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
= $langobj;}
38 function convert($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, &$n) {}
43 function getExtraHashOptions() {return '';}
44 function getParsedTitle() {return '';}
45 function markNoConversion($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; }
52 * Internationalisation code
56 var $mConverter, $mVariants, $mCode, $mLoaded = false;
57 var $mMagicExtensions = array(), $mMagicHookDone = false;
59 static public $mLocalisationKeys = array( 'fallback', 'namespaceNames',
60 'skinNames', 'mathNames',
61 'bookstoreList', 'magicWords', 'messages', 'rtl', 'digitTransformTable',
62 'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
63 'defaultUserOptionOverrides', 'linkTrail', 'namespaceAliases',
64 'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
65 'defaultDateFormat', 'extraUserToggles', 'specialPageAliases' );
67 static public $mMergeableMapKeys = array( 'messages', 'namespaceNames', 'mathNames',
68 'dateFormats', 'defaultUserOptionOverrides', 'magicWords' );
70 static public $mMergeableListKeys = array( 'extraUserToggles' );
72 static public $mMergeableAliasListKeys = array( 'specialPageAliases' );
74 static public $mLocalisationCache = array();
76 static public $mWeekdayMsgs = array(
77 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
81 static public $mWeekdayAbbrevMsgs = array(
82 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
85 static public $mMonthMsgs = array(
86 'january', 'february', 'march', 'april', 'may_long', 'june',
87 'july', 'august', 'september', 'october', 'november',
90 static public $mMonthGenMsgs = array(
91 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
92 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
95 static public $mMonthAbbrevMsgs = array(
96 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
97 'sep', 'oct', 'nov', 'dec'
100 static public $mIranianCalendarMonthMsgs = array(
101 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
102 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
103 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
104 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
107 static public $mHebrewCalendarMonthMsgs = array(
108 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
109 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
110 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
111 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
112 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
115 static public $mHebrewCalendarMonthGenMsgs = array(
116 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
117 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
118 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
119 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
120 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
124 * Create a language object for a given language code
126 static function factory( $code ) {
128 static $recursionLevel = 0;
130 if ( $code == 'en' ) {
133 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
134 // Preload base classes to work around APC/PHP5 bug
135 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
136 include_once("$IP/languages/classes/$class.deps.php");
138 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
139 include_once("$IP/languages/classes/$class.php");
143 if ( $recursionLevel > 5 ) {
144 throw new MWException( "Language fallback loop detected when creating class $class\n" );
147 if( ! class_exists( $class ) ) {
148 $fallback = Language
::getFallbackFor( $code );
150 $lang = Language
::factory( $fallback );
152 $lang->setCode( $code );
160 function __construct() {
161 $this->mConverter
= new FakeConverter($this);
162 // Set the code to the name of the descendant
163 if ( get_class( $this ) == 'Language' ) {
166 $this->mCode
= str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
171 * Hook which will be called if this is the content language.
172 * Descendants can use this to register hook functions or modify globals
174 function initContLang() {}
177 * @deprecated Use User::getDefaultOptions()
180 function getDefaultUserOptions() {
181 wfDeprecated( __METHOD__
);
182 return User
::getDefaultOptions();
185 function getFallbackLanguageCode() {
186 return self
::getFallbackFor( $this->mCode
);
190 * Exports $wgBookstoreListEn
193 function getBookstoreList() {
195 return $this->bookstoreList
;
201 function getNamespaces() {
203 return $this->namespaceNames
;
207 * A convenience function that returns the same thing as
208 * getNamespaces() except with the array values changed to ' '
209 * where it found '_', useful for producing output to be displayed
210 * e.g. in <select> forms.
214 function getFormattedNamespaces() {
215 $ns = $this->getNamespaces();
216 foreach($ns as $k => $v) {
217 $ns[$k] = strtr($v, '_', ' ');
223 * Get a namespace value by key
225 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
226 * echo $mw_ns; // prints 'MediaWiki'
229 * @param $index Int: the array key of the namespace to return
230 * @return mixed, string if the namespace value exists, otherwise false
232 function getNsText( $index ) {
233 $ns = $this->getNamespaces();
234 return isset( $ns[$index] ) ?
$ns[$index] : false;
238 * A convenience function that returns the same thing as
239 * getNsText() except with '_' changed to ' ', useful for
244 function getFormattedNsText( $index ) {
245 $ns = $this->getNsText( $index );
246 return strtr($ns, '_', ' ');
250 * Get a namespace key by value, case insensitive.
251 * Only matches namespace names for the current language, not the
252 * canonical ones defined in Namespace.php.
254 * @param $text String
255 * @return mixed An integer if $text is a valid value otherwise false
257 function getLocalNsIndex( $text ) {
259 $lctext = $this->lc($text);
260 return isset( $this->mNamespaceIds
[$lctext] ) ?
$this->mNamespaceIds
[$lctext] : false;
264 * Get a namespace key by value, case insensitive. Canonical namespace
265 * names override custom ones defined for the current language.
267 * @param $text String
268 * @return mixed An integer if $text is a valid value otherwise false
270 function getNsIndex( $text ) {
272 $lctext = $this->lc($text);
273 if( ( $ns = MWNamespace
::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
274 return isset( $this->mNamespaceIds
[$lctext] ) ?
$this->mNamespaceIds
[$lctext] : false;
278 * short names for language variants used for language conversion links.
280 * @param $code String
283 function getVariantname( $code ) {
284 return $this->getMessageFromDB( "variantname-$code" );
287 function specialPage( $name ) {
288 $aliases = $this->getSpecialPageAliases();
289 if ( isset( $aliases[$name][0] ) ) {
290 $name = $aliases[$name][0];
292 return $this->getNsText(NS_SPECIAL
) . ':' . $name;
295 function getQuickbarSettings() {
297 $this->getMessage( 'qbsettings-none' ),
298 $this->getMessage( 'qbsettings-fixedleft' ),
299 $this->getMessage( 'qbsettings-fixedright' ),
300 $this->getMessage( 'qbsettings-floatingleft' ),
301 $this->getMessage( 'qbsettings-floatingright' )
305 function getSkinNames() {
307 return $this->skinNames
;
310 function getMathNames() {
312 return $this->mathNames
;
315 function getDatePreferences() {
317 return $this->datePreferences
;
320 function getDateFormats() {
322 return $this->dateFormats
;
325 function getDefaultDateFormat() {
327 return $this->defaultDateFormat
;
330 function getDatePreferenceMigrationMap() {
332 return $this->datePreferenceMigrationMap
;
335 function getDefaultUserOptionOverrides() {
337 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
338 if (is_array($this->defaultUserOptionOverrides
)) {
339 return $this->defaultUserOptionOverrides
;
345 function getExtraUserToggles() {
347 return $this->extraUserToggles
;
350 function getUserToggle( $tog ) {
351 return $this->getMessageFromDB( "tog-$tog" );
355 * Get language names, indexed by code.
356 * If $customisedOnly is true, only returns codes with a messages file
358 public static function getLanguageNames( $customisedOnly = false ) {
359 global $wgLanguageNames, $wgExtraLanguageNames;
360 $allNames = $wgExtraLanguageNames +
$wgLanguageNames;
361 if ( !$customisedOnly ) {
367 $dir = opendir( "$IP/languages/messages" );
368 while( false !== ( $file = readdir( $dir ) ) ) {
370 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
371 $code = str_replace( '_', '-', strtolower( $m[1] ) );
372 if ( isset( $allNames[$code] ) ) {
373 $names[$code] = $allNames[$code];
382 * Ugly hack to get a message maybe from the MediaWiki namespace, if this
383 * language object is the content or user language.
385 function getMessageFromDB( $msg ) {
386 global $wgContLang, $wgLang;
387 if ( $wgContLang->getCode() == $this->getCode() ) {
389 return wfMsgForContent( $msg );
390 } elseif ( $wgLang->getCode() == $this->getCode() ) {
392 return wfMsg( $msg );
394 # Neither, get from localisation
395 return $this->getMessage( $msg );
399 function getLanguageName( $code ) {
400 $names = self
::getLanguageNames();
401 if ( !array_key_exists( $code, $names ) ) {
404 return $names[$code];
407 function getMonthName( $key ) {
408 return $this->getMessageFromDB( self
::$mMonthMsgs[$key-1] );
411 function getMonthNameGen( $key ) {
412 return $this->getMessageFromDB( self
::$mMonthGenMsgs[$key-1] );
415 function getMonthAbbreviation( $key ) {
416 return $this->getMessageFromDB( self
::$mMonthAbbrevMsgs[$key-1] );
419 function getWeekdayName( $key ) {
420 return $this->getMessageFromDB( self
::$mWeekdayMsgs[$key-1] );
423 function getWeekdayAbbreviation( $key ) {
424 return $this->getMessageFromDB( self
::$mWeekdayAbbrevMsgs[$key-1] );
427 function getIranianCalendarMonthName( $key ) {
428 return $this->getMessageFromDB( self
::$mIranianCalendarMonthMsgs[$key-1] );
431 function getHebrewCalendarMonthName( $key ) {
432 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthMsgs[$key-1] );
435 function getHebrewCalendarMonthNameGen( $key ) {
436 return $this->getMessageFromDB( self
::$mHebrewCalendarMonthGenMsgs[$key-1] );
441 * Used by date() and time() to adjust the time output.
443 * @param $ts Int the time in date('YmdHis') format
444 * @param $tz Mixed: adjust the time by this amount (default false, mean we
445 * get user timecorrection setting)
448 function userAdjust( $ts, $tz = false ) {
449 global $wgUser, $wgLocalTZoffset;
452 $tz = $wgUser->getOption( 'timecorrection' );
455 # minutes and hours differences:
460 # Global offset in minutes.
461 if( isset($wgLocalTZoffset) ) {
462 if( $wgLocalTZoffset >= 0 ) {
463 $hrDiff = floor($wgLocalTZoffset / 60);
465 $hrDiff = ceil($wgLocalTZoffset / 60);
467 $minDiff = $wgLocalTZoffset %
60;
469 } elseif ( strpos( $tz, ':' ) !== false ) {
470 $tzArray = explode( ':', $tz );
471 $hrDiff = intval($tzArray[0]);
472 $minDiff = intval($hrDiff < 0 ?
-$tzArray[1] : $tzArray[1]);
474 $hrDiff = intval( $tz );
477 # No difference ? Return time unchanged
478 if ( 0 == $hrDiff && 0 == $minDiff ) { return $ts; }
480 wfSuppressWarnings(); // E_STRICT system time bitching
481 # Generate an adjusted date
483 (int)substr( $ts, 8, 2) ) +
$hrDiff, # Hours
484 (int)substr( $ts, 10, 2 ) +
$minDiff, # Minutes
485 (int)substr( $ts, 12, 2 ), # Seconds
486 (int)substr( $ts, 4, 2 ), # Month
487 (int)substr( $ts, 6, 2 ), # Day
488 (int)substr( $ts, 0, 4 ) ); #Year
490 $date = date( 'YmdHis', $t );
497 * This is a workalike of PHP's date() function, but with better
498 * internationalisation, a reduced set of format characters, and a better
501 * Supported format characters are dDjlNwzWFmMntLYyaAgGhHiscrU. See the
502 * PHP manual for definitions. There are a number of extensions, which
505 * xn Do not translate digits of the next numeric format character
506 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
507 * xr Use roman numerals for the next numeric format character
508 * xh Use hebrew numerals for the next numeric format character
510 * xg Genitive month name
512 * xij j (day number) in Iranian calendar
513 * xiF F (month name) in Iranian calendar
514 * xin n (month number) in Iranian calendar
515 * xiY Y (full year) in Iranian calendar
517 * xjj j (day number) in Hebrew calendar
518 * xjF F (month name) in Hebrew calendar
519 * xjt t (days in month) in Hebrew calendar
520 * xjx xg (genitive month name) in Hebrew calendar
521 * xjn n (month number) in Hebrew calendar
522 * xjY Y (full year) in Hebrew calendar
524 * xkY Y (full year) in Thai solar calendar. Months and days are
525 * identical to the Gregorian calendar
527 * Characters enclosed in double quotes will be considered literal (with
528 * the quotes themselves removed). Unmatched quotes will be considered
529 * literal quotes. Example:
531 * "The month is" F => The month is January
534 * Backslash escaping is also supported.
536 * Input timestamp is assumed to be pre-normalized to the desired local
539 * @param $format String
540 * @param $ts String: 14-character timestamp
544 function sprintfDate( $format, $ts ) {
554 for ( $p = 0; $p < strlen( $format ); $p++
) {
557 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
558 $code .= $format[++
$p];
561 if ( ( $code === 'xi' ||
$code == 'xj' ||
$code == 'xk' ) && $p < strlen( $format ) - 1 ) {
562 $code .= $format[++
$p];
573 $rawToggle = !$rawToggle;
582 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
585 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
586 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
589 $num = substr( $ts, 6, 2 );
592 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
593 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) +
1 );
596 $num = intval( substr( $ts, 6, 2 ) );
599 if ( !$iranian ) $iranian = self
::tsToIranian( $ts );
603 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
607 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
608 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) +
1 );
611 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
612 $w = gmdate( 'w', $unix );
616 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
617 $num = gmdate( 'w', $unix );
620 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
621 $num = gmdate( 'z', $unix );
624 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
625 $num = gmdate( 'W', $unix );
628 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
631 if ( !$iranian ) $iranian = self
::tsToIranian( $ts );
632 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
635 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
636 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
639 $num = substr( $ts, 4, 2 );
642 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
645 $num = intval( substr( $ts, 4, 2 ) );
648 if ( !$iranian ) $iranian = self
::tsToIranian( $ts );
652 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
656 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
657 $num = gmdate( 't', $unix );
660 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
664 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
665 $num = gmdate( 'L', $unix );
668 $num = substr( $ts, 0, 4 );
671 if ( !$iranian ) $iranian = self
::tsToIranian( $ts );
675 if ( !$hebrew ) $hebrew = self
::tsToHebrew( $ts );
679 if ( !$thai ) $thai = self
::tsToThai( $ts );
683 $num = substr( $ts, 2, 2 );
686 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'am' : 'pm';
689 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ?
'AM' : 'PM';
692 $h = substr( $ts, 8, 2 );
693 $num = $h %
12 ?
$h %
12 : 12;
696 $num = intval( substr( $ts, 8, 2 ) );
699 $h = substr( $ts, 8, 2 );
700 $num = sprintf( '%02d', $h %
12 ?
$h %
12 : 12 );
703 $num = substr( $ts, 8, 2 );
706 $num = substr( $ts, 10, 2 );
709 $num = substr( $ts, 12, 2 );
712 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
713 $s .= gmdate( 'c', $unix );
716 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
717 $s .= gmdate( 'r', $unix );
720 if ( !$unix ) $unix = wfTimestamp( TS_UNIX
, $ts );
725 if ( $p < strlen( $format ) - 1 ) {
733 if ( $p < strlen( $format ) - 1 ) {
734 $endQuote = strpos( $format, '"', $p +
1 );
735 if ( $endQuote === false ) {
736 # No terminating quote, assume literal "
739 $s .= substr( $format, $p +
1, $endQuote - $p - 1 );
743 # Quote at end of string, assume literal "
750 if ( $num !== false ) {
751 if ( $rawToggle ||
$raw ) {
754 } elseif ( $roman ) {
755 $s .= self
::romanNumeral( $num );
757 } elseif( $hebrewNum ) {
758 $s .= self
::hebrewNumeral( $num );
761 $s .= $this->formatNum( $num, true );
769 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
770 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
772 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
773 * Gregorian dates to Iranian dates. Originally written in C, it
774 * is released under the terms of GNU Lesser General Public
775 * License. Conversion to PHP was performed by Niklas Laxström.
777 * Link: http://www.farsiweb.info/jalali/jalali.c
779 private static function tsToIranian( $ts ) {
780 $gy = substr( $ts, 0, 4 ) -1600;
781 $gm = substr( $ts, 4, 2 ) -1;
782 $gd = substr( $ts, 6, 2 ) -1;
784 # Days passed from the beginning (including leap years)
787 - floor(($gy+
99) / 100)
788 +
floor(($gy+
399) / 400);
791 // Add days of the past months of this year
792 for( $i = 0; $i < $gm; $i++
) {
793 $gDayNo +
= self
::$GREG_DAYS[$i];
797 if ( $gm > 1 && (($gy%4
===0 && $gy%100
!==0 ||
($gy%400
==0)))) {
801 // Days passed in current month
804 $jDayNo = $gDayNo - 79;
806 $jNp = floor($jDayNo / 12053);
809 $jy = 979 +
33*$jNp +
4*floor($jDayNo/1461);
812 if ( $jDayNo >= 366 ) {
813 $jy +
= floor(($jDayNo-1)/365);
814 $jDayNo = floor(($jDayNo-1)%365
);
817 for ( $i = 0; $i < 11 && $jDayNo >= self
::$IRANIAN_DAYS[$i]; $i++
) {
818 $jDayNo -= self
::$IRANIAN_DAYS[$i];
824 return array($jy, $jm, $jd);
828 * Converting Gregorian dates to Hebrew dates.
830 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
831 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
832 * to translate the relevant functions into PHP and release them under
835 private static function tsToHebrew( $ts ) {
837 $year = substr( $ts, 0, 4 );
838 $month = substr( $ts, 4, 2 );
839 $day = substr( $ts, 6, 2 );
841 # Calculate Hebrew year
842 $hebrewYear = $year +
3760;
844 # Month number when September = 1, August = 12
853 # Calculate day of year from 1 September
855 for( $i = 1; $i < $month; $i++
) {
859 # Check if the year is leap
860 if( $year %
400 == 0 ||
( $year %
4 == 0 && $year %
100 > 0 ) ) {
863 } elseif( $i == 8 ||
$i == 10 ||
$i == 1 ||
$i == 3 ) {
870 # Calculate the start of the Hebrew year
871 $start = self
::hebrewYearStart( $hebrewYear );
873 # Calculate next year's start
874 if( $dayOfYear <= $start ) {
875 # Day is before the start of the year - it is the previous year
881 # Add days since previous year's 1 September
883 if( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
887 # Start of the new (previous) year
888 $start = self
::hebrewYearStart( $hebrewYear );
891 $nextStart = self
::hebrewYearStart( $hebrewYear +
1 );
894 # Calculate Hebrew day of year
895 $hebrewDayOfYear = $dayOfYear - $start;
897 # Difference between year's days
898 $diff = $nextStart - $start;
899 # Add 12 (or 13 for leap years) days to ignore the difference between
900 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
901 # difference is only about the year type
902 if( ( $year %
400 == 0 ) ||
( $year %
100 != 0 && $year %
4 == 0 ) ) {
908 # Check the year pattern, and is leap year
909 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
910 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
912 $yearPattern = $diff %
30;
914 $isLeap = $diff >= 30;
916 # Calculate day in the month from number of day in the Hebrew year
917 # Don't check Adar - if the day is not in Adar, we will stop before;
918 # if it is in Adar, we will use it to check if it is Adar I or Adar II
919 $hebrewDay = $hebrewDayOfYear;
922 while( $hebrewMonth <= 12 ) {
923 # Calculate days in this month
924 if( $isLeap && $hebrewMonth == 6 ) {
925 # Adar in a leap year
927 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
929 if( $hebrewDay <= $days ) {
933 # Subtract the days of Adar I
937 if( $hebrewDay <= $days ) {
943 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
944 # Cheshvan in a complete year (otherwise as the rule below)
946 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
947 # Kislev in an incomplete year (otherwise as the rule below)
950 # Odd months have 30 days, even have 29
951 $days = 30 - ( $hebrewMonth - 1 ) %
2;
953 if( $hebrewDay <= $days ) {
954 # In the current month
957 # Subtract the days of the current month
959 # Try in the next month
964 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
968 * This calculates the Hebrew year start, as days since 1 September.
969 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
970 * Used for Hebrew date.
972 private static function hebrewYearStart( $year ) {
973 $a = intval( ( 12 * ( $year - 1 ) +
17 ) %
19 );
974 $b = intval( ( $year - 1 ) %
4 );
975 $m = 32.044093161144 +
1.5542417966212 * $a +
$b / 4.0 - 0.0031777940220923 * ( $year - 1 );
985 $c = intval( ( $Mar +
3 * ( $year - 1 ) +
5 * $b +
5 ) %
7);
986 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
988 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
990 } else if( $c == 2 ||
$c == 4 ||
$c == 6 ) {
994 $Mar +
= intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
999 * Algorithm to convert Gregorian dates to Thai solar dates.
1001 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1003 * @param $ts String: 14-character timestamp
1004 * @return array converted year, month, day
1006 private static function tsToThai( $ts ) {
1007 $gy = substr( $ts, 0, 4 );
1008 $gm = substr( $ts, 4, 2 );
1009 $gd = substr( $ts, 6, 2 );
1011 # Add 543 years to the Gregorian calendar
1012 # Months and days are identical
1013 $gy_thai = $gy +
543;
1015 return array( $gy_thai, $gm, $gd );
1020 * Roman number formatting up to 3000
1022 static function romanNumeral( $num ) {
1023 static $table = array(
1024 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1025 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1026 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1027 array( '', 'M', 'MM', 'MMM' )
1030 $num = intval( $num );
1031 if ( $num > 3000 ||
$num <= 0 ) {
1036 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1037 if ( $num >= $pow10 ) {
1038 $s .= $table[$i][floor($num / $pow10)];
1040 $num = $num %
$pow10;
1046 * Hebrew Gematria number formatting up to 9999
1048 static function hebrewNumeral( $num ) {
1049 static $table = array(
1050 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1051 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1052 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1053 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1056 $num = intval( $num );
1057 if ( $num > 9999 ||
$num <= 0 ) {
1062 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1063 if ( $num >= $pow10 ) {
1064 if ( $num == 15 ||
$num == 16 ) {
1065 $s .= $table[0][9] . $table[0][$num - 9];
1068 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1069 if( $pow10 == 1000 ) {
1074 $num = $num %
$pow10;
1076 if( strlen( $s ) == 2 ) {
1079 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1080 $str .= substr( $s, strlen( $s ) - 2, 2 );
1082 $start = substr( $str, 0, strlen( $str ) - 2 );
1083 $end = substr( $str, strlen( $str ) - 2 );
1086 $str = $start . 'ך';
1089 $str = $start . 'ם';
1092 $str = $start . 'ן';
1095 $str = $start . 'ף';
1098 $str = $start . 'ץ';
1105 * This is meant to be used by time(), date(), and timeanddate() to get
1106 * the date preference they're supposed to use, it should be used in
1110 * function timeanddate([...], $format = true) {
1111 * $datePreference = $this->dateFormat($format);
1116 * @param $usePrefs Mixed: if true, the user's preference is used
1117 * if false, the site/language default is used
1118 * if int/string, assumed to be a format.
1121 function dateFormat( $usePrefs = true ) {
1124 if( is_bool( $usePrefs ) ) {
1126 $datePreference = $wgUser->getDatePreference();
1128 $options = User
::getDefaultOptions();
1129 $datePreference = (string)$options['date'];
1132 $datePreference = (string)$usePrefs;
1136 if( $datePreference == '' ) {
1140 return $datePreference;
1144 * @param $ts Mixed: the time format which needs to be turned into a
1145 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1146 * @param $adj Bool: whether to adjust the time output according to the
1147 * user configured offset ($timecorrection)
1148 * @param $format Mixed: true to use user's date format preference
1149 * @param $timecorrection String: the time offset as returned by
1150 * validateTimeZone() in Special:Preferences
1153 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1156 $ts = $this->userAdjust( $ts, $timecorrection );
1159 $pref = $this->dateFormat( $format );
1160 if( $pref == 'default' ||
!isset( $this->dateFormats
["$pref date"] ) ) {
1161 $pref = $this->defaultDateFormat
;
1163 return $this->sprintfDate( $this->dateFormats
["$pref date"], $ts );
1167 * @param $ts Mixed: the time format which needs to be turned into a
1168 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1169 * @param $adj Bool: whether to adjust the time output according to the
1170 * user configured offset ($timecorrection)
1171 * @param $format Mixed: true to use user's date format preference
1172 * @param $timecorrection String: the time offset as returned by
1173 * validateTimeZone() in Special:Preferences
1176 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1179 $ts = $this->userAdjust( $ts, $timecorrection );
1182 $pref = $this->dateFormat( $format );
1183 if( $pref == 'default' ||
!isset( $this->dateFormats
["$pref time"] ) ) {
1184 $pref = $this->defaultDateFormat
;
1186 return $this->sprintfDate( $this->dateFormats
["$pref time"], $ts );
1190 * @param $ts Mixed: the time format which needs to be turned into a
1191 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1192 * @param $adj Bool: whether to adjust the time output according to the
1193 * user configured offset ($timecorrection)
1194 * @param $format Mixed: what format to return, if it's false output the
1195 * default one (default true)
1196 * @param $timecorrection String: the time offset as returned by
1197 * validateTimeZone() in Special:Preferences
1200 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1203 $ts = wfTimestamp( TS_MW
, $ts );
1206 $ts = $this->userAdjust( $ts, $timecorrection );
1209 $pref = $this->dateFormat( $format );
1210 if( $pref == 'default' ||
!isset( $this->dateFormats
["$pref both"] ) ) {
1211 $pref = $this->defaultDateFormat
;
1214 return $this->sprintfDate( $this->dateFormats
["$pref both"], $ts );
1217 function getMessage( $key ) {
1219 return isset( $this->messages
[$key] ) ?
$this->messages
[$key] : null;
1222 function getAllMessages() {
1224 return $this->messages
;
1227 function iconv( $in, $out, $string ) {
1228 # For most languages, this is a wrapper for iconv
1229 return iconv( $in, $out . '//IGNORE', $string );
1232 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1233 function ucwordbreaksCallbackAscii($matches){
1234 return $this->ucfirst($matches[1]);
1237 function ucwordbreaksCallbackMB($matches){
1238 return mb_strtoupper($matches[0]);
1241 function ucCallback($matches){
1242 list( $wikiUpperChars ) = self
::getCaseMaps();
1243 return strtr( $matches[1], $wikiUpperChars );
1246 function lcCallback($matches){
1247 list( , $wikiLowerChars ) = self
::getCaseMaps();
1248 return strtr( $matches[1], $wikiLowerChars );
1251 function ucwordsCallbackMB($matches){
1252 return mb_strtoupper($matches[0]);
1255 function ucwordsCallbackWiki($matches){
1256 list( $wikiUpperChars ) = self
::getCaseMaps();
1257 return strtr( $matches[0], $wikiUpperChars );
1260 function ucfirst( $str ) {
1261 if ( empty($str) ) return $str;
1262 if ( ord($str[0]) < 128 ) return ucfirst($str);
1263 else return self
::uc($str,true); // fall back to more complex logic in case of multibyte strings
1266 function uc( $str, $first = false ) {
1267 if ( function_exists( 'mb_strtoupper' ) ) {
1269 if ( self
::isMultibyte( $str ) ) {
1270 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1272 return ucfirst( $str );
1275 return self
::isMultibyte( $str ) ?
mb_strtoupper( $str ) : strtoupper( $str );
1278 if ( self
::isMultibyte( $str ) ) {
1279 list( $wikiUpperChars ) = $this->getCaseMaps();
1280 $x = $first ?
'^' : '';
1281 return preg_replace_callback(
1282 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1283 array($this,"ucCallback"),
1287 return $first ?
ucfirst( $str ) : strtoupper( $str );
1292 function lcfirst( $str ) {
1293 if ( empty($str) ) return $str;
1294 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1295 // editing string in place = cool
1296 $str[0]=strtolower($str[0]);
1299 else return self
::lc( $str, true );
1302 function lc( $str, $first = false ) {
1303 if ( function_exists( 'mb_strtolower' ) )
1305 if ( self
::isMultibyte( $str ) )
1306 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1308 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1310 return self
::isMultibyte( $str ) ?
mb_strtolower( $str ) : strtolower( $str );
1312 if ( self
::isMultibyte( $str ) ) {
1313 list( , $wikiLowerChars ) = self
::getCaseMaps();
1314 $x = $first ?
'^' : '';
1315 return preg_replace_callback(
1316 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1317 array($this,"lcCallback"),
1321 return $first ?
strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1324 function isMultibyte( $str ) {
1325 return (bool)preg_match( '/[\x80-\xff]/', $str );
1328 function ucwords($str) {
1329 if ( self
::isMultibyte( $str ) ) {
1330 $str = self
::lc($str);
1332 // regexp to find first letter in each word (i.e. after each space)
1333 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1335 // function to use to capitalize a single char
1336 if ( function_exists( 'mb_strtoupper' ) )
1337 return preg_replace_callback(
1339 array($this,"ucwordsCallbackMB"),
1343 return preg_replace_callback(
1345 array($this,"ucwordsCallbackWiki"),
1350 return ucwords( strtolower( $str ) );
1353 # capitalize words at word breaks
1354 function ucwordbreaks($str){
1355 if (self
::isMultibyte( $str ) ) {
1356 $str = self
::lc($str);
1358 // since \b doesn't work for UTF-8, we explicitely define word break chars
1359 $breaks= "[ \-\(\)\}\{\.,\?!]";
1361 // find first letter after word break
1362 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1364 if ( function_exists( 'mb_strtoupper' ) )
1365 return preg_replace_callback(
1367 array($this,"ucwordbreaksCallbackMB"),
1371 return preg_replace_callback(
1373 array($this,"ucwordsCallbackWiki"),
1378 return preg_replace_callback(
1379 '/\b([\w\x80-\xff]+)\b/',
1380 array($this,"ucwordbreaksCallbackAscii"),
1385 * Return a case-folded representation of $s
1387 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1388 * and $s2 are the same except for the case of their characters. It is not
1389 * necessary for the value returned to make sense when displayed.
1391 * Do *not* perform any other normalisation in this function. If a caller
1392 * uses this function when it should be using a more general normalisation
1393 * function, then fix the caller.
1395 function caseFold( $s ) {
1396 return $this->uc( $s );
1399 function checkTitleEncoding( $s ) {
1400 if( is_array( $s ) ) {
1401 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1403 # Check for non-UTF-8 URLs
1404 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1405 if(!$ishigh) return $s;
1407 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1408 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1409 if( $isutf8 ) return $s;
1411 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1414 function fallback8bitEncoding() {
1416 return $this->fallback8bitEncoding
;
1420 * Some languages have special punctuation to strip out
1421 * or characters which need to be converted for MySQL's
1422 * indexing to grok it correctly. Make such changes here.
1424 * @param $string String
1427 function stripForSearch( $string ) {
1429 if ( $wgDBtype != 'mysql' ) {
1433 # MySQL fulltext index doesn't grok utf-8, so we
1434 # need to fold cases and convert to hex
1436 wfProfileIn( __METHOD__
);
1437 if( function_exists( 'mb_strtolower' ) ) {
1438 $out = preg_replace(
1439 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1440 "'U8' . bin2hex( \"$1\" )",
1441 mb_strtolower( $string ) );
1443 list( , $wikiLowerChars ) = self
::getCaseMaps();
1444 $out = preg_replace(
1445 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1446 "'U8' . bin2hex( strtr( \"\$1\", \$wikiLowerChars ) )",
1449 wfProfileOut( __METHOD__
);
1453 function convertForSearchResult( $termsArray ) {
1454 # some languages, e.g. Chinese, need to do a conversion
1455 # in order for search results to be displayed correctly
1460 * Get the first character of a string.
1465 function firstChar( $s ) {
1467 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1468 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1470 if ( isset( $matches[1] ) ) {
1471 if ( strlen( $matches[1] ) != 3 ) {
1475 // Break down Hangul syllables to grab the first jamo
1476 $code = utf8ToCodepoint( $matches[1] );
1477 if ( $code < 0xac00 ||
0xd7a4 <= $code) {
1479 } elseif ( $code < 0xb098 ) {
1480 return "\xe3\x84\xb1";
1481 } elseif ( $code < 0xb2e4 ) {
1482 return "\xe3\x84\xb4";
1483 } elseif ( $code < 0xb77c ) {
1484 return "\xe3\x84\xb7";
1485 } elseif ( $code < 0xb9c8 ) {
1486 return "\xe3\x84\xb9";
1487 } elseif ( $code < 0xbc14 ) {
1488 return "\xe3\x85\x81";
1489 } elseif ( $code < 0xc0ac ) {
1490 return "\xe3\x85\x82";
1491 } elseif ( $code < 0xc544 ) {
1492 return "\xe3\x85\x85";
1493 } elseif ( $code < 0xc790 ) {
1494 return "\xe3\x85\x87";
1495 } elseif ( $code < 0xcc28 ) {
1496 return "\xe3\x85\x88";
1497 } elseif ( $code < 0xce74 ) {
1498 return "\xe3\x85\x8a";
1499 } elseif ( $code < 0xd0c0 ) {
1500 return "\xe3\x85\x8b";
1501 } elseif ( $code < 0xd30c ) {
1502 return "\xe3\x85\x8c";
1503 } elseif ( $code < 0xd558 ) {
1504 return "\xe3\x85\x8d";
1506 return "\xe3\x85\x8e";
1513 function initEncoding() {
1514 # Some languages may have an alternate char encoding option
1515 # (Esperanto X-coding, Japanese furigana conversion, etc)
1516 # If this language is used as the primary content language,
1517 # an override to the defaults can be set here on startup.
1520 function recodeForEdit( $s ) {
1521 # For some languages we'll want to explicitly specify
1522 # which characters make it into the edit box raw
1523 # or are converted in some way or another.
1524 # Note that if wgOutputEncoding is different from
1525 # wgInputEncoding, this text will be further converted
1526 # to wgOutputEncoding.
1527 global $wgEditEncoding;
1528 if( $wgEditEncoding == '' or
1529 $wgEditEncoding == 'UTF-8' ) {
1532 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1536 function recodeInput( $s ) {
1537 # Take the previous into account.
1538 global $wgEditEncoding;
1539 if($wgEditEncoding != "") {
1540 $enc = $wgEditEncoding;
1544 if( $enc == 'UTF-8' ) {
1547 return $this->iconv( $enc, 'UTF-8', $s );
1552 * For right-to-left language support
1562 * A hidden direction mark (LRM or RLM), depending on the language direction
1566 function getDirMark() {
1567 return $this->isRTL() ?
"\xE2\x80\x8F" : "\xE2\x80\x8E";
1571 * An arrow, depending on the language direction
1575 function getArrow() {
1576 return $this->isRTL() ?
'←' : '→';
1580 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1584 function linkPrefixExtension() {
1586 return $this->linkPrefixExtension
;
1589 function &getMagicWords() {
1591 return $this->magicWords
;
1594 # Fill a MagicWord object with data from here
1595 function getMagic( &$mw ) {
1596 if ( !$this->mMagicHookDone
) {
1597 $this->mMagicHookDone
= true;
1598 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions
, $this->getCode() ) );
1600 if ( isset( $this->mMagicExtensions
[$mw->mId
] ) ) {
1601 $rawEntry = $this->mMagicExtensions
[$mw->mId
];
1603 $magicWords =& $this->getMagicWords();
1604 if ( isset( $magicWords[$mw->mId
] ) ) {
1605 $rawEntry = $magicWords[$mw->mId
];
1607 # Fall back to English if local list is incomplete
1608 $magicWords =& Language
::getMagicWords();
1609 if ( !isset($magicWords[$mw->mId
]) ) { throw new MWException("Magic word not found" ); }
1610 $rawEntry = $magicWords[$mw->mId
];
1614 if( !is_array( $rawEntry ) ) {
1615 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1617 $mw->mCaseSensitive
= $rawEntry[0];
1618 $mw->mSynonyms
= array_slice( $rawEntry, 1 );
1623 * Add magic words to the extension array
1625 function addMagicWordsByLang( $newWords ) {
1626 $code = $this->getCode();
1627 $fallbackChain = array();
1628 while ( $code && !in_array( $code, $fallbackChain ) ) {
1629 $fallbackChain[] = $code;
1630 $code = self
::getFallbackFor( $code );
1632 if ( !in_array( 'en', $fallbackChain ) ) {
1633 $fallbackChain[] = 'en';
1635 $fallbackChain = array_reverse( $fallbackChain );
1636 foreach ( $fallbackChain as $code ) {
1637 if ( isset( $newWords[$code] ) ) {
1638 $this->mMagicExtensions
= $newWords[$code] +
$this->mMagicExtensions
;
1644 * Get special page names, as an associative array
1645 * case folded alias => real name
1647 function getSpecialPageAliases() {
1649 if ( !isset( $this->mExtendedSpecialPageAliases
) ) {
1650 $this->mExtendedSpecialPageAliases
= $this->specialPageAliases
;
1651 wfRunHooks( 'LanguageGetSpecialPageAliases',
1652 array( &$this->mExtendedSpecialPageAliases
, $this->getCode() ) );
1654 return $this->mExtendedSpecialPageAliases
;
1658 * Italic is unsuitable for some languages
1660 * @param $text String: the text to be emphasized.
1663 function emphasize( $text ) {
1664 return "<em>$text</em>";
1668 * Normally we output all numbers in plain en_US style, that is
1669 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1670 * point twohundredthirtyfive. However this is not sutable for all
1671 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1672 * Icelandic just want to use commas instead of dots, and dots instead
1673 * of commas like "293.291,235".
1675 * An example of this function being called:
1677 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1680 * See LanguageGu.php for the Gujarati implementation and
1681 * LanguageIs.php for the , => . and . => , implementation.
1683 * @todo check if it's viable to use localeconv() for the decimal
1685 * @param $number Mixed: the string to be formatted, should be an integer
1686 * or a floating point number.
1687 * @param $nocommafy Bool: set to true for special numbers like dates
1690 function formatNum( $number, $nocommafy = false ) {
1691 global $wgTranslateNumerals;
1693 $number = $this->commafy($number);
1694 $s = $this->separatorTransformTable();
1695 if (!is_null($s)) { $number = strtr($number, $s); }
1698 if ($wgTranslateNumerals) {
1699 $s = $this->digitTransformTable();
1700 if (!is_null($s)) { $number = strtr($number, $s); }
1706 function parseFormattedNumber( $number ) {
1707 $s = $this->digitTransformTable();
1708 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1710 $s = $this->separatorTransformTable();
1711 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1713 $number = strtr( $number, array (',' => '') );
1718 * Adds commas to a given number
1723 function commafy($_) {
1724 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1727 function digitTransformTable() {
1729 return $this->digitTransformTable
;
1732 function separatorTransformTable() {
1734 return $this->separatorTransformTable
;
1739 * For the credit list in includes/Credits.php (action=credits)
1744 function listToText( $l ) {
1747 for ($i = $m; $i >= 0; $i--) {
1750 } else if ($i == $m - 1) {
1751 $s = $l[$i] . ' ' . $this->getMessageFromDB( 'and' ) . ' ' . $s;
1753 $s = $l[$i] . ', ' . $s;
1760 * Truncate a string to a specified length in bytes, appending an optional
1761 * string (e.g. for ellipses)
1763 * The database offers limited byte lengths for some columns in the database;
1764 * multi-byte character sets mean we need to ensure that only whole characters
1765 * are included, otherwise broken characters can be passed to the user
1767 * If $length is negative, the string will be truncated from the beginning
1769 * @param $string String to truncate
1770 * @param $length Int: maximum length (excluding ellipses)
1771 * @param $ellipsis String to append to the truncated text
1774 function truncate( $string, $length, $ellipsis = "" ) {
1775 if( $length == 0 ) {
1778 if ( strlen( $string ) <= abs( $length ) ) {
1782 $string = substr( $string, 0, $length );
1783 $char = ord( $string[strlen( $string ) - 1] );
1785 if ($char >= 0xc0) {
1786 # We got the first byte only of a multibyte char; remove it.
1787 $string = substr( $string, 0, -1 );
1788 } elseif( $char >= 0x80 &&
1789 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
1790 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
1791 # We chopped in the middle of a character; remove it
1794 return $string . $ellipsis;
1796 $string = substr( $string, $length );
1797 $char = ord( $string[0] );
1798 if( $char >= 0x80 && $char < 0xc0 ) {
1799 # We chopped in the middle of a character; remove the whole thing
1800 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
1802 return $ellipsis . $string;
1807 * Grammatical transformations, needed for inflected languages
1808 * Invoked by putting {{grammar:case|word}} in a message
1810 * @param $word string
1811 * @param $case string
1814 function convertGrammar( $word, $case ) {
1815 global $wgGrammarForms;
1816 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
1817 return $wgGrammarForms[$this->getCode()][$case][$word];
1823 * Plural form transformations, needed for some languages.
1824 * For example, there are 3 form of plural in Russian and Polish,
1825 * depending on "count mod 10". See [[w:Plural]]
1826 * For English it is pretty simple.
1828 * Invoked by putting {{plural:count|wordform1|wordform2}}
1829 * or {{plural:count|wordform1|wordform2|wordform3}}
1831 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
1833 * @param $count Integer: non-localized number
1834 * @param $forms Array: different plural forms
1835 * @return string Correct form of plural for $count in this language
1837 function convertPlural( $count, $forms ) {
1838 if ( !count($forms) ) { return ''; }
1839 $forms = $this->preConvertPlural( $forms, 2 );
1841 return ( abs($count) == 1 ) ?
$forms[0] : $forms[1];
1845 * Checks that convertPlural was given an array and pads it to requested
1846 * amound of forms by copying the last one.
1848 * @param $count Integer: How many forms should there be at least
1849 * @param $forms Array of forms given to convertPlural
1850 * @return array Padded array of forms or an exception if not an array
1852 protected function preConvertPlural( /* Array */ $forms, $count ) {
1853 while ( count($forms) < $count ) {
1854 $forms[] = $forms[count($forms)-1];
1860 * For translaing of expiry times
1861 * @param $str String: the validated block time in English
1862 * @return Somehow translated block time
1863 * @see LanguageFi.php for example implementation
1865 function translateBlockExpiry( $str ) {
1867 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
1869 if ( $scBlockExpiryOptions == '-') {
1873 foreach (explode(',', $scBlockExpiryOptions) as $option) {
1874 if ( strpos($option, ":") === false )
1876 list($show, $value) = explode(":", $option);
1877 if ( strcmp ( $str, $value) == 0 ) {
1878 return htmlspecialchars( trim( $show ) );
1886 * languages like Chinese need to be segmented in order for the diff
1889 * @param $text String
1892 function segmentForDiff( $text ) {
1897 * and unsegment to show the result
1899 * @param $text String
1902 function unsegmentForDiff( $text ) {
1906 # convert text to different variants of a language.
1907 function convert( $text, $isTitle = false) {
1908 return $this->mConverter
->convert($text, $isTitle);
1911 # Convert text from within Parser
1912 function parserConvert( $text, &$parser ) {
1913 return $this->mConverter
->parserConvert( $text, $parser );
1916 # Check if this is a language with variants
1917 function hasVariants(){
1918 return sizeof($this->getVariants())>1;
1921 # Put custom tags (e.g. -{ }-) around math to prevent conversion
1922 function armourMath($text){
1923 return $this->mConverter
->armourMath($text);
1928 * Perform output conversion on a string, and encode for safe HTML output.
1929 * @param $text String
1930 * @param $isTitle Bool -- wtf?
1932 * @todo this should get integrated somewhere sane
1934 function convertHtml( $text, $isTitle = false ) {
1935 return htmlspecialchars( $this->convert( $text, $isTitle ) );
1938 function convertCategoryKey( $key ) {
1939 return $this->mConverter
->convertCategoryKey( $key );
1943 * get the list of variants supported by this langauge
1944 * see sample implementation in LanguageZh.php
1946 * @return array an array of language codes
1948 function getVariants() {
1949 return $this->mConverter
->getVariants();
1953 function getPreferredVariant( $fromUser = true ) {
1954 return $this->mConverter
->getPreferredVariant( $fromUser );
1958 * if a language supports multiple variants, it is
1959 * possible that non-existing link in one variant
1960 * actually exists in another variant. this function
1961 * tries to find it. See e.g. LanguageZh.php
1963 * @param $link String: the name of the link
1964 * @param $nt Mixed: the title object of the link
1965 * @return null the input parameters may be modified upon return
1967 function findVariantLink( &$link, &$nt ) {
1968 $this->mConverter
->findVariantLink($link, $nt);
1972 * If a language supports multiple variants, converts text
1973 * into an array of all possible variants of the text:
1974 * 'variant' => text in that variant
1977 function convertLinkToAllVariants($text){
1978 return $this->mConverter
->convertLinkToAllVariants($text);
1983 * returns language specific options used by User::getPageRenderHash()
1984 * for example, the preferred language variant
1988 function getExtraHashOptions() {
1989 return $this->mConverter
->getExtraHashOptions();
1993 * for languages that support multiple variants, the title of an
1994 * article may be displayed differently in different variants. this
1995 * function returns the apporiate title defined in the body of the article.
1999 function getParsedTitle() {
2000 return $this->mConverter
->getParsedTitle();
2004 * Enclose a string with the "no conversion" tag. This is used by
2005 * various functions in the Parser
2007 * @param $text String: text to be tagged for no conversion
2009 * @return string the tagged text
2011 function markNoConversion( $text, $noParse=false ) {
2012 return $this->mConverter
->markNoConversion( $text, $noParse );
2016 * A regular expression to match legal word-trailing characters
2017 * which should be merged onto a link of the form [[foo]]bar.
2021 function linkTrail() {
2023 return $this->linkTrail
;
2026 function getLangObj() {
2031 * Get the RFC 3066 code for this language object
2033 function getCode() {
2034 return $this->mCode
;
2037 function setCode( $code ) {
2038 $this->mCode
= $code;
2041 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2042 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2045 static function getMessagesFileName( $code ) {
2047 return self
::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2050 static function getClassFileName( $code ) {
2052 return self
::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2055 static function getLocalisationArray( $code, $disableCache = false ) {
2056 self
::loadLocalisation( $code, $disableCache );
2057 return self
::$mLocalisationCache[$code];
2061 * Load localisation data for a given code into the static cache
2063 * @return array Dependencies, map of filenames to mtimes
2065 static function loadLocalisation( $code, $disableCache = false ) {
2066 static $recursionGuard = array();
2067 global $wgMemc, $wgCheckSerialized;
2070 throw new MWException( "Invalid language code requested" );
2073 if ( !$disableCache ) {
2074 # Try the per-process cache
2075 if ( isset( self
::$mLocalisationCache[$code] ) ) {
2076 return self
::$mLocalisationCache[$code]['deps'];
2079 wfProfileIn( __METHOD__
);
2081 # Try the serialized directory
2082 $cache = wfGetPrecompiledData( self
::getFileName( "Messages", $code, '.ser' ) );
2084 if ( $wgCheckSerialized && self
::isLocalisationOutOfDate( $cache ) ) {
2086 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2088 self
::$mLocalisationCache[$code] = $cache;
2089 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2090 wfProfileOut( __METHOD__
);
2091 return self
::$mLocalisationCache[$code]['deps'];
2095 # Try the global cache
2096 $memcKey = wfMemcKey('localisation', $code );
2097 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2098 $cache = $wgMemc->get( $memcKey );
2100 if ( self
::isLocalisationOutOfDate( $cache ) ) {
2101 $wgMemc->delete( $memcKey );
2102 $wgMemc->delete( $fbMemcKey );
2104 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2106 self
::$mLocalisationCache[$code] = $cache;
2107 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2108 wfProfileOut( __METHOD__
);
2109 return $cache['deps'];
2113 wfProfileIn( __METHOD__
);
2116 # Default fallback, may be overridden when the messages file is included
2117 if ( $code != 'en' ) {
2123 # Load the primary localisation from the source file
2124 $filename = self
::getMessagesFileName( $code );
2125 if ( !file_exists( $filename ) ) {
2126 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2127 $cache = compact( self
::$mLocalisationKeys ); // Set correct fallback
2130 $deps = array( $filename => filemtime( $filename ) );
2131 require( $filename );
2132 $cache = compact( self
::$mLocalisationKeys );
2133 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2136 if ( !empty( $fallback ) ) {
2137 # Load the fallback localisation, with a circular reference guard
2138 if ( isset( $recursionGuard[$code] ) ) {
2139 throw new MWException( "Error: Circular fallback reference in language code $code" );
2141 $recursionGuard[$code] = true;
2142 $newDeps = self
::loadLocalisation( $fallback, $disableCache );
2143 unset( $recursionGuard[$code] );
2145 $secondary = self
::$mLocalisationCache[$fallback];
2146 $deps = array_merge( $deps, $newDeps );
2148 # Merge the fallback localisation with the current localisation
2149 foreach ( self
::$mLocalisationKeys as $key ) {
2150 if ( isset( $cache[$key] ) ) {
2151 if ( isset( $secondary[$key] ) ) {
2152 if ( in_array( $key, self
::$mMergeableMapKeys ) ) {
2153 $cache[$key] = $cache[$key] +
$secondary[$key];
2154 } elseif ( in_array( $key, self
::$mMergeableListKeys ) ) {
2155 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2156 } elseif ( in_array( $key, self
::$mMergeableAliasListKeys ) ) {
2157 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2161 $cache[$key] = $secondary[$key];
2165 # Merge bookstore lists if requested
2166 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2167 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2169 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2170 unset( $cache['bookstoreList']['inherit'] );
2174 # Add dependencies to the cache entry
2175 $cache['deps'] = $deps;
2177 # Replace spaces with underscores in namespace names
2178 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2180 # And do the same for specialpage aliases. $page is an array.
2181 foreach ( $cache['specialPageAliases'] as &$page ) {
2182 $page = str_replace( ' ', '_', $page );
2184 # Decouple the reference to prevent accidental damage
2187 # Save to both caches
2188 self
::$mLocalisationCache[$code] = $cache;
2189 if ( !$disableCache ) {
2190 $wgMemc->set( $memcKey, $cache );
2191 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2194 wfProfileOut( __METHOD__
);
2199 * Test if a given localisation cache is out of date with respect to the
2200 * source Messages files. This is done automatically for the global cache
2201 * in $wgMemc, but is only done on certain occasions for the serialized
2204 * @param $cache mixed Either a language code or a cache array
2206 static function isLocalisationOutOfDate( $cache ) {
2207 if ( !is_array( $cache ) ) {
2208 self
::loadLocalisation( $cache );
2209 $cache = self
::$mLocalisationCache[$cache];
2212 foreach ( $cache['deps'] as $file => $mtime ) {
2213 if ( !file_exists( $file ) ||
filemtime( $file ) > $mtime ) {
2222 * Get the fallback for a given language
2224 static function getFallbackFor( $code ) {
2226 if ( $code === 'en' ) return false;
2229 static $cache = array();
2231 if ( isset($cache[$code]) ) return $cache[$code];
2235 $memcKey = wfMemcKey( 'fallback', $code );
2236 $fbcode = $wgMemc->get( $memcKey );
2238 if ( is_string($fbcode) ) {
2239 // False is stored as a string to detect failures in memcache properly
2240 if ( $fbcode === '' ) $fbcode = false;
2242 // Update local cache and return
2243 $cache[$code] = $fbcode;
2247 // Nothing in caches, load and and update both caches
2248 self
::loadLocalisation( $code );
2249 $fbcode = self
::$mLocalisationCache[$code]['fallback'];
2251 $cache[$code] = $fbcode;
2252 $wgMemc->set( $memcKey, (string) $fbcode );
2258 * Get all messages for a given language
2260 static function getMessagesFor( $code ) {
2261 self
::loadLocalisation( $code );
2262 return self
::$mLocalisationCache[$code]['messages'];
2266 * Get a message for a given language
2268 static function getMessageFor( $key, $code ) {
2269 self
::loadLocalisation( $code );
2270 return isset( self
::$mLocalisationCache[$code]['messages'][$key] ) ? self
::$mLocalisationCache[$code]['messages'][$key] : null;
2274 * Load localisation data for this object
2277 if ( !$this->mLoaded
) {
2278 self
::loadLocalisation( $this->getCode() );
2279 $cache =& self
::$mLocalisationCache[$this->getCode()];
2280 foreach ( self
::$mLocalisationKeys as $key ) {
2281 $this->$key = $cache[$key];
2283 $this->mLoaded
= true;
2285 $this->fixUpSettings();
2290 * Do any necessary post-cache-load settings adjustment
2292 function fixUpSettings() {
2293 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2294 $wgNamespaceAliases, $wgAmericanDates;
2295 wfProfileIn( __METHOD__
);
2296 if ( $wgExtraNamespaces ) {
2297 $this->namespaceNames
= $wgExtraNamespaces +
$this->namespaceNames
;
2300 $this->namespaceNames
[NS_PROJECT
] = $wgMetaNamespace;
2301 if ( $wgMetaNamespaceTalk ) {
2302 $this->namespaceNames
[NS_PROJECT_TALK
] = $wgMetaNamespaceTalk;
2304 $talk = $this->namespaceNames
[NS_PROJECT_TALK
];
2305 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2307 # Allow grammar transformations
2308 # Allowing full message-style parsing would make simple requests
2309 # such as action=raw much more expensive than they need to be.
2310 # This will hopefully cover most cases.
2311 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2312 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2313 $talk = str_replace( ' ', '_', $talk );
2314 $this->namespaceNames
[NS_PROJECT_TALK
] = $talk;
2317 # The above mixing may leave namespaces out of canonical order.
2318 # Re-order by namespace ID number...
2319 ksort( $this->namespaceNames
);
2321 # Put namespace names and aliases into a hashtable.
2322 # If this is too slow, then we should arrange it so that it is done
2323 # before caching. The catch is that at pre-cache time, the above
2324 # class-specific fixup hasn't been done.
2325 $this->mNamespaceIds
= array();
2326 foreach ( $this->namespaceNames
as $index => $name ) {
2327 $this->mNamespaceIds
[$this->lc($name)] = $index;
2329 if ( $this->namespaceAliases
) {
2330 foreach ( $this->namespaceAliases
as $name => $index ) {
2331 $this->mNamespaceIds
[$this->lc($name)] = $index;
2334 if ( $wgNamespaceAliases ) {
2335 foreach ( $wgNamespaceAliases as $name => $index ) {
2336 $this->mNamespaceIds
[$this->lc($name)] = $index;
2340 if ( $this->defaultDateFormat
== 'dmy or mdy' ) {
2341 $this->defaultDateFormat
= $wgAmericanDates ?
'mdy' : 'dmy';
2343 wfProfileOut( __METHOD__
);
2346 function replaceGrammarInNamespace( $m ) {
2347 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2350 static function getCaseMaps() {
2351 static $wikiUpperChars, $wikiLowerChars;
2352 if ( isset( $wikiUpperChars ) ) {
2353 return array( $wikiUpperChars, $wikiLowerChars );
2356 wfProfileIn( __METHOD__
);
2357 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2358 if ( $arr === false ) {
2359 throw new MWException(
2360 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2363 wfProfileOut( __METHOD__
);
2364 return array( $wikiUpperChars, $wikiLowerChars );
2367 function formatTimePeriod( $seconds ) {
2368 if ( $seconds < 10 ) {
2369 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2370 } elseif ( $seconds < 60 ) {
2371 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2372 } elseif ( $seconds < 3600 ) {
2373 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2374 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2376 $hours = floor( $seconds / 3600 );
2377 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2378 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2379 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2380 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2381 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2385 function formatBitrate( $bps ) {
2386 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2388 return $this->formatNum( $bps ) . $units[0];
2390 $unitIndex = floor( log10( $bps ) / 3 );
2391 $mantissa = $bps / pow( 1000, $unitIndex );
2392 if ( $mantissa < 10 ) {
2393 $mantissa = round( $mantissa, 1 );
2395 $mantissa = round( $mantissa );
2397 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2401 * Format a size in bytes for output, using an appropriate
2402 * unit (B, KB, MB or GB) according to the magnitude in question
2404 * @param $size Size to format
2405 * @return string Plain text (not HTML)
2407 function formatSize( $size ) {
2408 // For small sizes no decimal places necessary
2410 if( $size > 1024 ) {
2411 $size = $size / 1024;
2412 if( $size > 1024 ) {
2413 $size = $size / 1024;
2414 // For MB and bigger two decimal places are smarter
2416 if( $size > 1024 ) {
2417 $size = $size / 1024;
2418 $msg = 'size-gigabytes';
2420 $msg = 'size-megabytes';
2423 $msg = 'size-kilobytes';
2426 $msg = 'size-bytes';
2428 $size = round( $size, $round );
2429 $text = $this->getMessageFromDB( $msg );
2430 return str_replace( '$1', $this->formatNum( $size ), $text );