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'
146 static public $durationIntervals = array(
147 'millennia' => 31557600000,
148 'centuries' => 3155760000,
149 'decades' => 315576000,
150 'years' => 31557600, // 86400 * 365.25
159 * Get a cached language object for a given language code
160 * @param $code String
163 static function factory( $code ) {
164 if ( !isset( self
::$mLangObjCache[$code] ) ) {
165 if ( count( self
::$mLangObjCache ) > 10 ) {
166 // Don't keep a billion objects around, that's stupid.
167 self
::$mLangObjCache = array();
169 self
::$mLangObjCache[$code] = self
::newFromCode( $code );
171 return self
::$mLangObjCache[$code];
175 * Create a language object for a given language code
176 * @param $code String
177 * @throws MWException
180 protected static function newFromCode( $code ) {
181 // Protect against path traversal below
182 if ( !Language
::isValidCode( $code )
183 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
185 throw new MWException( "Invalid language code \"$code\"" );
188 if ( !Language
::isValidBuiltInCode( $code ) ) {
189 // It's not possible to customise this code with class files, so
190 // just return a Language object. This is to support uselang= hacks.
191 $lang = new Language
;
192 $lang->setCode( $code );
196 // Check if there is a language class for the code
197 $class = self
::classFromCode( $code );
198 self
::preloadLanguageClass( $class );
199 if ( MWInit
::classExists( $class ) ) {
204 // Keep trying the fallback list until we find an existing class
205 $fallbacks = Language
::getFallbacksFor( $code );
206 foreach ( $fallbacks as $fallbackCode ) {
207 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
208 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
211 $class = self
::classFromCode( $fallbackCode );
212 self
::preloadLanguageClass( $class );
213 if ( MWInit
::classExists( $class ) ) {
214 $lang = Language
::newFromCode( $fallbackCode );
215 $lang->setCode( $code );
220 throw new MWException( "Invalid fallback sequence for language '$code'" );
224 * Returns true if a language code string is of a valid form, whether or
225 * not it exists. This includes codes which are used solely for
226 * customisation via the MediaWiki namespace.
228 * @param $code string
232 public static function isValidCode( $code ) {
234 strcspn( $code, ":/\\\000" ) === strlen( $code )
235 && !preg_match( Title
::getTitleInvalidRegex(), $code );
239 * Returns true if a language code is of a valid form for the purposes of
240 * internal customisation of MediaWiki, via Messages*.php.
242 * @param $code string
247 public static function isValidBuiltInCode( $code ) {
249 if( !is_string($code) ) {
250 $type = gettype( $code );
251 if( $type === 'object' ) {
252 $addmsg = " of class " . get_class( $code );
256 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
259 return preg_match( '/^[a-z0-9-]+$/i', $code );
264 * @return String Name of the language class
266 public static function classFromCode( $code ) {
267 if ( $code == 'en' ) {
270 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
275 * Includes language class files
277 * @param $class string Name of the language class
279 public static function preloadLanguageClass( $class ) {
282 if ( $class === 'Language' ) {
286 if ( !defined( 'MW_COMPILED' ) ) {
287 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
288 include_once( "$IP/languages/classes/$class.php" );
294 * Get the LocalisationCache instance
296 * @return LocalisationCache
298 public static function getLocalisationCache() {
299 if ( is_null( self
::$dataCache ) ) {
300 global $wgLocalisationCacheConf;
301 $class = $wgLocalisationCacheConf['class'];
302 self
::$dataCache = new $class( $wgLocalisationCacheConf );
304 return self
::$dataCache;
307 function __construct() {
308 $this->mConverter
= new FakeConverter( $this );
309 // Set the code to the name of the descendant
310 if ( get_class( $this ) == 'Language' ) {
313 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
315 self
::getLocalisationCache();
319 * Reduce memory usage
321 function __destruct() {
322 foreach ( $this as $name => $value ) {
323 unset( $this->$name );
328 * Hook which will be called if this is the content language.
329 * Descendants can use this to register hook functions or modify globals
331 function initContLang() { }
334 * Same as getFallbacksFor for current language.
336 * @deprecated in 1.19
338 function getFallbackLanguageCode() {
339 wfDeprecated( __METHOD__
);
340 return self
::getFallbackFor( $this->mCode
);
347 function getFallbackLanguages() {
348 return self
::getFallbacksFor( $this->mCode
);
352 * Exports $wgBookstoreListEn
355 function getBookstoreList() {
356 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
362 public function getNamespaces() {
363 if ( is_null( $this->namespaceNames
) ) {
364 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
366 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
367 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
369 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
371 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
372 if ( $wgMetaNamespaceTalk ) {
373 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
375 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
376 $this->namespaceNames
[NS_PROJECT_TALK
] =
377 $this->fixVariableInNamespace( $talk );
380 # Sometimes a language will be localised but not actually exist on this wiki.
381 foreach ( $this->namespaceNames
as $key => $text ) {
382 if ( !isset( $validNamespaces[$key] ) ) {
383 unset( $this->namespaceNames
[$key] );
387 # The above mixing may leave namespaces out of canonical order.
388 # Re-order by namespace ID number...
389 ksort( $this->namespaceNames
);
391 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
393 return $this->namespaceNames
;
397 * Arbitrarily set all of the namespace names at once. Mainly used for testing
398 * @param $namespaces Array of namespaces (id => name)
400 public function setNamespaces( array $namespaces ) {
401 $this->namespaceNames
= $namespaces;
405 * A convenience function that returns the same thing as
406 * getNamespaces() except with the array values changed to ' '
407 * where it found '_', useful for producing output to be displayed
408 * e.g. in <select> forms.
412 function getFormattedNamespaces() {
413 $ns = $this->getNamespaces();
414 foreach ( $ns as $k => $v ) {
415 $ns[$k] = strtr( $v, '_', ' ' );
421 * Get a namespace value by key
423 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
424 * echo $mw_ns; // prints 'MediaWiki'
427 * @param $index Int: the array key of the namespace to return
428 * @return mixed, string if the namespace value exists, otherwise false
430 function getNsText( $index ) {
431 $ns = $this->getNamespaces();
432 return isset( $ns[$index] ) ?
$ns[$index] : false;
436 * A convenience function that returns the same thing as
437 * getNsText() except with '_' changed to ' ', useful for
440 * @param $index string
444 function getFormattedNsText( $index ) {
445 $ns = $this->getNsText( $index );
446 return strtr( $ns, '_', ' ' );
450 * Returns gender-dependent namespace alias if available.
451 * @param $index Int: namespace index
452 * @param $gender String: gender key (male, female... )
456 function getGenderNsText( $index, $gender ) {
457 global $wgExtraGenderNamespaces;
459 $ns = $wgExtraGenderNamespaces + self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
460 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
464 * Whether this language makes distinguishes genders for example in
469 function needsGenderDistinction() {
470 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
471 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
472 // $wgExtraGenderNamespaces overrides everything
474 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
475 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
476 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
479 // Check what is in i18n files
480 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
481 return count( $aliases ) > 0;
486 * Get a namespace key by value, case insensitive.
487 * Only matches namespace names for the current language, not the
488 * canonical ones defined in Namespace.php.
490 * @param $text String
491 * @return mixed An integer if $text is a valid value otherwise false
493 function getLocalNsIndex( $text ) {
494 $lctext = $this->lc( $text );
495 $ids = $this->getNamespaceIds();
496 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
502 function getNamespaceAliases() {
503 if ( is_null( $this->namespaceAliases
) ) {
504 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
508 foreach ( $aliases as $name => $index ) {
509 if ( $index === NS_PROJECT_TALK
) {
510 unset( $aliases[$name] );
511 $name = $this->fixVariableInNamespace( $name );
512 $aliases[$name] = $index;
517 global $wgExtraGenderNamespaces;
518 $genders = $wgExtraGenderNamespaces +
(array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
519 foreach ( $genders as $index => $forms ) {
520 foreach ( $forms as $alias ) {
521 $aliases[$alias] = $index;
525 $this->namespaceAliases
= $aliases;
527 return $this->namespaceAliases
;
533 function getNamespaceIds() {
534 if ( is_null( $this->mNamespaceIds
) ) {
535 global $wgNamespaceAliases;
536 # Put namespace names and aliases into a hashtable.
537 # If this is too slow, then we should arrange it so that it is done
538 # before caching. The catch is that at pre-cache time, the above
539 # class-specific fixup hasn't been done.
540 $this->mNamespaceIds
= array();
541 foreach ( $this->getNamespaces() as $index => $name ) {
542 $this->mNamespaceIds
[$this->lc( $name )] = $index;
544 foreach ( $this->getNamespaceAliases() as $name => $index ) {
545 $this->mNamespaceIds
[$this->lc( $name )] = $index;
547 if ( $wgNamespaceAliases ) {
548 foreach ( $wgNamespaceAliases as $name => $index ) {
549 $this->mNamespaceIds
[$this->lc( $name )] = $index;
553 return $this->mNamespaceIds
;
557 * Get a namespace key by value, case insensitive. Canonical namespace
558 * names override custom ones defined for the current language.
560 * @param $text String
561 * @return mixed An integer if $text is a valid value otherwise false
563 function getNsIndex( $text ) {
564 $lctext = $this->lc( $text );
565 $ns = MWNamespace
::getCanonicalIndex( $lctext );
566 if ( $ns !== null ) {
569 $ids = $this->getNamespaceIds();
570 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
574 * short names for language variants used for language conversion links.
576 * @param $code String
577 * @param $usemsg bool Use the "variantname-xyz" message if it exists
580 function getVariantname( $code, $usemsg = true ) {
581 $msg = "variantname-$code";
582 if ( $usemsg && wfMessage( $msg )->exists() ) {
583 return $this->getMessageFromDB( $msg );
585 $name = self
::fetchLanguageName( $code );
587 return $name; # if it's defined as a language name, show that
589 # otherwise, output the language code
595 * @param $name string
598 function specialPage( $name ) {
599 $aliases = $this->getSpecialPageAliases();
600 if ( isset( $aliases[$name][0] ) ) {
601 $name = $aliases[$name][0];
603 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
609 function getQuickbarSettings() {
611 $this->getMessage( 'qbsettings-none' ),
612 $this->getMessage( 'qbsettings-fixedleft' ),
613 $this->getMessage( 'qbsettings-fixedright' ),
614 $this->getMessage( 'qbsettings-floatingleft' ),
615 $this->getMessage( 'qbsettings-floatingright' ),
616 $this->getMessage( 'qbsettings-directionality' )
623 function getDatePreferences() {
624 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
630 function getDateFormats() {
631 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
635 * @return array|string
637 function getDefaultDateFormat() {
638 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
639 if ( $df === 'dmy or mdy' ) {
640 global $wgAmericanDates;
641 return $wgAmericanDates ?
'mdy' : 'dmy';
650 function getDatePreferenceMigrationMap() {
651 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
658 function getImageFile( $image ) {
659 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
665 function getExtraUserToggles() {
666 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
673 function getUserToggle( $tog ) {
674 return $this->getMessageFromDB( "tog-$tog" );
678 * Get native language names, indexed by code.
679 * Only those defined in MediaWiki, no other data like CLDR.
680 * If $customisedOnly is true, only returns codes with a messages file
682 * @param $customisedOnly bool
685 * @deprecated in 1.20, use fetchLanguageNames()
687 public static function getLanguageNames( $customisedOnly = false ) {
688 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
692 * Get translated language names. This is done on best effort and
693 * by default this is exactly the same as Language::getLanguageNames.
694 * The CLDR extension provides translated names.
695 * @param $code String Language code.
696 * @return Array language code => language name
698 * @deprecated in 1.20, use fetchLanguageNames()
700 public static function getTranslatedLanguageNames( $code ) {
701 return self
::fetchLanguageNames( $code, 'all' );
705 * Get an array of language names, indexed by code.
706 * @param $inLanguage null|string: Code of language in which to return the names
707 * Use null for autonyms (native names)
708 * @param $include string:
709 * 'all' all available languages
710 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
711 * 'mwfile' only if the language is in 'mw' *and* has a message file
712 * @return array: language code => language name
715 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
716 global $wgExtraLanguageNames;
717 static $coreLanguageNames;
719 if ( $coreLanguageNames === null ) {
720 include( MWInit
::compiledPath( 'languages/Names.php' ) );
726 # TODO: also include when $inLanguage is null, when this code is more efficient
727 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
730 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
731 foreach ( $mwNames as $mwCode => $mwName ) {
732 # - Prefer own MediaWiki native name when not using the hook
733 # TODO: prefer it always to make it consistent, but casing is different in CLDR
734 # - For other names just add if not added through the hook
735 if ( ( $mwCode === $inLanguage && !$inLanguage ) ||
!isset( $names[$mwCode] ) ) {
736 $names[$mwCode] = $mwName;
740 if ( $include === 'all' ) {
745 $coreCodes = array_keys( $mwNames );
746 foreach( $coreCodes as $coreCode ) {
747 $returnMw[$coreCode] = $names[$coreCode];
750 if( $include === 'mwfile' ) {
751 $namesMwFile = array();
752 # We do this using a foreach over the codes instead of a directory
753 # loop so that messages files in extensions will work correctly.
754 foreach ( $returnMw as $code => $value ) {
755 if ( is_readable( self
::getMessagesFileName( $code ) ) ) {
756 $namesMwFile[$code] = $names[$code];
761 # 'mw' option; default if it's not one of the other two options (all/mwfile)
766 * @param $code string: The code of the language for which to get the name
767 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
768 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
769 * @return string: Language name or empty
772 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
773 $array = self
::fetchLanguageNames( $inLanguage, $include );
774 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
778 * Get a message from the MediaWiki namespace.
780 * @param $msg String: message name
783 function getMessageFromDB( $msg ) {
784 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
788 * Get the native language name of $code.
789 * Only if defined in MediaWiki, no other data like CLDR.
790 * @param $code string
792 * @deprecated in 1.20, use fetchLanguageName()
794 function getLanguageName( $code ) {
795 return self
::fetchLanguageName( $code );
802 function getMonthName( $key ) {
803 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
809 function getMonthNamesArray() {
810 $monthNames = array( '' );
811 for ( $i = 1; $i < 13; $i++
) {
812 $monthNames[] = $this->getMonthName( $i );
821 function getMonthNameGen( $key ) {
822 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
829 function getMonthAbbreviation( $key ) {
830 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
836 function getMonthAbbreviationsArray() {
837 $monthNames = array( '' );
838 for ( $i = 1; $i < 13; $i++
) {
839 $monthNames[] = $this->getMonthAbbreviation( $i );
848 function getWeekdayName( $key ) {
849 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
856 function getWeekdayAbbreviation( $key ) {
857 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
864 function getIranianCalendarMonthName( $key ) {
865 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
872 function getHebrewCalendarMonthName( $key ) {
873 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
880 function getHebrewCalendarMonthNameGen( $key ) {
881 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
888 function getHijriCalendarMonthName( $key ) {
889 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
893 * This is a workalike of PHP's date() function, but with better
894 * internationalisation, a reduced set of format characters, and a better
897 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
898 * PHP manual for definitions. There are a number of extensions, which
901 * xn Do not translate digits of the next numeric format character
902 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
903 * xr Use roman numerals for the next numeric format character
904 * xh Use hebrew numerals for the next numeric format character
906 * xg Genitive month name
908 * xij j (day number) in Iranian calendar
909 * xiF F (month name) in Iranian calendar
910 * xin n (month number) in Iranian calendar
911 * xiy y (two digit year) in Iranian calendar
912 * xiY Y (full year) in Iranian calendar
914 * xjj j (day number) in Hebrew calendar
915 * xjF F (month name) in Hebrew calendar
916 * xjt t (days in month) in Hebrew calendar
917 * xjx xg (genitive month name) in Hebrew calendar
918 * xjn n (month number) in Hebrew calendar
919 * xjY Y (full year) in Hebrew calendar
921 * xmj j (day number) in Hijri calendar
922 * xmF F (month name) in Hijri calendar
923 * xmn n (month number) in Hijri calendar
924 * xmY Y (full year) in Hijri calendar
926 * xkY Y (full year) in Thai solar calendar. Months and days are
927 * identical to the Gregorian calendar
928 * xoY Y (full year) in Minguo calendar or Juche year.
929 * Months and days are identical to the
931 * xtY Y (full year) in Japanese nengo. Months and days are
932 * identical to the Gregorian calendar
934 * Characters enclosed in double quotes will be considered literal (with
935 * the quotes themselves removed). Unmatched quotes will be considered
936 * literal quotes. Example:
938 * "The month is" F => The month is January
941 * Backslash escaping is also supported.
943 * Input timestamp is assumed to be pre-normalized to the desired local
946 * @param $format String
947 * @param $ts String: 14-character timestamp
950 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
954 function sprintfDate( $format, $ts ) {
967 for ( $p = 0; $p < strlen( $format ); $p++
) {
970 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
971 $code .= $format[++
$p];
974 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
975 $code .= $format[++
$p];
986 $rawToggle = !$rawToggle;
995 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
998 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
999 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1002 $num = substr( $ts, 6, 2 );
1005 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
1006 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) +
1 );
1009 $num = intval( substr( $ts, 6, 2 ) );
1013 $iranian = self
::tsToIranian( $ts );
1019 $hijri = self
::tsToHijri( $ts );
1025 $hebrew = self
::tsToHebrew( $ts );
1031 $unix = wfTimestamp( TS_UNIX
, $ts );
1033 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) +
1 );
1037 $unix = wfTimestamp( TS_UNIX
, $ts );
1039 $w = gmdate( 'w', $unix );
1044 $unix = wfTimestamp( TS_UNIX
, $ts );
1046 $num = gmdate( 'w', $unix );
1050 $unix = wfTimestamp( TS_UNIX
, $ts );
1052 $num = gmdate( 'z', $unix );
1056 $unix = wfTimestamp( TS_UNIX
, $ts );
1058 $num = gmdate( 'W', $unix );
1061 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1065 $iranian = self
::tsToIranian( $ts );
1067 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1071 $hijri = self
::tsToHijri( $ts );
1073 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1077 $hebrew = self
::tsToHebrew( $ts );
1079 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1082 $num = substr( $ts, 4, 2 );
1085 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1088 $num = intval( substr( $ts, 4, 2 ) );
1092 $iranian = self
::tsToIranian( $ts );
1098 $hijri = self
::tsToHijri ( $ts );
1104 $hebrew = self
::tsToHebrew( $ts );
1110 $unix = wfTimestamp( TS_UNIX
, $ts );
1112 $num = gmdate( 't', $unix );
1116 $hebrew = self
::tsToHebrew( $ts );
1122 $unix = wfTimestamp( TS_UNIX
, $ts );
1124 $num = gmdate( 'L', $unix );
1128 $unix = wfTimestamp( TS_UNIX
, $ts );
1130 $num = gmdate( 'o', $unix );
1133 $num = substr( $ts, 0, 4 );
1137 $iranian = self
::tsToIranian( $ts );
1143 $hijri = self
::tsToHijri( $ts );
1149 $hebrew = self
::tsToHebrew( $ts );
1155 $thai = self
::tsToYear( $ts, 'thai' );
1161 $minguo = self
::tsToYear( $ts, 'minguo' );
1167 $tenno = self
::tsToYear( $ts, 'tenno' );
1172 $num = substr( $ts, 2, 2 );
1176 $iranian = self
::tsToIranian( $ts );
1178 $num = substr( $iranian[0], -2 );
1181 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1184 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1187 $h = substr( $ts, 8, 2 );
1188 $num = $h %
12 ?
$h %
12 : 12;
1191 $num = intval( substr( $ts, 8, 2 ) );
1194 $h = substr( $ts, 8, 2 );
1195 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1198 $num = substr( $ts, 8, 2 );
1201 $num = substr( $ts, 10, 2 );
1204 $num = substr( $ts, 12, 2 );
1208 $unix = wfTimestamp( TS_UNIX
, $ts );
1210 $s .= gmdate( 'c', $unix );
1214 $unix = wfTimestamp( TS_UNIX
, $ts );
1216 $s .= gmdate( 'r', $unix );
1220 $unix = wfTimestamp( TS_UNIX
, $ts );
1225 # Backslash escaping
1226 if ( $p < strlen( $format ) - 1 ) {
1227 $s .= $format[++
$p];
1234 if ( $p < strlen( $format ) - 1 ) {
1235 $endQuote = strpos( $format, '"', $p +
1 );
1236 if ( $endQuote === false ) {
1237 # No terminating quote, assume literal "
1240 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1244 # Quote at end of string, assume literal "
1251 if ( $num !== false ) {
1252 if ( $rawToggle ||
$raw ) {
1255 } elseif ( $roman ) {
1256 $s .= self
::romanNumeral( $num );
1258 } elseif ( $hebrewNum ) {
1259 $s .= self
::hebrewNumeral( $num );
1262 $s .= $this->formatNum( $num, true );
1269 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1270 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1273 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1274 * Gregorian dates to Iranian dates. Originally written in C, it
1275 * is released under the terms of GNU Lesser General Public
1276 * License. Conversion to PHP was performed by Niklas Laxström.
1278 * Link: http://www.farsiweb.info/jalali/jalali.c
1284 private static function tsToIranian( $ts ) {
1285 $gy = substr( $ts, 0, 4 ) -1600;
1286 $gm = substr( $ts, 4, 2 ) -1;
1287 $gd = substr( $ts, 6, 2 ) -1;
1289 # Days passed from the beginning (including leap years)
1291 +
floor( ( $gy +
3 ) / 4 )
1292 - floor( ( $gy +
99 ) / 100 )
1293 +
floor( ( $gy +
399 ) / 400 );
1295 // Add days of the past months of this year
1296 for ( $i = 0; $i < $gm; $i++
) {
1297 $gDayNo +
= self
::$GREG_DAYS[$i];
1301 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1305 // Days passed in current month
1306 $gDayNo +
= (int)$gd;
1308 $jDayNo = $gDayNo - 79;
1310 $jNp = floor( $jDayNo / 12053 );
1313 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1316 if ( $jDayNo >= 366 ) {
1317 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1318 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1321 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1322 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1328 return array( $jy, $jm, $jd );
1332 * Converting Gregorian dates to Hijri dates.
1334 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1336 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1342 private static function tsToHijri( $ts ) {
1343 $year = substr( $ts, 0, 4 );
1344 $month = substr( $ts, 4, 2 );
1345 $day = substr( $ts, 6, 2 );
1353 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1354 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1357 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1358 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1359 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1362 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1363 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1366 $zl = $zjd -1948440 +
10632;
1367 $zn = (int)( ( $zl - 1 ) / 10631 );
1368 $zl = $zl - 10631 * $zn +
354;
1369 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1370 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1371 $zm = (int)( ( 24 * $zl ) / 709 );
1372 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1373 $zy = 30 * $zn +
$zj - 30;
1375 return array( $zy, $zm, $zd );
1379 * Converting Gregorian dates to Hebrew dates.
1381 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1382 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1383 * to translate the relevant functions into PHP and release them under
1386 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1387 * and Adar II is 14. In a non-leap year, Adar is 6.
1393 private static function tsToHebrew( $ts ) {
1395 $year = substr( $ts, 0, 4 );
1396 $month = substr( $ts, 4, 2 );
1397 $day = substr( $ts, 6, 2 );
1399 # Calculate Hebrew year
1400 $hebrewYear = $year +
3760;
1402 # Month number when September = 1, August = 12
1404 if ( $month > 12 ) {
1411 # Calculate day of year from 1 September
1413 for ( $i = 1; $i < $month; $i++
) {
1417 # Check if the year is leap
1418 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1421 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1428 # Calculate the start of the Hebrew year
1429 $start = self
::hebrewYearStart( $hebrewYear );
1431 # Calculate next year's start
1432 if ( $dayOfYear <= $start ) {
1433 # Day is before the start of the year - it is the previous year
1435 $nextStart = $start;
1439 # Add days since previous year's 1 September
1441 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1445 # Start of the new (previous) year
1446 $start = self
::hebrewYearStart( $hebrewYear );
1449 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1452 # Calculate Hebrew day of year
1453 $hebrewDayOfYear = $dayOfYear - $start;
1455 # Difference between year's days
1456 $diff = $nextStart - $start;
1457 # Add 12 (or 13 for leap years) days to ignore the difference between
1458 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1459 # difference is only about the year type
1460 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1466 # Check the year pattern, and is leap year
1467 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1468 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1469 # and non-leap years
1470 $yearPattern = $diff %
30;
1471 # Check if leap year
1472 $isLeap = $diff >= 30;
1474 # Calculate day in the month from number of day in the Hebrew year
1475 # Don't check Adar - if the day is not in Adar, we will stop before;
1476 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1477 $hebrewDay = $hebrewDayOfYear;
1480 while ( $hebrewMonth <= 12 ) {
1481 # Calculate days in this month
1482 if ( $isLeap && $hebrewMonth == 6 ) {
1483 # Adar in a leap year
1485 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1487 if ( $hebrewDay <= $days ) {
1491 # Subtract the days of Adar I
1492 $hebrewDay -= $days;
1495 if ( $hebrewDay <= $days ) {
1501 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1502 # Cheshvan in a complete year (otherwise as the rule below)
1504 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1505 # Kislev in an incomplete year (otherwise as the rule below)
1508 # Odd months have 30 days, even have 29
1509 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1511 if ( $hebrewDay <= $days ) {
1512 # In the current month
1515 # Subtract the days of the current month
1516 $hebrewDay -= $days;
1517 # Try in the next month
1522 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1526 * This calculates the Hebrew year start, as days since 1 September.
1527 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1528 * Used for Hebrew date.
1534 private static function hebrewYearStart( $year ) {
1535 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1536 $b = intval( ( $year - 1 ) %
4 );
1537 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1541 $Mar = intval( $m );
1547 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1548 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1550 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1552 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1556 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1561 * Algorithm to convert Gregorian dates to Thai solar dates,
1562 * Minguo dates or Minguo dates.
1564 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1565 * http://en.wikipedia.org/wiki/Minguo_calendar
1566 * http://en.wikipedia.org/wiki/Japanese_era_name
1568 * @param $ts String: 14-character timestamp
1569 * @param $cName String: calender name
1570 * @return Array: converted year, month, day
1572 private static function tsToYear( $ts, $cName ) {
1573 $gy = substr( $ts, 0, 4 );
1574 $gm = substr( $ts, 4, 2 );
1575 $gd = substr( $ts, 6, 2 );
1577 if ( !strcmp( $cName, 'thai' ) ) {
1579 # Add 543 years to the Gregorian calendar
1580 # Months and days are identical
1581 $gy_offset = $gy +
543;
1582 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1584 # Deduct 1911 years from the Gregorian calendar
1585 # Months and days are identical
1586 $gy_offset = $gy - 1911;
1587 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1588 # Nengō dates up to Meiji period
1589 # Deduct years from the Gregorian calendar
1590 # depending on the nengo periods
1591 # Months and days are identical
1592 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1594 $gy_gannen = $gy - 1868 +
1;
1595 $gy_offset = $gy_gannen;
1596 if ( $gy_gannen == 1 ) {
1599 $gy_offset = '明治' . $gy_offset;
1601 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1602 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1603 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1604 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1605 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1609 $gy_gannen = $gy - 1912 +
1;
1610 $gy_offset = $gy_gannen;
1611 if ( $gy_gannen == 1 ) {
1614 $gy_offset = '大正' . $gy_offset;
1616 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1617 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1618 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1622 $gy_gannen = $gy - 1926 +
1;
1623 $gy_offset = $gy_gannen;
1624 if ( $gy_gannen == 1 ) {
1627 $gy_offset = '昭和' . $gy_offset;
1630 $gy_gannen = $gy - 1989 +
1;
1631 $gy_offset = $gy_gannen;
1632 if ( $gy_gannen == 1 ) {
1635 $gy_offset = '平成' . $gy_offset;
1641 return array( $gy_offset, $gm, $gd );
1645 * Roman number formatting up to 3000
1651 static function romanNumeral( $num ) {
1652 static $table = array(
1653 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1654 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1655 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1656 array( '', 'M', 'MM', 'MMM' )
1659 $num = intval( $num );
1660 if ( $num > 3000 ||
$num <= 0 ) {
1665 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1666 if ( $num >= $pow10 ) {
1667 $s .= $table[$i][(int)floor( $num / $pow10 )];
1669 $num = $num %
$pow10;
1675 * Hebrew Gematria number formatting up to 9999
1681 static function hebrewNumeral( $num ) {
1682 static $table = array(
1683 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1684 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1685 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1686 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1689 $num = intval( $num );
1690 if ( $num > 9999 ||
$num <= 0 ) {
1695 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1696 if ( $num >= $pow10 ) {
1697 if ( $num == 15 ||
$num == 16 ) {
1698 $s .= $table[0][9] . $table[0][$num - 9];
1701 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1702 if ( $pow10 == 1000 ) {
1707 $num = $num %
$pow10;
1709 if ( strlen( $s ) == 2 ) {
1712 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1713 $str .= substr( $s, strlen( $s ) - 2, 2 );
1715 $start = substr( $str, 0, strlen( $str ) - 2 );
1716 $end = substr( $str, strlen( $str ) - 2 );
1719 $str = $start . 'ך';
1722 $str = $start . 'ם';
1725 $str = $start . 'ן';
1728 $str = $start . 'ף';
1731 $str = $start . 'ץ';
1738 * Used by date() and time() to adjust the time output.
1740 * @param $ts Int the time in date('YmdHis') format
1741 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1742 * get user timecorrection setting)
1745 function userAdjust( $ts, $tz = false ) {
1746 global $wgUser, $wgLocalTZoffset;
1748 if ( $tz === false ) {
1749 $tz = $wgUser->getOption( 'timecorrection' );
1752 $data = explode( '|', $tz, 3 );
1754 if ( $data[0] == 'ZoneInfo' ) {
1755 wfSuppressWarnings();
1756 $userTZ = timezone_open( $data[2] );
1757 wfRestoreWarnings();
1758 if ( $userTZ !== false ) {
1759 $date = date_create( $ts, timezone_open( 'UTC' ) );
1760 date_timezone_set( $date, $userTZ );
1761 $date = date_format( $date, 'YmdHis' );
1764 # Unrecognized timezone, default to 'Offset' with the stored offset.
1765 $data[0] = 'Offset';
1769 if ( $data[0] == 'System' ||
$tz == '' ) {
1770 # Global offset in minutes.
1771 if ( isset( $wgLocalTZoffset ) ) {
1772 $minDiff = $wgLocalTZoffset;
1774 } elseif ( $data[0] == 'Offset' ) {
1775 $minDiff = intval( $data[1] );
1777 $data = explode( ':', $tz );
1778 if ( count( $data ) == 2 ) {
1779 $data[0] = intval( $data[0] );
1780 $data[1] = intval( $data[1] );
1781 $minDiff = abs( $data[0] ) * 60 +
$data[1];
1782 if ( $data[0] < 0 ) {
1783 $minDiff = -$minDiff;
1786 $minDiff = intval( $data[0] ) * 60;
1790 # No difference ? Return time unchanged
1791 if ( 0 == $minDiff ) {
1795 wfSuppressWarnings(); // E_STRICT system time bitching
1796 # Generate an adjusted date; take advantage of the fact that mktime
1797 # will normalize out-of-range values so we don't have to split $minDiff
1798 # into hours and minutes.
1800 (int)substr( $ts, 8, 2 ) ), # Hours
1801 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
1802 (int)substr( $ts, 12, 2 ), # Seconds
1803 (int)substr( $ts, 4, 2 ), # Month
1804 (int)substr( $ts, 6, 2 ), # Day
1805 (int)substr( $ts, 0, 4 ) ); # Year
1807 $date = date( 'YmdHis', $t );
1808 wfRestoreWarnings();
1814 * This is meant to be used by time(), date(), and timeanddate() to get
1815 * the date preference they're supposed to use, it should be used in
1819 * function timeanddate([...], $format = true) {
1820 * $datePreference = $this->dateFormat($format);
1825 * @param $usePrefs Mixed: if true, the user's preference is used
1826 * if false, the site/language default is used
1827 * if int/string, assumed to be a format.
1830 function dateFormat( $usePrefs = true ) {
1833 if ( is_bool( $usePrefs ) ) {
1835 $datePreference = $wgUser->getDatePreference();
1837 $datePreference = (string)User
::getDefaultOption( 'date' );
1840 $datePreference = (string)$usePrefs;
1844 if ( $datePreference == '' ) {
1848 return $datePreference;
1852 * Get a format string for a given type and preference
1853 * @param $type string May be date, time or both
1854 * @param $pref string The format name as it appears in Messages*.php
1858 function getDateFormatString( $type, $pref ) {
1859 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
1860 if ( $pref == 'default' ) {
1861 $pref = $this->getDefaultDateFormat();
1862 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1864 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1865 if ( is_null( $df ) ) {
1866 $pref = $this->getDefaultDateFormat();
1867 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1870 $this->dateFormatStrings
[$type][$pref] = $df;
1872 return $this->dateFormatStrings
[$type][$pref];
1876 * @param $ts Mixed: the time format which needs to be turned into a
1877 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1878 * @param $adj Bool: whether to adjust the time output according to the
1879 * user configured offset ($timecorrection)
1880 * @param $format Mixed: true to use user's date format preference
1881 * @param $timecorrection String|bool the time offset as returned by
1882 * validateTimeZone() in Special:Preferences
1885 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1886 $ts = wfTimestamp( TS_MW
, $ts );
1888 $ts = $this->userAdjust( $ts, $timecorrection );
1890 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1891 return $this->sprintfDate( $df, $ts );
1895 * @param $ts Mixed: the time format which needs to be turned into a
1896 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1897 * @param $adj Bool: whether to adjust the time output according to the
1898 * user configured offset ($timecorrection)
1899 * @param $format Mixed: true to use user's date format preference
1900 * @param $timecorrection String|bool the time offset as returned by
1901 * validateTimeZone() in Special:Preferences
1904 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1905 $ts = wfTimestamp( TS_MW
, $ts );
1907 $ts = $this->userAdjust( $ts, $timecorrection );
1909 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1910 return $this->sprintfDate( $df, $ts );
1914 * @param $ts Mixed: the time format which needs to be turned into a
1915 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1916 * @param $adj Bool: whether to adjust the time output according to the
1917 * user configured offset ($timecorrection)
1918 * @param $format Mixed: what format to return, if it's false output the
1919 * default one (default true)
1920 * @param $timecorrection String|bool the time offset as returned by
1921 * validateTimeZone() in Special:Preferences
1924 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1925 $ts = wfTimestamp( TS_MW
, $ts );
1927 $ts = $this->userAdjust( $ts, $timecorrection );
1929 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1930 return $this->sprintfDate( $df, $ts );
1934 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
1938 * @param integer $seconds The amount of seconds.
1939 * @param array $chosenIntervals The intervals to enable.
1943 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
1944 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
1946 $segments = array();
1948 foreach ( $intervals as $intervalName => $intervalValue ) {
1949 $message = new Message( 'duration-' . $intervalName, array( $intervalValue ) );
1950 $segments[] = $message->inLanguage( $this )->escaped();
1953 return $this->listToText( $segments );
1957 * Takes a number of seconds and returns an array with a set of corresponding intervals.
1958 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
1962 * @param integer $seconds The amount of seconds.
1963 * @param array $chosenIntervals The intervals to enable.
1967 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
1968 if ( empty( $chosenIntervals ) ) {
1969 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
1972 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
1973 $sortedNames = array_keys( $intervals );
1974 $smallestInterval = array_pop( $sortedNames );
1976 $segments = array();
1978 foreach ( $intervals as $name => $length ) {
1979 $value = floor( $seconds / $length );
1981 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
1982 $seconds -= $value * $length;
1983 $segments[$name] = $value;
1991 * Internal helper function for userDate(), userTime() and userTimeAndDate()
1993 * @param $type String: can be 'date', 'time' or 'both'
1994 * @param $ts Mixed: the time format which needs to be turned into a
1995 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1996 * @param $user User object used to get preferences for timezone and format
1997 * @param $options Array, can contain the following keys:
1998 * - 'timecorrection': time correction, can have the following values:
1999 * - true: use user's preference
2000 * - false: don't use time correction
2001 * - integer: value of time correction in minutes
2002 * - 'format': format to use, can have the following values:
2003 * - true: use user's preference
2004 * - false: use default preference
2005 * - string: format to use
2009 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2010 $ts = wfTimestamp( TS_MW
, $ts );
2011 $options +
= array( 'timecorrection' => true, 'format' => true );
2012 if ( $options['timecorrection'] !== false ) {
2013 if ( $options['timecorrection'] === true ) {
2014 $offset = $user->getOption( 'timecorrection' );
2016 $offset = $options['timecorrection'];
2018 $ts = $this->userAdjust( $ts, $offset );
2020 if ( $options['format'] === true ) {
2021 $format = $user->getDatePreference();
2023 $format = $options['format'];
2025 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2026 return $this->sprintfDate( $df, $ts );
2030 * Get the formatted date for the given timestamp and formatted for
2033 * @param $ts Mixed: the time format which needs to be turned into a
2034 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2035 * @param $user User object used to get preferences for timezone and format
2036 * @param $options Array, can contain the following keys:
2037 * - 'timecorrection': time correction, can have the following values:
2038 * - true: use user's preference
2039 * - false: don't use time correction
2040 * - integer: value of time correction in minutes
2041 * - 'format': format to use, can have the following values:
2042 * - true: use user's preference
2043 * - false: use default preference
2044 * - string: format to use
2048 public function userDate( $ts, User
$user, array $options = array() ) {
2049 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2053 * Get the formatted time for the given timestamp and formatted for
2056 * @param $ts Mixed: the time format which needs to be turned into a
2057 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2058 * @param $user User object used to get preferences for timezone and format
2059 * @param $options Array, can contain the following keys:
2060 * - 'timecorrection': time correction, can have the following values:
2061 * - true: use user's preference
2062 * - false: don't use time correction
2063 * - integer: value of time correction in minutes
2064 * - 'format': format to use, can have the following values:
2065 * - true: use user's preference
2066 * - false: use default preference
2067 * - string: format to use
2071 public function userTime( $ts, User
$user, array $options = array() ) {
2072 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2076 * Get the formatted date and time for the given timestamp and formatted for
2079 * @param $ts Mixed: the time format which needs to be turned into a
2080 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2081 * @param $user User object used to get preferences for timezone and format
2082 * @param $options Array, can contain the following keys:
2083 * - 'timecorrection': time correction, can have the following values:
2084 * - true: use user's preference
2085 * - false: don't use time correction
2086 * - integer: value of time correction in minutes
2087 * - 'format': format to use, can have the following values:
2088 * - true: use user's preference
2089 * - false: use default preference
2090 * - string: format to use
2094 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2095 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2099 * @param $key string
2100 * @return array|null
2102 function getMessage( $key ) {
2103 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2109 function getAllMessages() {
2110 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2119 function iconv( $in, $out, $string ) {
2120 # This is a wrapper for iconv in all languages except esperanto,
2121 # which does some nasty x-conversions beforehand
2123 # Even with //IGNORE iconv can whine about illegal characters in
2124 # *input* string. We just ignore those too.
2125 # REF: http://bugs.php.net/bug.php?id=37166
2126 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2127 wfSuppressWarnings();
2128 $text = iconv( $in, $out . '//IGNORE', $string );
2129 wfRestoreWarnings();
2133 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2136 * @param $matches array
2137 * @return mixed|string
2139 function ucwordbreaksCallbackAscii( $matches ) {
2140 return $this->ucfirst( $matches[1] );
2144 * @param $matches array
2147 function ucwordbreaksCallbackMB( $matches ) {
2148 return mb_strtoupper( $matches[0] );
2152 * @param $matches array
2155 function ucCallback( $matches ) {
2156 list( $wikiUpperChars ) = self
::getCaseMaps();
2157 return strtr( $matches[1], $wikiUpperChars );
2161 * @param $matches array
2164 function lcCallback( $matches ) {
2165 list( , $wikiLowerChars ) = self
::getCaseMaps();
2166 return strtr( $matches[1], $wikiLowerChars );
2170 * @param $matches array
2173 function ucwordsCallbackMB( $matches ) {
2174 return mb_strtoupper( $matches[0] );
2178 * @param $matches array
2181 function ucwordsCallbackWiki( $matches ) {
2182 list( $wikiUpperChars ) = self
::getCaseMaps();
2183 return strtr( $matches[0], $wikiUpperChars );
2187 * Make a string's first character uppercase
2189 * @param $str string
2193 function ucfirst( $str ) {
2195 if ( $o < 96 ) { // if already uppercase...
2197 } elseif ( $o < 128 ) {
2198 return ucfirst( $str ); // use PHP's ucfirst()
2200 // fall back to more complex logic in case of multibyte strings
2201 return $this->uc( $str, true );
2206 * Convert a string to uppercase
2208 * @param $str string
2209 * @param $first bool
2213 function uc( $str, $first = false ) {
2214 if ( function_exists( 'mb_strtoupper' ) ) {
2216 if ( $this->isMultibyte( $str ) ) {
2217 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2219 return ucfirst( $str );
2222 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2225 if ( $this->isMultibyte( $str ) ) {
2226 $x = $first ?
'^' : '';
2227 return preg_replace_callback(
2228 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2229 array( $this, 'ucCallback' ),
2233 return $first ?
ucfirst( $str ) : strtoupper( $str );
2239 * @param $str string
2240 * @return mixed|string
2242 function lcfirst( $str ) {
2245 return strval( $str );
2246 } elseif ( $o >= 128 ) {
2247 return $this->lc( $str, true );
2248 } elseif ( $o > 96 ) {
2251 $str[0] = strtolower( $str[0] );
2257 * @param $str string
2258 * @param $first bool
2259 * @return mixed|string
2261 function lc( $str, $first = false ) {
2262 if ( function_exists( 'mb_strtolower' ) ) {
2264 if ( $this->isMultibyte( $str ) ) {
2265 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2267 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2270 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2273 if ( $this->isMultibyte( $str ) ) {
2274 $x = $first ?
'^' : '';
2275 return preg_replace_callback(
2276 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2277 array( $this, 'lcCallback' ),
2281 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2287 * @param $str string
2290 function isMultibyte( $str ) {
2291 return (bool)preg_match( '/[\x80-\xff]/', $str );
2295 * @param $str string
2296 * @return mixed|string
2298 function ucwords( $str ) {
2299 if ( $this->isMultibyte( $str ) ) {
2300 $str = $this->lc( $str );
2302 // regexp to find first letter in each word (i.e. after each space)
2303 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2305 // function to use to capitalize a single char
2306 if ( function_exists( 'mb_strtoupper' ) ) {
2307 return preg_replace_callback(
2309 array( $this, 'ucwordsCallbackMB' ),
2313 return preg_replace_callback(
2315 array( $this, 'ucwordsCallbackWiki' ),
2320 return ucwords( strtolower( $str ) );
2325 * capitalize words at word breaks
2327 * @param $str string
2330 function ucwordbreaks( $str ) {
2331 if ( $this->isMultibyte( $str ) ) {
2332 $str = $this->lc( $str );
2334 // since \b doesn't work for UTF-8, we explicitely define word break chars
2335 $breaks = "[ \-\(\)\}\{\.,\?!]";
2337 // find first letter after word break
2338 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2340 if ( function_exists( 'mb_strtoupper' ) ) {
2341 return preg_replace_callback(
2343 array( $this, 'ucwordbreaksCallbackMB' ),
2347 return preg_replace_callback(
2349 array( $this, 'ucwordsCallbackWiki' ),
2354 return preg_replace_callback(
2355 '/\b([\w\x80-\xff]+)\b/',
2356 array( $this, 'ucwordbreaksCallbackAscii' ),
2363 * Return a case-folded representation of $s
2365 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2366 * and $s2 are the same except for the case of their characters. It is not
2367 * necessary for the value returned to make sense when displayed.
2369 * Do *not* perform any other normalisation in this function. If a caller
2370 * uses this function when it should be using a more general normalisation
2371 * function, then fix the caller.
2377 function caseFold( $s ) {
2378 return $this->uc( $s );
2385 function checkTitleEncoding( $s ) {
2386 if ( is_array( $s ) ) {
2387 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2389 # Check for non-UTF-8 URLs
2390 $ishigh = preg_match( '/[\x80-\xff]/', $s );
2395 if ( function_exists( 'mb_check_encoding' ) ) {
2396 $isutf8 = mb_check_encoding( $s, 'UTF-8' );
2398 $isutf8 = preg_match( '/^(?>[\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2399 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
2405 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2411 function fallback8bitEncoding() {
2412 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2416 * Most writing systems use whitespace to break up words.
2417 * Some languages such as Chinese don't conventionally do this,
2418 * which requires special handling when breaking up words for
2423 function hasWordBreaks() {
2428 * Some languages such as Chinese require word segmentation,
2429 * Specify such segmentation when overridden in derived class.
2431 * @param $string String
2434 function segmentByWord( $string ) {
2439 * Some languages have special punctuation need to be normalized.
2440 * Make such changes here.
2442 * @param $string String
2445 function normalizeForSearch( $string ) {
2446 return self
::convertDoubleWidth( $string );
2450 * convert double-width roman characters to single-width.
2451 * range: ff00-ff5f ~= 0020-007f
2453 * @param $string string
2457 protected static function convertDoubleWidth( $string ) {
2458 static $full = null;
2459 static $half = null;
2461 if ( $full === null ) {
2462 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2463 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2464 $full = str_split( $fullWidth, 3 );
2465 $half = str_split( $halfWidth );
2468 $string = str_replace( $full, $half, $string );
2473 * @param $string string
2474 * @param $pattern string
2477 protected static function insertSpace( $string, $pattern ) {
2478 $string = preg_replace( $pattern, " $1 ", $string );
2479 $string = preg_replace( '/ +/', ' ', $string );
2484 * @param $termsArray array
2487 function convertForSearchResult( $termsArray ) {
2488 # some languages, e.g. Chinese, need to do a conversion
2489 # in order for search results to be displayed correctly
2494 * Get the first character of a string.
2499 function firstChar( $s ) {
2502 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2503 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2508 if ( isset( $matches[1] ) ) {
2509 if ( strlen( $matches[1] ) != 3 ) {
2513 // Break down Hangul syllables to grab the first jamo
2514 $code = utf8ToCodepoint( $matches[1] );
2515 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2517 } elseif ( $code < 0xb098 ) {
2518 return "\xe3\x84\xb1";
2519 } elseif ( $code < 0xb2e4 ) {
2520 return "\xe3\x84\xb4";
2521 } elseif ( $code < 0xb77c ) {
2522 return "\xe3\x84\xb7";
2523 } elseif ( $code < 0xb9c8 ) {
2524 return "\xe3\x84\xb9";
2525 } elseif ( $code < 0xbc14 ) {
2526 return "\xe3\x85\x81";
2527 } elseif ( $code < 0xc0ac ) {
2528 return "\xe3\x85\x82";
2529 } elseif ( $code < 0xc544 ) {
2530 return "\xe3\x85\x85";
2531 } elseif ( $code < 0xc790 ) {
2532 return "\xe3\x85\x87";
2533 } elseif ( $code < 0xcc28 ) {
2534 return "\xe3\x85\x88";
2535 } elseif ( $code < 0xce74 ) {
2536 return "\xe3\x85\x8a";
2537 } elseif ( $code < 0xd0c0 ) {
2538 return "\xe3\x85\x8b";
2539 } elseif ( $code < 0xd30c ) {
2540 return "\xe3\x85\x8c";
2541 } elseif ( $code < 0xd558 ) {
2542 return "\xe3\x85\x8d";
2544 return "\xe3\x85\x8e";
2551 function initEncoding() {
2552 # Some languages may have an alternate char encoding option
2553 # (Esperanto X-coding, Japanese furigana conversion, etc)
2554 # If this language is used as the primary content language,
2555 # an override to the defaults can be set here on startup.
2562 function recodeForEdit( $s ) {
2563 # For some languages we'll want to explicitly specify
2564 # which characters make it into the edit box raw
2565 # or are converted in some way or another.
2566 global $wgEditEncoding;
2567 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2570 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2578 function recodeInput( $s ) {
2579 # Take the previous into account.
2580 global $wgEditEncoding;
2581 if ( $wgEditEncoding != '' ) {
2582 $enc = $wgEditEncoding;
2586 if ( $enc == 'UTF-8' ) {
2589 return $this->iconv( $enc, 'UTF-8', $s );
2594 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2595 * also cleans up certain backwards-compatible sequences, converting them
2596 * to the modern Unicode equivalent.
2598 * This is language-specific for performance reasons only.
2604 function normalize( $s ) {
2605 global $wgAllUnicodeFixes;
2606 $s = UtfNormal
::cleanUp( $s );
2607 if ( $wgAllUnicodeFixes ) {
2608 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2609 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2616 * Transform a string using serialized data stored in the given file (which
2617 * must be in the serialized subdirectory of $IP). The file contains pairs
2618 * mapping source characters to destination characters.
2620 * The data is cached in process memory. This will go faster if you have the
2621 * FastStringSearch extension.
2623 * @param $file string
2624 * @param $string string
2626 * @throws MWException
2629 function transformUsingPairFile( $file, $string ) {
2630 if ( !isset( $this->transformData
[$file] ) ) {
2631 $data = wfGetPrecompiledData( $file );
2632 if ( $data === false ) {
2633 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
2635 $this->transformData
[$file] = new ReplacementArray( $data );
2637 return $this->transformData
[$file]->replace( $string );
2641 * For right-to-left language support
2646 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
2650 * Return the correct HTML 'dir' attribute value for this language.
2654 return $this->isRTL() ?
'rtl' : 'ltr';
2658 * Return 'left' or 'right' as appropriate alignment for line-start
2659 * for this language's text direction.
2661 * Should be equivalent to CSS3 'start' text-align value....
2665 function alignStart() {
2666 return $this->isRTL() ?
'right' : 'left';
2670 * Return 'right' or 'left' as appropriate alignment for line-end
2671 * for this language's text direction.
2673 * Should be equivalent to CSS3 'end' text-align value....
2677 function alignEnd() {
2678 return $this->isRTL() ?
'left' : 'right';
2682 * A hidden direction mark (LRM or RLM), depending on the language direction.
2683 * Unlike getDirMark(), this function returns the character as an HTML entity.
2684 * This function should be used when the output is guaranteed to be HTML,
2685 * because it makes the output HTML source code more readable. When
2686 * the output is plain text or can be escaped, getDirMark() should be used.
2688 * @param $opposite Boolean Get the direction mark opposite to your language
2691 function getDirMarkEntity( $opposite = false ) {
2692 if ( $opposite ) { return $this->isRTL() ?
'‎' : '‏'; }
2693 return $this->isRTL() ?
'‏' : '‎';
2697 * A hidden direction mark (LRM or RLM), depending on the language direction.
2698 * This function produces them as invisible Unicode characters and
2699 * the output may be hard to read and debug, so it should only be used
2700 * when the output is plain text or can be escaped. When the output is
2701 * HTML, use getDirMarkEntity() instead.
2703 * @param $opposite Boolean Get the direction mark opposite to your language
2706 function getDirMark( $opposite = false ) {
2707 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2708 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2709 if ( $opposite ) { return $this->isRTL() ?
$lrm : $rlm; }
2710 return $this->isRTL() ?
$rlm : $lrm;
2716 function capitalizeAllNouns() {
2717 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2721 * An arrow, depending on the language direction.
2723 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2726 function getArrow( $direction = 'forwards' ) {
2727 switch ( $direction ) {
2729 return $this->isRTL() ?
'←' : '→';
2731 return $this->isRTL() ?
'→' : '←';
2744 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2748 function linkPrefixExtension() {
2749 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2755 function getMagicWords() {
2756 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2759 protected function doMagicHook() {
2760 if ( $this->mMagicHookDone
) {
2763 $this->mMagicHookDone
= true;
2764 wfProfileIn( 'LanguageGetMagic' );
2765 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2766 wfProfileOut( 'LanguageGetMagic' );
2770 * Fill a MagicWord object with data from here
2774 function getMagic( $mw ) {
2775 $this->doMagicHook();
2777 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
2778 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
2780 $magicWords = $this->getMagicWords();
2781 if ( isset( $magicWords[$mw->mId
] ) ) {
2782 $rawEntry = $magicWords[$mw->mId
];
2788 if ( !is_array( $rawEntry ) ) {
2789 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
2791 $mw->mCaseSensitive
= $rawEntry[0];
2792 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
2797 * Add magic words to the extension array
2799 * @param $newWords array
2801 function addMagicWordsByLang( $newWords ) {
2802 $fallbackChain = $this->getFallbackLanguages();
2803 $fallbackChain = array_reverse( $fallbackChain );
2804 foreach ( $fallbackChain as $code ) {
2805 if ( isset( $newWords[$code] ) ) {
2806 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
2812 * Get special page names, as an associative array
2813 * case folded alias => real name
2815 function getSpecialPageAliases() {
2816 // Cache aliases because it may be slow to load them
2817 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
2819 $this->mExtendedSpecialPageAliases
=
2820 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
2821 wfRunHooks( 'LanguageGetSpecialPageAliases',
2822 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
2825 return $this->mExtendedSpecialPageAliases
;
2829 * Italic is unsuitable for some languages
2831 * @param $text String: the text to be emphasized.
2834 function emphasize( $text ) {
2835 return "<em>$text</em>";
2839 * Normally we output all numbers in plain en_US style, that is
2840 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2841 * point twohundredthirtyfive. However this is not suitable for all
2842 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2843 * Icelandic just want to use commas instead of dots, and dots instead
2844 * of commas like "293.291,235".
2846 * An example of this function being called:
2848 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2851 * See LanguageGu.php for the Gujarati implementation and
2852 * $separatorTransformTable on MessageIs.php for
2853 * the , => . and . => , implementation.
2855 * @todo check if it's viable to use localeconv() for the decimal
2857 * @param $number Mixed: the string to be formatted, should be an integer
2858 * or a floating point number.
2859 * @param $nocommafy Bool: set to true for special numbers like dates
2862 public function formatNum( $number, $nocommafy = false ) {
2863 global $wgTranslateNumerals;
2864 if ( !$nocommafy ) {
2865 $number = $this->commafy( $number );
2866 $s = $this->separatorTransformTable();
2868 $number = strtr( $number, $s );
2872 if ( $wgTranslateNumerals ) {
2873 $s = $this->digitTransformTable();
2875 $number = strtr( $number, $s );
2883 * @param $number string
2886 function parseFormattedNumber( $number ) {
2887 $s = $this->digitTransformTable();
2889 $number = strtr( $number, array_flip( $s ) );
2892 $s = $this->separatorTransformTable();
2894 $number = strtr( $number, array_flip( $s ) );
2897 $number = strtr( $number, array( ',' => '' ) );
2902 * Adds commas to a given number
2907 function commafy( $_ ) {
2908 $digitGroupingPattern = $this->digitGroupingPattern();
2909 if ( $_ === null ) {
2913 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
2914 // default grouping is at thousands, use the same for ###,###,### pattern too.
2915 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2917 // Ref: http://cldr.unicode.org/translation/number-patterns
2919 if ( intval( $_ ) < 0 ) {
2920 // For negative numbers apply the algorithm like positive number and add sign.
2922 $_ = substr( $_, 1 );
2924 $numberpart = array();
2925 $decimalpart = array();
2926 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
2927 preg_match( "/\d+/", $_, $numberpart );
2928 preg_match( "/\.\d*/", $_, $decimalpart );
2929 $groupedNumber = ( count( $decimalpart ) > 0 ) ?
$decimalpart[0]:"";
2930 if ( $groupedNumber === $_ ) {
2931 // the string does not have any number part. Eg: .12345
2932 return $sign . $groupedNumber;
2934 $start = $end = strlen( $numberpart[0] );
2935 while ( $start > 0 ) {
2936 $match = $matches[0][$numMatches -1] ;
2937 $matchLen = strlen( $match );
2938 $start = $end - $matchLen;
2942 $groupedNumber = substr( $_ , $start, $end -$start ) . $groupedNumber ;
2944 if ( $numMatches > 1 ) {
2945 // use the last pattern for the rest of the number
2949 $groupedNumber = "," . $groupedNumber;
2952 return $sign . $groupedNumber;
2958 function digitGroupingPattern() {
2959 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
2965 function digitTransformTable() {
2966 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
2972 function separatorTransformTable() {
2973 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
2977 * Take a list of strings and build a locale-friendly comma-separated
2978 * list, using the local comma-separator message.
2979 * The last two strings are chained with an "and".
2984 function listToText( array $l ) {
2986 $m = count( $l ) - 1;
2988 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2990 for ( $i = $m; $i >= 0; $i-- ) {
2993 } elseif ( $i == $m - 1 ) {
2994 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2996 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
3004 * Take a list of strings and build a locale-friendly comma-separated
3005 * list, using the local comma-separator message.
3006 * @param $list array of strings to put in a comma list
3009 function commaList( array $list ) {
3013 array( 'parsemag', 'escapenoentities', 'language' => $this )
3020 * Take a list of strings and build a locale-friendly semicolon-separated
3021 * list, using the local semicolon-separator message.
3022 * @param $list array of strings to put in a semicolon list
3025 function semicolonList( array $list ) {
3028 'semicolon-separator',
3029 array( 'parsemag', 'escapenoentities', 'language' => $this )
3036 * Same as commaList, but separate it with the pipe instead.
3037 * @param $list array of strings to put in a pipe list
3040 function pipeList( array $list ) {
3044 array( 'escapenoentities', 'language' => $this )
3051 * Truncate a string to a specified length in bytes, appending an optional
3052 * string (e.g. for ellipses)
3054 * The database offers limited byte lengths for some columns in the database;
3055 * multi-byte character sets mean we need to ensure that only whole characters
3056 * are included, otherwise broken characters can be passed to the user
3058 * If $length is negative, the string will be truncated from the beginning
3060 * @param $string String to truncate
3061 * @param $length Int: maximum length (including ellipses)
3062 * @param $ellipsis String to append to the truncated text
3063 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3064 * $adjustLength was introduced in 1.18, before that behaved as if false.
3067 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3068 # Use the localized ellipsis character
3069 if ( $ellipsis == '...' ) {
3070 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
3072 # Check if there is no need to truncate
3073 if ( $length == 0 ) {
3074 return $ellipsis; // convention
3075 } elseif ( strlen( $string ) <= abs( $length ) ) {
3076 return $string; // no need to truncate
3078 $stringOriginal = $string;
3079 # If ellipsis length is >= $length then we can't apply $adjustLength
3080 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3081 $string = $ellipsis; // this can be slightly unexpected
3082 # Otherwise, truncate and add ellipsis...
3084 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3085 if ( $length > 0 ) {
3086 $length -= $eLength;
3087 $string = substr( $string, 0, $length ); // xyz...
3088 $string = $this->removeBadCharLast( $string );
3089 $string = $string . $ellipsis;
3091 $length +
= $eLength;
3092 $string = substr( $string, $length ); // ...xyz
3093 $string = $this->removeBadCharFirst( $string );
3094 $string = $ellipsis . $string;
3097 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3098 # This check is *not* redundant if $adjustLength, due to the single case where
3099 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3100 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3103 return $stringOriginal;
3108 * Remove bytes that represent an incomplete Unicode character
3109 * at the end of string (e.g. bytes of the char are missing)
3111 * @param $string String
3114 protected function removeBadCharLast( $string ) {
3115 if ( $string != '' ) {
3116 $char = ord( $string[strlen( $string ) - 1] );
3118 if ( $char >= 0xc0 ) {
3119 # We got the first byte only of a multibyte char; remove it.
3120 $string = substr( $string, 0, -1 );
3121 } elseif ( $char >= 0x80 &&
3122 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3123 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
3125 # We chopped in the middle of a character; remove it
3133 * Remove bytes that represent an incomplete Unicode character
3134 * at the start of string (e.g. bytes of the char are missing)
3136 * @param $string String
3139 protected function removeBadCharFirst( $string ) {
3140 if ( $string != '' ) {
3141 $char = ord( $string[0] );
3142 if ( $char >= 0x80 && $char < 0xc0 ) {
3143 # We chopped in the middle of a character; remove the whole thing
3144 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3151 * Truncate a string of valid HTML to a specified length in bytes,
3152 * appending an optional string (e.g. for ellipses), and return valid HTML
3154 * This is only intended for styled/linked text, such as HTML with
3155 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3156 * Also, this will not detect things like "display:none" CSS.
3158 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3160 * @param string $text HTML string to truncate
3161 * @param int $length (zero/positive) Maximum length (including ellipses)
3162 * @param string $ellipsis String to append to the truncated text
3165 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3166 # Use the localized ellipsis character
3167 if ( $ellipsis == '...' ) {
3168 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
3170 # Check if there is clearly no need to truncate
3171 if ( $length <= 0 ) {
3172 return $ellipsis; // no text shown, nothing to format (convention)
3173 } elseif ( strlen( $text ) <= $length ) {
3174 return $text; // string short enough even *with* HTML (short-circuit)
3177 $dispLen = 0; // innerHTML legth so far
3178 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3179 $tagType = 0; // 0-open, 1-close
3180 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3181 $entityState = 0; // 0-not entity, 1-entity
3182 $tag = $ret = ''; // accumulated tag name, accumulated result string
3183 $openTags = array(); // open tag stack
3184 $maybeState = null; // possible truncation state
3186 $textLen = strlen( $text );
3187 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3188 for ( $pos = 0; true; ++
$pos ) {
3189 # Consider truncation once the display length has reached the maximim.
3190 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3191 # Check that we're not in the middle of a bracket/entity...
3192 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3193 if ( !$testingEllipsis ) {
3194 $testingEllipsis = true;
3195 # Save where we are; we will truncate here unless there turn out to
3196 # be so few remaining characters that truncation is not necessary.
3197 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3198 $maybeState = array( $ret, $openTags ); // save state
3200 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3201 # String in fact does need truncation, the truncation point was OK.
3202 list( $ret, $openTags ) = $maybeState; // reload state
3203 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3204 $ret .= $ellipsis; // add ellipsis
3208 if ( $pos >= $textLen ) break; // extra iteration just for above checks
3210 # Read the next char...
3212 $lastCh = $pos ?
$text[$pos - 1] : '';
3213 $ret .= $ch; // add to result string
3215 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3216 $entityState = 0; // for bad HTML
3217 $bracketState = 1; // tag started (checking for backslash)
3218 } elseif ( $ch == '>' ) {
3219 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3220 $entityState = 0; // for bad HTML
3221 $bracketState = 0; // out of brackets
3222 } elseif ( $bracketState == 1 ) {
3224 $tagType = 1; // close tag (e.g. "</span>")
3226 $tagType = 0; // open tag (e.g. "<span>")
3229 $bracketState = 2; // building tag name
3230 } elseif ( $bracketState == 2 ) {
3234 // Name found (e.g. "<a href=..."), add on tag attributes...
3235 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3237 } elseif ( $bracketState == 0 ) {
3238 if ( $entityState ) {
3241 $dispLen++
; // entity is one displayed char
3244 if ( $neLength == 0 && !$maybeState ) {
3245 // Save state without $ch. We want to *hit* the first
3246 // display char (to get tags) but not *use* it if truncating.
3247 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3250 $entityState = 1; // entity found, (e.g. " ")
3252 $dispLen++
; // this char is displayed
3253 // Add the next $max display text chars after this in one swoop...
3254 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3255 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3256 $dispLen +
= $skipped;
3262 // Close the last tag if left unclosed by bad HTML
3263 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3264 while ( count( $openTags ) > 0 ) {
3265 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3271 * truncateHtml() helper function
3272 * like strcspn() but adds the skipped chars to $ret
3281 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3282 if ( $len === null ) {
3283 $len = -1; // -1 means "no limit" for strcspn
3284 } elseif ( $len < 0 ) {
3288 if ( $start < strlen( $text ) ) {
3289 $skipCount = strcspn( $text, $search, $start, $len );
3290 $ret .= substr( $text, $start, $skipCount );
3296 * truncateHtml() helper function
3297 * (a) push or pop $tag from $openTags as needed
3298 * (b) clear $tag value
3299 * @param &$tag string Current HTML tag name we are looking at
3300 * @param $tagType int (0-open tag, 1-close tag)
3301 * @param $lastCh string Character before the '>' that ended this tag
3302 * @param &$openTags array Open tag stack (not accounting for $tag)
3304 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3305 $tag = ltrim( $tag );
3307 if ( $tagType == 0 && $lastCh != '/' ) {
3308 $openTags[] = $tag; // tag opened (didn't close itself)
3309 } elseif ( $tagType == 1 ) {
3310 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3311 array_pop( $openTags ); // tag closed
3319 * Grammatical transformations, needed for inflected languages
3320 * Invoked by putting {{grammar:case|word}} in a message
3322 * @param $word string
3323 * @param $case string
3326 function convertGrammar( $word, $case ) {
3327 global $wgGrammarForms;
3328 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3329 return $wgGrammarForms[$this->getCode()][$case][$word];
3334 * Get the grammar forms for the content language
3335 * @return array of grammar forms
3338 function getGrammarForms() {
3339 global $wgGrammarForms;
3340 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3341 return $wgGrammarForms[$this->getCode()];
3346 * Provides an alternative text depending on specified gender.
3347 * Usage {{gender:username|masculine|feminine|neutral}}.
3348 * username is optional, in which case the gender of current user is used,
3349 * but only in (some) interface messages; otherwise default gender is used.
3351 * If no forms are given, an empty string is returned. If only one form is
3352 * given, it will be returned unconditionally. These details are implied by
3353 * the caller and cannot be overridden in subclasses.
3355 * If more than one form is given, the default is to use the neutral one
3356 * if it is specified, and to use the masculine one otherwise. These
3357 * details can be overridden in subclasses.
3359 * @param $gender string
3360 * @param $forms array
3364 function gender( $gender, $forms ) {
3365 if ( !count( $forms ) ) {
3368 $forms = $this->preConvertPlural( $forms, 2 );
3369 if ( $gender === 'male' ) {
3372 if ( $gender === 'female' ) {
3375 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3379 * Plural form transformations, needed for some languages.
3380 * For example, there are 3 form of plural in Russian and Polish,
3381 * depending on "count mod 10". See [[w:Plural]]
3382 * For English it is pretty simple.
3384 * Invoked by putting {{plural:count|wordform1|wordform2}}
3385 * or {{plural:count|wordform1|wordform2|wordform3}}
3387 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3389 * @param $count Integer: non-localized number
3390 * @param $forms Array: different plural forms
3391 * @return string Correct form of plural for $count in this language
3393 function convertPlural( $count, $forms ) {
3394 if ( !count( $forms ) ) {
3397 $forms = $this->preConvertPlural( $forms, 2 );
3399 return ( $count == 1 ) ?
$forms[0] : $forms[1];
3403 * Checks that convertPlural was given an array and pads it to requested
3404 * amount of forms by copying the last one.
3406 * @param $count Integer: How many forms should there be at least
3407 * @param $forms Array of forms given to convertPlural
3408 * @return array Padded array of forms or an exception if not an array
3410 protected function preConvertPlural( /* Array */ $forms, $count ) {
3411 while ( count( $forms ) < $count ) {
3412 $forms[] = $forms[count( $forms ) - 1];
3418 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3419 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3420 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3421 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3424 * @param $str String: the validated block duration in English
3425 * @return string Somehow translated block duration
3426 * @see LanguageFi.php for example implementation
3428 function translateBlockExpiry( $str ) {
3429 $duration = SpecialBlock
::getSuggestedDurations( $this );
3430 foreach ( $duration as $show => $value ) {
3431 if ( strcmp( $str, $value ) == 0 ) {
3432 return htmlspecialchars( trim( $show ) );
3436 // Since usually only infinite or indefinite is only on list, so try
3437 // equivalents if still here.
3438 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3439 if ( in_array( $str, $indefs ) ) {
3440 foreach ( $indefs as $val ) {
3441 $show = array_search( $val, $duration, true );
3442 if ( $show !== false ) {
3443 return htmlspecialchars( trim( $show ) );
3447 // If all else fails, return the original string.
3452 * languages like Chinese need to be segmented in order for the diff
3455 * @param $text String
3458 public function segmentForDiff( $text ) {
3463 * and unsegment to show the result
3465 * @param $text String
3468 public function unsegmentForDiff( $text ) {
3473 * Return the LanguageConverter used in the Language
3476 * @return LanguageConverter
3478 public function getConverter() {
3479 return $this->mConverter
;
3483 * convert text to all supported variants
3485 * @param $text string
3488 public function autoConvertToAllVariants( $text ) {
3489 return $this->mConverter
->autoConvertToAllVariants( $text );
3493 * convert text to different variants of a language.
3495 * @param $text string
3498 public function convert( $text ) {
3499 return $this->mConverter
->convert( $text );
3503 * Convert a Title object to a string in the preferred variant
3505 * @param $title Title
3508 public function convertTitle( $title ) {
3509 return $this->mConverter
->convertTitle( $title );
3513 * Check if this is a language with variants
3517 public function hasVariants() {
3518 return sizeof( $this->getVariants() ) > 1;
3522 * Check if the language has the specific variant
3525 * @param $variant string
3528 public function hasVariant( $variant ) {
3529 return (bool)$this->mConverter
->validateVariant( $variant );
3533 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3535 * @param $text string
3538 public function armourMath( $text ) {
3539 return $this->mConverter
->armourMath( $text );
3543 * Perform output conversion on a string, and encode for safe HTML output.
3544 * @param $text String text to be converted
3545 * @param $isTitle Bool whether this conversion is for the article title
3547 * @todo this should get integrated somewhere sane
3549 public function convertHtml( $text, $isTitle = false ) {
3550 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3554 * @param $key string
3557 public function convertCategoryKey( $key ) {
3558 return $this->mConverter
->convertCategoryKey( $key );
3562 * Get the list of variants supported by this language
3563 * see sample implementation in LanguageZh.php
3565 * @return array an array of language codes
3567 public function getVariants() {
3568 return $this->mConverter
->getVariants();
3574 public function getPreferredVariant() {
3575 return $this->mConverter
->getPreferredVariant();
3581 public function getDefaultVariant() {
3582 return $this->mConverter
->getDefaultVariant();
3588 public function getURLVariant() {
3589 return $this->mConverter
->getURLVariant();
3593 * If a language supports multiple variants, it is
3594 * possible that non-existing link in one variant
3595 * actually exists in another variant. this function
3596 * tries to find it. See e.g. LanguageZh.php
3598 * @param $link String: the name of the link
3599 * @param $nt Mixed: the title object of the link
3600 * @param $ignoreOtherCond Boolean: to disable other conditions when
3601 * we need to transclude a template or update a category's link
3602 * @return null the input parameters may be modified upon return
3604 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3605 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
3609 * If a language supports multiple variants, converts text
3610 * into an array of all possible variants of the text:
3611 * 'variant' => text in that variant
3613 * @deprecated since 1.17 Use autoConvertToAllVariants()
3615 * @param $text string
3619 public function convertLinkToAllVariants( $text ) {
3620 return $this->mConverter
->convertLinkToAllVariants( $text );
3624 * returns language specific options used by User::getPageRenderHash()
3625 * for example, the preferred language variant
3629 function getExtraHashOptions() {
3630 return $this->mConverter
->getExtraHashOptions();
3634 * For languages that support multiple variants, the title of an
3635 * article may be displayed differently in different variants. this
3636 * function returns the apporiate title defined in the body of the article.
3640 public function getParsedTitle() {
3641 return $this->mConverter
->getParsedTitle();
3645 * Enclose a string with the "no conversion" tag. This is used by
3646 * various functions in the Parser
3648 * @param $text String: text to be tagged for no conversion
3649 * @param $noParse bool
3650 * @return string the tagged text
3652 public function markNoConversion( $text, $noParse = false ) {
3653 return $this->mConverter
->markNoConversion( $text, $noParse );
3657 * A regular expression to match legal word-trailing characters
3658 * which should be merged onto a link of the form [[foo]]bar.
3662 public function linkTrail() {
3663 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
3669 function getLangObj() {
3674 * Get the RFC 3066 code for this language object
3678 public function getCode() {
3679 return $this->mCode
;
3683 * Get the code in Bcp47 format which we can use
3684 * inside of html lang="" tags.
3688 public function getHtmlCode() {
3689 if ( is_null( $this->mHtmlCode
) ) {
3690 $this->mHtmlCode
= wfBCP47( $this->getCode() );
3692 return $this->mHtmlCode
;
3696 * @param $code string
3698 public function setCode( $code ) {
3699 $this->mCode
= $code;
3700 // Ensure we don't leave an incorrect html code lying around
3701 $this->mHtmlCode
= null;
3705 * Get the name of a file for a certain language code
3706 * @param $prefix string Prepend this to the filename
3707 * @param $code string Language code
3708 * @param $suffix string Append this to the filename
3709 * @throws MWException
3710 * @return string $prefix . $mangledCode . $suffix
3712 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
3713 // Protect against path traversal
3714 if ( !Language
::isValidCode( $code )
3715 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
3717 throw new MWException( "Invalid language code \"$code\"" );
3720 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
3724 * Get the language code from a file name. Inverse of getFileName()
3725 * @param $filename string $prefix . $languageCode . $suffix
3726 * @param $prefix string Prefix before the language code
3727 * @param $suffix string Suffix after the language code
3728 * @return string Language code, or false if $prefix or $suffix isn't found
3730 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
3732 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
3733 preg_quote( $suffix, '/' ) . '/', $filename, $m );
3734 if ( !count( $m ) ) {
3737 return str_replace( '_', '-', strtolower( $m[1] ) );
3741 * @param $code string
3744 public static function getMessagesFileName( $code ) {
3746 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
3747 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
3752 * @param $code string
3755 public static function getClassFileName( $code ) {
3757 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
3761 * Get the first fallback for a given language.
3763 * @param $code string
3765 * @return bool|string
3767 public static function getFallbackFor( $code ) {
3768 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
3771 $fallbacks = self
::getFallbacksFor( $code );
3772 $first = array_shift( $fallbacks );
3778 * Get the ordered list of fallback languages.
3781 * @param $code string Language code
3784 public static function getFallbacksFor( $code ) {
3785 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
3788 $v = self
::getLocalisationCache()->getItem( $code, 'fallback' );
3789 $v = array_map( 'trim', explode( ',', $v ) );
3790 if ( $v[count( $v ) - 1] !== 'en' ) {
3798 * Get all messages for a given language
3799 * WARNING: this may take a long time. If you just need all message *keys*
3800 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
3802 * @param $code string
3806 public static function getMessagesFor( $code ) {
3807 return self
::getLocalisationCache()->getItem( $code, 'messages' );
3811 * Get a message for a given language
3813 * @param $key string
3814 * @param $code string
3818 public static function getMessageFor( $key, $code ) {
3819 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
3823 * Get all message keys for a given language. This is a faster alternative to
3824 * array_keys( Language::getMessagesFor( $code ) )
3827 * @param $code string Language code
3828 * @return array of message keys (strings)
3830 public static function getMessageKeysFor( $code ) {
3831 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
3838 function fixVariableInNamespace( $talk ) {
3839 if ( strpos( $talk, '$1' ) === false ) {
3843 global $wgMetaNamespace;
3844 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
3846 # Allow grammar transformations
3847 # Allowing full message-style parsing would make simple requests
3848 # such as action=raw much more expensive than they need to be.
3849 # This will hopefully cover most cases.
3850 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
3851 array( &$this, 'replaceGrammarInNamespace' ), $talk );
3852 return str_replace( ' ', '_', $talk );
3859 function replaceGrammarInNamespace( $m ) {
3860 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
3864 * @throws MWException
3867 static function getCaseMaps() {
3868 static $wikiUpperChars, $wikiLowerChars;
3869 if ( isset( $wikiUpperChars ) ) {
3870 return array( $wikiUpperChars, $wikiLowerChars );
3873 wfProfileIn( __METHOD__
);
3874 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
3875 if ( $arr === false ) {
3876 throw new MWException(
3877 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
3879 $wikiUpperChars = $arr['wikiUpperChars'];
3880 $wikiLowerChars = $arr['wikiLowerChars'];
3881 wfProfileOut( __METHOD__
);
3882 return array( $wikiUpperChars, $wikiLowerChars );
3886 * Decode an expiry (block, protection, etc) which has come from the DB
3888 * @FIXME: why are we returnings DBMS-dependent strings???
3890 * @param $expiry String: Database expiry String
3891 * @param $format Bool|Int true to process using language functions, or TS_ constant
3892 * to return the expiry in a given timestamp
3895 public function formatExpiry( $expiry, $format = true ) {
3896 static $infinity, $infinityMsg;
3897 if ( $infinity === null ) {
3898 $infinityMsg = wfMessage( 'infiniteblock' );
3899 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
3902 if ( $expiry == '' ||
$expiry == $infinity ) {
3903 return $format === true
3907 return $format === true
3908 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
3909 : wfTimestamp( $format, $expiry );
3915 * @param $seconds int|float
3916 * @param $format Array Optional
3917 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
3918 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
3919 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
3920 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
3923 function formatTimePeriod( $seconds, $format = array() ) {
3924 if ( !is_array( $format ) ) {
3925 $format = array( 'avoid' => $format ); // For backwards compatibility
3927 if ( !isset( $format['avoid'] ) ) {
3928 $format['avoid'] = false;
3930 if ( !isset( $format['noabbrevs' ] ) ) {
3931 $format['noabbrevs'] = false;
3933 $secondsMsg = wfMessage(
3934 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
3935 $minutesMsg = wfMessage(
3936 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
3937 $hoursMsg = wfMessage(
3938 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
3939 $daysMsg = wfMessage(
3940 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
3942 if ( round( $seconds * 10 ) < 100 ) {
3943 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
3944 $s = $secondsMsg->params( $s )->text();
3945 } elseif ( round( $seconds ) < 60 ) {
3946 $s = $this->formatNum( round( $seconds ) );
3947 $s = $secondsMsg->params( $s )->text();
3948 } elseif ( round( $seconds ) < 3600 ) {
3949 $minutes = floor( $seconds / 60 );
3950 $secondsPart = round( fmod( $seconds, 60 ) );
3951 if ( $secondsPart == 60 ) {
3955 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3957 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3958 } elseif ( round( $seconds ) <= 2 * 86400 ) {
3959 $hours = floor( $seconds / 3600 );
3960 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
3961 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
3962 if ( $secondsPart == 60 ) {
3966 if ( $minutes == 60 ) {
3970 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
3972 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3973 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
3974 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3977 $days = floor( $seconds / 86400 );
3978 if ( $format['avoid'] === 'avoidminutes' ) {
3979 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
3980 if ( $hours == 24 ) {
3984 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3986 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
3987 } elseif ( $format['avoid'] === 'avoidseconds' ) {
3988 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
3989 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
3990 if ( $minutes == 60 ) {
3994 if ( $hours == 24 ) {
3998 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4000 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4002 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4004 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4006 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4013 * Format a bitrate for output, using an appropriate
4014 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4016 * This use base 1000. For base 1024 use formatSize(), for another base
4017 * see formatComputingNumbers()
4022 function formatBitrate( $bps ) {
4023 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4027 * @param $size int Size of the unit
4028 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4029 * @param $messageKey string Message key to be uesd
4032 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4034 return str_replace( '$1', $this->formatNum( $size ),
4035 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4038 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4041 $maxIndex = count( $sizes ) - 1;
4042 while ( $size >= $boundary && $index < $maxIndex ) {
4047 // For small sizes no decimal places necessary
4050 // For MB and bigger two decimal places are smarter
4053 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4055 $size = round( $size, $round );
4056 $text = $this->getMessageFromDB( $msg );
4057 return str_replace( '$1', $this->formatNum( $size ), $text );
4061 * Format a size in bytes for output, using an appropriate
4062 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4064 * This method use base 1024. For base 1000 use formatBitrate(), for
4065 * another base see formatComputingNumbers()
4067 * @param $size int Size to format
4068 * @return string Plain text (not HTML)
4070 function formatSize( $size ) {
4071 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4075 * Make a list item, used by various special pages
4077 * @param $page String Page link
4078 * @param $details String Text between brackets
4079 * @param $oppositedm Boolean Add the direction mark opposite to your
4080 * language, to display text properly
4083 function specialList( $page, $details, $oppositedm = true ) {
4084 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) .
4085 $this->getDirMark();
4086 $details = $details ?
$dirmark . $this->getMessageFromDB( 'word-separator' ) .
4087 wfMsgExt( 'parentheses', array( 'escape', 'replaceafter', 'language' => $this ), $details ) : '';
4088 return $page . $details;
4092 * Generate (prev x| next x) (20|50|100...) type links for paging
4094 * @param $title Title object to link
4095 * @param $offset Integer offset parameter
4096 * @param $limit Integer limit parameter
4097 * @param $query String optional URL query parameter string
4098 * @param $atend Bool optional param for specified if this is the last page
4101 public function viewPrevNext( Title
$title, $offset, $limit, array $query = array(), $atend = false ) {
4102 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4104 # Make 'previous' link
4105 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4106 if ( $offset > 0 ) {
4107 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4108 $query, $prev, 'prevn-title', 'mw-prevlink' );
4110 $plink = htmlspecialchars( $prev );
4114 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4116 $nlink = htmlspecialchars( $next );
4118 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4119 $query, $next, 'prevn-title', 'mw-nextlink' );
4122 # Make links to set number of items per page
4123 $numLinks = array();
4124 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4125 $numLinks[] = $this->numLink( $title, $offset, $num,
4126 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4129 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4130 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4134 * Helper function for viewPrevNext() that generates links
4136 * @param $title Title object to link
4137 * @param $offset Integer offset parameter
4138 * @param $limit Integer limit parameter
4139 * @param $query Array extra query parameters
4140 * @param $link String text to use for the link; will be escaped
4141 * @param $tooltipMsg String name of the message to use as tooltip
4142 * @param $class String value of the "class" attribute of the link
4143 * @return String HTML fragment
4145 private function numLink( Title
$title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4146 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4147 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4148 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4149 'title' => $tooltip, 'class' => $class ), $link );
4153 * Get the conversion rule title, if any.
4157 public function getConvRuleTitle() {
4158 return $this->mConverter
->getConvRuleTitle();