3 * Internationalisation code.
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
25 * @defgroup Language Language
28 if ( !defined( 'MEDIAWIKI' ) ) {
29 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
34 global $wgLanguageNames;
35 require_once( dirname( __FILE__
) . '/Names.php' );
37 if ( function_exists( 'mb_strtoupper' ) ) {
38 mb_internal_encoding( 'UTF-8' );
42 * a fake language converter
52 function __construct( $langobj ) { $this->mLang
= $langobj; }
53 function autoConvertToAllVariants( $text ) { return array( $this->mLang
->getCode() => $text ); }
54 function convert( $t ) { return $t; }
55 function convertTo( $text, $variant ) { return $text; }
56 function convertTitle( $t ) { return $t->getPrefixedText(); }
57 function getVariants() { return array( $this->mLang
->getCode() ); }
58 function getPreferredVariant() { return $this->mLang
->getCode(); }
59 function getDefaultVariant() { return $this->mLang
->getCode(); }
60 function getURLVariant() { return ''; }
61 function getConvRuleTitle() { return false; }
62 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
63 function getExtraHashOptions() { return ''; }
64 function getParsedTitle() { return ''; }
65 function markNoConversion( $text, $noParse = false ) { return $text; }
66 function convertCategoryKey( $key ) { return $key; }
67 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
68 function armourMath( $text ) { return $text; }
72 * Internationalisation code
78 * @var LanguageConverter
82 var $mVariants, $mCode, $mLoaded = false;
83 var $mMagicExtensions = array(), $mMagicHookDone = false;
84 private $mHtmlCode = null;
86 var $dateFormatStrings = array();
87 var $mExtendedSpecialPageAliases;
89 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
92 * ReplacementArray object caches
94 var $transformData = array();
97 * @var LocalisationCache
99 static public $dataCache;
101 static public $mLangObjCache = array();
103 static public $mWeekdayMsgs = array(
104 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
108 static public $mWeekdayAbbrevMsgs = array(
109 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
112 static public $mMonthMsgs = array(
113 'january', 'february', 'march', 'april', 'may_long', 'june',
114 'july', 'august', 'september', 'october', 'november',
117 static public $mMonthGenMsgs = array(
118 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
119 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
122 static public $mMonthAbbrevMsgs = array(
123 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
124 'sep', 'oct', 'nov', 'dec'
127 static public $mIranianCalendarMonthMsgs = array(
128 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
129 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
130 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
131 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
134 static public $mHebrewCalendarMonthMsgs = array(
135 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
136 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
137 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
138 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
139 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
142 static public $mHebrewCalendarMonthGenMsgs = array(
143 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
144 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
145 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
146 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
147 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
150 static public $mHijriCalendarMonthMsgs = array(
151 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
152 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
153 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
154 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
161 static public $durationIntervals = array(
162 'millennia' => 31557600000,
163 'centuries' => 3155760000,
164 'decades' => 315576000,
165 'years' => 31557600, // 86400 * 365.25
174 * Get a cached language object for a given language code
175 * @param $code String
178 static function factory( $code ) {
179 if ( !isset( self
::$mLangObjCache[$code] ) ) {
180 if ( count( self
::$mLangObjCache ) > 10 ) {
181 // Don't keep a billion objects around, that's stupid.
182 self
::$mLangObjCache = array();
184 self
::$mLangObjCache[$code] = self
::newFromCode( $code );
186 return self
::$mLangObjCache[$code];
190 * Create a language object for a given language code
191 * @param $code String
192 * @throws MWException
195 protected static function newFromCode( $code ) {
196 // Protect against path traversal below
197 if ( !Language
::isValidCode( $code )
198 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
200 throw new MWException( "Invalid language code \"$code\"" );
203 if ( !Language
::isValidBuiltInCode( $code ) ) {
204 // It's not possible to customise this code with class files, so
205 // just return a Language object. This is to support uselang= hacks.
206 $lang = new Language
;
207 $lang->setCode( $code );
211 // Check if there is a language class for the code
212 $class = self
::classFromCode( $code );
213 self
::preloadLanguageClass( $class );
214 if ( MWInit
::classExists( $class ) ) {
219 // Keep trying the fallback list until we find an existing class
220 $fallbacks = Language
::getFallbacksFor( $code );
221 foreach ( $fallbacks as $fallbackCode ) {
222 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
223 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
226 $class = self
::classFromCode( $fallbackCode );
227 self
::preloadLanguageClass( $class );
228 if ( MWInit
::classExists( $class ) ) {
229 $lang = Language
::newFromCode( $fallbackCode );
230 $lang->setCode( $code );
235 throw new MWException( "Invalid fallback sequence for language '$code'" );
239 * Returns true if a language code string is of a valid form, whether or
240 * not it exists. This includes codes which are used solely for
241 * customisation via the MediaWiki namespace.
243 * @param $code string
247 public static function isValidCode( $code ) {
249 // People think language codes are html safe, so enforce it.
250 // Ideally we should only allow a-zA-Z0-9-
251 // but, .+ and other chars are often used for {{int:}} hacks
252 // see bugs 37564, 37587, 36938
253 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
254 && !preg_match( Title
::getTitleInvalidRegex(), $code );
258 * Returns true if a language code is of a valid form for the purposes of
259 * internal customisation of MediaWiki, via Messages*.php.
261 * @param $code string
263 * @throws MWException
267 public static function isValidBuiltInCode( $code ) {
269 if( !is_string($code) ) {
270 $type = gettype( $code );
271 if( $type === 'object' ) {
272 $addmsg = " of class " . get_class( $code );
276 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
279 return preg_match( '/^[a-z0-9-]+$/i', $code );
284 * @return String Name of the language class
286 public static function classFromCode( $code ) {
287 if ( $code == 'en' ) {
290 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
295 * Includes language class files
297 * @param $class string Name of the language class
299 public static function preloadLanguageClass( $class ) {
302 if ( $class === 'Language' ) {
306 if ( !defined( 'MW_COMPILED' ) ) {
307 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
308 include_once( "$IP/languages/classes/$class.php" );
314 * Get the LocalisationCache instance
316 * @return LocalisationCache
318 public static function getLocalisationCache() {
319 if ( is_null( self
::$dataCache ) ) {
320 global $wgLocalisationCacheConf;
321 $class = $wgLocalisationCacheConf['class'];
322 self
::$dataCache = new $class( $wgLocalisationCacheConf );
324 return self
::$dataCache;
327 function __construct() {
328 $this->mConverter
= new FakeConverter( $this );
329 // Set the code to the name of the descendant
330 if ( get_class( $this ) == 'Language' ) {
333 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
335 self
::getLocalisationCache();
339 * Reduce memory usage
341 function __destruct() {
342 foreach ( $this as $name => $value ) {
343 unset( $this->$name );
348 * Hook which will be called if this is the content language.
349 * Descendants can use this to register hook functions or modify globals
351 function initContLang() { }
354 * Same as getFallbacksFor for current language.
356 * @deprecated in 1.19
358 function getFallbackLanguageCode() {
359 wfDeprecated( __METHOD__
);
360 return self
::getFallbackFor( $this->mCode
);
367 function getFallbackLanguages() {
368 return self
::getFallbacksFor( $this->mCode
);
372 * Exports $wgBookstoreListEn
375 function getBookstoreList() {
376 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
382 public function getNamespaces() {
383 if ( is_null( $this->namespaceNames
) ) {
384 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
386 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
387 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
389 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
391 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
392 if ( $wgMetaNamespaceTalk ) {
393 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
395 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
396 $this->namespaceNames
[NS_PROJECT_TALK
] =
397 $this->fixVariableInNamespace( $talk );
400 # Sometimes a language will be localised but not actually exist on this wiki.
401 foreach ( $this->namespaceNames
as $key => $text ) {
402 if ( !isset( $validNamespaces[$key] ) ) {
403 unset( $this->namespaceNames
[$key] );
407 # The above mixing may leave namespaces out of canonical order.
408 # Re-order by namespace ID number...
409 ksort( $this->namespaceNames
);
411 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
413 return $this->namespaceNames
;
417 * Arbitrarily set all of the namespace names at once. Mainly used for testing
418 * @param $namespaces Array of namespaces (id => name)
420 public function setNamespaces( array $namespaces ) {
421 $this->namespaceNames
= $namespaces;
422 $this->mNamespaceIds
= null;
426 * Resets all of the namespace caches. Mainly used for testing
428 public function resetNamespaces( ) {
429 $this->namespaceNames
= null;
430 $this->mNamespaceIds
= null;
431 $this->namespaceAliases
= null;
435 * A convenience function that returns the same thing as
436 * getNamespaces() except with the array values changed to ' '
437 * where it found '_', useful for producing output to be displayed
438 * e.g. in <select> forms.
442 function getFormattedNamespaces() {
443 $ns = $this->getNamespaces();
444 foreach ( $ns as $k => $v ) {
445 $ns[$k] = strtr( $v, '_', ' ' );
451 * Get a namespace value by key
453 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
454 * echo $mw_ns; // prints 'MediaWiki'
457 * @param $index Int: the array key of the namespace to return
458 * @return mixed, string if the namespace value exists, otherwise false
460 function getNsText( $index ) {
461 $ns = $this->getNamespaces();
462 return isset( $ns[$index] ) ?
$ns[$index] : false;
466 * A convenience function that returns the same thing as
467 * getNsText() except with '_' changed to ' ', useful for
470 * @param $index string
474 function getFormattedNsText( $index ) {
475 $ns = $this->getNsText( $index );
476 return strtr( $ns, '_', ' ' );
480 * Returns gender-dependent namespace alias if available.
481 * @param $index Int: namespace index
482 * @param $gender String: gender key (male, female... )
486 function getGenderNsText( $index, $gender ) {
487 global $wgExtraGenderNamespaces;
489 $ns = $wgExtraGenderNamespaces + self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
490 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
494 * Whether this language makes distinguishes genders for example in
499 function needsGenderDistinction() {
500 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
501 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
502 // $wgExtraGenderNamespaces overrides everything
504 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
505 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
506 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
509 // Check what is in i18n files
510 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
511 return count( $aliases ) > 0;
516 * Get a namespace key by value, case insensitive.
517 * Only matches namespace names for the current language, not the
518 * canonical ones defined in Namespace.php.
520 * @param $text String
521 * @return mixed An integer if $text is a valid value otherwise false
523 function getLocalNsIndex( $text ) {
524 $lctext = $this->lc( $text );
525 $ids = $this->getNamespaceIds();
526 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
532 function getNamespaceAliases() {
533 if ( is_null( $this->namespaceAliases
) ) {
534 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
538 foreach ( $aliases as $name => $index ) {
539 if ( $index === NS_PROJECT_TALK
) {
540 unset( $aliases[$name] );
541 $name = $this->fixVariableInNamespace( $name );
542 $aliases[$name] = $index;
547 global $wgExtraGenderNamespaces;
548 $genders = $wgExtraGenderNamespaces +
(array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
549 foreach ( $genders as $index => $forms ) {
550 foreach ( $forms as $alias ) {
551 $aliases[$alias] = $index;
555 $this->namespaceAliases
= $aliases;
557 return $this->namespaceAliases
;
563 function getNamespaceIds() {
564 if ( is_null( $this->mNamespaceIds
) ) {
565 global $wgNamespaceAliases;
566 # Put namespace names and aliases into a hashtable.
567 # If this is too slow, then we should arrange it so that it is done
568 # before caching. The catch is that at pre-cache time, the above
569 # class-specific fixup hasn't been done.
570 $this->mNamespaceIds
= array();
571 foreach ( $this->getNamespaces() as $index => $name ) {
572 $this->mNamespaceIds
[$this->lc( $name )] = $index;
574 foreach ( $this->getNamespaceAliases() as $name => $index ) {
575 $this->mNamespaceIds
[$this->lc( $name )] = $index;
577 if ( $wgNamespaceAliases ) {
578 foreach ( $wgNamespaceAliases as $name => $index ) {
579 $this->mNamespaceIds
[$this->lc( $name )] = $index;
583 return $this->mNamespaceIds
;
587 * Get a namespace key by value, case insensitive. Canonical namespace
588 * names override custom ones defined for the current language.
590 * @param $text String
591 * @return mixed An integer if $text is a valid value otherwise false
593 function getNsIndex( $text ) {
594 $lctext = $this->lc( $text );
595 $ns = MWNamespace
::getCanonicalIndex( $lctext );
596 if ( $ns !== null ) {
599 $ids = $this->getNamespaceIds();
600 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
604 * short names for language variants used for language conversion links.
606 * @param $code String
607 * @param $usemsg bool Use the "variantname-xyz" message if it exists
610 function getVariantname( $code, $usemsg = true ) {
611 $msg = "variantname-$code";
612 if ( $usemsg && wfMessage( $msg )->exists() ) {
613 return $this->getMessageFromDB( $msg );
615 $name = self
::fetchLanguageName( $code );
617 return $name; # if it's defined as a language name, show that
619 # otherwise, output the language code
625 * @param $name string
628 function specialPage( $name ) {
629 $aliases = $this->getSpecialPageAliases();
630 if ( isset( $aliases[$name][0] ) ) {
631 $name = $aliases[$name][0];
633 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
639 function getQuickbarSettings() {
641 $this->getMessage( 'qbsettings-none' ),
642 $this->getMessage( 'qbsettings-fixedleft' ),
643 $this->getMessage( 'qbsettings-fixedright' ),
644 $this->getMessage( 'qbsettings-floatingleft' ),
645 $this->getMessage( 'qbsettings-floatingright' ),
646 $this->getMessage( 'qbsettings-directionality' )
653 function getDatePreferences() {
654 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
660 function getDateFormats() {
661 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
665 * @return array|string
667 function getDefaultDateFormat() {
668 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
669 if ( $df === 'dmy or mdy' ) {
670 global $wgAmericanDates;
671 return $wgAmericanDates ?
'mdy' : 'dmy';
680 function getDatePreferenceMigrationMap() {
681 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
688 function getImageFile( $image ) {
689 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
695 function getExtraUserToggles() {
696 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
703 function getUserToggle( $tog ) {
704 return $this->getMessageFromDB( "tog-$tog" );
708 * Get native language names, indexed by code.
709 * Only those defined in MediaWiki, no other data like CLDR.
710 * If $customisedOnly is true, only returns codes with a messages file
712 * @param $customisedOnly bool
715 * @deprecated in 1.20, use fetchLanguageNames()
717 public static function getLanguageNames( $customisedOnly = false ) {
718 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
722 * Get translated language names. This is done on best effort and
723 * by default this is exactly the same as Language::getLanguageNames.
724 * The CLDR extension provides translated names.
725 * @param $code String Language code.
726 * @return Array language code => language name
728 * @deprecated in 1.20, use fetchLanguageNames()
730 public static function getTranslatedLanguageNames( $code ) {
731 return self
::fetchLanguageNames( $code, 'all' );
735 * Get an array of language names, indexed by code.
736 * @param $inLanguage null|string: Code of language in which to return the names
737 * Use null for autonyms (native names)
738 * @param $include string:
739 * 'all' all available languages
740 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
741 * 'mwfile' only if the language is in 'mw' *and* has a message file
742 * @return array: language code => language name
745 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
746 global $wgExtraLanguageNames;
747 static $coreLanguageNames;
749 if ( $coreLanguageNames === null ) {
750 include( MWInit
::compiledPath( 'languages/Names.php' ) );
756 # TODO: also include when $inLanguage is null, when this code is more efficient
757 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
760 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
761 foreach ( $mwNames as $mwCode => $mwName ) {
762 # - Prefer own MediaWiki native name when not using the hook
763 # - For other names just add if not added through the hook
764 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
765 $names[$mwCode] = $mwName;
769 if ( $include === 'all' ) {
774 $coreCodes = array_keys( $mwNames );
775 foreach( $coreCodes as $coreCode ) {
776 $returnMw[$coreCode] = $names[$coreCode];
779 if( $include === 'mwfile' ) {
780 $namesMwFile = array();
781 # We do this using a foreach over the codes instead of a directory
782 # loop so that messages files in extensions will work correctly.
783 foreach ( $returnMw as $code => $value ) {
784 if ( is_readable( self
::getMessagesFileName( $code ) ) ) {
785 $namesMwFile[$code] = $names[$code];
790 # 'mw' option; default if it's not one of the other two options (all/mwfile)
795 * @param $code string: The code of the language for which to get the name
796 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
797 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
798 * @return string: Language name or empty
801 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
802 $array = self
::fetchLanguageNames( $inLanguage, $include );
803 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
807 * Get a message from the MediaWiki namespace.
809 * @param $msg String: message name
812 function getMessageFromDB( $msg ) {
813 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
817 * Get the native language name of $code.
818 * Only if defined in MediaWiki, no other data like CLDR.
819 * @param $code string
821 * @deprecated in 1.20, use fetchLanguageName()
823 function getLanguageName( $code ) {
824 return self
::fetchLanguageName( $code );
831 function getMonthName( $key ) {
832 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
838 function getMonthNamesArray() {
839 $monthNames = array( '' );
840 for ( $i = 1; $i < 13; $i++
) {
841 $monthNames[] = $this->getMonthName( $i );
850 function getMonthNameGen( $key ) {
851 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
858 function getMonthAbbreviation( $key ) {
859 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
865 function getMonthAbbreviationsArray() {
866 $monthNames = array( '' );
867 for ( $i = 1; $i < 13; $i++
) {
868 $monthNames[] = $this->getMonthAbbreviation( $i );
877 function getWeekdayName( $key ) {
878 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
885 function getWeekdayAbbreviation( $key ) {
886 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
893 function getIranianCalendarMonthName( $key ) {
894 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
901 function getHebrewCalendarMonthName( $key ) {
902 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
909 function getHebrewCalendarMonthNameGen( $key ) {
910 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
917 function getHijriCalendarMonthName( $key ) {
918 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
922 * This is a workalike of PHP's date() function, but with better
923 * internationalisation, a reduced set of format characters, and a better
926 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
927 * PHP manual for definitions. There are a number of extensions, which
930 * xn Do not translate digits of the next numeric format character
931 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
932 * xr Use roman numerals for the next numeric format character
933 * xh Use hebrew numerals for the next numeric format character
935 * xg Genitive month name
937 * xij j (day number) in Iranian calendar
938 * xiF F (month name) in Iranian calendar
939 * xin n (month number) in Iranian calendar
940 * xiy y (two digit year) in Iranian calendar
941 * xiY Y (full year) in Iranian calendar
943 * xjj j (day number) in Hebrew calendar
944 * xjF F (month name) in Hebrew calendar
945 * xjt t (days in month) in Hebrew calendar
946 * xjx xg (genitive month name) in Hebrew calendar
947 * xjn n (month number) in Hebrew calendar
948 * xjY Y (full year) in Hebrew calendar
950 * xmj j (day number) in Hijri calendar
951 * xmF F (month name) in Hijri calendar
952 * xmn n (month number) in Hijri calendar
953 * xmY Y (full year) in Hijri calendar
955 * xkY Y (full year) in Thai solar calendar. Months and days are
956 * identical to the Gregorian calendar
957 * xoY Y (full year) in Minguo calendar or Juche year.
958 * Months and days are identical to the
960 * xtY Y (full year) in Japanese nengo. Months and days are
961 * identical to the Gregorian calendar
963 * Characters enclosed in double quotes will be considered literal (with
964 * the quotes themselves removed). Unmatched quotes will be considered
965 * literal quotes. Example:
967 * "The month is" F => The month is January
970 * Backslash escaping is also supported.
972 * Input timestamp is assumed to be pre-normalized to the desired local
975 * @param $format String
976 * @param $ts String: 14-character timestamp
979 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
983 function sprintfDate( $format, $ts ) {
996 for ( $p = 0; $p < strlen( $format ); $p++
) {
999 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
1000 $code .= $format[++
$p];
1003 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
1004 $code .= $format[++
$p];
1015 $rawToggle = !$rawToggle;
1024 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1027 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
1028 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1031 $num = substr( $ts, 6, 2 );
1034 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
1035 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) +
1 );
1038 $num = intval( substr( $ts, 6, 2 ) );
1042 $iranian = self
::tsToIranian( $ts );
1048 $hijri = self
::tsToHijri( $ts );
1054 $hebrew = self
::tsToHebrew( $ts );
1060 $unix = wfTimestamp( TS_UNIX
, $ts );
1062 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) +
1 );
1066 $unix = wfTimestamp( TS_UNIX
, $ts );
1068 $w = gmdate( 'w', $unix );
1073 $unix = wfTimestamp( TS_UNIX
, $ts );
1075 $num = gmdate( 'w', $unix );
1079 $unix = wfTimestamp( TS_UNIX
, $ts );
1081 $num = gmdate( 'z', $unix );
1085 $unix = wfTimestamp( TS_UNIX
, $ts );
1087 $num = gmdate( 'W', $unix );
1090 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1094 $iranian = self
::tsToIranian( $ts );
1096 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1100 $hijri = self
::tsToHijri( $ts );
1102 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1106 $hebrew = self
::tsToHebrew( $ts );
1108 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1111 $num = substr( $ts, 4, 2 );
1114 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1117 $num = intval( substr( $ts, 4, 2 ) );
1121 $iranian = self
::tsToIranian( $ts );
1127 $hijri = self
::tsToHijri ( $ts );
1133 $hebrew = self
::tsToHebrew( $ts );
1139 $unix = wfTimestamp( TS_UNIX
, $ts );
1141 $num = gmdate( 't', $unix );
1145 $hebrew = self
::tsToHebrew( $ts );
1151 $unix = wfTimestamp( TS_UNIX
, $ts );
1153 $num = gmdate( 'L', $unix );
1157 $unix = wfTimestamp( TS_UNIX
, $ts );
1159 $num = gmdate( 'o', $unix );
1162 $num = substr( $ts, 0, 4 );
1166 $iranian = self
::tsToIranian( $ts );
1172 $hijri = self
::tsToHijri( $ts );
1178 $hebrew = self
::tsToHebrew( $ts );
1184 $thai = self
::tsToYear( $ts, 'thai' );
1190 $minguo = self
::tsToYear( $ts, 'minguo' );
1196 $tenno = self
::tsToYear( $ts, 'tenno' );
1201 $num = substr( $ts, 2, 2 );
1205 $iranian = self
::tsToIranian( $ts );
1207 $num = substr( $iranian[0], -2 );
1210 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1213 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1216 $h = substr( $ts, 8, 2 );
1217 $num = $h %
12 ?
$h %
12 : 12;
1220 $num = intval( substr( $ts, 8, 2 ) );
1223 $h = substr( $ts, 8, 2 );
1224 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1227 $num = substr( $ts, 8, 2 );
1230 $num = substr( $ts, 10, 2 );
1233 $num = substr( $ts, 12, 2 );
1237 $unix = wfTimestamp( TS_UNIX
, $ts );
1239 $s .= gmdate( 'c', $unix );
1243 $unix = wfTimestamp( TS_UNIX
, $ts );
1245 $s .= gmdate( 'r', $unix );
1249 $unix = wfTimestamp( TS_UNIX
, $ts );
1254 # Backslash escaping
1255 if ( $p < strlen( $format ) - 1 ) {
1256 $s .= $format[++
$p];
1263 if ( $p < strlen( $format ) - 1 ) {
1264 $endQuote = strpos( $format, '"', $p +
1 );
1265 if ( $endQuote === false ) {
1266 # No terminating quote, assume literal "
1269 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1273 # Quote at end of string, assume literal "
1280 if ( $num !== false ) {
1281 if ( $rawToggle ||
$raw ) {
1284 } elseif ( $roman ) {
1285 $s .= Language
::romanNumeral( $num );
1287 } elseif ( $hebrewNum ) {
1288 $s .= self
::hebrewNumeral( $num );
1291 $s .= $this->formatNum( $num, true );
1298 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1299 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1302 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1303 * Gregorian dates to Iranian dates. Originally written in C, it
1304 * is released under the terms of GNU Lesser General Public
1305 * License. Conversion to PHP was performed by Niklas Laxström.
1307 * Link: http://www.farsiweb.info/jalali/jalali.c
1313 private static function tsToIranian( $ts ) {
1314 $gy = substr( $ts, 0, 4 ) -1600;
1315 $gm = substr( $ts, 4, 2 ) -1;
1316 $gd = substr( $ts, 6, 2 ) -1;
1318 # Days passed from the beginning (including leap years)
1320 +
floor( ( $gy +
3 ) / 4 )
1321 - floor( ( $gy +
99 ) / 100 )
1322 +
floor( ( $gy +
399 ) / 400 );
1324 // Add days of the past months of this year
1325 for ( $i = 0; $i < $gm; $i++
) {
1326 $gDayNo +
= self
::$GREG_DAYS[$i];
1330 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1334 // Days passed in current month
1335 $gDayNo +
= (int)$gd;
1337 $jDayNo = $gDayNo - 79;
1339 $jNp = floor( $jDayNo / 12053 );
1342 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1345 if ( $jDayNo >= 366 ) {
1346 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1347 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1350 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1351 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1357 return array( $jy, $jm, $jd );
1361 * Converting Gregorian dates to Hijri dates.
1363 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1365 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1371 private static function tsToHijri( $ts ) {
1372 $year = substr( $ts, 0, 4 );
1373 $month = substr( $ts, 4, 2 );
1374 $day = substr( $ts, 6, 2 );
1382 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1383 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1386 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1387 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1388 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1391 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1392 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1395 $zl = $zjd -1948440 +
10632;
1396 $zn = (int)( ( $zl - 1 ) / 10631 );
1397 $zl = $zl - 10631 * $zn +
354;
1398 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1399 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1400 $zm = (int)( ( 24 * $zl ) / 709 );
1401 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1402 $zy = 30 * $zn +
$zj - 30;
1404 return array( $zy, $zm, $zd );
1408 * Converting Gregorian dates to Hebrew dates.
1410 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1411 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1412 * to translate the relevant functions into PHP and release them under
1415 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1416 * and Adar II is 14. In a non-leap year, Adar is 6.
1422 private static function tsToHebrew( $ts ) {
1424 $year = substr( $ts, 0, 4 );
1425 $month = substr( $ts, 4, 2 );
1426 $day = substr( $ts, 6, 2 );
1428 # Calculate Hebrew year
1429 $hebrewYear = $year +
3760;
1431 # Month number when September = 1, August = 12
1433 if ( $month > 12 ) {
1440 # Calculate day of year from 1 September
1442 for ( $i = 1; $i < $month; $i++
) {
1446 # Check if the year is leap
1447 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1450 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1457 # Calculate the start of the Hebrew year
1458 $start = self
::hebrewYearStart( $hebrewYear );
1460 # Calculate next year's start
1461 if ( $dayOfYear <= $start ) {
1462 # Day is before the start of the year - it is the previous year
1464 $nextStart = $start;
1468 # Add days since previous year's 1 September
1470 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1474 # Start of the new (previous) year
1475 $start = self
::hebrewYearStart( $hebrewYear );
1478 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1481 # Calculate Hebrew day of year
1482 $hebrewDayOfYear = $dayOfYear - $start;
1484 # Difference between year's days
1485 $diff = $nextStart - $start;
1486 # Add 12 (or 13 for leap years) days to ignore the difference between
1487 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1488 # difference is only about the year type
1489 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1495 # Check the year pattern, and is leap year
1496 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1497 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1498 # and non-leap years
1499 $yearPattern = $diff %
30;
1500 # Check if leap year
1501 $isLeap = $diff >= 30;
1503 # Calculate day in the month from number of day in the Hebrew year
1504 # Don't check Adar - if the day is not in Adar, we will stop before;
1505 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1506 $hebrewDay = $hebrewDayOfYear;
1509 while ( $hebrewMonth <= 12 ) {
1510 # Calculate days in this month
1511 if ( $isLeap && $hebrewMonth == 6 ) {
1512 # Adar in a leap year
1514 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1516 if ( $hebrewDay <= $days ) {
1520 # Subtract the days of Adar I
1521 $hebrewDay -= $days;
1524 if ( $hebrewDay <= $days ) {
1530 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1531 # Cheshvan in a complete year (otherwise as the rule below)
1533 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1534 # Kislev in an incomplete year (otherwise as the rule below)
1537 # Odd months have 30 days, even have 29
1538 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1540 if ( $hebrewDay <= $days ) {
1541 # In the current month
1544 # Subtract the days of the current month
1545 $hebrewDay -= $days;
1546 # Try in the next month
1551 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1555 * This calculates the Hebrew year start, as days since 1 September.
1556 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1557 * Used for Hebrew date.
1563 private static function hebrewYearStart( $year ) {
1564 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1565 $b = intval( ( $year - 1 ) %
4 );
1566 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1570 $Mar = intval( $m );
1576 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1577 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1579 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1581 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1585 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1590 * Algorithm to convert Gregorian dates to Thai solar dates,
1591 * Minguo dates or Minguo dates.
1593 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1594 * http://en.wikipedia.org/wiki/Minguo_calendar
1595 * http://en.wikipedia.org/wiki/Japanese_era_name
1597 * @param $ts String: 14-character timestamp
1598 * @param $cName String: calender name
1599 * @return Array: converted year, month, day
1601 private static function tsToYear( $ts, $cName ) {
1602 $gy = substr( $ts, 0, 4 );
1603 $gm = substr( $ts, 4, 2 );
1604 $gd = substr( $ts, 6, 2 );
1606 if ( !strcmp( $cName, 'thai' ) ) {
1608 # Add 543 years to the Gregorian calendar
1609 # Months and days are identical
1610 $gy_offset = $gy +
543;
1611 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1613 # Deduct 1911 years from the Gregorian calendar
1614 # Months and days are identical
1615 $gy_offset = $gy - 1911;
1616 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1617 # Nengō dates up to Meiji period
1618 # Deduct years from the Gregorian calendar
1619 # depending on the nengo periods
1620 # Months and days are identical
1621 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1623 $gy_gannen = $gy - 1868 +
1;
1624 $gy_offset = $gy_gannen;
1625 if ( $gy_gannen == 1 ) {
1628 $gy_offset = '明治' . $gy_offset;
1630 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1631 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1632 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1633 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1634 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1638 $gy_gannen = $gy - 1912 +
1;
1639 $gy_offset = $gy_gannen;
1640 if ( $gy_gannen == 1 ) {
1643 $gy_offset = '大正' . $gy_offset;
1645 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1646 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1647 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1651 $gy_gannen = $gy - 1926 +
1;
1652 $gy_offset = $gy_gannen;
1653 if ( $gy_gannen == 1 ) {
1656 $gy_offset = '昭和' . $gy_offset;
1659 $gy_gannen = $gy - 1989 +
1;
1660 $gy_offset = $gy_gannen;
1661 if ( $gy_gannen == 1 ) {
1664 $gy_offset = '平成' . $gy_offset;
1670 return array( $gy_offset, $gm, $gd );
1674 * Roman number formatting up to 10000
1680 static function romanNumeral( $num ) {
1681 static $table = array(
1682 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1683 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1684 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1685 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM', 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1688 $num = intval( $num );
1689 if ( $num > 10000 ||
$num <= 0 ) {
1694 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1695 if ( $num >= $pow10 ) {
1696 $s .= $table[$i][(int)floor( $num / $pow10 )];
1698 $num = $num %
$pow10;
1704 * Hebrew Gematria number formatting up to 9999
1710 static function hebrewNumeral( $num ) {
1711 static $table = array(
1712 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1713 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1714 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1715 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1718 $num = intval( $num );
1719 if ( $num > 9999 ||
$num <= 0 ) {
1724 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1725 if ( $num >= $pow10 ) {
1726 if ( $num == 15 ||
$num == 16 ) {
1727 $s .= $table[0][9] . $table[0][$num - 9];
1730 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1731 if ( $pow10 == 1000 ) {
1736 $num = $num %
$pow10;
1738 if ( strlen( $s ) == 2 ) {
1741 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1742 $str .= substr( $s, strlen( $s ) - 2, 2 );
1744 $start = substr( $str, 0, strlen( $str ) - 2 );
1745 $end = substr( $str, strlen( $str ) - 2 );
1748 $str = $start . 'ך';
1751 $str = $start . 'ם';
1754 $str = $start . 'ן';
1757 $str = $start . 'ף';
1760 $str = $start . 'ץ';
1767 * Used by date() and time() to adjust the time output.
1769 * @param $ts Int the time in date('YmdHis') format
1770 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1771 * get user timecorrection setting)
1774 function userAdjust( $ts, $tz = false ) {
1775 global $wgUser, $wgLocalTZoffset;
1777 if ( $tz === false ) {
1778 $tz = $wgUser->getOption( 'timecorrection' );
1781 $data = explode( '|', $tz, 3 );
1783 if ( $data[0] == 'ZoneInfo' ) {
1784 wfSuppressWarnings();
1785 $userTZ = timezone_open( $data[2] );
1786 wfRestoreWarnings();
1787 if ( $userTZ !== false ) {
1788 $date = date_create( $ts, timezone_open( 'UTC' ) );
1789 date_timezone_set( $date, $userTZ );
1790 $date = date_format( $date, 'YmdHis' );
1793 # Unrecognized timezone, default to 'Offset' with the stored offset.
1794 $data[0] = 'Offset';
1798 if ( $data[0] == 'System' ||
$tz == '' ) {
1799 # Global offset in minutes.
1800 if ( isset( $wgLocalTZoffset ) ) {
1801 $minDiff = $wgLocalTZoffset;
1803 } elseif ( $data[0] == 'Offset' ) {
1804 $minDiff = intval( $data[1] );
1806 $data = explode( ':', $tz );
1807 if ( count( $data ) == 2 ) {
1808 $data[0] = intval( $data[0] );
1809 $data[1] = intval( $data[1] );
1810 $minDiff = abs( $data[0] ) * 60 +
$data[1];
1811 if ( $data[0] < 0 ) {
1812 $minDiff = -$minDiff;
1815 $minDiff = intval( $data[0] ) * 60;
1819 # No difference ? Return time unchanged
1820 if ( 0 == $minDiff ) {
1824 wfSuppressWarnings(); // E_STRICT system time bitching
1825 # Generate an adjusted date; take advantage of the fact that mktime
1826 # will normalize out-of-range values so we don't have to split $minDiff
1827 # into hours and minutes.
1829 (int)substr( $ts, 8, 2 ) ), # Hours
1830 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
1831 (int)substr( $ts, 12, 2 ), # Seconds
1832 (int)substr( $ts, 4, 2 ), # Month
1833 (int)substr( $ts, 6, 2 ), # Day
1834 (int)substr( $ts, 0, 4 ) ); # Year
1836 $date = date( 'YmdHis', $t );
1837 wfRestoreWarnings();
1843 * This is meant to be used by time(), date(), and timeanddate() to get
1844 * the date preference they're supposed to use, it should be used in
1848 * function timeanddate([...], $format = true) {
1849 * $datePreference = $this->dateFormat($format);
1854 * @param $usePrefs Mixed: if true, the user's preference is used
1855 * if false, the site/language default is used
1856 * if int/string, assumed to be a format.
1859 function dateFormat( $usePrefs = true ) {
1862 if ( is_bool( $usePrefs ) ) {
1864 $datePreference = $wgUser->getDatePreference();
1866 $datePreference = (string)User
::getDefaultOption( 'date' );
1869 $datePreference = (string)$usePrefs;
1873 if ( $datePreference == '' ) {
1877 return $datePreference;
1881 * Get a format string for a given type and preference
1882 * @param $type string May be date, time or both
1883 * @param $pref string The format name as it appears in Messages*.php
1887 function getDateFormatString( $type, $pref ) {
1888 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
1889 if ( $pref == 'default' ) {
1890 $pref = $this->getDefaultDateFormat();
1891 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1893 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1894 if ( is_null( $df ) ) {
1895 $pref = $this->getDefaultDateFormat();
1896 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1899 $this->dateFormatStrings
[$type][$pref] = $df;
1901 return $this->dateFormatStrings
[$type][$pref];
1905 * @param $ts Mixed: the time format which needs to be turned into a
1906 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1907 * @param $adj Bool: whether to adjust the time output according to the
1908 * user configured offset ($timecorrection)
1909 * @param $format Mixed: true to use user's date format preference
1910 * @param $timecorrection String|bool the time offset as returned by
1911 * validateTimeZone() in Special:Preferences
1914 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1915 $ts = wfTimestamp( TS_MW
, $ts );
1917 $ts = $this->userAdjust( $ts, $timecorrection );
1919 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1920 return $this->sprintfDate( $df, $ts );
1924 * @param $ts Mixed: the time format which needs to be turned into a
1925 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1926 * @param $adj Bool: whether to adjust the time output according to the
1927 * user configured offset ($timecorrection)
1928 * @param $format Mixed: true to use user's date format preference
1929 * @param $timecorrection String|bool the time offset as returned by
1930 * validateTimeZone() in Special:Preferences
1933 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1934 $ts = wfTimestamp( TS_MW
, $ts );
1936 $ts = $this->userAdjust( $ts, $timecorrection );
1938 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1939 return $this->sprintfDate( $df, $ts );
1943 * @param $ts Mixed: the time format which needs to be turned into a
1944 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1945 * @param $adj Bool: whether to adjust the time output according to the
1946 * user configured offset ($timecorrection)
1947 * @param $format Mixed: what format to return, if it's false output the
1948 * default one (default true)
1949 * @param $timecorrection String|bool the time offset as returned by
1950 * validateTimeZone() in Special:Preferences
1953 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1954 $ts = wfTimestamp( TS_MW
, $ts );
1956 $ts = $this->userAdjust( $ts, $timecorrection );
1958 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1959 return $this->sprintfDate( $df, $ts );
1963 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
1967 * @param integer $seconds The amount of seconds.
1968 * @param array $chosenIntervals The intervals to enable.
1972 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
1973 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
1975 $segments = array();
1977 foreach ( $intervals as $intervalName => $intervalValue ) {
1978 $message = new Message( 'duration-' . $intervalName, array( $intervalValue ) );
1979 $segments[] = $message->inLanguage( $this )->escaped();
1982 return $this->listToText( $segments );
1986 * Takes a number of seconds and returns an array with a set of corresponding intervals.
1987 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
1991 * @param integer $seconds The amount of seconds.
1992 * @param array $chosenIntervals The intervals to enable.
1996 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
1997 if ( empty( $chosenIntervals ) ) {
1998 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
2001 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2002 $sortedNames = array_keys( $intervals );
2003 $smallestInterval = array_pop( $sortedNames );
2005 $segments = array();
2007 foreach ( $intervals as $name => $length ) {
2008 $value = floor( $seconds / $length );
2010 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2011 $seconds -= $value * $length;
2012 $segments[$name] = $value;
2020 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2022 * @param $type String: can be 'date', 'time' or 'both'
2023 * @param $ts Mixed: the time format which needs to be turned into a
2024 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2025 * @param $user User object used to get preferences for timezone and format
2026 * @param $options Array, can contain the following keys:
2027 * - 'timecorrection': time correction, can have the following values:
2028 * - true: use user's preference
2029 * - false: don't use time correction
2030 * - integer: value of time correction in minutes
2031 * - 'format': format to use, can have the following values:
2032 * - true: use user's preference
2033 * - false: use default preference
2034 * - string: format to use
2038 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2039 $ts = wfTimestamp( TS_MW
, $ts );
2040 $options +
= array( 'timecorrection' => true, 'format' => true );
2041 if ( $options['timecorrection'] !== false ) {
2042 if ( $options['timecorrection'] === true ) {
2043 $offset = $user->getOption( 'timecorrection' );
2045 $offset = $options['timecorrection'];
2047 $ts = $this->userAdjust( $ts, $offset );
2049 if ( $options['format'] === true ) {
2050 $format = $user->getDatePreference();
2052 $format = $options['format'];
2054 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2055 return $this->sprintfDate( $df, $ts );
2059 * Get the formatted date for the given timestamp and formatted for
2062 * @param $ts Mixed: the time format which needs to be turned into a
2063 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2064 * @param $user User object used to get preferences for timezone and format
2065 * @param $options Array, can contain the following keys:
2066 * - 'timecorrection': time correction, can have the following values:
2067 * - true: use user's preference
2068 * - false: don't use time correction
2069 * - integer: value of time correction in minutes
2070 * - 'format': format to use, can have the following values:
2071 * - true: use user's preference
2072 * - false: use default preference
2073 * - string: format to use
2077 public function userDate( $ts, User
$user, array $options = array() ) {
2078 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2082 * Get the formatted time for the given timestamp and formatted for
2085 * @param $ts Mixed: the time format which needs to be turned into a
2086 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2087 * @param $user User object used to get preferences for timezone and format
2088 * @param $options Array, can contain the following keys:
2089 * - 'timecorrection': time correction, can have the following values:
2090 * - true: use user's preference
2091 * - false: don't use time correction
2092 * - integer: value of time correction in minutes
2093 * - 'format': format to use, can have the following values:
2094 * - true: use user's preference
2095 * - false: use default preference
2096 * - string: format to use
2100 public function userTime( $ts, User
$user, array $options = array() ) {
2101 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2105 * Get the formatted date and time for the given timestamp and formatted for
2108 * @param $ts Mixed: the time format which needs to be turned into a
2109 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2110 * @param $user User object used to get preferences for timezone and format
2111 * @param $options Array, can contain the following keys:
2112 * - 'timecorrection': time correction, can have the following values:
2113 * - true: use user's preference
2114 * - false: don't use time correction
2115 * - integer: value of time correction in minutes
2116 * - 'format': format to use, can have the following values:
2117 * - true: use user's preference
2118 * - false: use default preference
2119 * - string: format to use
2123 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2124 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2128 * @param $key string
2129 * @return array|null
2131 function getMessage( $key ) {
2132 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2138 function getAllMessages() {
2139 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2148 function iconv( $in, $out, $string ) {
2149 # This is a wrapper for iconv in all languages except esperanto,
2150 # which does some nasty x-conversions beforehand
2152 # Even with //IGNORE iconv can whine about illegal characters in
2153 # *input* string. We just ignore those too.
2154 # REF: http://bugs.php.net/bug.php?id=37166
2155 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2156 wfSuppressWarnings();
2157 $text = iconv( $in, $out . '//IGNORE', $string );
2158 wfRestoreWarnings();
2162 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2165 * @param $matches array
2166 * @return mixed|string
2168 function ucwordbreaksCallbackAscii( $matches ) {
2169 return $this->ucfirst( $matches[1] );
2173 * @param $matches array
2176 function ucwordbreaksCallbackMB( $matches ) {
2177 return mb_strtoupper( $matches[0] );
2181 * @param $matches array
2184 function ucCallback( $matches ) {
2185 list( $wikiUpperChars ) = self
::getCaseMaps();
2186 return strtr( $matches[1], $wikiUpperChars );
2190 * @param $matches array
2193 function lcCallback( $matches ) {
2194 list( , $wikiLowerChars ) = self
::getCaseMaps();
2195 return strtr( $matches[1], $wikiLowerChars );
2199 * @param $matches array
2202 function ucwordsCallbackMB( $matches ) {
2203 return mb_strtoupper( $matches[0] );
2207 * @param $matches array
2210 function ucwordsCallbackWiki( $matches ) {
2211 list( $wikiUpperChars ) = self
::getCaseMaps();
2212 return strtr( $matches[0], $wikiUpperChars );
2216 * Make a string's first character uppercase
2218 * @param $str string
2222 function ucfirst( $str ) {
2224 if ( $o < 96 ) { // if already uppercase...
2226 } elseif ( $o < 128 ) {
2227 return ucfirst( $str ); // use PHP's ucfirst()
2229 // fall back to more complex logic in case of multibyte strings
2230 return $this->uc( $str, true );
2235 * Convert a string to uppercase
2237 * @param $str string
2238 * @param $first bool
2242 function uc( $str, $first = false ) {
2243 if ( function_exists( 'mb_strtoupper' ) ) {
2245 if ( $this->isMultibyte( $str ) ) {
2246 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2248 return ucfirst( $str );
2251 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2254 if ( $this->isMultibyte( $str ) ) {
2255 $x = $first ?
'^' : '';
2256 return preg_replace_callback(
2257 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2258 array( $this, 'ucCallback' ),
2262 return $first ?
ucfirst( $str ) : strtoupper( $str );
2268 * @param $str string
2269 * @return mixed|string
2271 function lcfirst( $str ) {
2274 return strval( $str );
2275 } elseif ( $o >= 128 ) {
2276 return $this->lc( $str, true );
2277 } elseif ( $o > 96 ) {
2280 $str[0] = strtolower( $str[0] );
2286 * @param $str string
2287 * @param $first bool
2288 * @return mixed|string
2290 function lc( $str, $first = false ) {
2291 if ( function_exists( 'mb_strtolower' ) ) {
2293 if ( $this->isMultibyte( $str ) ) {
2294 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2296 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2299 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2302 if ( $this->isMultibyte( $str ) ) {
2303 $x = $first ?
'^' : '';
2304 return preg_replace_callback(
2305 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2306 array( $this, 'lcCallback' ),
2310 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2316 * @param $str string
2319 function isMultibyte( $str ) {
2320 return (bool)preg_match( '/[\x80-\xff]/', $str );
2324 * @param $str string
2325 * @return mixed|string
2327 function ucwords( $str ) {
2328 if ( $this->isMultibyte( $str ) ) {
2329 $str = $this->lc( $str );
2331 // regexp to find first letter in each word (i.e. after each space)
2332 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2334 // function to use to capitalize a single char
2335 if ( function_exists( 'mb_strtoupper' ) ) {
2336 return preg_replace_callback(
2338 array( $this, 'ucwordsCallbackMB' ),
2342 return preg_replace_callback(
2344 array( $this, 'ucwordsCallbackWiki' ),
2349 return ucwords( strtolower( $str ) );
2354 * capitalize words at word breaks
2356 * @param $str string
2359 function ucwordbreaks( $str ) {
2360 if ( $this->isMultibyte( $str ) ) {
2361 $str = $this->lc( $str );
2363 // since \b doesn't work for UTF-8, we explicitely define word break chars
2364 $breaks = "[ \-\(\)\}\{\.,\?!]";
2366 // find first letter after word break
2367 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2369 if ( function_exists( 'mb_strtoupper' ) ) {
2370 return preg_replace_callback(
2372 array( $this, 'ucwordbreaksCallbackMB' ),
2376 return preg_replace_callback(
2378 array( $this, 'ucwordsCallbackWiki' ),
2383 return preg_replace_callback(
2384 '/\b([\w\x80-\xff]+)\b/',
2385 array( $this, 'ucwordbreaksCallbackAscii' ),
2392 * Return a case-folded representation of $s
2394 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2395 * and $s2 are the same except for the case of their characters. It is not
2396 * necessary for the value returned to make sense when displayed.
2398 * Do *not* perform any other normalisation in this function. If a caller
2399 * uses this function when it should be using a more general normalisation
2400 * function, then fix the caller.
2406 function caseFold( $s ) {
2407 return $this->uc( $s );
2414 function checkTitleEncoding( $s ) {
2415 if ( is_array( $s ) ) {
2416 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2418 # Check for non-UTF-8 URLs
2419 $ishigh = preg_match( '/[\x80-\xff]/', $s );
2424 if ( function_exists( 'mb_check_encoding' ) ) {
2425 $isutf8 = mb_check_encoding( $s, 'UTF-8' );
2427 $isutf8 = preg_match( '/^(?>[\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2428 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
2434 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2440 function fallback8bitEncoding() {
2441 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2445 * Most writing systems use whitespace to break up words.
2446 * Some languages such as Chinese don't conventionally do this,
2447 * which requires special handling when breaking up words for
2452 function hasWordBreaks() {
2457 * Some languages such as Chinese require word segmentation,
2458 * Specify such segmentation when overridden in derived class.
2460 * @param $string String
2463 function segmentByWord( $string ) {
2468 * Some languages have special punctuation need to be normalized.
2469 * Make such changes here.
2471 * @param $string String
2474 function normalizeForSearch( $string ) {
2475 return self
::convertDoubleWidth( $string );
2479 * convert double-width roman characters to single-width.
2480 * range: ff00-ff5f ~= 0020-007f
2482 * @param $string string
2486 protected static function convertDoubleWidth( $string ) {
2487 static $full = null;
2488 static $half = null;
2490 if ( $full === null ) {
2491 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2492 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2493 $full = str_split( $fullWidth, 3 );
2494 $half = str_split( $halfWidth );
2497 $string = str_replace( $full, $half, $string );
2502 * @param $string string
2503 * @param $pattern string
2506 protected static function insertSpace( $string, $pattern ) {
2507 $string = preg_replace( $pattern, " $1 ", $string );
2508 $string = preg_replace( '/ +/', ' ', $string );
2513 * @param $termsArray array
2516 function convertForSearchResult( $termsArray ) {
2517 # some languages, e.g. Chinese, need to do a conversion
2518 # in order for search results to be displayed correctly
2523 * Get the first character of a string.
2528 function firstChar( $s ) {
2531 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2532 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2537 if ( isset( $matches[1] ) ) {
2538 if ( strlen( $matches[1] ) != 3 ) {
2542 // Break down Hangul syllables to grab the first jamo
2543 $code = utf8ToCodepoint( $matches[1] );
2544 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2546 } elseif ( $code < 0xb098 ) {
2547 return "\xe3\x84\xb1";
2548 } elseif ( $code < 0xb2e4 ) {
2549 return "\xe3\x84\xb4";
2550 } elseif ( $code < 0xb77c ) {
2551 return "\xe3\x84\xb7";
2552 } elseif ( $code < 0xb9c8 ) {
2553 return "\xe3\x84\xb9";
2554 } elseif ( $code < 0xbc14 ) {
2555 return "\xe3\x85\x81";
2556 } elseif ( $code < 0xc0ac ) {
2557 return "\xe3\x85\x82";
2558 } elseif ( $code < 0xc544 ) {
2559 return "\xe3\x85\x85";
2560 } elseif ( $code < 0xc790 ) {
2561 return "\xe3\x85\x87";
2562 } elseif ( $code < 0xcc28 ) {
2563 return "\xe3\x85\x88";
2564 } elseif ( $code < 0xce74 ) {
2565 return "\xe3\x85\x8a";
2566 } elseif ( $code < 0xd0c0 ) {
2567 return "\xe3\x85\x8b";
2568 } elseif ( $code < 0xd30c ) {
2569 return "\xe3\x85\x8c";
2570 } elseif ( $code < 0xd558 ) {
2571 return "\xe3\x85\x8d";
2573 return "\xe3\x85\x8e";
2580 function initEncoding() {
2581 # Some languages may have an alternate char encoding option
2582 # (Esperanto X-coding, Japanese furigana conversion, etc)
2583 # If this language is used as the primary content language,
2584 # an override to the defaults can be set here on startup.
2591 function recodeForEdit( $s ) {
2592 # For some languages we'll want to explicitly specify
2593 # which characters make it into the edit box raw
2594 # or are converted in some way or another.
2595 global $wgEditEncoding;
2596 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2599 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2607 function recodeInput( $s ) {
2608 # Take the previous into account.
2609 global $wgEditEncoding;
2610 if ( $wgEditEncoding != '' ) {
2611 $enc = $wgEditEncoding;
2615 if ( $enc == 'UTF-8' ) {
2618 return $this->iconv( $enc, 'UTF-8', $s );
2623 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2624 * also cleans up certain backwards-compatible sequences, converting them
2625 * to the modern Unicode equivalent.
2627 * This is language-specific for performance reasons only.
2633 function normalize( $s ) {
2634 global $wgAllUnicodeFixes;
2635 $s = UtfNormal
::cleanUp( $s );
2636 if ( $wgAllUnicodeFixes ) {
2637 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2638 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2645 * Transform a string using serialized data stored in the given file (which
2646 * must be in the serialized subdirectory of $IP). The file contains pairs
2647 * mapping source characters to destination characters.
2649 * The data is cached in process memory. This will go faster if you have the
2650 * FastStringSearch extension.
2652 * @param $file string
2653 * @param $string string
2655 * @throws MWException
2658 function transformUsingPairFile( $file, $string ) {
2659 if ( !isset( $this->transformData
[$file] ) ) {
2660 $data = wfGetPrecompiledData( $file );
2661 if ( $data === false ) {
2662 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
2664 $this->transformData
[$file] = new ReplacementArray( $data );
2666 return $this->transformData
[$file]->replace( $string );
2670 * For right-to-left language support
2675 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
2679 * Return the correct HTML 'dir' attribute value for this language.
2683 return $this->isRTL() ?
'rtl' : 'ltr';
2687 * Return 'left' or 'right' as appropriate alignment for line-start
2688 * for this language's text direction.
2690 * Should be equivalent to CSS3 'start' text-align value....
2694 function alignStart() {
2695 return $this->isRTL() ?
'right' : 'left';
2699 * Return 'right' or 'left' as appropriate alignment for line-end
2700 * for this language's text direction.
2702 * Should be equivalent to CSS3 'end' text-align value....
2706 function alignEnd() {
2707 return $this->isRTL() ?
'left' : 'right';
2711 * A hidden direction mark (LRM or RLM), depending on the language direction.
2712 * Unlike getDirMark(), this function returns the character as an HTML entity.
2713 * This function should be used when the output is guaranteed to be HTML,
2714 * because it makes the output HTML source code more readable. When
2715 * the output is plain text or can be escaped, getDirMark() should be used.
2717 * @param $opposite Boolean Get the direction mark opposite to your language
2720 function getDirMarkEntity( $opposite = false ) {
2721 if ( $opposite ) { return $this->isRTL() ?
'‎' : '‏'; }
2722 return $this->isRTL() ?
'‏' : '‎';
2726 * A hidden direction mark (LRM or RLM), depending on the language direction.
2727 * This function produces them as invisible Unicode characters and
2728 * the output may be hard to read and debug, so it should only be used
2729 * when the output is plain text or can be escaped. When the output is
2730 * HTML, use getDirMarkEntity() instead.
2732 * @param $opposite Boolean Get the direction mark opposite to your language
2735 function getDirMark( $opposite = false ) {
2736 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2737 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2738 if ( $opposite ) { return $this->isRTL() ?
$lrm : $rlm; }
2739 return $this->isRTL() ?
$rlm : $lrm;
2745 function capitalizeAllNouns() {
2746 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2750 * An arrow, depending on the language direction.
2752 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2755 function getArrow( $direction = 'forwards' ) {
2756 switch ( $direction ) {
2758 return $this->isRTL() ?
'←' : '→';
2760 return $this->isRTL() ?
'→' : '←';
2773 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2777 function linkPrefixExtension() {
2778 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2784 function getMagicWords() {
2785 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2788 protected function doMagicHook() {
2789 if ( $this->mMagicHookDone
) {
2792 $this->mMagicHookDone
= true;
2793 wfProfileIn( 'LanguageGetMagic' );
2794 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2795 wfProfileOut( 'LanguageGetMagic' );
2799 * Fill a MagicWord object with data from here
2803 function getMagic( $mw ) {
2804 $this->doMagicHook();
2806 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
2807 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
2809 $magicWords = $this->getMagicWords();
2810 if ( isset( $magicWords[$mw->mId
] ) ) {
2811 $rawEntry = $magicWords[$mw->mId
];
2817 if ( !is_array( $rawEntry ) ) {
2818 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
2820 $mw->mCaseSensitive
= $rawEntry[0];
2821 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
2826 * Add magic words to the extension array
2828 * @param $newWords array
2830 function addMagicWordsByLang( $newWords ) {
2831 $fallbackChain = $this->getFallbackLanguages();
2832 $fallbackChain = array_reverse( $fallbackChain );
2833 foreach ( $fallbackChain as $code ) {
2834 if ( isset( $newWords[$code] ) ) {
2835 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
2841 * Get special page names, as an associative array
2842 * case folded alias => real name
2844 function getSpecialPageAliases() {
2845 // Cache aliases because it may be slow to load them
2846 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
2848 $this->mExtendedSpecialPageAliases
=
2849 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
2850 wfRunHooks( 'LanguageGetSpecialPageAliases',
2851 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
2854 return $this->mExtendedSpecialPageAliases
;
2858 * Italic is unsuitable for some languages
2860 * @param $text String: the text to be emphasized.
2863 function emphasize( $text ) {
2864 return "<em>$text</em>";
2868 * Normally we output all numbers in plain en_US style, that is
2869 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2870 * point twohundredthirtyfive. However this is not suitable for all
2871 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2872 * Icelandic just want to use commas instead of dots, and dots instead
2873 * of commas like "293.291,235".
2875 * An example of this function being called:
2877 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2880 * See LanguageGu.php for the Gujarati implementation and
2881 * $separatorTransformTable on MessageIs.php for
2882 * the , => . and . => , implementation.
2884 * @todo check if it's viable to use localeconv() for the decimal
2886 * @param $number Mixed: the string to be formatted, should be an integer
2887 * or a floating point number.
2888 * @param $nocommafy Bool: set to true for special numbers like dates
2891 public function formatNum( $number, $nocommafy = false ) {
2892 global $wgTranslateNumerals;
2893 if ( !$nocommafy ) {
2894 $number = $this->commafy( $number );
2895 $s = $this->separatorTransformTable();
2897 $number = strtr( $number, $s );
2901 if ( $wgTranslateNumerals ) {
2902 $s = $this->digitTransformTable();
2904 $number = strtr( $number, $s );
2912 * @param $number string
2915 function parseFormattedNumber( $number ) {
2916 $s = $this->digitTransformTable();
2918 $number = strtr( $number, array_flip( $s ) );
2921 $s = $this->separatorTransformTable();
2923 $number = strtr( $number, array_flip( $s ) );
2926 $number = strtr( $number, array( ',' => '' ) );
2931 * Adds commas to a given number
2936 function commafy( $_ ) {
2937 $digitGroupingPattern = $this->digitGroupingPattern();
2938 if ( $_ === null ) {
2942 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
2943 // default grouping is at thousands, use the same for ###,###,### pattern too.
2944 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2946 // Ref: http://cldr.unicode.org/translation/number-patterns
2948 if ( intval( $_ ) < 0 ) {
2949 // For negative numbers apply the algorithm like positive number and add sign.
2951 $_ = substr( $_, 1 );
2953 $numberpart = array();
2954 $decimalpart = array();
2955 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
2956 preg_match( "/\d+/", $_, $numberpart );
2957 preg_match( "/\.\d*/", $_, $decimalpart );
2958 $groupedNumber = ( count( $decimalpart ) > 0 ) ?
$decimalpart[0]:"";
2959 if ( $groupedNumber === $_ ) {
2960 // the string does not have any number part. Eg: .12345
2961 return $sign . $groupedNumber;
2963 $start = $end = strlen( $numberpart[0] );
2964 while ( $start > 0 ) {
2965 $match = $matches[0][$numMatches -1] ;
2966 $matchLen = strlen( $match );
2967 $start = $end - $matchLen;
2971 $groupedNumber = substr( $_ , $start, $end -$start ) . $groupedNumber ;
2973 if ( $numMatches > 1 ) {
2974 // use the last pattern for the rest of the number
2978 $groupedNumber = "," . $groupedNumber;
2981 return $sign . $groupedNumber;
2987 function digitGroupingPattern() {
2988 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
2994 function digitTransformTable() {
2995 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3001 function separatorTransformTable() {
3002 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3006 * Take a list of strings and build a locale-friendly comma-separated
3007 * list, using the local comma-separator message.
3008 * The last two strings are chained with an "and".
3013 function listToText( array $l ) {
3015 $m = count( $l ) - 1;
3017 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
3019 for ( $i = $m; $i >= 0; $i-- ) {
3022 } elseif ( $i == $m - 1 ) {
3023 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
3025 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
3033 * Take a list of strings and build a locale-friendly comma-separated
3034 * list, using the local comma-separator message.
3035 * @param $list array of strings to put in a comma list
3038 function commaList( array $list ) {
3042 array( 'parsemag', 'escapenoentities', 'language' => $this )
3049 * Take a list of strings and build a locale-friendly semicolon-separated
3050 * list, using the local semicolon-separator message.
3051 * @param $list array of strings to put in a semicolon list
3054 function semicolonList( array $list ) {
3057 'semicolon-separator',
3058 array( 'parsemag', 'escapenoentities', 'language' => $this )
3065 * Same as commaList, but separate it with the pipe instead.
3066 * @param $list array of strings to put in a pipe list
3069 function pipeList( array $list ) {
3073 array( 'escapenoentities', 'language' => $this )
3080 * Truncate a string to a specified length in bytes, appending an optional
3081 * string (e.g. for ellipses)
3083 * The database offers limited byte lengths for some columns in the database;
3084 * multi-byte character sets mean we need to ensure that only whole characters
3085 * are included, otherwise broken characters can be passed to the user
3087 * If $length is negative, the string will be truncated from the beginning
3089 * @param $string String to truncate
3090 * @param $length Int: maximum length (including ellipses)
3091 * @param $ellipsis String to append to the truncated text
3092 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3093 * $adjustLength was introduced in 1.18, before that behaved as if false.
3096 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3097 # Use the localized ellipsis character
3098 if ( $ellipsis == '...' ) {
3099 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
3101 # Check if there is no need to truncate
3102 if ( $length == 0 ) {
3103 return $ellipsis; // convention
3104 } elseif ( strlen( $string ) <= abs( $length ) ) {
3105 return $string; // no need to truncate
3107 $stringOriginal = $string;
3108 # If ellipsis length is >= $length then we can't apply $adjustLength
3109 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3110 $string = $ellipsis; // this can be slightly unexpected
3111 # Otherwise, truncate and add ellipsis...
3113 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3114 if ( $length > 0 ) {
3115 $length -= $eLength;
3116 $string = substr( $string, 0, $length ); // xyz...
3117 $string = $this->removeBadCharLast( $string );
3118 $string = $string . $ellipsis;
3120 $length +
= $eLength;
3121 $string = substr( $string, $length ); // ...xyz
3122 $string = $this->removeBadCharFirst( $string );
3123 $string = $ellipsis . $string;
3126 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3127 # This check is *not* redundant if $adjustLength, due to the single case where
3128 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3129 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3132 return $stringOriginal;
3137 * Remove bytes that represent an incomplete Unicode character
3138 * at the end of string (e.g. bytes of the char are missing)
3140 * @param $string String
3143 protected function removeBadCharLast( $string ) {
3144 if ( $string != '' ) {
3145 $char = ord( $string[strlen( $string ) - 1] );
3147 if ( $char >= 0xc0 ) {
3148 # We got the first byte only of a multibyte char; remove it.
3149 $string = substr( $string, 0, -1 );
3150 } elseif ( $char >= 0x80 &&
3151 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3152 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
3154 # We chopped in the middle of a character; remove it
3162 * Remove bytes that represent an incomplete Unicode character
3163 * at the start of string (e.g. bytes of the char are missing)
3165 * @param $string String
3168 protected function removeBadCharFirst( $string ) {
3169 if ( $string != '' ) {
3170 $char = ord( $string[0] );
3171 if ( $char >= 0x80 && $char < 0xc0 ) {
3172 # We chopped in the middle of a character; remove the whole thing
3173 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3180 * Truncate a string of valid HTML to a specified length in bytes,
3181 * appending an optional string (e.g. for ellipses), and return valid HTML
3183 * This is only intended for styled/linked text, such as HTML with
3184 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3185 * Also, this will not detect things like "display:none" CSS.
3187 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3189 * @param string $text HTML string to truncate
3190 * @param int $length (zero/positive) Maximum length (including ellipses)
3191 * @param string $ellipsis String to append to the truncated text
3194 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3195 # Use the localized ellipsis character
3196 if ( $ellipsis == '...' ) {
3197 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
3199 # Check if there is clearly no need to truncate
3200 if ( $length <= 0 ) {
3201 return $ellipsis; // no text shown, nothing to format (convention)
3202 } elseif ( strlen( $text ) <= $length ) {
3203 return $text; // string short enough even *with* HTML (short-circuit)
3206 $dispLen = 0; // innerHTML legth so far
3207 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3208 $tagType = 0; // 0-open, 1-close
3209 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3210 $entityState = 0; // 0-not entity, 1-entity
3211 $tag = $ret = ''; // accumulated tag name, accumulated result string
3212 $openTags = array(); // open tag stack
3213 $maybeState = null; // possible truncation state
3215 $textLen = strlen( $text );
3216 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3217 for ( $pos = 0; true; ++
$pos ) {
3218 # Consider truncation once the display length has reached the maximim.
3219 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3220 # Check that we're not in the middle of a bracket/entity...
3221 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3222 if ( !$testingEllipsis ) {
3223 $testingEllipsis = true;
3224 # Save where we are; we will truncate here unless there turn out to
3225 # be so few remaining characters that truncation is not necessary.
3226 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3227 $maybeState = array( $ret, $openTags ); // save state
3229 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3230 # String in fact does need truncation, the truncation point was OK.
3231 list( $ret, $openTags ) = $maybeState; // reload state
3232 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3233 $ret .= $ellipsis; // add ellipsis
3237 if ( $pos >= $textLen ) break; // extra iteration just for above checks
3239 # Read the next char...
3241 $lastCh = $pos ?
$text[$pos - 1] : '';
3242 $ret .= $ch; // add to result string
3244 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3245 $entityState = 0; // for bad HTML
3246 $bracketState = 1; // tag started (checking for backslash)
3247 } elseif ( $ch == '>' ) {
3248 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3249 $entityState = 0; // for bad HTML
3250 $bracketState = 0; // out of brackets
3251 } elseif ( $bracketState == 1 ) {
3253 $tagType = 1; // close tag (e.g. "</span>")
3255 $tagType = 0; // open tag (e.g. "<span>")
3258 $bracketState = 2; // building tag name
3259 } elseif ( $bracketState == 2 ) {
3263 // Name found (e.g. "<a href=..."), add on tag attributes...
3264 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3266 } elseif ( $bracketState == 0 ) {
3267 if ( $entityState ) {
3270 $dispLen++
; // entity is one displayed char
3273 if ( $neLength == 0 && !$maybeState ) {
3274 // Save state without $ch. We want to *hit* the first
3275 // display char (to get tags) but not *use* it if truncating.
3276 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3279 $entityState = 1; // entity found, (e.g. " ")
3281 $dispLen++
; // this char is displayed
3282 // Add the next $max display text chars after this in one swoop...
3283 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3284 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3285 $dispLen +
= $skipped;
3291 // Close the last tag if left unclosed by bad HTML
3292 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3293 while ( count( $openTags ) > 0 ) {
3294 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3300 * truncateHtml() helper function
3301 * like strcspn() but adds the skipped chars to $ret
3310 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3311 if ( $len === null ) {
3312 $len = -1; // -1 means "no limit" for strcspn
3313 } elseif ( $len < 0 ) {
3317 if ( $start < strlen( $text ) ) {
3318 $skipCount = strcspn( $text, $search, $start, $len );
3319 $ret .= substr( $text, $start, $skipCount );
3325 * truncateHtml() helper function
3326 * (a) push or pop $tag from $openTags as needed
3327 * (b) clear $tag value
3328 * @param &$tag string Current HTML tag name we are looking at
3329 * @param $tagType int (0-open tag, 1-close tag)
3330 * @param $lastCh string Character before the '>' that ended this tag
3331 * @param &$openTags array Open tag stack (not accounting for $tag)
3333 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3334 $tag = ltrim( $tag );
3336 if ( $tagType == 0 && $lastCh != '/' ) {
3337 $openTags[] = $tag; // tag opened (didn't close itself)
3338 } elseif ( $tagType == 1 ) {
3339 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3340 array_pop( $openTags ); // tag closed
3348 * Grammatical transformations, needed for inflected languages
3349 * Invoked by putting {{grammar:case|word}} in a message
3351 * @param $word string
3352 * @param $case string
3355 function convertGrammar( $word, $case ) {
3356 global $wgGrammarForms;
3357 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3358 return $wgGrammarForms[$this->getCode()][$case][$word];
3363 * Get the grammar forms for the content language
3364 * @return array of grammar forms
3367 function getGrammarForms() {
3368 global $wgGrammarForms;
3369 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3370 return $wgGrammarForms[$this->getCode()];
3375 * Provides an alternative text depending on specified gender.
3376 * Usage {{gender:username|masculine|feminine|neutral}}.
3377 * username is optional, in which case the gender of current user is used,
3378 * but only in (some) interface messages; otherwise default gender is used.
3380 * If no forms are given, an empty string is returned. If only one form is
3381 * given, it will be returned unconditionally. These details are implied by
3382 * the caller and cannot be overridden in subclasses.
3384 * If more than one form is given, the default is to use the neutral one
3385 * if it is specified, and to use the masculine one otherwise. These
3386 * details can be overridden in subclasses.
3388 * @param $gender string
3389 * @param $forms array
3393 function gender( $gender, $forms ) {
3394 if ( !count( $forms ) ) {
3397 $forms = $this->preConvertPlural( $forms, 2 );
3398 if ( $gender === 'male' ) {
3401 if ( $gender === 'female' ) {
3404 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3408 * Plural form transformations, needed for some languages.
3409 * For example, there are 3 form of plural in Russian and Polish,
3410 * depending on "count mod 10". See [[w:Plural]]
3411 * For English it is pretty simple.
3413 * Invoked by putting {{plural:count|wordform1|wordform2}}
3414 * or {{plural:count|wordform1|wordform2|wordform3}}
3416 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3418 * @param $count Integer: non-localized number
3419 * @param $forms Array: different plural forms
3420 * @return string Correct form of plural for $count in this language
3422 function convertPlural( $count, $forms ) {
3423 if ( !count( $forms ) ) {
3426 $forms = $this->preConvertPlural( $forms, 2 );
3428 return ( $count == 1 ) ?
$forms[0] : $forms[1];
3432 * Checks that convertPlural was given an array and pads it to requested
3433 * amount of forms by copying the last one.
3435 * @param $count Integer: How many forms should there be at least
3436 * @param $forms Array of forms given to convertPlural
3437 * @return array Padded array of forms or an exception if not an array
3439 protected function preConvertPlural( /* Array */ $forms, $count ) {
3440 while ( count( $forms ) < $count ) {
3441 $forms[] = $forms[count( $forms ) - 1];
3447 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3448 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3449 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3450 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3453 * @param $str String: the validated block duration in English
3454 * @return string Somehow translated block duration
3455 * @see LanguageFi.php for example implementation
3457 function translateBlockExpiry( $str ) {
3458 $duration = SpecialBlock
::getSuggestedDurations( $this );
3459 foreach ( $duration as $show => $value ) {
3460 if ( strcmp( $str, $value ) == 0 ) {
3461 return htmlspecialchars( trim( $show ) );
3465 // Since usually only infinite or indefinite is only on list, so try
3466 // equivalents if still here.
3467 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3468 if ( in_array( $str, $indefs ) ) {
3469 foreach ( $indefs as $val ) {
3470 $show = array_search( $val, $duration, true );
3471 if ( $show !== false ) {
3472 return htmlspecialchars( trim( $show ) );
3476 // If all else fails, return the original string.
3481 * languages like Chinese need to be segmented in order for the diff
3484 * @param $text String
3487 public function segmentForDiff( $text ) {
3492 * and unsegment to show the result
3494 * @param $text String
3497 public function unsegmentForDiff( $text ) {
3502 * Return the LanguageConverter used in the Language
3505 * @return LanguageConverter
3507 public function getConverter() {
3508 return $this->mConverter
;
3512 * convert text to all supported variants
3514 * @param $text string
3517 public function autoConvertToAllVariants( $text ) {
3518 return $this->mConverter
->autoConvertToAllVariants( $text );
3522 * convert text to different variants of a language.
3524 * @param $text string
3527 public function convert( $text ) {
3528 return $this->mConverter
->convert( $text );
3532 * Convert a Title object to a string in the preferred variant
3534 * @param $title Title
3537 public function convertTitle( $title ) {
3538 return $this->mConverter
->convertTitle( $title );
3542 * Check if this is a language with variants
3546 public function hasVariants() {
3547 return sizeof( $this->getVariants() ) > 1;
3551 * Check if the language has the specific variant
3554 * @param $variant string
3557 public function hasVariant( $variant ) {
3558 return (bool)$this->mConverter
->validateVariant( $variant );
3562 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3564 * @param $text string
3567 public function armourMath( $text ) {
3568 return $this->mConverter
->armourMath( $text );
3572 * Perform output conversion on a string, and encode for safe HTML output.
3573 * @param $text String text to be converted
3574 * @param $isTitle Bool whether this conversion is for the article title
3576 * @todo this should get integrated somewhere sane
3578 public function convertHtml( $text, $isTitle = false ) {
3579 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3583 * @param $key string
3586 public function convertCategoryKey( $key ) {
3587 return $this->mConverter
->convertCategoryKey( $key );
3591 * Get the list of variants supported by this language
3592 * see sample implementation in LanguageZh.php
3594 * @return array an array of language codes
3596 public function getVariants() {
3597 return $this->mConverter
->getVariants();
3603 public function getPreferredVariant() {
3604 return $this->mConverter
->getPreferredVariant();
3610 public function getDefaultVariant() {
3611 return $this->mConverter
->getDefaultVariant();
3617 public function getURLVariant() {
3618 return $this->mConverter
->getURLVariant();
3622 * If a language supports multiple variants, it is
3623 * possible that non-existing link in one variant
3624 * actually exists in another variant. this function
3625 * tries to find it. See e.g. LanguageZh.php
3627 * @param $link String: the name of the link
3628 * @param $nt Mixed: the title object of the link
3629 * @param $ignoreOtherCond Boolean: to disable other conditions when
3630 * we need to transclude a template or update a category's link
3631 * @return null the input parameters may be modified upon return
3633 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3634 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
3638 * If a language supports multiple variants, converts text
3639 * into an array of all possible variants of the text:
3640 * 'variant' => text in that variant
3642 * @deprecated since 1.17 Use autoConvertToAllVariants()
3644 * @param $text string
3648 public function convertLinkToAllVariants( $text ) {
3649 return $this->mConverter
->convertLinkToAllVariants( $text );
3653 * returns language specific options used by User::getPageRenderHash()
3654 * for example, the preferred language variant
3658 function getExtraHashOptions() {
3659 return $this->mConverter
->getExtraHashOptions();
3663 * For languages that support multiple variants, the title of an
3664 * article may be displayed differently in different variants. this
3665 * function returns the apporiate title defined in the body of the article.
3669 public function getParsedTitle() {
3670 return $this->mConverter
->getParsedTitle();
3674 * Enclose a string with the "no conversion" tag. This is used by
3675 * various functions in the Parser
3677 * @param $text String: text to be tagged for no conversion
3678 * @param $noParse bool
3679 * @return string the tagged text
3681 public function markNoConversion( $text, $noParse = false ) {
3682 return $this->mConverter
->markNoConversion( $text, $noParse );
3686 * A regular expression to match legal word-trailing characters
3687 * which should be merged onto a link of the form [[foo]]bar.
3691 public function linkTrail() {
3692 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
3698 function getLangObj() {
3703 * Get the RFC 3066 code for this language object
3705 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3706 * htmlspecialchars() or similar
3710 public function getCode() {
3711 return $this->mCode
;
3715 * Get the code in Bcp47 format which we can use
3716 * inside of html lang="" tags.
3718 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3719 * htmlspecialchars() or similar.
3724 public function getHtmlCode() {
3725 if ( is_null( $this->mHtmlCode
) ) {
3726 $this->mHtmlCode
= wfBCP47( $this->getCode() );
3728 return $this->mHtmlCode
;
3732 * @param $code string
3734 public function setCode( $code ) {
3735 $this->mCode
= $code;
3736 // Ensure we don't leave an incorrect html code lying around
3737 $this->mHtmlCode
= null;
3741 * Get the name of a file for a certain language code
3742 * @param $prefix string Prepend this to the filename
3743 * @param $code string Language code
3744 * @param $suffix string Append this to the filename
3745 * @throws MWException
3746 * @return string $prefix . $mangledCode . $suffix
3748 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
3749 // Protect against path traversal
3750 if ( !Language
::isValidCode( $code )
3751 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
3753 throw new MWException( "Invalid language code \"$code\"" );
3756 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
3760 * Get the language code from a file name. Inverse of getFileName()
3761 * @param $filename string $prefix . $languageCode . $suffix
3762 * @param $prefix string Prefix before the language code
3763 * @param $suffix string Suffix after the language code
3764 * @return string Language code, or false if $prefix or $suffix isn't found
3766 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
3768 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
3769 preg_quote( $suffix, '/' ) . '/', $filename, $m );
3770 if ( !count( $m ) ) {
3773 return str_replace( '_', '-', strtolower( $m[1] ) );
3777 * @param $code string
3780 public static function getMessagesFileName( $code ) {
3782 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
3783 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
3788 * @param $code string
3791 public static function getClassFileName( $code ) {
3793 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
3797 * Get the first fallback for a given language.
3799 * @param $code string
3801 * @return bool|string
3803 public static function getFallbackFor( $code ) {
3804 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
3807 $fallbacks = self
::getFallbacksFor( $code );
3808 $first = array_shift( $fallbacks );
3814 * Get the ordered list of fallback languages.
3817 * @param $code string Language code
3820 public static function getFallbacksFor( $code ) {
3821 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
3824 $v = self
::getLocalisationCache()->getItem( $code, 'fallback' );
3825 $v = array_map( 'trim', explode( ',', $v ) );
3826 if ( $v[count( $v ) - 1] !== 'en' ) {
3834 * Get all messages for a given language
3835 * WARNING: this may take a long time. If you just need all message *keys*
3836 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
3838 * @param $code string
3842 public static function getMessagesFor( $code ) {
3843 return self
::getLocalisationCache()->getItem( $code, 'messages' );
3847 * Get a message for a given language
3849 * @param $key string
3850 * @param $code string
3854 public static function getMessageFor( $key, $code ) {
3855 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
3859 * Get all message keys for a given language. This is a faster alternative to
3860 * array_keys( Language::getMessagesFor( $code ) )
3863 * @param $code string Language code
3864 * @return array of message keys (strings)
3866 public static function getMessageKeysFor( $code ) {
3867 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
3874 function fixVariableInNamespace( $talk ) {
3875 if ( strpos( $talk, '$1' ) === false ) {
3879 global $wgMetaNamespace;
3880 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
3882 # Allow grammar transformations
3883 # Allowing full message-style parsing would make simple requests
3884 # such as action=raw much more expensive than they need to be.
3885 # This will hopefully cover most cases.
3886 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
3887 array( &$this, 'replaceGrammarInNamespace' ), $talk );
3888 return str_replace( ' ', '_', $talk );
3895 function replaceGrammarInNamespace( $m ) {
3896 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
3900 * @throws MWException
3903 static function getCaseMaps() {
3904 static $wikiUpperChars, $wikiLowerChars;
3905 if ( isset( $wikiUpperChars ) ) {
3906 return array( $wikiUpperChars, $wikiLowerChars );
3909 wfProfileIn( __METHOD__
);
3910 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
3911 if ( $arr === false ) {
3912 throw new MWException(
3913 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
3915 $wikiUpperChars = $arr['wikiUpperChars'];
3916 $wikiLowerChars = $arr['wikiLowerChars'];
3917 wfProfileOut( __METHOD__
);
3918 return array( $wikiUpperChars, $wikiLowerChars );
3922 * Decode an expiry (block, protection, etc) which has come from the DB
3924 * @todo FIXME: why are we returnings DBMS-dependent strings???
3926 * @param $expiry String: Database expiry String
3927 * @param $format Bool|Int true to process using language functions, or TS_ constant
3928 * to return the expiry in a given timestamp
3931 public function formatExpiry( $expiry, $format = true ) {
3932 static $infinity, $infinityMsg;
3933 if ( $infinity === null ) {
3934 $infinityMsg = wfMessage( 'infiniteblock' );
3935 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
3938 if ( $expiry == '' ||
$expiry == $infinity ) {
3939 return $format === true
3943 return $format === true
3944 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
3945 : wfTimestamp( $format, $expiry );
3951 * @param $seconds int|float
3952 * @param $format Array Optional
3953 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
3954 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
3955 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
3956 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
3959 function formatTimePeriod( $seconds, $format = array() ) {
3960 if ( !is_array( $format ) ) {
3961 $format = array( 'avoid' => $format ); // For backwards compatibility
3963 if ( !isset( $format['avoid'] ) ) {
3964 $format['avoid'] = false;
3966 if ( !isset( $format['noabbrevs' ] ) ) {
3967 $format['noabbrevs'] = false;
3969 $secondsMsg = wfMessage(
3970 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
3971 $minutesMsg = wfMessage(
3972 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
3973 $hoursMsg = wfMessage(
3974 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
3975 $daysMsg = wfMessage(
3976 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
3978 if ( round( $seconds * 10 ) < 100 ) {
3979 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
3980 $s = $secondsMsg->params( $s )->text();
3981 } elseif ( round( $seconds ) < 60 ) {
3982 $s = $this->formatNum( round( $seconds ) );
3983 $s = $secondsMsg->params( $s )->text();
3984 } elseif ( round( $seconds ) < 3600 ) {
3985 $minutes = floor( $seconds / 60 );
3986 $secondsPart = round( fmod( $seconds, 60 ) );
3987 if ( $secondsPart == 60 ) {
3991 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3993 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3994 } elseif ( round( $seconds ) <= 2 * 86400 ) {
3995 $hours = floor( $seconds / 3600 );
3996 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
3997 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
3998 if ( $secondsPart == 60 ) {
4002 if ( $minutes == 60 ) {
4006 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4008 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4009 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4010 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4013 $days = floor( $seconds / 86400 );
4014 if ( $format['avoid'] === 'avoidminutes' ) {
4015 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4016 if ( $hours == 24 ) {
4020 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4022 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4023 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4024 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4025 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4026 if ( $minutes == 60 ) {
4030 if ( $hours == 24 ) {
4034 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4036 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4038 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4040 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4042 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4049 * Format a bitrate for output, using an appropriate
4050 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4052 * This use base 1000. For base 1024 use formatSize(), for another base
4053 * see formatComputingNumbers()
4058 function formatBitrate( $bps ) {
4059 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4063 * @param $size int Size of the unit
4064 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4065 * @param $messageKey string Message key to be uesd
4068 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4070 return str_replace( '$1', $this->formatNum( $size ),
4071 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4074 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4077 $maxIndex = count( $sizes ) - 1;
4078 while ( $size >= $boundary && $index < $maxIndex ) {
4083 // For small sizes no decimal places necessary
4086 // For MB and bigger two decimal places are smarter
4089 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4091 $size = round( $size, $round );
4092 $text = $this->getMessageFromDB( $msg );
4093 return str_replace( '$1', $this->formatNum( $size ), $text );
4097 * Format a size in bytes for output, using an appropriate
4098 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4100 * This method use base 1024. For base 1000 use formatBitrate(), for
4101 * another base see formatComputingNumbers()
4103 * @param $size int Size to format
4104 * @return string Plain text (not HTML)
4106 function formatSize( $size ) {
4107 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4111 * Make a list item, used by various special pages
4113 * @param $page String Page link
4114 * @param $details String Text between brackets
4115 * @param $oppositedm Boolean Add the direction mark opposite to your
4116 * language, to display text properly
4119 function specialList( $page, $details, $oppositedm = true ) {
4120 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) .
4121 $this->getDirMark();
4122 $details = $details ?
$dirmark . $this->getMessageFromDB( 'word-separator' ) .
4123 wfMsgExt( 'parentheses', array( 'escape', 'replaceafter', 'language' => $this ), $details ) : '';
4124 return $page . $details;
4128 * Generate (prev x| next x) (20|50|100...) type links for paging
4130 * @param $title Title object to link
4131 * @param $offset Integer offset parameter
4132 * @param $limit Integer limit parameter
4133 * @param $query array|String optional URL query parameter string
4134 * @param $atend Bool optional param for specified if this is the last page
4137 public function viewPrevNext( Title
$title, $offset, $limit, array $query = array(), $atend = false ) {
4138 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4140 # Make 'previous' link
4141 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4142 if ( $offset > 0 ) {
4143 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4144 $query, $prev, 'prevn-title', 'mw-prevlink' );
4146 $plink = htmlspecialchars( $prev );
4150 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4152 $nlink = htmlspecialchars( $next );
4154 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4155 $query, $next, 'prevn-title', 'mw-nextlink' );
4158 # Make links to set number of items per page
4159 $numLinks = array();
4160 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4161 $numLinks[] = $this->numLink( $title, $offset, $num,
4162 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4165 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4166 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4170 * Helper function for viewPrevNext() that generates links
4172 * @param $title Title object to link
4173 * @param $offset Integer offset parameter
4174 * @param $limit Integer limit parameter
4175 * @param $query Array extra query parameters
4176 * @param $link String text to use for the link; will be escaped
4177 * @param $tooltipMsg String name of the message to use as tooltip
4178 * @param $class String value of the "class" attribute of the link
4179 * @return String HTML fragment
4181 private function numLink( Title
$title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4182 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4183 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4184 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4185 'title' => $tooltip, 'class' => $class ), $link );
4189 * Get the conversion rule title, if any.
4193 public function getConvRuleTitle() {
4194 return $this->mConverter
->getConvRuleTitle();