* Extend language::getLanguageName to return localized language names if available
[mediawiki.git] / languages / Language.php
blobeaec2b671eaaa81de6dbb18c1d141d7ac22fd754
1 <?php
2 /**
3 * @defgroup Language Language
5 * @file
6 * @ingroup Language
7 */
9 if( !defined( 'MEDIAWIKI' ) ) {
10 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
11 exit( 1 );
14 # Read language names
15 global $wgLanguageNames;
16 require_once( dirname(__FILE__) . '/Names.php' ) ;
18 global $wgInputEncoding, $wgOutputEncoding;
20 /**
21 * These are always UTF-8, they exist only for backwards compatibility
23 $wgInputEncoding = "UTF-8";
24 $wgOutputEncoding = "UTF-8";
26 if( function_exists( 'mb_strtoupper' ) ) {
27 mb_internal_encoding('UTF-8');
30 /**
31 * a fake language converter
33 * @ingroup Language
35 class FakeConverter {
36 var $mLang;
37 function FakeConverter($langobj) {$this->mLang = $langobj;}
38 function convert($t, $i) {return $t;}
39 function parserConvert($t, $p) {return $t;}
40 function getVariants() { return array( $this->mLang->getCode() ); }
41 function getPreferredVariant() {return $this->mLang->getCode(); }
42 function findVariantLink(&$l, &$n, $forTemplate = false) {}
43 function getExtraHashOptions() {return '';}
44 function getParsedTitle() {return '';}
45 function markNoConversion($text, $noParse=false) {return $text;}
46 function convertCategoryKey( $key ) {return $key; }
47 function convertLinkToAllVariants($text){ return array( $this->mLang->getCode() => $text); }
48 function armourMath($text){ return $text; }
51 /**
52 * Internationalisation code
53 * @ingroup Language
55 class Language {
56 var $mConverter, $mVariants, $mCode, $mLoaded = false;
57 var $mMagicExtensions = array(), $mMagicHookDone = false;
59 static public $mLocalisationKeys = array(
60 'fallback', 'namespaceNames', 'mathNames', 'bookstoreList',
61 'magicWords', 'messages', 'rtl', 'digitTransformTable',
62 'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
63 'defaultUserOptionOverrides', 'linkTrail', 'namespaceAliases',
64 'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
65 'defaultDateFormat', 'extraUserToggles', 'specialPageAliases',
66 'imageFiles'
69 static public $mMergeableMapKeys = array( 'messages', 'namespaceNames', 'mathNames',
70 'dateFormats', 'defaultUserOptionOverrides', 'magicWords', 'imageFiles' );
72 static public $mMergeableListKeys = array( 'extraUserToggles' );
74 static public $mMergeableAliasListKeys = array( 'specialPageAliases' );
76 static public $mLocalisationCache = array();
77 static public $mLangObjCache = array();
79 static public $mWeekdayMsgs = array(
80 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
81 'friday', 'saturday'
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',
91 'december'
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',
96 'december-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
136 static function factory( $code ) {
137 if ( !isset( self::$mLangObjCache[$code] ) ) {
138 if( count( self::$mLangObjCache ) > 10 ) {
139 // Don't keep a billion objects around, that's stupid.
140 self::$mLangObjCache = array();
142 self::$mLangObjCache[$code] = self::newFromCode( $code );
144 return self::$mLangObjCache[$code];
148 * Create a language object for a given language code
150 protected static function newFromCode( $code ) {
151 global $IP;
152 static $recursionLevel = 0;
153 if ( $code == 'en' ) {
154 $class = 'Language';
155 } else {
156 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
157 // Preload base classes to work around APC/PHP5 bug
158 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
159 include_once("$IP/languages/classes/$class.deps.php");
161 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
162 include_once("$IP/languages/classes/$class.php");
166 if ( $recursionLevel > 5 ) {
167 throw new MWException( "Language fallback loop detected when creating class $class\n" );
170 if( ! class_exists( $class ) ) {
171 $fallback = Language::getFallbackFor( $code );
172 ++$recursionLevel;
173 $lang = Language::newFromCode( $fallback );
174 --$recursionLevel;
175 $lang->setCode( $code );
176 } else {
177 $lang = new $class;
179 return $lang;
182 function __construct() {
183 $this->mConverter = new FakeConverter($this);
184 // Set the code to the name of the descendant
185 if ( get_class( $this ) == 'Language' ) {
186 $this->mCode = 'en';
187 } else {
188 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
193 * Reduce memory usage
195 function __destruct() {
196 foreach ( $this as $name => $value ) {
197 unset( $this->$name );
202 * Hook which will be called if this is the content language.
203 * Descendants can use this to register hook functions or modify globals
205 function initContLang() {}
208 * @deprecated Use User::getDefaultOptions()
209 * @return array
211 function getDefaultUserOptions() {
212 wfDeprecated( __METHOD__ );
213 return User::getDefaultOptions();
216 function getFallbackLanguageCode() {
217 return self::getFallbackFor( $this->mCode );
221 * Exports $wgBookstoreListEn
222 * @return array
224 function getBookstoreList() {
225 $this->load();
226 return $this->bookstoreList;
230 * @return array
232 function getNamespaces() {
233 $this->load();
234 return $this->namespaceNames;
238 * A convenience function that returns the same thing as
239 * getNamespaces() except with the array values changed to ' '
240 * where it found '_', useful for producing output to be displayed
241 * e.g. in <select> forms.
243 * @return array
245 function getFormattedNamespaces() {
246 $ns = $this->getNamespaces();
247 foreach($ns as $k => $v) {
248 $ns[$k] = strtr($v, '_', ' ');
250 return $ns;
254 * Get a namespace value by key
255 * <code>
256 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
257 * echo $mw_ns; // prints 'MediaWiki'
258 * </code>
260 * @param $index Int: the array key of the namespace to return
261 * @return mixed, string if the namespace value exists, otherwise false
263 function getNsText( $index ) {
264 $ns = $this->getNamespaces();
265 return isset( $ns[$index] ) ? $ns[$index] : false;
269 * A convenience function that returns the same thing as
270 * getNsText() except with '_' changed to ' ', useful for
271 * producing output.
273 * @return array
275 function getFormattedNsText( $index ) {
276 $ns = $this->getNsText( $index );
277 return strtr($ns, '_', ' ');
281 * Get a namespace key by value, case insensitive.
282 * Only matches namespace names for the current language, not the
283 * canonical ones defined in Namespace.php.
285 * @param $text String
286 * @return mixed An integer if $text is a valid value otherwise false
288 function getLocalNsIndex( $text ) {
289 $this->load();
290 $lctext = $this->lc($text);
291 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
295 * Get a namespace key by value, case insensitive. Canonical namespace
296 * names override custom ones defined for the current language.
298 * @param $text String
299 * @return mixed An integer if $text is a valid value otherwise false
301 function getNsIndex( $text ) {
302 $this->load();
303 $lctext = $this->lc($text);
304 if( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
305 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
309 * short names for language variants used for language conversion links.
311 * @param $code String
312 * @return string
314 function getVariantname( $code ) {
315 return $this->getMessageFromDB( "variantname-$code" );
318 function specialPage( $name ) {
319 $aliases = $this->getSpecialPageAliases();
320 if ( isset( $aliases[$name][0] ) ) {
321 $name = $aliases[$name][0];
323 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
326 function getQuickbarSettings() {
327 return array(
328 $this->getMessage( 'qbsettings-none' ),
329 $this->getMessage( 'qbsettings-fixedleft' ),
330 $this->getMessage( 'qbsettings-fixedright' ),
331 $this->getMessage( 'qbsettings-floatingleft' ),
332 $this->getMessage( 'qbsettings-floatingright' )
336 function getMathNames() {
337 $this->load();
338 return $this->mathNames;
341 function getDatePreferences() {
342 $this->load();
343 return $this->datePreferences;
346 function getDateFormats() {
347 $this->load();
348 return $this->dateFormats;
351 function getDefaultDateFormat() {
352 $this->load();
353 return $this->defaultDateFormat;
356 function getDatePreferenceMigrationMap() {
357 $this->load();
358 return $this->datePreferenceMigrationMap;
361 function getImageFile( $image ) {
362 $this->load();
363 return $this->imageFiles[$image];
366 function getDefaultUserOptionOverrides() {
367 $this->load();
368 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
369 if (is_array($this->defaultUserOptionOverrides)) {
370 return $this->defaultUserOptionOverrides;
371 } else {
372 return array();
376 function getExtraUserToggles() {
377 $this->load();
378 return $this->extraUserToggles;
381 function getUserToggle( $tog ) {
382 return $this->getMessageFromDB( "tog-$tog" );
386 * Get language names, indexed by code.
387 * If $customisedOnly is true, only returns codes with a messages file
389 public static function getLanguageNames( $customisedOnly = false ) {
390 global $wgLanguageNames, $wgExtraLanguageNames;
391 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
392 if ( !$customisedOnly ) {
393 return $allNames;
396 global $IP;
397 $names = array();
398 $dir = opendir( "$IP/languages/messages" );
399 while( false !== ( $file = readdir( $dir ) ) ) {
400 $m = array();
401 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
402 $code = str_replace( '_', '-', strtolower( $m[1] ) );
403 if ( isset( $allNames[$code] ) ) {
404 $names[$code] = $allNames[$code];
408 closedir( $dir );
409 return $names;
413 * Get a message from the MediaWiki namespace.
415 * @param $msg String: message name
416 * @return string
418 function getMessageFromDB( $msg ) {
419 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
423 * Get a language name
425 * @param $code String language code
426 * @return $localized boolean gets the localized language name
428 function getLanguageName( $code, $localized = false ) {
429 $names = self::getLanguageNames();
430 if ( !array_key_exists( $code, $names ) ) {
431 return '';
433 if( $localized ) {
434 $languageNames = array();
435 wfRunHooks( 'LanguageGetLocalizedLanguageNames', array( &$languageNames, $this->getCode() ) );
436 return isset( $languageNames[$code] ) ? $languageNames[$code] : $names[$code];
437 } else {
438 return $names[$code];
442 function getLanguageNameLocalized( $code ) {
443 return self::getLanguageName( $code, true );
446 function getMonthName( $key ) {
447 return $this->getMessageFromDB( self::$mMonthMsgs[$key-1] );
450 function getMonthNameGen( $key ) {
451 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key-1] );
454 function getMonthAbbreviation( $key ) {
455 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key-1] );
458 function getWeekdayName( $key ) {
459 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key-1] );
462 function getWeekdayAbbreviation( $key ) {
463 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key-1] );
466 function getIranianCalendarMonthName( $key ) {
467 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key-1] );
470 function getHebrewCalendarMonthName( $key ) {
471 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key-1] );
474 function getHebrewCalendarMonthNameGen( $key ) {
475 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key-1] );
478 function getHijriCalendarMonthName( $key ) {
479 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key-1] );
483 * Used by date() and time() to adjust the time output.
485 * @param $ts Int the time in date('YmdHis') format
486 * @param $tz Mixed: adjust the time by this amount (default false, mean we
487 * get user timecorrection setting)
488 * @return int
490 function userAdjust( $ts, $tz = false ) {
491 global $wgUser, $wgLocalTZoffset;
493 if ( $tz === false ) {
494 $tz = $wgUser->getOption( 'timecorrection' );
497 $data = explode( '|', $tz, 3 );
499 if ( $data[0] == 'ZoneInfo' ) {
500 if ( function_exists( 'timezone_open' ) && @timezone_open( $data[2] ) !== false ) {
501 $date = date_create( $ts, timezone_open( 'UTC' ) );
502 date_timezone_set( $date, timezone_open( $data[2] ) );
503 $date = date_format( $date, 'YmdHis' );
504 return $date;
506 # Unrecognized timezone, default to 'Offset' with the stored offset.
507 $data[0] = 'Offset';
510 $minDiff = 0;
511 if ( $data[0] == 'System' || $tz == '' ) {
512 # Global offset in minutes.
513 if( isset($wgLocalTZoffset) ) $minDiff = $wgLocalTZoffset;
514 } else if ( $data[0] == 'Offset' ) {
515 $minDiff = intval( $data[1] );
516 } else {
517 $data = explode( ':', $tz );
518 if( count( $data ) == 2 ) {
519 $data[0] = intval( $data[0] );
520 $data[1] = intval( $data[1] );
521 $minDiff = abs( $data[0] ) * 60 + $data[1];
522 if ( $data[0] < 0 ) $minDiff = -$minDiff;
523 } else {
524 $minDiff = intval( $data[0] ) * 60;
528 # No difference ? Return time unchanged
529 if ( 0 == $minDiff ) return $ts;
531 wfSuppressWarnings(); // E_STRICT system time bitching
532 # Generate an adjusted date; take advantage of the fact that mktime
533 # will normalize out-of-range values so we don't have to split $minDiff
534 # into hours and minutes.
535 $t = mktime( (
536 (int)substr( $ts, 8, 2) ), # Hours
537 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
538 (int)substr( $ts, 12, 2 ), # Seconds
539 (int)substr( $ts, 4, 2 ), # Month
540 (int)substr( $ts, 6, 2 ), # Day
541 (int)substr( $ts, 0, 4 ) ); #Year
543 $date = date( 'YmdHis', $t );
544 wfRestoreWarnings();
546 return $date;
550 * This is a workalike of PHP's date() function, but with better
551 * internationalisation, a reduced set of format characters, and a better
552 * escaping format.
554 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
555 * PHP manual for definitions. "o" format character is supported since
556 * PHP 5.1.0, previous versions return literal o.
557 * There are a number of extensions, which start with "x":
559 * xn Do not translate digits of the next numeric format character
560 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
561 * xr Use roman numerals for the next numeric format character
562 * xh Use hebrew numerals for the next numeric format character
563 * xx Literal x
564 * xg Genitive month name
566 * xij j (day number) in Iranian calendar
567 * xiF F (month name) in Iranian calendar
568 * xin n (month number) in Iranian calendar
569 * xiY Y (full year) in Iranian calendar
571 * xjj j (day number) in Hebrew calendar
572 * xjF F (month name) in Hebrew calendar
573 * xjt t (days in month) in Hebrew calendar
574 * xjx xg (genitive month name) in Hebrew calendar
575 * xjn n (month number) in Hebrew calendar
576 * xjY Y (full year) in Hebrew calendar
578 * xmj j (day number) in Hijri calendar
579 * xmF F (month name) in Hijri calendar
580 * xmn n (month number) in Hijri calendar
581 * xmY Y (full year) in Hijri calendar
583 * xkY Y (full year) in Thai solar calendar. Months and days are
584 * identical to the Gregorian calendar
586 * Characters enclosed in double quotes will be considered literal (with
587 * the quotes themselves removed). Unmatched quotes will be considered
588 * literal quotes. Example:
590 * "The month is" F => The month is January
591 * i's" => 20'11"
593 * Backslash escaping is also supported.
595 * Input timestamp is assumed to be pre-normalized to the desired local
596 * time zone, if any.
598 * @param $format String
599 * @param $ts String: 14-character timestamp
600 * YYYYMMDDHHMMSS
601 * 01234567890123
602 * @todo emulation of "o" format character for PHP pre 5.1.0
603 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
605 function sprintfDate( $format, $ts ) {
606 $s = '';
607 $raw = false;
608 $roman = false;
609 $hebrewNum = false;
610 $unix = false;
611 $rawToggle = false;
612 $iranian = false;
613 $hebrew = false;
614 $hijri = false;
615 $thai = false;
616 for ( $p = 0; $p < strlen( $format ); $p++ ) {
617 $num = false;
618 $code = $format[$p];
619 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
620 $code .= $format[++$p];
623 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' ) && $p < strlen( $format ) - 1 ) {
624 $code .= $format[++$p];
627 switch ( $code ) {
628 case 'xx':
629 $s .= 'x';
630 break;
631 case 'xn':
632 $raw = true;
633 break;
634 case 'xN':
635 $rawToggle = !$rawToggle;
636 break;
637 case 'xr':
638 $roman = true;
639 break;
640 case 'xh':
641 $hebrewNum = true;
642 break;
643 case 'xg':
644 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
645 break;
646 case 'xjx':
647 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
648 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
649 break;
650 case 'd':
651 $num = substr( $ts, 6, 2 );
652 break;
653 case 'D':
654 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
655 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
656 break;
657 case 'j':
658 $num = intval( substr( $ts, 6, 2 ) );
659 break;
660 case 'xij':
661 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
662 $num = $iranian[2];
663 break;
664 case 'xmj':
665 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
666 $num = $hijri[2];
667 break;
668 case 'xjj':
669 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
670 $num = $hebrew[2];
671 break;
672 case 'l':
673 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
674 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
675 break;
676 case 'N':
677 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
678 $w = gmdate( 'w', $unix );
679 $num = $w ? $w : 7;
680 break;
681 case 'w':
682 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
683 $num = gmdate( 'w', $unix );
684 break;
685 case 'z':
686 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
687 $num = gmdate( 'z', $unix );
688 break;
689 case 'W':
690 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
691 $num = gmdate( 'W', $unix );
692 break;
693 case 'F':
694 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
695 break;
696 case 'xiF':
697 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
698 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
699 break;
700 case 'xmF':
701 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
702 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
703 break;
704 case 'xjF':
705 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
706 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
707 break;
708 case 'm':
709 $num = substr( $ts, 4, 2 );
710 break;
711 case 'M':
712 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
713 break;
714 case 'n':
715 $num = intval( substr( $ts, 4, 2 ) );
716 break;
717 case 'xin':
718 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
719 $num = $iranian[1];
720 break;
721 case 'xmn':
722 if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
723 $num = $hijri[1];
724 break;
725 case 'xjn':
726 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
727 $num = $hebrew[1];
728 break;
729 case 't':
730 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
731 $num = gmdate( 't', $unix );
732 break;
733 case 'xjt':
734 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
735 $num = $hebrew[3];
736 break;
737 case 'L':
738 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
739 $num = gmdate( 'L', $unix );
740 break;
741 # 'o' is supported since PHP 5.1.0
742 # return literal if not supported
743 # TODO: emulation for pre 5.1.0 versions
744 case 'o':
745 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
746 if ( version_compare(PHP_VERSION, '5.1.0') === 1 )
747 $num = date( 'o', $unix );
748 else
749 $s .= 'o';
750 break;
751 case 'Y':
752 $num = substr( $ts, 0, 4 );
753 break;
754 case 'xiY':
755 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
756 $num = $iranian[0];
757 break;
758 case 'xmY':
759 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
760 $num = $hijri[0];
761 break;
762 case 'xjY':
763 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
764 $num = $hebrew[0];
765 break;
766 case 'xkY':
767 if ( !$thai ) $thai = self::tsToThai( $ts );
768 $num = $thai[0];
769 break;
770 case 'y':
771 $num = substr( $ts, 2, 2 );
772 break;
773 case 'a':
774 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
775 break;
776 case 'A':
777 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
778 break;
779 case 'g':
780 $h = substr( $ts, 8, 2 );
781 $num = $h % 12 ? $h % 12 : 12;
782 break;
783 case 'G':
784 $num = intval( substr( $ts, 8, 2 ) );
785 break;
786 case 'h':
787 $h = substr( $ts, 8, 2 );
788 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
789 break;
790 case 'H':
791 $num = substr( $ts, 8, 2 );
792 break;
793 case 'i':
794 $num = substr( $ts, 10, 2 );
795 break;
796 case 's':
797 $num = substr( $ts, 12, 2 );
798 break;
799 case 'c':
800 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
801 $s .= gmdate( 'c', $unix );
802 break;
803 case 'r':
804 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
805 $s .= gmdate( 'r', $unix );
806 break;
807 case 'U':
808 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
809 $num = $unix;
810 break;
811 case '\\':
812 # Backslash escaping
813 if ( $p < strlen( $format ) - 1 ) {
814 $s .= $format[++$p];
815 } else {
816 $s .= '\\';
818 break;
819 case '"':
820 # Quoted literal
821 if ( $p < strlen( $format ) - 1 ) {
822 $endQuote = strpos( $format, '"', $p + 1 );
823 if ( $endQuote === false ) {
824 # No terminating quote, assume literal "
825 $s .= '"';
826 } else {
827 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
828 $p = $endQuote;
830 } else {
831 # Quote at end of string, assume literal "
832 $s .= '"';
834 break;
835 default:
836 $s .= $format[$p];
838 if ( $num !== false ) {
839 if ( $rawToggle || $raw ) {
840 $s .= $num;
841 $raw = false;
842 } elseif ( $roman ) {
843 $s .= self::romanNumeral( $num );
844 $roman = false;
845 } elseif( $hebrewNum ) {
846 $s .= self::hebrewNumeral( $num );
847 $hebrewNum = false;
848 } else {
849 $s .= $this->formatNum( $num, true );
851 $num = false;
854 return $s;
857 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
858 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
860 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
861 * Gregorian dates to Iranian dates. Originally written in C, it
862 * is released under the terms of GNU Lesser General Public
863 * License. Conversion to PHP was performed by Niklas Laxström.
865 * Link: http://www.farsiweb.info/jalali/jalali.c
867 private static function tsToIranian( $ts ) {
868 $gy = substr( $ts, 0, 4 ) -1600;
869 $gm = substr( $ts, 4, 2 ) -1;
870 $gd = substr( $ts, 6, 2 ) -1;
872 # Days passed from the beginning (including leap years)
873 $gDayNo = 365*$gy
874 + floor(($gy+3) / 4)
875 - floor(($gy+99) / 100)
876 + floor(($gy+399) / 400);
879 // Add days of the past months of this year
880 for( $i = 0; $i < $gm; $i++ ) {
881 $gDayNo += self::$GREG_DAYS[$i];
884 // Leap years
885 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
886 $gDayNo++;
889 // Days passed in current month
890 $gDayNo += $gd;
892 $jDayNo = $gDayNo - 79;
894 $jNp = floor($jDayNo / 12053);
895 $jDayNo %= 12053;
897 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
898 $jDayNo %= 1461;
900 if ( $jDayNo >= 366 ) {
901 $jy += floor(($jDayNo-1)/365);
902 $jDayNo = floor(($jDayNo-1)%365);
905 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
906 $jDayNo -= self::$IRANIAN_DAYS[$i];
909 $jm= $i+1;
910 $jd= $jDayNo+1;
912 return array($jy, $jm, $jd);
915 * Converting Gregorian dates to Hijri dates.
917 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
919 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
921 private static function tsToHijri ( $ts ) {
922 $year = substr( $ts, 0, 4 );
923 $month = substr( $ts, 4, 2 );
924 $day = substr( $ts, 6, 2 );
926 $zyr = $year;
927 $zd=$day;
928 $zm=$month;
929 $zy=$zyr;
933 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
937 $zjd=(int)((1461*($zy + 4800 + (int)( ($zm-14) /12) ))/4) + (int)((367*($zm-2-12*((int)(($zm-14)/12))))/12)-(int)((3*(int)(( ($zy+4900+(int)(($zm-14)/12))/100)))/4)+$zd-32075;
939 else
941 $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
944 $zl=$zjd-1948440+10632;
945 $zn=(int)(($zl-1)/10631);
946 $zl=$zl-10631*$zn+354;
947 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
948 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
949 $zm=(int)((24*$zl)/709);
950 $zd=$zl-(int)((709*$zm)/24);
951 $zy=30*$zn+$zj-30;
953 return array ($zy, $zm, $zd);
957 * Converting Gregorian dates to Hebrew dates.
959 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
960 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
961 * to translate the relevant functions into PHP and release them under
962 * GNU GPL.
964 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
965 * and Adar II is 14. In a non-leap year, Adar is 6.
967 private static function tsToHebrew( $ts ) {
968 # Parse date
969 $year = substr( $ts, 0, 4 );
970 $month = substr( $ts, 4, 2 );
971 $day = substr( $ts, 6, 2 );
973 # Calculate Hebrew year
974 $hebrewYear = $year + 3760;
976 # Month number when September = 1, August = 12
977 $month += 4;
978 if( $month > 12 ) {
979 # Next year
980 $month -= 12;
981 $year++;
982 $hebrewYear++;
985 # Calculate day of year from 1 September
986 $dayOfYear = $day;
987 for( $i = 1; $i < $month; $i++ ) {
988 if( $i == 6 ) {
989 # February
990 $dayOfYear += 28;
991 # Check if the year is leap
992 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
993 $dayOfYear++;
995 } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
996 $dayOfYear += 30;
997 } else {
998 $dayOfYear += 31;
1002 # Calculate the start of the Hebrew year
1003 $start = self::hebrewYearStart( $hebrewYear );
1005 # Calculate next year's start
1006 if( $dayOfYear <= $start ) {
1007 # Day is before the start of the year - it is the previous year
1008 # Next year's start
1009 $nextStart = $start;
1010 # Previous year
1011 $year--;
1012 $hebrewYear--;
1013 # Add days since previous year's 1 September
1014 $dayOfYear += 365;
1015 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1016 # Leap year
1017 $dayOfYear++;
1019 # Start of the new (previous) year
1020 $start = self::hebrewYearStart( $hebrewYear );
1021 } else {
1022 # Next year's start
1023 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1026 # Calculate Hebrew day of year
1027 $hebrewDayOfYear = $dayOfYear - $start;
1029 # Difference between year's days
1030 $diff = $nextStart - $start;
1031 # Add 12 (or 13 for leap years) days to ignore the difference between
1032 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1033 # difference is only about the year type
1034 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1035 $diff += 13;
1036 } else {
1037 $diff += 12;
1040 # Check the year pattern, and is leap year
1041 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1042 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1043 # and non-leap years
1044 $yearPattern = $diff % 30;
1045 # Check if leap year
1046 $isLeap = $diff >= 30;
1048 # Calculate day in the month from number of day in the Hebrew year
1049 # Don't check Adar - if the day is not in Adar, we will stop before;
1050 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1051 $hebrewDay = $hebrewDayOfYear;
1052 $hebrewMonth = 1;
1053 $days = 0;
1054 while( $hebrewMonth <= 12 ) {
1055 # Calculate days in this month
1056 if( $isLeap && $hebrewMonth == 6 ) {
1057 # Adar in a leap year
1058 if( $isLeap ) {
1059 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1060 $days = 30;
1061 if( $hebrewDay <= $days ) {
1062 # Day in Adar I
1063 $hebrewMonth = 13;
1064 } else {
1065 # Subtract the days of Adar I
1066 $hebrewDay -= $days;
1067 # Try Adar II
1068 $days = 29;
1069 if( $hebrewDay <= $days ) {
1070 # Day in Adar II
1071 $hebrewMonth = 14;
1075 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1076 # Cheshvan in a complete year (otherwise as the rule below)
1077 $days = 30;
1078 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1079 # Kislev in an incomplete year (otherwise as the rule below)
1080 $days = 29;
1081 } else {
1082 # Odd months have 30 days, even have 29
1083 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1085 if( $hebrewDay <= $days ) {
1086 # In the current month
1087 break;
1088 } else {
1089 # Subtract the days of the current month
1090 $hebrewDay -= $days;
1091 # Try in the next month
1092 $hebrewMonth++;
1096 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1100 * This calculates the Hebrew year start, as days since 1 September.
1101 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1102 * Used for Hebrew date.
1104 private static function hebrewYearStart( $year ) {
1105 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1106 $b = intval( ( $year - 1 ) % 4 );
1107 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1108 if( $m < 0 ) {
1109 $m--;
1111 $Mar = intval( $m );
1112 if( $m < 0 ) {
1113 $m++;
1115 $m -= $Mar;
1117 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1118 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1119 $Mar++;
1120 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1121 $Mar += 2;
1122 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1123 $Mar++;
1126 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1127 return $Mar;
1131 * Algorithm to convert Gregorian dates to Thai solar dates.
1133 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1135 * @param $ts String: 14-character timestamp
1136 * @return array converted year, month, day
1138 private static function tsToThai( $ts ) {
1139 $gy = substr( $ts, 0, 4 );
1140 $gm = substr( $ts, 4, 2 );
1141 $gd = substr( $ts, 6, 2 );
1143 # Add 543 years to the Gregorian calendar
1144 # Months and days are identical
1145 $gy_thai = $gy + 543;
1147 return array( $gy_thai, $gm, $gd );
1152 * Roman number formatting up to 3000
1154 static function romanNumeral( $num ) {
1155 static $table = array(
1156 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1157 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1158 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1159 array( '', 'M', 'MM', 'MMM' )
1162 $num = intval( $num );
1163 if ( $num > 3000 || $num <= 0 ) {
1164 return $num;
1167 $s = '';
1168 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1169 if ( $num >= $pow10 ) {
1170 $s .= $table[$i][floor($num / $pow10)];
1172 $num = $num % $pow10;
1174 return $s;
1178 * Hebrew Gematria number formatting up to 9999
1180 static function hebrewNumeral( $num ) {
1181 static $table = array(
1182 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1183 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1184 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1185 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1188 $num = intval( $num );
1189 if ( $num > 9999 || $num <= 0 ) {
1190 return $num;
1193 $s = '';
1194 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1195 if ( $num >= $pow10 ) {
1196 if ( $num == 15 || $num == 16 ) {
1197 $s .= $table[0][9] . $table[0][$num - 9];
1198 $num = 0;
1199 } else {
1200 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1201 if( $pow10 == 1000 ) {
1202 $s .= "'";
1206 $num = $num % $pow10;
1208 if( strlen( $s ) == 2 ) {
1209 $str = $s . "'";
1210 } else {
1211 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1212 $str .= substr( $s, strlen( $s ) - 2, 2 );
1214 $start = substr( $str, 0, strlen( $str ) - 2 );
1215 $end = substr( $str, strlen( $str ) - 2 );
1216 switch( $end ) {
1217 case 'כ':
1218 $str = $start . 'ך';
1219 break;
1220 case 'מ':
1221 $str = $start . 'ם';
1222 break;
1223 case 'נ':
1224 $str = $start . 'ן';
1225 break;
1226 case 'פ':
1227 $str = $start . 'ף';
1228 break;
1229 case 'צ':
1230 $str = $start . 'ץ';
1231 break;
1233 return $str;
1237 * This is meant to be used by time(), date(), and timeanddate() to get
1238 * the date preference they're supposed to use, it should be used in
1239 * all children.
1241 *<code>
1242 * function timeanddate([...], $format = true) {
1243 * $datePreference = $this->dateFormat($format);
1244 * [...]
1246 *</code>
1248 * @param $usePrefs Mixed: if true, the user's preference is used
1249 * if false, the site/language default is used
1250 * if int/string, assumed to be a format.
1251 * @return string
1253 function dateFormat( $usePrefs = true ) {
1254 global $wgUser;
1256 if( is_bool( $usePrefs ) ) {
1257 if( $usePrefs ) {
1258 $datePreference = $wgUser->getDatePreference();
1259 } else {
1260 $options = User::getDefaultOptions();
1261 $datePreference = (string)$options['date'];
1263 } else {
1264 $datePreference = (string)$usePrefs;
1267 // return int
1268 if( $datePreference == '' ) {
1269 return 'default';
1272 return $datePreference;
1276 * @param $ts Mixed: the time format which needs to be turned into a
1277 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1278 * @param $adj Bool: whether to adjust the time output according to the
1279 * user configured offset ($timecorrection)
1280 * @param $format Mixed: true to use user's date format preference
1281 * @param $timecorrection String: the time offset as returned by
1282 * validateTimeZone() in Special:Preferences
1283 * @return string
1285 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1286 $this->load();
1287 if ( $adj ) {
1288 $ts = $this->userAdjust( $ts, $timecorrection );
1291 $pref = $this->dateFormat( $format );
1292 if( $pref == 'default' || !isset( $this->dateFormats["$pref date"] ) ) {
1293 $pref = $this->defaultDateFormat;
1295 return $this->sprintfDate( $this->dateFormats["$pref date"], $ts );
1299 * @param $ts Mixed: the time format which needs to be turned into a
1300 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1301 * @param $adj Bool: whether to adjust the time output according to the
1302 * user configured offset ($timecorrection)
1303 * @param $format Mixed: true to use user's date format preference
1304 * @param $timecorrection String: the time offset as returned by
1305 * validateTimeZone() in Special:Preferences
1306 * @return string
1308 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1309 $this->load();
1310 if ( $adj ) {
1311 $ts = $this->userAdjust( $ts, $timecorrection );
1314 $pref = $this->dateFormat( $format );
1315 if( $pref == 'default' || !isset( $this->dateFormats["$pref time"] ) ) {
1316 $pref = $this->defaultDateFormat;
1318 return $this->sprintfDate( $this->dateFormats["$pref time"], $ts );
1322 * @param $ts Mixed: the time format which needs to be turned into a
1323 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1324 * @param $adj Bool: whether to adjust the time output according to the
1325 * user configured offset ($timecorrection)
1326 * @param $format Mixed: what format to return, if it's false output the
1327 * default one (default true)
1328 * @param $timecorrection String: the time offset as returned by
1329 * validateTimeZone() in Special:Preferences
1330 * @return string
1332 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1333 $this->load();
1335 $ts = wfTimestamp( TS_MW, $ts );
1337 if ( $adj ) {
1338 $ts = $this->userAdjust( $ts, $timecorrection );
1341 $pref = $this->dateFormat( $format );
1342 if( $pref == 'default' || !isset( $this->dateFormats["$pref both"] ) ) {
1343 $pref = $this->defaultDateFormat;
1346 return $this->sprintfDate( $this->dateFormats["$pref both"], $ts );
1349 function getMessage( $key ) {
1350 $this->load();
1351 return isset( $this->messages[$key] ) ? $this->messages[$key] : null;
1354 function getAllMessages() {
1355 $this->load();
1356 return $this->messages;
1359 function iconv( $in, $out, $string ) {
1360 # For most languages, this is a wrapper for iconv
1361 return iconv( $in, $out . '//IGNORE', $string );
1364 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1365 function ucwordbreaksCallbackAscii($matches){
1366 return $this->ucfirst($matches[1]);
1369 function ucwordbreaksCallbackMB($matches){
1370 return mb_strtoupper($matches[0]);
1373 function ucCallback($matches){
1374 list( $wikiUpperChars ) = self::getCaseMaps();
1375 return strtr( $matches[1], $wikiUpperChars );
1378 function lcCallback($matches){
1379 list( , $wikiLowerChars ) = self::getCaseMaps();
1380 return strtr( $matches[1], $wikiLowerChars );
1383 function ucwordsCallbackMB($matches){
1384 return mb_strtoupper($matches[0]);
1387 function ucwordsCallbackWiki($matches){
1388 list( $wikiUpperChars ) = self::getCaseMaps();
1389 return strtr( $matches[0], $wikiUpperChars );
1392 function ucfirst( $str ) {
1393 if ( empty($str) ) return $str;
1394 if ( ord($str[0]) < 128 ) return ucfirst($str);
1395 else return self::uc($str,true); // fall back to more complex logic in case of multibyte strings
1398 function uc( $str, $first = false ) {
1399 if ( function_exists( 'mb_strtoupper' ) ) {
1400 if ( $first ) {
1401 if ( self::isMultibyte( $str ) ) {
1402 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1403 } else {
1404 return ucfirst( $str );
1406 } else {
1407 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1409 } else {
1410 if ( self::isMultibyte( $str ) ) {
1411 list( $wikiUpperChars ) = $this->getCaseMaps();
1412 $x = $first ? '^' : '';
1413 return preg_replace_callback(
1414 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1415 array($this,"ucCallback"),
1416 $str
1418 } else {
1419 return $first ? ucfirst( $str ) : strtoupper( $str );
1424 function lcfirst( $str ) {
1425 if ( empty($str) ) return $str;
1426 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1427 // editing string in place = cool
1428 $str[0]=strtolower($str[0]);
1429 return $str;
1431 else return self::lc( $str, true );
1434 function lc( $str, $first = false ) {
1435 if ( function_exists( 'mb_strtolower' ) )
1436 if ( $first )
1437 if ( self::isMultibyte( $str ) )
1438 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1439 else
1440 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1441 else
1442 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1443 else
1444 if ( self::isMultibyte( $str ) ) {
1445 list( , $wikiLowerChars ) = self::getCaseMaps();
1446 $x = $first ? '^' : '';
1447 return preg_replace_callback(
1448 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1449 array($this,"lcCallback"),
1450 $str
1452 } else
1453 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1456 function isMultibyte( $str ) {
1457 return (bool)preg_match( '/[\x80-\xff]/', $str );
1460 function ucwords($str) {
1461 if ( self::isMultibyte( $str ) ) {
1462 $str = self::lc($str);
1464 // regexp to find first letter in each word (i.e. after each space)
1465 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1467 // function to use to capitalize a single char
1468 if ( function_exists( 'mb_strtoupper' ) )
1469 return preg_replace_callback(
1470 $replaceRegexp,
1471 array($this,"ucwordsCallbackMB"),
1472 $str
1474 else
1475 return preg_replace_callback(
1476 $replaceRegexp,
1477 array($this,"ucwordsCallbackWiki"),
1478 $str
1481 else
1482 return ucwords( strtolower( $str ) );
1485 # capitalize words at word breaks
1486 function ucwordbreaks($str){
1487 if (self::isMultibyte( $str ) ) {
1488 $str = self::lc($str);
1490 // since \b doesn't work for UTF-8, we explicitely define word break chars
1491 $breaks= "[ \-\(\)\}\{\.,\?!]";
1493 // find first letter after word break
1494 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1496 if ( function_exists( 'mb_strtoupper' ) )
1497 return preg_replace_callback(
1498 $replaceRegexp,
1499 array($this,"ucwordbreaksCallbackMB"),
1500 $str
1502 else
1503 return preg_replace_callback(
1504 $replaceRegexp,
1505 array($this,"ucwordsCallbackWiki"),
1506 $str
1509 else
1510 return preg_replace_callback(
1511 '/\b([\w\x80-\xff]+)\b/',
1512 array($this,"ucwordbreaksCallbackAscii"),
1513 $str );
1517 * Return a case-folded representation of $s
1519 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1520 * and $s2 are the same except for the case of their characters. It is not
1521 * necessary for the value returned to make sense when displayed.
1523 * Do *not* perform any other normalisation in this function. If a caller
1524 * uses this function when it should be using a more general normalisation
1525 * function, then fix the caller.
1527 function caseFold( $s ) {
1528 return $this->uc( $s );
1531 function checkTitleEncoding( $s ) {
1532 if( is_array( $s ) ) {
1533 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1535 # Check for non-UTF-8 URLs
1536 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1537 if(!$ishigh) return $s;
1539 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1540 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1541 if( $isutf8 ) return $s;
1543 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1546 function fallback8bitEncoding() {
1547 $this->load();
1548 return $this->fallback8bitEncoding;
1552 * Some languages have special punctuation to strip out
1553 * or characters which need to be converted for MySQL's
1554 * indexing to grok it correctly. Make such changes here.
1556 * @param $string String
1557 * @return String
1559 function stripForSearch( $string ) {
1560 global $wgDBtype;
1561 if ( $wgDBtype != 'mysql' ) {
1562 return $string;
1566 wfProfileIn( __METHOD__ );
1568 // MySQL fulltext index doesn't grok utf-8, so we
1569 // need to fold cases and convert to hex
1570 $out = preg_replace_callback(
1571 "/([\\xc0-\\xff][\\x80-\\xbf]*)/",
1572 array( $this, 'stripForSearchCallback' ),
1573 $this->lc( $string ) );
1575 // And to add insult to injury, the default indexing
1576 // ignores short words... Pad them so we can pass them
1577 // through without reconfiguring the server...
1578 $minLength = $this->minSearchLength();
1579 if( $minLength > 1 ) {
1580 $n = $minLength-1;
1581 $out = preg_replace(
1582 "/\b(\w{1,$n})\b/",
1583 "$1U800",
1584 $out );
1587 // Periods within things like hostnames and IP addresses
1588 // are also important -- we want a search for "example.com"
1589 // or "192.168.1.1" to work sanely.
1591 // MySQL's search seems to ignore them, so you'd match on
1592 // "example.wikipedia.com" and "192.168.83.1" as well.
1593 $out = preg_replace(
1594 "/(\w)\.(\w|\*)/u",
1595 "$1U82e$2",
1596 $out );
1598 wfProfileOut( __METHOD__ );
1599 return $out;
1603 * Armor a case-folded UTF-8 string to get through MySQL's
1604 * fulltext search without being mucked up by funny charset
1605 * settings or anything else of the sort.
1607 protected function stripForSearchCallback( $matches ) {
1608 return 'U8' . bin2hex( $matches[1] );
1612 * Check MySQL server's ft_min_word_len setting so we know
1613 * if we need to pad short words...
1615 protected function minSearchLength() {
1616 if( !isset( $this->minSearchLength ) ) {
1617 $sql = "show global variables like 'ft\\_min\\_word\\_len'";
1618 $dbr = wfGetDB( DB_SLAVE );
1619 $result = $dbr->query( $sql );
1620 $row = $result->fetchObject();
1621 $result->free();
1623 if( $row && $row->Variable_name == 'ft_min_word_len' ) {
1624 $this->minSearchLength = intval( $row->Value );
1625 } else {
1626 $this->minSearchLength = 0;
1629 return $this->minSearchLength;
1632 function convertForSearchResult( $termsArray ) {
1633 # some languages, e.g. Chinese, need to do a conversion
1634 # in order for search results to be displayed correctly
1635 return $termsArray;
1639 * Get the first character of a string.
1641 * @param $s string
1642 * @return string
1644 function firstChar( $s ) {
1645 $matches = array();
1646 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1647 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1649 if ( isset( $matches[1] ) ) {
1650 if ( strlen( $matches[1] ) != 3 ) {
1651 return $matches[1];
1654 // Break down Hangul syllables to grab the first jamo
1655 $code = utf8ToCodepoint( $matches[1] );
1656 if ( $code < 0xac00 || 0xd7a4 <= $code) {
1657 return $matches[1];
1658 } elseif ( $code < 0xb098 ) {
1659 return "\xe3\x84\xb1";
1660 } elseif ( $code < 0xb2e4 ) {
1661 return "\xe3\x84\xb4";
1662 } elseif ( $code < 0xb77c ) {
1663 return "\xe3\x84\xb7";
1664 } elseif ( $code < 0xb9c8 ) {
1665 return "\xe3\x84\xb9";
1666 } elseif ( $code < 0xbc14 ) {
1667 return "\xe3\x85\x81";
1668 } elseif ( $code < 0xc0ac ) {
1669 return "\xe3\x85\x82";
1670 } elseif ( $code < 0xc544 ) {
1671 return "\xe3\x85\x85";
1672 } elseif ( $code < 0xc790 ) {
1673 return "\xe3\x85\x87";
1674 } elseif ( $code < 0xcc28 ) {
1675 return "\xe3\x85\x88";
1676 } elseif ( $code < 0xce74 ) {
1677 return "\xe3\x85\x8a";
1678 } elseif ( $code < 0xd0c0 ) {
1679 return "\xe3\x85\x8b";
1680 } elseif ( $code < 0xd30c ) {
1681 return "\xe3\x85\x8c";
1682 } elseif ( $code < 0xd558 ) {
1683 return "\xe3\x85\x8d";
1684 } else {
1685 return "\xe3\x85\x8e";
1687 } else {
1688 return "";
1692 function initEncoding() {
1693 # Some languages may have an alternate char encoding option
1694 # (Esperanto X-coding, Japanese furigana conversion, etc)
1695 # If this language is used as the primary content language,
1696 # an override to the defaults can be set here on startup.
1699 function recodeForEdit( $s ) {
1700 # For some languages we'll want to explicitly specify
1701 # which characters make it into the edit box raw
1702 # or are converted in some way or another.
1703 # Note that if wgOutputEncoding is different from
1704 # wgInputEncoding, this text will be further converted
1705 # to wgOutputEncoding.
1706 global $wgEditEncoding;
1707 if( $wgEditEncoding == '' or
1708 $wgEditEncoding == 'UTF-8' ) {
1709 return $s;
1710 } else {
1711 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1715 function recodeInput( $s ) {
1716 # Take the previous into account.
1717 global $wgEditEncoding;
1718 if($wgEditEncoding != "") {
1719 $enc = $wgEditEncoding;
1720 } else {
1721 $enc = 'UTF-8';
1723 if( $enc == 'UTF-8' ) {
1724 return $s;
1725 } else {
1726 return $this->iconv( $enc, 'UTF-8', $s );
1731 * For right-to-left language support
1733 * @return bool
1735 function isRTL() {
1736 $this->load();
1737 return $this->rtl;
1741 * A hidden direction mark (LRM or RLM), depending on the language direction
1743 * @return string
1745 function getDirMark() {
1746 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1750 * An arrow, depending on the language direction
1752 * @return string
1754 function getArrow() {
1755 return $this->isRTL() ? '←' : '→';
1759 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1761 * @return bool
1763 function linkPrefixExtension() {
1764 $this->load();
1765 return $this->linkPrefixExtension;
1768 function &getMagicWords() {
1769 $this->load();
1770 return $this->magicWords;
1773 # Fill a MagicWord object with data from here
1774 function getMagic( &$mw ) {
1775 if ( !$this->mMagicHookDone ) {
1776 $this->mMagicHookDone = true;
1777 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1779 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1780 $rawEntry = $this->mMagicExtensions[$mw->mId];
1781 } else {
1782 $magicWords =& $this->getMagicWords();
1783 if ( isset( $magicWords[$mw->mId] ) ) {
1784 $rawEntry = $magicWords[$mw->mId];
1785 } else {
1786 # Fall back to English if local list is incomplete
1787 $magicWords =& Language::getMagicWords();
1788 if ( !isset($magicWords[$mw->mId]) ) {
1789 throw new MWException("Magic word '{$mw->mId}' not found" );
1791 $rawEntry = $magicWords[$mw->mId];
1795 if( !is_array( $rawEntry ) ) {
1796 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1797 } else {
1798 $mw->mCaseSensitive = $rawEntry[0];
1799 $mw->mSynonyms = array_slice( $rawEntry, 1 );
1804 * Add magic words to the extension array
1806 function addMagicWordsByLang( $newWords ) {
1807 $code = $this->getCode();
1808 $fallbackChain = array();
1809 while ( $code && !in_array( $code, $fallbackChain ) ) {
1810 $fallbackChain[] = $code;
1811 $code = self::getFallbackFor( $code );
1813 if ( !in_array( 'en', $fallbackChain ) ) {
1814 $fallbackChain[] = 'en';
1816 $fallbackChain = array_reverse( $fallbackChain );
1817 foreach ( $fallbackChain as $code ) {
1818 if ( isset( $newWords[$code] ) ) {
1819 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1825 * Get special page names, as an associative array
1826 * case folded alias => real name
1828 function getSpecialPageAliases() {
1829 $this->load();
1831 // Cache aliases because it may be slow to load them
1832 if ( !isset( $this->mExtendedSpecialPageAliases ) ) {
1834 // Initialise array
1835 $this->mExtendedSpecialPageAliases = $this->specialPageAliases;
1837 global $wgExtensionAliasesFiles;
1838 foreach ( $wgExtensionAliasesFiles as $file ) {
1840 // Fail fast
1841 if ( !file_exists($file) )
1842 throw new MWException( "Aliases file does not exist: $file" );
1844 $aliases = array();
1845 require($file);
1847 // Check the availability of aliases
1848 if ( !isset($aliases['en']) )
1849 throw new MWException( "Malformed aliases file: $file" );
1851 // Merge all aliases in fallback chain
1852 $code = $this->getCode();
1853 do {
1854 if ( !isset($aliases[$code]) ) continue;
1856 $aliases[$code] = $this->fixSpecialPageAliases( $aliases[$code] );
1857 /* Merge the aliases, THIS will break if there is special page name
1858 * which looks like a numerical key, thanks to PHP...
1859 * See the array_merge_recursive manual entry */
1860 $this->mExtendedSpecialPageAliases = array_merge_recursive(
1861 $this->mExtendedSpecialPageAliases, $aliases[$code] );
1863 } while ( $code = self::getFallbackFor( $code ) );
1866 wfRunHooks( 'LanguageGetSpecialPageAliases',
1867 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1870 return $this->mExtendedSpecialPageAliases;
1874 * Function to fix special page aliases. Will convert the first letter to
1875 * upper case and spaces to underscores. Can be given a full aliases array,
1876 * in which case it will recursively fix all aliases.
1878 public function fixSpecialPageAliases( $mixed ) {
1879 // Work recursively until in string level
1880 if ( is_array($mixed) ) {
1881 $callback = array( $this, 'fixSpecialPageAliases' );
1882 return array_map( $callback, $mixed );
1884 return str_replace( ' ', '_', $this->ucfirst( $mixed ) );
1888 * Italic is unsuitable for some languages
1890 * @param $text String: the text to be emphasized.
1891 * @return string
1893 function emphasize( $text ) {
1894 return "<em>$text</em>";
1898 * Normally we output all numbers in plain en_US style, that is
1899 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1900 * point twohundredthirtyfive. However this is not sutable for all
1901 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1902 * Icelandic just want to use commas instead of dots, and dots instead
1903 * of commas like "293.291,235".
1905 * An example of this function being called:
1906 * <code>
1907 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1908 * </code>
1910 * See LanguageGu.php for the Gujarati implementation and
1911 * $separatorTransformTable on MessageIs.php for
1912 * the , => . and . => , implementation.
1914 * @todo check if it's viable to use localeconv() for the decimal
1915 * separator thing.
1916 * @param $number Mixed: the string to be formatted, should be an integer
1917 * or a floating point number.
1918 * @param $nocommafy Bool: set to true for special numbers like dates
1919 * @return string
1921 function formatNum( $number, $nocommafy = false ) {
1922 global $wgTranslateNumerals;
1923 if (!$nocommafy) {
1924 $number = $this->commafy($number);
1925 $s = $this->separatorTransformTable();
1926 if ($s) { $number = strtr($number, $s); }
1929 if ($wgTranslateNumerals) {
1930 $s = $this->digitTransformTable();
1931 if ($s) { $number = strtr($number, $s); }
1934 return $number;
1937 function parseFormattedNumber( $number ) {
1938 $s = $this->digitTransformTable();
1939 if ($s) { $number = strtr($number, array_flip($s)); }
1941 $s = $this->separatorTransformTable();
1942 if ($s) { $number = strtr($number, array_flip($s)); }
1944 $number = strtr( $number, array (',' => '') );
1945 return $number;
1949 * Adds commas to a given number
1951 * @param $_ mixed
1952 * @return string
1954 function commafy($_) {
1955 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1958 function digitTransformTable() {
1959 $this->load();
1960 return $this->digitTransformTable;
1963 function separatorTransformTable() {
1964 $this->load();
1965 return $this->separatorTransformTable;
1970 * Take a list of strings and build a locale-friendly comma-separated
1971 * list, using the local comma-separator message.
1972 * The last two strings are chained with an "and".
1974 * @param $l Array
1975 * @return string
1977 function listToText( $l ) {
1978 $s = '';
1979 $m = count( $l ) - 1;
1980 if( $m == 1 ) {
1981 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
1983 else {
1984 for ( $i = $m; $i >= 0; $i-- ) {
1985 if ( $i == $m ) {
1986 $s = $l[$i];
1987 } else if( $i == $m - 1 ) {
1988 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
1989 } else {
1990 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
1993 return $s;
1998 * Take a list of strings and build a locale-friendly comma-separated
1999 * list, using the local comma-separator message.
2000 * @param $list array of strings to put in a comma list
2001 * @return string
2003 function commaList( $list ) {
2004 return implode(
2005 $list,
2006 wfMsgExt( 'comma-separator', array( 'escapenoentities', 'language' => $this ) ) );
2010 * Same as commaList, but separate it with the pipe instead.
2011 * @param $list array of strings to put in a pipe list
2012 * @return string
2014 function pipeList( $list ) {
2015 return implode(
2016 $list,
2017 wfMsgExt( 'pipe-separator', array( 'escapenoentities', 'language' => $this ) ) );
2021 * Truncate a string to a specified length in bytes, appending an optional
2022 * string (e.g. for ellipses)
2024 * The database offers limited byte lengths for some columns in the database;
2025 * multi-byte character sets mean we need to ensure that only whole characters
2026 * are included, otherwise broken characters can be passed to the user
2028 * If $length is negative, the string will be truncated from the beginning
2030 * @param $string String to truncate
2031 * @param $length Int: maximum length (excluding ellipses)
2032 * @param $ellipsis String to append to the truncated text
2033 * @return string
2035 function truncate( $string, $length, $ellipsis = "" ) {
2036 if( $length == 0 ) {
2037 return $ellipsis;
2039 if ( strlen( $string ) <= abs( $length ) ) {
2040 return $string;
2042 if( $length > 0 ) {
2043 $string = substr( $string, 0, $length );
2044 $char = ord( $string[strlen( $string ) - 1] );
2045 $m = array();
2046 if ($char >= 0xc0) {
2047 # We got the first byte only of a multibyte char; remove it.
2048 $string = substr( $string, 0, -1 );
2049 } elseif( $char >= 0x80 &&
2050 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2051 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
2052 # We chopped in the middle of a character; remove it
2053 $string = $m[1];
2055 return $string . $ellipsis;
2056 } else {
2057 $string = substr( $string, $length );
2058 $char = ord( $string[0] );
2059 if( $char >= 0x80 && $char < 0xc0 ) {
2060 # We chopped in the middle of a character; remove the whole thing
2061 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2063 return $ellipsis . $string;
2068 * Grammatical transformations, needed for inflected languages
2069 * Invoked by putting {{grammar:case|word}} in a message
2071 * @param $word string
2072 * @param $case string
2073 * @return string
2075 function convertGrammar( $word, $case ) {
2076 global $wgGrammarForms;
2077 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
2078 return $wgGrammarForms[$this->getCode()][$case][$word];
2080 return $word;
2084 * Plural form transformations, needed for some languages.
2085 * For example, there are 3 form of plural in Russian and Polish,
2086 * depending on "count mod 10". See [[w:Plural]]
2087 * For English it is pretty simple.
2089 * Invoked by putting {{plural:count|wordform1|wordform2}}
2090 * or {{plural:count|wordform1|wordform2|wordform3}}
2092 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2094 * @param $count Integer: non-localized number
2095 * @param $forms Array: different plural forms
2096 * @return string Correct form of plural for $count in this language
2098 function convertPlural( $count, $forms ) {
2099 if ( !count($forms) ) { return ''; }
2100 $forms = $this->preConvertPlural( $forms, 2 );
2102 return ( $count == 1 ) ? $forms[0] : $forms[1];
2106 * Checks that convertPlural was given an array and pads it to requested
2107 * amound of forms by copying the last one.
2109 * @param $count Integer: How many forms should there be at least
2110 * @param $forms Array of forms given to convertPlural
2111 * @return array Padded array of forms or an exception if not an array
2113 protected function preConvertPlural( /* Array */ $forms, $count ) {
2114 while ( count($forms) < $count ) {
2115 $forms[] = $forms[count($forms)-1];
2117 return $forms;
2121 * For translaing of expiry times
2122 * @param $str String: the validated block time in English
2123 * @return Somehow translated block time
2124 * @see LanguageFi.php for example implementation
2126 function translateBlockExpiry( $str ) {
2128 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2130 if ( $scBlockExpiryOptions == '-') {
2131 return $str;
2134 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2135 if ( strpos($option, ":") === false )
2136 continue;
2137 list($show, $value) = explode(":", $option);
2138 if ( strcmp ( $str, $value) == 0 ) {
2139 return htmlspecialchars( trim( $show ) );
2143 return $str;
2147 * languages like Chinese need to be segmented in order for the diff
2148 * to be of any use
2150 * @param $text String
2151 * @return String
2153 function segmentForDiff( $text ) {
2154 return $text;
2158 * and unsegment to show the result
2160 * @param $text String
2161 * @return String
2163 function unsegmentForDiff( $text ) {
2164 return $text;
2167 # convert text to different variants of a language.
2168 function convert( $text, $isTitle = false) {
2169 return $this->mConverter->convert($text, $isTitle);
2172 # Convert text from within Parser
2173 function parserConvert( $text, &$parser ) {
2174 return $this->mConverter->parserConvert( $text, $parser );
2177 # Check if this is a language with variants
2178 function hasVariants(){
2179 return sizeof($this->getVariants())>1;
2182 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2183 function armourMath($text){
2184 return $this->mConverter->armourMath($text);
2189 * Perform output conversion on a string, and encode for safe HTML output.
2190 * @param $text String
2191 * @param $isTitle Bool -- wtf?
2192 * @return string
2193 * @todo this should get integrated somewhere sane
2195 function convertHtml( $text, $isTitle = false ) {
2196 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2199 function convertCategoryKey( $key ) {
2200 return $this->mConverter->convertCategoryKey( $key );
2204 * get the list of variants supported by this langauge
2205 * see sample implementation in LanguageZh.php
2207 * @return array an array of language codes
2209 function getVariants() {
2210 return $this->mConverter->getVariants();
2214 function getPreferredVariant( $fromUser = true ) {
2215 return $this->mConverter->getPreferredVariant( $fromUser );
2219 * if a language supports multiple variants, it is
2220 * possible that non-existing link in one variant
2221 * actually exists in another variant. this function
2222 * tries to find it. See e.g. LanguageZh.php
2224 * @param $link String: the name of the link
2225 * @param $nt Mixed: the title object of the link
2226 * @return null the input parameters may be modified upon return
2228 function findVariantLink( &$link, &$nt, $forTemplate = false ) {
2229 $this->mConverter->findVariantLink($link, $nt, $forTemplate );
2233 * If a language supports multiple variants, converts text
2234 * into an array of all possible variants of the text:
2235 * 'variant' => text in that variant
2238 function convertLinkToAllVariants($text){
2239 return $this->mConverter->convertLinkToAllVariants($text);
2244 * returns language specific options used by User::getPageRenderHash()
2245 * for example, the preferred language variant
2247 * @return string
2249 function getExtraHashOptions() {
2250 return $this->mConverter->getExtraHashOptions();
2254 * for languages that support multiple variants, the title of an
2255 * article may be displayed differently in different variants. this
2256 * function returns the apporiate title defined in the body of the article.
2258 * @return string
2260 function getParsedTitle() {
2261 return $this->mConverter->getParsedTitle();
2265 * Enclose a string with the "no conversion" tag. This is used by
2266 * various functions in the Parser
2268 * @param $text String: text to be tagged for no conversion
2269 * @param $noParse
2270 * @return string the tagged text
2272 function markNoConversion( $text, $noParse=false ) {
2273 return $this->mConverter->markNoConversion( $text, $noParse );
2277 * A regular expression to match legal word-trailing characters
2278 * which should be merged onto a link of the form [[foo]]bar.
2280 * @return string
2282 function linkTrail() {
2283 $this->load();
2284 return $this->linkTrail;
2287 function getLangObj() {
2288 return $this;
2292 * Get the RFC 3066 code for this language object
2294 function getCode() {
2295 return $this->mCode;
2298 function setCode( $code ) {
2299 $this->mCode = $code;
2302 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2303 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2306 static function getMessagesFileName( $code ) {
2307 global $IP;
2308 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2311 static function getClassFileName( $code ) {
2312 global $IP;
2313 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2316 static function getLocalisationArray( $code, $disableCache = false ) {
2317 self::loadLocalisation( $code, $disableCache );
2318 return self::$mLocalisationCache[$code];
2322 * Load localisation data for a given code into the static cache
2324 * @return array Dependencies, map of filenames to mtimes
2326 static function loadLocalisation( $code, $disableCache = false ) {
2327 static $recursionGuard = array();
2328 global $wgMemc, $wgEnableSerializedMessages, $wgCheckSerialized;
2330 if ( !$code ) {
2331 throw new MWException( "Invalid language code requested" );
2334 if ( !$disableCache ) {
2335 # Try the per-process cache
2336 if ( isset( self::$mLocalisationCache[$code] ) ) {
2337 return self::$mLocalisationCache[$code]['deps'];
2340 wfProfileIn( __METHOD__ );
2342 # Try the serialized directory
2343 if( $wgEnableSerializedMessages ) {
2344 $cache = wfGetPrecompiledData( self::getFileName( "Messages", $code, '.ser' ) );
2345 if ( $cache ) {
2346 if ( $wgCheckSerialized && self::isLocalisationOutOfDate( $cache ) ) {
2347 $cache = false;
2348 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2349 } else {
2350 self::$mLocalisationCache[$code] = $cache;
2351 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2352 wfProfileOut( __METHOD__ );
2353 return self::$mLocalisationCache[$code]['deps'];
2358 # Try the global cache
2359 $memcKey = wfMemcKey('localisation', $code );
2360 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2361 $cache = $wgMemc->get( $memcKey );
2362 if ( $cache ) {
2363 if ( self::isLocalisationOutOfDate( $cache ) ) {
2364 $wgMemc->delete( $memcKey );
2365 $wgMemc->delete( $fbMemcKey );
2366 $cache = false;
2367 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2368 } else {
2369 self::$mLocalisationCache[$code] = $cache;
2370 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2371 wfProfileOut( __METHOD__ );
2372 return $cache['deps'];
2375 } else {
2376 wfProfileIn( __METHOD__ );
2379 # Default fallback, may be overridden when the messages file is included
2380 if ( $code != 'en' ) {
2381 $fallback = 'en';
2382 } else {
2383 $fallback = false;
2386 # Load the primary localisation from the source file
2387 $filename = self::getMessagesFileName( $code );
2388 if ( !file_exists( $filename ) ) {
2389 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2390 $cache = compact( self::$mLocalisationKeys ); // Set correct fallback
2391 $deps = array();
2392 } else {
2393 $deps = array( $filename => filemtime( $filename ) );
2394 require( $filename );
2395 $cache = compact( self::$mLocalisationKeys );
2396 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2399 if ( !empty( $fallback ) ) {
2400 # Load the fallback localisation, with a circular reference guard
2401 if ( isset( $recursionGuard[$code] ) ) {
2402 throw new MWException( "Error: Circular fallback reference in language code $code" );
2404 $recursionGuard[$code] = true;
2405 $newDeps = self::loadLocalisation( $fallback, $disableCache );
2406 unset( $recursionGuard[$code] );
2408 $secondary = self::$mLocalisationCache[$fallback];
2409 $deps = array_merge( $deps, $newDeps );
2411 # Merge the fallback localisation with the current localisation
2412 foreach ( self::$mLocalisationKeys as $key ) {
2413 if ( isset( $cache[$key] ) ) {
2414 if ( isset( $secondary[$key] ) ) {
2415 if ( in_array( $key, self::$mMergeableMapKeys ) ) {
2416 $cache[$key] = $cache[$key] + $secondary[$key];
2417 } elseif ( in_array( $key, self::$mMergeableListKeys ) ) {
2418 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2419 } elseif ( in_array( $key, self::$mMergeableAliasListKeys ) ) {
2420 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2423 } else {
2424 $cache[$key] = $secondary[$key];
2428 # Merge bookstore lists if requested
2429 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2430 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2432 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2433 unset( $cache['bookstoreList']['inherit'] );
2437 # Add dependencies to the cache entry
2438 $cache['deps'] = $deps;
2440 # Replace spaces with underscores in namespace names
2441 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2443 # And do the same for specialpage aliases. $page is an array.
2444 foreach ( $cache['specialPageAliases'] as &$page ) {
2445 $page = str_replace( ' ', '_', $page );
2447 # Decouple the reference to prevent accidental damage
2448 unset($page);
2450 # Save to both caches
2451 self::$mLocalisationCache[$code] = $cache;
2452 if ( !$disableCache ) {
2453 $wgMemc->set( $memcKey, $cache );
2454 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2457 wfProfileOut( __METHOD__ );
2458 return $deps;
2462 * Test if a given localisation cache is out of date with respect to the
2463 * source Messages files. This is done automatically for the global cache
2464 * in $wgMemc, but is only done on certain occasions for the serialized
2465 * data file.
2467 * @param $cache mixed Either a language code or a cache array
2469 static function isLocalisationOutOfDate( $cache ) {
2470 if ( !is_array( $cache ) ) {
2471 self::loadLocalisation( $cache );
2472 $cache = self::$mLocalisationCache[$cache];
2474 $expired = false;
2475 foreach ( $cache['deps'] as $file => $mtime ) {
2476 if ( !file_exists( $file ) || filemtime( $file ) > $mtime ) {
2477 $expired = true;
2478 break;
2481 return $expired;
2485 * Get the fallback for a given language
2487 static function getFallbackFor( $code ) {
2488 // Shortcut
2489 if ( $code === 'en' ) return false;
2491 // Local cache
2492 static $cache = array();
2493 // Quick return
2494 if ( isset($cache[$code]) ) return $cache[$code];
2496 // Try memcache
2497 global $wgMemc;
2498 $memcKey = wfMemcKey( 'fallback', $code );
2499 $fbcode = $wgMemc->get( $memcKey );
2501 if ( is_string($fbcode) ) {
2502 // False is stored as a string to detect failures in memcache properly
2503 if ( $fbcode === '' ) $fbcode = false;
2505 // Update local cache and return
2506 $cache[$code] = $fbcode;
2507 return $fbcode;
2510 // Nothing in caches, load and and update both caches
2511 self::loadLocalisation( $code );
2512 $fbcode = self::$mLocalisationCache[$code]['fallback'];
2514 $cache[$code] = $fbcode;
2515 $wgMemc->set( $memcKey, (string) $fbcode );
2517 return $fbcode;
2520 /**
2521 * Get all messages for a given language
2523 static function getMessagesFor( $code ) {
2524 self::loadLocalisation( $code );
2525 return self::$mLocalisationCache[$code]['messages'];
2528 /**
2529 * Get a message for a given language
2531 static function getMessageFor( $key, $code ) {
2532 self::loadLocalisation( $code );
2533 return isset( self::$mLocalisationCache[$code]['messages'][$key] ) ? self::$mLocalisationCache[$code]['messages'][$key] : null;
2537 * Load localisation data for this object
2539 function load() {
2540 if ( !$this->mLoaded ) {
2541 self::loadLocalisation( $this->getCode() );
2542 $cache =& self::$mLocalisationCache[$this->getCode()];
2543 foreach ( self::$mLocalisationKeys as $key ) {
2544 $this->$key = $cache[$key];
2546 $this->mLoaded = true;
2548 $this->fixUpSettings();
2553 * Do any necessary post-cache-load settings adjustment
2555 function fixUpSettings() {
2556 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2557 $wgNamespaceAliases, $wgAmericanDates;
2558 wfProfileIn( __METHOD__ );
2559 if ( $wgExtraNamespaces ) {
2560 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
2563 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
2564 if ( $wgMetaNamespaceTalk ) {
2565 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
2566 } else {
2567 $talk = $this->namespaceNames[NS_PROJECT_TALK];
2568 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2570 # Allow grammar transformations
2571 # Allowing full message-style parsing would make simple requests
2572 # such as action=raw much more expensive than they need to be.
2573 # This will hopefully cover most cases.
2574 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2575 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2576 $talk = str_replace( ' ', '_', $talk );
2577 $this->namespaceNames[NS_PROJECT_TALK] = $talk;
2580 # The above mixing may leave namespaces out of canonical order.
2581 # Re-order by namespace ID number...
2582 ksort( $this->namespaceNames );
2584 # Put namespace names and aliases into a hashtable.
2585 # If this is too slow, then we should arrange it so that it is done
2586 # before caching. The catch is that at pre-cache time, the above
2587 # class-specific fixup hasn't been done.
2588 $this->mNamespaceIds = array();
2589 foreach ( $this->namespaceNames as $index => $name ) {
2590 $this->mNamespaceIds[$this->lc($name)] = $index;
2592 if ( $this->namespaceAliases ) {
2593 foreach ( $this->namespaceAliases as $name => $index ) {
2594 $this->mNamespaceIds[$this->lc($name)] = $index;
2597 if ( $wgNamespaceAliases ) {
2598 foreach ( $wgNamespaceAliases as $name => $index ) {
2599 $this->mNamespaceIds[$this->lc($name)] = $index;
2603 if ( $this->defaultDateFormat == 'dmy or mdy' ) {
2604 $this->defaultDateFormat = $wgAmericanDates ? 'mdy' : 'dmy';
2606 wfProfileOut( __METHOD__ );
2609 function replaceGrammarInNamespace( $m ) {
2610 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2613 static function getCaseMaps() {
2614 static $wikiUpperChars, $wikiLowerChars;
2615 if ( isset( $wikiUpperChars ) ) {
2616 return array( $wikiUpperChars, $wikiLowerChars );
2619 wfProfileIn( __METHOD__ );
2620 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2621 if ( $arr === false ) {
2622 throw new MWException(
2623 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2625 extract( $arr );
2626 wfProfileOut( __METHOD__ );
2627 return array( $wikiUpperChars, $wikiLowerChars );
2630 function formatTimePeriod( $seconds ) {
2631 if ( $seconds < 10 ) {
2632 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2633 } elseif ( $seconds < 60 ) {
2634 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2635 } elseif ( $seconds < 3600 ) {
2636 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2637 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2638 } else {
2639 $hours = floor( $seconds / 3600 );
2640 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2641 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2642 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2643 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2644 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2648 function formatBitrate( $bps ) {
2649 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2650 if ( $bps <= 0 ) {
2651 return $this->formatNum( $bps ) . $units[0];
2653 $unitIndex = floor( log10( $bps ) / 3 );
2654 $mantissa = $bps / pow( 1000, $unitIndex );
2655 if ( $mantissa < 10 ) {
2656 $mantissa = round( $mantissa, 1 );
2657 } else {
2658 $mantissa = round( $mantissa );
2660 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2664 * Format a size in bytes for output, using an appropriate
2665 * unit (B, KB, MB or GB) according to the magnitude in question
2667 * @param $size Size to format
2668 * @return string Plain text (not HTML)
2670 function formatSize( $size ) {
2671 // For small sizes no decimal places necessary
2672 $round = 0;
2673 if( $size > 1024 ) {
2674 $size = $size / 1024;
2675 if( $size > 1024 ) {
2676 $size = $size / 1024;
2677 // For MB and bigger two decimal places are smarter
2678 $round = 2;
2679 if( $size > 1024 ) {
2680 $size = $size / 1024;
2681 $msg = 'size-gigabytes';
2682 } else {
2683 $msg = 'size-megabytes';
2685 } else {
2686 $msg = 'size-kilobytes';
2688 } else {
2689 $msg = 'size-bytes';
2691 $size = round( $size, $round );
2692 $text = $this->getMessageFromDB( $msg );
2693 return str_replace( '$1', $this->formatNum( $size ), $text );