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";
34 global $wgLanguageNames;
35 require_once( dirname( __FILE__
) . '/Names.php' );
37 if ( function_exists( 'mb_strtoupper' ) ) {
38 mb_internal_encoding( 'UTF-8' );
42 * a fake language converter
52 function __construct( $langobj ) { $this->mLang
= $langobj; }
53 function autoConvertToAllVariants( $text ) { return array( $this->mLang
->getCode() => $text ); }
54 function convert( $t ) { return $t; }
55 function convertTo( $text, $variant ) { return $text; }
56 function convertTitle( $t ) { return $t->getPrefixedText(); }
57 function getVariants() { return array( $this->mLang
->getCode() ); }
58 function getPreferredVariant() { return $this->mLang
->getCode(); }
59 function getDefaultVariant() { return $this->mLang
->getCode(); }
60 function getURLVariant() { return ''; }
61 function getConvRuleTitle() { return false; }
62 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
63 function getExtraHashOptions() { return ''; }
64 function getParsedTitle() { return ''; }
65 function markNoConversion( $text, $noParse = false ) { return $text; }
66 function convertCategoryKey( $key ) { return $key; }
67 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
68 function armourMath( $text ) { return $text; }
72 * Internationalisation code
78 * @var LanguageConverter
82 var $mVariants, $mCode, $mLoaded = false;
83 var $mMagicExtensions = array(), $mMagicHookDone = false;
84 private $mHtmlCode = null;
86 var $dateFormatStrings = array();
87 var $mExtendedSpecialPageAliases;
89 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
92 * ReplacementArray object caches
94 var $transformData = array();
97 * @var LocalisationCache
99 static public $dataCache;
101 static public $mLangObjCache = array();
103 static public $mWeekdayMsgs = array(
104 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
108 static public $mWeekdayAbbrevMsgs = array(
109 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
112 static public $mMonthMsgs = array(
113 'january', 'february', 'march', 'april', 'may_long', 'june',
114 'july', 'august', 'september', 'october', 'november',
117 static public $mMonthGenMsgs = array(
118 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
119 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
122 static public $mMonthAbbrevMsgs = array(
123 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
124 'sep', 'oct', 'nov', 'dec'
127 static public $mIranianCalendarMonthMsgs = array(
128 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
129 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
130 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
131 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
134 static public $mHebrewCalendarMonthMsgs = array(
135 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
136 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
137 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
138 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
139 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
142 static public $mHebrewCalendarMonthGenMsgs = array(
143 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
144 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
145 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
146 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
147 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
150 static public $mHijriCalendarMonthMsgs = array(
151 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
152 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
153 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
154 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
161 static public $durationIntervals = array(
162 'millennia' => 31557600000,
163 'centuries' => 3155760000,
164 'decades' => 315576000,
165 'years' => 31557600, // 86400 * 365.25
174 * Get a cached language object for a given language code
175 * @param $code String
178 static function factory( $code ) {
179 if ( !isset( self
::$mLangObjCache[$code] ) ) {
180 if ( count( self
::$mLangObjCache ) > 10 ) {
181 // Don't keep a billion objects around, that's stupid.
182 self
::$mLangObjCache = array();
184 self
::$mLangObjCache[$code] = self
::newFromCode( $code );
186 return self
::$mLangObjCache[$code];
190 * Create a language object for a given language code
191 * @param $code String
192 * @throws MWException
195 protected static function newFromCode( $code ) {
196 // Protect against path traversal below
197 if ( !Language
::isValidCode( $code )
198 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
200 throw new MWException( "Invalid language code \"$code\"" );
203 if ( !Language
::isValidBuiltInCode( $code ) ) {
204 // It's not possible to customise this code with class files, so
205 // just return a Language object. This is to support uselang= hacks.
206 $lang = new Language
;
207 $lang->setCode( $code );
211 // Check if there is a language class for the code
212 $class = self
::classFromCode( $code );
213 self
::preloadLanguageClass( $class );
214 if ( MWInit
::classExists( $class ) ) {
219 // Keep trying the fallback list until we find an existing class
220 $fallbacks = Language
::getFallbacksFor( $code );
221 foreach ( $fallbacks as $fallbackCode ) {
222 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
223 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
226 $class = self
::classFromCode( $fallbackCode );
227 self
::preloadLanguageClass( $class );
228 if ( MWInit
::classExists( $class ) ) {
229 $lang = Language
::newFromCode( $fallbackCode );
230 $lang->setCode( $code );
235 throw new MWException( "Invalid fallback sequence for language '$code'" );
239 * Returns true if a language code string is of a valid form, whether or
240 * not it exists. This includes codes which are used solely for
241 * customisation via the MediaWiki namespace.
243 * @param $code string
247 public static function isValidCode( $code ) {
249 strcspn( $code, ":/\\\000" ) === strlen( $code )
250 && !preg_match( Title
::getTitleInvalidRegex(), $code );
254 * Returns true if a language code is of a valid form for the purposes of
255 * internal customisation of MediaWiki, via Messages*.php.
257 * @param $code string
259 * @throws MWException
263 public static function isValidBuiltInCode( $code ) {
265 if( !is_string($code) ) {
266 $type = gettype( $code );
267 if( $type === 'object' ) {
268 $addmsg = " of class " . get_class( $code );
272 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
275 return preg_match( '/^[a-z0-9-]+$/i', $code );
280 * @return String Name of the language class
282 public static function classFromCode( $code ) {
283 if ( $code == 'en' ) {
286 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
291 * Includes language class files
293 * @param $class string Name of the language class
295 public static function preloadLanguageClass( $class ) {
298 if ( $class === 'Language' ) {
302 if ( !defined( 'MW_COMPILED' ) ) {
303 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
304 include_once( "$IP/languages/classes/$class.php" );
310 * Get the LocalisationCache instance
312 * @return LocalisationCache
314 public static function getLocalisationCache() {
315 if ( is_null( self
::$dataCache ) ) {
316 global $wgLocalisationCacheConf;
317 $class = $wgLocalisationCacheConf['class'];
318 self
::$dataCache = new $class( $wgLocalisationCacheConf );
320 return self
::$dataCache;
323 function __construct() {
324 $this->mConverter
= new FakeConverter( $this );
325 // Set the code to the name of the descendant
326 if ( get_class( $this ) == 'Language' ) {
329 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
331 self
::getLocalisationCache();
335 * Reduce memory usage
337 function __destruct() {
338 foreach ( $this as $name => $value ) {
339 unset( $this->$name );
344 * Hook which will be called if this is the content language.
345 * Descendants can use this to register hook functions or modify globals
347 function initContLang() { }
350 * Same as getFallbacksFor for current language.
352 * @deprecated in 1.19
354 function getFallbackLanguageCode() {
355 wfDeprecated( __METHOD__
);
356 return self
::getFallbackFor( $this->mCode
);
363 function getFallbackLanguages() {
364 return self
::getFallbacksFor( $this->mCode
);
368 * Exports $wgBookstoreListEn
371 function getBookstoreList() {
372 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
378 public function getNamespaces() {
379 if ( is_null( $this->namespaceNames
) ) {
380 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
382 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
383 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
385 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
387 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
388 if ( $wgMetaNamespaceTalk ) {
389 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
391 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
392 $this->namespaceNames
[NS_PROJECT_TALK
] =
393 $this->fixVariableInNamespace( $talk );
396 # Sometimes a language will be localised but not actually exist on this wiki.
397 foreach ( $this->namespaceNames
as $key => $text ) {
398 if ( !isset( $validNamespaces[$key] ) ) {
399 unset( $this->namespaceNames
[$key] );
403 # The above mixing may leave namespaces out of canonical order.
404 # Re-order by namespace ID number...
405 ksort( $this->namespaceNames
);
407 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
409 return $this->namespaceNames
;
413 * Arbitrarily set all of the namespace names at once. Mainly used for testing
414 * @param $namespaces Array of namespaces (id => name)
416 public function setNamespaces( array $namespaces ) {
417 $this->namespaceNames
= $namespaces;
418 $this->mNamespaceIds
= null;
422 * Resets all of the namespace caches. Mainly used for testing
424 public function resetNamespaces( ) {
425 $this->namespaceNames
= null;
426 $this->mNamespaceIds
= null;
427 $this->namespaceAliases
= null;
431 * A convenience function that returns the same thing as
432 * getNamespaces() except with the array values changed to ' '
433 * where it found '_', useful for producing output to be displayed
434 * e.g. in <select> forms.
438 function getFormattedNamespaces() {
439 $ns = $this->getNamespaces();
440 foreach ( $ns as $k => $v ) {
441 $ns[$k] = strtr( $v, '_', ' ' );
447 * Get a namespace value by key
449 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
450 * echo $mw_ns; // prints 'MediaWiki'
453 * @param $index Int: the array key of the namespace to return
454 * @return mixed, string if the namespace value exists, otherwise false
456 function getNsText( $index ) {
457 $ns = $this->getNamespaces();
458 return isset( $ns[$index] ) ?
$ns[$index] : false;
462 * A convenience function that returns the same thing as
463 * getNsText() except with '_' changed to ' ', useful for
466 * @param $index string
470 function getFormattedNsText( $index ) {
471 $ns = $this->getNsText( $index );
472 return strtr( $ns, '_', ' ' );
476 * Returns gender-dependent namespace alias if available.
477 * @param $index Int: namespace index
478 * @param $gender String: gender key (male, female... )
482 function getGenderNsText( $index, $gender ) {
483 global $wgExtraGenderNamespaces;
485 $ns = $wgExtraGenderNamespaces + self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
486 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
490 * Whether this language makes distinguishes genders for example in
495 function needsGenderDistinction() {
496 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
497 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
498 // $wgExtraGenderNamespaces overrides everything
500 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
501 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
502 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
505 // Check what is in i18n files
506 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
507 return count( $aliases ) > 0;
512 * Get a namespace key by value, case insensitive.
513 * Only matches namespace names for the current language, not the
514 * canonical ones defined in Namespace.php.
516 * @param $text String
517 * @return mixed An integer if $text is a valid value otherwise false
519 function getLocalNsIndex( $text ) {
520 $lctext = $this->lc( $text );
521 $ids = $this->getNamespaceIds();
522 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
528 function getNamespaceAliases() {
529 if ( is_null( $this->namespaceAliases
) ) {
530 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
534 foreach ( $aliases as $name => $index ) {
535 if ( $index === NS_PROJECT_TALK
) {
536 unset( $aliases[$name] );
537 $name = $this->fixVariableInNamespace( $name );
538 $aliases[$name] = $index;
543 global $wgExtraGenderNamespaces;
544 $genders = $wgExtraGenderNamespaces +
(array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
545 foreach ( $genders as $index => $forms ) {
546 foreach ( $forms as $alias ) {
547 $aliases[$alias] = $index;
551 $this->namespaceAliases
= $aliases;
553 return $this->namespaceAliases
;
559 function getNamespaceIds() {
560 if ( is_null( $this->mNamespaceIds
) ) {
561 global $wgNamespaceAliases;
562 # Put namespace names and aliases into a hashtable.
563 # If this is too slow, then we should arrange it so that it is done
564 # before caching. The catch is that at pre-cache time, the above
565 # class-specific fixup hasn't been done.
566 $this->mNamespaceIds
= array();
567 foreach ( $this->getNamespaces() as $index => $name ) {
568 $this->mNamespaceIds
[$this->lc( $name )] = $index;
570 foreach ( $this->getNamespaceAliases() as $name => $index ) {
571 $this->mNamespaceIds
[$this->lc( $name )] = $index;
573 if ( $wgNamespaceAliases ) {
574 foreach ( $wgNamespaceAliases as $name => $index ) {
575 $this->mNamespaceIds
[$this->lc( $name )] = $index;
579 return $this->mNamespaceIds
;
583 * Get a namespace key by value, case insensitive. Canonical namespace
584 * names override custom ones defined for the current language.
586 * @param $text String
587 * @return mixed An integer if $text is a valid value otherwise false
589 function getNsIndex( $text ) {
590 $lctext = $this->lc( $text );
591 $ns = MWNamespace
::getCanonicalIndex( $lctext );
592 if ( $ns !== null ) {
595 $ids = $this->getNamespaceIds();
596 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
600 * short names for language variants used for language conversion links.
602 * @param $code String
603 * @param $usemsg bool Use the "variantname-xyz" message if it exists
606 function getVariantname( $code, $usemsg = true ) {
607 $msg = "variantname-$code";
608 if ( $usemsg && wfMessage( $msg )->exists() ) {
609 return $this->getMessageFromDB( $msg );
611 $name = self
::fetchLanguageName( $code );
613 return $name; # if it's defined as a language name, show that
615 # otherwise, output the language code
621 * @param $name string
624 function specialPage( $name ) {
625 $aliases = $this->getSpecialPageAliases();
626 if ( isset( $aliases[$name][0] ) ) {
627 $name = $aliases[$name][0];
629 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
635 function getQuickbarSettings() {
637 $this->getMessage( 'qbsettings-none' ),
638 $this->getMessage( 'qbsettings-fixedleft' ),
639 $this->getMessage( 'qbsettings-fixedright' ),
640 $this->getMessage( 'qbsettings-floatingleft' ),
641 $this->getMessage( 'qbsettings-floatingright' ),
642 $this->getMessage( 'qbsettings-directionality' )
649 function getDatePreferences() {
650 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
656 function getDateFormats() {
657 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
661 * @return array|string
663 function getDefaultDateFormat() {
664 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
665 if ( $df === 'dmy or mdy' ) {
666 global $wgAmericanDates;
667 return $wgAmericanDates ?
'mdy' : 'dmy';
676 function getDatePreferenceMigrationMap() {
677 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
684 function getImageFile( $image ) {
685 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
691 function getExtraUserToggles() {
692 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
699 function getUserToggle( $tog ) {
700 return $this->getMessageFromDB( "tog-$tog" );
704 * Get native language names, indexed by code.
705 * Only those defined in MediaWiki, no other data like CLDR.
706 * If $customisedOnly is true, only returns codes with a messages file
708 * @param $customisedOnly bool
711 * @deprecated in 1.20, use fetchLanguageNames()
713 public static function getLanguageNames( $customisedOnly = false ) {
714 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
718 * Get translated language names. This is done on best effort and
719 * by default this is exactly the same as Language::getLanguageNames.
720 * The CLDR extension provides translated names.
721 * @param $code String Language code.
722 * @return Array language code => language name
724 * @deprecated in 1.20, use fetchLanguageNames()
726 public static function getTranslatedLanguageNames( $code ) {
727 return self
::fetchLanguageNames( $code, 'all' );
731 * Get an array of language names, indexed by code.
732 * @param $inLanguage null|string: Code of language in which to return the names
733 * Use null for autonyms (native names)
734 * @param $include string:
735 * 'all' all available languages
736 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
737 * 'mwfile' only if the language is in 'mw' *and* has a message file
738 * @return array: language code => language name
741 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
742 global $wgExtraLanguageNames;
743 static $coreLanguageNames;
745 if ( $coreLanguageNames === null ) {
746 include( MWInit
::compiledPath( 'languages/Names.php' ) );
752 # TODO: also include when $inLanguage is null, when this code is more efficient
753 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
756 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
757 foreach ( $mwNames as $mwCode => $mwName ) {
758 # - Prefer own MediaWiki native name when not using the hook
759 # - For other names just add if not added through the hook
760 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
761 $names[$mwCode] = $mwName;
765 if ( $include === 'all' ) {
770 $coreCodes = array_keys( $mwNames );
771 foreach( $coreCodes as $coreCode ) {
772 $returnMw[$coreCode] = $names[$coreCode];
775 if( $include === 'mwfile' ) {
776 $namesMwFile = array();
777 # We do this using a foreach over the codes instead of a directory
778 # loop so that messages files in extensions will work correctly.
779 foreach ( $returnMw as $code => $value ) {
780 if ( is_readable( self
::getMessagesFileName( $code ) ) ) {
781 $namesMwFile[$code] = $names[$code];
786 # 'mw' option; default if it's not one of the other two options (all/mwfile)
791 * @param $code string: The code of the language for which to get the name
792 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
793 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
794 * @return string: Language name or empty
797 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
798 $array = self
::fetchLanguageNames( $inLanguage, $include );
799 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
803 * Get a message from the MediaWiki namespace.
805 * @param $msg String: message name
808 function getMessageFromDB( $msg ) {
809 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
813 * Get the native language name of $code.
814 * Only if defined in MediaWiki, no other data like CLDR.
815 * @param $code string
817 * @deprecated in 1.20, use fetchLanguageName()
819 function getLanguageName( $code ) {
820 return self
::fetchLanguageName( $code );
827 function getMonthName( $key ) {
828 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
834 function getMonthNamesArray() {
835 $monthNames = array( '' );
836 for ( $i = 1; $i < 13; $i++
) {
837 $monthNames[] = $this->getMonthName( $i );
846 function getMonthNameGen( $key ) {
847 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
854 function getMonthAbbreviation( $key ) {
855 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
861 function getMonthAbbreviationsArray() {
862 $monthNames = array( '' );
863 for ( $i = 1; $i < 13; $i++
) {
864 $monthNames[] = $this->getMonthAbbreviation( $i );
873 function getWeekdayName( $key ) {
874 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
881 function getWeekdayAbbreviation( $key ) {
882 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
889 function getIranianCalendarMonthName( $key ) {
890 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
897 function getHebrewCalendarMonthName( $key ) {
898 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
905 function getHebrewCalendarMonthNameGen( $key ) {
906 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
913 function getHijriCalendarMonthName( $key ) {
914 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
918 * This is a workalike of PHP's date() function, but with better
919 * internationalisation, a reduced set of format characters, and a better
922 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
923 * PHP manual for definitions. There are a number of extensions, which
926 * xn Do not translate digits of the next numeric format character
927 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
928 * xr Use roman numerals for the next numeric format character
929 * xh Use hebrew numerals for the next numeric format character
931 * xg Genitive month name
933 * xij j (day number) in Iranian calendar
934 * xiF F (month name) in Iranian calendar
935 * xin n (month number) in Iranian calendar
936 * xiy y (two digit year) in Iranian calendar
937 * xiY Y (full year) in Iranian calendar
939 * xjj j (day number) in Hebrew calendar
940 * xjF F (month name) in Hebrew calendar
941 * xjt t (days in month) in Hebrew calendar
942 * xjx xg (genitive month name) in Hebrew calendar
943 * xjn n (month number) in Hebrew calendar
944 * xjY Y (full year) in Hebrew calendar
946 * xmj j (day number) in Hijri calendar
947 * xmF F (month name) in Hijri calendar
948 * xmn n (month number) in Hijri calendar
949 * xmY Y (full year) in Hijri calendar
951 * xkY Y (full year) in Thai solar calendar. Months and days are
952 * identical to the Gregorian calendar
953 * xoY Y (full year) in Minguo calendar or Juche year.
954 * Months and days are identical to the
956 * xtY Y (full year) in Japanese nengo. Months and days are
957 * identical to the Gregorian calendar
959 * Characters enclosed in double quotes will be considered literal (with
960 * the quotes themselves removed). Unmatched quotes will be considered
961 * literal quotes. Example:
963 * "The month is" F => The month is January
966 * Backslash escaping is also supported.
968 * Input timestamp is assumed to be pre-normalized to the desired local
971 * @param $format String
972 * @param $ts String: 14-character timestamp
975 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
979 function sprintfDate( $format, $ts ) {
992 for ( $p = 0; $p < strlen( $format ); $p++
) {
995 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
996 $code .= $format[++
$p];
999 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
1000 $code .= $format[++
$p];
1011 $rawToggle = !$rawToggle;
1020 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1023 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
1024 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1027 $num = substr( $ts, 6, 2 );
1030 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
1031 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) +
1 );
1034 $num = intval( substr( $ts, 6, 2 ) );
1038 $iranian = self
::tsToIranian( $ts );
1044 $hijri = self
::tsToHijri( $ts );
1050 $hebrew = self
::tsToHebrew( $ts );
1056 $unix = wfTimestamp( TS_UNIX
, $ts );
1058 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) +
1 );
1062 $unix = wfTimestamp( TS_UNIX
, $ts );
1064 $w = gmdate( 'w', $unix );
1069 $unix = wfTimestamp( TS_UNIX
, $ts );
1071 $num = gmdate( 'w', $unix );
1075 $unix = wfTimestamp( TS_UNIX
, $ts );
1077 $num = gmdate( 'z', $unix );
1081 $unix = wfTimestamp( TS_UNIX
, $ts );
1083 $num = gmdate( 'W', $unix );
1086 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1090 $iranian = self
::tsToIranian( $ts );
1092 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1096 $hijri = self
::tsToHijri( $ts );
1098 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1102 $hebrew = self
::tsToHebrew( $ts );
1104 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1107 $num = substr( $ts, 4, 2 );
1110 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1113 $num = intval( substr( $ts, 4, 2 ) );
1117 $iranian = self
::tsToIranian( $ts );
1123 $hijri = self
::tsToHijri ( $ts );
1129 $hebrew = self
::tsToHebrew( $ts );
1135 $unix = wfTimestamp( TS_UNIX
, $ts );
1137 $num = gmdate( 't', $unix );
1141 $hebrew = self
::tsToHebrew( $ts );
1147 $unix = wfTimestamp( TS_UNIX
, $ts );
1149 $num = gmdate( 'L', $unix );
1153 $unix = wfTimestamp( TS_UNIX
, $ts );
1155 $num = gmdate( 'o', $unix );
1158 $num = substr( $ts, 0, 4 );
1162 $iranian = self
::tsToIranian( $ts );
1168 $hijri = self
::tsToHijri( $ts );
1174 $hebrew = self
::tsToHebrew( $ts );
1180 $thai = self
::tsToYear( $ts, 'thai' );
1186 $minguo = self
::tsToYear( $ts, 'minguo' );
1192 $tenno = self
::tsToYear( $ts, 'tenno' );
1197 $num = substr( $ts, 2, 2 );
1201 $iranian = self
::tsToIranian( $ts );
1203 $num = substr( $iranian[0], -2 );
1206 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1209 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1212 $h = substr( $ts, 8, 2 );
1213 $num = $h %
12 ?
$h %
12 : 12;
1216 $num = intval( substr( $ts, 8, 2 ) );
1219 $h = substr( $ts, 8, 2 );
1220 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1223 $num = substr( $ts, 8, 2 );
1226 $num = substr( $ts, 10, 2 );
1229 $num = substr( $ts, 12, 2 );
1233 $unix = wfTimestamp( TS_UNIX
, $ts );
1235 $s .= gmdate( 'c', $unix );
1239 $unix = wfTimestamp( TS_UNIX
, $ts );
1241 $s .= gmdate( 'r', $unix );
1245 $unix = wfTimestamp( TS_UNIX
, $ts );
1250 # Backslash escaping
1251 if ( $p < strlen( $format ) - 1 ) {
1252 $s .= $format[++
$p];
1259 if ( $p < strlen( $format ) - 1 ) {
1260 $endQuote = strpos( $format, '"', $p +
1 );
1261 if ( $endQuote === false ) {
1262 # No terminating quote, assume literal "
1265 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1269 # Quote at end of string, assume literal "
1276 if ( $num !== false ) {
1277 if ( $rawToggle ||
$raw ) {
1280 } elseif ( $roman ) {
1281 $s .= self
::romanNumeral( $num );
1283 } elseif ( $hebrewNum ) {
1284 $s .= self
::hebrewNumeral( $num );
1287 $s .= $this->formatNum( $num, true );
1294 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1295 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1298 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1299 * Gregorian dates to Iranian dates. Originally written in C, it
1300 * is released under the terms of GNU Lesser General Public
1301 * License. Conversion to PHP was performed by Niklas Laxström.
1303 * Link: http://www.farsiweb.info/jalali/jalali.c
1309 private static function tsToIranian( $ts ) {
1310 $gy = substr( $ts, 0, 4 ) -1600;
1311 $gm = substr( $ts, 4, 2 ) -1;
1312 $gd = substr( $ts, 6, 2 ) -1;
1314 # Days passed from the beginning (including leap years)
1316 +
floor( ( $gy +
3 ) / 4 )
1317 - floor( ( $gy +
99 ) / 100 )
1318 +
floor( ( $gy +
399 ) / 400 );
1320 // Add days of the past months of this year
1321 for ( $i = 0; $i < $gm; $i++
) {
1322 $gDayNo +
= self
::$GREG_DAYS[$i];
1326 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1330 // Days passed in current month
1331 $gDayNo +
= (int)$gd;
1333 $jDayNo = $gDayNo - 79;
1335 $jNp = floor( $jDayNo / 12053 );
1338 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1341 if ( $jDayNo >= 366 ) {
1342 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1343 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1346 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1347 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1353 return array( $jy, $jm, $jd );
1357 * Converting Gregorian dates to Hijri dates.
1359 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1361 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1367 private static function tsToHijri( $ts ) {
1368 $year = substr( $ts, 0, 4 );
1369 $month = substr( $ts, 4, 2 );
1370 $day = substr( $ts, 6, 2 );
1378 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1379 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1382 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1383 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1384 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1387 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1388 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1391 $zl = $zjd -1948440 +
10632;
1392 $zn = (int)( ( $zl - 1 ) / 10631 );
1393 $zl = $zl - 10631 * $zn +
354;
1394 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1395 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1396 $zm = (int)( ( 24 * $zl ) / 709 );
1397 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1398 $zy = 30 * $zn +
$zj - 30;
1400 return array( $zy, $zm, $zd );
1404 * Converting Gregorian dates to Hebrew dates.
1406 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1407 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1408 * to translate the relevant functions into PHP and release them under
1411 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1412 * and Adar II is 14. In a non-leap year, Adar is 6.
1418 private static function tsToHebrew( $ts ) {
1420 $year = substr( $ts, 0, 4 );
1421 $month = substr( $ts, 4, 2 );
1422 $day = substr( $ts, 6, 2 );
1424 # Calculate Hebrew year
1425 $hebrewYear = $year +
3760;
1427 # Month number when September = 1, August = 12
1429 if ( $month > 12 ) {
1436 # Calculate day of year from 1 September
1438 for ( $i = 1; $i < $month; $i++
) {
1442 # Check if the year is leap
1443 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1446 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1453 # Calculate the start of the Hebrew year
1454 $start = self
::hebrewYearStart( $hebrewYear );
1456 # Calculate next year's start
1457 if ( $dayOfYear <= $start ) {
1458 # Day is before the start of the year - it is the previous year
1460 $nextStart = $start;
1464 # Add days since previous year's 1 September
1466 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1470 # Start of the new (previous) year
1471 $start = self
::hebrewYearStart( $hebrewYear );
1474 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1477 # Calculate Hebrew day of year
1478 $hebrewDayOfYear = $dayOfYear - $start;
1480 # Difference between year's days
1481 $diff = $nextStart - $start;
1482 # Add 12 (or 13 for leap years) days to ignore the difference between
1483 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1484 # difference is only about the year type
1485 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1491 # Check the year pattern, and is leap year
1492 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1493 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1494 # and non-leap years
1495 $yearPattern = $diff %
30;
1496 # Check if leap year
1497 $isLeap = $diff >= 30;
1499 # Calculate day in the month from number of day in the Hebrew year
1500 # Don't check Adar - if the day is not in Adar, we will stop before;
1501 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1502 $hebrewDay = $hebrewDayOfYear;
1505 while ( $hebrewMonth <= 12 ) {
1506 # Calculate days in this month
1507 if ( $isLeap && $hebrewMonth == 6 ) {
1508 # Adar in a leap year
1510 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1512 if ( $hebrewDay <= $days ) {
1516 # Subtract the days of Adar I
1517 $hebrewDay -= $days;
1520 if ( $hebrewDay <= $days ) {
1526 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1527 # Cheshvan in a complete year (otherwise as the rule below)
1529 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1530 # Kislev in an incomplete year (otherwise as the rule below)
1533 # Odd months have 30 days, even have 29
1534 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1536 if ( $hebrewDay <= $days ) {
1537 # In the current month
1540 # Subtract the days of the current month
1541 $hebrewDay -= $days;
1542 # Try in the next month
1547 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1551 * This calculates the Hebrew year start, as days since 1 September.
1552 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1553 * Used for Hebrew date.
1559 private static function hebrewYearStart( $year ) {
1560 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1561 $b = intval( ( $year - 1 ) %
4 );
1562 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1566 $Mar = intval( $m );
1572 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1573 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1575 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1577 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1581 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1586 * Algorithm to convert Gregorian dates to Thai solar dates,
1587 * Minguo dates or Minguo dates.
1589 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1590 * http://en.wikipedia.org/wiki/Minguo_calendar
1591 * http://en.wikipedia.org/wiki/Japanese_era_name
1593 * @param $ts String: 14-character timestamp
1594 * @param $cName String: calender name
1595 * @return Array: converted year, month, day
1597 private static function tsToYear( $ts, $cName ) {
1598 $gy = substr( $ts, 0, 4 );
1599 $gm = substr( $ts, 4, 2 );
1600 $gd = substr( $ts, 6, 2 );
1602 if ( !strcmp( $cName, 'thai' ) ) {
1604 # Add 543 years to the Gregorian calendar
1605 # Months and days are identical
1606 $gy_offset = $gy +
543;
1607 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1609 # Deduct 1911 years from the Gregorian calendar
1610 # Months and days are identical
1611 $gy_offset = $gy - 1911;
1612 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1613 # Nengō dates up to Meiji period
1614 # Deduct years from the Gregorian calendar
1615 # depending on the nengo periods
1616 # Months and days are identical
1617 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1619 $gy_gannen = $gy - 1868 +
1;
1620 $gy_offset = $gy_gannen;
1621 if ( $gy_gannen == 1 ) {
1624 $gy_offset = '明治' . $gy_offset;
1626 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1627 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1628 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1629 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1630 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1634 $gy_gannen = $gy - 1912 +
1;
1635 $gy_offset = $gy_gannen;
1636 if ( $gy_gannen == 1 ) {
1639 $gy_offset = '大正' . $gy_offset;
1641 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1642 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1643 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1647 $gy_gannen = $gy - 1926 +
1;
1648 $gy_offset = $gy_gannen;
1649 if ( $gy_gannen == 1 ) {
1652 $gy_offset = '昭和' . $gy_offset;
1655 $gy_gannen = $gy - 1989 +
1;
1656 $gy_offset = $gy_gannen;
1657 if ( $gy_gannen == 1 ) {
1660 $gy_offset = '平成' . $gy_offset;
1666 return array( $gy_offset, $gm, $gd );
1670 * Roman number formatting up to 3000
1676 static function romanNumeral( $num ) {
1677 static $table = array(
1678 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1679 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1680 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1681 array( '', 'M', 'MM', 'MMM' )
1684 $num = intval( $num );
1685 if ( $num > 3000 ||
$num <= 0 ) {
1690 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1691 if ( $num >= $pow10 ) {
1692 $s .= $table[$i][(int)floor( $num / $pow10 )];
1694 $num = $num %
$pow10;
1700 * Hebrew Gematria number formatting up to 9999
1706 static function hebrewNumeral( $num ) {
1707 static $table = array(
1708 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1709 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1710 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1711 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1714 $num = intval( $num );
1715 if ( $num > 9999 ||
$num <= 0 ) {
1720 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1721 if ( $num >= $pow10 ) {
1722 if ( $num == 15 ||
$num == 16 ) {
1723 $s .= $table[0][9] . $table[0][$num - 9];
1726 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1727 if ( $pow10 == 1000 ) {
1732 $num = $num %
$pow10;
1734 if ( strlen( $s ) == 2 ) {
1737 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1738 $str .= substr( $s, strlen( $s ) - 2, 2 );
1740 $start = substr( $str, 0, strlen( $str ) - 2 );
1741 $end = substr( $str, strlen( $str ) - 2 );
1744 $str = $start . 'ך';
1747 $str = $start . 'ם';
1750 $str = $start . 'ן';
1753 $str = $start . 'ף';
1756 $str = $start . 'ץ';
1763 * Used by date() and time() to adjust the time output.
1765 * @param $ts Int the time in date('YmdHis') format
1766 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1767 * get user timecorrection setting)
1770 function userAdjust( $ts, $tz = false ) {
1771 global $wgUser, $wgLocalTZoffset;
1773 if ( $tz === false ) {
1774 $tz = $wgUser->getOption( 'timecorrection' );
1777 $data = explode( '|', $tz, 3 );
1779 if ( $data[0] == 'ZoneInfo' ) {
1780 wfSuppressWarnings();
1781 $userTZ = timezone_open( $data[2] );
1782 wfRestoreWarnings();
1783 if ( $userTZ !== false ) {
1784 $date = date_create( $ts, timezone_open( 'UTC' ) );
1785 date_timezone_set( $date, $userTZ );
1786 $date = date_format( $date, 'YmdHis' );
1789 # Unrecognized timezone, default to 'Offset' with the stored offset.
1790 $data[0] = 'Offset';
1794 if ( $data[0] == 'System' ||
$tz == '' ) {
1795 # Global offset in minutes.
1796 if ( isset( $wgLocalTZoffset ) ) {
1797 $minDiff = $wgLocalTZoffset;
1799 } elseif ( $data[0] == 'Offset' ) {
1800 $minDiff = intval( $data[1] );
1802 $data = explode( ':', $tz );
1803 if ( count( $data ) == 2 ) {
1804 $data[0] = intval( $data[0] );
1805 $data[1] = intval( $data[1] );
1806 $minDiff = abs( $data[0] ) * 60 +
$data[1];
1807 if ( $data[0] < 0 ) {
1808 $minDiff = -$minDiff;
1811 $minDiff = intval( $data[0] ) * 60;
1815 # No difference ? Return time unchanged
1816 if ( 0 == $minDiff ) {
1820 wfSuppressWarnings(); // E_STRICT system time bitching
1821 # Generate an adjusted date; take advantage of the fact that mktime
1822 # will normalize out-of-range values so we don't have to split $minDiff
1823 # into hours and minutes.
1825 (int)substr( $ts, 8, 2 ) ), # Hours
1826 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
1827 (int)substr( $ts, 12, 2 ), # Seconds
1828 (int)substr( $ts, 4, 2 ), # Month
1829 (int)substr( $ts, 6, 2 ), # Day
1830 (int)substr( $ts, 0, 4 ) ); # Year
1832 $date = date( 'YmdHis', $t );
1833 wfRestoreWarnings();
1839 * This is meant to be used by time(), date(), and timeanddate() to get
1840 * the date preference they're supposed to use, it should be used in
1844 * function timeanddate([...], $format = true) {
1845 * $datePreference = $this->dateFormat($format);
1850 * @param $usePrefs Mixed: if true, the user's preference is used
1851 * if false, the site/language default is used
1852 * if int/string, assumed to be a format.
1855 function dateFormat( $usePrefs = true ) {
1858 if ( is_bool( $usePrefs ) ) {
1860 $datePreference = $wgUser->getDatePreference();
1862 $datePreference = (string)User
::getDefaultOption( 'date' );
1865 $datePreference = (string)$usePrefs;
1869 if ( $datePreference == '' ) {
1873 return $datePreference;
1877 * Get a format string for a given type and preference
1878 * @param $type string May be date, time or both
1879 * @param $pref string The format name as it appears in Messages*.php
1883 function getDateFormatString( $type, $pref ) {
1884 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
1885 if ( $pref == 'default' ) {
1886 $pref = $this->getDefaultDateFormat();
1887 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1889 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1890 if ( is_null( $df ) ) {
1891 $pref = $this->getDefaultDateFormat();
1892 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1895 $this->dateFormatStrings
[$type][$pref] = $df;
1897 return $this->dateFormatStrings
[$type][$pref];
1901 * @param $ts Mixed: the time format which needs to be turned into a
1902 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1903 * @param $adj Bool: whether to adjust the time output according to the
1904 * user configured offset ($timecorrection)
1905 * @param $format Mixed: true to use user's date format preference
1906 * @param $timecorrection String|bool the time offset as returned by
1907 * validateTimeZone() in Special:Preferences
1910 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1911 $ts = wfTimestamp( TS_MW
, $ts );
1913 $ts = $this->userAdjust( $ts, $timecorrection );
1915 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1916 return $this->sprintfDate( $df, $ts );
1920 * @param $ts Mixed: the time format which needs to be turned into a
1921 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1922 * @param $adj Bool: whether to adjust the time output according to the
1923 * user configured offset ($timecorrection)
1924 * @param $format Mixed: true to use user's date format preference
1925 * @param $timecorrection String|bool the time offset as returned by
1926 * validateTimeZone() in Special:Preferences
1929 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1930 $ts = wfTimestamp( TS_MW
, $ts );
1932 $ts = $this->userAdjust( $ts, $timecorrection );
1934 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1935 return $this->sprintfDate( $df, $ts );
1939 * @param $ts Mixed: the time format which needs to be turned into a
1940 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1941 * @param $adj Bool: whether to adjust the time output according to the
1942 * user configured offset ($timecorrection)
1943 * @param $format Mixed: what format to return, if it's false output the
1944 * default one (default true)
1945 * @param $timecorrection String|bool the time offset as returned by
1946 * validateTimeZone() in Special:Preferences
1949 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1950 $ts = wfTimestamp( TS_MW
, $ts );
1952 $ts = $this->userAdjust( $ts, $timecorrection );
1954 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1955 return $this->sprintfDate( $df, $ts );
1959 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
1963 * @param integer $seconds The amount of seconds.
1964 * @param array $chosenIntervals The intervals to enable.
1968 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
1969 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
1971 $segments = array();
1973 foreach ( $intervals as $intervalName => $intervalValue ) {
1974 $message = new Message( 'duration-' . $intervalName, array( $intervalValue ) );
1975 $segments[] = $message->inLanguage( $this )->escaped();
1978 return $this->listToText( $segments );
1982 * Takes a number of seconds and returns an array with a set of corresponding intervals.
1983 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
1987 * @param integer $seconds The amount of seconds.
1988 * @param array $chosenIntervals The intervals to enable.
1992 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
1993 if ( empty( $chosenIntervals ) ) {
1994 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
1997 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
1998 $sortedNames = array_keys( $intervals );
1999 $smallestInterval = array_pop( $sortedNames );
2001 $segments = array();
2003 foreach ( $intervals as $name => $length ) {
2004 $value = floor( $seconds / $length );
2006 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2007 $seconds -= $value * $length;
2008 $segments[$name] = $value;
2016 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2018 * @param $type String: can be 'date', 'time' or 'both'
2019 * @param $ts Mixed: the time format which needs to be turned into a
2020 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2021 * @param $user User object used to get preferences for timezone and format
2022 * @param $options Array, can contain the following keys:
2023 * - 'timecorrection': time correction, can have the following values:
2024 * - true: use user's preference
2025 * - false: don't use time correction
2026 * - integer: value of time correction in minutes
2027 * - 'format': format to use, can have the following values:
2028 * - true: use user's preference
2029 * - false: use default preference
2030 * - string: format to use
2034 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2035 $ts = wfTimestamp( TS_MW
, $ts );
2036 $options +
= array( 'timecorrection' => true, 'format' => true );
2037 if ( $options['timecorrection'] !== false ) {
2038 if ( $options['timecorrection'] === true ) {
2039 $offset = $user->getOption( 'timecorrection' );
2041 $offset = $options['timecorrection'];
2043 $ts = $this->userAdjust( $ts, $offset );
2045 if ( $options['format'] === true ) {
2046 $format = $user->getDatePreference();
2048 $format = $options['format'];
2050 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2051 return $this->sprintfDate( $df, $ts );
2055 * Get the formatted date for the given timestamp and formatted for
2058 * @param $ts Mixed: the time format which needs to be turned into a
2059 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2060 * @param $user User object used to get preferences for timezone and format
2061 * @param $options Array, can contain the following keys:
2062 * - 'timecorrection': time correction, can have the following values:
2063 * - true: use user's preference
2064 * - false: don't use time correction
2065 * - integer: value of time correction in minutes
2066 * - 'format': format to use, can have the following values:
2067 * - true: use user's preference
2068 * - false: use default preference
2069 * - string: format to use
2073 public function userDate( $ts, User
$user, array $options = array() ) {
2074 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2078 * Get the formatted time for the given timestamp and formatted for
2081 * @param $ts Mixed: the time format which needs to be turned into a
2082 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2083 * @param $user User object used to get preferences for timezone and format
2084 * @param $options Array, can contain the following keys:
2085 * - 'timecorrection': time correction, can have the following values:
2086 * - true: use user's preference
2087 * - false: don't use time correction
2088 * - integer: value of time correction in minutes
2089 * - 'format': format to use, can have the following values:
2090 * - true: use user's preference
2091 * - false: use default preference
2092 * - string: format to use
2096 public function userTime( $ts, User
$user, array $options = array() ) {
2097 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2101 * Get the formatted date and time for the given timestamp and formatted for
2104 * @param $ts Mixed: the time format which needs to be turned into a
2105 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2106 * @param $user User object used to get preferences for timezone and format
2107 * @param $options Array, can contain the following keys:
2108 * - 'timecorrection': time correction, can have the following values:
2109 * - true: use user's preference
2110 * - false: don't use time correction
2111 * - integer: value of time correction in minutes
2112 * - 'format': format to use, can have the following values:
2113 * - true: use user's preference
2114 * - false: use default preference
2115 * - string: format to use
2119 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2120 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2124 * @param $key string
2125 * @return array|null
2127 function getMessage( $key ) {
2128 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2134 function getAllMessages() {
2135 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2144 function iconv( $in, $out, $string ) {
2145 # This is a wrapper for iconv in all languages except esperanto,
2146 # which does some nasty x-conversions beforehand
2148 # Even with //IGNORE iconv can whine about illegal characters in
2149 # *input* string. We just ignore those too.
2150 # REF: http://bugs.php.net/bug.php?id=37166
2151 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2152 wfSuppressWarnings();
2153 $text = iconv( $in, $out . '//IGNORE', $string );
2154 wfRestoreWarnings();
2158 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2161 * @param $matches array
2162 * @return mixed|string
2164 function ucwordbreaksCallbackAscii( $matches ) {
2165 return $this->ucfirst( $matches[1] );
2169 * @param $matches array
2172 function ucwordbreaksCallbackMB( $matches ) {
2173 return mb_strtoupper( $matches[0] );
2177 * @param $matches array
2180 function ucCallback( $matches ) {
2181 list( $wikiUpperChars ) = self
::getCaseMaps();
2182 return strtr( $matches[1], $wikiUpperChars );
2186 * @param $matches array
2189 function lcCallback( $matches ) {
2190 list( , $wikiLowerChars ) = self
::getCaseMaps();
2191 return strtr( $matches[1], $wikiLowerChars );
2195 * @param $matches array
2198 function ucwordsCallbackMB( $matches ) {
2199 return mb_strtoupper( $matches[0] );
2203 * @param $matches array
2206 function ucwordsCallbackWiki( $matches ) {
2207 list( $wikiUpperChars ) = self
::getCaseMaps();
2208 return strtr( $matches[0], $wikiUpperChars );
2212 * Make a string's first character uppercase
2214 * @param $str string
2218 function ucfirst( $str ) {
2220 if ( $o < 96 ) { // if already uppercase...
2222 } elseif ( $o < 128 ) {
2223 return ucfirst( $str ); // use PHP's ucfirst()
2225 // fall back to more complex logic in case of multibyte strings
2226 return $this->uc( $str, true );
2231 * Convert a string to uppercase
2233 * @param $str string
2234 * @param $first bool
2238 function uc( $str, $first = false ) {
2239 if ( function_exists( 'mb_strtoupper' ) ) {
2241 if ( $this->isMultibyte( $str ) ) {
2242 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2244 return ucfirst( $str );
2247 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2250 if ( $this->isMultibyte( $str ) ) {
2251 $x = $first ?
'^' : '';
2252 return preg_replace_callback(
2253 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2254 array( $this, 'ucCallback' ),
2258 return $first ?
ucfirst( $str ) : strtoupper( $str );
2264 * @param $str string
2265 * @return mixed|string
2267 function lcfirst( $str ) {
2270 return strval( $str );
2271 } elseif ( $o >= 128 ) {
2272 return $this->lc( $str, true );
2273 } elseif ( $o > 96 ) {
2276 $str[0] = strtolower( $str[0] );
2282 * @param $str string
2283 * @param $first bool
2284 * @return mixed|string
2286 function lc( $str, $first = false ) {
2287 if ( function_exists( 'mb_strtolower' ) ) {
2289 if ( $this->isMultibyte( $str ) ) {
2290 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2292 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2295 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2298 if ( $this->isMultibyte( $str ) ) {
2299 $x = $first ?
'^' : '';
2300 return preg_replace_callback(
2301 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2302 array( $this, 'lcCallback' ),
2306 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2312 * @param $str string
2315 function isMultibyte( $str ) {
2316 return (bool)preg_match( '/[\x80-\xff]/', $str );
2320 * @param $str string
2321 * @return mixed|string
2323 function ucwords( $str ) {
2324 if ( $this->isMultibyte( $str ) ) {
2325 $str = $this->lc( $str );
2327 // regexp to find first letter in each word (i.e. after each space)
2328 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2330 // function to use to capitalize a single char
2331 if ( function_exists( 'mb_strtoupper' ) ) {
2332 return preg_replace_callback(
2334 array( $this, 'ucwordsCallbackMB' ),
2338 return preg_replace_callback(
2340 array( $this, 'ucwordsCallbackWiki' ),
2345 return ucwords( strtolower( $str ) );
2350 * capitalize words at word breaks
2352 * @param $str string
2355 function ucwordbreaks( $str ) {
2356 if ( $this->isMultibyte( $str ) ) {
2357 $str = $this->lc( $str );
2359 // since \b doesn't work for UTF-8, we explicitely define word break chars
2360 $breaks = "[ \-\(\)\}\{\.,\?!]";
2362 // find first letter after word break
2363 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2365 if ( function_exists( 'mb_strtoupper' ) ) {
2366 return preg_replace_callback(
2368 array( $this, 'ucwordbreaksCallbackMB' ),
2372 return preg_replace_callback(
2374 array( $this, 'ucwordsCallbackWiki' ),
2379 return preg_replace_callback(
2380 '/\b([\w\x80-\xff]+)\b/',
2381 array( $this, 'ucwordbreaksCallbackAscii' ),
2388 * Return a case-folded representation of $s
2390 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2391 * and $s2 are the same except for the case of their characters. It is not
2392 * necessary for the value returned to make sense when displayed.
2394 * Do *not* perform any other normalisation in this function. If a caller
2395 * uses this function when it should be using a more general normalisation
2396 * function, then fix the caller.
2402 function caseFold( $s ) {
2403 return $this->uc( $s );
2410 function checkTitleEncoding( $s ) {
2411 if ( is_array( $s ) ) {
2412 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2414 # Check for non-UTF-8 URLs
2415 $ishigh = preg_match( '/[\x80-\xff]/', $s );
2420 if ( function_exists( 'mb_check_encoding' ) ) {
2421 $isutf8 = mb_check_encoding( $s, 'UTF-8' );
2423 $isutf8 = preg_match( '/^(?>[\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2424 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
2430 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2436 function fallback8bitEncoding() {
2437 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2441 * Most writing systems use whitespace to break up words.
2442 * Some languages such as Chinese don't conventionally do this,
2443 * which requires special handling when breaking up words for
2448 function hasWordBreaks() {
2453 * Some languages such as Chinese require word segmentation,
2454 * Specify such segmentation when overridden in derived class.
2456 * @param $string String
2459 function segmentByWord( $string ) {
2464 * Some languages have special punctuation need to be normalized.
2465 * Make such changes here.
2467 * @param $string String
2470 function normalizeForSearch( $string ) {
2471 return self
::convertDoubleWidth( $string );
2475 * convert double-width roman characters to single-width.
2476 * range: ff00-ff5f ~= 0020-007f
2478 * @param $string string
2482 protected static function convertDoubleWidth( $string ) {
2483 static $full = null;
2484 static $half = null;
2486 if ( $full === null ) {
2487 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2488 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2489 $full = str_split( $fullWidth, 3 );
2490 $half = str_split( $halfWidth );
2493 $string = str_replace( $full, $half, $string );
2498 * @param $string string
2499 * @param $pattern string
2502 protected static function insertSpace( $string, $pattern ) {
2503 $string = preg_replace( $pattern, " $1 ", $string );
2504 $string = preg_replace( '/ +/', ' ', $string );
2509 * @param $termsArray array
2512 function convertForSearchResult( $termsArray ) {
2513 # some languages, e.g. Chinese, need to do a conversion
2514 # in order for search results to be displayed correctly
2519 * Get the first character of a string.
2524 function firstChar( $s ) {
2527 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2528 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2533 if ( isset( $matches[1] ) ) {
2534 if ( strlen( $matches[1] ) != 3 ) {
2538 // Break down Hangul syllables to grab the first jamo
2539 $code = utf8ToCodepoint( $matches[1] );
2540 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2542 } elseif ( $code < 0xb098 ) {
2543 return "\xe3\x84\xb1";
2544 } elseif ( $code < 0xb2e4 ) {
2545 return "\xe3\x84\xb4";
2546 } elseif ( $code < 0xb77c ) {
2547 return "\xe3\x84\xb7";
2548 } elseif ( $code < 0xb9c8 ) {
2549 return "\xe3\x84\xb9";
2550 } elseif ( $code < 0xbc14 ) {
2551 return "\xe3\x85\x81";
2552 } elseif ( $code < 0xc0ac ) {
2553 return "\xe3\x85\x82";
2554 } elseif ( $code < 0xc544 ) {
2555 return "\xe3\x85\x85";
2556 } elseif ( $code < 0xc790 ) {
2557 return "\xe3\x85\x87";
2558 } elseif ( $code < 0xcc28 ) {
2559 return "\xe3\x85\x88";
2560 } elseif ( $code < 0xce74 ) {
2561 return "\xe3\x85\x8a";
2562 } elseif ( $code < 0xd0c0 ) {
2563 return "\xe3\x85\x8b";
2564 } elseif ( $code < 0xd30c ) {
2565 return "\xe3\x85\x8c";
2566 } elseif ( $code < 0xd558 ) {
2567 return "\xe3\x85\x8d";
2569 return "\xe3\x85\x8e";
2576 function initEncoding() {
2577 # Some languages may have an alternate char encoding option
2578 # (Esperanto X-coding, Japanese furigana conversion, etc)
2579 # If this language is used as the primary content language,
2580 # an override to the defaults can be set here on startup.
2587 function recodeForEdit( $s ) {
2588 # For some languages we'll want to explicitly specify
2589 # which characters make it into the edit box raw
2590 # or are converted in some way or another.
2591 global $wgEditEncoding;
2592 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2595 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2603 function recodeInput( $s ) {
2604 # Take the previous into account.
2605 global $wgEditEncoding;
2606 if ( $wgEditEncoding != '' ) {
2607 $enc = $wgEditEncoding;
2611 if ( $enc == 'UTF-8' ) {
2614 return $this->iconv( $enc, 'UTF-8', $s );
2619 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2620 * also cleans up certain backwards-compatible sequences, converting them
2621 * to the modern Unicode equivalent.
2623 * This is language-specific for performance reasons only.
2629 function normalize( $s ) {
2630 global $wgAllUnicodeFixes;
2631 $s = UtfNormal
::cleanUp( $s );
2632 if ( $wgAllUnicodeFixes ) {
2633 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2634 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2641 * Transform a string using serialized data stored in the given file (which
2642 * must be in the serialized subdirectory of $IP). The file contains pairs
2643 * mapping source characters to destination characters.
2645 * The data is cached in process memory. This will go faster if you have the
2646 * FastStringSearch extension.
2648 * @param $file string
2649 * @param $string string
2651 * @throws MWException
2654 function transformUsingPairFile( $file, $string ) {
2655 if ( !isset( $this->transformData
[$file] ) ) {
2656 $data = wfGetPrecompiledData( $file );
2657 if ( $data === false ) {
2658 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
2660 $this->transformData
[$file] = new ReplacementArray( $data );
2662 return $this->transformData
[$file]->replace( $string );
2666 * For right-to-left language support
2671 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
2675 * Return the correct HTML 'dir' attribute value for this language.
2679 return $this->isRTL() ?
'rtl' : 'ltr';
2683 * Return 'left' or 'right' as appropriate alignment for line-start
2684 * for this language's text direction.
2686 * Should be equivalent to CSS3 'start' text-align value....
2690 function alignStart() {
2691 return $this->isRTL() ?
'right' : 'left';
2695 * Return 'right' or 'left' as appropriate alignment for line-end
2696 * for this language's text direction.
2698 * Should be equivalent to CSS3 'end' text-align value....
2702 function alignEnd() {
2703 return $this->isRTL() ?
'left' : 'right';
2707 * A hidden direction mark (LRM or RLM), depending on the language direction.
2708 * Unlike getDirMark(), this function returns the character as an HTML entity.
2709 * This function should be used when the output is guaranteed to be HTML,
2710 * because it makes the output HTML source code more readable. When
2711 * the output is plain text or can be escaped, getDirMark() should be used.
2713 * @param $opposite Boolean Get the direction mark opposite to your language
2716 function getDirMarkEntity( $opposite = false ) {
2717 if ( $opposite ) { return $this->isRTL() ?
'‎' : '‏'; }
2718 return $this->isRTL() ?
'‏' : '‎';
2722 * A hidden direction mark (LRM or RLM), depending on the language direction.
2723 * This function produces them as invisible Unicode characters and
2724 * the output may be hard to read and debug, so it should only be used
2725 * when the output is plain text or can be escaped. When the output is
2726 * HTML, use getDirMarkEntity() instead.
2728 * @param $opposite Boolean Get the direction mark opposite to your language
2731 function getDirMark( $opposite = false ) {
2732 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2733 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2734 if ( $opposite ) { return $this->isRTL() ?
$lrm : $rlm; }
2735 return $this->isRTL() ?
$rlm : $lrm;
2741 function capitalizeAllNouns() {
2742 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2746 * An arrow, depending on the language direction.
2748 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2751 function getArrow( $direction = 'forwards' ) {
2752 switch ( $direction ) {
2754 return $this->isRTL() ?
'←' : '→';
2756 return $this->isRTL() ?
'→' : '←';
2769 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2773 function linkPrefixExtension() {
2774 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2780 function getMagicWords() {
2781 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2784 protected function doMagicHook() {
2785 if ( $this->mMagicHookDone
) {
2788 $this->mMagicHookDone
= true;
2789 wfProfileIn( 'LanguageGetMagic' );
2790 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2791 wfProfileOut( 'LanguageGetMagic' );
2795 * Fill a MagicWord object with data from here
2799 function getMagic( $mw ) {
2800 $this->doMagicHook();
2802 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
2803 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
2805 $magicWords = $this->getMagicWords();
2806 if ( isset( $magicWords[$mw->mId
] ) ) {
2807 $rawEntry = $magicWords[$mw->mId
];
2813 if ( !is_array( $rawEntry ) ) {
2814 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
2816 $mw->mCaseSensitive
= $rawEntry[0];
2817 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
2822 * Add magic words to the extension array
2824 * @param $newWords array
2826 function addMagicWordsByLang( $newWords ) {
2827 $fallbackChain = $this->getFallbackLanguages();
2828 $fallbackChain = array_reverse( $fallbackChain );
2829 foreach ( $fallbackChain as $code ) {
2830 if ( isset( $newWords[$code] ) ) {
2831 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
2837 * Get special page names, as an associative array
2838 * case folded alias => real name
2840 function getSpecialPageAliases() {
2841 // Cache aliases because it may be slow to load them
2842 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
2844 $this->mExtendedSpecialPageAliases
=
2845 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
2846 wfRunHooks( 'LanguageGetSpecialPageAliases',
2847 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
2850 return $this->mExtendedSpecialPageAliases
;
2854 * Italic is unsuitable for some languages
2856 * @param $text String: the text to be emphasized.
2859 function emphasize( $text ) {
2860 return "<em>$text</em>";
2864 * Normally we output all numbers in plain en_US style, that is
2865 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2866 * point twohundredthirtyfive. However this is not suitable for all
2867 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2868 * Icelandic just want to use commas instead of dots, and dots instead
2869 * of commas like "293.291,235".
2871 * An example of this function being called:
2873 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2876 * See LanguageGu.php for the Gujarati implementation and
2877 * $separatorTransformTable on MessageIs.php for
2878 * the , => . and . => , implementation.
2880 * @todo check if it's viable to use localeconv() for the decimal
2882 * @param $number Mixed: the string to be formatted, should be an integer
2883 * or a floating point number.
2884 * @param $nocommafy Bool: set to true for special numbers like dates
2887 public function formatNum( $number, $nocommafy = false ) {
2888 global $wgTranslateNumerals;
2889 if ( !$nocommafy ) {
2890 $number = $this->commafy( $number );
2891 $s = $this->separatorTransformTable();
2893 $number = strtr( $number, $s );
2897 if ( $wgTranslateNumerals ) {
2898 $s = $this->digitTransformTable();
2900 $number = strtr( $number, $s );
2908 * @param $number string
2911 function parseFormattedNumber( $number ) {
2912 $s = $this->digitTransformTable();
2914 $number = strtr( $number, array_flip( $s ) );
2917 $s = $this->separatorTransformTable();
2919 $number = strtr( $number, array_flip( $s ) );
2922 $number = strtr( $number, array( ',' => '' ) );
2927 * Adds commas to a given number
2932 function commafy( $_ ) {
2933 $digitGroupingPattern = $this->digitGroupingPattern();
2934 if ( $_ === null ) {
2938 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
2939 // default grouping is at thousands, use the same for ###,###,### pattern too.
2940 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2942 // Ref: http://cldr.unicode.org/translation/number-patterns
2944 if ( intval( $_ ) < 0 ) {
2945 // For negative numbers apply the algorithm like positive number and add sign.
2947 $_ = substr( $_, 1 );
2949 $numberpart = array();
2950 $decimalpart = array();
2951 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
2952 preg_match( "/\d+/", $_, $numberpart );
2953 preg_match( "/\.\d*/", $_, $decimalpart );
2954 $groupedNumber = ( count( $decimalpart ) > 0 ) ?
$decimalpart[0]:"";
2955 if ( $groupedNumber === $_ ) {
2956 // the string does not have any number part. Eg: .12345
2957 return $sign . $groupedNumber;
2959 $start = $end = strlen( $numberpart[0] );
2960 while ( $start > 0 ) {
2961 $match = $matches[0][$numMatches -1] ;
2962 $matchLen = strlen( $match );
2963 $start = $end - $matchLen;
2967 $groupedNumber = substr( $_ , $start, $end -$start ) . $groupedNumber ;
2969 if ( $numMatches > 1 ) {
2970 // use the last pattern for the rest of the number
2974 $groupedNumber = "," . $groupedNumber;
2977 return $sign . $groupedNumber;
2983 function digitGroupingPattern() {
2984 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
2990 function digitTransformTable() {
2991 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
2997 function separatorTransformTable() {
2998 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3002 * Take a list of strings and build a locale-friendly comma-separated
3003 * list, using the local comma-separator message.
3004 * The last two strings are chained with an "and".
3009 function listToText( array $l ) {
3011 $m = count( $l ) - 1;
3013 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
3015 for ( $i = $m; $i >= 0; $i-- ) {
3018 } elseif ( $i == $m - 1 ) {
3019 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
3021 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
3029 * Take a list of strings and build a locale-friendly comma-separated
3030 * list, using the local comma-separator message.
3031 * @param $list array of strings to put in a comma list
3034 function commaList( array $list ) {
3038 array( 'parsemag', 'escapenoentities', 'language' => $this )
3045 * Take a list of strings and build a locale-friendly semicolon-separated
3046 * list, using the local semicolon-separator message.
3047 * @param $list array of strings to put in a semicolon list
3050 function semicolonList( array $list ) {
3053 'semicolon-separator',
3054 array( 'parsemag', 'escapenoentities', 'language' => $this )
3061 * Same as commaList, but separate it with the pipe instead.
3062 * @param $list array of strings to put in a pipe list
3065 function pipeList( array $list ) {
3069 array( 'escapenoentities', 'language' => $this )
3076 * Truncate a string to a specified length in bytes, appending an optional
3077 * string (e.g. for ellipses)
3079 * The database offers limited byte lengths for some columns in the database;
3080 * multi-byte character sets mean we need to ensure that only whole characters
3081 * are included, otherwise broken characters can be passed to the user
3083 * If $length is negative, the string will be truncated from the beginning
3085 * @param $string String to truncate
3086 * @param $length Int: maximum length (including ellipses)
3087 * @param $ellipsis String to append to the truncated text
3088 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3089 * $adjustLength was introduced in 1.18, before that behaved as if false.
3092 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3093 # Use the localized ellipsis character
3094 if ( $ellipsis == '...' ) {
3095 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
3097 # Check if there is no need to truncate
3098 if ( $length == 0 ) {
3099 return $ellipsis; // convention
3100 } elseif ( strlen( $string ) <= abs( $length ) ) {
3101 return $string; // no need to truncate
3103 $stringOriginal = $string;
3104 # If ellipsis length is >= $length then we can't apply $adjustLength
3105 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3106 $string = $ellipsis; // this can be slightly unexpected
3107 # Otherwise, truncate and add ellipsis...
3109 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3110 if ( $length > 0 ) {
3111 $length -= $eLength;
3112 $string = substr( $string, 0, $length ); // xyz...
3113 $string = $this->removeBadCharLast( $string );
3114 $string = $string . $ellipsis;
3116 $length +
= $eLength;
3117 $string = substr( $string, $length ); // ...xyz
3118 $string = $this->removeBadCharFirst( $string );
3119 $string = $ellipsis . $string;
3122 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3123 # This check is *not* redundant if $adjustLength, due to the single case where
3124 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3125 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3128 return $stringOriginal;
3133 * Remove bytes that represent an incomplete Unicode character
3134 * at the end of string (e.g. bytes of the char are missing)
3136 * @param $string String
3139 protected function removeBadCharLast( $string ) {
3140 if ( $string != '' ) {
3141 $char = ord( $string[strlen( $string ) - 1] );
3143 if ( $char >= 0xc0 ) {
3144 # We got the first byte only of a multibyte char; remove it.
3145 $string = substr( $string, 0, -1 );
3146 } elseif ( $char >= 0x80 &&
3147 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3148 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
3150 # We chopped in the middle of a character; remove it
3158 * Remove bytes that represent an incomplete Unicode character
3159 * at the start of string (e.g. bytes of the char are missing)
3161 * @param $string String
3164 protected function removeBadCharFirst( $string ) {
3165 if ( $string != '' ) {
3166 $char = ord( $string[0] );
3167 if ( $char >= 0x80 && $char < 0xc0 ) {
3168 # We chopped in the middle of a character; remove the whole thing
3169 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3176 * Truncate a string of valid HTML to a specified length in bytes,
3177 * appending an optional string (e.g. for ellipses), and return valid HTML
3179 * This is only intended for styled/linked text, such as HTML with
3180 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3181 * Also, this will not detect things like "display:none" CSS.
3183 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3185 * @param string $text HTML string to truncate
3186 * @param int $length (zero/positive) Maximum length (including ellipses)
3187 * @param string $ellipsis String to append to the truncated text
3190 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3191 # Use the localized ellipsis character
3192 if ( $ellipsis == '...' ) {
3193 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
3195 # Check if there is clearly no need to truncate
3196 if ( $length <= 0 ) {
3197 return $ellipsis; // no text shown, nothing to format (convention)
3198 } elseif ( strlen( $text ) <= $length ) {
3199 return $text; // string short enough even *with* HTML (short-circuit)
3202 $dispLen = 0; // innerHTML legth so far
3203 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3204 $tagType = 0; // 0-open, 1-close
3205 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3206 $entityState = 0; // 0-not entity, 1-entity
3207 $tag = $ret = ''; // accumulated tag name, accumulated result string
3208 $openTags = array(); // open tag stack
3209 $maybeState = null; // possible truncation state
3211 $textLen = strlen( $text );
3212 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3213 for ( $pos = 0; true; ++
$pos ) {
3214 # Consider truncation once the display length has reached the maximim.
3215 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3216 # Check that we're not in the middle of a bracket/entity...
3217 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3218 if ( !$testingEllipsis ) {
3219 $testingEllipsis = true;
3220 # Save where we are; we will truncate here unless there turn out to
3221 # be so few remaining characters that truncation is not necessary.
3222 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3223 $maybeState = array( $ret, $openTags ); // save state
3225 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3226 # String in fact does need truncation, the truncation point was OK.
3227 list( $ret, $openTags ) = $maybeState; // reload state
3228 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3229 $ret .= $ellipsis; // add ellipsis
3233 if ( $pos >= $textLen ) break; // extra iteration just for above checks
3235 # Read the next char...
3237 $lastCh = $pos ?
$text[$pos - 1] : '';
3238 $ret .= $ch; // add to result string
3240 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3241 $entityState = 0; // for bad HTML
3242 $bracketState = 1; // tag started (checking for backslash)
3243 } elseif ( $ch == '>' ) {
3244 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3245 $entityState = 0; // for bad HTML
3246 $bracketState = 0; // out of brackets
3247 } elseif ( $bracketState == 1 ) {
3249 $tagType = 1; // close tag (e.g. "</span>")
3251 $tagType = 0; // open tag (e.g. "<span>")
3254 $bracketState = 2; // building tag name
3255 } elseif ( $bracketState == 2 ) {
3259 // Name found (e.g. "<a href=..."), add on tag attributes...
3260 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3262 } elseif ( $bracketState == 0 ) {
3263 if ( $entityState ) {
3266 $dispLen++
; // entity is one displayed char
3269 if ( $neLength == 0 && !$maybeState ) {
3270 // Save state without $ch. We want to *hit* the first
3271 // display char (to get tags) but not *use* it if truncating.
3272 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3275 $entityState = 1; // entity found, (e.g. " ")
3277 $dispLen++
; // this char is displayed
3278 // Add the next $max display text chars after this in one swoop...
3279 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3280 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3281 $dispLen +
= $skipped;
3287 // Close the last tag if left unclosed by bad HTML
3288 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3289 while ( count( $openTags ) > 0 ) {
3290 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3296 * truncateHtml() helper function
3297 * like strcspn() but adds the skipped chars to $ret
3306 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3307 if ( $len === null ) {
3308 $len = -1; // -1 means "no limit" for strcspn
3309 } elseif ( $len < 0 ) {
3313 if ( $start < strlen( $text ) ) {
3314 $skipCount = strcspn( $text, $search, $start, $len );
3315 $ret .= substr( $text, $start, $skipCount );
3321 * truncateHtml() helper function
3322 * (a) push or pop $tag from $openTags as needed
3323 * (b) clear $tag value
3324 * @param &$tag string Current HTML tag name we are looking at
3325 * @param $tagType int (0-open tag, 1-close tag)
3326 * @param $lastCh string Character before the '>' that ended this tag
3327 * @param &$openTags array Open tag stack (not accounting for $tag)
3329 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3330 $tag = ltrim( $tag );
3332 if ( $tagType == 0 && $lastCh != '/' ) {
3333 $openTags[] = $tag; // tag opened (didn't close itself)
3334 } elseif ( $tagType == 1 ) {
3335 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3336 array_pop( $openTags ); // tag closed
3344 * Grammatical transformations, needed for inflected languages
3345 * Invoked by putting {{grammar:case|word}} in a message
3347 * @param $word string
3348 * @param $case string
3351 function convertGrammar( $word, $case ) {
3352 global $wgGrammarForms;
3353 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3354 return $wgGrammarForms[$this->getCode()][$case][$word];
3359 * Get the grammar forms for the content language
3360 * @return array of grammar forms
3363 function getGrammarForms() {
3364 global $wgGrammarForms;
3365 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3366 return $wgGrammarForms[$this->getCode()];
3371 * Provides an alternative text depending on specified gender.
3372 * Usage {{gender:username|masculine|feminine|neutral}}.
3373 * username is optional, in which case the gender of current user is used,
3374 * but only in (some) interface messages; otherwise default gender is used.
3376 * If no forms are given, an empty string is returned. If only one form is
3377 * given, it will be returned unconditionally. These details are implied by
3378 * the caller and cannot be overridden in subclasses.
3380 * If more than one form is given, the default is to use the neutral one
3381 * if it is specified, and to use the masculine one otherwise. These
3382 * details can be overridden in subclasses.
3384 * @param $gender string
3385 * @param $forms array
3389 function gender( $gender, $forms ) {
3390 if ( !count( $forms ) ) {
3393 $forms = $this->preConvertPlural( $forms, 2 );
3394 if ( $gender === 'male' ) {
3397 if ( $gender === 'female' ) {
3400 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3404 * Plural form transformations, needed for some languages.
3405 * For example, there are 3 form of plural in Russian and Polish,
3406 * depending on "count mod 10". See [[w:Plural]]
3407 * For English it is pretty simple.
3409 * Invoked by putting {{plural:count|wordform1|wordform2}}
3410 * or {{plural:count|wordform1|wordform2|wordform3}}
3412 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3414 * @param $count Integer: non-localized number
3415 * @param $forms Array: different plural forms
3416 * @return string Correct form of plural for $count in this language
3418 function convertPlural( $count, $forms ) {
3419 if ( !count( $forms ) ) {
3422 $forms = $this->preConvertPlural( $forms, 2 );
3424 return ( $count == 1 ) ?
$forms[0] : $forms[1];
3428 * Checks that convertPlural was given an array and pads it to requested
3429 * amount of forms by copying the last one.
3431 * @param $count Integer: How many forms should there be at least
3432 * @param $forms Array of forms given to convertPlural
3433 * @return array Padded array of forms or an exception if not an array
3435 protected function preConvertPlural( /* Array */ $forms, $count ) {
3436 while ( count( $forms ) < $count ) {
3437 $forms[] = $forms[count( $forms ) - 1];
3443 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3444 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3445 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3446 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3449 * @param $str String: the validated block duration in English
3450 * @return string Somehow translated block duration
3451 * @see LanguageFi.php for example implementation
3453 function translateBlockExpiry( $str ) {
3454 $duration = SpecialBlock
::getSuggestedDurations( $this );
3455 foreach ( $duration as $show => $value ) {
3456 if ( strcmp( $str, $value ) == 0 ) {
3457 return htmlspecialchars( trim( $show ) );
3461 // Since usually only infinite or indefinite is only on list, so try
3462 // equivalents if still here.
3463 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3464 if ( in_array( $str, $indefs ) ) {
3465 foreach ( $indefs as $val ) {
3466 $show = array_search( $val, $duration, true );
3467 if ( $show !== false ) {
3468 return htmlspecialchars( trim( $show ) );
3472 // If all else fails, return the original string.
3477 * languages like Chinese need to be segmented in order for the diff
3480 * @param $text String
3483 public function segmentForDiff( $text ) {
3488 * and unsegment to show the result
3490 * @param $text String
3493 public function unsegmentForDiff( $text ) {
3498 * Return the LanguageConverter used in the Language
3501 * @return LanguageConverter
3503 public function getConverter() {
3504 return $this->mConverter
;
3508 * convert text to all supported variants
3510 * @param $text string
3513 public function autoConvertToAllVariants( $text ) {
3514 return $this->mConverter
->autoConvertToAllVariants( $text );
3518 * convert text to different variants of a language.
3520 * @param $text string
3523 public function convert( $text ) {
3524 return $this->mConverter
->convert( $text );
3528 * Convert a Title object to a string in the preferred variant
3530 * @param $title Title
3533 public function convertTitle( $title ) {
3534 return $this->mConverter
->convertTitle( $title );
3538 * Check if this is a language with variants
3542 public function hasVariants() {
3543 return sizeof( $this->getVariants() ) > 1;
3547 * Check if the language has the specific variant
3550 * @param $variant string
3553 public function hasVariant( $variant ) {
3554 return (bool)$this->mConverter
->validateVariant( $variant );
3558 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3560 * @param $text string
3563 public function armourMath( $text ) {
3564 return $this->mConverter
->armourMath( $text );
3568 * Perform output conversion on a string, and encode for safe HTML output.
3569 * @param $text String text to be converted
3570 * @param $isTitle Bool whether this conversion is for the article title
3572 * @todo this should get integrated somewhere sane
3574 public function convertHtml( $text, $isTitle = false ) {
3575 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3579 * @param $key string
3582 public function convertCategoryKey( $key ) {
3583 return $this->mConverter
->convertCategoryKey( $key );
3587 * Get the list of variants supported by this language
3588 * see sample implementation in LanguageZh.php
3590 * @return array an array of language codes
3592 public function getVariants() {
3593 return $this->mConverter
->getVariants();
3599 public function getPreferredVariant() {
3600 return $this->mConverter
->getPreferredVariant();
3606 public function getDefaultVariant() {
3607 return $this->mConverter
->getDefaultVariant();
3613 public function getURLVariant() {
3614 return $this->mConverter
->getURLVariant();
3618 * If a language supports multiple variants, it is
3619 * possible that non-existing link in one variant
3620 * actually exists in another variant. this function
3621 * tries to find it. See e.g. LanguageZh.php
3623 * @param $link String: the name of the link
3624 * @param $nt Mixed: the title object of the link
3625 * @param $ignoreOtherCond Boolean: to disable other conditions when
3626 * we need to transclude a template or update a category's link
3627 * @return null the input parameters may be modified upon return
3629 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3630 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
3634 * If a language supports multiple variants, converts text
3635 * into an array of all possible variants of the text:
3636 * 'variant' => text in that variant
3638 * @deprecated since 1.17 Use autoConvertToAllVariants()
3640 * @param $text string
3644 public function convertLinkToAllVariants( $text ) {
3645 return $this->mConverter
->convertLinkToAllVariants( $text );
3649 * returns language specific options used by User::getPageRenderHash()
3650 * for example, the preferred language variant
3654 function getExtraHashOptions() {
3655 return $this->mConverter
->getExtraHashOptions();
3659 * For languages that support multiple variants, the title of an
3660 * article may be displayed differently in different variants. this
3661 * function returns the apporiate title defined in the body of the article.
3665 public function getParsedTitle() {
3666 return $this->mConverter
->getParsedTitle();
3670 * Enclose a string with the "no conversion" tag. This is used by
3671 * various functions in the Parser
3673 * @param $text String: text to be tagged for no conversion
3674 * @param $noParse bool
3675 * @return string the tagged text
3677 public function markNoConversion( $text, $noParse = false ) {
3678 return $this->mConverter
->markNoConversion( $text, $noParse );
3682 * A regular expression to match legal word-trailing characters
3683 * which should be merged onto a link of the form [[foo]]bar.
3687 public function linkTrail() {
3688 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
3694 function getLangObj() {
3699 * Get the RFC 3066 code for this language object
3701 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3702 * htmlspecialchars() or similar
3706 public function getCode() {
3707 return $this->mCode
;
3711 * Get the code in Bcp47 format which we can use
3712 * inside of html lang="" tags.
3714 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3715 * htmlspecialchars() or similar.
3720 public function getHtmlCode() {
3721 if ( is_null( $this->mHtmlCode
) ) {
3722 $this->mHtmlCode
= wfBCP47( $this->getCode() );
3724 return $this->mHtmlCode
;
3728 * @param $code string
3730 public function setCode( $code ) {
3731 $this->mCode
= $code;
3732 // Ensure we don't leave an incorrect html code lying around
3733 $this->mHtmlCode
= null;
3737 * Get the name of a file for a certain language code
3738 * @param $prefix string Prepend this to the filename
3739 * @param $code string Language code
3740 * @param $suffix string Append this to the filename
3741 * @throws MWException
3742 * @return string $prefix . $mangledCode . $suffix
3744 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
3745 // Protect against path traversal
3746 if ( !Language
::isValidCode( $code )
3747 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
3749 throw new MWException( "Invalid language code \"$code\"" );
3752 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
3756 * Get the language code from a file name. Inverse of getFileName()
3757 * @param $filename string $prefix . $languageCode . $suffix
3758 * @param $prefix string Prefix before the language code
3759 * @param $suffix string Suffix after the language code
3760 * @return string Language code, or false if $prefix or $suffix isn't found
3762 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
3764 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
3765 preg_quote( $suffix, '/' ) . '/', $filename, $m );
3766 if ( !count( $m ) ) {
3769 return str_replace( '_', '-', strtolower( $m[1] ) );
3773 * @param $code string
3776 public static function getMessagesFileName( $code ) {
3778 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
3779 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
3784 * @param $code string
3787 public static function getClassFileName( $code ) {
3789 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
3793 * Get the first fallback for a given language.
3795 * @param $code string
3797 * @return bool|string
3799 public static function getFallbackFor( $code ) {
3800 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
3803 $fallbacks = self
::getFallbacksFor( $code );
3804 $first = array_shift( $fallbacks );
3810 * Get the ordered list of fallback languages.
3813 * @param $code string Language code
3816 public static function getFallbacksFor( $code ) {
3817 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
3820 $v = self
::getLocalisationCache()->getItem( $code, 'fallback' );
3821 $v = array_map( 'trim', explode( ',', $v ) );
3822 if ( $v[count( $v ) - 1] !== 'en' ) {
3830 * Get all messages for a given language
3831 * WARNING: this may take a long time. If you just need all message *keys*
3832 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
3834 * @param $code string
3838 public static function getMessagesFor( $code ) {
3839 return self
::getLocalisationCache()->getItem( $code, 'messages' );
3843 * Get a message for a given language
3845 * @param $key string
3846 * @param $code string
3850 public static function getMessageFor( $key, $code ) {
3851 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
3855 * Get all message keys for a given language. This is a faster alternative to
3856 * array_keys( Language::getMessagesFor( $code ) )
3859 * @param $code string Language code
3860 * @return array of message keys (strings)
3862 public static function getMessageKeysFor( $code ) {
3863 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
3870 function fixVariableInNamespace( $talk ) {
3871 if ( strpos( $talk, '$1' ) === false ) {
3875 global $wgMetaNamespace;
3876 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
3878 # Allow grammar transformations
3879 # Allowing full message-style parsing would make simple requests
3880 # such as action=raw much more expensive than they need to be.
3881 # This will hopefully cover most cases.
3882 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
3883 array( &$this, 'replaceGrammarInNamespace' ), $talk );
3884 return str_replace( ' ', '_', $talk );
3891 function replaceGrammarInNamespace( $m ) {
3892 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
3896 * @throws MWException
3899 static function getCaseMaps() {
3900 static $wikiUpperChars, $wikiLowerChars;
3901 if ( isset( $wikiUpperChars ) ) {
3902 return array( $wikiUpperChars, $wikiLowerChars );
3905 wfProfileIn( __METHOD__
);
3906 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
3907 if ( $arr === false ) {
3908 throw new MWException(
3909 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
3911 $wikiUpperChars = $arr['wikiUpperChars'];
3912 $wikiLowerChars = $arr['wikiLowerChars'];
3913 wfProfileOut( __METHOD__
);
3914 return array( $wikiUpperChars, $wikiLowerChars );
3918 * Decode an expiry (block, protection, etc) which has come from the DB
3920 * @FIXME: why are we returnings DBMS-dependent strings???
3922 * @param $expiry String: Database expiry String
3923 * @param $format Bool|Int true to process using language functions, or TS_ constant
3924 * to return the expiry in a given timestamp
3927 public function formatExpiry( $expiry, $format = true ) {
3928 static $infinity, $infinityMsg;
3929 if ( $infinity === null ) {
3930 $infinityMsg = wfMessage( 'infiniteblock' );
3931 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
3934 if ( $expiry == '' ||
$expiry == $infinity ) {
3935 return $format === true
3939 return $format === true
3940 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
3941 : wfTimestamp( $format, $expiry );
3947 * @param $seconds int|float
3948 * @param $format Array Optional
3949 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
3950 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
3951 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
3952 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
3955 function formatTimePeriod( $seconds, $format = array() ) {
3956 if ( !is_array( $format ) ) {
3957 $format = array( 'avoid' => $format ); // For backwards compatibility
3959 if ( !isset( $format['avoid'] ) ) {
3960 $format['avoid'] = false;
3962 if ( !isset( $format['noabbrevs' ] ) ) {
3963 $format['noabbrevs'] = false;
3965 $secondsMsg = wfMessage(
3966 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
3967 $minutesMsg = wfMessage(
3968 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
3969 $hoursMsg = wfMessage(
3970 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
3971 $daysMsg = wfMessage(
3972 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
3974 if ( round( $seconds * 10 ) < 100 ) {
3975 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
3976 $s = $secondsMsg->params( $s )->text();
3977 } elseif ( round( $seconds ) < 60 ) {
3978 $s = $this->formatNum( round( $seconds ) );
3979 $s = $secondsMsg->params( $s )->text();
3980 } elseif ( round( $seconds ) < 3600 ) {
3981 $minutes = floor( $seconds / 60 );
3982 $secondsPart = round( fmod( $seconds, 60 ) );
3983 if ( $secondsPart == 60 ) {
3987 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3989 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3990 } elseif ( round( $seconds ) <= 2 * 86400 ) {
3991 $hours = floor( $seconds / 3600 );
3992 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
3993 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
3994 if ( $secondsPart == 60 ) {
3998 if ( $minutes == 60 ) {
4002 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4004 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4005 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4006 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4009 $days = floor( $seconds / 86400 );
4010 if ( $format['avoid'] === 'avoidminutes' ) {
4011 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4012 if ( $hours == 24 ) {
4016 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4018 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4019 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4020 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4021 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4022 if ( $minutes == 60 ) {
4026 if ( $hours == 24 ) {
4030 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4032 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4034 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4036 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4038 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4045 * Format a bitrate for output, using an appropriate
4046 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4048 * This use base 1000. For base 1024 use formatSize(), for another base
4049 * see formatComputingNumbers()
4054 function formatBitrate( $bps ) {
4055 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4059 * @param $size int Size of the unit
4060 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4061 * @param $messageKey string Message key to be uesd
4064 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4066 return str_replace( '$1', $this->formatNum( $size ),
4067 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4070 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4073 $maxIndex = count( $sizes ) - 1;
4074 while ( $size >= $boundary && $index < $maxIndex ) {
4079 // For small sizes no decimal places necessary
4082 // For MB and bigger two decimal places are smarter
4085 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4087 $size = round( $size, $round );
4088 $text = $this->getMessageFromDB( $msg );
4089 return str_replace( '$1', $this->formatNum( $size ), $text );
4093 * Format a size in bytes for output, using an appropriate
4094 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4096 * This method use base 1024. For base 1000 use formatBitrate(), for
4097 * another base see formatComputingNumbers()
4099 * @param $size int Size to format
4100 * @return string Plain text (not HTML)
4102 function formatSize( $size ) {
4103 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4107 * Make a list item, used by various special pages
4109 * @param $page String Page link
4110 * @param $details String Text between brackets
4111 * @param $oppositedm Boolean Add the direction mark opposite to your
4112 * language, to display text properly
4115 function specialList( $page, $details, $oppositedm = true ) {
4116 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) .
4117 $this->getDirMark();
4118 $details = $details ?
$dirmark . $this->getMessageFromDB( 'word-separator' ) .
4119 wfMsgExt( 'parentheses', array( 'escape', 'replaceafter', 'language' => $this ), $details ) : '';
4120 return $page . $details;
4124 * Generate (prev x| next x) (20|50|100...) type links for paging
4126 * @param $title Title object to link
4127 * @param $offset Integer offset parameter
4128 * @param $limit Integer limit parameter
4129 * @param $query array|String optional URL query parameter string
4130 * @param $atend Bool optional param for specified if this is the last page
4133 public function viewPrevNext( Title
$title, $offset, $limit, array $query = array(), $atend = false ) {
4134 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4136 # Make 'previous' link
4137 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4138 if ( $offset > 0 ) {
4139 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4140 $query, $prev, 'prevn-title', 'mw-prevlink' );
4142 $plink = htmlspecialchars( $prev );
4146 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4148 $nlink = htmlspecialchars( $next );
4150 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4151 $query, $next, 'prevn-title', 'mw-nextlink' );
4154 # Make links to set number of items per page
4155 $numLinks = array();
4156 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4157 $numLinks[] = $this->numLink( $title, $offset, $num,
4158 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4161 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4162 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4166 * Helper function for viewPrevNext() that generates links
4168 * @param $title Title object to link
4169 * @param $offset Integer offset parameter
4170 * @param $limit Integer limit parameter
4171 * @param $query Array extra query parameters
4172 * @param $link String text to use for the link; will be escaped
4173 * @param $tooltipMsg String name of the message to use as tooltip
4174 * @param $class String value of the "class" attribute of the link
4175 * @return String HTML fragment
4177 private function numLink( Title
$title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4178 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4179 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4180 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4181 'title' => $tooltip, 'class' => $class ), $link );
4185 * Get the conversion rule title, if any.
4189 public function getConvRuleTitle() {
4190 return $this->mConverter
->getConvRuleTitle();