AbstractOperations.cpp 31 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738
  1. /*
  2. * Copyright (c) 2021, Tim Flynn <trflynn89@pm.me>
  3. *
  4. * SPDX-License-Identifier: BSD-2-Clause
  5. */
  6. #include <AK/AllOf.h>
  7. #include <AK/AnyOf.h>
  8. #include <AK/Array.h>
  9. #include <AK/CharacterTypes.h>
  10. #include <AK/Find.h>
  11. #include <AK/Function.h>
  12. #include <AK/QuickSort.h>
  13. #include <AK/TypeCasts.h>
  14. #include <LibJS/Runtime/Array.h>
  15. #include <LibJS/Runtime/GlobalObject.h>
  16. #include <LibJS/Runtime/Intl/AbstractOperations.h>
  17. #include <LibJS/Runtime/Intl/Locale.h>
  18. #include <LibUnicode/Locale.h>
  19. namespace JS::Intl {
  20. // 6.2.2 IsStructurallyValidLanguageTag ( locale ), https://tc39.es/ecma402/#sec-isstructurallyvalidlanguagetag
  21. Optional<Unicode::LocaleID> is_structurally_valid_language_tag(StringView locale)
  22. {
  23. auto contains_duplicate_variant = [](auto& variants) {
  24. if (variants.is_empty())
  25. return false;
  26. quick_sort(variants);
  27. for (size_t i = 0; i < variants.size() - 1; ++i) {
  28. if (variants[i].equals_ignoring_case(variants[i + 1]))
  29. return true;
  30. }
  31. return false;
  32. };
  33. // IsStructurallyValidLanguageTag returns true if all of the following conditions hold, false otherwise:
  34. // locale can be generated from the EBNF grammar for unicode_locale_id in Unicode Technical Standard #35 LDML § 3.2 Unicode Locale Identifier;
  35. auto locale_id = Unicode::parse_unicode_locale_id(locale);
  36. if (!locale_id.has_value())
  37. return {};
  38. // locale does not use any of the backwards compatibility syntax described in Unicode Technical Standard #35 LDML § 3.3 BCP 47 Conformance;
  39. // https://unicode.org/reports/tr35/#BCP_47_Conformance
  40. if (locale.contains('_') || locale_id->language_id.is_root || !locale_id->language_id.language.has_value())
  41. return {};
  42. // the unicode_language_id within locale contains no duplicate unicode_variant_subtag subtags; and
  43. if (contains_duplicate_variant(locale_id->language_id.variants))
  44. return {};
  45. // if locale contains an extensions* component, that component
  46. Vector<char> unique_keys;
  47. for (auto& extension : locale_id->extensions) {
  48. // does not contain any other_extensions components with duplicate [alphanum-[tTuUxX]] subtags,
  49. // contains at most one unicode_locale_extensions component,
  50. // contains at most one transformed_extensions component, and
  51. char key = extension.visit(
  52. [](Unicode::LocaleExtension const&) { return 'u'; },
  53. [](Unicode::TransformedExtension const&) { return 't'; },
  54. [](Unicode::OtherExtension const& ext) { return static_cast<char>(to_ascii_lowercase(ext.key)); });
  55. if (unique_keys.contains_slow(key))
  56. return {};
  57. unique_keys.append(key);
  58. // if a transformed_extensions component that contains a tlang component is present, then
  59. // the tlang component contains no duplicate unicode_variant_subtag subtags.
  60. if (auto* transformed = extension.get_pointer<Unicode::TransformedExtension>()) {
  61. auto& language = transformed->language;
  62. if (language.has_value() && contains_duplicate_variant(language->variants))
  63. return {};
  64. }
  65. }
  66. return locale_id;
  67. }
  68. // 6.2.3 CanonicalizeUnicodeLocaleId ( locale ), https://tc39.es/ecma402/#sec-canonicalizeunicodelocaleid
  69. String canonicalize_unicode_locale_id(Unicode::LocaleID& locale)
  70. {
  71. // Note: This implementation differs from the spec in how Step 3 is implemented. The spec assumes
  72. // the input to this method is a string, and is written such that operations are performed on parts
  73. // of that string. LibUnicode gives us the parsed locale in a structure, so we can mutate that
  74. // structure directly. From a footnote in the spec:
  75. //
  76. // The third step of this algorithm ensures that a Unicode locale extension sequence in the
  77. // returned language tag contains:
  78. // * only the first instance of any attribute duplicated in the input, and
  79. // * only the first keyword for a given key in the input.
  80. for (auto& extension : locale.extensions) {
  81. if (!extension.has<Unicode::LocaleExtension>())
  82. continue;
  83. auto& locale_extension = extension.get<Unicode::LocaleExtension>();
  84. auto attributes = move(locale_extension.attributes);
  85. for (auto& attribute : attributes) {
  86. if (!locale_extension.attributes.contains_slow(attribute))
  87. locale_extension.attributes.append(move(attribute));
  88. }
  89. auto keywords = move(locale_extension.keywords);
  90. for (auto& keyword : keywords) {
  91. if (!any_of(locale_extension.keywords, [&](auto const& k) { return k.key == keyword.key; }))
  92. locale_extension.keywords.append(move(keyword));
  93. }
  94. break;
  95. }
  96. // 1. Let localeId be the string locale after performing the algorithm to transform it to canonical syntax per Unicode Technical Standard #35 LDML § 3.2.1 Canonical Unicode Locale Identifiers.
  97. // 2. Let localeId be the string localeId after performing the algorithm to transform it to canonical form.
  98. auto locale_id = Unicode::canonicalize_unicode_locale_id(locale);
  99. VERIFY(locale_id.has_value());
  100. // 4. Return localeId.
  101. return locale_id.release_value();
  102. }
  103. // 6.3.1 IsWellFormedCurrencyCode ( currency ), https://tc39.es/ecma402/#sec-canonicalcodefordisplaynames
  104. bool is_well_formed_currency_code(StringView currency)
  105. {
  106. // 1. Let normalized be the result of mapping currency to upper case as described in 6.1.
  107. // 2. If the number of elements in normalized is not 3, return false.
  108. if (currency.length() != 3)
  109. return false;
  110. // 3. If normalized contains any character that is not in the range "A" to "Z" (U+0041 to U+005A), return false.
  111. if (!all_of(currency, is_ascii_alpha))
  112. return false;
  113. // 4. Return true.
  114. return true;
  115. }
  116. // 6.5.1 IsWellFormedUnitIdentifier ( unitIdentifier ), https://tc39.es/ecma402/#sec-iswellformedunitidentifier
  117. bool is_well_formed_unit_identifier(StringView unit_identifier)
  118. {
  119. // 6.5.2 IsSanctionedSimpleUnitIdentifier ( unitIdentifier ), https://tc39.es/ecma402/#sec-iswellformedunitidentifier
  120. constexpr auto is_sanctioned_simple_unit_identifier = [](StringView unit_identifier) {
  121. // 1. If unitIdentifier is listed in Table 2 below, return true.
  122. // 2. Else, return false.
  123. static constexpr auto sanctioned_units = AK::Array { "acre"sv, "bit"sv, "byte"sv, "celsius"sv, "centimeter"sv, "day"sv, "degree"sv, "fahrenheit"sv, "fluid-ounce"sv, "foot"sv, "gallon"sv, "gigabit"sv, "gigabyte"sv, "gram"sv, "hectare"sv, "hour"sv, "inch"sv, "kilobit"sv, "kilobyte"sv, "kilogram"sv, "kilometer"sv, "liter"sv, "megabit"sv, "megabyte"sv, "meter"sv, "mile"sv, "mile-scandinavian"sv, "milliliter"sv, "millimeter"sv, "millisecond"sv, "minute"sv, "month"sv, "ounce"sv, "percent"sv, "petabyte"sv, "pound"sv, "second"sv, "stone"sv, "terabit"sv, "terabyte"sv, "week"sv, "yard"sv, "year"sv };
  124. return find(sanctioned_units.begin(), sanctioned_units.end(), unit_identifier) != sanctioned_units.end();
  125. };
  126. // 1. If the result of IsSanctionedSimpleUnitIdentifier(unitIdentifier) is true, then
  127. if (is_sanctioned_simple_unit_identifier(unit_identifier)) {
  128. // a. Return true.
  129. return true;
  130. }
  131. auto indices = unit_identifier.find_all("-per-"sv);
  132. // 2. If the substring "-per-" does not occur exactly once in unitIdentifier, then
  133. if (indices.size() != 1) {
  134. // a. Return false.
  135. return false;
  136. }
  137. // 3. Let numerator be the substring of unitIdentifier from the beginning to just before "-per-".
  138. auto numerator = unit_identifier.substring_view(0, indices[0]);
  139. // 4. If the result of IsSanctionedSimpleUnitIdentifier(numerator) is false, then
  140. if (!is_sanctioned_simple_unit_identifier(numerator)) {
  141. // a. Return false.
  142. return false;
  143. }
  144. // 5. Let denominator be the substring of unitIdentifier from just after "-per-" to the end.
  145. auto denominator = unit_identifier.substring_view(indices[0] + 5);
  146. // 6. If the result of IsSanctionedSimpleUnitIdentifier(denominator) is false, then
  147. if (!is_sanctioned_simple_unit_identifier(denominator)) {
  148. // a. Return false.
  149. return false;
  150. }
  151. // 7. Return true.
  152. return true;
  153. }
  154. // 9.2.1 CanonicalizeLocaleList ( locales ), https://tc39.es/ecma402/#sec-canonicalizelocalelist
  155. ThrowCompletionOr<Vector<String>> canonicalize_locale_list(GlobalObject& global_object, Value locales)
  156. {
  157. auto& vm = global_object.vm();
  158. // 1. If locales is undefined, then
  159. if (locales.is_undefined()) {
  160. // a. Return a new empty List.
  161. return Vector<String> {};
  162. }
  163. // 2. Let seen be a new empty List.
  164. Vector<String> seen;
  165. Object* object = nullptr;
  166. // 3. If Type(locales) is String or Type(locales) is Object and locales has an [[InitializedLocale]] internal slot, then
  167. if (locales.is_string() || (locales.is_object() && is<Locale>(locales.as_object()))) {
  168. // a. Let O be CreateArrayFromList(« locales »).
  169. object = Array::create_from(global_object, { locales });
  170. }
  171. // 4. Else,
  172. else {
  173. // a. Let O be ? ToObject(locales).
  174. object = TRY(locales.to_object(global_object));
  175. }
  176. // 5. Let len be ? ToLength(? Get(O, "length")).
  177. auto length_value = TRY(object->get(vm.names.length));
  178. auto length = TRY(length_value.to_length(global_object));
  179. // 6. Let k be 0.
  180. // 7. Repeat, while k < len,
  181. for (size_t k = 0; k < length; ++k) {
  182. // a. Let Pk be ToString(k).
  183. auto property_key = PropertyKey { k };
  184. // b. Let kPresent be ? HasProperty(O, Pk).
  185. auto key_present = TRY(object->has_property(property_key));
  186. // c. If kPresent is true, then
  187. if (key_present) {
  188. // i. Let kValue be ? Get(O, Pk).
  189. auto key_value = TRY(object->get(property_key));
  190. // ii. If Type(kValue) is not String or Object, throw a TypeError exception.
  191. if (!key_value.is_string() && !key_value.is_object())
  192. return vm.throw_completion<TypeError>(global_object, ErrorType::NotAnObjectOrString, key_value.to_string_without_side_effects());
  193. String tag;
  194. // iii. If Type(kValue) is Object and kValue has an [[InitializedLocale]] internal slot, then
  195. if (key_value.is_object() && is<Locale>(key_value.as_object())) {
  196. // 1. Let tag be kValue.[[Locale]].
  197. tag = static_cast<Locale const&>(key_value.as_object()).locale();
  198. }
  199. // iv. Else,
  200. else {
  201. // 1. Let tag be ? ToString(kValue).
  202. tag = TRY(key_value.to_string(global_object));
  203. }
  204. // v. If IsStructurallyValidLanguageTag(tag) is false, throw a RangeError exception.
  205. auto locale_id = is_structurally_valid_language_tag(tag);
  206. if (!locale_id.has_value())
  207. return vm.throw_completion<RangeError>(global_object, ErrorType::IntlInvalidLanguageTag, tag);
  208. // vi. Let canonicalizedTag be CanonicalizeUnicodeLocaleId(tag).
  209. auto canonicalized_tag = JS::Intl::canonicalize_unicode_locale_id(*locale_id);
  210. // vii. If canonicalizedTag is not an element of seen, append canonicalizedTag as the last element of seen.
  211. if (!seen.contains_slow(canonicalized_tag))
  212. seen.append(move(canonicalized_tag));
  213. }
  214. // d. Increase k by 1.
  215. }
  216. return seen;
  217. }
  218. // 9.2.2 BestAvailableLocale ( availableLocales, locale ), https://tc39.es/ecma402/#sec-bestavailablelocale
  219. Optional<String> best_available_locale(StringView locale)
  220. {
  221. // 1. Let candidate be locale.
  222. StringView candidate = locale;
  223. // 2. Repeat,
  224. while (true) {
  225. // a. If availableLocales contains an element equal to candidate, return candidate.
  226. if (Unicode::is_locale_available(candidate))
  227. return candidate;
  228. // b. Let pos be the character index of the last occurrence of "-" (U+002D) within candidate. If that character does not occur, return undefined.
  229. auto pos = candidate.find_last('-');
  230. if (!pos.has_value())
  231. return {};
  232. // c. If pos ≥ 2 and the character "-" occurs at index pos-2 of candidate, decrease pos by 2.
  233. if ((*pos >= 2) && (candidate[*pos - 2] == '-'))
  234. pos = *pos - 2;
  235. // d. Let candidate be the substring of candidate from position 0, inclusive, to position pos, exclusive.
  236. candidate = candidate.substring_view(0, *pos);
  237. }
  238. }
  239. struct MatcherResult {
  240. String locale;
  241. Vector<Unicode::Extension> extensions {};
  242. };
  243. // 9.2.3 LookupMatcher ( availableLocales, requestedLocales ), https://tc39.es/ecma402/#sec-lookupmatcher
  244. static MatcherResult lookup_matcher(Vector<String> const& requested_locales)
  245. {
  246. // 1. Let result be a new Record.
  247. MatcherResult result {};
  248. // 2. For each element locale of requestedLocales, do
  249. for (auto const& locale : requested_locales) {
  250. auto locale_id = Unicode::parse_unicode_locale_id(locale);
  251. VERIFY(locale_id.has_value());
  252. // a. Let noExtensionsLocale be the String value that is locale with any Unicode locale extension sequences removed.
  253. auto extensions = locale_id->remove_extension_type<Unicode::LocaleExtension>();
  254. auto no_extensions_locale = locale_id->to_string();
  255. // b. Let availableLocale be BestAvailableLocale(availableLocales, noExtensionsLocale).
  256. auto available_locale = best_available_locale(no_extensions_locale);
  257. // c. If availableLocale is not undefined, then
  258. if (available_locale.has_value()) {
  259. // i. Set result.[[locale]] to availableLocale.
  260. result.locale = available_locale.release_value();
  261. // ii. If locale and noExtensionsLocale are not the same String value, then
  262. if (locale != no_extensions_locale) {
  263. // 1. Let extension be the String value consisting of the substring of the Unicode locale extension sequence within locale.
  264. // 2. Set result.[[extension]] to extension.
  265. result.extensions.extend(move(extensions));
  266. }
  267. // iii. Return result.
  268. return result;
  269. }
  270. }
  271. // 3. Let defLocale be DefaultLocale().
  272. // 4. Set result.[[locale]] to defLocale.
  273. result.locale = Unicode::default_locale();
  274. // 5. Return result.
  275. return result;
  276. }
  277. // 9.2.4 BestFitMatcher ( availableLocales, requestedLocales ), https://tc39.es/ecma402/#sec-bestfitmatcher
  278. static MatcherResult best_fit_matcher(Vector<String> const& requested_locales)
  279. {
  280. // The algorithm is implementation dependent, but should produce results that a typical user of the requested locales would
  281. // perceive as at least as good as those produced by the LookupMatcher abstract operation.
  282. return lookup_matcher(requested_locales);
  283. }
  284. // 9.2.6 InsertUnicodeExtensionAndCanonicalize ( locale, extension ), https://tc39.es/ecma402/#sec-insert-unicode-extension-and-canonicalize
  285. String insert_unicode_extension_and_canonicalize(Unicode::LocaleID locale, Unicode::LocaleExtension extension)
  286. {
  287. // Note: This implementation differs from the spec in how the extension is inserted. The spec assumes
  288. // the input to this method is a string, and is written such that operations are performed on parts
  289. // of that string. LibUnicode gives us the parsed locale in a structure, so we can mutate that
  290. // structure directly.
  291. locale.extensions.append(move(extension));
  292. return JS::Intl::canonicalize_unicode_locale_id(locale);
  293. }
  294. template<typename T>
  295. static auto& find_key_in_value(T& value, StringView key)
  296. {
  297. if (key == "ca"sv)
  298. return value.ca;
  299. if (key == "hc"sv)
  300. return value.hc;
  301. if (key == "nu"sv)
  302. return value.nu;
  303. // If you hit this point, you must add any missing keys from [[RelevantExtensionKeys]] to LocaleOptions and LocaleResult.
  304. VERIFY_NOT_REACHED();
  305. }
  306. // 9.2.7 ResolveLocale ( availableLocales, requestedLocales, options, relevantExtensionKeys, localeData ), https://tc39.es/ecma402/#sec-resolvelocale
  307. LocaleResult resolve_locale(Vector<String> const& requested_locales, LocaleOptions const& options, Span<StringView const> relevant_extension_keys)
  308. {
  309. // 1. Let matcher be options.[[localeMatcher]].
  310. auto const& matcher = options.locale_matcher;
  311. MatcherResult matcher_result;
  312. // 2. If matcher is "lookup", then
  313. if (matcher.is_string() && (matcher.as_string().string() == "lookup"sv)) {
  314. // a. Let r be LookupMatcher(availableLocales, requestedLocales).
  315. matcher_result = lookup_matcher(requested_locales);
  316. }
  317. // 3. Else,
  318. else {
  319. // a. Let r be BestFitMatcher(availableLocales, requestedLocales).
  320. matcher_result = best_fit_matcher(requested_locales);
  321. }
  322. // 4. Let foundLocale be r.[[locale]].
  323. auto found_locale = move(matcher_result.locale);
  324. // 5. Let result be a new Record.
  325. LocaleResult result {};
  326. // 6. Set result.[[dataLocale]] to foundLocale.
  327. result.data_locale = found_locale;
  328. // 7. If r has an [[extension]] field, then
  329. Vector<Unicode::Keyword> keywords;
  330. for (auto& extension : matcher_result.extensions) {
  331. if (!extension.has<Unicode::LocaleExtension>())
  332. continue;
  333. // a. Let components be ! UnicodeExtensionComponents(r.[[extension]]).
  334. auto& components = extension.get<Unicode::LocaleExtension>();
  335. // b. Let keywords be components.[[Keywords]].
  336. keywords = move(components.keywords);
  337. break;
  338. }
  339. // 8. Let supportedExtension be "-u".
  340. Unicode::LocaleExtension supported_extension {};
  341. // 9. For each element key of relevantExtensionKeys, do
  342. for (auto const& key : relevant_extension_keys) {
  343. // a. Let foundLocaleData be localeData.[[<foundLocale>]].
  344. // b. Assert: Type(foundLocaleData) is Record.
  345. // c. Let keyLocaleData be foundLocaleData.[[<key>]].
  346. // d. Assert: Type(keyLocaleData) is List.
  347. auto key_locale_data = Unicode::get_locale_key_mapping(found_locale, key);
  348. // e. Let value be keyLocaleData[0].
  349. // f. Assert: Type(value) is either String or Null.
  350. Optional<String> value;
  351. if (!key_locale_data.is_empty())
  352. value = key_locale_data[0];
  353. // g. Let supportedExtensionAddition be "".
  354. Optional<Unicode::Keyword> supported_extension_addition {};
  355. // h. If r has an [[extension]] field, then
  356. for (auto& entry : keywords) {
  357. // i. If keywords contains an element whose [[Key]] is the same as key, then
  358. if (entry.key != key)
  359. continue;
  360. // 1. Let entry be the element of keywords whose [[Key]] is the same as key.
  361. // 2. Let requestedValue be entry.[[Value]].
  362. auto requested_value = entry.value;
  363. // 3. If requestedValue is not the empty String, then
  364. if (!requested_value.is_empty()) {
  365. // a. If keyLocaleData contains requestedValue, then
  366. if (key_locale_data.contains_slow(requested_value)) {
  367. // i. Let value be requestedValue.
  368. value = move(requested_value);
  369. // ii. Let supportedExtensionAddition be the string-concatenation of "-", key, "-", and value.
  370. supported_extension_addition = Unicode::Keyword { key, move(entry.value) };
  371. }
  372. }
  373. // 4. Else if keyLocaleData contains "true", then
  374. else if (key_locale_data.contains_slow("true"sv)) {
  375. // a. Let value be "true".
  376. value = "true"sv;
  377. // b. Let supportedExtensionAddition be the string-concatenation of "-" and key.
  378. supported_extension_addition = Unicode::Keyword { key, {} };
  379. }
  380. break;
  381. }
  382. // i. If options has a field [[<key>]], then
  383. // i. Let optionsValue be options.[[<key>]].
  384. // ii. Assert: Type(optionsValue) is either String, Undefined, or Null.
  385. auto options_value = find_key_in_value(options, key);
  386. // iii. If Type(optionsValue) is String, then
  387. if (options_value.has_value()) {
  388. // 1. Let optionsValue be the string optionsValue after performing the algorithm steps to transform Unicode extension values to canonical syntax per Unicode Technical Standard #35 LDML § 3.2.1 Canonical Unicode Locale Identifiers, treating key as ukey and optionsValue as uvalue productions.
  389. // 2. Let optionsValue be the string optionsValue after performing the algorithm steps to replace Unicode extension values with their canonical form per Unicode Technical Standard #35 LDML § 3.2.1 Canonical Unicode Locale Identifiers, treating key as ukey and optionsValue as uvalue productions.
  390. Unicode::canonicalize_unicode_extension_values(key, *options_value, true);
  391. // 3. If optionsValue is the empty String, then
  392. if (options_value->is_empty()) {
  393. // a. Let optionsValue be "true".
  394. options_value = "true"sv;
  395. }
  396. }
  397. // iv. If keyLocaleData contains optionsValue, then
  398. if (options_value.has_value() && key_locale_data.contains_slow(*options_value)) {
  399. // 1. If SameValue(optionsValue, value) is false, then
  400. if (options_value != value) {
  401. // a. Let value be optionsValue.
  402. value = move(options_value);
  403. // b. Let supportedExtensionAddition be "".
  404. supported_extension_addition.clear();
  405. }
  406. }
  407. // j. Set result.[[<key>]] to value.
  408. find_key_in_value(result, key) = move(value);
  409. // k. Append supportedExtensionAddition to supportedExtension.
  410. if (supported_extension_addition.has_value())
  411. supported_extension.keywords.append(supported_extension_addition.release_value());
  412. }
  413. // 10. If the number of elements in supportedExtension is greater than 2, then
  414. if (!supported_extension.keywords.is_empty()) {
  415. auto locale_id = Unicode::parse_unicode_locale_id(found_locale);
  416. VERIFY(locale_id.has_value());
  417. // a. Let foundLocale be InsertUnicodeExtensionAndCanonicalize(foundLocale, supportedExtension).
  418. found_locale = insert_unicode_extension_and_canonicalize(locale_id.release_value(), move(supported_extension));
  419. }
  420. // 11. Set result.[[locale]] to foundLocale.
  421. result.locale = move(found_locale);
  422. // 12. Return result.
  423. return result;
  424. }
  425. // 9.2.8 LookupSupportedLocales ( availableLocales, requestedLocales ), https://tc39.es/ecma402/#sec-lookupsupportedlocales
  426. Vector<String> lookup_supported_locales(Vector<String> const& requested_locales)
  427. {
  428. // 1. Let subset be a new empty List.
  429. Vector<String> subset;
  430. // 2. For each element locale of requestedLocales, do
  431. for (auto const& locale : requested_locales) {
  432. auto locale_id = Unicode::parse_unicode_locale_id(locale);
  433. VERIFY(locale_id.has_value());
  434. // a. Let noExtensionsLocale be the String value that is locale with any Unicode locale extension sequences removed.
  435. locale_id->remove_extension_type<Unicode::LocaleExtension>();
  436. auto no_extensions_locale = locale_id->to_string();
  437. // b. Let availableLocale be BestAvailableLocale(availableLocales, noExtensionsLocale).
  438. auto available_locale = best_available_locale(no_extensions_locale);
  439. // c. If availableLocale is not undefined, append locale to the end of subset.
  440. if (available_locale.has_value())
  441. subset.append(locale);
  442. }
  443. // 3. Return subset.
  444. return subset;
  445. }
  446. // 9.2.9 BestFitSupportedLocales ( availableLocales, requestedLocales ), https://tc39.es/ecma402/#sec-bestfitsupportedlocales
  447. Vector<String> best_fit_supported_locales(Vector<String> const& requested_locales)
  448. {
  449. // The BestFitSupportedLocales abstract operation returns the subset of the provided BCP 47
  450. // language priority list requestedLocales for which availableLocales has a matching locale
  451. // when using the Best Fit Matcher algorithm. Locales appear in the same order in the returned
  452. // list as in requestedLocales. The steps taken are implementation dependent.
  453. // :yakbrain:
  454. return lookup_supported_locales(requested_locales);
  455. }
  456. // 9.2.10 SupportedLocales ( availableLocales, requestedLocales, options ), https://tc39.es/ecma402/#sec-supportedlocales
  457. ThrowCompletionOr<Array*> supported_locales(GlobalObject& global_object, Vector<String> const& requested_locales, Value options)
  458. {
  459. auto& vm = global_object.vm();
  460. // 1. Set options to ? CoerceOptionsToObject(options).
  461. auto* options_object = TRY(coerce_options_to_object(global_object, options));
  462. // 2. Let matcher be ? GetOption(options, "localeMatcher", "string", « "lookup", "best fit" », "best fit").
  463. auto matcher = TRY(get_option(global_object, *options_object, vm.names.localeMatcher, Value::Type::String, { "lookup"sv, "best fit"sv }, "best fit"sv));
  464. Vector<String> supported_locales;
  465. // 3. If matcher is "best fit", then
  466. if (matcher.as_string().string() == "best fit"sv) {
  467. // a. Let supportedLocales be BestFitSupportedLocales(availableLocales, requestedLocales).
  468. supported_locales = best_fit_supported_locales(requested_locales);
  469. }
  470. // 4. Else,
  471. else {
  472. // a. Let supportedLocales be LookupSupportedLocales(availableLocales, requestedLocales).
  473. supported_locales = lookup_supported_locales(requested_locales);
  474. }
  475. // 5. Return CreateArrayFromList(supportedLocales).
  476. return Array::create_from<String>(global_object, supported_locales, [&vm](auto& locale) { return js_string(vm, locale); });
  477. }
  478. // 9.2.12 CoerceOptionsToObject ( options ), https://tc39.es/ecma402/#sec-coerceoptionstoobject
  479. ThrowCompletionOr<Object*> coerce_options_to_object(GlobalObject& global_object, Value options)
  480. {
  481. // 1. If options is undefined, then
  482. if (options.is_undefined()) {
  483. // a. Return ! OrdinaryObjectCreate(null).
  484. return Object::create(global_object, nullptr);
  485. }
  486. // 2. Return ? ToObject(options).
  487. return TRY(options.to_object(global_object));
  488. }
  489. // 9.2.13 GetOption ( options, property, type, values, fallback ), https://tc39.es/ecma402/#sec-getoption
  490. ThrowCompletionOr<Value> get_option(GlobalObject& global_object, Object const& options, PropertyKey const& property, Value::Type type, Span<StringView const> values, Fallback fallback)
  491. {
  492. auto& vm = global_object.vm();
  493. // 1. Assert: Type(options) is Object.
  494. // 2. Let value be ? Get(options, property).
  495. auto value = TRY(options.get(property));
  496. // 3. If value is undefined, return fallback.
  497. if (value.is_undefined()) {
  498. return fallback.visit(
  499. [](Empty) { return js_undefined(); },
  500. [](bool f) { return Value(f); },
  501. [&vm](StringView f) { return Value(js_string(vm, f)); });
  502. }
  503. // 4. Assert: type is "boolean" or "string".
  504. VERIFY((type == Value::Type::Boolean) || (type == Value::Type::String));
  505. // 5. If type is "boolean", then
  506. if (type == Value::Type::Boolean) {
  507. // a. Set value to ! ToBoolean(value).
  508. value = Value(value.to_boolean());
  509. }
  510. // 6. If type is "string", then
  511. else {
  512. // a. Set value to ? ToString(value).
  513. value = TRY(value.to_primitive_string(global_object));
  514. }
  515. // 7. If values is not undefined and values does not contain an element equal to value, throw a RangeError exception.
  516. if (!values.is_empty()) {
  517. // Note: Every location in the spec that invokes GetOption with type=boolean also has values=undefined.
  518. VERIFY(value.is_string());
  519. if (!values.contains_slow(value.as_string().string()))
  520. return vm.throw_completion<RangeError>(global_object, ErrorType::OptionIsNotValidValue, value.to_string_without_side_effects(), property.as_string());
  521. }
  522. // 8. Return value.
  523. return value;
  524. }
  525. // 9.2.14 DefaultNumberOption ( value, minimum, maximum, fallback ), https://tc39.es/ecma402/#sec-defaultnumberoption
  526. ThrowCompletionOr<Optional<int>> default_number_option(GlobalObject& global_object, Value value, int minimum, int maximum, Optional<int> fallback)
  527. {
  528. auto& vm = global_object.vm();
  529. // 1. If value is undefined, return fallback.
  530. if (value.is_undefined())
  531. return fallback;
  532. // 2. Set value to ? ToNumber(value).
  533. value = TRY(value.to_number(global_object));
  534. // 3. If value is NaN or less than minimum or greater than maximum, throw a RangeError exception.
  535. if (value.is_nan() || (value.as_double() < minimum) || (value.as_double() > maximum))
  536. return vm.throw_completion<RangeError>(global_object, ErrorType::IntlNumberIsNaNOrOutOfRange, value, minimum, maximum);
  537. // 4. Return floor(value).
  538. return floor(value.as_double());
  539. }
  540. // 9.2.15 GetNumberOption ( options, property, minimum, maximum, fallback ), https://tc39.es/ecma402/#sec-getnumberoption
  541. ThrowCompletionOr<Optional<int>> get_number_option(GlobalObject& global_object, Object const& options, PropertyKey const& property, int minimum, int maximum, Optional<int> fallback)
  542. {
  543. // 1. Assert: Type(options) is Object.
  544. // 2. Let value be ? Get(options, property).
  545. auto value = TRY(options.get(property));
  546. // 3. Return ? DefaultNumberOption(value, minimum, maximum, fallback).
  547. return default_number_option(global_object, value, minimum, maximum, move(fallback));
  548. }
  549. // 9.2.16 PartitionPattern ( pattern ), https://tc39.es/ecma402/#sec-partitionpattern
  550. Vector<PatternPartition> partition_pattern(StringView pattern)
  551. {
  552. // 1. Let result be a new empty List.
  553. Vector<PatternPartition> result;
  554. // 2. Let beginIndex be ! StringIndexOf(pattern, "{", 0).
  555. auto begin_index = pattern.find('{', 0);
  556. // 3. Let endIndex be 0.
  557. size_t end_index = 0;
  558. // 4. Let nextIndex be 0.
  559. size_t next_index = 0;
  560. // 5. Let length be the number of code units in pattern.
  561. // 6. Repeat, while beginIndex is an integer index into pattern,
  562. while (begin_index.has_value()) {
  563. // a. Set endIndex to ! StringIndexOf(pattern, "}", beginIndex).
  564. end_index = pattern.find('}', *begin_index).value();
  565. // b. Assert: endIndex is greater than beginIndex.
  566. VERIFY(end_index > *begin_index);
  567. // c. If beginIndex is greater than nextIndex, then
  568. if (*begin_index > next_index) {
  569. // i. Let literal be a substring of pattern from position nextIndex, inclusive, to position beginIndex, exclusive.
  570. auto literal = pattern.substring_view(next_index, *begin_index - next_index);
  571. // ii. Append a new Record { [[Type]]: "literal", [[Value]]: literal } as the last element of the list result.
  572. result.append({ "literal"sv, literal });
  573. }
  574. // d. Let p be the substring of pattern from position beginIndex, exclusive, to position endIndex, exclusive.
  575. auto partition = pattern.substring_view(*begin_index + 1, end_index - *begin_index - 1);
  576. // e. Append a new Record { [[Type]]: p, [[Value]]: undefined } as the last element of the list result.
  577. result.append({ partition, {} });
  578. // f. Set nextIndex to endIndex + 1.
  579. next_index = end_index + 1;
  580. // g. Set beginIndex to ! StringIndexOf(pattern, "{", nextIndex).
  581. begin_index = pattern.find('{', next_index);
  582. }
  583. // 7. If nextIndex is less than length, then
  584. if (next_index < pattern.length()) {
  585. // a. Let literal be the substring of pattern from position nextIndex, inclusive, to position length, exclusive.
  586. auto literal = pattern.substring_view(next_index);
  587. // b. Append a new Record { [[Type]]: "literal", [[Value]]: literal } as the last element of the list result.
  588. result.append({ "literal"sv, literal });
  589. }
  590. // 8. Return result.
  591. return result;
  592. }
  593. }