| 1 | // © 2019 and later: Unicode, Inc. and others. |
| 2 | // License & terms of use: http://www.unicode.org/copyright.html |
| 3 | |
| 4 | // localematcher.h |
| 5 | // created: 2019may08 Markus W. Scherer |
| 6 | |
| 7 | #ifndef __LOCALEMATCHER_H__ |
| 8 | #define __LOCALEMATCHER_H__ |
| 9 | |
| 10 | #include "unicode/utypes.h" |
| 11 | |
| 12 | #if U_SHOW_CPLUSPLUS_API |
| 13 | |
| 14 | #include "unicode/locid.h" |
| 15 | #include "unicode/stringpiece.h" |
| 16 | #include "unicode/uobject.h" |
| 17 | |
| 18 | /** |
| 19 | * \file |
| 20 | * \brief C++ API: Locale matcher: User's desired locales vs. application's supported locales. |
| 21 | */ |
| 22 | |
| 23 | /** |
| 24 | * Builder option for whether the language subtag or the script subtag is most important. |
| 25 | * |
| 26 | * @see LocaleMatcher::Builder#setFavorSubtag(ULocMatchFavorSubtag) |
| 27 | * @stable ICU 65 |
| 28 | */ |
| 29 | enum ULocMatchFavorSubtag { |
| 30 | /** |
| 31 | * Language differences are most important, then script differences, then region differences. |
| 32 | * (This is the default behavior.) |
| 33 | * |
| 34 | * @stable ICU 65 |
| 35 | */ |
| 36 | ULOCMATCH_FAVOR_LANGUAGE, |
| 37 | /** |
| 38 | * Makes script differences matter relatively more than language differences. |
| 39 | * |
| 40 | * @stable ICU 65 |
| 41 | */ |
| 42 | ULOCMATCH_FAVOR_SCRIPT |
| 43 | }; |
| 44 | #ifndef U_IN_DOXYGEN |
| 45 | typedef enum ULocMatchFavorSubtag ULocMatchFavorSubtag; |
| 46 | #endif |
| 47 | |
| 48 | /** |
| 49 | * Builder option for whether all desired locales are treated equally or |
| 50 | * earlier ones are preferred. |
| 51 | * |
| 52 | * @see LocaleMatcher::Builder#setDemotionPerDesiredLocale(ULocMatchDemotion) |
| 53 | * @stable ICU 65 |
| 54 | */ |
| 55 | enum ULocMatchDemotion { |
| 56 | /** |
| 57 | * All desired locales are treated equally. |
| 58 | * |
| 59 | * @stable ICU 65 |
| 60 | */ |
| 61 | ULOCMATCH_DEMOTION_NONE, |
| 62 | /** |
| 63 | * Earlier desired locales are preferred. |
| 64 | * |
| 65 | * <p>From each desired locale to the next, |
| 66 | * the distance to any supported locale is increased by an additional amount |
| 67 | * which is at least as large as most region mismatches. |
| 68 | * A later desired locale has to have a better match with some supported locale |
| 69 | * due to more than merely having the same region subtag. |
| 70 | * |
| 71 | * <p>For example: <code>Supported={en, sv} desired=[en-GB, sv]</code> |
| 72 | * yields <code>Result(en-GB, en)</code> because |
| 73 | * with the demotion of sv its perfect match is no better than |
| 74 | * the region distance between the earlier desired locale en-GB and en=en-US. |
| 75 | * |
| 76 | * <p>Notes: |
| 77 | * <ul> |
| 78 | * <li>In some cases, language and/or script differences can be as small as |
| 79 | * the typical region difference. (Example: sr-Latn vs. sr-Cyrl) |
| 80 | * <li>It is possible for certain region differences to be larger than usual, |
| 81 | * and larger than the demotion. |
| 82 | * (As of CLDR 35 there is no such case, but |
| 83 | * this is possible in future versions of the data.) |
| 84 | * </ul> |
| 85 | * |
| 86 | * @stable ICU 65 |
| 87 | */ |
| 88 | ULOCMATCH_DEMOTION_REGION |
| 89 | }; |
| 90 | #ifndef U_IN_DOXYGEN |
| 91 | typedef enum ULocMatchDemotion ULocMatchDemotion; |
| 92 | #endif |
| 93 | |
| 94 | /** |
| 95 | * Builder option for whether to include or ignore one-way (fallback) match data. |
| 96 | * The LocaleMatcher uses CLDR languageMatch data which includes fallback (oneway=true) entries. |
| 97 | * Sometimes it is desirable to ignore those. |
| 98 | * |
| 99 | * <p>For example, consider a web application with the UI in a given language, |
| 100 | * with a link to another, related web app. |
| 101 | * The link should include the UI language, and the target server may also use |
| 102 | * the client’s Accept-Language header data. |
| 103 | * The target server has its own list of supported languages. |
| 104 | * One may want to favor UI language consistency, that is, |
| 105 | * if there is a decent match for the original UI language, we want to use it, |
| 106 | * but not if it is merely a fallback. |
| 107 | * |
| 108 | * @see LocaleMatcher::Builder#setDirection(ULocMatchDirection) |
| 109 | * @stable ICU 67 |
| 110 | */ |
| 111 | enum ULocMatchDirection { |
| 112 | /** |
| 113 | * Locale matching includes one-way matches such as Breton→French. (default) |
| 114 | * |
| 115 | * @stable ICU 67 |
| 116 | */ |
| 117 | ULOCMATCH_DIRECTION_WITH_ONE_WAY, |
| 118 | /** |
| 119 | * Locale matching limited to two-way matches including e.g. Danish↔Norwegian |
| 120 | * but ignoring one-way matches. |
| 121 | * |
| 122 | * @stable ICU 67 |
| 123 | */ |
| 124 | ULOCMATCH_DIRECTION_ONLY_TWO_WAY |
| 125 | }; |
| 126 | #ifndef U_IN_DOXYGEN |
| 127 | typedef enum ULocMatchDirection ULocMatchDirection; |
| 128 | #endif |
| 129 | |
| 130 | struct UHashtable; |
| 131 | |
| 132 | U_NAMESPACE_BEGIN |
| 133 | |
| 134 | struct LSR; |
| 135 | |
| 136 | class LocaleDistance; |
| 137 | class LocaleLsrIterator; |
| 138 | class UVector; |
| 139 | class XLikelySubtags; |
| 140 | |
| 141 | /** |
| 142 | * Immutable class that picks the best match between a user's desired locales and |
| 143 | * an application's supported locales. |
| 144 | * Movable but not copyable. |
| 145 | * |
| 146 | * <p>Example: |
| 147 | * <pre> |
| 148 | * UErrorCode errorCode = U_ZERO_ERROR; |
| 149 | * LocaleMatcher matcher = LocaleMatcher::Builder().setSupportedLocales("fr, en-GB, en").build(errorCode); |
| 150 | * Locale *bestSupported = matcher.getBestLocale(Locale.US, errorCode); // "en" |
| 151 | * </pre> |
| 152 | * |
| 153 | * <p>A matcher takes into account when languages are close to one another, |
| 154 | * such as Danish and Norwegian, |
| 155 | * and when regional variants are close, like en-GB and en-AU as opposed to en-US. |
| 156 | * |
| 157 | * <p>If there are multiple supported locales with the same (language, script, region) |
| 158 | * likely subtags, then the current implementation returns the first of those locales. |
| 159 | * It ignores variant subtags (except for pseudolocale variants) and extensions. |
| 160 | * This may change in future versions. |
| 161 | * |
| 162 | * <p>For example, the current implementation does not distinguish between |
| 163 | * de, de-DE, de-Latn, de-1901, de-u-co-phonebk. |
| 164 | * |
| 165 | * <p>If you prefer one equivalent locale over another, then provide only the preferred one, |
| 166 | * or place it earlier in the list of supported locales. |
| 167 | * |
| 168 | * <p>Otherwise, the order of supported locales may have no effect on the best-match results. |
| 169 | * The current implementation compares each desired locale with supported locales |
| 170 | * in the following order: |
| 171 | * 1. Default locale, if supported; |
| 172 | * 2. CLDR "paradigm locales" like en-GB and es-419; |
| 173 | * 3. other supported locales. |
| 174 | * This may change in future versions. |
| 175 | * |
| 176 | * <p>Often a product will just need one matcher instance, built with the languages |
| 177 | * that it supports. However, it may want multiple instances with different |
| 178 | * default languages based on additional information, such as the domain. |
| 179 | * |
| 180 | * <p>This class is not intended for public subclassing. |
| 181 | * |
| 182 | * @stable ICU 65 |
| 183 | */ |
| 184 | class U_COMMON_API LocaleMatcher : public UMemory { |
| 185 | public: |
| 186 | /** |
| 187 | * Data for the best-matching pair of a desired and a supported locale. |
| 188 | * Movable but not copyable. |
| 189 | * |
| 190 | * @stable ICU 65 |
| 191 | */ |
| 192 | class U_COMMON_API Result : public UMemory { |
| 193 | public: |
| 194 | /** |
| 195 | * Move constructor; might modify the source. |
| 196 | * This object will have the same contents that the source object had. |
| 197 | * |
| 198 | * @param src Result to move contents from. |
| 199 | * @stable ICU 65 |
| 200 | */ |
| 201 | Result(Result &&src) noexcept; |
| 202 | |
| 203 | /** |
| 204 | * Destructor. |
| 205 | * |
| 206 | * @stable ICU 65 |
| 207 | */ |
| 208 | ~Result(); |
| 209 | |
| 210 | /** |
| 211 | * Move assignment; might modify the source. |
| 212 | * This object will have the same contents that the source object had. |
| 213 | * |
| 214 | * @param src Result to move contents from. |
| 215 | * @stable ICU 65 |
| 216 | */ |
| 217 | Result &operator=(Result &&src) noexcept; |
| 218 | |
| 219 | /** |
| 220 | * Returns the best-matching desired locale. |
| 221 | * nullptr if the list of desired locales is empty or if none matched well enough. |
| 222 | * |
| 223 | * @return the best-matching desired locale, or nullptr. |
| 224 | * @stable ICU 65 |
| 225 | */ |
| 226 | inline const Locale *getDesiredLocale() const { return desiredLocale; } |
| 227 | |
| 228 | /** |
| 229 | * Returns the best-matching supported locale. |
| 230 | * If none matched well enough, this is the default locale. |
| 231 | * The default locale is nullptr if Builder::setNoDefaultLocale() was called, |
| 232 | * or if the list of supported locales is empty and no explicit default locale is set. |
| 233 | * |
| 234 | * @return the best-matching supported locale, or nullptr. |
| 235 | * @stable ICU 65 |
| 236 | */ |
| 237 | inline const Locale *getSupportedLocale() const { return supportedLocale; } |
| 238 | |
| 239 | /** |
| 240 | * Returns the index of the best-matching desired locale in the input Iterable order. |
| 241 | * -1 if the list of desired locales is empty or if none matched well enough. |
| 242 | * |
| 243 | * @return the index of the best-matching desired locale, or -1. |
| 244 | * @stable ICU 65 |
| 245 | */ |
| 246 | inline int32_t getDesiredIndex() const { return desiredIndex; } |
| 247 | |
| 248 | /** |
| 249 | * Returns the index of the best-matching supported locale in the |
| 250 | * constructor’s or builder’s input order (“set” Collection plus “added” locales). |
| 251 | * If the matcher was built from a locale list string, then the iteration order is that |
| 252 | * of a LocalePriorityList built from the same string. |
| 253 | * -1 if the list of supported locales is empty or if none matched well enough. |
| 254 | * |
| 255 | * @return the index of the best-matching supported locale, or -1. |
| 256 | * @stable ICU 65 |
| 257 | */ |
| 258 | inline int32_t getSupportedIndex() const { return supportedIndex; } |
| 259 | |
| 260 | /** |
| 261 | * Takes the best-matching supported locale and adds relevant fields of the |
| 262 | * best-matching desired locale, such as the -t- and -u- extensions. |
| 263 | * May replace some fields of the supported locale. |
| 264 | * The result is the locale that should be used for date and number formatting, collation, etc. |
| 265 | * Returns the root locale if getSupportedLocale() returns nullptr. |
| 266 | * |
| 267 | * <p>Example: desired=ar-SA-u-nu-latn, supported=ar-EG, resolved locale=ar-SA-u-nu-latn |
| 268 | * |
| 269 | * @return a locale combining the best-matching desired and supported locales. |
| 270 | * @stable ICU 65 |
| 271 | */ |
| 272 | Locale makeResolvedLocale(UErrorCode &errorCode) const; |
| 273 | |
| 274 | private: |
| 275 | Result(const Locale *desired, const Locale *supported, |
| 276 | int32_t desIndex, int32_t suppIndex, UBool owned) : |
| 277 | desiredLocale(desired), supportedLocale(supported), |
| 278 | desiredIndex(desIndex), supportedIndex(suppIndex), |
| 279 | desiredIsOwned(owned) {} |
| 280 | |
| 281 | Result(const Result &other) = delete; |
| 282 | Result &operator=(const Result &other) = delete; |
| 283 | |
| 284 | const Locale *desiredLocale; |
| 285 | const Locale *supportedLocale; |
| 286 | int32_t desiredIndex; |
| 287 | int32_t supportedIndex; |
| 288 | UBool desiredIsOwned; |
| 289 | |
| 290 | friend class LocaleMatcher; |
| 291 | }; |
| 292 | |
| 293 | /** |
| 294 | * LocaleMatcher builder. |
| 295 | * Movable but not copyable. |
| 296 | * |
| 297 | * @stable ICU 65 |
| 298 | */ |
| 299 | class U_COMMON_API Builder : public UMemory { |
| 300 | public: |
| 301 | /** |
| 302 | * Constructs a builder used in chaining parameters for building a LocaleMatcher. |
| 303 | * |
| 304 | * @return a new Builder object |
| 305 | * @stable ICU 65 |
| 306 | */ |
| 307 | Builder() {} |
| 308 | |
| 309 | /** |
| 310 | * Move constructor; might modify the source. |
| 311 | * This builder will have the same contents that the source builder had. |
| 312 | * |
| 313 | * @param src Builder to move contents from. |
| 314 | * @stable ICU 65 |
| 315 | */ |
| 316 | Builder(Builder &&src) noexcept; |
| 317 | |
| 318 | /** |
| 319 | * Destructor. |
| 320 | * |
| 321 | * @stable ICU 65 |
| 322 | */ |
| 323 | ~Builder(); |
| 324 | |
| 325 | /** |
| 326 | * Move assignment; might modify the source. |
| 327 | * This builder will have the same contents that the source builder had. |
| 328 | * |
| 329 | * @param src Builder to move contents from. |
| 330 | * @stable ICU 65 |
| 331 | */ |
| 332 | Builder &operator=(Builder &&src) noexcept; |
| 333 | |
| 334 | /** |
| 335 | * Parses an Accept-Language string |
| 336 | * (<a href="https://tools.ietf.org/html/rfc2616#section-14.4">RFC 2616 Section 14.4</a>), |
| 337 | * such as "af, en, fr;q=0.9", and sets the supported locales accordingly. |
| 338 | * Allows whitespace in more places but does not allow "*". |
| 339 | * Clears any previously set/added supported locales first. |
| 340 | * |
| 341 | * @param locales the Accept-Language string of locales to set |
| 342 | * @return this Builder object |
| 343 | * @stable ICU 65 |
| 344 | */ |
| 345 | Builder &setSupportedLocalesFromListString(StringPiece locales); |
| 346 | |
| 347 | /** |
| 348 | * Copies the supported locales, preserving iteration order. |
| 349 | * Clears any previously set/added supported locales first. |
| 350 | * Duplicates are allowed, and are not removed. |
| 351 | * |
| 352 | * @param locales the list of locale |
| 353 | * @return this Builder object |
| 354 | * @stable ICU 65 |
| 355 | */ |
| 356 | Builder &setSupportedLocales(Locale::Iterator &locales); |
| 357 | |
| 358 | /** |
| 359 | * Copies the supported locales from the begin/end range, preserving iteration order. |
| 360 | * Clears any previously set/added supported locales first. |
| 361 | * Duplicates are allowed, and are not removed. |
| 362 | * |
| 363 | * Each of the iterator parameter values must be an |
| 364 | * input iterator whose value is convertible to const Locale &. |
| 365 | * |
| 366 | * @param begin Start of range. |
| 367 | * @param end Exclusive end of range. |
| 368 | * @return this Builder object |
| 369 | * @stable ICU 65 |
| 370 | */ |
| 371 | template<typename Iter> |
| 372 | Builder &setSupportedLocales(Iter begin, Iter end) { |
| 373 | if (U_FAILURE(errorCode_)) { return *this; } |
| 374 | clearSupportedLocales(); |
| 375 | while (begin != end) { |
| 376 | addSupportedLocale(*begin++); |
| 377 | } |
| 378 | return *this; |
| 379 | } |
| 380 | |
| 381 | /** |
| 382 | * Copies the supported locales from the begin/end range, preserving iteration order. |
| 383 | * Calls the converter to convert each *begin to a Locale or const Locale &. |
| 384 | * Clears any previously set/added supported locales first. |
| 385 | * Duplicates are allowed, and are not removed. |
| 386 | * |
| 387 | * Each of the iterator parameter values must be an |
| 388 | * input iterator whose value is convertible to const Locale &. |
| 389 | * |
| 390 | * @param begin Start of range. |
| 391 | * @param end Exclusive end of range. |
| 392 | * @param converter Converter from *begin to const Locale & or compatible. |
| 393 | * @return this Builder object |
| 394 | * @stable ICU 65 |
| 395 | */ |
| 396 | template<typename Iter, typename Conv> |
| 397 | Builder &setSupportedLocalesViaConverter(Iter begin, Iter end, Conv converter) { |
| 398 | if (U_FAILURE(errorCode_)) { return *this; } |
| 399 | clearSupportedLocales(); |
| 400 | while (begin != end) { |
| 401 | addSupportedLocale(converter(*begin++)); |
| 402 | } |
| 403 | return *this; |
| 404 | } |
| 405 | |
| 406 | /** |
| 407 | * Adds another supported locale. |
| 408 | * Duplicates are allowed, and are not removed. |
| 409 | * |
| 410 | * @param locale another locale |
| 411 | * @return this Builder object |
| 412 | * @stable ICU 65 |
| 413 | */ |
| 414 | Builder &addSupportedLocale(const Locale &locale); |
| 415 | |
| 416 | /** |
| 417 | * Sets no default locale. |
| 418 | * There will be no explicit or implicit default locale. |
| 419 | * If there is no good match, then the matcher will return nullptr for the |
| 420 | * best supported locale. |
| 421 | * |
| 422 | * @stable ICU 68 |
| 423 | */ |
| 424 | Builder &setNoDefaultLocale(); |
| 425 | |
| 426 | /** |
| 427 | * Sets the default locale; if nullptr, or if it is not set explicitly, |
| 428 | * then the first supported locale is used as the default locale. |
| 429 | * There is no default locale at all (nullptr will be returned instead) |
| 430 | * if setNoDefaultLocale() is called. |
| 431 | * |
| 432 | * @param defaultLocale the default locale (will be copied) |
| 433 | * @return this Builder object |
| 434 | * @stable ICU 65 |
| 435 | */ |
| 436 | Builder &setDefaultLocale(const Locale *defaultLocale); |
| 437 | |
| 438 | /** |
| 439 | * If ULOCMATCH_FAVOR_SCRIPT, then the language differences are smaller than script |
| 440 | * differences. |
| 441 | * This is used in situations (such as maps) where |
| 442 | * it is better to fall back to the same script than a similar language. |
| 443 | * |
| 444 | * @param subtag the subtag to favor |
| 445 | * @return this Builder object |
| 446 | * @stable ICU 65 |
| 447 | */ |
| 448 | Builder &setFavorSubtag(ULocMatchFavorSubtag subtag); |
| 449 | |
| 450 | /** |
| 451 | * Option for whether all desired locales are treated equally or |
| 452 | * earlier ones are preferred (this is the default). |
| 453 | * |
| 454 | * @param demotion the demotion per desired locale to set. |
| 455 | * @return this Builder object |
| 456 | * @stable ICU 65 |
| 457 | */ |
| 458 | Builder &setDemotionPerDesiredLocale(ULocMatchDemotion demotion); |
| 459 | |
| 460 | /** |
| 461 | * Option for whether to include or ignore one-way (fallback) match data. |
| 462 | * By default, they are included. |
| 463 | * |
| 464 | * @param matchDirection the match direction to set. |
| 465 | * @return this Builder object |
| 466 | * @stable ICU 67 |
| 467 | */ |
| 468 | Builder &setDirection(ULocMatchDirection matchDirection) { |
| 469 | if (U_SUCCESS(errorCode_)) { |
| 470 | direction_ = matchDirection; |
| 471 | } |
| 472 | return *this; |
| 473 | } |
| 474 | |
| 475 | /** |
| 476 | * Sets the maximum distance for an acceptable match. |
| 477 | * The matcher will return a match for a pair of locales only if |
| 478 | * they match at least as well as the pair given here. |
| 479 | * |
| 480 | * For example, setMaxDistance(en-US, en-GB) limits matches to ones where the |
| 481 | * (desired, support) locales have a distance no greater than a region subtag difference. |
| 482 | * This is much stricter than the CLDR default. |
| 483 | * |
| 484 | * The details of locale matching are subject to changes in |
| 485 | * CLDR data and in the algorithm. |
| 486 | * Specifying a maximum distance in relative terms via a sample pair of locales |
| 487 | * insulates from changes that affect all distance metrics similarly, |
| 488 | * but some changes will necessarily affect relative distances between |
| 489 | * different pairs of locales. |
| 490 | * |
| 491 | * @param desired the desired locale for distance comparison. |
| 492 | * @param supported the supported locale for distance comparison. |
| 493 | * @return this Builder object |
| 494 | * @stable ICU 68 |
| 495 | */ |
| 496 | Builder &setMaxDistance(const Locale &desired, const Locale &supported); |
| 497 | |
| 498 | /** |
| 499 | * Sets the UErrorCode if an error occurred while setting parameters. |
| 500 | * Preserves older error codes in the outErrorCode. |
| 501 | * |
| 502 | * @param outErrorCode Set to an error code if it does not contain one already |
| 503 | * and an error occurred while setting parameters. |
| 504 | * Otherwise unchanged. |
| 505 | * @return true if U_FAILURE(outErrorCode) |
| 506 | * @stable ICU 65 |
| 507 | */ |
| 508 | UBool copyErrorTo(UErrorCode &outErrorCode) const; |
| 509 | |
| 510 | /** |
| 511 | * Builds and returns a new locale matcher. |
| 512 | * This builder can continue to be used. |
| 513 | * |
| 514 | * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test, |
| 515 | * or else the function returns immediately. Check for U_FAILURE() |
| 516 | * on output or use with function chaining. (See User Guide for details.) |
| 517 | * @return LocaleMatcher |
| 518 | * @stable ICU 65 |
| 519 | */ |
| 520 | LocaleMatcher build(UErrorCode &errorCode) const; |
| 521 | |
| 522 | private: |
| 523 | friend class LocaleMatcher; |
| 524 | |
| 525 | Builder(const Builder &other) = delete; |
| 526 | Builder &operator=(const Builder &other) = delete; |
| 527 | |
| 528 | void clearSupportedLocales(); |
| 529 | bool ensureSupportedLocaleVector(); |
| 530 | |
| 531 | UErrorCode errorCode_ = U_ZERO_ERROR; |
| 532 | UVector *supportedLocales_ = nullptr; |
| 533 | int32_t thresholdDistance_ = -1; |
| 534 | ULocMatchDemotion demotion_ = ULOCMATCH_DEMOTION_REGION; |
| 535 | Locale *defaultLocale_ = nullptr; |
| 536 | bool withDefault_ = true; |
| 537 | ULocMatchFavorSubtag favor_ = ULOCMATCH_FAVOR_LANGUAGE; |
| 538 | ULocMatchDirection direction_ = ULOCMATCH_DIRECTION_WITH_ONE_WAY; |
| 539 | Locale *maxDistanceDesired_ = nullptr; |
| 540 | Locale *maxDistanceSupported_ = nullptr; |
| 541 | }; |
| 542 | |
| 543 | // FYI No public LocaleMatcher constructors in C++; use the Builder. |
| 544 | |
| 545 | /** |
| 546 | * Move copy constructor; might modify the source. |
| 547 | * This matcher will have the same settings that the source matcher had. |
| 548 | * @param src source matcher |
| 549 | * @stable ICU 65 |
| 550 | */ |
| 551 | LocaleMatcher(LocaleMatcher &&src) noexcept; |
| 552 | |
| 553 | /** |
| 554 | * Destructor. |
| 555 | * @stable ICU 65 |
| 556 | */ |
| 557 | ~LocaleMatcher(); |
| 558 | |
| 559 | /** |
| 560 | * Move assignment operator; might modify the source. |
| 561 | * This matcher will have the same settings that the source matcher had. |
| 562 | * The behavior is undefined if *this and src are the same object. |
| 563 | * @param src source matcher |
| 564 | * @return *this |
| 565 | * @stable ICU 65 |
| 566 | */ |
| 567 | LocaleMatcher &operator=(LocaleMatcher &&src) noexcept; |
| 568 | |
| 569 | /** |
| 570 | * Returns the supported locale which best matches the desired locale. |
| 571 | * |
| 572 | * @param desiredLocale Typically a user's language. |
| 573 | * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test, |
| 574 | * or else the function returns immediately. Check for U_FAILURE() |
| 575 | * on output or use with function chaining. (See User Guide for details.) |
| 576 | * @return the best-matching supported locale. |
| 577 | * @stable ICU 65 |
| 578 | */ |
| 579 | const Locale *getBestMatch(const Locale &desiredLocale, UErrorCode &errorCode) const; |
| 580 | |
| 581 | /** |
| 582 | * Returns the supported locale which best matches one of the desired locales. |
| 583 | * |
| 584 | * @param desiredLocales Typically a user's languages, in order of preference (descending). |
| 585 | * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test, |
| 586 | * or else the function returns immediately. Check for U_FAILURE() |
| 587 | * on output or use with function chaining. (See User Guide for details.) |
| 588 | * @return the best-matching supported locale. |
| 589 | * @stable ICU 65 |
| 590 | */ |
| 591 | const Locale *getBestMatch(Locale::Iterator &desiredLocales, UErrorCode &errorCode) const; |
| 592 | |
| 593 | /** |
| 594 | * Parses an Accept-Language string |
| 595 | * (<a href="https://tools.ietf.org/html/rfc2616#section-14.4">RFC 2616 Section 14.4</a>), |
| 596 | * such as "af, en, fr;q=0.9", |
| 597 | * and returns the supported locale which best matches one of the desired locales. |
| 598 | * Allows whitespace in more places but does not allow "*". |
| 599 | * |
| 600 | * @param desiredLocaleList Typically a user's languages, as an Accept-Language string. |
| 601 | * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test, |
| 602 | * or else the function returns immediately. Check for U_FAILURE() |
| 603 | * on output or use with function chaining. (See User Guide for details.) |
| 604 | * @return the best-matching supported locale. |
| 605 | * @stable ICU 65 |
| 606 | */ |
| 607 | const Locale *getBestMatchForListString(StringPiece desiredLocaleList, UErrorCode &errorCode) const; |
| 608 | |
| 609 | /** |
| 610 | * Returns the best match between the desired locale and the supported locales. |
| 611 | * If the result's desired locale is not nullptr, then it is the address of the input locale. |
| 612 | * It has not been cloned. |
| 613 | * |
| 614 | * @param desiredLocale Typically a user's language. |
| 615 | * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test, |
| 616 | * or else the function returns immediately. Check for U_FAILURE() |
| 617 | * on output or use with function chaining. (See User Guide for details.) |
| 618 | * @return the best-matching pair of the desired and a supported locale. |
| 619 | * @stable ICU 65 |
| 620 | */ |
| 621 | Result getBestMatchResult(const Locale &desiredLocale, UErrorCode &errorCode) const; |
| 622 | |
| 623 | /** |
| 624 | * Returns the best match between the desired and supported locales. |
| 625 | * If the result's desired locale is not nullptr, then it is a clone of |
| 626 | * the best-matching desired locale. The Result object owns the clone. |
| 627 | * |
| 628 | * @param desiredLocales Typically a user's languages, in order of preference (descending). |
| 629 | * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test, |
| 630 | * or else the function returns immediately. Check for U_FAILURE() |
| 631 | * on output or use with function chaining. (See User Guide for details.) |
| 632 | * @return the best-matching pair of a desired and a supported locale. |
| 633 | * @stable ICU 65 |
| 634 | */ |
| 635 | Result getBestMatchResult(Locale::Iterator &desiredLocales, UErrorCode &errorCode) const; |
| 636 | |
| 637 | /** |
| 638 | * Returns true if the pair of locales matches acceptably. |
| 639 | * This is influenced by Builder options such as setDirection(), setFavorSubtag(), |
| 640 | * and setMaxDistance(). |
| 641 | * |
| 642 | * @param desired The desired locale. |
| 643 | * @param supported The supported locale. |
| 644 | * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test, |
| 645 | * or else the function returns immediately. Check for U_FAILURE() |
| 646 | * on output or use with function chaining. (See User Guide for details.) |
| 647 | * @return true if the pair of locales matches acceptably. |
| 648 | * @stable ICU 68 |
| 649 | */ |
| 650 | UBool isMatch(const Locale &desired, const Locale &supported, UErrorCode &errorCode) const; |
| 651 | |
| 652 | #ifndef U_HIDE_INTERNAL_API |
| 653 | /** |
| 654 | * Returns a fraction between 0 and 1, where 1 means that the languages are a |
| 655 | * perfect match, and 0 means that they are completely different. |
| 656 | * |
| 657 | * <p>This is mostly an implementation detail, and the precise values may change over time. |
| 658 | * The implementation may use either the maximized forms or the others ones, or both. |
| 659 | * The implementation may or may not rely on the forms to be consistent with each other. |
| 660 | * |
| 661 | * <p>Callers should construct and use a matcher rather than match pairs of locales directly. |
| 662 | * |
| 663 | * @param desired Desired locale. |
| 664 | * @param supported Supported locale. |
| 665 | * @param errorCode ICU error code. Its input value must pass the U_SUCCESS() test, |
| 666 | * or else the function returns immediately. Check for U_FAILURE() |
| 667 | * on output or use with function chaining. (See User Guide for details.) |
| 668 | * @return value between 0 and 1, inclusive. |
| 669 | * @internal (has a known user) |
| 670 | */ |
| 671 | double internalMatch(const Locale &desired, const Locale &supported, UErrorCode &errorCode) const; |
| 672 | #endif // U_HIDE_INTERNAL_API |
| 673 | |
| 674 | private: |
| 675 | LocaleMatcher(const Builder &builder, UErrorCode &errorCode); |
| 676 | LocaleMatcher(const LocaleMatcher &other) = delete; |
| 677 | LocaleMatcher &operator=(const LocaleMatcher &other) = delete; |
| 678 | |
| 679 | int32_t putIfAbsent(const LSR &lsr, int32_t i, int32_t suppLength, UErrorCode &errorCode); |
| 680 | |
| 681 | int32_t getBestSuppIndex(LSR desiredLSR, LocaleLsrIterator *remainingIter, UErrorCode &errorCode) const; |
| 682 | |
| 683 | const XLikelySubtags &likelySubtags; |
| 684 | const LocaleDistance &localeDistance; |
| 685 | int32_t thresholdDistance; |
| 686 | int32_t demotionPerDesiredLocale; |
| 687 | ULocMatchFavorSubtag favorSubtag; |
| 688 | ULocMatchDirection direction; |
| 689 | |
| 690 | // These are in input order. |
| 691 | const Locale ** supportedLocales; |
| 692 | LSR *lsrs; |
| 693 | int32_t supportedLocalesLength; |
| 694 | // These are in preference order: 1. Default locale 2. paradigm locales 3. others. |
| 695 | UHashtable *supportedLsrToIndex; // Map<LSR, Integer> |
| 696 | // Array versions of the supportedLsrToIndex keys and values. |
| 697 | // The distance lookup loops over the supportedLSRs and returns the index of the best match. |
| 698 | const LSR **supportedLSRs; |
| 699 | int32_t *supportedIndexes; |
| 700 | int32_t supportedLSRsLength; |
| 701 | Locale *ownedDefaultLocale; |
| 702 | const Locale *defaultLocale; |
| 703 | }; |
| 704 | |
| 705 | U_NAMESPACE_END |
| 706 | |
| 707 | #endif // U_SHOW_CPLUSPLUS_API |
| 708 | #endif // __LOCALEMATCHER_H__ |
| 709 | |