2021-08-24 22:58:45 -04:00
/*
2023-01-12 10:22:37 -05:00
* Copyright ( c ) 2021 - 2023 , Tim Flynn < trflynn89 @ serenityos . org >
2021-08-24 22:58:45 -04:00
*
* SPDX - License - Identifier : BSD - 2 - Clause
*/
# include <AK/AllOf.h>
2021-08-28 16:04:55 -04:00
# include <AK/AnyOf.h>
2021-08-24 22:58:45 -04:00
# include <AK/CharacterTypes.h>
2021-09-10 11:04:54 -04:00
# include <AK/Find.h>
2021-09-04 17:08:57 +01:00
# include <AK/Function.h>
2021-08-24 22:58:45 -04:00
# include <AK/QuickSort.h>
# include <AK/TypeCasts.h>
# include <LibJS/Runtime/Array.h>
# include <LibJS/Runtime/GlobalObject.h>
# include <LibJS/Runtime/Intl/AbstractOperations.h>
2021-09-01 22:26:06 -04:00
# include <LibJS/Runtime/Intl/Locale.h>
2022-09-02 12:11:30 -04:00
# include <LibLocale/Locale.h>
2021-08-24 22:58:45 -04:00
namespace JS : : Intl {
// 6.2.2 IsStructurallyValidLanguageTag ( locale ), https://tc39.es/ecma402/#sec-isstructurallyvalidlanguagetag
2023-01-19 11:39:05 -05:00
ThrowCompletionOr < Optional < : : Locale : : LocaleID > > is_structurally_valid_language_tag ( VM & vm , StringView locale )
2021-08-24 22:58:45 -04:00
{
2023-01-19 11:39:05 -05:00
auto contains_duplicate_variant = [ & ] ( auto & variants ) - > ThrowCompletionOr < bool > {
2021-08-28 14:46:36 -04:00
if ( variants . is_empty ( ) )
return false ;
quick_sort ( variants ) ;
for ( size_t i = 0 ; i < variants . size ( ) - 1 ; + + i ) {
2023-03-08 13:03:04 -05:00
if ( variants [ i ] . equals_ignoring_case ( variants [ i + 1 ] ) )
2021-08-28 14:46:36 -04:00
return true ;
}
return false ;
} ;
2021-08-24 22:58:45 -04:00
// IsStructurallyValidLanguageTag returns true if all of the following conditions hold, false otherwise:
// locale can be generated from the EBNF grammar for unicode_locale_id in Unicode Technical Standard #35 LDML § 3.2 Unicode Locale Identifier;
2023-01-19 11:39:05 -05:00
auto locale_id = TRY_OR_THROW_OOM ( vm , : : Locale : : parse_unicode_locale_id ( locale ) ) ;
2021-08-24 22:58:45 -04:00
if ( ! locale_id . has_value ( ) )
2023-01-19 17:11:00 -05:00
return OptionalNone { } ;
2021-08-24 22:58:45 -04:00
// locale does not use any of the backwards compatibility syntax described in Unicode Technical Standard #35 LDML § 3.3 BCP 47 Conformance;
// https://unicode.org/reports/tr35/#BCP_47_Conformance
if ( locale . contains ( ' _ ' ) | | locale_id - > language_id . is_root | | ! locale_id - > language_id . language . has_value ( ) )
2023-01-19 17:11:00 -05:00
return OptionalNone { } ;
2021-08-24 22:58:45 -04:00
// the unicode_language_id within locale contains no duplicate unicode_variant_subtag subtags; and
2023-01-19 11:39:05 -05:00
if ( TRY ( contains_duplicate_variant ( locale_id - > language_id . variants ) ) )
2023-01-19 17:11:00 -05:00
return OptionalNone { } ;
2021-08-24 22:58:45 -04:00
2021-08-28 14:46:36 -04:00
// if locale contains an extensions* component, that component
Vector < char > unique_keys ;
for ( auto & extension : locale_id - > extensions ) {
// does not contain any other_extensions components with duplicate [alphanum-[tTuUxX]] subtags,
// contains at most one unicode_locale_extensions component,
// contains at most one transformed_extensions component, and
char key = extension . visit (
2022-09-02 12:01:10 -04:00
[ ] ( : : Locale : : LocaleExtension const & ) { return ' u ' ; } ,
[ ] ( : : Locale : : TransformedExtension const & ) { return ' t ' ; } ,
[ ] ( : : Locale : : OtherExtension const & ext ) { return static_cast < char > ( to_ascii_lowercase ( ext . key ) ) ; } ) ;
2021-08-28 14:46:36 -04:00
if ( unique_keys . contains_slow ( key ) )
2023-01-19 17:11:00 -05:00
return OptionalNone { } ;
2023-01-19 11:39:05 -05:00
TRY_OR_THROW_OOM ( vm , unique_keys . try_append ( key ) ) ;
2021-08-28 14:46:36 -04:00
// if a transformed_extensions component that contains a tlang component is present, then
// the tlang component contains no duplicate unicode_variant_subtag subtags.
2022-09-02 12:01:10 -04:00
if ( auto * transformed = extension . get_pointer < : : Locale : : TransformedExtension > ( ) ) {
2021-08-28 14:46:36 -04:00
auto & language = transformed - > language ;
2023-01-19 11:39:05 -05:00
if ( language . has_value ( ) & & TRY ( contains_duplicate_variant ( language - > variants ) ) )
return Optional < : : Locale : : LocaleID > { } ;
2021-08-24 22:58:45 -04:00
}
}
return locale_id ;
}
// 6.2.3 CanonicalizeUnicodeLocaleId ( locale ), https://tc39.es/ecma402/#sec-canonicalizeunicodelocaleid
2023-01-19 13:13:57 -05:00
ThrowCompletionOr < String > canonicalize_unicode_locale_id ( VM & vm , : : Locale : : LocaleID & locale )
2021-08-24 22:58:45 -04:00
{
2021-08-28 16:04:55 -04:00
// Note: This implementation differs from the spec in how Step 3 is implemented. The spec assumes
// the input to this method is a string, and is written such that operations are performed on parts
// of that string. LibUnicode gives us the parsed locale in a structure, so we can mutate that
// structure directly. From a footnote in the spec:
//
// The third step of this algorithm ensures that a Unicode locale extension sequence in the
// returned language tag contains:
// * only the first instance of any attribute duplicated in the input, and
// * only the first keyword for a given key in the input.
for ( auto & extension : locale . extensions ) {
2022-09-02 12:01:10 -04:00
if ( ! extension . has < : : Locale : : LocaleExtension > ( ) )
2021-08-28 16:04:55 -04:00
continue ;
2022-09-02 12:01:10 -04:00
auto & locale_extension = extension . get < : : Locale : : LocaleExtension > ( ) ;
2021-08-28 16:04:55 -04:00
auto attributes = move ( locale_extension . attributes ) ;
for ( auto & attribute : attributes ) {
if ( ! locale_extension . attributes . contains_slow ( attribute ) )
2023-01-19 11:39:05 -05:00
TRY_OR_THROW_OOM ( vm , locale_extension . attributes . try_append ( move ( attribute ) ) ) ;
2021-08-28 16:04:55 -04:00
}
auto keywords = move ( locale_extension . keywords ) ;
for ( auto & keyword : keywords ) {
if ( ! any_of ( locale_extension . keywords , [ & ] ( auto const & k ) { return k . key = = keyword . key ; } ) )
2023-01-19 11:39:05 -05:00
TRY_OR_THROW_OOM ( vm , locale_extension . keywords . try_append ( move ( keyword ) ) ) ;
2021-08-28 16:04:55 -04:00
}
break ;
}
2021-08-24 22:58:45 -04:00
// 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.
// 2. Let localeId be the string localeId after performing the algorithm to transform it to canonical form.
2023-01-19 11:39:05 -05:00
auto locale_id = TRY_OR_THROW_OOM ( vm , : : Locale : : canonicalize_unicode_locale_id ( locale ) ) ;
2021-08-24 22:58:45 -04:00
VERIFY ( locale_id . has_value ( ) ) ;
// 4. Return localeId.
2023-01-19 13:13:57 -05:00
return locale_id . release_value ( ) ;
2021-08-24 22:58:45 -04:00
}
2022-03-28 09:53:06 -04:00
// 6.3.1 IsWellFormedCurrencyCode ( currency ), https://tc39.es/ecma402/#sec-iswellformedcurrencycode
2021-09-10 11:04:54 -04:00
bool is_well_formed_currency_code ( StringView currency )
2021-08-24 22:58:45 -04:00
{
2022-03-28 09:53:06 -04:00
// 1. If the length of currency is not 3, return false.
2021-08-24 22:58:45 -04:00
if ( currency . length ( ) ! = 3 )
return false ;
2022-03-28 09:53:06 -04:00
// 2. Let normalized be the ASCII-uppercase of currency.
// 3. If normalized contains any code unit outside of 0x0041 through 0x005A (corresponding to Unicode characters LATIN CAPITAL LETTER A through LATIN CAPITAL LETTER Z), return false.
2021-08-24 22:58:45 -04:00
if ( ! all_of ( currency , is_ascii_alpha ) )
return false ;
// 4. Return true.
return true ;
}
2021-09-10 11:04:54 -04:00
// 6.5.1 IsWellFormedUnitIdentifier ( unitIdentifier ), https://tc39.es/ecma402/#sec-iswellformedunitidentifier
bool is_well_formed_unit_identifier ( StringView unit_identifier )
{
2022-03-28 12:05:18 -04:00
// 6.5.2 IsSanctionedSingleUnitIdentifier ( unitIdentifier ), https://tc39.es/ecma402/#sec-issanctionedsingleunitidentifier
constexpr auto is_sanctioned_single_unit_identifier = [ ] ( StringView unit_identifier ) {
2021-09-10 11:04:54 -04:00
// 1. If unitIdentifier is listed in Table 2 below, return true.
// 2. Else, return false.
2022-03-28 12:05:18 -04:00
static constexpr auto sanctioned_units = sanctioned_single_unit_identifiers ( ) ;
2021-09-10 11:04:54 -04:00
return find ( sanctioned_units . begin ( ) , sanctioned_units . end ( ) , unit_identifier ) ! = sanctioned_units . end ( ) ;
} ;
2022-03-28 12:05:18 -04:00
// 1. If ! IsSanctionedSingleUnitIdentifier(unitIdentifier) is true, then
if ( is_sanctioned_single_unit_identifier ( unit_identifier ) ) {
2021-09-10 11:04:54 -04:00
// a. Return true.
return true ;
}
2022-05-02 20:54:39 +02:00
// 2. Let i be StringIndexOf(unitIdentifier, "-per-", 0).
2021-09-10 11:04:54 -04:00
auto indices = unit_identifier . find_all ( " -per- " sv ) ;
2022-05-02 20:54:39 +02:00
// 3. If i is -1 or StringIndexOf(unitIdentifier, "-per-", i + 1) is not -1, then
2021-09-10 11:04:54 -04:00
if ( indices . size ( ) ! = 1 ) {
// a. Return false.
return false ;
}
2022-03-28 12:05:18 -04:00
// 4. Assert: The five-character substring "-per-" occurs exactly once in unitIdentifier, at index i.
// NOTE: We skip this because the indices vector being of size 1 already verifies this invariant.
2021-09-10 11:04:54 -04:00
2022-03-28 12:05:18 -04:00
// 5. Let numerator be the substring of unitIdentifier from 0 to i.
auto numerator = unit_identifier . substring_view ( 0 , indices [ 0 ] ) ;
2021-09-10 11:04:54 -04:00
2022-03-28 12:05:18 -04:00
// 6. Let denominator be the substring of unitIdentifier from i + 5.
2021-09-10 11:04:54 -04:00
auto denominator = unit_identifier . substring_view ( indices [ 0 ] + 5 ) ;
2022-03-28 12:05:18 -04:00
// 7. If ! IsSanctionedSingleUnitIdentifier(numerator) and ! IsSanctionedSingleUnitIdentifier(denominator) are both true, then
if ( is_sanctioned_single_unit_identifier ( numerator ) & & is_sanctioned_single_unit_identifier ( denominator ) ) {
// a. Return true.
return true ;
2021-09-10 11:04:54 -04:00
}
2022-03-28 12:05:18 -04:00
// 8. Return false.
return false ;
2021-09-10 11:04:54 -04:00
}
2021-08-24 22:58:45 -04:00
// 9.2.1 CanonicalizeLocaleList ( locales ), https://tc39.es/ecma402/#sec-canonicalizelocalelist
2023-01-19 13:13:57 -05:00
ThrowCompletionOr < Vector < String > > canonicalize_locale_list ( VM & vm , Value locales )
2021-08-24 22:58:45 -04:00
{
2022-08-20 08:25:24 +01:00
auto & realm = * vm . current_realm ( ) ;
2021-08-24 22:58:45 -04:00
// 1. If locales is undefined, then
if ( locales . is_undefined ( ) ) {
// a. Return a new empty List.
2023-01-19 13:13:57 -05:00
return Vector < String > { } ;
2021-08-24 22:58:45 -04:00
}
// 2. Let seen be a new empty List.
2023-01-19 13:13:57 -05:00
Vector < String > seen ;
2021-08-24 22:58:45 -04:00
Object * object = nullptr ;
// 3. If Type(locales) is String or Type(locales) is Object and locales has an [[InitializedLocale]] internal slot, then
2021-09-01 22:26:06 -04:00
if ( locales . is_string ( ) | | ( locales . is_object ( ) & & is < Locale > ( locales . as_object ( ) ) ) ) {
2022-05-02 20:54:39 +02:00
// a. Let O be CreateArrayFromList(« locales »).
2022-08-16 00:20:49 +01:00
object = Array : : create_from ( realm , { locales } ) ;
2021-08-24 22:58:45 -04:00
}
// 4. Else,
else {
// a. Let O be ? ToObject(locales).
2022-08-21 14:00:56 +01:00
object = TRY ( locales . to_object ( vm ) ) ;
2021-08-24 22:58:45 -04:00
}
// 5. Let len be ? ToLength(? Get(O, "length")).
2021-10-17 23:57:37 +03:00
auto length_value = TRY ( object - > get ( vm . names . length ) ) ;
2022-08-21 14:00:56 +01:00
auto length = TRY ( length_value . to_length ( vm ) ) ;
2021-08-24 22:58:45 -04:00
// 6. Let k be 0.
// 7. Repeat, while k < len,
for ( size_t k = 0 ; k < length ; + + k ) {
// a. Let Pk be ToString(k).
2021-10-24 16:01:24 +02:00
auto property_key = PropertyKey { k } ;
2021-08-24 22:58:45 -04:00
// b. Let kPresent be ? HasProperty(O, Pk).
2021-10-03 02:00:39 +01:00
auto key_present = TRY ( object - > has_property ( property_key ) ) ;
2021-08-24 22:58:45 -04:00
// c. If kPresent is true, then
if ( key_present ) {
// i. Let kValue be ? Get(O, Pk).
2021-10-02 23:52:27 +01:00
auto key_value = TRY ( object - > get ( property_key ) ) ;
2021-08-24 22:58:45 -04:00
// ii. If Type(kValue) is not String or Object, throw a TypeError exception.
2021-09-18 19:24:28 +03:00
if ( ! key_value . is_string ( ) & & ! key_value . is_object ( ) )
2023-01-19 13:13:57 -05:00
return vm . throw_completion < TypeError > ( ErrorType : : NotAnObjectOrString , key_value ) ;
2021-08-24 22:58:45 -04:00
2023-01-19 13:13:57 -05:00
String tag ;
2021-09-01 22:26:06 -04:00
2021-08-24 22:58:45 -04:00
// iii. If Type(kValue) is Object and kValue has an [[InitializedLocale]] internal slot, then
2021-09-01 22:26:06 -04:00
if ( key_value . is_object ( ) & & is < Locale > ( key_value . as_object ( ) ) ) {
// 1. Let tag be kValue.[[Locale]].
tag = static_cast < Locale const & > ( key_value . as_object ( ) ) . locale ( ) ;
}
2021-08-24 22:58:45 -04:00
// iv. Else,
2021-09-01 22:26:06 -04:00
else {
// 1. Let tag be ? ToString(kValue).
2023-01-19 13:13:57 -05:00
tag = TRY ( key_value . to_string ( vm ) ) ;
2021-09-01 22:26:06 -04:00
}
2021-08-24 22:58:45 -04:00
2022-03-28 08:30:48 -04:00
// v. If ! IsStructurallyValidLanguageTag(tag) is false, throw a RangeError exception.
2023-01-20 14:26:48 -05:00
auto locale_id = MUST_OR_THROW_OOM ( is_structurally_valid_language_tag ( vm , tag ) ) ;
2021-09-18 19:24:28 +03:00
if ( ! locale_id . has_value ( ) )
2022-08-16 20:33:17 +01:00
return vm . throw_completion < RangeError > ( ErrorType : : IntlInvalidLanguageTag , tag ) ;
2021-08-24 22:58:45 -04:00
2022-03-28 08:30:48 -04:00
// vi. Let canonicalizedTag be ! CanonicalizeUnicodeLocaleId(tag).
2023-01-20 14:26:48 -05:00
auto canonicalized_tag = MUST_OR_THROW_OOM ( canonicalize_unicode_locale_id ( vm , * locale_id ) ) ;
2021-08-24 22:58:45 -04:00
// vii. If canonicalizedTag is not an element of seen, append canonicalizedTag as the last element of seen.
if ( ! seen . contains_slow ( canonicalized_tag ) )
2023-01-19 11:39:05 -05:00
TRY_OR_THROW_OOM ( vm , seen . try_append ( move ( canonicalized_tag ) ) ) ;
2021-08-24 22:58:45 -04:00
}
// d. Increase k by 1.
}
return seen ;
}
// 9.2.2 BestAvailableLocale ( availableLocales, locale ), https://tc39.es/ecma402/#sec-bestavailablelocale
2023-01-19 13:13:57 -05:00
Optional < StringView > best_available_locale ( StringView locale )
2021-08-24 22:58:45 -04:00
{
// 1. Let candidate be locale.
StringView candidate = locale ;
// 2. Repeat,
while ( true ) {
2023-01-30 10:35:38 -05:00
// a. If availableLocales contains candidate, return candidate.
2022-09-02 12:01:10 -04:00
if ( : : Locale : : is_locale_available ( candidate ) )
2021-08-24 22:58:45 -04:00
return candidate ;
// b. Let pos be the character index of the last occurrence of "-" (U+002D) within candidate. If that character does not occur, return undefined.
auto pos = candidate . find_last ( ' - ' ) ;
if ( ! pos . has_value ( ) )
return { } ;
// c. If pos ≥ 2 and the character "-" occurs at index pos-2 of candidate, decrease pos by 2.
if ( ( * pos > = 2 ) & & ( candidate [ * pos - 2 ] = = ' - ' ) )
pos = * pos - 2 ;
// d. Let candidate be the substring of candidate from position 0, inclusive, to position pos, exclusive.
candidate = candidate . substring_view ( 0 , * pos ) ;
}
}
2021-08-30 08:51:52 -04:00
struct MatcherResult {
2023-01-19 13:13:57 -05:00
String locale ;
2022-09-02 12:01:10 -04:00
Vector < : : Locale : : Extension > extensions { } ;
2021-08-30 08:51:52 -04:00
} ;
2021-08-24 22:58:45 -04:00
// 9.2.3 LookupMatcher ( availableLocales, requestedLocales ), https://tc39.es/ecma402/#sec-lookupmatcher
2023-01-19 13:13:57 -05:00
static ThrowCompletionOr < MatcherResult > lookup_matcher ( VM & vm , Vector < String > const & requested_locales )
2021-08-24 22:58:45 -04:00
{
// 1. Let result be a new Record.
2021-08-30 08:51:52 -04:00
MatcherResult result { } ;
2021-08-24 22:58:45 -04:00
// 2. For each element locale of requestedLocales, do
for ( auto const & locale : requested_locales ) {
2023-01-19 11:39:05 -05:00
auto locale_id = TRY_OR_THROW_OOM ( vm , : : Locale : : parse_unicode_locale_id ( locale ) ) ;
2021-08-30 08:51:52 -04:00
VERIFY ( locale_id . has_value ( ) ) ;
2021-08-24 22:58:45 -04:00
// a. Let noExtensionsLocale be the String value that is locale with any Unicode locale extension sequences removed.
2022-09-02 12:01:10 -04:00
auto extensions = locale_id - > remove_extension_type < : : Locale : : LocaleExtension > ( ) ;
2023-01-19 13:13:57 -05:00
auto no_extensions_locale = TRY_OR_THROW_OOM ( vm , locale_id - > to_string ( ) ) ;
2021-08-24 22:58:45 -04:00
2022-03-28 08:30:48 -04:00
// b. Let availableLocale be ! BestAvailableLocale(availableLocales, noExtensionsLocale).
2021-08-24 22:58:45 -04:00
auto available_locale = best_available_locale ( no_extensions_locale ) ;
// c. If availableLocale is not undefined, then
if ( available_locale . has_value ( ) ) {
// i. Set result.[[locale]] to availableLocale.
2023-01-19 13:13:57 -05:00
result . locale = TRY_OR_THROW_OOM ( vm , String : : from_utf8 ( * available_locale ) ) ;
2021-08-24 22:58:45 -04:00
// ii. If locale and noExtensionsLocale are not the same String value, then
2021-08-30 08:51:52 -04:00
if ( locale ! = no_extensions_locale ) {
// 1. Let extension be the String value consisting of the substring of the Unicode locale extension sequence within locale.
// 2. Set result.[[extension]] to extension.
2021-09-05 18:31:22 -04:00
result . extensions . extend ( move ( extensions ) ) ;
2021-08-30 08:51:52 -04:00
}
2021-08-24 22:58:45 -04:00
// iii. Return result.
return result ;
}
}
2022-03-28 08:30:48 -04:00
// 3. Let defLocale be ! DefaultLocale().
2021-08-24 22:58:45 -04:00
// 4. Set result.[[locale]] to defLocale.
2023-01-19 13:13:57 -05:00
result . locale = TRY_OR_THROW_OOM ( vm , String : : from_utf8 ( : : Locale : : default_locale ( ) ) ) ;
2021-08-24 22:58:45 -04:00
// 5. Return result.
return result ;
}
// 9.2.4 BestFitMatcher ( availableLocales, requestedLocales ), https://tc39.es/ecma402/#sec-bestfitmatcher
2023-01-19 13:13:57 -05:00
static ThrowCompletionOr < MatcherResult > best_fit_matcher ( VM & vm , Vector < String > const & requested_locales )
2021-08-24 22:58:45 -04:00
{
// The algorithm is implementation dependent, but should produce results that a typical user of the requested locales would
// perceive as at least as good as those produced by the LookupMatcher abstract operation.
2023-01-19 11:39:05 -05:00
return lookup_matcher ( vm , requested_locales ) ;
2021-08-24 22:58:45 -04:00
}
2021-09-01 22:08:15 -04:00
// 9.2.6 InsertUnicodeExtensionAndCanonicalize ( locale, extension ), https://tc39.es/ecma402/#sec-insert-unicode-extension-and-canonicalize
2023-01-19 13:13:57 -05:00
ThrowCompletionOr < String > insert_unicode_extension_and_canonicalize ( VM & vm , : : Locale : : LocaleID locale , : : Locale : : LocaleExtension extension )
2021-09-01 22:08:15 -04:00
{
// Note: This implementation differs from the spec in how the extension is inserted. The spec assumes
// the input to this method is a string, and is written such that operations are performed on parts
// of that string. LibUnicode gives us the parsed locale in a structure, so we can mutate that
// structure directly.
2023-01-19 11:39:05 -05:00
TRY_OR_THROW_OOM ( vm , locale . extensions . try_append ( move ( extension ) ) ) ;
2023-01-20 14:26:48 -05:00
return MUST_OR_THROW_OOM ( canonicalize_unicode_locale_id ( vm , locale ) ) ;
2021-09-01 22:08:15 -04:00
}
2021-09-10 10:01:41 -04:00
template < typename T >
static auto & find_key_in_value ( T & value , StringView key )
{
2021-11-28 17:55:47 -05:00
if ( key = = " ca " sv )
return value . ca ;
2022-01-29 11:30:46 -05:00
if ( key = = " co " sv )
return value . co ;
2021-11-28 17:55:47 -05:00
if ( key = = " hc " sv )
return value . hc ;
2022-01-29 11:30:46 -05:00
if ( key = = " kf " sv )
return value . kf ;
if ( key = = " kn " sv )
return value . kn ;
2021-09-10 11:04:54 -04:00
if ( key = = " nu " sv )
return value . nu ;
2021-09-10 10:01:41 -04:00
// If you hit this point, you must add any missing keys from [[RelevantExtensionKeys]] to LocaleOptions and LocaleResult.
VERIFY_NOT_REACHED ( ) ;
}
2021-08-24 22:58:45 -04:00
// 9.2.7 ResolveLocale ( availableLocales, requestedLocales, options, relevantExtensionKeys, localeData ), https://tc39.es/ecma402/#sec-resolvelocale
2023-02-05 19:02:54 +00:00
ThrowCompletionOr < LocaleResult > resolve_locale ( VM & vm , Vector < String > const & requested_locales , LocaleOptions const & options , ReadonlySpan < StringView > relevant_extension_keys )
2021-08-24 22:58:45 -04:00
{
// 1. Let matcher be options.[[localeMatcher]].
auto const & matcher = options . locale_matcher ;
2021-08-30 08:51:52 -04:00
MatcherResult matcher_result ;
2021-08-24 22:58:45 -04:00
// 2. If matcher is "lookup", then
2023-01-15 10:31:39 -05:00
if ( matcher . is_string ( ) & & ( TRY ( matcher . as_string ( ) . utf8_string_view ( ) ) = = " lookup " sv ) ) {
2022-03-28 08:30:48 -04:00
// a. Let r be ! LookupMatcher(availableLocales, requestedLocales).
2023-01-20 14:26:48 -05:00
matcher_result = MUST_OR_THROW_OOM ( lookup_matcher ( vm , requested_locales ) ) ;
2021-08-24 22:58:45 -04:00
}
// 3. Else,
else {
2022-03-28 08:30:48 -04:00
// a. Let r be ! BestFitMatcher(availableLocales, requestedLocales).
2023-01-20 14:26:48 -05:00
matcher_result = MUST_OR_THROW_OOM ( best_fit_matcher ( vm , requested_locales ) ) ;
2021-08-24 22:58:45 -04:00
}
2021-08-30 08:51:52 -04:00
2021-08-24 22:58:45 -04:00
// 4. Let foundLocale be r.[[locale]].
2021-08-30 08:51:52 -04:00
auto found_locale = move ( matcher_result . locale ) ;
2021-08-24 22:58:45 -04:00
// 5. Let result be a new Record.
2021-08-30 08:51:52 -04:00
LocaleResult result { } ;
2021-08-24 22:58:45 -04:00
// 6. Set result.[[dataLocale]] to foundLocale.
2021-09-10 10:01:41 -04:00
result . data_locale = found_locale ;
2021-08-24 22:58:45 -04:00
// 7. If r has an [[extension]] field, then
2022-09-02 12:01:10 -04:00
Vector < : : Locale : : Keyword > keywords ;
2021-09-10 10:01:41 -04:00
for ( auto & extension : matcher_result . extensions ) {
2022-09-02 12:01:10 -04:00
if ( ! extension . has < : : Locale : : LocaleExtension > ( ) )
2021-09-10 10:01:41 -04:00
continue ;
// a. Let components be ! UnicodeExtensionComponents(r.[[extension]]).
2022-09-02 12:01:10 -04:00
auto & components = extension . get < : : Locale : : LocaleExtension > ( ) ;
2021-09-10 10:01:41 -04:00
// b. Let keywords be components.[[Keywords]].
keywords = move ( components . keywords ) ;
break ;
}
2021-08-24 22:58:45 -04:00
// 8. Let supportedExtension be "-u".
2022-09-02 12:01:10 -04:00
: : Locale : : LocaleExtension supported_extension { } ;
2021-09-10 10:01:41 -04:00
2021-08-24 22:58:45 -04:00
// 9. For each element key of relevantExtensionKeys, do
2021-09-10 10:01:41 -04:00
for ( auto const & key : relevant_extension_keys ) {
// a. Let foundLocaleData be localeData.[[<foundLocale>]].
// b. Assert: Type(foundLocaleData) is Record.
// c. Let keyLocaleData be foundLocaleData.[[<key>]].
// d. Assert: Type(keyLocaleData) is List.
2022-09-02 12:01:10 -04:00
auto key_locale_data = : : Locale : : get_available_keyword_values ( key ) ;
2021-09-10 10:01:41 -04:00
// e. Let value be keyLocaleData[0].
// f. Assert: Type(value) is either String or Null.
2022-07-14 13:55:13 -04:00
// NOTE: ECMA-402 assumes keyLocaleData is sorted by locale preference. Our list is sorted
// alphabetically, so we get the locale's preferred value from LibUnicode.
2023-01-19 13:13:57 -05:00
Optional < String > value ;
2023-01-27 11:12:01 -05:00
if ( auto preference = TRY_OR_THROW_OOM ( vm , : : Locale : : get_preferred_keyword_value_for_locale ( found_locale , key ) ) ; preference . has_value ( ) )
2023-01-19 13:13:57 -05:00
value = TRY_OR_THROW_OOM ( vm , String : : from_utf8 ( * preference ) ) ;
2021-09-10 10:01:41 -04:00
// g. Let supportedExtensionAddition be "".
2022-09-02 12:01:10 -04:00
Optional < : : Locale : : Keyword > supported_extension_addition { } ;
2021-09-10 10:01:41 -04:00
// h. If r has an [[extension]] field, then
for ( auto & entry : keywords ) {
// i. If keywords contains an element whose [[Key]] is the same as key, then
if ( entry . key ! = key )
continue ;
// 1. Let entry be the element of keywords whose [[Key]] is the same as key.
// 2. Let requestedValue be entry.[[Value]].
auto requested_value = entry . value ;
// 3. If requestedValue is not the empty String, then
if ( ! requested_value . is_empty ( ) ) {
// a. If keyLocaleData contains requestedValue, then
if ( key_locale_data . contains_slow ( requested_value ) ) {
// i. Let value be requestedValue.
2023-01-19 13:13:57 -05:00
value = move ( requested_value ) ;
2021-09-10 10:01:41 -04:00
// ii. Let supportedExtensionAddition be the string-concatenation of "-", key, "-", and value.
2023-01-19 11:39:05 -05:00
supported_extension_addition = : : Locale : : Keyword { TRY_OR_THROW_OOM ( vm , String : : from_utf8 ( key ) ) , move ( entry . value ) } ;
2021-09-10 10:01:41 -04:00
}
}
// 4. Else if keyLocaleData contains "true", then
else if ( key_locale_data . contains_slow ( " true " sv ) ) {
// a. Let value be "true".
2023-08-07 11:12:38 +02:00
value = " true " _string ;
2021-09-10 10:01:41 -04:00
// b. Let supportedExtensionAddition be the string-concatenation of "-" and key.
2023-01-19 11:39:05 -05:00
supported_extension_addition = : : Locale : : Keyword { TRY_OR_THROW_OOM ( vm , String : : from_utf8 ( key ) ) , { } } ;
2021-09-10 10:01:41 -04:00
}
break ;
}
// i. If options has a field [[<key>]], then
// i. Let optionsValue be options.[[<key>]].
// ii. Assert: Type(optionsValue) is either String, Undefined, or Null.
auto options_value = find_key_in_value ( options , key ) ;
// iii. If Type(optionsValue) is String, then
if ( options_value . has_value ( ) ) {
// 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.
// 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.
2023-01-19 13:13:57 -05:00
TRY_OR_THROW_OOM ( vm , : : Locale : : canonicalize_unicode_extension_values ( key , * options_value , true ) ) ;
2021-09-10 10:01:41 -04:00
// 3. If optionsValue is the empty String, then
if ( options_value - > is_empty ( ) ) {
// a. Let optionsValue be "true".
2023-08-07 11:12:38 +02:00
options_value = " true " _string ;
2021-09-10 10:01:41 -04:00
}
}
2023-01-12 10:53:04 -05:00
// iv. If SameValue(optionsValue, value) is false and keyLocaleData contains optionsValue, then
if ( options_value . has_value ( ) & & ( options_value ! = value ) & & key_locale_data . contains_slow ( * options_value ) ) {
// 1. Let value be optionsValue.
value = move ( options_value ) ;
// 2. Let supportedExtensionAddition be "".
supported_extension_addition . clear ( ) ;
2021-09-10 10:01:41 -04:00
}
// j. Set result.[[<key>]] to value.
find_key_in_value ( result , key ) = move ( value ) ;
2023-01-12 10:53:04 -05:00
// k. Set supportedExtension to the string-concatenation of supportedExtension and supportedExtensionAddition.
2021-09-10 10:01:41 -04:00
if ( supported_extension_addition . has_value ( ) )
2023-01-19 11:39:05 -05:00
TRY_OR_THROW_OOM ( vm , supported_extension . keywords . try_append ( supported_extension_addition . release_value ( ) ) ) ;
2021-09-10 10:01:41 -04:00
}
2023-01-12 10:53:04 -05:00
// 10. If supportedExtension is not "-u", then
2021-09-10 10:01:41 -04:00
if ( ! supported_extension . keywords . is_empty ( ) ) {
2023-01-19 11:39:05 -05:00
auto locale_id = TRY_OR_THROW_OOM ( vm , : : Locale : : parse_unicode_locale_id ( found_locale ) ) ;
2021-09-10 10:01:41 -04:00
VERIFY ( locale_id . has_value ( ) ) ;
2023-01-12 10:53:04 -05:00
// a. Set foundLocale to InsertUnicodeExtensionAndCanonicalize(foundLocale, supportedExtension).
2023-01-20 14:26:48 -05:00
found_locale = MUST_OR_THROW_OOM ( insert_unicode_extension_and_canonicalize ( vm , locale_id . release_value ( ) , move ( supported_extension ) ) ) ;
2021-09-10 10:01:41 -04:00
}
2021-08-30 08:51:52 -04:00
2021-08-24 22:58:45 -04:00
// 11. Set result.[[locale]] to foundLocale.
2021-08-30 08:51:52 -04:00
result . locale = move ( found_locale ) ;
2021-08-24 22:58:45 -04:00
// 12. Return result.
return result ;
}
2021-09-04 17:08:57 +01:00
// 9.2.8 LookupSupportedLocales ( availableLocales, requestedLocales ), https://tc39.es/ecma402/#sec-lookupsupportedlocales
2023-01-19 13:13:57 -05:00
static ThrowCompletionOr < Vector < String > > lookup_supported_locales ( VM & vm , Vector < String > const & requested_locales )
2021-09-04 17:08:57 +01:00
{
// 1. Let subset be a new empty List.
2023-01-19 13:13:57 -05:00
Vector < String > subset ;
2021-09-04 17:08:57 +01:00
// 2. For each element locale of requestedLocales, do
for ( auto const & locale : requested_locales ) {
2023-01-19 11:39:05 -05:00
auto locale_id = TRY_OR_THROW_OOM ( vm , : : Locale : : parse_unicode_locale_id ( locale ) ) ;
2021-09-04 17:08:57 +01:00
VERIFY ( locale_id . has_value ( ) ) ;
// a. Let noExtensionsLocale be the String value that is locale with any Unicode locale extension sequences removed.
2022-09-02 12:01:10 -04:00
locale_id - > remove_extension_type < : : Locale : : LocaleExtension > ( ) ;
2023-01-19 13:13:57 -05:00
auto no_extensions_locale = TRY_OR_THROW_OOM ( vm , locale_id - > to_string ( ) ) ;
2021-09-04 17:08:57 +01:00
2022-03-28 08:30:48 -04:00
// b. Let availableLocale be ! BestAvailableLocale(availableLocales, noExtensionsLocale).
2021-09-04 17:08:57 +01:00
auto available_locale = best_available_locale ( no_extensions_locale ) ;
// c. If availableLocale is not undefined, append locale to the end of subset.
if ( available_locale . has_value ( ) )
2023-01-19 11:39:05 -05:00
TRY_OR_THROW_OOM ( vm , subset . try_append ( locale ) ) ;
2021-09-04 17:08:57 +01:00
}
// 3. Return subset.
return subset ;
}
// 9.2.9 BestFitSupportedLocales ( availableLocales, requestedLocales ), https://tc39.es/ecma402/#sec-bestfitsupportedlocales
2023-01-19 13:13:57 -05:00
static ThrowCompletionOr < Vector < String > > best_fit_supported_locales ( VM & vm , Vector < String > const & requested_locales )
2021-09-04 17:08:57 +01:00
{
// The BestFitSupportedLocales abstract operation returns the subset of the provided BCP 47
// language priority list requestedLocales for which availableLocales has a matching locale
// when using the Best Fit Matcher algorithm. Locales appear in the same order in the returned
// list as in requestedLocales. The steps taken are implementation dependent.
// :yakbrain:
2023-01-19 11:39:05 -05:00
return lookup_supported_locales ( vm , requested_locales ) ;
2021-09-04 17:08:57 +01:00
}
// 9.2.10 SupportedLocales ( availableLocales, requestedLocales, options ), https://tc39.es/ecma402/#sec-supportedlocales
2023-01-19 13:13:57 -05:00
ThrowCompletionOr < Array * > supported_locales ( VM & vm , Vector < String > const & requested_locales , Value options )
2021-09-04 17:08:57 +01:00
{
2022-08-20 08:25:24 +01:00
auto & realm = * vm . current_realm ( ) ;
2021-09-04 17:08:57 +01:00
// 1. Set options to ? CoerceOptionsToObject(options).
2022-08-20 08:25:24 +01:00
auto * options_object = TRY ( coerce_options_to_object ( vm , options ) ) ;
2021-09-04 17:08:57 +01:00
2023-01-12 10:22:37 -05:00
// 2. Let matcher be ? GetOption(options, "localeMatcher", string, « "lookup", "best fit" », "best fit").
2022-08-20 08:52:42 +01:00
auto matcher = TRY ( get_option ( vm , * options_object , vm . names . localeMatcher , OptionType : : String , { " lookup " sv , " best fit " sv } , " best fit " sv ) ) ;
2021-09-04 17:08:57 +01:00
2023-01-19 13:13:57 -05:00
Vector < String > supported_locales ;
2021-09-04 17:08:57 +01:00
// 3. If matcher is "best fit", then
2023-01-15 10:31:39 -05:00
if ( TRY ( matcher . as_string ( ) . utf8_string_view ( ) ) = = " best fit " sv ) {
2021-09-04 17:08:57 +01:00
// a. Let supportedLocales be BestFitSupportedLocales(availableLocales, requestedLocales).
2023-01-19 11:39:05 -05:00
supported_locales = TRY ( best_fit_supported_locales ( vm , requested_locales ) ) ;
2021-09-04 17:08:57 +01:00
}
// 4. Else,
else {
// a. Let supportedLocales be LookupSupportedLocales(availableLocales, requestedLocales).
2023-01-19 11:39:05 -05:00
supported_locales = TRY ( lookup_supported_locales ( vm , requested_locales ) ) ;
2021-09-04 17:08:57 +01:00
}
2022-05-02 20:54:39 +02:00
// 5. Return CreateArrayFromList(supportedLocales).
2023-01-19 13:13:57 -05:00
return Array : : create_from < String > ( realm , supported_locales , [ & vm ] ( auto & locale ) { return PrimitiveString : : create ( vm , move ( locale ) ) ; } ) . ptr ( ) ;
2021-09-04 17:08:57 +01:00
}
2021-09-01 22:08:15 -04:00
// 9.2.12 CoerceOptionsToObject ( options ), https://tc39.es/ecma402/#sec-coerceoptionstoobject
2022-08-20 08:25:24 +01:00
ThrowCompletionOr < Object * > coerce_options_to_object ( VM & vm , Value options )
2021-09-01 22:08:15 -04:00
{
2022-08-20 08:25:24 +01:00
auto & realm = * vm . current_realm ( ) ;
2022-08-16 00:20:49 +01:00
2021-09-01 22:08:15 -04:00
// 1. If options is undefined, then
if ( options . is_undefined ( ) ) {
2022-05-02 20:54:39 +02:00
// a. Return OrdinaryObjectCreate(null).
2022-12-13 20:49:50 +00:00
return Object : : create ( realm , nullptr ) . ptr ( ) ;
2021-09-01 22:08:15 -04:00
}
// 2. Return ? ToObject(options).
2023-04-13 15:26:41 +02:00
return TRY ( options . to_object ( vm ) ) . ptr ( ) ;
2021-09-01 22:08:15 -04:00
}
2022-06-14 23:03:25 +01:00
// NOTE: 9.2.13 GetOption has been removed and is being pulled in from ECMA-262 in the Temporal proposal.
2021-08-24 22:58:45 -04:00
2023-04-11 08:45:51 -04:00
// 9.2.14 GetBooleanOrStringNumberFormatOption ( options, property, stringValues, fallback ), https://tc39.es/ecma402/#sec-getbooleanorstringnumberformatoption
2023-02-05 19:02:54 +00:00
ThrowCompletionOr < StringOrBoolean > get_boolean_or_string_number_format_option ( VM & vm , Object const & options , PropertyKey const & property , ReadonlySpan < StringView > string_values , StringOrBoolean fallback )
2022-07-12 13:18:23 -04:00
{
// 1. Let value be ? Get(options, property).
auto value = TRY ( options . get ( property ) ) ;
// 2. If value is undefined, return fallback.
if ( value . is_undefined ( ) )
return fallback ;
2023-01-30 10:36:47 -05:00
// 3. If value is true, return true.
2022-07-12 13:18:23 -04:00
if ( value . is_boolean ( ) & & value . as_bool ( ) )
2023-01-30 10:36:47 -05:00
return StringOrBoolean { true } ;
2022-07-12 13:18:23 -04:00
2023-01-30 10:36:47 -05:00
// 4. If ToBoolean(value) is false, return false.
if ( ! value . to_boolean ( ) )
return StringOrBoolean { false } ;
2022-07-12 13:18:23 -04:00
2023-01-30 10:36:47 -05:00
// 5. Let value be ? ToString(value).
2023-01-15 10:31:39 -05:00
auto value_string = TRY ( value . to_string ( vm ) ) ;
2022-07-12 13:18:23 -04:00
2023-01-30 10:36:47 -05:00
// 6. If stringValues does not contain value, throw a RangeError exception.
auto it = find ( string_values . begin ( ) , string_values . end ( ) , value_string . bytes_as_string_view ( ) ) ;
if ( it = = string_values . end ( ) )
2022-09-13 11:00:58 -04:00
return vm . throw_completion < RangeError > ( ErrorType : : OptionIsNotValidValue , value_string , property . as_string ( ) ) ;
2022-07-12 13:18:23 -04:00
2023-01-30 10:36:47 -05:00
// 7. Return value.
2022-07-12 13:18:23 -04:00
return StringOrBoolean { * it } ;
}
2023-04-11 08:45:51 -04:00
// 9.2.15 DefaultNumberOption ( value, minimum, maximum, fallback ), https://tc39.es/ecma402/#sec-defaultnumberoption
2022-08-20 08:25:24 +01:00
ThrowCompletionOr < Optional < int > > default_number_option ( VM & vm , Value value , int minimum , int maximum , Optional < int > fallback )
2021-09-10 11:04:54 -04:00
{
// 1. If value is undefined, return fallback.
if ( value . is_undefined ( ) )
return fallback ;
2021-11-24 08:51:03 -05:00
// 2. Set value to ? ToNumber(value).
2022-08-21 14:00:56 +01:00
value = TRY ( value . to_number ( vm ) ) ;
2021-09-10 11:04:54 -04:00
// 3. If value is NaN or less than minimum or greater than maximum, throw a RangeError exception.
2021-09-18 19:37:39 +03:00
if ( value . is_nan ( ) | | ( value . as_double ( ) < minimum ) | | ( value . as_double ( ) > maximum ) )
2022-08-16 20:33:17 +01:00
return vm . throw_completion < RangeError > ( ErrorType : : IntlNumberIsNaNOrOutOfRange , value , minimum , maximum ) ;
2021-09-10 11:04:54 -04:00
// 4. Return floor(value).
2021-10-20 19:17:45 +01:00
return floor ( value . as_double ( ) ) ;
2021-09-10 11:04:54 -04:00
}
2023-04-11 08:45:51 -04:00
// 9.2.16 GetNumberOption ( options, property, minimum, maximum, fallback ), https://tc39.es/ecma402/#sec-getnumberoption
2022-08-20 08:25:24 +01:00
ThrowCompletionOr < Optional < int > > get_number_option ( VM & vm , Object const & options , PropertyKey const & property , int minimum , int maximum , Optional < int > fallback )
2021-09-10 11:04:54 -04:00
{
// 1. Assert: Type(options) is Object.
2021-10-02 23:52:27 +01:00
2021-09-10 11:04:54 -04:00
// 2. Let value be ? Get(options, property).
2021-10-02 23:52:27 +01:00
auto value = TRY ( options . get ( property ) ) ;
2021-09-10 11:04:54 -04:00
// 3. Return ? DefaultNumberOption(value, minimum, maximum, fallback).
2022-08-20 08:25:24 +01:00
return default_number_option ( vm , value , minimum , maximum , move ( fallback ) ) ;
2021-09-10 11:04:54 -04:00
}
2023-04-11 08:45:51 -04:00
// 9.2.17 PartitionPattern ( pattern ), https://tc39.es/ecma402/#sec-partitionpattern
2023-01-19 12:26:37 -05:00
ThrowCompletionOr < Vector < PatternPartition > > partition_pattern ( VM & vm , StringView pattern )
2021-09-06 14:37:23 -04:00
{
// 1. Let result be a new empty List.
Vector < PatternPartition > result ;
2022-05-02 20:54:39 +02:00
// 2. Let beginIndex be StringIndexOf(pattern, "{", 0).
2021-09-06 14:37:23 -04:00
auto begin_index = pattern . find ( ' { ' , 0 ) ;
// 3. Let endIndex be 0.
size_t end_index = 0 ;
// 4. Let nextIndex be 0.
size_t next_index = 0 ;
// 5. Let length be the number of code units in pattern.
// 6. Repeat, while beginIndex is an integer index into pattern,
while ( begin_index . has_value ( ) ) {
2022-05-02 20:54:39 +02:00
// a. Set endIndex to StringIndexOf(pattern, "}", beginIndex).
2021-09-06 14:37:23 -04:00
end_index = pattern . find ( ' } ' , * begin_index ) . value ( ) ;
// b. Assert: endIndex is greater than beginIndex.
VERIFY ( end_index > * begin_index ) ;
// c. If beginIndex is greater than nextIndex, then
if ( * begin_index > next_index ) {
// i. Let literal be a substring of pattern from position nextIndex, inclusive, to position beginIndex, exclusive.
auto literal = pattern . substring_view ( next_index , * begin_index - next_index ) ;
// ii. Append a new Record { [[Type]]: "literal", [[Value]]: literal } as the last element of the list result.
2023-01-21 10:34:07 -05:00
TRY_OR_THROW_OOM ( vm , result . try_append ( { " literal " sv , TRY_OR_THROW_OOM ( vm , String : : from_utf8 ( literal ) ) } ) ) ;
2021-09-06 14:37:23 -04:00
}
// d. Let p be the substring of pattern from position beginIndex, exclusive, to position endIndex, exclusive.
auto partition = pattern . substring_view ( * begin_index + 1 , end_index - * begin_index - 1 ) ;
// e. Append a new Record { [[Type]]: p, [[Value]]: undefined } as the last element of the list result.
2023-01-19 12:26:37 -05:00
TRY_OR_THROW_OOM ( vm , result . try_append ( { partition , { } } ) ) ;
2021-09-06 14:37:23 -04:00
// f. Set nextIndex to endIndex + 1.
next_index = end_index + 1 ;
2022-05-02 20:54:39 +02:00
// g. Set beginIndex to StringIndexOf(pattern, "{", nextIndex).
2021-09-06 14:37:23 -04:00
begin_index = pattern . find ( ' { ' , next_index ) ;
}
// 7. If nextIndex is less than length, then
if ( next_index < pattern . length ( ) ) {
// a. Let literal be the substring of pattern from position nextIndex, inclusive, to position length, exclusive.
auto literal = pattern . substring_view ( next_index ) ;
// b. Append a new Record { [[Type]]: "literal", [[Value]]: literal } as the last element of the list result.
2023-01-21 10:34:07 -05:00
TRY_OR_THROW_OOM ( vm , result . try_append ( { " literal " sv , TRY_OR_THROW_OOM ( vm , String : : from_utf8 ( literal ) ) } ) ) ;
2021-09-06 14:37:23 -04:00
}
// 8. Return result.
return result ;
}
2021-08-24 22:58:45 -04:00
}