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 * a fake language converter
47 function __construct( $langobj ) { $this->mLang
= $langobj; }
48 function autoConvert( $text, $variant = false ) { return $text; }
49 function autoConvertToAllVariants( $text ) { return array( $this->mLang
->getCode() => $text ); }
50 function convert( $t ) { return $t; }
51 function convertTo( $text, $variant ) { return $text; }
52 function convertTitle( $t ) { return $t->getPrefixedText(); }
53 function convertNamespace( $ns ) { return $this->mLang
->getFormattedNsText( $ns ); }
54 function getVariants() { return array( $this->mLang
->getCode() ); }
55 function getVariantFallbacks( $variant ) { return $this->mLang
->getCode(); }
56 function getPreferredVariant() { return $this->mLang
->getCode(); }
57 function getDefaultVariant() { return $this->mLang
->getCode(); }
58 function getURLVariant() { return ''; }
59 function getConvRuleTitle() { return false; }
60 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
61 function getExtraHashOptions() { return ''; }
62 function getParsedTitle() { return ''; }
63 function markNoConversion( $text, $noParse = false ) { return $text; }
64 function convertCategoryKey( $key ) { return $key; }
65 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
66 /** @deprecated since 1.22 is no longer used */
67 function armourMath( $text ) { return $text; }
68 function validateVariant( $variant = null ) { return $variant === $this->mLang
->getCode() ?
$variant : null; }
69 function translate( $text, $variant ) { return $text; }
73 * Internationalisation code
79 * @var LanguageConverter
83 public $mVariants, $mCode, $mLoaded = false;
84 public $mMagicExtensions = array(), $mMagicHookDone = false;
85 private $mHtmlCode = null, $mParentLanguage = false;
87 public $dateFormatStrings = array();
88 public $mExtendedSpecialPageAliases;
90 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
93 * ReplacementArray object caches
95 public $transformData = array();
98 * @var LocalisationCache
100 static public $dataCache;
102 static public $mLangObjCache = array();
104 static public $mWeekdayMsgs = array(
105 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
109 static public $mWeekdayAbbrevMsgs = array(
110 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
113 static public $mMonthMsgs = array(
114 'january', 'february', 'march', 'april', 'may_long', 'june',
115 'july', 'august', 'september', 'october', 'november',
118 static public $mMonthGenMsgs = array(
119 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
120 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
123 static public $mMonthAbbrevMsgs = array(
124 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
125 'sep', 'oct', 'nov', 'dec'
128 static public $mIranianCalendarMonthMsgs = array(
129 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
130 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
131 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
132 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
135 static public $mHebrewCalendarMonthMsgs = array(
136 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
137 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
138 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
139 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
140 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
143 static public $mHebrewCalendarMonthGenMsgs = array(
144 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
145 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
146 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
147 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
148 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
151 static public $mHijriCalendarMonthMsgs = array(
152 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
153 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
154 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
155 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
162 static public $durationIntervals = array(
163 'millennia' => 31556952000,
164 'centuries' => 3155695200,
165 'decades' => 315569520,
166 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
175 * Cache for language fallbacks.
176 * @see Language::getFallbacksIncludingSiteLanguage
180 static private $fallbackLanguageCache = array();
183 * Get a cached or new language object for a given language code
184 * @param $code String
187 static function factory( $code ) {
188 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
190 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
191 $code = $wgDummyLanguageCodes[$code];
194 // get the language object to process
195 $langObj = isset( self
::$mLangObjCache[$code] )
196 ? self
::$mLangObjCache[$code]
197 : self
::newFromCode( $code );
199 // merge the language object in to get it up front in the cache
200 self
::$mLangObjCache = array_merge( array( $code => $langObj ), self
::$mLangObjCache );
201 // get rid of the oldest ones in case we have an overflow
202 self
::$mLangObjCache = array_slice( self
::$mLangObjCache, 0, $wgLangObjCacheSize, true );
208 * Create a language object for a given language code
209 * @param $code String
210 * @throws MWException
213 protected static function newFromCode( $code ) {
214 // Protect against path traversal below
215 if ( !Language
::isValidCode( $code )
216 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
218 throw new MWException( "Invalid language code \"$code\"" );
221 if ( !Language
::isValidBuiltInCode( $code ) ) {
222 // It's not possible to customise this code with class files, so
223 // just return a Language object. This is to support uselang= hacks.
224 $lang = new Language
;
225 $lang->setCode( $code );
229 // Check if there is a language class for the code
230 $class = self
::classFromCode( $code );
231 self
::preloadLanguageClass( $class );
232 if ( class_exists( $class ) ) {
237 // Keep trying the fallback list until we find an existing class
238 $fallbacks = Language
::getFallbacksFor( $code );
239 foreach ( $fallbacks as $fallbackCode ) {
240 if ( !Language
::isValidBuiltInCode( $fallbackCode ) ) {
241 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
244 $class = self
::classFromCode( $fallbackCode );
245 self
::preloadLanguageClass( $class );
246 if ( class_exists( $class ) ) {
247 $lang = Language
::newFromCode( $fallbackCode );
248 $lang->setCode( $code );
253 throw new MWException( "Invalid fallback sequence for language '$code'" );
257 * Checks whether any localisation is available for that language tag
258 * in MediaWiki (MessagesXx.php exists).
260 * @param string $code Language tag (in lower case)
261 * @return bool Whether language is supported
264 public static function isSupportedLanguage( $code ) {
265 return $code === strtolower( $code ) && is_readable( self
::getMessagesFileName( $code ) );
269 * Returns true if a language code string is a well-formed language tag
270 * according to RFC 5646.
271 * This function only checks well-formedness; it doesn't check that
272 * language, script or variant codes actually exist in the repositories.
274 * Based on regexes by Mark Davis of the Unicode Consortium:
275 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
277 * @param $code string
278 * @param $lenient boolean Whether to allow '_' as separator. The default is only '-'.
283 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
286 $alphanum = '[a-z0-9]';
287 $x = 'x'; # private use singleton
288 $singleton = '[a-wy-z]'; # other singleton
289 $s = $lenient ?
'[-_]' : '-';
291 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
292 $script = "$alpha{4}"; # ISO 15924
293 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
294 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
295 $extension = "$singleton(?:$s$alphanum{2,8})+";
296 $privateUse = "$x(?:$s$alphanum{1,8})+";
298 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
299 # Since these are limited, this is safe even later changes to the registry --
300 # the only oddity is that it might change the type of the tag, and thus
301 # the results from the capturing groups.
302 # http://www.iana.org/assignments/language-subtag-registry
304 $grandfathered = "en{$s}GB{$s}oed"
305 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
306 . "|no{$s}(?:bok|nyn)"
307 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
308 . "|zh{$s}min{$s}nan";
310 $variantList = "$variant(?:$s$variant)*";
311 $extensionList = "$extension(?:$s$extension)*";
313 $langtag = "(?:($language)"
316 . "(?:$s$variantList)?"
317 . "(?:$s$extensionList)?"
318 . "(?:$s$privateUse)?)";
320 # The final breakdown, with capturing groups for each of these components
321 # The variants, extensions, grandfathered, and private-use may have interior '-'
323 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
325 return (bool)preg_match( "/$root/", strtolower( $code ) );
329 * Returns true if a language code string is of a valid form, whether or
330 * not it exists. This includes codes which are used solely for
331 * customisation via the MediaWiki namespace.
333 * @param $code string
337 public static function isValidCode( $code ) {
338 static $cache = array();
339 if ( isset( $cache[$code] ) ) {
340 return $cache[$code];
342 // People think language codes are html safe, so enforce it.
343 // Ideally we should only allow a-zA-Z0-9-
344 // but, .+ and other chars are often used for {{int:}} hacks
345 // see bugs 37564, 37587, 36938
347 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
348 && !preg_match( Title
::getTitleInvalidRegex(), $code );
350 return $cache[$code];
354 * Returns true if a language code is of a valid form for the purposes of
355 * internal customisation of MediaWiki, via Messages*.php.
357 * @param $code string
359 * @throws MWException
363 public static function isValidBuiltInCode( $code ) {
365 if ( !is_string( $code ) ) {
366 if ( is_object( $code ) ) {
367 $addmsg = " of class " . get_class( $code );
371 $type = gettype( $code );
372 throw new MWException( __METHOD__
. " must be passed a string, $type given$addmsg" );
375 return (bool)preg_match( '/^[a-z0-9-]{2,}$/i', $code );
379 * Returns true if a language code is an IETF tag known to MediaWiki.
381 * @param $code string
386 public static function isKnownLanguageTag( $tag ) {
387 static $coreLanguageNames;
389 // Quick escape for invalid input to avoid exceptions down the line
390 // when code tries to process tags which are not valid at all.
391 if ( !self
::isValidBuiltInCode( $tag ) ) {
395 if ( $coreLanguageNames === null ) {
397 include "$IP/languages/Names.php";
400 if ( isset( $coreLanguageNames[$tag] )
401 || self
::fetchLanguageName( $tag, $tag ) !== ''
411 * @return String Name of the language class
413 public static function classFromCode( $code ) {
414 if ( $code == 'en' ) {
417 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
422 * Includes language class files
424 * @param $class string Name of the language class
426 public static function preloadLanguageClass( $class ) {
429 if ( $class === 'Language' ) {
433 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
434 include_once "$IP/languages/classes/$class.php";
439 * Get the LocalisationCache instance
441 * @return LocalisationCache
443 public static function getLocalisationCache() {
444 if ( is_null( self
::$dataCache ) ) {
445 global $wgLocalisationCacheConf;
446 $class = $wgLocalisationCacheConf['class'];
447 self
::$dataCache = new $class( $wgLocalisationCacheConf );
449 return self
::$dataCache;
452 function __construct() {
453 $this->mConverter
= new FakeConverter( $this );
454 // Set the code to the name of the descendant
455 if ( get_class( $this ) == 'Language' ) {
458 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
460 self
::getLocalisationCache();
464 * Reduce memory usage
466 function __destruct() {
467 foreach ( $this as $name => $value ) {
468 unset( $this->$name );
473 * Hook which will be called if this is the content language.
474 * Descendants can use this to register hook functions or modify globals
476 function initContLang() { }
479 * Same as getFallbacksFor for current language.
481 * @deprecated in 1.19
483 function getFallbackLanguageCode() {
484 wfDeprecated( __METHOD__
, '1.19' );
485 return self
::getFallbackFor( $this->mCode
);
492 function getFallbackLanguages() {
493 return self
::getFallbacksFor( $this->mCode
);
497 * Exports $wgBookstoreListEn
500 function getBookstoreList() {
501 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
505 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
506 * available in localised form, it will be included in English.
510 public function getNamespaces() {
511 if ( is_null( $this->namespaceNames
) ) {
512 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
514 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
515 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
517 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
519 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
520 if ( $wgMetaNamespaceTalk ) {
521 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
523 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
524 $this->namespaceNames
[NS_PROJECT_TALK
] =
525 $this->fixVariableInNamespace( $talk );
528 # Sometimes a language will be localised but not actually exist on this wiki.
529 foreach ( $this->namespaceNames
as $key => $text ) {
530 if ( !isset( $validNamespaces[$key] ) ) {
531 unset( $this->namespaceNames
[$key] );
535 # The above mixing may leave namespaces out of canonical order.
536 # Re-order by namespace ID number...
537 ksort( $this->namespaceNames
);
539 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames
) );
541 return $this->namespaceNames
;
545 * Arbitrarily set all of the namespace names at once. Mainly used for testing
546 * @param $namespaces Array of namespaces (id => name)
548 public function setNamespaces( array $namespaces ) {
549 $this->namespaceNames
= $namespaces;
550 $this->mNamespaceIds
= null;
554 * Resets all of the namespace caches. Mainly used for testing
556 public function resetNamespaces() {
557 $this->namespaceNames
= null;
558 $this->mNamespaceIds
= null;
559 $this->namespaceAliases
= null;
563 * A convenience function that returns the same thing as
564 * getNamespaces() except with the array values changed to ' '
565 * where it found '_', useful for producing output to be displayed
566 * e.g. in <select> forms.
570 function getFormattedNamespaces() {
571 $ns = $this->getNamespaces();
572 foreach ( $ns as $k => $v ) {
573 $ns[$k] = strtr( $v, '_', ' ' );
579 * Get a namespace value by key
581 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
582 * echo $mw_ns; // prints 'MediaWiki'
585 * @param $index Int: the array key of the namespace to return
586 * @return mixed, string if the namespace value exists, otherwise false
588 function getNsText( $index ) {
589 $ns = $this->getNamespaces();
590 return isset( $ns[$index] ) ?
$ns[$index] : false;
594 * A convenience function that returns the same thing as
595 * getNsText() except with '_' changed to ' ', useful for
599 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
600 * echo $mw_ns; // prints 'MediaWiki talk'
603 * @param int $index The array key of the namespace to return
604 * @return string Namespace name without underscores (empty string if namespace does not exist)
606 function getFormattedNsText( $index ) {
607 $ns = $this->getNsText( $index );
608 return strtr( $ns, '_', ' ' );
612 * Returns gender-dependent namespace alias if available.
613 * @param $index Int: namespace index
614 * @param $gender String: gender key (male, female... )
618 function getGenderNsText( $index, $gender ) {
619 global $wgExtraGenderNamespaces;
621 $ns = $wgExtraGenderNamespaces + self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
622 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
626 * Whether this language makes distinguishes genders for example in
631 function needsGenderDistinction() {
632 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
633 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
634 // $wgExtraGenderNamespaces overrides everything
636 } elseif ( isset( $wgExtraNamespaces[NS_USER
] ) && isset( $wgExtraNamespaces[NS_USER_TALK
] ) ) {
637 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
638 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
641 // Check what is in i18n files
642 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
643 return count( $aliases ) > 0;
648 * Get a namespace key by value, case insensitive.
649 * Only matches namespace names for the current language, not the
650 * canonical ones defined in Namespace.php.
652 * @param $text String
653 * @return mixed An integer if $text is a valid value otherwise false
655 function getLocalNsIndex( $text ) {
656 $lctext = $this->lc( $text );
657 $ids = $this->getNamespaceIds();
658 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
664 function getNamespaceAliases() {
665 if ( is_null( $this->namespaceAliases
) ) {
666 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
670 foreach ( $aliases as $name => $index ) {
671 if ( $index === NS_PROJECT_TALK
) {
672 unset( $aliases[$name] );
673 $name = $this->fixVariableInNamespace( $name );
674 $aliases[$name] = $index;
679 global $wgExtraGenderNamespaces;
680 $genders = $wgExtraGenderNamespaces +
(array)self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
681 foreach ( $genders as $index => $forms ) {
682 foreach ( $forms as $alias ) {
683 $aliases[$alias] = $index;
687 # Also add converted namespace names as aliases, to avoid confusion.
688 $convertedNames = array();
689 foreach ( $this->getVariants() as $variant ) {
690 if ( $variant === $this->mCode
) {
693 foreach ( $this->getNamespaces() as $ns => $_ ) {
694 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
698 $this->namespaceAliases
= $aliases +
$convertedNames;
700 return $this->namespaceAliases
;
706 function getNamespaceIds() {
707 if ( is_null( $this->mNamespaceIds
) ) {
708 global $wgNamespaceAliases;
709 # Put namespace names and aliases into a hashtable.
710 # If this is too slow, then we should arrange it so that it is done
711 # before caching. The catch is that at pre-cache time, the above
712 # class-specific fixup hasn't been done.
713 $this->mNamespaceIds
= array();
714 foreach ( $this->getNamespaces() as $index => $name ) {
715 $this->mNamespaceIds
[$this->lc( $name )] = $index;
717 foreach ( $this->getNamespaceAliases() as $name => $index ) {
718 $this->mNamespaceIds
[$this->lc( $name )] = $index;
720 if ( $wgNamespaceAliases ) {
721 foreach ( $wgNamespaceAliases as $name => $index ) {
722 $this->mNamespaceIds
[$this->lc( $name )] = $index;
726 return $this->mNamespaceIds
;
730 * Get a namespace key by value, case insensitive. Canonical namespace
731 * names override custom ones defined for the current language.
733 * @param $text String
734 * @return mixed An integer if $text is a valid value otherwise false
736 function getNsIndex( $text ) {
737 $lctext = $this->lc( $text );
738 $ns = MWNamespace
::getCanonicalIndex( $lctext );
739 if ( $ns !== null ) {
742 $ids = $this->getNamespaceIds();
743 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
747 * short names for language variants used for language conversion links.
749 * @param $code String
750 * @param $usemsg bool Use the "variantname-xyz" message if it exists
753 function getVariantname( $code, $usemsg = true ) {
754 $msg = "variantname-$code";
755 if ( $usemsg && wfMessage( $msg )->exists() ) {
756 return $this->getMessageFromDB( $msg );
758 $name = self
::fetchLanguageName( $code );
760 return $name; # if it's defined as a language name, show that
762 # otherwise, output the language code
768 * @param $name string
771 function specialPage( $name ) {
772 $aliases = $this->getSpecialPageAliases();
773 if ( isset( $aliases[$name][0] ) ) {
774 $name = $aliases[$name][0];
776 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
782 function getDatePreferences() {
783 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
789 function getDateFormats() {
790 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
794 * @return array|string
796 function getDefaultDateFormat() {
797 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
798 if ( $df === 'dmy or mdy' ) {
799 global $wgAmericanDates;
800 return $wgAmericanDates ?
'mdy' : 'dmy';
809 function getDatePreferenceMigrationMap() {
810 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
817 function getImageFile( $image ) {
818 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
824 function getExtraUserToggles() {
825 return (array)self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
832 function getUserToggle( $tog ) {
833 return $this->getMessageFromDB( "tog-$tog" );
837 * Get native language names, indexed by code.
838 * Only those defined in MediaWiki, no other data like CLDR.
839 * If $customisedOnly is true, only returns codes with a messages file
841 * @param $customisedOnly bool
844 * @deprecated in 1.20, use fetchLanguageNames()
846 public static function getLanguageNames( $customisedOnly = false ) {
847 return self
::fetchLanguageNames( null, $customisedOnly ?
'mwfile' : 'mw' );
851 * Get translated language names. This is done on best effort and
852 * by default this is exactly the same as Language::getLanguageNames.
853 * The CLDR extension provides translated names.
854 * @param $code String Language code.
855 * @return Array language code => language name
857 * @deprecated in 1.20, use fetchLanguageNames()
859 public static function getTranslatedLanguageNames( $code ) {
860 return self
::fetchLanguageNames( $code, 'all' );
864 * Get an array of language names, indexed by code.
865 * @param $inLanguage null|string: Code of language in which to return the names
866 * Use null for autonyms (native names)
867 * @param $include string:
868 * 'all' all available languages
869 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
870 * 'mwfile' only if the language is in 'mw' *and* has a message file
871 * @return array: language code => language name
874 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
875 global $wgExtraLanguageNames;
876 static $coreLanguageNames;
878 if ( $coreLanguageNames === null ) {
880 include "$IP/languages/Names.php";
886 # TODO: also include when $inLanguage is null, when this code is more efficient
887 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
890 $mwNames = $wgExtraLanguageNames +
$coreLanguageNames;
891 foreach ( $mwNames as $mwCode => $mwName ) {
892 # - Prefer own MediaWiki native name when not using the hook
893 # - For other names just add if not added through the hook
894 if ( $mwCode === $inLanguage ||
!isset( $names[$mwCode] ) ) {
895 $names[$mwCode] = $mwName;
899 if ( $include === 'all' ) {
904 $coreCodes = array_keys( $mwNames );
905 foreach ( $coreCodes as $coreCode ) {
906 $returnMw[$coreCode] = $names[$coreCode];
909 if ( $include === 'mwfile' ) {
910 $namesMwFile = array();
911 # We do this using a foreach over the codes instead of a directory
912 # loop so that messages files in extensions will work correctly.
913 foreach ( $returnMw as $code => $value ) {
914 if ( is_readable( self
::getMessagesFileName( $code ) ) ) {
915 $namesMwFile[$code] = $names[$code];
920 # 'mw' option; default if it's not one of the other two options (all/mwfile)
925 * @param $code string: The code of the language for which to get the name
926 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
927 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
928 * @return string: Language name or empty
931 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
932 $array = self
::fetchLanguageNames( $inLanguage, $include );
933 return !array_key_exists( $code, $array ) ?
'' : $array[$code];
937 * Get a message from the MediaWiki namespace.
939 * @param $msg String: message name
942 function getMessageFromDB( $msg ) {
943 return wfMessage( $msg )->inLanguage( $this )->text();
947 * Get the native language name of $code.
948 * Only if defined in MediaWiki, no other data like CLDR.
949 * @param $code string
951 * @deprecated in 1.20, use fetchLanguageName()
953 function getLanguageName( $code ) {
954 return self
::fetchLanguageName( $code );
961 function getMonthName( $key ) {
962 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
968 function getMonthNamesArray() {
969 $monthNames = array( '' );
970 for ( $i = 1; $i < 13; $i++
) {
971 $monthNames[] = $this->getMonthName( $i );
980 function getMonthNameGen( $key ) {
981 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
988 function getMonthAbbreviation( $key ) {
989 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
995 function getMonthAbbreviationsArray() {
996 $monthNames = array( '' );
997 for ( $i = 1; $i < 13; $i++
) {
998 $monthNames[] = $this->getMonthAbbreviation( $i );
1004 * @param $key string
1007 function getWeekdayName( $key ) {
1008 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
1012 * @param $key string
1015 function getWeekdayAbbreviation( $key ) {
1016 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
1020 * @param $key string
1023 function getIranianCalendarMonthName( $key ) {
1024 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
1028 * @param $key string
1031 function getHebrewCalendarMonthName( $key ) {
1032 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
1036 * @param $key string
1039 function getHebrewCalendarMonthNameGen( $key ) {
1040 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1044 * @param $key string
1047 function getHijriCalendarMonthName( $key ) {
1048 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
1052 * This is a workalike of PHP's date() function, but with better
1053 * internationalisation, a reduced set of format characters, and a better
1056 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1057 * the PHP manual for definitions. There are a number of extensions, which
1060 * xn Do not translate digits of the next numeric format character
1061 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1062 * xr Use roman numerals for the next numeric format character
1063 * xh Use hebrew numerals for the next numeric format character
1065 * xg Genitive month name
1067 * xij j (day number) in Iranian calendar
1068 * xiF F (month name) in Iranian calendar
1069 * xin n (month number) in Iranian calendar
1070 * xiy y (two digit year) in Iranian calendar
1071 * xiY Y (full year) in Iranian calendar
1073 * xjj j (day number) in Hebrew calendar
1074 * xjF F (month name) in Hebrew calendar
1075 * xjt t (days in month) in Hebrew calendar
1076 * xjx xg (genitive month name) in Hebrew calendar
1077 * xjn n (month number) in Hebrew calendar
1078 * xjY Y (full year) in Hebrew calendar
1080 * xmj j (day number) in Hijri calendar
1081 * xmF F (month name) in Hijri calendar
1082 * xmn n (month number) in Hijri calendar
1083 * xmY Y (full year) in Hijri calendar
1085 * xkY Y (full year) in Thai solar calendar. Months and days are
1086 * identical to the Gregorian calendar
1087 * xoY Y (full year) in Minguo calendar or Juche year.
1088 * Months and days are identical to the
1089 * Gregorian calendar
1090 * xtY Y (full year) in Japanese nengo. Months and days are
1091 * identical to the Gregorian calendar
1093 * Characters enclosed in double quotes will be considered literal (with
1094 * the quotes themselves removed). Unmatched quotes will be considered
1095 * literal quotes. Example:
1097 * "The month is" F => The month is January
1100 * Backslash escaping is also supported.
1102 * Input timestamp is assumed to be pre-normalized to the desired local
1103 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1104 * $ts is UTC if $zone is not given.
1106 * @param $format String
1107 * @param $ts String: 14-character timestamp
1110 * @param $zone DateTimeZone: Timezone of $ts
1111 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1113 * @throws MWException
1116 function sprintfDate( $format, $ts, DateTimeZone
$zone = null ) {
1121 $dateTimeObj = false;
1130 if ( strlen( $ts ) !== 14 ) {
1131 throw new MWException( __METHOD__
. ": The timestamp $ts should have 14 characters" );
1134 if ( !ctype_digit( $ts ) ) {
1135 throw new MWException( __METHOD__
. ": The timestamp $ts should be a number" );
1138 for ( $p = 0; $p < strlen( $format ); $p++
) {
1140 $code = $format[$p];
1141 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
1142 $code .= $format[++
$p];
1145 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
1146 $code .= $format[++
$p];
1157 $rawToggle = !$rawToggle;
1166 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1170 $hebrew = self
::tsToHebrew( $ts );
1172 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1175 $num = substr( $ts, 6, 2 );
1178 if ( !$dateTimeObj ) {
1179 $dateTimeObj = DateTime
::createFromFormat(
1180 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1183 $s .= $this->getWeekdayAbbreviation( $dateTimeObj->format( 'w' ) +
1 );
1186 $num = intval( substr( $ts, 6, 2 ) );
1190 $iranian = self
::tsToIranian( $ts );
1196 $hijri = self
::tsToHijri( $ts );
1202 $hebrew = self
::tsToHebrew( $ts );
1207 if ( !$dateTimeObj ) {
1208 $dateTimeObj = DateTime
::createFromFormat(
1209 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1212 $s .= $this->getWeekdayName( $dateTimeObj->format( 'w' ) +
1 );
1215 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1219 $iranian = self
::tsToIranian( $ts );
1221 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1225 $hijri = self
::tsToHijri( $ts );
1227 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1231 $hebrew = self
::tsToHebrew( $ts );
1233 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1236 $num = substr( $ts, 4, 2 );
1239 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1242 $num = intval( substr( $ts, 4, 2 ) );
1246 $iranian = self
::tsToIranian( $ts );
1252 $hijri = self
::tsToHijri ( $ts );
1258 $hebrew = self
::tsToHebrew( $ts );
1264 $hebrew = self
::tsToHebrew( $ts );
1269 $num = substr( $ts, 0, 4 );
1273 $iranian = self
::tsToIranian( $ts );
1279 $hijri = self
::tsToHijri( $ts );
1285 $hebrew = self
::tsToHebrew( $ts );
1291 $thai = self
::tsToYear( $ts, 'thai' );
1297 $minguo = self
::tsToYear( $ts, 'minguo' );
1303 $tenno = self
::tsToYear( $ts, 'tenno' );
1308 $num = substr( $ts, 2, 2 );
1312 $iranian = self
::tsToIranian( $ts );
1314 $num = substr( $iranian[0], -2 );
1317 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1320 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1323 $h = substr( $ts, 8, 2 );
1324 $num = $h %
12 ?
$h %
12 : 12;
1327 $num = intval( substr( $ts, 8, 2 ) );
1330 $h = substr( $ts, 8, 2 );
1331 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1334 $num = substr( $ts, 8, 2 );
1337 $num = substr( $ts, 10, 2 );
1340 $num = substr( $ts, 12, 2 );
1348 // Pass through string from $dateTimeObj->format()
1349 if ( !$dateTimeObj ) {
1350 $dateTimeObj = DateTime
::createFromFormat(
1351 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1354 $s .= $dateTimeObj->format( $code );
1366 // Pass through number from $dateTimeObj->format()
1367 if ( !$dateTimeObj ) {
1368 $dateTimeObj = DateTime
::createFromFormat(
1369 'YmdHis', $ts, $zone ?
: new DateTimeZone( 'UTC' )
1372 $num = $dateTimeObj->format( $code );
1375 # Backslash escaping
1376 if ( $p < strlen( $format ) - 1 ) {
1377 $s .= $format[++
$p];
1384 if ( $p < strlen( $format ) - 1 ) {
1385 $endQuote = strpos( $format, '"', $p +
1 );
1386 if ( $endQuote === false ) {
1387 # No terminating quote, assume literal "
1390 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1394 # Quote at end of string, assume literal "
1401 if ( $num !== false ) {
1402 if ( $rawToggle ||
$raw ) {
1405 } elseif ( $roman ) {
1406 $s .= Language
::romanNumeral( $num );
1408 } elseif ( $hebrewNum ) {
1409 $s .= self
::hebrewNumeral( $num );
1412 $s .= $this->formatNum( $num, true );
1419 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1420 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1423 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1424 * Gregorian dates to Iranian dates. Originally written in C, it
1425 * is released under the terms of GNU Lesser General Public
1426 * License. Conversion to PHP was performed by Niklas Laxström.
1428 * Link: http://www.farsiweb.info/jalali/jalali.c
1434 private static function tsToIranian( $ts ) {
1435 $gy = substr( $ts, 0, 4 ) -1600;
1436 $gm = substr( $ts, 4, 2 ) -1;
1437 $gd = substr( $ts, 6, 2 ) -1;
1439 # Days passed from the beginning (including leap years)
1441 +
floor( ( $gy +
3 ) / 4 )
1442 - floor( ( $gy +
99 ) / 100 )
1443 +
floor( ( $gy +
399 ) / 400 );
1445 // Add days of the past months of this year
1446 for ( $i = 0; $i < $gm; $i++
) {
1447 $gDayNo +
= self
::$GREG_DAYS[$i];
1451 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1455 // Days passed in current month
1456 $gDayNo +
= (int)$gd;
1458 $jDayNo = $gDayNo - 79;
1460 $jNp = floor( $jDayNo / 12053 );
1463 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1466 if ( $jDayNo >= 366 ) {
1467 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1468 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1471 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1472 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1478 return array( $jy, $jm, $jd );
1482 * Converting Gregorian dates to Hijri dates.
1484 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1486 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1492 private static function tsToHijri( $ts ) {
1493 $year = substr( $ts, 0, 4 );
1494 $month = substr( $ts, 4, 2 );
1495 $day = substr( $ts, 6, 2 );
1503 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1504 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1507 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1508 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1509 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1512 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1513 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1516 $zl = $zjd -1948440 +
10632;
1517 $zn = (int)( ( $zl - 1 ) / 10631 );
1518 $zl = $zl - 10631 * $zn +
354;
1519 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1520 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1521 $zm = (int)( ( 24 * $zl ) / 709 );
1522 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1523 $zy = 30 * $zn +
$zj - 30;
1525 return array( $zy, $zm, $zd );
1529 * Converting Gregorian dates to Hebrew dates.
1531 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1532 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1533 * to translate the relevant functions into PHP and release them under
1536 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1537 * and Adar II is 14. In a non-leap year, Adar is 6.
1543 private static function tsToHebrew( $ts ) {
1545 $year = substr( $ts, 0, 4 );
1546 $month = substr( $ts, 4, 2 );
1547 $day = substr( $ts, 6, 2 );
1549 # Calculate Hebrew year
1550 $hebrewYear = $year +
3760;
1552 # Month number when September = 1, August = 12
1554 if ( $month > 12 ) {
1561 # Calculate day of year from 1 September
1563 for ( $i = 1; $i < $month; $i++
) {
1567 # Check if the year is leap
1568 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1571 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1578 # Calculate the start of the Hebrew year
1579 $start = self
::hebrewYearStart( $hebrewYear );
1581 # Calculate next year's start
1582 if ( $dayOfYear <= $start ) {
1583 # Day is before the start of the year - it is the previous year
1585 $nextStart = $start;
1589 # Add days since previous year's 1 September
1591 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1595 # Start of the new (previous) year
1596 $start = self
::hebrewYearStart( $hebrewYear );
1599 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1602 # Calculate Hebrew day of year
1603 $hebrewDayOfYear = $dayOfYear - $start;
1605 # Difference between year's days
1606 $diff = $nextStart - $start;
1607 # Add 12 (or 13 for leap years) days to ignore the difference between
1608 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1609 # difference is only about the year type
1610 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1616 # Check the year pattern, and is leap year
1617 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1618 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1619 # and non-leap years
1620 $yearPattern = $diff %
30;
1621 # Check if leap year
1622 $isLeap = $diff >= 30;
1624 # Calculate day in the month from number of day in the Hebrew year
1625 # Don't check Adar - if the day is not in Adar, we will stop before;
1626 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1627 $hebrewDay = $hebrewDayOfYear;
1630 while ( $hebrewMonth <= 12 ) {
1631 # Calculate days in this month
1632 if ( $isLeap && $hebrewMonth == 6 ) {
1633 # Adar in a leap year
1635 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1637 if ( $hebrewDay <= $days ) {
1641 # Subtract the days of Adar I
1642 $hebrewDay -= $days;
1645 if ( $hebrewDay <= $days ) {
1651 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1652 # Cheshvan in a complete year (otherwise as the rule below)
1654 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1655 # Kislev in an incomplete year (otherwise as the rule below)
1658 # Odd months have 30 days, even have 29
1659 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1661 if ( $hebrewDay <= $days ) {
1662 # In the current month
1665 # Subtract the days of the current month
1666 $hebrewDay -= $days;
1667 # Try in the next month
1672 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1676 * This calculates the Hebrew year start, as days since 1 September.
1677 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1678 * Used for Hebrew date.
1684 private static function hebrewYearStart( $year ) {
1685 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1686 $b = intval( ( $year - 1 ) %
4 );
1687 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1691 $Mar = intval( $m );
1697 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1698 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1700 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1702 } elseif ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1706 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1711 * Algorithm to convert Gregorian dates to Thai solar dates,
1712 * Minguo dates or Minguo dates.
1714 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1715 * http://en.wikipedia.org/wiki/Minguo_calendar
1716 * http://en.wikipedia.org/wiki/Japanese_era_name
1718 * @param $ts String: 14-character timestamp
1719 * @param $cName String: calender name
1720 * @return Array: converted year, month, day
1722 private static function tsToYear( $ts, $cName ) {
1723 $gy = substr( $ts, 0, 4 );
1724 $gm = substr( $ts, 4, 2 );
1725 $gd = substr( $ts, 6, 2 );
1727 if ( !strcmp( $cName, 'thai' ) ) {
1729 # Add 543 years to the Gregorian calendar
1730 # Months and days are identical
1731 $gy_offset = $gy +
543;
1732 } elseif ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1734 # Deduct 1911 years from the Gregorian calendar
1735 # Months and days are identical
1736 $gy_offset = $gy - 1911;
1737 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1738 # Nengō dates up to Meiji period
1739 # Deduct years from the Gregorian calendar
1740 # depending on the nengo periods
1741 # Months and days are identical
1742 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1744 $gy_gannen = $gy - 1868 +
1;
1745 $gy_offset = $gy_gannen;
1746 if ( $gy_gannen == 1 ) {
1749 $gy_offset = '明治' . $gy_offset;
1751 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1752 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1753 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1754 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1755 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1759 $gy_gannen = $gy - 1912 +
1;
1760 $gy_offset = $gy_gannen;
1761 if ( $gy_gannen == 1 ) {
1764 $gy_offset = '大正' . $gy_offset;
1766 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1767 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1768 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1772 $gy_gannen = $gy - 1926 +
1;
1773 $gy_offset = $gy_gannen;
1774 if ( $gy_gannen == 1 ) {
1777 $gy_offset = '昭和' . $gy_offset;
1780 $gy_gannen = $gy - 1989 +
1;
1781 $gy_offset = $gy_gannen;
1782 if ( $gy_gannen == 1 ) {
1785 $gy_offset = '平成' . $gy_offset;
1791 return array( $gy_offset, $gm, $gd );
1795 * Roman number formatting up to 10000
1801 static function romanNumeral( $num ) {
1802 static $table = array(
1803 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1804 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1805 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1806 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM', 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1809 $num = intval( $num );
1810 if ( $num > 10000 ||
$num <= 0 ) {
1815 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1816 if ( $num >= $pow10 ) {
1817 $s .= $table[$i][(int)floor( $num / $pow10 )];
1819 $num = $num %
$pow10;
1825 * Hebrew Gematria number formatting up to 9999
1831 static function hebrewNumeral( $num ) {
1832 static $table = array(
1833 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1834 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1835 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1836 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1839 $num = intval( $num );
1840 if ( $num > 9999 ||
$num <= 0 ) {
1845 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1846 if ( $num >= $pow10 ) {
1847 if ( $num == 15 ||
$num == 16 ) {
1848 $s .= $table[0][9] . $table[0][$num - 9];
1851 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1852 if ( $pow10 == 1000 ) {
1857 $num = $num %
$pow10;
1859 if ( strlen( $s ) == 2 ) {
1862 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1863 $str .= substr( $s, strlen( $s ) - 2, 2 );
1865 $start = substr( $str, 0, strlen( $str ) - 2 );
1866 $end = substr( $str, strlen( $str ) - 2 );
1869 $str = $start . 'ך';
1872 $str = $start . 'ם';
1875 $str = $start . 'ן';
1878 $str = $start . 'ף';
1881 $str = $start . 'ץ';
1888 * Used by date() and time() to adjust the time output.
1890 * @param $ts Int the time in date('YmdHis') format
1891 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1892 * get user timecorrection setting)
1895 function userAdjust( $ts, $tz = false ) {
1896 global $wgUser, $wgLocalTZoffset;
1898 if ( $tz === false ) {
1899 $tz = $wgUser->getOption( 'timecorrection' );
1902 $data = explode( '|', $tz, 3 );
1904 if ( $data[0] == 'ZoneInfo' ) {
1905 wfSuppressWarnings();
1906 $userTZ = timezone_open( $data[2] );
1907 wfRestoreWarnings();
1908 if ( $userTZ !== false ) {
1909 $date = date_create( $ts, timezone_open( 'UTC' ) );
1910 date_timezone_set( $date, $userTZ );
1911 $date = date_format( $date, 'YmdHis' );
1914 # Unrecognized timezone, default to 'Offset' with the stored offset.
1915 $data[0] = 'Offset';
1919 if ( $data[0] == 'System' ||
$tz == '' ) {
1920 # Global offset in minutes.
1921 if ( isset( $wgLocalTZoffset ) ) {
1922 $minDiff = $wgLocalTZoffset;
1924 } elseif ( $data[0] == 'Offset' ) {
1925 $minDiff = intval( $data[1] );
1927 $data = explode( ':', $tz );
1928 if ( count( $data ) == 2 ) {
1929 $data[0] = intval( $data[0] );
1930 $data[1] = intval( $data[1] );
1931 $minDiff = abs( $data[0] ) * 60 +
$data[1];
1932 if ( $data[0] < 0 ) {
1933 $minDiff = -$minDiff;
1936 $minDiff = intval( $data[0] ) * 60;
1940 # No difference ? Return time unchanged
1941 if ( 0 == $minDiff ) {
1945 wfSuppressWarnings(); // E_STRICT system time bitching
1946 # Generate an adjusted date; take advantage of the fact that mktime
1947 # will normalize out-of-range values so we don't have to split $minDiff
1948 # into hours and minutes.
1950 (int)substr( $ts, 8, 2 ) ), # Hours
1951 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
1952 (int)substr( $ts, 12, 2 ), # Seconds
1953 (int)substr( $ts, 4, 2 ), # Month
1954 (int)substr( $ts, 6, 2 ), # Day
1955 (int)substr( $ts, 0, 4 ) ); # Year
1957 $date = date( 'YmdHis', $t );
1958 wfRestoreWarnings();
1964 * This is meant to be used by time(), date(), and timeanddate() to get
1965 * the date preference they're supposed to use, it should be used in
1969 * function timeanddate([...], $format = true) {
1970 * $datePreference = $this->dateFormat($format);
1975 * @param $usePrefs Mixed: if true, the user's preference is used
1976 * if false, the site/language default is used
1977 * if int/string, assumed to be a format.
1980 function dateFormat( $usePrefs = true ) {
1983 if ( is_bool( $usePrefs ) ) {
1985 $datePreference = $wgUser->getDatePreference();
1987 $datePreference = (string)User
::getDefaultOption( 'date' );
1990 $datePreference = (string)$usePrefs;
1994 if ( $datePreference == '' ) {
1998 return $datePreference;
2002 * Get a format string for a given type and preference
2003 * @param $type string May be date, time or both
2004 * @param $pref string The format name as it appears in Messages*.php
2006 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2010 function getDateFormatString( $type, $pref ) {
2011 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
2012 if ( $pref == 'default' ) {
2013 $pref = $this->getDefaultDateFormat();
2014 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2016 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2018 if ( $type === 'pretty' && $df === null ) {
2019 $df = $this->getDateFormatString( 'date', $pref );
2022 if ( $df === null ) {
2023 $pref = $this->getDefaultDateFormat();
2024 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
2027 $this->dateFormatStrings
[$type][$pref] = $df;
2029 return $this->dateFormatStrings
[$type][$pref];
2033 * @param $ts Mixed: the time format which needs to be turned into a
2034 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2035 * @param $adj Bool: whether to adjust the time output according to the
2036 * user configured offset ($timecorrection)
2037 * @param $format Mixed: true to use user's date format preference
2038 * @param $timecorrection String|bool the time offset as returned by
2039 * validateTimeZone() in Special:Preferences
2042 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2043 $ts = wfTimestamp( TS_MW
, $ts );
2045 $ts = $this->userAdjust( $ts, $timecorrection );
2047 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2048 return $this->sprintfDate( $df, $ts );
2052 * @param $ts Mixed: the time format which needs to be turned into a
2053 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2054 * @param $adj Bool: whether to adjust the time output according to the
2055 * user configured offset ($timecorrection)
2056 * @param $format Mixed: true to use user's date format preference
2057 * @param $timecorrection String|bool the time offset as returned by
2058 * validateTimeZone() in Special:Preferences
2061 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2062 $ts = wfTimestamp( TS_MW
, $ts );
2064 $ts = $this->userAdjust( $ts, $timecorrection );
2066 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2067 return $this->sprintfDate( $df, $ts );
2071 * @param $ts Mixed: the time format which needs to be turned into a
2072 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2073 * @param $adj Bool: whether to adjust the time output according to the
2074 * user configured offset ($timecorrection)
2075 * @param $format Mixed: what format to return, if it's false output the
2076 * default one (default true)
2077 * @param $timecorrection String|bool the time offset as returned by
2078 * validateTimeZone() in Special:Preferences
2081 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2082 $ts = wfTimestamp( TS_MW
, $ts );
2084 $ts = $this->userAdjust( $ts, $timecorrection );
2086 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2087 return $this->sprintfDate( $df, $ts );
2091 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2095 * @param integer $seconds The amount of seconds.
2096 * @param array $chosenIntervals The intervals to enable.
2100 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2101 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2103 $segments = array();
2105 foreach ( $intervals as $intervalName => $intervalValue ) {
2106 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2107 // duration-years, duration-decades, duration-centuries, duration-millennia
2108 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2109 $segments[] = $message->inLanguage( $this )->escaped();
2112 return $this->listToText( $segments );
2116 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2117 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2121 * @param integer $seconds The amount of seconds.
2122 * @param array $chosenIntervals The intervals to enable.
2126 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2127 if ( empty( $chosenIntervals ) ) {
2128 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
2131 $intervals = array_intersect_key( self
::$durationIntervals, array_flip( $chosenIntervals ) );
2132 $sortedNames = array_keys( $intervals );
2133 $smallestInterval = array_pop( $sortedNames );
2135 $segments = array();
2137 foreach ( $intervals as $name => $length ) {
2138 $value = floor( $seconds / $length );
2140 if ( $value > 0 ||
( $name == $smallestInterval && empty( $segments ) ) ) {
2141 $seconds -= $value * $length;
2142 $segments[$name] = $value;
2150 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2152 * @param $type String: can be 'date', 'time' or 'both'
2153 * @param $ts Mixed: the time format which needs to be turned into a
2154 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2155 * @param $user User object used to get preferences for timezone and format
2156 * @param $options Array, can contain the following keys:
2157 * - 'timecorrection': time correction, can have the following values:
2158 * - true: use user's preference
2159 * - false: don't use time correction
2160 * - integer: value of time correction in minutes
2161 * - 'format': format to use, can have the following values:
2162 * - true: use user's preference
2163 * - false: use default preference
2164 * - string: format to use
2168 private function internalUserTimeAndDate( $type, $ts, User
$user, array $options ) {
2169 $ts = wfTimestamp( TS_MW
, $ts );
2170 $options +
= array( 'timecorrection' => true, 'format' => true );
2171 if ( $options['timecorrection'] !== false ) {
2172 if ( $options['timecorrection'] === true ) {
2173 $offset = $user->getOption( 'timecorrection' );
2175 $offset = $options['timecorrection'];
2177 $ts = $this->userAdjust( $ts, $offset );
2179 if ( $options['format'] === true ) {
2180 $format = $user->getDatePreference();
2182 $format = $options['format'];
2184 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2185 return $this->sprintfDate( $df, $ts );
2189 * Get the formatted date for the given timestamp and formatted for
2192 * @param $ts Mixed: the time format which needs to be turned into a
2193 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2194 * @param $user User object used to get preferences for timezone and format
2195 * @param $options Array, can contain the following keys:
2196 * - 'timecorrection': time correction, can have the following values:
2197 * - true: use user's preference
2198 * - false: don't use time correction
2199 * - integer: value of time correction in minutes
2200 * - 'format': format to use, can have the following values:
2201 * - true: use user's preference
2202 * - false: use default preference
2203 * - string: format to use
2207 public function userDate( $ts, User
$user, array $options = array() ) {
2208 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2212 * Get the formatted time for the given timestamp and formatted for
2215 * @param $ts Mixed: the time format which needs to be turned into a
2216 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2217 * @param $user User object used to get preferences for timezone and format
2218 * @param $options Array, can contain the following keys:
2219 * - 'timecorrection': time correction, can have the following values:
2220 * - true: use user's preference
2221 * - false: don't use time correction
2222 * - integer: value of time correction in minutes
2223 * - 'format': format to use, can have the following values:
2224 * - true: use user's preference
2225 * - false: use default preference
2226 * - string: format to use
2230 public function userTime( $ts, User
$user, array $options = array() ) {
2231 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2235 * Get the formatted date and time for the given timestamp and formatted for
2238 * @param $ts Mixed: the time format which needs to be turned into a
2239 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2240 * @param $user User object used to get preferences for timezone and format
2241 * @param $options Array, can contain the following keys:
2242 * - 'timecorrection': time correction, can have the following values:
2243 * - true: use user's preference
2244 * - false: don't use time correction
2245 * - integer: value of time correction in minutes
2246 * - 'format': format to use, can have the following values:
2247 * - true: use user's preference
2248 * - false: use default preference
2249 * - string: format to use
2253 public function userTimeAndDate( $ts, User
$user, array $options = array() ) {
2254 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2258 * Convert an MWTimestamp into a pretty human-readable timestamp using
2259 * the given user preferences and relative base time.
2261 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2262 * on your timestamp object, which will then call this function. Calling
2263 * this function directly will cause hooks to be skipped over.
2265 * @see MWTimestamp::getHumanTimestamp
2266 * @param MWTimestamp $ts Timestamp to prettify
2267 * @param MWTimestamp $relativeTo Base timestamp
2268 * @param User $user User preferences to use
2269 * @return string Human timestamp
2272 public function getHumanTimestamp( MWTimestamp
$ts, MWTimestamp
$relativeTo, User
$user ) {
2273 $diff = $ts->diff( $relativeTo );
2274 $diffDay = (bool)( (int)$ts->timestamp
->format( 'w' ) - (int)$relativeTo->timestamp
->format( 'w' ) );
2275 $days = $diff->days ?
: (int)$diffDay;
2276 if ( $diff->invert ||
$days > 5 && $ts->timestamp
->format( 'Y' ) !== $relativeTo->timestamp
->format( 'Y' ) ) {
2277 // Timestamps are in different years: use full timestamp
2278 // Also do full timestamp for future dates
2280 * @FIXME Add better handling of future timestamps.
2282 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?
: 'default' );
2283 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2284 } elseif ( $days > 5 ) {
2285 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2286 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?
: 'default' );
2287 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) );
2288 } elseif ( $days > 1 ) {
2289 // Timestamp within the past week: show the day of the week and time
2290 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2291 $weekday = self
::$mWeekdayMsgs[$ts->timestamp
->format( 'w' )];
2293 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2294 $ts = wfMessage( "$weekday-at" )
2295 ->inLanguage( $this )
2296 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2298 } elseif ( $days == 1 ) {
2299 // Timestamp was yesterday: say 'yesterday' and the time.
2300 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2301 $ts = wfMessage( 'yesterday-at' )
2302 ->inLanguage( $this )
2303 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2305 } elseif ( $diff->h
> 1 ||
$diff->h
== 1 && $diff->i
> 30 ) {
2306 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2307 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?
: 'default' );
2308 $ts = wfMessage( 'today-at' )
2309 ->inLanguage( $this )
2310 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW
) ) )
2313 // From here on in, the timestamp was soon enough ago so that we can simply say
2314 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2315 } elseif ( $diff->h
== 1 ) {
2316 // Less than 90 minutes, but more than an hour ago.
2317 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2318 } elseif ( $diff->i
>= 1 ) {
2319 // A few minutes ago.
2320 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i
)->text();
2321 } elseif ( $diff->s
>= 30 ) {
2322 // Less than a minute, but more than 30 sec ago.
2323 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s
)->text();
2325 // Less than 30 seconds ago.
2326 $ts = wfMessage( 'just-now' )->text();
2333 * @param $key string
2334 * @return array|null
2336 function getMessage( $key ) {
2337 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
2343 function getAllMessages() {
2344 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
2353 function iconv( $in, $out, $string ) {
2354 # This is a wrapper for iconv in all languages except esperanto,
2355 # which does some nasty x-conversions beforehand
2357 # Even with //IGNORE iconv can whine about illegal characters in
2358 # *input* string. We just ignore those too.
2359 # REF: http://bugs.php.net/bug.php?id=37166
2360 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2361 wfSuppressWarnings();
2362 $text = iconv( $in, $out . '//IGNORE', $string );
2363 wfRestoreWarnings();
2367 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2370 * @param $matches array
2371 * @return mixed|string
2373 function ucwordbreaksCallbackAscii( $matches ) {
2374 return $this->ucfirst( $matches[1] );
2378 * @param $matches array
2381 function ucwordbreaksCallbackMB( $matches ) {
2382 return mb_strtoupper( $matches[0] );
2386 * @param $matches array
2389 function ucCallback( $matches ) {
2390 list( $wikiUpperChars ) = self
::getCaseMaps();
2391 return strtr( $matches[1], $wikiUpperChars );
2395 * @param $matches array
2398 function lcCallback( $matches ) {
2399 list( , $wikiLowerChars ) = self
::getCaseMaps();
2400 return strtr( $matches[1], $wikiLowerChars );
2404 * @param $matches array
2407 function ucwordsCallbackMB( $matches ) {
2408 return mb_strtoupper( $matches[0] );
2412 * @param $matches array
2415 function ucwordsCallbackWiki( $matches ) {
2416 list( $wikiUpperChars ) = self
::getCaseMaps();
2417 return strtr( $matches[0], $wikiUpperChars );
2421 * Make a string's first character uppercase
2423 * @param $str string
2427 function ucfirst( $str ) {
2429 if ( $o < 96 ) { // if already uppercase...
2431 } elseif ( $o < 128 ) {
2432 return ucfirst( $str ); // use PHP's ucfirst()
2434 // fall back to more complex logic in case of multibyte strings
2435 return $this->uc( $str, true );
2440 * Convert a string to uppercase
2442 * @param $str string
2443 * @param $first bool
2447 function uc( $str, $first = false ) {
2448 if ( function_exists( 'mb_strtoupper' ) ) {
2450 if ( $this->isMultibyte( $str ) ) {
2451 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2453 return ucfirst( $str );
2456 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
2459 if ( $this->isMultibyte( $str ) ) {
2460 $x = $first ?
'^' : '';
2461 return preg_replace_callback(
2462 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2463 array( $this, 'ucCallback' ),
2467 return $first ?
ucfirst( $str ) : strtoupper( $str );
2473 * @param $str string
2474 * @return mixed|string
2476 function lcfirst( $str ) {
2479 return strval( $str );
2480 } elseif ( $o >= 128 ) {
2481 return $this->lc( $str, true );
2482 } elseif ( $o > 96 ) {
2485 $str[0] = strtolower( $str[0] );
2491 * @param $str string
2492 * @param $first bool
2493 * @return mixed|string
2495 function lc( $str, $first = false ) {
2496 if ( function_exists( 'mb_strtolower' ) ) {
2498 if ( $this->isMultibyte( $str ) ) {
2499 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2501 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2504 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
2507 if ( $this->isMultibyte( $str ) ) {
2508 $x = $first ?
'^' : '';
2509 return preg_replace_callback(
2510 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2511 array( $this, 'lcCallback' ),
2515 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2521 * @param $str string
2524 function isMultibyte( $str ) {
2525 return (bool)preg_match( '/[\x80-\xff]/', $str );
2529 * @param $str string
2530 * @return mixed|string
2532 function ucwords( $str ) {
2533 if ( $this->isMultibyte( $str ) ) {
2534 $str = $this->lc( $str );
2536 // regexp to find first letter in each word (i.e. after each space)
2537 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2539 // function to use to capitalize a single char
2540 if ( function_exists( 'mb_strtoupper' ) ) {
2541 return preg_replace_callback(
2543 array( $this, 'ucwordsCallbackMB' ),
2547 return preg_replace_callback(
2549 array( $this, 'ucwordsCallbackWiki' ),
2554 return ucwords( strtolower( $str ) );
2559 * capitalize words at word breaks
2561 * @param $str string
2564 function ucwordbreaks( $str ) {
2565 if ( $this->isMultibyte( $str ) ) {
2566 $str = $this->lc( $str );
2568 // since \b doesn't work for UTF-8, we explicitely define word break chars
2569 $breaks = "[ \-\(\)\}\{\.,\?!]";
2571 // find first letter after word break
2572 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2574 if ( function_exists( 'mb_strtoupper' ) ) {
2575 return preg_replace_callback(
2577 array( $this, 'ucwordbreaksCallbackMB' ),
2581 return preg_replace_callback(
2583 array( $this, 'ucwordsCallbackWiki' ),
2588 return preg_replace_callback(
2589 '/\b([\w\x80-\xff]+)\b/',
2590 array( $this, 'ucwordbreaksCallbackAscii' ),
2597 * Return a case-folded representation of $s
2599 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2600 * and $s2 are the same except for the case of their characters. It is not
2601 * necessary for the value returned to make sense when displayed.
2603 * Do *not* perform any other normalisation in this function. If a caller
2604 * uses this function when it should be using a more general normalisation
2605 * function, then fix the caller.
2611 function caseFold( $s ) {
2612 return $this->uc( $s );
2619 function checkTitleEncoding( $s ) {
2620 if ( is_array( $s ) ) {
2621 throw new MWException( 'Given array to checkTitleEncoding.' );
2623 if ( StringUtils
::isUtf8( $s ) ) {
2627 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2633 function fallback8bitEncoding() {
2634 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
2638 * Most writing systems use whitespace to break up words.
2639 * Some languages such as Chinese don't conventionally do this,
2640 * which requires special handling when breaking up words for
2645 function hasWordBreaks() {
2650 * Some languages such as Chinese require word segmentation,
2651 * Specify such segmentation when overridden in derived class.
2653 * @param $string String
2656 function segmentByWord( $string ) {
2661 * Some languages have special punctuation need to be normalized.
2662 * Make such changes here.
2664 * @param $string String
2667 function normalizeForSearch( $string ) {
2668 return self
::convertDoubleWidth( $string );
2672 * convert double-width roman characters to single-width.
2673 * range: ff00-ff5f ~= 0020-007f
2675 * @param $string string
2679 protected static function convertDoubleWidth( $string ) {
2680 static $full = null;
2681 static $half = null;
2683 if ( $full === null ) {
2684 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2685 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2686 $full = str_split( $fullWidth, 3 );
2687 $half = str_split( $halfWidth );
2690 $string = str_replace( $full, $half, $string );
2695 * @param $string string
2696 * @param $pattern string
2699 protected static function insertSpace( $string, $pattern ) {
2700 $string = preg_replace( $pattern, " $1 ", $string );
2701 $string = preg_replace( '/ +/', ' ', $string );
2706 * @param $termsArray array
2709 function convertForSearchResult( $termsArray ) {
2710 # some languages, e.g. Chinese, need to do a conversion
2711 # in order for search results to be displayed correctly
2716 * Get the first character of a string.
2721 function firstChar( $s ) {
2724 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2725 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2730 if ( isset( $matches[1] ) ) {
2731 if ( strlen( $matches[1] ) != 3 ) {
2735 // Break down Hangul syllables to grab the first jamo
2736 $code = utf8ToCodepoint( $matches[1] );
2737 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
2739 } elseif ( $code < 0xb098 ) {
2740 return "\xe3\x84\xb1";
2741 } elseif ( $code < 0xb2e4 ) {
2742 return "\xe3\x84\xb4";
2743 } elseif ( $code < 0xb77c ) {
2744 return "\xe3\x84\xb7";
2745 } elseif ( $code < 0xb9c8 ) {
2746 return "\xe3\x84\xb9";
2747 } elseif ( $code < 0xbc14 ) {
2748 return "\xe3\x85\x81";
2749 } elseif ( $code < 0xc0ac ) {
2750 return "\xe3\x85\x82";
2751 } elseif ( $code < 0xc544 ) {
2752 return "\xe3\x85\x85";
2753 } elseif ( $code < 0xc790 ) {
2754 return "\xe3\x85\x87";
2755 } elseif ( $code < 0xcc28 ) {
2756 return "\xe3\x85\x88";
2757 } elseif ( $code < 0xce74 ) {
2758 return "\xe3\x85\x8a";
2759 } elseif ( $code < 0xd0c0 ) {
2760 return "\xe3\x85\x8b";
2761 } elseif ( $code < 0xd30c ) {
2762 return "\xe3\x85\x8c";
2763 } elseif ( $code < 0xd558 ) {
2764 return "\xe3\x85\x8d";
2766 return "\xe3\x85\x8e";
2773 function initEncoding() {
2774 # Some languages may have an alternate char encoding option
2775 # (Esperanto X-coding, Japanese furigana conversion, etc)
2776 # If this language is used as the primary content language,
2777 # an override to the defaults can be set here on startup.
2784 function recodeForEdit( $s ) {
2785 # For some languages we'll want to explicitly specify
2786 # which characters make it into the edit box raw
2787 # or are converted in some way or another.
2788 global $wgEditEncoding;
2789 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2792 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2800 function recodeInput( $s ) {
2801 # Take the previous into account.
2802 global $wgEditEncoding;
2803 if ( $wgEditEncoding != '' ) {
2804 $enc = $wgEditEncoding;
2808 if ( $enc == 'UTF-8' ) {
2811 return $this->iconv( $enc, 'UTF-8', $s );
2816 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2817 * also cleans up certain backwards-compatible sequences, converting them
2818 * to the modern Unicode equivalent.
2820 * This is language-specific for performance reasons only.
2826 function normalize( $s ) {
2827 global $wgAllUnicodeFixes;
2828 $s = UtfNormal
::cleanUp( $s );
2829 if ( $wgAllUnicodeFixes ) {
2830 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2831 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2838 * Transform a string using serialized data stored in the given file (which
2839 * must be in the serialized subdirectory of $IP). The file contains pairs
2840 * mapping source characters to destination characters.
2842 * The data is cached in process memory. This will go faster if you have the
2843 * FastStringSearch extension.
2845 * @param $file string
2846 * @param $string string
2848 * @throws MWException
2851 function transformUsingPairFile( $file, $string ) {
2852 if ( !isset( $this->transformData
[$file] ) ) {
2853 $data = wfGetPrecompiledData( $file );
2854 if ( $data === false ) {
2855 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
2857 $this->transformData
[$file] = new ReplacementArray( $data );
2859 return $this->transformData
[$file]->replace( $string );
2863 * For right-to-left language support
2868 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
2872 * Return the correct HTML 'dir' attribute value for this language.
2876 return $this->isRTL() ?
'rtl' : 'ltr';
2880 * Return 'left' or 'right' as appropriate alignment for line-start
2881 * for this language's text direction.
2883 * Should be equivalent to CSS3 'start' text-align value....
2887 function alignStart() {
2888 return $this->isRTL() ?
'right' : 'left';
2892 * Return 'right' or 'left' as appropriate alignment for line-end
2893 * for this language's text direction.
2895 * Should be equivalent to CSS3 'end' text-align value....
2899 function alignEnd() {
2900 return $this->isRTL() ?
'left' : 'right';
2904 * A hidden direction mark (LRM or RLM), depending on the language direction.
2905 * Unlike getDirMark(), this function returns the character as an HTML entity.
2906 * This function should be used when the output is guaranteed to be HTML,
2907 * because it makes the output HTML source code more readable. When
2908 * the output is plain text or can be escaped, getDirMark() should be used.
2910 * @param $opposite Boolean Get the direction mark opposite to your language
2914 function getDirMarkEntity( $opposite = false ) {
2916 return $this->isRTL() ?
'‎' : '‏';
2918 return $this->isRTL() ?
'‏' : '‎';
2922 * A hidden direction mark (LRM or RLM), depending on the language direction.
2923 * This function produces them as invisible Unicode characters and
2924 * the output may be hard to read and debug, so it should only be used
2925 * when the output is plain text or can be escaped. When the output is
2926 * HTML, use getDirMarkEntity() instead.
2928 * @param $opposite Boolean Get the direction mark opposite to your language
2931 function getDirMark( $opposite = false ) {
2932 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2933 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2935 return $this->isRTL() ?
$lrm : $rlm;
2937 return $this->isRTL() ?
$rlm : $lrm;
2943 function capitalizeAllNouns() {
2944 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2948 * An arrow, depending on the language direction.
2950 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2953 function getArrow( $direction = 'forwards' ) {
2954 switch ( $direction ) {
2956 return $this->isRTL() ?
'←' : '→';
2958 return $this->isRTL() ?
'→' : '←';
2971 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2975 function linkPrefixExtension() {
2976 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2980 * Get all magic words from cache.
2983 function getMagicWords() {
2984 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2988 * Run the LanguageGetMagic hook once.
2990 protected function doMagicHook() {
2991 if ( $this->mMagicHookDone
) {
2994 $this->mMagicHookDone
= true;
2995 wfProfileIn( 'LanguageGetMagic' );
2996 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2997 wfProfileOut( 'LanguageGetMagic' );
3001 * Fill a MagicWord object with data from here
3005 function getMagic( $mw ) {
3006 // Saves a function call
3007 if ( ! $this->mMagicHookDone
) {
3008 $this->doMagicHook();
3011 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
3012 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
3014 $rawEntry = self
::$dataCache->getSubitem(
3015 $this->mCode
, 'magicWords', $mw->mId
);
3018 if ( !is_array( $rawEntry ) ) {
3019 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3021 $mw->mCaseSensitive
= $rawEntry[0];
3022 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
3027 * Add magic words to the extension array
3029 * @param $newWords array
3031 function addMagicWordsByLang( $newWords ) {
3032 $fallbackChain = $this->getFallbackLanguages();
3033 $fallbackChain = array_reverse( $fallbackChain );
3034 foreach ( $fallbackChain as $code ) {
3035 if ( isset( $newWords[$code] ) ) {
3036 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
3042 * Get special page names, as an associative array
3043 * case folded alias => real name
3045 function getSpecialPageAliases() {
3046 // Cache aliases because it may be slow to load them
3047 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
3049 $this->mExtendedSpecialPageAliases
=
3050 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
3051 wfRunHooks( 'LanguageGetSpecialPageAliases',
3052 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
3055 return $this->mExtendedSpecialPageAliases
;
3059 * Italic is unsuitable for some languages
3061 * @param $text String: the text to be emphasized.
3064 function emphasize( $text ) {
3065 return "<em>$text</em>";
3069 * Normally we output all numbers in plain en_US style, that is
3070 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3071 * point twohundredthirtyfive. However this is not suitable for all
3072 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3073 * Icelandic just want to use commas instead of dots, and dots instead
3074 * of commas like "293.291,235".
3076 * An example of this function being called:
3078 * wfMessage( 'message' )->numParams( $num )->text()
3081 * See LanguageGu.php for the Gujarati implementation and
3082 * $separatorTransformTable on MessageIs.php for
3083 * the , => . and . => , implementation.
3085 * @todo check if it's viable to use localeconv() for the decimal
3087 * @param $number Mixed: the string to be formatted, should be an integer
3088 * or a floating point number.
3089 * @param $nocommafy Bool: set to true for special numbers like dates
3092 public function formatNum( $number, $nocommafy = false ) {
3093 global $wgTranslateNumerals;
3094 if ( !$nocommafy ) {
3095 $number = $this->commafy( $number );
3096 $s = $this->separatorTransformTable();
3098 $number = strtr( $number, $s );
3102 if ( $wgTranslateNumerals ) {
3103 $s = $this->digitTransformTable();
3105 $number = strtr( $number, $s );
3113 * Front-end for non-commafied formatNum
3115 * @param mixed $number the string to be formatted, should be an integer
3116 * or a floating point number.
3120 public function formatNumNoSeparators( $number ) {
3121 return $this->formatNum( $number, true );
3125 * @param $number string
3128 function parseFormattedNumber( $number ) {
3129 $s = $this->digitTransformTable();
3131 $number = strtr( $number, array_flip( $s ) );
3134 $s = $this->separatorTransformTable();
3136 $number = strtr( $number, array_flip( $s ) );
3139 $number = strtr( $number, array( ',' => '' ) );
3144 * Adds commas to a given number
3146 * @param $number mixed
3149 function commafy( $number ) {
3150 $digitGroupingPattern = $this->digitGroupingPattern();
3151 if ( $number === null ) {
3155 if ( !$digitGroupingPattern ||
$digitGroupingPattern === "###,###,###" ) {
3156 // default grouping is at thousands, use the same for ###,###,### pattern too.
3157 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3159 // Ref: http://cldr.unicode.org/translation/number-patterns
3161 if ( intval( $number ) < 0 ) {
3162 // For negative numbers apply the algorithm like positive number and add sign.
3164 $number = substr( $number, 1 );
3166 $integerPart = array();
3167 $decimalPart = array();
3168 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3169 preg_match( "/\d+/", $number, $integerPart );
3170 preg_match( "/\.\d*/", $number, $decimalPart );
3171 $groupedNumber = ( count( $decimalPart ) > 0 ) ?
$decimalPart[0] : "";
3172 if ( $groupedNumber === $number ) {
3173 // the string does not have any number part. Eg: .12345
3174 return $sign . $groupedNumber;
3176 $start = $end = strlen( $integerPart[0] );
3177 while ( $start > 0 ) {
3178 $match = $matches[0][$numMatches - 1];
3179 $matchLen = strlen( $match );
3180 $start = $end - $matchLen;
3184 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3186 if ( $numMatches > 1 ) {
3187 // use the last pattern for the rest of the number
3191 $groupedNumber = "," . $groupedNumber;
3194 return $sign . $groupedNumber;
3201 function digitGroupingPattern() {
3202 return self
::$dataCache->getItem( $this->mCode
, 'digitGroupingPattern' );
3208 function digitTransformTable() {
3209 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
3215 function separatorTransformTable() {
3216 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
3220 * Take a list of strings and build a locale-friendly comma-separated
3221 * list, using the local comma-separator message.
3222 * The last two strings are chained with an "and".
3223 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3228 function listToText( array $l ) {
3229 $m = count( $l ) - 1;
3234 $and = $this->getMessageFromDB( 'and' );
3235 $space = $this->getMessageFromDB( 'word-separator' );
3237 $comma = $this->getMessageFromDB( 'comma-separator' );
3241 for ( $i = $m - 1; $i >= 0; $i-- ) {
3242 if ( $i == $m - 1 ) {
3243 $s = $l[$i] . $and . $space . $s;
3245 $s = $l[$i] . $comma . $s;
3252 * Take a list of strings and build a locale-friendly comma-separated
3253 * list, using the local comma-separator message.
3254 * @param $list array of strings to put in a comma list
3257 function commaList( array $list ) {
3259 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3265 * Take a list of strings and build a locale-friendly semicolon-separated
3266 * list, using the local semicolon-separator message.
3267 * @param $list array of strings to put in a semicolon list
3270 function semicolonList( array $list ) {
3272 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3278 * Same as commaList, but separate it with the pipe instead.
3279 * @param $list array of strings to put in a pipe list
3282 function pipeList( array $list ) {
3284 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3290 * Truncate a string to a specified length in bytes, appending an optional
3291 * string (e.g. for ellipses)
3293 * The database offers limited byte lengths for some columns in the database;
3294 * multi-byte character sets mean we need to ensure that only whole characters
3295 * are included, otherwise broken characters can be passed to the user
3297 * If $length is negative, the string will be truncated from the beginning
3299 * @param $string String to truncate
3300 * @param $length Int: maximum length (including ellipses)
3301 * @param $ellipsis String to append to the truncated text
3302 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3303 * $adjustLength was introduced in 1.18, before that behaved as if false.
3306 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3307 # Use the localized ellipsis character
3308 if ( $ellipsis == '...' ) {
3309 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3311 # Check if there is no need to truncate
3312 if ( $length == 0 ) {
3313 return $ellipsis; // convention
3314 } elseif ( strlen( $string ) <= abs( $length ) ) {
3315 return $string; // no need to truncate
3317 $stringOriginal = $string;
3318 # If ellipsis length is >= $length then we can't apply $adjustLength
3319 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3320 $string = $ellipsis; // this can be slightly unexpected
3321 # Otherwise, truncate and add ellipsis...
3323 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
3324 if ( $length > 0 ) {
3325 $length -= $eLength;
3326 $string = substr( $string, 0, $length ); // xyz...
3327 $string = $this->removeBadCharLast( $string );
3328 $string = rtrim( $string );
3329 $string = $string . $ellipsis;
3331 $length +
= $eLength;
3332 $string = substr( $string, $length ); // ...xyz
3333 $string = $this->removeBadCharFirst( $string );
3334 $string = ltrim( $string );
3335 $string = $ellipsis . $string;
3338 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3339 # This check is *not* redundant if $adjustLength, due to the single case where
3340 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3341 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3344 return $stringOriginal;
3349 * Remove bytes that represent an incomplete Unicode character
3350 * at the end of string (e.g. bytes of the char are missing)
3352 * @param $string String
3355 protected function removeBadCharLast( $string ) {
3356 if ( $string != '' ) {
3357 $char = ord( $string[strlen( $string ) - 1] );
3359 if ( $char >= 0xc0 ) {
3360 # We got the first byte only of a multibyte char; remove it.
3361 $string = substr( $string, 0, -1 );
3362 } elseif ( $char >= 0x80 &&
3363 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3364 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3366 # We chopped in the middle of a character; remove it
3374 * Remove bytes that represent an incomplete Unicode character
3375 * at the start of string (e.g. bytes of the char are missing)
3377 * @param $string String
3380 protected function removeBadCharFirst( $string ) {
3381 if ( $string != '' ) {
3382 $char = ord( $string[0] );
3383 if ( $char >= 0x80 && $char < 0xc0 ) {
3384 # We chopped in the middle of a character; remove the whole thing
3385 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3392 * Truncate a string of valid HTML to a specified length in bytes,
3393 * appending an optional string (e.g. for ellipses), and return valid HTML
3395 * This is only intended for styled/linked text, such as HTML with
3396 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3397 * Also, this will not detect things like "display:none" CSS.
3399 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3401 * @param string $text HTML string to truncate
3402 * @param int $length (zero/positive) Maximum length (including ellipses)
3403 * @param string $ellipsis String to append to the truncated text
3406 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3407 # Use the localized ellipsis character
3408 if ( $ellipsis == '...' ) {
3409 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3411 # Check if there is clearly no need to truncate
3412 if ( $length <= 0 ) {
3413 return $ellipsis; // no text shown, nothing to format (convention)
3414 } elseif ( strlen( $text ) <= $length ) {
3415 return $text; // string short enough even *with* HTML (short-circuit)
3418 $dispLen = 0; // innerHTML legth so far
3419 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3420 $tagType = 0; // 0-open, 1-close
3421 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3422 $entityState = 0; // 0-not entity, 1-entity
3423 $tag = $ret = ''; // accumulated tag name, accumulated result string
3424 $openTags = array(); // open tag stack
3425 $maybeState = null; // possible truncation state
3427 $textLen = strlen( $text );
3428 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3429 for ( $pos = 0; true; ++
$pos ) {
3430 # Consider truncation once the display length has reached the maximim.
3431 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3432 # Check that we're not in the middle of a bracket/entity...
3433 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3434 if ( !$testingEllipsis ) {
3435 $testingEllipsis = true;
3436 # Save where we are; we will truncate here unless there turn out to
3437 # be so few remaining characters that truncation is not necessary.
3438 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3439 $maybeState = array( $ret, $openTags ); // save state
3441 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3442 # String in fact does need truncation, the truncation point was OK.
3443 list( $ret, $openTags ) = $maybeState; // reload state
3444 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3445 $ret .= $ellipsis; // add ellipsis
3449 if ( $pos >= $textLen ) {
3450 break; // extra iteration just for above checks
3453 # Read the next char...
3455 $lastCh = $pos ?
$text[$pos - 1] : '';
3456 $ret .= $ch; // add to result string
3458 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3459 $entityState = 0; // for bad HTML
3460 $bracketState = 1; // tag started (checking for backslash)
3461 } elseif ( $ch == '>' ) {
3462 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3463 $entityState = 0; // for bad HTML
3464 $bracketState = 0; // out of brackets
3465 } elseif ( $bracketState == 1 ) {
3467 $tagType = 1; // close tag (e.g. "</span>")
3469 $tagType = 0; // open tag (e.g. "<span>")
3472 $bracketState = 2; // building tag name
3473 } elseif ( $bracketState == 2 ) {
3477 // Name found (e.g. "<a href=..."), add on tag attributes...
3478 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
3480 } elseif ( $bracketState == 0 ) {
3481 if ( $entityState ) {
3484 $dispLen++
; // entity is one displayed char
3487 if ( $neLength == 0 && !$maybeState ) {
3488 // Save state without $ch. We want to *hit* the first
3489 // display char (to get tags) but not *use* it if truncating.
3490 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3493 $entityState = 1; // entity found, (e.g. " ")
3495 $dispLen++
; // this char is displayed
3496 // Add the next $max display text chars after this in one swoop...
3497 $max = ( $testingEllipsis ?
$length : $neLength ) - $dispLen;
3498 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
3499 $dispLen +
= $skipped;
3505 // Close the last tag if left unclosed by bad HTML
3506 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3507 while ( count( $openTags ) > 0 ) {
3508 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3514 * truncateHtml() helper function
3515 * like strcspn() but adds the skipped chars to $ret
3524 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3525 if ( $len === null ) {
3526 $len = -1; // -1 means "no limit" for strcspn
3527 } elseif ( $len < 0 ) {
3531 if ( $start < strlen( $text ) ) {
3532 $skipCount = strcspn( $text, $search, $start, $len );
3533 $ret .= substr( $text, $start, $skipCount );
3539 * truncateHtml() helper function
3540 * (a) push or pop $tag from $openTags as needed
3541 * (b) clear $tag value
3542 * @param &$tag string Current HTML tag name we are looking at
3543 * @param $tagType int (0-open tag, 1-close tag)
3544 * @param $lastCh string Character before the '>' that ended this tag
3545 * @param &$openTags array Open tag stack (not accounting for $tag)
3547 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3548 $tag = ltrim( $tag );
3550 if ( $tagType == 0 && $lastCh != '/' ) {
3551 $openTags[] = $tag; // tag opened (didn't close itself)
3552 } elseif ( $tagType == 1 ) {
3553 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3554 array_pop( $openTags ); // tag closed
3562 * Grammatical transformations, needed for inflected languages
3563 * Invoked by putting {{grammar:case|word}} in a message
3565 * @param $word string
3566 * @param $case string
3569 function convertGrammar( $word, $case ) {
3570 global $wgGrammarForms;
3571 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3572 return $wgGrammarForms[$this->getCode()][$case][$word];
3577 * Get the grammar forms for the content language
3578 * @return array of grammar forms
3581 function getGrammarForms() {
3582 global $wgGrammarForms;
3583 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3584 return $wgGrammarForms[$this->getCode()];
3589 * Provides an alternative text depending on specified gender.
3590 * Usage {{gender:username|masculine|feminine|neutral}}.
3591 * username is optional, in which case the gender of current user is used,
3592 * but only in (some) interface messages; otherwise default gender is used.
3594 * If no forms are given, an empty string is returned. If only one form is
3595 * given, it will be returned unconditionally. These details are implied by
3596 * the caller and cannot be overridden in subclasses.
3598 * If more than one form is given, the default is to use the neutral one
3599 * if it is specified, and to use the masculine one otherwise. These
3600 * details can be overridden in subclasses.
3602 * @param $gender string
3603 * @param $forms array
3607 function gender( $gender, $forms ) {
3608 if ( !count( $forms ) ) {
3611 $forms = $this->preConvertPlural( $forms, 2 );
3612 if ( $gender === 'male' ) {
3615 if ( $gender === 'female' ) {
3618 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
3622 * Plural form transformations, needed for some languages.
3623 * For example, there are 3 form of plural in Russian and Polish,
3624 * depending on "count mod 10". See [[w:Plural]]
3625 * For English it is pretty simple.
3627 * Invoked by putting {{plural:count|wordform1|wordform2}}
3628 * or {{plural:count|wordform1|wordform2|wordform3}}
3630 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3632 * @param $count Integer: non-localized number
3633 * @param $forms Array: different plural forms
3634 * @return string Correct form of plural for $count in this language
3636 function convertPlural( $count, $forms ) {
3637 // Handle explicit n=pluralform cases
3638 foreach ( $forms as $index => $form ) {
3639 if ( preg_match( '/^\d+=/i', $form ) ) {
3640 $pos = strpos( $form, '=' );
3641 if ( substr( $form, 0, $pos ) === (string)$count ) {
3642 return substr( $form, $pos +
1 );
3644 unset( $forms[$index] );
3648 $forms = array_values( $forms );
3649 if ( !count( $forms ) ) {
3653 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3654 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3655 return $forms[$pluralForm];
3659 * Checks that convertPlural was given an array and pads it to requested
3660 * amount of forms by copying the last one.
3662 * @param $count Integer: How many forms should there be at least
3663 * @param $forms Array of forms given to convertPlural
3664 * @return array Padded array of forms or an exception if not an array
3666 protected function preConvertPlural( /* Array */ $forms, $count ) {
3667 while ( count( $forms ) < $count ) {
3668 $forms[] = $forms[count( $forms ) - 1];
3674 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3675 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3676 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3677 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3680 * @param $str String: the validated block duration in English
3681 * @return string Somehow translated block duration
3682 * @see LanguageFi.php for example implementation
3684 function translateBlockExpiry( $str ) {
3685 $duration = SpecialBlock
::getSuggestedDurations( $this );
3686 foreach ( $duration as $show => $value ) {
3687 if ( strcmp( $str, $value ) == 0 ) {
3688 return htmlspecialchars( trim( $show ) );
3692 // Since usually only infinite or indefinite is only on list, so try
3693 // equivalents if still here.
3694 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3695 if ( in_array( $str, $indefs ) ) {
3696 foreach ( $indefs as $val ) {
3697 $show = array_search( $val, $duration, true );
3698 if ( $show !== false ) {
3699 return htmlspecialchars( trim( $show ) );
3704 // If all else fails, return a standard duration or timestamp description.
3705 $time = strtotime( $str, 0 );
3706 if ( $time === false ) { // Unknown format. Return it as-is in case.
3708 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3709 // $time is relative to 0 so it's a duration length.
3710 return $this->formatDuration( $time );
3711 } else { // It's an absolute timestamp.
3712 if ( $time === 0 ) {
3713 // wfTimestamp() handles 0 as current time instead of epoch.
3714 return $this->timeanddate( '19700101000000' );
3716 return $this->timeanddate( $time );
3722 * languages like Chinese need to be segmented in order for the diff
3725 * @param $text String
3728 public function segmentForDiff( $text ) {
3733 * and unsegment to show the result
3735 * @param $text String
3738 public function unsegmentForDiff( $text ) {
3743 * Return the LanguageConverter used in the Language
3746 * @return LanguageConverter
3748 public function getConverter() {
3749 return $this->mConverter
;
3753 * convert text to all supported variants
3755 * @param $text string
3758 public function autoConvertToAllVariants( $text ) {
3759 return $this->mConverter
->autoConvertToAllVariants( $text );
3763 * convert text to different variants of a language.
3765 * @param $text string
3768 public function convert( $text ) {
3769 return $this->mConverter
->convert( $text );
3773 * Convert a Title object to a string in the preferred variant
3775 * @param $title Title
3778 public function convertTitle( $title ) {
3779 return $this->mConverter
->convertTitle( $title );
3783 * Convert a namespace index to a string in the preferred variant
3788 public function convertNamespace( $ns ) {
3789 return $this->mConverter
->convertNamespace( $ns );
3793 * Check if this is a language with variants
3797 public function hasVariants() {
3798 return count( $this->getVariants() ) > 1;
3802 * Check if the language has the specific variant
3805 * @param $variant string
3808 public function hasVariant( $variant ) {
3809 return (bool)$this->mConverter
->validateVariant( $variant );
3813 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3815 * @param $text string
3817 * @deprecated since 1.22 is no longer used
3819 public function armourMath( $text ) {
3820 return $this->mConverter
->armourMath( $text );
3824 * Perform output conversion on a string, and encode for safe HTML output.
3825 * @param $text String text to be converted
3826 * @param $isTitle Bool whether this conversion is for the article title
3828 * @todo this should get integrated somewhere sane
3830 public function convertHtml( $text, $isTitle = false ) {
3831 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3835 * @param $key string
3838 public function convertCategoryKey( $key ) {
3839 return $this->mConverter
->convertCategoryKey( $key );
3843 * Get the list of variants supported by this language
3844 * see sample implementation in LanguageZh.php
3846 * @return array an array of language codes
3848 public function getVariants() {
3849 return $this->mConverter
->getVariants();
3855 public function getPreferredVariant() {
3856 return $this->mConverter
->getPreferredVariant();
3862 public function getDefaultVariant() {
3863 return $this->mConverter
->getDefaultVariant();
3869 public function getURLVariant() {
3870 return $this->mConverter
->getURLVariant();
3874 * If a language supports multiple variants, it is
3875 * possible that non-existing link in one variant
3876 * actually exists in another variant. this function
3877 * tries to find it. See e.g. LanguageZh.php
3879 * @param $link String: the name of the link
3880 * @param $nt Mixed: the title object of the link
3881 * @param $ignoreOtherCond Boolean: to disable other conditions when
3882 * we need to transclude a template or update a category's link
3883 * @return null the input parameters may be modified upon return
3885 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3886 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
3890 * If a language supports multiple variants, converts text
3891 * into an array of all possible variants of the text:
3892 * 'variant' => text in that variant
3894 * @deprecated since 1.17 Use autoConvertToAllVariants()
3896 * @param $text string
3900 public function convertLinkToAllVariants( $text ) {
3901 return $this->mConverter
->convertLinkToAllVariants( $text );
3905 * returns language specific options used by User::getPageRenderHash()
3906 * for example, the preferred language variant
3910 function getExtraHashOptions() {
3911 return $this->mConverter
->getExtraHashOptions();
3915 * For languages that support multiple variants, the title of an
3916 * article may be displayed differently in different variants. this
3917 * function returns the apporiate title defined in the body of the article.
3921 public function getParsedTitle() {
3922 return $this->mConverter
->getParsedTitle();
3926 * Prepare external link text for conversion. When the text is
3927 * a URL, it shouldn't be converted, and it'll be wrapped in
3928 * the "raw" tag (-{R| }-) to prevent conversion.
3930 * This function is called "markNoConversion" for historical
3933 * @param $text String: text to be used for external link
3934 * @param $noParse bool: wrap it without confirming it's a real URL first
3935 * @return string the tagged text
3937 public function markNoConversion( $text, $noParse = false ) {
3938 // Excluding protocal-relative URLs may avoid many false positives.
3939 if ( $noParse ||
preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
3940 return $this->mConverter
->markNoConversion( $text );
3947 * A regular expression to match legal word-trailing characters
3948 * which should be merged onto a link of the form [[foo]]bar.
3952 public function linkTrail() {
3953 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
3957 * A regular expression character set to match legal word-prefixing
3958 * characters which should be merged onto a link of the form foo[[bar]].
3962 public function linkPrefixCharset() {
3963 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixCharset' );
3969 function getLangObj() {
3974 * Get the "parent" language which has a converter to convert a "compatible" language
3975 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
3977 * @return Language|null
3980 public function getParentLanguage() {
3981 if ( $this->mParentLanguage
!== false ) {
3982 return $this->mParentLanguage
;
3985 $pieces = explode( '-', $this->getCode() );
3987 if ( !in_array( $code, LanguageConverter
::$languagesWithVariants ) ) {
3988 $this->mParentLanguage
= null;
3991 $lang = Language
::factory( $code );
3992 if ( !$lang->hasVariant( $this->getCode() ) ) {
3993 $this->mParentLanguage
= null;
3997 $this->mParentLanguage
= $lang;
4002 * Get the RFC 3066 code for this language object
4004 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4005 * htmlspecialchars() or similar
4009 public function getCode() {
4010 return $this->mCode
;
4014 * Get the code in Bcp47 format which we can use
4015 * inside of html lang="" tags.
4017 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4018 * htmlspecialchars() or similar.
4023 public function getHtmlCode() {
4024 if ( is_null( $this->mHtmlCode
) ) {
4025 $this->mHtmlCode
= wfBCP47( $this->getCode() );
4027 return $this->mHtmlCode
;
4031 * @param $code string
4033 public function setCode( $code ) {
4034 $this->mCode
= $code;
4035 // Ensure we don't leave incorrect cached data lying around
4036 $this->mHtmlCode
= null;
4037 $this->mParentLanguage
= false;
4041 * Get the name of a file for a certain language code
4042 * @param $prefix string Prepend this to the filename
4043 * @param $code string Language code
4044 * @param $suffix string Append this to the filename
4045 * @throws MWException
4046 * @return string $prefix . $mangledCode . $suffix
4048 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4049 // Protect against path traversal
4050 if ( !Language
::isValidCode( $code )
4051 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
4053 throw new MWException( "Invalid language code \"$code\"" );
4056 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4060 * Get the language code from a file name. Inverse of getFileName()
4061 * @param $filename string $prefix . $languageCode . $suffix
4062 * @param $prefix string Prefix before the language code
4063 * @param $suffix string Suffix after the language code
4064 * @return string Language code, or false if $prefix or $suffix isn't found
4066 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4068 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4069 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4070 if ( !count( $m ) ) {
4073 return str_replace( '_', '-', strtolower( $m[1] ) );
4077 * @param $code string
4080 public static function getMessagesFileName( $code ) {
4082 $file = self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4083 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
4088 * @param $code string
4091 public static function getClassFileName( $code ) {
4093 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4097 * Get the first fallback for a given language.
4099 * @param $code string
4101 * @return bool|string
4103 public static function getFallbackFor( $code ) {
4104 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4107 $fallbacks = self
::getFallbacksFor( $code );
4108 $first = array_shift( $fallbacks );
4114 * Get the ordered list of fallback languages.
4117 * @param $code string Language code
4120 public static function getFallbacksFor( $code ) {
4121 if ( $code === 'en' ||
!Language
::isValidBuiltInCode( $code ) ) {
4124 $v = self
::getLocalisationCache()->getItem( $code, 'fallback' );
4125 $v = array_map( 'trim', explode( ',', $v ) );
4126 if ( $v[count( $v ) - 1] !== 'en' ) {
4134 * Get the ordered list of fallback languages, ending with the fallback
4135 * language chain for the site language.
4138 * @param string $code Language code
4139 * @return array array( fallbacks, site fallbacks )
4141 public static function getFallbacksIncludingSiteLanguage( $code ) {
4142 global $wgLanguageCode;
4144 // Usually, we will only store a tiny number of fallback chains, so we
4145 // keep them in static memory.
4146 $cacheKey = "{$code}-{$wgLanguageCode}";
4148 if ( !array_key_exists( $cacheKey, self
::$fallbackLanguageCache ) ) {
4149 $fallbacks = self
::getFallbacksFor( $code );
4151 // Append the site's fallback chain, including the site language itself
4152 $siteFallbacks = self
::getFallbacksFor( $wgLanguageCode );
4153 array_unshift( $siteFallbacks, $wgLanguageCode );
4155 // Eliminate any languages already included in the chain
4156 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4158 self
::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4160 return self
::$fallbackLanguageCache[$cacheKey];
4164 * Get all messages for a given language
4165 * WARNING: this may take a long time. If you just need all message *keys*
4166 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4168 * @param $code string
4172 public static function getMessagesFor( $code ) {
4173 return self
::getLocalisationCache()->getItem( $code, 'messages' );
4177 * Get a message for a given language
4179 * @param $key string
4180 * @param $code string
4184 public static function getMessageFor( $key, $code ) {
4185 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4189 * Get all message keys for a given language. This is a faster alternative to
4190 * array_keys( Language::getMessagesFor( $code ) )
4193 * @param $code string Language code
4194 * @return array of message keys (strings)
4196 public static function getMessageKeysFor( $code ) {
4197 return self
::getLocalisationCache()->getSubItemList( $code, 'messages' );
4204 function fixVariableInNamespace( $talk ) {
4205 if ( strpos( $talk, '$1' ) === false ) {
4209 global $wgMetaNamespace;
4210 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4212 # Allow grammar transformations
4213 # Allowing full message-style parsing would make simple requests
4214 # such as action=raw much more expensive than they need to be.
4215 # This will hopefully cover most cases.
4216 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4217 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4218 return str_replace( ' ', '_', $talk );
4225 function replaceGrammarInNamespace( $m ) {
4226 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4230 * @throws MWException
4233 static function getCaseMaps() {
4234 static $wikiUpperChars, $wikiLowerChars;
4235 if ( isset( $wikiUpperChars ) ) {
4236 return array( $wikiUpperChars, $wikiLowerChars );
4239 wfProfileIn( __METHOD__
);
4240 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4241 if ( $arr === false ) {
4242 throw new MWException(
4243 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4245 $wikiUpperChars = $arr['wikiUpperChars'];
4246 $wikiLowerChars = $arr['wikiLowerChars'];
4247 wfProfileOut( __METHOD__
);
4248 return array( $wikiUpperChars, $wikiLowerChars );
4252 * Decode an expiry (block, protection, etc) which has come from the DB
4254 * @todo FIXME: why are we returnings DBMS-dependent strings???
4256 * @param $expiry String: Database expiry String
4257 * @param $format Bool|Int true to process using language functions, or TS_ constant
4258 * to return the expiry in a given timestamp
4262 public function formatExpiry( $expiry, $format = true ) {
4264 if ( $infinity === null ) {
4265 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
4268 if ( $expiry == '' ||
$expiry == $infinity ) {
4269 return $format === true
4270 ?
$this->getMessageFromDB( 'infiniteblock' )
4273 return $format === true
4274 ?
$this->timeanddate( $expiry, /* User preference timezone */ true )
4275 : wfTimestamp( $format, $expiry );
4281 * @param $seconds int|float
4282 * @param $format Array Optional
4283 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
4284 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
4285 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
4286 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
4289 function formatTimePeriod( $seconds, $format = array() ) {
4290 if ( !is_array( $format ) ) {
4291 $format = array( 'avoid' => $format ); // For backwards compatibility
4293 if ( !isset( $format['avoid'] ) ) {
4294 $format['avoid'] = false;
4296 if ( !isset( $format['noabbrevs' ] ) ) {
4297 $format['noabbrevs'] = false;
4299 $secondsMsg = wfMessage(
4300 $format['noabbrevs'] ?
'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4301 $minutesMsg = wfMessage(
4302 $format['noabbrevs'] ?
'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4303 $hoursMsg = wfMessage(
4304 $format['noabbrevs'] ?
'hours' : 'hours-abbrev' )->inLanguage( $this );
4305 $daysMsg = wfMessage(
4306 $format['noabbrevs'] ?
'days' : 'days-abbrev' )->inLanguage( $this );
4308 if ( round( $seconds * 10 ) < 100 ) {
4309 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4310 $s = $secondsMsg->params( $s )->text();
4311 } elseif ( round( $seconds ) < 60 ) {
4312 $s = $this->formatNum( round( $seconds ) );
4313 $s = $secondsMsg->params( $s )->text();
4314 } elseif ( round( $seconds ) < 3600 ) {
4315 $minutes = floor( $seconds / 60 );
4316 $secondsPart = round( fmod( $seconds, 60 ) );
4317 if ( $secondsPart == 60 ) {
4321 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4323 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4324 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4325 $hours = floor( $seconds / 3600 );
4326 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4327 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4328 if ( $secondsPart == 60 ) {
4332 if ( $minutes == 60 ) {
4336 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4338 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4339 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4340 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4343 $days = floor( $seconds / 86400 );
4344 if ( $format['avoid'] === 'avoidminutes' ) {
4345 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4346 if ( $hours == 24 ) {
4350 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4352 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4353 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4354 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4355 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4356 if ( $minutes == 60 ) {
4360 if ( $hours == 24 ) {
4364 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4366 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4368 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4370 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4372 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4379 * Format a bitrate for output, using an appropriate
4380 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4382 * This use base 1000. For base 1024 use formatSize(), for another base
4383 * see formatComputingNumbers()
4388 function formatBitrate( $bps ) {
4389 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4393 * @param $size int Size of the unit
4394 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4395 * @param $messageKey string Message key to be uesd
4398 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4400 return str_replace( '$1', $this->formatNum( $size ),
4401 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4404 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4407 $maxIndex = count( $sizes ) - 1;
4408 while ( $size >= $boundary && $index < $maxIndex ) {
4413 // For small sizes no decimal places necessary
4416 // For MB and bigger two decimal places are smarter
4419 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4421 $size = round( $size, $round );
4422 $text = $this->getMessageFromDB( $msg );
4423 return str_replace( '$1', $this->formatNum( $size ), $text );
4427 * Format a size in bytes for output, using an appropriate
4428 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4430 * This method use base 1024. For base 1000 use formatBitrate(), for
4431 * another base see formatComputingNumbers()
4433 * @param $size int Size to format
4434 * @return string Plain text (not HTML)
4436 function formatSize( $size ) {
4437 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4441 * Make a list item, used by various special pages
4443 * @param $page String Page link
4444 * @param $details String Text between brackets
4445 * @param $oppositedm Boolean Add the direction mark opposite to your
4446 * language, to display text properly
4449 function specialList( $page, $details, $oppositedm = true ) {
4450 $dirmark = ( $oppositedm ?
$this->getDirMark( true ) : '' ) .
4451 $this->getDirMark();
4452 $details = $details ?
$dirmark . $this->getMessageFromDB( 'word-separator' ) .
4453 wfMessage( 'parentheses' )->rawParams( $details )->inLanguage( $this )->escaped() : '';
4454 return $page . $details;
4458 * Generate (prev x| next x) (20|50|100...) type links for paging
4460 * @param $title Title object to link
4461 * @param $offset Integer offset parameter
4462 * @param $limit Integer limit parameter
4463 * @param $query array|String optional URL query parameter string
4464 * @param $atend Bool optional param for specified if this is the last page
4467 public function viewPrevNext( Title
$title, $offset, $limit, array $query = array(), $atend = false ) {
4468 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4470 # Make 'previous' link
4471 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4472 if ( $offset > 0 ) {
4473 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4474 $query, $prev, 'prevn-title', 'mw-prevlink' );
4476 $plink = htmlspecialchars( $prev );
4480 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4482 $nlink = htmlspecialchars( $next );
4484 $nlink = $this->numLink( $title, $offset +
$limit, $limit,
4485 $query, $next, 'nextn-title', 'mw-nextlink' );
4488 # Make links to set number of items per page
4489 $numLinks = array();
4490 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4491 $numLinks[] = $this->numLink( $title, $offset, $num,
4492 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4495 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4496 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4500 * Helper function for viewPrevNext() that generates links
4502 * @param $title Title object to link
4503 * @param $offset Integer offset parameter
4504 * @param $limit Integer limit parameter
4505 * @param $query Array extra query parameters
4506 * @param $link String text to use for the link; will be escaped
4507 * @param $tooltipMsg String name of the message to use as tooltip
4508 * @param $class String value of the "class" attribute of the link
4509 * @return String HTML fragment
4511 private function numLink( Title
$title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4512 $query = array( 'limit' => $limit, 'offset' => $offset ) +
$query;
4513 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4514 return Html
::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4515 'title' => $tooltip, 'class' => $class ), $link );
4519 * Get the conversion rule title, if any.
4523 public function getConvRuleTitle() {
4524 return $this->mConverter
->getConvRuleTitle();
4528 * Get the compiled plural rules for the language
4530 * @return array Associative array with plural form, and plural rule as key-value pairs
4532 public function getCompiledPluralRules() {
4533 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'compiledPluralRules' );
4534 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4535 if ( !$pluralRules ) {
4536 foreach ( $fallbacks as $fallbackCode ) {
4537 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4538 if ( $pluralRules ) {
4543 return $pluralRules;
4547 * Get the plural rules for the language
4549 * @return array Associative array with plural form number and plural rule as key-value pairs
4551 public function getPluralRules() {
4552 $pluralRules = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRules' );
4553 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4554 if ( !$pluralRules ) {
4555 foreach ( $fallbacks as $fallbackCode ) {
4556 $pluralRules = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4557 if ( $pluralRules ) {
4562 return $pluralRules;
4566 * Get the plural rule types for the language
4568 * @return array Associative array with plural form number and plural rule type as key-value pairs
4570 public function getPluralRuleTypes() {
4571 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $this->mCode
), 'pluralRuleTypes' );
4572 $fallbacks = Language
::getFallbacksFor( $this->mCode
);
4573 if ( !$pluralRuleTypes ) {
4574 foreach ( $fallbacks as $fallbackCode ) {
4575 $pluralRuleTypes = self
::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4576 if ( $pluralRuleTypes ) {
4581 return $pluralRuleTypes;
4585 * Find the index number of the plural rule appropriate for the given number
4586 * @return int The index number of the plural rule
4588 public function getPluralRuleIndexNumber( $number ) {
4589 $pluralRules = $this->getCompiledPluralRules();
4590 $form = CLDRPluralRuleEvaluator
::evaluateCompiled( $number, $pluralRules );
4595 * Find the plural rule type appropriate for the given number
4596 * For example, if the language is set to Arabic, getPluralType(5) should
4599 * @return string The name of the plural rule type, e.g. one, two, few, many
4601 public function getPluralRuleType( $number ) {
4602 $index = $this->getPluralRuleIndexNumber( $number );
4603 $pluralRuleTypes = $this->getPluralRuleTypes();
4604 if ( isset( $pluralRuleTypes[$index] ) ) {
4605 return $pluralRuleTypes[$index];