3 * Internationalisation code
10 * @defgroup Language Language
13 if ( !defined( 'MEDIAWIKI' ) ) {
14 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
19 global $wgLanguageNames;
20 require_once( dirname( __FILE__
) . '/Names.php' );
22 if ( function_exists( 'mb_strtoupper' ) ) {
23 mb_internal_encoding( 'UTF-8' );
27 * a fake language converter
33 function __construct( $langobj ) { $this->mLang
= $langobj; }
34 function autoConvertToAllVariants( $text ) { return array( $this->mLang
->getCode() => $text ); }
35 function convert( $t ) { return $t; }
36 function convertTitle( $t ) { return $t->getPrefixedText(); }
37 function getVariants() { return array( $this->mLang
->getCode() ); }
38 function getPreferredVariant() { return $this->mLang
->getCode(); }
39 function getDefaultVariant() { return $this->mLang
->getCode(); }
40 function getURLVariant() { return ''; }
41 function getConvRuleTitle() { return false; }
42 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
43 function getExtraHashOptions() { return ''; }
44 function getParsedTitle() { return ''; }
45 function markNoConversion( $text, $noParse = false ) { return $text; }
46 function convertCategoryKey( $key ) { return $key; }
47 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
48 function armourMath( $text ) { return $text; }
52 * Internationalisation code
56 var $mConverter, $mVariants, $mCode, $mLoaded = false;
57 var $mMagicExtensions = array(), $mMagicHookDone = false;
59 var $mNamespaceIds, $namespaceNames, $namespaceAliases;
60 var $dateFormatStrings = array();
61 var $mExtendedSpecialPageAliases;
64 * ReplacementArray object caches
66 var $transformData = array();
69 * @var LocalisationCache
71 static public $dataCache;
73 static public $mLangObjCache = array();
75 static public $mWeekdayMsgs = array(
76 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
80 static public $mWeekdayAbbrevMsgs = array(
81 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
84 static public $mMonthMsgs = array(
85 'january', 'february', 'march', 'april', 'may_long', 'june',
86 'july', 'august', 'september', 'october', 'november',
89 static public $mMonthGenMsgs = array(
90 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
91 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
94 static public $mMonthAbbrevMsgs = array(
95 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
96 'sep', 'oct', 'nov', 'dec'
99 static public $mIranianCalendarMonthMsgs = array(
100 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
101 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
102 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
103 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
106 static public $mHebrewCalendarMonthMsgs = array(
107 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
108 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
109 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
110 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
111 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
114 static public $mHebrewCalendarMonthGenMsgs = array(
115 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
116 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
117 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
118 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
119 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
122 static public $mHijriCalendarMonthMsgs = array(
123 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
124 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
125 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
126 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
130 * Get a cached language object for a given language code
131 * @param $code String
134 static function factory( $code ) {
135 if ( !isset( self
::$mLangObjCache[$code] ) ) {
136 if ( count( self
::$mLangObjCache ) > 10 ) {
137 // Don't keep a billion objects around, that's stupid.
138 self
::$mLangObjCache = array();
140 self
::$mLangObjCache[$code] = self
::newFromCode( $code );
142 return self
::$mLangObjCache[$code];
146 * Create a language object for a given language code
147 * @param $code String
150 protected static function newFromCode( $code ) {
152 static $recursionLevel = 0;
154 // Protect against path traversal below
155 if ( !Language
::isValidCode( $code )
156 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
158 throw new MWException( "Invalid language code \"$code\"" );
161 if ( !Language
::isValidBuiltInCode( $code ) ) {
162 // It's not possible to customise this code with class files, so
163 // just return a Language object. This is to support uselang= hacks.
164 $lang = new Language
;
165 $lang->setCode( $code );
169 if ( $code == 'en' ) {
172 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
173 if ( !defined( 'MW_COMPILED' ) ) {
174 // Preload base classes to work around APC/PHP5 bug
175 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
176 include_once( "$IP/languages/classes/$class.deps.php" );
178 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
179 include_once( "$IP/languages/classes/$class.php" );
184 if ( $recursionLevel > 5 ) {
185 throw new MWException( "Language fallback loop detected when creating class $class\n" );
188 if ( !MWInit
::classExists( $class ) ) {
189 $fallback = Language
::getFallbackFor( $code );
191 $lang = Language
::newFromCode( $fallback );
193 $lang->setCode( $code );
201 * Returns true if a language code string is of a valid form, whether or
202 * not it exists. This includes codes which are used solely for
203 * customisation via the MediaWiki namespace.
205 public static function isValidCode( $code ) {
207 strcspn( $code, ":/\\\000" ) === strlen( $code )
208 && !preg_match( Title
::getTitleInvalidRegex(), $code );
212 * Returns true if a language code is of a valid form for the purposes of
213 * internal customisation of MediaWiki, via Messages*.php.
215 public static function isValidBuiltInCode( $code ) {
216 return preg_match( '/^[a-z0-9-]*$/i', $code );
220 * Get the LocalisationCache instance
222 * @return LocalisationCache
224 public static function getLocalisationCache() {
225 if ( is_null( self
::$dataCache ) ) {
226 global $wgLocalisationCacheConf;
227 $class = $wgLocalisationCacheConf['class'];
228 self
::$dataCache = new $class( $wgLocalisationCacheConf );
230 return self
::$dataCache;
233 function __construct() {
234 $this->mConverter
= new FakeConverter( $this );
235 // Set the code to the name of the descendant
236 if ( get_class( $this ) == 'Language' ) {
239 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
241 self
::getLocalisationCache();
245 * Reduce memory usage
247 function __destruct() {
248 foreach ( $this as $name => $value ) {
249 unset( $this->$name );
254 * Hook which will be called if this is the content language.
255 * Descendants can use this to register hook functions or modify globals
257 function initContLang() { }
259 function getFallbackLanguageCode() {
260 if ( $this->mCode
=== 'en' ) {
263 return self
::$dataCache->getItem( $this->mCode
, 'fallback' );
268 * Exports $wgBookstoreListEn
271 function getBookstoreList() {
272 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
278 function getNamespaces() {
279 if ( is_null( $this->namespaceNames
) ) {
280 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
282 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
283 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
285 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
287 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
288 if ( $wgMetaNamespaceTalk ) {
289 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
291 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
292 $this->namespaceNames
[NS_PROJECT_TALK
] =
293 $this->fixVariableInNamespace( $talk );
296 # Sometimes a language will be localised but not actually exist on this wiki.
297 foreach( $this->namespaceNames
as $key => $text ) {
298 if ( !isset( $validNamespaces[$key] ) ) {
299 unset( $this->namespaceNames
[$key] );
303 # The above mixing may leave namespaces out of canonical order.
304 # Re-order by namespace ID number...
305 ksort( $this->namespaceNames
);
307 return $this->namespaceNames
;
311 * A convenience function that returns the same thing as
312 * getNamespaces() except with the array values changed to ' '
313 * where it found '_', useful for producing output to be displayed
314 * e.g. in <select> forms.
318 function getFormattedNamespaces() {
319 $ns = $this->getNamespaces();
320 foreach ( $ns as $k => $v ) {
321 $ns[$k] = strtr( $v, '_', ' ' );
327 * Get a namespace value by key
329 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
330 * echo $mw_ns; // prints 'MediaWiki'
333 * @param $index Int: the array key of the namespace to return
334 * @return mixed, string if the namespace value exists, otherwise false
336 function getNsText( $index ) {
337 $ns = $this->getNamespaces();
338 return isset( $ns[$index] ) ?
$ns[$index] : false;
342 * A convenience function that returns the same thing as
343 * getNsText() except with '_' changed to ' ', useful for
348 function getFormattedNsText( $index ) {
349 $ns = $this->getNsText( $index );
350 return strtr( $ns, '_', ' ' );
354 * Returns gender-dependent namespace alias if available.
355 * @param $index Int: namespace index
356 * @param $gender String: gender key (male, female... )
360 function getGenderNsText( $index, $gender ) {
361 $ns = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
362 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
366 * Whether this language makes distinguishes genders for example in
371 function needsGenderDistinction() {
372 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
373 return count( $aliases ) > 0;
377 * Get a namespace key by value, case insensitive.
378 * Only matches namespace names for the current language, not the
379 * canonical ones defined in Namespace.php.
381 * @param $text String
382 * @return mixed An integer if $text is a valid value otherwise false
384 function getLocalNsIndex( $text ) {
385 $lctext = $this->lc( $text );
386 $ids = $this->getNamespaceIds();
387 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
390 function getNamespaceAliases() {
391 if ( is_null( $this->namespaceAliases
) ) {
392 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
396 foreach ( $aliases as $name => $index ) {
397 if ( $index === NS_PROJECT_TALK
) {
398 unset( $aliases[$name] );
399 $name = $this->fixVariableInNamespace( $name );
400 $aliases[$name] = $index;
405 $genders = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
406 foreach ( $genders as $index => $forms ) {
407 foreach ( $forms as $alias ) {
408 $aliases[$alias] = $index;
412 $this->namespaceAliases
= $aliases;
414 return $this->namespaceAliases
;
417 function getNamespaceIds() {
418 if ( is_null( $this->mNamespaceIds
) ) {
419 global $wgNamespaceAliases;
420 # Put namespace names and aliases into a hashtable.
421 # If this is too slow, then we should arrange it so that it is done
422 # before caching. The catch is that at pre-cache time, the above
423 # class-specific fixup hasn't been done.
424 $this->mNamespaceIds
= array();
425 foreach ( $this->getNamespaces() as $index => $name ) {
426 $this->mNamespaceIds
[$this->lc( $name )] = $index;
428 foreach ( $this->getNamespaceAliases() as $name => $index ) {
429 $this->mNamespaceIds
[$this->lc( $name )] = $index;
431 if ( $wgNamespaceAliases ) {
432 foreach ( $wgNamespaceAliases as $name => $index ) {
433 $this->mNamespaceIds
[$this->lc( $name )] = $index;
437 return $this->mNamespaceIds
;
442 * Get a namespace key by value, case insensitive. Canonical namespace
443 * names override custom ones defined for the current language.
445 * @param $text String
446 * @return mixed An integer if $text is a valid value otherwise false
448 function getNsIndex( $text ) {
449 $lctext = $this->lc( $text );
450 if ( ( $ns = MWNamespace
::getCanonicalIndex( $lctext ) ) !== null ) {
453 $ids = $this->getNamespaceIds();
454 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
458 * short names for language variants used for language conversion links.
460 * @param $code String
463 function getVariantname( $code ) {
464 return $this->getMessageFromDB( "variantname-$code" );
467 function specialPage( $name ) {
468 $aliases = $this->getSpecialPageAliases();
469 if ( isset( $aliases[$name][0] ) ) {
470 $name = $aliases[$name][0];
472 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
475 function getQuickbarSettings() {
477 $this->getMessage( 'qbsettings-none' ),
478 $this->getMessage( 'qbsettings-fixedleft' ),
479 $this->getMessage( 'qbsettings-fixedright' ),
480 $this->getMessage( 'qbsettings-floatingleft' ),
481 $this->getMessage( 'qbsettings-floatingright' )
485 function getDatePreferences() {
486 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
489 function getDateFormats() {
490 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
493 function getDefaultDateFormat() {
494 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
495 if ( $df === 'dmy or mdy' ) {
496 global $wgAmericanDates;
497 return $wgAmericanDates ?
'mdy' : 'dmy';
503 function getDatePreferenceMigrationMap() {
504 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
507 function getImageFile( $image ) {
508 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
511 function getDefaultUserOptionOverrides() {
512 return self
::$dataCache->getItem( $this->mCode
, 'defaultUserOptionOverrides' );
515 function getExtraUserToggles() {
516 return self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
519 function getUserToggle( $tog ) {
520 return $this->getMessageFromDB( "tog-$tog" );
524 * Get language names, indexed by code.
525 * If $customisedOnly is true, only returns codes with a messages file
527 public static function getLanguageNames( $customisedOnly = false ) {
528 global $wgExtraLanguageNames;
529 static $coreLanguageNames;
531 if ( $coreLanguageNames === null ) {
532 include( MWInit
::compiledPath( 'languages/Names.php' ) );
535 $allNames = $wgExtraLanguageNames +
$coreLanguageNames;
536 if ( !$customisedOnly ) {
542 $dir = opendir( "$IP/languages/messages" );
543 while ( false !== ( $file = readdir( $dir ) ) ) {
544 $code = self
::getCodeFromFileName( $file, 'Messages' );
545 if ( $code && isset( $allNames[$code] ) ) {
546 $names[$code] = $allNames[$code];
554 * Get translated language names. This is done on best effort and
555 * by default this is exactly the same as Language::getLanguageNames.
556 * The CLDR extension provides translated names.
557 * @param $code String Language code.
558 * @return Array language code => language name
561 public static function getTranslatedLanguageNames( $code ) {
563 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $code ) );
565 foreach ( self
::getLanguageNames() as $code => $name ) {
566 if ( !isset( $names[$code] ) ) $names[$code] = $name;
573 * Get a message from the MediaWiki namespace.
575 * @param $msg String: message name
578 function getMessageFromDB( $msg ) {
579 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
582 function getLanguageName( $code ) {
583 $names = self
::getLanguageNames();
584 if ( !array_key_exists( $code, $names ) ) {
587 return $names[$code];
590 function getMonthName( $key ) {
591 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
594 function getMonthNamesArray() {
595 $monthNames = array( '' );
596 for ( $i=1; $i < 13; $i++
) {
597 $monthNames[] = $this->getMonthName( $i );
602 function getMonthNameGen( $key ) {
603 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
606 function getMonthAbbreviation( $key ) {
607 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
610 function getMonthAbbreviationsArray() {
611 $monthNames = array('');
612 for ( $i=1; $i < 13; $i++
) {
613 $monthNames[] = $this->getMonthAbbreviation( $i );
618 function getWeekdayName( $key ) {
619 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
622 function getWeekdayAbbreviation( $key ) {
623 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
626 function getIranianCalendarMonthName( $key ) {
627 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
630 function getHebrewCalendarMonthName( $key ) {
631 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
634 function getHebrewCalendarMonthNameGen( $key ) {
635 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
638 function getHijriCalendarMonthName( $key ) {
639 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
643 * Used by date() and time() to adjust the time output.
645 * @param $ts Int the time in date('YmdHis') format
646 * @param $tz Mixed: adjust the time by this amount (default false, mean we
647 * get user timecorrection setting)
650 function userAdjust( $ts, $tz = false ) {
651 global $wgUser, $wgLocalTZoffset;
653 if ( $tz === false ) {
654 $tz = $wgUser->getOption( 'timecorrection' );
657 $data = explode( '|', $tz, 3 );
659 if ( $data[0] == 'ZoneInfo' ) {
660 if ( function_exists( 'timezone_open' ) && @timezone_open
( $data[2] ) !== false ) {
661 $date = date_create( $ts, timezone_open( 'UTC' ) );
662 date_timezone_set( $date, timezone_open( $data[2] ) );
663 $date = date_format( $date, 'YmdHis' );
666 # Unrecognized timezone, default to 'Offset' with the stored offset.
671 if ( $data[0] == 'System' ||
$tz == '' ) {
672 # Global offset in minutes.
673 if ( isset( $wgLocalTZoffset ) ) {
674 $minDiff = $wgLocalTZoffset;
676 } else if ( $data[0] == 'Offset' ) {
677 $minDiff = intval( $data[1] );
679 $data = explode( ':', $tz );
680 if ( count( $data ) == 2 ) {
681 $data[0] = intval( $data[0] );
682 $data[1] = intval( $data[1] );
683 $minDiff = abs( $data[0] ) * 60 +
$data[1];
684 if ( $data[0] < 0 ) {
685 $minDiff = -$minDiff;
688 $minDiff = intval( $data[0] ) * 60;
692 # No difference ? Return time unchanged
693 if ( 0 == $minDiff ) {
697 wfSuppressWarnings(); // E_STRICT system time bitching
698 # Generate an adjusted date; take advantage of the fact that mktime
699 # will normalize out-of-range values so we don't have to split $minDiff
700 # into hours and minutes.
702 (int)substr( $ts, 8, 2 ) ), # Hours
703 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
704 (int)substr( $ts, 12, 2 ), # Seconds
705 (int)substr( $ts, 4, 2 ), # Month
706 (int)substr( $ts, 6, 2 ), # Day
707 (int)substr( $ts, 0, 4 ) ); # Year
709 $date = date( 'YmdHis', $t );
716 * This is a workalike of PHP's date() function, but with better
717 * internationalisation, a reduced set of format characters, and a better
720 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
721 * PHP manual for definitions. There are a number of extensions, which
724 * xn Do not translate digits of the next numeric format character
725 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
726 * xr Use roman numerals for the next numeric format character
727 * xh Use hebrew numerals for the next numeric format character
729 * xg Genitive month name
731 * xij j (day number) in Iranian calendar
732 * xiF F (month name) in Iranian calendar
733 * xin n (month number) in Iranian calendar
734 * xiY Y (full year) in Iranian calendar
736 * xjj j (day number) in Hebrew calendar
737 * xjF F (month name) in Hebrew calendar
738 * xjt t (days in month) in Hebrew calendar
739 * xjx xg (genitive month name) in Hebrew calendar
740 * xjn n (month number) in Hebrew calendar
741 * xjY Y (full year) in Hebrew calendar
743 * xmj j (day number) in Hijri calendar
744 * xmF F (month name) in Hijri calendar
745 * xmn n (month number) in Hijri calendar
746 * xmY Y (full year) in Hijri calendar
748 * xkY Y (full year) in Thai solar calendar. Months and days are
749 * identical to the Gregorian calendar
750 * xoY Y (full year) in Minguo calendar or Juche year.
751 * Months and days are identical to the
753 * xtY Y (full year) in Japanese nengo. Months and days are
754 * identical to the Gregorian calendar
756 * Characters enclosed in double quotes will be considered literal (with
757 * the quotes themselves removed). Unmatched quotes will be considered
758 * literal quotes. Example:
760 * "The month is" F => The month is January
763 * Backslash escaping is also supported.
765 * Input timestamp is assumed to be pre-normalized to the desired local
768 * @param $format String
769 * @param $ts String: 14-character timestamp
772 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
774 function sprintfDate( $format, $ts ) {
787 for ( $p = 0; $p < strlen( $format ); $p++
) {
790 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
791 $code .= $format[++
$p];
794 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
795 $code .= $format[++
$p];
806 $rawToggle = !$rawToggle;
815 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
818 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
819 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
822 $num = substr( $ts, 6, 2 );
825 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
826 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) +
1 );
829 $num = intval( substr( $ts, 6, 2 ) );
833 $iranian = self
::tsToIranian( $ts );
839 $hijri = self
::tsToHijri( $ts );
845 $hebrew = self
::tsToHebrew( $ts );
851 $unix = wfTimestamp( TS_UNIX
, $ts );
853 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) +
1 );
857 $unix = wfTimestamp( TS_UNIX
, $ts );
859 $w = gmdate( 'w', $unix );
864 $unix = wfTimestamp( TS_UNIX
, $ts );
866 $num = gmdate( 'w', $unix );
870 $unix = wfTimestamp( TS_UNIX
, $ts );
872 $num = gmdate( 'z', $unix );
876 $unix = wfTimestamp( TS_UNIX
, $ts );
878 $num = gmdate( 'W', $unix );
881 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
885 $iranian = self
::tsToIranian( $ts );
887 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
891 $hijri = self
::tsToHijri( $ts );
893 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
897 $hebrew = self
::tsToHebrew( $ts );
899 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
902 $num = substr( $ts, 4, 2 );
905 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
908 $num = intval( substr( $ts, 4, 2 ) );
912 $iranian = self
::tsToIranian( $ts );
918 $hijri = self
::tsToHijri ( $ts );
924 $hebrew = self
::tsToHebrew( $ts );
930 $unix = wfTimestamp( TS_UNIX
, $ts );
932 $num = gmdate( 't', $unix );
936 $hebrew = self
::tsToHebrew( $ts );
942 $unix = wfTimestamp( TS_UNIX
, $ts );
944 $num = gmdate( 'L', $unix );
948 $unix = wfTimestamp( TS_UNIX
, $ts );
950 $num = date( 'o', $unix );
953 $num = substr( $ts, 0, 4 );
957 $iranian = self
::tsToIranian( $ts );
963 $hijri = self
::tsToHijri( $ts );
969 $hebrew = self
::tsToHebrew( $ts );
975 $thai = self
::tsToYear( $ts, 'thai' );
981 $minguo = self
::tsToYear( $ts, 'minguo' );
987 $tenno = self
::tsToYear( $ts, 'tenno' );
992 $num = substr( $ts, 2, 2 );
995 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
998 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1001 $h = substr( $ts, 8, 2 );
1002 $num = $h %
12 ?
$h %
12 : 12;
1005 $num = intval( substr( $ts, 8, 2 ) );
1008 $h = substr( $ts, 8, 2 );
1009 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1012 $num = substr( $ts, 8, 2 );
1015 $num = substr( $ts, 10, 2 );
1018 $num = substr( $ts, 12, 2 );
1022 $unix = wfTimestamp( TS_UNIX
, $ts );
1024 $s .= gmdate( 'c', $unix );
1028 $unix = wfTimestamp( TS_UNIX
, $ts );
1030 $s .= gmdate( 'r', $unix );
1034 $unix = wfTimestamp( TS_UNIX
, $ts );
1039 # Backslash escaping
1040 if ( $p < strlen( $format ) - 1 ) {
1041 $s .= $format[++
$p];
1048 if ( $p < strlen( $format ) - 1 ) {
1049 $endQuote = strpos( $format, '"', $p +
1 );
1050 if ( $endQuote === false ) {
1051 # No terminating quote, assume literal "
1054 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1058 # Quote at end of string, assume literal "
1065 if ( $num !== false ) {
1066 if ( $rawToggle ||
$raw ) {
1069 } elseif ( $roman ) {
1070 $s .= self
::romanNumeral( $num );
1072 } elseif ( $hebrewNum ) {
1073 $s .= self
::hebrewNumeral( $num );
1076 $s .= $this->formatNum( $num, true );
1083 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1084 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1086 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1087 * Gregorian dates to Iranian dates. Originally written in C, it
1088 * is released under the terms of GNU Lesser General Public
1089 * License. Conversion to PHP was performed by Niklas Laxström.
1091 * Link: http://www.farsiweb.info/jalali/jalali.c
1093 private static function tsToIranian( $ts ) {
1094 $gy = substr( $ts, 0, 4 ) -1600;
1095 $gm = substr( $ts, 4, 2 ) -1;
1096 $gd = (int)substr( $ts, 6, 2 ) -1;
1098 # Days passed from the beginning (including leap years)
1100 +
floor( ( $gy +
3 ) / 4 )
1101 - floor( ( $gy +
99 ) / 100 )
1102 +
floor( ( $gy +
399 ) / 400 );
1105 // Add days of the past months of this year
1106 for ( $i = 0; $i < $gm; $i++
) {
1107 $gDayNo +
= self
::$GREG_DAYS[$i];
1111 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1115 // Days passed in current month
1118 $jDayNo = $gDayNo - 79;
1120 $jNp = floor( $jDayNo / 12053 );
1123 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1126 if ( $jDayNo >= 366 ) {
1127 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1128 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1131 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1132 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1138 return array( $jy, $jm, $jd );
1142 * Converting Gregorian dates to Hijri dates.
1144 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1146 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1148 private static function tsToHijri( $ts ) {
1149 $year = substr( $ts, 0, 4 );
1150 $month = substr( $ts, 4, 2 );
1151 $day = substr( $ts, 6, 2 );
1159 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1160 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1163 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1164 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1165 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1168 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1169 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1172 $zl = $zjd -1948440 +
10632;
1173 $zn = (int)( ( $zl - 1 ) / 10631 );
1174 $zl = $zl - 10631 * $zn +
354;
1175 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1176 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1177 $zm = (int)( ( 24 * $zl ) / 709 );
1178 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1179 $zy = 30 * $zn +
$zj - 30;
1181 return array( $zy, $zm, $zd );
1185 * Converting Gregorian dates to Hebrew dates.
1187 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1188 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1189 * to translate the relevant functions into PHP and release them under
1192 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1193 * and Adar II is 14. In a non-leap year, Adar is 6.
1195 private static function tsToHebrew( $ts ) {
1197 $year = substr( $ts, 0, 4 );
1198 $month = substr( $ts, 4, 2 );
1199 $day = substr( $ts, 6, 2 );
1201 # Calculate Hebrew year
1202 $hebrewYear = $year +
3760;
1204 # Month number when September = 1, August = 12
1206 if ( $month > 12 ) {
1213 # Calculate day of year from 1 September
1215 for ( $i = 1; $i < $month; $i++
) {
1219 # Check if the year is leap
1220 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1223 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1230 # Calculate the start of the Hebrew year
1231 $start = self
::hebrewYearStart( $hebrewYear );
1233 # Calculate next year's start
1234 if ( $dayOfYear <= $start ) {
1235 # Day is before the start of the year - it is the previous year
1237 $nextStart = $start;
1241 # Add days since previous year's 1 September
1243 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1247 # Start of the new (previous) year
1248 $start = self
::hebrewYearStart( $hebrewYear );
1251 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1254 # Calculate Hebrew day of year
1255 $hebrewDayOfYear = $dayOfYear - $start;
1257 # Difference between year's days
1258 $diff = $nextStart - $start;
1259 # Add 12 (or 13 for leap years) days to ignore the difference between
1260 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1261 # difference is only about the year type
1262 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1268 # Check the year pattern, and is leap year
1269 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1270 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1271 # and non-leap years
1272 $yearPattern = $diff %
30;
1273 # Check if leap year
1274 $isLeap = $diff >= 30;
1276 # Calculate day in the month from number of day in the Hebrew year
1277 # Don't check Adar - if the day is not in Adar, we will stop before;
1278 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1279 $hebrewDay = $hebrewDayOfYear;
1282 while ( $hebrewMonth <= 12 ) {
1283 # Calculate days in this month
1284 if ( $isLeap && $hebrewMonth == 6 ) {
1285 # Adar in a leap year
1287 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1289 if ( $hebrewDay <= $days ) {
1293 # Subtract the days of Adar I
1294 $hebrewDay -= $days;
1297 if ( $hebrewDay <= $days ) {
1303 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1304 # Cheshvan in a complete year (otherwise as the rule below)
1306 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1307 # Kislev in an incomplete year (otherwise as the rule below)
1310 # Odd months have 30 days, even have 29
1311 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1313 if ( $hebrewDay <= $days ) {
1314 # In the current month
1317 # Subtract the days of the current month
1318 $hebrewDay -= $days;
1319 # Try in the next month
1324 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1328 * This calculates the Hebrew year start, as days since 1 September.
1329 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1330 * Used for Hebrew date.
1332 private static function hebrewYearStart( $year ) {
1333 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1334 $b = intval( ( $year - 1 ) %
4 );
1335 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1339 $Mar = intval( $m );
1345 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1346 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1348 } else if ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1350 } else if ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1354 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1359 * Algorithm to convert Gregorian dates to Thai solar dates,
1360 * Minguo dates or Minguo dates.
1362 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1363 * http://en.wikipedia.org/wiki/Minguo_calendar
1364 * http://en.wikipedia.org/wiki/Japanese_era_name
1366 * @param $ts String: 14-character timestamp
1367 * @param $cName String: calender name
1368 * @return Array: converted year, month, day
1370 private static function tsToYear( $ts, $cName ) {
1371 $gy = substr( $ts, 0, 4 );
1372 $gm = substr( $ts, 4, 2 );
1373 $gd = substr( $ts, 6, 2 );
1375 if ( !strcmp( $cName, 'thai' ) ) {
1377 # Add 543 years to the Gregorian calendar
1378 # Months and days are identical
1379 $gy_offset = $gy +
543;
1380 } else if ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1382 # Deduct 1911 years from the Gregorian calendar
1383 # Months and days are identical
1384 $gy_offset = $gy - 1911;
1385 } else if ( !strcmp( $cName, 'tenno' ) ) {
1386 # Nengō dates up to Meiji period
1387 # Deduct years from the Gregorian calendar
1388 # depending on the nengo periods
1389 # Months and days are identical
1390 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1392 $gy_gannen = $gy - 1868 +
1;
1393 $gy_offset = $gy_gannen;
1394 if ( $gy_gannen == 1 ) {
1397 $gy_offset = '明治' . $gy_offset;
1399 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1400 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1401 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1402 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1403 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1407 $gy_gannen = $gy - 1912 +
1;
1408 $gy_offset = $gy_gannen;
1409 if ( $gy_gannen == 1 ) {
1412 $gy_offset = '大正' . $gy_offset;
1414 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1415 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1416 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1420 $gy_gannen = $gy - 1926 +
1;
1421 $gy_offset = $gy_gannen;
1422 if ( $gy_gannen == 1 ) {
1425 $gy_offset = '昭和' . $gy_offset;
1428 $gy_gannen = $gy - 1989 +
1;
1429 $gy_offset = $gy_gannen;
1430 if ( $gy_gannen == 1 ) {
1433 $gy_offset = '平成' . $gy_offset;
1439 return array( $gy_offset, $gm, $gd );
1443 * Roman number formatting up to 3000
1445 static function romanNumeral( $num ) {
1446 static $table = array(
1447 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1448 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1449 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1450 array( '', 'M', 'MM', 'MMM' )
1453 $num = intval( $num );
1454 if ( $num > 3000 ||
$num <= 0 ) {
1459 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1460 if ( $num >= $pow10 ) {
1461 $s .= $table[$i][floor( $num / $pow10 )];
1463 $num = $num %
$pow10;
1469 * Hebrew Gematria number formatting up to 9999
1471 static function hebrewNumeral( $num ) {
1472 static $table = array(
1473 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1474 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1475 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1476 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1479 $num = intval( $num );
1480 if ( $num > 9999 ||
$num <= 0 ) {
1485 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1486 if ( $num >= $pow10 ) {
1487 if ( $num == 15 ||
$num == 16 ) {
1488 $s .= $table[0][9] . $table[0][$num - 9];
1491 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1492 if ( $pow10 == 1000 ) {
1497 $num = $num %
$pow10;
1499 if ( strlen( $s ) == 2 ) {
1502 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1503 $str .= substr( $s, strlen( $s ) - 2, 2 );
1505 $start = substr( $str, 0, strlen( $str ) - 2 );
1506 $end = substr( $str, strlen( $str ) - 2 );
1509 $str = $start . 'ך';
1512 $str = $start . 'ם';
1515 $str = $start . 'ן';
1518 $str = $start . 'ף';
1521 $str = $start . 'ץ';
1528 * This is meant to be used by time(), date(), and timeanddate() to get
1529 * the date preference they're supposed to use, it should be used in
1533 * function timeanddate([...], $format = true) {
1534 * $datePreference = $this->dateFormat($format);
1539 * @param $usePrefs Mixed: if true, the user's preference is used
1540 * if false, the site/language default is used
1541 * if int/string, assumed to be a format.
1544 function dateFormat( $usePrefs = true ) {
1547 if ( is_bool( $usePrefs ) ) {
1549 $datePreference = $wgUser->getDatePreference();
1551 $datePreference = (string)User
::getDefaultOption( 'date' );
1554 $datePreference = (string)$usePrefs;
1558 if ( $datePreference == '' ) {
1562 return $datePreference;
1566 * Get a format string for a given type and preference
1567 * @param $type May be date, time or both
1568 * @param $pref The format name as it appears in Messages*.php
1570 function getDateFormatString( $type, $pref ) {
1571 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
1572 if ( $pref == 'default' ) {
1573 $pref = $this->getDefaultDateFormat();
1574 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1576 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1577 if ( is_null( $df ) ) {
1578 $pref = $this->getDefaultDateFormat();
1579 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1582 $this->dateFormatStrings
[$type][$pref] = $df;
1584 return $this->dateFormatStrings
[$type][$pref];
1588 * @param $ts Mixed: the time format which needs to be turned into a
1589 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1590 * @param $adj Bool: whether to adjust the time output according to the
1591 * user configured offset ($timecorrection)
1592 * @param $format Mixed: true to use user's date format preference
1593 * @param $timecorrection String: the time offset as returned by
1594 * validateTimeZone() in Special:Preferences
1597 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1598 $ts = wfTimestamp( TS_MW
, $ts );
1600 $ts = $this->userAdjust( $ts, $timecorrection );
1602 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1603 return $this->sprintfDate( $df, $ts );
1607 * @param $ts Mixed: the time format which needs to be turned into a
1608 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1609 * @param $adj Bool: whether to adjust the time output according to the
1610 * user configured offset ($timecorrection)
1611 * @param $format Mixed: true to use user's date format preference
1612 * @param $timecorrection String: the time offset as returned by
1613 * validateTimeZone() in Special:Preferences
1616 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1617 $ts = wfTimestamp( TS_MW
, $ts );
1619 $ts = $this->userAdjust( $ts, $timecorrection );
1621 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1622 return $this->sprintfDate( $df, $ts );
1626 * @param $ts Mixed: the time format which needs to be turned into a
1627 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1628 * @param $adj Bool: whether to adjust the time output according to the
1629 * user configured offset ($timecorrection)
1630 * @param $format Mixed: what format to return, if it's false output the
1631 * default one (default true)
1632 * @param $timecorrection String: the time offset as returned by
1633 * validateTimeZone() in Special:Preferences
1636 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1637 $ts = wfTimestamp( TS_MW
, $ts );
1639 $ts = $this->userAdjust( $ts, $timecorrection );
1641 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1642 return $this->sprintfDate( $df, $ts );
1645 function getMessage( $key ) {
1646 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
1649 function getAllMessages() {
1650 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
1653 function iconv( $in, $out, $string ) {
1654 # This is a wrapper for iconv in all languages except esperanto,
1655 # which does some nasty x-conversions beforehand
1657 # Even with //IGNORE iconv can whine about illegal characters in
1658 # *input* string. We just ignore those too.
1659 # REF: http://bugs.php.net/bug.php?id=37166
1660 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1661 wfSuppressWarnings();
1662 $text = iconv( $in, $out . '//IGNORE', $string );
1663 wfRestoreWarnings();
1667 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1668 function ucwordbreaksCallbackAscii( $matches ) {
1669 return $this->ucfirst( $matches[1] );
1672 function ucwordbreaksCallbackMB( $matches ) {
1673 return mb_strtoupper( $matches[0] );
1676 function ucCallback( $matches ) {
1677 list( $wikiUpperChars ) = self
::getCaseMaps();
1678 return strtr( $matches[1], $wikiUpperChars );
1681 function lcCallback( $matches ) {
1682 list( , $wikiLowerChars ) = self
::getCaseMaps();
1683 return strtr( $matches[1], $wikiLowerChars );
1686 function ucwordsCallbackMB( $matches ) {
1687 return mb_strtoupper( $matches[0] );
1690 function ucwordsCallbackWiki( $matches ) {
1691 list( $wikiUpperChars ) = self
::getCaseMaps();
1692 return strtr( $matches[0], $wikiUpperChars );
1696 * Make a string's first character uppercase
1698 function ucfirst( $str ) {
1700 if ( $o < 96 ) { // if already uppercase...
1702 } elseif ( $o < 128 ) {
1703 return ucfirst( $str ); // use PHP's ucfirst()
1705 // fall back to more complex logic in case of multibyte strings
1706 return $this->uc( $str, true );
1711 * Convert a string to uppercase
1713 function uc( $str, $first = false ) {
1714 if ( function_exists( 'mb_strtoupper' ) ) {
1716 if ( $this->isMultibyte( $str ) ) {
1717 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1719 return ucfirst( $str );
1722 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
1725 if ( $this->isMultibyte( $str ) ) {
1726 $x = $first ?
'^' : '';
1727 return preg_replace_callback(
1728 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1729 array( $this, 'ucCallback' ),
1733 return $first ?
ucfirst( $str ) : strtoupper( $str );
1738 function lcfirst( $str ) {
1741 return strval( $str );
1742 } elseif ( $o >= 128 ) {
1743 return $this->lc( $str, true );
1744 } elseif ( $o > 96 ) {
1747 $str[0] = strtolower( $str[0] );
1752 function lc( $str, $first = false ) {
1753 if ( function_exists( 'mb_strtolower' ) ) {
1755 if ( $this->isMultibyte( $str ) ) {
1756 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1758 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1761 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
1764 if ( $this->isMultibyte( $str ) ) {
1765 $x = $first ?
'^' : '';
1766 return preg_replace_callback(
1767 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1768 array( $this, 'lcCallback' ),
1772 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1777 function isMultibyte( $str ) {
1778 return (bool)preg_match( '/[\x80-\xff]/', $str );
1781 function ucwords( $str ) {
1782 if ( $this->isMultibyte( $str ) ) {
1783 $str = $this->lc( $str );
1785 // regexp to find first letter in each word (i.e. after each space)
1786 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1788 // function to use to capitalize a single char
1789 if ( function_exists( 'mb_strtoupper' ) ) {
1790 return preg_replace_callback(
1792 array( $this, 'ucwordsCallbackMB' ),
1796 return preg_replace_callback(
1798 array( $this, 'ucwordsCallbackWiki' ),
1803 return ucwords( strtolower( $str ) );
1807 # capitalize words at word breaks
1808 function ucwordbreaks( $str ) {
1809 if ( $this->isMultibyte( $str ) ) {
1810 $str = $this->lc( $str );
1812 // since \b doesn't work for UTF-8, we explicitely define word break chars
1813 $breaks = "[ \-\(\)\}\{\.,\?!]";
1815 // find first letter after word break
1816 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1818 if ( function_exists( 'mb_strtoupper' ) ) {
1819 return preg_replace_callback(
1821 array( $this, 'ucwordbreaksCallbackMB' ),
1825 return preg_replace_callback(
1827 array( $this, 'ucwordsCallbackWiki' ),
1832 return preg_replace_callback(
1833 '/\b([\w\x80-\xff]+)\b/',
1834 array( $this, 'ucwordbreaksCallbackAscii' ),
1841 * Return a case-folded representation of $s
1843 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1844 * and $s2 are the same except for the case of their characters. It is not
1845 * necessary for the value returned to make sense when displayed.
1847 * Do *not* perform any other normalisation in this function. If a caller
1848 * uses this function when it should be using a more general normalisation
1849 * function, then fix the caller.
1851 function caseFold( $s ) {
1852 return $this->uc( $s );
1855 function checkTitleEncoding( $s ) {
1856 if ( is_array( $s ) ) {
1857 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1859 # Check for non-UTF-8 URLs
1860 $ishigh = preg_match( '/[\x80-\xff]/', $s );
1865 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1866 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1871 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
1874 function fallback8bitEncoding() {
1875 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
1879 * Most writing systems use whitespace to break up words.
1880 * Some languages such as Chinese don't conventionally do this,
1881 * which requires special handling when breaking up words for
1884 function hasWordBreaks() {
1889 * Some languages such as Chinese require word segmentation,
1890 * Specify such segmentation when overridden in derived class.
1892 * @param $string String
1895 function segmentByWord( $string ) {
1900 * Some languages have special punctuation need to be normalized.
1901 * Make such changes here.
1903 * @param $string String
1906 function normalizeForSearch( $string ) {
1907 return self
::convertDoubleWidth( $string );
1911 * convert double-width roman characters to single-width.
1912 * range: ff00-ff5f ~= 0020-007f
1914 protected static function convertDoubleWidth( $string ) {
1915 static $full = null;
1916 static $half = null;
1918 if ( $full === null ) {
1919 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1920 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1921 $full = str_split( $fullWidth, 3 );
1922 $half = str_split( $halfWidth );
1925 $string = str_replace( $full, $half, $string );
1929 protected static function insertSpace( $string, $pattern ) {
1930 $string = preg_replace( $pattern, " $1 ", $string );
1931 $string = preg_replace( '/ +/', ' ', $string );
1935 function convertForSearchResult( $termsArray ) {
1936 # some languages, e.g. Chinese, need to do a conversion
1937 # in order for search results to be displayed correctly
1942 * Get the first character of a string.
1947 function firstChar( $s ) {
1950 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1951 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
1956 if ( isset( $matches[1] ) ) {
1957 if ( strlen( $matches[1] ) != 3 ) {
1961 // Break down Hangul syllables to grab the first jamo
1962 $code = utf8ToCodepoint( $matches[1] );
1963 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
1965 } elseif ( $code < 0xb098 ) {
1966 return "\xe3\x84\xb1";
1967 } elseif ( $code < 0xb2e4 ) {
1968 return "\xe3\x84\xb4";
1969 } elseif ( $code < 0xb77c ) {
1970 return "\xe3\x84\xb7";
1971 } elseif ( $code < 0xb9c8 ) {
1972 return "\xe3\x84\xb9";
1973 } elseif ( $code < 0xbc14 ) {
1974 return "\xe3\x85\x81";
1975 } elseif ( $code < 0xc0ac ) {
1976 return "\xe3\x85\x82";
1977 } elseif ( $code < 0xc544 ) {
1978 return "\xe3\x85\x85";
1979 } elseif ( $code < 0xc790 ) {
1980 return "\xe3\x85\x87";
1981 } elseif ( $code < 0xcc28 ) {
1982 return "\xe3\x85\x88";
1983 } elseif ( $code < 0xce74 ) {
1984 return "\xe3\x85\x8a";
1985 } elseif ( $code < 0xd0c0 ) {
1986 return "\xe3\x85\x8b";
1987 } elseif ( $code < 0xd30c ) {
1988 return "\xe3\x85\x8c";
1989 } elseif ( $code < 0xd558 ) {
1990 return "\xe3\x85\x8d";
1992 return "\xe3\x85\x8e";
1999 function initEncoding() {
2000 # Some languages may have an alternate char encoding option
2001 # (Esperanto X-coding, Japanese furigana conversion, etc)
2002 # If this language is used as the primary content language,
2003 # an override to the defaults can be set here on startup.
2006 function recodeForEdit( $s ) {
2007 # For some languages we'll want to explicitly specify
2008 # which characters make it into the edit box raw
2009 # or are converted in some way or another.
2010 # Note that if wgOutputEncoding is different from
2011 # wgInputEncoding, this text will be further converted
2012 # to wgOutputEncoding.
2013 global $wgEditEncoding;
2014 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2017 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2021 function recodeInput( $s ) {
2022 # Take the previous into account.
2023 global $wgEditEncoding;
2024 if ( $wgEditEncoding != '' ) {
2025 $enc = $wgEditEncoding;
2029 if ( $enc == 'UTF-8' ) {
2032 return $this->iconv( $enc, 'UTF-8', $s );
2037 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2038 * also cleans up certain backwards-compatible sequences, converting them
2039 * to the modern Unicode equivalent.
2041 * This is language-specific for performance reasons only.
2043 function normalize( $s ) {
2044 global $wgAllUnicodeFixes;
2045 $s = UtfNormal
::cleanUp( $s );
2046 if ( $wgAllUnicodeFixes ) {
2047 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2048 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2055 * Transform a string using serialized data stored in the given file (which
2056 * must be in the serialized subdirectory of $IP). The file contains pairs
2057 * mapping source characters to destination characters.
2059 * The data is cached in process memory. This will go faster if you have the
2060 * FastStringSearch extension.
2062 function transformUsingPairFile( $file, $string ) {
2063 if ( !isset( $this->transformData
[$file] ) ) {
2064 $data = wfGetPrecompiledData( $file );
2065 if ( $data === false ) {
2066 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
2068 $this->transformData
[$file] = new ReplacementArray( $data );
2070 return $this->transformData
[$file]->replace( $string );
2074 * For right-to-left language support
2079 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
2083 * Return the correct HTML 'dir' attribute value for this language.
2087 return $this->isRTL() ?
'rtl' : 'ltr';
2091 * Return 'left' or 'right' as appropriate alignment for line-start
2092 * for this language's text direction.
2094 * Should be equivalent to CSS3 'start' text-align value....
2098 function alignStart() {
2099 return $this->isRTL() ?
'right' : 'left';
2103 * Return 'right' or 'left' as appropriate alignment for line-end
2104 * for this language's text direction.
2106 * Should be equivalent to CSS3 'end' text-align value....
2110 function alignEnd() {
2111 return $this->isRTL() ?
'left' : 'right';
2115 * A hidden direction mark (LRM or RLM), depending on the language direction
2119 function getDirMark() {
2120 return $this->isRTL() ?
"\xE2\x80\x8F" : "\xE2\x80\x8E";
2123 function capitalizeAllNouns() {
2124 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2128 * An arrow, depending on the language direction
2132 function getArrow() {
2133 return $this->isRTL() ?
'←' : '→';
2137 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2141 function linkPrefixExtension() {
2142 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2145 function getMagicWords() {
2146 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2149 protected function doMagicHook() {
2150 if ( $this->mMagicHookDone
) {
2153 $this->mMagicHookDone
= true;
2154 wfProfileIn( 'LanguageGetMagic' );
2155 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2156 wfProfileOut( 'LanguageGetMagic' );
2159 # Fill a MagicWord object with data from here
2160 function getMagic( $mw ) {
2161 $this->doMagicHook();
2163 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
2164 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
2166 $magicWords = $this->getMagicWords();
2167 if ( isset( $magicWords[$mw->mId
] ) ) {
2168 $rawEntry = $magicWords[$mw->mId
];
2174 if ( !is_array( $rawEntry ) ) {
2175 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
2177 $mw->mCaseSensitive
= $rawEntry[0];
2178 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
2183 * Add magic words to the extension array
2185 function addMagicWordsByLang( $newWords ) {
2186 $code = $this->getCode();
2187 $fallbackChain = array();
2188 while ( $code && !in_array( $code, $fallbackChain ) ) {
2189 $fallbackChain[] = $code;
2190 $code = self
::getFallbackFor( $code );
2192 if ( !in_array( 'en', $fallbackChain ) ) {
2193 $fallbackChain[] = 'en';
2195 $fallbackChain = array_reverse( $fallbackChain );
2196 foreach ( $fallbackChain as $code ) {
2197 if ( isset( $newWords[$code] ) ) {
2198 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
2204 * Get special page names, as an associative array
2205 * case folded alias => real name
2207 function getSpecialPageAliases() {
2208 // Cache aliases because it may be slow to load them
2209 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
2211 $this->mExtendedSpecialPageAliases
=
2212 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
2213 wfRunHooks( 'LanguageGetSpecialPageAliases',
2214 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
2217 return $this->mExtendedSpecialPageAliases
;
2221 * Italic is unsuitable for some languages
2223 * @param $text String: the text to be emphasized.
2226 function emphasize( $text ) {
2227 return "<em>$text</em>";
2231 * Normally we output all numbers in plain en_US style, that is
2232 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2233 * point twohundredthirtyfive. However this is not sutable for all
2234 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2235 * Icelandic just want to use commas instead of dots, and dots instead
2236 * of commas like "293.291,235".
2238 * An example of this function being called:
2240 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2243 * See LanguageGu.php for the Gujarati implementation and
2244 * $separatorTransformTable on MessageIs.php for
2245 * the , => . and . => , implementation.
2247 * @todo check if it's viable to use localeconv() for the decimal
2249 * @param $number Mixed: the string to be formatted, should be an integer
2250 * or a floating point number.
2251 * @param $nocommafy Bool: set to true for special numbers like dates
2254 function formatNum( $number, $nocommafy = false ) {
2255 global $wgTranslateNumerals;
2256 if ( !$nocommafy ) {
2257 $number = $this->commafy( $number );
2258 $s = $this->separatorTransformTable();
2260 $number = strtr( $number, $s );
2264 if ( $wgTranslateNumerals ) {
2265 $s = $this->digitTransformTable();
2267 $number = strtr( $number, $s );
2274 function parseFormattedNumber( $number ) {
2275 $s = $this->digitTransformTable();
2277 $number = strtr( $number, array_flip( $s ) );
2280 $s = $this->separatorTransformTable();
2282 $number = strtr( $number, array_flip( $s ) );
2285 $number = strtr( $number, array( ',' => '' ) );
2290 * Adds commas to a given number
2295 function commafy( $_ ) {
2296 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2299 function digitTransformTable() {
2300 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
2303 function separatorTransformTable() {
2304 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
2308 * Take a list of strings and build a locale-friendly comma-separated
2309 * list, using the local comma-separator message.
2310 * The last two strings are chained with an "and".
2315 function listToText( $l ) {
2317 $m = count( $l ) - 1;
2319 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2321 for ( $i = $m; $i >= 0; $i-- ) {
2324 } else if ( $i == $m - 1 ) {
2325 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2327 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2335 * Take a list of strings and build a locale-friendly comma-separated
2336 * list, using the local comma-separator message.
2337 * @param $list array of strings to put in a comma list
2340 function commaList( $list ) {
2345 array( 'parsemag', 'escapenoentities', 'language' => $this )
2351 * Take a list of strings and build a locale-friendly semicolon-separated
2352 * list, using the local semicolon-separator message.
2353 * @param $list array of strings to put in a semicolon list
2356 function semicolonList( $list ) {
2360 'semicolon-separator',
2361 array( 'parsemag', 'escapenoentities', 'language' => $this )
2367 * Same as commaList, but separate it with the pipe instead.
2368 * @param $list array of strings to put in a pipe list
2371 function pipeList( $list ) {
2376 array( 'escapenoentities', 'language' => $this )
2382 * Truncate a string to a specified length in bytes, appending an optional
2383 * string (e.g. for ellipses)
2385 * The database offers limited byte lengths for some columns in the database;
2386 * multi-byte character sets mean we need to ensure that only whole characters
2387 * are included, otherwise broken characters can be passed to the user
2389 * If $length is negative, the string will be truncated from the beginning
2391 * @param $string String to truncate
2392 * @param $length Int: maximum length (including ellipses)
2393 * @param $ellipsis String to append to the truncated text
2394 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
2395 * $adjustLength was introduced in 1.18, before that behaved as if false.
2398 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
2399 # Use the localized ellipsis character
2400 if ( $ellipsis == '...' ) {
2401 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2403 # Check if there is no need to truncate
2404 if ( $length == 0 ) {
2405 return $ellipsis; // convention
2406 } elseif ( strlen( $string ) <= abs( $length ) ) {
2407 return $string; // no need to truncate
2409 $stringOriginal = $string;
2410 # If ellipsis length is >= $length then we can't apply $adjustLength
2411 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
2412 $string = $ellipsis; // this can be slightly unexpected
2413 # Otherwise, truncate and add ellipsis...
2415 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
2416 if ( $length > 0 ) {
2417 $length -= $eLength;
2418 $string = substr( $string, 0, $length ); // xyz...
2419 $string = $this->removeBadCharLast( $string );
2420 $string = $string . $ellipsis;
2422 $length +
= $eLength;
2423 $string = substr( $string, $length ); // ...xyz
2424 $string = $this->removeBadCharFirst( $string );
2425 $string = $ellipsis . $string;
2428 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
2429 # This check is *not* redundant if $adjustLength, due to the single case where
2430 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
2431 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
2434 return $stringOriginal;
2439 * Remove bytes that represent an incomplete Unicode character
2440 * at the end of string (e.g. bytes of the char are missing)
2442 * @param $string String
2445 protected function removeBadCharLast( $string ) {
2446 if ( $string != '' ) {
2447 $char = ord( $string[strlen( $string ) - 1] );
2449 if ( $char >= 0xc0 ) {
2450 # We got the first byte only of a multibyte char; remove it.
2451 $string = substr( $string, 0, -1 );
2452 } elseif ( $char >= 0x80 &&
2453 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2454 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
2456 # We chopped in the middle of a character; remove it
2464 * Remove bytes that represent an incomplete Unicode character
2465 * at the start of string (e.g. bytes of the char are missing)
2467 * @param $string String
2470 protected function removeBadCharFirst( $string ) {
2471 if ( $string != '' ) {
2472 $char = ord( $string[0] );
2473 if ( $char >= 0x80 && $char < 0xc0 ) {
2474 # We chopped in the middle of a character; remove the whole thing
2475 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2482 * Truncate a string of valid HTML to a specified length in bytes,
2483 * appending an optional string (e.g. for ellipses), and return valid HTML
2485 * This is only intended for styled/linked text, such as HTML with
2486 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
2487 * Also, this will not detect things like "display:none" CSS.
2489 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
2491 * @param string $text HTML string to truncate
2492 * @param int $length (zero/positive) Maximum length (including ellipses)
2493 * @param string $ellipsis String to append to the truncated text
2496 function truncateHtml( $text, $length, $ellipsis = '...' ) {
2497 # Use the localized ellipsis character
2498 if ( $ellipsis == '...' ) {
2499 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2501 # Check if there is clearly no need to truncate
2502 if ( $length <= 0 ) {
2503 return $ellipsis; // no text shown, nothing to format (convention)
2504 } elseif ( strlen( $text ) <= $length ) {
2505 return $text; // string short enough even *with* HTML (short-circuit)
2508 $displayLen = 0; // innerHTML legth so far
2509 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
2510 $tagType = 0; // 0-open, 1-close
2511 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
2512 $entityState = 0; // 0-not entity, 1-entity
2513 $tag = $ret = $pRet = ''; // accumulated tag name, accumulated result string
2514 $openTags = array(); // open tag stack
2515 $pOpenTags = array();
2517 $textLen = strlen( $text );
2518 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
2519 for ( $pos = 0; true; ++
$pos ) {
2520 # Consider truncation once the display length has reached the maximim.
2521 # Check that we're not in the middle of a bracket/entity...
2522 if ( $displayLen >= $neLength && $bracketState == 0 && $entityState == 0 ) {
2523 if ( !$testingEllipsis ) {
2524 $testingEllipsis = true;
2525 # Save where we are; we will truncate here unless there turn out to
2526 # be so few remaining characters that truncation is not necessary.
2527 $pOpenTags = $openTags; // save state
2528 $pRet = $ret; // save state
2529 } elseif ( $displayLen > $length && $displayLen > strlen( $ellipsis ) ) {
2530 # String in fact does need truncation, the truncation point was OK.
2531 $openTags = $pOpenTags; // reload state
2532 $ret = $this->removeBadCharLast( $pRet ); // reload state, multi-byte char fix
2533 $ret .= $ellipsis; // add ellipsis
2537 if ( $pos >= $textLen ) break; // extra iteration just for above checks
2539 # Read the next char...
2541 $lastCh = $pos ?
$text[$pos - 1] : '';
2542 $ret .= $ch; // add to result string
2544 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
2545 $entityState = 0; // for bad HTML
2546 $bracketState = 1; // tag started (checking for backslash)
2547 } elseif ( $ch == '>' ) {
2548 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
2549 $entityState = 0; // for bad HTML
2550 $bracketState = 0; // out of brackets
2551 } elseif ( $bracketState == 1 ) {
2553 $tagType = 1; // close tag (e.g. "</span>")
2555 $tagType = 0; // open tag (e.g. "<span>")
2558 $bracketState = 2; // building tag name
2559 } elseif ( $bracketState == 2 ) {
2563 // Name found (e.g. "<a href=..."), add on tag attributes...
2564 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
2566 } elseif ( $bracketState == 0 ) {
2567 if ( $entityState ) {
2570 $displayLen++
; // entity is one displayed char
2574 $entityState = 1; // entity found, (e.g. " ")
2576 $displayLen++
; // this char is displayed
2577 // Add the next $max display text chars after this in one swoop...
2578 $max = ( $testingEllipsis ?
$length : $neLength ) - $displayLen;
2579 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
2580 $displayLen +
= $skipped;
2586 if ( $displayLen == 0 ) {
2587 return ''; // no text shown, nothing to format
2589 // Close the last tag if left unclosed by bad HTML
2590 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
2591 while ( count( $openTags ) > 0 ) {
2592 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
2597 // truncateHtml() helper function
2598 // like strcspn() but adds the skipped chars to $ret
2599 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
2600 if ( $len === null ) {
2601 $len = -1; // -1 means "no limit" for strcspn
2602 } elseif ( $len < 0 ) {
2606 if ( $start < strlen( $text ) ) {
2607 $skipCount = strcspn( $text, $search, $start, $len );
2608 $ret .= substr( $text, $start, $skipCount );
2614 * truncateHtml() helper function
2615 * (a) push or pop $tag from $openTags as needed
2616 * (b) clear $tag value
2617 * @param String &$tag Current HTML tag name we are looking at
2618 * @param int $tagType (0-open tag, 1-close tag)
2619 * @param char $lastCh Character before the '>' that ended this tag
2620 * @param array &$openTags Open tag stack (not accounting for $tag)
2622 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
2623 $tag = ltrim( $tag );
2625 if ( $tagType == 0 && $lastCh != '/' ) {
2626 $openTags[] = $tag; // tag opened (didn't close itself)
2627 } else if ( $tagType == 1 ) {
2628 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
2629 array_pop( $openTags ); // tag closed
2637 * Grammatical transformations, needed for inflected languages
2638 * Invoked by putting {{grammar:case|word}} in a message
2640 * @param $word string
2641 * @param $case string
2644 function convertGrammar( $word, $case ) {
2645 global $wgGrammarForms;
2646 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
2647 return $wgGrammarForms[$this->getCode()][$case][$word];
2653 * Provides an alternative text depending on specified gender.
2654 * Usage {{gender:username|masculine|feminine|neutral}}.
2655 * username is optional, in which case the gender of current user is used,
2656 * but only in (some) interface messages; otherwise default gender is used.
2657 * If second or third parameter are not specified, masculine is used.
2658 * These details may be overriden per language.
2660 function gender( $gender, $forms ) {
2661 if ( !count( $forms ) ) {
2664 $forms = $this->preConvertPlural( $forms, 2 );
2665 if ( $gender === 'male' ) {
2668 if ( $gender === 'female' ) {
2671 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
2675 * Plural form transformations, needed for some languages.
2676 * For example, there are 3 form of plural in Russian and Polish,
2677 * depending on "count mod 10". See [[w:Plural]]
2678 * For English it is pretty simple.
2680 * Invoked by putting {{plural:count|wordform1|wordform2}}
2681 * or {{plural:count|wordform1|wordform2|wordform3}}
2683 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2685 * @param $count Integer: non-localized number
2686 * @param $forms Array: different plural forms
2687 * @return string Correct form of plural for $count in this language
2689 function convertPlural( $count, $forms ) {
2690 if ( !count( $forms ) ) {
2693 $forms = $this->preConvertPlural( $forms, 2 );
2695 return ( $count == 1 ) ?
$forms[0] : $forms[1];
2699 * Checks that convertPlural was given an array and pads it to requested
2700 * amount of forms by copying the last one.
2702 * @param $count Integer: How many forms should there be at least
2703 * @param $forms Array of forms given to convertPlural
2704 * @return array Padded array of forms or an exception if not an array
2706 protected function preConvertPlural( /* Array */ $forms, $count ) {
2707 while ( count( $forms ) < $count ) {
2708 $forms[] = $forms[count( $forms ) - 1];
2714 * Maybe translate block durations. Note that this function is somewhat misnamed: it
2715 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
2716 * (which is an absolute timestamp).
2717 * @param $str String: the validated block duration in English
2718 * @return Somehow translated block duration
2719 * @see LanguageFi.php for example implementation
2721 function translateBlockExpiry( $str ) {
2722 $duration = SpecialBlock
::getSuggestedDurations( $this );
2723 foreach( $duration as $show => $value ){
2724 if ( strcmp( $str, $value ) == 0 ) {
2725 return htmlspecialchars( trim( $show ) );
2729 // Since usually only infinite or indefinite is only on list, so try
2730 // equivalents if still here.
2731 $indefs = array( 'infinite', 'infinity', 'indefinite' );
2732 if ( in_array( $str, $indefs ) ) {
2733 foreach( $indefs as $val ) {
2734 $show = array_search( $val, $duration, true );
2735 if ( $show !== false ) {
2736 return htmlspecialchars( trim( $show ) );
2740 // If all else fails, return the original string.
2745 * languages like Chinese need to be segmented in order for the diff
2748 * @param $text String
2751 function segmentForDiff( $text ) {
2756 * and unsegment to show the result
2758 * @param $text String
2761 function unsegmentForDiff( $text ) {
2765 # convert text to all supported variants
2766 function autoConvertToAllVariants( $text ) {
2767 return $this->mConverter
->autoConvertToAllVariants( $text );
2770 # convert text to different variants of a language.
2771 function convert( $text ) {
2772 return $this->mConverter
->convert( $text );
2775 # Convert a Title object to a string in the preferred variant
2776 function convertTitle( $title ) {
2777 return $this->mConverter
->convertTitle( $title );
2780 # Check if this is a language with variants
2781 function hasVariants() {
2782 return sizeof( $this->getVariants() ) > 1;
2785 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2786 function armourMath( $text ) {
2787 return $this->mConverter
->armourMath( $text );
2791 * Perform output conversion on a string, and encode for safe HTML output.
2792 * @param $text String text to be converted
2793 * @param $isTitle Bool whether this conversion is for the article title
2795 * @todo this should get integrated somewhere sane
2797 function convertHtml( $text, $isTitle = false ) {
2798 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2801 function convertCategoryKey( $key ) {
2802 return $this->mConverter
->convertCategoryKey( $key );
2806 * Get the list of variants supported by this language
2807 * see sample implementation in LanguageZh.php
2809 * @return array an array of language codes
2811 function getVariants() {
2812 return $this->mConverter
->getVariants();
2815 function getPreferredVariant() {
2816 return $this->mConverter
->getPreferredVariant();
2819 function getDefaultVariant() {
2820 return $this->mConverter
->getDefaultVariant();
2823 function getURLVariant() {
2824 return $this->mConverter
->getURLVariant();
2828 * If a language supports multiple variants, it is
2829 * possible that non-existing link in one variant
2830 * actually exists in another variant. this function
2831 * tries to find it. See e.g. LanguageZh.php
2833 * @param $link String: the name of the link
2834 * @param $nt Mixed: the title object of the link
2835 * @param $ignoreOtherCond Boolean: to disable other conditions when
2836 * we need to transclude a template or update a category's link
2837 * @return null the input parameters may be modified upon return
2839 function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
2840 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
2844 * If a language supports multiple variants, converts text
2845 * into an array of all possible variants of the text:
2846 * 'variant' => text in that variant
2848 * @deprecated since 1.17 Use autoConvertToAllVariants()
2850 function convertLinkToAllVariants( $text ) {
2851 return $this->mConverter
->convertLinkToAllVariants( $text );
2855 * returns language specific options used by User::getPageRenderHash()
2856 * for example, the preferred language variant
2860 function getExtraHashOptions() {
2861 return $this->mConverter
->getExtraHashOptions();
2865 * For languages that support multiple variants, the title of an
2866 * article may be displayed differently in different variants. this
2867 * function returns the apporiate title defined in the body of the article.
2871 function getParsedTitle() {
2872 return $this->mConverter
->getParsedTitle();
2876 * Enclose a string with the "no conversion" tag. This is used by
2877 * various functions in the Parser
2879 * @param $text String: text to be tagged for no conversion
2881 * @return string the tagged text
2883 function markNoConversion( $text, $noParse = false ) {
2884 return $this->mConverter
->markNoConversion( $text, $noParse );
2888 * A regular expression to match legal word-trailing characters
2889 * which should be merged onto a link of the form [[foo]]bar.
2893 function linkTrail() {
2894 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
2897 function getLangObj() {
2902 * Get the RFC 3066 code for this language object
2904 function getCode() {
2905 return $this->mCode
;
2908 function setCode( $code ) {
2909 $this->mCode
= $code;
2913 * Get the name of a file for a certain language code
2914 * @param $prefix string Prepend this to the filename
2915 * @param $code string Language code
2916 * @param $suffix string Append this to the filename
2917 * @return string $prefix . $mangledCode . $suffix
2919 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2920 // Protect against path traversal
2921 if ( !Language
::isValidCode( $code )
2922 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
2924 throw new MWException( "Invalid language code \"$code\"" );
2927 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2931 * Get the language code from a file name. Inverse of getFileName()
2932 * @param $filename string $prefix . $languageCode . $suffix
2933 * @param $prefix string Prefix before the language code
2934 * @param $suffix string Suffix after the language code
2935 * @return Language code, or false if $prefix or $suffix isn't found
2937 static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
2939 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
2940 preg_quote( $suffix, '/' ) . '/', $filename, $m );
2941 if ( !count( $m ) ) {
2944 return str_replace( '_', '-', strtolower( $m[1] ) );
2947 static function getMessagesFileName( $code ) {
2949 return self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2952 static function getClassFileName( $code ) {
2954 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2958 * Get the fallback for a given language
2960 static function getFallbackFor( $code ) {
2961 if ( $code === 'en' ) {
2965 return self
::getLocalisationCache()->getItem( $code, 'fallback' );
2970 * Get all messages for a given language
2971 * WARNING: this may take a long time
2973 static function getMessagesFor( $code ) {
2974 return self
::getLocalisationCache()->getItem( $code, 'messages' );
2978 * Get a message for a given language
2980 static function getMessageFor( $key, $code ) {
2981 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
2984 function fixVariableInNamespace( $talk ) {
2985 if ( strpos( $talk, '$1' ) === false ) {
2989 global $wgMetaNamespace;
2990 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2992 # Allow grammar transformations
2993 # Allowing full message-style parsing would make simple requests
2994 # such as action=raw much more expensive than they need to be.
2995 # This will hopefully cover most cases.
2996 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2997 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2998 return str_replace( ' ', '_', $talk );
3001 function replaceGrammarInNamespace( $m ) {
3002 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
3005 static function getCaseMaps() {
3006 static $wikiUpperChars, $wikiLowerChars;
3007 if ( isset( $wikiUpperChars ) ) {
3008 return array( $wikiUpperChars, $wikiLowerChars );
3011 wfProfileIn( __METHOD__
);
3012 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
3013 if ( $arr === false ) {
3014 throw new MWException(
3015 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
3017 $wikiUpperChars = $arr['wikiUpperChars'];
3018 $wikiLowerChars = $arr['wikiLowerChars'];
3019 wfProfileOut( __METHOD__
);
3020 return array( $wikiUpperChars, $wikiLowerChars );
3024 * Decode an expiry (block, protection, etc) which has come from the DB
3026 * @param $expiry String: Database expiry String
3027 * @param $format Bool|Int true to process using language functions, or TS_ constant
3028 * to return the expiry in a given timestamp
3031 public function formatExpiry( $expiry, $format = true ) {
3032 static $infinity, $infinityMsg;
3033 if( $infinity === null ){
3034 $infinityMsg = wfMessage( 'infiniteblock' );
3035 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
3038 if ( $expiry == '' ||
$expiry == $infinity ) {
3039 return $format === true
3043 return $format === true
3044 ?
$this->timeanddate( $expiry )
3045 : wfTimestamp( $format, $expiry );
3051 * @param $seconds String
3054 function formatTimePeriod( $seconds ) {
3055 if ( round( $seconds * 10 ) < 100 ) {
3056 return $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
3057 } elseif ( round( $seconds ) < 60 ) {
3058 return $this->formatNum( round( $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
3059 } elseif ( round( $seconds ) < 3600 ) {
3060 $minutes = floor( $seconds / 60 );
3061 $secondsPart = round( fmod( $seconds, 60 ) );
3062 if ( $secondsPart == 60 ) {
3066 return $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
3067 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
3069 $hours = floor( $seconds / 3600 );
3070 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
3071 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
3072 if ( $secondsPart == 60 ) {
3076 if ( $minutes == 60 ) {
3080 return $this->formatNum( $hours ) . $this->getMessageFromDB( 'hours-abbrev' ) . ' ' .
3081 $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
3082 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
3086 function formatBitrate( $bps ) {
3087 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
3089 return $this->formatNum( $bps ) . $units[0];
3091 $unitIndex = floor( log10( $bps ) / 3 );
3092 $mantissa = $bps / pow( 1000, $unitIndex );
3093 if ( $mantissa < 10 ) {
3094 $mantissa = round( $mantissa, 1 );
3096 $mantissa = round( $mantissa );
3098 return $this->formatNum( $mantissa ) . $units[$unitIndex];
3102 * Format a size in bytes for output, using an appropriate
3103 * unit (B, KB, MB or GB) according to the magnitude in question
3105 * @param $size Size to format
3106 * @return string Plain text (not HTML)
3108 function formatSize( $size ) {
3109 // For small sizes no decimal places necessary
3111 if ( $size > 1024 ) {
3112 $size = $size / 1024;
3113 if ( $size > 1024 ) {
3114 $size = $size / 1024;
3115 // For MB and bigger two decimal places are smarter
3117 if ( $size > 1024 ) {
3118 $size = $size / 1024;
3119 $msg = 'size-gigabytes';
3121 $msg = 'size-megabytes';
3124 $msg = 'size-kilobytes';
3127 $msg = 'size-bytes';
3129 $size = round( $size, $round );
3130 $text = $this->getMessageFromDB( $msg );
3131 return str_replace( '$1', $this->formatNum( $size ), $text );
3135 * Get the conversion rule title, if any.
3137 function getConvRuleTitle() {
3138 return $this->mConverter
->getConvRuleTitle();