3 * Internationalisation code
10 * @defgroup Language Language
13 if ( !defined( 'MEDIAWIKI' ) ) {
14 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
19 global $wgLanguageNames;
20 require_once( dirname( __FILE__
) . '/Names.php' );
22 if ( function_exists( 'mb_strtoupper' ) ) {
23 mb_internal_encoding( 'UTF-8' );
27 * a fake language converter
37 function __construct( $langobj ) { $this->mLang
= $langobj; }
38 function autoConvertToAllVariants( $text ) { return array( $this->mLang
->getCode() => $text ); }
39 function convert( $t ) { return $t; }
40 function convertTo( $text, $variant ) { return $text; }
41 function convertTitle( $t ) { return $t->getPrefixedText(); }
42 function getVariants() { return array( $this->mLang
->getCode() ); }
43 function getPreferredVariant() { return $this->mLang
->getCode(); }
44 function getDefaultVariant() { return $this->mLang
->getCode(); }
45 function getURLVariant() { return ''; }
46 function getConvRuleTitle() { return false; }
47 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
48 function getExtraHashOptions() { return ''; }
49 function getParsedTitle() { return ''; }
50 function markNoConversion( $text, $noParse = false ) { return $text; }
51 function convertCategoryKey( $key ) { return $key; }
52 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
53 function armourMath( $text ) { return $text; }
57 * Internationalisation code
63 * @var LanguageConverter
67 var $mVariants, $mCode, $mLoaded = false;
68 var $mMagicExtensions = array(), $mMagicHookDone = false;
69 private $mHtmlCode = null;
71 var $dateFormatStrings = array();
72 var $mExtendedSpecialPageAliases;
74 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
77 * ReplacementArray object caches
79 var $transformData = array();
82 * @var LocalisationCache
84 static public $dataCache;
86 static public $mLangObjCache = array();
88 static public $mWeekdayMsgs = array(
89 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
93 static public $mWeekdayAbbrevMsgs = array(
94 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
97 static public $mMonthMsgs = array(
98 'january', 'february', 'march', 'april', 'may_long', 'june',
99 'july', 'august', 'september', 'october', 'november',
102 static public $mMonthGenMsgs = array(
103 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
104 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
107 static public $mMonthAbbrevMsgs = array(
108 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
109 'sep', 'oct', 'nov', 'dec'
112 static public $mIranianCalendarMonthMsgs = array(
113 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
114 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
115 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
116 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
119 static public $mHebrewCalendarMonthMsgs = array(
120 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
121 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
122 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
123 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
124 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
127 static public $mHebrewCalendarMonthGenMsgs = array(
128 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
129 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
130 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
131 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
132 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
135 static public $mHijriCalendarMonthMsgs = array(
136 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
137 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
138 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
139 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
143 * Get a cached language object for a given language code
144 * @param $code String
147 static function factory( $code ) {
148 if ( !isset( self
::$mLangObjCache[$code] ) ) {
149 if ( count( self
::$mLangObjCache ) > 10 ) {
150 // Don't keep a billion objects around, that's stupid.
151 self
::$mLangObjCache = array();
153 self
::$mLangObjCache[$code] = self
::newFromCode( $code );
155 return self
::$mLangObjCache[$code];
159 * Create a language object for a given language code
160 * @param $code String
161 * @throws MWException
164 protected static function newFromCode( $code ) {
165 // Protect against path traversal below
166 if ( !Language
::isValidCode( $code )
167 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
169 throw new MWException( "Invalid language code \"$code\"" );
172 if ( !Language
::isValidBuiltInCode( $code ) ) {
173 // It's not possible to customise this code with class files, so
174 // just return a Language object. This is to support uselang= hacks.
175 $lang = new Language
;
176 $lang->setCode( $code );
180 // Check if there is a language class for the code
181 $class = self
::classFromCode( $code );
182 self
::preloadLanguageClass( $class );
183 if ( MWInit
::classExists( $class ) ) {
188 // Keep trying the fallback list until we find an existing class
189 $fallbacks = Language
::getFallbacksFor( $code );
190 foreach ( $fallbacks as $fallbackCode ) {
191 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
192 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
195 $class = self
::classFromCode( $fallbackCode );
196 self
::preloadLanguageClass( $class );
197 if ( MWInit
::classExists( $class ) ) {
198 $lang = Language
::newFromCode( $fallbackCode );
199 $lang->setCode( $code );
204 throw new MWException( "Invalid fallback sequence for language '$code'" );
208 * Returns true if a language code string is of a valid form, whether or
209 * not it exists. This includes codes which are used solely for
210 * customisation via the MediaWiki namespace.
212 * @param $code string
216 public static function isValidCode( $code ) {
218 strcspn( $code, ":/\\\000" ) === strlen( $code )
219 && !preg_match( Title
::getTitleInvalidRegex(), $code );
223 * Returns true if a language code is of a valid form for the purposes of
224 * internal customisation of MediaWiki, via Messages*.php.
226 * @param $code string
231 public static function isValidBuiltInCode( $code ) {
232 return preg_match( '/^[a-z0-9-]+$/i', $code );
237 * @return String Name of the language class
239 public static function classFromCode( $code ) {
240 if ( $code == 'en' ) {
243 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
248 * Includes language class files
250 * @param $class string Name of the language class
252 public static function preloadLanguageClass( $class ) {
255 if ( $class === 'Language' ) {
259 if ( !defined( 'MW_COMPILED' ) ) {
260 // Preload base classes to work around APC/PHP5 bug
261 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
262 include_once( "$IP/languages/classes/$class.deps.php" );
264 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
265 include_once( "$IP/languages/classes/$class.php" );
271 * Get the LocalisationCache instance
273 * @return LocalisationCache
275 public static function getLocalisationCache() {
276 if ( is_null( self
::$dataCache ) ) {
277 global $wgLocalisationCacheConf;
278 $class = $wgLocalisationCacheConf['class'];
279 self
::$dataCache = new $class( $wgLocalisationCacheConf );
281 return self
::$dataCache;
284 function __construct() {
285 $this->mConverter
= new FakeConverter( $this );
286 // Set the code to the name of the descendant
287 if ( get_class( $this ) == 'Language' ) {
290 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
292 self
::getLocalisationCache();
296 * Reduce memory usage
298 function __destruct() {
299 foreach ( $this as $name => $value ) {
300 unset( $this->$name );
305 * Hook which will be called if this is the content language.
306 * Descendants can use this to register hook functions or modify globals
308 function initContLang() { }
311 * Same as getFallbacksFor for current language.
313 * @deprecated in 1.19
315 function getFallbackLanguageCode() {
316 wfDeprecated( __METHOD__
);
317 return self
::getFallbackFor( $this->mCode
);
324 function getFallbackLanguages() {
325 return self
::getFallbacksFor( $this->mCode
);
329 * Exports $wgBookstoreListEn
332 function getBookstoreList() {
333 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
339 public function getNamespaces() {
340 if ( is_null( $this->namespaceNames
) ) {
341 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
343 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
344 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
346 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
348 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
349 if ( $wgMetaNamespaceTalk ) {
350 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
352 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
353 $this->namespaceNames
[NS_PROJECT_TALK
] =
354 $this->fixVariableInNamespace( $talk );
357 # Sometimes a language will be localised but not actually exist on this wiki.
358 foreach ( $this->namespaceNames
as $key => $text ) {
359 if ( !isset( $validNamespaces[$key] ) ) {
360 unset( $this->namespaceNames
[$key] );
364 # The above mixing may leave namespaces out of canonical order.
365 # Re-order by namespace ID number...
366 ksort( $this->namespaceNames
);
368 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
370 return $this->namespaceNames
;
374 * Arbitrarily set all of the namespace names at once. Mainly used for testing
375 * @param $namespaces Array of namespaces (id => name)
377 public function setNamespaces( array $namespaces ) {
378 $this->namespaceNames
= $namespaces;
382 * A convenience function that returns the same thing as
383 * getNamespaces() except with the array values changed to ' '
384 * where it found '_', useful for producing output to be displayed
385 * e.g. in <select> forms.
389 function getFormattedNamespaces() {
390 $ns = $this->getNamespaces();
391 foreach ( $ns as $k => $v ) {
392 $ns[$k] = strtr( $v, '_', ' ' );
398 * Get a namespace value by key
400 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
401 * echo $mw_ns; // prints 'MediaWiki'
404 * @param $index Int: the array key of the namespace to return
405 * @return mixed, string if the namespace value exists, otherwise false
407 function getNsText( $index ) {
408 $ns = $this->getNamespaces();
409 return isset( $ns[$index] ) ?
$ns[$index] : false;
413 * A convenience function that returns the same thing as
414 * getNsText() except with '_' changed to ' ', useful for
417 * @param $index string
421 function getFormattedNsText( $index ) {
422 $ns = $this->getNsText( $index );
423 return strtr( $ns, '_', ' ' );
427 * Returns gender-dependent namespace alias if available.
428 * @param $index Int: namespace index
429 * @param $gender String: gender key (male, female... )
433 function getGenderNsText( $index, $gender ) {
434 global $wgExtraGenderNamespaces;
436 $ns = $wgExtraGenderNamespaces + self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
437 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
441 * Whether this language makes distinguishes genders for example in
446 function needsGenderDistinction() {
447 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
448 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
449 // $wgExtraGenderNamespaces overrides everything
451 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
452 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
453 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
456 // Check what is in i18n files
457 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
458 return count( $aliases ) > 0;
463 * Get a namespace key by value, case insensitive.
464 * Only matches namespace names for the current language, not the
465 * canonical ones defined in Namespace.php.
467 * @param $text String
468 * @return mixed An integer if $text is a valid value otherwise false
470 function getLocalNsIndex( $text ) {
471 $lctext = $this->lc( $text );
472 $ids = $this->getNamespaceIds();
473 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
479 function getNamespaceAliases() {
480 if ( is_null( $this->namespaceAliases
) ) {
481 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
485 foreach ( $aliases as $name => $index ) {
486 if ( $index === NS_PROJECT_TALK
) {
487 unset( $aliases[$name] );
488 $name = $this->fixVariableInNamespace( $name );
489 $aliases[$name] = $index;
494 global $wgExtraGenderNamespaces;
495 $genders = $wgExtraGenderNamespaces +
(array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
496 foreach ( $genders as $index => $forms ) {
497 foreach ( $forms as $alias ) {
498 $aliases[$alias] = $index;
502 $this->namespaceAliases
= $aliases;
504 return $this->namespaceAliases
;
510 function getNamespaceIds() {
511 if ( is_null( $this->mNamespaceIds
) ) {
512 global $wgNamespaceAliases;
513 # Put namespace names and aliases into a hashtable.
514 # If this is too slow, then we should arrange it so that it is done
515 # before caching. The catch is that at pre-cache time, the above
516 # class-specific fixup hasn't been done.
517 $this->mNamespaceIds
= array();
518 foreach ( $this->getNamespaces() as $index => $name ) {
519 $this->mNamespaceIds
[$this->lc( $name )] = $index;
521 foreach ( $this->getNamespaceAliases() as $name => $index ) {
522 $this->mNamespaceIds
[$this->lc( $name )] = $index;
524 if ( $wgNamespaceAliases ) {
525 foreach ( $wgNamespaceAliases as $name => $index ) {
526 $this->mNamespaceIds
[$this->lc( $name )] = $index;
530 return $this->mNamespaceIds
;
534 * Get a namespace key by value, case insensitive. Canonical namespace
535 * names override custom ones defined for the current language.
537 * @param $text String
538 * @return mixed An integer if $text is a valid value otherwise false
540 function getNsIndex( $text ) {
541 $lctext = $this->lc( $text );
542 $ns = MWNamespace
::getCanonicalIndex( $lctext );
543 if ( $ns !== null ) {
546 $ids = $this->getNamespaceIds();
547 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
551 * short names for language variants used for language conversion links.
553 * @param $code String
554 * @param $usemsg bool Use the "variantname-xyz" message if it exists
557 function getVariantname( $code, $usemsg = true ) {
558 $msg = "variantname-$code";
559 if ( $usemsg && wfMessage( $msg )->exists() ) {
560 return $this->getMessageFromDB( $msg );
562 $name = self
::fetchLanguageName( $code );
564 return $name; # if it's defined as a language name, show that
566 # otherwise, output the language code
572 * @param $name string
575 function specialPage( $name ) {
576 $aliases = $this->getSpecialPageAliases();
577 if ( isset( $aliases[$name][0] ) ) {
578 $name = $aliases[$name][0];
580 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
586 function getQuickbarSettings() {
588 $this->getMessage( 'qbsettings-none' ),
589 $this->getMessage( 'qbsettings-fixedleft' ),
590 $this->getMessage( 'qbsettings-fixedright' ),
591 $this->getMessage( 'qbsettings-floatingleft' ),
592 $this->getMessage( 'qbsettings-floatingright' ),
593 $this->getMessage( 'qbsettings-directionality' )
600 function getDatePreferences() {
601 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
607 function getDateFormats() {
608 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
612 * @return array|string
614 function getDefaultDateFormat() {
615 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
616 if ( $df === 'dmy or mdy' ) {
617 global $wgAmericanDates;
618 return $wgAmericanDates ?
'mdy' : 'dmy';
627 function getDatePreferenceMigrationMap() {
628 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
635 function getImageFile( $image ) {
636 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
642 function getExtraUserToggles() {
643 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
650 function getUserToggle( $tog ) {
651 return $this->getMessageFromDB( "tog-$tog" );
655 * Get native language names, indexed by code.
656 * Only those defined in MediaWiki, no other data like CLDR.
657 * If $customisedOnly is true, only returns codes with a messages file
659 * @param $customisedOnly bool
662 * @deprecated in 1.20, use fetchLanguageNames()
664 public static function getLanguageNames( $customisedOnly = false ) {
665 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
669 * Get translated language names. This is done on best effort and
670 * by default this is exactly the same as Language::getLanguageNames.
671 * The CLDR extension provides translated names.
672 * @param $code String Language code.
673 * @return Array language code => language name
675 * @deprecated in 1.20, use fetchLanguageNames()
677 public static function getTranslatedLanguageNames( $code ) {
678 return self
::fetchLanguageNames( $code, 'all' );
682 * Get an array of language names, indexed by code.
683 * @param $inLanguage null|string: Code of language in which to return the names
684 * Use null for autonyms (native names)
685 * @param $include string:
686 * 'all' all available languages
687 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames
688 * 'mwfile' only if the language is in 'mw' *and* has a message file
689 * @return array|bool: language code => language name, false if $include is wrong
692 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
693 global $wgExtraLanguageNames;
694 static $coreLanguageNames;
696 if ( $coreLanguageNames === null ) {
697 include( MWInit
::compiledPath( 'languages/Names.php' ) );
703 # TODO: also include when $inLanguage is null, when this code is more efficient
704 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
707 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
708 foreach ( $mwNames as $mwCode => $mwName ) {
709 # - Prefer own MediaWiki native name when not using the hook
710 # TODO: prefer it always to make it consistent, but casing is different in CLDR
711 # - For other names just add if not added through the hook
712 if ( ( $mwCode === $inLanguage && !$inLanguage ) ||
!isset( $names[$mwCode] ) ) {
713 $names[$mwCode] = $mwName;
717 if ( $include === 'all' ) {
722 $coreCodes = array_keys( $mwNames );
723 foreach( $coreCodes as $coreCode ) {
724 $returnMw[$coreCode] = $names[$coreCode];
727 if( $include === 'mw' ) {
729 } elseif( $include === 'mwfile' ) {
730 $namesMwFile = array();
731 # We do this using a foreach over the codes instead of a directory
732 # loop so that messages files in extensions will work correctly.
733 foreach ( $returnMw as $code => $value ) {
734 if ( is_readable( self
::getMessagesFileName( $code ) ) ) {
735 $namesMwFile[$code] = $names[$code];
744 * @param $code string: The code of the language for which to get the name
745 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
746 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
747 * @return string: Language name or empty
750 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
751 $array = self
::fetchLanguageNames( $inLanguage, $include );
752 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
756 * Get a message from the MediaWiki namespace.
758 * @param $msg String: message name
761 function getMessageFromDB( $msg ) {
762 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
766 * Get the native language name of $code.
767 * Only if defined in MediaWiki, no other data like CLDR.
768 * @param $code string
770 * @deprecated in 1.20, use fetchLanguageName()
772 function getLanguageName( $code ) {
773 return self
::fetchLanguageName( $code );
780 function getMonthName( $key ) {
781 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
787 function getMonthNamesArray() {
788 $monthNames = array( '' );
789 for ( $i = 1; $i < 13; $i++
) {
790 $monthNames[] = $this->getMonthName( $i );
799 function getMonthNameGen( $key ) {
800 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
807 function getMonthAbbreviation( $key ) {
808 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
814 function getMonthAbbreviationsArray() {
815 $monthNames = array( '' );
816 for ( $i = 1; $i < 13; $i++
) {
817 $monthNames[] = $this->getMonthAbbreviation( $i );
826 function getWeekdayName( $key ) {
827 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
834 function getWeekdayAbbreviation( $key ) {
835 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
842 function getIranianCalendarMonthName( $key ) {
843 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
850 function getHebrewCalendarMonthName( $key ) {
851 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
858 function getHebrewCalendarMonthNameGen( $key ) {
859 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
866 function getHijriCalendarMonthName( $key ) {
867 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
871 * This is a workalike of PHP's date() function, but with better
872 * internationalisation, a reduced set of format characters, and a better
875 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
876 * PHP manual for definitions. There are a number of extensions, which
879 * xn Do not translate digits of the next numeric format character
880 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
881 * xr Use roman numerals for the next numeric format character
882 * xh Use hebrew numerals for the next numeric format character
884 * xg Genitive month name
886 * xij j (day number) in Iranian calendar
887 * xiF F (month name) in Iranian calendar
888 * xin n (month number) in Iranian calendar
889 * xiy y (two digit year) in Iranian calendar
890 * xiY Y (full year) in Iranian calendar
892 * xjj j (day number) in Hebrew calendar
893 * xjF F (month name) in Hebrew calendar
894 * xjt t (days in month) in Hebrew calendar
895 * xjx xg (genitive month name) in Hebrew calendar
896 * xjn n (month number) in Hebrew calendar
897 * xjY Y (full year) in Hebrew calendar
899 * xmj j (day number) in Hijri calendar
900 * xmF F (month name) in Hijri calendar
901 * xmn n (month number) in Hijri calendar
902 * xmY Y (full year) in Hijri calendar
904 * xkY Y (full year) in Thai solar calendar. Months and days are
905 * identical to the Gregorian calendar
906 * xoY Y (full year) in Minguo calendar or Juche year.
907 * Months and days are identical to the
909 * xtY Y (full year) in Japanese nengo. Months and days are
910 * identical to the Gregorian calendar
912 * Characters enclosed in double quotes will be considered literal (with
913 * the quotes themselves removed). Unmatched quotes will be considered
914 * literal quotes. Example:
916 * "The month is" F => The month is January
919 * Backslash escaping is also supported.
921 * Input timestamp is assumed to be pre-normalized to the desired local
924 * @param $format String
925 * @param $ts String: 14-character timestamp
928 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
932 function sprintfDate( $format, $ts ) {
945 for ( $p = 0; $p < strlen( $format ); $p++
) {
948 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
949 $code .= $format[++
$p];
952 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
953 $code .= $format[++
$p];
964 $rawToggle = !$rawToggle;
973 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
976 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
977 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
980 $num = substr( $ts, 6, 2 );
983 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
984 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) +
1 );
987 $num = intval( substr( $ts, 6, 2 ) );
991 $iranian = self
::tsToIranian( $ts );
997 $hijri = self
::tsToHijri( $ts );
1003 $hebrew = self
::tsToHebrew( $ts );
1009 $unix = wfTimestamp( TS_UNIX
, $ts );
1011 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) +
1 );
1015 $unix = wfTimestamp( TS_UNIX
, $ts );
1017 $w = gmdate( 'w', $unix );
1022 $unix = wfTimestamp( TS_UNIX
, $ts );
1024 $num = gmdate( 'w', $unix );
1028 $unix = wfTimestamp( TS_UNIX
, $ts );
1030 $num = gmdate( 'z', $unix );
1034 $unix = wfTimestamp( TS_UNIX
, $ts );
1036 $num = gmdate( 'W', $unix );
1039 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1043 $iranian = self
::tsToIranian( $ts );
1045 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1049 $hijri = self
::tsToHijri( $ts );
1051 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1055 $hebrew = self
::tsToHebrew( $ts );
1057 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1060 $num = substr( $ts, 4, 2 );
1063 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1066 $num = intval( substr( $ts, 4, 2 ) );
1070 $iranian = self
::tsToIranian( $ts );
1076 $hijri = self
::tsToHijri ( $ts );
1082 $hebrew = self
::tsToHebrew( $ts );
1088 $unix = wfTimestamp( TS_UNIX
, $ts );
1090 $num = gmdate( 't', $unix );
1094 $hebrew = self
::tsToHebrew( $ts );
1100 $unix = wfTimestamp( TS_UNIX
, $ts );
1102 $num = gmdate( 'L', $unix );
1106 $unix = wfTimestamp( TS_UNIX
, $ts );
1108 $num = gmdate( 'o', $unix );
1111 $num = substr( $ts, 0, 4 );
1115 $iranian = self
::tsToIranian( $ts );
1121 $hijri = self
::tsToHijri( $ts );
1127 $hebrew = self
::tsToHebrew( $ts );
1133 $thai = self
::tsToYear( $ts, 'thai' );
1139 $minguo = self
::tsToYear( $ts, 'minguo' );
1145 $tenno = self
::tsToYear( $ts, 'tenno' );
1150 $num = substr( $ts, 2, 2 );
1154 $iranian = self
::tsToIranian( $ts );
1156 $num = substr( $iranian[0], -2 );
1159 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1162 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1165 $h = substr( $ts, 8, 2 );
1166 $num = $h %
12 ?
$h %
12 : 12;
1169 $num = intval( substr( $ts, 8, 2 ) );
1172 $h = substr( $ts, 8, 2 );
1173 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1176 $num = substr( $ts, 8, 2 );
1179 $num = substr( $ts, 10, 2 );
1182 $num = substr( $ts, 12, 2 );
1186 $unix = wfTimestamp( TS_UNIX
, $ts );
1188 $s .= gmdate( 'c', $unix );
1192 $unix = wfTimestamp( TS_UNIX
, $ts );
1194 $s .= gmdate( 'r', $unix );
1198 $unix = wfTimestamp( TS_UNIX
, $ts );
1203 # Backslash escaping
1204 if ( $p < strlen( $format ) - 1 ) {
1205 $s .= $format[++
$p];
1212 if ( $p < strlen( $format ) - 1 ) {
1213 $endQuote = strpos( $format, '"', $p +
1 );
1214 if ( $endQuote === false ) {
1215 # No terminating quote, assume literal "
1218 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1222 # Quote at end of string, assume literal "
1229 if ( $num !== false ) {
1230 if ( $rawToggle ||
$raw ) {
1233 } elseif ( $roman ) {
1234 $s .= self
::romanNumeral( $num );
1236 } elseif ( $hebrewNum ) {
1237 $s .= self
::hebrewNumeral( $num );
1240 $s .= $this->formatNum( $num, true );
1247 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1248 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1251 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1252 * Gregorian dates to Iranian dates. Originally written in C, it
1253 * is released under the terms of GNU Lesser General Public
1254 * License. Conversion to PHP was performed by Niklas Laxström.
1256 * Link: http://www.farsiweb.info/jalali/jalali.c
1262 private static function tsToIranian( $ts ) {
1263 $gy = substr( $ts, 0, 4 ) -1600;
1264 $gm = substr( $ts, 4, 2 ) -1;
1265 $gd = substr( $ts, 6, 2 ) -1;
1267 # Days passed from the beginning (including leap years)
1269 +
floor( ( $gy +
3 ) / 4 )
1270 - floor( ( $gy +
99 ) / 100 )
1271 +
floor( ( $gy +
399 ) / 400 );
1273 // Add days of the past months of this year
1274 for ( $i = 0; $i < $gm; $i++
) {
1275 $gDayNo +
= self
::$GREG_DAYS[$i];
1279 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1283 // Days passed in current month
1284 $gDayNo +
= (int)$gd;
1286 $jDayNo = $gDayNo - 79;
1288 $jNp = floor( $jDayNo / 12053 );
1291 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1294 if ( $jDayNo >= 366 ) {
1295 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1296 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1299 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1300 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1306 return array( $jy, $jm, $jd );
1310 * Converting Gregorian dates to Hijri dates.
1312 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1314 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1320 private static function tsToHijri( $ts ) {
1321 $year = substr( $ts, 0, 4 );
1322 $month = substr( $ts, 4, 2 );
1323 $day = substr( $ts, 6, 2 );
1331 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1332 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1335 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1336 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1337 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1340 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1341 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1344 $zl = $zjd -1948440 +
10632;
1345 $zn = (int)( ( $zl - 1 ) / 10631 );
1346 $zl = $zl - 10631 * $zn +
354;
1347 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1348 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1349 $zm = (int)( ( 24 * $zl ) / 709 );
1350 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1351 $zy = 30 * $zn +
$zj - 30;
1353 return array( $zy, $zm, $zd );
1357 * Converting Gregorian dates to Hebrew dates.
1359 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1360 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1361 * to translate the relevant functions into PHP and release them under
1364 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1365 * and Adar II is 14. In a non-leap year, Adar is 6.
1371 private static function tsToHebrew( $ts ) {
1373 $year = substr( $ts, 0, 4 );
1374 $month = substr( $ts, 4, 2 );
1375 $day = substr( $ts, 6, 2 );
1377 # Calculate Hebrew year
1378 $hebrewYear = $year +
3760;
1380 # Month number when September = 1, August = 12
1382 if ( $month > 12 ) {
1389 # Calculate day of year from 1 September
1391 for ( $i = 1; $i < $month; $i++
) {
1395 # Check if the year is leap
1396 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1399 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1406 # Calculate the start of the Hebrew year
1407 $start = self
::hebrewYearStart( $hebrewYear );
1409 # Calculate next year's start
1410 if ( $dayOfYear <= $start ) {
1411 # Day is before the start of the year - it is the previous year
1413 $nextStart = $start;
1417 # Add days since previous year's 1 September
1419 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1423 # Start of the new (previous) year
1424 $start = self
::hebrewYearStart( $hebrewYear );
1427 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1430 # Calculate Hebrew day of year
1431 $hebrewDayOfYear = $dayOfYear - $start;
1433 # Difference between year's days
1434 $diff = $nextStart - $start;
1435 # Add 12 (or 13 for leap years) days to ignore the difference between
1436 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1437 # difference is only about the year type
1438 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1444 # Check the year pattern, and is leap year
1445 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1446 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1447 # and non-leap years
1448 $yearPattern = $diff %
30;
1449 # Check if leap year
1450 $isLeap = $diff >= 30;
1452 # Calculate day in the month from number of day in the Hebrew year
1453 # Don't check Adar - if the day is not in Adar, we will stop before;
1454 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1455 $hebrewDay = $hebrewDayOfYear;
1458 while ( $hebrewMonth <= 12 ) {
1459 # Calculate days in this month
1460 if ( $isLeap && $hebrewMonth == 6 ) {
1461 # Adar in a leap year
1463 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1465 if ( $hebrewDay <= $days ) {
1469 # Subtract the days of Adar I
1470 $hebrewDay -= $days;
1473 if ( $hebrewDay <= $days ) {
1479 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1480 # Cheshvan in a complete year (otherwise as the rule below)
1482 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1483 # Kislev in an incomplete year (otherwise as the rule below)
1486 # Odd months have 30 days, even have 29
1487 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1489 if ( $hebrewDay <= $days ) {
1490 # In the current month
1493 # Subtract the days of the current month
1494 $hebrewDay -= $days;
1495 # Try in the next month
1500 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1504 * This calculates the Hebrew year start, as days since 1 September.
1505 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1506 * Used for Hebrew date.
1512 private static function hebrewYearStart( $year ) {
1513 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1514 $b = intval( ( $year - 1 ) %
4 );
1515 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1519 $Mar = intval( $m );
1525 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1526 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1528 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1530 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1534 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1539 * Algorithm to convert Gregorian dates to Thai solar dates,
1540 * Minguo dates or Minguo dates.
1542 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1543 * http://en.wikipedia.org/wiki/Minguo_calendar
1544 * http://en.wikipedia.org/wiki/Japanese_era_name
1546 * @param $ts String: 14-character timestamp
1547 * @param $cName String: calender name
1548 * @return Array: converted year, month, day
1550 private static function tsToYear( $ts, $cName ) {
1551 $gy = substr( $ts, 0, 4 );
1552 $gm = substr( $ts, 4, 2 );
1553 $gd = substr( $ts, 6, 2 );
1555 if ( !strcmp( $cName, 'thai' ) ) {
1557 # Add 543 years to the Gregorian calendar
1558 # Months and days are identical
1559 $gy_offset = $gy +
543;
1560 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1562 # Deduct 1911 years from the Gregorian calendar
1563 # Months and days are identical
1564 $gy_offset = $gy - 1911;
1565 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1566 # Nengō dates up to Meiji period
1567 # Deduct years from the Gregorian calendar
1568 # depending on the nengo periods
1569 # Months and days are identical
1570 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1572 $gy_gannen = $gy - 1868 +
1;
1573 $gy_offset = $gy_gannen;
1574 if ( $gy_gannen == 1 ) {
1577 $gy_offset = '明治' . $gy_offset;
1579 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1580 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1581 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1582 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1583 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1587 $gy_gannen = $gy - 1912 +
1;
1588 $gy_offset = $gy_gannen;
1589 if ( $gy_gannen == 1 ) {
1592 $gy_offset = '大正' . $gy_offset;
1594 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1595 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1596 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1600 $gy_gannen = $gy - 1926 +
1;
1601 $gy_offset = $gy_gannen;
1602 if ( $gy_gannen == 1 ) {
1605 $gy_offset = '昭和' . $gy_offset;
1608 $gy_gannen = $gy - 1989 +
1;
1609 $gy_offset = $gy_gannen;
1610 if ( $gy_gannen == 1 ) {
1613 $gy_offset = '平成' . $gy_offset;
1619 return array( $gy_offset, $gm, $gd );
1623 * Roman number formatting up to 3000
1629 static function romanNumeral( $num ) {
1630 static $table = array(
1631 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1632 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1633 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1634 array( '', 'M', 'MM', 'MMM' )
1637 $num = intval( $num );
1638 if ( $num > 3000 ||
$num <= 0 ) {
1643 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1644 if ( $num >= $pow10 ) {
1645 $s .= $table[$i][(int)floor( $num / $pow10 )];
1647 $num = $num %
$pow10;
1653 * Hebrew Gematria number formatting up to 9999
1659 static function hebrewNumeral( $num ) {
1660 static $table = array(
1661 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1662 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1663 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1664 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1667 $num = intval( $num );
1668 if ( $num > 9999 ||
$num <= 0 ) {
1673 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1674 if ( $num >= $pow10 ) {
1675 if ( $num == 15 ||
$num == 16 ) {
1676 $s .= $table[0][9] . $table[0][$num - 9];
1679 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1680 if ( $pow10 == 1000 ) {
1685 $num = $num %
$pow10;
1687 if ( strlen( $s ) == 2 ) {
1690 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1691 $str .= substr( $s, strlen( $s ) - 2, 2 );
1693 $start = substr( $str, 0, strlen( $str ) - 2 );
1694 $end = substr( $str, strlen( $str ) - 2 );
1697 $str = $start . 'ך';
1700 $str = $start . 'ם';
1703 $str = $start . 'ן';
1706 $str = $start . 'ף';
1709 $str = $start . 'ץ';
1716 * Used by date() and time() to adjust the time output.
1718 * @param $ts Int the time in date('YmdHis') format
1719 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1720 * get user timecorrection setting)
1723 function userAdjust( $ts, $tz = false ) {
1724 global $wgUser, $wgLocalTZoffset;
1726 if ( $tz === false ) {
1727 $tz = $wgUser->getOption( 'timecorrection' );
1730 $data = explode( '|', $tz, 3 );
1732 if ( $data[0] == 'ZoneInfo' ) {
1733 wfSuppressWarnings();
1734 $userTZ = timezone_open( $data[2] );
1735 wfRestoreWarnings();
1736 if ( $userTZ !== false ) {
1737 $date = date_create( $ts, timezone_open( 'UTC' ) );
1738 date_timezone_set( $date, $userTZ );
1739 $date = date_format( $date, 'YmdHis' );
1742 # Unrecognized timezone, default to 'Offset' with the stored offset.
1743 $data[0] = 'Offset';
1747 if ( $data[0] == 'System' ||
$tz == '' ) {
1748 # Global offset in minutes.
1749 if ( isset( $wgLocalTZoffset ) ) {
1750 $minDiff = $wgLocalTZoffset;
1752 } elseif ( $data[0] == 'Offset' ) {
1753 $minDiff = intval( $data[1] );
1755 $data = explode( ':', $tz );
1756 if ( count( $data ) == 2 ) {
1757 $data[0] = intval( $data[0] );
1758 $data[1] = intval( $data[1] );
1759 $minDiff = abs( $data[0] ) * 60 +
$data[1];
1760 if ( $data[0] < 0 ) {
1761 $minDiff = -$minDiff;
1764 $minDiff = intval( $data[0] ) * 60;
1768 # No difference ? Return time unchanged
1769 if ( 0 == $minDiff ) {
1773 wfSuppressWarnings(); // E_STRICT system time bitching
1774 # Generate an adjusted date; take advantage of the fact that mktime
1775 # will normalize out-of-range values so we don't have to split $minDiff
1776 # into hours and minutes.
1778 (int)substr( $ts, 8, 2 ) ), # Hours
1779 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
1780 (int)substr( $ts, 12, 2 ), # Seconds
1781 (int)substr( $ts, 4, 2 ), # Month
1782 (int)substr( $ts, 6, 2 ), # Day
1783 (int)substr( $ts, 0, 4 ) ); # Year
1785 $date = date( 'YmdHis', $t );
1786 wfRestoreWarnings();
1792 * This is meant to be used by time(), date(), and timeanddate() to get
1793 * the date preference they're supposed to use, it should be used in
1797 * function timeanddate([...], $format = true) {
1798 * $datePreference = $this->dateFormat($format);
1803 * @param $usePrefs Mixed: if true, the user's preference is used
1804 * if false, the site/language default is used
1805 * if int/string, assumed to be a format.
1808 function dateFormat( $usePrefs = true ) {
1811 if ( is_bool( $usePrefs ) ) {
1813 $datePreference = $wgUser->getDatePreference();
1815 $datePreference = (string)User
::getDefaultOption( 'date' );
1818 $datePreference = (string)$usePrefs;
1822 if ( $datePreference == '' ) {
1826 return $datePreference;
1830 * Get a format string for a given type and preference
1831 * @param $type string May be date, time or both
1832 * @param $pref string The format name as it appears in Messages*.php
1836 function getDateFormatString( $type, $pref ) {
1837 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
1838 if ( $pref == 'default' ) {
1839 $pref = $this->getDefaultDateFormat();
1840 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1842 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1843 if ( is_null( $df ) ) {
1844 $pref = $this->getDefaultDateFormat();
1845 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1848 $this->dateFormatStrings
[$type][$pref] = $df;
1850 return $this->dateFormatStrings
[$type][$pref];
1854 * @param $ts Mixed: the time format which needs to be turned into a
1855 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1856 * @param $adj Bool: whether to adjust the time output according to the
1857 * user configured offset ($timecorrection)
1858 * @param $format Mixed: true to use user's date format preference
1859 * @param $timecorrection String|bool the time offset as returned by
1860 * validateTimeZone() in Special:Preferences
1863 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1864 $ts = wfTimestamp( TS_MW
, $ts );
1866 $ts = $this->userAdjust( $ts, $timecorrection );
1868 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1869 return $this->sprintfDate( $df, $ts );
1873 * @param $ts Mixed: the time format which needs to be turned into a
1874 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1875 * @param $adj Bool: whether to adjust the time output according to the
1876 * user configured offset ($timecorrection)
1877 * @param $format Mixed: true to use user's date format preference
1878 * @param $timecorrection String|bool the time offset as returned by
1879 * validateTimeZone() in Special:Preferences
1882 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1883 $ts = wfTimestamp( TS_MW
, $ts );
1885 $ts = $this->userAdjust( $ts, $timecorrection );
1887 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1888 return $this->sprintfDate( $df, $ts );
1892 * @param $ts Mixed: the time format which needs to be turned into a
1893 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1894 * @param $adj Bool: whether to adjust the time output according to the
1895 * user configured offset ($timecorrection)
1896 * @param $format Mixed: what format to return, if it's false output the
1897 * default one (default true)
1898 * @param $timecorrection String|bool the time offset as returned by
1899 * validateTimeZone() in Special:Preferences
1902 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1903 $ts = wfTimestamp( TS_MW
, $ts );
1905 $ts = $this->userAdjust( $ts, $timecorrection );
1907 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1908 return $this->sprintfDate( $df, $ts );
1912 * Internal helper function for userDate(), userTime() and userTimeAndDate()
1914 * @param $type String: can be 'date', 'time' or 'both'
1915 * @param $ts Mixed: the time format which needs to be turned into a
1916 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1917 * @param $user User object used to get preferences for timezone and format
1918 * @param $options Array, can contain the following keys:
1919 * - 'timecorrection': time correction, can have the following values:
1920 * - true: use user's preference
1921 * - false: don't use time correction
1922 * - integer: value of time correction in minutes
1923 * - 'format': format to use, can have the following values:
1924 * - true: use user's preference
1925 * - false: use default preference
1926 * - string: format to use
1930 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
1931 $ts = wfTimestamp( TS_MW
, $ts );
1932 $options +
= array( 'timecorrection' => true, 'format' => true );
1933 if ( $options['timecorrection'] !== false ) {
1934 if ( $options['timecorrection'] === true ) {
1935 $offset = $user->getOption( 'timecorrection' );
1937 $offset = $options['timecorrection'];
1939 $ts = $this->userAdjust( $ts, $offset );
1941 if ( $options['format'] === true ) {
1942 $format = $user->getDatePreference();
1944 $format = $options['format'];
1946 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
1947 return $this->sprintfDate( $df, $ts );
1951 * Get the formatted date for the given timestamp and formatted for
1954 * @param $ts Mixed: the time format which needs to be turned into a
1955 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1956 * @param $user User object used to get preferences for timezone and format
1957 * @param $options Array, can contain the following keys:
1958 * - 'timecorrection': time correction, can have the following values:
1959 * - true: use user's preference
1960 * - false: don't use time correction
1961 * - integer: value of time correction in minutes
1962 * - 'format': format to use, can have the following values:
1963 * - true: use user's preference
1964 * - false: use default preference
1965 * - string: format to use
1969 public function userDate( $ts, User
$user, array $options = array() ) {
1970 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
1974 * Get the formatted time for the given timestamp and formatted for
1977 * @param $ts Mixed: the time format which needs to be turned into a
1978 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1979 * @param $user User object used to get preferences for timezone and format
1980 * @param $options Array, can contain the following keys:
1981 * - 'timecorrection': time correction, can have the following values:
1982 * - true: use user's preference
1983 * - false: don't use time correction
1984 * - integer: value of time correction in minutes
1985 * - 'format': format to use, can have the following values:
1986 * - true: use user's preference
1987 * - false: use default preference
1988 * - string: format to use
1992 public function userTime( $ts, User
$user, array $options = array() ) {
1993 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
1997 * Get the formatted date and time for the given timestamp and formatted for
2000 * @param $ts Mixed: the time format which needs to be turned into a
2001 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2002 * @param $user User object used to get preferences for timezone and format
2003 * @param $options Array, can contain the following keys:
2004 * - 'timecorrection': time correction, can have the following values:
2005 * - true: use user's preference
2006 * - false: don't use time correction
2007 * - integer: value of time correction in minutes
2008 * - 'format': format to use, can have the following values:
2009 * - true: use user's preference
2010 * - false: use default preference
2011 * - string: format to use
2015 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2016 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2020 * @param $key string
2021 * @return array|null
2023 function getMessage( $key ) {
2024 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2030 function getAllMessages() {
2031 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2040 function iconv( $in, $out, $string ) {
2041 # This is a wrapper for iconv in all languages except esperanto,
2042 # which does some nasty x-conversions beforehand
2044 # Even with //IGNORE iconv can whine about illegal characters in
2045 # *input* string. We just ignore those too.
2046 # REF: http://bugs.php.net/bug.php?id=37166
2047 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2048 wfSuppressWarnings();
2049 $text = iconv( $in, $out . '//IGNORE', $string );
2050 wfRestoreWarnings();
2054 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2057 * @param $matches array
2058 * @return mixed|string
2060 function ucwordbreaksCallbackAscii( $matches ) {
2061 return $this->ucfirst( $matches[1] );
2065 * @param $matches array
2068 function ucwordbreaksCallbackMB( $matches ) {
2069 return mb_strtoupper( $matches[0] );
2073 * @param $matches array
2076 function ucCallback( $matches ) {
2077 list( $wikiUpperChars ) = self
::getCaseMaps();
2078 return strtr( $matches[1], $wikiUpperChars );
2082 * @param $matches array
2085 function lcCallback( $matches ) {
2086 list( , $wikiLowerChars ) = self
::getCaseMaps();
2087 return strtr( $matches[1], $wikiLowerChars );
2091 * @param $matches array
2094 function ucwordsCallbackMB( $matches ) {
2095 return mb_strtoupper( $matches[0] );
2099 * @param $matches array
2102 function ucwordsCallbackWiki( $matches ) {
2103 list( $wikiUpperChars ) = self
::getCaseMaps();
2104 return strtr( $matches[0], $wikiUpperChars );
2108 * Make a string's first character uppercase
2110 * @param $str string
2114 function ucfirst( $str ) {
2116 if ( $o < 96 ) { // if already uppercase...
2118 } elseif ( $o < 128 ) {
2119 return ucfirst( $str ); // use PHP's ucfirst()
2121 // fall back to more complex logic in case of multibyte strings
2122 return $this->uc( $str, true );
2127 * Convert a string to uppercase
2129 * @param $str string
2130 * @param $first bool
2134 function uc( $str, $first = false ) {
2135 if ( function_exists( 'mb_strtoupper' ) ) {
2137 if ( $this->isMultibyte( $str ) ) {
2138 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2140 return ucfirst( $str );
2143 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2146 if ( $this->isMultibyte( $str ) ) {
2147 $x = $first ?
'^' : '';
2148 return preg_replace_callback(
2149 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2150 array( $this, 'ucCallback' ),
2154 return $first ?
ucfirst( $str ) : strtoupper( $str );
2160 * @param $str string
2161 * @return mixed|string
2163 function lcfirst( $str ) {
2166 return strval( $str );
2167 } elseif ( $o >= 128 ) {
2168 return $this->lc( $str, true );
2169 } elseif ( $o > 96 ) {
2172 $str[0] = strtolower( $str[0] );
2178 * @param $str string
2179 * @param $first bool
2180 * @return mixed|string
2182 function lc( $str, $first = false ) {
2183 if ( function_exists( 'mb_strtolower' ) ) {
2185 if ( $this->isMultibyte( $str ) ) {
2186 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2188 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2191 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2194 if ( $this->isMultibyte( $str ) ) {
2195 $x = $first ?
'^' : '';
2196 return preg_replace_callback(
2197 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2198 array( $this, 'lcCallback' ),
2202 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2208 * @param $str string
2211 function isMultibyte( $str ) {
2212 return (bool)preg_match( '/[\x80-\xff]/', $str );
2216 * @param $str string
2217 * @return mixed|string
2219 function ucwords( $str ) {
2220 if ( $this->isMultibyte( $str ) ) {
2221 $str = $this->lc( $str );
2223 // regexp to find first letter in each word (i.e. after each space)
2224 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2226 // function to use to capitalize a single char
2227 if ( function_exists( 'mb_strtoupper' ) ) {
2228 return preg_replace_callback(
2230 array( $this, 'ucwordsCallbackMB' ),
2234 return preg_replace_callback(
2236 array( $this, 'ucwordsCallbackWiki' ),
2241 return ucwords( strtolower( $str ) );
2246 * capitalize words at word breaks
2248 * @param $str string
2251 function ucwordbreaks( $str ) {
2252 if ( $this->isMultibyte( $str ) ) {
2253 $str = $this->lc( $str );
2255 // since \b doesn't work for UTF-8, we explicitely define word break chars
2256 $breaks = "[ \-\(\)\}\{\.,\?!]";
2258 // find first letter after word break
2259 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2261 if ( function_exists( 'mb_strtoupper' ) ) {
2262 return preg_replace_callback(
2264 array( $this, 'ucwordbreaksCallbackMB' ),
2268 return preg_replace_callback(
2270 array( $this, 'ucwordsCallbackWiki' ),
2275 return preg_replace_callback(
2276 '/\b([\w\x80-\xff]+)\b/',
2277 array( $this, 'ucwordbreaksCallbackAscii' ),
2284 * Return a case-folded representation of $s
2286 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2287 * and $s2 are the same except for the case of their characters. It is not
2288 * necessary for the value returned to make sense when displayed.
2290 * Do *not* perform any other normalisation in this function. If a caller
2291 * uses this function when it should be using a more general normalisation
2292 * function, then fix the caller.
2298 function caseFold( $s ) {
2299 return $this->uc( $s );
2306 function checkTitleEncoding( $s ) {
2307 if ( is_array( $s ) ) {
2308 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2310 # Check for non-UTF-8 URLs
2311 $ishigh = preg_match( '/[\x80-\xff]/', $s );
2316 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2317 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
2322 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2328 function fallback8bitEncoding() {
2329 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2333 * Most writing systems use whitespace to break up words.
2334 * Some languages such as Chinese don't conventionally do this,
2335 * which requires special handling when breaking up words for
2340 function hasWordBreaks() {
2345 * Some languages such as Chinese require word segmentation,
2346 * Specify such segmentation when overridden in derived class.
2348 * @param $string String
2351 function segmentByWord( $string ) {
2356 * Some languages have special punctuation need to be normalized.
2357 * Make such changes here.
2359 * @param $string String
2362 function normalizeForSearch( $string ) {
2363 return self
::convertDoubleWidth( $string );
2367 * convert double-width roman characters to single-width.
2368 * range: ff00-ff5f ~= 0020-007f
2370 * @param $string string
2374 protected static function convertDoubleWidth( $string ) {
2375 static $full = null;
2376 static $half = null;
2378 if ( $full === null ) {
2379 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2380 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2381 $full = str_split( $fullWidth, 3 );
2382 $half = str_split( $halfWidth );
2385 $string = str_replace( $full, $half, $string );
2390 * @param $string string
2391 * @param $pattern string
2394 protected static function insertSpace( $string, $pattern ) {
2395 $string = preg_replace( $pattern, " $1 ", $string );
2396 $string = preg_replace( '/ +/', ' ', $string );
2401 * @param $termsArray array
2404 function convertForSearchResult( $termsArray ) {
2405 # some languages, e.g. Chinese, need to do a conversion
2406 # in order for search results to be displayed correctly
2411 * Get the first character of a string.
2416 function firstChar( $s ) {
2419 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2420 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2425 if ( isset( $matches[1] ) ) {
2426 if ( strlen( $matches[1] ) != 3 ) {
2430 // Break down Hangul syllables to grab the first jamo
2431 $code = utf8ToCodepoint( $matches[1] );
2432 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2434 } elseif ( $code < 0xb098 ) {
2435 return "\xe3\x84\xb1";
2436 } elseif ( $code < 0xb2e4 ) {
2437 return "\xe3\x84\xb4";
2438 } elseif ( $code < 0xb77c ) {
2439 return "\xe3\x84\xb7";
2440 } elseif ( $code < 0xb9c8 ) {
2441 return "\xe3\x84\xb9";
2442 } elseif ( $code < 0xbc14 ) {
2443 return "\xe3\x85\x81";
2444 } elseif ( $code < 0xc0ac ) {
2445 return "\xe3\x85\x82";
2446 } elseif ( $code < 0xc544 ) {
2447 return "\xe3\x85\x85";
2448 } elseif ( $code < 0xc790 ) {
2449 return "\xe3\x85\x87";
2450 } elseif ( $code < 0xcc28 ) {
2451 return "\xe3\x85\x88";
2452 } elseif ( $code < 0xce74 ) {
2453 return "\xe3\x85\x8a";
2454 } elseif ( $code < 0xd0c0 ) {
2455 return "\xe3\x85\x8b";
2456 } elseif ( $code < 0xd30c ) {
2457 return "\xe3\x85\x8c";
2458 } elseif ( $code < 0xd558 ) {
2459 return "\xe3\x85\x8d";
2461 return "\xe3\x85\x8e";
2468 function initEncoding() {
2469 # Some languages may have an alternate char encoding option
2470 # (Esperanto X-coding, Japanese furigana conversion, etc)
2471 # If this language is used as the primary content language,
2472 # an override to the defaults can be set here on startup.
2479 function recodeForEdit( $s ) {
2480 # For some languages we'll want to explicitly specify
2481 # which characters make it into the edit box raw
2482 # or are converted in some way or another.
2483 global $wgEditEncoding;
2484 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2487 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2495 function recodeInput( $s ) {
2496 # Take the previous into account.
2497 global $wgEditEncoding;
2498 if ( $wgEditEncoding != '' ) {
2499 $enc = $wgEditEncoding;
2503 if ( $enc == 'UTF-8' ) {
2506 return $this->iconv( $enc, 'UTF-8', $s );
2511 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2512 * also cleans up certain backwards-compatible sequences, converting them
2513 * to the modern Unicode equivalent.
2515 * This is language-specific for performance reasons only.
2521 function normalize( $s ) {
2522 global $wgAllUnicodeFixes;
2523 $s = UtfNormal
::cleanUp( $s );
2524 if ( $wgAllUnicodeFixes ) {
2525 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2526 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2533 * Transform a string using serialized data stored in the given file (which
2534 * must be in the serialized subdirectory of $IP). The file contains pairs
2535 * mapping source characters to destination characters.
2537 * The data is cached in process memory. This will go faster if you have the
2538 * FastStringSearch extension.
2540 * @param $file string
2541 * @param $string string
2543 * @throws MWException
2546 function transformUsingPairFile( $file, $string ) {
2547 if ( !isset( $this->transformData
[$file] ) ) {
2548 $data = wfGetPrecompiledData( $file );
2549 if ( $data === false ) {
2550 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
2552 $this->transformData
[$file] = new ReplacementArray( $data );
2554 return $this->transformData
[$file]->replace( $string );
2558 * For right-to-left language support
2563 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
2567 * Return the correct HTML 'dir' attribute value for this language.
2571 return $this->isRTL() ?
'rtl' : 'ltr';
2575 * Return 'left' or 'right' as appropriate alignment for line-start
2576 * for this language's text direction.
2578 * Should be equivalent to CSS3 'start' text-align value....
2582 function alignStart() {
2583 return $this->isRTL() ?
'right' : 'left';
2587 * Return 'right' or 'left' as appropriate alignment for line-end
2588 * for this language's text direction.
2590 * Should be equivalent to CSS3 'end' text-align value....
2594 function alignEnd() {
2595 return $this->isRTL() ?
'left' : 'right';
2599 * A hidden direction mark (LRM or RLM), depending on the language direction.
2600 * Unlike getDirMark(), this function returns the character as an HTML entity.
2601 * This function should be used when the output is guaranteed to be HTML,
2602 * because it makes the output HTML source code more readable. When
2603 * the output is plain text or can be escaped, getDirMark() should be used.
2605 * @param $opposite Boolean Get the direction mark opposite to your language
2608 function getDirMarkEntity( $opposite = false ) {
2609 if ( $opposite ) { return $this->isRTL() ?
'‎' : '‏'; }
2610 return $this->isRTL() ?
'‏' : '‎';
2614 * A hidden direction mark (LRM or RLM), depending on the language direction.
2615 * This function produces them as invisible Unicode characters and
2616 * the output may be hard to read and debug, so it should only be used
2617 * when the output is plain text or can be escaped. When the output is
2618 * HTML, use getDirMarkEntity() instead.
2620 * @param $opposite Boolean Get the direction mark opposite to your language
2623 function getDirMark( $opposite = false ) {
2624 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2625 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2626 if ( $opposite ) { return $this->isRTL() ?
$lrm : $rlm; }
2627 return $this->isRTL() ?
$rlm : $lrm;
2633 function capitalizeAllNouns() {
2634 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2638 * An arrow, depending on the language direction
2642 function getArrow() {
2643 return $this->isRTL() ?
'←' : '→';
2647 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2651 function linkPrefixExtension() {
2652 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2658 function getMagicWords() {
2659 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2662 protected function doMagicHook() {
2663 if ( $this->mMagicHookDone
) {
2666 $this->mMagicHookDone
= true;
2667 wfProfileIn( 'LanguageGetMagic' );
2668 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2669 wfProfileOut( 'LanguageGetMagic' );
2673 * Fill a MagicWord object with data from here
2677 function getMagic( $mw ) {
2678 $this->doMagicHook();
2680 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
2681 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
2683 $magicWords = $this->getMagicWords();
2684 if ( isset( $magicWords[$mw->mId
] ) ) {
2685 $rawEntry = $magicWords[$mw->mId
];
2691 if ( !is_array( $rawEntry ) ) {
2692 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
2694 $mw->mCaseSensitive
= $rawEntry[0];
2695 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
2700 * Add magic words to the extension array
2702 * @param $newWords array
2704 function addMagicWordsByLang( $newWords ) {
2705 $fallbackChain = $this->getFallbackLanguages();
2706 $fallbackChain = array_reverse( $fallbackChain );
2707 foreach ( $fallbackChain as $code ) {
2708 if ( isset( $newWords[$code] ) ) {
2709 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
2715 * Get special page names, as an associative array
2716 * case folded alias => real name
2718 function getSpecialPageAliases() {
2719 // Cache aliases because it may be slow to load them
2720 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
2722 $this->mExtendedSpecialPageAliases
=
2723 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
2724 wfRunHooks( 'LanguageGetSpecialPageAliases',
2725 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
2728 return $this->mExtendedSpecialPageAliases
;
2732 * Italic is unsuitable for some languages
2734 * @param $text String: the text to be emphasized.
2737 function emphasize( $text ) {
2738 return "<em>$text</em>";
2742 * Normally we output all numbers in plain en_US style, that is
2743 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2744 * point twohundredthirtyfive. However this is not suitable for all
2745 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2746 * Icelandic just want to use commas instead of dots, and dots instead
2747 * of commas like "293.291,235".
2749 * An example of this function being called:
2751 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2754 * See LanguageGu.php for the Gujarati implementation and
2755 * $separatorTransformTable on MessageIs.php for
2756 * the , => . and . => , implementation.
2758 * @todo check if it's viable to use localeconv() for the decimal
2760 * @param $number Mixed: the string to be formatted, should be an integer
2761 * or a floating point number.
2762 * @param $nocommafy Bool: set to true for special numbers like dates
2765 public function formatNum( $number, $nocommafy = false ) {
2766 global $wgTranslateNumerals;
2767 if ( !$nocommafy ) {
2768 $number = $this->commafy( $number );
2769 $s = $this->separatorTransformTable();
2771 $number = strtr( $number, $s );
2775 if ( $wgTranslateNumerals ) {
2776 $s = $this->digitTransformTable();
2778 $number = strtr( $number, $s );
2786 * @param $number string
2789 function parseFormattedNumber( $number ) {
2790 $s = $this->digitTransformTable();
2792 $number = strtr( $number, array_flip( $s ) );
2795 $s = $this->separatorTransformTable();
2797 $number = strtr( $number, array_flip( $s ) );
2800 $number = strtr( $number, array( ',' => '' ) );
2805 * Adds commas to a given number
2810 function commafy( $_ ) {
2811 $digitGroupingPattern = $this->digitGroupingPattern();
2812 if ( $_ === null ) {
2816 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
2817 // default grouping is at thousands, use the same for ###,###,### pattern too.
2818 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2820 // Ref: http://cldr.unicode.org/translation/number-patterns
2822 if ( intval( $_ ) < 0 ) {
2823 // For negative numbers apply the algorithm like positive number and add sign.
2825 $_ = substr( $_, 1 );
2827 $numberpart = array();
2828 $decimalpart = array();
2829 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
2830 preg_match( "/\d+/", $_, $numberpart );
2831 preg_match( "/\.\d*/", $_, $decimalpart );
2832 $groupedNumber = ( count( $decimalpart ) > 0 ) ?
$decimalpart[0]:"";
2833 if ( $groupedNumber === $_ ) {
2834 // the string does not have any number part. Eg: .12345
2835 return $sign . $groupedNumber;
2837 $start = $end = strlen( $numberpart[0] );
2838 while ( $start > 0 ) {
2839 $match = $matches[0][$numMatches -1] ;
2840 $matchLen = strlen( $match );
2841 $start = $end - $matchLen;
2845 $groupedNumber = substr( $_ , $start, $end -$start ) . $groupedNumber ;
2847 if ( $numMatches > 1 ) {
2848 // use the last pattern for the rest of the number
2852 $groupedNumber = "," . $groupedNumber;
2855 return $sign . $groupedNumber;
2861 function digitGroupingPattern() {
2862 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
2868 function digitTransformTable() {
2869 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
2875 function separatorTransformTable() {
2876 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
2880 * Take a list of strings and build a locale-friendly comma-separated
2881 * list, using the local comma-separator message.
2882 * The last two strings are chained with an "and".
2887 function listToText( array $l ) {
2889 $m = count( $l ) - 1;
2891 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2893 for ( $i = $m; $i >= 0; $i-- ) {
2896 } elseif ( $i == $m - 1 ) {
2897 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2899 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2907 * Take a list of strings and build a locale-friendly comma-separated
2908 * list, using the local comma-separator message.
2909 * @param $list array of strings to put in a comma list
2912 function commaList( array $list ) {
2916 array( 'parsemag', 'escapenoentities', 'language' => $this )
2923 * Take a list of strings and build a locale-friendly semicolon-separated
2924 * list, using the local semicolon-separator message.
2925 * @param $list array of strings to put in a semicolon list
2928 function semicolonList( array $list ) {
2931 'semicolon-separator',
2932 array( 'parsemag', 'escapenoentities', 'language' => $this )
2939 * Same as commaList, but separate it with the pipe instead.
2940 * @param $list array of strings to put in a pipe list
2943 function pipeList( array $list ) {
2947 array( 'escapenoentities', 'language' => $this )
2954 * Truncate a string to a specified length in bytes, appending an optional
2955 * string (e.g. for ellipses)
2957 * The database offers limited byte lengths for some columns in the database;
2958 * multi-byte character sets mean we need to ensure that only whole characters
2959 * are included, otherwise broken characters can be passed to the user
2961 * If $length is negative, the string will be truncated from the beginning
2963 * @param $string String to truncate
2964 * @param $length Int: maximum length (including ellipses)
2965 * @param $ellipsis String to append to the truncated text
2966 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
2967 * $adjustLength was introduced in 1.18, before that behaved as if false.
2970 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
2971 # Use the localized ellipsis character
2972 if ( $ellipsis == '...' ) {
2973 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2975 # Check if there is no need to truncate
2976 if ( $length == 0 ) {
2977 return $ellipsis; // convention
2978 } elseif ( strlen( $string ) <= abs( $length ) ) {
2979 return $string; // no need to truncate
2981 $stringOriginal = $string;
2982 # If ellipsis length is >= $length then we can't apply $adjustLength
2983 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
2984 $string = $ellipsis; // this can be slightly unexpected
2985 # Otherwise, truncate and add ellipsis...
2987 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
2988 if ( $length > 0 ) {
2989 $length -= $eLength;
2990 $string = substr( $string, 0, $length ); // xyz...
2991 $string = $this->removeBadCharLast( $string );
2992 $string = $string . $ellipsis;
2994 $length +
= $eLength;
2995 $string = substr( $string, $length ); // ...xyz
2996 $string = $this->removeBadCharFirst( $string );
2997 $string = $ellipsis . $string;
3000 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3001 # This check is *not* redundant if $adjustLength, due to the single case where
3002 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3003 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3006 return $stringOriginal;
3011 * Remove bytes that represent an incomplete Unicode character
3012 * at the end of string (e.g. bytes of the char are missing)
3014 * @param $string String
3017 protected function removeBadCharLast( $string ) {
3018 if ( $string != '' ) {
3019 $char = ord( $string[strlen( $string ) - 1] );
3021 if ( $char >= 0xc0 ) {
3022 # We got the first byte only of a multibyte char; remove it.
3023 $string = substr( $string, 0, -1 );
3024 } elseif ( $char >= 0x80 &&
3025 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3026 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
3028 # We chopped in the middle of a character; remove it
3036 * Remove bytes that represent an incomplete Unicode character
3037 * at the start of string (e.g. bytes of the char are missing)
3039 * @param $string String
3042 protected function removeBadCharFirst( $string ) {
3043 if ( $string != '' ) {
3044 $char = ord( $string[0] );
3045 if ( $char >= 0x80 && $char < 0xc0 ) {
3046 # We chopped in the middle of a character; remove the whole thing
3047 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3054 * Truncate a string of valid HTML to a specified length in bytes,
3055 * appending an optional string (e.g. for ellipses), and return valid HTML
3057 * This is only intended for styled/linked text, such as HTML with
3058 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3059 * Also, this will not detect things like "display:none" CSS.
3061 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3063 * @param string $text HTML string to truncate
3064 * @param int $length (zero/positive) Maximum length (including ellipses)
3065 * @param string $ellipsis String to append to the truncated text
3068 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3069 # Use the localized ellipsis character
3070 if ( $ellipsis == '...' ) {
3071 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
3073 # Check if there is clearly no need to truncate
3074 if ( $length <= 0 ) {
3075 return $ellipsis; // no text shown, nothing to format (convention)
3076 } elseif ( strlen( $text ) <= $length ) {
3077 return $text; // string short enough even *with* HTML (short-circuit)
3080 $dispLen = 0; // innerHTML legth so far
3081 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3082 $tagType = 0; // 0-open, 1-close
3083 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3084 $entityState = 0; // 0-not entity, 1-entity
3085 $tag = $ret = ''; // accumulated tag name, accumulated result string
3086 $openTags = array(); // open tag stack
3087 $maybeState = null; // possible truncation state
3089 $textLen = strlen( $text );
3090 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3091 for ( $pos = 0; true; ++
$pos ) {
3092 # Consider truncation once the display length has reached the maximim.
3093 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3094 # Check that we're not in the middle of a bracket/entity...
3095 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3096 if ( !$testingEllipsis ) {
3097 $testingEllipsis = true;
3098 # Save where we are; we will truncate here unless there turn out to
3099 # be so few remaining characters that truncation is not necessary.
3100 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3101 $maybeState = array( $ret, $openTags ); // save state
3103 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3104 # String in fact does need truncation, the truncation point was OK.
3105 list( $ret, $openTags ) = $maybeState; // reload state
3106 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3107 $ret .= $ellipsis; // add ellipsis
3111 if ( $pos >= $textLen ) break; // extra iteration just for above checks
3113 # Read the next char...
3115 $lastCh = $pos ?
$text[$pos - 1] : '';
3116 $ret .= $ch; // add to result string
3118 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3119 $entityState = 0; // for bad HTML
3120 $bracketState = 1; // tag started (checking for backslash)
3121 } elseif ( $ch == '>' ) {
3122 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3123 $entityState = 0; // for bad HTML
3124 $bracketState = 0; // out of brackets
3125 } elseif ( $bracketState == 1 ) {
3127 $tagType = 1; // close tag (e.g. "</span>")
3129 $tagType = 0; // open tag (e.g. "<span>")
3132 $bracketState = 2; // building tag name
3133 } elseif ( $bracketState == 2 ) {
3137 // Name found (e.g. "<a href=..."), add on tag attributes...
3138 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3140 } elseif ( $bracketState == 0 ) {
3141 if ( $entityState ) {
3144 $dispLen++
; // entity is one displayed char
3147 if ( $neLength == 0 && !$maybeState ) {
3148 // Save state without $ch. We want to *hit* the first
3149 // display char (to get tags) but not *use* it if truncating.
3150 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3153 $entityState = 1; // entity found, (e.g. " ")
3155 $dispLen++
; // this char is displayed
3156 // Add the next $max display text chars after this in one swoop...
3157 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3158 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3159 $dispLen +
= $skipped;
3165 // Close the last tag if left unclosed by bad HTML
3166 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3167 while ( count( $openTags ) > 0 ) {
3168 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3174 * truncateHtml() helper function
3175 * like strcspn() but adds the skipped chars to $ret
3184 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3185 if ( $len === null ) {
3186 $len = -1; // -1 means "no limit" for strcspn
3187 } elseif ( $len < 0 ) {
3191 if ( $start < strlen( $text ) ) {
3192 $skipCount = strcspn( $text, $search, $start, $len );
3193 $ret .= substr( $text, $start, $skipCount );
3199 * truncateHtml() helper function
3200 * (a) push or pop $tag from $openTags as needed
3201 * (b) clear $tag value
3202 * @param &$tag string Current HTML tag name we are looking at
3203 * @param $tagType int (0-open tag, 1-close tag)
3204 * @param $lastCh string Character before the '>' that ended this tag
3205 * @param &$openTags array Open tag stack (not accounting for $tag)
3207 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3208 $tag = ltrim( $tag );
3210 if ( $tagType == 0 && $lastCh != '/' ) {
3211 $openTags[] = $tag; // tag opened (didn't close itself)
3212 } elseif ( $tagType == 1 ) {
3213 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3214 array_pop( $openTags ); // tag closed
3222 * Grammatical transformations, needed for inflected languages
3223 * Invoked by putting {{grammar:case|word}} in a message
3225 * @param $word string
3226 * @param $case string
3229 function convertGrammar( $word, $case ) {
3230 global $wgGrammarForms;
3231 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3232 return $wgGrammarForms[$this->getCode()][$case][$word];
3237 * Get the grammar forms for the content language
3238 * @return array of grammar forms
3241 function getGrammarForms() {
3242 global $wgGrammarForms;
3243 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3244 return $wgGrammarForms[$this->getCode()];
3249 * Provides an alternative text depending on specified gender.
3250 * Usage {{gender:username|masculine|feminine|neutral}}.
3251 * username is optional, in which case the gender of current user is used,
3252 * but only in (some) interface messages; otherwise default gender is used.
3254 * If no forms are given, an empty string is returned. If only one form is
3255 * given, it will be returned unconditionally. These details are implied by
3256 * the caller and cannot be overridden in subclasses.
3258 * If more than one form is given, the default is to use the neutral one
3259 * if it is specified, and to use the masculine one otherwise. These
3260 * details can be overridden in subclasses.
3262 * @param $gender string
3263 * @param $forms array
3267 function gender( $gender, $forms ) {
3268 if ( !count( $forms ) ) {
3271 $forms = $this->preConvertPlural( $forms, 2 );
3272 if ( $gender === 'male' ) {
3275 if ( $gender === 'female' ) {
3278 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3282 * Plural form transformations, needed for some languages.
3283 * For example, there are 3 form of plural in Russian and Polish,
3284 * depending on "count mod 10". See [[w:Plural]]
3285 * For English it is pretty simple.
3287 * Invoked by putting {{plural:count|wordform1|wordform2}}
3288 * or {{plural:count|wordform1|wordform2|wordform3}}
3290 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3292 * @param $count Integer: non-localized number
3293 * @param $forms Array: different plural forms
3294 * @return string Correct form of plural for $count in this language
3296 function convertPlural( $count, $forms ) {
3297 if ( !count( $forms ) ) {
3300 $forms = $this->preConvertPlural( $forms, 2 );
3302 return ( $count == 1 ) ?
$forms[0] : $forms[1];
3306 * Checks that convertPlural was given an array and pads it to requested
3307 * amount of forms by copying the last one.
3309 * @param $count Integer: How many forms should there be at least
3310 * @param $forms Array of forms given to convertPlural
3311 * @return array Padded array of forms or an exception if not an array
3313 protected function preConvertPlural( /* Array */ $forms, $count ) {
3314 while ( count( $forms ) < $count ) {
3315 $forms[] = $forms[count( $forms ) - 1];
3321 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3322 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3323 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3324 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3327 * @param $str String: the validated block duration in English
3328 * @return string Somehow translated block duration
3329 * @see LanguageFi.php for example implementation
3331 function translateBlockExpiry( $str ) {
3332 $duration = SpecialBlock
::getSuggestedDurations( $this );
3333 foreach ( $duration as $show => $value ) {
3334 if ( strcmp( $str, $value ) == 0 ) {
3335 return htmlspecialchars( trim( $show ) );
3339 // Since usually only infinite or indefinite is only on list, so try
3340 // equivalents if still here.
3341 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3342 if ( in_array( $str, $indefs ) ) {
3343 foreach ( $indefs as $val ) {
3344 $show = array_search( $val, $duration, true );
3345 if ( $show !== false ) {
3346 return htmlspecialchars( trim( $show ) );
3350 // If all else fails, return the original string.
3355 * languages like Chinese need to be segmented in order for the diff
3358 * @param $text String
3361 public function segmentForDiff( $text ) {
3366 * and unsegment to show the result
3368 * @param $text String
3371 public function unsegmentForDiff( $text ) {
3376 * Return the LanguageConverter used in the Language
3379 * @return LanguageConverter
3381 public function getConverter() {
3382 return $this->mConverter
;
3386 * convert text to all supported variants
3388 * @param $text string
3391 public function autoConvertToAllVariants( $text ) {
3392 return $this->mConverter
->autoConvertToAllVariants( $text );
3396 * convert text to different variants of a language.
3398 * @param $text string
3401 public function convert( $text ) {
3402 return $this->mConverter
->convert( $text );
3406 * Convert a Title object to a string in the preferred variant
3408 * @param $title Title
3411 public function convertTitle( $title ) {
3412 return $this->mConverter
->convertTitle( $title );
3416 * Check if this is a language with variants
3420 public function hasVariants() {
3421 return sizeof( $this->getVariants() ) > 1;
3425 * Check if the language has the specific variant
3428 * @param $variant string
3431 public function hasVariant( $variant ) {
3432 return (bool)$this->mConverter
->validateVariant( $variant );
3436 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3438 * @param $text string
3441 public function armourMath( $text ) {
3442 return $this->mConverter
->armourMath( $text );
3446 * Perform output conversion on a string, and encode for safe HTML output.
3447 * @param $text String text to be converted
3448 * @param $isTitle Bool whether this conversion is for the article title
3450 * @todo this should get integrated somewhere sane
3452 public function convertHtml( $text, $isTitle = false ) {
3453 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3457 * @param $key string
3460 public function convertCategoryKey( $key ) {
3461 return $this->mConverter
->convertCategoryKey( $key );
3465 * Get the list of variants supported by this language
3466 * see sample implementation in LanguageZh.php
3468 * @return array an array of language codes
3470 public function getVariants() {
3471 return $this->mConverter
->getVariants();
3477 public function getPreferredVariant() {
3478 return $this->mConverter
->getPreferredVariant();
3484 public function getDefaultVariant() {
3485 return $this->mConverter
->getDefaultVariant();
3491 public function getURLVariant() {
3492 return $this->mConverter
->getURLVariant();
3496 * If a language supports multiple variants, it is
3497 * possible that non-existing link in one variant
3498 * actually exists in another variant. this function
3499 * tries to find it. See e.g. LanguageZh.php
3501 * @param $link String: the name of the link
3502 * @param $nt Mixed: the title object of the link
3503 * @param $ignoreOtherCond Boolean: to disable other conditions when
3504 * we need to transclude a template or update a category's link
3505 * @return null the input parameters may be modified upon return
3507 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3508 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
3512 * If a language supports multiple variants, converts text
3513 * into an array of all possible variants of the text:
3514 * 'variant' => text in that variant
3516 * @deprecated since 1.17 Use autoConvertToAllVariants()
3518 * @param $text string
3522 public function convertLinkToAllVariants( $text ) {
3523 return $this->mConverter
->convertLinkToAllVariants( $text );
3527 * returns language specific options used by User::getPageRenderHash()
3528 * for example, the preferred language variant
3532 function getExtraHashOptions() {
3533 return $this->mConverter
->getExtraHashOptions();
3537 * For languages that support multiple variants, the title of an
3538 * article may be displayed differently in different variants. this
3539 * function returns the apporiate title defined in the body of the article.
3543 public function getParsedTitle() {
3544 return $this->mConverter
->getParsedTitle();
3548 * Enclose a string with the "no conversion" tag. This is used by
3549 * various functions in the Parser
3551 * @param $text String: text to be tagged for no conversion
3552 * @param $noParse bool
3553 * @return string the tagged text
3555 public function markNoConversion( $text, $noParse = false ) {
3556 return $this->mConverter
->markNoConversion( $text, $noParse );
3560 * A regular expression to match legal word-trailing characters
3561 * which should be merged onto a link of the form [[foo]]bar.
3565 public function linkTrail() {
3566 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
3572 function getLangObj() {
3577 * Get the RFC 3066 code for this language object
3581 public function getCode() {
3582 return $this->mCode
;
3586 * Get the code in Bcp47 format which we can use
3587 * inside of html lang="" tags.
3591 public function getHtmlCode() {
3592 if ( is_null( $this->mHtmlCode
) ) {
3593 $this->mHtmlCode
= wfBCP47( $this->getCode() );
3595 return $this->mHtmlCode
;
3599 * @param $code string
3601 public function setCode( $code ) {
3602 $this->mCode
= $code;
3603 // Ensure we don't leave an incorrect html code lying around
3604 $this->mHtmlCode
= null;
3608 * Get the name of a file for a certain language code
3609 * @param $prefix string Prepend this to the filename
3610 * @param $code string Language code
3611 * @param $suffix string Append this to the filename
3612 * @throws MWException
3613 * @return string $prefix . $mangledCode . $suffix
3615 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
3616 // Protect against path traversal
3617 if ( !Language
::isValidCode( $code )
3618 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
3620 throw new MWException( "Invalid language code \"$code\"" );
3623 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
3627 * Get the language code from a file name. Inverse of getFileName()
3628 * @param $filename string $prefix . $languageCode . $suffix
3629 * @param $prefix string Prefix before the language code
3630 * @param $suffix string Suffix after the language code
3631 * @return string Language code, or false if $prefix or $suffix isn't found
3633 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
3635 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
3636 preg_quote( $suffix, '/' ) . '/', $filename, $m );
3637 if ( !count( $m ) ) {
3640 return str_replace( '_', '-', strtolower( $m[1] ) );
3644 * @param $code string
3647 public static function getMessagesFileName( $code ) {
3649 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
3650 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
3655 * @param $code string
3658 public static function getClassFileName( $code ) {
3660 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
3664 * Get the first fallback for a given language.
3666 * @param $code string
3668 * @return bool|string
3670 public static function getFallbackFor( $code ) {
3671 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
3674 $fallbacks = self
::getFallbacksFor( $code );
3675 $first = array_shift( $fallbacks );
3681 * Get the ordered list of fallback languages.
3684 * @param $code string Language code
3687 public static function getFallbacksFor( $code ) {
3688 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
3691 $v = self
::getLocalisationCache()->getItem( $code, 'fallback' );
3692 $v = array_map( 'trim', explode( ',', $v ) );
3693 if ( $v[count( $v ) - 1] !== 'en' ) {
3701 * Get all messages for a given language
3702 * WARNING: this may take a long time. If you just need all message *keys*
3703 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
3705 * @param $code string
3709 public static function getMessagesFor( $code ) {
3710 return self
::getLocalisationCache()->getItem( $code, 'messages' );
3714 * Get a message for a given language
3716 * @param $key string
3717 * @param $code string
3721 public static function getMessageFor( $key, $code ) {
3722 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
3726 * Get all message keys for a given language. This is a faster alternative to
3727 * array_keys( Language::getMessagesFor( $code ) )
3730 * @param $code string Language code
3731 * @return array of message keys (strings)
3733 public static function getMessageKeysFor( $code ) {
3734 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
3741 function fixVariableInNamespace( $talk ) {
3742 if ( strpos( $talk, '$1' ) === false ) {
3746 global $wgMetaNamespace;
3747 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
3749 # Allow grammar transformations
3750 # Allowing full message-style parsing would make simple requests
3751 # such as action=raw much more expensive than they need to be.
3752 # This will hopefully cover most cases.
3753 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
3754 array( &$this, 'replaceGrammarInNamespace' ), $talk );
3755 return str_replace( ' ', '_', $talk );
3762 function replaceGrammarInNamespace( $m ) {
3763 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
3767 * @throws MWException
3770 static function getCaseMaps() {
3771 static $wikiUpperChars, $wikiLowerChars;
3772 if ( isset( $wikiUpperChars ) ) {
3773 return array( $wikiUpperChars, $wikiLowerChars );
3776 wfProfileIn( __METHOD__
);
3777 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
3778 if ( $arr === false ) {
3779 throw new MWException(
3780 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
3782 $wikiUpperChars = $arr['wikiUpperChars'];
3783 $wikiLowerChars = $arr['wikiLowerChars'];
3784 wfProfileOut( __METHOD__
);
3785 return array( $wikiUpperChars, $wikiLowerChars );
3789 * Decode an expiry (block, protection, etc) which has come from the DB
3791 * @FIXME: why are we returnings DBMS-dependent strings???
3793 * @param $expiry String: Database expiry String
3794 * @param $format Bool|Int true to process using language functions, or TS_ constant
3795 * to return the expiry in a given timestamp
3798 public function formatExpiry( $expiry, $format = true ) {
3799 static $infinity, $infinityMsg;
3800 if ( $infinity === null ) {
3801 $infinityMsg = wfMessage( 'infiniteblock' );
3802 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
3805 if ( $expiry == '' ||
$expiry == $infinity ) {
3806 return $format === true
3810 return $format === true
3811 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
3812 : wfTimestamp( $format, $expiry );
3818 * @param $seconds int|float
3819 * @param $format Array Optional
3820 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
3821 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
3822 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
3823 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
3826 function formatTimePeriod( $seconds, $format = array() ) {
3827 if ( !is_array( $format ) ) {
3828 $format = array( 'avoid' => $format ); // For backwards compatibility
3830 if ( !isset( $format['avoid'] ) ) {
3831 $format['avoid'] = false;
3833 if ( !isset( $format['noabbrevs' ] ) ) {
3834 $format['noabbrevs'] = false;
3836 $secondsMsg = wfMessage(
3837 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
3838 $minutesMsg = wfMessage(
3839 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
3840 $hoursMsg = wfMessage(
3841 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
3842 $daysMsg = wfMessage(
3843 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
3845 if ( round( $seconds * 10 ) < 100 ) {
3846 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
3847 $s = $secondsMsg->params( $s )->text();
3848 } elseif ( round( $seconds ) < 60 ) {
3849 $s = $this->formatNum( round( $seconds ) );
3850 $s = $secondsMsg->params( $s )->text();
3851 } elseif ( round( $seconds ) < 3600 ) {
3852 $minutes = floor( $seconds / 60 );
3853 $secondsPart = round( fmod( $seconds, 60 ) );
3854 if ( $secondsPart == 60 ) {
3858 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3860 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3861 } elseif ( round( $seconds ) <= 2 * 86400 ) {
3862 $hours = floor( $seconds / 3600 );
3863 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
3864 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
3865 if ( $secondsPart == 60 ) {
3869 if ( $minutes == 60 ) {
3873 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
3875 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3876 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
3877 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3880 $days = floor( $seconds / 86400 );
3881 if ( $format['avoid'] === 'avoidminutes' ) {
3882 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
3883 if ( $hours == 24 ) {
3887 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3889 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
3890 } elseif ( $format['avoid'] === 'avoidseconds' ) {
3891 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
3892 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
3893 if ( $minutes == 60 ) {
3897 if ( $hours == 24 ) {
3901 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3903 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
3905 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3907 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3909 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
3916 * Format a bitrate for output, using an appropriate
3917 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
3919 * This use base 1000. For base 1024 use formatSize(), for another base
3920 * see formatComputingNumbers()
3925 function formatBitrate( $bps ) {
3926 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
3930 * @param $size int Size of the unit
3931 * @param $boundary int Size boundary (1000, or 1024 in most cases)
3932 * @param $messageKey string Message key to be uesd
3935 function formatComputingNumbers( $size, $boundary, $messageKey ) {
3937 return str_replace( '$1', $this->formatNum( $size ),
3938 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
3941 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
3944 $maxIndex = count( $sizes ) - 1;
3945 while ( $size >= $boundary && $index < $maxIndex ) {
3950 // For small sizes no decimal places necessary
3953 // For MB and bigger two decimal places are smarter
3956 $msg = str_replace( '$1', $sizes[$index], $messageKey );
3958 $size = round( $size, $round );
3959 $text = $this->getMessageFromDB( $msg );
3960 return str_replace( '$1', $this->formatNum( $size ), $text );
3964 * Format a size in bytes for output, using an appropriate
3965 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
3967 * This method use base 1024. For base 1000 use formatBitrate(), for
3968 * another base see formatComputingNumbers()
3970 * @param $size int Size to format
3971 * @return string Plain text (not HTML)
3973 function formatSize( $size ) {
3974 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
3978 * Make a list item, used by various special pages
3980 * @param $page String Page link
3981 * @param $details String Text between brackets
3982 * @param $oppositedm Boolean Add the direction mark opposite to your
3983 * language, to display text properly
3986 function specialList( $page, $details, $oppositedm = true ) {
3987 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) .
3988 $this->getDirMark();
3989 $details = $details ?
$dirmark . $this->getMessageFromDB( 'word-separator' ) .
3990 wfMsgExt( 'parentheses', array( 'escape', 'replaceafter', 'language' => $this ), $details ) : '';
3991 return $page . $details;
3995 * Generate (prev x| next x) (20|50|100...) type links for paging
3997 * @param $title Title object to link
3998 * @param $offset Integer offset parameter
3999 * @param $limit Integer limit parameter
4000 * @param $query String optional URL query parameter string
4001 * @param $atend Bool optional param for specified if this is the last page
4004 public function viewPrevNext( Title
$title, $offset, $limit, array $query = array(), $atend = false ) {
4005 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4007 # Make 'previous' link
4008 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4009 if ( $offset > 0 ) {
4010 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4011 $query, $prev, 'prevn-title', 'mw-prevlink' );
4013 $plink = htmlspecialchars( $prev );
4017 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4019 $nlink = htmlspecialchars( $next );
4021 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4022 $query, $next, 'prevn-title', 'mw-nextlink' );
4025 # Make links to set number of items per page
4026 $numLinks = array();
4027 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4028 $numLinks[] = $this->numLink( $title, $offset, $num,
4029 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4032 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4033 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4037 * Helper function for viewPrevNext() that generates links
4039 * @param $title Title object to link
4040 * @param $offset Integer offset parameter
4041 * @param $limit Integer limit parameter
4042 * @param $query Array extra query parameters
4043 * @param $link String text to use for the link; will be escaped
4044 * @param $tooltipMsg String name of the message to use as tooltip
4045 * @param $class String value of the "class" attribute of the link
4046 * @return String HTML fragment
4048 private function numLink( Title
$title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4049 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4050 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4051 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4052 'title' => $tooltip, 'class' => $class ), $link );
4056 * Get the conversion rule title, if any.
4060 public function getConvRuleTitle() {
4061 return $this->mConverter
->getConvRuleTitle();