3 * Internationalisation code.
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
25 * @defgroup Language Language
28 if ( !defined( 'MEDIAWIKI' ) ) {
29 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
33 if ( function_exists( 'mb_strtoupper' ) ) {
34 mb_internal_encoding( 'UTF-8' );
38 * Internationalisation code
43 * @var LanguageConverter
47 public $mVariants, $mCode, $mLoaded = false;
48 public $mMagicExtensions = array(), $mMagicHookDone = false;
49 private $mHtmlCode = null, $mParentLanguage = false;
51 public $dateFormatStrings = array();
52 public $mExtendedSpecialPageAliases;
54 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
57 * ReplacementArray object caches
59 public $transformData = array();
62 * @var LocalisationCache
64 static public $dataCache;
66 static public $mLangObjCache = array();
68 static public $mWeekdayMsgs = array(
69 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
73 static public $mWeekdayAbbrevMsgs = array(
74 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
77 static public $mMonthMsgs = array(
78 'january', 'february', 'march', 'april', 'may_long', 'june',
79 'july', 'august', 'september', 'october', 'november',
82 static public $mMonthGenMsgs = array(
83 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
84 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
87 static public $mMonthAbbrevMsgs = array(
88 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
89 'sep', 'oct', 'nov', 'dec'
92 static public $mIranianCalendarMonthMsgs = array(
93 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
94 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
95 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
96 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
99 static public $mHebrewCalendarMonthMsgs = array(
100 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
101 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
102 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
103 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
104 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
107 static public $mHebrewCalendarMonthGenMsgs = array(
108 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
109 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
110 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
111 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
112 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
115 static public $mHijriCalendarMonthMsgs = array(
116 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
117 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
118 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
119 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
126 static public $durationIntervals = array(
127 'millennia' => 31556952000,
128 'centuries' => 3155695200,
129 'decades' => 315569520,
130 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
139 * Cache for language fallbacks.
140 * @see Language::getFallbacksIncludingSiteLanguage
144 static private $fallbackLanguageCache = array();
147 * Cache for language names
148 * @var MapCacheLRU|null
150 static private $languageNameCache;
153 * Get a cached or new language object for a given language code
154 * @param string $code
157 static function factory( $code ) {
158 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
160 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
161 $code = $wgDummyLanguageCodes[$code];
164 // get the language object to process
165 $langObj = isset( self
::$mLangObjCache[$code] )
166 ? self
::$mLangObjCache[$code]
167 : self
::newFromCode( $code );
169 // merge the language object in to get it up front in the cache
170 self
::$mLangObjCache = array_merge( array( $code => $langObj ), self
::$mLangObjCache );
171 // get rid of the oldest ones in case we have an overflow
172 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
178 * Create a language object for a given language code
179 * @param string $code
180 * @throws MWException
183 protected static function newFromCode( $code ) {
184 // Protect against path traversal below
185 if ( !Language
::isValidCode( $code )
186 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code )
188 throw new MWException( "Invalid language code \"$code\"" );
191 if ( !Language
::isValidBuiltInCode( $code ) ) {
192 // It's not possible to customise this code with class files, so
193 // just return a Language object. This is to support uselang= hacks.
194 $lang = new Language
;
195 $lang->setCode( $code );
199 // Check if there is a language class for the code
200 $class = self
::classFromCode( $code );
201 self
::preloadLanguageClass( $class );
202 if ( class_exists( $class ) ) {
207 // Keep trying the fallback list until we find an existing class
208 $fallbacks = Language
::getFallbacksFor( $code );
209 foreach ( $fallbacks as $fallbackCode ) {
210 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
211 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
214 $class = self
::classFromCode( $fallbackCode );
215 self
::preloadLanguageClass( $class );
216 if ( class_exists( $class ) ) {
217 $lang = Language
::newFromCode( $fallbackCode );
218 $lang->setCode( $code );
223 throw new MWException( "Invalid fallback sequence for language '$code'" );
227 * Checks whether any localisation is available for that language tag
228 * in MediaWiki (MessagesXx.php exists).
230 * @param string $code Language tag (in lower case)
231 * @return bool Whether language is supported
234 public static function isSupportedLanguage( $code ) {
235 return self
::isValidBuiltInCode( $code )
236 && ( is_readable( self
::getMessagesFileName( $code ) )
237 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
242 * Returns true if a language code string is a well-formed language tag
243 * according to RFC 5646.
244 * This function only checks well-formedness; it doesn't check that
245 * language, script or variant codes actually exist in the repositories.
247 * Based on regexes by Mark Davis of the Unicode Consortium:
248 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
250 * @param string $code
251 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
256 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
259 $alphanum = '[a-z0-9]';
260 $x = 'x'; # private use singleton
261 $singleton = '[a-wy-z]'; # other singleton
262 $s = $lenient ?
'[-_]' : '-';
264 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
265 $script = "$alpha{4}"; # ISO 15924
266 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
267 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
268 $extension = "$singleton(?:$s$alphanum{2,8})+";
269 $privateUse = "$x(?:$s$alphanum{1,8})+";
271 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
272 # Since these are limited, this is safe even later changes to the registry --
273 # the only oddity is that it might change the type of the tag, and thus
274 # the results from the capturing groups.
275 # http://www.iana.org/assignments/language-subtag-registry
277 $grandfathered = "en{$s}GB{$s}oed"
278 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
279 . "|no{$s}(?:bok|nyn)"
280 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
281 . "|zh{$s}min{$s}nan";
283 $variantList = "$variant(?:$s$variant)*";
284 $extensionList = "$extension(?:$s$extension)*";
286 $langtag = "(?:($language)"
289 . "(?:$s$variantList)?"
290 . "(?:$s$extensionList)?"
291 . "(?:$s$privateUse)?)";
293 # The final breakdown, with capturing groups for each of these components
294 # The variants, extensions, grandfathered, and private-use may have interior '-'
296 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
298 return (bool)preg_match( "/$root/", strtolower( $code ) );
302 * Returns true if a language code string is of a valid form, whether or
303 * not it exists. This includes codes which are used solely for
304 * customisation via the MediaWiki namespace.
306 * @param string $code
310 public static function isValidCode( $code ) {
311 static $cache = array();
312 if ( isset( $cache[$code] ) ) {
313 return $cache[$code];
315 // People think language codes are html safe, so enforce it.
316 // Ideally we should only allow a-zA-Z0-9-
317 // but, .+ and other chars are often used for {{int:}} hacks
318 // see bugs 37564, 37587, 36938
320 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
321 && !preg_match( MediaWikiTitleCodec
::getTitleInvalidRegex(), $code );
323 return $cache[$code];
327 * Returns true if a language code is of a valid form for the purposes of
328 * internal customisation of MediaWiki, via Messages*.php or *.json.
330 * @param string $code
332 * @throws MWException
336 public static function isValidBuiltInCode( $code ) {
338 if ( !is_string( $code ) ) {
339 if ( is_object( $code ) ) {
340 $addmsg = " of class " . get_class( $code );
344 $type = gettype( $code );
345 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
348 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
352 * Returns true if a language code is an IETF tag known to MediaWiki.
359 public static function isKnownLanguageTag( $tag ) {
360 static $coreLanguageNames;
362 // Quick escape for invalid input to avoid exceptions down the line
363 // when code tries to process tags which are not valid at all.
364 if ( !self
::isValidBuiltInCode( $tag ) ) {
368 if ( $coreLanguageNames === null ) {
370 include "$IP/languages/Names.php";
373 if ( isset( $coreLanguageNames[$tag] )
374 || self
::fetchLanguageName( $tag, $tag ) !== ''
383 * @param string $code
384 * @return string Name of the language class
386 public static function classFromCode( $code ) {
387 if ( $code == 'en' ) {
390 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
395 * Includes language class files
397 * @param string $class Name of the language class
399 public static function preloadLanguageClass( $class ) {
402 if ( $class === 'Language' ) {
406 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
407 include_once "$IP/languages/classes/$class.php";
412 * Get the LocalisationCache instance
414 * @return LocalisationCache
416 public static function getLocalisationCache() {
417 if ( is_null( self
::$dataCache ) ) {
418 global $wgLocalisationCacheConf;
419 $class = $wgLocalisationCacheConf['class'];
420 self
::$dataCache = new $class( $wgLocalisationCacheConf );
422 return self
::$dataCache;
425 function __construct() {
426 $this->mConverter
= new FakeConverter( $this );
427 // Set the code to the name of the descendant
428 if ( get_class( $this ) == 'Language' ) {
431 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
433 self
::getLocalisationCache();
437 * Reduce memory usage
439 function __destruct() {
440 foreach ( $this as $name => $value ) {
441 unset( $this->$name );
446 * Hook which will be called if this is the content language.
447 * Descendants can use this to register hook functions or modify globals
449 function initContLang() {
456 function getFallbackLanguages() {
457 return self
::getFallbacksFor( $this->mCode
);
461 * Exports $wgBookstoreListEn
464 function getBookstoreList() {
465 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
469 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
470 * available in localised form, it will be included in English.
474 public function getNamespaces() {
475 if ( is_null( $this->namespaceNames
) ) {
476 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
478 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
479 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
481 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
483 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
484 if ( $wgMetaNamespaceTalk ) {
485 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
487 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
488 $this->namespaceNames
[NS_PROJECT_TALK
] =
489 $this->fixVariableInNamespace( $talk );
492 # Sometimes a language will be localised but not actually exist on this wiki.
493 foreach ( $this->namespaceNames
as $key => $text ) {
494 if ( !isset( $validNamespaces[$key] ) ) {
495 unset( $this->namespaceNames
[$key] );
499 # The above mixing may leave namespaces out of canonical order.
500 # Re-order by namespace ID number...
501 ksort( $this->namespaceNames
);
503 Hooks
::run( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
506 return $this->namespaceNames
;
510 * Arbitrarily set all of the namespace names at once. Mainly used for testing
511 * @param array $namespaces Array of namespaces (id => name)
513 public function setNamespaces( array $namespaces ) {
514 $this->namespaceNames
= $namespaces;
515 $this->mNamespaceIds
= null;
519 * Resets all of the namespace caches. Mainly used for testing
521 public function resetNamespaces() {
522 $this->namespaceNames
= null;
523 $this->mNamespaceIds
= null;
524 $this->namespaceAliases
= null;
528 * A convenience function that returns the same thing as
529 * getNamespaces() except with the array values changed to ' '
530 * where it found '_', useful for producing output to be displayed
531 * e.g. in <select> forms.
535 function getFormattedNamespaces() {
536 $ns = $this->getNamespaces();
537 foreach ( $ns as $k => $v ) {
538 $ns[$k] = strtr( $v, '_', ' ' );
544 * Get a namespace value by key
546 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
547 * echo $mw_ns; // prints 'MediaWiki'
550 * @param int $index The array key of the namespace to return
551 * @return string|bool String if the namespace value exists, otherwise false
553 function getNsText( $index ) {
554 $ns = $this->getNamespaces();
556 return isset( $ns[$index] ) ?
$ns[$index] : false;
560 * A convenience function that returns the same thing as
561 * getNsText() except with '_' changed to ' ', useful for
565 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
566 * echo $mw_ns; // prints 'MediaWiki talk'
569 * @param int $index The array key of the namespace to return
570 * @return string Namespace name without underscores (empty string if namespace does not exist)
572 function getFormattedNsText( $index ) {
573 $ns = $this->getNsText( $index );
575 return strtr( $ns, '_', ' ' );
579 * Returns gender-dependent namespace alias if available.
580 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
581 * @param int $index Namespace index
582 * @param string $gender Gender key (male, female... )
586 function getGenderNsText( $index, $gender ) {
587 global $wgExtraGenderNamespaces;
589 $ns = $wgExtraGenderNamespaces +
590 self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
592 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
596 * Whether this language uses gender-dependent namespace aliases.
597 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
601 function needsGenderDistinction() {
602 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
603 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
604 // $wgExtraGenderNamespaces overrides everything
606 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
607 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
608 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
611 // Check what is in i18n files
612 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
613 return count( $aliases ) > 0;
618 * Get a namespace key by value, case insensitive.
619 * Only matches namespace names for the current language, not the
620 * canonical ones defined in Namespace.php.
622 * @param string $text
623 * @return int|bool An integer if $text is a valid value otherwise false
625 function getLocalNsIndex( $text ) {
626 $lctext = $this->lc( $text );
627 $ids = $this->getNamespaceIds();
628 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
634 function getNamespaceAliases() {
635 if ( is_null( $this->namespaceAliases
) ) {
636 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
640 foreach ( $aliases as $name => $index ) {
641 if ( $index === NS_PROJECT_TALK
) {
642 unset( $aliases[$name] );
643 $name = $this->fixVariableInNamespace( $name );
644 $aliases[$name] = $index;
649 global $wgExtraGenderNamespaces;
650 $genders = $wgExtraGenderNamespaces +
651 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
652 foreach ( $genders as $index => $forms ) {
653 foreach ( $forms as $alias ) {
654 $aliases[$alias] = $index;
658 # Also add converted namespace names as aliases, to avoid confusion.
659 $convertedNames = array();
660 foreach ( $this->getVariants() as $variant ) {
661 if ( $variant === $this->mCode
) {
664 foreach ( $this->getNamespaces() as $ns => $_ ) {
665 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
669 $this->namespaceAliases
= $aliases +
$convertedNames;
672 return $this->namespaceAliases
;
678 function getNamespaceIds() {
679 if ( is_null( $this->mNamespaceIds
) ) {
680 global $wgNamespaceAliases;
681 # Put namespace names and aliases into a hashtable.
682 # If this is too slow, then we should arrange it so that it is done
683 # before caching. The catch is that at pre-cache time, the above
684 # class-specific fixup hasn't been done.
685 $this->mNamespaceIds
= array();
686 foreach ( $this->getNamespaces() as $index => $name ) {
687 $this->mNamespaceIds
[$this->lc( $name )] = $index;
689 foreach ( $this->getNamespaceAliases() as $name => $index ) {
690 $this->mNamespaceIds
[$this->lc( $name )] = $index;
692 if ( $wgNamespaceAliases ) {
693 foreach ( $wgNamespaceAliases as $name => $index ) {
694 $this->mNamespaceIds
[$this->lc( $name )] = $index;
698 return $this->mNamespaceIds
;
702 * Get a namespace key by value, case insensitive. Canonical namespace
703 * names override custom ones defined for the current language.
705 * @param string $text
706 * @return int|bool An integer if $text is a valid value otherwise false
708 function getNsIndex( $text ) {
709 $lctext = $this->lc( $text );
710 $ns = MWNamespace
::getCanonicalIndex( $lctext );
711 if ( $ns !== null ) {
714 $ids = $this->getNamespaceIds();
715 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
719 * short names for language variants used for language conversion links.
721 * @param string $code
722 * @param bool $usemsg Use the "variantname-xyz" message if it exists
725 function getVariantname( $code, $usemsg = true ) {
726 $msg = "variantname-$code";
727 if ( $usemsg && wfMessage( $msg )->exists() ) {
728 return $this->getMessageFromDB( $msg );
730 $name = self
::fetchLanguageName( $code );
732 return $name; # if it's defined as a language name, show that
734 # otherwise, output the language code
740 * @deprecated since 1.24, doesn't handle conflicting aliases. Use
741 * SpecialPageFactory::getLocalNameFor instead.
742 * @param string $name
745 function specialPage( $name ) {
746 $aliases = $this->getSpecialPageAliases();
747 if ( isset( $aliases[$name][0] ) ) {
748 $name = $aliases[$name][0];
750 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
756 function getDatePreferences() {
757 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
763 function getDateFormats() {
764 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
768 * @return array|string
770 function getDefaultDateFormat() {
771 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
772 if ( $df === 'dmy or mdy' ) {
773 global $wgAmericanDates;
774 return $wgAmericanDates ?
'mdy' : 'dmy';
783 function getDatePreferenceMigrationMap() {
784 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
788 * @param string $image
791 function getImageFile( $image ) {
792 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
799 function getImageFiles() {
800 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
806 function getExtraUserToggles() {
807 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
814 function getUserToggle( $tog ) {
815 return $this->getMessageFromDB( "tog-$tog" );
819 * Get native language names, indexed by code.
820 * Only those defined in MediaWiki, no other data like CLDR.
821 * If $customisedOnly is true, only returns codes with a messages file
823 * @param bool $customisedOnly
826 * @deprecated since 1.20, use fetchLanguageNames()
828 public static function getLanguageNames( $customisedOnly = false ) {
829 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
833 * Get translated language names. This is done on best effort and
834 * by default this is exactly the same as Language::getLanguageNames.
835 * The CLDR extension provides translated names.
836 * @param string $code Language code.
837 * @return array Language code => language name
839 * @deprecated since 1.20, use fetchLanguageNames()
841 public static function getTranslatedLanguageNames( $code ) {
842 return self
::fetchLanguageNames( $code, 'all' );
846 * Get an array of language names, indexed by code.
847 * @param null|string $inLanguage Code of language in which to return the names
848 * Use null for autonyms (native names)
849 * @param string $include One of:
850 * 'all' all available languages
851 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
852 * 'mwfile' only if the language is in 'mw' *and* has a message file
853 * @return array Language code => language name
856 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
857 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
858 $cacheKey .= ":$include";
859 if ( self
::$languageNameCache === null ) {
860 self
::$languageNameCache = new MapCacheLRU( 20 );
862 if ( self
::$languageNameCache->has( $cacheKey ) ) {
863 $ret = self
::$languageNameCache->get( $cacheKey );
865 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
866 self
::$languageNameCache->set( $cacheKey, $ret );
872 * Uncached helper for fetchLanguageNames
873 * @param null|string $inLanguage Code of language in which to return the names
874 * Use null for autonyms (native names)
875 * @param string $include One of:
876 * 'all' all available languages
877 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
878 * 'mwfile' only if the language is in 'mw' *and* has a message file
879 * @return array Language code => language name
881 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
882 global $wgExtraLanguageNames;
883 static $coreLanguageNames;
885 if ( $coreLanguageNames === null ) {
887 include "$IP/languages/Names.php";
890 // If passed an invalid language code to use, fallback to en
891 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
898 # TODO: also include when $inLanguage is null, when this code is more efficient
899 Hooks
::run( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
902 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
903 foreach ( $mwNames as $mwCode => $mwName ) {
904 # - Prefer own MediaWiki native name when not using the hook
905 # - For other names just add if not added through the hook
906 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
907 $names[$mwCode] = $mwName;
911 if ( $include === 'all' ) {
917 $coreCodes = array_keys( $mwNames );
918 foreach ( $coreCodes as $coreCode ) {
919 $returnMw[$coreCode] = $names[$coreCode];
922 if ( $include === 'mwfile' ) {
923 $namesMwFile = array();
924 # We do this using a foreach over the codes instead of a directory
925 # loop so that messages files in extensions will work correctly.
926 foreach ( $returnMw as $code => $value ) {
927 if ( is_readable( self
::getMessagesFileName( $code ) )
928 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
930 $namesMwFile[$code] = $names[$code];
934 ksort( $namesMwFile );
939 # 'mw' option; default if it's not one of the other two options (all/mwfile)
944 * @param string $code The code of the language for which to get the name
945 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
946 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
947 * @return string Language name or empty
950 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
951 $code = strtolower( $code );
952 $array = self
::fetchLanguageNames( $inLanguage, $include );
953 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
957 * Get a message from the MediaWiki namespace.
959 * @param string $msg Message name
962 function getMessageFromDB( $msg ) {
963 return $this->msg( $msg )->text();
967 * Get message object in this language. Only for use inside this class.
969 * @param string $msg Message name
972 protected function msg( $msg ) {
973 return wfMessage( $msg )->inLanguage( $this );
977 * Get the native language name of $code.
978 * Only if defined in MediaWiki, no other data like CLDR.
979 * @param string $code
981 * @deprecated since 1.20, use fetchLanguageName()
983 function getLanguageName( $code ) {
984 return self
::fetchLanguageName( $code );
991 function getMonthName( $key ) {
992 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
998 function getMonthNamesArray() {
999 $monthNames = array( '' );
1000 for ( $i = 1; $i < 13; $i++
) {
1001 $monthNames[] = $this->getMonthName( $i );
1007 * @param string $key
1010 function getMonthNameGen( $key ) {
1011 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1015 * @param string $key
1018 function getMonthAbbreviation( $key ) {
1019 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1025 function getMonthAbbreviationsArray() {
1026 $monthNames = array( '' );
1027 for ( $i = 1; $i < 13; $i++
) {
1028 $monthNames[] = $this->getMonthAbbreviation( $i );
1034 * @param string $key
1037 function getWeekdayName( $key ) {
1038 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1042 * @param string $key
1045 function getWeekdayAbbreviation( $key ) {
1046 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1050 * @param string $key
1053 function getIranianCalendarMonthName( $key ) {
1054 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1058 * @param string $key
1061 function getHebrewCalendarMonthName( $key ) {
1062 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1066 * @param string $key
1069 function getHebrewCalendarMonthNameGen( $key ) {
1070 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1074 * @param string $key
1077 function getHijriCalendarMonthName( $key ) {
1078 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1082 * Pass through result from $dateTimeObj->format()
1083 * @param DateTime|bool|null &$dateTimeObj
1085 * @param DateTimeZone|bool|null $zone
1086 * @param string $code
1089 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1090 if ( !$dateTimeObj ) {
1091 $dateTimeObj = DateTime
::createFromFormat(
1092 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1095 return $dateTimeObj->format( $code );
1099 * This is a workalike of PHP's date() function, but with better
1100 * internationalisation, a reduced set of format characters, and a better
1103 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1104 * the PHP manual for definitions. There are a number of extensions, which
1107 * xn Do not translate digits of the next numeric format character
1108 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1109 * xr Use roman numerals for the next numeric format character
1110 * xh Use hebrew numerals for the next numeric format character
1112 * xg Genitive month name
1114 * xij j (day number) in Iranian calendar
1115 * xiF F (month name) in Iranian calendar
1116 * xin n (month number) in Iranian calendar
1117 * xiy y (two digit year) in Iranian calendar
1118 * xiY Y (full year) in Iranian calendar
1120 * xjj j (day number) in Hebrew calendar
1121 * xjF F (month name) in Hebrew calendar
1122 * xjt t (days in month) in Hebrew calendar
1123 * xjx xg (genitive month name) in Hebrew calendar
1124 * xjn n (month number) in Hebrew calendar
1125 * xjY Y (full year) in Hebrew calendar
1127 * xmj j (day number) in Hijri calendar
1128 * xmF F (month name) in Hijri calendar
1129 * xmn n (month number) in Hijri calendar
1130 * xmY Y (full year) in Hijri calendar
1132 * xkY Y (full year) in Thai solar calendar. Months and days are
1133 * identical to the Gregorian calendar
1134 * xoY Y (full year) in Minguo calendar or Juche year.
1135 * Months and days are identical to the
1136 * Gregorian calendar
1137 * xtY Y (full year) in Japanese nengo. Months and days are
1138 * identical to the Gregorian calendar
1140 * Characters enclosed in double quotes will be considered literal (with
1141 * the quotes themselves removed). Unmatched quotes will be considered
1142 * literal quotes. Example:
1144 * "The month is" F => The month is January
1147 * Backslash escaping is also supported.
1149 * Input timestamp is assumed to be pre-normalized to the desired local
1150 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1151 * $ts is UTC if $zone is not given.
1153 * @param string $format
1154 * @param string $ts 14-character timestamp
1157 * @param DateTimeZone $zone Timezone of $ts
1158 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1159 * Only makes sense if $ts is the current time.
1160 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1162 * @throws MWException
1165 function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = null ) {
1170 $dateTimeObj = false;
1179 $usedSecond = false;
1180 $usedMinute = false;
1187 $usedISOYear = false;
1188 $usedIsLeapYear = false;
1190 $usedHebrewMonth = false;
1191 $usedIranianMonth = false;
1192 $usedHijriMonth = false;
1193 $usedHebrewYear = false;
1194 $usedIranianYear = false;
1195 $usedHijriYear = false;
1196 $usedTennoYear = false;
1198 if ( strlen( $ts ) !== 14 ) {
1199 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1202 if ( !ctype_digit( $ts ) ) {
1203 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1206 $formatLength = strlen( $format );
1207 for ( $p = 0; $p < $formatLength; $p++
) {
1209 $code = $format[$p];
1210 if ( $code == 'x' && $p < $formatLength - 1 ) {
1211 $code .= $format[++
$p];
1214 if ( ( $code === 'xi'
1220 && $p < $formatLength - 1 ) {
1221 $code .= $format[++
$p];
1232 $rawToggle = !$rawToggle;
1242 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1245 $usedHebrewMonth = true;
1247 $hebrew = self
::tsToHebrew( $ts );
1249 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1253 $num = substr( $ts, 6, 2 );
1257 $s .= $this->getWeekdayAbbreviation( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1 );
1261 $num = intval( substr( $ts, 6, 2 ) );
1266 $iranian = self
::tsToIranian( $ts );
1273 $hijri = self
::tsToHijri( $ts );
1280 $hebrew = self
::tsToHebrew( $ts );
1286 $s .= $this->getWeekdayName( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1 );
1290 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1293 $usedIranianMonth = true;
1295 $iranian = self
::tsToIranian( $ts );
1297 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1300 $usedHijriMonth = true;
1302 $hijri = self
::tsToHijri( $ts );
1304 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1307 $usedHebrewMonth = true;
1309 $hebrew = self
::tsToHebrew( $ts );
1311 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1315 $num = substr( $ts, 4, 2 );
1319 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1323 $num = intval( substr( $ts, 4, 2 ) );
1326 $usedIranianMonth = true;
1328 $iranian = self
::tsToIranian( $ts );
1333 $usedHijriMonth = true;
1335 $hijri = self
::tsToHijri ( $ts );
1340 $usedHebrewMonth = true;
1342 $hebrew = self
::tsToHebrew( $ts );
1347 $usedHebrewMonth = true;
1349 $hebrew = self
::tsToHebrew( $ts );
1355 $num = substr( $ts, 0, 4 );
1358 $usedIranianYear = true;
1360 $iranian = self
::tsToIranian( $ts );
1365 $usedHijriYear = true;
1367 $hijri = self
::tsToHijri( $ts );
1372 $usedHebrewYear = true;
1374 $hebrew = self
::tsToHebrew( $ts );
1381 $thai = self
::tsToYear( $ts, 'thai' );
1388 $minguo = self
::tsToYear( $ts, 'minguo' );
1393 $usedTennoYear = true;
1395 $tenno = self
::tsToYear( $ts, 'tenno' );
1401 $num = substr( $ts, 2, 2 );
1404 $usedIranianYear = true;
1406 $iranian = self
::tsToIranian( $ts );
1408 $num = substr( $iranian[0], -2 );
1412 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1416 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1420 $h = substr( $ts, 8, 2 );
1421 $num = $h %
12 ?
$h %
12 : 12;
1425 $num = intval( substr( $ts, 8, 2 ) );
1429 $h = substr( $ts, 8, 2 );
1430 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1434 $num = substr( $ts, 8, 2 );
1438 $num = substr( $ts, 10, 2 );
1442 $num = substr( $ts, 12, 2 );
1452 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1458 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1462 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1466 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1469 $usedIsLeapYear = true;
1470 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1473 $usedISOYear = true;
1474 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1481 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1484 # Backslash escaping
1485 if ( $p < $formatLength - 1 ) {
1486 $s .= $format[++
$p];
1493 if ( $p < $formatLength - 1 ) {
1494 $endQuote = strpos( $format, '"', $p +
1 );
1495 if ( $endQuote === false ) {
1496 # No terminating quote, assume literal "
1499 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1503 # Quote at end of string, assume literal "
1510 if ( $num !== false ) {
1511 if ( $rawToggle ||
$raw ) {
1514 } elseif ( $roman ) {
1515 $s .= Language
::romanNumeral( $num );
1517 } elseif ( $hebrewNum ) {
1518 $s .= self
::hebrewNumeral( $num );
1521 $s .= $this->formatNum( $num, true );
1526 if ( $usedSecond ) {
1528 } elseif ( $usedMinute ) {
1529 $ttl = 60 - substr( $ts, 12, 2 );
1530 } elseif ( $usedHour ) {
1531 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1532 } elseif ( $usedAMPM ) {
1533 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1534 } elseif ( $usedDay ||
$usedHebrewMonth ||
$usedIranianMonth ||
$usedHijriMonth ||
$usedHebrewYear ||
$usedIranianYear ||
$usedHijriYear ||
$usedTennoYear ) {
1535 // @todo Someone who understands the non-Gregorian calendars should write proper logic for them
1536 // so that they don't need purged every day.
1537 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1539 $possibleTtls = array();
1540 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1542 $possibleTtls[] = ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
$timeRemainingInDay;
1543 } elseif ( $usedISOYear ) {
1544 // December 28th falls on the last ISO week of the year, every year.
1545 // The last ISO week of a year can be 52 or 53.
1546 $lastWeekOfISOYear = DateTime
::createFromFormat( 'Ymd', substr( $ts, 0, 4 ) . '1228', $zone ?
: new DateTimeZone( 'UTC' ) )->format( 'W' );
1547 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1548 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1549 $timeRemainingInWeek = ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
$timeRemainingInDay;
1550 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1554 $possibleTtls[] = ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) - substr( $ts, 6, 2 ) ) * 86400 +
$timeRemainingInDay;
1555 } elseif ( $usedYear ) {
1556 $possibleTtls[] = ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1557 +
$timeRemainingInDay;
1558 } elseif ( $usedIsLeapYear ) {
1559 $year = substr( $ts, 0, 4 );
1560 $timeRemainingInYear = ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1561 +
$timeRemainingInDay;
1563 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1564 // this isn't a leap year. see when the next one starts
1565 $nextCandidate = $year - $mod +
4;
1566 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1567 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
$timeRemainingInYear;
1569 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
$timeRemainingInYear;
1572 // this is a leap year, so the next year isn't
1573 $possibleTtls[] = $timeRemainingInYear;
1577 if ( $possibleTtls ) {
1578 $ttl = min( $possibleTtls );
1585 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1586 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1589 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1590 * Gregorian dates to Iranian dates. Originally written in C, it
1591 * is released under the terms of GNU Lesser General Public
1592 * License. Conversion to PHP was performed by Niklas Laxström.
1594 * Link: http://www.farsiweb.info/jalali/jalali.c
1600 private static function tsToIranian( $ts ) {
1601 $gy = substr( $ts, 0, 4 ) -1600;
1602 $gm = substr( $ts, 4, 2 ) -1;
1603 $gd = substr( $ts, 6, 2 ) -1;
1605 # Days passed from the beginning (including leap years)
1607 +
floor( ( $gy +
3 ) / 4 )
1608 - floor( ( $gy +
99 ) / 100 )
1609 +
floor( ( $gy +
399 ) / 400 );
1611 // Add days of the past months of this year
1612 for ( $i = 0; $i < $gm; $i++
) {
1613 $gDayNo +
= self
::$GREG_DAYS[$i];
1617 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1621 // Days passed in current month
1622 $gDayNo +
= (int)$gd;
1624 $jDayNo = $gDayNo - 79;
1626 $jNp = floor( $jDayNo / 12053 );
1629 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1632 if ( $jDayNo >= 366 ) {
1633 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1634 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1637 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1638 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1644 return array( $jy, $jm, $jd );
1648 * Converting Gregorian dates to Hijri dates.
1650 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1652 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1658 private static function tsToHijri( $ts ) {
1659 $year = substr( $ts, 0, 4 );
1660 $month = substr( $ts, 4, 2 );
1661 $day = substr( $ts, 6, 2 );
1669 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1670 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1672 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1673 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1674 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1677 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1678 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1681 $zl = $zjd -1948440 +
10632;
1682 $zn = (int)( ( $zl - 1 ) / 10631 );
1683 $zl = $zl - 10631 * $zn +
354;
1684 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1685 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1686 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1687 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1688 $zm = (int)( ( 24 * $zl ) / 709 );
1689 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1690 $zy = 30 * $zn +
$zj - 30;
1692 return array( $zy, $zm, $zd );
1696 * Converting Gregorian dates to Hebrew dates.
1698 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1699 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1700 * to translate the relevant functions into PHP and release them under
1703 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1704 * and Adar II is 14. In a non-leap year, Adar is 6.
1710 private static function tsToHebrew( $ts ) {
1712 $year = substr( $ts, 0, 4 );
1713 $month = substr( $ts, 4, 2 );
1714 $day = substr( $ts, 6, 2 );
1716 # Calculate Hebrew year
1717 $hebrewYear = $year +
3760;
1719 # Month number when September = 1, August = 12
1721 if ( $month > 12 ) {
1728 # Calculate day of year from 1 September
1730 for ( $i = 1; $i < $month; $i++
) {
1734 # Check if the year is leap
1735 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1738 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1745 # Calculate the start of the Hebrew year
1746 $start = self
::hebrewYearStart( $hebrewYear );
1748 # Calculate next year's start
1749 if ( $dayOfYear <= $start ) {
1750 # Day is before the start of the year - it is the previous year
1752 $nextStart = $start;
1756 # Add days since previous year's 1 September
1758 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1762 # Start of the new (previous) year
1763 $start = self
::hebrewYearStart( $hebrewYear );
1766 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1769 # Calculate Hebrew day of year
1770 $hebrewDayOfYear = $dayOfYear - $start;
1772 # Difference between year's days
1773 $diff = $nextStart - $start;
1774 # Add 12 (or 13 for leap years) days to ignore the difference between
1775 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1776 # difference is only about the year type
1777 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1783 # Check the year pattern, and is leap year
1784 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1785 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1786 # and non-leap years
1787 $yearPattern = $diff %
30;
1788 # Check if leap year
1789 $isLeap = $diff >= 30;
1791 # Calculate day in the month from number of day in the Hebrew year
1792 # Don't check Adar - if the day is not in Adar, we will stop before;
1793 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1794 $hebrewDay = $hebrewDayOfYear;
1797 while ( $hebrewMonth <= 12 ) {
1798 # Calculate days in this month
1799 if ( $isLeap && $hebrewMonth == 6 ) {
1800 # Adar in a leap year
1802 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1804 if ( $hebrewDay <= $days ) {
1808 # Subtract the days of Adar I
1809 $hebrewDay -= $days;
1812 if ( $hebrewDay <= $days ) {
1818 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1819 # Cheshvan in a complete year (otherwise as the rule below)
1821 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1822 # Kislev in an incomplete year (otherwise as the rule below)
1825 # Odd months have 30 days, even have 29
1826 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1828 if ( $hebrewDay <= $days ) {
1829 # In the current month
1832 # Subtract the days of the current month
1833 $hebrewDay -= $days;
1834 # Try in the next month
1839 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1843 * This calculates the Hebrew year start, as days since 1 September.
1844 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1845 * Used for Hebrew date.
1851 private static function hebrewYearStart( $year ) {
1852 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1853 $b = intval( ( $year - 1 ) %
4 );
1854 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1858 $Mar = intval( $m );
1864 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1865 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1867 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1869 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1873 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1878 * Algorithm to convert Gregorian dates to Thai solar dates,
1879 * Minguo dates or Minguo dates.
1881 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1882 * http://en.wikipedia.org/wiki/Minguo_calendar
1883 * http://en.wikipedia.org/wiki/Japanese_era_name
1885 * @param string $ts 14-character timestamp
1886 * @param string $cName Calender name
1887 * @return array Converted year, month, day
1889 private static function tsToYear( $ts, $cName ) {
1890 $gy = substr( $ts, 0, 4 );
1891 $gm = substr( $ts, 4, 2 );
1892 $gd = substr( $ts, 6, 2 );
1894 if ( !strcmp( $cName, 'thai' ) ) {
1896 # Add 543 years to the Gregorian calendar
1897 # Months and days are identical
1898 $gy_offset = $gy +
543;
1899 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1901 # Deduct 1911 years from the Gregorian calendar
1902 # Months and days are identical
1903 $gy_offset = $gy - 1911;
1904 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1905 # Nengō dates up to Meiji period
1906 # Deduct years from the Gregorian calendar
1907 # depending on the nengo periods
1908 # Months and days are identical
1910 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1911 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1914 $gy_gannen = $gy - 1868 +
1;
1915 $gy_offset = $gy_gannen;
1916 if ( $gy_gannen == 1 ) {
1919 $gy_offset = '明治' . $gy_offset;
1921 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1922 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1923 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1924 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1925 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1928 $gy_gannen = $gy - 1912 +
1;
1929 $gy_offset = $gy_gannen;
1930 if ( $gy_gannen == 1 ) {
1933 $gy_offset = '大正' . $gy_offset;
1935 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1936 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1937 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1940 $gy_gannen = $gy - 1926 +
1;
1941 $gy_offset = $gy_gannen;
1942 if ( $gy_gannen == 1 ) {
1945 $gy_offset = '昭和' . $gy_offset;
1948 $gy_gannen = $gy - 1989 +
1;
1949 $gy_offset = $gy_gannen;
1950 if ( $gy_gannen == 1 ) {
1953 $gy_offset = '平成' . $gy_offset;
1959 return array( $gy_offset, $gm, $gd );
1963 * Roman number formatting up to 10000
1969 static function romanNumeral( $num ) {
1970 static $table = array(
1971 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1972 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1973 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1974 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1975 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1978 $num = intval( $num );
1979 if ( $num > 10000 ||
$num <= 0 ) {
1984 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1985 if ( $num >= $pow10 ) {
1986 $s .= $table[$i][(int)floor( $num / $pow10 )];
1988 $num = $num %
$pow10;
1994 * Hebrew Gematria number formatting up to 9999
2000 static function hebrewNumeral( $num ) {
2001 static $table = array(
2002 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
2003 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
2004 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
2005 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
2008 $num = intval( $num );
2009 if ( $num > 9999 ||
$num <= 0 ) {
2014 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2015 if ( $num >= $pow10 ) {
2016 if ( $num == 15 ||
$num == 16 ) {
2017 $s .= $table[0][9] . $table[0][$num - 9];
2020 $s .= $table[$i][intval( ( $num / $pow10 ) )];
2021 if ( $pow10 == 1000 ) {
2026 $num = $num %
$pow10;
2028 if ( strlen( $s ) == 2 ) {
2031 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
2032 $str .= substr( $s, strlen( $s ) - 2, 2 );
2034 $start = substr( $str, 0, strlen( $str ) - 2 );
2035 $end = substr( $str, strlen( $str ) - 2 );
2038 $str = $start . 'ך';
2041 $str = $start . 'ם';
2044 $str = $start . 'ן';
2047 $str = $start . 'ף';
2050 $str = $start . 'ץ';
2057 * Used by date() and time() to adjust the time output.
2059 * @param string $ts The time in date('YmdHis') format
2060 * @param mixed $tz Adjust the time by this amount (default false, mean we
2061 * get user timecorrection setting)
2064 function userAdjust( $ts, $tz = false ) {
2065 global $wgUser, $wgLocalTZoffset;
2067 if ( $tz === false ) {
2068 $tz = $wgUser->getOption( 'timecorrection' );
2071 $data = explode( '|', $tz, 3 );
2073 if ( $data[0] == 'ZoneInfo' ) {
2074 wfSuppressWarnings();
2075 $userTZ = timezone_open( $data[2] );
2076 wfRestoreWarnings();
2077 if ( $userTZ !== false ) {
2078 $date = date_create( $ts, timezone_open( 'UTC' ) );
2079 date_timezone_set( $date, $userTZ );
2080 $date = date_format( $date, 'YmdHis' );
2083 # Unrecognized timezone, default to 'Offset' with the stored offset.
2084 $data[0] = 'Offset';
2087 if ( $data[0] == 'System' ||
$tz == '' ) {
2088 # Global offset in minutes.
2089 $minDiff = $wgLocalTZoffset;
2090 } elseif ( $data[0] == 'Offset' ) {
2091 $minDiff = intval( $data[1] );
2093 $data = explode( ':', $tz );
2094 if ( count( $data ) == 2 ) {
2095 $data[0] = intval( $data[0] );
2096 $data[1] = intval( $data[1] );
2097 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2098 if ( $data[0] < 0 ) {
2099 $minDiff = -$minDiff;
2102 $minDiff = intval( $data[0] ) * 60;
2106 # No difference ? Return time unchanged
2107 if ( 0 == $minDiff ) {
2111 wfSuppressWarnings(); // E_STRICT system time bitching
2112 # Generate an adjusted date; take advantage of the fact that mktime
2113 # will normalize out-of-range values so we don't have to split $minDiff
2114 # into hours and minutes.
2116 (int)substr( $ts, 8, 2 ) ), # Hours
2117 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2118 (int)substr( $ts, 12, 2 ), # Seconds
2119 (int)substr( $ts, 4, 2 ), # Month
2120 (int)substr( $ts, 6, 2 ), # Day
2121 (int)substr( $ts, 0, 4 ) ); # Year
2123 $date = date( 'YmdHis', $t );
2124 wfRestoreWarnings();
2130 * This is meant to be used by time(), date(), and timeanddate() to get
2131 * the date preference they're supposed to use, it should be used in
2135 * function timeanddate([...], $format = true) {
2136 * $datePreference = $this->dateFormat($format);
2141 * @param int|string|bool $usePrefs If true, the user's preference is used
2142 * if false, the site/language default is used
2143 * if int/string, assumed to be a format.
2146 function dateFormat( $usePrefs = true ) {
2149 if ( is_bool( $usePrefs ) ) {
2151 $datePreference = $wgUser->getDatePreference();
2153 $datePreference = (string)User
::getDefaultOption( 'date' );
2156 $datePreference = (string)$usePrefs;
2160 if ( $datePreference == '' ) {
2164 return $datePreference;
2168 * Get a format string for a given type and preference
2169 * @param string $type May be date, time or both
2170 * @param string $pref The format name as it appears in Messages*.php
2172 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2176 function getDateFormatString( $type, $pref ) {
2177 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2178 if ( $pref == 'default' ) {
2179 $pref = $this->getDefaultDateFormat();
2180 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2182 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2184 if ( $type === 'pretty' && $df === null ) {
2185 $df = $this->getDateFormatString( 'date', $pref );
2188 if ( $df === null ) {
2189 $pref = $this->getDefaultDateFormat();
2190 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2193 $this->dateFormatStrings
[$type][$pref] = $df;
2195 return $this->dateFormatStrings
[$type][$pref];
2199 * @param string $ts The time format which needs to be turned into a
2200 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2201 * @param bool $adj Whether to adjust the time output according to the
2202 * user configured offset ($timecorrection)
2203 * @param mixed $format True to use user's date format preference
2204 * @param string|bool $timecorrection The time offset as returned by
2205 * validateTimeZone() in Special:Preferences
2208 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2209 $ts = wfTimestamp( TS_MW
, $ts );
2211 $ts = $this->userAdjust( $ts, $timecorrection );
2213 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2214 return $this->sprintfDate( $df, $ts );
2218 * @param string $ts The time format which needs to be turned into a
2219 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2220 * @param bool $adj Whether to adjust the time output according to the
2221 * user configured offset ($timecorrection)
2222 * @param mixed $format True to use user's date format preference
2223 * @param string|bool $timecorrection The time offset as returned by
2224 * validateTimeZone() in Special:Preferences
2227 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2228 $ts = wfTimestamp( TS_MW
, $ts );
2230 $ts = $this->userAdjust( $ts, $timecorrection );
2232 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2233 return $this->sprintfDate( $df, $ts );
2237 * @param string $ts The time format which needs to be turned into a
2238 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2239 * @param bool $adj Whether to adjust the time output according to the
2240 * user configured offset ($timecorrection)
2241 * @param mixed $format What format to return, if it's false output the
2242 * default one (default true)
2243 * @param string|bool $timecorrection The time offset as returned by
2244 * validateTimeZone() in Special:Preferences
2247 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2248 $ts = wfTimestamp( TS_MW
, $ts );
2250 $ts = $this->userAdjust( $ts, $timecorrection );
2252 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2253 return $this->sprintfDate( $df, $ts );
2257 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2261 * @param int $seconds The amount of seconds.
2262 * @param array $chosenIntervals The intervals to enable.
2266 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2267 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2269 $segments = array();
2271 foreach ( $intervals as $intervalName => $intervalValue ) {
2272 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2273 // duration-years, duration-decades, duration-centuries, duration-millennia
2274 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2275 $segments[] = $message->inLanguage( $this )->escaped();
2278 return $this->listToText( $segments );
2282 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2283 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2287 * @param int $seconds The amount of seconds.
2288 * @param array $chosenIntervals The intervals to enable.
2292 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2293 if ( empty( $chosenIntervals ) ) {
2294 $chosenIntervals = array(
2306 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2307 $sortedNames = array_keys( $intervals );
2308 $smallestInterval = array_pop( $sortedNames );
2310 $segments = array();
2312 foreach ( $intervals as $name => $length ) {
2313 $value = floor( $seconds / $length );
2315 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2316 $seconds -= $value * $length;
2317 $segments[$name] = $value;
2325 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2327 * @param string $type Can be 'date', 'time' or 'both'
2328 * @param string $ts The time format which needs to be turned into a
2329 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2330 * @param User $user User object used to get preferences for timezone and format
2331 * @param array $options Array, can contain the following keys:
2332 * - 'timecorrection': time correction, can have the following values:
2333 * - true: use user's preference
2334 * - false: don't use time correction
2335 * - int: value of time correction in minutes
2336 * - 'format': format to use, can have the following values:
2337 * - true: use user's preference
2338 * - false: use default preference
2339 * - string: format to use
2343 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2344 $ts = wfTimestamp( TS_MW
, $ts );
2345 $options +
= array( 'timecorrection' => true, 'format' => true );
2346 if ( $options['timecorrection'] !== false ) {
2347 if ( $options['timecorrection'] === true ) {
2348 $offset = $user->getOption( 'timecorrection' );
2350 $offset = $options['timecorrection'];
2352 $ts = $this->userAdjust( $ts, $offset );
2354 if ( $options['format'] === true ) {
2355 $format = $user->getDatePreference();
2357 $format = $options['format'];
2359 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2360 return $this->sprintfDate( $df, $ts );
2364 * Get the formatted date for the given timestamp and formatted for
2367 * @param mixed $ts Mixed: the time format which needs to be turned into a
2368 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2369 * @param User $user User object used to get preferences for timezone and format
2370 * @param array $options Array, can contain the following keys:
2371 * - 'timecorrection': time correction, can have the following values:
2372 * - true: use user's preference
2373 * - false: don't use time correction
2374 * - int: value of time correction in minutes
2375 * - 'format': format to use, can have the following values:
2376 * - true: use user's preference
2377 * - false: use default preference
2378 * - string: format to use
2382 public function userDate( $ts, User
$user, array $options = array() ) {
2383 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2387 * Get the formatted time for the given timestamp and formatted for
2390 * @param mixed $ts The time format which needs to be turned into a
2391 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2392 * @param User $user User object used to get preferences for timezone and format
2393 * @param array $options Array, can contain the following keys:
2394 * - 'timecorrection': time correction, can have the following values:
2395 * - true: use user's preference
2396 * - false: don't use time correction
2397 * - int: value of time correction in minutes
2398 * - 'format': format to use, can have the following values:
2399 * - true: use user's preference
2400 * - false: use default preference
2401 * - string: format to use
2405 public function userTime( $ts, User
$user, array $options = array() ) {
2406 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2410 * Get the formatted date and time for the given timestamp and formatted for
2413 * @param mixed $ts The time format which needs to be turned into a
2414 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2415 * @param User $user User object used to get preferences for timezone and format
2416 * @param array $options Array, can contain the following keys:
2417 * - 'timecorrection': time correction, can have the following values:
2418 * - true: use user's preference
2419 * - false: don't use time correction
2420 * - int: value of time correction in minutes
2421 * - 'format': format to use, can have the following values:
2422 * - true: use user's preference
2423 * - false: use default preference
2424 * - string: format to use
2428 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2429 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2433 * Convert an MWTimestamp into a pretty human-readable timestamp using
2434 * the given user preferences and relative base time.
2436 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2437 * on your timestamp object, which will then call this function. Calling
2438 * this function directly will cause hooks to be skipped over.
2440 * @see MWTimestamp::getHumanTimestamp
2441 * @param MWTimestamp $ts Timestamp to prettify
2442 * @param MWTimestamp $relativeTo Base timestamp
2443 * @param User $user User preferences to use
2444 * @return string Human timestamp
2447 public function getHumanTimestamp( MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user ) {
2448 $diff = $ts->diff( $relativeTo );
2449 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2450 (int)$relativeTo->timestamp
->format( 'w' ) );
2451 $days = $diff->days ?
: (int)$diffDay;
2452 if ( $diff->invert ||
$days > 5
2453 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2455 // Timestamps are in different years: use full timestamp
2456 // Also do full timestamp for future dates
2458 * @todo FIXME: Add better handling of future timestamps.
2460 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2461 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2462 } elseif ( $days > 5 ) {
2463 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2464 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2465 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2466 } elseif ( $days > 1 ) {
2467 // Timestamp within the past week: show the day of the week and time
2468 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2469 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2471 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2472 $ts = wfMessage( "$weekday-at" )
2473 ->inLanguage( $this )
2474 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2476 } elseif ( $days == 1 ) {
2477 // Timestamp was yesterday: say 'yesterday' and the time.
2478 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2479 $ts = wfMessage( 'yesterday-at' )
2480 ->inLanguage( $this )
2481 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2483 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2484 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2485 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2486 $ts = wfMessage( 'today-at' )
2487 ->inLanguage( $this )
2488 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2491 // From here on in, the timestamp was soon enough ago so that we can simply say
2492 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2493 } elseif ( $diff->h
== 1 ) {
2494 // Less than 90 minutes, but more than an hour ago.
2495 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2496 } elseif ( $diff->i
>= 1 ) {
2497 // A few minutes ago.
2498 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2499 } elseif ( $diff->s
>= 30 ) {
2500 // Less than a minute, but more than 30 sec ago.
2501 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2503 // Less than 30 seconds ago.
2504 $ts = wfMessage( 'just-now' )->text();
2511 * @param string $key
2512 * @return array|null
2514 function getMessage( $key ) {
2515 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2521 function getAllMessages() {
2522 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2527 * @param string $out
2528 * @param string $string
2531 function iconv( $in, $out, $string ) {
2532 # This is a wrapper for iconv in all languages except esperanto,
2533 # which does some nasty x-conversions beforehand
2535 # Even with //IGNORE iconv can whine about illegal characters in
2536 # *input* string. We just ignore those too.
2537 # REF: http://bugs.php.net/bug.php?id=37166
2538 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2539 wfSuppressWarnings();
2540 $text = iconv( $in, $out . '//IGNORE', $string );
2541 wfRestoreWarnings();
2545 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2548 * @param array $matches
2549 * @return mixed|string
2551 function ucwordbreaksCallbackAscii( $matches ) {
2552 return $this->ucfirst( $matches[1] );
2556 * @param array $matches
2559 function ucwordbreaksCallbackMB( $matches ) {
2560 return mb_strtoupper( $matches[0] );
2564 * @param array $matches
2567 function ucCallback( $matches ) {
2568 list( $wikiUpperChars ) = self
::getCaseMaps();
2569 return strtr( $matches[1], $wikiUpperChars );
2573 * @param array $matches
2576 function lcCallback( $matches ) {
2577 list( , $wikiLowerChars ) = self
::getCaseMaps();
2578 return strtr( $matches[1], $wikiLowerChars );
2582 * @param array $matches
2585 function ucwordsCallbackMB( $matches ) {
2586 return mb_strtoupper( $matches[0] );
2590 * @param array $matches
2593 function ucwordsCallbackWiki( $matches ) {
2594 list( $wikiUpperChars ) = self
::getCaseMaps();
2595 return strtr( $matches[0], $wikiUpperChars );
2599 * Make a string's first character uppercase
2601 * @param string $str
2605 function ucfirst( $str ) {
2607 if ( $o < 96 ) { // if already uppercase...
2609 } elseif ( $o < 128 ) {
2610 return ucfirst( $str ); // use PHP's ucfirst()
2612 // fall back to more complex logic in case of multibyte strings
2613 return $this->uc( $str, true );
2618 * Convert a string to uppercase
2620 * @param string $str
2621 * @param bool $first
2625 function uc( $str, $first = false ) {
2626 if ( function_exists( 'mb_strtoupper' ) ) {
2628 if ( $this->isMultibyte( $str ) ) {
2629 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2631 return ucfirst( $str );
2634 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2637 if ( $this->isMultibyte( $str ) ) {
2638 $x = $first ?
'^' : '';
2639 return preg_replace_callback(
2640 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2641 array( $this, 'ucCallback' ),
2645 return $first ?
ucfirst( $str ) : strtoupper( $str );
2651 * @param string $str
2652 * @return mixed|string
2654 function lcfirst( $str ) {
2657 return strval( $str );
2658 } elseif ( $o >= 128 ) {
2659 return $this->lc( $str, true );
2660 } elseif ( $o > 96 ) {
2663 $str[0] = strtolower( $str[0] );
2669 * @param string $str
2670 * @param bool $first
2671 * @return mixed|string
2673 function lc( $str, $first = false ) {
2674 if ( function_exists( 'mb_strtolower' ) ) {
2676 if ( $this->isMultibyte( $str ) ) {
2677 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2679 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2682 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2685 if ( $this->isMultibyte( $str ) ) {
2686 $x = $first ?
'^' : '';
2687 return preg_replace_callback(
2688 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2689 array( $this, 'lcCallback' ),
2693 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2699 * @param string $str
2702 function isMultibyte( $str ) {
2703 return (bool)preg_match( '/[\x80-\xff]/', $str );
2707 * @param string $str
2708 * @return mixed|string
2710 function ucwords( $str ) {
2711 if ( $this->isMultibyte( $str ) ) {
2712 $str = $this->lc( $str );
2714 // regexp to find first letter in each word (i.e. after each space)
2715 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2717 // function to use to capitalize a single char
2718 if ( function_exists( 'mb_strtoupper' ) ) {
2719 return preg_replace_callback(
2721 array( $this, 'ucwordsCallbackMB' ),
2725 return preg_replace_callback(
2727 array( $this, 'ucwordsCallbackWiki' ),
2732 return ucwords( strtolower( $str ) );
2737 * capitalize words at word breaks
2739 * @param string $str
2742 function ucwordbreaks( $str ) {
2743 if ( $this->isMultibyte( $str ) ) {
2744 $str = $this->lc( $str );
2746 // since \b doesn't work for UTF-8, we explicitely define word break chars
2747 $breaks = "[ \-\(\)\}\{\.,\?!]";
2749 // find first letter after word break
2750 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2751 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2753 if ( function_exists( 'mb_strtoupper' ) ) {
2754 return preg_replace_callback(
2756 array( $this, 'ucwordbreaksCallbackMB' ),
2760 return preg_replace_callback(
2762 array( $this, 'ucwordsCallbackWiki' ),
2767 return preg_replace_callback(
2768 '/\b([\w\x80-\xff]+)\b/',
2769 array( $this, 'ucwordbreaksCallbackAscii' ),
2776 * Return a case-folded representation of $s
2778 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2779 * and $s2 are the same except for the case of their characters. It is not
2780 * necessary for the value returned to make sense when displayed.
2782 * Do *not* perform any other normalisation in this function. If a caller
2783 * uses this function when it should be using a more general normalisation
2784 * function, then fix the caller.
2790 function caseFold( $s ) {
2791 return $this->uc( $s );
2798 function checkTitleEncoding( $s ) {
2799 if ( is_array( $s ) ) {
2800 throw new MWException( 'Given array to checkTitleEncoding.' );
2802 if ( StringUtils
::isUtf8( $s ) ) {
2806 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2812 function fallback8bitEncoding() {
2813 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2817 * Most writing systems use whitespace to break up words.
2818 * Some languages such as Chinese don't conventionally do this,
2819 * which requires special handling when breaking up words for
2824 function hasWordBreaks() {
2829 * Some languages such as Chinese require word segmentation,
2830 * Specify such segmentation when overridden in derived class.
2832 * @param string $string
2835 function segmentByWord( $string ) {
2840 * Some languages have special punctuation need to be normalized.
2841 * Make such changes here.
2843 * @param string $string
2846 function normalizeForSearch( $string ) {
2847 return self
::convertDoubleWidth( $string );
2851 * convert double-width roman characters to single-width.
2852 * range: ff00-ff5f ~= 0020-007f
2854 * @param string $string
2858 protected static function convertDoubleWidth( $string ) {
2859 static $full = null;
2860 static $half = null;
2862 if ( $full === null ) {
2863 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2864 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2865 $full = str_split( $fullWidth, 3 );
2866 $half = str_split( $halfWidth );
2869 $string = str_replace( $full, $half, $string );
2874 * @param string $string
2875 * @param string $pattern
2878 protected static function insertSpace( $string, $pattern ) {
2879 $string = preg_replace( $pattern, " $1 ", $string );
2880 $string = preg_replace( '/ +/', ' ', $string );
2885 * @param array $termsArray
2888 function convertForSearchResult( $termsArray ) {
2889 # some languages, e.g. Chinese, need to do a conversion
2890 # in order for search results to be displayed correctly
2895 * Get the first character of a string.
2900 function firstChar( $s ) {
2903 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2904 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2909 if ( isset( $matches[1] ) ) {
2910 if ( strlen( $matches[1] ) != 3 ) {
2914 // Break down Hangul syllables to grab the first jamo
2915 $code = utf8ToCodepoint( $matches[1] );
2916 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2918 } elseif ( $code < 0xb098 ) {
2919 return "\xe3\x84\xb1";
2920 } elseif ( $code < 0xb2e4 ) {
2921 return "\xe3\x84\xb4";
2922 } elseif ( $code < 0xb77c ) {
2923 return "\xe3\x84\xb7";
2924 } elseif ( $code < 0xb9c8 ) {
2925 return "\xe3\x84\xb9";
2926 } elseif ( $code < 0xbc14 ) {
2927 return "\xe3\x85\x81";
2928 } elseif ( $code < 0xc0ac ) {
2929 return "\xe3\x85\x82";
2930 } elseif ( $code < 0xc544 ) {
2931 return "\xe3\x85\x85";
2932 } elseif ( $code < 0xc790 ) {
2933 return "\xe3\x85\x87";
2934 } elseif ( $code < 0xcc28 ) {
2935 return "\xe3\x85\x88";
2936 } elseif ( $code < 0xce74 ) {
2937 return "\xe3\x85\x8a";
2938 } elseif ( $code < 0xd0c0 ) {
2939 return "\xe3\x85\x8b";
2940 } elseif ( $code < 0xd30c ) {
2941 return "\xe3\x85\x8c";
2942 } elseif ( $code < 0xd558 ) {
2943 return "\xe3\x85\x8d";
2945 return "\xe3\x85\x8e";
2952 function initEncoding() {
2953 # Some languages may have an alternate char encoding option
2954 # (Esperanto X-coding, Japanese furigana conversion, etc)
2955 # If this language is used as the primary content language,
2956 # an override to the defaults can be set here on startup.
2963 function recodeForEdit( $s ) {
2964 # For some languages we'll want to explicitly specify
2965 # which characters make it into the edit box raw
2966 # or are converted in some way or another.
2967 global $wgEditEncoding;
2968 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2971 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2979 function recodeInput( $s ) {
2980 # Take the previous into account.
2981 global $wgEditEncoding;
2982 if ( $wgEditEncoding != '' ) {
2983 $enc = $wgEditEncoding;
2987 if ( $enc == 'UTF-8' ) {
2990 return $this->iconv( $enc, 'UTF-8', $s );
2995 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2996 * also cleans up certain backwards-compatible sequences, converting them
2997 * to the modern Unicode equivalent.
2999 * This is language-specific for performance reasons only.
3005 function normalize( $s ) {
3006 global $wgAllUnicodeFixes;
3007 $s = UtfNormal
::cleanUp( $s );
3008 if ( $wgAllUnicodeFixes ) {
3009 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3010 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3017 * Transform a string using serialized data stored in the given file (which
3018 * must be in the serialized subdirectory of $IP). The file contains pairs
3019 * mapping source characters to destination characters.
3021 * The data is cached in process memory. This will go faster if you have the
3022 * FastStringSearch extension.
3024 * @param string $file
3025 * @param string $string
3027 * @throws MWException
3030 function transformUsingPairFile( $file, $string ) {
3031 if ( !isset( $this->transformData
[$file] ) ) {
3032 $data = wfGetPrecompiledData( $file );
3033 if ( $data === false ) {
3034 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3036 $this->transformData
[$file] = new ReplacementArray( $data );
3038 return $this->transformData
[$file]->replace( $string );
3042 * For right-to-left language support
3047 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3051 * Return the correct HTML 'dir' attribute value for this language.
3055 return $this->isRTL() ?
'rtl' : 'ltr';
3059 * Return 'left' or 'right' as appropriate alignment for line-start
3060 * for this language's text direction.
3062 * Should be equivalent to CSS3 'start' text-align value....
3066 function alignStart() {
3067 return $this->isRTL() ?
'right' : 'left';
3071 * Return 'right' or 'left' as appropriate alignment for line-end
3072 * for this language's text direction.
3074 * Should be equivalent to CSS3 'end' text-align value....
3078 function alignEnd() {
3079 return $this->isRTL() ?
'left' : 'right';
3083 * A hidden direction mark (LRM or RLM), depending on the language direction.
3084 * Unlike getDirMark(), this function returns the character as an HTML entity.
3085 * This function should be used when the output is guaranteed to be HTML,
3086 * because it makes the output HTML source code more readable. When
3087 * the output is plain text or can be escaped, getDirMark() should be used.
3089 * @param bool $opposite Get the direction mark opposite to your language
3093 function getDirMarkEntity( $opposite = false ) {
3095 return $this->isRTL() ?
'‎' : '‏';
3097 return $this->isRTL() ?
'‏' : '‎';
3101 * A hidden direction mark (LRM or RLM), depending on the language direction.
3102 * This function produces them as invisible Unicode characters and
3103 * the output may be hard to read and debug, so it should only be used
3104 * when the output is plain text or can be escaped. When the output is
3105 * HTML, use getDirMarkEntity() instead.
3107 * @param bool $opposite Get the direction mark opposite to your language
3110 function getDirMark( $opposite = false ) {
3111 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3112 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3114 return $this->isRTL() ?
$lrm : $rlm;
3116 return $this->isRTL() ?
$rlm : $lrm;
3122 function capitalizeAllNouns() {
3123 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3127 * An arrow, depending on the language direction.
3129 * @param string $direction The direction of the arrow: forwards (default),
3130 * backwards, left, right, up, down.
3133 function getArrow( $direction = 'forwards' ) {
3134 switch ( $direction ) {
3136 return $this->isRTL() ?
'←' : '→';
3138 return $this->isRTL() ?
'→' : '←';
3151 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3155 function linkPrefixExtension() {
3156 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3160 * Get all magic words from cache.
3163 function getMagicWords() {
3164 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3168 * Run the LanguageGetMagic hook once.
3170 protected function doMagicHook() {
3171 if ( $this->mMagicHookDone
) {
3174 $this->mMagicHookDone
= true;
3175 Hooks
::run( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
3179 * Fill a MagicWord object with data from here
3181 * @param MagicWord $mw
3183 function getMagic( $mw ) {
3184 // Saves a function call
3185 if ( !$this->mMagicHookDone
) {
3186 $this->doMagicHook();
3189 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3190 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3192 $rawEntry = self
::$dataCache->getSubitem(
3193 $this->mCode
, 'magicWords', $mw->mId
);
3196 if ( !is_array( $rawEntry ) ) {
3197 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3199 $mw->mCaseSensitive
= $rawEntry[0];
3200 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3205 * Add magic words to the extension array
3207 * @param array $newWords
3209 function addMagicWordsByLang( $newWords ) {
3210 $fallbackChain = $this->getFallbackLanguages();
3211 $fallbackChain = array_reverse( $fallbackChain );
3212 foreach ( $fallbackChain as $code ) {
3213 if ( isset( $newWords[$code] ) ) {
3214 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3220 * Get special page names, as an associative array
3221 * canonical name => array of valid names, including aliases
3224 function getSpecialPageAliases() {
3225 // Cache aliases because it may be slow to load them
3226 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3228 $this->mExtendedSpecialPageAliases
=
3229 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3230 Hooks
::run( 'LanguageGetSpecialPageAliases',
3231 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
3234 return $this->mExtendedSpecialPageAliases
;
3238 * Italic is unsuitable for some languages
3240 * @param string $text The text to be emphasized.
3243 function emphasize( $text ) {
3244 return "<em>$text</em>";
3248 * Normally we output all numbers in plain en_US style, that is
3249 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3250 * point twohundredthirtyfive. However this is not suitable for all
3251 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3252 * Icelandic just want to use commas instead of dots, and dots instead
3253 * of commas like "293.291,235".
3255 * An example of this function being called:
3257 * wfMessage( 'message' )->numParams( $num )->text()
3260 * See $separatorTransformTable on MessageIs.php for
3261 * the , => . and . => , implementation.
3263 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3264 * @param int|float $number The string to be formatted, should be an integer
3265 * or a floating point number.
3266 * @param bool $nocommafy Set to true for special numbers like dates
3269 public function formatNum( $number, $nocommafy = false ) {
3270 global $wgTranslateNumerals;
3271 if ( !$nocommafy ) {
3272 $number = $this->commafy( $number );
3273 $s = $this->separatorTransformTable();
3275 $number = strtr( $number, $s );
3279 if ( $wgTranslateNumerals ) {
3280 $s = $this->digitTransformTable();
3282 $number = strtr( $number, $s );
3290 * Front-end for non-commafied formatNum
3292 * @param int|float $number The string to be formatted, should be an integer
3293 * or a floating point number.
3297 public function formatNumNoSeparators( $number ) {
3298 return $this->formatNum( $number, true );
3302 * @param string $number
3305 public function parseFormattedNumber( $number ) {
3306 $s = $this->digitTransformTable();
3308 // eliminate empty array values such as ''. (bug 64347)
3309 $s = array_filter( $s );
3310 $number = strtr( $number, array_flip( $s ) );
3313 $s = $this->separatorTransformTable();
3315 // eliminate empty array values such as ''. (bug 64347)
3316 $s = array_filter( $s );
3317 $number = strtr( $number, array_flip( $s ) );
3320 $number = strtr( $number, array( ',' => '' ) );
3325 * Adds commas to a given number
3327 * @param mixed $number
3330 function commafy( $number ) {
3331 $digitGroupingPattern = $this->digitGroupingPattern();
3332 if ( $number === null ) {
3336 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3337 // default grouping is at thousands, use the same for ###,###,### pattern too.
3338 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3340 // Ref: http://cldr.unicode.org/translation/number-patterns
3342 if ( intval( $number ) < 0 ) {
3343 // For negative numbers apply the algorithm like positive number and add sign.
3345 $number = substr( $number, 1 );
3347 $integerPart = array();
3348 $decimalPart = array();
3349 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3350 preg_match( "/\d+/", $number, $integerPart );
3351 preg_match( "/\.\d*/", $number, $decimalPart );
3352 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3353 if ( $groupedNumber === $number ) {
3354 // the string does not have any number part. Eg: .12345
3355 return $sign . $groupedNumber;
3357 $start = $end = ($integerPart) ?
strlen( $integerPart[0] ) : 0;
3358 while ( $start > 0 ) {
3359 $match = $matches[0][$numMatches - 1];
3360 $matchLen = strlen( $match );
3361 $start = $end - $matchLen;
3365 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3367 if ( $numMatches > 1 ) {
3368 // use the last pattern for the rest of the number
3372 $groupedNumber = "," . $groupedNumber;
3375 return $sign . $groupedNumber;
3382 function digitGroupingPattern() {
3383 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3389 function digitTransformTable() {
3390 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3396 function separatorTransformTable() {
3397 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3401 * Take a list of strings and build a locale-friendly comma-separated
3402 * list, using the local comma-separator message.
3403 * The last two strings are chained with an "and".
3404 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3406 * @param string[] $l
3409 function listToText( array $l ) {
3410 $m = count( $l ) - 1;
3415 $and = $this->msg( 'and' )->escaped();
3416 $space = $this->msg( 'word-separator' )->escaped();
3418 $comma = $this->msg( 'comma-separator' )->escaped();
3422 for ( $i = $m - 1; $i >= 0; $i-- ) {
3423 if ( $i == $m - 1 ) {
3424 $s = $l[$i] . $and . $space . $s;
3426 $s = $l[$i] . $comma . $s;
3433 * Take a list of strings and build a locale-friendly comma-separated
3434 * list, using the local comma-separator message.
3435 * @param string[] $list Array of strings to put in a comma list
3438 function commaList( array $list ) {
3440 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3446 * Take a list of strings and build a locale-friendly semicolon-separated
3447 * list, using the local semicolon-separator message.
3448 * @param string[] $list Array of strings to put in a semicolon list
3451 function semicolonList( array $list ) {
3453 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3459 * Same as commaList, but separate it with the pipe instead.
3460 * @param string[] $list Array of strings to put in a pipe list
3463 function pipeList( array $list ) {
3465 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3471 * Truncate a string to a specified length in bytes, appending an optional
3472 * string (e.g. for ellipses)
3474 * The database offers limited byte lengths for some columns in the database;
3475 * multi-byte character sets mean we need to ensure that only whole characters
3476 * are included, otherwise broken characters can be passed to the user
3478 * If $length is negative, the string will be truncated from the beginning
3480 * @param string $string String to truncate
3481 * @param int $length Maximum length (including ellipses)
3482 * @param string $ellipsis String to append to the truncated text
3483 * @param bool $adjustLength Subtract length of ellipsis from $length.
3484 * $adjustLength was introduced in 1.18, before that behaved as if false.
3487 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3488 # Use the localized ellipsis character
3489 if ( $ellipsis == '...' ) {
3490 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3492 # Check if there is no need to truncate
3493 if ( $length == 0 ) {
3494 return $ellipsis; // convention
3495 } elseif ( strlen( $string ) <= abs( $length ) ) {
3496 return $string; // no need to truncate
3498 $stringOriginal = $string;
3499 # If ellipsis length is >= $length then we can't apply $adjustLength
3500 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3501 $string = $ellipsis; // this can be slightly unexpected
3502 # Otherwise, truncate and add ellipsis...
3504 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3505 if ( $length > 0 ) {
3506 $length -= $eLength;
3507 $string = substr( $string, 0, $length ); // xyz...
3508 $string = $this->removeBadCharLast( $string );
3509 $string = rtrim( $string );
3510 $string = $string . $ellipsis;
3512 $length +
= $eLength;
3513 $string = substr( $string, $length ); // ...xyz
3514 $string = $this->removeBadCharFirst( $string );
3515 $string = ltrim( $string );
3516 $string = $ellipsis . $string;
3519 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3520 # This check is *not* redundant if $adjustLength, due to the single case where
3521 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3522 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3525 return $stringOriginal;
3530 * Remove bytes that represent an incomplete Unicode character
3531 * at the end of string (e.g. bytes of the char are missing)
3533 * @param string $string
3536 protected function removeBadCharLast( $string ) {
3537 if ( $string != '' ) {
3538 $char = ord( $string[strlen( $string ) - 1] );
3540 if ( $char >= 0xc0 ) {
3541 # We got the first byte only of a multibyte char; remove it.
3542 $string = substr( $string, 0, -1 );
3543 } elseif ( $char >= 0x80 &&
3544 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3545 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3547 # We chopped in the middle of a character; remove it
3555 * Remove bytes that represent an incomplete Unicode character
3556 * at the start of string (e.g. bytes of the char are missing)
3558 * @param string $string
3561 protected function removeBadCharFirst( $string ) {
3562 if ( $string != '' ) {
3563 $char = ord( $string[0] );
3564 if ( $char >= 0x80 && $char < 0xc0 ) {
3565 # We chopped in the middle of a character; remove the whole thing
3566 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3573 * Truncate a string of valid HTML to a specified length in bytes,
3574 * appending an optional string (e.g. for ellipses), and return valid HTML
3576 * This is only intended for styled/linked text, such as HTML with
3577 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3578 * Also, this will not detect things like "display:none" CSS.
3580 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3582 * @param string $text HTML string to truncate
3583 * @param int $length (zero/positive) Maximum length (including ellipses)
3584 * @param string $ellipsis String to append to the truncated text
3587 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3588 # Use the localized ellipsis character
3589 if ( $ellipsis == '...' ) {
3590 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3592 # Check if there is clearly no need to truncate
3593 if ( $length <= 0 ) {
3594 return $ellipsis; // no text shown, nothing to format (convention)
3595 } elseif ( strlen( $text ) <= $length ) {
3596 return $text; // string short enough even *with* HTML (short-circuit)
3599 $dispLen = 0; // innerHTML legth so far
3600 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3601 $tagType = 0; // 0-open, 1-close
3602 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3603 $entityState = 0; // 0-not entity, 1-entity
3604 $tag = $ret = ''; // accumulated tag name, accumulated result string
3605 $openTags = array(); // open tag stack
3606 $maybeState = null; // possible truncation state
3608 $textLen = strlen( $text );
3609 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3610 for ( $pos = 0; true; ++
$pos ) {
3611 # Consider truncation once the display length has reached the maximim.
3612 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3613 # Check that we're not in the middle of a bracket/entity...
3614 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3615 if ( !$testingEllipsis ) {
3616 $testingEllipsis = true;
3617 # Save where we are; we will truncate here unless there turn out to
3618 # be so few remaining characters that truncation is not necessary.
3619 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3620 $maybeState = array( $ret, $openTags ); // save state
3622 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3623 # String in fact does need truncation, the truncation point was OK.
3624 list( $ret, $openTags ) = $maybeState; // reload state
3625 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3626 $ret .= $ellipsis; // add ellipsis
3630 if ( $pos >= $textLen ) {
3631 break; // extra iteration just for above checks
3634 # Read the next char...
3636 $lastCh = $pos ?
$text[$pos - 1] : '';
3637 $ret .= $ch; // add to result string
3639 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3640 $entityState = 0; // for bad HTML
3641 $bracketState = 1; // tag started (checking for backslash)
3642 } elseif ( $ch == '>' ) {
3643 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3644 $entityState = 0; // for bad HTML
3645 $bracketState = 0; // out of brackets
3646 } elseif ( $bracketState == 1 ) {
3648 $tagType = 1; // close tag (e.g. "</span>")
3650 $tagType = 0; // open tag (e.g. "<span>")
3653 $bracketState = 2; // building tag name
3654 } elseif ( $bracketState == 2 ) {
3658 // Name found (e.g. "<a href=..."), add on tag attributes...
3659 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3661 } elseif ( $bracketState == 0 ) {
3662 if ( $entityState ) {
3665 $dispLen++
; // entity is one displayed char
3668 if ( $neLength == 0 && !$maybeState ) {
3669 // Save state without $ch. We want to *hit* the first
3670 // display char (to get tags) but not *use* it if truncating.
3671 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3674 $entityState = 1; // entity found, (e.g. " ")
3676 $dispLen++
; // this char is displayed
3677 // Add the next $max display text chars after this in one swoop...
3678 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3679 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3680 $dispLen +
= $skipped;
3686 // Close the last tag if left unclosed by bad HTML
3687 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3688 while ( count( $openTags ) > 0 ) {
3689 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3695 * truncateHtml() helper function
3696 * like strcspn() but adds the skipped chars to $ret
3698 * @param string $ret
3699 * @param string $text
3700 * @param string $search
3702 * @param null|int $len
3705 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3706 if ( $len === null ) {
3707 $len = -1; // -1 means "no limit" for strcspn
3708 } elseif ( $len < 0 ) {
3712 if ( $start < strlen( $text ) ) {
3713 $skipCount = strcspn( $text, $search, $start, $len );
3714 $ret .= substr( $text, $start, $skipCount );
3720 * truncateHtml() helper function
3721 * (a) push or pop $tag from $openTags as needed
3722 * (b) clear $tag value
3723 * @param string &$tag Current HTML tag name we are looking at
3724 * @param int $tagType (0-open tag, 1-close tag)
3725 * @param string $lastCh Character before the '>' that ended this tag
3726 * @param array &$openTags Open tag stack (not accounting for $tag)
3728 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3729 $tag = ltrim( $tag );
3731 if ( $tagType == 0 && $lastCh != '/' ) {
3732 $openTags[] = $tag; // tag opened (didn't close itself)
3733 } elseif ( $tagType == 1 ) {
3734 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3735 array_pop( $openTags ); // tag closed
3743 * Grammatical transformations, needed for inflected languages
3744 * Invoked by putting {{grammar:case|word}} in a message
3746 * @param string $word
3747 * @param string $case
3750 function convertGrammar( $word, $case ) {
3751 global $wgGrammarForms;
3752 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3753 return $wgGrammarForms[$this->getCode()][$case][$word];
3759 * Get the grammar forms for the content language
3760 * @return array Array of grammar forms
3763 function getGrammarForms() {
3764 global $wgGrammarForms;
3765 if ( isset( $wgGrammarForms[$this->getCode()] )
3766 && is_array( $wgGrammarForms[$this->getCode()] )
3768 return $wgGrammarForms[$this->getCode()];
3774 * Provides an alternative text depending on specified gender.
3775 * Usage {{gender:username|masculine|feminine|unknown}}.
3776 * username is optional, in which case the gender of current user is used,
3777 * but only in (some) interface messages; otherwise default gender is used.
3779 * If no forms are given, an empty string is returned. If only one form is
3780 * given, it will be returned unconditionally. These details are implied by
3781 * the caller and cannot be overridden in subclasses.
3783 * If three forms are given, the default is to use the third (unknown) form.
3784 * If fewer than three forms are given, the default is to use the first (masculine) form.
3785 * These details can be overridden in subclasses.
3787 * @param string $gender
3788 * @param array $forms
3792 function gender( $gender, $forms ) {
3793 if ( !count( $forms ) ) {
3796 $forms = $this->preConvertPlural( $forms, 2 );
3797 if ( $gender === 'male' ) {
3800 if ( $gender === 'female' ) {
3803 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3807 * Plural form transformations, needed for some languages.
3808 * For example, there are 3 form of plural in Russian and Polish,
3809 * depending on "count mod 10". See [[w:Plural]]
3810 * For English it is pretty simple.
3812 * Invoked by putting {{plural:count|wordform1|wordform2}}
3813 * or {{plural:count|wordform1|wordform2|wordform3}}
3815 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3817 * @param int $count Non-localized number
3818 * @param array $forms Different plural forms
3819 * @return string Correct form of plural for $count in this language
3821 function convertPlural( $count, $forms ) {
3822 // Handle explicit n=pluralform cases
3823 $forms = $this->handleExplicitPluralForms( $count, $forms );
3824 if ( is_string( $forms ) ) {
3827 if ( !count( $forms ) ) {
3831 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3832 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3833 return $forms[$pluralForm];
3837 * Handles explicit plural forms for Language::convertPlural()
3839 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3840 * If an explicitly defined plural form matches the $count, then
3841 * string value returned, otherwise array returned for further consideration
3842 * by CLDR rules or overridden convertPlural().
3846 * @param int $count Non-localized number
3847 * @param array $forms Different plural forms
3849 * @return array|string
3851 protected function handleExplicitPluralForms( $count, array $forms ) {
3852 foreach ( $forms as $index => $form ) {
3853 if ( preg_match( '/\d+=/i', $form ) ) {
3854 $pos = strpos( $form, '=' );
3855 if ( substr( $form, 0, $pos ) === (string)$count ) {
3856 return substr( $form, $pos +
1 );
3858 unset( $forms[$index] );
3861 return array_values( $forms );
3865 * Checks that convertPlural was given an array and pads it to requested
3866 * amount of forms by copying the last one.
3868 * @param array $forms Array of forms given to convertPlural
3869 * @param int $count How many forms should there be at least
3870 * @return array Padded array of forms or an exception if not an array
3872 protected function preConvertPlural( /* Array */ $forms, $count ) {
3873 while ( count( $forms ) < $count ) {
3874 $forms[] = $forms[count( $forms ) - 1];
3880 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3881 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3882 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3883 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3886 * @param string $str The validated block duration in English
3887 * @return string Somehow translated block duration
3888 * @see LanguageFi.php for example implementation
3890 function translateBlockExpiry( $str ) {
3891 $duration = SpecialBlock
::getSuggestedDurations( $this );
3892 foreach ( $duration as $show => $value ) {
3893 if ( strcmp( $str, $value ) == 0 ) {
3894 return htmlspecialchars( trim( $show ) );
3898 // Since usually only infinite or indefinite is only on list, so try
3899 // equivalents if still here.
3900 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3901 if ( in_array( $str, $indefs ) ) {
3902 foreach ( $indefs as $val ) {
3903 $show = array_search( $val, $duration, true );
3904 if ( $show !== false ) {
3905 return htmlspecialchars( trim( $show ) );
3910 // If all else fails, return a standard duration or timestamp description.
3911 $time = strtotime( $str, 0 );
3912 if ( $time === false ) { // Unknown format. Return it as-is in case.
3914 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3915 // $time is relative to 0 so it's a duration length.
3916 return $this->formatDuration( $time );
3917 } else { // It's an absolute timestamp.
3918 if ( $time === 0 ) {
3919 // wfTimestamp() handles 0 as current time instead of epoch.
3920 return $this->timeanddate( '19700101000000' );
3922 return $this->timeanddate( $time );
3928 * languages like Chinese need to be segmented in order for the diff
3931 * @param string $text
3934 public function segmentForDiff( $text ) {
3939 * and unsegment to show the result
3941 * @param string $text
3944 public function unsegmentForDiff( $text ) {
3949 * Return the LanguageConverter used in the Language
3952 * @return LanguageConverter
3954 public function getConverter() {
3955 return $this->mConverter
;
3959 * convert text to all supported variants
3961 * @param string $text
3964 public function autoConvertToAllVariants( $text ) {
3965 return $this->mConverter
->autoConvertToAllVariants( $text );
3969 * convert text to different variants of a language.
3971 * @param string $text
3974 public function convert( $text ) {
3975 return $this->mConverter
->convert( $text );
3979 * Convert a Title object to a string in the preferred variant
3981 * @param Title $title
3984 public function convertTitle( $title ) {
3985 return $this->mConverter
->convertTitle( $title );
3989 * Convert a namespace index to a string in the preferred variant
3994 public function convertNamespace( $ns ) {
3995 return $this->mConverter
->convertNamespace( $ns );
3999 * Check if this is a language with variants
4003 public function hasVariants() {
4004 return count( $this->getVariants() ) > 1;
4008 * Check if the language has the specific variant
4011 * @param string $variant
4014 public function hasVariant( $variant ) {
4015 return (bool)$this->mConverter
->validateVariant( $variant );
4019 * Put custom tags (e.g. -{ }-) around math to prevent conversion
4021 * @param string $text
4023 * @deprecated since 1.22 is no longer used
4025 public function armourMath( $text ) {
4026 return $this->mConverter
->armourMath( $text );
4030 * Perform output conversion on a string, and encode for safe HTML output.
4031 * @param string $text Text to be converted
4032 * @param bool $isTitle Whether this conversion is for the article title
4034 * @todo this should get integrated somewhere sane
4036 public function convertHtml( $text, $isTitle = false ) {
4037 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4041 * @param string $key
4044 public function convertCategoryKey( $key ) {
4045 return $this->mConverter
->convertCategoryKey( $key );
4049 * Get the list of variants supported by this language
4050 * see sample implementation in LanguageZh.php
4052 * @return array An array of language codes
4054 public function getVariants() {
4055 return $this->mConverter
->getVariants();
4061 public function getPreferredVariant() {
4062 return $this->mConverter
->getPreferredVariant();
4068 public function getDefaultVariant() {
4069 return $this->mConverter
->getDefaultVariant();
4075 public function getURLVariant() {
4076 return $this->mConverter
->getURLVariant();
4080 * If a language supports multiple variants, it is
4081 * possible that non-existing link in one variant
4082 * actually exists in another variant. this function
4083 * tries to find it. See e.g. LanguageZh.php
4084 * The input parameters may be modified upon return
4086 * @param string &$link The name of the link
4087 * @param Title &$nt The title object of the link
4088 * @param bool $ignoreOtherCond To disable other conditions when
4089 * we need to transclude a template or update a category's link
4091 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4092 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4096 * returns language specific options used by User::getPageRenderHash()
4097 * for example, the preferred language variant
4101 function getExtraHashOptions() {
4102 return $this->mConverter
->getExtraHashOptions();
4106 * For languages that support multiple variants, the title of an
4107 * article may be displayed differently in different variants. this
4108 * function returns the apporiate title defined in the body of the article.
4112 public function getParsedTitle() {
4113 return $this->mConverter
->getParsedTitle();
4117 * Prepare external link text for conversion. When the text is
4118 * a URL, it shouldn't be converted, and it'll be wrapped in
4119 * the "raw" tag (-{R| }-) to prevent conversion.
4121 * This function is called "markNoConversion" for historical
4124 * @param string $text Text to be used for external link
4125 * @param bool $noParse Wrap it without confirming it's a real URL first
4126 * @return string The tagged text
4128 public function markNoConversion( $text, $noParse = false ) {
4129 // Excluding protocal-relative URLs may avoid many false positives.
4130 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4131 return $this->mConverter
->markNoConversion( $text );
4138 * A regular expression to match legal word-trailing characters
4139 * which should be merged onto a link of the form [[foo]]bar.
4143 public function linkTrail() {
4144 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4148 * A regular expression character set to match legal word-prefixing
4149 * characters which should be merged onto a link of the form foo[[bar]].
4153 public function linkPrefixCharset() {
4154 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4158 * @deprecated since 1.24, will be removed in 1.25
4161 function getLangObj() {
4162 wfDeprecated( __METHOD__
, '1.24' );
4167 * Get the "parent" language which has a converter to convert a "compatible" language
4168 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4170 * @return Language|null
4173 public function getParentLanguage() {
4174 if ( $this->mParentLanguage
!== false ) {
4175 return $this->mParentLanguage
;
4178 $pieces = explode( '-', $this->getCode() );
4180 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4181 $this->mParentLanguage
= null;
4184 $lang = Language
::factory( $code );
4185 if ( !$lang->hasVariant( $this->getCode() ) ) {
4186 $this->mParentLanguage
= null;
4190 $this->mParentLanguage
= $lang;
4195 * Get the RFC 3066 code for this language object
4197 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4198 * htmlspecialchars() or similar
4202 public function getCode() {
4203 return $this->mCode
;
4207 * Get the code in Bcp47 format which we can use
4208 * inside of html lang="" tags.
4210 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4211 * htmlspecialchars() or similar.
4216 public function getHtmlCode() {
4217 if ( is_null( $this->mHtmlCode
) ) {
4218 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4220 return $this->mHtmlCode
;
4224 * @param string $code
4226 public function setCode( $code ) {
4227 $this->mCode
= $code;
4228 // Ensure we don't leave incorrect cached data lying around
4229 $this->mHtmlCode
= null;
4230 $this->mParentLanguage
= false;
4234 * Get the name of a file for a certain language code
4235 * @param string $prefix Prepend this to the filename
4236 * @param string $code Language code
4237 * @param string $suffix Append this to the filename
4238 * @throws MWException
4239 * @return string $prefix . $mangledCode . $suffix
4241 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4242 if ( !self
::isValidBuiltInCode( $code ) ) {
4243 throw new MWException( "Invalid language code \"$code\"" );
4246 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4250 * Get the language code from a file name. Inverse of getFileName()
4251 * @param string $filename $prefix . $languageCode . $suffix
4252 * @param string $prefix Prefix before the language code
4253 * @param string $suffix Suffix after the language code
4254 * @return string Language code, or false if $prefix or $suffix isn't found
4256 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4258 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4259 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4260 if ( !count( $m ) ) {
4263 return str_replace( '_', '-', strtolower( $m[1] ) );
4267 * @param string $code
4270 public static function getMessagesFileName( $code ) {
4272 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4273 Hooks
::run( 'Language::getMessagesFileName', array( $code, &$file ) );
4278 * @param string $code
4282 public static function getJsonMessagesFileName( $code ) {
4285 if ( !self
::isValidBuiltInCode( $code ) ) {
4286 throw new MWException( "Invalid language code \"$code\"" );
4289 return "$IP/languages/i18n/$code.json";
4293 * @param string $code
4296 public static function getClassFileName( $code ) {
4298 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4302 * Get the first fallback for a given language.
4304 * @param string $code
4306 * @return bool|string
4308 public static function getFallbackFor( $code ) {
4309 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4312 $fallbacks = self
::getFallbacksFor( $code );
4313 $first = array_shift( $fallbacks );
4319 * Get the ordered list of fallback languages.
4322 * @param string $code Language code
4325 public static function getFallbacksFor( $code ) {
4326 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4329 $v = self
::getLocalisationCache()->getItem( $code, 'fallback' );
4330 $v = array_map( 'trim', explode( ',', $v ) );
4331 if ( $v[count( $v ) - 1] !== 'en' ) {
4339 * Get the ordered list of fallback languages, ending with the fallback
4340 * language chain for the site language.
4343 * @param string $code Language code
4344 * @return array Array( fallbacks, site fallbacks )
4346 public static function getFallbacksIncludingSiteLanguage( $code ) {
4347 global $wgLanguageCode;
4349 // Usually, we will only store a tiny number of fallback chains, so we
4350 // keep them in static memory.
4351 $cacheKey = "{$code}-{$wgLanguageCode}";
4353 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4354 $fallbacks = self
::getFallbacksFor( $code );
4356 // Append the site's fallback chain, including the site language itself
4357 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4358 array_unshift( $siteFallbacks, $wgLanguageCode );
4360 // Eliminate any languages already included in the chain
4361 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4363 self
::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4365 return self
::$fallbackLanguageCache[$cacheKey];
4369 * Get all messages for a given language
4370 * WARNING: this may take a long time. If you just need all message *keys*
4371 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4373 * @param string $code
4377 public static function getMessagesFor( $code ) {
4378 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4382 * Get a message for a given language
4384 * @param string $key
4385 * @param string $code
4389 public static function getMessageFor( $key, $code ) {
4390 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4394 * Get all message keys for a given language. This is a faster alternative to
4395 * array_keys( Language::getMessagesFor( $code ) )
4398 * @param string $code Language code
4399 * @return array Array of message keys (strings)
4401 public static function getMessageKeysFor( $code ) {
4402 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4406 * @param string $talk
4409 function fixVariableInNamespace( $talk ) {
4410 if ( strpos( $talk, '$1' ) === false ) {
4414 global $wgMetaNamespace;
4415 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4417 # Allow grammar transformations
4418 # Allowing full message-style parsing would make simple requests
4419 # such as action=raw much more expensive than they need to be.
4420 # This will hopefully cover most cases.
4421 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4422 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4423 return str_replace( ' ', '_', $talk );
4430 function replaceGrammarInNamespace( $m ) {
4431 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4435 * @throws MWException
4438 static function getCaseMaps() {
4439 static $wikiUpperChars, $wikiLowerChars;
4440 if ( isset( $wikiUpperChars ) ) {
4441 return array( $wikiUpperChars, $wikiLowerChars );
4444 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4445 if ( $arr === false ) {
4446 throw new MWException(
4447 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4449 $wikiUpperChars = $arr['wikiUpperChars'];
4450 $wikiLowerChars = $arr['wikiLowerChars'];
4451 return array( $wikiUpperChars, $wikiLowerChars );
4455 * Decode an expiry (block, protection, etc) which has come from the DB
4457 * @todo FIXME: why are we returnings DBMS-dependent strings???
4459 * @param string $expiry Database expiry String
4460 * @param bool|int $format True to process using language functions, or TS_ constant
4461 * to return the expiry in a given timestamp
4465 public function formatExpiry( $expiry, $format = true ) {
4467 if ( $infinity === null ) {
4468 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
4471 if ( $expiry == '' ||
$expiry == $infinity ) {
4472 return $format === true
4473 ?
$this->getMessageFromDB( 'infiniteblock' )
4476 return $format === true
4477 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4478 : wfTimestamp( $format, $expiry );
4484 * @param int|float $seconds
4485 * @param array $format Optional
4486 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4487 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4488 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4490 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4491 * or 'avoidminutes'.
4494 function formatTimePeriod( $seconds, $format = array() ) {
4495 if ( !is_array( $format ) ) {
4496 $format = array( 'avoid' => $format ); // For backwards compatibility
4498 if ( !isset( $format['avoid'] ) ) {
4499 $format['avoid'] = false;
4501 if ( !isset( $format['noabbrevs'] ) ) {
4502 $format['noabbrevs'] = false;
4504 $secondsMsg = wfMessage(
4505 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4506 $minutesMsg = wfMessage(
4507 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4508 $hoursMsg = wfMessage(
4509 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4510 $daysMsg = wfMessage(
4511 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4513 if ( round( $seconds * 10 ) < 100 ) {
4514 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4515 $s = $secondsMsg->params( $s )->text();
4516 } elseif ( round( $seconds ) < 60 ) {
4517 $s = $this->formatNum( round( $seconds ) );
4518 $s = $secondsMsg->params( $s )->text();
4519 } elseif ( round( $seconds ) < 3600 ) {
4520 $minutes = floor( $seconds / 60 );
4521 $secondsPart = round( fmod( $seconds, 60 ) );
4522 if ( $secondsPart == 60 ) {
4526 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4528 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4529 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4530 $hours = floor( $seconds / 3600 );
4531 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4532 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4533 if ( $secondsPart == 60 ) {
4537 if ( $minutes == 60 ) {
4541 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4543 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4544 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4545 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4548 $days = floor( $seconds / 86400 );
4549 if ( $format['avoid'] === 'avoidminutes' ) {
4550 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4551 if ( $hours == 24 ) {
4555 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4557 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4558 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4559 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4560 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4561 if ( $minutes == 60 ) {
4565 if ( $hours == 24 ) {
4569 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4571 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4573 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4575 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4577 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4584 * Format a bitrate for output, using an appropriate
4585 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4586 * the magnitude in question.
4588 * This use base 1000. For base 1024 use formatSize(), for another base
4589 * see formatComputingNumbers().
4594 function formatBitrate( $bps ) {
4595 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4599 * @param int $size Size of the unit
4600 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4601 * @param string $messageKey Message key to be uesd
4604 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4606 return str_replace( '$1', $this->formatNum( $size ),
4607 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4610 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4613 $maxIndex = count( $sizes ) - 1;
4614 while ( $size >= $boundary && $index < $maxIndex ) {
4619 // For small sizes no decimal places necessary
4622 // For MB and bigger two decimal places are smarter
4625 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4627 $size = round( $size, $round );
4628 $text = $this->getMessageFromDB( $msg );
4629 return str_replace( '$1', $this->formatNum( $size ), $text );
4633 * Format a size in bytes for output, using an appropriate
4634 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4636 * This method use base 1024. For base 1000 use formatBitrate(), for
4637 * another base see formatComputingNumbers()
4639 * @param int $size Size to format
4640 * @return string Plain text (not HTML)
4642 function formatSize( $size ) {
4643 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4647 * Make a list item, used by various special pages
4649 * @param string $page Page link
4650 * @param string $details HTML safe text between brackets
4651 * @param bool $oppositedm Add the direction mark opposite to your
4652 * language, to display text properly
4653 * @return HTML escaped string
4655 function specialList( $page, $details, $oppositedm = true ) {
4660 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) . $this->getDirMark();
4664 $this->msg( 'word-separator' )->escaped() .
4665 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4669 * Generate (prev x| next x) (20|50|100...) type links for paging
4671 * @param Title $title Title object to link
4672 * @param int $offset
4674 * @param array $query Optional URL query parameter string
4675 * @param bool $atend Optional param for specified if this is the last page
4678 public function viewPrevNext( Title
$title, $offset, $limit,
4679 array $query = array(), $atend = false
4681 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4683 # Make 'previous' link
4684 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4685 if ( $offset > 0 ) {
4686 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4687 $query, $prev, 'prevn-title', 'mw-prevlink' );
4689 $plink = htmlspecialchars( $prev );
4693 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4695 $nlink = htmlspecialchars( $next );
4697 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4698 $query, $next, 'nextn-title', 'mw-nextlink' );
4701 # Make links to set number of items per page
4702 $numLinks = array();
4703 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4704 $numLinks[] = $this->numLink( $title, $offset, $num,
4705 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4708 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4709 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4713 * Helper function for viewPrevNext() that generates links
4715 * @param Title $title Title object to link
4716 * @param int $offset
4718 * @param array $query Extra query parameters
4719 * @param string $link Text to use for the link; will be escaped
4720 * @param string $tooltipMsg Name of the message to use as tooltip
4721 * @param string $class Value of the "class" attribute of the link
4722 * @return string HTML fragment
4724 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4727 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4728 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4729 ->numParams( $limit )->text();
4731 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4732 'title' => $tooltip, 'class' => $class ), $link );
4736 * Get the conversion rule title, if any.
4740 public function getConvRuleTitle() {
4741 return $this->mConverter
->getConvRuleTitle();
4745 * Get the compiled plural rules for the language
4747 * @return array Associative array with plural form, and plural rule as key-value pairs
4749 public function getCompiledPluralRules() {
4750 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4751 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4752 if ( !$pluralRules ) {
4753 foreach ( $fallbacks as $fallbackCode ) {
4754 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4755 if ( $pluralRules ) {
4760 return $pluralRules;
4764 * Get the plural rules for the language
4766 * @return array Associative array with plural form number and plural rule as key-value pairs
4768 public function getPluralRules() {
4769 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4770 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4771 if ( !$pluralRules ) {
4772 foreach ( $fallbacks as $fallbackCode ) {
4773 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4774 if ( $pluralRules ) {
4779 return $pluralRules;
4783 * Get the plural rule types for the language
4785 * @return array Associative array with plural form number and plural rule type as key-value pairs
4787 public function getPluralRuleTypes() {
4788 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4789 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4790 if ( !$pluralRuleTypes ) {
4791 foreach ( $fallbacks as $fallbackCode ) {
4792 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4793 if ( $pluralRuleTypes ) {
4798 return $pluralRuleTypes;
4802 * Find the index number of the plural rule appropriate for the given number
4803 * @param int $number
4804 * @return int The index number of the plural rule
4806 public function getPluralRuleIndexNumber( $number ) {
4807 $pluralRules = $this->getCompiledPluralRules();
4808 $form = CLDRPluralRuleEvaluator
::evaluateCompiled( $number, $pluralRules );
4813 * Find the plural rule type appropriate for the given number
4814 * For example, if the language is set to Arabic, getPluralType(5) should
4817 * @param int $number
4818 * @return string The name of the plural rule type, e.g. one, two, few, many
4820 public function getPluralRuleType( $number ) {
4821 $index = $this->getPluralRuleIndexNumber( $number );
4822 $pluralRuleTypes = $this->getPluralRuleTypes();
4823 if ( isset( $pluralRuleTypes[$index] ) ) {
4824 return $pluralRuleTypes[$index];