 57 | #include "umutex.h" |
| 58 | #include "uniquecharstr.h" |
| 59 | #include "ustr_imp.h" |
| 60 | #include "uvector.h" |
| 61 | |
| 62 | U_NAMESPACE_BEGINnamespace icu_77 { |
| 63 | |
| 64 | static Locale *gLocaleCache = nullptr; |
| 65 | static UInitOnce gLocaleCacheInitOnce {}; |
| 66 | |
| 67 | // gDefaultLocaleMutex protects all access to gDefaultLocalesHashT and gDefaultLocale. |
| 68 | static UMutex gDefaultLocaleMutex; |
| 69 | static UHashtable *gDefaultLocalesHashT = nullptr; |
| 70 | static Locale *gDefaultLocale = nullptr; |
| 71 | |
| 72 | /** |
| 73 | * \def ULOC_STRING_LIMIT |
| 74 | * strings beyond this value crash in CharString |
| 75 | */ |
| 76 | #define ULOC_STRING_LIMIT357913941 357913941 |
| 77 | |
| 78 | U_NAMESPACE_END} |
| 79 | |
| 80 | typedef enum ELocalePos { |
| 81 | eENGLISH, |
| 82 | eFRENCH, |
| 83 | eGERMAN, |
| 84 | eITALIAN, |
| 85 | eJAPANESE, |
| 86 | eKOREAN, |
| 87 | eCHINESE, |
| 88 | |
| 89 | eFRANCE, |
| 90 | eGERMANY, |
| 91 | eITALY, |
| 92 | eJAPAN, |
| 93 | eKOREA, |
| 94 | eCHINA, /* Alias for PRC */ |
| 95 | eTAIWAN, |
| 96 | eUK, |
| 97 | eUS, |
| 98 | eCANADA, |
| 99 | eCANADA_FRENCH, |
| 100 | eROOT, |
| 101 | |
| 102 | |
| 103 | //eDEFAULT, |
| 104 | eMAX_LOCALES |
| 105 | } ELocalePos; |
| 106 | |
| 107 | namespace { |
| 108 | |
| 109 | // |
| 110 | // Deleter function for Locales owned by the default Locale hash table/ |
| 111 | // |
| 112 | void U_CALLCONV |
| 113 | deleteLocale(void *obj) { |
| 114 | delete static_cast<icu::Locale*>(obj); |
| 115 | } |
| 116 | |
| 117 | UBool U_CALLCONV locale_cleanup() |
| 118 | { |
| 119 | U_NAMESPACE_USEusing namespace icu_77; |
| 120 | |
| 121 | delete [] gLocaleCache; |
| 122 | gLocaleCache = nullptr; |
| 123 | gLocaleCacheInitOnce.reset(); |
| 124 | |
| 125 | if (gDefaultLocalesHashT) { |
| 126 | uhash_closeuhash_close_77(gDefaultLocalesHashT); // Automatically deletes all elements, using deleter func. |
| 127 | gDefaultLocalesHashT = nullptr; |
| 128 | } |
| 129 | gDefaultLocale = nullptr; |
| 130 | return true; |
| 131 | } |
| 132 | |
| 133 | void U_CALLCONV locale_init(UErrorCode &status) { |
| 134 | U_NAMESPACE_USEusing namespace icu_77; |
| 135 | |
| 136 | U_ASSERT(gLocaleCache == nullptr)(static_cast <bool> (gLocaleCache == nullptr) ? void (0 ) : __assert_fail ("gLocaleCache == nullptr", __builtin_FILE ( ), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 137 | gLocaleCache = new Locale[static_cast<int>(eMAX_LOCALES)]; |
| 138 | if (gLocaleCache == nullptr) { |
| 139 | status = U_MEMORY_ALLOCATION_ERROR; |
| 140 | return; |
| 141 | } |
| 142 | ucln_common_registerCleanupucln_common_registerCleanup_77(UCLN_COMMON_LOCALE, locale_cleanup); |
| 143 | gLocaleCache[eROOT] = Locale(""); |
| 144 | gLocaleCache[eENGLISH] = Locale("en"); |
| 145 | gLocaleCache[eFRENCH] = Locale("fr"); |
| 146 | gLocaleCache[eGERMAN] = Locale("de"); |
| 147 | gLocaleCache[eITALIAN] = Locale("it"); |
| 148 | gLocaleCache[eJAPANESE] = Locale("ja"); |
| 149 | gLocaleCache[eKOREAN] = Locale("ko"); |
| 150 | gLocaleCache[eCHINESE] = Locale("zh"); |
| 151 | gLocaleCache[eFRANCE] = Locale("fr", "FR"); |
| 152 | gLocaleCache[eGERMANY] = Locale("de", "DE"); |
| 153 | gLocaleCache[eITALY] = Locale("it", "IT"); |
| 154 | gLocaleCache[eJAPAN] = Locale("ja", "JP"); |
| 155 | gLocaleCache[eKOREA] = Locale("ko", "KR"); |
| 156 | gLocaleCache[eCHINA] = Locale("zh", "CN"); |
| 157 | gLocaleCache[eTAIWAN] = Locale("zh", "TW"); |
| 158 | gLocaleCache[eUK] = Locale("en", "GB"); |
| 159 | gLocaleCache[eUS] = Locale("en", "US"); |
| 160 | gLocaleCache[eCANADA] = Locale("en", "CA"); |
| 161 | gLocaleCache[eCANADA_FRENCH] = Locale("fr", "CA"); |
| 162 | } |
| 163 | |
| 164 | } // namespace |
| 165 | |
| 166 | U_NAMESPACE_BEGINnamespace icu_77 { |
| 167 | |
| 168 | Locale *locale_set_default_internal(const char *id, UErrorCode& status) { |
| 169 | // Synchronize this entire function. |
| 170 | Mutex lock(&gDefaultLocaleMutex); |
| 171 | |
| 172 | UBool canonicalize = false; |
| 173 | |
| 174 | // If given a nullptr string for the locale id, grab the default |
| 175 | // name from the system. |
| 176 | // (Different from most other locale APIs, where a null name means use |
| 177 | // the current ICU default locale.) |
| 178 | if (id == nullptr) { |
| 179 | id = uprv_getDefaultLocaleIDuprv_getDefaultLocaleID_77(); // This function not thread safe? TODO: verify. |
| 180 | canonicalize = true; // always canonicalize host ID |
| 181 | } |
| 182 | |
| 183 | CharString localeNameBuf = |
| 184 | canonicalize ? ulocimp_canonicalizeulocimp_canonicalize_77(id, status) : ulocimp_getNameulocimp_getName_77(id, status); |
| 185 | |
| 186 | if (U_FAILURE(status)) { |
| 187 | return gDefaultLocale; |
| 188 | } |
| 189 | |
| 190 | if (gDefaultLocalesHashT == nullptr) { |
| 191 | gDefaultLocalesHashT = uhash_openuhash_open_77(uhash_hashCharsuhash_hashChars_77, uhash_compareCharsuhash_compareChars_77, nullptr, &status); |
| 192 | if (U_FAILURE(status)) { |
| 193 | return gDefaultLocale; |
| 194 | } |
| 195 | uhash_setValueDeleteruhash_setValueDeleter_77(gDefaultLocalesHashT, deleteLocale); |
| 196 | ucln_common_registerCleanupucln_common_registerCleanup_77(UCLN_COMMON_LOCALE, locale_cleanup); |
| 197 | } |
| 198 | |
| 199 | Locale* newDefault = static_cast<Locale*>(uhash_getuhash_get_77(gDefaultLocalesHashT, localeNameBuf.data())); |
| 200 | if (newDefault == nullptr) { |
| 201 | newDefault = new Locale(Locale::eBOGUS); |
| 202 | if (newDefault == nullptr) { |
| 203 | status = U_MEMORY_ALLOCATION_ERROR; |
| 204 | return gDefaultLocale; |
| 205 | } |
| 206 | newDefault->init(localeNameBuf.data(), false); |
| 207 | uhash_putuhash_put_77(gDefaultLocalesHashT, const_cast<char*>(newDefault->getName()), newDefault, &status); |
| 208 | if (U_FAILURE(status)) { |
| 209 | return gDefaultLocale; |
| 210 | } |
| 211 | } |
| 212 | gDefaultLocale = newDefault; |
| 213 | return gDefaultLocale; |
| 214 | } |
| 215 | |
| 216 | U_NAMESPACE_END} |
| 217 | |
| 218 | /* sfb 07/21/99 */ |
| 219 | U_CFUNCextern "C" void |
| 220 | locale_set_defaultlocale_set_default_77(const char *id) |
| 221 | { |
| 222 | U_NAMESPACE_USEusing namespace icu_77; |
| 223 | UErrorCode status = U_ZERO_ERROR; |
| 224 | locale_set_default_internal(id, status); |
| 225 | } |
| 226 | /* end */ |
| 227 | |
| 228 | U_CFUNCextern "C" const char * |
| 229 | locale_get_defaultlocale_get_default_77() |
| 230 | { |
| 231 | U_NAMESPACE_USEusing namespace icu_77; |
| 232 | return Locale::getDefault().getName(); |
| 233 | } |
| 234 | |
| 235 | |
| 236 | U_NAMESPACE_BEGINnamespace icu_77 { |
| 237 | |
| 238 | UOBJECT_DEFINE_RTTI_IMPLEMENTATION(Locale)UClassID Locale::getStaticClassID() { static char classID = 0 ; return (UClassID)&classID; } UClassID Locale::getDynamicClassID () const { return Locale::getStaticClassID(); } |
| 239 | |
| 240 | /*Character separating the posix id fields*/ |
| 241 | // '_' |
| 242 | // In the platform codepage. |
| 243 | #define SEP_CHAR'_' '_' |
| 244 | #define NULL_CHAR'\0' '\0' |
| 245 | |
| 246 | Locale::~Locale() |
| 247 | { |
| 248 | if ((baseName != fullName) && (baseName != fullNameBuffer)) { |
| 249 | uprv_freeuprv_free_77(baseName); |
| 250 | } |
| 251 | baseName = nullptr; |
| 252 | /*if fullName is on the heap, we free it*/ |
| 253 | if (fullName != fullNameBuffer) |
| 254 | { |
| 255 | uprv_freeuprv_free_77(fullName); |
| 256 | fullName = nullptr; |
| 257 | } |
| 258 | } |
| 259 | |
| 260 | Locale::Locale() |
| 261 | : UObject(), fullName(fullNameBuffer), baseName(nullptr) |
| 262 | { |
| 263 | init(nullptr, false); |
| 264 | } |
| 265 | |
| 266 | /* |
| 267 | * Internal constructor to allow construction of a locale object with |
| 268 | * NO side effects. (Default constructor tries to get |
| 269 | * the default locale.) |
| 270 | */ |
| 271 | Locale::Locale(Locale::ELocaleType) |
| 272 | : UObject(), fullName(fullNameBuffer), baseName(nullptr) |
| 273 | { |
| 274 | setToBogus(); |
| 275 | } |
| 276 | |
| 277 | |
| 278 | Locale::Locale( const char * newLanguage, |
| 279 | const char * newCountry, |
| 280 | const char * newVariant, |
| 281 | const char * newKeywords) |
| 282 | : UObject(), fullName(fullNameBuffer), baseName(nullptr) |
| 283 | { |
| 284 | if( (newLanguage==nullptr) && (newCountry == nullptr) && (newVariant == nullptr) ) |
| 285 | { |
| 286 | init(nullptr, false); /* shortcut */ |
| 287 | } |
| 288 | else |
| 289 | { |
| 290 | UErrorCode status = U_ZERO_ERROR; |
| 291 | int32_t lsize = 0; |
| 292 | int32_t csize = 0; |
| 293 | int32_t vsize = 0; |
| 294 | int32_t ksize = 0; |
| 295 | |
| 296 | // Check the sizes of the input strings. |
| 297 | |
| 298 | // Language |
| 299 | if ( newLanguage != nullptr ) |
| 300 | { |
| 301 | lsize = static_cast<int32_t>(uprv_strlen(newLanguage):: strlen(newLanguage)); |
| 302 | if ( lsize < 0 || lsize > ULOC_STRING_LIMIT357913941 ) { // int32 wrap |
| 303 | setToBogus(); |
| 304 | return; |
| 305 | } |
| 306 | } |
| 307 | |
| 308 | CharString togo(newLanguage, lsize, status); // start with newLanguage |
| 309 | |
| 310 | // _Country |
| 311 | if ( newCountry != nullptr ) |
| 312 | { |
| 313 | csize = static_cast<int32_t>(uprv_strlen(newCountry):: strlen(newCountry)); |
| 314 | if ( csize < 0 || csize > ULOC_STRING_LIMIT357913941 ) { // int32 wrap |
| 315 | setToBogus(); |
| 316 | return; |
| 317 | } |
| 318 | } |
| 319 | |
| 320 | // _Variant |
| 321 | if ( newVariant != nullptr ) |
| 322 | { |
| 323 | // remove leading _'s |
| 324 | while(newVariant[0] == SEP_CHAR'_') |
| 325 | { |
| 326 | newVariant++; |
| 327 | } |
| 328 | |
| 329 | // remove trailing _'s |
| 330 | vsize = static_cast<int32_t>(uprv_strlen(newVariant):: strlen(newVariant)); |
| 331 | if ( vsize < 0 || vsize > ULOC_STRING_LIMIT357913941 ) { // int32 wrap |
| 332 | setToBogus(); |
| 333 | return; |
| 334 | } |
| 335 | while( (vsize>1) && (newVariant[vsize-1] == SEP_CHAR'_') ) |
| 336 | { |
| 337 | vsize--; |
| 338 | } |
| 339 | } |
| 340 | |
| 341 | if ( newKeywords != nullptr) |
| 342 | { |
| 343 | ksize = static_cast<int32_t>(uprv_strlen(newKeywords):: strlen(newKeywords)); |
| 344 | if ( ksize < 0 || ksize > ULOC_STRING_LIMIT357913941 ) { |
| 345 | setToBogus(); |
| 346 | return; |
| 347 | } |
| 348 | } |
| 349 | |
| 350 | // We've checked the input sizes, now build up the full locale string.. |
| 351 | |
| 352 | // newLanguage is already copied |
| 353 | |
| 354 | if ( ( vsize != 0 ) || (csize != 0) ) // at least: __v |
| 355 | { // ^ |
| 356 | togo.append(SEP_CHAR'_', status); |
| 357 | } |
| 358 | |
| 359 | if ( csize != 0 ) |
| 360 | { |
| 361 | togo.append(newCountry, status); |
| 362 | } |
| 363 | |
| 364 | if ( vsize != 0) |
| 365 | { |
| 366 | togo.append(SEP_CHAR'_', status) |
| 367 | .append(newVariant, vsize, status); |
| 368 | } |
| 369 | |
| 370 | if ( ksize != 0) |
| 371 | { |
| 372 | if (uprv_strchr(newKeywords, '='):: strchr(newKeywords, '=')) { |
| 373 | togo.append('@', status); /* keyword parsing */ |
| 374 | } |
| 375 | else { |
| 376 | togo.append('_', status); /* Variant parsing with a script */ |
| 377 | if ( vsize == 0) { |
| 378 | togo.append('_', status); /* No country found */ |
| 379 | } |
| 380 | } |
| 381 | togo.append(newKeywords, status); |
| 382 | } |
| 383 | |
| 384 | if (U_FAILURE(status)) { |
| 385 | // Something went wrong with appending, etc. |
| 386 | setToBogus(); |
| 387 | return; |
| 388 | } |
| 389 | // Parse it, because for example 'language' might really be a complete |
| 390 | // string. |
| 391 | init(togo.data(), false); |
| 392 | } |
| 393 | } |
| 394 | |
| 395 | Locale::Locale(const Locale &other) |
| 396 | : UObject(other), fullName(fullNameBuffer), baseName(nullptr) |
| 397 | { |
| 398 | *this = other; |
| 399 | } |
| 400 | |
| 401 | Locale::Locale(Locale&& other) noexcept |
| 402 | : UObject(other), fullName(fullNameBuffer), baseName(fullName) { |
| 403 | *this = std::move(other); |
| 404 | } |
| 405 | |
| 406 | Locale& Locale::operator=(const Locale& other) { |
| 407 | if (this == &other) { |
| 408 | return *this; |
| 409 | } |
| 410 | |
| 411 | setToBogus(); |
| 412 | |
| 413 | if (other.fullName == other.fullNameBuffer) { |
| 414 | uprv_strcpy(fullNameBuffer, other.fullNameBuffer):: strcpy(fullNameBuffer, other.fullNameBuffer); |
| 415 | } else if (other.fullName == nullptr) { |
| 416 | fullName = nullptr; |
| 417 | } else { |
| 418 | fullName = uprv_strdupuprv_strdup_77(other.fullName); |
| 419 | if (fullName == nullptr) return *this; |
| 420 | } |
| 421 | |
| 422 | if (other.baseName == other.fullName) { |
| 423 | baseName = fullName; |
| 424 | } else if (other.baseName != nullptr) { |
| 425 | baseName = uprv_strdupuprv_strdup_77(other.baseName); |
| 426 | if (baseName == nullptr) return *this; |
| 427 | } |
| 428 | |
| 429 | uprv_strcpy(language, other.language):: strcpy(language, other.language); |
| 430 | uprv_strcpy(script, other.script):: strcpy(script, other.script); |
| 431 | uprv_strcpy(country, other.country):: strcpy(country, other.country); |
| 432 | |
| 433 | variantBegin = other.variantBegin; |
| 434 | fIsBogus = other.fIsBogus; |
| 435 | |
| 436 | return *this; |
| 437 | } |
| 438 | |
| 439 | Locale& Locale::operator=(Locale&& other) noexcept { |
| 440 | if ((baseName != fullName) && (baseName != fullNameBuffer)) uprv_freeuprv_free_77(baseName); |
| 441 | if (fullName != fullNameBuffer) uprv_freeuprv_free_77(fullName); |
| 442 | |
| 443 | if (other.fullName == other.fullNameBuffer || other.baseName == other.fullNameBuffer) { |
| 444 | uprv_strcpy(fullNameBuffer, other.fullNameBuffer):: strcpy(fullNameBuffer, other.fullNameBuffer); |
| 445 | } |
| 446 | if (other.fullName == other.fullNameBuffer) { |
| 447 | fullName = fullNameBuffer; |
| 448 | } else { |
| 449 | fullName = other.fullName; |
| 450 | } |
| 451 | |
| 452 | if (other.baseName == other.fullNameBuffer) { |
| 453 | baseName = fullNameBuffer; |
| 454 | } else if (other.baseName == other.fullName) { |
| 455 | baseName = fullName; |
| 456 | } else { |
| 457 | baseName = other.baseName; |
| 458 | } |
| 459 | |
| 460 | uprv_strcpy(language, other.language):: strcpy(language, other.language); |
| 461 | uprv_strcpy(script, other.script):: strcpy(script, other.script); |
| 462 | uprv_strcpy(country, other.country):: strcpy(country, other.country); |
| 463 | |
| 464 | variantBegin = other.variantBegin; |
| 465 | fIsBogus = other.fIsBogus; |
| 466 | |
| 467 | other.baseName = other.fullName = other.fullNameBuffer; |
| 468 | |
| 469 | return *this; |
| 470 | } |
| 471 | |
| 472 | Locale * |
| 473 | Locale::clone() const { |
| 474 | return new Locale(*this); |
| 475 | } |
| 476 | |
| 477 | bool |
| 478 | Locale::operator==( const Locale& other) const |
| 479 | { |
| 480 | return (uprv_strcmp(other.fullName, fullName):: strcmp(other.fullName, fullName) == 0); |
| 481 | } |
| 482 | |
| 483 | namespace { |
| 484 | |
| 485 | UInitOnce gKnownCanonicalizedInitOnce {}; |
| 486 | UHashtable *gKnownCanonicalized = nullptr; |
| 487 | |
| 488 | constexpr const char* KNOWN_CANONICALIZED[] = { |
| 489 | "c", |
| 490 | // Commonly used locales known are already canonicalized |
| 491 | "af", "af_ZA", "am", "am_ET", "ar", "ar_001", "as", "as_IN", "az", "az_AZ", |
| 492 | "be", "be_BY", "bg", "bg_BG", "bn", "bn_IN", "bs", "bs_BA", "ca", "ca_ES", |
| 493 | "cs", "cs_CZ", "cy", "cy_GB", "da", "da_DK", "de", "de_DE", "el", "el_GR", |
| 494 | "en", "en_GB", "en_US", "es", "es_419", "es_ES", "et", "et_EE", "eu", |
| 495 | "eu_ES", "fa", "fa_IR", "fi", "fi_FI", "fil", "fil_PH", "fr", "fr_FR", |
| 496 | "ga", "ga_IE", "gl", "gl_ES", "gu", "gu_IN", "he", "he_IL", "hi", "hi_IN", |
| 497 | "hr", "hr_HR", "hu", "hu_HU", "hy", "hy_AM", "id", "id_ID", "is", "is_IS", |
| 498 | "it", "it_IT", "ja", "ja_JP", "jv", "jv_ID", "ka", "ka_GE", "kk", "kk_KZ", |
| 499 | "km", "km_KH", "kn", "kn_IN", "ko", "ko_KR", "ky", "ky_KG", "lo", "lo_LA", |
| 500 | "lt", "lt_LT", "lv", "lv_LV", "mk", "mk_MK", "ml", "ml_IN", "mn", "mn_MN", |
| 501 | "mr", "mr_IN", "ms", "ms_MY", "my", "my_MM", "nb", "nb_NO", "ne", "ne_NP", |
| 502 | "nl", "nl_NL", "no", "or", "or_IN", "pa", "pa_IN", "pl", "pl_PL", "ps", "ps_AF", |
| 503 | "pt", "pt_BR", "pt_PT", "ro", "ro_RO", "ru", "ru_RU", "sd", "sd_IN", "si", |
| 504 | "si_LK", "sk", "sk_SK", "sl", "sl_SI", "so", "so_SO", "sq", "sq_AL", "sr", |
| 505 | "sr_Cyrl_RS", "sr_Latn", "sr_RS", "sv", "sv_SE", "sw", "sw_TZ", "ta", |
| 506 | "ta_IN", "te", "te_IN", "th", "th_TH", "tk", "tk_TM", "tr", "tr_TR", "uk", |
| 507 | "uk_UA", "ur", "ur_PK", "uz", "uz_UZ", "vi", "vi_VN", "yue", "yue_Hant", |
| 508 | "yue_Hant_HK", "yue_HK", "zh", "zh_CN", "zh_Hans", "zh_Hans_CN", "zh_Hant", |
| 509 | "zh_Hant_TW", "zh_TW", "zu", "zu_ZA" |
| 510 | }; |
| 511 | |
| 512 | UBool U_CALLCONV cleanupKnownCanonicalized() { |
| 513 | gKnownCanonicalizedInitOnce.reset(); |
| 514 | if (gKnownCanonicalized) { uhash_closeuhash_close_77(gKnownCanonicalized); } |
| 515 | return true; |
| 516 | } |
| 517 | |
| 518 | void U_CALLCONV loadKnownCanonicalized(UErrorCode &status) { |
| 519 | ucln_common_registerCleanupucln_common_registerCleanup_77(UCLN_COMMON_LOCALE_KNOWN_CANONICALIZED, |
| 520 | cleanupKnownCanonicalized); |
| 521 | LocalUHashtablePointer newKnownCanonicalizedMap( |
| 522 | uhash_openuhash_open_77(uhash_hashCharsuhash_hashChars_77, uhash_compareCharsuhash_compareChars_77, nullptr, &status)); |
| 523 | for (int32_t i = 0; |
| 524 | U_SUCCESS(status) && i < UPRV_LENGTHOF(KNOWN_CANONICALIZED)(int32_t)(sizeof(KNOWN_CANONICALIZED)/sizeof((KNOWN_CANONICALIZED )[0])); |
| 525 | i++) { |
| 526 | uhash_putiuhash_puti_77(newKnownCanonicalizedMap.getAlias(), |
| 527 | (void*)KNOWN_CANONICALIZED[i], |
| 528 | 1, &status); |
| 529 | } |
| 530 | if (U_FAILURE(status)) { |
| 531 | return; |
| 532 | } |
| 533 | |
| 534 | gKnownCanonicalized = newKnownCanonicalizedMap.orphan(); |
| 535 | } |
| 536 | |
| 537 | class AliasData; |
| 538 | |
| 539 | /** |
| 540 | * A Builder class to build the alias data. |
| 541 | */ |
| 542 | class AliasDataBuilder { |
| 543 | public: |
| 544 | AliasDataBuilder() { |
| 545 | } |
| 546 | |
| 547 | // Build the AliasData from resource. |
| 548 | AliasData* build(UErrorCode &status); |
| 549 | |
| 550 | private: |
| 551 | void readAlias(UResourceBundle* alias, |
| 552 | UniqueCharStrings* strings, |
| 553 | LocalMemory<const char*>& types, |
| 554 | LocalMemory<int32_t>& replacementIndexes, |
| 555 | int32_t &length, |
| 556 | void (*checkType)(const char* type), |
| 557 | void (*checkReplacement)(const UChar* replacement), |
| 558 | UErrorCode &status); |
| 559 | |
| 560 | // Read the languageAlias data from alias to |
| 561 | // strings+types+replacementIndexes |
| 562 | // The number of record will be stored into length. |
| 563 | // Allocate length items for types, to store the type field. |
| 564 | // Allocate length items for replacementIndexes, |
| 565 | // to store the index in the strings for the replacement script. |
| 566 | void readLanguageAlias(UResourceBundle* alias, |
| 567 | UniqueCharStrings* strings, |
| 568 | LocalMemory<const char*>& types, |
| 569 | LocalMemory<int32_t>& replacementIndexes, |
| 570 | int32_t &length, |
| 571 | UErrorCode &status); |
| 572 | |
| 573 | // Read the scriptAlias data from alias to |
| 574 | // strings+types+replacementIndexes |
| 575 | // Allocate length items for types, to store the type field. |
| 576 | // Allocate length items for replacementIndexes, |
| 577 | // to store the index in the strings for the replacement script. |
| 578 | void readScriptAlias(UResourceBundle* alias, |
| 579 | UniqueCharStrings* strings, |
| 580 | LocalMemory<const char*>& types, |
| 581 | LocalMemory<int32_t>& replacementIndexes, |
| 582 | int32_t &length, UErrorCode &status); |
| 583 | |
| 584 | // Read the territoryAlias data from alias to |
| 585 | // strings+types+replacementIndexes |
| 586 | // Allocate length items for types, to store the type field. |
| 587 | // Allocate length items for replacementIndexes, |
| 588 | // to store the index in the strings for the replacement script. |
| 589 | void readTerritoryAlias(UResourceBundle* alias, |
| 590 | UniqueCharStrings* strings, |
| 591 | LocalMemory<const char*>& types, |
| 592 | LocalMemory<int32_t>& replacementIndexes, |
| 593 | int32_t &length, UErrorCode &status); |
| 594 | |
| 595 | // Read the variantAlias data from alias to |
| 596 | // strings+types+replacementIndexes |
| 597 | // Allocate length items for types, to store the type field. |
| 598 | // Allocate length items for replacementIndexes, |
| 599 | // to store the index in the strings for the replacement variant. |
| 600 | void readVariantAlias(UResourceBundle* alias, |
| 601 | UniqueCharStrings* strings, |
| 602 | LocalMemory<const char*>& types, |
| 603 | LocalMemory<int32_t>& replacementIndexes, |
| 604 | int32_t &length, UErrorCode &status); |
| 605 | |
| 606 | // Read the subdivisionAlias data from alias to |
| 607 | // strings+types+replacementIndexes |
| 608 | // Allocate length items for types, to store the type field. |
| 609 | // Allocate length items for replacementIndexes, |
| 610 | // to store the index in the strings for the replacement variant. |
| 611 | void readSubdivisionAlias(UResourceBundle* alias, |
| 612 | UniqueCharStrings* strings, |
| 613 | LocalMemory<const char*>& types, |
| 614 | LocalMemory<int32_t>& replacementIndexes, |
| 615 | int32_t &length, UErrorCode &status); |
| 616 | }; |
| 617 | |
| 618 | /** |
| 619 | * A class to hold the Alias Data. |
| 620 | */ |
| 621 | class AliasData : public UMemory { |
| 622 | public: |
| 623 | static const AliasData* singleton(UErrorCode& status) { |
| 624 | if (U_FAILURE(status)) { |
| 625 | // Do not get into loadData if the status already has error. |
| 626 | return nullptr; |
| 627 | } |
| 628 | umtx_initOnce(AliasData::gInitOnce, &AliasData::loadData, status); |
| 629 | return gSingleton; |
| 630 | } |
| 631 | |
| 632 | const CharStringMap& languageMap() const { return language; } |
| 633 | const CharStringMap& scriptMap() const { return script; } |
| 634 | const CharStringMap& territoryMap() const { return territory; } |
| 635 | const CharStringMap& variantMap() const { return variant; } |
| 636 | const CharStringMap& subdivisionMap() const { return subdivision; } |
| 637 | |
| 638 | static void U_CALLCONV loadData(UErrorCode &status); |
| 639 | static UBool U_CALLCONV cleanup(); |
| 640 | |
| 641 | static UInitOnce gInitOnce; |
| 642 | |
| 643 | private: |
| 644 | AliasData(CharStringMap languageMap, |
| 645 | CharStringMap scriptMap, |
| 646 | CharStringMap territoryMap, |
| 647 | CharStringMap variantMap, |
| 648 | CharStringMap subdivisionMap, |
| 649 | CharString* strings) |
| 650 | : language(std::move(languageMap)), |
| 651 | script(std::move(scriptMap)), |
| 652 | territory(std::move(territoryMap)), |
| 653 | variant(std::move(variantMap)), |
| 654 | subdivision(std::move(subdivisionMap)), |
| 655 | strings(strings) { |
| 656 | } |
| 657 | |
| 658 | ~AliasData() { |
| 659 | delete strings; |
| 660 | } |
| 661 | |
| 662 | static const AliasData* gSingleton; |
| 663 | |
| 664 | CharStringMap language; |
| 665 | CharStringMap script; |
| 666 | CharStringMap territory; |
| 667 | CharStringMap variant; |
| 668 | CharStringMap subdivision; |
| 669 | CharString* strings; |
| 670 | |
| 671 | friend class AliasDataBuilder; |
| 672 | }; |
| 673 | |
| 674 | |
| 675 | const AliasData* AliasData::gSingleton = nullptr; |
| 676 | UInitOnce AliasData::gInitOnce {}; |
| 677 | |
| 678 | UBool U_CALLCONV |
| 679 | AliasData::cleanup() |
| 680 | { |
| 681 | gInitOnce.reset(); |
| 682 | delete gSingleton; |
| 683 | return true; |
| 684 | } |
| 685 | |
| 686 | void |
| 687 | AliasDataBuilder::readAlias( |
| 688 | UResourceBundle* alias, |
| 689 | UniqueCharStrings* strings, |
| 690 | LocalMemory<const char*>& types, |
| 691 | LocalMemory<int32_t>& replacementIndexes, |
| 692 | int32_t &length, |
| 693 | void (*checkType)(const char* type), |
| 694 | void (*checkReplacement)(const UChar* replacement), |
| 695 | UErrorCode &status) { |
| 696 | if (U_FAILURE(status)) { |
| 697 | return; |
| 698 | } |
| 699 | length = ures_getSizeures_getSize_77(alias); |
| 700 | const char** rawTypes = types.allocateInsteadAndCopy(length); |
| 701 | if (rawTypes == nullptr) { |
| 702 | status = U_MEMORY_ALLOCATION_ERROR; |
| 703 | return; |
| 704 | } |
| 705 | int32_t* rawIndexes = replacementIndexes.allocateInsteadAndCopy(length); |
| 706 | if (rawIndexes == nullptr) { |
| 707 | status = U_MEMORY_ALLOCATION_ERROR; |
| 708 | return; |
| 709 | } |
| 710 | for (int i = 0; U_SUCCESS(status) && ures_hasNextures_hasNext_77(alias); i++) { |
| 711 | LocalUResourceBundlePointer res( |
| 712 | ures_getNextResourceures_getNextResource_77(alias, nullptr, &status)); |
| 713 | const char* aliasFrom = ures_getKeyures_getKey_77(res.getAlias()); |
| 714 | const UChar* aliasTo = |
| 715 | ures_getStringByKeyures_getStringByKey_77(res.getAlias(), "replacement", nullptr, &status); |
| 716 | if (U_FAILURE(status)) return; |
| 717 | |
| 718 | checkType(aliasFrom); |
| 719 | checkReplacement(aliasTo); |
| 720 | |
| 721 | rawTypes[i] = aliasFrom; |
| 722 | rawIndexes[i] = strings->add(aliasTo, status); |
| 723 | } |
| 724 | } |
| 725 | |
| 726 | /** |
| 727 | * Read the languageAlias data from alias to strings+types+replacementIndexes. |
| 728 | * Allocate length items for types, to store the type field. Allocate length |
| 729 | * items for replacementIndexes, to store the index in the strings for the |
| 730 | * replacement language. |
| 731 | */ |
| 732 | void |
| 733 | AliasDataBuilder::readLanguageAlias( |
| 734 | UResourceBundle* alias, |
| 735 | UniqueCharStrings* strings, |
| 736 | LocalMemory<const char*>& types, |
| 737 | LocalMemory<int32_t>& replacementIndexes, |
| 738 | int32_t &length, |
| 739 | UErrorCode &status) |
| 740 | { |
| 741 | return readAlias( |
| 742 | alias, strings, types, replacementIndexes, length, |
| 743 | #if U_DEBUG1 |
| 744 | [](const char* type) { |
| 745 | // Assert the aliasFrom only contains the following possibilities |
| 746 | // language_REGION_variant |
| 747 | // language_REGION |
| 748 | // language_variant |
| 749 | // language |
| 750 | // und_variant |
| 751 | Locale test(type); |
| 752 | // Assert no script in aliasFrom |
| 753 | U_ASSERT(test.getScript()[0] == '\0')(static_cast <bool> (test.getScript()[0] == '\0') ? void (0) : __assert_fail ("test.getScript()[0] == '\\0'", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 754 | // Assert when language is und, no REGION in aliasFrom. |
| 755 | U_ASSERT(test.getLanguage()[0] != '\0' || test.getCountry()[0] == '\0')(static_cast <bool> (test.getLanguage()[0] != '\0' || test .getCountry()[0] == '\0') ? void (0) : __assert_fail ("test.getLanguage()[0] != '\\0' || test.getCountry()[0] == '\\0'" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 756 | }, |
| 757 | #else |
| 758 | [](const char*) {}, |
| 759 | #endif |
| 760 | [](const UChar*) {}, status); |
| 761 | } |
| 762 | |
| 763 | /** |
| 764 | * Read the scriptAlias data from alias to strings+types+replacementIndexes. |
| 765 | * Allocate length items for types, to store the type field. Allocate length |
| 766 | * items for replacementIndexes, to store the index in the strings for the |
| 767 | * replacement script. |
| 768 | */ |
| 769 | void |
| 770 | AliasDataBuilder::readScriptAlias( |
| 771 | UResourceBundle* alias, |
| 772 | UniqueCharStrings* strings, |
| 773 | LocalMemory<const char*>& types, |
| 774 | LocalMemory<int32_t>& replacementIndexes, |
| 775 | int32_t &length, |
| 776 | UErrorCode &status) |
| 777 | { |
| 778 | return readAlias( |
| 779 | alias, strings, types, replacementIndexes, length, |
| 780 | #if U_DEBUG1 |
| 781 | [](const char* type) { |
| 782 | U_ASSERT(uprv_strlen(type) == 4)(static_cast <bool> (:: strlen(type) == 4) ? void (0) : __assert_fail (":: strlen(type) == 4", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 783 | }, |
| 784 | [](const UChar* replacement) { |
| 785 | U_ASSERT(u_strlen(replacement) == 4)(static_cast <bool> (u_strlen_77(replacement) == 4) ? void (0) : __assert_fail ("u_strlen_77(replacement) == 4", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 786 | }, |
| 787 | #else |
| 788 | [](const char*) {}, |
| 789 | [](const UChar*) { }, |
| 790 | #endif |
| 791 | status); |
| 792 | } |
| 793 | |
| 794 | /** |
| 795 | * Read the territoryAlias data from alias to strings+types+replacementIndexes. |
| 796 | * Allocate length items for types, to store the type field. Allocate length |
| 797 | * items for replacementIndexes, to store the index in the strings for the |
| 798 | * replacement regions. |
| 799 | */ |
| 800 | void |
| 801 | AliasDataBuilder::readTerritoryAlias( |
| 802 | UResourceBundle* alias, |
| 803 | UniqueCharStrings* strings, |
| 804 | LocalMemory<const char*>& types, |
| 805 | LocalMemory<int32_t>& replacementIndexes, |
| 806 | int32_t &length, |
| 807 | UErrorCode &status) |
| 808 | { |
| 809 | return readAlias( |
| 810 | alias, strings, types, replacementIndexes, length, |
| 811 | #if U_DEBUG1 |
| 812 | [](const char* type) { |
| 813 | U_ASSERT(uprv_strlen(type) == 2 || uprv_strlen(type) == 3)(static_cast <bool> (:: strlen(type) == 2 || :: strlen( type) == 3) ? void (0) : __assert_fail (":: strlen(type) == 2 || :: strlen(type) == 3" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 814 | }, |
| 815 | #else |
| 816 | [](const char*) {}, |
| 817 | #endif |
| 818 | [](const UChar*) { }, |
| 819 | status); |
| 820 | } |
| 821 | |
| 822 | /** |
| 823 | * Read the variantAlias data from alias to strings+types+replacementIndexes. |
| 824 | * Allocate length items for types, to store the type field. Allocate length |
| 825 | * items for replacementIndexes, to store the index in the strings for the |
| 826 | * replacement variant. |
| 827 | */ |
| 828 | void |
| 829 | AliasDataBuilder::readVariantAlias( |
| 830 | UResourceBundle* alias, |
| 831 | UniqueCharStrings* strings, |
| 832 | LocalMemory<const char*>& types, |
| 833 | LocalMemory<int32_t>& replacementIndexes, |
| 834 | int32_t &length, |
| 835 | UErrorCode &status) |
| 836 | { |
| 837 | return readAlias( |
| 838 | alias, strings, types, replacementIndexes, length, |
| 839 | #if U_DEBUG1 |
| 840 | [](const char* type) { |
| 841 | U_ASSERT(uprv_strlen(type) >= 4 && uprv_strlen(type) <= 8)(static_cast <bool> (:: strlen(type) >= 4 && :: strlen(type) <= 8) ? void (0) : __assert_fail (":: strlen(type) >= 4 && :: strlen(type) <= 8" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 842 | U_ASSERT(uprv_strlen(type) != 4 ||(static_cast <bool> (:: strlen(type) != 4 || (type[0] >= '0' && type[0] <= '9')) ? void (0) : __assert_fail (":: strlen(type) != 4 || (type[0] >= '0' && type[0] <= '9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )) |
| 843 | (type[0] >= '0' && type[0] <= '9'))(static_cast <bool> (:: strlen(type) != 4 || (type[0] >= '0' && type[0] <= '9')) ? void (0) : __assert_fail (":: strlen(type) != 4 || (type[0] >= '0' && type[0] <= '9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 844 | }, |
| 845 | [](const UChar* replacement) { |
| 846 | int32_t len = u_strlenu_strlen_77(replacement); |
| 847 | U_ASSERT(len >= 4 && len <= 8)(static_cast <bool> (len >= 4 && len <= 8 ) ? void (0) : __assert_fail ("len >= 4 && len <= 8" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 848 | U_ASSERT(len != 4 ||(static_cast <bool> (len != 4 || (*replacement >= u'0' && *replacement <= u'9')) ? void (0) : __assert_fail ("len != 4 || (*replacement >= u'0' && *replacement <= u'9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )) |
| 849 | (*replacement >= u'0' &&(static_cast <bool> (len != 4 || (*replacement >= u'0' && *replacement <= u'9')) ? void (0) : __assert_fail ("len != 4 || (*replacement >= u'0' && *replacement <= u'9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )) |
| 850 | *replacement <= u'9'))(static_cast <bool> (len != 4 || (*replacement >= u'0' && *replacement <= u'9')) ? void (0) : __assert_fail ("len != 4 || (*replacement >= u'0' && *replacement <= u'9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 851 | }, |
| 852 | #else |
| 853 | [](const char*) {}, |
| 854 | [](const UChar*) { }, |
| 855 | #endif |
| 856 | status); |
| 857 | } |
| 858 | |
| 859 | /** |
| 860 | * Read the subdivisionAlias data from alias to strings+types+replacementIndexes. |
| 861 | * Allocate length items for types, to store the type field. Allocate length |
| 862 | * items for replacementIndexes, to store the index in the strings for the |
| 863 | * replacement regions. |
| 864 | */ |
| 865 | void |
| 866 | AliasDataBuilder::readSubdivisionAlias( |
| 867 | UResourceBundle* alias, |
| 868 | UniqueCharStrings* strings, |
| 869 | LocalMemory<const char*>& types, |
| 870 | LocalMemory<int32_t>& replacementIndexes, |
| 871 | int32_t &length, |
| 872 | UErrorCode &status) |
| 873 | { |
| 874 | return readAlias( |
| 875 | alias, strings, types, replacementIndexes, length, |
| 876 | #if U_DEBUG1 |
| 877 | [](const char* type) { |
| 878 | U_ASSERT(uprv_strlen(type) >= 3 && uprv_strlen(type) <= 8)(static_cast <bool> (:: strlen(type) >= 3 && :: strlen(type) <= 8) ? void (0) : __assert_fail (":: strlen(type) >= 3 && :: strlen(type) <= 8" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 879 | }, |
| 880 | #else |
| 881 | [](const char*) {}, |
| 882 | #endif |
| 883 | [](const UChar*) { }, |
| 884 | status); |
| 885 | } |
| 886 | |
| 887 | /** |
| 888 | * Initializes the alias data from the ICU resource bundles. The alias data |
| 889 | * contains alias of language, country, script and variants. |
| 890 | * |
| 891 | * If the alias data has already loaded, then this method simply returns without |
| 892 | * doing anything meaningful. |
| 893 | */ |
| 894 | void U_CALLCONV |
| 895 | AliasData::loadData(UErrorCode &status) |
| 896 | { |
| 897 | #ifdef LOCALE_CANONICALIZATION_DEBUG |
| 898 | UDate start = uprv_getRawUTCtimeuprv_getRawUTCtime_77(); |
| 899 | #endif // LOCALE_CANONICALIZATION_DEBUG |
| 900 | ucln_common_registerCleanupucln_common_registerCleanup_77(UCLN_COMMON_LOCALE_ALIAS, cleanup); |
| 901 | AliasDataBuilder builder; |
| 902 | gSingleton = builder.build(status); |
| 903 | #ifdef LOCALE_CANONICALIZATION_DEBUG |
| 904 | UDate end = uprv_getRawUTCtimeuprv_getRawUTCtime_77(); |
| 905 | printf("AliasData::loadData took total %f ms\n", end - start); |
| 906 | #endif // LOCALE_CANONICALIZATION_DEBUG |
| 907 | } |
| 908 | |
| 909 | /** |
| 910 | * Build the alias data from resources. |
| 911 | */ |
| 912 | AliasData* |
| 913 | AliasDataBuilder::build(UErrorCode &status) { |
| 914 | if (U_FAILURE(status)) { return nullptr; } |
| 915 | |
| 916 | LocalUResourceBundlePointer metadata( |
| 917 | ures_openDirectures_openDirect_77(nullptr, "metadata", &status)); |
| 918 | LocalUResourceBundlePointer metadataAlias( |
| 919 | ures_getByKeyures_getByKey_77(metadata.getAlias(), "alias", nullptr, &status)); |
| 920 | LocalUResourceBundlePointer languageAlias( |
| 921 | ures_getByKeyures_getByKey_77(metadataAlias.getAlias(), "language", nullptr, &status)); |
| 922 | LocalUResourceBundlePointer scriptAlias( |
| 923 | ures_getByKeyures_getByKey_77(metadataAlias.getAlias(), "script", nullptr, &status)); |
| 924 | LocalUResourceBundlePointer territoryAlias( |
| 925 | ures_getByKeyures_getByKey_77(metadataAlias.getAlias(), "territory", nullptr, &status)); |
| 926 | LocalUResourceBundlePointer variantAlias( |
| 927 | ures_getByKeyures_getByKey_77(metadataAlias.getAlias(), "variant", nullptr, &status)); |
| 928 | LocalUResourceBundlePointer subdivisionAlias( |
| 929 | ures_getByKeyures_getByKey_77(metadataAlias.getAlias(), "subdivision", nullptr, &status)); |
| 930 | |
| 931 | if (U_FAILURE(status)) { |
| 932 | return nullptr; |
| 933 | } |
| 934 | int32_t languagesLength = 0, scriptLength = 0, territoryLength = 0, |
| 935 | variantLength = 0, subdivisionLength = 0; |
| 936 | |
| 937 | // Read the languageAlias into languageTypes, languageReplacementIndexes |
| 938 | // and strings |
| 939 | UniqueCharStrings strings(status); |
| 940 | LocalMemory<const char*> languageTypes; |
| 941 | LocalMemory<int32_t> languageReplacementIndexes; |
| 942 | readLanguageAlias(languageAlias.getAlias(), |
| 943 | &strings, |
| 944 | languageTypes, |
| 945 | languageReplacementIndexes, |
| 946 | languagesLength, |
| 947 | status); |
| 948 | |
| 949 | // Read the scriptAlias into scriptTypes, scriptReplacementIndexes |
| 950 | // and strings |
| 951 | LocalMemory<const char*> scriptTypes; |
| 952 | LocalMemory<int32_t> scriptReplacementIndexes; |
| 953 | readScriptAlias(scriptAlias.getAlias(), |
| 954 | &strings, |
| 955 | scriptTypes, |
| 956 | scriptReplacementIndexes, |
| 957 | scriptLength, |
| 958 | status); |
| 959 | |
| 960 | // Read the territoryAlias into territoryTypes, territoryReplacementIndexes |
| 961 | // and strings |
| 962 | LocalMemory<const char*> territoryTypes; |
| 963 | LocalMemory<int32_t> territoryReplacementIndexes; |
| 964 | readTerritoryAlias(territoryAlias.getAlias(), |
| 965 | &strings, |
| 966 | territoryTypes, |
| 967 | territoryReplacementIndexes, |
| 968 | territoryLength, status); |
| 969 | |
| 970 | // Read the variantAlias into variantTypes, variantReplacementIndexes |
| 971 | // and strings |
| 972 | LocalMemory<const char*> variantTypes; |
| 973 | LocalMemory<int32_t> variantReplacementIndexes; |
| 974 | readVariantAlias(variantAlias.getAlias(), |
| 975 | &strings, |
| 976 | variantTypes, |
| 977 | variantReplacementIndexes, |
| 978 | variantLength, status); |
| 979 | |
| 980 | // Read the subdivisionAlias into subdivisionTypes, subdivisionReplacementIndexes |
| 981 | // and strings |
| 982 | LocalMemory<const char*> subdivisionTypes; |
| 983 | LocalMemory<int32_t> subdivisionReplacementIndexes; |
| 984 | readSubdivisionAlias(subdivisionAlias.getAlias(), |
| 985 | &strings, |
| 986 | subdivisionTypes, |
| 987 | subdivisionReplacementIndexes, |
| 988 | subdivisionLength, status); |
| 989 | |
| 990 | if (U_FAILURE(status)) { |
| 991 | return nullptr; |
| 992 | } |
| 993 | |
| 994 | // We can only use strings after freeze it. |
| 995 | strings.freeze(); |
| 996 | |
| 997 | // Build the languageMap from languageTypes & languageReplacementIndexes |
| 998 | CharStringMap languageMap(490, status); |
| 999 | for (int32_t i = 0; U_SUCCESS(status) && i < languagesLength; i++) { |
| 1000 | languageMap.put(languageTypes[i], |
| 1001 | strings.get(languageReplacementIndexes[i]), |
| 1002 | status); |
| 1003 | } |
| 1004 | |
| 1005 | // Build the scriptMap from scriptTypes & scriptReplacementIndexes |
| 1006 | CharStringMap scriptMap(1, status); |
| 1007 | for (int32_t i = 0; U_SUCCESS(status) && i < scriptLength; i++) { |
| 1008 | scriptMap.put(scriptTypes[i], |
| 1009 | strings.get(scriptReplacementIndexes[i]), |
| 1010 | status); |
| 1011 | } |
| 1012 | |
| 1013 | // Build the territoryMap from territoryTypes & territoryReplacementIndexes |
| 1014 | CharStringMap territoryMap(650, status); |
| 1015 | for (int32_t i = 0; U_SUCCESS(status) && i < territoryLength; i++) { |
| 1016 | territoryMap.put(territoryTypes[i], |
| 1017 | strings.get(territoryReplacementIndexes[i]), |
| 1018 | status); |
| 1019 | } |
| 1020 | |
| 1021 | // Build the variantMap from variantTypes & variantReplacementIndexes. |
| 1022 | CharStringMap variantMap(2, status); |
| 1023 | for (int32_t i = 0; U_SUCCESS(status) && i < variantLength; i++) { |
| 1024 | variantMap.put(variantTypes[i], |
| 1025 | strings.get(variantReplacementIndexes[i]), |
| 1026 | status); |
| 1027 | } |
| 1028 | |
| 1029 | // Build the subdivisionMap from subdivisionTypes & subdivisionReplacementIndexes. |
| 1030 | CharStringMap subdivisionMap(2, status); |
| 1031 | for (int32_t i = 0; U_SUCCESS(status) && i < subdivisionLength; i++) { |
| 1032 | subdivisionMap.put(subdivisionTypes[i], |
| 1033 | strings.get(subdivisionReplacementIndexes[i]), |
| 1034 | status); |
| 1035 | } |
| 1036 | |
| 1037 | if (U_FAILURE(status)) { |
| 1038 | return nullptr; |
| 1039 | } |
| 1040 | |
| 1041 | // copy hashtables |
| 1042 | auto *data = new AliasData( |
| 1043 | std::move(languageMap), |
| 1044 | std::move(scriptMap), |
| 1045 | std::move(territoryMap), |
| 1046 | std::move(variantMap), |
| 1047 | std::move(subdivisionMap), |
| 1048 | strings.orphanCharStrings()); |
| 1049 | |
| 1050 | if (data == nullptr) { |
| 1051 | status = U_MEMORY_ALLOCATION_ERROR; |
| 1052 | } |
| 1053 | return data; |
| 1054 | } |
| 1055 | |
| 1056 | /** |
| 1057 | * A class that find the replacement values of locale fields by using AliasData. |
| 1058 | */ |
| 1059 | class AliasReplacer { |
| 1060 | public: |
| 1061 | AliasReplacer(UErrorCode& status) : |
| 1062 | language(nullptr), script(nullptr), region(nullptr), |
| 1063 | extensions(nullptr), |
| 1064 | // store value in variants only once |
| 1065 | variants(nullptr, |
| 1066 | ([](UElement e1, UElement e2) -> UBool { |
| 1067 | return 0==uprv_strcmp((const char*)e1.pointer,:: strcmp((const char*)e1.pointer, (const char*)e2.pointer) |
| 1068 | (const char*)e2.pointer):: strcmp((const char*)e1.pointer, (const char*)e2.pointer);}), |
| 1069 | status), |
| 1070 | data(nullptr) { |
| 1071 | } |
| 1072 | ~AliasReplacer() { |
| 1073 | } |
| 1074 | |
| 1075 | // Check the fields inside locale, if need to replace fields, |
| 1076 | // place the the replaced locale ID in out and return true. |
| 1077 | // Otherwise return false for no replacement or error. |
| 1078 | bool replace( |
| 1079 | const Locale& locale, CharString& out, UErrorCode& status); |
| 1080 | |
| 1081 | private: |
| 1082 | const char* language; |
| 1083 | const char* script; |
| 1084 | const char* region; |
| 1085 | const char* extensions; |
| 1086 | UVector variants; |
| 1087 | |
| 1088 | const AliasData* data; |
| 1089 | |
| 1090 | inline bool notEmpty(const char* str) { |
| 1091 | return str && str[0] != NULL_CHAR'\0'; |
| 1092 | } |
| 1093 | |
| 1094 | /** |
| 1095 | * If replacement is neither null nor empty and input is either null or empty, |
| 1096 | * return replacement. |
| 1097 | * If replacement is neither null nor empty but input is not empty, return input. |
| 1098 | * If replacement is either null or empty and type is either null or empty, |
| 1099 | * return input. |
| 1100 | * Otherwise return null. |
| 1101 | * replacement input type return |
| 1102 | * AAA nullptr * AAA |
| 1103 | * AAA BBB * BBB |
| 1104 | * nullptr || "" CCC nullptr CCC |
| 1105 | * nullptr || "" * DDD nullptr |
| 1106 | */ |
| 1107 | inline const char* deleteOrReplace( |
| 1108 | const char* input, const char* type, const char* replacement) { |
| 1109 | return notEmpty(replacement) ? |
| 1110 | ((input == nullptr) ? replacement : input) : |
| 1111 | ((type == nullptr) ? input : nullptr); |
| 1112 | } |
| 1113 | |
| 1114 | inline bool same(const char* a, const char* b) { |
| 1115 | if (a == nullptr && b == nullptr) { |
| 1116 | return true; |
| 1117 | } |
| 1118 | if ((a == nullptr && b != nullptr) || |
| 1119 | (a != nullptr && b == nullptr)) { |
| 1120 | return false; |
| 1121 | } |
| 1122 | return uprv_strcmp(a, b):: strcmp(a, b) == 0; |
| 1123 | } |
| 1124 | |
| 1125 | // Gather fields and generate locale ID into out. |
| 1126 | CharString& outputToString(CharString& out, UErrorCode& status); |
| 1127 | |
| 1128 | // Generate the lookup key. |
| 1129 | CharString& generateKey(const char* language, const char* region, |
| 1130 | const char* variant, CharString& out, |
| 1131 | UErrorCode& status); |
| 1132 | |
| 1133 | void parseLanguageReplacement(const char* replacement, |
| 1134 | const char*& replaceLanguage, |
| 1135 | const char*& replaceScript, |
| 1136 | const char*& replaceRegion, |
| 1137 | const char*& replaceVariant, |
| 1138 | const char*& replaceExtensions, |
| 1139 | UVector& toBeFreed, |
| 1140 | UErrorCode& status); |
| 1141 | |
| 1142 | // Replace by using languageAlias. |
| 1143 | bool replaceLanguage(bool checkLanguage, bool checkRegion, |
| 1144 | bool checkVariants, UVector& toBeFreed, |
| 1145 | UErrorCode& status); |
| 1146 | |
| 1147 | // Replace by using territoryAlias. |
| 1148 | bool replaceTerritory(UVector& toBeFreed, UErrorCode& status); |
| 1149 | |
| 1150 | // Replace by using scriptAlias. |
| 1151 | bool replaceScript(UErrorCode& status); |
| 1152 | |
| 1153 | // Replace by using variantAlias. |
| 1154 | bool replaceVariant(UErrorCode& status); |
| 1155 | |
| 1156 | // Replace by using subdivisionAlias. |
| 1157 | bool replaceSubdivision(StringPiece subdivision, |
| 1158 | CharString& output, UErrorCode& status); |
| 1159 | |
| 1160 | // Replace transformed extensions. |
| 1161 | bool replaceTransformedExtensions( |
| 1162 | CharString& transformedExtensions, CharString& output, UErrorCode& status); |
| 1163 | }; |
| 1164 | |
| 1165 | CharString& |
| 1166 | AliasReplacer::generateKey( |
| 1167 | const char* language, const char* region, const char* variant, |
| 1168 | CharString& out, UErrorCode& status) |
| 1169 | { |
| 1170 | if (U_FAILURE(status)) { return out; } |
| 1171 | out.append(language, status); |
| 1172 | if (notEmpty(region)) { |
| 1173 | out.append(SEP_CHAR'_', status) |
| 1174 | .append(region, status); |
| 1175 | } |
| 1176 | if (notEmpty(variant)) { |
| 1177 | out.append(SEP_CHAR'_', status) |
| 1178 | .append(variant, status); |
| 1179 | } |
| 1180 | return out; |
| 1181 | } |
| 1182 | |
| 1183 | void |
| 1184 | AliasReplacer::parseLanguageReplacement( |
| 1185 | const char* replacement, |
| 1186 | const char*& replacedLanguage, |
| 1187 | const char*& replacedScript, |
| 1188 | const char*& replacedRegion, |
| 1189 | const char*& replacedVariant, |
| 1190 | const char*& replacedExtensions, |
| 1191 | UVector& toBeFreed, |
| 1192 | UErrorCode& status) |
| 1193 | { |
| 1194 | if (U_FAILURE(status)) { |
| 1195 | return; |
| 1196 | } |
| 1197 | replacedScript = replacedRegion = replacedVariant |
| 1198 | = replacedExtensions = nullptr; |
| 1199 | if (uprv_strchr(replacement, '_'):: strchr(replacement, '_') == nullptr) { |
| 1200 | replacedLanguage = replacement; |
| 1201 | // reach the end, just return it. |
| 1202 | return; |
| 1203 | } |
| 1204 | // We have multiple field so we have to allocate and parse |
| 1205 | CharString* str = |
| 1206 | new CharString(replacement, static_cast<int32_t>(uprv_strlen(replacement):: strlen(replacement)), status); |
| 1207 | LocalPointer<CharString> lpStr(str, status); |
| 1208 | toBeFreed.adoptElement(lpStr.orphan(), status); |
| 1209 | if (U_FAILURE(status)) { |
| 1210 | return; |
| 1211 | } |
| 1212 | char* data = str->data(); |
| 1213 | replacedLanguage = (const char*) data; |
| 1214 | char* endOfField = uprv_strchr(data, '_'):: strchr(data, '_'); |
| 1215 | *endOfField = '\0'; // null terminiate it. |
| 1216 | endOfField++; |
| 1217 | const char* start = endOfField; |
| 1218 | endOfField = const_cast<char*>(uprv_strchr(start, '_'):: strchr(start, '_')); |
| 1219 | size_t len = 0; |
| 1220 | if (endOfField == nullptr) { |
| 1221 | len = uprv_strlen(start):: strlen(start); |
| 1222 | } else { |
| 1223 | len = endOfField - start; |
| 1224 | *endOfField = '\0'; // null terminiate it. |
| 1225 | } |
| 1226 | if (len == 4 && uprv_isASCIILetteruprv_isASCIILetter_77(*start)) { |
| 1227 | // Got a script |
| 1228 | replacedScript = start; |
| 1229 | if (endOfField == nullptr) { |
| 1230 | return; |
| 1231 | } |
| 1232 | start = endOfField++; |
| 1233 | endOfField = const_cast<char*>(uprv_strchr(start, '_'):: strchr(start, '_')); |
| 1234 | if (endOfField == nullptr) { |
| 1235 | len = uprv_strlen(start):: strlen(start); |
| 1236 | } else { |
| 1237 | len = endOfField - start; |
| 1238 | *endOfField = '\0'; // null terminiate it. |
| 1239 | } |
| 1240 | } |
| 1241 | if (len >= 2 && len <= 3) { |
| 1242 | // Got a region |
| 1243 | replacedRegion = start; |
| 1244 | if (endOfField == nullptr) { |
| 1245 | return; |
| 1246 | } |
| 1247 | start = endOfField++; |
| 1248 | endOfField = const_cast<char*>(uprv_strchr(start, '_'):: strchr(start, '_')); |
| 1249 | if (endOfField == nullptr) { |
| 1250 | len = uprv_strlen(start):: strlen(start); |
| 1251 | } else { |
| 1252 | len = endOfField - start; |
| 1253 | *endOfField = '\0'; // null terminiate it. |
| 1254 | } |
| 1255 | } |
| 1256 | if (len >= 4) { |
| 1257 | // Got a variant |
| 1258 | replacedVariant = start; |
| 1259 | if (endOfField == nullptr) { |
| 1260 | return; |
| 1261 | } |
| 1262 | start = endOfField++; |
| 1263 | } |
| 1264 | replacedExtensions = start; |
| 1265 | } |
| 1266 | |
| 1267 | bool |
| 1268 | AliasReplacer::replaceLanguage( |
| 1269 | bool checkLanguage, bool checkRegion, |
| 1270 | bool checkVariants, UVector& toBeFreed, UErrorCode& status) |
| 1271 | { |
| 1272 | if (U_FAILURE(status)) { |
| 1273 | return false; |
| 1274 | } |
| 1275 | if ( (checkRegion && region == nullptr) || |
| 1276 | (checkVariants && variants.size() == 0)) { |
| 1277 | // Nothing to search. |
| 1278 | return false; |
| 1279 | } |
| 1280 | int32_t variant_size = checkVariants ? variants.size() : 1; |
| 1281 | // Since we may have more than one variant, we need to loop through them. |
| 1282 | const char* searchLanguage = checkLanguage ? language : "und"; |
| 1283 | const char* searchRegion = checkRegion ? region : nullptr; |
| 1284 | const char* searchVariant = nullptr; |
| 1285 | for (int32_t variant_index = 0; |
| 1286 | variant_index < variant_size; |
| 1287 | variant_index++) { |
| 1288 | if (checkVariants) { |
| 1289 | U_ASSERT(variant_index < variant_size)(static_cast <bool> (variant_index < variant_size) ? void (0) : __assert_fail ("variant_index < variant_size", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 1290 | searchVariant = static_cast<const char*>(variants.elementAt(variant_index)); |
| 1291 | } |
| 1292 | |
| 1293 | if (searchVariant != nullptr && uprv_strlen(searchVariant):: strlen(searchVariant) < 4) { |
| 1294 | // Do not consider ill-formed variant subtag. |
| 1295 | searchVariant = nullptr; |
| 1296 | } |
| 1297 | CharString typeKey; |
| 1298 | generateKey(searchLanguage, searchRegion, searchVariant, typeKey, |
| 1299 | status); |
| 1300 | if (U_FAILURE(status)) { |
| 1301 | return false; |
| 1302 | } |
| 1303 | const char *replacement = data->languageMap().get(typeKey.data()); |
| 1304 | if (replacement == nullptr) { |
| 1305 | // Found no replacement data. |
| 1306 | continue; |
| 1307 | } |
| 1308 | |
| 1309 | const char* replacedLanguage = nullptr; |
| 1310 | const char* replacedScript = nullptr; |
| 1311 | const char* replacedRegion = nullptr; |
| 1312 | const char* replacedVariant = nullptr; |
| 1313 | const char* replacedExtensions = nullptr; |
| 1314 | parseLanguageReplacement(replacement, |
| 1315 | replacedLanguage, |
| 1316 | replacedScript, |
| 1317 | replacedRegion, |
| 1318 | replacedVariant, |
| 1319 | replacedExtensions, |
| 1320 | toBeFreed, |
| 1321 | status); |
| 1322 | replacedLanguage = |
| 1323 | (replacedLanguage != nullptr && uprv_strcmp(replacedLanguage, "und"):: strcmp(replacedLanguage, "und") == 0) ? |
| 1324 | language : replacedLanguage; |
| 1325 | replacedScript = deleteOrReplace(script, nullptr, replacedScript); |
| 1326 | replacedRegion = deleteOrReplace(region, searchRegion, replacedRegion); |
| 1327 | replacedVariant = deleteOrReplace( |
| 1328 | searchVariant, searchVariant, replacedVariant); |
| 1329 | |
| 1330 | if ( same(language, replacedLanguage) && |
| 1331 | same(script, replacedScript) && |
| 1332 | same(region, replacedRegion) && |
| 1333 | same(searchVariant, replacedVariant) && |
| 1334 | replacedExtensions == nullptr) { |
| 1335 | // Replacement produce no changes. |
| 1336 | continue; |
| 1337 | } |
| 1338 | |
| 1339 | language = replacedLanguage; |
| 1340 | region = replacedRegion; |
| 1341 | script = replacedScript; |
| 1342 | if (searchVariant != nullptr) { |
| 1343 | if (notEmpty(replacedVariant)) { |
| 1344 | variants.setElementAt((void*)replacedVariant, variant_index); |
| 1345 | } else { |
| 1346 | variants.removeElementAt(variant_index); |
| 1347 | } |
| 1348 | } |
| 1349 | if (replacedExtensions != nullptr) { |
| 1350 | // DO NOTHING |
| 1351 | // UTS35 does not specify what should we do if we have extensions in the |
| 1352 | // replacement. Currently we know only the following 4 "BCP47 LegacyRules" have |
| 1353 | // extensions in them languageAlias: |
| 1354 | // i_default => en_x_i_default |
| 1355 | // i_enochian => und_x_i_enochian |
| 1356 | // i_mingo => see_x_i_mingo |
| 1357 | // zh_min => nan_x_zh_min |
| 1358 | // But all of them are already changed by code inside ultag_parse() before |
| 1359 | // hitting this code. |
| 1360 | } |
| 1361 | |
| 1362 | // Something changed by language alias data. |
| 1363 | return true; |
| 1364 | } |
| 1365 | // Nothing changed by language alias data. |
| 1366 | return false; |
| 1367 | } |
| 1368 | |
| 1369 | bool |
| 1370 | AliasReplacer::replaceTerritory(UVector& toBeFreed, UErrorCode& status) |
| 1371 | { |
| 1372 | if (U_FAILURE(status)) { |
| 1373 | return false; |
| 1374 | } |
| 1375 | if (region == nullptr) { |
| 1376 | // No region to search. |
| 1377 | return false; |
| 1378 | } |
| 1379 | const char *replacement = data->territoryMap().get(region); |
| 1380 | if (replacement == nullptr) { |
| 1381 | // Found no replacement data for this region. |
| 1382 | return false; |
| 1383 | } |
| 1384 | const char* replacedRegion = replacement; |
| 1385 | const char* firstSpace = uprv_strchr(replacement, ' '):: strchr(replacement, ' '); |
| 1386 | if (firstSpace != nullptr) { |
| 1387 | // If there are are more than one region in the replacement. |
| 1388 | // We need to check which one match based on the language. |
| 1389 | // Cannot use nullptr for language because that will construct |
| 1390 | // the default locale, in that case, use "und" to get the correct |
| 1391 | // locale. |
| 1392 | Locale l = LocaleBuilder() |
| 1393 | .setLanguage(language == nullptr ? "und" : language) |
| 1394 | .setScript(script) |
| 1395 | .build(status); |
| 1396 | l.addLikelySubtags(status); |
| 1397 | const char* likelyRegion = l.getCountry(); |
| 1398 | LocalPointer<CharString> item; |
| 1399 | if (likelyRegion != nullptr && uprv_strlen(likelyRegion):: strlen(likelyRegion) > 0) { |
| 1400 | size_t len = uprv_strlen(likelyRegion):: strlen(likelyRegion); |
| 1401 | const char* foundInReplacement = uprv_strstr(replacement,:: strstr(replacement, likelyRegion) |
| 1402 | likelyRegion):: strstr(replacement, likelyRegion); |
| 1403 | if (foundInReplacement != nullptr) { |
| 1404 | // Assuming the case there are no three letter region code in |
| 1405 | // the replacement of territoryAlias |
| 1406 | U_ASSERT(foundInReplacement == replacement ||(static_cast <bool> (foundInReplacement == replacement || *(foundInReplacement-1) == ' ') ? void (0) : __assert_fail ( "foundInReplacement == replacement || *(foundInReplacement-1) == ' '" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )) |
| 1407 | *(foundInReplacement-1) == ' ')(static_cast <bool> (foundInReplacement == replacement || *(foundInReplacement-1) == ' ') ? void (0) : __assert_fail ( "foundInReplacement == replacement || *(foundInReplacement-1) == ' '" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 1408 | U_ASSERT(foundInReplacement[len] == ' ' ||(static_cast <bool> (foundInReplacement[len] == ' ' || foundInReplacement [len] == '\0') ? void (0) : __assert_fail ("foundInReplacement[len] == ' ' || foundInReplacement[len] == '\\0'" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )) |
| 1409 | foundInReplacement[len] == '\0')(static_cast <bool> (foundInReplacement[len] == ' ' || foundInReplacement [len] == '\0') ? void (0) : __assert_fail ("foundInReplacement[len] == ' ' || foundInReplacement[len] == '\\0'" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 1410 | item.adoptInsteadAndCheckErrorCode( |
| 1411 | new CharString(foundInReplacement, static_cast<int32_t>(len), status), status); |
| 1412 | } |
| 1413 | } |
| 1414 | if (item.isNull() && U_SUCCESS(status)) { |
| 1415 | item.adoptInsteadAndCheckErrorCode( |
| 1416 | new CharString(replacement, |
| 1417 | static_cast<int32_t>(firstSpace - replacement), status), status); |
| 1418 | } |
| 1419 | if (U_FAILURE(status)) { return false; } |
| 1420 | replacedRegion = item->data(); |
| 1421 | toBeFreed.adoptElement(item.orphan(), status); |
| 1422 | if (U_FAILURE(status)) { return false; } |
| 1423 | } |
| 1424 | U_ASSERT(!same(region, replacedRegion))(static_cast <bool> (!same(region, replacedRegion)) ? void (0) : __assert_fail ("!same(region, replacedRegion)", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1425 | region = replacedRegion; |
| 1426 | // The region is changed by data in territory alias. |
| 1427 | return true; |
| 1428 | } |
| 1429 | |
| 1430 | bool |
| 1431 | AliasReplacer::replaceScript(UErrorCode& status) |
| 1432 | { |
| 1433 | if (U_FAILURE(status)) { |
| 1434 | return false; |
| 1435 | } |
| 1436 | if (script == nullptr) { |
| 1437 | // No script to search. |
| 1438 | return false; |
| 1439 | } |
| 1440 | const char *replacement = data->scriptMap().get(script); |
| 1441 | if (replacement == nullptr) { |
| 1442 | // Found no replacement data for this script. |
| 1443 | return false; |
| 1444 | } |
| 1445 | U_ASSERT(!same(script, replacement))(static_cast <bool> (!same(script, replacement)) ? void (0) : __assert_fail ("!same(script, replacement)", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1446 | script = replacement; |
| 1447 | // The script is changed by data in script alias. |
| 1448 | return true; |
| 1449 | } |
| 1450 | |
| 1451 | bool |
| 1452 | AliasReplacer::replaceVariant(UErrorCode& status) |
| 1453 | { |
| 1454 | if (U_FAILURE(status)) { |
| 1455 | return false; |
| 1456 | } |
| 1457 | // Since we may have more than one variant, we need to loop through them. |
| 1458 | for (int32_t i = 0; i < variants.size(); i++) { |
| 1459 | const char* variant = static_cast<const char*>(variants.elementAt(i)); |
| 1460 | const char *replacement = data->variantMap().get(variant); |
| 1461 | if (replacement == nullptr) { |
| 1462 | // Found no replacement data for this variant. |
| 1463 | continue; |
| 1464 | } |
| 1465 | U_ASSERT((uprv_strlen(replacement) >= 5 &&(static_cast <bool> ((:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement [0] <= '9')) ? void (0) : __assert_fail ("(:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement[0] <= '9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )) |
| 1466 | uprv_strlen(replacement) <= 8) ||(static_cast <bool> ((:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement [0] <= '9')) ? void (0) : __assert_fail ("(:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement[0] <= '9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )) |
| 1467 | (uprv_strlen(replacement) == 4 &&(static_cast <bool> ((:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement [0] <= '9')) ? void (0) : __assert_fail ("(:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement[0] <= '9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )) |
| 1468 | replacement[0] >= '0' &&(static_cast <bool> ((:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement [0] <= '9')) ? void (0) : __assert_fail ("(:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement[0] <= '9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )) |
| 1469 | replacement[0] <= '9'))(static_cast <bool> ((:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement [0] <= '9')) ? void (0) : __assert_fail ("(:: strlen(replacement) >= 5 && :: strlen(replacement) <= 8) || (:: strlen(replacement) == 4 && replacement[0] >= '0' && replacement[0] <= '9')" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 1470 | if (!same(variant, replacement)) { |
| 1471 | variants.setElementAt((void*)replacement, i); |
| 1472 | // Special hack to handle hepburn-heploc => alalc97 |
| 1473 | if (uprv_strcmp(variant, "heploc"):: strcmp(variant, "heploc") == 0) { |
| 1474 | for (int32_t j = 0; j < variants.size(); j++) { |
| 1475 | if (uprv_strcmp((const char*)(variants.elementAt(j)),:: strcmp((const char*)(variants.elementAt(j)), "hepburn") |
| 1476 | "hepburn"):: strcmp((const char*)(variants.elementAt(j)), "hepburn") == 0) { |
| 1477 | variants.removeElementAt(j); |
| 1478 | } |
| 1479 | } |
| 1480 | } |
| 1481 | return true; |
| 1482 | } |
| 1483 | } |
| 1484 | return false; |
| 1485 | } |
| 1486 | |
| 1487 | bool |
| 1488 | AliasReplacer::replaceSubdivision( |
| 1489 | StringPiece subdivision, CharString& output, UErrorCode& status) |
| 1490 | { |
| 1491 | if (U_FAILURE(status)) { |
| 1492 | return false; |
| 1493 | } |
| 1494 | const char *replacement = data->subdivisionMap().get(subdivision.data()); |
| 1495 | if (replacement != nullptr) { |
| 1496 | const char* firstSpace = uprv_strchr(replacement, ' '):: strchr(replacement, ' '); |
| 1497 | // Found replacement data for this subdivision. |
| 1498 | size_t len = (firstSpace != nullptr) ? |
| 1499 | (firstSpace - replacement) : uprv_strlen(replacement):: strlen(replacement); |
| 1500 | if (2 <= len && len <= 8) { |
| 1501 | output.append(replacement, static_cast<int32_t>(len), status); |
| 1502 | if (2 == len) { |
| 1503 | // Add 'zzzz' based on changes to UTS #35 for CLDR-14312. |
| 1504 | output.append("zzzz", 4, status); |
| 1505 | } |
| 1506 | } |
| 1507 | return true; |
| 1508 | } |
| 1509 | return false; |
| 1510 | } |
| 1511 | |
| 1512 | bool |
| 1513 | AliasReplacer::replaceTransformedExtensions( |
| 1514 | CharString& transformedExtensions, CharString& output, UErrorCode& status) |
| 1515 | { |
| 1516 | // The content of the transformedExtensions will be modified in this |
| 1517 | // function to NUL-terminating (tkey-tvalue) pairs. |
| 1518 | if (U_FAILURE(status)) { |
| 1519 | return false; |
| 1520 | } |
| 1521 | int32_t len = transformedExtensions.length(); |
| 1522 | const char* str = transformedExtensions.data(); |
| 1523 | const char* tkey = ultag_getTKeyStartultag_getTKeyStart_77(str); |
| 1524 | int32_t tlangLen = (tkey == str) ? 0 : |
| 1525 | ((tkey == nullptr) ? len : static_cast<int32_t>((tkey - str - 1))); |
| 1526 | if (tlangLen > 0) { |
| 1527 | Locale tlang = LocaleBuilder() |
| 1528 | .setLanguageTag(StringPiece(str, tlangLen)) |
| 1529 | .build(status); |
| 1530 | tlang.canonicalize(status); |
| 1531 | output = tlang.toLanguageTag<CharString>(status); |
| 1532 | if (U_FAILURE(status)) { |
| 1533 | return false; |
| 1534 | } |
| 1535 | T_CString_toLowerCaseT_CString_toLowerCase_77(output.data()); |
| 1536 | } |
| 1537 | if (tkey != nullptr) { |
| 1538 | // We need to sort the tfields by tkey |
| 1539 | UVector tfields(status); |
| 1540 | if (U_FAILURE(status)) { |
| 1541 | return false; |
| 1542 | } |
| 1543 | do { |
| 1544 | const char* tvalue = uprv_strchr(tkey, '-'):: strchr(tkey, '-'); |
| 1545 | if (tvalue == nullptr) { |
| 1546 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 1547 | return false; |
| 1548 | } |
| 1549 | const char* nextTKey = ultag_getTKeyStartultag_getTKeyStart_77(tvalue); |
| 1550 | if (nextTKey != nullptr) { |
| 1551 | *const_cast<char*>(nextTKey - 1) = '\0'; // NUL terminate tvalue |
| 1552 | } |
| 1553 | tfields.insertElementAt((void*)tkey, tfields.size(), status); |
| 1554 | if (U_FAILURE(status)) { |
| 1555 | return false; |
| 1556 | } |
| 1557 | tkey = nextTKey; |
| 1558 | } while (tkey != nullptr); |
| 1559 | tfields.sort([](UElement e1, UElement e2) -> int32_t { |
| 1560 | return uprv_strcmp((const char*)e1.pointer, (const char*)e2.pointer):: strcmp((const char*)e1.pointer, (const char*)e2.pointer); |
| 1561 | }, status); |
| 1562 | for (int32_t i = 0; i < tfields.size(); i++) { |
| 1563 | if (output.length() > 0) { |
| 1564 | output.append('-', status); |
| 1565 | } |
| 1566 | const char* tfield = static_cast<const char*>(tfields.elementAt(i)); |
| 1567 | const char* tvalue = uprv_strchr(tfield, '-'):: strchr(tfield, '-'); |
| 1568 | if (tvalue == nullptr) { |
| 1569 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 1570 | return false; |
| 1571 | } |
| 1572 | // Split the "tkey-tvalue" pair string so that we can canonicalize the tvalue. |
| 1573 | *const_cast<char*>(tvalue++) = '\0'; // NUL terminate tkey |
| 1574 | output.append(tfield, status).append('-', status); |
| 1575 | std::optional<std::string_view> bcpTValue = ulocimp_toBcpTypeulocimp_toBcpType_77(tfield, tvalue); |
| 1576 | output.append(bcpTValue.has_value() ? *bcpTValue : tvalue, status); |
| 1577 | } |
| 1578 | } |
| 1579 | if (U_FAILURE(status)) { |
| 1580 | return false; |
| 1581 | } |
| 1582 | return true; |
| 1583 | } |
| 1584 | |
| 1585 | CharString& |
| 1586 | AliasReplacer::outputToString( |
| 1587 | CharString& out, UErrorCode& status) |
| 1588 | { |
| 1589 | if (U_FAILURE(status)) { return out; } |
| 1590 | out.append(language, status); |
| 1591 | if (notEmpty(script)) { |
| 1592 | out.append(SEP_CHAR'_', status) |
| 1593 | .append(script, status); |
| 1594 | } |
| 1595 | if (notEmpty(region)) { |
| 1596 | out.append(SEP_CHAR'_', status) |
| 1597 | .append(region, status); |
| 1598 | } |
| 1599 | if (variants.size() > 0) { |
| 1600 | if (!notEmpty(script) && !notEmpty(region)) { |
| 1601 | out.append(SEP_CHAR'_', status); |
| 1602 | } |
| 1603 | variants.sort([](UElement e1, UElement e2) -> int32_t { |
| 1604 | return uprv_strcmp((const char*)e1.pointer, (const char*)e2.pointer):: strcmp((const char*)e1.pointer, (const char*)e2.pointer); |
| 1605 | }, status); |
| 1606 | int32_t variantsStart = out.length(); |
| 1607 | for (int32_t i = 0; i < variants.size(); i++) { |
| 1608 | out.append(SEP_CHAR'_', status) |
| 1609 | .append(static_cast<const char*>(variants.elementAt(i)), |
| 1610 | status); |
| 1611 | } |
| 1612 | T_CString_toUpperCaseT_CString_toUpperCase_77(out.data() + variantsStart); |
| 1613 | } |
| 1614 | if (notEmpty(extensions)) { |
| 1615 | CharString tmp("und_", status); |
| 1616 | tmp.append(extensions, status); |
| 1617 | Locale tmpLocale(tmp.data()); |
| 1618 | // only support x extension inside CLDR for now. |
| 1619 | U_ASSERT(extensions[0] == 'x')(static_cast <bool> (extensions[0] == 'x') ? void (0) : __assert_fail ("extensions[0] == 'x'", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1620 | out.append(tmpLocale.getName() + 1, status); |
| 1621 | } |
| 1622 | return out; |
| 1623 | } |
| 1624 | |
| 1625 | bool |
| 1626 | AliasReplacer::replace(const Locale& locale, CharString& out, UErrorCode& status) |
| 1627 | { |
| 1628 | data = AliasData::singleton(status); |
| 1629 | if (U_FAILURE(status)) { |
| 1630 | return false; |
| 1631 | } |
| 1632 | U_ASSERT(data != nullptr)(static_cast <bool> (data != nullptr) ? void (0) : __assert_fail ("data != nullptr", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1633 | out.clear(); |
| 1634 | language = locale.getLanguage(); |
| 1635 | if (!notEmpty(language)) { |
| 1636 | language = nullptr; |
| 1637 | } |
| 1638 | script = locale.getScript(); |
| 1639 | if (!notEmpty(script)) { |
| 1640 | script = nullptr; |
| 1641 | } |
| 1642 | region = locale.getCountry(); |
| 1643 | if (!notEmpty(region)) { |
| 1644 | region = nullptr; |
| 1645 | } |
| 1646 | const char* variantsStr = locale.getVariant(); |
| 1647 | CharString variantsBuff(variantsStr, -1, status); |
| 1648 | if (!variantsBuff.isEmpty()) { |
| 1649 | if (U_FAILURE(status)) { return false; } |
| 1650 | char* start = variantsBuff.data(); |
| 1651 | T_CString_toLowerCaseT_CString_toLowerCase_77(start); |
| 1652 | char* end; |
| 1653 | while ((end = uprv_strchr(start, SEP_CHAR):: strchr(start, '_')) != nullptr && |
| 1654 | U_SUCCESS(status)) { |
| 1655 | *end = NULL_CHAR'\0'; // null terminate inside variantsBuff |
| 1656 | // do not add "" or duplicate data to variants |
| 1657 | if (*start && !variants.contains(start)) { |
| 1658 | variants.addElement(start, status); |
| 1659 | } |
| 1660 | start = end + 1; |
| 1661 | } |
| 1662 | // do not add "" or duplicate data to variants |
| 1663 | if (*start && !variants.contains(start)) { |
| 1664 | variants.addElement(start, status); |
| 1665 | } |
| 1666 | } |
| 1667 | if (U_FAILURE(status)) { return false; } |
| 1668 | |
| 1669 | // Sort the variants |
| 1670 | variants.sort([](UElement e1, UElement e2) -> int32_t { |
| 1671 | return uprv_strcmp((const char*)e1.pointer, (const char*)e2.pointer):: strcmp((const char*)e1.pointer, (const char*)e2.pointer); |
| 1672 | }, status); |
| 1673 | |
| 1674 | // A changed count to assert when loop too many times. |
| 1675 | int changed = 0; |
| 1676 | // A UVector to to hold CharString allocated by the replace* method |
| 1677 | // and freed when out of scope from his function. |
| 1678 | UVector stringsToBeFreed([](void *obj) { delete static_cast<CharString*>(obj); }, |
| 1679 | nullptr, 10, status); |
| 1680 | while (U_SUCCESS(status)) { |
| 1681 | // Something wrong with the data cause looping here more than 10 times |
| 1682 | // already. |
| 1683 | U_ASSERT(changed < 5)(static_cast <bool> (changed < 5) ? void (0) : __assert_fail ("changed < 5", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1684 | // From observation of key in data/misc/metadata.txt |
| 1685 | // we know currently we only need to search in the following combination |
| 1686 | // of fields for type in languageAlias: |
| 1687 | // * lang_region_variant |
| 1688 | // * lang_region |
| 1689 | // * lang_variant |
| 1690 | // * lang |
| 1691 | // * und_variant |
| 1692 | // This assumption is ensured by the U_ASSERT in readLanguageAlias |
| 1693 | // |
| 1694 | // lang REGION variant |
| 1695 | if ( replaceLanguage(true, true, true, stringsToBeFreed, status) || |
| 1696 | replaceLanguage(true, true, false, stringsToBeFreed, status) || |
| 1697 | replaceLanguage(true, false, true, stringsToBeFreed, status) || |
| 1698 | replaceLanguage(true, false, false, stringsToBeFreed, status) || |
| 1699 | replaceLanguage(false,false, true, stringsToBeFreed, status) || |
| 1700 | replaceTerritory(stringsToBeFreed, status) || |
| 1701 | replaceScript(status) || |
| 1702 | replaceVariant(status)) { |
| 1703 | // Some values in data is changed, try to match from the beginning |
| 1704 | // again. |
| 1705 | changed++; |
| 1706 | continue; |
| 1707 | } |
| 1708 | // Nothing changed. Break out. |
| 1709 | break; |
| 1710 | } // while(1) |
| 1711 | |
| 1712 | if (U_FAILURE(status)) { return false; } |
| 1713 | // Nothing changed and we know the order of the variants are not change |
| 1714 | // because we have no variant or only one. |
| 1715 | const char* extensionsStr = locale_getKeywordsStartlocale_getKeywordsStart_77(locale.getName()); |
| 1716 | if (changed == 0 && variants.size() <= 1 && extensionsStr == nullptr) { |
| 1717 | return false; |
| 1718 | } |
| 1719 | outputToString(out, status); |
| 1720 | if (U_FAILURE(status)) { |
| 1721 | return false; |
| 1722 | } |
| 1723 | if (extensionsStr != nullptr) { |
| 1724 | changed = 0; |
| 1725 | Locale temp(locale); |
| 1726 | LocalPointer<icu::StringEnumeration> iter(locale.createKeywords(status)); |
| 1727 | if (U_SUCCESS(status) && !iter.isNull()) { |
| 1728 | const char* key; |
| 1729 | while ((key = iter->next(nullptr, status)) != nullptr) { |
| 1730 | if (uprv_strcmp("sd", key):: strcmp("sd", key) == 0 || uprv_strcmp("rg", key):: strcmp("rg", key) == 0 || |
| 1731 | uprv_strcmp("t", key):: strcmp("t", key) == 0) { |
| 1732 | auto value = locale.getKeywordValue<CharString>(key, status); |
| 1733 | if (U_FAILURE(status)) { |
| 1734 | status = U_ZERO_ERROR; |
| 1735 | continue; |
| 1736 | } |
| 1737 | CharString replacement; |
| 1738 | if (uprv_strlen(key):: strlen(key) == 2) { |
| 1739 | if (replaceSubdivision(value.toStringPiece(), replacement, status)) { |
| 1740 | changed++; |
| 1741 | temp.setKeywordValue(key, replacement.data(), status); |
| 1742 | } |
| 1743 | } else { |
| 1744 | U_ASSERT(uprv_strcmp(key, "t") == 0)(static_cast <bool> (:: strcmp(key, "t") == 0) ? void ( 0) : __assert_fail (":: strcmp(key, \"t\") == 0", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1745 | if (replaceTransformedExtensions(value, replacement, status)) { |
| 1746 | changed++; |
| 1747 | temp.setKeywordValue(key, replacement.data(), status); |
| 1748 | } |
| 1749 | } |
| 1750 | if (U_FAILURE(status)) { |
| 1751 | return false; |
| 1752 | } |
| 1753 | } |
| 1754 | } |
| 1755 | } |
| 1756 | if (changed != 0) { |
| 1757 | extensionsStr = locale_getKeywordsStartlocale_getKeywordsStart_77(temp.getName()); |
| 1758 | } |
| 1759 | out.append(extensionsStr, status); |
| 1760 | } |
| 1761 | if (U_FAILURE(status)) { |
| 1762 | return false; |
| 1763 | } |
| 1764 | // If the tag is not changed, return. |
| 1765 | if (uprv_strcmp(out.data(), locale.getName()):: strcmp(out.data(), locale.getName()) == 0) { |
| 1766 | out.clear(); |
| 1767 | return false; |
| 1768 | } |
| 1769 | return true; |
| 1770 | } |
| 1771 | |
| 1772 | // Return true if the locale is changed during canonicalization. |
| 1773 | // The replaced value then will be put into out. |
| 1774 | bool |
| 1775 | canonicalizeLocale(const Locale& locale, CharString& out, UErrorCode& status) |
| 1776 | { |
| 1777 | if (U_FAILURE(status)) { return false; } |
| 1778 | AliasReplacer replacer(status); |
| 1779 | return replacer.replace(locale, out, status); |
| 1780 | } |
| 1781 | |
| 1782 | // Function to optimize for known cases without so we can skip the loading |
| 1783 | // of resources in the startup time until we really need it. |
| 1784 | bool |
| 1785 | isKnownCanonicalizedLocale(const char* locale, UErrorCode& status) |
| 1786 | { |
| 1787 | if (U_FAILURE(status)) { return false; } |
| 1788 | |
| 1789 | if ( uprv_strcmp(locale, "c"):: strcmp(locale, "c") == 0 || |
| 1790 | uprv_strcmp(locale, "en"):: strcmp(locale, "en") == 0 || |
| 1791 | uprv_strcmp(locale, "en_US"):: strcmp(locale, "en_US") == 0) { |
| 1792 | return true; |
| 1793 | } |
| 1794 | |
| 1795 | // common well-known Canonicalized. |
| 1796 | umtx_initOnce(gKnownCanonicalizedInitOnce, |
| 1797 | &loadKnownCanonicalized, status); |
| 1798 | if (U_FAILURE(status)) { |
| 1799 | return false; |
| 1800 | } |
| 1801 | U_ASSERT(gKnownCanonicalized != nullptr)(static_cast <bool> (gKnownCanonicalized != nullptr) ? void (0) : __assert_fail ("gKnownCanonicalized != nullptr", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1802 | return uhash_getiuhash_geti_77(gKnownCanonicalized, locale) != 0; |
| 1803 | } |
| 1804 | |
| 1805 | } // namespace |
| 1806 | |
| 1807 | U_NAMESPACE_END} |
| 1808 | |
| 1809 | // Function for testing. |
| 1810 | U_EXPORT const char* const* |
| 1811 | ulocimp_getKnownCanonicalizedLocaleForTestulocimp_getKnownCanonicalizedLocaleForTest_77(int32_t& length) |
| 1812 | { |
| 1813 | U_NAMESPACE_USEusing namespace icu_77; |
| 1814 | length = UPRV_LENGTHOF(KNOWN_CANONICALIZED)(int32_t)(sizeof(KNOWN_CANONICALIZED)/sizeof((KNOWN_CANONICALIZED )[0])); |
| 1815 | return KNOWN_CANONICALIZED; |
| 1816 | } |
| 1817 | |
| 1818 | // Function for testing. |
| 1819 | U_EXPORT bool |
| 1820 | ulocimp_isCanonicalizedLocaleForTestulocimp_isCanonicalizedLocaleForTest_77(const char* localeName) |
| 1821 | { |
| 1822 | U_NAMESPACE_USEusing namespace icu_77; |
| 1823 | Locale l(localeName); |
| 1824 | UErrorCode status = U_ZERO_ERROR; |
| 1825 | CharString temp; |
| 1826 | return !canonicalizeLocale(l, temp, status) && U_SUCCESS(status); |
| 1827 | } |
| 1828 | |
| 1829 | U_NAMESPACE_BEGINnamespace icu_77 { |
| 1830 | |
| 1831 | Locale& Locale::init(const char* localeID, UBool canonicalize) |
| 1832 | { |
| 1833 | return localeID == nullptr ? *this = getDefault() : init(StringPiece{localeID}, canonicalize); |
| 1834 | } |
| 1835 | |
| 1836 | /*This function initializes a Locale from a C locale ID*/ |
| 1837 | Locale& Locale::init(StringPiece localeID, UBool canonicalize) |
| 1838 | { |
| 1839 | fIsBogus = false; |
| 1840 | /* Free our current storage */ |
| 1841 | if ((baseName != fullName) && (baseName != fullNameBuffer)) { |
| 1842 | uprv_freeuprv_free_77(baseName); |
| 1843 | } |
| 1844 | baseName = nullptr; |
| 1845 | if(fullName != fullNameBuffer) { |
| 1846 | uprv_freeuprv_free_77(fullName); |
| 1847 | fullName = fullNameBuffer; |
| 1848 | } |
| 1849 | |
| 1850 | // not a loop: |
| 1851 | // just an easy way to have a common error-exit |
| 1852 | // without goto and without another function |
| 1853 | do { |
| 1854 | char *separator; |
| 1855 | char *field[5] = {nullptr}; |
| 1856 | int32_t fieldLen[5] = {0}; |
| 1857 | int32_t fieldIdx; |
| 1858 | int32_t variantField; |
| 1859 | int32_t length; |
| 1860 | UErrorCode err; |
| 1861 | |
| 1862 | /* preset all fields to empty */ |
| 1863 | language[0] = script[0] = country[0] = 0; |
| 1864 | |
| 1865 | const auto parse = [canonicalize](std::string_view localeID, |
| 1866 | char* name, |
| 1867 | int32_t nameCapacity, |
| 1868 | UErrorCode& status) { |
| 1869 | return ByteSinkUtil::viaByteSinkToTerminatedChars( |
| 1870 | name, nameCapacity, |
| 1871 | [&](ByteSink& sink, UErrorCode& status) { |
| 1872 | if (canonicalize) { |
| 1873 | ulocimp_canonicalizeulocimp_canonicalize_77(localeID, sink, status); |
| 1874 | } else { |
| 1875 | ulocimp_getNameulocimp_getName_77(localeID, sink, status); |
| 1876 | } |
| 1877 | }, |
| 1878 | status); |
| 1879 | }; |
| 1880 | |
| 1881 | // "canonicalize" the locale ID to ICU/Java format |
| 1882 | err = U_ZERO_ERROR; |
| 1883 | length = parse(localeID, fullName, sizeof fullNameBuffer, err); |
| 1884 | |
| 1885 | if (err == U_BUFFER_OVERFLOW_ERROR || length >= static_cast<int32_t>(sizeof(fullNameBuffer))) { |
| 1886 | U_ASSERT(baseName == nullptr)(static_cast <bool> (baseName == nullptr) ? void (0) : __assert_fail ("baseName == nullptr", __builtin_FILE (), __builtin_LINE () , __extension__ __PRETTY_FUNCTION__)); |
| 1887 | /*Go to heap for the fullName if necessary*/ |
| 1888 | char* newFullName = static_cast<char*>(uprv_mallocuprv_malloc_77(sizeof(char) * (length + 1))); |
| 1889 | if (newFullName == nullptr) { |
| 1890 | break; // error: out of memory |
| 1891 | } |
| 1892 | fullName = newFullName; |
| 1893 | err = U_ZERO_ERROR; |
| 1894 | length = parse(localeID, fullName, length + 1, err); |
| 1895 | } |
| 1896 | if(U_FAILURE(err) || err == U_STRING_NOT_TERMINATED_WARNING) { |
| 1897 | /* should never occur */ |
| 1898 | break; |
| 1899 | } |
| 1900 | |
| 1901 | variantBegin = length; |
| 1902 | |
| 1903 | /* after uloc_getName/canonicalize() we know that only '_' are separators */ |
| 1904 | /* But _ could also appeared in timezone such as "en@timezone=America/Los_Angeles" */ |
| 1905 | separator = field[0] = fullName; |
Value stored to 'separator' is never read | |
| 1906 | fieldIdx = 1; |
| 1907 | char* at = uprv_strchr(fullName, '@'):: strchr(fullName, '@'); |
| 1908 | while ((separator = uprv_strchr(field[fieldIdx-1], SEP_CHAR):: strchr(field[fieldIdx-1], '_')) != nullptr && |
| 1909 | fieldIdx < UPRV_LENGTHOF(field)(int32_t)(sizeof(field)/sizeof((field)[0]))-1 && |
| 1910 | (at == nullptr || separator < at)) { |
| 1911 | field[fieldIdx] = separator + 1; |
| 1912 | fieldLen[fieldIdx - 1] = static_cast<int32_t>(separator - field[fieldIdx - 1]); |
| 1913 | fieldIdx++; |
| 1914 | } |
| 1915 | // variant may contain @foo or .foo POSIX cruft; remove it |
| 1916 | separator = uprv_strchr(field[fieldIdx-1], '@'):: strchr(field[fieldIdx-1], '@'); |
| 1917 | char* sep2 = uprv_strchr(field[fieldIdx-1], '.'):: strchr(field[fieldIdx-1], '.'); |
| 1918 | if (separator!=nullptr || sep2!=nullptr) { |
| 1919 | if (separator==nullptr || (sep2!=nullptr && separator > sep2)) { |
| 1920 | separator = sep2; |
| 1921 | } |
| 1922 | fieldLen[fieldIdx - 1] = static_cast<int32_t>(separator - field[fieldIdx - 1]); |
| 1923 | } else { |
| 1924 | fieldLen[fieldIdx - 1] = length - static_cast<int32_t>(field[fieldIdx - 1] - fullName); |
| 1925 | } |
| 1926 | |
| 1927 | if (fieldLen[0] >= static_cast<int32_t>(sizeof(language))) |
| 1928 | { |
| 1929 | break; // error: the language field is too long |
| 1930 | } |
| 1931 | |
| 1932 | variantField = 1; /* Usually the 2nd one, except when a script or country is also used. */ |
| 1933 | if (fieldLen[0] > 0) { |
| 1934 | /* We have a language */ |
| 1935 | uprv_memcpy(language, fullName, fieldLen[0])do { clang diagnostic push
clang diagnostic ignored "-Waddress" (static_cast <bool> (language != __null) ? void (0) : __assert_fail ("language != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); (static_cast <bool > (fullName != __null) ? void (0) : __assert_fail ("fullName != __null" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); clang diagnostic pop :: memcpy(language, fullName, fieldLen [0]); } while (false); |
| 1936 | language[fieldLen[0]] = 0; |
| 1937 | } |
| 1938 | if (fieldLen[1] == 4 && uprv_isASCIILetteruprv_isASCIILetter_77(field[1][0]) && |
| 1939 | uprv_isASCIILetteruprv_isASCIILetter_77(field[1][1]) && uprv_isASCIILetteruprv_isASCIILetter_77(field[1][2]) && |
| 1940 | uprv_isASCIILetteruprv_isASCIILetter_77(field[1][3])) { |
| 1941 | /* We have at least a script */ |
| 1942 | uprv_memcpy(script, field[1], fieldLen[1])do { clang diagnostic push
clang diagnostic ignored "-Waddress" (static_cast <bool> (script != __null) ? void (0) : __assert_fail ("script != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); (static_cast <bool> (field[1] != __null) ? void (0) : __assert_fail ("field[1] != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); clang diagnostic pop :: memcpy(script, field[1], fieldLen[1]); } while (false); |
| 1943 | script[fieldLen[1]] = 0; |
| 1944 | variantField++; |
| 1945 | } |
| 1946 | |
| 1947 | if (fieldLen[variantField] == 2 || fieldLen[variantField] == 3) { |
| 1948 | /* We have a country */ |
| 1949 | uprv_memcpy(country, field[variantField], fieldLen[variantField])do { clang diagnostic push
clang diagnostic ignored "-Waddress" (static_cast <bool> (country != __null) ? void (0) : __assert_fail ("country != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); (static_cast <bool> (field[variantField ] != __null) ? void (0) : __assert_fail ("field[variantField] != __null" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); clang diagnostic pop :: memcpy(country, field[variantField ], fieldLen[variantField]); } while (false); |
| 1950 | country[fieldLen[variantField]] = 0; |
| 1951 | variantField++; |
| 1952 | } else if (fieldLen[variantField] == 0) { |
| 1953 | variantField++; /* script or country empty but variant in next field (i.e. en__POSIX) */ |
| 1954 | } |
| 1955 | |
| 1956 | if (fieldLen[variantField] > 0) { |
| 1957 | /* We have a variant */ |
| 1958 | variantBegin = static_cast<int32_t>(field[variantField] - fullName); |
| 1959 | } |
| 1960 | |
| 1961 | err = U_ZERO_ERROR; |
| 1962 | initBaseName(err); |
| 1963 | if (U_FAILURE(err)) { |
| 1964 | break; |
| 1965 | } |
| 1966 | |
| 1967 | if (canonicalize) { |
| 1968 | if (!isKnownCanonicalizedLocale(fullName, err)) { |
| 1969 | CharString replaced; |
| 1970 | // Not sure it is already canonicalized |
| 1971 | if (canonicalizeLocale(*this, replaced, err)) { |
| 1972 | U_ASSERT(U_SUCCESS(err))(static_cast <bool> (U_SUCCESS(err)) ? void (0) : __assert_fail ("U_SUCCESS(err)", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); |
| 1973 | // If need replacement, call init again. |
| 1974 | init(replaced.data(), false); |
| 1975 | } |
| 1976 | if (U_FAILURE(err)) { |
| 1977 | break; |
| 1978 | } |
| 1979 | } |
| 1980 | } // if (canonicalize) { |
| 1981 | |
| 1982 | // successful end of init() |
| 1983 | return *this; |
| 1984 | } while(0); /*loop doesn't iterate*/ |
| 1985 | |
| 1986 | // when an error occurs, then set this object to "bogus" (there is no UErrorCode here) |
| 1987 | setToBogus(); |
| 1988 | |
| 1989 | return *this; |
| 1990 | } |
| 1991 | |
| 1992 | /* |
| 1993 | * Set up the base name. |
| 1994 | * If there are no key words, it's exactly the full name. |
| 1995 | * If key words exist, it's the full name truncated at the '@' character. |
| 1996 | * Need to set up both at init() and after setting a keyword. |
| 1997 | */ |
| 1998 | void |
| 1999 | Locale::initBaseName(UErrorCode &status) { |
| 2000 | if (U_FAILURE(status)) { |
| 2001 | return; |
| 2002 | } |
| 2003 | U_ASSERT(baseName==nullptr || baseName==fullName)(static_cast <bool> (baseName==nullptr || baseName==fullName ) ? void (0) : __assert_fail ("baseName==nullptr || baseName==fullName" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 2004 | const char *atPtr = uprv_strchr(fullName, '@'):: strchr(fullName, '@'); |
| 2005 | const char *eqPtr = uprv_strchr(fullName, '='):: strchr(fullName, '='); |
| 2006 | if (atPtr && eqPtr && atPtr < eqPtr) { |
| 2007 | // Key words exist. |
| 2008 | int32_t baseNameLength = static_cast<int32_t>(atPtr - fullName); |
| 2009 | char* newBaseName = static_cast<char*>(uprv_mallocuprv_malloc_77(baseNameLength + 1)); |
| 2010 | if (newBaseName == nullptr) { |
| 2011 | status = U_MEMORY_ALLOCATION_ERROR; |
| 2012 | return; |
| 2013 | } |
| 2014 | baseName = newBaseName; |
| 2015 | uprv_strncpy(baseName, fullName, baseNameLength):: strncpy(baseName, fullName, baseNameLength); |
| 2016 | baseName[baseNameLength] = 0; |
| 2017 | |
| 2018 | // The original computation of variantBegin leaves it equal to the length |
| 2019 | // of fullName if there is no variant. It should instead be |
| 2020 | // the length of the baseName. |
| 2021 | if (variantBegin > baseNameLength) { |
| 2022 | variantBegin = baseNameLength; |
| 2023 | } |
| 2024 | } else { |
| 2025 | baseName = fullName; |
| 2026 | } |
| 2027 | } |
| 2028 | |
| 2029 | |
| 2030 | int32_t |
| 2031 | Locale::hashCode() const |
| 2032 | { |
| 2033 | return ustr_hashCharsNustr_hashCharsN_77(fullName, static_cast<int32_t>(uprv_strlen(fullName):: strlen(fullName))); |
| 2034 | } |
| 2035 | |
| 2036 | void |
| 2037 | Locale::setToBogus() { |
| 2038 | /* Free our current storage */ |
| 2039 | if((baseName != fullName) && (baseName != fullNameBuffer)) { |
| 2040 | uprv_freeuprv_free_77(baseName); |
| 2041 | } |
| 2042 | baseName = nullptr; |
| 2043 | if(fullName != fullNameBuffer) { |
| 2044 | uprv_freeuprv_free_77(fullName); |
| 2045 | fullName = fullNameBuffer; |
| 2046 | } |
| 2047 | *fullNameBuffer = 0; |
| 2048 | *language = 0; |
| 2049 | *script = 0; |
| 2050 | *country = 0; |
| 2051 | fIsBogus = true; |
| 2052 | variantBegin = 0; |
| 2053 | } |
| 2054 | |
| 2055 | const Locale& U_EXPORT2 |
| 2056 | Locale::getDefault() |
| 2057 | { |
| 2058 | { |
| 2059 | Mutex lock(&gDefaultLocaleMutex); |
| 2060 | if (gDefaultLocale != nullptr) { |
| 2061 | return *gDefaultLocale; |
| 2062 | } |
| 2063 | } |
| 2064 | UErrorCode status = U_ZERO_ERROR; |
| 2065 | return *locale_set_default_internal(nullptr, status); |
| 2066 | } |
| 2067 | |
| 2068 | |
| 2069 | |
| 2070 | void U_EXPORT2 |
| 2071 | Locale::setDefault( const Locale& newLocale, |
| 2072 | UErrorCode& status) |
| 2073 | { |
| 2074 | if (U_FAILURE(status)) { |
| 2075 | return; |
| 2076 | } |
| 2077 | |
| 2078 | /* Set the default from the full name string of the supplied locale. |
| 2079 | * This is a convenient way to access the default locale caching mechanisms. |
| 2080 | */ |
| 2081 | const char *localeID = newLocale.getName(); |
| 2082 | locale_set_default_internal(localeID, status); |
| 2083 | } |
| 2084 | |
| 2085 | void |
| 2086 | Locale::addLikelySubtags(UErrorCode& status) { |
| 2087 | if (U_FAILURE(status)) { |
| 2088 | return; |
| 2089 | } |
| 2090 | |
| 2091 | CharString maximizedLocaleID = ulocimp_addLikelySubtagsulocimp_addLikelySubtags_77(fullName, status); |
| 2092 | |
| 2093 | if (U_FAILURE(status)) { |
| 2094 | return; |
| 2095 | } |
| 2096 | |
| 2097 | init(maximizedLocaleID.data(), /*canonicalize=*/false); |
| 2098 | if (isBogus()) { |
| 2099 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2100 | } |
| 2101 | } |
| 2102 | |
| 2103 | void |
| 2104 | Locale::minimizeSubtags(UErrorCode& status) { |
| 2105 | Locale::minimizeSubtags(false, status); |
| 2106 | } |
| 2107 | void |
| 2108 | Locale::minimizeSubtags(bool favorScript, UErrorCode& status) { |
| 2109 | if (U_FAILURE(status)) { |
| 2110 | return; |
| 2111 | } |
| 2112 | |
| 2113 | CharString minimizedLocaleID = ulocimp_minimizeSubtagsulocimp_minimizeSubtags_77(fullName, favorScript, status); |
| 2114 | |
| 2115 | if (U_FAILURE(status)) { |
| 2116 | return; |
| 2117 | } |
| 2118 | |
| 2119 | init(minimizedLocaleID.data(), /*canonicalize=*/false); |
| 2120 | if (isBogus()) { |
| 2121 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2122 | } |
| 2123 | } |
| 2124 | |
| 2125 | void |
| 2126 | Locale::canonicalize(UErrorCode& status) { |
| 2127 | if (U_FAILURE(status)) { |
| 2128 | return; |
| 2129 | } |
| 2130 | if (isBogus()) { |
| 2131 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2132 | return; |
| 2133 | } |
| 2134 | CharString uncanonicalized(fullName, status); |
| 2135 | if (U_FAILURE(status)) { |
| 2136 | return; |
| 2137 | } |
| 2138 | init(uncanonicalized.data(), /*canonicalize=*/true); |
| 2139 | if (isBogus()) { |
| 2140 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2141 | } |
| 2142 | } |
| 2143 | |
| 2144 | Locale U_EXPORT2 |
| 2145 | Locale::forLanguageTag(StringPiece tag, UErrorCode& status) |
| 2146 | { |
| 2147 | Locale result(Locale::eBOGUS); |
| 2148 | |
| 2149 | if (U_FAILURE(status)) { |
| 2150 | return result; |
| 2151 | } |
| 2152 | |
| 2153 | // If a BCP 47 language tag is passed as the language parameter to the |
| 2154 | // normal Locale constructor, it will actually fall back to invoking |
| 2155 | // uloc_forLanguageTag() to parse it if it somehow is able to detect that |
| 2156 | // the string actually is BCP 47. This works well for things like strings |
| 2157 | // using BCP 47 extensions, but it does not at all work for things like |
| 2158 | // legacy language tags (marked as “Type: grandfathered” in BCP 47, |
| 2159 | // e.g., "en-GB-oed") which are possible to also |
| 2160 | // interpret as ICU locale IDs and because of that won't trigger the BCP 47 |
| 2161 | // parsing. Therefore the code here explicitly calls uloc_forLanguageTag() |
| 2162 | // and then Locale::init(), instead of just calling the normal constructor. |
| 2163 | |
| 2164 | int32_t parsedLength; |
| 2165 | CharString localeID = ulocimp_forLanguageTagulocimp_forLanguageTag_77( |
| 2166 | tag.data(), |
| 2167 | tag.length(), |
| 2168 | &parsedLength, |
| 2169 | status); |
| 2170 | |
| 2171 | if (U_FAILURE(status)) { |
| 2172 | return result; |
| 2173 | } |
| 2174 | |
| 2175 | if (parsedLength != tag.size()) { |
| 2176 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2177 | return result; |
| 2178 | } |
| 2179 | |
| 2180 | result.init(localeID.data(), /*canonicalize=*/false); |
| 2181 | if (result.isBogus()) { |
| 2182 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2183 | } |
| 2184 | return result; |
| 2185 | } |
| 2186 | |
| 2187 | void |
| 2188 | Locale::toLanguageTag(ByteSink& sink, UErrorCode& status) const |
| 2189 | { |
| 2190 | if (U_FAILURE(status)) { |
| 2191 | return; |
| 2192 | } |
| 2193 | |
| 2194 | if (fIsBogus) { |
| 2195 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2196 | return; |
| 2197 | } |
| 2198 | |
| 2199 | ulocimp_toLanguageTagulocimp_toLanguageTag_77(fullName, sink, /*strict=*/false, status); |
| 2200 | } |
| 2201 | |
| 2202 | Locale U_EXPORT2 |
| 2203 | Locale::createFromName (const char *name) |
| 2204 | { |
| 2205 | if (name) { |
| 2206 | Locale l(""); |
| 2207 | l.init(name, false); |
| 2208 | return l; |
| 2209 | } |
| 2210 | else { |
| 2211 | return getDefault(); |
| 2212 | } |
| 2213 | } |
| 2214 | |
| 2215 | Locale U_EXPORT2 |
| 2216 | Locale::createFromName(StringPiece name) { |
| 2217 | Locale loc(""); |
| 2218 | loc.init(name, false); |
| 2219 | return loc; |
| 2220 | } |
| 2221 | |
| 2222 | Locale U_EXPORT2 |
| 2223 | Locale::createCanonical(const char* name) { |
| 2224 | Locale loc(""); |
| 2225 | loc.init(name, true); |
| 2226 | return loc; |
| 2227 | } |
| 2228 | |
| 2229 | const char * |
| 2230 | Locale::getISO3Language() const |
| 2231 | { |
| 2232 | return uloc_getISO3Languageuloc_getISO3Language_77(fullName); |
| 2233 | } |
| 2234 | |
| 2235 | |
| 2236 | const char * |
| 2237 | Locale::getISO3Country() const |
| 2238 | { |
| 2239 | return uloc_getISO3Countryuloc_getISO3Country_77(fullName); |
| 2240 | } |
| 2241 | |
| 2242 | /** |
| 2243 | * Return the LCID value as specified in the "LocaleID" resource for this |
| 2244 | * locale. The LocaleID must be expressed as a hexadecimal number, from |
| 2245 | * one to four digits. If the LocaleID resource is not present, or is |
| 2246 | * in an incorrect format, 0 is returned. The LocaleID is for use in |
| 2247 | * Windows (it is an LCID), but is available on all platforms. |
| 2248 | */ |
| 2249 | uint32_t |
| 2250 | Locale::getLCID() const |
| 2251 | { |
| 2252 | return uloc_getLCIDuloc_getLCID_77(fullName); |
| 2253 | } |
| 2254 | |
| 2255 | const char* const* U_EXPORT2 Locale::getISOCountries() |
| 2256 | { |
| 2257 | return uloc_getISOCountriesuloc_getISOCountries_77(); |
| 2258 | } |
| 2259 | |
| 2260 | const char* const* U_EXPORT2 Locale::getISOLanguages() |
| 2261 | { |
| 2262 | return uloc_getISOLanguagesuloc_getISOLanguages_77(); |
| 2263 | } |
| 2264 | |
| 2265 | // Set the locale's data based on a posix id. |
| 2266 | void Locale::setFromPOSIXID(const char *posixID) |
| 2267 | { |
| 2268 | init(posixID, true); |
| 2269 | } |
| 2270 | |
| 2271 | const Locale & U_EXPORT2 |
| 2272 | Locale::getRoot() |
| 2273 | { |
| 2274 | return getLocale(eROOT); |
| 2275 | } |
| 2276 | |
| 2277 | const Locale & U_EXPORT2 |
| 2278 | Locale::getEnglish() |
| 2279 | { |
| 2280 | return getLocale(eENGLISH); |
| 2281 | } |
| 2282 | |
| 2283 | const Locale & U_EXPORT2 |
| 2284 | Locale::getFrench() |
| 2285 | { |
| 2286 | return getLocale(eFRENCH); |
| 2287 | } |
| 2288 | |
| 2289 | const Locale & U_EXPORT2 |
| 2290 | Locale::getGerman() |
| 2291 | { |
| 2292 | return getLocale(eGERMAN); |
| 2293 | } |
| 2294 | |
| 2295 | const Locale & U_EXPORT2 |
| 2296 | Locale::getItalian() |
| 2297 | { |
| 2298 | return getLocale(eITALIAN); |
| 2299 | } |
| 2300 | |
| 2301 | const Locale & U_EXPORT2 |
| 2302 | Locale::getJapanese() |
| 2303 | { |
| 2304 | return getLocale(eJAPANESE); |
| 2305 | } |
| 2306 | |
| 2307 | const Locale & U_EXPORT2 |
| 2308 | Locale::getKorean() |
| 2309 | { |
| 2310 | return getLocale(eKOREAN); |
| 2311 | } |
| 2312 | |
| 2313 | const Locale & U_EXPORT2 |
| 2314 | Locale::getChinese() |
| 2315 | { |
| 2316 | return getLocale(eCHINESE); |
| 2317 | } |
| 2318 | |
| 2319 | const Locale & U_EXPORT2 |
| 2320 | Locale::getSimplifiedChinese() |
| 2321 | { |
| 2322 | return getLocale(eCHINA); |
| 2323 | } |
| 2324 | |
| 2325 | const Locale & U_EXPORT2 |
| 2326 | Locale::getTraditionalChinese() |
| 2327 | { |
| 2328 | return getLocale(eTAIWAN); |
| 2329 | } |
| 2330 | |
| 2331 | |
| 2332 | const Locale & U_EXPORT2 |
| 2333 | Locale::getFrance() |
| 2334 | { |
| 2335 | return getLocale(eFRANCE); |
| 2336 | } |
| 2337 | |
| 2338 | const Locale & U_EXPORT2 |
| 2339 | Locale::getGermany() |
| 2340 | { |
| 2341 | return getLocale(eGERMANY); |
| 2342 | } |
| 2343 | |
| 2344 | const Locale & U_EXPORT2 |
| 2345 | Locale::getItaly() |
| 2346 | { |
| 2347 | return getLocale(eITALY); |
| 2348 | } |
| 2349 | |
| 2350 | const Locale & U_EXPORT2 |
| 2351 | Locale::getJapan() |
| 2352 | { |
| 2353 | return getLocale(eJAPAN); |
| 2354 | } |
| 2355 | |
| 2356 | const Locale & U_EXPORT2 |
| 2357 | Locale::getKorea() |
| 2358 | { |
| 2359 | return getLocale(eKOREA); |
| 2360 | } |
| 2361 | |
| 2362 | const Locale & U_EXPORT2 |
| 2363 | Locale::getChina() |
| 2364 | { |
| 2365 | return getLocale(eCHINA); |
| 2366 | } |
| 2367 | |
| 2368 | const Locale & U_EXPORT2 |
| 2369 | Locale::getPRC() |
| 2370 | { |
| 2371 | return getLocale(eCHINA); |
| 2372 | } |
| 2373 | |
| 2374 | const Locale & U_EXPORT2 |
| 2375 | Locale::getTaiwan() |
| 2376 | { |
| 2377 | return getLocale(eTAIWAN); |
| 2378 | } |
| 2379 | |
| 2380 | const Locale & U_EXPORT2 |
| 2381 | Locale::getUK() |
| 2382 | { |
| 2383 | return getLocale(eUK); |
| 2384 | } |
| 2385 | |
| 2386 | const Locale & U_EXPORT2 |
| 2387 | Locale::getUS() |
| 2388 | { |
| 2389 | return getLocale(eUS); |
| 2390 | } |
| 2391 | |
| 2392 | const Locale & U_EXPORT2 |
| 2393 | Locale::getCanada() |
| 2394 | { |
| 2395 | return getLocale(eCANADA); |
| 2396 | } |
| 2397 | |
| 2398 | const Locale & U_EXPORT2 |
| 2399 | Locale::getCanadaFrench() |
| 2400 | { |
| 2401 | return getLocale(eCANADA_FRENCH); |
| 2402 | } |
| 2403 | |
| 2404 | const Locale & |
| 2405 | Locale::getLocale(int locid) |
| 2406 | { |
| 2407 | Locale *localeCache = getLocaleCache(); |
| 2408 | U_ASSERT((locid < eMAX_LOCALES)&&(locid>=0))(static_cast <bool> ((locid < eMAX_LOCALES)&& (locid>=0)) ? void (0) : __assert_fail ("(locid < eMAX_LOCALES)&&(locid>=0)" , __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__ )); |
| 2409 | if (localeCache == nullptr) { |
| 2410 | // Failure allocating the locale cache. |
| 2411 | // The best we can do is return a nullptr reference. |
| 2412 | locid = 0; |
| 2413 | } |
| 2414 | return localeCache[locid]; /*operating on nullptr*/ |
| 2415 | } |
| 2416 | |
| 2417 | /* |
| 2418 | This function is defined this way in order to get around static |
| 2419 | initialization and static destruction. |
| 2420 | */ |
| 2421 | Locale * |
| 2422 | Locale::getLocaleCache() |
| 2423 | { |
| 2424 | UErrorCode status = U_ZERO_ERROR; |
| 2425 | umtx_initOnce(gLocaleCacheInitOnce, locale_init, status); |
| 2426 | return gLocaleCache; |
| 2427 | } |
| 2428 | |
| 2429 | class KeywordEnumeration : public StringEnumeration { |
| 2430 | protected: |
| 2431 | CharString keywords; |
| 2432 | private: |
| 2433 | const char *current; |
| 2434 | static const char fgClassID; |
| 2435 | |
| 2436 | public: |
| 2437 | static UClassID U_EXPORT2 getStaticClassID() { return (UClassID)&fgClassID; } |
| 2438 | virtual UClassID getDynamicClassID() const override { return getStaticClassID(); } |
| 2439 | public: |
| 2440 | KeywordEnumeration(const char *keys, int32_t keywordLen, int32_t currentIndex, UErrorCode &status) |
| 2441 | : keywords(), current(keywords.data()) { |
| 2442 | if(U_SUCCESS(status) && keywordLen != 0) { |
| 2443 | if(keys == nullptr || keywordLen < 0) { |
| 2444 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2445 | } else { |
| 2446 | keywords.append(keys, keywordLen, status); |
| 2447 | current = keywords.data() + currentIndex; |
| 2448 | } |
| 2449 | } |
| 2450 | } |
| 2451 | |
| 2452 | virtual ~KeywordEnumeration(); |
| 2453 | |
| 2454 | virtual StringEnumeration * clone() const override |
| 2455 | { |
| 2456 | UErrorCode status = U_ZERO_ERROR; |
| 2457 | return new KeywordEnumeration( |
| 2458 | keywords.data(), keywords.length(), |
| 2459 | static_cast<int32_t>(current - keywords.data()), status); |
| 2460 | } |
| 2461 | |
| 2462 | virtual int32_t count(UErrorCode& status) const override { |
| 2463 | if (U_FAILURE(status)) { return 0; } |
| 2464 | const char *kw = keywords.data(); |
| 2465 | int32_t result = 0; |
| 2466 | while(*kw) { |
| 2467 | result++; |
| 2468 | kw += uprv_strlen(kw):: strlen(kw)+1; |
| 2469 | } |
| 2470 | return result; |
| 2471 | } |
| 2472 | |
| 2473 | virtual const char* next(int32_t* resultLength, UErrorCode& status) override { |
| 2474 | const char* result; |
| 2475 | int32_t len; |
| 2476 | if(U_SUCCESS(status) && *current != 0) { |
| 2477 | result = current; |
| 2478 | len = static_cast<int32_t>(uprv_strlen(current):: strlen(current)); |
| 2479 | current += len+1; |
| 2480 | if(resultLength != nullptr) { |
| 2481 | *resultLength = len; |
| 2482 | } |
| 2483 | } else { |
| 2484 | if(resultLength != nullptr) { |
| 2485 | *resultLength = 0; |
| 2486 | } |
| 2487 | result = nullptr; |
| 2488 | } |
| 2489 | return result; |
| 2490 | } |
| 2491 | |
| 2492 | virtual const UnicodeString* snext(UErrorCode& status) override { |
| 2493 | if (U_FAILURE(status)) { return nullptr; } |
| 2494 | int32_t resultLength = 0; |
| 2495 | const char *s = next(&resultLength, status); |
| 2496 | return setChars(s, resultLength, status); |
| 2497 | } |
| 2498 | |
| 2499 | virtual void reset(UErrorCode& status) override { |
| 2500 | if (U_FAILURE(status)) { return; } |
| 2501 | current = keywords.data(); |
| 2502 | } |
| 2503 | }; |
| 2504 | |
| 2505 | const char KeywordEnumeration::fgClassID = '\0'; |
| 2506 | |
| 2507 | // Out-of-line virtual destructor to serve as the "key function". |
| 2508 | KeywordEnumeration::~KeywordEnumeration() = default; |
| 2509 | |
| 2510 | // A wrapper around KeywordEnumeration that calls uloc_toUnicodeLocaleKey() in |
| 2511 | // the next() method for each keyword before returning it. |
| 2512 | class UnicodeKeywordEnumeration : public KeywordEnumeration { |
| 2513 | public: |
| 2514 | using KeywordEnumeration::KeywordEnumeration; |
| 2515 | virtual ~UnicodeKeywordEnumeration(); |
| 2516 | |
| 2517 | virtual const char* next(int32_t* resultLength, UErrorCode& status) override { |
| 2518 | const char* legacy_key = KeywordEnumeration::next(nullptr, status); |
| 2519 | while (U_SUCCESS(status) && legacy_key != nullptr) { |
| 2520 | const char* key = uloc_toUnicodeLocaleKeyuloc_toUnicodeLocaleKey_77(legacy_key); |
| 2521 | if (key != nullptr) { |
| 2522 | if (resultLength != nullptr) { |
| 2523 | *resultLength = static_cast<int32_t>(uprv_strlen(key):: strlen(key)); |
| 2524 | } |
| 2525 | return key; |
| 2526 | } |
| 2527 | // Not a Unicode keyword, could be a t, x or other, continue to look at the next one. |
| 2528 | legacy_key = KeywordEnumeration::next(nullptr, status); |
| 2529 | } |
| 2530 | if (resultLength != nullptr) *resultLength = 0; |
| 2531 | return nullptr; |
| 2532 | } |
| 2533 | virtual int32_t count(UErrorCode& status) const override { |
| 2534 | if (U_FAILURE(status)) { return 0; } |
| 2535 | const char *kw = keywords.data(); |
| 2536 | int32_t result = 0; |
| 2537 | while(*kw) { |
| 2538 | if (uloc_toUnicodeLocaleKeyuloc_toUnicodeLocaleKey_77(kw) != nullptr) { |
| 2539 | result++; |
| 2540 | } |
| 2541 | kw += uprv_strlen(kw):: strlen(kw)+1; |
| 2542 | } |
| 2543 | return result; |
| 2544 | } |
| 2545 | }; |
| 2546 | |
| 2547 | // Out-of-line virtual destructor to serve as the "key function". |
| 2548 | UnicodeKeywordEnumeration::~UnicodeKeywordEnumeration() = default; |
| 2549 | |
| 2550 | StringEnumeration * |
| 2551 | Locale::createKeywords(UErrorCode &status) const |
| 2552 | { |
| 2553 | StringEnumeration *result = nullptr; |
| 2554 | |
| 2555 | if (U_FAILURE(status)) { |
| 2556 | return result; |
| 2557 | } |
| 2558 | |
| 2559 | const char* variantStart = uprv_strchr(fullName, '@'):: strchr(fullName, '@'); |
| 2560 | const char* assignment = uprv_strchr(fullName, '='):: strchr(fullName, '='); |
| 2561 | if(variantStart) { |
| 2562 | if(assignment > variantStart) { |
| 2563 | CharString keywords = ulocimp_getKeywordsulocimp_getKeywords_77(variantStart + 1, '@', false, status); |
| 2564 | if (U_SUCCESS(status) && !keywords.isEmpty()) { |
| 2565 | result = new KeywordEnumeration(keywords.data(), keywords.length(), 0, status); |
| 2566 | if (!result) { |
| 2567 | status = U_MEMORY_ALLOCATION_ERROR; |
| 2568 | } |
| 2569 | } |
| 2570 | } else { |
| 2571 | status = U_INVALID_FORMAT_ERROR; |
| 2572 | } |
| 2573 | } |
| 2574 | return result; |
| 2575 | } |
| 2576 | |
| 2577 | StringEnumeration * |
| 2578 | Locale::createUnicodeKeywords(UErrorCode &status) const |
| 2579 | { |
| 2580 | StringEnumeration *result = nullptr; |
| 2581 | |
| 2582 | if (U_FAILURE(status)) { |
| 2583 | return result; |
| 2584 | } |
| 2585 | |
| 2586 | const char* variantStart = uprv_strchr(fullName, '@'):: strchr(fullName, '@'); |
| 2587 | const char* assignment = uprv_strchr(fullName, '='):: strchr(fullName, '='); |
| 2588 | if(variantStart) { |
| 2589 | if(assignment > variantStart) { |
| 2590 | CharString keywords = ulocimp_getKeywordsulocimp_getKeywords_77(variantStart + 1, '@', false, status); |
| 2591 | if (U_SUCCESS(status) && !keywords.isEmpty()) { |
| 2592 | result = new UnicodeKeywordEnumeration(keywords.data(), keywords.length(), 0, status); |
| 2593 | if (!result) { |
| 2594 | status = U_MEMORY_ALLOCATION_ERROR; |
| 2595 | } |
| 2596 | } |
| 2597 | } else { |
| 2598 | status = U_INVALID_FORMAT_ERROR; |
| 2599 | } |
| 2600 | } |
| 2601 | return result; |
| 2602 | } |
| 2603 | |
| 2604 | int32_t |
| 2605 | Locale::getKeywordValue(const char* keywordName, char *buffer, int32_t bufLen, UErrorCode &status) const |
| 2606 | { |
| 2607 | return uloc_getKeywordValueuloc_getKeywordValue_77(fullName, keywordName, buffer, bufLen, &status); |
| 2608 | } |
| 2609 | |
| 2610 | void |
| 2611 | Locale::getKeywordValue(StringPiece keywordName, ByteSink& sink, UErrorCode& status) const { |
| 2612 | if (U_FAILURE(status)) { |
| 2613 | return; |
| 2614 | } |
| 2615 | |
| 2616 | if (fIsBogus) { |
| 2617 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2618 | return; |
| 2619 | } |
| 2620 | |
| 2621 | ulocimp_getKeywordValueulocimp_getKeywordValue_77(fullName, keywordName, sink, status); |
| 2622 | } |
| 2623 | |
| 2624 | void |
| 2625 | Locale::getUnicodeKeywordValue(StringPiece keywordName, |
| 2626 | ByteSink& sink, |
| 2627 | UErrorCode& status) const { |
| 2628 | if (U_FAILURE(status)) { |
| 2629 | return; |
| 2630 | } |
| 2631 | |
| 2632 | std::optional<std::string_view> legacy_key = ulocimp_toLegacyKeyWithFallbackulocimp_toLegacyKeyWithFallback_77(keywordName); |
| 2633 | if (!legacy_key.has_value()) { |
| 2634 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2635 | return; |
| 2636 | } |
| 2637 | |
| 2638 | auto legacy_value = getKeywordValue<CharString>(*legacy_key, status); |
| 2639 | |
| 2640 | if (U_FAILURE(status)) { |
| 2641 | return; |
| 2642 | } |
| 2643 | |
| 2644 | std::optional<std::string_view> unicode_value = |
| 2645 | ulocimp_toBcpTypeWithFallbackulocimp_toBcpTypeWithFallback_77(keywordName, legacy_value.toStringPiece()); |
| 2646 | if (!unicode_value.has_value()) { |
| 2647 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2648 | return; |
| 2649 | } |
| 2650 | |
| 2651 | sink.Append(unicode_value->data(), static_cast<int32_t>(unicode_value->size())); |
| 2652 | } |
| 2653 | |
| 2654 | void |
| 2655 | Locale::setKeywordValue(StringPiece keywordName, |
| 2656 | StringPiece keywordValue, |
| 2657 | UErrorCode& status) { |
| 2658 | if (U_FAILURE(status)) { return; } |
| 2659 | if (keywordName.empty()) { |
| 2660 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2661 | return; |
| 2662 | } |
| 2663 | if (status == U_STRING_NOT_TERMINATED_WARNING) { |
| 2664 | status = U_ZERO_ERROR; |
| 2665 | } |
| 2666 | |
| 2667 | int32_t length = static_cast<int32_t>(uprv_strlen(fullName):: strlen(fullName)); |
| 2668 | int32_t capacity = fullName == fullNameBuffer ? ULOC_FULLNAME_CAPACITY157 : length + 1; |
| 2669 | |
| 2670 | const char* start = locale_getKeywordsStartlocale_getKeywordsStart_77(fullName); |
| 2671 | int32_t offset = start == nullptr ? length : start - fullName; |
| 2672 | |
| 2673 | for (;;) { |
| 2674 | // Remove -1 from the capacity so that this function can guarantee NUL termination. |
| 2675 | CheckedArrayByteSink sink(fullName + offset, capacity - offset - 1); |
| 2676 | |
| 2677 | int32_t reslen = ulocimp_setKeywordValueulocimp_setKeywordValue_77( |
| 2678 | {fullName + offset, static_cast<std::string_view::size_type>(length - offset)}, |
| 2679 | keywordName, |
| 2680 | keywordValue, |
| 2681 | sink, |
| 2682 | status); |
| 2683 | |
| 2684 | if (status == U_BUFFER_OVERFLOW_ERROR) { |
| 2685 | capacity = reslen + offset + 1; |
| 2686 | char* newFullName = static_cast<char*>(uprv_mallocuprv_malloc_77(capacity)); |
| 2687 | if (newFullName == nullptr) { |
| 2688 | status = U_MEMORY_ALLOCATION_ERROR; |
| 2689 | return; |
| 2690 | } |
| 2691 | uprv_memcpy(newFullName, fullName, length + 1)do { clang diagnostic push
clang diagnostic ignored "-Waddress" (static_cast <bool> (newFullName != __null) ? void (0 ) : __assert_fail ("newFullName != __null", __builtin_FILE () , __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); (static_cast <bool> (fullName != __null) ? void (0) : __assert_fail ("fullName != __null", __builtin_FILE (), __builtin_LINE (), __extension__ __PRETTY_FUNCTION__)); clang diagnostic pop :: memcpy(newFullName, fullName, length + 1); } while (false); |
| 2692 | if (fullName != fullNameBuffer) { |
| 2693 | if (baseName == fullName) { |
| 2694 | baseName = newFullName; // baseName should not point to freed memory. |
| 2695 | } |
| 2696 | // if fullName is already on the heap, need to free it. |
| 2697 | uprv_freeuprv_free_77(fullName); |
| 2698 | } |
| 2699 | fullName = newFullName; |
| 2700 | status = U_ZERO_ERROR; |
| 2701 | continue; |
| 2702 | } |
| 2703 | |
| 2704 | if (U_FAILURE(status)) { return; } |
| 2705 | u_terminateCharsu_terminateChars_77(fullName, capacity, reslen + offset, &status); |
| 2706 | break; |
| 2707 | } |
| 2708 | |
| 2709 | if (baseName == fullName) { |
| 2710 | // May have added the first keyword, meaning that the fullName is no longer also the baseName. |
| 2711 | initBaseName(status); |
| 2712 | } |
| 2713 | } |
| 2714 | |
| 2715 | void |
| 2716 | Locale::setUnicodeKeywordValue(StringPiece keywordName, |
| 2717 | StringPiece keywordValue, |
| 2718 | UErrorCode& status) { |
| 2719 | if (U_FAILURE(status)) { |
| 2720 | return; |
| 2721 | } |
| 2722 | |
| 2723 | std::optional<std::string_view> legacy_key = ulocimp_toLegacyKeyWithFallbackulocimp_toLegacyKeyWithFallback_77(keywordName); |
| 2724 | if (!legacy_key.has_value()) { |
| 2725 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2726 | return; |
| 2727 | } |
| 2728 | |
| 2729 | std::string_view value; |
| 2730 | |
| 2731 | if (!keywordValue.empty()) { |
| 2732 | std::optional<std::string_view> legacy_value = |
| 2733 | ulocimp_toLegacyTypeWithFallbackulocimp_toLegacyTypeWithFallback_77(keywordName, keywordValue); |
| 2734 | if (!legacy_value.has_value()) { |
| 2735 | status = U_ILLEGAL_ARGUMENT_ERROR; |
| 2736 | return; |
| 2737 | } |
| 2738 | value = *legacy_value; |
| 2739 | } |
| 2740 | |
| 2741 | setKeywordValue(*legacy_key, value, status); |
| 2742 | } |
| 2743 | |
| 2744 | const char * |
| 2745 | Locale::getBaseName() const { |
| 2746 | return baseName; |
| 2747 | } |
| 2748 | |
| 2749 | Locale::Iterator::~Iterator() = default; |
| 2750 | |
| 2751 | //eof |
| 2752 | U_NAMESPACE_END} |