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 __DIR__
. '/Names.php';
37 if ( function_exists( 'mb_strtoupper' ) ) {
38 mb_internal_encoding( 'UTF-8' );
42 * a fake language converter
51 function __construct( $langobj ) { $this->mLang
= $langobj; }
52 function autoConvertToAllVariants( $text ) { return array( $this->mLang
->getCode() => $text ); }
53 function convert( $t ) { return $t; }
54 function convertTo( $text, $variant ) { return $text; }
55 function convertTitle( $t ) { return $t->getPrefixedText(); }
56 function convertNamespace( $ns ) { return $this->mLang
->getFormattedNsText( $ns ); }
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 public $mVariants, $mCode, $mLoaded = false;
83 public $mMagicExtensions = array(), $mMagicHookDone = false;
84 private $mHtmlCode = null, $mParentLanguage = false;
86 public $dateFormatStrings = array();
87 public $mExtendedSpecialPageAliases;
89 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
92 * ReplacementArray object caches
94 public $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' => 31556952000,
163 'centuries' => 3155695200,
164 'decades' => 315569520,
165 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
174 * Cache for language fallbacks.
175 * @see Language::getFallbacksIncludingSiteLanguage
179 static private $fallbackLanguageCache = array();
182 * Get a cached or new language object for a given language code
183 * @param $code String
186 static function factory( $code ) {
187 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
189 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
190 $code = $wgDummyLanguageCodes[$code];
193 // get the language object to process
194 $langObj = isset( self
::$mLangObjCache[$code] )
195 ? self
::$mLangObjCache[$code]
196 : self
::newFromCode( $code );
198 // merge the language object in to get it up front in the cache
199 self
::$mLangObjCache = array_merge( array( $code => $langObj ), self
::$mLangObjCache );
200 // get rid of the oldest ones in case we have an overflow
201 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
207 * Create a language object for a given language code
208 * @param $code String
209 * @throws MWException
212 protected static function newFromCode( $code ) {
213 // Protect against path traversal below
214 if ( !Language
::isValidCode( $code )
215 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
217 throw new MWException( "Invalid language code \"$code\"" );
220 if ( !Language
::isValidBuiltInCode( $code ) ) {
221 // It's not possible to customise this code with class files, so
222 // just return a Language object. This is to support uselang= hacks.
223 $lang = new Language
;
224 $lang->setCode( $code );
228 // Check if there is a language class for the code
229 $class = self
::classFromCode( $code );
230 self
::preloadLanguageClass( $class );
231 if ( MWInit
::classExists( $class ) ) {
236 // Keep trying the fallback list until we find an existing class
237 $fallbacks = Language
::getFallbacksFor( $code );
238 foreach ( $fallbacks as $fallbackCode ) {
239 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
240 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
243 $class = self
::classFromCode( $fallbackCode );
244 self
::preloadLanguageClass( $class );
245 if ( MWInit
::classExists( $class ) ) {
246 $lang = Language
::newFromCode( $fallbackCode );
247 $lang->setCode( $code );
252 throw new MWException( "Invalid fallback sequence for language '$code'" );
256 * Checks whether any localisation is available for that language tag
257 * in MediaWiki (MessagesXx.php exists).
259 * @param string $code Language tag (in lower case)
260 * @return bool Whether language is supported
263 public static function isSupportedLanguage( $code ) {
264 return $code === strtolower( $code ) && is_readable( self
::getMessagesFileName( $code ) );
268 * Returns true if a language code string is a well-formed language tag
269 * according to RFC 5646.
270 * This function only checks well-formedness; it doesn't check that
271 * language, script or variant codes actually exist in the repositories.
273 * Based on regexes by Mark Davis of the Unicode Consortium:
274 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
276 * @param $code string
277 * @param $lenient boolean Whether to allow '_' as separator. The default is only '-'.
282 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
285 $alphanum = '[a-z0-9]';
286 $x = 'x'; # private use singleton
287 $singleton = '[a-wy-z]'; # other singleton
288 $s = $lenient ?
'[-_]' : '-';
290 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
291 $script = "$alpha{4}"; # ISO 15924
292 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
293 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
294 $extension = "$singleton(?:$s$alphanum{2,8})+";
295 $privateUse = "$x(?:$s$alphanum{1,8})+";
297 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
298 # Since these are limited, this is safe even later changes to the registry --
299 # the only oddity is that it might change the type of the tag, and thus
300 # the results from the capturing groups.
301 # http://www.iana.org/assignments/language-subtag-registry
303 $grandfathered = "en{$s}GB{$s}oed"
304 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
305 . "|no{$s}(?:bok|nyn)"
306 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
307 . "|zh{$s}min{$s}nan";
309 $variantList = "$variant(?:$s$variant)*";
310 $extensionList = "$extension(?:$s$extension)*";
312 $langtag = "(?:($language)"
315 . "(?:$s$variantList)?"
316 . "(?:$s$extensionList)?"
317 . "(?:$s$privateUse)?)";
319 # The final breakdown, with capturing groups for each of these components
320 # The variants, extensions, grandfathered, and private-use may have interior '-'
322 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
324 return (bool)preg_match( "/$root/", strtolower( $code ) );
328 * Returns true if a language code string is of a valid form, whether or
329 * not it exists. This includes codes which are used solely for
330 * customisation via the MediaWiki namespace.
332 * @param $code string
336 public static function isValidCode( $code ) {
337 static $cache = array();
338 if ( isset( $cache[$code] ) ) {
339 return $cache[$code];
341 // People think language codes are html safe, so enforce it.
342 // Ideally we should only allow a-zA-Z0-9-
343 // but, .+ and other chars are often used for {{int:}} hacks
344 // see bugs 37564, 37587, 36938
346 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
347 && !preg_match( Title
::getTitleInvalidRegex(), $code );
349 return $cache[$code];
353 * Returns true if a language code is of a valid form for the purposes of
354 * internal customisation of MediaWiki, via Messages*.php.
356 * @param $code string
358 * @throws MWException
362 public static function isValidBuiltInCode( $code ) {
364 if ( !is_string( $code ) ) {
365 if ( is_object( $code ) ) {
366 $addmsg = " of class " . get_class( $code );
370 $type = gettype( $code );
371 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
374 return (bool)preg_match( '/^[a-z0-9-]{2,}$/i', $code );
378 * Returns true if a language code is an IETF tag known to MediaWiki.
380 * @param $code string
385 public static function isKnownLanguageTag( $tag ) {
386 static $coreLanguageNames;
388 // Quick escape for invalid input to avoid exceptions down the line
389 // when code tries to process tags which are not valid at all.
390 if ( !self
::isValidBuiltInCode( $tag ) ) {
394 if ( $coreLanguageNames === null ) {
395 include MWInit
::compiledPath( 'languages/Names.php' );
398 if ( isset( $coreLanguageNames[$tag] )
399 || self
::fetchLanguageName( $tag, $tag ) !== ''
409 * @return String Name of the language class
411 public static function classFromCode( $code ) {
412 if ( $code == 'en' ) {
415 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
420 * Includes language class files
422 * @param $class string Name of the language class
424 public static function preloadLanguageClass( $class ) {
427 if ( $class === 'Language' ) {
431 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
432 include_once "$IP/languages/classes/$class.php";
437 * Get the LocalisationCache instance
439 * @return LocalisationCache
441 public static function getLocalisationCache() {
442 if ( is_null( self
::$dataCache ) ) {
443 global $wgLocalisationCacheConf;
444 $class = $wgLocalisationCacheConf['class'];
445 self
::$dataCache = new $class( $wgLocalisationCacheConf );
447 return self
::$dataCache;
450 function __construct() {
451 $this->mConverter
= new FakeConverter( $this );
452 // Set the code to the name of the descendant
453 if ( get_class( $this ) == 'Language' ) {
456 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
458 self
::getLocalisationCache();
462 * Reduce memory usage
464 function __destruct() {
465 foreach ( $this as $name => $value ) {
466 unset( $this->$name );
471 * Hook which will be called if this is the content language.
472 * Descendants can use this to register hook functions or modify globals
474 function initContLang() { }
477 * Same as getFallbacksFor for current language.
479 * @deprecated in 1.19
481 function getFallbackLanguageCode() {
482 wfDeprecated( __METHOD__
, '1.19' );
483 return self
::getFallbackFor( $this->mCode
);
490 function getFallbackLanguages() {
491 return self
::getFallbacksFor( $this->mCode
);
495 * Exports $wgBookstoreListEn
498 function getBookstoreList() {
499 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
503 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
504 * available in localised form, it will be included in English.
508 public function getNamespaces() {
509 if ( is_null( $this->namespaceNames
) ) {
510 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
512 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
513 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
515 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
517 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
518 if ( $wgMetaNamespaceTalk ) {
519 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
521 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
522 $this->namespaceNames
[NS_PROJECT_TALK
] =
523 $this->fixVariableInNamespace( $talk );
526 # Sometimes a language will be localised but not actually exist on this wiki.
527 foreach ( $this->namespaceNames
as $key => $text ) {
528 if ( !isset( $validNamespaces[$key] ) ) {
529 unset( $this->namespaceNames
[$key] );
533 # The above mixing may leave namespaces out of canonical order.
534 # Re-order by namespace ID number...
535 ksort( $this->namespaceNames
);
537 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
539 return $this->namespaceNames
;
543 * Arbitrarily set all of the namespace names at once. Mainly used for testing
544 * @param $namespaces Array of namespaces (id => name)
546 public function setNamespaces( array $namespaces ) {
547 $this->namespaceNames
= $namespaces;
548 $this->mNamespaceIds
= null;
552 * Resets all of the namespace caches. Mainly used for testing
554 public function resetNamespaces() {
555 $this->namespaceNames
= null;
556 $this->mNamespaceIds
= null;
557 $this->namespaceAliases
= null;
561 * A convenience function that returns the same thing as
562 * getNamespaces() except with the array values changed to ' '
563 * where it found '_', useful for producing output to be displayed
564 * e.g. in <select> forms.
568 function getFormattedNamespaces() {
569 $ns = $this->getNamespaces();
570 foreach ( $ns as $k => $v ) {
571 $ns[$k] = strtr( $v, '_', ' ' );
577 * Get a namespace value by key
579 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
580 * echo $mw_ns; // prints 'MediaWiki'
583 * @param $index Int: the array key of the namespace to return
584 * @return mixed, string if the namespace value exists, otherwise false
586 function getNsText( $index ) {
587 $ns = $this->getNamespaces();
588 return isset( $ns[$index] ) ?
$ns[$index] : false;
592 * A convenience function that returns the same thing as
593 * getNsText() except with '_' changed to ' ', useful for
597 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
598 * echo $mw_ns; // prints 'MediaWiki talk'
601 * @param int $index The array key of the namespace to return
602 * @return string Namespace name without underscores (empty string if namespace does not exist)
604 function getFormattedNsText( $index ) {
605 $ns = $this->getNsText( $index );
606 return strtr( $ns, '_', ' ' );
610 * Returns gender-dependent namespace alias if available.
611 * @param $index Int: namespace index
612 * @param $gender String: gender key (male, female... )
616 function getGenderNsText( $index, $gender ) {
617 global $wgExtraGenderNamespaces;
619 $ns = $wgExtraGenderNamespaces + self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
620 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
624 * Whether this language makes distinguishes genders for example in
629 function needsGenderDistinction() {
630 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
631 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
632 // $wgExtraGenderNamespaces overrides everything
634 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
635 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
636 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
639 // Check what is in i18n files
640 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
641 return count( $aliases ) > 0;
646 * Get a namespace key by value, case insensitive.
647 * Only matches namespace names for the current language, not the
648 * canonical ones defined in Namespace.php.
650 * @param $text String
651 * @return mixed An integer if $text is a valid value otherwise false
653 function getLocalNsIndex( $text ) {
654 $lctext = $this->lc( $text );
655 $ids = $this->getNamespaceIds();
656 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
662 function getNamespaceAliases() {
663 if ( is_null( $this->namespaceAliases
) ) {
664 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
668 foreach ( $aliases as $name => $index ) {
669 if ( $index === NS_PROJECT_TALK
) {
670 unset( $aliases[$name] );
671 $name = $this->fixVariableInNamespace( $name );
672 $aliases[$name] = $index;
677 global $wgExtraGenderNamespaces;
678 $genders = $wgExtraGenderNamespaces +
(array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
679 foreach ( $genders as $index => $forms ) {
680 foreach ( $forms as $alias ) {
681 $aliases[$alias] = $index;
685 $this->namespaceAliases
= $aliases;
687 return $this->namespaceAliases
;
693 function getNamespaceIds() {
694 if ( is_null( $this->mNamespaceIds
) ) {
695 global $wgNamespaceAliases;
696 # Put namespace names and aliases into a hashtable.
697 # If this is too slow, then we should arrange it so that it is done
698 # before caching. The catch is that at pre-cache time, the above
699 # class-specific fixup hasn't been done.
700 $this->mNamespaceIds
= array();
701 foreach ( $this->getNamespaces() as $index => $name ) {
702 $this->mNamespaceIds
[$this->lc( $name )] = $index;
704 foreach ( $this->getNamespaceAliases() as $name => $index ) {
705 $this->mNamespaceIds
[$this->lc( $name )] = $index;
707 if ( $wgNamespaceAliases ) {
708 foreach ( $wgNamespaceAliases as $name => $index ) {
709 $this->mNamespaceIds
[$this->lc( $name )] = $index;
713 return $this->mNamespaceIds
;
717 * Get a namespace key by value, case insensitive. Canonical namespace
718 * names override custom ones defined for the current language.
720 * @param $text String
721 * @return mixed An integer if $text is a valid value otherwise false
723 function getNsIndex( $text ) {
724 $lctext = $this->lc( $text );
725 $ns = MWNamespace
::getCanonicalIndex( $lctext );
726 if ( $ns !== null ) {
729 $ids = $this->getNamespaceIds();
730 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
734 * short names for language variants used for language conversion links.
736 * @param $code String
737 * @param $usemsg bool Use the "variantname-xyz" message if it exists
740 function getVariantname( $code, $usemsg = true ) {
741 $msg = "variantname-$code";
742 if ( $usemsg && wfMessage( $msg )->exists() ) {
743 return $this->getMessageFromDB( $msg );
745 $name = self
::fetchLanguageName( $code );
747 return $name; # if it's defined as a language name, show that
749 # otherwise, output the language code
755 * @param $name string
758 function specialPage( $name ) {
759 $aliases = $this->getSpecialPageAliases();
760 if ( isset( $aliases[$name][0] ) ) {
761 $name = $aliases[$name][0];
763 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
769 function getDatePreferences() {
770 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
776 function getDateFormats() {
777 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
781 * @return array|string
783 function getDefaultDateFormat() {
784 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
785 if ( $df === 'dmy or mdy' ) {
786 global $wgAmericanDates;
787 return $wgAmericanDates ?
'mdy' : 'dmy';
796 function getDatePreferenceMigrationMap() {
797 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
804 function getImageFile( $image ) {
805 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
811 function getExtraUserToggles() {
812 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
819 function getUserToggle( $tog ) {
820 return $this->getMessageFromDB( "tog-$tog" );
824 * Get native language names, indexed by code.
825 * Only those defined in MediaWiki, no other data like CLDR.
826 * If $customisedOnly is true, only returns codes with a messages file
828 * @param $customisedOnly bool
831 * @deprecated in 1.20, use fetchLanguageNames()
833 public static function getLanguageNames( $customisedOnly = false ) {
834 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
838 * Get translated language names. This is done on best effort and
839 * by default this is exactly the same as Language::getLanguageNames.
840 * The CLDR extension provides translated names.
841 * @param $code String Language code.
842 * @return Array language code => language name
844 * @deprecated in 1.20, use fetchLanguageNames()
846 public static function getTranslatedLanguageNames( $code ) {
847 return self
::fetchLanguageNames( $code, 'all' );
851 * Get an array of language names, indexed by code.
852 * @param $inLanguage null|string: Code of language in which to return the names
853 * Use null for autonyms (native names)
854 * @param $include string:
855 * 'all' all available languages
856 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
857 * 'mwfile' only if the language is in 'mw' *and* has a message file
858 * @return array: language code => language name
861 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
862 global $wgExtraLanguageNames;
863 static $coreLanguageNames;
865 if ( $coreLanguageNames === null ) {
866 include MWInit
::compiledPath( 'languages/Names.php' );
872 # TODO: also include when $inLanguage is null, when this code is more efficient
873 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
876 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
877 foreach ( $mwNames as $mwCode => $mwName ) {
878 # - Prefer own MediaWiki native name when not using the hook
879 # - For other names just add if not added through the hook
880 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
881 $names[$mwCode] = $mwName;
885 if ( $include === 'all' ) {
890 $coreCodes = array_keys( $mwNames );
891 foreach ( $coreCodes as $coreCode ) {
892 $returnMw[$coreCode] = $names[$coreCode];
895 if ( $include === 'mwfile' ) {
896 $namesMwFile = array();
897 # We do this using a foreach over the codes instead of a directory
898 # loop so that messages files in extensions will work correctly.
899 foreach ( $returnMw as $code => $value ) {
900 if ( is_readable( self
::getMessagesFileName( $code ) ) ) {
901 $namesMwFile[$code] = $names[$code];
906 # 'mw' option; default if it's not one of the other two options (all/mwfile)
911 * @param $code string: The code of the language for which to get the name
912 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
913 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
914 * @return string: Language name or empty
917 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
918 $array = self
::fetchLanguageNames( $inLanguage, $include );
919 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
923 * Get a message from the MediaWiki namespace.
925 * @param $msg String: message name
928 function getMessageFromDB( $msg ) {
929 return wfMessage( $msg )->inLanguage( $this )->text();
933 * Get the native language name of $code.
934 * Only if defined in MediaWiki, no other data like CLDR.
935 * @param $code string
937 * @deprecated in 1.20, use fetchLanguageName()
939 function getLanguageName( $code ) {
940 return self
::fetchLanguageName( $code );
947 function getMonthName( $key ) {
948 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
954 function getMonthNamesArray() {
955 $monthNames = array( '' );
956 for ( $i = 1; $i < 13; $i++
) {
957 $monthNames[] = $this->getMonthName( $i );
966 function getMonthNameGen( $key ) {
967 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
974 function getMonthAbbreviation( $key ) {
975 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
981 function getMonthAbbreviationsArray() {
982 $monthNames = array( '' );
983 for ( $i = 1; $i < 13; $i++
) {
984 $monthNames[] = $this->getMonthAbbreviation( $i );
993 function getWeekdayName( $key ) {
994 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1001 function getWeekdayAbbreviation( $key ) {
1002 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1006 * @param $key string
1009 function getIranianCalendarMonthName( $key ) {
1010 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1014 * @param $key string
1017 function getHebrewCalendarMonthName( $key ) {
1018 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1022 * @param $key string
1025 function getHebrewCalendarMonthNameGen( $key ) {
1026 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1030 * @param $key string
1033 function getHijriCalendarMonthName( $key ) {
1034 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1038 * This is a workalike of PHP's date() function, but with better
1039 * internationalisation, a reduced set of format characters, and a better
1042 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1043 * the PHP manual for definitions. There are a number of extensions, which
1046 * xn Do not translate digits of the next numeric format character
1047 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1048 * xr Use roman numerals for the next numeric format character
1049 * xh Use hebrew numerals for the next numeric format character
1051 * xg Genitive month name
1053 * xij j (day number) in Iranian calendar
1054 * xiF F (month name) in Iranian calendar
1055 * xin n (month number) in Iranian calendar
1056 * xiy y (two digit year) in Iranian calendar
1057 * xiY Y (full year) in Iranian calendar
1059 * xjj j (day number) in Hebrew calendar
1060 * xjF F (month name) in Hebrew calendar
1061 * xjt t (days in month) in Hebrew calendar
1062 * xjx xg (genitive month name) in Hebrew calendar
1063 * xjn n (month number) in Hebrew calendar
1064 * xjY Y (full year) in Hebrew calendar
1066 * xmj j (day number) in Hijri calendar
1067 * xmF F (month name) in Hijri calendar
1068 * xmn n (month number) in Hijri calendar
1069 * xmY Y (full year) in Hijri calendar
1071 * xkY Y (full year) in Thai solar calendar. Months and days are
1072 * identical to the Gregorian calendar
1073 * xoY Y (full year) in Minguo calendar or Juche year.
1074 * Months and days are identical to the
1075 * Gregorian calendar
1076 * xtY Y (full year) in Japanese nengo. Months and days are
1077 * identical to the Gregorian calendar
1079 * Characters enclosed in double quotes will be considered literal (with
1080 * the quotes themselves removed). Unmatched quotes will be considered
1081 * literal quotes. Example:
1083 * "The month is" F => The month is January
1086 * Backslash escaping is also supported.
1088 * Input timestamp is assumed to be pre-normalized to the desired local
1089 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1090 * $ts is UTC if $zone is not given.
1092 * @param $format String
1093 * @param $ts String: 14-character timestamp
1096 * @param $zone DateTimeZone: Timezone of $ts
1097 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1099 * @throws MWException
1102 function sprintfDate( $format, $ts, DateTimeZone
$zone = null ) {
1107 $dateTimeObj = false;
1116 if ( strlen( $ts ) !== 14 ) {
1117 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1120 if ( !ctype_digit( $ts ) ) {
1121 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1124 for ( $p = 0; $p < strlen( $format ); $p++
) {
1126 $code = $format[$p];
1127 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
1128 $code .= $format[++
$p];
1131 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
1132 $code .= $format[++
$p];
1143 $rawToggle = !$rawToggle;
1152 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1156 $hebrew = self
::tsToHebrew( $ts );
1158 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1161 $num = substr( $ts, 6, 2 );
1164 if ( !$dateTimeObj ) {
1165 $dateTimeObj = DateTime
::createFromFormat(
1166 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1169 $s .= $this->getWeekdayAbbreviation( $dateTimeObj->format( 'w' ) +
1 );
1172 $num = intval( substr( $ts, 6, 2 ) );
1176 $iranian = self
::tsToIranian( $ts );
1182 $hijri = self
::tsToHijri( $ts );
1188 $hebrew = self
::tsToHebrew( $ts );
1193 if ( !$dateTimeObj ) {
1194 $dateTimeObj = DateTime
::createFromFormat(
1195 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1198 $s .= $this->getWeekdayName( $dateTimeObj->format( 'w' ) +
1 );
1201 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1205 $iranian = self
::tsToIranian( $ts );
1207 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1211 $hijri = self
::tsToHijri( $ts );
1213 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1217 $hebrew = self
::tsToHebrew( $ts );
1219 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1222 $num = substr( $ts, 4, 2 );
1225 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1228 $num = intval( substr( $ts, 4, 2 ) );
1232 $iranian = self
::tsToIranian( $ts );
1238 $hijri = self
::tsToHijri ( $ts );
1244 $hebrew = self
::tsToHebrew( $ts );
1250 $hebrew = self
::tsToHebrew( $ts );
1255 $num = substr( $ts, 0, 4 );
1259 $iranian = self
::tsToIranian( $ts );
1265 $hijri = self
::tsToHijri( $ts );
1271 $hebrew = self
::tsToHebrew( $ts );
1277 $thai = self
::tsToYear( $ts, 'thai' );
1283 $minguo = self
::tsToYear( $ts, 'minguo' );
1289 $tenno = self
::tsToYear( $ts, 'tenno' );
1294 $num = substr( $ts, 2, 2 );
1298 $iranian = self
::tsToIranian( $ts );
1300 $num = substr( $iranian[0], -2 );
1303 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1306 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1309 $h = substr( $ts, 8, 2 );
1310 $num = $h %
12 ?
$h %
12 : 12;
1313 $num = intval( substr( $ts, 8, 2 ) );
1316 $h = substr( $ts, 8, 2 );
1317 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1320 $num = substr( $ts, 8, 2 );
1323 $num = substr( $ts, 10, 2 );
1326 $num = substr( $ts, 12, 2 );
1334 // Pass through string from $dateTimeObj->format()
1335 if ( !$dateTimeObj ) {
1336 $dateTimeObj = DateTime
::createFromFormat(
1337 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1340 $s .= $dateTimeObj->format( $code );
1352 // Pass through number from $dateTimeObj->format()
1353 if ( !$dateTimeObj ) {
1354 $dateTimeObj = DateTime
::createFromFormat(
1355 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1358 $num = $dateTimeObj->format( $code );
1361 # Backslash escaping
1362 if ( $p < strlen( $format ) - 1 ) {
1363 $s .= $format[++
$p];
1370 if ( $p < strlen( $format ) - 1 ) {
1371 $endQuote = strpos( $format, '"', $p +
1 );
1372 if ( $endQuote === false ) {
1373 # No terminating quote, assume literal "
1376 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1380 # Quote at end of string, assume literal "
1387 if ( $num !== false ) {
1388 if ( $rawToggle ||
$raw ) {
1391 } elseif ( $roman ) {
1392 $s .= Language
::romanNumeral( $num );
1394 } elseif ( $hebrewNum ) {
1395 $s .= self
::hebrewNumeral( $num );
1398 $s .= $this->formatNum( $num, true );
1405 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1406 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1409 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1410 * Gregorian dates to Iranian dates. Originally written in C, it
1411 * is released under the terms of GNU Lesser General Public
1412 * License. Conversion to PHP was performed by Niklas Laxström.
1414 * Link: http://www.farsiweb.info/jalali/jalali.c
1420 private static function tsToIranian( $ts ) {
1421 $gy = substr( $ts, 0, 4 ) -1600;
1422 $gm = substr( $ts, 4, 2 ) -1;
1423 $gd = substr( $ts, 6, 2 ) -1;
1425 # Days passed from the beginning (including leap years)
1427 +
floor( ( $gy +
3 ) / 4 )
1428 - floor( ( $gy +
99 ) / 100 )
1429 +
floor( ( $gy +
399 ) / 400 );
1431 // Add days of the past months of this year
1432 for ( $i = 0; $i < $gm; $i++
) {
1433 $gDayNo +
= self
::$GREG_DAYS[$i];
1437 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1441 // Days passed in current month
1442 $gDayNo +
= (int)$gd;
1444 $jDayNo = $gDayNo - 79;
1446 $jNp = floor( $jDayNo / 12053 );
1449 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1452 if ( $jDayNo >= 366 ) {
1453 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1454 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1457 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1458 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1464 return array( $jy, $jm, $jd );
1468 * Converting Gregorian dates to Hijri dates.
1470 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1472 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1478 private static function tsToHijri( $ts ) {
1479 $year = substr( $ts, 0, 4 );
1480 $month = substr( $ts, 4, 2 );
1481 $day = substr( $ts, 6, 2 );
1489 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1490 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1493 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1494 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1495 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1498 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1499 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1502 $zl = $zjd -1948440 +
10632;
1503 $zn = (int)( ( $zl - 1 ) / 10631 );
1504 $zl = $zl - 10631 * $zn +
354;
1505 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1506 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1507 $zm = (int)( ( 24 * $zl ) / 709 );
1508 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1509 $zy = 30 * $zn +
$zj - 30;
1511 return array( $zy, $zm, $zd );
1515 * Converting Gregorian dates to Hebrew dates.
1517 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1518 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1519 * to translate the relevant functions into PHP and release them under
1522 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1523 * and Adar II is 14. In a non-leap year, Adar is 6.
1529 private static function tsToHebrew( $ts ) {
1531 $year = substr( $ts, 0, 4 );
1532 $month = substr( $ts, 4, 2 );
1533 $day = substr( $ts, 6, 2 );
1535 # Calculate Hebrew year
1536 $hebrewYear = $year +
3760;
1538 # Month number when September = 1, August = 12
1540 if ( $month > 12 ) {
1547 # Calculate day of year from 1 September
1549 for ( $i = 1; $i < $month; $i++
) {
1553 # Check if the year is leap
1554 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1557 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1564 # Calculate the start of the Hebrew year
1565 $start = self
::hebrewYearStart( $hebrewYear );
1567 # Calculate next year's start
1568 if ( $dayOfYear <= $start ) {
1569 # Day is before the start of the year - it is the previous year
1571 $nextStart = $start;
1575 # Add days since previous year's 1 September
1577 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1581 # Start of the new (previous) year
1582 $start = self
::hebrewYearStart( $hebrewYear );
1585 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1588 # Calculate Hebrew day of year
1589 $hebrewDayOfYear = $dayOfYear - $start;
1591 # Difference between year's days
1592 $diff = $nextStart - $start;
1593 # Add 12 (or 13 for leap years) days to ignore the difference between
1594 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1595 # difference is only about the year type
1596 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1602 # Check the year pattern, and is leap year
1603 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1604 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1605 # and non-leap years
1606 $yearPattern = $diff %
30;
1607 # Check if leap year
1608 $isLeap = $diff >= 30;
1610 # Calculate day in the month from number of day in the Hebrew year
1611 # Don't check Adar - if the day is not in Adar, we will stop before;
1612 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1613 $hebrewDay = $hebrewDayOfYear;
1616 while ( $hebrewMonth <= 12 ) {
1617 # Calculate days in this month
1618 if ( $isLeap && $hebrewMonth == 6 ) {
1619 # Adar in a leap year
1621 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1623 if ( $hebrewDay <= $days ) {
1627 # Subtract the days of Adar I
1628 $hebrewDay -= $days;
1631 if ( $hebrewDay <= $days ) {
1637 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1638 # Cheshvan in a complete year (otherwise as the rule below)
1640 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1641 # Kislev in an incomplete year (otherwise as the rule below)
1644 # Odd months have 30 days, even have 29
1645 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1647 if ( $hebrewDay <= $days ) {
1648 # In the current month
1651 # Subtract the days of the current month
1652 $hebrewDay -= $days;
1653 # Try in the next month
1658 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1662 * This calculates the Hebrew year start, as days since 1 September.
1663 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1664 * Used for Hebrew date.
1670 private static function hebrewYearStart( $year ) {
1671 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1672 $b = intval( ( $year - 1 ) %
4 );
1673 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1677 $Mar = intval( $m );
1683 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1684 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1686 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1688 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1692 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1697 * Algorithm to convert Gregorian dates to Thai solar dates,
1698 * Minguo dates or Minguo dates.
1700 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1701 * http://en.wikipedia.org/wiki/Minguo_calendar
1702 * http://en.wikipedia.org/wiki/Japanese_era_name
1704 * @param $ts String: 14-character timestamp
1705 * @param $cName String: calender name
1706 * @return Array: converted year, month, day
1708 private static function tsToYear( $ts, $cName ) {
1709 $gy = substr( $ts, 0, 4 );
1710 $gm = substr( $ts, 4, 2 );
1711 $gd = substr( $ts, 6, 2 );
1713 if ( !strcmp( $cName, 'thai' ) ) {
1715 # Add 543 years to the Gregorian calendar
1716 # Months and days are identical
1717 $gy_offset = $gy +
543;
1718 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1720 # Deduct 1911 years from the Gregorian calendar
1721 # Months and days are identical
1722 $gy_offset = $gy - 1911;
1723 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1724 # Nengō dates up to Meiji period
1725 # Deduct years from the Gregorian calendar
1726 # depending on the nengo periods
1727 # Months and days are identical
1728 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1730 $gy_gannen = $gy - 1868 +
1;
1731 $gy_offset = $gy_gannen;
1732 if ( $gy_gannen == 1 ) {
1735 $gy_offset = '明治' . $gy_offset;
1737 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1738 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1739 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1740 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1741 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1745 $gy_gannen = $gy - 1912 +
1;
1746 $gy_offset = $gy_gannen;
1747 if ( $gy_gannen == 1 ) {
1750 $gy_offset = '大正' . $gy_offset;
1752 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1753 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1754 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1758 $gy_gannen = $gy - 1926 +
1;
1759 $gy_offset = $gy_gannen;
1760 if ( $gy_gannen == 1 ) {
1763 $gy_offset = '昭和' . $gy_offset;
1766 $gy_gannen = $gy - 1989 +
1;
1767 $gy_offset = $gy_gannen;
1768 if ( $gy_gannen == 1 ) {
1771 $gy_offset = '平成' . $gy_offset;
1777 return array( $gy_offset, $gm, $gd );
1781 * Roman number formatting up to 10000
1787 static function romanNumeral( $num ) {
1788 static $table = array(
1789 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1790 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1791 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1792 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM', 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1795 $num = intval( $num );
1796 if ( $num > 10000 ||
$num <= 0 ) {
1801 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1802 if ( $num >= $pow10 ) {
1803 $s .= $table[$i][(int)floor( $num / $pow10 )];
1805 $num = $num %
$pow10;
1811 * Hebrew Gematria number formatting up to 9999
1817 static function hebrewNumeral( $num ) {
1818 static $table = array(
1819 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1820 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1821 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1822 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1825 $num = intval( $num );
1826 if ( $num > 9999 ||
$num <= 0 ) {
1831 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1832 if ( $num >= $pow10 ) {
1833 if ( $num == 15 ||
$num == 16 ) {
1834 $s .= $table[0][9] . $table[0][$num - 9];
1837 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1838 if ( $pow10 == 1000 ) {
1843 $num = $num %
$pow10;
1845 if ( strlen( $s ) == 2 ) {
1848 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1849 $str .= substr( $s, strlen( $s ) - 2, 2 );
1851 $start = substr( $str, 0, strlen( $str ) - 2 );
1852 $end = substr( $str, strlen( $str ) - 2 );
1855 $str = $start . 'ך';
1858 $str = $start . 'ם';
1861 $str = $start . 'ן';
1864 $str = $start . 'ף';
1867 $str = $start . 'ץ';
1874 * Used by date() and time() to adjust the time output.
1876 * @param $ts Int the time in date('YmdHis') format
1877 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1878 * get user timecorrection setting)
1881 function userAdjust( $ts, $tz = false ) {
1882 global $wgUser, $wgLocalTZoffset;
1884 if ( $tz === false ) {
1885 $tz = $wgUser->getOption( 'timecorrection' );
1888 $data = explode( '|', $tz, 3 );
1890 if ( $data[0] == 'ZoneInfo' ) {
1891 wfSuppressWarnings();
1892 $userTZ = timezone_open( $data[2] );
1893 wfRestoreWarnings();
1894 if ( $userTZ !== false ) {
1895 $date = date_create( $ts, timezone_open( 'UTC' ) );
1896 date_timezone_set( $date, $userTZ );
1897 $date = date_format( $date, 'YmdHis' );
1900 # Unrecognized timezone, default to 'Offset' with the stored offset.
1901 $data[0] = 'Offset';
1905 if ( $data[0] == 'System' ||
$tz == '' ) {
1906 # Global offset in minutes.
1907 if ( isset( $wgLocalTZoffset ) ) {
1908 $minDiff = $wgLocalTZoffset;
1910 } elseif ( $data[0] == 'Offset' ) {
1911 $minDiff = intval( $data[1] );
1913 $data = explode( ':', $tz );
1914 if ( count( $data ) == 2 ) {
1915 $data[0] = intval( $data[0] );
1916 $data[1] = intval( $data[1] );
1917 $minDiff = abs( $data[0] ) * 60 +
$data[1];
1918 if ( $data[0] < 0 ) {
1919 $minDiff = -$minDiff;
1922 $minDiff = intval( $data[0] ) * 60;
1926 # No difference ? Return time unchanged
1927 if ( 0 == $minDiff ) {
1931 wfSuppressWarnings(); // E_STRICT system time bitching
1932 # Generate an adjusted date; take advantage of the fact that mktime
1933 # will normalize out-of-range values so we don't have to split $minDiff
1934 # into hours and minutes.
1936 (int)substr( $ts, 8, 2 ) ), # Hours
1937 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
1938 (int)substr( $ts, 12, 2 ), # Seconds
1939 (int)substr( $ts, 4, 2 ), # Month
1940 (int)substr( $ts, 6, 2 ), # Day
1941 (int)substr( $ts, 0, 4 ) ); # Year
1943 $date = date( 'YmdHis', $t );
1944 wfRestoreWarnings();
1950 * This is meant to be used by time(), date(), and timeanddate() to get
1951 * the date preference they're supposed to use, it should be used in
1955 * function timeanddate([...], $format = true) {
1956 * $datePreference = $this->dateFormat($format);
1961 * @param $usePrefs Mixed: if true, the user's preference is used
1962 * if false, the site/language default is used
1963 * if int/string, assumed to be a format.
1966 function dateFormat( $usePrefs = true ) {
1969 if ( is_bool( $usePrefs ) ) {
1971 $datePreference = $wgUser->getDatePreference();
1973 $datePreference = (string)User
::getDefaultOption( 'date' );
1976 $datePreference = (string)$usePrefs;
1980 if ( $datePreference == '' ) {
1984 return $datePreference;
1988 * Get a format string for a given type and preference
1989 * @param $type string May be date, time or both
1990 * @param $pref string The format name as it appears in Messages*.php
1992 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
1996 function getDateFormatString( $type, $pref ) {
1997 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
1998 if ( $pref == 'default' ) {
1999 $pref = $this->getDefaultDateFormat();
2000 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2002 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2004 if ( $type === 'pretty' && $df === null ) {
2005 $df = $this->getDateFormatString( 'date', $pref );
2008 if ( $df === null ) {
2009 $pref = $this->getDefaultDateFormat();
2010 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2013 $this->dateFormatStrings
[$type][$pref] = $df;
2015 return $this->dateFormatStrings
[$type][$pref];
2019 * @param $ts Mixed: the time format which needs to be turned into a
2020 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2021 * @param $adj Bool: whether to adjust the time output according to the
2022 * user configured offset ($timecorrection)
2023 * @param $format Mixed: true to use user's date format preference
2024 * @param $timecorrection String|bool the time offset as returned by
2025 * validateTimeZone() in Special:Preferences
2028 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2029 $ts = wfTimestamp( TS_MW
, $ts );
2031 $ts = $this->userAdjust( $ts, $timecorrection );
2033 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2034 return $this->sprintfDate( $df, $ts );
2038 * @param $ts Mixed: the time format which needs to be turned into a
2039 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2040 * @param $adj Bool: whether to adjust the time output according to the
2041 * user configured offset ($timecorrection)
2042 * @param $format Mixed: true to use user's date format preference
2043 * @param $timecorrection String|bool the time offset as returned by
2044 * validateTimeZone() in Special:Preferences
2047 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2048 $ts = wfTimestamp( TS_MW
, $ts );
2050 $ts = $this->userAdjust( $ts, $timecorrection );
2052 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2053 return $this->sprintfDate( $df, $ts );
2057 * @param $ts Mixed: the time format which needs to be turned into a
2058 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2059 * @param $adj Bool: whether to adjust the time output according to the
2060 * user configured offset ($timecorrection)
2061 * @param $format Mixed: what format to return, if it's false output the
2062 * default one (default true)
2063 * @param $timecorrection String|bool the time offset as returned by
2064 * validateTimeZone() in Special:Preferences
2067 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2068 $ts = wfTimestamp( TS_MW
, $ts );
2070 $ts = $this->userAdjust( $ts, $timecorrection );
2072 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2073 return $this->sprintfDate( $df, $ts );
2077 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2081 * @param integer $seconds The amount of seconds.
2082 * @param array $chosenIntervals The intervals to enable.
2086 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2087 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2089 $segments = array();
2091 foreach ( $intervals as $intervalName => $intervalValue ) {
2092 // Messages: duration-centuries, duration-decades, duration-years, duration-days,
2093 // duration-hours, duration-minutes, duration-seconds
2094 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2095 $segments[] = $message->inLanguage( $this )->escaped();
2098 return $this->listToText( $segments );
2102 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2103 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2107 * @param integer $seconds The amount of seconds.
2108 * @param array $chosenIntervals The intervals to enable.
2112 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2113 if ( empty( $chosenIntervals ) ) {
2114 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
2117 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2118 $sortedNames = array_keys( $intervals );
2119 $smallestInterval = array_pop( $sortedNames );
2121 $segments = array();
2123 foreach ( $intervals as $name => $length ) {
2124 $value = floor( $seconds / $length );
2126 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2127 $seconds -= $value * $length;
2128 $segments[$name] = $value;
2136 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2138 * @param $type String: can be 'date', 'time' or 'both'
2139 * @param $ts Mixed: the time format which needs to be turned into a
2140 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2141 * @param $user User object used to get preferences for timezone and format
2142 * @param $options Array, can contain the following keys:
2143 * - 'timecorrection': time correction, can have the following values:
2144 * - true: use user's preference
2145 * - false: don't use time correction
2146 * - integer: value of time correction in minutes
2147 * - 'format': format to use, can have the following values:
2148 * - true: use user's preference
2149 * - false: use default preference
2150 * - string: format to use
2154 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2155 $ts = wfTimestamp( TS_MW
, $ts );
2156 $options +
= array( 'timecorrection' => true, 'format' => true );
2157 if ( $options['timecorrection'] !== false ) {
2158 if ( $options['timecorrection'] === true ) {
2159 $offset = $user->getOption( 'timecorrection' );
2161 $offset = $options['timecorrection'];
2163 $ts = $this->userAdjust( $ts, $offset );
2165 if ( $options['format'] === true ) {
2166 $format = $user->getDatePreference();
2168 $format = $options['format'];
2170 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2171 return $this->sprintfDate( $df, $ts );
2175 * Get the formatted date for the given timestamp and formatted for
2178 * @param $ts Mixed: the time format which needs to be turned into a
2179 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2180 * @param $user User object used to get preferences for timezone and format
2181 * @param $options Array, can contain the following keys:
2182 * - 'timecorrection': time correction, can have the following values:
2183 * - true: use user's preference
2184 * - false: don't use time correction
2185 * - integer: value of time correction in minutes
2186 * - 'format': format to use, can have the following values:
2187 * - true: use user's preference
2188 * - false: use default preference
2189 * - string: format to use
2193 public function userDate( $ts, User
$user, array $options = array() ) {
2194 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2198 * Get the formatted time for the given timestamp and formatted for
2201 * @param $ts Mixed: the time format which needs to be turned into a
2202 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2203 * @param $user User object used to get preferences for timezone and format
2204 * @param $options Array, can contain the following keys:
2205 * - 'timecorrection': time correction, can have the following values:
2206 * - true: use user's preference
2207 * - false: don't use time correction
2208 * - integer: value of time correction in minutes
2209 * - 'format': format to use, can have the following values:
2210 * - true: use user's preference
2211 * - false: use default preference
2212 * - string: format to use
2216 public function userTime( $ts, User
$user, array $options = array() ) {
2217 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2221 * Get the formatted date and time for the given timestamp and formatted for
2224 * @param $ts Mixed: the time format which needs to be turned into a
2225 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2226 * @param $user User object used to get preferences for timezone and format
2227 * @param $options Array, can contain the following keys:
2228 * - 'timecorrection': time correction, can have the following values:
2229 * - true: use user's preference
2230 * - false: don't use time correction
2231 * - integer: value of time correction in minutes
2232 * - 'format': format to use, can have the following values:
2233 * - true: use user's preference
2234 * - false: use default preference
2235 * - string: format to use
2239 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2240 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2244 * Convert an MWTimestamp into a pretty human-readable timestamp using
2245 * the given user preferences and relative base time.
2247 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2248 * on your timestamp object, which will then call this function. Calling
2249 * this function directly will cause hooks to be skipped over.
2251 * @see MWTimestamp::getHumanTimestamp
2252 * @param MWTimestamp $ts Timestamp to prettify
2253 * @param MWTimestamp $relativeTo Base timestamp
2254 * @param User $user User preferences to use
2255 * @return string Human timestamp
2258 public function getHumanTimestamp( MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user ) {
2259 $diff = $ts->diff( $relativeTo );
2260 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) - (int)$relativeTo->timestamp
->format( 'w' ) );
2261 $days = $diff->days ?
: (int)$diffDay;
2262 if ( $diff->invert ||
$days > 5 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' ) ) {
2263 // Timestamps are in different years: use full timestamp
2264 // Also do full timestamp for future dates
2266 * @FIXME Add better handling of future timestamps.
2268 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2269 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2270 } elseif ( $days > 5 ) {
2271 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2272 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2273 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2274 } elseif ( $days > 1 ) {
2275 // Timestamp within the past week: show the day of the week and time
2276 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2277 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2278 $ts = wfMessage( "$weekday-at" )
2279 ->inLanguage( $this )
2280 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2282 } elseif ( $days == 1 ) {
2283 // Timestamp was yesterday: say 'yesterday' and the time.
2284 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2285 $ts = wfMessage( 'yesterday-at' )
2286 ->inLanguage( $this )
2287 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2289 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2290 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2291 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2292 $ts = wfMessage( 'today-at' )
2293 ->inLanguage( $this )
2294 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2297 // From here on in, the timestamp was soon enough ago so that we can simply say
2298 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2299 } elseif ( $diff->h
== 1 ) {
2300 // Less than 90 minutes, but more than an hour ago.
2301 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2302 } elseif ( $diff->i
>= 1 ) {
2303 // A few minutes ago.
2304 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2305 } elseif ( $diff->s
>= 30 ) {
2306 // Less than a minute, but more than 30 sec ago.
2307 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2309 // Less than 30 seconds ago.
2310 $ts = wfMessage( 'just-now' )->text();
2317 * @param $key string
2318 * @return array|null
2320 function getMessage( $key ) {
2321 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2327 function getAllMessages() {
2328 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2337 function iconv( $in, $out, $string ) {
2338 # This is a wrapper for iconv in all languages except esperanto,
2339 # which does some nasty x-conversions beforehand
2341 # Even with //IGNORE iconv can whine about illegal characters in
2342 # *input* string. We just ignore those too.
2343 # REF: http://bugs.php.net/bug.php?id=37166
2344 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2345 wfSuppressWarnings();
2346 $text = iconv( $in, $out . '//IGNORE', $string );
2347 wfRestoreWarnings();
2351 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2354 * @param $matches array
2355 * @return mixed|string
2357 function ucwordbreaksCallbackAscii( $matches ) {
2358 return $this->ucfirst( $matches[1] );
2362 * @param $matches array
2365 function ucwordbreaksCallbackMB( $matches ) {
2366 return mb_strtoupper( $matches[0] );
2370 * @param $matches array
2373 function ucCallback( $matches ) {
2374 list( $wikiUpperChars ) = self
::getCaseMaps();
2375 return strtr( $matches[1], $wikiUpperChars );
2379 * @param $matches array
2382 function lcCallback( $matches ) {
2383 list( , $wikiLowerChars ) = self
::getCaseMaps();
2384 return strtr( $matches[1], $wikiLowerChars );
2388 * @param $matches array
2391 function ucwordsCallbackMB( $matches ) {
2392 return mb_strtoupper( $matches[0] );
2396 * @param $matches array
2399 function ucwordsCallbackWiki( $matches ) {
2400 list( $wikiUpperChars ) = self
::getCaseMaps();
2401 return strtr( $matches[0], $wikiUpperChars );
2405 * Make a string's first character uppercase
2407 * @param $str string
2411 function ucfirst( $str ) {
2413 if ( $o < 96 ) { // if already uppercase...
2415 } elseif ( $o < 128 ) {
2416 return ucfirst( $str ); // use PHP's ucfirst()
2418 // fall back to more complex logic in case of multibyte strings
2419 return $this->uc( $str, true );
2424 * Convert a string to uppercase
2426 * @param $str string
2427 * @param $first bool
2431 function uc( $str, $first = false ) {
2432 if ( function_exists( 'mb_strtoupper' ) ) {
2434 if ( $this->isMultibyte( $str ) ) {
2435 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2437 return ucfirst( $str );
2440 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2443 if ( $this->isMultibyte( $str ) ) {
2444 $x = $first ?
'^' : '';
2445 return preg_replace_callback(
2446 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2447 array( $this, 'ucCallback' ),
2451 return $first ?
ucfirst( $str ) : strtoupper( $str );
2457 * @param $str string
2458 * @return mixed|string
2460 function lcfirst( $str ) {
2463 return strval( $str );
2464 } elseif ( $o >= 128 ) {
2465 return $this->lc( $str, true );
2466 } elseif ( $o > 96 ) {
2469 $str[0] = strtolower( $str[0] );
2475 * @param $str string
2476 * @param $first bool
2477 * @return mixed|string
2479 function lc( $str, $first = false ) {
2480 if ( function_exists( 'mb_strtolower' ) ) {
2482 if ( $this->isMultibyte( $str ) ) {
2483 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2485 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2488 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2491 if ( $this->isMultibyte( $str ) ) {
2492 $x = $first ?
'^' : '';
2493 return preg_replace_callback(
2494 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2495 array( $this, 'lcCallback' ),
2499 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2505 * @param $str string
2508 function isMultibyte( $str ) {
2509 return (bool)preg_match( '/[\x80-\xff]/', $str );
2513 * @param $str string
2514 * @return mixed|string
2516 function ucwords( $str ) {
2517 if ( $this->isMultibyte( $str ) ) {
2518 $str = $this->lc( $str );
2520 // regexp to find first letter in each word (i.e. after each space)
2521 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2523 // function to use to capitalize a single char
2524 if ( function_exists( 'mb_strtoupper' ) ) {
2525 return preg_replace_callback(
2527 array( $this, 'ucwordsCallbackMB' ),
2531 return preg_replace_callback(
2533 array( $this, 'ucwordsCallbackWiki' ),
2538 return ucwords( strtolower( $str ) );
2543 * capitalize words at word breaks
2545 * @param $str string
2548 function ucwordbreaks( $str ) {
2549 if ( $this->isMultibyte( $str ) ) {
2550 $str = $this->lc( $str );
2552 // since \b doesn't work for UTF-8, we explicitely define word break chars
2553 $breaks = "[ \-\(\)\}\{\.,\?!]";
2555 // find first letter after word break
2556 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2558 if ( function_exists( 'mb_strtoupper' ) ) {
2559 return preg_replace_callback(
2561 array( $this, 'ucwordbreaksCallbackMB' ),
2565 return preg_replace_callback(
2567 array( $this, 'ucwordsCallbackWiki' ),
2572 return preg_replace_callback(
2573 '/\b([\w\x80-\xff]+)\b/',
2574 array( $this, 'ucwordbreaksCallbackAscii' ),
2581 * Return a case-folded representation of $s
2583 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2584 * and $s2 are the same except for the case of their characters. It is not
2585 * necessary for the value returned to make sense when displayed.
2587 * Do *not* perform any other normalisation in this function. If a caller
2588 * uses this function when it should be using a more general normalisation
2589 * function, then fix the caller.
2595 function caseFold( $s ) {
2596 return $this->uc( $s );
2603 function checkTitleEncoding( $s ) {
2604 if ( is_array( $s ) ) {
2605 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2607 if ( StringUtils
::isUtf8( $s ) ) {
2611 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2617 function fallback8bitEncoding() {
2618 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2622 * Most writing systems use whitespace to break up words.
2623 * Some languages such as Chinese don't conventionally do this,
2624 * which requires special handling when breaking up words for
2629 function hasWordBreaks() {
2634 * Some languages such as Chinese require word segmentation,
2635 * Specify such segmentation when overridden in derived class.
2637 * @param $string String
2640 function segmentByWord( $string ) {
2645 * Some languages have special punctuation need to be normalized.
2646 * Make such changes here.
2648 * @param $string String
2651 function normalizeForSearch( $string ) {
2652 return self
::convertDoubleWidth( $string );
2656 * convert double-width roman characters to single-width.
2657 * range: ff00-ff5f ~= 0020-007f
2659 * @param $string string
2663 protected static function convertDoubleWidth( $string ) {
2664 static $full = null;
2665 static $half = null;
2667 if ( $full === null ) {
2668 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2669 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2670 $full = str_split( $fullWidth, 3 );
2671 $half = str_split( $halfWidth );
2674 $string = str_replace( $full, $half, $string );
2679 * @param $string string
2680 * @param $pattern string
2683 protected static function insertSpace( $string, $pattern ) {
2684 $string = preg_replace( $pattern, " $1 ", $string );
2685 $string = preg_replace( '/ +/', ' ', $string );
2690 * @param $termsArray array
2693 function convertForSearchResult( $termsArray ) {
2694 # some languages, e.g. Chinese, need to do a conversion
2695 # in order for search results to be displayed correctly
2700 * Get the first character of a string.
2705 function firstChar( $s ) {
2708 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2709 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2714 if ( isset( $matches[1] ) ) {
2715 if ( strlen( $matches[1] ) != 3 ) {
2719 // Break down Hangul syllables to grab the first jamo
2720 $code = utf8ToCodepoint( $matches[1] );
2721 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2723 } elseif ( $code < 0xb098 ) {
2724 return "\xe3\x84\xb1";
2725 } elseif ( $code < 0xb2e4 ) {
2726 return "\xe3\x84\xb4";
2727 } elseif ( $code < 0xb77c ) {
2728 return "\xe3\x84\xb7";
2729 } elseif ( $code < 0xb9c8 ) {
2730 return "\xe3\x84\xb9";
2731 } elseif ( $code < 0xbc14 ) {
2732 return "\xe3\x85\x81";
2733 } elseif ( $code < 0xc0ac ) {
2734 return "\xe3\x85\x82";
2735 } elseif ( $code < 0xc544 ) {
2736 return "\xe3\x85\x85";
2737 } elseif ( $code < 0xc790 ) {
2738 return "\xe3\x85\x87";
2739 } elseif ( $code < 0xcc28 ) {
2740 return "\xe3\x85\x88";
2741 } elseif ( $code < 0xce74 ) {
2742 return "\xe3\x85\x8a";
2743 } elseif ( $code < 0xd0c0 ) {
2744 return "\xe3\x85\x8b";
2745 } elseif ( $code < 0xd30c ) {
2746 return "\xe3\x85\x8c";
2747 } elseif ( $code < 0xd558 ) {
2748 return "\xe3\x85\x8d";
2750 return "\xe3\x85\x8e";
2757 function initEncoding() {
2758 # Some languages may have an alternate char encoding option
2759 # (Esperanto X-coding, Japanese furigana conversion, etc)
2760 # If this language is used as the primary content language,
2761 # an override to the defaults can be set here on startup.
2768 function recodeForEdit( $s ) {
2769 # For some languages we'll want to explicitly specify
2770 # which characters make it into the edit box raw
2771 # or are converted in some way or another.
2772 global $wgEditEncoding;
2773 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2776 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2784 function recodeInput( $s ) {
2785 # Take the previous into account.
2786 global $wgEditEncoding;
2787 if ( $wgEditEncoding != '' ) {
2788 $enc = $wgEditEncoding;
2792 if ( $enc == 'UTF-8' ) {
2795 return $this->iconv( $enc, 'UTF-8', $s );
2800 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2801 * also cleans up certain backwards-compatible sequences, converting them
2802 * to the modern Unicode equivalent.
2804 * This is language-specific for performance reasons only.
2810 function normalize( $s ) {
2811 global $wgAllUnicodeFixes;
2812 $s = UtfNormal
::cleanUp( $s );
2813 if ( $wgAllUnicodeFixes ) {
2814 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2815 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2822 * Transform a string using serialized data stored in the given file (which
2823 * must be in the serialized subdirectory of $IP). The file contains pairs
2824 * mapping source characters to destination characters.
2826 * The data is cached in process memory. This will go faster if you have the
2827 * FastStringSearch extension.
2829 * @param $file string
2830 * @param $string string
2832 * @throws MWException
2835 function transformUsingPairFile( $file, $string ) {
2836 if ( !isset( $this->transformData
[$file] ) ) {
2837 $data = wfGetPrecompiledData( $file );
2838 if ( $data === false ) {
2839 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
2841 $this->transformData
[$file] = new ReplacementArray( $data );
2843 return $this->transformData
[$file]->replace( $string );
2847 * For right-to-left language support
2852 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
2856 * Return the correct HTML 'dir' attribute value for this language.
2860 return $this->isRTL() ?
'rtl' : 'ltr';
2864 * Return 'left' or 'right' as appropriate alignment for line-start
2865 * for this language's text direction.
2867 * Should be equivalent to CSS3 'start' text-align value....
2871 function alignStart() {
2872 return $this->isRTL() ?
'right' : 'left';
2876 * Return 'right' or 'left' as appropriate alignment for line-end
2877 * for this language's text direction.
2879 * Should be equivalent to CSS3 'end' text-align value....
2883 function alignEnd() {
2884 return $this->isRTL() ?
'left' : 'right';
2888 * A hidden direction mark (LRM or RLM), depending on the language direction.
2889 * Unlike getDirMark(), this function returns the character as an HTML entity.
2890 * This function should be used when the output is guaranteed to be HTML,
2891 * because it makes the output HTML source code more readable. When
2892 * the output is plain text or can be escaped, getDirMark() should be used.
2894 * @param $opposite Boolean Get the direction mark opposite to your language
2898 function getDirMarkEntity( $opposite = false ) {
2900 return $this->isRTL() ?
'‎' : '‏';
2902 return $this->isRTL() ?
'‏' : '‎';
2906 * A hidden direction mark (LRM or RLM), depending on the language direction.
2907 * This function produces them as invisible Unicode characters and
2908 * the output may be hard to read and debug, so it should only be used
2909 * when the output is plain text or can be escaped. When the output is
2910 * HTML, use getDirMarkEntity() instead.
2912 * @param $opposite Boolean Get the direction mark opposite to your language
2915 function getDirMark( $opposite = false ) {
2916 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2917 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2919 return $this->isRTL() ?
$lrm : $rlm;
2921 return $this->isRTL() ?
$rlm : $lrm;
2927 function capitalizeAllNouns() {
2928 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2932 * An arrow, depending on the language direction.
2934 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2937 function getArrow( $direction = 'forwards' ) {
2938 switch ( $direction ) {
2940 return $this->isRTL() ?
'←' : '→';
2942 return $this->isRTL() ?
'→' : '←';
2955 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2959 function linkPrefixExtension() {
2960 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2964 * Get all magic words from cache.
2967 function getMagicWords() {
2968 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2972 * Run the LanguageGetMagic hook once.
2974 protected function doMagicHook() {
2975 if ( $this->mMagicHookDone
) {
2978 $this->mMagicHookDone
= true;
2979 wfProfileIn( 'LanguageGetMagic' );
2980 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2981 wfProfileOut( 'LanguageGetMagic' );
2985 * Fill a MagicWord object with data from here
2989 function getMagic( $mw ) {
2990 // Saves a function call
2991 if ( ! $this->mMagicHookDone
) {
2992 $this->doMagicHook();
2995 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
2996 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
2998 $rawEntry = self
::$dataCache->getSubitem(
2999 $this->mCode
, 'magicWords', $mw->mId
);
3002 if ( !is_array( $rawEntry ) ) {
3003 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3005 $mw->mCaseSensitive
= $rawEntry[0];
3006 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3011 * Add magic words to the extension array
3013 * @param $newWords array
3015 function addMagicWordsByLang( $newWords ) {
3016 $fallbackChain = $this->getFallbackLanguages();
3017 $fallbackChain = array_reverse( $fallbackChain );
3018 foreach ( $fallbackChain as $code ) {
3019 if ( isset( $newWords[$code] ) ) {
3020 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3026 * Get special page names, as an associative array
3027 * case folded alias => real name
3029 function getSpecialPageAliases() {
3030 // Cache aliases because it may be slow to load them
3031 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3033 $this->mExtendedSpecialPageAliases
=
3034 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3035 wfRunHooks( 'LanguageGetSpecialPageAliases',
3036 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
3039 return $this->mExtendedSpecialPageAliases
;
3043 * Italic is unsuitable for some languages
3045 * @param $text String: the text to be emphasized.
3048 function emphasize( $text ) {
3049 return "<em>$text</em>";
3053 * Normally we output all numbers in plain en_US style, that is
3054 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3055 * point twohundredthirtyfive. However this is not suitable for all
3056 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
3057 * Icelandic just want to use commas instead of dots, and dots instead
3058 * of commas like "293.291,235".
3060 * An example of this function being called:
3062 * wfMessage( 'message' )->numParams( $num )->text()
3065 * See LanguageGu.php for the Gujarati implementation and
3066 * $separatorTransformTable on MessageIs.php for
3067 * the , => . and . => , implementation.
3069 * @todo check if it's viable to use localeconv() for the decimal
3071 * @param $number Mixed: the string to be formatted, should be an integer
3072 * or a floating point number.
3073 * @param $nocommafy Bool: set to true for special numbers like dates
3076 public function formatNum( $number, $nocommafy = false ) {
3077 global $wgTranslateNumerals;
3078 if ( !$nocommafy ) {
3079 $number = $this->commafy( $number );
3080 $s = $this->separatorTransformTable();
3082 $number = strtr( $number, $s );
3086 if ( $wgTranslateNumerals ) {
3087 $s = $this->digitTransformTable();
3089 $number = strtr( $number, $s );
3097 * Front-end for non-commafied formatNum
3099 * @param mixed $number the string to be formatted, should be an integer
3100 * or a floating point number.
3104 public function formatNumNoSeparators( $number ) {
3105 return $this->formatNum( $number, true );
3109 * @param $number string
3112 function parseFormattedNumber( $number ) {
3113 $s = $this->digitTransformTable();
3115 $number = strtr( $number, array_flip( $s ) );
3118 $s = $this->separatorTransformTable();
3120 $number = strtr( $number, array_flip( $s ) );
3123 $number = strtr( $number, array( ',' => '' ) );
3128 * Adds commas to a given number
3130 * @param $number mixed
3133 function commafy( $number ) {
3134 $digitGroupingPattern = $this->digitGroupingPattern();
3135 if ( $number === null ) {
3139 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3140 // default grouping is at thousands, use the same for ###,###,### pattern too.
3141 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3143 // Ref: http://cldr.unicode.org/translation/number-patterns
3145 if ( intval( $number ) < 0 ) {
3146 // For negative numbers apply the algorithm like positive number and add sign.
3148 $number = substr( $number, 1 );
3150 $integerPart = array();
3151 $decimalPart = array();
3152 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3153 preg_match( "/\d+/", $number, $integerPart );
3154 preg_match( "/\.\d*/", $number, $decimalPart );
3155 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3156 if ( $groupedNumber === $number ) {
3157 // the string does not have any number part. Eg: .12345
3158 return $sign . $groupedNumber;
3160 $start = $end = strlen( $integerPart[0] );
3161 while ( $start > 0 ) {
3162 $match = $matches[0][$numMatches - 1];
3163 $matchLen = strlen( $match );
3164 $start = $end - $matchLen;
3168 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3170 if ( $numMatches > 1 ) {
3171 // use the last pattern for the rest of the number
3175 $groupedNumber = "," . $groupedNumber;
3178 return $sign . $groupedNumber;
3185 function digitGroupingPattern() {
3186 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3192 function digitTransformTable() {
3193 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3199 function separatorTransformTable() {
3200 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3204 * Take a list of strings and build a locale-friendly comma-separated
3205 * list, using the local comma-separator message.
3206 * The last two strings are chained with an "and".
3207 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3212 function listToText( array $l ) {
3213 $m = count( $l ) - 1;
3218 $and = $this->getMessageFromDB( 'and' );
3219 $space = $this->getMessageFromDB( 'word-separator' );
3221 $comma = $this->getMessageFromDB( 'comma-separator' );
3225 for ( $i = $m - 1; $i >= 0; $i-- ) {
3226 if ( $i == $m - 1 ) {
3227 $s = $l[$i] . $and . $space . $s;
3229 $s = $l[$i] . $comma . $s;
3236 * Take a list of strings and build a locale-friendly comma-separated
3237 * list, using the local comma-separator message.
3238 * @param $list array of strings to put in a comma list
3241 function commaList( array $list ) {
3243 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3249 * Take a list of strings and build a locale-friendly semicolon-separated
3250 * list, using the local semicolon-separator message.
3251 * @param $list array of strings to put in a semicolon list
3254 function semicolonList( array $list ) {
3256 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3262 * Same as commaList, but separate it with the pipe instead.
3263 * @param $list array of strings to put in a pipe list
3266 function pipeList( array $list ) {
3268 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3274 * Truncate a string to a specified length in bytes, appending an optional
3275 * string (e.g. for ellipses)
3277 * The database offers limited byte lengths for some columns in the database;
3278 * multi-byte character sets mean we need to ensure that only whole characters
3279 * are included, otherwise broken characters can be passed to the user
3281 * If $length is negative, the string will be truncated from the beginning
3283 * @param $string String to truncate
3284 * @param $length Int: maximum length (including ellipses)
3285 * @param $ellipsis String to append to the truncated text
3286 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3287 * $adjustLength was introduced in 1.18, before that behaved as if false.
3290 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3291 # Use the localized ellipsis character
3292 if ( $ellipsis == '...' ) {
3293 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3295 # Check if there is no need to truncate
3296 if ( $length == 0 ) {
3297 return $ellipsis; // convention
3298 } elseif ( strlen( $string ) <= abs( $length ) ) {
3299 return $string; // no need to truncate
3301 $stringOriginal = $string;
3302 # If ellipsis length is >= $length then we can't apply $adjustLength
3303 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3304 $string = $ellipsis; // this can be slightly unexpected
3305 # Otherwise, truncate and add ellipsis...
3307 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3308 if ( $length > 0 ) {
3309 $length -= $eLength;
3310 $string = substr( $string, 0, $length ); // xyz...
3311 $string = $this->removeBadCharLast( $string );
3312 $string = $string . $ellipsis;
3314 $length +
= $eLength;
3315 $string = substr( $string, $length ); // ...xyz
3316 $string = $this->removeBadCharFirst( $string );
3317 $string = $ellipsis . $string;
3320 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3321 # This check is *not* redundant if $adjustLength, due to the single case where
3322 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3323 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3326 return $stringOriginal;
3331 * Remove bytes that represent an incomplete Unicode character
3332 * at the end of string (e.g. bytes of the char are missing)
3334 * @param $string String
3337 protected function removeBadCharLast( $string ) {
3338 if ( $string != '' ) {
3339 $char = ord( $string[strlen( $string ) - 1] );
3341 if ( $char >= 0xc0 ) {
3342 # We got the first byte only of a multibyte char; remove it.
3343 $string = substr( $string, 0, -1 );
3344 } elseif ( $char >= 0x80 &&
3345 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3346 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3348 # We chopped in the middle of a character; remove it
3356 * Remove bytes that represent an incomplete Unicode character
3357 * at the start of string (e.g. bytes of the char are missing)
3359 * @param $string String
3362 protected function removeBadCharFirst( $string ) {
3363 if ( $string != '' ) {
3364 $char = ord( $string[0] );
3365 if ( $char >= 0x80 && $char < 0xc0 ) {
3366 # We chopped in the middle of a character; remove the whole thing
3367 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3374 * Truncate a string of valid HTML to a specified length in bytes,
3375 * appending an optional string (e.g. for ellipses), and return valid HTML
3377 * This is only intended for styled/linked text, such as HTML with
3378 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3379 * Also, this will not detect things like "display:none" CSS.
3381 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3383 * @param string $text HTML string to truncate
3384 * @param int $length (zero/positive) Maximum length (including ellipses)
3385 * @param string $ellipsis String to append to the truncated text
3388 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3389 # Use the localized ellipsis character
3390 if ( $ellipsis == '...' ) {
3391 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3393 # Check if there is clearly no need to truncate
3394 if ( $length <= 0 ) {
3395 return $ellipsis; // no text shown, nothing to format (convention)
3396 } elseif ( strlen( $text ) <= $length ) {
3397 return $text; // string short enough even *with* HTML (short-circuit)
3400 $dispLen = 0; // innerHTML legth so far
3401 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3402 $tagType = 0; // 0-open, 1-close
3403 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3404 $entityState = 0; // 0-not entity, 1-entity
3405 $tag = $ret = ''; // accumulated tag name, accumulated result string
3406 $openTags = array(); // open tag stack
3407 $maybeState = null; // possible truncation state
3409 $textLen = strlen( $text );
3410 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3411 for ( $pos = 0; true; ++
$pos ) {
3412 # Consider truncation once the display length has reached the maximim.
3413 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3414 # Check that we're not in the middle of a bracket/entity...
3415 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3416 if ( !$testingEllipsis ) {
3417 $testingEllipsis = true;
3418 # Save where we are; we will truncate here unless there turn out to
3419 # be so few remaining characters that truncation is not necessary.
3420 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3421 $maybeState = array( $ret, $openTags ); // save state
3423 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3424 # String in fact does need truncation, the truncation point was OK.
3425 list( $ret, $openTags ) = $maybeState; // reload state
3426 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3427 $ret .= $ellipsis; // add ellipsis
3431 if ( $pos >= $textLen ) {
3432 break; // extra iteration just for above checks
3435 # Read the next char...
3437 $lastCh = $pos ?
$text[$pos - 1] : '';
3438 $ret .= $ch; // add to result string
3440 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3441 $entityState = 0; // for bad HTML
3442 $bracketState = 1; // tag started (checking for backslash)
3443 } elseif ( $ch == '>' ) {
3444 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3445 $entityState = 0; // for bad HTML
3446 $bracketState = 0; // out of brackets
3447 } elseif ( $bracketState == 1 ) {
3449 $tagType = 1; // close tag (e.g. "</span>")
3451 $tagType = 0; // open tag (e.g. "<span>")
3454 $bracketState = 2; // building tag name
3455 } elseif ( $bracketState == 2 ) {
3459 // Name found (e.g. "<a href=..."), add on tag attributes...
3460 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3462 } elseif ( $bracketState == 0 ) {
3463 if ( $entityState ) {
3466 $dispLen++
; // entity is one displayed char
3469 if ( $neLength == 0 && !$maybeState ) {
3470 // Save state without $ch. We want to *hit* the first
3471 // display char (to get tags) but not *use* it if truncating.
3472 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3475 $entityState = 1; // entity found, (e.g. " ")
3477 $dispLen++
; // this char is displayed
3478 // Add the next $max display text chars after this in one swoop...
3479 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3480 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3481 $dispLen +
= $skipped;
3487 // Close the last tag if left unclosed by bad HTML
3488 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3489 while ( count( $openTags ) > 0 ) {
3490 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3496 * truncateHtml() helper function
3497 * like strcspn() but adds the skipped chars to $ret
3506 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3507 if ( $len === null ) {
3508 $len = -1; // -1 means "no limit" for strcspn
3509 } elseif ( $len < 0 ) {
3513 if ( $start < strlen( $text ) ) {
3514 $skipCount = strcspn( $text, $search, $start, $len );
3515 $ret .= substr( $text, $start, $skipCount );
3521 * truncateHtml() helper function
3522 * (a) push or pop $tag from $openTags as needed
3523 * (b) clear $tag value
3524 * @param &$tag string Current HTML tag name we are looking at
3525 * @param $tagType int (0-open tag, 1-close tag)
3526 * @param $lastCh string Character before the '>' that ended this tag
3527 * @param &$openTags array Open tag stack (not accounting for $tag)
3529 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3530 $tag = ltrim( $tag );
3532 if ( $tagType == 0 && $lastCh != '/' ) {
3533 $openTags[] = $tag; // tag opened (didn't close itself)
3534 } elseif ( $tagType == 1 ) {
3535 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3536 array_pop( $openTags ); // tag closed
3544 * Grammatical transformations, needed for inflected languages
3545 * Invoked by putting {{grammar:case|word}} in a message
3547 * @param $word string
3548 * @param $case string
3551 function convertGrammar( $word, $case ) {
3552 global $wgGrammarForms;
3553 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3554 return $wgGrammarForms[$this->getCode()][$case][$word];
3559 * Get the grammar forms for the content language
3560 * @return array of grammar forms
3563 function getGrammarForms() {
3564 global $wgGrammarForms;
3565 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3566 return $wgGrammarForms[$this->getCode()];
3571 * Provides an alternative text depending on specified gender.
3572 * Usage {{gender:username|masculine|feminine|neutral}}.
3573 * username is optional, in which case the gender of current user is used,
3574 * but only in (some) interface messages; otherwise default gender is used.
3576 * If no forms are given, an empty string is returned. If only one form is
3577 * given, it will be returned unconditionally. These details are implied by
3578 * the caller and cannot be overridden in subclasses.
3580 * If more than one form is given, the default is to use the neutral one
3581 * if it is specified, and to use the masculine one otherwise. These
3582 * details can be overridden in subclasses.
3584 * @param $gender string
3585 * @param $forms array
3589 function gender( $gender, $forms ) {
3590 if ( !count( $forms ) ) {
3593 $forms = $this->preConvertPlural( $forms, 2 );
3594 if ( $gender === 'male' ) {
3597 if ( $gender === 'female' ) {
3600 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3604 * Plural form transformations, needed for some languages.
3605 * For example, there are 3 form of plural in Russian and Polish,
3606 * depending on "count mod 10". See [[w:Plural]]
3607 * For English it is pretty simple.
3609 * Invoked by putting {{plural:count|wordform1|wordform2}}
3610 * or {{plural:count|wordform1|wordform2|wordform3}}
3612 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3614 * @param $count Integer: non-localized number
3615 * @param $forms Array: different plural forms
3616 * @return string Correct form of plural for $count in this language
3618 function convertPlural( $count, $forms ) {
3619 // Handle explicit n=pluralform cases
3620 foreach ( $forms as $index => $form ) {
3621 if ( preg_match( '/^\d+=/i', $form ) ) {
3622 $pos = strpos( $form, '=' );
3623 if ( substr( $form, 0, $pos ) === (string) $count ) {
3624 return substr( $form, $pos +
1 );
3626 unset( $forms[$index] );
3630 $forms = array_values( $forms );
3631 if ( !count( $forms ) ) {
3635 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3636 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3637 return $forms[$pluralForm];
3641 * Checks that convertPlural was given an array and pads it to requested
3642 * amount of forms by copying the last one.
3644 * @param $count Integer: How many forms should there be at least
3645 * @param $forms Array of forms given to convertPlural
3646 * @return array Padded array of forms or an exception if not an array
3648 protected function preConvertPlural( /* Array */ $forms, $count ) {
3649 while ( count( $forms ) < $count ) {
3650 $forms[] = $forms[count( $forms ) - 1];
3656 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3657 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3658 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3659 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3662 * @param $str String: the validated block duration in English
3663 * @return string Somehow translated block duration
3664 * @see LanguageFi.php for example implementation
3666 function translateBlockExpiry( $str ) {
3667 $duration = SpecialBlock
::getSuggestedDurations( $this );
3668 foreach ( $duration as $show => $value ) {
3669 if ( strcmp( $str, $value ) == 0 ) {
3670 return htmlspecialchars( trim( $show ) );
3674 // Since usually only infinite or indefinite is only on list, so try
3675 // equivalents if still here.
3676 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3677 if ( in_array( $str, $indefs ) ) {
3678 foreach ( $indefs as $val ) {
3679 $show = array_search( $val, $duration, true );
3680 if ( $show !== false ) {
3681 return htmlspecialchars( trim( $show ) );
3686 // If all else fails, return a standard duration or timestamp description.
3687 $time = strtotime( $str, 0 );
3688 if ( $time === false ) { // Unknown format. Return it as-is in case.
3690 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3691 // $time is relative to 0 so it's a duration length.
3692 return $this->formatDuration( $time );
3693 } else { // It's an absolute timestamp.
3694 if ( $time === 0 ) {
3695 // wfTimestamp() handles 0 as current time instead of epoch.
3696 return $this->timeanddate( '19700101000000' );
3698 return $this->timeanddate( $time );
3704 * languages like Chinese need to be segmented in order for the diff
3707 * @param $text String
3710 public function segmentForDiff( $text ) {
3715 * and unsegment to show the result
3717 * @param $text String
3720 public function unsegmentForDiff( $text ) {
3725 * Return the LanguageConverter used in the Language
3728 * @return LanguageConverter
3730 public function getConverter() {
3731 return $this->mConverter
;
3735 * convert text to all supported variants
3737 * @param $text string
3740 public function autoConvertToAllVariants( $text ) {
3741 return $this->mConverter
->autoConvertToAllVariants( $text );
3745 * convert text to different variants of a language.
3747 * @param $text string
3750 public function convert( $text ) {
3751 return $this->mConverter
->convert( $text );
3755 * Convert a Title object to a string in the preferred variant
3757 * @param $title Title
3760 public function convertTitle( $title ) {
3761 return $this->mConverter
->convertTitle( $title );
3765 * Convert a namespace index to a string in the preferred variant
3770 public function convertNamespace( $ns ) {
3771 return $this->mConverter
->convertNamespace( $ns );
3775 * Check if this is a language with variants
3779 public function hasVariants() {
3780 return count( $this->getVariants() ) > 1;
3784 * Check if the language has the specific variant
3787 * @param $variant string
3790 public function hasVariant( $variant ) {
3791 return (bool)$this->mConverter
->validateVariant( $variant );
3795 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3797 * @param $text string
3800 public function armourMath( $text ) {
3801 return $this->mConverter
->armourMath( $text );
3805 * Perform output conversion on a string, and encode for safe HTML output.
3806 * @param $text String text to be converted
3807 * @param $isTitle Bool whether this conversion is for the article title
3809 * @todo this should get integrated somewhere sane
3811 public function convertHtml( $text, $isTitle = false ) {
3812 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3816 * @param $key string
3819 public function convertCategoryKey( $key ) {
3820 return $this->mConverter
->convertCategoryKey( $key );
3824 * Get the list of variants supported by this language
3825 * see sample implementation in LanguageZh.php
3827 * @return array an array of language codes
3829 public function getVariants() {
3830 return $this->mConverter
->getVariants();
3836 public function getPreferredVariant() {
3837 return $this->mConverter
->getPreferredVariant();
3843 public function getDefaultVariant() {
3844 return $this->mConverter
->getDefaultVariant();
3850 public function getURLVariant() {
3851 return $this->mConverter
->getURLVariant();
3855 * If a language supports multiple variants, it is
3856 * possible that non-existing link in one variant
3857 * actually exists in another variant. this function
3858 * tries to find it. See e.g. LanguageZh.php
3860 * @param $link String: the name of the link
3861 * @param $nt Mixed: the title object of the link
3862 * @param $ignoreOtherCond Boolean: to disable other conditions when
3863 * we need to transclude a template or update a category's link
3864 * @return null the input parameters may be modified upon return
3866 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3867 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
3871 * If a language supports multiple variants, converts text
3872 * into an array of all possible variants of the text:
3873 * 'variant' => text in that variant
3875 * @deprecated since 1.17 Use autoConvertToAllVariants()
3877 * @param $text string
3881 public function convertLinkToAllVariants( $text ) {
3882 return $this->mConverter
->convertLinkToAllVariants( $text );
3886 * returns language specific options used by User::getPageRenderHash()
3887 * for example, the preferred language variant
3891 function getExtraHashOptions() {
3892 return $this->mConverter
->getExtraHashOptions();
3896 * For languages that support multiple variants, the title of an
3897 * article may be displayed differently in different variants. this
3898 * function returns the apporiate title defined in the body of the article.
3902 public function getParsedTitle() {
3903 return $this->mConverter
->getParsedTitle();
3907 * Prepare external link text for conversion. When the text is
3908 * a URL, it shouldn't be converted, and it'll be wrapped in
3909 * the "raw" tag (-{R| }-) to prevent conversion.
3911 * This function is called "markNoConversion" for historical
3914 * @param $text String: text to be used for external link
3915 * @param $noParse bool: wrap it without confirming it's a real URL first
3916 * @return string the tagged text
3918 public function markNoConversion( $text, $noParse = false ) {
3919 // Excluding protocal-relative URLs may avoid many false positives.
3920 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
3921 return $this->mConverter
->markNoConversion( $text );
3928 * A regular expression to match legal word-trailing characters
3929 * which should be merged onto a link of the form [[foo]]bar.
3933 public function linkTrail() {
3934 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
3940 function getLangObj() {
3945 * Get the "parent" language which has a converter to convert a "compatible" language
3946 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
3948 * @return Language|null
3951 public function getParentLanguage() {
3952 if ( $this->mParentLanguage
!== false ) {
3953 return $this->mParentLanguage
;
3956 $pieces = explode( '-', $this->getCode() );
3958 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
3959 $this->mParentLanguage
= null;
3962 $lang = Language
::factory( $code );
3963 if ( !$lang->hasVariant( $this->getCode() ) ) {
3964 $this->mParentLanguage
= null;
3968 $this->mParentLanguage
= $lang;
3973 * Get the RFC 3066 code for this language object
3975 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3976 * htmlspecialchars() or similar
3980 public function getCode() {
3981 return $this->mCode
;
3985 * Get the code in Bcp47 format which we can use
3986 * inside of html lang="" tags.
3988 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3989 * htmlspecialchars() or similar.
3994 public function getHtmlCode() {
3995 if ( is_null( $this->mHtmlCode
) ) {
3996 $this->mHtmlCode
= wfBCP47( $this->getCode() );
3998 return $this->mHtmlCode
;
4002 * @param $code string
4004 public function setCode( $code ) {
4005 $this->mCode
= $code;
4006 // Ensure we don't leave incorrect cached data lying around
4007 $this->mHtmlCode
= null;
4008 $this->mParentLanguage
= false;
4012 * Get the name of a file for a certain language code
4013 * @param $prefix string Prepend this to the filename
4014 * @param $code string Language code
4015 * @param $suffix string Append this to the filename
4016 * @throws MWException
4017 * @return string $prefix . $mangledCode . $suffix
4019 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4020 // Protect against path traversal
4021 if ( !Language
::isValidCode( $code )
4022 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
4024 throw new MWException( "Invalid language code \"$code\"" );
4027 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4031 * Get the language code from a file name. Inverse of getFileName()
4032 * @param $filename string $prefix . $languageCode . $suffix
4033 * @param $prefix string Prefix before the language code
4034 * @param $suffix string Suffix after the language code
4035 * @return string Language code, or false if $prefix or $suffix isn't found
4037 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4039 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4040 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4041 if ( !count( $m ) ) {
4044 return str_replace( '_', '-', strtolower( $m[1] ) );
4048 * @param $code string
4051 public static function getMessagesFileName( $code ) {
4053 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4054 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
4059 * @param $code string
4062 public static function getClassFileName( $code ) {
4064 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4068 * Get the first fallback for a given language.
4070 * @param $code string
4072 * @return bool|string
4074 public static function getFallbackFor( $code ) {
4075 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4078 $fallbacks = self
::getFallbacksFor( $code );
4079 $first = array_shift( $fallbacks );
4085 * Get the ordered list of fallback languages.
4088 * @param $code string Language code
4091 public static function getFallbacksFor( $code ) {
4092 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4095 $v = self
::getLocalisationCache()->getItem( $code, 'fallback' );
4096 $v = array_map( 'trim', explode( ',', $v ) );
4097 if ( $v[count( $v ) - 1] !== 'en' ) {
4105 * Get the ordered list of fallback languages, ending with the fallback
4106 * language chain for the site language.
4109 * @param string $code Language code
4110 * @return array array( fallbacks, site fallbacks )
4112 public static function getFallbacksIncludingSiteLanguage( $code ) {
4113 global $wgLanguageCode;
4115 // Usually, we will only store a tiny number of fallback chains, so we
4116 // keep them in static memory.
4117 $cacheKey = "{$code}-{$wgLanguageCode}";
4119 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4120 $fallbacks = self
::getFallbacksFor( $code );
4122 // Append the site's fallback chain, including the site language itself
4123 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4124 array_unshift( $siteFallbacks, $wgLanguageCode );
4126 // Eliminate any languages already included in the chain
4127 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4129 self
::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4131 return self
::$fallbackLanguageCache[$cacheKey];
4135 * Get all messages for a given language
4136 * WARNING: this may take a long time. If you just need all message *keys*
4137 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4139 * @param $code string
4143 public static function getMessagesFor( $code ) {
4144 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4148 * Get a message for a given language
4150 * @param $key string
4151 * @param $code string
4155 public static function getMessageFor( $key, $code ) {
4156 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4160 * Get all message keys for a given language. This is a faster alternative to
4161 * array_keys( Language::getMessagesFor( $code ) )
4164 * @param $code string Language code
4165 * @return array of message keys (strings)
4167 public static function getMessageKeysFor( $code ) {
4168 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4175 function fixVariableInNamespace( $talk ) {
4176 if ( strpos( $talk, '$1' ) === false ) {
4180 global $wgMetaNamespace;
4181 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4183 # Allow grammar transformations
4184 # Allowing full message-style parsing would make simple requests
4185 # such as action=raw much more expensive than they need to be.
4186 # This will hopefully cover most cases.
4187 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4188 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4189 return str_replace( ' ', '_', $talk );
4196 function replaceGrammarInNamespace( $m ) {
4197 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4201 * @throws MWException
4204 static function getCaseMaps() {
4205 static $wikiUpperChars, $wikiLowerChars;
4206 if ( isset( $wikiUpperChars ) ) {
4207 return array( $wikiUpperChars, $wikiLowerChars );
4210 wfProfileIn( __METHOD__
);
4211 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4212 if ( $arr === false ) {
4213 throw new MWException(
4214 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4216 $wikiUpperChars = $arr['wikiUpperChars'];
4217 $wikiLowerChars = $arr['wikiLowerChars'];
4218 wfProfileOut( __METHOD__
);
4219 return array( $wikiUpperChars, $wikiLowerChars );
4223 * Decode an expiry (block, protection, etc) which has come from the DB
4225 * @todo FIXME: why are we returnings DBMS-dependent strings???
4227 * @param $expiry String: Database expiry String
4228 * @param $format Bool|Int true to process using language functions, or TS_ constant
4229 * to return the expiry in a given timestamp
4233 public function formatExpiry( $expiry, $format = true ) {
4235 if ( $infinity === null ) {
4236 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
4239 if ( $expiry == '' ||
$expiry == $infinity ) {
4240 return $format === true
4241 ?
$this->getMessageFromDB( 'infiniteblock' )
4244 return $format === true
4245 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4246 : wfTimestamp( $format, $expiry );
4252 * @param $seconds int|float
4253 * @param $format Array Optional
4254 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
4255 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
4256 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
4257 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
4260 function formatTimePeriod( $seconds, $format = array() ) {
4261 if ( !is_array( $format ) ) {
4262 $format = array( 'avoid' => $format ); // For backwards compatibility
4264 if ( !isset( $format['avoid'] ) ) {
4265 $format['avoid'] = false;
4267 if ( !isset( $format['noabbrevs' ] ) ) {
4268 $format['noabbrevs'] = false;
4270 $secondsMsg = wfMessage(
4271 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4272 $minutesMsg = wfMessage(
4273 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4274 $hoursMsg = wfMessage(
4275 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4276 $daysMsg = wfMessage(
4277 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4279 if ( round( $seconds * 10 ) < 100 ) {
4280 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4281 $s = $secondsMsg->params( $s )->text();
4282 } elseif ( round( $seconds ) < 60 ) {
4283 $s = $this->formatNum( round( $seconds ) );
4284 $s = $secondsMsg->params( $s )->text();
4285 } elseif ( round( $seconds ) < 3600 ) {
4286 $minutes = floor( $seconds / 60 );
4287 $secondsPart = round( fmod( $seconds, 60 ) );
4288 if ( $secondsPart == 60 ) {
4292 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4294 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4295 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4296 $hours = floor( $seconds / 3600 );
4297 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4298 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4299 if ( $secondsPart == 60 ) {
4303 if ( $minutes == 60 ) {
4307 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4309 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4310 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4311 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4314 $days = floor( $seconds / 86400 );
4315 if ( $format['avoid'] === 'avoidminutes' ) {
4316 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4317 if ( $hours == 24 ) {
4321 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4323 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4324 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4325 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4326 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4327 if ( $minutes == 60 ) {
4331 if ( $hours == 24 ) {
4335 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4337 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4339 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4341 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4343 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4350 * Format a bitrate for output, using an appropriate
4351 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4353 * This use base 1000. For base 1024 use formatSize(), for another base
4354 * see formatComputingNumbers()
4359 function formatBitrate( $bps ) {
4360 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4364 * @param $size int Size of the unit
4365 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4366 * @param $messageKey string Message key to be uesd
4369 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4371 return str_replace( '$1', $this->formatNum( $size ),
4372 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4375 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4378 $maxIndex = count( $sizes ) - 1;
4379 while ( $size >= $boundary && $index < $maxIndex ) {
4384 // For small sizes no decimal places necessary
4387 // For MB and bigger two decimal places are smarter
4390 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4392 $size = round( $size, $round );
4393 $text = $this->getMessageFromDB( $msg );
4394 return str_replace( '$1', $this->formatNum( $size ), $text );
4398 * Format a size in bytes for output, using an appropriate
4399 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4401 * This method use base 1024. For base 1000 use formatBitrate(), for
4402 * another base see formatComputingNumbers()
4404 * @param $size int Size to format
4405 * @return string Plain text (not HTML)
4407 function formatSize( $size ) {
4408 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4412 * Make a list item, used by various special pages
4414 * @param $page String Page link
4415 * @param $details String Text between brackets
4416 * @param $oppositedm Boolean Add the direction mark opposite to your
4417 * language, to display text properly
4420 function specialList( $page, $details, $oppositedm = true ) {
4421 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) .
4422 $this->getDirMark();
4423 $details = $details ?
$dirmark . $this->getMessageFromDB( 'word-separator' ) .
4424 wfMessage( 'parentheses' )->rawParams( $details )->inLanguage( $this )->escaped() : '';
4425 return $page . $details;
4429 * Generate (prev x| next x) (20|50|100...) type links for paging
4431 * @param $title Title object to link
4432 * @param $offset Integer offset parameter
4433 * @param $limit Integer limit parameter
4434 * @param $query array|String optional URL query parameter string
4435 * @param $atend Bool optional param for specified if this is the last page
4438 public function viewPrevNext( Title
$title, $offset, $limit, array $query = array(), $atend = false ) {
4439 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4441 # Make 'previous' link
4442 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4443 if ( $offset > 0 ) {
4444 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4445 $query, $prev, 'prevn-title', 'mw-prevlink' );
4447 $plink = htmlspecialchars( $prev );
4451 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4453 $nlink = htmlspecialchars( $next );
4455 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4456 $query, $next, 'prevn-title', 'mw-nextlink' );
4459 # Make links to set number of items per page
4460 $numLinks = array();
4461 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4462 $numLinks[] = $this->numLink( $title, $offset, $num,
4463 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4466 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4467 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4471 * Helper function for viewPrevNext() that generates links
4473 * @param $title Title object to link
4474 * @param $offset Integer offset parameter
4475 * @param $limit Integer limit parameter
4476 * @param $query Array extra query parameters
4477 * @param $link String text to use for the link; will be escaped
4478 * @param $tooltipMsg String name of the message to use as tooltip
4479 * @param $class String value of the "class" attribute of the link
4480 * @return String HTML fragment
4482 private function numLink( Title
$title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4483 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4484 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4485 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4486 'title' => $tooltip, 'class' => $class ), $link );
4490 * Get the conversion rule title, if any.
4494 public function getConvRuleTitle() {
4495 return $this->mConverter
->getConvRuleTitle();
4499 * Get the compiled plural rules for the language
4501 * @return array Associative array with plural form, and plural rule as key-value pairs
4503 public function getCompiledPluralRules() {
4504 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4505 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4506 if ( !$pluralRules ) {
4507 foreach ( $fallbacks as $fallbackCode ) {
4508 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4509 if ( $pluralRules ) {
4514 return $pluralRules;
4518 * Get the plural rules for the language
4520 * @return array Associative array with plural form number and plural rule as key-value pairs
4522 public function getPluralRules() {
4523 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4524 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4525 if ( !$pluralRules ) {
4526 foreach ( $fallbacks as $fallbackCode ) {
4527 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4528 if ( $pluralRules ) {
4533 return $pluralRules;
4537 * Get the plural rule types for the language
4539 * @return array Associative array with plural form number and plural rule type as key-value pairs
4541 public function getPluralRuleTypes() {
4542 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4543 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4544 if ( !$pluralRuleTypes ) {
4545 foreach ( $fallbacks as $fallbackCode ) {
4546 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4547 if ( $pluralRuleTypes ) {
4552 return $pluralRuleTypes;
4556 * Find the index number of the plural rule appropriate for the given number
4557 * @return int The index number of the plural rule
4559 public function getPluralRuleIndexNumber( $number ) {
4560 $pluralRules = $this->getCompiledPluralRules();
4561 $form = CLDRPluralRuleEvaluator
::evaluateCompiled( $number, $pluralRules );
4566 * Find the plural rule type appropriate for the given number
4567 * For example, if the language is set to Arabic, getPluralType(5) should
4570 * @return string The name of the plural rule type, e.g. one, two, few, many
4572 public function getPluralRuleType( $number ) {
4573 $index = $this->getPluralRuleIndexNumber( $number );
4574 $pluralRuleTypes = $this->getPluralRuleTypes();
4575 if ( isset( $pluralRuleTypes[$index] ) ) {
4576 return $pluralRuleTypes[$index];