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";
33 if ( function_exists( 'mb_strtoupper' ) ) {
34 mb_internal_encoding( 'UTF-8' );
38 * Internationalisation code
43 * @var LanguageConverter
47 public $mVariants, $mCode, $mLoaded = false;
48 public $mMagicExtensions = array(), $mMagicHookDone = false;
49 private $mHtmlCode = null, $mParentLanguage = false;
51 public $dateFormatStrings = array();
52 public $mExtendedSpecialPageAliases;
54 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
57 * ReplacementArray object caches
59 public $transformData = array();
62 * @var LocalisationCache
64 static public $dataCache;
66 static public $mLangObjCache = array();
68 static public $mWeekdayMsgs = array(
69 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
73 static public $mWeekdayAbbrevMsgs = array(
74 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
77 static public $mMonthMsgs = array(
78 'january', 'february', 'march', 'april', 'may_long', 'june',
79 'july', 'august', 'september', 'october', 'november',
82 static public $mMonthGenMsgs = array(
83 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
84 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
87 static public $mMonthAbbrevMsgs = array(
88 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
89 'sep', 'oct', 'nov', 'dec'
92 static public $mIranianCalendarMonthMsgs = array(
93 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
94 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
95 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
96 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
99 static public $mHebrewCalendarMonthMsgs = array(
100 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
101 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
102 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
103 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
104 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
107 static public $mHebrewCalendarMonthGenMsgs = array(
108 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
109 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
110 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
111 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
112 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
115 static public $mHijriCalendarMonthMsgs = array(
116 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
117 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
118 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
119 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
126 static public $durationIntervals = array(
127 'millennia' => 31556952000,
128 'centuries' => 3155695200,
129 'decades' => 315569520,
130 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
139 * Cache for language fallbacks.
140 * @see Language::getFallbacksIncludingSiteLanguage
144 static private $fallbackLanguageCache = array();
147 * Cache for language names
148 * @var MapCacheLRU|null
150 static private $languageNameCache;
153 * Get a cached or new language object for a given language code
154 * @param string $code
157 static function factory( $code ) {
158 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
160 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
161 $code = $wgDummyLanguageCodes[$code];
164 // get the language object to process
165 $langObj = isset( self
::$mLangObjCache[$code] )
166 ? self
::$mLangObjCache[$code]
167 : self
::newFromCode( $code );
169 // merge the language object in to get it up front in the cache
170 self
::$mLangObjCache = array_merge( array( $code => $langObj ), self
::$mLangObjCache );
171 // get rid of the oldest ones in case we have an overflow
172 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
178 * Create a language object for a given language code
179 * @param string $code
180 * @throws MWException
183 protected static function newFromCode( $code ) {
184 // Protect against path traversal below
185 if ( !Language
::isValidCode( $code )
186 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code )
188 throw new MWException( "Invalid language code \"$code\"" );
191 if ( !Language
::isValidBuiltInCode( $code ) ) {
192 // It's not possible to customise this code with class files, so
193 // just return a Language object. This is to support uselang= hacks.
194 $lang = new Language
;
195 $lang->setCode( $code );
199 // Check if there is a language class for the code
200 $class = self
::classFromCode( $code );
201 self
::preloadLanguageClass( $class );
202 if ( class_exists( $class ) ) {
207 // Keep trying the fallback list until we find an existing class
208 $fallbacks = Language
::getFallbacksFor( $code );
209 foreach ( $fallbacks as $fallbackCode ) {
210 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
211 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
214 $class = self
::classFromCode( $fallbackCode );
215 self
::preloadLanguageClass( $class );
216 if ( class_exists( $class ) ) {
217 $lang = Language
::newFromCode( $fallbackCode );
218 $lang->setCode( $code );
223 throw new MWException( "Invalid fallback sequence for language '$code'" );
227 * Checks whether any localisation is available for that language tag
228 * in MediaWiki (MessagesXx.php exists).
230 * @param string $code Language tag (in lower case)
231 * @return bool Whether language is supported
234 public static function isSupportedLanguage( $code ) {
235 return self
::isValidBuiltInCode( $code )
236 && ( is_readable( self
::getMessagesFileName( $code ) )
237 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
242 * Returns true if a language code string is a well-formed language tag
243 * according to RFC 5646.
244 * This function only checks well-formedness; it doesn't check that
245 * language, script or variant codes actually exist in the repositories.
247 * Based on regexes by Mark Davis of the Unicode Consortium:
248 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
250 * @param string $code
251 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
256 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
259 $alphanum = '[a-z0-9]';
260 $x = 'x'; # private use singleton
261 $singleton = '[a-wy-z]'; # other singleton
262 $s = $lenient ?
'[-_]' : '-';
264 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
265 $script = "$alpha{4}"; # ISO 15924
266 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
267 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
268 $extension = "$singleton(?:$s$alphanum{2,8})+";
269 $privateUse = "$x(?:$s$alphanum{1,8})+";
271 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
272 # Since these are limited, this is safe even later changes to the registry --
273 # the only oddity is that it might change the type of the tag, and thus
274 # the results from the capturing groups.
275 # http://www.iana.org/assignments/language-subtag-registry
277 $grandfathered = "en{$s}GB{$s}oed"
278 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
279 . "|no{$s}(?:bok|nyn)"
280 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
281 . "|zh{$s}min{$s}nan";
283 $variantList = "$variant(?:$s$variant)*";
284 $extensionList = "$extension(?:$s$extension)*";
286 $langtag = "(?:($language)"
289 . "(?:$s$variantList)?"
290 . "(?:$s$extensionList)?"
291 . "(?:$s$privateUse)?)";
293 # The final breakdown, with capturing groups for each of these components
294 # The variants, extensions, grandfathered, and private-use may have interior '-'
296 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
298 return (bool)preg_match( "/$root/", strtolower( $code ) );
302 * Returns true if a language code string is of a valid form, whether or
303 * not it exists. This includes codes which are used solely for
304 * customisation via the MediaWiki namespace.
306 * @param string $code
310 public static function isValidCode( $code ) {
311 static $cache = array();
312 if ( isset( $cache[$code] ) ) {
313 return $cache[$code];
315 // People think language codes are html safe, so enforce it.
316 // Ideally we should only allow a-zA-Z0-9-
317 // but, .+ and other chars are often used for {{int:}} hacks
318 // see bugs 37564, 37587, 36938
320 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
321 && !preg_match( Title
::getTitleInvalidRegex(), $code );
323 return $cache[$code];
327 * Returns true if a language code is of a valid form for the purposes of
328 * internal customisation of MediaWiki, via Messages*.php or *.json.
330 * @param string $code
332 * @throws MWException
336 public static function isValidBuiltInCode( $code ) {
338 if ( !is_string( $code ) ) {
339 if ( is_object( $code ) ) {
340 $addmsg = " of class " . get_class( $code );
344 $type = gettype( $code );
345 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
348 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
352 * Returns true if a language code is an IETF tag known to MediaWiki.
359 public static function isKnownLanguageTag( $tag ) {
360 static $coreLanguageNames;
362 // Quick escape for invalid input to avoid exceptions down the line
363 // when code tries to process tags which are not valid at all.
364 if ( !self
::isValidBuiltInCode( $tag ) ) {
368 if ( $coreLanguageNames === null ) {
370 include "$IP/languages/Names.php";
373 if ( isset( $coreLanguageNames[$tag] )
374 || self
::fetchLanguageName( $tag, $tag ) !== ''
383 * @param string $code
384 * @return string Name of the language class
386 public static function classFromCode( $code ) {
387 if ( $code == 'en' ) {
390 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
395 * Includes language class files
397 * @param string $class Name of the language class
399 public static function preloadLanguageClass( $class ) {
402 if ( $class === 'Language' ) {
406 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
407 include_once "$IP/languages/classes/$class.php";
412 * Get the LocalisationCache instance
414 * @return LocalisationCache
416 public static function getLocalisationCache() {
417 if ( is_null( self
::$dataCache ) ) {
418 global $wgLocalisationCacheConf;
419 $class = $wgLocalisationCacheConf['class'];
420 self
::$dataCache = new $class( $wgLocalisationCacheConf );
422 return self
::$dataCache;
425 function __construct() {
426 $this->mConverter
= new FakeConverter( $this );
427 // Set the code to the name of the descendant
428 if ( get_class( $this ) == 'Language' ) {
431 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
433 self
::getLocalisationCache();
437 * Reduce memory usage
439 function __destruct() {
440 foreach ( $this as $name => $value ) {
441 unset( $this->$name );
446 * Hook which will be called if this is the content language.
447 * Descendants can use this to register hook functions or modify globals
449 function initContLang() {
456 function getFallbackLanguages() {
457 return self
::getFallbacksFor( $this->mCode
);
461 * Exports $wgBookstoreListEn
464 function getBookstoreList() {
465 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
469 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
470 * available in localised form, it will be included in English.
474 public function getNamespaces() {
475 if ( is_null( $this->namespaceNames
) ) {
476 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
478 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
479 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
481 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
483 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
484 if ( $wgMetaNamespaceTalk ) {
485 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
487 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
488 $this->namespaceNames
[NS_PROJECT_TALK
] =
489 $this->fixVariableInNamespace( $talk );
492 # Sometimes a language will be localised but not actually exist on this wiki.
493 foreach ( $this->namespaceNames
as $key => $text ) {
494 if ( !isset( $validNamespaces[$key] ) ) {
495 unset( $this->namespaceNames
[$key] );
499 # The above mixing may leave namespaces out of canonical order.
500 # Re-order by namespace ID number...
501 ksort( $this->namespaceNames
);
503 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
506 return $this->namespaceNames
;
510 * Arbitrarily set all of the namespace names at once. Mainly used for testing
511 * @param array $namespaces Array of namespaces (id => name)
513 public function setNamespaces( array $namespaces ) {
514 $this->namespaceNames
= $namespaces;
515 $this->mNamespaceIds
= null;
519 * Resets all of the namespace caches. Mainly used for testing
521 public function resetNamespaces() {
522 $this->namespaceNames
= null;
523 $this->mNamespaceIds
= null;
524 $this->namespaceAliases
= null;
528 * A convenience function that returns the same thing as
529 * getNamespaces() except with the array values changed to ' '
530 * where it found '_', useful for producing output to be displayed
531 * e.g. in <select> forms.
535 function getFormattedNamespaces() {
536 $ns = $this->getNamespaces();
537 foreach ( $ns as $k => $v ) {
538 $ns[$k] = strtr( $v, '_', ' ' );
544 * Get a namespace value by key
546 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
547 * echo $mw_ns; // prints 'MediaWiki'
550 * @param int $index The array key of the namespace to return
551 * @return string|bool String if the namespace value exists, otherwise false
553 function getNsText( $index ) {
554 $ns = $this->getNamespaces();
556 return isset( $ns[$index] ) ?
$ns[$index] : false;
560 * A convenience function that returns the same thing as
561 * getNsText() except with '_' changed to ' ', useful for
565 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
566 * echo $mw_ns; // prints 'MediaWiki talk'
569 * @param int $index The array key of the namespace to return
570 * @return string Namespace name without underscores (empty string if namespace does not exist)
572 function getFormattedNsText( $index ) {
573 $ns = $this->getNsText( $index );
575 return strtr( $ns, '_', ' ' );
579 * Returns gender-dependent namespace alias if available.
580 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
581 * @param int $index Namespace index
582 * @param string $gender Gender key (male, female... )
586 function getGenderNsText( $index, $gender ) {
587 global $wgExtraGenderNamespaces;
589 $ns = $wgExtraGenderNamespaces +
590 self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
592 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
596 * Whether this language uses gender-dependent namespace aliases.
597 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
601 function needsGenderDistinction() {
602 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
603 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
604 // $wgExtraGenderNamespaces overrides everything
606 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
607 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
608 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
611 // Check what is in i18n files
612 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
613 return count( $aliases ) > 0;
618 * Get a namespace key by value, case insensitive.
619 * Only matches namespace names for the current language, not the
620 * canonical ones defined in Namespace.php.
622 * @param string $text
623 * @return int|bool An integer if $text is a valid value otherwise false
625 function getLocalNsIndex( $text ) {
626 $lctext = $this->lc( $text );
627 $ids = $this->getNamespaceIds();
628 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
634 function getNamespaceAliases() {
635 if ( is_null( $this->namespaceAliases
) ) {
636 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
640 foreach ( $aliases as $name => $index ) {
641 if ( $index === NS_PROJECT_TALK
) {
642 unset( $aliases[$name] );
643 $name = $this->fixVariableInNamespace( $name );
644 $aliases[$name] = $index;
649 global $wgExtraGenderNamespaces;
650 $genders = $wgExtraGenderNamespaces +
651 (array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
652 foreach ( $genders as $index => $forms ) {
653 foreach ( $forms as $alias ) {
654 $aliases[$alias] = $index;
658 # Also add converted namespace names as aliases, to avoid confusion.
659 $convertedNames = array();
660 foreach ( $this->getVariants() as $variant ) {
661 if ( $variant === $this->mCode
) {
664 foreach ( $this->getNamespaces() as $ns => $_ ) {
665 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
669 $this->namespaceAliases
= $aliases +
$convertedNames;
672 return $this->namespaceAliases
;
678 function getNamespaceIds() {
679 if ( is_null( $this->mNamespaceIds
) ) {
680 global $wgNamespaceAliases;
681 # Put namespace names and aliases into a hashtable.
682 # If this is too slow, then we should arrange it so that it is done
683 # before caching. The catch is that at pre-cache time, the above
684 # class-specific fixup hasn't been done.
685 $this->mNamespaceIds
= array();
686 foreach ( $this->getNamespaces() as $index => $name ) {
687 $this->mNamespaceIds
[$this->lc( $name )] = $index;
689 foreach ( $this->getNamespaceAliases() as $name => $index ) {
690 $this->mNamespaceIds
[$this->lc( $name )] = $index;
692 if ( $wgNamespaceAliases ) {
693 foreach ( $wgNamespaceAliases as $name => $index ) {
694 $this->mNamespaceIds
[$this->lc( $name )] = $index;
698 return $this->mNamespaceIds
;
702 * Get a namespace key by value, case insensitive. Canonical namespace
703 * names override custom ones defined for the current language.
705 * @param string $text
706 * @return int|bool An integer if $text is a valid value otherwise false
708 function getNsIndex( $text ) {
709 $lctext = $this->lc( $text );
710 $ns = MWNamespace
::getCanonicalIndex( $lctext );
711 if ( $ns !== null ) {
714 $ids = $this->getNamespaceIds();
715 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
719 * short names for language variants used for language conversion links.
721 * @param string $code
722 * @param bool $usemsg Use the "variantname-xyz" message if it exists
725 function getVariantname( $code, $usemsg = true ) {
726 $msg = "variantname-$code";
727 if ( $usemsg && wfMessage( $msg )->exists() ) {
728 return $this->getMessageFromDB( $msg );
730 $name = self
::fetchLanguageName( $code );
732 return $name; # if it's defined as a language name, show that
734 # otherwise, output the language code
740 * @param string $name
743 function specialPage( $name ) {
744 $aliases = $this->getSpecialPageAliases();
745 if ( isset( $aliases[$name][0] ) ) {
746 $name = $aliases[$name][0];
748 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
754 function getDatePreferences() {
755 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
761 function getDateFormats() {
762 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
766 * @return array|string
768 function getDefaultDateFormat() {
769 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
770 if ( $df === 'dmy or mdy' ) {
771 global $wgAmericanDates;
772 return $wgAmericanDates ?
'mdy' : 'dmy';
781 function getDatePreferenceMigrationMap() {
782 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
786 * @param string $image
789 function getImageFile( $image ) {
790 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
797 function getImageFiles() {
798 return self
::$dataCache->getItem( $this->mCode
, 'imageFiles' );
804 function getExtraUserToggles() {
805 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
812 function getUserToggle( $tog ) {
813 return $this->getMessageFromDB( "tog-$tog" );
817 * Get native language names, indexed by code.
818 * Only those defined in MediaWiki, no other data like CLDR.
819 * If $customisedOnly is true, only returns codes with a messages file
821 * @param bool $customisedOnly
824 * @deprecated since 1.20, use fetchLanguageNames()
826 public static function getLanguageNames( $customisedOnly = false ) {
827 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
831 * Get translated language names. This is done on best effort and
832 * by default this is exactly the same as Language::getLanguageNames.
833 * The CLDR extension provides translated names.
834 * @param string $code Language code.
835 * @return array Language code => language name
837 * @deprecated since 1.20, use fetchLanguageNames()
839 public static function getTranslatedLanguageNames( $code ) {
840 return self
::fetchLanguageNames( $code, 'all' );
844 * Get an array of language names, indexed by code.
845 * @param null|string $inLanguage Code of language in which to return the names
846 * Use null for autonyms (native names)
847 * @param string $include One of:
848 * 'all' all available languages
849 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
850 * 'mwfile' only if the language is in 'mw' *and* has a message file
851 * @return array Language code => language name
854 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
855 wfProfileIn( __METHOD__
);
856 $cacheKey = $inLanguage === null ?
'null' : $inLanguage;
857 $cacheKey .= ":$include";
858 if ( self
::$languageNameCache === null ) {
859 self
::$languageNameCache = new MapCacheLRU( 20 );
861 if ( self
::$languageNameCache->has( $cacheKey ) ) {
862 $ret = self
::$languageNameCache->get( $cacheKey );
864 $ret = self
::fetchLanguageNamesUncached( $inLanguage, $include );
865 self
::$languageNameCache->set( $cacheKey, $ret );
867 wfProfileOut( __METHOD__
);
872 * Uncached helper for fetchLanguageNames
873 * @param null|string $inLanguage Code of language in which to return the names
874 * Use null for autonyms (native names)
875 * @param string $include One of:
876 * 'all' all available languages
877 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
878 * 'mwfile' only if the language is in 'mw' *and* has a message file
879 * @return array Language code => language name
881 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
882 global $wgExtraLanguageNames;
883 static $coreLanguageNames;
885 if ( $coreLanguageNames === null ) {
887 include "$IP/languages/Names.php";
890 // If passed an invalid language code to use, fallback to en
891 if ( $inLanguage !== null && !Language
::isValidCode( $inLanguage ) ) {
898 # TODO: also include when $inLanguage is null, when this code is more efficient
899 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
902 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
903 foreach ( $mwNames as $mwCode => $mwName ) {
904 # - Prefer own MediaWiki native name when not using the hook
905 # - For other names just add if not added through the hook
906 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
907 $names[$mwCode] = $mwName;
911 if ( $include === 'all' ) {
916 $coreCodes = array_keys( $mwNames );
917 foreach ( $coreCodes as $coreCode ) {
918 $returnMw[$coreCode] = $names[$coreCode];
921 if ( $include === 'mwfile' ) {
922 $namesMwFile = array();
923 # We do this using a foreach over the codes instead of a directory
924 # loop so that messages files in extensions will work correctly.
925 foreach ( $returnMw as $code => $value ) {
926 if ( is_readable( self
::getMessagesFileName( $code ) )
927 ||
is_readable( self
::getJsonMessagesFileName( $code ) )
929 $namesMwFile[$code] = $names[$code];
936 # 'mw' option; default if it's not one of the other two options (all/mwfile)
941 * @param string $code The code of the language for which to get the name
942 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
943 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
944 * @return string Language name or empty
947 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
948 $code = strtolower( $code );
949 $array = self
::fetchLanguageNames( $inLanguage, $include );
950 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
954 * Get a message from the MediaWiki namespace.
956 * @param string $msg Message name
959 function getMessageFromDB( $msg ) {
960 return wfMessage( $msg )->inLanguage( $this )->text();
964 * Get the native language name of $code.
965 * Only if defined in MediaWiki, no other data like CLDR.
966 * @param string $code
968 * @deprecated since 1.20, use fetchLanguageName()
970 function getLanguageName( $code ) {
971 return self
::fetchLanguageName( $code );
978 function getMonthName( $key ) {
979 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
985 function getMonthNamesArray() {
986 $monthNames = array( '' );
987 for ( $i = 1; $i < 13; $i++
) {
988 $monthNames[] = $this->getMonthName( $i );
997 function getMonthNameGen( $key ) {
998 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
1002 * @param string $key
1005 function getMonthAbbreviation( $key ) {
1006 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
1012 function getMonthAbbreviationsArray() {
1013 $monthNames = array( '' );
1014 for ( $i = 1; $i < 13; $i++
) {
1015 $monthNames[] = $this->getMonthAbbreviation( $i );
1021 * @param string $key
1024 function getWeekdayName( $key ) {
1025 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1029 * @param string $key
1032 function getWeekdayAbbreviation( $key ) {
1033 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1037 * @param string $key
1040 function getIranianCalendarMonthName( $key ) {
1041 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1045 * @param string $key
1048 function getHebrewCalendarMonthName( $key ) {
1049 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1053 * @param string $key
1056 function getHebrewCalendarMonthNameGen( $key ) {
1057 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1061 * @param string $key
1064 function getHijriCalendarMonthName( $key ) {
1065 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1069 * Pass through result from $dateTimeObj->format()
1070 * @param DateTime|bool|null &$dateTimeObj
1072 * @param DateTimeZone|bool|null $zone
1073 * @param string $code
1076 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1077 if ( !$dateTimeObj ) {
1078 $dateTimeObj = DateTime
::createFromFormat(
1079 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1082 return $dateTimeObj->format( $code );
1086 * This is a workalike of PHP's date() function, but with better
1087 * internationalisation, a reduced set of format characters, and a better
1090 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1091 * the PHP manual for definitions. There are a number of extensions, which
1094 * xn Do not translate digits of the next numeric format character
1095 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1096 * xr Use roman numerals for the next numeric format character
1097 * xh Use hebrew numerals for the next numeric format character
1099 * xg Genitive month name
1101 * xij j (day number) in Iranian calendar
1102 * xiF F (month name) in Iranian calendar
1103 * xin n (month number) in Iranian calendar
1104 * xiy y (two digit year) in Iranian calendar
1105 * xiY Y (full year) in Iranian calendar
1107 * xjj j (day number) in Hebrew calendar
1108 * xjF F (month name) in Hebrew calendar
1109 * xjt t (days in month) in Hebrew calendar
1110 * xjx xg (genitive month name) in Hebrew calendar
1111 * xjn n (month number) in Hebrew calendar
1112 * xjY Y (full year) in Hebrew calendar
1114 * xmj j (day number) in Hijri calendar
1115 * xmF F (month name) in Hijri calendar
1116 * xmn n (month number) in Hijri calendar
1117 * xmY Y (full year) in Hijri calendar
1119 * xkY Y (full year) in Thai solar calendar. Months and days are
1120 * identical to the Gregorian calendar
1121 * xoY Y (full year) in Minguo calendar or Juche year.
1122 * Months and days are identical to the
1123 * Gregorian calendar
1124 * xtY Y (full year) in Japanese nengo. Months and days are
1125 * identical to the Gregorian calendar
1127 * Characters enclosed in double quotes will be considered literal (with
1128 * the quotes themselves removed). Unmatched quotes will be considered
1129 * literal quotes. Example:
1131 * "The month is" F => The month is January
1134 * Backslash escaping is also supported.
1136 * Input timestamp is assumed to be pre-normalized to the desired local
1137 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1138 * $ts is UTC if $zone is not given.
1140 * @param string $format
1141 * @param string $ts 14-character timestamp
1144 * @param DateTimeZone $zone Timezone of $ts
1145 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1146 * Only makes sense if $ts is the current time.
1147 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1149 * @throws MWException
1152 function sprintfDate( $format, $ts, DateTimeZone
$zone = null, &$ttl = null ) {
1157 $dateTimeObj = false;
1166 $usedSecond = false;
1167 $usedMinute = false;
1174 $usedISOYear = false;
1175 $usedIsLeapYear = false;
1177 $usedHebrewMonth = false;
1178 $usedIranianMonth = false;
1179 $usedHijriMonth = false;
1180 $usedHebrewYear = false;
1181 $usedIranianYear = false;
1182 $usedHijriYear = false;
1183 $usedTennoYear = false;
1185 if ( strlen( $ts ) !== 14 ) {
1186 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1189 if ( !ctype_digit( $ts ) ) {
1190 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1193 $formatLength = strlen( $format );
1194 for ( $p = 0; $p < $formatLength; $p++
) {
1196 $code = $format[$p];
1197 if ( $code == 'x' && $p < $formatLength - 1 ) {
1198 $code .= $format[++
$p];
1201 if ( ( $code === 'xi'
1207 && $p < $formatLength - 1 ) {
1208 $code .= $format[++
$p];
1219 $rawToggle = !$rawToggle;
1229 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1232 $usedHebrewMonth = true;
1234 $hebrew = self
::tsToHebrew( $ts );
1236 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1240 $num = substr( $ts, 6, 2 );
1244 $s .= $this->getWeekdayAbbreviation( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1 );
1248 $num = intval( substr( $ts, 6, 2 ) );
1253 $iranian = self
::tsToIranian( $ts );
1260 $hijri = self
::tsToHijri( $ts );
1267 $hebrew = self
::tsToHebrew( $ts );
1273 $s .= $this->getWeekdayName( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) +
1 );
1277 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1280 $usedIranianMonth = true;
1282 $iranian = self
::tsToIranian( $ts );
1284 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1287 $usedHijriMonth = true;
1289 $hijri = self
::tsToHijri( $ts );
1291 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1294 $usedHebrewMonth = true;
1296 $hebrew = self
::tsToHebrew( $ts );
1298 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1302 $num = substr( $ts, 4, 2 );
1306 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1310 $num = intval( substr( $ts, 4, 2 ) );
1313 $usedIranianMonth = true;
1315 $iranian = self
::tsToIranian( $ts );
1320 $usedHijriMonth = true;
1322 $hijri = self
::tsToHijri ( $ts );
1327 $usedHebrewMonth = true;
1329 $hebrew = self
::tsToHebrew( $ts );
1334 $usedHebrewMonth = true;
1336 $hebrew = self
::tsToHebrew( $ts );
1342 $num = substr( $ts, 0, 4 );
1345 $usedIranianYear = true;
1347 $iranian = self
::tsToIranian( $ts );
1352 $usedHijriYear = true;
1354 $hijri = self
::tsToHijri( $ts );
1359 $usedHebrewYear = true;
1361 $hebrew = self
::tsToHebrew( $ts );
1368 $thai = self
::tsToYear( $ts, 'thai' );
1375 $minguo = self
::tsToYear( $ts, 'minguo' );
1380 $usedTennoYear = true;
1382 $tenno = self
::tsToYear( $ts, 'tenno' );
1388 $num = substr( $ts, 2, 2 );
1391 $usedIranianYear = true;
1393 $iranian = self
::tsToIranian( $ts );
1395 $num = substr( $iranian[0], -2 );
1399 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1403 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1407 $h = substr( $ts, 8, 2 );
1408 $num = $h %
12 ?
$h %
12 : 12;
1412 $num = intval( substr( $ts, 8, 2 ) );
1416 $h = substr( $ts, 8, 2 );
1417 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1421 $num = substr( $ts, 8, 2 );
1425 $num = substr( $ts, 10, 2 );
1429 $num = substr( $ts, 12, 2 );
1439 $s .= Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1445 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1449 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1453 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1456 $usedIsLeapYear = true;
1457 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1460 $usedISOYear = true;
1461 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1468 $num = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1471 # Backslash escaping
1472 if ( $p < $formatLength - 1 ) {
1473 $s .= $format[++
$p];
1480 if ( $p < $formatLength - 1 ) {
1481 $endQuote = strpos( $format, '"', $p +
1 );
1482 if ( $endQuote === false ) {
1483 # No terminating quote, assume literal "
1486 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1490 # Quote at end of string, assume literal "
1497 if ( $num !== false ) {
1498 if ( $rawToggle ||
$raw ) {
1501 } elseif ( $roman ) {
1502 $s .= Language
::romanNumeral( $num );
1504 } elseif ( $hebrewNum ) {
1505 $s .= self
::hebrewNumeral( $num );
1508 $s .= $this->formatNum( $num, true );
1513 if ( $usedSecond ) {
1515 } elseif ( $usedMinute ) {
1516 $ttl = 60 - substr( $ts, 12, 2 );
1517 } elseif ( $usedHour ) {
1518 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1519 } elseif ( $usedAMPM ) {
1520 $ttl = 43200 - ( substr( $ts, 8, 2 ) %
12 ) * 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1521 } elseif ( $usedDay ||
$usedHebrewMonth ||
$usedIranianMonth ||
$usedHijriMonth ||
$usedHebrewYear ||
$usedIranianYear ||
$usedHijriYear ||
$usedTennoYear ) {
1522 // @todo Someone who understands the non-Gregorian calendars should write proper logic for them
1523 // so that they don't need purged every day.
1524 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1526 $possibleTtls = array();
1527 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1529 $possibleTtls[] = ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
$timeRemainingInDay;
1530 } elseif ( $usedISOYear ) {
1531 // December 28th falls on the last ISO week of the year, every year.
1532 // The last ISO week of a year can be 52 or 53.
1533 $lastWeekOfISOYear = DateTime
::createFromFormat( 'Ymd', substr( $ts, 0, 4 ) . '1228', $zone ?
: new DateTimeZone( 'UTC' ) )->format( 'W' );
1534 $currentISOWeek = Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1535 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1536 $timeRemainingInWeek = ( 7 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
$timeRemainingInDay;
1537 $possibleTtls[] = $weeksRemaining * 604800 +
$timeRemainingInWeek;
1541 $possibleTtls[] = ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) - substr( $ts, 6, 2 ) ) * 86400 +
$timeRemainingInDay;
1542 } elseif ( $usedYear ) {
1543 $possibleTtls[] = ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1544 +
$timeRemainingInDay;
1545 } elseif ( $usedIsLeapYear ) {
1546 $year = substr( $ts, 0, 4 );
1547 $timeRemainingInYear = ( Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) +
364 - Language
::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1548 +
$timeRemainingInDay;
1550 if ( $mod ||
( !( $year %
100 ) && $year %
400 ) ) {
1551 // this isn't a leap year. see when the next one starts
1552 $nextCandidate = $year - $mod +
4;
1553 if ( $nextCandidate %
100 ||
!( $nextCandidate %
400 ) ) {
1554 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
$timeRemainingInYear;
1556 $possibleTtls[] = ( $nextCandidate - $year +
3 ) * 365 * 86400 +
$timeRemainingInYear;
1559 // this is a leap year, so the next year isn't
1560 $possibleTtls[] = $timeRemainingInYear;
1564 if ( $possibleTtls ) {
1565 $ttl = min( $possibleTtls );
1572 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1573 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1576 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1577 * Gregorian dates to Iranian dates. Originally written in C, it
1578 * is released under the terms of GNU Lesser General Public
1579 * License. Conversion to PHP was performed by Niklas Laxström.
1581 * Link: http://www.farsiweb.info/jalali/jalali.c
1587 private static function tsToIranian( $ts ) {
1588 $gy = substr( $ts, 0, 4 ) -1600;
1589 $gm = substr( $ts, 4, 2 ) -1;
1590 $gd = substr( $ts, 6, 2 ) -1;
1592 # Days passed from the beginning (including leap years)
1594 +
floor( ( $gy +
3 ) / 4 )
1595 - floor( ( $gy +
99 ) / 100 )
1596 +
floor( ( $gy +
399 ) / 400 );
1598 // Add days of the past months of this year
1599 for ( $i = 0; $i < $gm; $i++
) {
1600 $gDayNo +
= self
::$GREG_DAYS[$i];
1604 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1608 // Days passed in current month
1609 $gDayNo +
= (int)$gd;
1611 $jDayNo = $gDayNo - 79;
1613 $jNp = floor( $jDayNo / 12053 );
1616 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1619 if ( $jDayNo >= 366 ) {
1620 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1621 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1624 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1625 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1631 return array( $jy, $jm, $jd );
1635 * Converting Gregorian dates to Hijri dates.
1637 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1639 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1645 private static function tsToHijri( $ts ) {
1646 $year = substr( $ts, 0, 4 );
1647 $month = substr( $ts, 4, 2 );
1648 $day = substr( $ts, 6, 2 );
1656 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1657 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1659 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1660 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1661 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1664 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1665 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1668 $zl = $zjd -1948440 +
10632;
1669 $zn = (int)( ( $zl - 1 ) / 10631 );
1670 $zl = $zl - 10631 * $zn +
354;
1671 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1672 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1673 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1674 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1675 $zm = (int)( ( 24 * $zl ) / 709 );
1676 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1677 $zy = 30 * $zn +
$zj - 30;
1679 return array( $zy, $zm, $zd );
1683 * Converting Gregorian dates to Hebrew dates.
1685 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1686 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1687 * to translate the relevant functions into PHP and release them under
1690 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1691 * and Adar II is 14. In a non-leap year, Adar is 6.
1697 private static function tsToHebrew( $ts ) {
1699 $year = substr( $ts, 0, 4 );
1700 $month = substr( $ts, 4, 2 );
1701 $day = substr( $ts, 6, 2 );
1703 # Calculate Hebrew year
1704 $hebrewYear = $year +
3760;
1706 # Month number when September = 1, August = 12
1708 if ( $month > 12 ) {
1715 # Calculate day of year from 1 September
1717 for ( $i = 1; $i < $month; $i++
) {
1721 # Check if the year is leap
1722 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1725 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1732 # Calculate the start of the Hebrew year
1733 $start = self
::hebrewYearStart( $hebrewYear );
1735 # Calculate next year's start
1736 if ( $dayOfYear <= $start ) {
1737 # Day is before the start of the year - it is the previous year
1739 $nextStart = $start;
1743 # Add days since previous year's 1 September
1745 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1749 # Start of the new (previous) year
1750 $start = self
::hebrewYearStart( $hebrewYear );
1753 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1756 # Calculate Hebrew day of year
1757 $hebrewDayOfYear = $dayOfYear - $start;
1759 # Difference between year's days
1760 $diff = $nextStart - $start;
1761 # Add 12 (or 13 for leap years) days to ignore the difference between
1762 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1763 # difference is only about the year type
1764 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1770 # Check the year pattern, and is leap year
1771 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1772 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1773 # and non-leap years
1774 $yearPattern = $diff %
30;
1775 # Check if leap year
1776 $isLeap = $diff >= 30;
1778 # Calculate day in the month from number of day in the Hebrew year
1779 # Don't check Adar - if the day is not in Adar, we will stop before;
1780 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1781 $hebrewDay = $hebrewDayOfYear;
1784 while ( $hebrewMonth <= 12 ) {
1785 # Calculate days in this month
1786 if ( $isLeap && $hebrewMonth == 6 ) {
1787 # Adar in a leap year
1789 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1791 if ( $hebrewDay <= $days ) {
1795 # Subtract the days of Adar I
1796 $hebrewDay -= $days;
1799 if ( $hebrewDay <= $days ) {
1805 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1806 # Cheshvan in a complete year (otherwise as the rule below)
1808 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1809 # Kislev in an incomplete year (otherwise as the rule below)
1812 # Odd months have 30 days, even have 29
1813 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1815 if ( $hebrewDay <= $days ) {
1816 # In the current month
1819 # Subtract the days of the current month
1820 $hebrewDay -= $days;
1821 # Try in the next month
1826 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1830 * This calculates the Hebrew year start, as days since 1 September.
1831 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1832 * Used for Hebrew date.
1838 private static function hebrewYearStart( $year ) {
1839 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1840 $b = intval( ( $year - 1 ) %
4 );
1841 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1845 $Mar = intval( $m );
1851 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1852 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1854 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1856 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1860 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1865 * Algorithm to convert Gregorian dates to Thai solar dates,
1866 * Minguo dates or Minguo dates.
1868 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1869 * http://en.wikipedia.org/wiki/Minguo_calendar
1870 * http://en.wikipedia.org/wiki/Japanese_era_name
1872 * @param string $ts 14-character timestamp
1873 * @param string $cName Calender name
1874 * @return array Converted year, month, day
1876 private static function tsToYear( $ts, $cName ) {
1877 $gy = substr( $ts, 0, 4 );
1878 $gm = substr( $ts, 4, 2 );
1879 $gd = substr( $ts, 6, 2 );
1881 if ( !strcmp( $cName, 'thai' ) ) {
1883 # Add 543 years to the Gregorian calendar
1884 # Months and days are identical
1885 $gy_offset = $gy +
543;
1886 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1888 # Deduct 1911 years from the Gregorian calendar
1889 # Months and days are identical
1890 $gy_offset = $gy - 1911;
1891 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1892 # Nengō dates up to Meiji period
1893 # Deduct years from the Gregorian calendar
1894 # depending on the nengo periods
1895 # Months and days are identical
1897 ||
( ( $gy == 1912 ) && ( $gm < 7 ) )
1898 ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1901 $gy_gannen = $gy - 1868 +
1;
1902 $gy_offset = $gy_gannen;
1903 if ( $gy_gannen == 1 ) {
1906 $gy_offset = '明治' . $gy_offset;
1908 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1909 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1910 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1911 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1912 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1915 $gy_gannen = $gy - 1912 +
1;
1916 $gy_offset = $gy_gannen;
1917 if ( $gy_gannen == 1 ) {
1920 $gy_offset = '大正' . $gy_offset;
1922 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1923 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1924 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1927 $gy_gannen = $gy - 1926 +
1;
1928 $gy_offset = $gy_gannen;
1929 if ( $gy_gannen == 1 ) {
1932 $gy_offset = '昭和' . $gy_offset;
1935 $gy_gannen = $gy - 1989 +
1;
1936 $gy_offset = $gy_gannen;
1937 if ( $gy_gannen == 1 ) {
1940 $gy_offset = '平成' . $gy_offset;
1946 return array( $gy_offset, $gm, $gd );
1950 * Roman number formatting up to 10000
1956 static function romanNumeral( $num ) {
1957 static $table = array(
1958 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1959 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1960 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1961 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
1962 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1965 $num = intval( $num );
1966 if ( $num > 10000 ||
$num <= 0 ) {
1971 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1972 if ( $num >= $pow10 ) {
1973 $s .= $table[$i][(int)floor( $num / $pow10 )];
1975 $num = $num %
$pow10;
1981 * Hebrew Gematria number formatting up to 9999
1987 static function hebrewNumeral( $num ) {
1988 static $table = array(
1989 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1990 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1991 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1992 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1995 $num = intval( $num );
1996 if ( $num > 9999 ||
$num <= 0 ) {
2001 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2002 if ( $num >= $pow10 ) {
2003 if ( $num == 15 ||
$num == 16 ) {
2004 $s .= $table[0][9] . $table[0][$num - 9];
2007 $s .= $table[$i][intval( ( $num / $pow10 ) )];
2008 if ( $pow10 == 1000 ) {
2013 $num = $num %
$pow10;
2015 if ( strlen( $s ) == 2 ) {
2018 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
2019 $str .= substr( $s, strlen( $s ) - 2, 2 );
2021 $start = substr( $str, 0, strlen( $str ) - 2 );
2022 $end = substr( $str, strlen( $str ) - 2 );
2025 $str = $start . 'ך';
2028 $str = $start . 'ם';
2031 $str = $start . 'ן';
2034 $str = $start . 'ף';
2037 $str = $start . 'ץ';
2044 * Used by date() and time() to adjust the time output.
2046 * @param string $ts The time in date('YmdHis') format
2047 * @param mixed $tz Adjust the time by this amount (default false, mean we
2048 * get user timecorrection setting)
2051 function userAdjust( $ts, $tz = false ) {
2052 global $wgUser, $wgLocalTZoffset;
2054 if ( $tz === false ) {
2055 $tz = $wgUser->getOption( 'timecorrection' );
2058 $data = explode( '|', $tz, 3 );
2060 if ( $data[0] == 'ZoneInfo' ) {
2061 wfSuppressWarnings();
2062 $userTZ = timezone_open( $data[2] );
2063 wfRestoreWarnings();
2064 if ( $userTZ !== false ) {
2065 $date = date_create( $ts, timezone_open( 'UTC' ) );
2066 date_timezone_set( $date, $userTZ );
2067 $date = date_format( $date, 'YmdHis' );
2070 # Unrecognized timezone, default to 'Offset' with the stored offset.
2071 $data[0] = 'Offset';
2074 if ( $data[0] == 'System' ||
$tz == '' ) {
2075 # Global offset in minutes.
2076 $minDiff = $wgLocalTZoffset;
2077 } elseif ( $data[0] == 'Offset' ) {
2078 $minDiff = intval( $data[1] );
2080 $data = explode( ':', $tz );
2081 if ( count( $data ) == 2 ) {
2082 $data[0] = intval( $data[0] );
2083 $data[1] = intval( $data[1] );
2084 $minDiff = abs( $data[0] ) * 60 +
$data[1];
2085 if ( $data[0] < 0 ) {
2086 $minDiff = -$minDiff;
2089 $minDiff = intval( $data[0] ) * 60;
2093 # No difference ? Return time unchanged
2094 if ( 0 == $minDiff ) {
2098 wfSuppressWarnings(); // E_STRICT system time bitching
2099 # Generate an adjusted date; take advantage of the fact that mktime
2100 # will normalize out-of-range values so we don't have to split $minDiff
2101 # into hours and minutes.
2103 (int)substr( $ts, 8, 2 ) ), # Hours
2104 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
2105 (int)substr( $ts, 12, 2 ), # Seconds
2106 (int)substr( $ts, 4, 2 ), # Month
2107 (int)substr( $ts, 6, 2 ), # Day
2108 (int)substr( $ts, 0, 4 ) ); # Year
2110 $date = date( 'YmdHis', $t );
2111 wfRestoreWarnings();
2117 * This is meant to be used by time(), date(), and timeanddate() to get
2118 * the date preference they're supposed to use, it should be used in
2122 * function timeanddate([...], $format = true) {
2123 * $datePreference = $this->dateFormat($format);
2128 * @param int|string|bool $usePrefs If true, the user's preference is used
2129 * if false, the site/language default is used
2130 * if int/string, assumed to be a format.
2133 function dateFormat( $usePrefs = true ) {
2136 if ( is_bool( $usePrefs ) ) {
2138 $datePreference = $wgUser->getDatePreference();
2140 $datePreference = (string)User
::getDefaultOption( 'date' );
2143 $datePreference = (string)$usePrefs;
2147 if ( $datePreference == '' ) {
2151 return $datePreference;
2155 * Get a format string for a given type and preference
2156 * @param string $type May be date, time or both
2157 * @param string $pref The format name as it appears in Messages*.php
2159 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2163 function getDateFormatString( $type, $pref ) {
2164 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2165 if ( $pref == 'default' ) {
2166 $pref = $this->getDefaultDateFormat();
2167 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2169 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2171 if ( $type === 'pretty' && $df === null ) {
2172 $df = $this->getDateFormatString( 'date', $pref );
2175 if ( $df === null ) {
2176 $pref = $this->getDefaultDateFormat();
2177 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2180 $this->dateFormatStrings
[$type][$pref] = $df;
2182 return $this->dateFormatStrings
[$type][$pref];
2186 * @param string $ts The time format which needs to be turned into a
2187 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2188 * @param bool $adj Whether to adjust the time output according to the
2189 * user configured offset ($timecorrection)
2190 * @param mixed $format True to use user's date format preference
2191 * @param string|bool $timecorrection The time offset as returned by
2192 * validateTimeZone() in Special:Preferences
2195 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2196 $ts = wfTimestamp( TS_MW
, $ts );
2198 $ts = $this->userAdjust( $ts, $timecorrection );
2200 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2201 return $this->sprintfDate( $df, $ts );
2205 * @param string $ts The time format which needs to be turned into a
2206 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2207 * @param bool $adj Whether to adjust the time output according to the
2208 * user configured offset ($timecorrection)
2209 * @param mixed $format True to use user's date format preference
2210 * @param string|bool $timecorrection The time offset as returned by
2211 * validateTimeZone() in Special:Preferences
2214 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2215 $ts = wfTimestamp( TS_MW
, $ts );
2217 $ts = $this->userAdjust( $ts, $timecorrection );
2219 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2220 return $this->sprintfDate( $df, $ts );
2224 * @param string $ts The time format which needs to be turned into a
2225 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2226 * @param bool $adj Whether to adjust the time output according to the
2227 * user configured offset ($timecorrection)
2228 * @param mixed $format What format to return, if it's false output the
2229 * default one (default true)
2230 * @param string|bool $timecorrection The time offset as returned by
2231 * validateTimeZone() in Special:Preferences
2234 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2235 $ts = wfTimestamp( TS_MW
, $ts );
2237 $ts = $this->userAdjust( $ts, $timecorrection );
2239 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2240 return $this->sprintfDate( $df, $ts );
2244 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2248 * @param int $seconds The amount of seconds.
2249 * @param array $chosenIntervals The intervals to enable.
2253 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2254 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2256 $segments = array();
2258 foreach ( $intervals as $intervalName => $intervalValue ) {
2259 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2260 // duration-years, duration-decades, duration-centuries, duration-millennia
2261 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2262 $segments[] = $message->inLanguage( $this )->escaped();
2265 return $this->listToText( $segments );
2269 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2270 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2274 * @param int $seconds The amount of seconds.
2275 * @param array $chosenIntervals The intervals to enable.
2279 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2280 if ( empty( $chosenIntervals ) ) {
2281 $chosenIntervals = array(
2293 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2294 $sortedNames = array_keys( $intervals );
2295 $smallestInterval = array_pop( $sortedNames );
2297 $segments = array();
2299 foreach ( $intervals as $name => $length ) {
2300 $value = floor( $seconds / $length );
2302 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2303 $seconds -= $value * $length;
2304 $segments[$name] = $value;
2312 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2314 * @param string $type Can be 'date', 'time' or 'both'
2315 * @param string $ts The time format which needs to be turned into a
2316 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2317 * @param User $user User object used to get preferences for timezone and format
2318 * @param array $options Array, can contain the following keys:
2319 * - 'timecorrection': time correction, can have the following values:
2320 * - true: use user's preference
2321 * - false: don't use time correction
2322 * - int: value of time correction in minutes
2323 * - 'format': format to use, can have the following values:
2324 * - true: use user's preference
2325 * - false: use default preference
2326 * - string: format to use
2330 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2331 $ts = wfTimestamp( TS_MW
, $ts );
2332 $options +
= array( 'timecorrection' => true, 'format' => true );
2333 if ( $options['timecorrection'] !== false ) {
2334 if ( $options['timecorrection'] === true ) {
2335 $offset = $user->getOption( 'timecorrection' );
2337 $offset = $options['timecorrection'];
2339 $ts = $this->userAdjust( $ts, $offset );
2341 if ( $options['format'] === true ) {
2342 $format = $user->getDatePreference();
2344 $format = $options['format'];
2346 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2347 return $this->sprintfDate( $df, $ts );
2351 * Get the formatted date for the given timestamp and formatted for
2354 * @param mixed $ts Mixed: the time format which needs to be turned into a
2355 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2356 * @param User $user User object used to get preferences for timezone and format
2357 * @param array $options Array, can contain the following keys:
2358 * - 'timecorrection': time correction, can have the following values:
2359 * - true: use user's preference
2360 * - false: don't use time correction
2361 * - int: value of time correction in minutes
2362 * - 'format': format to use, can have the following values:
2363 * - true: use user's preference
2364 * - false: use default preference
2365 * - string: format to use
2369 public function userDate( $ts, User
$user, array $options = array() ) {
2370 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2374 * Get the formatted time for the given timestamp and formatted for
2377 * @param mixed $ts The time format which needs to be turned into a
2378 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2379 * @param User $user User object used to get preferences for timezone and format
2380 * @param array $options Array, can contain the following keys:
2381 * - 'timecorrection': time correction, can have the following values:
2382 * - true: use user's preference
2383 * - false: don't use time correction
2384 * - int: value of time correction in minutes
2385 * - 'format': format to use, can have the following values:
2386 * - true: use user's preference
2387 * - false: use default preference
2388 * - string: format to use
2392 public function userTime( $ts, User
$user, array $options = array() ) {
2393 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2397 * Get the formatted date and time for the given timestamp and formatted for
2400 * @param mixed $ts The time format which needs to be turned into a
2401 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2402 * @param User $user User object used to get preferences for timezone and format
2403 * @param array $options Array, can contain the following keys:
2404 * - 'timecorrection': time correction, can have the following values:
2405 * - true: use user's preference
2406 * - false: don't use time correction
2407 * - int: value of time correction in minutes
2408 * - 'format': format to use, can have the following values:
2409 * - true: use user's preference
2410 * - false: use default preference
2411 * - string: format to use
2415 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2416 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2420 * Convert an MWTimestamp into a pretty human-readable timestamp using
2421 * the given user preferences and relative base time.
2423 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2424 * on your timestamp object, which will then call this function. Calling
2425 * this function directly will cause hooks to be skipped over.
2427 * @see MWTimestamp::getHumanTimestamp
2428 * @param MWTimestamp $ts Timestamp to prettify
2429 * @param MWTimestamp $relativeTo Base timestamp
2430 * @param User $user User preferences to use
2431 * @return string Human timestamp
2434 public function getHumanTimestamp( MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user ) {
2435 $diff = $ts->diff( $relativeTo );
2436 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) -
2437 (int)$relativeTo->timestamp
->format( 'w' ) );
2438 $days = $diff->days ?
: (int)$diffDay;
2439 if ( $diff->invert ||
$days > 5
2440 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' )
2442 // Timestamps are in different years: use full timestamp
2443 // Also do full timestamp for future dates
2445 * @todo FIXME: Add better handling of future timestamps.
2447 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2448 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2449 } elseif ( $days > 5 ) {
2450 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2451 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2452 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2453 } elseif ( $days > 1 ) {
2454 // Timestamp within the past week: show the day of the week and time
2455 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2456 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2458 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2459 $ts = wfMessage( "$weekday-at" )
2460 ->inLanguage( $this )
2461 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2463 } elseif ( $days == 1 ) {
2464 // Timestamp was yesterday: say 'yesterday' and the time.
2465 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2466 $ts = wfMessage( 'yesterday-at' )
2467 ->inLanguage( $this )
2468 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2470 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2471 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2472 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2473 $ts = wfMessage( 'today-at' )
2474 ->inLanguage( $this )
2475 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2478 // From here on in, the timestamp was soon enough ago so that we can simply say
2479 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2480 } elseif ( $diff->h
== 1 ) {
2481 // Less than 90 minutes, but more than an hour ago.
2482 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2483 } elseif ( $diff->i
>= 1 ) {
2484 // A few minutes ago.
2485 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2486 } elseif ( $diff->s
>= 30 ) {
2487 // Less than a minute, but more than 30 sec ago.
2488 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2490 // Less than 30 seconds ago.
2491 $ts = wfMessage( 'just-now' )->text();
2498 * @param string $key
2499 * @return array|null
2501 function getMessage( $key ) {
2502 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2508 function getAllMessages() {
2509 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2514 * @param string $out
2515 * @param string $string
2518 function iconv( $in, $out, $string ) {
2519 # This is a wrapper for iconv in all languages except esperanto,
2520 # which does some nasty x-conversions beforehand
2522 # Even with //IGNORE iconv can whine about illegal characters in
2523 # *input* string. We just ignore those too.
2524 # REF: http://bugs.php.net/bug.php?id=37166
2525 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2526 wfSuppressWarnings();
2527 $text = iconv( $in, $out . '//IGNORE', $string );
2528 wfRestoreWarnings();
2532 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2535 * @param array $matches
2536 * @return mixed|string
2538 function ucwordbreaksCallbackAscii( $matches ) {
2539 return $this->ucfirst( $matches[1] );
2543 * @param array $matches
2546 function ucwordbreaksCallbackMB( $matches ) {
2547 return mb_strtoupper( $matches[0] );
2551 * @param array $matches
2554 function ucCallback( $matches ) {
2555 list( $wikiUpperChars ) = self
::getCaseMaps();
2556 return strtr( $matches[1], $wikiUpperChars );
2560 * @param array $matches
2563 function lcCallback( $matches ) {
2564 list( , $wikiLowerChars ) = self
::getCaseMaps();
2565 return strtr( $matches[1], $wikiLowerChars );
2569 * @param array $matches
2572 function ucwordsCallbackMB( $matches ) {
2573 return mb_strtoupper( $matches[0] );
2577 * @param array $matches
2580 function ucwordsCallbackWiki( $matches ) {
2581 list( $wikiUpperChars ) = self
::getCaseMaps();
2582 return strtr( $matches[0], $wikiUpperChars );
2586 * Make a string's first character uppercase
2588 * @param string $str
2592 function ucfirst( $str ) {
2594 if ( $o < 96 ) { // if already uppercase...
2596 } elseif ( $o < 128 ) {
2597 return ucfirst( $str ); // use PHP's ucfirst()
2599 // fall back to more complex logic in case of multibyte strings
2600 return $this->uc( $str, true );
2605 * Convert a string to uppercase
2607 * @param string $str
2608 * @param bool $first
2612 function uc( $str, $first = false ) {
2613 if ( function_exists( 'mb_strtoupper' ) ) {
2615 if ( $this->isMultibyte( $str ) ) {
2616 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2618 return ucfirst( $str );
2621 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2624 if ( $this->isMultibyte( $str ) ) {
2625 $x = $first ?
'^' : '';
2626 return preg_replace_callback(
2627 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2628 array( $this, 'ucCallback' ),
2632 return $first ?
ucfirst( $str ) : strtoupper( $str );
2638 * @param string $str
2639 * @return mixed|string
2641 function lcfirst( $str ) {
2644 return strval( $str );
2645 } elseif ( $o >= 128 ) {
2646 return $this->lc( $str, true );
2647 } elseif ( $o > 96 ) {
2650 $str[0] = strtolower( $str[0] );
2656 * @param string $str
2657 * @param bool $first
2658 * @return mixed|string
2660 function lc( $str, $first = false ) {
2661 if ( function_exists( 'mb_strtolower' ) ) {
2663 if ( $this->isMultibyte( $str ) ) {
2664 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2666 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2669 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2672 if ( $this->isMultibyte( $str ) ) {
2673 $x = $first ?
'^' : '';
2674 return preg_replace_callback(
2675 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2676 array( $this, 'lcCallback' ),
2680 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2686 * @param string $str
2689 function isMultibyte( $str ) {
2690 return (bool)preg_match( '/[\x80-\xff]/', $str );
2694 * @param string $str
2695 * @return mixed|string
2697 function ucwords( $str ) {
2698 if ( $this->isMultibyte( $str ) ) {
2699 $str = $this->lc( $str );
2701 // regexp to find first letter in each word (i.e. after each space)
2702 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2704 // function to use to capitalize a single char
2705 if ( function_exists( 'mb_strtoupper' ) ) {
2706 return preg_replace_callback(
2708 array( $this, 'ucwordsCallbackMB' ),
2712 return preg_replace_callback(
2714 array( $this, 'ucwordsCallbackWiki' ),
2719 return ucwords( strtolower( $str ) );
2724 * capitalize words at word breaks
2726 * @param string $str
2729 function ucwordbreaks( $str ) {
2730 if ( $this->isMultibyte( $str ) ) {
2731 $str = $this->lc( $str );
2733 // since \b doesn't work for UTF-8, we explicitely define word break chars
2734 $breaks = "[ \-\(\)\}\{\.,\?!]";
2736 // find first letter after word break
2737 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2738 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2740 if ( function_exists( 'mb_strtoupper' ) ) {
2741 return preg_replace_callback(
2743 array( $this, 'ucwordbreaksCallbackMB' ),
2747 return preg_replace_callback(
2749 array( $this, 'ucwordsCallbackWiki' ),
2754 return preg_replace_callback(
2755 '/\b([\w\x80-\xff]+)\b/',
2756 array( $this, 'ucwordbreaksCallbackAscii' ),
2763 * Return a case-folded representation of $s
2765 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2766 * and $s2 are the same except for the case of their characters. It is not
2767 * necessary for the value returned to make sense when displayed.
2769 * Do *not* perform any other normalisation in this function. If a caller
2770 * uses this function when it should be using a more general normalisation
2771 * function, then fix the caller.
2777 function caseFold( $s ) {
2778 return $this->uc( $s );
2785 function checkTitleEncoding( $s ) {
2786 if ( is_array( $s ) ) {
2787 throw new MWException( 'Given array to checkTitleEncoding.' );
2789 if ( StringUtils
::isUtf8( $s ) ) {
2793 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2799 function fallback8bitEncoding() {
2800 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2804 * Most writing systems use whitespace to break up words.
2805 * Some languages such as Chinese don't conventionally do this,
2806 * which requires special handling when breaking up words for
2811 function hasWordBreaks() {
2816 * Some languages such as Chinese require word segmentation,
2817 * Specify such segmentation when overridden in derived class.
2819 * @param string $string
2822 function segmentByWord( $string ) {
2827 * Some languages have special punctuation need to be normalized.
2828 * Make such changes here.
2830 * @param string $string
2833 function normalizeForSearch( $string ) {
2834 return self
::convertDoubleWidth( $string );
2838 * convert double-width roman characters to single-width.
2839 * range: ff00-ff5f ~= 0020-007f
2841 * @param string $string
2845 protected static function convertDoubleWidth( $string ) {
2846 static $full = null;
2847 static $half = null;
2849 if ( $full === null ) {
2850 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2851 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2852 $full = str_split( $fullWidth, 3 );
2853 $half = str_split( $halfWidth );
2856 $string = str_replace( $full, $half, $string );
2861 * @param string $string
2862 * @param string $pattern
2865 protected static function insertSpace( $string, $pattern ) {
2866 $string = preg_replace( $pattern, " $1 ", $string );
2867 $string = preg_replace( '/ +/', ' ', $string );
2872 * @param array $termsArray
2875 function convertForSearchResult( $termsArray ) {
2876 # some languages, e.g. Chinese, need to do a conversion
2877 # in order for search results to be displayed correctly
2882 * Get the first character of a string.
2887 function firstChar( $s ) {
2890 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2891 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2896 if ( isset( $matches[1] ) ) {
2897 if ( strlen( $matches[1] ) != 3 ) {
2901 // Break down Hangul syllables to grab the first jamo
2902 $code = utf8ToCodepoint( $matches[1] );
2903 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2905 } elseif ( $code < 0xb098 ) {
2906 return "\xe3\x84\xb1";
2907 } elseif ( $code < 0xb2e4 ) {
2908 return "\xe3\x84\xb4";
2909 } elseif ( $code < 0xb77c ) {
2910 return "\xe3\x84\xb7";
2911 } elseif ( $code < 0xb9c8 ) {
2912 return "\xe3\x84\xb9";
2913 } elseif ( $code < 0xbc14 ) {
2914 return "\xe3\x85\x81";
2915 } elseif ( $code < 0xc0ac ) {
2916 return "\xe3\x85\x82";
2917 } elseif ( $code < 0xc544 ) {
2918 return "\xe3\x85\x85";
2919 } elseif ( $code < 0xc790 ) {
2920 return "\xe3\x85\x87";
2921 } elseif ( $code < 0xcc28 ) {
2922 return "\xe3\x85\x88";
2923 } elseif ( $code < 0xce74 ) {
2924 return "\xe3\x85\x8a";
2925 } elseif ( $code < 0xd0c0 ) {
2926 return "\xe3\x85\x8b";
2927 } elseif ( $code < 0xd30c ) {
2928 return "\xe3\x85\x8c";
2929 } elseif ( $code < 0xd558 ) {
2930 return "\xe3\x85\x8d";
2932 return "\xe3\x85\x8e";
2939 function initEncoding() {
2940 # Some languages may have an alternate char encoding option
2941 # (Esperanto X-coding, Japanese furigana conversion, etc)
2942 # If this language is used as the primary content language,
2943 # an override to the defaults can be set here on startup.
2950 function recodeForEdit( $s ) {
2951 # For some languages we'll want to explicitly specify
2952 # which characters make it into the edit box raw
2953 # or are converted in some way or another.
2954 global $wgEditEncoding;
2955 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2958 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2966 function recodeInput( $s ) {
2967 # Take the previous into account.
2968 global $wgEditEncoding;
2969 if ( $wgEditEncoding != '' ) {
2970 $enc = $wgEditEncoding;
2974 if ( $enc == 'UTF-8' ) {
2977 return $this->iconv( $enc, 'UTF-8', $s );
2982 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2983 * also cleans up certain backwards-compatible sequences, converting them
2984 * to the modern Unicode equivalent.
2986 * This is language-specific for performance reasons only.
2992 function normalize( $s ) {
2993 global $wgAllUnicodeFixes;
2994 $s = UtfNormal
::cleanUp( $s );
2995 if ( $wgAllUnicodeFixes ) {
2996 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2997 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3004 * Transform a string using serialized data stored in the given file (which
3005 * must be in the serialized subdirectory of $IP). The file contains pairs
3006 * mapping source characters to destination characters.
3008 * The data is cached in process memory. This will go faster if you have the
3009 * FastStringSearch extension.
3011 * @param string $file
3012 * @param string $string
3014 * @throws MWException
3017 function transformUsingPairFile( $file, $string ) {
3018 if ( !isset( $this->transformData
[$file] ) ) {
3019 $data = wfGetPrecompiledData( $file );
3020 if ( $data === false ) {
3021 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
3023 $this->transformData
[$file] = new ReplacementArray( $data );
3025 return $this->transformData
[$file]->replace( $string );
3029 * For right-to-left language support
3034 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
3038 * Return the correct HTML 'dir' attribute value for this language.
3042 return $this->isRTL() ?
'rtl' : 'ltr';
3046 * Return 'left' or 'right' as appropriate alignment for line-start
3047 * for this language's text direction.
3049 * Should be equivalent to CSS3 'start' text-align value....
3053 function alignStart() {
3054 return $this->isRTL() ?
'right' : 'left';
3058 * Return 'right' or 'left' as appropriate alignment for line-end
3059 * for this language's text direction.
3061 * Should be equivalent to CSS3 'end' text-align value....
3065 function alignEnd() {
3066 return $this->isRTL() ?
'left' : 'right';
3070 * A hidden direction mark (LRM or RLM), depending on the language direction.
3071 * Unlike getDirMark(), this function returns the character as an HTML entity.
3072 * This function should be used when the output is guaranteed to be HTML,
3073 * because it makes the output HTML source code more readable. When
3074 * the output is plain text or can be escaped, getDirMark() should be used.
3076 * @param bool $opposite Get the direction mark opposite to your language
3080 function getDirMarkEntity( $opposite = false ) {
3082 return $this->isRTL() ?
'‎' : '‏';
3084 return $this->isRTL() ?
'‏' : '‎';
3088 * A hidden direction mark (LRM or RLM), depending on the language direction.
3089 * This function produces them as invisible Unicode characters and
3090 * the output may be hard to read and debug, so it should only be used
3091 * when the output is plain text or can be escaped. When the output is
3092 * HTML, use getDirMarkEntity() instead.
3094 * @param bool $opposite Get the direction mark opposite to your language
3097 function getDirMark( $opposite = false ) {
3098 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3099 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3101 return $this->isRTL() ?
$lrm : $rlm;
3103 return $this->isRTL() ?
$rlm : $lrm;
3109 function capitalizeAllNouns() {
3110 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
3114 * An arrow, depending on the language direction.
3116 * @param string $direction The direction of the arrow: forwards (default),
3117 * backwards, left, right, up, down.
3120 function getArrow( $direction = 'forwards' ) {
3121 switch ( $direction ) {
3123 return $this->isRTL() ?
'←' : '→';
3125 return $this->isRTL() ?
'→' : '←';
3138 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3142 function linkPrefixExtension() {
3143 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
3147 * Get all magic words from cache.
3150 function getMagicWords() {
3151 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
3155 * Run the LanguageGetMagic hook once.
3157 protected function doMagicHook() {
3158 if ( $this->mMagicHookDone
) {
3161 $this->mMagicHookDone
= true;
3162 wfProfileIn( 'LanguageGetMagic' );
3163 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
3164 wfProfileOut( 'LanguageGetMagic' );
3168 * Fill a MagicWord object with data from here
3170 * @param MagicWord $mw
3172 function getMagic( $mw ) {
3173 // Saves a function call
3174 if ( !$this->mMagicHookDone
) {
3175 $this->doMagicHook();
3178 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3179 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3181 $rawEntry = self
::$dataCache->getSubitem(
3182 $this->mCode
, 'magicWords', $mw->mId
);
3185 if ( !is_array( $rawEntry ) ) {
3186 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3188 $mw->mCaseSensitive
= $rawEntry[0];
3189 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3194 * Add magic words to the extension array
3196 * @param array $newWords
3198 function addMagicWordsByLang( $newWords ) {
3199 $fallbackChain = $this->getFallbackLanguages();
3200 $fallbackChain = array_reverse( $fallbackChain );
3201 foreach ( $fallbackChain as $code ) {
3202 if ( isset( $newWords[$code] ) ) {
3203 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3209 * Get special page names, as an associative array
3210 * case folded alias => real name
3213 function getSpecialPageAliases() {
3214 // Cache aliases because it may be slow to load them
3215 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3217 $this->mExtendedSpecialPageAliases
=
3218 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3219 wfRunHooks( 'LanguageGetSpecialPageAliases',
3220 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
3223 return $this->mExtendedSpecialPageAliases
;
3227 * Italic is unsuitable for some languages
3229 * @param string $text The text to be emphasized.
3232 function emphasize( $text ) {
3233 return "<em>$text</em>";
3237 * Normally we output all numbers in plain en_US style, that is
3238 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3239 * point twohundredthirtyfive. However this is not suitable for all
3240 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3241 * Icelandic just want to use commas instead of dots, and dots instead
3242 * of commas like "293.291,235".
3244 * An example of this function being called:
3246 * wfMessage( 'message' )->numParams( $num )->text()
3249 * See $separatorTransformTable on MessageIs.php for
3250 * the , => . and . => , implementation.
3252 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3253 * @param int|float $number The string to be formatted, should be an integer
3254 * or a floating point number.
3255 * @param bool $nocommafy Set to true for special numbers like dates
3258 public function formatNum( $number, $nocommafy = false ) {
3259 global $wgTranslateNumerals;
3260 if ( !$nocommafy ) {
3261 $number = $this->commafy( $number );
3262 $s = $this->separatorTransformTable();
3264 $number = strtr( $number, $s );
3268 if ( $wgTranslateNumerals ) {
3269 $s = $this->digitTransformTable();
3271 $number = strtr( $number, $s );
3279 * Front-end for non-commafied formatNum
3281 * @param int|float $number The string to be formatted, should be an integer
3282 * or a floating point number.
3286 public function formatNumNoSeparators( $number ) {
3287 return $this->formatNum( $number, true );
3291 * @param string $number
3294 public function parseFormattedNumber( $number ) {
3295 $s = $this->digitTransformTable();
3297 // eliminate empty array values such as ''. (bug 64347)
3298 $s = array_filter( $s );
3299 $number = strtr( $number, array_flip( $s ) );
3302 $s = $this->separatorTransformTable();
3304 // eliminate empty array values such as ''. (bug 64347)
3305 $s = array_filter( $s );
3306 $number = strtr( $number, array_flip( $s ) );
3309 $number = strtr( $number, array( ',' => '' ) );
3314 * Adds commas to a given number
3316 * @param mixed $number
3319 function commafy( $number ) {
3320 $digitGroupingPattern = $this->digitGroupingPattern();
3321 if ( $number === null ) {
3325 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3326 // default grouping is at thousands, use the same for ###,###,### pattern too.
3327 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3329 // Ref: http://cldr.unicode.org/translation/number-patterns
3331 if ( intval( $number ) < 0 ) {
3332 // For negative numbers apply the algorithm like positive number and add sign.
3334 $number = substr( $number, 1 );
3336 $integerPart = array();
3337 $decimalPart = array();
3338 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3339 preg_match( "/\d+/", $number, $integerPart );
3340 preg_match( "/\.\d*/", $number, $decimalPart );
3341 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3342 if ( $groupedNumber === $number ) {
3343 // the string does not have any number part. Eg: .12345
3344 return $sign . $groupedNumber;
3346 $start = $end = strlen( $integerPart[0] );
3347 while ( $start > 0 ) {
3348 $match = $matches[0][$numMatches - 1];
3349 $matchLen = strlen( $match );
3350 $start = $end - $matchLen;
3354 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3356 if ( $numMatches > 1 ) {
3357 // use the last pattern for the rest of the number
3361 $groupedNumber = "," . $groupedNumber;
3364 return $sign . $groupedNumber;
3371 function digitGroupingPattern() {
3372 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3378 function digitTransformTable() {
3379 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3385 function separatorTransformTable() {
3386 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3390 * Take a list of strings and build a locale-friendly comma-separated
3391 * list, using the local comma-separator message.
3392 * The last two strings are chained with an "and".
3393 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3395 * @param string[] $l
3398 function listToText( array $l ) {
3399 $m = count( $l ) - 1;
3404 $and = $this->getMessageFromDB( 'and' );
3405 $space = $this->getMessageFromDB( 'word-separator' );
3407 $comma = $this->getMessageFromDB( 'comma-separator' );
3411 for ( $i = $m - 1; $i >= 0; $i-- ) {
3412 if ( $i == $m - 1 ) {
3413 $s = $l[$i] . $and . $space . $s;
3415 $s = $l[$i] . $comma . $s;
3422 * Take a list of strings and build a locale-friendly comma-separated
3423 * list, using the local comma-separator message.
3424 * @param string[] $list Array of strings to put in a comma list
3427 function commaList( array $list ) {
3429 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3435 * Take a list of strings and build a locale-friendly semicolon-separated
3436 * list, using the local semicolon-separator message.
3437 * @param string[] $list Array of strings to put in a semicolon list
3440 function semicolonList( array $list ) {
3442 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3448 * Same as commaList, but separate it with the pipe instead.
3449 * @param string[] $list Array of strings to put in a pipe list
3452 function pipeList( array $list ) {
3454 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3460 * Truncate a string to a specified length in bytes, appending an optional
3461 * string (e.g. for ellipses)
3463 * The database offers limited byte lengths for some columns in the database;
3464 * multi-byte character sets mean we need to ensure that only whole characters
3465 * are included, otherwise broken characters can be passed to the user
3467 * If $length is negative, the string will be truncated from the beginning
3469 * @param string $string String to truncate
3470 * @param int $length Maximum length (including ellipses)
3471 * @param string $ellipsis String to append to the truncated text
3472 * @param bool $adjustLength Subtract length of ellipsis from $length.
3473 * $adjustLength was introduced in 1.18, before that behaved as if false.
3476 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3477 # Use the localized ellipsis character
3478 if ( $ellipsis == '...' ) {
3479 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3481 # Check if there is no need to truncate
3482 if ( $length == 0 ) {
3483 return $ellipsis; // convention
3484 } elseif ( strlen( $string ) <= abs( $length ) ) {
3485 return $string; // no need to truncate
3487 $stringOriginal = $string;
3488 # If ellipsis length is >= $length then we can't apply $adjustLength
3489 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3490 $string = $ellipsis; // this can be slightly unexpected
3491 # Otherwise, truncate and add ellipsis...
3493 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3494 if ( $length > 0 ) {
3495 $length -= $eLength;
3496 $string = substr( $string, 0, $length ); // xyz...
3497 $string = $this->removeBadCharLast( $string );
3498 $string = rtrim( $string );
3499 $string = $string . $ellipsis;
3501 $length +
= $eLength;
3502 $string = substr( $string, $length ); // ...xyz
3503 $string = $this->removeBadCharFirst( $string );
3504 $string = ltrim( $string );
3505 $string = $ellipsis . $string;
3508 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3509 # This check is *not* redundant if $adjustLength, due to the single case where
3510 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3511 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3514 return $stringOriginal;
3519 * Remove bytes that represent an incomplete Unicode character
3520 * at the end of string (e.g. bytes of the char are missing)
3522 * @param string $string
3525 protected function removeBadCharLast( $string ) {
3526 if ( $string != '' ) {
3527 $char = ord( $string[strlen( $string ) - 1] );
3529 if ( $char >= 0xc0 ) {
3530 # We got the first byte only of a multibyte char; remove it.
3531 $string = substr( $string, 0, -1 );
3532 } elseif ( $char >= 0x80 &&
3533 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3534 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3536 # We chopped in the middle of a character; remove it
3544 * Remove bytes that represent an incomplete Unicode character
3545 * at the start of string (e.g. bytes of the char are missing)
3547 * @param string $string
3550 protected function removeBadCharFirst( $string ) {
3551 if ( $string != '' ) {
3552 $char = ord( $string[0] );
3553 if ( $char >= 0x80 && $char < 0xc0 ) {
3554 # We chopped in the middle of a character; remove the whole thing
3555 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3562 * Truncate a string of valid HTML to a specified length in bytes,
3563 * appending an optional string (e.g. for ellipses), and return valid HTML
3565 * This is only intended for styled/linked text, such as HTML with
3566 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3567 * Also, this will not detect things like "display:none" CSS.
3569 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3571 * @param string $text HTML string to truncate
3572 * @param int $length (zero/positive) Maximum length (including ellipses)
3573 * @param string $ellipsis String to append to the truncated text
3576 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3577 # Use the localized ellipsis character
3578 if ( $ellipsis == '...' ) {
3579 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3581 # Check if there is clearly no need to truncate
3582 if ( $length <= 0 ) {
3583 return $ellipsis; // no text shown, nothing to format (convention)
3584 } elseif ( strlen( $text ) <= $length ) {
3585 return $text; // string short enough even *with* HTML (short-circuit)
3588 $dispLen = 0; // innerHTML legth so far
3589 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3590 $tagType = 0; // 0-open, 1-close
3591 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3592 $entityState = 0; // 0-not entity, 1-entity
3593 $tag = $ret = ''; // accumulated tag name, accumulated result string
3594 $openTags = array(); // open tag stack
3595 $maybeState = null; // possible truncation state
3597 $textLen = strlen( $text );
3598 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3599 for ( $pos = 0; true; ++
$pos ) {
3600 # Consider truncation once the display length has reached the maximim.
3601 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3602 # Check that we're not in the middle of a bracket/entity...
3603 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3604 if ( !$testingEllipsis ) {
3605 $testingEllipsis = true;
3606 # Save where we are; we will truncate here unless there turn out to
3607 # be so few remaining characters that truncation is not necessary.
3608 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3609 $maybeState = array( $ret, $openTags ); // save state
3611 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3612 # String in fact does need truncation, the truncation point was OK.
3613 list( $ret, $openTags ) = $maybeState; // reload state
3614 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3615 $ret .= $ellipsis; // add ellipsis
3619 if ( $pos >= $textLen ) {
3620 break; // extra iteration just for above checks
3623 # Read the next char...
3625 $lastCh = $pos ?
$text[$pos - 1] : '';
3626 $ret .= $ch; // add to result string
3628 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3629 $entityState = 0; // for bad HTML
3630 $bracketState = 1; // tag started (checking for backslash)
3631 } elseif ( $ch == '>' ) {
3632 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3633 $entityState = 0; // for bad HTML
3634 $bracketState = 0; // out of brackets
3635 } elseif ( $bracketState == 1 ) {
3637 $tagType = 1; // close tag (e.g. "</span>")
3639 $tagType = 0; // open tag (e.g. "<span>")
3642 $bracketState = 2; // building tag name
3643 } elseif ( $bracketState == 2 ) {
3647 // Name found (e.g. "<a href=..."), add on tag attributes...
3648 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3650 } elseif ( $bracketState == 0 ) {
3651 if ( $entityState ) {
3654 $dispLen++
; // entity is one displayed char
3657 if ( $neLength == 0 && !$maybeState ) {
3658 // Save state without $ch. We want to *hit* the first
3659 // display char (to get tags) but not *use* it if truncating.
3660 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3663 $entityState = 1; // entity found, (e.g. " ")
3665 $dispLen++
; // this char is displayed
3666 // Add the next $max display text chars after this in one swoop...
3667 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3668 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3669 $dispLen +
= $skipped;
3675 // Close the last tag if left unclosed by bad HTML
3676 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3677 while ( count( $openTags ) > 0 ) {
3678 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3684 * truncateHtml() helper function
3685 * like strcspn() but adds the skipped chars to $ret
3687 * @param string $ret
3688 * @param string $text
3689 * @param string $search
3691 * @param null|int $len
3694 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3695 if ( $len === null ) {
3696 $len = -1; // -1 means "no limit" for strcspn
3697 } elseif ( $len < 0 ) {
3701 if ( $start < strlen( $text ) ) {
3702 $skipCount = strcspn( $text, $search, $start, $len );
3703 $ret .= substr( $text, $start, $skipCount );
3709 * truncateHtml() helper function
3710 * (a) push or pop $tag from $openTags as needed
3711 * (b) clear $tag value
3712 * @param string &$tag Current HTML tag name we are looking at
3713 * @param int $tagType (0-open tag, 1-close tag)
3714 * @param string $lastCh Character before the '>' that ended this tag
3715 * @param array &$openTags Open tag stack (not accounting for $tag)
3717 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3718 $tag = ltrim( $tag );
3720 if ( $tagType == 0 && $lastCh != '/' ) {
3721 $openTags[] = $tag; // tag opened (didn't close itself)
3722 } elseif ( $tagType == 1 ) {
3723 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3724 array_pop( $openTags ); // tag closed
3732 * Grammatical transformations, needed for inflected languages
3733 * Invoked by putting {{grammar:case|word}} in a message
3735 * @param string $word
3736 * @param string $case
3739 function convertGrammar( $word, $case ) {
3740 global $wgGrammarForms;
3741 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3742 return $wgGrammarForms[$this->getCode()][$case][$word];
3748 * Get the grammar forms for the content language
3749 * @return array Array of grammar forms
3752 function getGrammarForms() {
3753 global $wgGrammarForms;
3754 if ( isset( $wgGrammarForms[$this->getCode()] )
3755 && is_array( $wgGrammarForms[$this->getCode()] )
3757 return $wgGrammarForms[$this->getCode()];
3763 * Provides an alternative text depending on specified gender.
3764 * Usage {{gender:username|masculine|feminine|unknown}}.
3765 * username is optional, in which case the gender of current user is used,
3766 * but only in (some) interface messages; otherwise default gender is used.
3768 * If no forms are given, an empty string is returned. If only one form is
3769 * given, it will be returned unconditionally. These details are implied by
3770 * the caller and cannot be overridden in subclasses.
3772 * If three forms are given, the default is to use the third (unknown) form.
3773 * If fewer than three forms are given, the default is to use the first (masculine) form.
3774 * These details can be overridden in subclasses.
3776 * @param string $gender
3777 * @param array $forms
3781 function gender( $gender, $forms ) {
3782 if ( !count( $forms ) ) {
3785 $forms = $this->preConvertPlural( $forms, 2 );
3786 if ( $gender === 'male' ) {
3789 if ( $gender === 'female' ) {
3792 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3796 * Plural form transformations, needed for some languages.
3797 * For example, there are 3 form of plural in Russian and Polish,
3798 * depending on "count mod 10". See [[w:Plural]]
3799 * For English it is pretty simple.
3801 * Invoked by putting {{plural:count|wordform1|wordform2}}
3802 * or {{plural:count|wordform1|wordform2|wordform3}}
3804 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3806 * @param int $count Non-localized number
3807 * @param array $forms Different plural forms
3808 * @return string Correct form of plural for $count in this language
3810 function convertPlural( $count, $forms ) {
3811 // Handle explicit n=pluralform cases
3812 $forms = $this->handleExplicitPluralForms( $count, $forms );
3813 if ( is_string( $forms ) ) {
3816 if ( !count( $forms ) ) {
3820 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3821 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3822 return $forms[$pluralForm];
3826 * Handles explicit plural forms for Language::convertPlural()
3828 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3829 * If an explicitly defined plural form matches the $count, then
3830 * string value returned, otherwise array returned for further consideration
3831 * by CLDR rules or overridden convertPlural().
3835 * @param int $count Non-localized number
3836 * @param array $forms Different plural forms
3838 * @return array|string
3840 protected function handleExplicitPluralForms( $count, array $forms ) {
3841 foreach ( $forms as $index => $form ) {
3842 if ( preg_match( '/\d+=/i', $form ) ) {
3843 $pos = strpos( $form, '=' );
3844 if ( substr( $form, 0, $pos ) === (string)$count ) {
3845 return substr( $form, $pos +
1 );
3847 unset( $forms[$index] );
3850 return array_values( $forms );
3854 * Checks that convertPlural was given an array and pads it to requested
3855 * amount of forms by copying the last one.
3857 * @param array $forms Array of forms given to convertPlural
3858 * @param int $count How many forms should there be at least
3859 * @return array Padded array of forms or an exception if not an array
3861 protected function preConvertPlural( /* Array */ $forms, $count ) {
3862 while ( count( $forms ) < $count ) {
3863 $forms[] = $forms[count( $forms ) - 1];
3869 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3870 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3871 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3872 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3875 * @param string $str The validated block duration in English
3876 * @return string Somehow translated block duration
3877 * @see LanguageFi.php for example implementation
3879 function translateBlockExpiry( $str ) {
3880 $duration = SpecialBlock
::getSuggestedDurations( $this );
3881 foreach ( $duration as $show => $value ) {
3882 if ( strcmp( $str, $value ) == 0 ) {
3883 return htmlspecialchars( trim( $show ) );
3887 // Since usually only infinite or indefinite is only on list, so try
3888 // equivalents if still here.
3889 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3890 if ( in_array( $str, $indefs ) ) {
3891 foreach ( $indefs as $val ) {
3892 $show = array_search( $val, $duration, true );
3893 if ( $show !== false ) {
3894 return htmlspecialchars( trim( $show ) );
3899 // If all else fails, return a standard duration or timestamp description.
3900 $time = strtotime( $str, 0 );
3901 if ( $time === false ) { // Unknown format. Return it as-is in case.
3903 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3904 // $time is relative to 0 so it's a duration length.
3905 return $this->formatDuration( $time );
3906 } else { // It's an absolute timestamp.
3907 if ( $time === 0 ) {
3908 // wfTimestamp() handles 0 as current time instead of epoch.
3909 return $this->timeanddate( '19700101000000' );
3911 return $this->timeanddate( $time );
3917 * languages like Chinese need to be segmented in order for the diff
3920 * @param string $text
3923 public function segmentForDiff( $text ) {
3928 * and unsegment to show the result
3930 * @param string $text
3933 public function unsegmentForDiff( $text ) {
3938 * Return the LanguageConverter used in the Language
3941 * @return LanguageConverter
3943 public function getConverter() {
3944 return $this->mConverter
;
3948 * convert text to all supported variants
3950 * @param string $text
3953 public function autoConvertToAllVariants( $text ) {
3954 return $this->mConverter
->autoConvertToAllVariants( $text );
3958 * convert text to different variants of a language.
3960 * @param string $text
3963 public function convert( $text ) {
3964 return $this->mConverter
->convert( $text );
3968 * Convert a Title object to a string in the preferred variant
3970 * @param Title $title
3973 public function convertTitle( $title ) {
3974 return $this->mConverter
->convertTitle( $title );
3978 * Convert a namespace index to a string in the preferred variant
3983 public function convertNamespace( $ns ) {
3984 return $this->mConverter
->convertNamespace( $ns );
3988 * Check if this is a language with variants
3992 public function hasVariants() {
3993 return count( $this->getVariants() ) > 1;
3997 * Check if the language has the specific variant
4000 * @param string $variant
4003 public function hasVariant( $variant ) {
4004 return (bool)$this->mConverter
->validateVariant( $variant );
4008 * Put custom tags (e.g. -{ }-) around math to prevent conversion
4010 * @param string $text
4012 * @deprecated since 1.22 is no longer used
4014 public function armourMath( $text ) {
4015 return $this->mConverter
->armourMath( $text );
4019 * Perform output conversion on a string, and encode for safe HTML output.
4020 * @param string $text Text to be converted
4021 * @param bool $isTitle Whether this conversion is for the article title
4023 * @todo this should get integrated somewhere sane
4025 public function convertHtml( $text, $isTitle = false ) {
4026 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4030 * @param string $key
4033 public function convertCategoryKey( $key ) {
4034 return $this->mConverter
->convertCategoryKey( $key );
4038 * Get the list of variants supported by this language
4039 * see sample implementation in LanguageZh.php
4041 * @return array An array of language codes
4043 public function getVariants() {
4044 return $this->mConverter
->getVariants();
4050 public function getPreferredVariant() {
4051 return $this->mConverter
->getPreferredVariant();
4057 public function getDefaultVariant() {
4058 return $this->mConverter
->getDefaultVariant();
4064 public function getURLVariant() {
4065 return $this->mConverter
->getURLVariant();
4069 * If a language supports multiple variants, it is
4070 * possible that non-existing link in one variant
4071 * actually exists in another variant. this function
4072 * tries to find it. See e.g. LanguageZh.php
4073 * The input parameters may be modified upon return
4075 * @param string &$link The name of the link
4076 * @param Title &$nt The title object of the link
4077 * @param bool $ignoreOtherCond To disable other conditions when
4078 * we need to transclude a template or update a category's link
4080 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4081 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
4085 * returns language specific options used by User::getPageRenderHash()
4086 * for example, the preferred language variant
4090 function getExtraHashOptions() {
4091 return $this->mConverter
->getExtraHashOptions();
4095 * For languages that support multiple variants, the title of an
4096 * article may be displayed differently in different variants. this
4097 * function returns the apporiate title defined in the body of the article.
4101 public function getParsedTitle() {
4102 return $this->mConverter
->getParsedTitle();
4106 * Prepare external link text for conversion. When the text is
4107 * a URL, it shouldn't be converted, and it'll be wrapped in
4108 * the "raw" tag (-{R| }-) to prevent conversion.
4110 * This function is called "markNoConversion" for historical
4113 * @param string $text Text to be used for external link
4114 * @param bool $noParse Wrap it without confirming it's a real URL first
4115 * @return string The tagged text
4117 public function markNoConversion( $text, $noParse = false ) {
4118 // Excluding protocal-relative URLs may avoid many false positives.
4119 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4120 return $this->mConverter
->markNoConversion( $text );
4127 * A regular expression to match legal word-trailing characters
4128 * which should be merged onto a link of the form [[foo]]bar.
4132 public function linkTrail() {
4133 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
4137 * A regular expression character set to match legal word-prefixing
4138 * characters which should be merged onto a link of the form foo[[bar]].
4142 public function linkPrefixCharset() {
4143 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
4147 * @deprecated since 1.24, will be removed in 1.25
4150 function getLangObj() {
4151 wfDeprecated( __METHOD__
, '1.24' );
4156 * Get the "parent" language which has a converter to convert a "compatible" language
4157 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4159 * @return Language|null
4162 public function getParentLanguage() {
4163 if ( $this->mParentLanguage
!== false ) {
4164 return $this->mParentLanguage
;
4167 $pieces = explode( '-', $this->getCode() );
4169 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
4170 $this->mParentLanguage
= null;
4173 $lang = Language
::factory( $code );
4174 if ( !$lang->hasVariant( $this->getCode() ) ) {
4175 $this->mParentLanguage
= null;
4179 $this->mParentLanguage
= $lang;
4184 * Get the RFC 3066 code for this language object
4186 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4187 * htmlspecialchars() or similar
4191 public function getCode() {
4192 return $this->mCode
;
4196 * Get the code in Bcp47 format which we can use
4197 * inside of html lang="" tags.
4199 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4200 * htmlspecialchars() or similar.
4205 public function getHtmlCode() {
4206 if ( is_null( $this->mHtmlCode
) ) {
4207 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4209 return $this->mHtmlCode
;
4213 * @param string $code
4215 public function setCode( $code ) {
4216 $this->mCode
= $code;
4217 // Ensure we don't leave incorrect cached data lying around
4218 $this->mHtmlCode
= null;
4219 $this->mParentLanguage
= false;
4223 * Get the name of a file for a certain language code
4224 * @param string $prefix Prepend this to the filename
4225 * @param string $code Language code
4226 * @param string $suffix Append this to the filename
4227 * @throws MWException
4228 * @return string $prefix . $mangledCode . $suffix
4230 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4231 if ( !self
::isValidBuiltInCode( $code ) ) {
4232 throw new MWException( "Invalid language code \"$code\"" );
4235 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4239 * Get the language code from a file name. Inverse of getFileName()
4240 * @param string $filename $prefix . $languageCode . $suffix
4241 * @param string $prefix Prefix before the language code
4242 * @param string $suffix Suffix after the language code
4243 * @return string Language code, or false if $prefix or $suffix isn't found
4245 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4247 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4248 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4249 if ( !count( $m ) ) {
4252 return str_replace( '_', '-', strtolower( $m[1] ) );
4256 * @param string $code
4259 public static function getMessagesFileName( $code ) {
4261 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4262 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
4267 * @param string $code
4271 public static function getJsonMessagesFileName( $code ) {
4274 if ( !self
::isValidBuiltInCode( $code ) ) {
4275 throw new MWException( "Invalid language code \"$code\"" );
4278 return "$IP/languages/i18n/$code.json";
4282 * @param string $code
4285 public static function getClassFileName( $code ) {
4287 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4291 * Get the first fallback for a given language.
4293 * @param string $code
4295 * @return bool|string
4297 public static function getFallbackFor( $code ) {
4298 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4301 $fallbacks = self
::getFallbacksFor( $code );
4302 $first = array_shift( $fallbacks );
4308 * Get the ordered list of fallback languages.
4311 * @param string $code Language code
4314 public static function getFallbacksFor( $code ) {
4315 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4318 $v = self
::getLocalisationCache()->getItem( $code, 'fallback' );
4319 $v = array_map( 'trim', explode( ',', $v ) );
4320 if ( $v[count( $v ) - 1] !== 'en' ) {
4328 * Get the ordered list of fallback languages, ending with the fallback
4329 * language chain for the site language.
4332 * @param string $code Language code
4333 * @return array Array( fallbacks, site fallbacks )
4335 public static function getFallbacksIncludingSiteLanguage( $code ) {
4336 global $wgLanguageCode;
4338 // Usually, we will only store a tiny number of fallback chains, so we
4339 // keep them in static memory.
4340 $cacheKey = "{$code}-{$wgLanguageCode}";
4342 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4343 $fallbacks = self
::getFallbacksFor( $code );
4345 // Append the site's fallback chain, including the site language itself
4346 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4347 array_unshift( $siteFallbacks, $wgLanguageCode );
4349 // Eliminate any languages already included in the chain
4350 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4352 self
::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4354 return self
::$fallbackLanguageCache[$cacheKey];
4358 * Get all messages for a given language
4359 * WARNING: this may take a long time. If you just need all message *keys*
4360 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4362 * @param string $code
4366 public static function getMessagesFor( $code ) {
4367 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4371 * Get a message for a given language
4373 * @param string $key
4374 * @param string $code
4378 public static function getMessageFor( $key, $code ) {
4379 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4383 * Get all message keys for a given language. This is a faster alternative to
4384 * array_keys( Language::getMessagesFor( $code ) )
4387 * @param string $code Language code
4388 * @return array Array of message keys (strings)
4390 public static function getMessageKeysFor( $code ) {
4391 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4395 * @param string $talk
4398 function fixVariableInNamespace( $talk ) {
4399 if ( strpos( $talk, '$1' ) === false ) {
4403 global $wgMetaNamespace;
4404 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4406 # Allow grammar transformations
4407 # Allowing full message-style parsing would make simple requests
4408 # such as action=raw much more expensive than they need to be.
4409 # This will hopefully cover most cases.
4410 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4411 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4412 return str_replace( ' ', '_', $talk );
4419 function replaceGrammarInNamespace( $m ) {
4420 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4424 * @throws MWException
4427 static function getCaseMaps() {
4428 static $wikiUpperChars, $wikiLowerChars;
4429 if ( isset( $wikiUpperChars ) ) {
4430 return array( $wikiUpperChars, $wikiLowerChars );
4433 wfProfileIn( __METHOD__
);
4434 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4435 if ( $arr === false ) {
4436 throw new MWException(
4437 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4439 $wikiUpperChars = $arr['wikiUpperChars'];
4440 $wikiLowerChars = $arr['wikiLowerChars'];
4441 wfProfileOut( __METHOD__
);
4442 return array( $wikiUpperChars, $wikiLowerChars );
4446 * Decode an expiry (block, protection, etc) which has come from the DB
4448 * @todo FIXME: why are we returnings DBMS-dependent strings???
4450 * @param string $expiry Database expiry String
4451 * @param bool|int $format True to process using language functions, or TS_ constant
4452 * to return the expiry in a given timestamp
4456 public function formatExpiry( $expiry, $format = true ) {
4458 if ( $infinity === null ) {
4459 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
4462 if ( $expiry == '' ||
$expiry == $infinity ) {
4463 return $format === true
4464 ?
$this->getMessageFromDB( 'infiniteblock' )
4467 return $format === true
4468 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4469 : wfTimestamp( $format, $expiry );
4475 * @param int|float $seconds
4476 * @param array $format Optional
4477 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4478 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4479 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4481 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4482 * or 'avoidminutes'.
4485 function formatTimePeriod( $seconds, $format = array() ) {
4486 if ( !is_array( $format ) ) {
4487 $format = array( 'avoid' => $format ); // For backwards compatibility
4489 if ( !isset( $format['avoid'] ) ) {
4490 $format['avoid'] = false;
4492 if ( !isset( $format['noabbrevs'] ) ) {
4493 $format['noabbrevs'] = false;
4495 $secondsMsg = wfMessage(
4496 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4497 $minutesMsg = wfMessage(
4498 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4499 $hoursMsg = wfMessage(
4500 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4501 $daysMsg = wfMessage(
4502 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4504 if ( round( $seconds * 10 ) < 100 ) {
4505 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4506 $s = $secondsMsg->params( $s )->text();
4507 } elseif ( round( $seconds ) < 60 ) {
4508 $s = $this->formatNum( round( $seconds ) );
4509 $s = $secondsMsg->params( $s )->text();
4510 } elseif ( round( $seconds ) < 3600 ) {
4511 $minutes = floor( $seconds / 60 );
4512 $secondsPart = round( fmod( $seconds, 60 ) );
4513 if ( $secondsPart == 60 ) {
4517 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4519 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4520 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4521 $hours = floor( $seconds / 3600 );
4522 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4523 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4524 if ( $secondsPart == 60 ) {
4528 if ( $minutes == 60 ) {
4532 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4534 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4535 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4536 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4539 $days = floor( $seconds / 86400 );
4540 if ( $format['avoid'] === 'avoidminutes' ) {
4541 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4542 if ( $hours == 24 ) {
4546 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4548 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4549 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4550 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4551 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4552 if ( $minutes == 60 ) {
4556 if ( $hours == 24 ) {
4560 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4562 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4564 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4566 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4568 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4575 * Format a bitrate for output, using an appropriate
4576 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4577 * the magnitude in question.
4579 * This use base 1000. For base 1024 use formatSize(), for another base
4580 * see formatComputingNumbers().
4585 function formatBitrate( $bps ) {
4586 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4590 * @param int $size Size of the unit
4591 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4592 * @param string $messageKey Message key to be uesd
4595 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4597 return str_replace( '$1', $this->formatNum( $size ),
4598 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4601 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4604 $maxIndex = count( $sizes ) - 1;
4605 while ( $size >= $boundary && $index < $maxIndex ) {
4610 // For small sizes no decimal places necessary
4613 // For MB and bigger two decimal places are smarter
4616 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4618 $size = round( $size, $round );
4619 $text = $this->getMessageFromDB( $msg );
4620 return str_replace( '$1', $this->formatNum( $size ), $text );
4624 * Format a size in bytes for output, using an appropriate
4625 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4627 * This method use base 1024. For base 1000 use formatBitrate(), for
4628 * another base see formatComputingNumbers()
4630 * @param int $size Size to format
4631 * @return string Plain text (not HTML)
4633 function formatSize( $size ) {
4634 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4638 * Make a list item, used by various special pages
4640 * @param string $page Page link
4641 * @param string $details Text between brackets
4642 * @param bool $oppositedm Add the direction mark opposite to your
4643 * language, to display text properly
4646 function specialList( $page, $details, $oppositedm = true ) {
4647 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) .
4648 $this->getDirMark();
4649 $details = $details ?
$dirmark . $this->getMessageFromDB( 'word-separator' ) .
4650 wfMessage( 'parentheses' )->rawParams( $details )->inLanguage( $this )->escaped() : '';
4651 return $page . $details;
4655 * Generate (prev x| next x) (20|50|100...) type links for paging
4657 * @param Title $title Title object to link
4658 * @param int $offset
4660 * @param array $query Optional URL query parameter string
4661 * @param bool $atend Optional param for specified if this is the last page
4664 public function viewPrevNext( Title
$title, $offset, $limit,
4665 array $query = array(), $atend = false
4667 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4669 # Make 'previous' link
4670 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4671 if ( $offset > 0 ) {
4672 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4673 $query, $prev, 'prevn-title', 'mw-prevlink' );
4675 $plink = htmlspecialchars( $prev );
4679 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4681 $nlink = htmlspecialchars( $next );
4683 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4684 $query, $next, 'nextn-title', 'mw-nextlink' );
4687 # Make links to set number of items per page
4688 $numLinks = array();
4689 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4690 $numLinks[] = $this->numLink( $title, $offset, $num,
4691 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4694 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4695 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4699 * Helper function for viewPrevNext() that generates links
4701 * @param Title $title Title object to link
4702 * @param int $offset
4704 * @param array $query Extra query parameters
4705 * @param string $link Text to use for the link; will be escaped
4706 * @param string $tooltipMsg Name of the message to use as tooltip
4707 * @param string $class Value of the "class" attribute of the link
4708 * @return string HTML fragment
4710 private function numLink( Title
$title, $offset, $limit, array $query, $link,
4713 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4714 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4715 ->numParams( $limit )->text();
4717 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4718 'title' => $tooltip, 'class' => $class ), $link );
4722 * Get the conversion rule title, if any.
4726 public function getConvRuleTitle() {
4727 return $this->mConverter
->getConvRuleTitle();
4731 * Get the compiled plural rules for the language
4733 * @return array Associative array with plural form, and plural rule as key-value pairs
4735 public function getCompiledPluralRules() {
4736 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4737 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4738 if ( !$pluralRules ) {
4739 foreach ( $fallbacks as $fallbackCode ) {
4740 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4741 if ( $pluralRules ) {
4746 return $pluralRules;
4750 * Get the plural rules for the language
4752 * @return array Associative array with plural form number and plural rule as key-value pairs
4754 public function getPluralRules() {
4755 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4756 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4757 if ( !$pluralRules ) {
4758 foreach ( $fallbacks as $fallbackCode ) {
4759 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4760 if ( $pluralRules ) {
4765 return $pluralRules;
4769 * Get the plural rule types for the language
4771 * @return array Associative array with plural form number and plural rule type as key-value pairs
4773 public function getPluralRuleTypes() {
4774 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4775 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4776 if ( !$pluralRuleTypes ) {
4777 foreach ( $fallbacks as $fallbackCode ) {
4778 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4779 if ( $pluralRuleTypes ) {
4784 return $pluralRuleTypes;
4788 * Find the index number of the plural rule appropriate for the given number
4789 * @param int $number
4790 * @return int The index number of the plural rule
4792 public function getPluralRuleIndexNumber( $number ) {
4793 $pluralRules = $this->getCompiledPluralRules();
4794 $form = CLDRPluralRuleEvaluator
::evaluateCompiled( $number, $pluralRules );
4799 * Find the plural rule type appropriate for the given number
4800 * For example, if the language is set to Arabic, getPluralType(5) should
4803 * @param int $number
4804 * @return string The name of the plural rule type, e.g. one, two, few, many
4806 public function getPluralRuleType( $number ) {
4807 $index = $this->getPluralRuleIndexNumber( $number );
4808 $pluralRuleTypes = $this->getPluralRuleTypes();
4809 if ( isset( $pluralRuleTypes[$index] ) ) {
4810 return $pluralRuleTypes[$index];