3 * @defgroup Language Language
9 if( !defined( 'MEDIAWIKI' ) ) {
10 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
15 global $wgLanguageNames;
16 require_once( dirname( __FILE__
) . '/Names.php' );
18 global $wgInputEncoding, $wgOutputEncoding;
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' );
31 * a fake language converter
37 function FakeConverter( $langobj ) { $this->mLang
= $langobj; }
38 function autoConvertToAllVariants( $text ) { return $text; }
39 function convert( $t ) { return $t; }
40 function convertTitle( $t ) { return $t->getPrefixedText(); }
41 function getVariants() { return array( $this->mLang
->getCode() ); }
42 function getPreferredVariant() { return $this->mLang
->getCode(); }
43 function getConvRuleTitle() { return false; }
44 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) {}
45 function getExtraHashOptions() { return ''; }
46 function getParsedTitle() { return ''; }
47 function markNoConversion( $text, $noParse = false ) { return $text; }
48 function convertCategoryKey( $key ) { return $key; }
49 function convertLinkToAllVariants( $text ) { return array( $this->mLang
->getCode() => $text ); }
50 function armourMath( $text ) { return $text; }
54 * Internationalisation code
58 var $mConverter, $mVariants, $mCode, $mLoaded = false;
59 var $mMagicExtensions = array(), $mMagicHookDone = false;
61 var $mNamespaceIds, $namespaceNames, $namespaceAliases;
62 var $dateFormatStrings = array();
63 var $mExtendedSpecialPageAliases;
66 * ReplacementArray object caches
68 var $transformData = array();
70 static public $dataCache;
71 static public $mLangObjCache = array();
73 static public $mWeekdayMsgs = array(
74 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
78 static public $mWeekdayAbbrevMsgs = array(
79 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
82 static public $mMonthMsgs = array(
83 'january', 'february', 'march', 'april', 'may_long', 'june',
84 'july', 'august', 'september', 'october', 'november',
87 static public $mMonthGenMsgs = array(
88 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
89 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
92 static public $mMonthAbbrevMsgs = array(
93 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
94 'sep', 'oct', 'nov', 'dec'
97 static public $mIranianCalendarMonthMsgs = array(
98 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
99 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
100 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
101 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
104 static public $mHebrewCalendarMonthMsgs = array(
105 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
106 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
107 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
108 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
109 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
112 static public $mHebrewCalendarMonthGenMsgs = array(
113 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
114 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
115 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
116 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
117 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
120 static public $mHijriCalendarMonthMsgs = array(
121 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
122 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
123 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
124 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
128 * Get a cached language object for a given language code
130 static function factory( $code ) {
131 if ( !isset( self
::$mLangObjCache[$code] ) ) {
132 if( count( self
::$mLangObjCache ) > 10 ) {
133 // Don't keep a billion objects around, that's stupid.
134 self
::$mLangObjCache = array();
136 self
::$mLangObjCache[$code] = self
::newFromCode( $code );
138 return self
::$mLangObjCache[$code];
142 * Create a language object for a given language code
144 protected static function newFromCode( $code ) {
146 static $recursionLevel = 0;
147 if ( $code == 'en' ) {
150 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
151 // Preload base classes to work around APC/PHP5 bug
152 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
153 include_once( "$IP/languages/classes/$class.deps.php" );
155 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
156 include_once( "$IP/languages/classes/$class.php" );
160 if ( $recursionLevel > 5 ) {
161 throw new MWException( "Language fallback loop detected when creating class $class\n" );
164 if( !class_exists( $class ) ) {
165 $fallback = Language
::getFallbackFor( $code );
167 $lang = Language
::newFromCode( $fallback );
169 $lang->setCode( $code );
177 * Get the LocalisationCache instance
179 public static function getLocalisationCache() {
180 if ( is_null( self
::$dataCache ) ) {
181 global $wgLocalisationCacheConf;
182 $class = $wgLocalisationCacheConf['class'];
183 self
::$dataCache = new $class( $wgLocalisationCacheConf );
185 return self
::$dataCache;
188 function __construct() {
189 $this->mConverter
= new FakeConverter( $this );
190 // Set the code to the name of the descendant
191 if ( get_class( $this ) == 'Language' ) {
194 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
196 self
::getLocalisationCache();
200 * Reduce memory usage
202 function __destruct() {
203 foreach ( $this as $name => $value ) {
204 unset( $this->$name );
209 * Hook which will be called if this is the content language.
210 * Descendants can use this to register hook functions or modify globals
212 function initContLang() {}
215 * @deprecated Use User::getDefaultOptions()
218 function getDefaultUserOptions() {
219 wfDeprecated( __METHOD__
);
220 return User
::getDefaultOptions();
223 function getFallbackLanguageCode() {
224 if ( $this->mCode
=== 'en' ) {
227 return self
::$dataCache->getItem( $this->mCode
, 'fallback' );
232 * Exports $wgBookstoreListEn
235 function getBookstoreList() {
236 return self
::$dataCache->getItem( $this->mCode
, 'bookstoreList' );
242 function getNamespaces() {
243 if ( is_null( $this->namespaceNames
) ) {
244 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk;
246 $this->namespaceNames
= self
::$dataCache->getItem( $this->mCode
, 'namespaceNames' );
247 if ( $wgExtraNamespaces ) {
248 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames
;
251 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
252 if ( $wgMetaNamespaceTalk ) {
253 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
255 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
256 $this->namespaceNames
[NS_PROJECT_TALK
] =
257 $this->fixVariableInNamespace( $talk );
260 # The above mixing may leave namespaces out of canonical order.
261 # Re-order by namespace ID number...
262 ksort( $this->namespaceNames
);
264 return $this->namespaceNames
;
268 * A convenience function that returns the same thing as
269 * getNamespaces() except with the array values changed to ' '
270 * where it found '_', useful for producing output to be displayed
271 * e.g. in <select> forms.
275 function getFormattedNamespaces() {
276 $ns = $this->getNamespaces();
277 foreach( $ns as $k => $v ) {
278 $ns[$k] = strtr( $v, '_', ' ' );
284 * Get a namespace value by key
286 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
287 * echo $mw_ns; // prints 'MediaWiki'
290 * @param $index Int: the array key of the namespace to return
291 * @return mixed, string if the namespace value exists, otherwise false
293 function getNsText( $index ) {
294 $ns = $this->getNamespaces();
295 return isset( $ns[$index] ) ?
$ns[$index] : false;
299 * A convenience function that returns the same thing as
300 * getNsText() except with '_' changed to ' ', useful for
305 function getFormattedNsText( $index ) {
306 $ns = $this->getNsText( $index );
307 return strtr( $ns, '_', ' ' );
311 * Get a namespace key by value, case insensitive.
312 * Only matches namespace names for the current language, not the
313 * canonical ones defined in Namespace.php.
315 * @param $text String
316 * @return mixed An integer if $text is a valid value otherwise false
318 function getLocalNsIndex( $text ) {
319 $lctext = $this->lc( $text );
320 $ids = $this->getNamespaceIds();
321 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
324 function getNamespaceAliases() {
325 if ( is_null( $this->namespaceAliases
) ) {
326 $aliases = self
::$dataCache->getItem( $this->mCode
, 'namespaceAliases' );
330 foreach ( $aliases as $name => $index ) {
331 if ( $index === NS_PROJECT_TALK
) {
332 unset( $aliases[$name] );
333 $name = $this->fixVariableInNamespace( $name );
334 $aliases[$name] = $index;
338 $this->namespaceAliases
= $aliases;
340 return $this->namespaceAliases
;
343 function getNamespaceIds() {
344 if ( is_null( $this->mNamespaceIds
) ) {
345 global $wgNamespaceAliases;
346 # Put namespace names and aliases into a hashtable.
347 # If this is too slow, then we should arrange it so that it is done
348 # before caching. The catch is that at pre-cache time, the above
349 # class-specific fixup hasn't been done.
350 $this->mNamespaceIds
= array();
351 foreach ( $this->getNamespaces() as $index => $name ) {
352 $this->mNamespaceIds
[$this->lc( $name )] = $index;
354 foreach ( $this->getNamespaceAliases() as $name => $index ) {
355 $this->mNamespaceIds
[$this->lc( $name )] = $index;
357 if ( $wgNamespaceAliases ) {
358 foreach ( $wgNamespaceAliases as $name => $index ) {
359 $this->mNamespaceIds
[$this->lc( $name )] = $index;
363 return $this->mNamespaceIds
;
368 * Get a namespace key by value, case insensitive. Canonical namespace
369 * names override custom ones defined for the current language.
371 * @param $text String
372 * @return mixed An integer if $text is a valid value otherwise false
374 function getNsIndex( $text ) {
375 $lctext = $this->lc( $text );
376 if ( ( $ns = MWNamespace
::getCanonicalIndex( $lctext ) ) !== null ) {
379 $ids = $this->getNamespaceIds();
380 return isset( $ids[$lctext] ) ?
$ids[$lctext] : false;
384 * short names for language variants used for language conversion links.
386 * @param $code String
389 function getVariantname( $code ) {
390 return $this->getMessageFromDB( "variantname-$code" );
393 function specialPage( $name ) {
394 $aliases = $this->getSpecialPageAliases();
395 if ( isset( $aliases[$name][0] ) ) {
396 $name = $aliases[$name][0];
398 return $this->getNsText( NS_SPECIAL
) . ':' . $name;
401 function getQuickbarSettings() {
403 $this->getMessage( 'qbsettings-none' ),
404 $this->getMessage( 'qbsettings-fixedleft' ),
405 $this->getMessage( 'qbsettings-fixedright' ),
406 $this->getMessage( 'qbsettings-floatingleft' ),
407 $this->getMessage( 'qbsettings-floatingright' )
411 function getMathNames() {
412 return self
::$dataCache->getItem( $this->mCode
, 'mathNames' );
415 function getDatePreferences() {
416 return self
::$dataCache->getItem( $this->mCode
, 'datePreferences' );
419 function getDateFormats() {
420 return self
::$dataCache->getItem( $this->mCode
, 'dateFormats' );
423 function getDefaultDateFormat() {
424 $df = self
::$dataCache->getItem( $this->mCode
, 'defaultDateFormat' );
425 if ( $df === 'dmy or mdy' ) {
426 global $wgAmericanDates;
427 return $wgAmericanDates ?
'mdy' : 'dmy';
433 function getDatePreferenceMigrationMap() {
434 return self
::$dataCache->getItem( $this->mCode
, 'datePreferenceMigrationMap' );
437 function getImageFile( $image ) {
438 return self
::$dataCache->getSubitem( $this->mCode
, 'imageFiles', $image );
441 function getDefaultUserOptionOverrides() {
442 return self
::$dataCache->getItem( $this->mCode
, 'defaultUserOptionOverrides' );
445 function getExtraUserToggles() {
446 return self
::$dataCache->getItem( $this->mCode
, 'extraUserToggles' );
449 function getUserToggle( $tog ) {
450 return $this->getMessageFromDB( "tog-$tog" );
454 * Get language names, indexed by code.
455 * If $customisedOnly is true, only returns codes with a messages file
457 public static function getLanguageNames( $customisedOnly = false ) {
458 global $wgLanguageNames, $wgExtraLanguageNames;
459 $allNames = $wgExtraLanguageNames +
$wgLanguageNames;
460 if ( !$customisedOnly ) {
466 $dir = opendir( "$IP/languages/messages" );
467 while( false !== ( $file = readdir( $dir ) ) ) {
468 $code = self
::getCodeFromFileName( $file, 'Messages' );
469 if ( $code && isset( $allNames[$code] ) ) {
470 $names[$code] = $allNames[$code];
478 * Get a message from the MediaWiki namespace.
480 * @param $msg String: message name
483 function getMessageFromDB( $msg ) {
484 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
487 function getLanguageName( $code ) {
488 $names = self
::getLanguageNames();
489 if ( !array_key_exists( $code, $names ) ) {
492 return $names[$code];
495 function getMonthName( $key ) {
496 return $this->getMessageFromDB( self
::$mMonthMsgs[$key - 1] );
499 function getMonthNameGen( $key ) {
500 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key - 1] );
503 function getMonthAbbreviation( $key ) {
504 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key - 1] );
507 function getWeekdayName( $key ) {
508 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key - 1] );
511 function getWeekdayAbbreviation( $key ) {
512 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key - 1] );
515 function getIranianCalendarMonthName( $key ) {
516 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key - 1] );
519 function getHebrewCalendarMonthName( $key ) {
520 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key - 1] );
523 function getHebrewCalendarMonthNameGen( $key ) {
524 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key - 1] );
527 function getHijriCalendarMonthName( $key ) {
528 return $this->getMessageFromDB( self
::$mHijriCalendarMonthMsgs[$key - 1] );
532 * Used by date() and time() to adjust the time output.
534 * @param $ts Int the time in date('YmdHis') format
535 * @param $tz Mixed: adjust the time by this amount (default false, mean we
536 * get user timecorrection setting)
539 function userAdjust( $ts, $tz = false ) {
540 global $wgUser, $wgLocalTZoffset;
542 if ( $tz === false ) {
543 $tz = $wgUser->getOption( 'timecorrection' );
546 $data = explode( '|', $tz, 3 );
548 if ( $data[0] == 'ZoneInfo' ) {
549 if ( function_exists( 'timezone_open' ) && @timezone_open
( $data[2] ) !== false ) {
550 $date = date_create( $ts, timezone_open( 'UTC' ) );
551 date_timezone_set( $date, timezone_open( $data[2] ) );
552 $date = date_format( $date, 'YmdHis' );
555 # Unrecognized timezone, default to 'Offset' with the stored offset.
560 if ( $data[0] == 'System' ||
$tz == '' ) {
561 # Global offset in minutes.
562 if( isset( $wgLocalTZoffset ) ) {
563 $minDiff = $wgLocalTZoffset;
565 } else if ( $data[0] == 'Offset' ) {
566 $minDiff = intval( $data[1] );
568 $data = explode( ':', $tz );
569 if( count( $data ) == 2 ) {
570 $data[0] = intval( $data[0] );
571 $data[1] = intval( $data[1] );
572 $minDiff = abs( $data[0] ) * 60 +
$data[1];
573 if ( $data[0] < 0 ) {
574 $minDiff = -$minDiff;
577 $minDiff = intval( $data[0] ) * 60;
581 # No difference ? Return time unchanged
582 if ( 0 == $minDiff ) {
586 wfSuppressWarnings(); // E_STRICT system time bitching
587 # Generate an adjusted date; take advantage of the fact that mktime
588 # will normalize out-of-range values so we don't have to split $minDiff
589 # into hours and minutes.
591 (int)substr( $ts, 8, 2) ), # Hours
592 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
593 (int)substr( $ts, 12, 2 ), # Seconds
594 (int)substr( $ts, 4, 2 ), # Month
595 (int)substr( $ts, 6, 2 ), # Day
596 (int)substr( $ts, 0, 4 ) ); #Year
598 $date = date( 'YmdHis', $t );
605 * This is a workalike of PHP's date() function, but with better
606 * internationalisation, a reduced set of format characters, and a better
609 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
610 * PHP manual for definitions. "o" format character is supported since
611 * PHP 5.1.0, previous versions return literal o.
612 * There are a number of extensions, which start with "x":
614 * xn Do not translate digits of the next numeric format character
615 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
616 * xr Use roman numerals for the next numeric format character
617 * xh Use hebrew numerals for the next numeric format character
619 * xg Genitive month name
621 * xij j (day number) in Iranian calendar
622 * xiF F (month name) in Iranian calendar
623 * xin n (month number) in Iranian calendar
624 * xiY Y (full year) in Iranian calendar
626 * xjj j (day number) in Hebrew calendar
627 * xjF F (month name) in Hebrew calendar
628 * xjt t (days in month) in Hebrew calendar
629 * xjx xg (genitive month name) in Hebrew calendar
630 * xjn n (month number) in Hebrew calendar
631 * xjY Y (full year) in Hebrew calendar
633 * xmj j (day number) in Hijri calendar
634 * xmF F (month name) in Hijri calendar
635 * xmn n (month number) in Hijri calendar
636 * xmY Y (full year) in Hijri calendar
638 * xkY Y (full year) in Thai solar calendar. Months and days are
639 * identical to the Gregorian calendar
640 * xoY Y (full year) in Minguo calendar or Juche year.
641 * Months and days are identical to the
643 * xtY Y (full year) in Japanese nengo. Months and days are
644 * identical to the Gregorian calendar
646 * Characters enclosed in double quotes will be considered literal (with
647 * the quotes themselves removed). Unmatched quotes will be considered
648 * literal quotes. Example:
650 * "The month is" F => The month is January
653 * Backslash escaping is also supported.
655 * Input timestamp is assumed to be pre-normalized to the desired local
658 * @param $format String
659 * @param $ts String: 14-character timestamp
662 * @todo emulation of "o" format character for PHP pre 5.1.0
663 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
665 function sprintfDate( $format, $ts ) {
678 for ( $p = 0; $p < strlen( $format ); $p++
) {
681 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
682 $code .= $format[++
$p];
685 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ||
$code == 'xm' ||
$code == 'xo' ||
$code == 'xt' ) && $p < strlen( $format ) - 1 ) {
686 $code .= $format[++
$p];
697 $rawToggle = !$rawToggle;
706 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
709 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
710 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
713 $num = substr( $ts, 6, 2 );
716 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
717 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) +
1 );
720 $num = intval( substr( $ts, 6, 2 ) );
724 $iranian = self
::tsToIranian( $ts );
730 $hijri = self
::tsToHijri( $ts );
736 $hebrew = self
::tsToHebrew( $ts );
742 $unix = wfTimestamp( TS_UNIX
, $ts );
744 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) +
1 );
748 $unix = wfTimestamp( TS_UNIX
, $ts );
750 $w = gmdate( 'w', $unix );
755 $unix = wfTimestamp( TS_UNIX
, $ts );
757 $num = gmdate( 'w', $unix );
761 $unix = wfTimestamp( TS_UNIX
, $ts );
763 $num = gmdate( 'z', $unix );
767 $unix = wfTimestamp( TS_UNIX
, $ts );
769 $num = gmdate( 'W', $unix );
772 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
776 $iranian = self
::tsToIranian( $ts );
778 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
782 $hijri = self
::tsToHijri( $ts );
784 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
788 $hebrew = self
::tsToHebrew( $ts );
790 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
793 $num = substr( $ts, 4, 2 );
796 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
799 $num = intval( substr( $ts, 4, 2 ) );
803 $iranian = self
::tsToIranian( $ts );
809 $hijri = self
::tsToHijri ( $ts );
815 $hebrew = self
::tsToHebrew( $ts );
821 $unix = wfTimestamp( TS_UNIX
, $ts );
823 $num = gmdate( 't', $unix );
827 $hebrew = self
::tsToHebrew( $ts );
833 $unix = wfTimestamp( TS_UNIX
, $ts );
835 $num = gmdate( 'L', $unix );
837 # 'o' is supported since PHP 5.1.0
838 # return literal if not supported
839 # TODO: emulation for pre 5.1.0 versions
842 $unix = wfTimestamp( TS_UNIX
, $ts );
844 if ( version_compare( PHP_VERSION
, '5.1.0' ) === 1 ) {
845 $num = date( 'o', $unix );
851 $num = substr( $ts, 0, 4 );
855 $iranian = self
::tsToIranian( $ts );
861 $hijri = self
::tsToHijri( $ts );
867 $hebrew = self
::tsToHebrew( $ts );
873 $thai = self
::tsToYear( $ts, 'thai' );
879 $minguo = self
::tsToYear( $ts, 'minguo' );
885 $tenno = self
::tsToYear( $ts, 'tenno' );
890 $num = substr( $ts, 2, 2 );
893 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
896 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
899 $h = substr( $ts, 8, 2 );
900 $num = $h %
12 ?
$h %
12 : 12;
903 $num = intval( substr( $ts, 8, 2 ) );
906 $h = substr( $ts, 8, 2 );
907 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
910 $num = substr( $ts, 8, 2 );
913 $num = substr( $ts, 10, 2 );
916 $num = substr( $ts, 12, 2 );
920 $unix = wfTimestamp( TS_UNIX
, $ts );
922 $s .= gmdate( 'c', $unix );
926 $unix = wfTimestamp( TS_UNIX
, $ts );
928 $s .= gmdate( 'r', $unix );
932 $unix = wfTimestamp( TS_UNIX
, $ts );
938 if ( $p < strlen( $format ) - 1 ) {
946 if ( $p < strlen( $format ) - 1 ) {
947 $endQuote = strpos( $format, '"', $p +
1 );
948 if ( $endQuote === false ) {
949 # No terminating quote, assume literal "
952 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
956 # Quote at end of string, assume literal "
963 if ( $num !== false ) {
964 if ( $rawToggle ||
$raw ) {
967 } elseif ( $roman ) {
968 $s .= self
::romanNumeral( $num );
970 } elseif( $hebrewNum ) {
971 $s .= self
::hebrewNumeral( $num );
974 $s .= $this->formatNum( $num, true );
982 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
983 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
985 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
986 * Gregorian dates to Iranian dates. Originally written in C, it
987 * is released under the terms of GNU Lesser General Public
988 * License. Conversion to PHP was performed by Niklas Laxström.
990 * Link: http://www.farsiweb.info/jalali/jalali.c
992 private static function tsToIranian( $ts ) {
993 $gy = substr( $ts, 0, 4 ) -1600;
994 $gm = substr( $ts, 4, 2 ) -1;
995 $gd = substr( $ts, 6, 2 ) -1;
997 # Days passed from the beginning (including leap years)
999 +
floor( ( $gy +
3 ) / 4 )
1000 - floor( ( $gy +
99 ) / 100 )
1001 +
floor( ( $gy +
399 ) / 400 );
1004 // Add days of the past months of this year
1005 for( $i = 0; $i < $gm; $i++
) {
1006 $gDayNo +
= self
::$GREG_DAYS[$i];
1010 if ( $gm > 1 && ( ( $gy%4
=== 0 && $gy%100
!== 0 ||
( $gy%400
== 0 ) ) ) ) {
1014 // Days passed in current month
1017 $jDayNo = $gDayNo - 79;
1019 $jNp = floor( $jDayNo / 12053 );
1022 $jy = 979 +
33 * $jNp +
4 * floor( $jDayNo / 1461 );
1025 if ( $jDayNo >= 366 ) {
1026 $jy +
= floor( ( $jDayNo - 1 ) / 365 );
1027 $jDayNo = floor( ( $jDayNo - 1 ) %
365 );
1030 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
1031 $jDayNo -= self
::$IRANIAN_DAYS[$i];
1037 return array( $jy, $jm, $jd );
1041 * Converting Gregorian dates to Hijri dates.
1043 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1045 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1047 private static function tsToHijri( $ts ) {
1048 $year = substr( $ts, 0, 4 );
1049 $month = substr( $ts, 4, 2 );
1050 $day = substr( $ts, 6, 2 );
1058 ( $zy > 1582 ) ||
( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1059 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1062 $zjd = (int)( ( 1461 * ( $zy +
4800 +
(int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1063 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1064 (int)( ( 3 * (int)( ( ( $zy +
4900 +
(int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1067 $zjd = 367 * $zy - (int)( ( 7 * ( $zy +
5001 +
(int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1068 (int)( ( 275 * $zm ) / 9 ) +
$zd +
1729777;
1071 $zl = $zjd-1948440 +
10632;
1072 $zn = (int)( ( $zl - 1 ) / 10631 );
1073 $zl = $zl - 10631 * $zn +
354;
1074 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ((int)( ( 50 * $zl ) / 17719 ) ) +
( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1075 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ((int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) +
29;
1076 $zm = (int)( ( 24 * $zl ) / 709 );
1077 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1078 $zy = 30 * $zn +
$zj - 30;
1080 return array( $zy, $zm, $zd );
1084 * Converting Gregorian dates to Hebrew dates.
1086 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1087 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1088 * to translate the relevant functions into PHP and release them under
1091 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1092 * and Adar II is 14. In a non-leap year, Adar is 6.
1094 private static function tsToHebrew( $ts ) {
1096 $year = substr( $ts, 0, 4 );
1097 $month = substr( $ts, 4, 2 );
1098 $day = substr( $ts, 6, 2 );
1100 # Calculate Hebrew year
1101 $hebrewYear = $year +
3760;
1103 # Month number when September = 1, August = 12
1112 # Calculate day of year from 1 September
1114 for( $i = 1; $i < $month; $i++
) {
1118 # Check if the year is leap
1119 if( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
1122 } elseif( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
1129 # Calculate the start of the Hebrew year
1130 $start = self
::hebrewYearStart( $hebrewYear );
1132 # Calculate next year's start
1133 if( $dayOfYear <= $start ) {
1134 # Day is before the start of the year - it is the previous year
1136 $nextStart = $start;
1140 # Add days since previous year's 1 September
1142 if( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1146 # Start of the new (previous) year
1147 $start = self
::hebrewYearStart( $hebrewYear );
1150 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
1153 # Calculate Hebrew day of year
1154 $hebrewDayOfYear = $dayOfYear - $start;
1156 # Difference between year's days
1157 $diff = $nextStart - $start;
1158 # Add 12 (or 13 for leap years) days to ignore the difference between
1159 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1160 # difference is only about the year type
1161 if( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
1167 # Check the year pattern, and is leap year
1168 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1169 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1170 # and non-leap years
1171 $yearPattern = $diff %
30;
1172 # Check if leap year
1173 $isLeap = $diff >= 30;
1175 # Calculate day in the month from number of day in the Hebrew year
1176 # Don't check Adar - if the day is not in Adar, we will stop before;
1177 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1178 $hebrewDay = $hebrewDayOfYear;
1181 while( $hebrewMonth <= 12 ) {
1182 # Calculate days in this month
1183 if( $isLeap && $hebrewMonth == 6 ) {
1184 # Adar in a leap year
1186 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1188 if( $hebrewDay <= $days ) {
1192 # Subtract the days of Adar I
1193 $hebrewDay -= $days;
1196 if( $hebrewDay <= $days ) {
1202 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1203 # Cheshvan in a complete year (otherwise as the rule below)
1205 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1206 # Kislev in an incomplete year (otherwise as the rule below)
1209 # Odd months have 30 days, even have 29
1210 $days = 30 - ( $hebrewMonth - 1 ) %
2;
1212 if( $hebrewDay <= $days ) {
1213 # In the current month
1216 # Subtract the days of the current month
1217 $hebrewDay -= $days;
1218 # Try in the next month
1223 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1227 * This calculates the Hebrew year start, as days since 1 September.
1228 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1229 * Used for Hebrew date.
1231 private static function hebrewYearStart( $year ) {
1232 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
1233 $b = intval( ( $year - 1 ) %
4 );
1234 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1238 $Mar = intval( $m );
1244 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7);
1245 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1247 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1249 } else if( $c == 2 ||
$c == 4 ||
$c == 6 ) {
1253 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1258 * Algorithm to convert Gregorian dates to Thai solar dates,
1259 * Minguo dates or Minguo dates.
1261 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1262 * http://en.wikipedia.org/wiki/Minguo_calendar
1263 * http://en.wikipedia.org/wiki/Japanese_era_name
1265 * @param $ts String: 14-character timestamp
1266 * @param $cName String: calender name
1267 * @return Array: converted year, month, day
1269 private static function tsToYear( $ts, $cName ) {
1270 $gy = substr( $ts, 0, 4 );
1271 $gm = substr( $ts, 4, 2 );
1272 $gd = substr( $ts, 6, 2 );
1274 if ( !strcmp( $cName, 'thai' ) ) {
1276 # Add 543 years to the Gregorian calendar
1277 # Months and days are identical
1278 $gy_offset = $gy +
543;
1279 } else if ( ( !strcmp( $cName, 'minguo' ) ) ||
!strcmp( $cName, 'juche' ) ) {
1281 # Deduct 1911 years from the Gregorian calendar
1282 # Months and days are identical
1283 $gy_offset = $gy - 1911;
1284 } else if ( !strcmp( $cName, 'tenno' ) ) {
1285 # Nengō dates up to Meiji period
1286 # Deduct years from the Gregorian calendar
1287 # depending on the nengo periods
1288 # Months and days are identical
1289 if ( ( $gy < 1912 ) ||
( ( $gy == 1912 ) && ( $gm < 7 ) ) ||
( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1291 $gy_gannen = $gy - 1868 +
1;
1292 $gy_offset = $gy_gannen;
1293 if ( $gy_gannen == 1 ) {
1296 $gy_offset = '明治' . $gy_offset;
1298 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1299 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1300 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1301 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1302 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1306 $gy_gannen = $gy - 1912 +
1;
1307 $gy_offset = $gy_gannen;
1308 if ( $gy_gannen == 1 ) {
1311 $gy_offset = '大正' . $gy_offset;
1313 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1314 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1315 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1319 $gy_gannen = $gy - 1926 +
1;
1320 $gy_offset = $gy_gannen;
1321 if ( $gy_gannen == 1 ) {
1324 $gy_offset = '昭和' . $gy_offset;
1327 $gy_gannen = $gy - 1989 +
1;
1328 $gy_offset = $gy_gannen;
1329 if ( $gy_gannen == 1 ) {
1332 $gy_offset = '平成' . $gy_offset;
1338 return array( $gy_offset, $gm, $gd );
1342 * Roman number formatting up to 3000
1344 static function romanNumeral( $num ) {
1345 static $table = array(
1346 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1347 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1348 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1349 array( '', 'M', 'MM', 'MMM' )
1352 $num = intval( $num );
1353 if ( $num > 3000 ||
$num <= 0 ) {
1358 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1359 if ( $num >= $pow10 ) {
1360 $s .= $table[$i][floor( $num / $pow10 )];
1362 $num = $num %
$pow10;
1368 * Hebrew Gematria number formatting up to 9999
1370 static function hebrewNumeral( $num ) {
1371 static $table = array(
1372 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1373 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1374 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1375 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1378 $num = intval( $num );
1379 if ( $num > 9999 ||
$num <= 0 ) {
1384 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1385 if ( $num >= $pow10 ) {
1386 if ( $num == 15 ||
$num == 16 ) {
1387 $s .= $table[0][9] . $table[0][$num - 9];
1390 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1391 if( $pow10 == 1000 ) {
1396 $num = $num %
$pow10;
1398 if( strlen( $s ) == 2 ) {
1401 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1402 $str .= substr( $s, strlen( $s ) - 2, 2 );
1404 $start = substr( $str, 0, strlen( $str ) - 2 );
1405 $end = substr( $str, strlen( $str ) - 2 );
1408 $str = $start . 'ך';
1411 $str = $start . 'ם';
1414 $str = $start . 'ן';
1417 $str = $start . 'ף';
1420 $str = $start . 'ץ';
1427 * This is meant to be used by time(), date(), and timeanddate() to get
1428 * the date preference they're supposed to use, it should be used in
1432 * function timeanddate([...], $format = true) {
1433 * $datePreference = $this->dateFormat($format);
1438 * @param $usePrefs Mixed: if true, the user's preference is used
1439 * if false, the site/language default is used
1440 * if int/string, assumed to be a format.
1443 function dateFormat( $usePrefs = true ) {
1446 if( is_bool( $usePrefs ) ) {
1448 $datePreference = $wgUser->getDatePreference();
1450 $datePreference = (string)User
::getDefaultOption( 'date' );
1453 $datePreference = (string)$usePrefs;
1457 if( $datePreference == '' ) {
1461 return $datePreference;
1465 * Get a format string for a given type and preference
1466 * @param $type May be date, time or both
1467 * @param $pref The format name as it appears in Messages*.php
1469 function getDateFormatString( $type, $pref ) {
1470 if ( !isset( $this->dateFormatStrings
[$type][$pref] ) ) {
1471 if ( $pref == 'default' ) {
1472 $pref = $this->getDefaultDateFormat();
1473 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1475 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1476 if ( is_null( $df ) ) {
1477 $pref = $this->getDefaultDateFormat();
1478 $df = self
::$dataCache->getSubitem( $this->mCode
, 'dateFormats', "$pref $type" );
1481 $this->dateFormatStrings
[$type][$pref] = $df;
1483 return $this->dateFormatStrings
[$type][$pref];
1487 * @param $ts Mixed: the time format which needs to be turned into a
1488 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1489 * @param $adj Bool: whether to adjust the time output according to the
1490 * user configured offset ($timecorrection)
1491 * @param $format Mixed: true to use user's date format preference
1492 * @param $timecorrection String: the time offset as returned by
1493 * validateTimeZone() in Special:Preferences
1496 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1498 $ts = $this->userAdjust( $ts, $timecorrection );
1500 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1501 return $this->sprintfDate( $df, $ts );
1505 * @param $ts Mixed: the time format which needs to be turned into a
1506 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1507 * @param $adj Bool: whether to adjust the time output according to the
1508 * user configured offset ($timecorrection)
1509 * @param $format Mixed: true to use user's date format preference
1510 * @param $timecorrection String: the time offset as returned by
1511 * validateTimeZone() in Special:Preferences
1514 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1516 $ts = $this->userAdjust( $ts, $timecorrection );
1518 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1519 return $this->sprintfDate( $df, $ts );
1523 * @param $ts Mixed: the time format which needs to be turned into a
1524 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1525 * @param $adj Bool: whether to adjust the time output according to the
1526 * user configured offset ($timecorrection)
1527 * @param $format Mixed: what format to return, if it's false output the
1528 * default one (default true)
1529 * @param $timecorrection String: the time offset as returned by
1530 * validateTimeZone() in Special:Preferences
1533 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1534 $ts = wfTimestamp( TS_MW
, $ts );
1536 $ts = $this->userAdjust( $ts, $timecorrection );
1538 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1539 return $this->sprintfDate( $df, $ts );
1542 function getMessage( $key ) {
1543 return self
::$dataCache->getSubitem( $this->mCode
, 'messages', $key );
1546 function getAllMessages() {
1547 return self
::$dataCache->getItem( $this->mCode
, 'messages' );
1550 function iconv( $in, $out, $string ) {
1551 # This is a wrapper for iconv in all languages except esperanto,
1552 # which does some nasty x-conversions beforehand
1554 # Even with //IGNORE iconv can whine about illegal characters in
1555 # *input* string. We just ignore those too.
1556 # REF: http://bugs.php.net/bug.php?id=37166
1557 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1558 wfSuppressWarnings();
1559 $text = iconv( $in, $out . '//IGNORE', $string );
1560 wfRestoreWarnings();
1564 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1565 function ucwordbreaksCallbackAscii( $matches ) {
1566 return $this->ucfirst( $matches[1] );
1569 function ucwordbreaksCallbackMB( $matches ) {
1570 return mb_strtoupper( $matches[0] );
1573 function ucCallback( $matches ) {
1574 list( $wikiUpperChars ) = self
::getCaseMaps();
1575 return strtr( $matches[1], $wikiUpperChars );
1578 function lcCallback( $matches ) {
1579 list( , $wikiLowerChars ) = self
::getCaseMaps();
1580 return strtr( $matches[1], $wikiLowerChars );
1583 function ucwordsCallbackMB( $matches ) {
1584 return mb_strtoupper( $matches[0] );
1587 function ucwordsCallbackWiki( $matches ) {
1588 list( $wikiUpperChars ) = self
::getCaseMaps();
1589 return strtr( $matches[0], $wikiUpperChars );
1592 function ucfirst( $str ) {
1596 } elseif ( $o < 128 ) {
1597 return ucfirst( $str );
1599 // fall back to more complex logic in case of multibyte strings
1600 return $this->uc( $str, true );
1604 function uc( $str, $first = false ) {
1605 if ( function_exists( 'mb_strtoupper' ) ) {
1607 if ( $this->isMultibyte( $str ) ) {
1608 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1610 return ucfirst( $str );
1613 return $this->isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
1616 if ( $this->isMultibyte( $str ) ) {
1617 list( $wikiUpperChars ) = $this->getCaseMaps();
1618 $x = $first ?
'^' : '';
1619 return preg_replace_callback(
1620 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1621 array( $this, 'ucCallback' ),
1625 return $first ?
ucfirst( $str ) : strtoupper( $str );
1630 function lcfirst( $str ) {
1633 return strval( $str );
1634 } elseif ( $o >= 128 ) {
1635 return $this->lc( $str, true );
1636 } elseif ( $o > 96 ) {
1639 $str[0] = strtolower( $str[0] );
1644 function lc( $str, $first = false ) {
1645 if ( function_exists( 'mb_strtolower' ) ) {
1647 if ( $this->isMultibyte( $str ) ) {
1648 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1650 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1653 return $this->isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
1656 if ( $this->isMultibyte( $str ) ) {
1657 list( , $wikiLowerChars ) = self
::getCaseMaps();
1658 $x = $first ?
'^' : '';
1659 return preg_replace_callback(
1660 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1661 array( $this, 'lcCallback' ),
1665 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1670 function isMultibyte( $str ) {
1671 return (bool)preg_match( '/[\x80-\xff]/', $str );
1674 function ucwords( $str ) {
1675 if ( $this->isMultibyte( $str ) ) {
1676 $str = $this->lc( $str );
1678 // regexp to find first letter in each word (i.e. after each space)
1679 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1681 // function to use to capitalize a single char
1682 if ( function_exists( 'mb_strtoupper' ) ) {
1683 return preg_replace_callback(
1685 array( $this, 'ucwordsCallbackMB' ),
1689 return preg_replace_callback(
1691 array( $this, 'ucwordsCallbackWiki' ),
1696 return ucwords( strtolower( $str ) );
1700 # capitalize words at word breaks
1701 function ucwordbreaks( $str ) {
1702 if ( $this->isMultibyte( $str ) ) {
1703 $str = $this->lc( $str );
1705 // since \b doesn't work for UTF-8, we explicitely define word break chars
1706 $breaks = "[ \-\(\)\}\{\.,\?!]";
1708 // find first letter after word break
1709 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1711 if ( function_exists( 'mb_strtoupper' ) ) {
1712 return preg_replace_callback(
1714 array( $this, 'ucwordbreaksCallbackMB' ),
1718 return preg_replace_callback(
1720 array( $this, 'ucwordsCallbackWiki' ),
1725 return preg_replace_callback(
1726 '/\b([\w\x80-\xff]+)\b/',
1727 array( $this, 'ucwordbreaksCallbackAscii' ),
1734 * Return a case-folded representation of $s
1736 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1737 * and $s2 are the same except for the case of their characters. It is not
1738 * necessary for the value returned to make sense when displayed.
1740 * Do *not* perform any other normalisation in this function. If a caller
1741 * uses this function when it should be using a more general normalisation
1742 * function, then fix the caller.
1744 function caseFold( $s ) {
1745 return $this->uc( $s );
1748 function checkTitleEncoding( $s ) {
1749 if( is_array( $s ) ) {
1750 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1752 # Check for non-UTF-8 URLs
1753 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1758 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1759 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1764 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
1767 function fallback8bitEncoding() {
1768 return self
::$dataCache->getItem( $this->mCode
, 'fallback8bitEncoding' );
1772 * Most writing systems use whitespace to break up words.
1773 * Some languages such as Chinese don't conventionally do this,
1774 * which requires special handling when breaking up words for
1777 function hasWordBreaks() {
1782 * Some languages such as Chinese require word segmentation,
1783 * Specify such segmentation when overridden in derived class.
1785 * @param $string String
1788 function segmentByWord( $string ) {
1793 * Some languages have special punctuation need to be normalized.
1794 * Make such changes here.
1796 * @param $string String
1799 function normalizeForSearch( $string ) {
1800 return self
::convertDoubleWidth($string);
1804 * convert double-width roman characters to single-width.
1805 * range: ff00-ff5f ~= 0020-007f
1807 protected static function convertDoubleWidth( $string ) {
1808 static $full = null;
1809 static $half = null;
1811 if( $full === null ) {
1812 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1813 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1814 $full = str_split( $fullWidth, 3 );
1815 $half = str_split( $halfWidth );
1818 $string = str_replace( $full, $half, $string );
1822 protected static function insertSpace( $string, $pattern ) {
1823 $string = preg_replace( $pattern, " $1 ", $string );
1824 $string = preg_replace( '/ +/', ' ', $string );
1828 function convertForSearchResult( $termsArray ) {
1829 # some languages, e.g. Chinese, need to do a conversion
1830 # in order for search results to be displayed correctly
1835 * Get the first character of a string.
1840 function firstChar( $s ) {
1843 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1844 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
1849 if ( isset( $matches[1] ) ) {
1850 if ( strlen( $matches[1] ) != 3 ) {
1854 // Break down Hangul syllables to grab the first jamo
1855 $code = utf8ToCodepoint( $matches[1] );
1856 if ( $code < 0xac00 ||
0xd7a4 <= $code ) {
1858 } elseif ( $code < 0xb098 ) {
1859 return "\xe3\x84\xb1";
1860 } elseif ( $code < 0xb2e4 ) {
1861 return "\xe3\x84\xb4";
1862 } elseif ( $code < 0xb77c ) {
1863 return "\xe3\x84\xb7";
1864 } elseif ( $code < 0xb9c8 ) {
1865 return "\xe3\x84\xb9";
1866 } elseif ( $code < 0xbc14 ) {
1867 return "\xe3\x85\x81";
1868 } elseif ( $code < 0xc0ac ) {
1869 return "\xe3\x85\x82";
1870 } elseif ( $code < 0xc544 ) {
1871 return "\xe3\x85\x85";
1872 } elseif ( $code < 0xc790 ) {
1873 return "\xe3\x85\x87";
1874 } elseif ( $code < 0xcc28 ) {
1875 return "\xe3\x85\x88";
1876 } elseif ( $code < 0xce74 ) {
1877 return "\xe3\x85\x8a";
1878 } elseif ( $code < 0xd0c0 ) {
1879 return "\xe3\x85\x8b";
1880 } elseif ( $code < 0xd30c ) {
1881 return "\xe3\x85\x8c";
1882 } elseif ( $code < 0xd558 ) {
1883 return "\xe3\x85\x8d";
1885 return "\xe3\x85\x8e";
1892 function initEncoding() {
1893 # Some languages may have an alternate char encoding option
1894 # (Esperanto X-coding, Japanese furigana conversion, etc)
1895 # If this language is used as the primary content language,
1896 # an override to the defaults can be set here on startup.
1899 function recodeForEdit( $s ) {
1900 # For some languages we'll want to explicitly specify
1901 # which characters make it into the edit box raw
1902 # or are converted in some way or another.
1903 # Note that if wgOutputEncoding is different from
1904 # wgInputEncoding, this text will be further converted
1905 # to wgOutputEncoding.
1906 global $wgEditEncoding;
1907 if( $wgEditEncoding == '' ||
$wgEditEncoding == 'UTF-8' ) {
1910 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1914 function recodeInput( $s ) {
1915 # Take the previous into account.
1916 global $wgEditEncoding;
1917 if( $wgEditEncoding != '' ) {
1918 $enc = $wgEditEncoding;
1922 if( $enc == 'UTF-8' ) {
1925 return $this->iconv( $enc, 'UTF-8', $s );
1930 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
1931 * also cleans up certain backwards-compatible sequences, converting them
1932 * to the modern Unicode equivalent.
1934 * This is language-specific for performance reasons only.
1936 function normalize( $s ) {
1937 global $wgAllUnicodeFixes;
1938 $s = UtfNormal
::cleanUp( $s );
1939 if ( $wgAllUnicodeFixes ) {
1940 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
1941 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
1948 * Transform a string using serialized data stored in the given file (which
1949 * must be in the serialized subdirectory of $IP). The file contains pairs
1950 * mapping source characters to destination characters.
1952 * The data is cached in process memory. This will go faster if you have the
1953 * FastStringSearch extension.
1955 function transformUsingPairFile( $file, $string ) {
1956 if ( !isset( $this->transformData
[$file] ) ) {
1957 $data = wfGetPrecompiledData( $file );
1958 if ( $data === false ) {
1959 throw new MWException( __METHOD__
. ": The transformation file $file is missing" );
1961 $this->transformData
[$file] = new ReplacementArray( $data );
1963 return $this->transformData
[$file]->replace( $string );
1967 * For right-to-left language support
1972 return self
::$dataCache->getItem( $this->mCode
, 'rtl' );
1976 * Return the correct HTML 'dir' attribute value for this language.
1980 return $this->isRTL() ?
'rtl' : 'ltr';
1984 * Return 'left' or 'right' as appropriate alignment for line-start
1985 * for this language's text direction.
1987 * Should be equivalent to CSS3 'start' text-align value....
1991 function alignStart() {
1992 return $this->isRTL() ?
'right' : 'left';
1996 * Return 'right' or 'left' as appropriate alignment for line-end
1997 * for this language's text direction.
1999 * Should be equivalent to CSS3 'end' text-align value....
2003 function alignEnd() {
2004 return $this->isRTL() ?
'left' : 'right';
2008 * A hidden direction mark (LRM or RLM), depending on the language direction
2012 function getDirMark() {
2013 return $this->isRTL() ?
"\xE2\x80\x8F" : "\xE2\x80\x8E";
2016 function capitalizeAllNouns() {
2017 return self
::$dataCache->getItem( $this->mCode
, 'capitalizeAllNouns' );
2021 * An arrow, depending on the language direction
2025 function getArrow() {
2026 return $this->isRTL() ?
'←' : '→';
2030 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2034 function linkPrefixExtension() {
2035 return self
::$dataCache->getItem( $this->mCode
, 'linkPrefixExtension' );
2038 function getMagicWords() {
2039 return self
::$dataCache->getItem( $this->mCode
, 'magicWords' );
2042 # Fill a MagicWord object with data from here
2043 function getMagic( $mw ) {
2044 if ( !$this->mMagicHookDone
) {
2045 $this->mMagicHookDone
= true;
2046 wfProfileIn( 'LanguageGetMagic' );
2047 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
2048 wfProfileOut( 'LanguageGetMagic' );
2050 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
2051 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
2053 $magicWords = $this->getMagicWords();
2054 if ( isset( $magicWords[$mw->mId
] ) ) {
2055 $rawEntry = $magicWords[$mw->mId
];
2061 if( !is_array( $rawEntry ) ) {
2062 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
2064 $mw->mCaseSensitive
= $rawEntry[0];
2065 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
2070 * Add magic words to the extension array
2072 function addMagicWordsByLang( $newWords ) {
2073 $code = $this->getCode();
2074 $fallbackChain = array();
2075 while ( $code && !in_array( $code, $fallbackChain ) ) {
2076 $fallbackChain[] = $code;
2077 $code = self
::getFallbackFor( $code );
2079 if ( !in_array( 'en', $fallbackChain ) ) {
2080 $fallbackChain[] = 'en';
2082 $fallbackChain = array_reverse( $fallbackChain );
2083 foreach ( $fallbackChain as $code ) {
2084 if ( isset( $newWords[$code] ) ) {
2085 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
2091 * Get special page names, as an associative array
2092 * case folded alias => real name
2094 function getSpecialPageAliases() {
2095 // Cache aliases because it may be slow to load them
2096 if ( is_null( $this->mExtendedSpecialPageAliases
) ) {
2098 $this->mExtendedSpecialPageAliases
=
2099 self
::$dataCache->getItem( $this->mCode
, 'specialPageAliases' );
2100 wfRunHooks( 'LanguageGetSpecialPageAliases',
2101 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
2104 return $this->mExtendedSpecialPageAliases
;
2108 * Italic is unsuitable for some languages
2110 * @param $text String: the text to be emphasized.
2113 function emphasize( $text ) {
2114 return "<em>$text</em>";
2118 * Normally we output all numbers in plain en_US style, that is
2119 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2120 * point twohundredthirtyfive. However this is not sutable for all
2121 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2122 * Icelandic just want to use commas instead of dots, and dots instead
2123 * of commas like "293.291,235".
2125 * An example of this function being called:
2127 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2130 * See LanguageGu.php for the Gujarati implementation and
2131 * $separatorTransformTable on MessageIs.php for
2132 * the , => . and . => , implementation.
2134 * @todo check if it's viable to use localeconv() for the decimal
2136 * @param $number Mixed: the string to be formatted, should be an integer
2137 * or a floating point number.
2138 * @param $nocommafy Bool: set to true for special numbers like dates
2141 function formatNum( $number, $nocommafy = false ) {
2142 global $wgTranslateNumerals;
2143 if ( !$nocommafy ) {
2144 $number = $this->commafy( $number );
2145 $s = $this->separatorTransformTable();
2147 $number = strtr( $number, $s );
2151 if ( $wgTranslateNumerals ) {
2152 $s = $this->digitTransformTable();
2154 $number = strtr( $number, $s );
2161 function parseFormattedNumber( $number ) {
2162 $s = $this->digitTransformTable();
2164 $number = strtr( $number, array_flip( $s ) );
2167 $s = $this->separatorTransformTable();
2169 $number = strtr( $number, array_flip( $s ) );
2172 $number = strtr( $number, array( ',' => '' ) );
2177 * Adds commas to a given number
2182 function commafy($_) {
2183 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2186 function digitTransformTable() {
2187 return self
::$dataCache->getItem( $this->mCode
, 'digitTransformTable' );
2190 function separatorTransformTable() {
2191 return self
::$dataCache->getItem( $this->mCode
, 'separatorTransformTable' );
2195 * Take a list of strings and build a locale-friendly comma-separated
2196 * list, using the local comma-separator message.
2197 * The last two strings are chained with an "and".
2202 function listToText( $l ) {
2204 $m = count( $l ) - 1;
2206 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2208 for ( $i = $m; $i >= 0; $i-- ) {
2211 } else if( $i == $m - 1 ) {
2212 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2214 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2222 * Take a list of strings and build a locale-friendly comma-separated
2223 * list, using the local comma-separator message.
2224 * @param $list array of strings to put in a comma list
2227 function commaList( $list ) {
2232 array( 'parsemag', 'escapenoentities', 'language' => $this )
2238 * Take a list of strings and build a locale-friendly semicolon-separated
2239 * list, using the local semicolon-separator message.
2240 * @param $list array of strings to put in a semicolon list
2243 function semicolonList( $list ) {
2247 'semicolon-separator',
2248 array( 'parsemag', 'escapenoentities', 'language' => $this )
2254 * Same as commaList, but separate it with the pipe instead.
2255 * @param $list array of strings to put in a pipe list
2258 function pipeList( $list ) {
2263 array( 'escapenoentities', 'language' => $this )
2269 * Truncate a string to a specified length in bytes, appending an optional
2270 * string (e.g. for ellipses)
2272 * The database offers limited byte lengths for some columns in the database;
2273 * multi-byte character sets mean we need to ensure that only whole characters
2274 * are included, otherwise broken characters can be passed to the user
2276 * If $length is negative, the string will be truncated from the beginning
2278 * @param $string String to truncate
2279 * @param $length Int: maximum length (excluding ellipses)
2280 * @param $ellipsis String to append to the truncated text
2283 function truncate( $string, $length, $ellipsis = '...' ) {
2284 # Use the localized ellipsis character
2285 if ( $ellipsis == '...' ) {
2286 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2288 # Check if there is no need to truncate
2289 if ( $length == 0 ) {
2291 } elseif ( strlen( $string ) <= abs( $length ) ) {
2294 $stringOriginal = $string;
2295 if ( $length > 0 ) {
2296 $string = substr( $string, 0, $length ); // xyz...
2297 $string = $this->removeBadCharLast( $string );
2298 $string = $string . $ellipsis;
2300 $string = substr( $string, $length ); // ...xyz
2301 $string = $this->removeBadCharFirst( $string );
2302 $string = $ellipsis . $string;
2304 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181)
2305 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
2308 return $stringOriginal;
2313 * Remove bytes that represent an incomplete Unicode character
2314 * at the end of string (e.g. bytes of the char are missing)
2316 * @param $string String
2319 protected function removeBadCharLast( $string ) {
2320 $char = ord( $string[strlen( $string ) - 1] );
2322 if ( $char >= 0xc0 ) {
2323 # We got the first byte only of a multibyte char; remove it.
2324 $string = substr( $string, 0, -1 );
2325 } elseif ( $char >= 0x80 &&
2326 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2327 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
2329 # We chopped in the middle of a character; remove it
2336 * Remove bytes that represent an incomplete Unicode character
2337 * at the start of string (e.g. bytes of the char are missing)
2339 * @param $string String
2342 protected function removeBadCharFirst( $string ) {
2343 $char = ord( $string[0] );
2344 if ( $char >= 0x80 && $char < 0xc0 ) {
2345 # We chopped in the middle of a character; remove the whole thing
2346 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2352 * Truncate a string of valid HTML to a specified length in bytes,
2353 * appending an optional string (e.g. for ellipses), and return valid HTML
2355 * This is only intended for styled/linked text, such as HTML with
2356 * tags like <span> and <a>, were the tags are self-contained (valid HTML)
2358 * Note: tries to fix broken HTML with MWTidy
2360 * @param string $text String to truncate
2361 * @param int $length (zero/positive) Maximum length (excluding ellipses)
2362 * @param string $ellipsis String to append to the truncated text
2365 function truncateHtml( $text, $length, $ellipsis = '...' ) {
2366 # Use the localized ellipsis character
2367 if ( $ellipsis == '...' ) {
2368 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2370 # Check if there is no need to truncate
2371 if ( $length <= 0 ) {
2372 return $ellipsis; // no text shown, nothing to format
2373 } elseif ( strlen( $text ) <= $length ) {
2374 return $text; // string short enough even *with* HTML
2376 $text = MWTidy
::tidy( $text ); // fix tags
2377 $displayLen = 0; // innerHTML legth so far
2378 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
2379 $tagType = 0; // 0-open, 1-close
2380 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
2381 $entityState = 0; // 0-not entity, 1-entity
2382 $tag = $ret = $ch = '';
2383 $openTags = array();
2384 $textLen = strlen($text);
2385 for( $pos = 0; $pos < $textLen; ++
$pos ) {
2387 $lastCh = $pos ?
$text[$pos - 1] : '';
2388 $ret .= $ch; // add to result string
2390 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
2391 $entityState = 0; // for bad HTML
2392 $bracketState = 1; // tag started (checking for backslash)
2393 } elseif ( $ch == '>' ) {
2394 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
2395 $entityState = 0; // for bad HTML
2396 $bracketState = 0; // out of brackets
2397 } elseif ( $bracketState == 1 ) {
2399 $tagType = 1; // close tag (e.g. "</span>")
2401 $tagType = 0; // open tag (e.g. "<span>")
2404 $bracketState = 2; // building tag name
2405 } elseif ( $bracketState == 2 ) {
2409 // Name found (e.g. "<a href=..."), add on tag attributes...
2410 $pos +
= $this->truncate_skip( $ret, $text, "<>", $pos +
1 );
2412 } elseif ( $bracketState == 0 ) {
2413 if ( $entityState ) {
2416 $displayLen++
; // entity is one displayed char
2420 $entityState = 1; // entity found, (e.g. " ")
2422 $displayLen++
; // this char is displayed
2423 // Add on the other display text after this...
2424 $skipped = $this->truncate_skip(
2425 $ret, $text, "<>&", $pos +
1, $length - $displayLen );
2426 $displayLen +
= $skipped;
2431 # Consider truncation once the display length has reached the maximim.
2432 # Double-check that we're not in the middle of a bracket/entity...
2433 if ( $displayLen >= $length && $bracketState == 0 && $entityState == 0 ) {
2434 if ( !$testingEllipsis ) {
2435 $testingEllipsis = true;
2436 # Save where we are; we will truncate here unless
2437 # the ellipsis actually makes the string longer.
2438 $pOpenTags = $openTags; // save state
2439 $pRet = $ret; // save state
2440 } elseif ( $displayLen > ( $length +
strlen( $ellipsis ) ) ) {
2441 # Ellipsis won't make string longer/equal, the truncation point was OK.
2442 $openTags = $pOpenTags; // reload state
2443 $ret = $this->removeBadCharLast( $pRet ); // reload state, multi-byte char fix
2444 $ret .= $ellipsis; // add ellipsis
2449 if ( $displayLen == 0 ) {
2450 return ''; // no text shown, nothing to format
2452 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags ); // for bad HTML
2453 while ( count( $openTags ) > 0 ) {
2454 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
2459 // truncateHtml() helper function
2460 // like strcspn() but adds the skipped chars to $ret
2461 private function truncate_skip( &$ret, $text, $search, $start, $len = -1 ) {
2463 if( $start < strlen( $text ) ) {
2464 $skipCount = strcspn( $text, $search, $start, $len );
2465 $ret .= substr( $text, $start, $skipCount );
2470 // truncateHtml() helper function
2471 // (a) push or pop $tag from $openTags as needed
2472 // (b) clear $tag value
2473 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
2474 $tag = ltrim( $tag );
2476 if( $tagType == 0 && $lastCh != '/' ) {
2477 $openTags[] = $tag; // tag opened (didn't close itself)
2478 } else if( $tagType == 1 ) {
2479 if( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
2480 array_pop( $openTags ); // tag closed
2488 * Grammatical transformations, needed for inflected languages
2489 * Invoked by putting {{grammar:case|word}} in a message
2491 * @param $word string
2492 * @param $case string
2495 function convertGrammar( $word, $case ) {
2496 global $wgGrammarForms;
2497 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
2498 return $wgGrammarForms[$this->getCode()][$case][$word];
2504 * Provides an alternative text depending on specified gender.
2505 * Usage {{gender:username|masculine|feminine|neutral}}.
2506 * username is optional, in which case the gender of current user is used,
2507 * but only in (some) interface messages; otherwise default gender is used.
2508 * If second or third parameter are not specified, masculine is used.
2509 * These details may be overriden per language.
2511 function gender( $gender, $forms ) {
2512 if ( !count( $forms ) ) {
2515 $forms = $this->preConvertPlural( $forms, 2 );
2516 if ( $gender === 'male' ) {
2519 if ( $gender === 'female' ) {
2522 return isset( $forms[2] ) ?
$forms[2] : $forms[0];
2526 * Plural form transformations, needed for some languages.
2527 * For example, there are 3 form of plural in Russian and Polish,
2528 * depending on "count mod 10". See [[w:Plural]]
2529 * For English it is pretty simple.
2531 * Invoked by putting {{plural:count|wordform1|wordform2}}
2532 * or {{plural:count|wordform1|wordform2|wordform3}}
2534 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2536 * @param $count Integer: non-localized number
2537 * @param $forms Array: different plural forms
2538 * @return string Correct form of plural for $count in this language
2540 function convertPlural( $count, $forms ) {
2541 if ( !count( $forms ) ) {
2544 $forms = $this->preConvertPlural( $forms, 2 );
2546 return ( $count == 1 ) ?
$forms[0] : $forms[1];
2550 * Checks that convertPlural was given an array and pads it to requested
2551 * amound of forms by copying the last one.
2553 * @param $count Integer: How many forms should there be at least
2554 * @param $forms Array of forms given to convertPlural
2555 * @return array Padded array of forms or an exception if not an array
2557 protected function preConvertPlural( /* Array */ $forms, $count ) {
2558 while ( count( $forms ) < $count ) {
2559 $forms[] = $forms[count( $forms ) - 1];
2565 * For translating of expiry times
2566 * @param $str String: the validated block time in English
2567 * @return Somehow translated block time
2568 * @see LanguageFi.php for example implementation
2570 function translateBlockExpiry( $str ) {
2571 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2573 if ( $scBlockExpiryOptions == '-' ) {
2577 foreach ( explode( ',', $scBlockExpiryOptions) as $option ) {
2578 if ( strpos( $option, ':' ) === false ) {
2581 list( $show, $value ) = explode( ':', $option );
2582 if ( strcmp( $str, $value ) == 0 ) {
2583 return htmlspecialchars( trim( $show ) );
2591 * languages like Chinese need to be segmented in order for the diff
2594 * @param $text String
2597 function segmentForDiff( $text ) {
2602 * and unsegment to show the result
2604 * @param $text String
2607 function unsegmentForDiff( $text ) {
2611 # convert text to all supported variants
2612 function autoConvertToAllVariants( $text ) {
2613 return $this->mConverter
->autoConvertToAllVariants( $text );
2616 # convert text to different variants of a language.
2617 function convert( $text ) {
2618 return $this->mConverter
->convert( $text );
2621 # Convert a Title object to a string in the preferred variant
2622 function convertTitle( $title ) {
2623 return $this->mConverter
->convertTitle( $title );
2626 # Check if this is a language with variants
2627 function hasVariants() {
2628 return sizeof( $this->getVariants() ) > 1;
2631 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2632 function armourMath( $text ) {
2633 return $this->mConverter
->armourMath( $text );
2637 * Perform output conversion on a string, and encode for safe HTML output.
2638 * @param $text String text to be converted
2639 * @param $isTitle Bool whether this conversion is for the article title
2641 * @todo this should get integrated somewhere sane
2643 function convertHtml( $text, $isTitle = false ) {
2644 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2647 function convertCategoryKey( $key ) {
2648 return $this->mConverter
->convertCategoryKey( $key );
2652 * Get the list of variants supported by this langauge
2653 * see sample implementation in LanguageZh.php
2655 * @return array an array of language codes
2657 function getVariants() {
2658 return $this->mConverter
->getVariants();
2661 function getPreferredVariant( $fromUser = true, $fromHeader = false ) {
2662 return $this->mConverter
->getPreferredVariant( $fromUser, $fromHeader );
2666 * If a language supports multiple variants, it is
2667 * possible that non-existing link in one variant
2668 * actually exists in another variant. this function
2669 * tries to find it. See e.g. LanguageZh.php
2671 * @param $link String: the name of the link
2672 * @param $nt Mixed: the title object of the link
2673 * @param $ignoreOtherCond Boolean: to disable other conditions when
2674 * we need to transclude a template or update a category's link
2675 * @return null the input parameters may be modified upon return
2677 function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
2678 $this->mConverter
->findVariantLink( $link, $nt, $ignoreOtherCond );
2682 * If a language supports multiple variants, converts text
2683 * into an array of all possible variants of the text:
2684 * 'variant' => text in that variant
2686 function convertLinkToAllVariants( $text ) {
2687 return $this->mConverter
->convertLinkToAllVariants( $text );
2691 * returns language specific options used by User::getPageRenderHash()
2692 * for example, the preferred language variant
2696 function getExtraHashOptions() {
2697 return $this->mConverter
->getExtraHashOptions();
2701 * For languages that support multiple variants, the title of an
2702 * article may be displayed differently in different variants. this
2703 * function returns the apporiate title defined in the body of the article.
2707 function getParsedTitle() {
2708 return $this->mConverter
->getParsedTitle();
2712 * Enclose a string with the "no conversion" tag. This is used by
2713 * various functions in the Parser
2715 * @param $text String: text to be tagged for no conversion
2717 * @return string the tagged text
2719 function markNoConversion( $text, $noParse = false ) {
2720 return $this->mConverter
->markNoConversion( $text, $noParse );
2724 * A regular expression to match legal word-trailing characters
2725 * which should be merged onto a link of the form [[foo]]bar.
2729 function linkTrail() {
2730 return self
::$dataCache->getItem( $this->mCode
, 'linkTrail' );
2733 function getLangObj() {
2738 * Get the RFC 3066 code for this language object
2740 function getCode() {
2741 return $this->mCode
;
2744 function setCode( $code ) {
2745 $this->mCode
= $code;
2749 * Get the name of a file for a certain language code
2750 * @param $prefix string Prepend this to the filename
2751 * @param $code string Language code
2752 * @param $suffix string Append this to the filename
2753 * @return string $prefix . $mangledCode . $suffix
2755 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2756 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2760 * Get the language code from a file name. Inverse of getFileName()
2761 * @param $filename string $prefix . $languageCode . $suffix
2762 * @param $prefix string Prefix before the language code
2763 * @param $suffix string Suffix after the language code
2764 * @return Language code, or false if $prefix or $suffix isn't found
2766 static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
2768 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
2769 preg_quote( $suffix, '/' ) . '/', $filename, $m );
2770 if ( !count( $m ) ) {
2773 return str_replace( '_', '-', strtolower( $m[1] ) );
2776 static function getMessagesFileName( $code ) {
2778 return self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2781 static function getClassFileName( $code ) {
2783 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2787 * Get the fallback for a given language
2789 static function getFallbackFor( $code ) {
2790 if ( $code === 'en' ) {
2794 return self
::getLocalisationCache()->getItem( $code, 'fallback' );
2799 * Get all messages for a given language
2800 * WARNING: this may take a long time
2802 static function getMessagesFor( $code ) {
2803 return self
::getLocalisationCache()->getItem( $code, 'messages' );
2807 * Get a message for a given language
2809 static function getMessageFor( $key, $code ) {
2810 return self
::getLocalisationCache()->getSubitem( $code, 'messages', $key );
2813 function fixVariableInNamespace( $talk ) {
2814 if ( strpos( $talk, '$1' ) === false ) {
2818 global $wgMetaNamespace;
2819 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2821 # Allow grammar transformations
2822 # Allowing full message-style parsing would make simple requests
2823 # such as action=raw much more expensive than they need to be.
2824 # This will hopefully cover most cases.
2825 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2826 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2827 return str_replace( ' ', '_', $talk );
2830 function replaceGrammarInNamespace( $m ) {
2831 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2834 static function getCaseMaps() {
2835 static $wikiUpperChars, $wikiLowerChars;
2836 if ( isset( $wikiUpperChars ) ) {
2837 return array( $wikiUpperChars, $wikiLowerChars );
2840 wfProfileIn( __METHOD__
);
2841 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2842 if ( $arr === false ) {
2843 throw new MWException(
2844 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2847 wfProfileOut( __METHOD__
);
2848 return array( $wikiUpperChars, $wikiLowerChars );
2851 function formatTimePeriod( $seconds ) {
2852 if ( $seconds < 10 ) {
2853 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2854 } elseif ( $seconds < 60 ) {
2855 return $this->formatNum( round( $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2856 } elseif ( $seconds < 3600 ) {
2857 $minutes = floor( $seconds / 60 );
2858 $secondsPart = round( fmod( $seconds, 60 ) );
2859 if ( $secondsPart == 60 ) {
2863 return $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2864 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2866 $hours = floor( $seconds / 3600 );
2867 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2868 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2869 if ( $secondsPart == 60 ) {
2873 if ( $minutes == 60 ) {
2877 return $this->formatNum( $hours ) . $this->getMessageFromDB( 'hours-abbrev' ) . ' ' .
2878 $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2879 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2883 function formatBitrate( $bps ) {
2884 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2886 return $this->formatNum( $bps ) . $units[0];
2888 $unitIndex = floor( log10( $bps ) / 3 );
2889 $mantissa = $bps / pow( 1000, $unitIndex );
2890 if ( $mantissa < 10 ) {
2891 $mantissa = round( $mantissa, 1 );
2893 $mantissa = round( $mantissa );
2895 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2899 * Format a size in bytes for output, using an appropriate
2900 * unit (B, KB, MB or GB) according to the magnitude in question
2902 * @param $size Size to format
2903 * @return string Plain text (not HTML)
2905 function formatSize( $size ) {
2906 // For small sizes no decimal places necessary
2908 if( $size > 1024 ) {
2909 $size = $size / 1024;
2910 if( $size > 1024 ) {
2911 $size = $size / 1024;
2912 // For MB and bigger two decimal places are smarter
2914 if( $size > 1024 ) {
2915 $size = $size / 1024;
2916 $msg = 'size-gigabytes';
2918 $msg = 'size-megabytes';
2921 $msg = 'size-kilobytes';
2924 $msg = 'size-bytes';
2926 $size = round( $size, $round );
2927 $text = $this->getMessageFromDB( $msg );
2928 return str_replace( '$1', $this->formatNum( $size ), $text );
2932 * Get the conversion rule title, if any.
2934 function getConvRuleTitle() {
2935 return $this->mConverter
->getConvRuleTitle();
2939 * Given a string, convert it to a (hopefully short) key that can be used
2940 * for efficient sorting. A binary sort according to the sortkeys
2941 * corresponds to a logical sort of the corresponding strings. Current
2942 * code expects that a null character should sort before all others, but
2943 * has no other particular expectations (and that one can be changed if
2946 * @param string $string UTF-8 string
2947 * @return string Binary sortkey
2949 public function convertToSortkey( $string ) {
2950 # Fake function for now
2951 return strtoupper( $string );
2955 * Does it make sense for lists to be split up into sections based on their
2956 * first letter? Logogram-based scripts probably want to return false.
2958 * TODO: Use this in CategoryPage.php.
2962 public function usesFirstLettersInLists() {
2967 * Given a string, return the logical "first letter" to be used for
2968 * grouping on category pages and so on. This has to be coordinated
2969 * carefully with convertToSortkey(), or else the sorted list might jump
2970 * back and forth between the same "initial letters" or other pathological
2971 * behavior. For instance, if you just return the first character, but "a"
2972 * sorts the same as "A" based on convertToSortkey(), then you might get a
2984 * etc., assuming for the sake of argument that $wgCapitalLinks is false.
2985 * Obviously, this is ignored if usesFirstLettersInLists() is false.
2987 * @param string $string UTF-8 string
2988 * @return string UTF-8 string corresponding to the first letter of input
2990 public function firstLetterForLists( $string ) {
2991 if ( $string[0] == "\0" ) {
2992 $string = substr( $string, 1 );
2994 return strtoupper( mb_substr( $string, 0, 1 ) );