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 $this->mLang
->getCode(); }
47 function getConvRuleTitle() { return false; }
48 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
49 function getExtraHashOptions() { return ''; }
50 function getParsedTitle() { return ''; }
51 function markNoConversion( $text, $noParse = false ) { return $text; }
52 function convertCategoryKey( $key ) { return $key; }
53 function convertLinkToAllVariants( $text ) { return autoConvertToAllVariants( $text ); }
54 function armourMath( $text ) { return $text; }
58 * Internationalisation code
62 var $mConverter, $mVariants, $mCode, $mLoaded = false;
63 var $mMagicExtensions = array(), $mMagicHookDone = false;
65 var $mNamespaceIds, $namespaceNames, $namespaceAliases;
66 var $dateFormatStrings = array();
67 var $mExtendedSpecialPageAliases;
70 * ReplacementArray object caches
72 var $transformData = array();
74 static public $dataCache;
75 static public $mLangObjCache = array();
77 static public $mWeekdayMsgs = array(
78 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
82 static public $mWeekdayAbbrevMsgs = array(
83 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
86 static public $mMonthMsgs = array(
87 'january', 'february', 'march', 'april', 'may_long', 'june',
88 'july', 'august', 'september', 'october', 'november',
91 static public $mMonthGenMsgs = array(
92 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
93 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
96 static public $mMonthAbbrevMsgs = array(
97 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
98 'sep', 'oct', 'nov', 'dec'
101 static public $mIranianCalendarMonthMsgs = array(
102 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
103 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
104 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
105 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
108 static public $mHebrewCalendarMonthMsgs = array(
109 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
110 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
111 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
112 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
113 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
116 static public $mHebrewCalendarMonthGenMsgs = array(
117 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
118 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
119 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
120 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
121 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
124 static public $mHijriCalendarMonthMsgs = array(
125 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
126 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
127 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
128 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
132 * Get a cached language object for a given language code
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
148 protected static function newFromCode( $code ) {
150 static $recursionLevel = 0;
151 if ( $code == 'en' ) {
154 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
155 // Preload base classes to work around APC/PHP5 bug
156 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
157 include_once( "$IP/languages/classes/$class.deps.php" );
159 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
160 include_once( "$IP/languages/classes/$class.php" );
164 if ( $recursionLevel > 5 ) {
165 throw new MWException( "Language fallback loop detected when creating class $class\n" );
168 if ( !class_exists( $class ) ) {
169 $fallback = Language
::getFallbackFor( $code );
171 $lang = Language
::newFromCode( $fallback );
173 $lang->setCode( $code );
181 * Get the LocalisationCache instance
183 public static function getLocalisationCache() {
184 if ( is_null( self
::$dataCache ) ) {
185 global $wgLocalisationCacheConf;
186 $class = $wgLocalisationCacheConf['class'];
187 self
::$dataCache = new $class( $wgLocalisationCacheConf );
189 return self
::$dataCache;
192 function __construct() {
193 $this->mConverter
= new FakeConverter( $this );
194 // Set the code to the name of the descendant
195 if ( get_class( $this ) == 'Language' ) {
198 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
200 self
::getLocalisationCache();
204 * Reduce memory usage
206 function __destruct() {
207 foreach ( $this as $name => $value ) {
208 unset( $this->$name );
213 * Hook which will be called if this is the content language.
214 * Descendants can use this to register hook functions or modify globals
216 function initContLang() { }
219 * @deprecated Use User::getDefaultOptions()
222 function getDefaultUserOptions() {
223 wfDeprecated( __METHOD__
);
224 return User
::getDefaultOptions();
227 function getFallbackLanguageCode() {
228 if ( $this->mCode
=== 'en' ) {
231 return self
::$dataCache->getItem( $this->mCode
, 'fallback' );
236 * Exports $wgBookstoreListEn
239 function getBookstoreList() {
240 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
246 function getNamespaces() {
247 if ( is_null( $this->namespaceNames
) ) {
248 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
250 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
251 $validNamespaces = MWNamespace
::getCanonicalNamespaces();
253 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames +
$validNamespaces;
255 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
256 if ( $wgMetaNamespaceTalk ) {
257 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
259 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
260 $this->namespaceNames
[NS_PROJECT_TALK
] =
261 $this->fixVariableInNamespace( $talk );
264 # Sometimes a language will be localised but not actually exist on this wiki.
265 foreach( $this->namespaceNames
as $key => $text ) {
266 if ( !isset( $validNamespaces[$key] ) ) {
267 unset( $this->namespaceNames
[$key] );
271 # The above mixing may leave namespaces out of canonical order.
272 # Re-order by namespace ID number...
273 ksort( $this->namespaceNames
);
275 return $this->namespaceNames
;
279 * A convenience function that returns the same thing as
280 * getNamespaces() except with the array values changed to ' '
281 * where it found '_', useful for producing output to be displayed
282 * e.g. in <select> forms.
286 function getFormattedNamespaces() {
287 $ns = $this->getNamespaces();
288 foreach ( $ns as $k => $v ) {
289 $ns[$k] = strtr( $v, '_', ' ' );
295 * Get a namespace value by key
297 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
298 * echo $mw_ns; // prints 'MediaWiki'
301 * @param $index Int: the array key of the namespace to return
302 * @return mixed, string if the namespace value exists, otherwise false
304 function getNsText( $index ) {
305 $ns = $this->getNamespaces();
306 return isset( $ns[$index] ) ?
$ns[$index] : false;
310 * A convenience function that returns the same thing as
311 * getNsText() except with '_' changed to ' ', useful for
316 function getFormattedNsText( $index ) {
317 $ns = $this->getNsText( $index );
318 return strtr( $ns, '_', ' ' );
322 * Get a namespace key by value, case insensitive.
323 * Only matches namespace names for the current language, not the
324 * canonical ones defined in Namespace.php.
326 * @param $text String
327 * @return mixed An integer if $text is a valid value otherwise false
329 function getLocalNsIndex( $text ) {
330 $lctext = $this->lc( $text );
331 $ids = $this->getNamespaceIds();
332 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
335 function getNamespaceAliases() {
336 if ( is_null( $this->namespaceAliases
) ) {
337 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
341 foreach ( $aliases as $name => $index ) {
342 if ( $index === NS_PROJECT_TALK
) {
343 unset( $aliases[$name] );
344 $name = $this->fixVariableInNamespace( $name );
345 $aliases[$name] = $index;
349 $this->namespaceAliases
= $aliases;
351 return $this->namespaceAliases
;
354 function getNamespaceIds() {
355 if ( is_null( $this->mNamespaceIds
) ) {
356 global $wgNamespaceAliases;
357 # Put namespace names and aliases into a hashtable.
358 # If this is too slow, then we should arrange it so that it is done
359 # before caching. The catch is that at pre-cache time, the above
360 # class-specific fixup hasn't been done.
361 $this->mNamespaceIds
= array();
362 foreach ( $this->getNamespaces() as $index => $name ) {
363 $this->mNamespaceIds
[$this->lc( $name )] = $index;
365 foreach ( $this->getNamespaceAliases() as $name => $index ) {
366 $this->mNamespaceIds
[$this->lc( $name )] = $index;
368 if ( $wgNamespaceAliases ) {
369 foreach ( $wgNamespaceAliases as $name => $index ) {
370 $this->mNamespaceIds
[$this->lc( $name )] = $index;
374 return $this->mNamespaceIds
;
379 * Get a namespace key by value, case insensitive. Canonical namespace
380 * names override custom ones defined for the current language.
382 * @param $text String
383 * @return mixed An integer if $text is a valid value otherwise false
385 function getNsIndex( $text ) {
386 $lctext = $this->lc( $text );
387 if ( ( $ns = MWNamespace
::getCanonicalIndex( $lctext ) ) !== null ) {
390 $ids = $this->getNamespaceIds();
391 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
395 * short names for language variants used for language conversion links.
397 * @param $code String
400 function getVariantname( $code ) {
401 return $this->getMessageFromDB( "variantname-$code" );
404 function specialPage( $name ) {
405 $aliases = $this->getSpecialPageAliases();
406 if ( isset( $aliases[$name][0] ) ) {
407 $name = $aliases[$name][0];
409 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
412 function getQuickbarSettings() {
414 $this->getMessage( 'qbsettings-none' ),
415 $this->getMessage( 'qbsettings-fixedleft' ),
416 $this->getMessage( 'qbsettings-fixedright' ),
417 $this->getMessage( 'qbsettings-floatingleft' ),
418 $this->getMessage( 'qbsettings-floatingright' )
422 function getMathNames() {
423 return self
::$dataCache->getItem( $this->mCode
, 'mathNames' );
426 function getDatePreferences() {
427 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
430 function getDateFormats() {
431 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
434 function getDefaultDateFormat() {
435 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
436 if ( $df === 'dmy or mdy' ) {
437 global $wgAmericanDates;
438 return $wgAmericanDates ?
'mdy' : 'dmy';
444 function getDatePreferenceMigrationMap() {
445 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
448 function getImageFile( $image ) {
449 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
452 function getDefaultUserOptionOverrides() {
453 return self
::$dataCache->getItem( $this->mCode
, 'defaultUserOptionOverrides' );
456 function getExtraUserToggles() {
457 return self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
460 function getUserToggle( $tog ) {
461 return $this->getMessageFromDB( "tog-$tog" );
465 * Get language names, indexed by code.
466 * If $customisedOnly is true, only returns codes with a messages file
468 public static function getLanguageNames( $customisedOnly = false ) {
469 global $wgLanguageNames, $wgExtraLanguageNames;
470 $allNames = $wgExtraLanguageNames +
$wgLanguageNames;
471 if ( !$customisedOnly ) {
477 $dir = opendir( "$IP/languages/messages" );
478 while ( false !== ( $file = readdir( $dir ) ) ) {
479 $code = self
::getCodeFromFileName( $file, 'Messages' );
480 if ( $code && isset( $allNames[$code] ) ) {
481 $names[$code] = $allNames[$code];
489 * Get a message from the MediaWiki namespace.
491 * @param $msg String: message name
494 function getMessageFromDB( $msg ) {
495 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
498 function getLanguageName( $code ) {
499 $names = self
::getLanguageNames();
500 if ( !array_key_exists( $code, $names ) ) {
503 return $names[$code];
506 function getMonthName( $key ) {
507 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
510 function getMonthNameGen( $key ) {
511 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
514 function getMonthAbbreviation( $key ) {
515 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
518 function getWeekdayName( $key ) {
519 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
522 function getWeekdayAbbreviation( $key ) {
523 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
526 function getIranianCalendarMonthName( $key ) {
527 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
530 function getHebrewCalendarMonthName( $key ) {
531 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
534 function getHebrewCalendarMonthNameGen( $key ) {
535 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
538 function getHijriCalendarMonthName( $key ) {
539 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
543 * Used by date() and time() to adjust the time output.
545 * @param $ts Int the time in date('YmdHis') format
546 * @param $tz Mixed: adjust the time by this amount (default false, mean we
547 * get user timecorrection setting)
550 function userAdjust( $ts, $tz = false ) {
551 global $wgUser, $wgLocalTZoffset;
553 if ( $tz === false ) {
554 $tz = $wgUser->getOption( 'timecorrection' );
557 $data = explode( '|', $tz, 3 );
559 if ( $data[0] == 'ZoneInfo' ) {
560 if ( function_exists( 'timezone_open' ) && @timezone_open
( $data[2] ) !== false ) {
561 $date = date_create( $ts, timezone_open( 'UTC' ) );
562 date_timezone_set( $date, timezone_open( $data[2] ) );
563 $date = date_format( $date, 'YmdHis' );
566 # Unrecognized timezone, default to 'Offset' with the stored offset.
571 if ( $data[0] == 'System' ||
$tz == '' ) {
572 # Global offset in minutes.
573 if ( isset( $wgLocalTZoffset ) ) {
574 $minDiff = $wgLocalTZoffset;
576 } else if ( $data[0] == 'Offset' ) {
577 $minDiff = intval( $data[1] );
579 $data = explode( ':', $tz );
580 if ( count( $data ) == 2 ) {
581 $data[0] = intval( $data[0] );
582 $data[1] = intval( $data[1] );
583 $minDiff = abs( $data[0] ) * 60 +
$data[1];
584 if ( $data[0] < 0 ) {
585 $minDiff = -$minDiff;
588 $minDiff = intval( $data[0] ) * 60;
592 # No difference ? Return time unchanged
593 if ( 0 == $minDiff ) {
597 wfSuppressWarnings(); // E_STRICT system time bitching
598 # Generate an adjusted date; take advantage of the fact that mktime
599 # will normalize out-of-range values so we don't have to split $minDiff
600 # into hours and minutes.
602 (int)substr( $ts, 8, 2 ) ), # Hours
603 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
604 (int)substr( $ts, 12, 2 ), # Seconds
605 (int)substr( $ts, 4, 2 ), # Month
606 (int)substr( $ts, 6, 2 ), # Day
607 (int)substr( $ts, 0, 4 ) ); # Year
609 $date = date( 'YmdHis', $t );
616 * This is a workalike of PHP's date() function, but with better
617 * internationalisation, a reduced set of format characters, and a better
620 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
621 * PHP manual for definitions. There are a number of extensions, which
624 * xn Do not translate digits of the next numeric format character
625 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
626 * xr Use roman numerals for the next numeric format character
627 * xh Use hebrew numerals for the next numeric format character
629 * xg Genitive month name
631 * xij j (day number) in Iranian calendar
632 * xiF F (month name) in Iranian calendar
633 * xin n (month number) in Iranian calendar
634 * xiY Y (full year) in Iranian calendar
636 * xjj j (day number) in Hebrew calendar
637 * xjF F (month name) in Hebrew calendar
638 * xjt t (days in month) in Hebrew calendar
639 * xjx xg (genitive month name) in Hebrew calendar
640 * xjn n (month number) in Hebrew calendar
641 * xjY Y (full year) in Hebrew calendar
643 * xmj j (day number) in Hijri calendar
644 * xmF F (month name) in Hijri calendar
645 * xmn n (month number) in Hijri calendar
646 * xmY Y (full year) in Hijri calendar
648 * xkY Y (full year) in Thai solar calendar. Months and days are
649 * identical to the Gregorian calendar
650 * xoY Y (full year) in Minguo calendar or Juche year.
651 * Months and days are identical to the
653 * xtY Y (full year) in Japanese nengo. Months and days are
654 * identical to the Gregorian calendar
656 * Characters enclosed in double quotes will be considered literal (with
657 * the quotes themselves removed). Unmatched quotes will be considered
658 * literal quotes. Example:
660 * "The month is" F => The month is January
663 * Backslash escaping is also supported.
665 * Input timestamp is assumed to be pre-normalized to the desired local
668 * @param $format String
669 * @param $ts String: 14-character timestamp
672 * @todo emulation of "o" format character for PHP pre 5.1.0
673 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
675 function sprintfDate( $format, $ts ) {
688 for ( $p = 0; $p < strlen( $format ); $p++
) {
691 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
692 $code .= $format[++
$p];
695 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
696 $code .= $format[++
$p];
707 $rawToggle = !$rawToggle;
716 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
719 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
720 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
723 $num = substr( $ts, 6, 2 );
726 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
727 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) +
1 );
730 $num = intval( substr( $ts, 6, 2 ) );
734 $iranian = self
::tsToIranian( $ts );
740 $hijri = self
::tsToHijri( $ts );
746 $hebrew = self
::tsToHebrew( $ts );
752 $unix = wfTimestamp( TS_UNIX
, $ts );
754 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) +
1 );
758 $unix = wfTimestamp( TS_UNIX
, $ts );
760 $w = gmdate( 'w', $unix );
765 $unix = wfTimestamp( TS_UNIX
, $ts );
767 $num = gmdate( 'w', $unix );
771 $unix = wfTimestamp( TS_UNIX
, $ts );
773 $num = gmdate( 'z', $unix );
777 $unix = wfTimestamp( TS_UNIX
, $ts );
779 $num = gmdate( 'W', $unix );
782 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
786 $iranian = self
::tsToIranian( $ts );
788 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
792 $hijri = self
::tsToHijri( $ts );
794 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
798 $hebrew = self
::tsToHebrew( $ts );
800 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
803 $num = substr( $ts, 4, 2 );
806 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
809 $num = intval( substr( $ts, 4, 2 ) );
813 $iranian = self
::tsToIranian( $ts );
819 $hijri = self
::tsToHijri ( $ts );
825 $hebrew = self
::tsToHebrew( $ts );
831 $unix = wfTimestamp( TS_UNIX
, $ts );
833 $num = gmdate( 't', $unix );
837 $hebrew = self
::tsToHebrew( $ts );
843 $unix = wfTimestamp( TS_UNIX
, $ts );
845 $num = gmdate( 'L', $unix );
847 # 'o' is supported since PHP 5.1.0
848 # return literal if not supported
849 # TODO: emulation for pre 5.1.0 versions
852 $unix = wfTimestamp( TS_UNIX
, $ts );
854 if ( version_compare( PHP_VERSION
, '5.1.0' ) === 1 ) {
855 $num = date( 'o', $unix );
861 $num = substr( $ts, 0, 4 );
865 $iranian = self
::tsToIranian( $ts );
871 $hijri = self
::tsToHijri( $ts );
877 $hebrew = self
::tsToHebrew( $ts );
883 $thai = self
::tsToYear( $ts, 'thai' );
889 $minguo = self
::tsToYear( $ts, 'minguo' );
895 $tenno = self
::tsToYear( $ts, 'tenno' );
900 $num = substr( $ts, 2, 2 );
903 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
906 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
909 $h = substr( $ts, 8, 2 );
910 $num = $h %
12 ?
$h %
12 : 12;
913 $num = intval( substr( $ts, 8, 2 ) );
916 $h = substr( $ts, 8, 2 );
917 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
920 $num = substr( $ts, 8, 2 );
923 $num = substr( $ts, 10, 2 );
926 $num = substr( $ts, 12, 2 );
930 $unix = wfTimestamp( TS_UNIX
, $ts );
932 $s .= gmdate( 'c', $unix );
936 $unix = wfTimestamp( TS_UNIX
, $ts );
938 $s .= gmdate( 'r', $unix );
942 $unix = wfTimestamp( TS_UNIX
, $ts );
948 if ( $p < strlen( $format ) - 1 ) {
956 if ( $p < strlen( $format ) - 1 ) {
957 $endQuote = strpos( $format, '"', $p +
1 );
958 if ( $endQuote === false ) {
959 # No terminating quote, assume literal "
962 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
966 # Quote at end of string, assume literal "
973 if ( $num !== false ) {
974 if ( $rawToggle ||
$raw ) {
977 } elseif ( $roman ) {
978 $s .= self
::romanNumeral( $num );
980 } elseif ( $hebrewNum ) {
981 $s .= self
::hebrewNumeral( $num );
984 $s .= $this->formatNum( $num, true );
992 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
993 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
995 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
996 * Gregorian dates to Iranian dates. Originally written in C, it
997 * is released under the terms of GNU Lesser General Public
998 * License. Conversion to PHP was performed by Niklas Laxström.
1000 * Link: http://www.farsiweb.info/jalali/jalali.c
1002 private static function tsToIranian( $ts ) {
1003 $gy = substr( $ts, 0, 4 ) -1600;
1004 $gm = substr( $ts, 4, 2 ) -1;
1005 $gd = substr( $ts, 6, 2 ) -1;
1007 # Days passed from the beginning (including leap years)
1009 +
floor( ( $gy +
3 ) / 4 )
1010 - floor( ( $gy +
99 ) / 100 )
1011 +
floor( ( $gy +
399 ) / 400 );
1014 // Add days of the past months of this year
1015 for ( $i = 0; $i < $gm; $i++
) {
1016 $gDayNo +
= self
::$GREG_DAYS[$i];
1020 if ( $gm > 1 && ( ( $gy %
4 === 0 && $gy %
100 !== 0 ||
( $gy %
400 == 0 ) ) ) ) {
1024 // Days passed in current month
1027 $jDayNo = $gDayNo - 79;
1029 $jNp = floor( $jDayNo / 12053 );
1032 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1035 if ( $jDayNo >= 366 ) {
1036 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1037 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1040 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1041 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1047 return array( $jy, $jm, $jd );
1051 * Converting Gregorian dates to Hijri dates.
1053 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1055 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1057 private static function tsToHijri( $ts ) {
1058 $year = substr( $ts, 0, 4 );
1059 $month = substr( $ts, 4, 2 );
1060 $day = substr( $ts, 6, 2 );
1068 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1069 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1072 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1073 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1074 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1077 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1078 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1081 $zl = $zjd -1948440 +
10632;
1082 $zn = (int)( ( $zl - 1 ) / 10631 );
1083 $zl = $zl - 10631 * $zn +
354;
1084 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1085 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1086 $zm = (int)( ( 24 * $zl ) / 709 );
1087 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1088 $zy = 30 * $zn +
$zj - 30;
1090 return array( $zy, $zm, $zd );
1094 * Converting Gregorian dates to Hebrew dates.
1096 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1097 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1098 * to translate the relevant functions into PHP and release them under
1101 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1102 * and Adar II is 14. In a non-leap year, Adar is 6.
1104 private static function tsToHebrew( $ts ) {
1106 $year = substr( $ts, 0, 4 );
1107 $month = substr( $ts, 4, 2 );
1108 $day = substr( $ts, 6, 2 );
1110 # Calculate Hebrew year
1111 $hebrewYear = $year +
3760;
1113 # Month number when September = 1, August = 12
1115 if ( $month > 12 ) {
1122 # Calculate day of year from 1 September
1124 for ( $i = 1; $i < $month; $i++
) {
1128 # Check if the year is leap
1129 if ( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1132 } elseif ( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1139 # Calculate the start of the Hebrew year
1140 $start = self
::hebrewYearStart( $hebrewYear );
1142 # Calculate next year's start
1143 if ( $dayOfYear <= $start ) {
1144 # Day is before the start of the year - it is the previous year
1146 $nextStart = $start;
1150 # Add days since previous year's 1 September
1152 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1156 # Start of the new (previous) year
1157 $start = self
::hebrewYearStart( $hebrewYear );
1160 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1163 # Calculate Hebrew day of year
1164 $hebrewDayOfYear = $dayOfYear - $start;
1166 # Difference between year's days
1167 $diff = $nextStart - $start;
1168 # Add 12 (or 13 for leap years) days to ignore the difference between
1169 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1170 # difference is only about the year type
1171 if ( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1177 # Check the year pattern, and is leap year
1178 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1179 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1180 # and non-leap years
1181 $yearPattern = $diff %
30;
1182 # Check if leap year
1183 $isLeap = $diff >= 30;
1185 # Calculate day in the month from number of day in the Hebrew year
1186 # Don't check Adar - if the day is not in Adar, we will stop before;
1187 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1188 $hebrewDay = $hebrewDayOfYear;
1191 while ( $hebrewMonth <= 12 ) {
1192 # Calculate days in this month
1193 if ( $isLeap && $hebrewMonth == 6 ) {
1194 # Adar in a leap year
1196 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1198 if ( $hebrewDay <= $days ) {
1202 # Subtract the days of Adar I
1203 $hebrewDay -= $days;
1206 if ( $hebrewDay <= $days ) {
1212 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1213 # Cheshvan in a complete year (otherwise as the rule below)
1215 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1216 # Kislev in an incomplete year (otherwise as the rule below)
1219 # Odd months have 30 days, even have 29
1220 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1222 if ( $hebrewDay <= $days ) {
1223 # In the current month
1226 # Subtract the days of the current month
1227 $hebrewDay -= $days;
1228 # Try in the next month
1233 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1237 * This calculates the Hebrew year start, as days since 1 September.
1238 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1239 * Used for Hebrew date.
1241 private static function hebrewYearStart( $year ) {
1242 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1243 $b = intval( ( $year - 1 ) %
4 );
1244 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1248 $Mar = intval( $m );
1254 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7 );
1255 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1257 } else if ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1259 } else if ( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1263 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1268 * Algorithm to convert Gregorian dates to Thai solar dates,
1269 * Minguo dates or Minguo dates.
1271 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1272 * http://en.wikipedia.org/wiki/Minguo_calendar
1273 * http://en.wikipedia.org/wiki/Japanese_era_name
1275 * @param $ts String: 14-character timestamp
1276 * @param $cName String: calender name
1277 * @return Array: converted year, month, day
1279 private static function tsToYear( $ts, $cName ) {
1280 $gy = substr( $ts, 0, 4 );
1281 $gm = substr( $ts, 4, 2 );
1282 $gd = substr( $ts, 6, 2 );
1284 if ( !strcmp( $cName, 'thai' ) ) {
1286 # Add 543 years to the Gregorian calendar
1287 # Months and days are identical
1288 $gy_offset = $gy +
543;
1289 } else if ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1291 # Deduct 1911 years from the Gregorian calendar
1292 # Months and days are identical
1293 $gy_offset = $gy - 1911;
1294 } else if ( !strcmp( $cName, 'tenno' ) ) {
1295 # Nengō dates up to Meiji period
1296 # Deduct years from the Gregorian calendar
1297 # depending on the nengo periods
1298 # Months and days are identical
1299 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1301 $gy_gannen = $gy - 1868 +
1;
1302 $gy_offset = $gy_gannen;
1303 if ( $gy_gannen == 1 ) {
1306 $gy_offset = '明治' . $gy_offset;
1308 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1309 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1310 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1311 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1312 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1316 $gy_gannen = $gy - 1912 +
1;
1317 $gy_offset = $gy_gannen;
1318 if ( $gy_gannen == 1 ) {
1321 $gy_offset = '大正' . $gy_offset;
1323 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1324 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1325 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1329 $gy_gannen = $gy - 1926 +
1;
1330 $gy_offset = $gy_gannen;
1331 if ( $gy_gannen == 1 ) {
1334 $gy_offset = '昭和' . $gy_offset;
1337 $gy_gannen = $gy - 1989 +
1;
1338 $gy_offset = $gy_gannen;
1339 if ( $gy_gannen == 1 ) {
1342 $gy_offset = '平成' . $gy_offset;
1348 return array( $gy_offset, $gm, $gd );
1352 * Roman number formatting up to 3000
1354 static function romanNumeral( $num ) {
1355 static $table = array(
1356 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1357 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1358 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1359 array( '', 'M', 'MM', 'MMM' )
1362 $num = intval( $num );
1363 if ( $num > 3000 ||
$num <= 0 ) {
1368 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1369 if ( $num >= $pow10 ) {
1370 $s .= $table[$i][floor( $num / $pow10 )];
1372 $num = $num %
$pow10;
1378 * Hebrew Gematria number formatting up to 9999
1380 static function hebrewNumeral( $num ) {
1381 static $table = array(
1382 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1383 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1384 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1385 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1388 $num = intval( $num );
1389 if ( $num > 9999 ||
$num <= 0 ) {
1394 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1395 if ( $num >= $pow10 ) {
1396 if ( $num == 15 ||
$num == 16 ) {
1397 $s .= $table[0][9] . $table[0][$num - 9];
1400 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1401 if ( $pow10 == 1000 ) {
1406 $num = $num %
$pow10;
1408 if ( strlen( $s ) == 2 ) {
1411 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1412 $str .= substr( $s, strlen( $s ) - 2, 2 );
1414 $start = substr( $str, 0, strlen( $str ) - 2 );
1415 $end = substr( $str, strlen( $str ) - 2 );
1418 $str = $start . 'ך';
1421 $str = $start . 'ם';
1424 $str = $start . 'ן';
1427 $str = $start . 'ף';
1430 $str = $start . 'ץ';
1437 * This is meant to be used by time(), date(), and timeanddate() to get
1438 * the date preference they're supposed to use, it should be used in
1442 * function timeanddate([...], $format = true) {
1443 * $datePreference = $this->dateFormat($format);
1448 * @param $usePrefs Mixed: if true, the user's preference is used
1449 * if false, the site/language default is used
1450 * if int/string, assumed to be a format.
1453 function dateFormat( $usePrefs = true ) {
1456 if ( is_bool( $usePrefs ) ) {
1458 $datePreference = $wgUser->getDatePreference();
1460 $datePreference = (string)User
::getDefaultOption( 'date' );
1463 $datePreference = (string)$usePrefs;
1467 if ( $datePreference == '' ) {
1471 return $datePreference;
1475 * Get a format string for a given type and preference
1476 * @param $type May be date, time or both
1477 * @param $pref The format name as it appears in Messages*.php
1479 function getDateFormatString( $type, $pref ) {
1480 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
1481 if ( $pref == 'default' ) {
1482 $pref = $this->getDefaultDateFormat();
1483 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1485 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1486 if ( is_null( $df ) ) {
1487 $pref = $this->getDefaultDateFormat();
1488 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1491 $this->dateFormatStrings
[$type][$pref] = $df;
1493 return $this->dateFormatStrings
[$type][$pref];
1497 * @param $ts Mixed: the time format which needs to be turned into a
1498 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1499 * @param $adj Bool: whether to adjust the time output according to the
1500 * user configured offset ($timecorrection)
1501 * @param $format Mixed: true to use user's date format preference
1502 * @param $timecorrection String: the time offset as returned by
1503 * validateTimeZone() in Special:Preferences
1506 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1508 $ts = $this->userAdjust( $ts, $timecorrection );
1510 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1511 return $this->sprintfDate( $df, $ts );
1515 * @param $ts Mixed: the time format which needs to be turned into a
1516 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1517 * @param $adj Bool: whether to adjust the time output according to the
1518 * user configured offset ($timecorrection)
1519 * @param $format Mixed: true to use user's date format preference
1520 * @param $timecorrection String: the time offset as returned by
1521 * validateTimeZone() in Special:Preferences
1524 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1526 $ts = $this->userAdjust( $ts, $timecorrection );
1528 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1529 return $this->sprintfDate( $df, $ts );
1533 * @param $ts Mixed: the time format which needs to be turned into a
1534 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1535 * @param $adj Bool: whether to adjust the time output according to the
1536 * user configured offset ($timecorrection)
1537 * @param $format Mixed: what format to return, if it's false output the
1538 * default one (default true)
1539 * @param $timecorrection String: the time offset as returned by
1540 * validateTimeZone() in Special:Preferences
1543 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1544 $ts = wfTimestamp( TS_MW
, $ts );
1546 $ts = $this->userAdjust( $ts, $timecorrection );
1548 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1549 return $this->sprintfDate( $df, $ts );
1552 function getMessage( $key ) {
1553 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
1556 function getAllMessages() {
1557 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
1560 function iconv( $in, $out, $string ) {
1561 # This is a wrapper for iconv in all languages except esperanto,
1562 # which does some nasty x-conversions beforehand
1564 # Even with //IGNORE iconv can whine about illegal characters in
1565 # *input* string. We just ignore those too.
1566 # REF: http://bugs.php.net/bug.php?id=37166
1567 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1568 wfSuppressWarnings();
1569 $text = iconv( $in, $out . '//IGNORE', $string );
1570 wfRestoreWarnings();
1574 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1575 function ucwordbreaksCallbackAscii( $matches ) {
1576 return $this->ucfirst( $matches[1] );
1579 function ucwordbreaksCallbackMB( $matches ) {
1580 return mb_strtoupper( $matches[0] );
1583 function ucCallback( $matches ) {
1584 list( $wikiUpperChars ) = self
::getCaseMaps();
1585 return strtr( $matches[1], $wikiUpperChars );
1588 function lcCallback( $matches ) {
1589 list( , $wikiLowerChars ) = self
::getCaseMaps();
1590 return strtr( $matches[1], $wikiLowerChars );
1593 function ucwordsCallbackMB( $matches ) {
1594 return mb_strtoupper( $matches[0] );
1597 function ucwordsCallbackWiki( $matches ) {
1598 list( $wikiUpperChars ) = self
::getCaseMaps();
1599 return strtr( $matches[0], $wikiUpperChars );
1603 * Make a string's first character uppercase
1605 function ucfirst( $str ) {
1607 if ( $o < 96 ) { // if already uppercase...
1609 } elseif ( $o < 128 ) {
1610 return ucfirst( $str ); // use PHP's ucfirst()
1612 // fall back to more complex logic in case of multibyte strings
1613 return $this->uc( $str, true );
1618 * Convert a string to uppercase
1620 function uc( $str, $first = false ) {
1621 if ( function_exists( 'mb_strtoupper' ) ) {
1623 if ( $this->isMultibyte( $str ) ) {
1624 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1626 return ucfirst( $str );
1629 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
1632 if ( $this->isMultibyte( $str ) ) {
1633 $x = $first ?
'^' : '';
1634 return preg_replace_callback(
1635 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1636 array( $this, 'ucCallback' ),
1640 return $first ?
ucfirst( $str ) : strtoupper( $str );
1645 function lcfirst( $str ) {
1648 return strval( $str );
1649 } elseif ( $o >= 128 ) {
1650 return $this->lc( $str, true );
1651 } elseif ( $o > 96 ) {
1654 $str[0] = strtolower( $str[0] );
1659 function lc( $str, $first = false ) {
1660 if ( function_exists( 'mb_strtolower' ) ) {
1662 if ( $this->isMultibyte( $str ) ) {
1663 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1665 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1668 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
1671 if ( $this->isMultibyte( $str ) ) {
1672 $x = $first ?
'^' : '';
1673 return preg_replace_callback(
1674 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1675 array( $this, 'lcCallback' ),
1679 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1684 function isMultibyte( $str ) {
1685 return (bool)preg_match( '/[\x80-\xff]/', $str );
1688 function ucwords( $str ) {
1689 if ( $this->isMultibyte( $str ) ) {
1690 $str = $this->lc( $str );
1692 // regexp to find first letter in each word (i.e. after each space)
1693 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1695 // function to use to capitalize a single char
1696 if ( function_exists( 'mb_strtoupper' ) ) {
1697 return preg_replace_callback(
1699 array( $this, 'ucwordsCallbackMB' ),
1703 return preg_replace_callback(
1705 array( $this, 'ucwordsCallbackWiki' ),
1710 return ucwords( strtolower( $str ) );
1714 # capitalize words at word breaks
1715 function ucwordbreaks( $str ) {
1716 if ( $this->isMultibyte( $str ) ) {
1717 $str = $this->lc( $str );
1719 // since \b doesn't work for UTF-8, we explicitely define word break chars
1720 $breaks = "[ \-\(\)\}\{\.,\?!]";
1722 // find first letter after word break
1723 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1725 if ( function_exists( 'mb_strtoupper' ) ) {
1726 return preg_replace_callback(
1728 array( $this, 'ucwordbreaksCallbackMB' ),
1732 return preg_replace_callback(
1734 array( $this, 'ucwordsCallbackWiki' ),
1739 return preg_replace_callback(
1740 '/\b([\w\x80-\xff]+)\b/',
1741 array( $this, 'ucwordbreaksCallbackAscii' ),
1748 * Return a case-folded representation of $s
1750 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1751 * and $s2 are the same except for the case of their characters. It is not
1752 * necessary for the value returned to make sense when displayed.
1754 * Do *not* perform any other normalisation in this function. If a caller
1755 * uses this function when it should be using a more general normalisation
1756 * function, then fix the caller.
1758 function caseFold( $s ) {
1759 return $this->uc( $s );
1762 function checkTitleEncoding( $s ) {
1763 if ( is_array( $s ) ) {
1764 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1766 # Check for non-UTF-8 URLs
1767 $ishigh = preg_match( '/[\x80-\xff]/', $s );
1772 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1773 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1778 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
1781 function fallback8bitEncoding() {
1782 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
1786 * Most writing systems use whitespace to break up words.
1787 * Some languages such as Chinese don't conventionally do this,
1788 * which requires special handling when breaking up words for
1791 function hasWordBreaks() {
1796 * Some languages such as Chinese require word segmentation,
1797 * Specify such segmentation when overridden in derived class.
1799 * @param $string String
1802 function segmentByWord( $string ) {
1807 * Some languages have special punctuation need to be normalized.
1808 * Make such changes here.
1810 * @param $string String
1813 function normalizeForSearch( $string ) {
1814 return self
::convertDoubleWidth( $string );
1818 * convert double-width roman characters to single-width.
1819 * range: ff00-ff5f ~= 0020-007f
1821 protected static function convertDoubleWidth( $string ) {
1822 static $full = null;
1823 static $half = null;
1825 if ( $full === null ) {
1826 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1827 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1828 $full = str_split( $fullWidth, 3 );
1829 $half = str_split( $halfWidth );
1832 $string = str_replace( $full, $half, $string );
1836 protected static function insertSpace( $string, $pattern ) {
1837 $string = preg_replace( $pattern, " $1 ", $string );
1838 $string = preg_replace( '/ +/', ' ', $string );
1842 function convertForSearchResult( $termsArray ) {
1843 # some languages, e.g. Chinese, need to do a conversion
1844 # in order for search results to be displayed correctly
1849 * Get the first character of a string.
1854 function firstChar( $s ) {
1857 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1858 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
1863 if ( isset( $matches[1] ) ) {
1864 if ( strlen( $matches[1] ) != 3 ) {
1868 // Break down Hangul syllables to grab the first jamo
1869 $code = utf8ToCodepoint( $matches[1] );
1870 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
1872 } elseif ( $code < 0xb098 ) {
1873 return "\xe3\x84\xb1";
1874 } elseif ( $code < 0xb2e4 ) {
1875 return "\xe3\x84\xb4";
1876 } elseif ( $code < 0xb77c ) {
1877 return "\xe3\x84\xb7";
1878 } elseif ( $code < 0xb9c8 ) {
1879 return "\xe3\x84\xb9";
1880 } elseif ( $code < 0xbc14 ) {
1881 return "\xe3\x85\x81";
1882 } elseif ( $code < 0xc0ac ) {
1883 return "\xe3\x85\x82";
1884 } elseif ( $code < 0xc544 ) {
1885 return "\xe3\x85\x85";
1886 } elseif ( $code < 0xc790 ) {
1887 return "\xe3\x85\x87";
1888 } elseif ( $code < 0xcc28 ) {
1889 return "\xe3\x85\x88";
1890 } elseif ( $code < 0xce74 ) {
1891 return "\xe3\x85\x8a";
1892 } elseif ( $code < 0xd0c0 ) {
1893 return "\xe3\x85\x8b";
1894 } elseif ( $code < 0xd30c ) {
1895 return "\xe3\x85\x8c";
1896 } elseif ( $code < 0xd558 ) {
1897 return "\xe3\x85\x8d";
1899 return "\xe3\x85\x8e";
1906 function initEncoding() {
1907 # Some languages may have an alternate char encoding option
1908 # (Esperanto X-coding, Japanese furigana conversion, etc)
1909 # If this language is used as the primary content language,
1910 # an override to the defaults can be set here on startup.
1913 function recodeForEdit( $s ) {
1914 # For some languages we'll want to explicitly specify
1915 # which characters make it into the edit box raw
1916 # or are converted in some way or another.
1917 # Note that if wgOutputEncoding is different from
1918 # wgInputEncoding, this text will be further converted
1919 # to wgOutputEncoding.
1920 global $wgEditEncoding;
1921 if ( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
1924 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1928 function recodeInput( $s ) {
1929 # Take the previous into account.
1930 global $wgEditEncoding;
1931 if ( $wgEditEncoding != '' ) {
1932 $enc = $wgEditEncoding;
1936 if ( $enc == 'UTF-8' ) {
1939 return $this->iconv( $enc, 'UTF-8', $s );
1944 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
1945 * also cleans up certain backwards-compatible sequences, converting them
1946 * to the modern Unicode equivalent.
1948 * This is language-specific for performance reasons only.
1950 function normalize( $s ) {
1951 global $wgAllUnicodeFixes;
1952 $s = UtfNormal
::cleanUp( $s );
1953 if ( $wgAllUnicodeFixes ) {
1954 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
1955 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
1962 * Transform a string using serialized data stored in the given file (which
1963 * must be in the serialized subdirectory of $IP). The file contains pairs
1964 * mapping source characters to destination characters.
1966 * The data is cached in process memory. This will go faster if you have the
1967 * FastStringSearch extension.
1969 function transformUsingPairFile( $file, $string ) {
1970 if ( !isset( $this->transformData
[$file] ) ) {
1971 $data = wfGetPrecompiledData( $file );
1972 if ( $data === false ) {
1973 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
1975 $this->transformData
[$file] = new ReplacementArray( $data );
1977 return $this->transformData
[$file]->replace( $string );
1981 * For right-to-left language support
1986 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
1990 * Return the correct HTML 'dir' attribute value for this language.
1994 return $this->isRTL() ?
'rtl' : 'ltr';
1998 * Return 'left' or 'right' as appropriate alignment for line-start
1999 * for this language's text direction.
2001 * Should be equivalent to CSS3 'start' text-align value....
2005 function alignStart() {
2006 return $this->isRTL() ?
'right' : 'left';
2010 * Return 'right' or 'left' as appropriate alignment for line-end
2011 * for this language's text direction.
2013 * Should be equivalent to CSS3 'end' text-align value....
2017 function alignEnd() {
2018 return $this->isRTL() ?
'left' : 'right';
2022 * A hidden direction mark (LRM or RLM), depending on the language direction
2026 function getDirMark() {
2027 return $this->isRTL() ?
"\xE2\x80\x8F" : "\xE2\x80\x8E";
2030 function capitalizeAllNouns() {
2031 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2035 * An arrow, depending on the language direction
2039 function getArrow() {
2040 return $this->isRTL() ?
'←' : '→';
2044 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2048 function linkPrefixExtension() {
2049 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2052 function getMagicWords() {
2053 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2056 # Fill a MagicWord object with data from here
2057 function getMagic( $mw ) {
2058 if ( !$this->mMagicHookDone
) {
2059 $this->mMagicHookDone
= true;
2060 wfProfileIn( 'LanguageGetMagic' );
2061 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2062 wfProfileOut( 'LanguageGetMagic' );
2064 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
2065 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
2067 $magicWords = $this->getMagicWords();
2068 if ( isset( $magicWords[$mw->mId
] ) ) {
2069 $rawEntry = $magicWords[$mw->mId
];
2075 if ( !is_array( $rawEntry ) ) {
2076 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
2078 $mw->mCaseSensitive
= $rawEntry[0];
2079 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
2084 * Add magic words to the extension array
2086 function addMagicWordsByLang( $newWords ) {
2087 $code = $this->getCode();
2088 $fallbackChain = array();
2089 while ( $code && !in_array( $code, $fallbackChain ) ) {
2090 $fallbackChain[] = $code;
2091 $code = self
::getFallbackFor( $code );
2093 if ( !in_array( 'en', $fallbackChain ) ) {
2094 $fallbackChain[] = 'en';
2096 $fallbackChain = array_reverse( $fallbackChain );
2097 foreach ( $fallbackChain as $code ) {
2098 if ( isset( $newWords[$code] ) ) {
2099 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
2105 * Get special page names, as an associative array
2106 * case folded alias => real name
2108 function getSpecialPageAliases() {
2109 // Cache aliases because it may be slow to load them
2110 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
2112 $this->mExtendedSpecialPageAliases
=
2113 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
2114 wfRunHooks( 'LanguageGetSpecialPageAliases',
2115 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
2118 return $this->mExtendedSpecialPageAliases
;
2122 * Italic is unsuitable for some languages
2124 * @param $text String: the text to be emphasized.
2127 function emphasize( $text ) {
2128 return "<em>$text</em>";
2132 * Normally we output all numbers in plain en_US style, that is
2133 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2134 * point twohundredthirtyfive. However this is not sutable for all
2135 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2136 * Icelandic just want to use commas instead of dots, and dots instead
2137 * of commas like "293.291,235".
2139 * An example of this function being called:
2141 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2144 * See LanguageGu.php for the Gujarati implementation and
2145 * $separatorTransformTable on MessageIs.php for
2146 * the , => . and . => , implementation.
2148 * @todo check if it's viable to use localeconv() for the decimal
2150 * @param $number Mixed: the string to be formatted, should be an integer
2151 * or a floating point number.
2152 * @param $nocommafy Bool: set to true for special numbers like dates
2155 function formatNum( $number, $nocommafy = false ) {
2156 global $wgTranslateNumerals;
2157 if ( !$nocommafy ) {
2158 $number = $this->commafy( $number );
2159 $s = $this->separatorTransformTable();
2161 $number = strtr( $number, $s );
2165 if ( $wgTranslateNumerals ) {
2166 $s = $this->digitTransformTable();
2168 $number = strtr( $number, $s );
2175 function parseFormattedNumber( $number ) {
2176 $s = $this->digitTransformTable();
2178 $number = strtr( $number, array_flip( $s ) );
2181 $s = $this->separatorTransformTable();
2183 $number = strtr( $number, array_flip( $s ) );
2186 $number = strtr( $number, array( ',' => '' ) );
2191 * Adds commas to a given number
2196 function commafy( $_ ) {
2197 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2200 function digitTransformTable() {
2201 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
2204 function separatorTransformTable() {
2205 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
2209 * Take a list of strings and build a locale-friendly comma-separated
2210 * list, using the local comma-separator message.
2211 * The last two strings are chained with an "and".
2216 function listToText( $l ) {
2218 $m = count( $l ) - 1;
2220 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2222 for ( $i = $m; $i >= 0; $i-- ) {
2225 } else if ( $i == $m - 1 ) {
2226 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2228 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2236 * Take a list of strings and build a locale-friendly comma-separated
2237 * list, using the local comma-separator message.
2238 * @param $list array of strings to put in a comma list
2241 function commaList( $list ) {
2246 array( 'parsemag', 'escapenoentities', 'language' => $this )
2252 * Take a list of strings and build a locale-friendly semicolon-separated
2253 * list, using the local semicolon-separator message.
2254 * @param $list array of strings to put in a semicolon list
2257 function semicolonList( $list ) {
2261 'semicolon-separator',
2262 array( 'parsemag', 'escapenoentities', 'language' => $this )
2268 * Same as commaList, but separate it with the pipe instead.
2269 * @param $list array of strings to put in a pipe list
2272 function pipeList( $list ) {
2277 array( 'escapenoentities', 'language' => $this )
2283 * Truncate a string to a specified length in bytes, appending an optional
2284 * string (e.g. for ellipses)
2286 * The database offers limited byte lengths for some columns in the database;
2287 * multi-byte character sets mean we need to ensure that only whole characters
2288 * are included, otherwise broken characters can be passed to the user
2290 * If $length is negative, the string will be truncated from the beginning
2292 * @param $string String to truncate
2293 * @param $length Int: maximum length (excluding ellipses)
2294 * @param $ellipsis String to append to the truncated text
2297 function truncate( $string, $length, $ellipsis = '...' ) {
2298 # Use the localized ellipsis character
2299 if ( $ellipsis == '...' ) {
2300 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2302 # Check if there is no need to truncate
2303 if ( $length == 0 ) {
2305 } elseif ( strlen( $string ) <= abs( $length ) ) {
2308 $stringOriginal = $string;
2309 if ( $length > 0 ) {
2310 $string = substr( $string, 0, $length ); // xyz...
2311 $string = $this->removeBadCharLast( $string );
2312 $string = $string . $ellipsis;
2314 $string = substr( $string, $length ); // ...xyz
2315 $string = $this->removeBadCharFirst( $string );
2316 $string = $ellipsis . $string;
2318 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181)
2319 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
2322 return $stringOriginal;
2327 * Remove bytes that represent an incomplete Unicode character
2328 * at the end of string (e.g. bytes of the char are missing)
2330 * @param $string String
2333 protected function removeBadCharLast( $string ) {
2334 $char = ord( $string[strlen( $string ) - 1] );
2336 if ( $char >= 0xc0 ) {
2337 # We got the first byte only of a multibyte char; remove it.
2338 $string = substr( $string, 0, -1 );
2339 } elseif ( $char >= 0x80 &&
2340 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2341 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
2343 # We chopped in the middle of a character; remove it
2350 * Remove bytes that represent an incomplete Unicode character
2351 * at the start of string (e.g. bytes of the char are missing)
2353 * @param $string String
2356 protected function removeBadCharFirst( $string ) {
2357 $char = ord( $string[0] );
2358 if ( $char >= 0x80 && $char < 0xc0 ) {
2359 # We chopped in the middle of a character; remove the whole thing
2360 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2366 * Truncate a string of valid HTML to a specified length in bytes,
2367 * appending an optional string (e.g. for ellipses), and return valid HTML
2369 * This is only intended for styled/linked text, such as HTML with
2370 * tags like <span> and <a>, were the tags are self-contained (valid HTML)
2372 * Note: tries to fix broken HTML with MWTidy
2374 * @param string $text String to truncate
2375 * @param int $length (zero/positive) Maximum length (excluding ellipses)
2376 * @param string $ellipsis String to append to the truncated text
2379 function truncateHtml( $text, $length, $ellipsis = '...' ) {
2380 # Use the localized ellipsis character
2381 if ( $ellipsis == '...' ) {
2382 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2384 # Check if there is no need to truncate
2385 if ( $length <= 0 ) {
2386 return $ellipsis; // no text shown, nothing to format
2387 } elseif ( strlen( $text ) <= $length ) {
2388 return $text; // string short enough even *with* HTML
2390 $text = MWTidy
::tidy( $text ); // fix tags
2391 $displayLen = 0; // innerHTML legth so far
2392 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
2393 $tagType = 0; // 0-open, 1-close
2394 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
2395 $entityState = 0; // 0-not entity, 1-entity
2396 $tag = $ret = $ch = '';
2397 $openTags = array();
2398 $textLen = strlen( $text );
2399 for ( $pos = 0; $pos < $textLen; ++
$pos ) {
2401 $lastCh = $pos ?
$text[$pos - 1] : '';
2402 $ret .= $ch; // add to result string
2404 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
2405 $entityState = 0; // for bad HTML
2406 $bracketState = 1; // tag started (checking for backslash)
2407 } elseif ( $ch == '>' ) {
2408 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
2409 $entityState = 0; // for bad HTML
2410 $bracketState = 0; // out of brackets
2411 } elseif ( $bracketState == 1 ) {
2413 $tagType = 1; // close tag (e.g. "</span>")
2415 $tagType = 0; // open tag (e.g. "<span>")
2418 $bracketState = 2; // building tag name
2419 } elseif ( $bracketState == 2 ) {
2423 // Name found (e.g. "<a href=..."), add on tag attributes...
2424 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
2426 } elseif ( $bracketState == 0 ) {
2427 if ( $entityState ) {
2430 $displayLen++
; // entity is one displayed char
2434 $entityState = 1; // entity found, (e.g. " ")
2436 $displayLen++
; // this char is displayed
2437 // Add on the other display text after this...
2438 $skipped = $this->truncate_skip(
2439 $ret, $text, "<>&", $pos +
1, $length - $displayLen );
2440 $displayLen +
= $skipped;
2445 # Consider truncation once the display length has reached the maximim.
2446 # Double-check that we're not in the middle of a bracket/entity...
2447 if ( $displayLen >= $length && $bracketState == 0 && $entityState == 0 ) {
2448 if ( !$testingEllipsis ) {
2449 $testingEllipsis = true;
2450 # Save where we are; we will truncate here unless
2451 # the ellipsis actually makes the string longer.
2452 $pOpenTags = $openTags; // save state
2453 $pRet = $ret; // save state
2454 } elseif ( $displayLen > ( $length +
strlen( $ellipsis ) ) ) {
2455 # Ellipsis won't make string longer/equal, the truncation point was OK.
2456 $openTags = $pOpenTags; // reload state
2457 $ret = $this->removeBadCharLast( $pRet ); // reload state, multi-byte char fix
2458 $ret .= $ellipsis; // add ellipsis
2463 if ( $displayLen == 0 ) {
2464 return ''; // no text shown, nothing to format
2466 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags ); // for bad HTML
2467 while ( count( $openTags ) > 0 ) {
2468 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
2473 // truncateHtml() helper function
2474 // like strcspn() but adds the skipped chars to $ret
2475 private function truncate_skip( &$ret, $text, $search, $start, $len = -1 ) {
2477 if ( $start < strlen( $text ) ) {
2478 $skipCount = strcspn( $text, $search, $start, $len );
2479 $ret .= substr( $text, $start, $skipCount );
2484 // truncateHtml() helper function
2485 // (a) push or pop $tag from $openTags as needed
2486 // (b) clear $tag value
2487 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
2488 $tag = ltrim( $tag );
2490 if ( $tagType == 0 && $lastCh != '/' ) {
2491 $openTags[] = $tag; // tag opened (didn't close itself)
2492 } else if ( $tagType == 1 ) {
2493 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
2494 array_pop( $openTags ); // tag closed
2502 * Grammatical transformations, needed for inflected languages
2503 * Invoked by putting {{grammar:case|word}} in a message
2505 * @param $word string
2506 * @param $case string
2509 function convertGrammar( $word, $case ) {
2510 global $wgGrammarForms;
2511 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
2512 return $wgGrammarForms[$this->getCode()][$case][$word];
2518 * Provides an alternative text depending on specified gender.
2519 * Usage {{gender:username|masculine|feminine|neutral}}.
2520 * username is optional, in which case the gender of current user is used,
2521 * but only in (some) interface messages; otherwise default gender is used.
2522 * If second or third parameter are not specified, masculine is used.
2523 * These details may be overriden per language.
2525 function gender( $gender, $forms ) {
2526 if ( !count( $forms ) ) {
2529 $forms = $this->preConvertPlural( $forms, 2 );
2530 if ( $gender === 'male' ) {
2533 if ( $gender === 'female' ) {
2536 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
2540 * Plural form transformations, needed for some languages.
2541 * For example, there are 3 form of plural in Russian and Polish,
2542 * depending on "count mod 10". See [[w:Plural]]
2543 * For English it is pretty simple.
2545 * Invoked by putting {{plural:count|wordform1|wordform2}}
2546 * or {{plural:count|wordform1|wordform2|wordform3}}
2548 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2550 * @param $count Integer: non-localized number
2551 * @param $forms Array: different plural forms
2552 * @return string Correct form of plural for $count in this language
2554 function convertPlural( $count, $forms ) {
2555 if ( !count( $forms ) ) {
2558 $forms = $this->preConvertPlural( $forms, 2 );
2560 return ( $count == 1 ) ?
$forms[0] : $forms[1];
2564 * Checks that convertPlural was given an array and pads it to requested
2565 * amound of forms by copying the last one.
2567 * @param $count Integer: How many forms should there be at least
2568 * @param $forms Array of forms given to convertPlural
2569 * @return array Padded array of forms or an exception if not an array
2571 protected function preConvertPlural( /* Array */ $forms, $count ) {
2572 while ( count( $forms ) < $count ) {
2573 $forms[] = $forms[count( $forms ) - 1];
2579 * For translating of expiry times
2580 * @param $str String: the validated block time in English
2581 * @return Somehow translated block time
2582 * @see LanguageFi.php for example implementation
2584 function translateBlockExpiry( $str ) {
2585 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2587 if ( $scBlockExpiryOptions == '-' ) {
2591 foreach ( explode( ',', $scBlockExpiryOptions ) as $option ) {
2592 if ( strpos( $option, ':' ) === false ) {
2595 list( $show, $value ) = explode( ':', $option );
2596 if ( strcmp( $str, $value ) == 0 ) {
2597 return htmlspecialchars( trim( $show ) );
2605 * languages like Chinese need to be segmented in order for the diff
2608 * @param $text String
2611 function segmentForDiff( $text ) {
2616 * and unsegment to show the result
2618 * @param $text String
2621 function unsegmentForDiff( $text ) {
2625 # convert text to all supported variants
2626 function autoConvertToAllVariants( $text ) {
2627 return $this->mConverter
->autoConvertToAllVariants( $text );
2630 # convert text to different variants of a language.
2631 function convert( $text ) {
2632 return $this->mConverter
->convert( $text );
2635 # Convert a Title object to a string in the preferred variant
2636 function convertTitle( $title ) {
2637 return $this->mConverter
->convertTitle( $title );
2640 # Check if this is a language with variants
2641 function hasVariants() {
2642 return sizeof( $this->getVariants() ) > 1;
2645 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2646 function armourMath( $text ) {
2647 return $this->mConverter
->armourMath( $text );
2651 * Perform output conversion on a string, and encode for safe HTML output.
2652 * @param $text String text to be converted
2653 * @param $isTitle Bool whether this conversion is for the article title
2655 * @todo this should get integrated somewhere sane
2657 function convertHtml( $text, $isTitle = false ) {
2658 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2661 function convertCategoryKey( $key ) {
2662 return $this->mConverter
->convertCategoryKey( $key );
2666 * Get the list of variants supported by this langauge
2667 * see sample implementation in LanguageZh.php
2669 * @return array an array of language codes
2671 function getVariants() {
2672 return $this->mConverter
->getVariants();
2675 function getPreferredVariant( $fromUser = true, $fromHeader = false ) {
2676 return $this->mConverter
->getPreferredVariant( $fromUser, $fromHeader );
2680 * If a language supports multiple variants, it is
2681 * possible that non-existing link in one variant
2682 * actually exists in another variant. this function
2683 * tries to find it. See e.g. LanguageZh.php
2685 * @param $link String: the name of the link
2686 * @param $nt Mixed: the title object of the link
2687 * @param $ignoreOtherCond Boolean: to disable other conditions when
2688 * we need to transclude a template or update a category's link
2689 * @return null the input parameters may be modified upon return
2691 function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
2692 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
2696 * If a language supports multiple variants, converts text
2697 * into an array of all possible variants of the text:
2698 * 'variant' => text in that variant
2700 function convertLinkToAllVariants( $text ) {
2701 return $this->mConverter
->convertLinkToAllVariants( $text );
2705 * returns language specific options used by User::getPageRenderHash()
2706 * for example, the preferred language variant
2710 function getExtraHashOptions() {
2711 return $this->mConverter
->getExtraHashOptions();
2715 * For languages that support multiple variants, the title of an
2716 * article may be displayed differently in different variants. this
2717 * function returns the apporiate title defined in the body of the article.
2721 function getParsedTitle() {
2722 return $this->mConverter
->getParsedTitle();
2726 * Enclose a string with the "no conversion" tag. This is used by
2727 * various functions in the Parser
2729 * @param $text String: text to be tagged for no conversion
2731 * @return string the tagged text
2733 function markNoConversion( $text, $noParse = false ) {
2734 return $this->mConverter
->markNoConversion( $text, $noParse );
2738 * A regular expression to match legal word-trailing characters
2739 * which should be merged onto a link of the form [[foo]]bar.
2743 function linkTrail() {
2744 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
2747 function getLangObj() {
2752 * Get the RFC 3066 code for this language object
2754 function getCode() {
2755 return $this->mCode
;
2758 function setCode( $code ) {
2759 $this->mCode
= $code;
2763 * Get the name of a file for a certain language code
2764 * @param $prefix string Prepend this to the filename
2765 * @param $code string Language code
2766 * @param $suffix string Append this to the filename
2767 * @return string $prefix . $mangledCode . $suffix
2769 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2770 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2774 * Get the language code from a file name. Inverse of getFileName()
2775 * @param $filename string $prefix . $languageCode . $suffix
2776 * @param $prefix string Prefix before the language code
2777 * @param $suffix string Suffix after the language code
2778 * @return Language code, or false if $prefix or $suffix isn't found
2780 static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
2782 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
2783 preg_quote( $suffix, '/' ) . '/', $filename, $m );
2784 if ( !count( $m ) ) {
2787 return str_replace( '_', '-', strtolower( $m[1] ) );
2790 static function getMessagesFileName( $code ) {
2792 return self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2795 static function getClassFileName( $code ) {
2797 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2801 * Get the fallback for a given language
2803 static function getFallbackFor( $code ) {
2804 if ( $code === 'en' ) {
2808 return self
::getLocalisationCache()->getItem( $code, 'fallback' );
2813 * Get all messages for a given language
2814 * WARNING: this may take a long time
2816 static function getMessagesFor( $code ) {
2817 return self
::getLocalisationCache()->getItem( $code, 'messages' );
2821 * Get a message for a given language
2823 static function getMessageFor( $key, $code ) {
2824 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
2827 function fixVariableInNamespace( $talk ) {
2828 if ( strpos( $talk, '$1' ) === false ) {
2832 global $wgMetaNamespace;
2833 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2835 # Allow grammar transformations
2836 # Allowing full message-style parsing would make simple requests
2837 # such as action=raw much more expensive than they need to be.
2838 # This will hopefully cover most cases.
2839 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2840 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2841 return str_replace( ' ', '_', $talk );
2844 function replaceGrammarInNamespace( $m ) {
2845 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2848 static function getCaseMaps() {
2849 static $wikiUpperChars, $wikiLowerChars;
2850 if ( isset( $wikiUpperChars ) ) {
2851 return array( $wikiUpperChars, $wikiLowerChars );
2854 wfProfileIn( __METHOD__
);
2855 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2856 if ( $arr === false ) {
2857 throw new MWException(
2858 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2861 wfProfileOut( __METHOD__
);
2862 return array( $wikiUpperChars, $wikiLowerChars );
2865 function formatTimePeriod( $seconds ) {
2866 if ( $seconds < 10 ) {
2867 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2868 } elseif ( $seconds < 60 ) {
2869 return $this->formatNum( round( $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2870 } elseif ( $seconds < 3600 ) {
2871 $minutes = floor( $seconds / 60 );
2872 $secondsPart = round( fmod( $seconds, 60 ) );
2873 if ( $secondsPart == 60 ) {
2877 return $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2878 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2880 $hours = floor( $seconds / 3600 );
2881 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2882 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2883 if ( $secondsPart == 60 ) {
2887 if ( $minutes == 60 ) {
2891 return $this->formatNum( $hours ) . $this->getMessageFromDB( 'hours-abbrev' ) . ' ' .
2892 $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2893 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2897 function formatBitrate( $bps ) {
2898 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2900 return $this->formatNum( $bps ) . $units[0];
2902 $unitIndex = floor( log10( $bps ) / 3 );
2903 $mantissa = $bps / pow( 1000, $unitIndex );
2904 if ( $mantissa < 10 ) {
2905 $mantissa = round( $mantissa, 1 );
2907 $mantissa = round( $mantissa );
2909 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2913 * Format a size in bytes for output, using an appropriate
2914 * unit (B, KB, MB or GB) according to the magnitude in question
2916 * @param $size Size to format
2917 * @return string Plain text (not HTML)
2919 function formatSize( $size ) {
2920 // For small sizes no decimal places necessary
2922 if ( $size > 1024 ) {
2923 $size = $size / 1024;
2924 if ( $size > 1024 ) {
2925 $size = $size / 1024;
2926 // For MB and bigger two decimal places are smarter
2928 if ( $size > 1024 ) {
2929 $size = $size / 1024;
2930 $msg = 'size-gigabytes';
2932 $msg = 'size-megabytes';
2935 $msg = 'size-kilobytes';
2938 $msg = 'size-bytes';
2940 $size = round( $size, $round );
2941 $text = $this->getMessageFromDB( $msg );
2942 return str_replace( '$1', $this->formatNum( $size ), $text );
2946 * Get the conversion rule title, if any.
2948 function getConvRuleTitle() {
2949 return $this->mConverter
->getConvRuleTitle();
2953 * Given a string, convert it to a (hopefully short) key that can be used
2954 * for efficient sorting. A binary sort according to the sortkeys
2955 * corresponds to a logical sort of the corresponding strings. Current
2956 * code expects that a null character should sort before all others, but
2957 * has no other particular expectations (and that one can be changed if
2960 * @param string $string UTF-8 string
2961 * @return string Binary sortkey
2963 public function convertToSortkey( $string ) {
2964 # Fake function for now
2965 return strtoupper( $string );
2969 * Given a string, return the logical "first letter" to be used for
2970 * grouping on category pages and so on. This has to be coordinated
2971 * carefully with convertToSortkey(), or else the sorted list might jump
2972 * back and forth between the same "initial letters" or other pathological
2973 * behavior. For instance, if you just return the first character, but "a"
2974 * sorts the same as "A" based on convertToSortkey(), then you might get a
2986 * etc., assuming for the sake of argument that $wgCapitalLinks is false.
2988 * @param string $string UTF-8 string
2989 * @return string UTF-8 string corresponding to the first letter of input
2991 public function firstLetterForLists( $string ) {
2992 if ( $string[0] == "\0" ) {
2993 $string = substr( $string, 1 );
2995 return strtoupper( $this->firstChar( $string ) );