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 global $wgInputEncoding, $wgOutputEncoding;
25 * These are always UTF-8, they exist only for backwards compatibility
27 $wgInputEncoding = 'UTF-8';
28 $wgOutputEncoding = 'UTF-8';
30 if ( function_exists( 'mb_strtoupper' ) ) {
31 mb_internal_encoding( 'UTF-8' );
35 * a fake language converter
41 function __construct( $langobj ) { $this->mLang
= $langobj; }
42 function autoConvertToAllVariants( $text ) { return array( $this->mLang
->getCode() => $text ); }
43 function convert( $t ) { return $t; }
44 function convertTitle( $t ) { return $t->getPrefixedText(); }
45 function getVariants() { return array( $this->mLang
->getCode() ); }
46 function getPreferredVariant() { return $this->mLang
->getCode(); }
47 function getDefaultVariant() { return $this->mLang
->getCode(); }
48 function getURLVariant() { return ''; }
49 function getConvRuleTitle() { return false; }
50 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
51 function getExtraHashOptions() { return ''; }
52 function getParsedTitle() { return ''; }
53 function markNoConversion( $text, $noParse = false ) { return $text; }
54 function convertCategoryKey( $key ) { return $key; }
55 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
56 function armourMath( $text ) { return $text; }
60 * Internationalisation code
64 var $mConverter, $mVariants, $mCode, $mLoaded = false;
65 var $mMagicExtensions = array(), $mMagicHookDone = false;
67 var $mNamespaceIds, $namespaceNames, $namespaceAliases;
68 var $dateFormatStrings = array();
69 var $mExtendedSpecialPageAliases;
72 * ReplacementArray object caches
74 var $transformData = array();
76 static public $dataCache;
77 static public $mLangObjCache = array();
79 static public $mWeekdayMsgs = array(
80 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
84 static public $mWeekdayAbbrevMsgs = array(
85 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
88 static public $mMonthMsgs = array(
89 'january', 'february', 'march', 'april', 'may_long', 'june',
90 'july', 'august', 'september', 'october', 'november',
93 static public $mMonthGenMsgs = array(
94 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
95 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
98 static public $mMonthAbbrevMsgs = array(
99 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
100 'sep', 'oct', 'nov', 'dec'
103 static public $mIranianCalendarMonthMsgs = array(
104 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
105 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
106 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
107 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
110 static public $mHebrewCalendarMonthMsgs = array(
111 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
112 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
113 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
114 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
115 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
118 static public $mHebrewCalendarMonthGenMsgs = array(
119 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
120 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
121 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
122 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
123 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
126 static public $mHijriCalendarMonthMsgs = array(
127 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
128 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
129 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
130 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
134 * Get a cached language object for a given language code
135 * @param $code String
138 static function factory( $code ) {
139 if ( !isset( self
::$mLangObjCache[$code] ) ) {
140 if ( count( self
::$mLangObjCache ) > 10 ) {
141 // Don't keep a billion objects around, that's stupid.
142 self
::$mLangObjCache = array();
144 self
::$mLangObjCache[$code] = self
::newFromCode( $code );
146 return self
::$mLangObjCache[$code];
150 * Create a language object for a given language code
151 * @param $code String
154 protected static function newFromCode( $code ) {
156 static $recursionLevel = 0;
158 // Protect against path traversal below
159 if ( !Language
::isValidCode( $code )
160 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
162 throw new MWException( "Invalid language code \"$code\"" );
165 if ( !Language
::isValidBuiltInCode( $code ) ) {
166 // It's not possible to customise this code with class files, so
167 // just return a Language object. This is to support uselang= hacks.
168 $lang = new Language
;
169 $lang->setCode( $code );
173 if ( $code == 'en' ) {
176 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
177 if ( !defined( 'MW_COMPILED' ) ) {
178 // Preload base classes to work around APC/PHP5 bug
179 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
180 include_once( "$IP/languages/classes/$class.deps.php" );
182 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
183 include_once( "$IP/languages/classes/$class.php" );
188 if ( $recursionLevel > 5 ) {
189 throw new MWException( "Language fallback loop detected when creating class $class\n" );
192 if ( !MWInit
::classExists( $class ) ) {
193 $fallback = Language
::getFallbackFor( $code );
195 $lang = Language
::newFromCode( $fallback );
197 $lang->setCode( $code );
205 * Returns true if a language code string is of a valid form, whether or
206 * not it exists. This includes codes which are used solely for
207 * customisation via the MediaWiki namespace.
209 public static function isValidCode( $code ) {
211 strcspn( $code, ":/\\\000" ) === strlen( $code )
212 && !preg_match( Title
::getTitleInvalidRegex(), $code );
216 * Returns true if a language code is of a valid form for the purposes of
217 * internal customisation of MediaWiki, via Messages*.php.
219 public static function isValidBuiltInCode( $code ) {
220 return preg_match( '/^[a-z0-9-]*$/i', $code );
224 * Get the LocalisationCache instance
226 * @return LocalisationCache
228 public static function getLocalisationCache() {
229 if ( is_null( self
::$dataCache ) ) {
230 global $wgLocalisationCacheConf;
231 $class = $wgLocalisationCacheConf['class'];
232 self
::$dataCache = new $class( $wgLocalisationCacheConf );
234 return self
::$dataCache;
237 function __construct() {
238 $this->mConverter
= new FakeConverter( $this );
239 // Set the code to the name of the descendant
240 if ( get_class( $this ) == 'Language' ) {
243 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
245 self
::getLocalisationCache();
249 * Reduce memory usage
251 function __destruct() {
252 foreach ( $this as $name => $value ) {
253 unset( $this->$name );
258 * Hook which will be called if this is the content language.
259 * Descendants can use this to register hook functions or modify globals
261 function initContLang() { }
264 * @deprecated Use User::getDefaultOptions()
267 function getDefaultUserOptions() {
268 wfDeprecated( __METHOD__
);
269 return User
::getDefaultOptions();
272 function getFallbackLanguageCode() {
273 if ( $this->mCode
=== 'en' ) {
276 return self
::$dataCache->getItem( $this->mCode
, 'fallback' );
281 * Exports $wgBookstoreListEn
284 function getBookstoreList() {
285 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
291 function getNamespaces() {
292 if ( is_null( $this->namespaceNames
) ) {
293 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
295 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
296 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
298 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
300 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
301 if ( $wgMetaNamespaceTalk ) {
302 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
304 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
305 $this->namespaceNames
[NS_PROJECT_TALK
] =
306 $this->fixVariableInNamespace( $talk );
309 # Sometimes a language will be localised but not actually exist on this wiki.
310 foreach( $this->namespaceNames
as $key => $text ) {
311 if ( !isset( $validNamespaces[$key] ) ) {
312 unset( $this->namespaceNames
[$key] );
316 # The above mixing may leave namespaces out of canonical order.
317 # Re-order by namespace ID number...
318 ksort( $this->namespaceNames
);
320 return $this->namespaceNames
;
324 * A convenience function that returns the same thing as
325 * getNamespaces() except with the array values changed to ' '
326 * where it found '_', useful for producing output to be displayed
327 * e.g. in <select> forms.
331 function getFormattedNamespaces() {
332 $ns = $this->getNamespaces();
333 foreach ( $ns as $k => $v ) {
334 $ns[$k] = strtr( $v, '_', ' ' );
340 * Get a namespace value by key
342 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
343 * echo $mw_ns; // prints 'MediaWiki'
346 * @param $index Int: the array key of the namespace to return
347 * @return mixed, string if the namespace value exists, otherwise false
349 function getNsText( $index ) {
350 $ns = $this->getNamespaces();
351 return isset( $ns[$index] ) ?
$ns[$index] : false;
355 * A convenience function that returns the same thing as
356 * getNsText() except with '_' changed to ' ', useful for
361 function getFormattedNsText( $index ) {
362 $ns = $this->getNsText( $index );
363 return strtr( $ns, '_', ' ' );
367 * Returns gender-dependent namespace alias if available.
368 * @param $index Int: namespace index
369 * @param $gender String: gender key (male, female... )
373 function getGenderNsText( $index, $gender ) {
374 $ns = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
375 return isset( $ns[$index][$gender] ) ?
$ns[$index][$gender] : $this->getNsText( $index );
379 * Whether this language makes distinguishes genders for example in
384 function needsGenderDistinction() {
385 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
386 return count( $aliases ) > 0;
390 * Get a namespace key by value, case insensitive.
391 * Only matches namespace names for the current language, not the
392 * canonical ones defined in Namespace.php.
394 * @param $text String
395 * @return mixed An integer if $text is a valid value otherwise false
397 function getLocalNsIndex( $text ) {
398 $lctext = $this->lc( $text );
399 $ids = $this->getNamespaceIds();
400 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
403 function getNamespaceAliases() {
404 if ( is_null( $this->namespaceAliases
) ) {
405 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
409 foreach ( $aliases as $name => $index ) {
410 if ( $index === NS_PROJECT_TALK
) {
411 unset( $aliases[$name] );
412 $name = $this->fixVariableInNamespace( $name );
413 $aliases[$name] = $index;
418 $genders = self
::$dataCache->getItem( $this->mCode
, 'namespaceGenderAliases' );
419 foreach ( $genders as $index => $forms ) {
420 foreach ( $forms as $alias ) {
421 $aliases[$alias] = $index;
425 $this->namespaceAliases
= $aliases;
427 return $this->namespaceAliases
;
430 function getNamespaceIds() {
431 if ( is_null( $this->mNamespaceIds
) ) {
432 global $wgNamespaceAliases;
433 # Put namespace names and aliases into a hashtable.
434 # If this is too slow, then we should arrange it so that it is done
435 # before caching. The catch is that at pre-cache time, the above
436 # class-specific fixup hasn't been done.
437 $this->mNamespaceIds
= array();
438 foreach ( $this->getNamespaces() as $index => $name ) {
439 $this->mNamespaceIds
[$this->lc( $name )] = $index;
441 foreach ( $this->getNamespaceAliases() as $name => $index ) {
442 $this->mNamespaceIds
[$this->lc( $name )] = $index;
444 if ( $wgNamespaceAliases ) {
445 foreach ( $wgNamespaceAliases as $name => $index ) {
446 $this->mNamespaceIds
[$this->lc( $name )] = $index;
450 return $this->mNamespaceIds
;
455 * Get a namespace key by value, case insensitive. Canonical namespace
456 * names override custom ones defined for the current language.
458 * @param $text String
459 * @return mixed An integer if $text is a valid value otherwise false
461 function getNsIndex( $text ) {
462 $lctext = $this->lc( $text );
463 if ( ( $ns = MWNamespace
::getCanonicalIndex( $lctext ) ) !== null ) {
466 $ids = $this->getNamespaceIds();
467 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
471 * short names for language variants used for language conversion links.
473 * @param $code String
476 function getVariantname( $code ) {
477 return $this->getMessageFromDB( "variantname-$code" );
480 function specialPage( $name ) {
481 $aliases = $this->getSpecialPageAliases();
482 if ( isset( $aliases[$name][0] ) ) {
483 $name = $aliases[$name][0];
485 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
488 function getQuickbarSettings() {
490 $this->getMessage( 'qbsettings-none' ),
491 $this->getMessage( 'qbsettings-fixedleft' ),
492 $this->getMessage( 'qbsettings-fixedright' ),
493 $this->getMessage( 'qbsettings-floatingleft' ),
494 $this->getMessage( 'qbsettings-floatingright' )
498 function getDatePreferences() {
499 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
502 function getDateFormats() {
503 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
506 function getDefaultDateFormat() {
507 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
508 if ( $df === 'dmy or mdy' ) {
509 global $wgAmericanDates;
510 return $wgAmericanDates ?
'mdy' : 'dmy';
516 function getDatePreferenceMigrationMap() {
517 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
520 function getImageFile( $image ) {
521 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
524 function getDefaultUserOptionOverrides() {
525 return self
::$dataCache->getItem( $this->mCode
, 'defaultUserOptionOverrides' );
528 function getExtraUserToggles() {
529 return self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
532 function getUserToggle( $tog ) {
533 return $this->getMessageFromDB( "tog-$tog" );
537 * Get language names, indexed by code.
538 * If $customisedOnly is true, only returns codes with a messages file
540 public static function getLanguageNames( $customisedOnly = false ) {
541 global $wgExtraLanguageNames;
542 static $coreLanguageNames;
544 if ( $coreLanguageNames === null ) {
545 include( MWInit
::compiledPath( 'languages/Names.php' ) );
548 $allNames = $wgExtraLanguageNames +
$coreLanguageNames;
549 if ( !$customisedOnly ) {
555 $dir = opendir( "$IP/languages/messages" );
556 while ( false !== ( $file = readdir( $dir ) ) ) {
557 $code = self
::getCodeFromFileName( $file, 'Messages' );
558 if ( $code && isset( $allNames[$code] ) ) {
559 $names[$code] = $allNames[$code];
567 * Get translated language names. This is done on best effort and
568 * by default this is exactly the same as Language::getLanguageNames.
569 * The CLDR extension provides translated names.
570 * @param $code String Language code.
571 * @return Array language code => language name
574 public static function getTranslatedLanguageNames( $code ) {
576 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $code ) );
578 foreach ( self
::getLanguageNames() as $code => $name ) {
579 if ( !isset( $names[$code] ) ) $names[$code] = $name;
586 * Get a message from the MediaWiki namespace.
588 * @param $msg String: message name
591 function getMessageFromDB( $msg ) {
592 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
595 function getLanguageName( $code ) {
596 $names = self
::getLanguageNames();
597 if ( !array_key_exists( $code, $names ) ) {
600 return $names[$code];
603 function getMonthName( $key ) {
604 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
607 function getMonthNamesArray() {
608 $monthNames = array( '' );
609 for ( $i=1; $i < 13; $i++
) {
610 $monthNames[] = $this->getMonthName( $i );
615 function getMonthNameGen( $key ) {
616 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
619 function getMonthAbbreviation( $key ) {
620 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
623 function getMonthAbbreviationsArray() {
624 $monthNames = array('');
625 for ( $i=1; $i < 13; $i++
) {
626 $monthNames[] = $this->getMonthAbbreviation( $i );
631 function getWeekdayName( $key ) {
632 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
635 function getWeekdayAbbreviation( $key ) {
636 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
639 function getIranianCalendarMonthName( $key ) {
640 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
643 function getHebrewCalendarMonthName( $key ) {
644 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
647 function getHebrewCalendarMonthNameGen( $key ) {
648 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
651 function getHijriCalendarMonthName( $key ) {
652 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
656 * Used by date() and time() to adjust the time output.
658 * @param $ts Int the time in date('YmdHis') format
659 * @param $tz Mixed: adjust the time by this amount (default false, mean we
660 * get user timecorrection setting)
663 function userAdjust( $ts, $tz = false ) {
664 global $wgUser, $wgLocalTZoffset;
666 if ( $tz === false ) {
667 $tz = $wgUser->getOption( 'timecorrection' );
670 $data = explode( '|', $tz, 3 );
672 if ( $data[0] == 'ZoneInfo' ) {
673 if ( function_exists( 'timezone_open' ) && @timezone_open
( $data[2] ) !== false ) {
674 $date = date_create( $ts, timezone_open( 'UTC' ) );
675 date_timezone_set( $date, timezone_open( $data[2] ) );
676 $date = date_format( $date, 'YmdHis' );
679 # Unrecognized timezone, default to 'Offset' with the stored offset.
684 if ( $data[0] == 'System' ||
$tz == '' ) {
685 # Global offset in minutes.
686 if ( isset( $wgLocalTZoffset ) ) {
687 $minDiff = $wgLocalTZoffset;
689 } else if ( $data[0] == 'Offset' ) {
690 $minDiff = intval( $data[1] );
692 $data = explode( ':', $tz );
693 if ( count( $data ) == 2 ) {
694 $data[0] = intval( $data[0] );
695 $data[1] = intval( $data[1] );
696 $minDiff = abs( $data[0] ) * 60 +
$data[1];
697 if ( $data[0] < 0 ) {
698 $minDiff = -$minDiff;
701 $minDiff = intval( $data[0] ) * 60;
705 # No difference ? Return time unchanged
706 if ( 0 == $minDiff ) {
710 wfSuppressWarnings(); // E_STRICT system time bitching
711 # Generate an adjusted date; take advantage of the fact that mktime
712 # will normalize out-of-range values so we don't have to split $minDiff
713 # into hours and minutes.
715 (int)substr( $ts, 8, 2 ) ), # Hours
716 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
717 (int)substr( $ts, 12, 2 ), # Seconds
718 (int)substr( $ts, 4, 2 ), # Month
719 (int)substr( $ts, 6, 2 ), # Day
720 (int)substr( $ts, 0, 4 ) ); # Year
722 $date = date( 'YmdHis', $t );
729 * This is a workalike of PHP's date() function, but with better
730 * internationalisation, a reduced set of format characters, and a better
733 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
734 * PHP manual for definitions. There are a number of extensions, which
737 * xn Do not translate digits of the next numeric format character
738 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
739 * xr Use roman numerals for the next numeric format character
740 * xh Use hebrew numerals for the next numeric format character
742 * xg Genitive month name
744 * xij j (day number) in Iranian calendar
745 * xiF F (month name) in Iranian calendar
746 * xin n (month number) in Iranian calendar
747 * xiY Y (full year) in Iranian calendar
749 * xjj j (day number) in Hebrew calendar
750 * xjF F (month name) in Hebrew calendar
751 * xjt t (days in month) in Hebrew calendar
752 * xjx xg (genitive month name) in Hebrew calendar
753 * xjn n (month number) in Hebrew calendar
754 * xjY Y (full year) in Hebrew calendar
756 * xmj j (day number) in Hijri calendar
757 * xmF F (month name) in Hijri calendar
758 * xmn n (month number) in Hijri calendar
759 * xmY Y (full year) in Hijri calendar
761 * xkY Y (full year) in Thai solar calendar. Months and days are
762 * identical to the Gregorian calendar
763 * xoY Y (full year) in Minguo calendar or Juche year.
764 * Months and days are identical to the
766 * xtY Y (full year) in Japanese nengo. Months and days are
767 * identical to the Gregorian calendar
769 * Characters enclosed in double quotes will be considered literal (with
770 * the quotes themselves removed). Unmatched quotes will be considered
771 * literal quotes. Example:
773 * "The month is" F => The month is January
776 * Backslash escaping is also supported.
778 * Input timestamp is assumed to be pre-normalized to the desired local
781 * @param $format String
782 * @param $ts String: 14-character timestamp
785 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
787 function sprintfDate( $format, $ts ) {
800 for ( $p = 0; $p < strlen( $format ); $p++
) {
803 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
804 $code .= $format[++
$p];
807 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
808 $code .= $format[++
$p];
819 $rawToggle = !$rawToggle;
828 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
831 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
832 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
835 $num = substr( $ts, 6, 2 );
838 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
839 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) +
1 );
842 $num = intval( substr( $ts, 6, 2 ) );
846 $iranian = self
::tsToIranian( $ts );
852 $hijri = self
::tsToHijri( $ts );
858 $hebrew = self
::tsToHebrew( $ts );
864 $unix = wfTimestamp( TS_UNIX
, $ts );
866 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) +
1 );
870 $unix = wfTimestamp( TS_UNIX
, $ts );
872 $w = gmdate( 'w', $unix );
877 $unix = wfTimestamp( TS_UNIX
, $ts );
879 $num = gmdate( 'w', $unix );
883 $unix = wfTimestamp( TS_UNIX
, $ts );
885 $num = gmdate( 'z', $unix );
889 $unix = wfTimestamp( TS_UNIX
, $ts );
891 $num = gmdate( 'W', $unix );
894 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
898 $iranian = self
::tsToIranian( $ts );
900 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
904 $hijri = self
::tsToHijri( $ts );
906 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
910 $hebrew = self
::tsToHebrew( $ts );
912 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
915 $num = substr( $ts, 4, 2 );
918 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
921 $num = intval( substr( $ts, 4, 2 ) );
925 $iranian = self
::tsToIranian( $ts );
931 $hijri = self
::tsToHijri ( $ts );
937 $hebrew = self
::tsToHebrew( $ts );
943 $unix = wfTimestamp( TS_UNIX
, $ts );
945 $num = gmdate( 't', $unix );
949 $hebrew = self
::tsToHebrew( $ts );
955 $unix = wfTimestamp( TS_UNIX
, $ts );
957 $num = gmdate( 'L', $unix );
961 $unix = wfTimestamp( TS_UNIX
, $ts );
963 $num = date( 'o', $unix );
966 $num = substr( $ts, 0, 4 );
970 $iranian = self
::tsToIranian( $ts );
976 $hijri = self
::tsToHijri( $ts );
982 $hebrew = self
::tsToHebrew( $ts );
988 $thai = self
::tsToYear( $ts, 'thai' );
994 $minguo = self
::tsToYear( $ts, 'minguo' );
1000 $tenno = self
::tsToYear( $ts, 'tenno' );
1005 $num = substr( $ts, 2, 2 );
1008 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
1011 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
1014 $h = substr( $ts, 8, 2 );
1015 $num = $h %
12 ?
$h %
12 : 12;
1018 $num = intval( substr( $ts, 8, 2 ) );
1021 $h = substr( $ts, 8, 2 );
1022 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
1025 $num = substr( $ts, 8, 2 );
1028 $num = substr( $ts, 10, 2 );
1031 $num = substr( $ts, 12, 2 );
1035 $unix = wfTimestamp( TS_UNIX
, $ts );
1037 $s .= gmdate( 'c', $unix );
1041 $unix = wfTimestamp( TS_UNIX
, $ts );
1043 $s .= gmdate( 'r', $unix );
1047 $unix = wfTimestamp( TS_UNIX
, $ts );
1052 # Backslash escaping
1053 if ( $p < strlen( $format ) - 1 ) {
1054 $s .= $format[++
$p];
1061 if ( $p < strlen( $format ) - 1 ) {
1062 $endQuote = strpos( $format, '"', $p +
1 );
1063 if ( $endQuote === false ) {
1064 # No terminating quote, assume literal "
1067 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
1071 # Quote at end of string, assume literal "
1078 if ( $num !== false ) {
1079 if ( $rawToggle ||
$raw ) {
1082 } elseif ( $roman ) {
1083 $s .= self
::romanNumeral( $num );
1085 } elseif ( $hebrewNum ) {
1086 $s .= self
::hebrewNumeral( $num );
1089 $s .= $this->formatNum( $num, true );
1096 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1097 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1099 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1100 * Gregorian dates to Iranian dates. Originally written in C, it
1101 * is released under the terms of GNU Lesser General Public
1102 * License. Conversion to PHP was performed by Niklas Laxström.
1104 * Link: http://www.farsiweb.info/jalali/jalali.c
1106 private static function tsToIranian( $ts ) {
1107 $gy = substr( $ts, 0, 4 ) -1600;
1108 $gm = substr( $ts, 4, 2 ) -1;
1109 $gd = (int)substr( $ts, 6, 2 ) -1;
1111 # Days passed from the beginning (including leap years)
1113 +
floor( ( $gy +
3 ) / 4 )
1114 - floor( ( $gy +
99 ) / 100 )
1115 +
floor( ( $gy +
399 ) / 400 );
1118 // Add days of the past months of this year
1119 for ( $i = 0; $i < $gm; $i++
) {
1120 $gDayNo +
= self
::$GREG_DAYS[$i];
1124 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1128 // Days passed in current month
1131 $jDayNo = $gDayNo - 79;
1133 $jNp = floor( $jDayNo / 12053 );
1136 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1139 if ( $jDayNo >= 366 ) {
1140 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1141 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1144 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1145 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1151 return array( $jy, $jm, $jd );
1155 * Converting Gregorian dates to Hijri dates.
1157 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1159 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1161 private static function tsToHijri( $ts ) {
1162 $year = substr( $ts, 0, 4 );
1163 $month = substr( $ts, 4, 2 );
1164 $day = substr( $ts, 6, 2 );
1172 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1173 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1176 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1177 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1178 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1181 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1182 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1185 $zl = $zjd -1948440 +
10632;
1186 $zn = (int)( ( $zl - 1 ) / 10631 );
1187 $zl = $zl - 10631 * $zn +
354;
1188 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1189 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1190 $zm = (int)( ( 24 * $zl ) / 709 );
1191 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1192 $zy = 30 * $zn +
$zj - 30;
1194 return array( $zy, $zm, $zd );
1198 * Converting Gregorian dates to Hebrew dates.
1200 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1201 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1202 * to translate the relevant functions into PHP and release them under
1205 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1206 * and Adar II is 14. In a non-leap year, Adar is 6.
1208 private static function tsToHebrew( $ts ) {
1210 $year = substr( $ts, 0, 4 );
1211 $month = substr( $ts, 4, 2 );
1212 $day = substr( $ts, 6, 2 );
1214 # Calculate Hebrew year
1215 $hebrewYear = $year +
3760;
1217 # Month number when September = 1, August = 12
1219 if ( $month > 12 ) {
1226 # Calculate day of year from 1 September
1228 for ( $i = 1; $i < $month; $i++
) {
1232 # Check if the year is leap
1233 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1236 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1243 # Calculate the start of the Hebrew year
1244 $start = self
::hebrewYearStart( $hebrewYear );
1246 # Calculate next year's start
1247 if ( $dayOfYear <= $start ) {
1248 # Day is before the start of the year - it is the previous year
1250 $nextStart = $start;
1254 # Add days since previous year's 1 September
1256 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1260 # Start of the new (previous) year
1261 $start = self
::hebrewYearStart( $hebrewYear );
1264 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1267 # Calculate Hebrew day of year
1268 $hebrewDayOfYear = $dayOfYear - $start;
1270 # Difference between year's days
1271 $diff = $nextStart - $start;
1272 # Add 12 (or 13 for leap years) days to ignore the difference between
1273 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1274 # difference is only about the year type
1275 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1281 # Check the year pattern, and is leap year
1282 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1283 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1284 # and non-leap years
1285 $yearPattern = $diff %
30;
1286 # Check if leap year
1287 $isLeap = $diff >= 30;
1289 # Calculate day in the month from number of day in the Hebrew year
1290 # Don't check Adar - if the day is not in Adar, we will stop before;
1291 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1292 $hebrewDay = $hebrewDayOfYear;
1295 while ( $hebrewMonth <= 12 ) {
1296 # Calculate days in this month
1297 if ( $isLeap && $hebrewMonth == 6 ) {
1298 # Adar in a leap year
1300 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1302 if ( $hebrewDay <= $days ) {
1306 # Subtract the days of Adar I
1307 $hebrewDay -= $days;
1310 if ( $hebrewDay <= $days ) {
1316 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1317 # Cheshvan in a complete year (otherwise as the rule below)
1319 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1320 # Kislev in an incomplete year (otherwise as the rule below)
1323 # Odd months have 30 days, even have 29
1324 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1326 if ( $hebrewDay <= $days ) {
1327 # In the current month
1330 # Subtract the days of the current month
1331 $hebrewDay -= $days;
1332 # Try in the next month
1337 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1341 * This calculates the Hebrew year start, as days since 1 September.
1342 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1343 * Used for Hebrew date.
1345 private static function hebrewYearStart( $year ) {
1346 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1347 $b = intval( ( $year - 1 ) %
4 );
1348 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1352 $Mar = intval( $m );
1358 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1359 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1361 } else if ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1363 } else if ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1367 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1372 * Algorithm to convert Gregorian dates to Thai solar dates,
1373 * Minguo dates or Minguo dates.
1375 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1376 * http://en.wikipedia.org/wiki/Minguo_calendar
1377 * http://en.wikipedia.org/wiki/Japanese_era_name
1379 * @param $ts String: 14-character timestamp
1380 * @param $cName String: calender name
1381 * @return Array: converted year, month, day
1383 private static function tsToYear( $ts, $cName ) {
1384 $gy = substr( $ts, 0, 4 );
1385 $gm = substr( $ts, 4, 2 );
1386 $gd = substr( $ts, 6, 2 );
1388 if ( !strcmp( $cName, 'thai' ) ) {
1390 # Add 543 years to the Gregorian calendar
1391 # Months and days are identical
1392 $gy_offset = $gy +
543;
1393 } else if ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1395 # Deduct 1911 years from the Gregorian calendar
1396 # Months and days are identical
1397 $gy_offset = $gy - 1911;
1398 } else if ( !strcmp( $cName, 'tenno' ) ) {
1399 # Nengō dates up to Meiji period
1400 # Deduct years from the Gregorian calendar
1401 # depending on the nengo periods
1402 # Months and days are identical
1403 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1405 $gy_gannen = $gy - 1868 +
1;
1406 $gy_offset = $gy_gannen;
1407 if ( $gy_gannen == 1 ) {
1410 $gy_offset = '明治' . $gy_offset;
1412 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1413 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1414 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1415 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1416 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1420 $gy_gannen = $gy - 1912 +
1;
1421 $gy_offset = $gy_gannen;
1422 if ( $gy_gannen == 1 ) {
1425 $gy_offset = '大正' . $gy_offset;
1427 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1428 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1429 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1433 $gy_gannen = $gy - 1926 +
1;
1434 $gy_offset = $gy_gannen;
1435 if ( $gy_gannen == 1 ) {
1438 $gy_offset = '昭和' . $gy_offset;
1441 $gy_gannen = $gy - 1989 +
1;
1442 $gy_offset = $gy_gannen;
1443 if ( $gy_gannen == 1 ) {
1446 $gy_offset = '平成' . $gy_offset;
1452 return array( $gy_offset, $gm, $gd );
1456 * Roman number formatting up to 3000
1458 static function romanNumeral( $num ) {
1459 static $table = array(
1460 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1461 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1462 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1463 array( '', 'M', 'MM', 'MMM' )
1466 $num = intval( $num );
1467 if ( $num > 3000 ||
$num <= 0 ) {
1472 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1473 if ( $num >= $pow10 ) {
1474 $s .= $table[$i][floor( $num / $pow10 )];
1476 $num = $num %
$pow10;
1482 * Hebrew Gematria number formatting up to 9999
1484 static function hebrewNumeral( $num ) {
1485 static $table = array(
1486 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1487 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1488 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1489 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1492 $num = intval( $num );
1493 if ( $num > 9999 ||
$num <= 0 ) {
1498 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1499 if ( $num >= $pow10 ) {
1500 if ( $num == 15 ||
$num == 16 ) {
1501 $s .= $table[0][9] . $table[0][$num - 9];
1504 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1505 if ( $pow10 == 1000 ) {
1510 $num = $num %
$pow10;
1512 if ( strlen( $s ) == 2 ) {
1515 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1516 $str .= substr( $s, strlen( $s ) - 2, 2 );
1518 $start = substr( $str, 0, strlen( $str ) - 2 );
1519 $end = substr( $str, strlen( $str ) - 2 );
1522 $str = $start . 'ך';
1525 $str = $start . 'ם';
1528 $str = $start . 'ן';
1531 $str = $start . 'ף';
1534 $str = $start . 'ץ';
1541 * This is meant to be used by time(), date(), and timeanddate() to get
1542 * the date preference they're supposed to use, it should be used in
1546 * function timeanddate([...], $format = true) {
1547 * $datePreference = $this->dateFormat($format);
1552 * @param $usePrefs Mixed: if true, the user's preference is used
1553 * if false, the site/language default is used
1554 * if int/string, assumed to be a format.
1557 function dateFormat( $usePrefs = true ) {
1560 if ( is_bool( $usePrefs ) ) {
1562 $datePreference = $wgUser->getDatePreference();
1564 $datePreference = (string)User
::getDefaultOption( 'date' );
1567 $datePreference = (string)$usePrefs;
1571 if ( $datePreference == '' ) {
1575 return $datePreference;
1579 * Get a format string for a given type and preference
1580 * @param $type May be date, time or both
1581 * @param $pref The format name as it appears in Messages*.php
1583 function getDateFormatString( $type, $pref ) {
1584 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
1585 if ( $pref == 'default' ) {
1586 $pref = $this->getDefaultDateFormat();
1587 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1589 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1590 if ( is_null( $df ) ) {
1591 $pref = $this->getDefaultDateFormat();
1592 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1595 $this->dateFormatStrings
[$type][$pref] = $df;
1597 return $this->dateFormatStrings
[$type][$pref];
1601 * @param $ts Mixed: the time format which needs to be turned into a
1602 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1603 * @param $adj Bool: whether to adjust the time output according to the
1604 * user configured offset ($timecorrection)
1605 * @param $format Mixed: true to use user's date format preference
1606 * @param $timecorrection String: the time offset as returned by
1607 * validateTimeZone() in Special:Preferences
1610 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1611 $ts = wfTimestamp( TS_MW
, $ts );
1613 $ts = $this->userAdjust( $ts, $timecorrection );
1615 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1616 return $this->sprintfDate( $df, $ts );
1620 * @param $ts Mixed: the time format which needs to be turned into a
1621 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1622 * @param $adj Bool: whether to adjust the time output according to the
1623 * user configured offset ($timecorrection)
1624 * @param $format Mixed: true to use user's date format preference
1625 * @param $timecorrection String: the time offset as returned by
1626 * validateTimeZone() in Special:Preferences
1629 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1630 $ts = wfTimestamp( TS_MW
, $ts );
1632 $ts = $this->userAdjust( $ts, $timecorrection );
1634 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1635 return $this->sprintfDate( $df, $ts );
1639 * @param $ts Mixed: the time format which needs to be turned into a
1640 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1641 * @param $adj Bool: whether to adjust the time output according to the
1642 * user configured offset ($timecorrection)
1643 * @param $format Mixed: what format to return, if it's false output the
1644 * default one (default true)
1645 * @param $timecorrection String: the time offset as returned by
1646 * validateTimeZone() in Special:Preferences
1649 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1650 $ts = wfTimestamp( TS_MW
, $ts );
1652 $ts = $this->userAdjust( $ts, $timecorrection );
1654 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1655 return $this->sprintfDate( $df, $ts );
1658 function getMessage( $key ) {
1659 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
1662 function getAllMessages() {
1663 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
1666 function iconv( $in, $out, $string ) {
1667 # This is a wrapper for iconv in all languages except esperanto,
1668 # which does some nasty x-conversions beforehand
1670 # Even with //IGNORE iconv can whine about illegal characters in
1671 # *input* string. We just ignore those too.
1672 # REF: http://bugs.php.net/bug.php?id=37166
1673 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1674 wfSuppressWarnings();
1675 $text = iconv( $in, $out . '//IGNORE', $string );
1676 wfRestoreWarnings();
1680 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1681 function ucwordbreaksCallbackAscii( $matches ) {
1682 return $this->ucfirst( $matches[1] );
1685 function ucwordbreaksCallbackMB( $matches ) {
1686 return mb_strtoupper( $matches[0] );
1689 function ucCallback( $matches ) {
1690 list( $wikiUpperChars ) = self
::getCaseMaps();
1691 return strtr( $matches[1], $wikiUpperChars );
1694 function lcCallback( $matches ) {
1695 list( , $wikiLowerChars ) = self
::getCaseMaps();
1696 return strtr( $matches[1], $wikiLowerChars );
1699 function ucwordsCallbackMB( $matches ) {
1700 return mb_strtoupper( $matches[0] );
1703 function ucwordsCallbackWiki( $matches ) {
1704 list( $wikiUpperChars ) = self
::getCaseMaps();
1705 return strtr( $matches[0], $wikiUpperChars );
1709 * Make a string's first character uppercase
1711 function ucfirst( $str ) {
1713 if ( $o < 96 ) { // if already uppercase...
1715 } elseif ( $o < 128 ) {
1716 return ucfirst( $str ); // use PHP's ucfirst()
1718 // fall back to more complex logic in case of multibyte strings
1719 return $this->uc( $str, true );
1724 * Convert a string to uppercase
1726 function uc( $str, $first = false ) {
1727 if ( function_exists( 'mb_strtoupper' ) ) {
1729 if ( $this->isMultibyte( $str ) ) {
1730 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1732 return ucfirst( $str );
1735 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
1738 if ( $this->isMultibyte( $str ) ) {
1739 $x = $first ?
'^' : '';
1740 return preg_replace_callback(
1741 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1742 array( $this, 'ucCallback' ),
1746 return $first ?
ucfirst( $str ) : strtoupper( $str );
1751 function lcfirst( $str ) {
1754 return strval( $str );
1755 } elseif ( $o >= 128 ) {
1756 return $this->lc( $str, true );
1757 } elseif ( $o > 96 ) {
1760 $str[0] = strtolower( $str[0] );
1765 function lc( $str, $first = false ) {
1766 if ( function_exists( 'mb_strtolower' ) ) {
1768 if ( $this->isMultibyte( $str ) ) {
1769 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1771 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1774 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
1777 if ( $this->isMultibyte( $str ) ) {
1778 $x = $first ?
'^' : '';
1779 return preg_replace_callback(
1780 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1781 array( $this, 'lcCallback' ),
1785 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1790 function isMultibyte( $str ) {
1791 return (bool)preg_match( '/[\x80-\xff]/', $str );
1794 function ucwords( $str ) {
1795 if ( $this->isMultibyte( $str ) ) {
1796 $str = $this->lc( $str );
1798 // regexp to find first letter in each word (i.e. after each space)
1799 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1801 // function to use to capitalize a single char
1802 if ( function_exists( 'mb_strtoupper' ) ) {
1803 return preg_replace_callback(
1805 array( $this, 'ucwordsCallbackMB' ),
1809 return preg_replace_callback(
1811 array( $this, 'ucwordsCallbackWiki' ),
1816 return ucwords( strtolower( $str ) );
1820 # capitalize words at word breaks
1821 function ucwordbreaks( $str ) {
1822 if ( $this->isMultibyte( $str ) ) {
1823 $str = $this->lc( $str );
1825 // since \b doesn't work for UTF-8, we explicitely define word break chars
1826 $breaks = "[ \-\(\)\}\{\.,\?!]";
1828 // find first letter after word break
1829 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1831 if ( function_exists( 'mb_strtoupper' ) ) {
1832 return preg_replace_callback(
1834 array( $this, 'ucwordbreaksCallbackMB' ),
1838 return preg_replace_callback(
1840 array( $this, 'ucwordsCallbackWiki' ),
1845 return preg_replace_callback(
1846 '/\b([\w\x80-\xff]+)\b/',
1847 array( $this, 'ucwordbreaksCallbackAscii' ),
1854 * Return a case-folded representation of $s
1856 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1857 * and $s2 are the same except for the case of their characters. It is not
1858 * necessary for the value returned to make sense when displayed.
1860 * Do *not* perform any other normalisation in this function. If a caller
1861 * uses this function when it should be using a more general normalisation
1862 * function, then fix the caller.
1864 function caseFold( $s ) {
1865 return $this->uc( $s );
1868 function checkTitleEncoding( $s ) {
1869 if ( is_array( $s ) ) {
1870 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1872 # Check for non-UTF-8 URLs
1873 $ishigh = preg_match( '/[\x80-\xff]/', $s );
1878 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1879 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1884 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
1887 function fallback8bitEncoding() {
1888 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
1892 * Most writing systems use whitespace to break up words.
1893 * Some languages such as Chinese don't conventionally do this,
1894 * which requires special handling when breaking up words for
1897 function hasWordBreaks() {
1902 * Some languages such as Chinese require word segmentation,
1903 * Specify such segmentation when overridden in derived class.
1905 * @param $string String
1908 function segmentByWord( $string ) {
1913 * Some languages have special punctuation need to be normalized.
1914 * Make such changes here.
1916 * @param $string String
1919 function normalizeForSearch( $string ) {
1920 return self
::convertDoubleWidth( $string );
1924 * convert double-width roman characters to single-width.
1925 * range: ff00-ff5f ~= 0020-007f
1927 protected static function convertDoubleWidth( $string ) {
1928 static $full = null;
1929 static $half = null;
1931 if ( $full === null ) {
1932 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1933 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1934 $full = str_split( $fullWidth, 3 );
1935 $half = str_split( $halfWidth );
1938 $string = str_replace( $full, $half, $string );
1942 protected static function insertSpace( $string, $pattern ) {
1943 $string = preg_replace( $pattern, " $1 ", $string );
1944 $string = preg_replace( '/ +/', ' ', $string );
1948 function convertForSearchResult( $termsArray ) {
1949 # some languages, e.g. Chinese, need to do a conversion
1950 # in order for search results to be displayed correctly
1955 * Get the first character of a string.
1960 function firstChar( $s ) {
1963 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1964 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
1969 if ( isset( $matches[1] ) ) {
1970 if ( strlen( $matches[1] ) != 3 ) {
1974 // Break down Hangul syllables to grab the first jamo
1975 $code = utf8ToCodepoint( $matches[1] );
1976 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
1978 } elseif ( $code < 0xb098 ) {
1979 return "\xe3\x84\xb1";
1980 } elseif ( $code < 0xb2e4 ) {
1981 return "\xe3\x84\xb4";
1982 } elseif ( $code < 0xb77c ) {
1983 return "\xe3\x84\xb7";
1984 } elseif ( $code < 0xb9c8 ) {
1985 return "\xe3\x84\xb9";
1986 } elseif ( $code < 0xbc14 ) {
1987 return "\xe3\x85\x81";
1988 } elseif ( $code < 0xc0ac ) {
1989 return "\xe3\x85\x82";
1990 } elseif ( $code < 0xc544 ) {
1991 return "\xe3\x85\x85";
1992 } elseif ( $code < 0xc790 ) {
1993 return "\xe3\x85\x87";
1994 } elseif ( $code < 0xcc28 ) {
1995 return "\xe3\x85\x88";
1996 } elseif ( $code < 0xce74 ) {
1997 return "\xe3\x85\x8a";
1998 } elseif ( $code < 0xd0c0 ) {
1999 return "\xe3\x85\x8b";
2000 } elseif ( $code < 0xd30c ) {
2001 return "\xe3\x85\x8c";
2002 } elseif ( $code < 0xd558 ) {
2003 return "\xe3\x85\x8d";
2005 return "\xe3\x85\x8e";
2012 function initEncoding() {
2013 # Some languages may have an alternate char encoding option
2014 # (Esperanto X-coding, Japanese furigana conversion, etc)
2015 # If this language is used as the primary content language,
2016 # an override to the defaults can be set here on startup.
2019 function recodeForEdit( $s ) {
2020 # For some languages we'll want to explicitly specify
2021 # which characters make it into the edit box raw
2022 # or are converted in some way or another.
2023 # Note that if wgOutputEncoding is different from
2024 # wgInputEncoding, this text will be further converted
2025 # to wgOutputEncoding.
2026 global $wgEditEncoding;
2027 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
2030 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2034 function recodeInput( $s ) {
2035 # Take the previous into account.
2036 global $wgEditEncoding;
2037 if ( $wgEditEncoding != '' ) {
2038 $enc = $wgEditEncoding;
2042 if ( $enc == 'UTF-8' ) {
2045 return $this->iconv( $enc, 'UTF-8', $s );
2050 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2051 * also cleans up certain backwards-compatible sequences, converting them
2052 * to the modern Unicode equivalent.
2054 * This is language-specific for performance reasons only.
2056 function normalize( $s ) {
2057 global $wgAllUnicodeFixes;
2058 $s = UtfNormal
::cleanUp( $s );
2059 if ( $wgAllUnicodeFixes ) {
2060 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2061 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2068 * Transform a string using serialized data stored in the given file (which
2069 * must be in the serialized subdirectory of $IP). The file contains pairs
2070 * mapping source characters to destination characters.
2072 * The data is cached in process memory. This will go faster if you have the
2073 * FastStringSearch extension.
2075 function transformUsingPairFile( $file, $string ) {
2076 if ( !isset( $this->transformData
[$file] ) ) {
2077 $data = wfGetPrecompiledData( $file );
2078 if ( $data === false ) {
2079 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
2081 $this->transformData
[$file] = new ReplacementArray( $data );
2083 return $this->transformData
[$file]->replace( $string );
2087 * For right-to-left language support
2092 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
2096 * Return the correct HTML 'dir' attribute value for this language.
2100 return $this->isRTL() ?
'rtl' : 'ltr';
2104 * Return 'left' or 'right' as appropriate alignment for line-start
2105 * for this language's text direction.
2107 * Should be equivalent to CSS3 'start' text-align value....
2111 function alignStart() {
2112 return $this->isRTL() ?
'right' : 'left';
2116 * Return 'right' or 'left' as appropriate alignment for line-end
2117 * for this language's text direction.
2119 * Should be equivalent to CSS3 'end' text-align value....
2123 function alignEnd() {
2124 return $this->isRTL() ?
'left' : 'right';
2128 * A hidden direction mark (LRM or RLM), depending on the language direction
2132 function getDirMark() {
2133 return $this->isRTL() ?
"\xE2\x80\x8F" : "\xE2\x80\x8E";
2136 function capitalizeAllNouns() {
2137 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2141 * An arrow, depending on the language direction
2145 function getArrow() {
2146 return $this->isRTL() ?
'←' : '→';
2150 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2154 function linkPrefixExtension() {
2155 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2158 function getMagicWords() {
2159 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2162 protected function doMagicHook() {
2163 if ( $this->mMagicHookDone
) {
2166 $this->mMagicHookDone
= true;
2167 wfProfileIn( 'LanguageGetMagic' );
2168 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2169 wfProfileOut( 'LanguageGetMagic' );
2172 # Fill a MagicWord object with data from here
2173 function getMagic( $mw ) {
2174 $this->doMagicHook();
2176 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
2177 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
2179 $magicWords = $this->getMagicWords();
2180 if ( isset( $magicWords[$mw->mId
] ) ) {
2181 $rawEntry = $magicWords[$mw->mId
];
2187 if ( !is_array( $rawEntry ) ) {
2188 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
2190 $mw->mCaseSensitive
= $rawEntry[0];
2191 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
2196 * Add magic words to the extension array
2198 function addMagicWordsByLang( $newWords ) {
2199 $code = $this->getCode();
2200 $fallbackChain = array();
2201 while ( $code && !in_array( $code, $fallbackChain ) ) {
2202 $fallbackChain[] = $code;
2203 $code = self
::getFallbackFor( $code );
2205 if ( !in_array( 'en', $fallbackChain ) ) {
2206 $fallbackChain[] = 'en';
2208 $fallbackChain = array_reverse( $fallbackChain );
2209 foreach ( $fallbackChain as $code ) {
2210 if ( isset( $newWords[$code] ) ) {
2211 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
2217 * Get special page names, as an associative array
2218 * case folded alias => real name
2220 function getSpecialPageAliases() {
2221 // Cache aliases because it may be slow to load them
2222 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
2224 $this->mExtendedSpecialPageAliases
=
2225 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
2226 wfRunHooks( 'LanguageGetSpecialPageAliases',
2227 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
2230 return $this->mExtendedSpecialPageAliases
;
2234 * Italic is unsuitable for some languages
2236 * @param $text String: the text to be emphasized.
2239 function emphasize( $text ) {
2240 return "<em>$text</em>";
2244 * Normally we output all numbers in plain en_US style, that is
2245 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2246 * point twohundredthirtyfive. However this is not sutable for all
2247 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2248 * Icelandic just want to use commas instead of dots, and dots instead
2249 * of commas like "293.291,235".
2251 * An example of this function being called:
2253 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2256 * See LanguageGu.php for the Gujarati implementation and
2257 * $separatorTransformTable on MessageIs.php for
2258 * the , => . and . => , implementation.
2260 * @todo check if it's viable to use localeconv() for the decimal
2262 * @param $number Mixed: the string to be formatted, should be an integer
2263 * or a floating point number.
2264 * @param $nocommafy Bool: set to true for special numbers like dates
2267 function formatNum( $number, $nocommafy = false ) {
2268 global $wgTranslateNumerals;
2269 if ( !$nocommafy ) {
2270 $number = $this->commafy( $number );
2271 $s = $this->separatorTransformTable();
2273 $number = strtr( $number, $s );
2277 if ( $wgTranslateNumerals ) {
2278 $s = $this->digitTransformTable();
2280 $number = strtr( $number, $s );
2287 function parseFormattedNumber( $number ) {
2288 $s = $this->digitTransformTable();
2290 $number = strtr( $number, array_flip( $s ) );
2293 $s = $this->separatorTransformTable();
2295 $number = strtr( $number, array_flip( $s ) );
2298 $number = strtr( $number, array( ',' => '' ) );
2303 * Adds commas to a given number
2308 function commafy( $_ ) {
2309 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2312 function digitTransformTable() {
2313 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
2316 function separatorTransformTable() {
2317 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
2321 * Take a list of strings and build a locale-friendly comma-separated
2322 * list, using the local comma-separator message.
2323 * The last two strings are chained with an "and".
2328 function listToText( $l ) {
2330 $m = count( $l ) - 1;
2332 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2334 for ( $i = $m; $i >= 0; $i-- ) {
2337 } else if ( $i == $m - 1 ) {
2338 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2340 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2348 * Take a list of strings and build a locale-friendly comma-separated
2349 * list, using the local comma-separator message.
2350 * @param $list array of strings to put in a comma list
2353 function commaList( $list ) {
2358 array( 'parsemag', 'escapenoentities', 'language' => $this )
2364 * Take a list of strings and build a locale-friendly semicolon-separated
2365 * list, using the local semicolon-separator message.
2366 * @param $list array of strings to put in a semicolon list
2369 function semicolonList( $list ) {
2373 'semicolon-separator',
2374 array( 'parsemag', 'escapenoentities', 'language' => $this )
2380 * Same as commaList, but separate it with the pipe instead.
2381 * @param $list array of strings to put in a pipe list
2384 function pipeList( $list ) {
2389 array( 'escapenoentities', 'language' => $this )
2395 * Truncate a string to a specified length in bytes, appending an optional
2396 * string (e.g. for ellipses)
2398 * The database offers limited byte lengths for some columns in the database;
2399 * multi-byte character sets mean we need to ensure that only whole characters
2400 * are included, otherwise broken characters can be passed to the user
2402 * If $length is negative, the string will be truncated from the beginning
2404 * @param $string String to truncate
2405 * @param $length Int: maximum length (including ellipses)
2406 * @param $ellipsis String to append to the truncated text
2407 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
2408 * $adjustLength was introduced in 1.18, before that behaved as if false.
2411 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
2412 # Use the localized ellipsis character
2413 if ( $ellipsis == '...' ) {
2414 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2416 # Check if there is no need to truncate
2417 if ( $length == 0 ) {
2418 return $ellipsis; // convention
2419 } elseif ( strlen( $string ) <= abs( $length ) ) {
2420 return $string; // no need to truncate
2422 $stringOriginal = $string;
2423 # If ellipsis length is >= $length then we can't apply $adjustLength
2424 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
2425 $string = $ellipsis; // this can be slightly unexpected
2426 # Otherwise, truncate and add ellipsis...
2428 $eLength = $adjustLength ?
strlen( $ellipsis ) : 0;
2429 if ( $length > 0 ) {
2430 $length -= $eLength;
2431 $string = substr( $string, 0, $length ); // xyz...
2432 $string = $this->removeBadCharLast( $string );
2433 $string = $string . $ellipsis;
2435 $length +
= $eLength;
2436 $string = substr( $string, $length ); // ...xyz
2437 $string = $this->removeBadCharFirst( $string );
2438 $string = $ellipsis . $string;
2441 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
2442 # This check is *not* redundant if $adjustLength, due to the single case where
2443 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
2444 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
2447 return $stringOriginal;
2452 * Remove bytes that represent an incomplete Unicode character
2453 * at the end of string (e.g. bytes of the char are missing)
2455 * @param $string String
2458 protected function removeBadCharLast( $string ) {
2459 if ( $string != '' ) {
2460 $char = ord( $string[strlen( $string ) - 1] );
2462 if ( $char >= 0xc0 ) {
2463 # We got the first byte only of a multibyte char; remove it.
2464 $string = substr( $string, 0, -1 );
2465 } elseif ( $char >= 0x80 &&
2466 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2467 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
2469 # We chopped in the middle of a character; remove it
2477 * Remove bytes that represent an incomplete Unicode character
2478 * at the start of string (e.g. bytes of the char are missing)
2480 * @param $string String
2483 protected function removeBadCharFirst( $string ) {
2484 if ( $string != '' ) {
2485 $char = ord( $string[0] );
2486 if ( $char >= 0x80 && $char < 0xc0 ) {
2487 # We chopped in the middle of a character; remove the whole thing
2488 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2495 * Truncate a string of valid HTML to a specified length in bytes,
2496 * appending an optional string (e.g. for ellipses), and return valid HTML
2498 * This is only intended for styled/linked text, such as HTML with
2499 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
2500 * Also, this will not detect things like "display:none" CSS.
2502 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
2504 * @param string $text HTML string to truncate
2505 * @param int $length (zero/positive) Maximum length (including ellipses)
2506 * @param string $ellipsis String to append to the truncated text
2509 function truncateHtml( $text, $length, $ellipsis = '...' ) {
2510 # Use the localized ellipsis character
2511 if ( $ellipsis == '...' ) {
2512 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2514 # Check if there is clearly no need to truncate
2515 if ( $length <= 0 ) {
2516 return $ellipsis; // no text shown, nothing to format (convention)
2517 } elseif ( strlen( $text ) <= $length ) {
2518 return $text; // string short enough even *with* HTML (short-circuit)
2521 $displayLen = 0; // innerHTML legth so far
2522 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
2523 $tagType = 0; // 0-open, 1-close
2524 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
2525 $entityState = 0; // 0-not entity, 1-entity
2526 $tag = $ret = $pRet = ''; // accumulated tag name, accumulated result string
2527 $openTags = array(); // open tag stack
2528 $pOpenTags = array();
2530 $textLen = strlen( $text );
2531 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
2532 for ( $pos = 0; true; ++
$pos ) {
2533 # Consider truncation once the display length has reached the maximim.
2534 # Check that we're not in the middle of a bracket/entity...
2535 if ( $displayLen >= $neLength && $bracketState == 0 && $entityState == 0 ) {
2536 if ( !$testingEllipsis ) {
2537 $testingEllipsis = true;
2538 # Save where we are; we will truncate here unless there turn out to
2539 # be so few remaining characters that truncation is not necessary.
2540 $pOpenTags = $openTags; // save state
2541 $pRet = $ret; // save state
2542 } elseif ( $displayLen > $length && $displayLen > strlen( $ellipsis ) ) {
2543 # String in fact does need truncation, the truncation point was OK.
2544 $openTags = $pOpenTags; // reload state
2545 $ret = $this->removeBadCharLast( $pRet ); // reload state, multi-byte char fix
2546 $ret .= $ellipsis; // add ellipsis
2550 if ( $pos >= $textLen ) break; // extra iteration just for above checks
2552 # Read the next char...
2554 $lastCh = $pos ?
$text[$pos - 1] : '';
2555 $ret .= $ch; // add to result string
2557 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
2558 $entityState = 0; // for bad HTML
2559 $bracketState = 1; // tag started (checking for backslash)
2560 } elseif ( $ch == '>' ) {
2561 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
2562 $entityState = 0; // for bad HTML
2563 $bracketState = 0; // out of brackets
2564 } elseif ( $bracketState == 1 ) {
2566 $tagType = 1; // close tag (e.g. "</span>")
2568 $tagType = 0; // open tag (e.g. "<span>")
2571 $bracketState = 2; // building tag name
2572 } elseif ( $bracketState == 2 ) {
2576 // Name found (e.g. "<a href=..."), add on tag attributes...
2577 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
2579 } elseif ( $bracketState == 0 ) {
2580 if ( $entityState ) {
2583 $displayLen++
; // entity is one displayed char
2587 $entityState = 1; // entity found, (e.g. " ")
2589 $displayLen++
; // this char is displayed
2590 // Add the next $max display text chars after this in one swoop...
2591 $max = ( $testingEllipsis ?
$length : $neLength ) - $displayLen;
2592 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos +
1, $max );
2593 $displayLen +
= $skipped;
2599 if ( $displayLen == 0 ) {
2600 return ''; // no text shown, nothing to format
2602 // Close the last tag if left unclosed by bad HTML
2603 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
2604 while ( count( $openTags ) > 0 ) {
2605 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
2610 // truncateHtml() helper function
2611 // like strcspn() but adds the skipped chars to $ret
2612 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
2613 if ( $len === null ) {
2614 $len = -1; // -1 means "no limit" for strcspn
2615 } elseif ( $len < 0 ) {
2619 if ( $start < strlen( $text ) ) {
2620 $skipCount = strcspn( $text, $search, $start, $len );
2621 $ret .= substr( $text, $start, $skipCount );
2627 * truncateHtml() helper function
2628 * (a) push or pop $tag from $openTags as needed
2629 * (b) clear $tag value
2630 * @param String &$tag Current HTML tag name we are looking at
2631 * @param int $tagType (0-open tag, 1-close tag)
2632 * @param char $lastCh Character before the '>' that ended this tag
2633 * @param array &$openTags Open tag stack (not accounting for $tag)
2635 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
2636 $tag = ltrim( $tag );
2638 if ( $tagType == 0 && $lastCh != '/' ) {
2639 $openTags[] = $tag; // tag opened (didn't close itself)
2640 } else if ( $tagType == 1 ) {
2641 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
2642 array_pop( $openTags ); // tag closed
2650 * Grammatical transformations, needed for inflected languages
2651 * Invoked by putting {{grammar:case|word}} in a message
2653 * @param $word string
2654 * @param $case string
2657 function convertGrammar( $word, $case ) {
2658 global $wgGrammarForms;
2659 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
2660 return $wgGrammarForms[$this->getCode()][$case][$word];
2666 * Provides an alternative text depending on specified gender.
2667 * Usage {{gender:username|masculine|feminine|neutral}}.
2668 * username is optional, in which case the gender of current user is used,
2669 * but only in (some) interface messages; otherwise default gender is used.
2670 * If second or third parameter are not specified, masculine is used.
2671 * These details may be overriden per language.
2673 function gender( $gender, $forms ) {
2674 if ( !count( $forms ) ) {
2677 $forms = $this->preConvertPlural( $forms, 2 );
2678 if ( $gender === 'male' ) {
2681 if ( $gender === 'female' ) {
2684 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
2688 * Plural form transformations, needed for some languages.
2689 * For example, there are 3 form of plural in Russian and Polish,
2690 * depending on "count mod 10". See [[w:Plural]]
2691 * For English it is pretty simple.
2693 * Invoked by putting {{plural:count|wordform1|wordform2}}
2694 * or {{plural:count|wordform1|wordform2|wordform3}}
2696 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2698 * @param $count Integer: non-localized number
2699 * @param $forms Array: different plural forms
2700 * @return string Correct form of plural for $count in this language
2702 function convertPlural( $count, $forms ) {
2703 if ( !count( $forms ) ) {
2706 $forms = $this->preConvertPlural( $forms, 2 );
2708 return ( $count == 1 ) ?
$forms[0] : $forms[1];
2712 * Checks that convertPlural was given an array and pads it to requested
2713 * amound of forms by copying the last one.
2715 * @param $count Integer: How many forms should there be at least
2716 * @param $forms Array of forms given to convertPlural
2717 * @return array Padded array of forms or an exception if not an array
2719 protected function preConvertPlural( /* Array */ $forms, $count ) {
2720 while ( count( $forms ) < $count ) {
2721 $forms[] = $forms[count( $forms ) - 1];
2727 * Maybe translate block durations. Note that this function is somewhat misnamed: it
2728 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
2729 * (which is an absolute timestamp).
2730 * @param $str String: the validated block duration in English
2731 * @return Somehow translated block duration
2732 * @see LanguageFi.php for example implementation
2734 function translateBlockExpiry( $str ) {
2735 foreach( SpecialBlock
::getSuggestedDurations( $this ) as $show => $value ){
2736 if ( strcmp( $str, $value ) == 0 ) {
2737 return htmlspecialchars( trim( $show ) );
2744 * languages like Chinese need to be segmented in order for the diff
2747 * @param $text String
2750 function segmentForDiff( $text ) {
2755 * and unsegment to show the result
2757 * @param $text String
2760 function unsegmentForDiff( $text ) {
2764 # convert text to all supported variants
2765 function autoConvertToAllVariants( $text ) {
2766 return $this->mConverter
->autoConvertToAllVariants( $text );
2769 # convert text to different variants of a language.
2770 function convert( $text ) {
2771 return $this->mConverter
->convert( $text );
2774 # Convert a Title object to a string in the preferred variant
2775 function convertTitle( $title ) {
2776 return $this->mConverter
->convertTitle( $title );
2779 # Check if this is a language with variants
2780 function hasVariants() {
2781 return sizeof( $this->getVariants() ) > 1;
2784 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2785 function armourMath( $text ) {
2786 return $this->mConverter
->armourMath( $text );
2790 * Perform output conversion on a string, and encode for safe HTML output.
2791 * @param $text String text to be converted
2792 * @param $isTitle Bool whether this conversion is for the article title
2794 * @todo this should get integrated somewhere sane
2796 function convertHtml( $text, $isTitle = false ) {
2797 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2800 function convertCategoryKey( $key ) {
2801 return $this->mConverter
->convertCategoryKey( $key );
2805 * Get the list of variants supported by this language
2806 * see sample implementation in LanguageZh.php
2808 * @return array an array of language codes
2810 function getVariants() {
2811 return $this->mConverter
->getVariants();
2814 function getPreferredVariant() {
2815 return $this->mConverter
->getPreferredVariant();
2818 function getDefaultVariant() {
2819 return $this->mConverter
->getDefaultVariant();
2822 function getURLVariant() {
2823 return $this->mConverter
->getURLVariant();
2827 * If a language supports multiple variants, it is
2828 * possible that non-existing link in one variant
2829 * actually exists in another variant. this function
2830 * tries to find it. See e.g. LanguageZh.php
2832 * @param $link String: the name of the link
2833 * @param $nt Mixed: the title object of the link
2834 * @param $ignoreOtherCond Boolean: to disable other conditions when
2835 * we need to transclude a template or update a category's link
2836 * @return null the input parameters may be modified upon return
2838 function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
2839 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
2843 * If a language supports multiple variants, converts text
2844 * into an array of all possible variants of the text:
2845 * 'variant' => text in that variant
2847 * @deprecated Use autoConvertToAllVariants()
2849 function convertLinkToAllVariants( $text ) {
2850 return $this->mConverter
->convertLinkToAllVariants( $text );
2854 * returns language specific options used by User::getPageRenderHash()
2855 * for example, the preferred language variant
2859 function getExtraHashOptions() {
2860 return $this->mConverter
->getExtraHashOptions();
2864 * For languages that support multiple variants, the title of an
2865 * article may be displayed differently in different variants. this
2866 * function returns the apporiate title defined in the body of the article.
2870 function getParsedTitle() {
2871 return $this->mConverter
->getParsedTitle();
2875 * Enclose a string with the "no conversion" tag. This is used by
2876 * various functions in the Parser
2878 * @param $text String: text to be tagged for no conversion
2880 * @return string the tagged text
2882 function markNoConversion( $text, $noParse = false ) {
2883 return $this->mConverter
->markNoConversion( $text, $noParse );
2887 * A regular expression to match legal word-trailing characters
2888 * which should be merged onto a link of the form [[foo]]bar.
2892 function linkTrail() {
2893 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
2896 function getLangObj() {
2901 * Get the RFC 3066 code for this language object
2903 function getCode() {
2904 return $this->mCode
;
2907 function setCode( $code ) {
2908 $this->mCode
= $code;
2912 * Get the name of a file for a certain language code
2913 * @param $prefix string Prepend this to the filename
2914 * @param $code string Language code
2915 * @param $suffix string Append this to the filename
2916 * @return string $prefix . $mangledCode . $suffix
2918 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2919 // Protect against path traversal
2920 if ( !Language
::isValidCode( $code )
2921 ||
strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
2923 throw new MWException( "Invalid language code \"$code\"" );
2926 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2930 * Get the language code from a file name. Inverse of getFileName()
2931 * @param $filename string $prefix . $languageCode . $suffix
2932 * @param $prefix string Prefix before the language code
2933 * @param $suffix string Suffix after the language code
2934 * @return Language code, or false if $prefix or $suffix isn't found
2936 static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
2938 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
2939 preg_quote( $suffix, '/' ) . '/', $filename, $m );
2940 if ( !count( $m ) ) {
2943 return str_replace( '_', '-', strtolower( $m[1] ) );
2946 static function getMessagesFileName( $code ) {
2948 return self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2951 static function getClassFileName( $code ) {
2953 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2957 * Get the fallback for a given language
2959 static function getFallbackFor( $code ) {
2960 if ( $code === 'en' ) {
2964 return self
::getLocalisationCache()->getItem( $code, 'fallback' );
2969 * Get all messages for a given language
2970 * WARNING: this may take a long time
2972 static function getMessagesFor( $code ) {
2973 return self
::getLocalisationCache()->getItem( $code, 'messages' );
2977 * Get a message for a given language
2979 static function getMessageFor( $key, $code ) {
2980 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
2983 function fixVariableInNamespace( $talk ) {
2984 if ( strpos( $talk, '$1' ) === false ) {
2988 global $wgMetaNamespace;
2989 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2991 # Allow grammar transformations
2992 # Allowing full message-style parsing would make simple requests
2993 # such as action=raw much more expensive than they need to be.
2994 # This will hopefully cover most cases.
2995 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2996 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2997 return str_replace( ' ', '_', $talk );
3000 function replaceGrammarInNamespace( $m ) {
3001 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
3004 static function getCaseMaps() {
3005 static $wikiUpperChars, $wikiLowerChars;
3006 if ( isset( $wikiUpperChars ) ) {
3007 return array( $wikiUpperChars, $wikiLowerChars );
3010 wfProfileIn( __METHOD__
);
3011 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
3012 if ( $arr === false ) {
3013 throw new MWException(
3014 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
3016 $wikiUpperChars = $arr['wikiUpperChars'];
3017 $wikiLowerChars = $arr['wikiLowerChars'];
3018 wfProfileOut( __METHOD__
);
3019 return array( $wikiUpperChars, $wikiLowerChars );
3023 * Decode an expiry (block, protection, etc) which has come from the DB
3025 * @param $expiry String: Database expiry String
3026 * @param $format Bool|Int true to process using language functions, or TS_ constant
3027 * to return the expiry in a given timestamp
3030 public function formatExpiry( $expiry, $format = true ) {
3031 static $infinity, $infinityMsg;
3032 if( $infinity === null ){
3033 $infinityMsg = wfMessage( 'infiniteblock' );
3034 $infinity = wfGetDB( DB_SLAVE
)->getInfinity();
3037 if ( $expiry == '' ||
$expiry == $infinity ) {
3038 return $format === true
3042 return $format === true
3043 ?
$this->timeanddate( $expiry )
3044 : wfTimestamp( $format, $expiry );
3050 * @param $seconds String
3053 function formatTimePeriod( $seconds ) {
3054 if ( round( $seconds * 10 ) < 100 ) {
3055 return $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
3056 } elseif ( round( $seconds ) < 60 ) {
3057 return $this->formatNum( round( $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
3058 } elseif ( round( $seconds ) < 3600 ) {
3059 $minutes = floor( $seconds / 60 );
3060 $secondsPart = round( fmod( $seconds, 60 ) );
3061 if ( $secondsPart == 60 ) {
3065 return $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
3066 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
3068 $hours = floor( $seconds / 3600 );
3069 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
3070 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
3071 if ( $secondsPart == 60 ) {
3075 if ( $minutes == 60 ) {
3079 return $this->formatNum( $hours ) . $this->getMessageFromDB( 'hours-abbrev' ) . ' ' .
3080 $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
3081 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
3085 function formatBitrate( $bps ) {
3086 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
3088 return $this->formatNum( $bps ) . $units[0];
3090 $unitIndex = floor( log10( $bps ) / 3 );
3091 $mantissa = $bps / pow( 1000, $unitIndex );
3092 if ( $mantissa < 10 ) {
3093 $mantissa = round( $mantissa, 1 );
3095 $mantissa = round( $mantissa );
3097 return $this->formatNum( $mantissa ) . $units[$unitIndex];
3101 * Format a size in bytes for output, using an appropriate
3102 * unit (B, KB, MB or GB) according to the magnitude in question
3104 * @param $size Size to format
3105 * @return string Plain text (not HTML)
3107 function formatSize( $size ) {
3108 // For small sizes no decimal places necessary
3110 if ( $size > 1024 ) {
3111 $size = $size / 1024;
3112 if ( $size > 1024 ) {
3113 $size = $size / 1024;
3114 // For MB and bigger two decimal places are smarter
3116 if ( $size > 1024 ) {
3117 $size = $size / 1024;
3118 $msg = 'size-gigabytes';
3120 $msg = 'size-megabytes';
3123 $msg = 'size-kilobytes';
3126 $msg = 'size-bytes';
3128 $size = round( $size, $round );
3129 $text = $this->getMessageFromDB( $msg );
3130 return str_replace( '$1', $this->formatNum( $size ), $text );
3134 * Get the conversion rule title, if any.
3136 function getConvRuleTitle() {
3137 return $this->mConverter
->getConvRuleTitle();