Add support for Higri (Islamic) calendar
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * @defgroup Language Language
4 *
5 * @file
6 * @ingroup Language
7 */
8
9 if( !defined( 'MEDIAWIKI' ) ) {
10 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
11 exit( 1 );
12 }
13
14 # Read language names
15 global $wgLanguageNames;
16 require_once( dirname(__FILE__) . '/Names.php' ) ;
17
18 global $wgInputEncoding, $wgOutputEncoding;
19
20 /**
21 * These are always UTF-8, they exist only for backwards compatibility
22 */
23 $wgInputEncoding = "UTF-8";
24 $wgOutputEncoding = "UTF-8";
25
26 if( function_exists( 'mb_strtoupper' ) ) {
27 mb_internal_encoding('UTF-8');
28 }
29
30 /**
31 * a fake language converter
32 *
33 * @ingroup Language
34 */
35 class FakeConverter {
36 var $mLang;
37 function FakeConverter($langobj) {$this->mLang = $langobj;}
38 function convert($t, $i) {return $t;}
39 function parserConvert($t, $p) {return $t;}
40 function getVariants() { return array( $this->mLang->getCode() ); }
41 function getPreferredVariant() {return $this->mLang->getCode(); }
42 function findVariantLink(&$l, &$n) {}
43 function getExtraHashOptions() {return '';}
44 function getParsedTitle() {return '';}
45 function markNoConversion($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; }
49 }
50
51 /**
52 * Internationalisation code
53 * @ingroup Language
54 */
55 class Language {
56 var $mConverter, $mVariants, $mCode, $mLoaded = false;
57 var $mMagicExtensions = array(), $mMagicHookDone = false;
58
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' );
66
67 static public $mMergeableMapKeys = array( 'messages', 'namespaceNames', 'mathNames',
68 'dateFormats', 'defaultUserOptionOverrides', 'magicWords' );
69
70 static public $mMergeableListKeys = array( 'extraUserToggles' );
71
72 static public $mMergeableAliasListKeys = array( 'specialPageAliases' );
73
74 static public $mLocalisationCache = array();
75
76 static public $mWeekdayMsgs = array(
77 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
78 'friday', 'saturday'
79 );
80
81 static public $mWeekdayAbbrevMsgs = array(
82 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
83 );
84
85 static public $mMonthMsgs = array(
86 'january', 'february', 'march', 'april', 'may_long', 'june',
87 'july', 'august', 'september', 'october', 'november',
88 'december'
89 );
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',
93 'december-gen'
94 );
95 static public $mMonthAbbrevMsgs = array(
96 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
97 'sep', 'oct', 'nov', 'dec'
98 );
99
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'
105 );
106
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'
113 );
114
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'
121 );
122
123 static public $mHijriCalendarMonthMsgs = array(
124 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
125 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
126 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
127 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
128 );
129
130 /**
131 * Create a language object for a given language code
132 */
133 static function factory( $code ) {
134 global $IP;
135 static $recursionLevel = 0;
136
137 if ( $code == 'en' ) {
138 $class = 'Language';
139 } else {
140 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
141 // Preload base classes to work around APC/PHP5 bug
142 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
143 include_once("$IP/languages/classes/$class.deps.php");
144 }
145 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
146 include_once("$IP/languages/classes/$class.php");
147 }
148 }
149
150 if ( $recursionLevel > 5 ) {
151 throw new MWException( "Language fallback loop detected when creating class $class\n" );
152 }
153
154 if( ! class_exists( $class ) ) {
155 $fallback = Language::getFallbackFor( $code );
156 ++$recursionLevel;
157 $lang = Language::factory( $fallback );
158 --$recursionLevel;
159 $lang->setCode( $code );
160 } else {
161 $lang = new $class;
162 }
163
164 return $lang;
165 }
166
167 function __construct() {
168 $this->mConverter = new FakeConverter($this);
169 // Set the code to the name of the descendant
170 if ( get_class( $this ) == 'Language' ) {
171 $this->mCode = 'en';
172 } else {
173 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
174 }
175 }
176
177 /**
178 * Hook which will be called if this is the content language.
179 * Descendants can use this to register hook functions or modify globals
180 */
181 function initContLang() {}
182
183 /**
184 * @deprecated Use User::getDefaultOptions()
185 * @return array
186 */
187 function getDefaultUserOptions() {
188 wfDeprecated( __METHOD__ );
189 return User::getDefaultOptions();
190 }
191
192 function getFallbackLanguageCode() {
193 return self::getFallbackFor( $this->mCode );
194 }
195
196 /**
197 * Exports $wgBookstoreListEn
198 * @return array
199 */
200 function getBookstoreList() {
201 $this->load();
202 return $this->bookstoreList;
203 }
204
205 /**
206 * @return array
207 */
208 function getNamespaces() {
209 $this->load();
210 return $this->namespaceNames;
211 }
212
213 /**
214 * A convenience function that returns the same thing as
215 * getNamespaces() except with the array values changed to ' '
216 * where it found '_', useful for producing output to be displayed
217 * e.g. in <select> forms.
218 *
219 * @return array
220 */
221 function getFormattedNamespaces() {
222 $ns = $this->getNamespaces();
223 foreach($ns as $k => $v) {
224 $ns[$k] = strtr($v, '_', ' ');
225 }
226 return $ns;
227 }
228
229 /**
230 * Get a namespace value by key
231 * <code>
232 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
233 * echo $mw_ns; // prints 'MediaWiki'
234 * </code>
235 *
236 * @param $index Int: the array key of the namespace to return
237 * @return mixed, string if the namespace value exists, otherwise false
238 */
239 function getNsText( $index ) {
240 $ns = $this->getNamespaces();
241 return isset( $ns[$index] ) ? $ns[$index] : false;
242 }
243
244 /**
245 * A convenience function that returns the same thing as
246 * getNsText() except with '_' changed to ' ', useful for
247 * producing output.
248 *
249 * @return array
250 */
251 function getFormattedNsText( $index ) {
252 $ns = $this->getNsText( $index );
253 return strtr($ns, '_', ' ');
254 }
255
256 /**
257 * Get a namespace key by value, case insensitive.
258 * Only matches namespace names for the current language, not the
259 * canonical ones defined in Namespace.php.
260 *
261 * @param $text String
262 * @return mixed An integer if $text is a valid value otherwise false
263 */
264 function getLocalNsIndex( $text ) {
265 $this->load();
266 $lctext = $this->lc($text);
267 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
268 }
269
270 /**
271 * Get a namespace key by value, case insensitive. Canonical namespace
272 * names override custom ones defined for the current language.
273 *
274 * @param $text String
275 * @return mixed An integer if $text is a valid value otherwise false
276 */
277 function getNsIndex( $text ) {
278 $this->load();
279 $lctext = $this->lc($text);
280 if( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
281 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
282 }
283
284 /**
285 * short names for language variants used for language conversion links.
286 *
287 * @param $code String
288 * @return string
289 */
290 function getVariantname( $code ) {
291 return $this->getMessageFromDB( "variantname-$code" );
292 }
293
294 function specialPage( $name ) {
295 $aliases = $this->getSpecialPageAliases();
296 if ( isset( $aliases[$name][0] ) ) {
297 $name = $aliases[$name][0];
298 }
299 return $this->getNsText(NS_SPECIAL) . ':' . $name;
300 }
301
302 function getQuickbarSettings() {
303 return array(
304 $this->getMessage( 'qbsettings-none' ),
305 $this->getMessage( 'qbsettings-fixedleft' ),
306 $this->getMessage( 'qbsettings-fixedright' ),
307 $this->getMessage( 'qbsettings-floatingleft' ),
308 $this->getMessage( 'qbsettings-floatingright' )
309 );
310 }
311
312 function getSkinNames() {
313 $this->load();
314 return $this->skinNames;
315 }
316
317 function getMathNames() {
318 $this->load();
319 return $this->mathNames;
320 }
321
322 function getDatePreferences() {
323 $this->load();
324 return $this->datePreferences;
325 }
326
327 function getDateFormats() {
328 $this->load();
329 return $this->dateFormats;
330 }
331
332 function getDefaultDateFormat() {
333 $this->load();
334 return $this->defaultDateFormat;
335 }
336
337 function getDatePreferenceMigrationMap() {
338 $this->load();
339 return $this->datePreferenceMigrationMap;
340 }
341
342 function getDefaultUserOptionOverrides() {
343 $this->load();
344 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
345 if (is_array($this->defaultUserOptionOverrides)) {
346 return $this->defaultUserOptionOverrides;
347 } else {
348 return array();
349 }
350 }
351
352 function getExtraUserToggles() {
353 $this->load();
354 return $this->extraUserToggles;
355 }
356
357 function getUserToggle( $tog ) {
358 return $this->getMessageFromDB( "tog-$tog" );
359 }
360
361 /**
362 * Get language names, indexed by code.
363 * If $customisedOnly is true, only returns codes with a messages file
364 */
365 public static function getLanguageNames( $customisedOnly = false ) {
366 global $wgLanguageNames, $wgExtraLanguageNames;
367 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
368 if ( !$customisedOnly ) {
369 return $allNames;
370 }
371
372 global $IP;
373 $names = array();
374 $dir = opendir( "$IP/languages/messages" );
375 while( false !== ( $file = readdir( $dir ) ) ) {
376 $m = array();
377 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
378 $code = str_replace( '_', '-', strtolower( $m[1] ) );
379 if ( isset( $allNames[$code] ) ) {
380 $names[$code] = $allNames[$code];
381 }
382 }
383 }
384 closedir( $dir );
385 return $names;
386 }
387
388 /**
389 * Ugly hack to get a message maybe from the MediaWiki namespace, if this
390 * language object is the content or user language.
391 */
392 function getMessageFromDB( $msg ) {
393 global $wgContLang, $wgLang;
394 if ( $wgContLang->getCode() == $this->getCode() ) {
395 # Content language
396 return wfMsgForContent( $msg );
397 } elseif ( $wgLang->getCode() == $this->getCode() ) {
398 # User language
399 return wfMsg( $msg );
400 } else {
401 # Neither, get from localisation
402 return $this->getMessage( $msg );
403 }
404 }
405
406 function getLanguageName( $code ) {
407 $names = self::getLanguageNames();
408 if ( !array_key_exists( $code, $names ) ) {
409 return '';
410 }
411 return $names[$code];
412 }
413
414 function getMonthName( $key ) {
415 return $this->getMessageFromDB( self::$mMonthMsgs[$key-1] );
416 }
417
418 function getMonthNameGen( $key ) {
419 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key-1] );
420 }
421
422 function getMonthAbbreviation( $key ) {
423 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key-1] );
424 }
425
426 function getWeekdayName( $key ) {
427 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key-1] );
428 }
429
430 function getWeekdayAbbreviation( $key ) {
431 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key-1] );
432 }
433
434 function getIranianCalendarMonthName( $key ) {
435 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key-1] );
436 }
437
438 function getHebrewCalendarMonthName( $key ) {
439 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key-1] );
440 }
441
442 function getHebrewCalendarMonthNameGen( $key ) {
443 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key-1] );
444 }
445
446 function getHijriCalendarMonthName( $key ) {
447 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key-1] );
448 }
449
450 /**
451 * Used by date() and time() to adjust the time output.
452 *
453 * @param $ts Int the time in date('YmdHis') format
454 * @param $tz Mixed: adjust the time by this amount (default false, mean we
455 * get user timecorrection setting)
456 * @return int
457 */
458 function userAdjust( $ts, $tz = false ) {
459 global $wgUser, $wgLocalTZoffset;
460
461 if (!$tz) {
462 $tz = $wgUser->getOption( 'timecorrection' );
463 }
464
465 # minutes and hours differences:
466 $minDiff = 0;
467 $hrDiff = 0;
468
469 if ( $tz === '' ) {
470 # Global offset in minutes.
471 if( isset($wgLocalTZoffset) ) {
472 if( $wgLocalTZoffset >= 0 ) {
473 $hrDiff = floor($wgLocalTZoffset / 60);
474 } else {
475 $hrDiff = ceil($wgLocalTZoffset / 60);
476 }
477 $minDiff = $wgLocalTZoffset % 60;
478 }
479 } elseif ( strpos( $tz, ':' ) !== false ) {
480 $tzArray = explode( ':', $tz );
481 $hrDiff = intval($tzArray[0]);
482 $minDiff = intval($hrDiff < 0 ? -$tzArray[1] : $tzArray[1]);
483 } else {
484 $hrDiff = intval( $tz );
485 }
486
487 # No difference ? Return time unchanged
488 if ( 0 == $hrDiff && 0 == $minDiff ) { return $ts; }
489
490 wfSuppressWarnings(); // E_STRICT system time bitching
491 # Generate an adjusted date
492 $t = mktime( (
493 (int)substr( $ts, 8, 2) ) + $hrDiff, # Hours
494 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
495 (int)substr( $ts, 12, 2 ), # Seconds
496 (int)substr( $ts, 4, 2 ), # Month
497 (int)substr( $ts, 6, 2 ), # Day
498 (int)substr( $ts, 0, 4 ) ); #Year
499
500 $date = date( 'YmdHis', $t );
501 wfRestoreWarnings();
502
503 return $date;
504 }
505
506 /**
507 * This is a workalike of PHP's date() function, but with better
508 * internationalisation, a reduced set of format characters, and a better
509 * escaping format.
510 *
511 * Supported format characters are dDjlNwzWFmMntLYyaAgGhHiscrU. See the
512 * PHP manual for definitions. There are a number of extensions, which
513 * start with "x":
514 *
515 * xn Do not translate digits of the next numeric format character
516 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
517 * xr Use roman numerals for the next numeric format character
518 * xh Use hebrew numerals for the next numeric format character
519 * xx Literal x
520 * xg Genitive month name
521 *
522 * xij j (day number) in Iranian calendar
523 * xiF F (month name) in Iranian calendar
524 * xin n (month number) in Iranian calendar
525 * xiY Y (full year) in Iranian calendar
526 *
527 * xjj j (day number) in Hebrew calendar
528 * xjF F (month name) in Hebrew calendar
529 * xjt t (days in month) in Hebrew calendar
530 * xjx xg (genitive month name) in Hebrew calendar
531 * xjn n (month number) in Hebrew calendar
532 * xjY Y (full year) in Hebrew calendar
533 *
534 * xmj j (day number) in Hijri calendar
535 * xmF F (month name) in Hijri calendar
536 * xmn n (month number) in Hijri calendar
537 * xmY Y (full year) in Hijri calendar
538 *
539 * xkY Y (full year) in Thai solar calendar. Months and days are
540 * identical to the Gregorian calendar
541 *
542 * Characters enclosed in double quotes will be considered literal (with
543 * the quotes themselves removed). Unmatched quotes will be considered
544 * literal quotes. Example:
545 *
546 * "The month is" F => The month is January
547 * i's" => 20'11"
548 *
549 * Backslash escaping is also supported.
550 *
551 * Input timestamp is assumed to be pre-normalized to the desired local
552 * time zone, if any.
553 *
554 * @param $format String
555 * @param $ts String: 14-character timestamp
556 * YYYYMMDDHHMMSS
557 * 01234567890123
558 */
559 function sprintfDate( $format, $ts ) {
560 $s = '';
561 $raw = false;
562 $roman = false;
563 $hebrewNum = false;
564 $unix = false;
565 $rawToggle = false;
566 $iranian = false;
567 $hebrew = false;
568 $hijri = false;
569 $thai = false;
570 for ( $p = 0; $p < strlen( $format ); $p++ ) {
571 $num = false;
572 $code = $format[$p];
573 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
574 $code .= $format[++$p];
575 }
576
577 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' ) && $p < strlen( $format ) - 1 ) {
578 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' ) && $p < strlen( $format ) - 1 ) {
579 $code .= $format[++$p];
580 }
581
582 switch ( $code ) {
583 case 'xx':
584 $s .= 'x';
585 break;
586 case 'xn':
587 $raw = true;
588 break;
589 case 'xN':
590 $rawToggle = !$rawToggle;
591 break;
592 case 'xr':
593 $roman = true;
594 break;
595 case 'xh':
596 $hebrewNum = true;
597 break;
598 case 'xg':
599 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
600 break;
601 case 'xjx':
602 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
603 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
604 break;
605 case 'd':
606 $num = substr( $ts, 6, 2 );
607 break;
608 case 'D':
609 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
610 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
611 break;
612 case 'j':
613 $num = intval( substr( $ts, 6, 2 ) );
614 break;
615 case 'xij':
616 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
617 $num = $iranian[2];
618 break;
619 case 'xmj':
620 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
621 $num = $hijri[2];
622 break;
623 case 'xjj':
624 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
625 $num = $hebrew[2];
626 break;
627 case 'l':
628 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
629 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
630 break;
631 case 'N':
632 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
633 $w = gmdate( 'w', $unix );
634 $num = $w ? $w : 7;
635 break;
636 case 'w':
637 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
638 $num = gmdate( 'w', $unix );
639 break;
640 case 'z':
641 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
642 $num = gmdate( 'z', $unix );
643 break;
644 case 'W':
645 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
646 $num = gmdate( 'W', $unix );
647 break;
648 case 'F':
649 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
650 break;
651 case 'xiF':
652 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
653 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
654 break;
655 case 'xmF':
656 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
657 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
658 break;
659 case 'xjF':
660 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
661 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
662 break;
663 case 'm':
664 $num = substr( $ts, 4, 2 );
665 break;
666 case 'M':
667 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
668 break;
669 case 'n':
670 $num = intval( substr( $ts, 4, 2 ) );
671 break;
672 case 'xin':
673 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
674 $num = $iranian[1];
675 break;
676 case 'xmn':
677 if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
678 $num = $hijri[1];
679 break;
680 case 'xjn':
681 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
682 $num = $hebrew[1];
683 break;
684 case 't':
685 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
686 $num = gmdate( 't', $unix );
687 break;
688 case 'xjt':
689 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
690 $num = $hebrew[3];
691 break;
692 case 'L':
693 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
694 $num = gmdate( 'L', $unix );
695 break;
696 case 'Y':
697 $num = substr( $ts, 0, 4 );
698 break;
699 case 'xiY':
700 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
701 $num = $iranian[0];
702 break;
703 case 'xmY':
704 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
705 $num = $hijri[0];
706 break;
707 case 'xjY':
708 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
709 $num = $hebrew[0];
710 break;
711 case 'xkY':
712 if ( !$thai ) $thai = self::tsToThai( $ts );
713 $num = $thai[0];
714 break;
715 case 'y':
716 $num = substr( $ts, 2, 2 );
717 break;
718 case 'a':
719 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
720 break;
721 case 'A':
722 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
723 break;
724 case 'g':
725 $h = substr( $ts, 8, 2 );
726 $num = $h % 12 ? $h % 12 : 12;
727 break;
728 case 'G':
729 $num = intval( substr( $ts, 8, 2 ) );
730 break;
731 case 'h':
732 $h = substr( $ts, 8, 2 );
733 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
734 break;
735 case 'H':
736 $num = substr( $ts, 8, 2 );
737 break;
738 case 'i':
739 $num = substr( $ts, 10, 2 );
740 break;
741 case 's':
742 $num = substr( $ts, 12, 2 );
743 break;
744 case 'c':
745 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
746 $s .= gmdate( 'c', $unix );
747 break;
748 case 'r':
749 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
750 $s .= gmdate( 'r', $unix );
751 break;
752 case 'U':
753 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
754 $num = $unix;
755 break;
756 case '\\':
757 # Backslash escaping
758 if ( $p < strlen( $format ) - 1 ) {
759 $s .= $format[++$p];
760 } else {
761 $s .= '\\';
762 }
763 break;
764 case '"':
765 # Quoted literal
766 if ( $p < strlen( $format ) - 1 ) {
767 $endQuote = strpos( $format, '"', $p + 1 );
768 if ( $endQuote === false ) {
769 # No terminating quote, assume literal "
770 $s .= '"';
771 } else {
772 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
773 $p = $endQuote;
774 }
775 } else {
776 # Quote at end of string, assume literal "
777 $s .= '"';
778 }
779 break;
780 default:
781 $s .= $format[$p];
782 }
783 if ( $num !== false ) {
784 if ( $rawToggle || $raw ) {
785 $s .= $num;
786 $raw = false;
787 } elseif ( $roman ) {
788 $s .= self::romanNumeral( $num );
789 $roman = false;
790 } elseif( $hebrewNum ) {
791 $s .= self::hebrewNumeral( $num );
792 $hebrewNum = false;
793 } else {
794 $s .= $this->formatNum( $num, true );
795 }
796 $num = false;
797 }
798 }
799 return $s;
800 }
801
802 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
803 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
804 /**
805 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
806 * Gregorian dates to Iranian dates. Originally written in C, it
807 * is released under the terms of GNU Lesser General Public
808 * License. Conversion to PHP was performed by Niklas Laxström.
809 *
810 * Link: http://www.farsiweb.info/jalali/jalali.c
811 */
812 private static function tsToIranian( $ts ) {
813 $gy = substr( $ts, 0, 4 ) -1600;
814 $gm = substr( $ts, 4, 2 ) -1;
815 $gd = substr( $ts, 6, 2 ) -1;
816
817 # Days passed from the beginning (including leap years)
818 $gDayNo = 365*$gy
819 + floor(($gy+3) / 4)
820 - floor(($gy+99) / 100)
821 + floor(($gy+399) / 400);
822
823
824 // Add days of the past months of this year
825 for( $i = 0; $i < $gm; $i++ ) {
826 $gDayNo += self::$GREG_DAYS[$i];
827 }
828
829 // Leap years
830 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
831 $gDayNo++;
832 }
833
834 // Days passed in current month
835 $gDayNo += $gd;
836
837 $jDayNo = $gDayNo - 79;
838
839 $jNp = floor($jDayNo / 12053);
840 $jDayNo %= 12053;
841
842 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
843 $jDayNo %= 1461;
844
845 if ( $jDayNo >= 366 ) {
846 $jy += floor(($jDayNo-1)/365);
847 $jDayNo = floor(($jDayNo-1)%365);
848 }
849
850 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
851 $jDayNo -= self::$IRANIAN_DAYS[$i];
852 }
853
854 $jm= $i+1;
855 $jd= $jDayNo+1;
856
857 return array($jy, $jm, $jd);
858 }
859 /**
860 * Converting Gregorian dates to Hijri dates.
861 *
862 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
863 *
864 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
865 */
866 private static function tsToHijri ( $ts ) {
867 $year = substr( $ts, 0, 4 );
868 $month = substr( $ts, 4, 2 );
869 $day = substr( $ts, 6, 2 );
870
871 $zyr = $year;
872 $zd=$day;
873 $zm=$month;
874 $zy=$zyr;
875
876
877
878 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
879 {
880
881
882 $zjd=(int)((1461*($zy + 4800 + (int)( ($zm-14) /12) ))/4) + (int)((367*($zm-2-12*((int)(($zm-14)/12))))/12)-(int)((3*(int)(( ($zy+4900+(int)(($zm-14)/12))/100)))/4)+$zd-32075;
883 }
884 else
885 {
886 $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
887 }
888
889 $zl=$zjd-1948440+10632;
890 $zn=(int)(($zl-1)/10631);
891 $zl=$zl-10631*$zn+354;
892 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
893 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
894 $zm=(int)((24*$zl)/709);
895 $zd=$zl-(int)((709*$zm)/24);
896 $zy=30*$zn+$zj-30;
897
898 return array ($zy, $zm, $zd);
899 }
900
901 /**
902 * Converting Gregorian dates to Hebrew dates.
903 *
904 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
905 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
906 * to translate the relevant functions into PHP and release them under
907 * GNU GPL.
908 */
909 private static function tsToHebrew( $ts ) {
910 # Parse date
911 $year = substr( $ts, 0, 4 );
912 $month = substr( $ts, 4, 2 );
913 $day = substr( $ts, 6, 2 );
914
915 # Calculate Hebrew year
916 $hebrewYear = $year + 3760;
917
918 # Month number when September = 1, August = 12
919 $month += 4;
920 if( $month > 12 ) {
921 # Next year
922 $month -= 12;
923 $year++;
924 $hebrewYear++;
925 }
926
927 # Calculate day of year from 1 September
928 $dayOfYear = $day;
929 for( $i = 1; $i < $month; $i++ ) {
930 if( $i == 6 ) {
931 # February
932 $dayOfYear += 28;
933 # Check if the year is leap
934 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
935 $dayOfYear++;
936 }
937 } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
938 $dayOfYear += 30;
939 } else {
940 $dayOfYear += 31;
941 }
942 }
943
944 # Calculate the start of the Hebrew year
945 $start = self::hebrewYearStart( $hebrewYear );
946
947 # Calculate next year's start
948 if( $dayOfYear <= $start ) {
949 # Day is before the start of the year - it is the previous year
950 # Next year's start
951 $nextStart = $start;
952 # Previous year
953 $year--;
954 $hebrewYear--;
955 # Add days since previous year's 1 September
956 $dayOfYear += 365;
957 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
958 # Leap year
959 $dayOfYear++;
960 }
961 # Start of the new (previous) year
962 $start = self::hebrewYearStart( $hebrewYear );
963 } else {
964 # Next year's start
965 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
966 }
967
968 # Calculate Hebrew day of year
969 $hebrewDayOfYear = $dayOfYear - $start;
970
971 # Difference between year's days
972 $diff = $nextStart - $start;
973 # Add 12 (or 13 for leap years) days to ignore the difference between
974 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
975 # difference is only about the year type
976 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
977 $diff += 13;
978 } else {
979 $diff += 12;
980 }
981
982 # Check the year pattern, and is leap year
983 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
984 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
985 # and non-leap years
986 $yearPattern = $diff % 30;
987 # Check if leap year
988 $isLeap = $diff >= 30;
989
990 # Calculate day in the month from number of day in the Hebrew year
991 # Don't check Adar - if the day is not in Adar, we will stop before;
992 # if it is in Adar, we will use it to check if it is Adar I or Adar II
993 $hebrewDay = $hebrewDayOfYear;
994 $hebrewMonth = 1;
995 $days = 0;
996 while( $hebrewMonth <= 12 ) {
997 # Calculate days in this month
998 if( $isLeap && $hebrewMonth == 6 ) {
999 # Adar in a leap year
1000 if( $isLeap ) {
1001 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1002 $days = 30;
1003 if( $hebrewDay <= $days ) {
1004 # Day in Adar I
1005 $hebrewMonth = 13;
1006 } else {
1007 # Subtract the days of Adar I
1008 $hebrewDay -= $days;
1009 # Try Adar II
1010 $days = 29;
1011 if( $hebrewDay <= $days ) {
1012 # Day in Adar II
1013 $hebrewMonth = 14;
1014 }
1015 }
1016 }
1017 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1018 # Cheshvan in a complete year (otherwise as the rule below)
1019 $days = 30;
1020 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1021 # Kislev in an incomplete year (otherwise as the rule below)
1022 $days = 29;
1023 } else {
1024 # Odd months have 30 days, even have 29
1025 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1026 }
1027 if( $hebrewDay <= $days ) {
1028 # In the current month
1029 break;
1030 } else {
1031 # Subtract the days of the current month
1032 $hebrewDay -= $days;
1033 # Try in the next month
1034 $hebrewMonth++;
1035 }
1036 }
1037
1038 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1039 }
1040
1041 /**
1042 * This calculates the Hebrew year start, as days since 1 September.
1043 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1044 * Used for Hebrew date.
1045 */
1046 private static function hebrewYearStart( $year ) {
1047 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1048 $b = intval( ( $year - 1 ) % 4 );
1049 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1050 if( $m < 0 ) {
1051 $m--;
1052 }
1053 $Mar = intval( $m );
1054 if( $m < 0 ) {
1055 $m++;
1056 }
1057 $m -= $Mar;
1058
1059 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1060 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1061 $Mar++;
1062 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1063 $Mar += 2;
1064 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1065 $Mar++;
1066 }
1067
1068 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1069 return $Mar;
1070 }
1071
1072 /**
1073 * Algorithm to convert Gregorian dates to Thai solar dates.
1074 *
1075 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1076 *
1077 * @param $ts String: 14-character timestamp
1078 * @return array converted year, month, day
1079 */
1080 private static function tsToThai( $ts ) {
1081 $gy = substr( $ts, 0, 4 );
1082 $gm = substr( $ts, 4, 2 );
1083 $gd = substr( $ts, 6, 2 );
1084
1085 # Add 543 years to the Gregorian calendar
1086 # Months and days are identical
1087 $gy_thai = $gy + 543;
1088
1089 return array( $gy_thai, $gm, $gd );
1090 }
1091
1092
1093 /**
1094 * Roman number formatting up to 3000
1095 */
1096 static function romanNumeral( $num ) {
1097 static $table = array(
1098 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1099 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1100 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1101 array( '', 'M', 'MM', 'MMM' )
1102 );
1103
1104 $num = intval( $num );
1105 if ( $num > 3000 || $num <= 0 ) {
1106 return $num;
1107 }
1108
1109 $s = '';
1110 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1111 if ( $num >= $pow10 ) {
1112 $s .= $table[$i][floor($num / $pow10)];
1113 }
1114 $num = $num % $pow10;
1115 }
1116 return $s;
1117 }
1118
1119 /**
1120 * Hebrew Gematria number formatting up to 9999
1121 */
1122 static function hebrewNumeral( $num ) {
1123 static $table = array(
1124 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1125 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1126 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1127 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1128 );
1129
1130 $num = intval( $num );
1131 if ( $num > 9999 || $num <= 0 ) {
1132 return $num;
1133 }
1134
1135 $s = '';
1136 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1137 if ( $num >= $pow10 ) {
1138 if ( $num == 15 || $num == 16 ) {
1139 $s .= $table[0][9] . $table[0][$num - 9];
1140 $num = 0;
1141 } else {
1142 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1143 if( $pow10 == 1000 ) {
1144 $s .= "'";
1145 }
1146 }
1147 }
1148 $num = $num % $pow10;
1149 }
1150 if( strlen( $s ) == 2 ) {
1151 $str = $s . "'";
1152 } else {
1153 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1154 $str .= substr( $s, strlen( $s ) - 2, 2 );
1155 }
1156 $start = substr( $str, 0, strlen( $str ) - 2 );
1157 $end = substr( $str, strlen( $str ) - 2 );
1158 switch( $end ) {
1159 case 'כ':
1160 $str = $start . 'ך';
1161 break;
1162 case 'מ':
1163 $str = $start . 'ם';
1164 break;
1165 case 'נ':
1166 $str = $start . 'ן';
1167 break;
1168 case 'פ':
1169 $str = $start . 'ף';
1170 break;
1171 case 'צ':
1172 $str = $start . 'ץ';
1173 break;
1174 }
1175 return $str;
1176 }
1177
1178 /**
1179 * This is meant to be used by time(), date(), and timeanddate() to get
1180 * the date preference they're supposed to use, it should be used in
1181 * all children.
1182 *
1183 *<code>
1184 * function timeanddate([...], $format = true) {
1185 * $datePreference = $this->dateFormat($format);
1186 * [...]
1187 * }
1188 *</code>
1189 *
1190 * @param $usePrefs Mixed: if true, the user's preference is used
1191 * if false, the site/language default is used
1192 * if int/string, assumed to be a format.
1193 * @return string
1194 */
1195 function dateFormat( $usePrefs = true ) {
1196 global $wgUser;
1197
1198 if( is_bool( $usePrefs ) ) {
1199 if( $usePrefs ) {
1200 $datePreference = $wgUser->getDatePreference();
1201 } else {
1202 $options = User::getDefaultOptions();
1203 $datePreference = (string)$options['date'];
1204 }
1205 } else {
1206 $datePreference = (string)$usePrefs;
1207 }
1208
1209 // return int
1210 if( $datePreference == '' ) {
1211 return 'default';
1212 }
1213
1214 return $datePreference;
1215 }
1216
1217 /**
1218 * @param $ts Mixed: the time format which needs to be turned into a
1219 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1220 * @param $adj Bool: whether to adjust the time output according to the
1221 * user configured offset ($timecorrection)
1222 * @param $format Mixed: true to use user's date format preference
1223 * @param $timecorrection String: the time offset as returned by
1224 * validateTimeZone() in Special:Preferences
1225 * @return string
1226 */
1227 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1228 $this->load();
1229 if ( $adj ) {
1230 $ts = $this->userAdjust( $ts, $timecorrection );
1231 }
1232
1233 $pref = $this->dateFormat( $format );
1234 if( $pref == 'default' || !isset( $this->dateFormats["$pref date"] ) ) {
1235 $pref = $this->defaultDateFormat;
1236 }
1237 return $this->sprintfDate( $this->dateFormats["$pref date"], $ts );
1238 }
1239
1240 /**
1241 * @param $ts Mixed: the time format which needs to be turned into a
1242 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1243 * @param $adj Bool: whether to adjust the time output according to the
1244 * user configured offset ($timecorrection)
1245 * @param $format Mixed: true to use user's date format preference
1246 * @param $timecorrection String: the time offset as returned by
1247 * validateTimeZone() in Special:Preferences
1248 * @return string
1249 */
1250 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1251 $this->load();
1252 if ( $adj ) {
1253 $ts = $this->userAdjust( $ts, $timecorrection );
1254 }
1255
1256 $pref = $this->dateFormat( $format );
1257 if( $pref == 'default' || !isset( $this->dateFormats["$pref time"] ) ) {
1258 $pref = $this->defaultDateFormat;
1259 }
1260 return $this->sprintfDate( $this->dateFormats["$pref time"], $ts );
1261 }
1262
1263 /**
1264 * @param $ts Mixed: the time format which needs to be turned into a
1265 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1266 * @param $adj Bool: whether to adjust the time output according to the
1267 * user configured offset ($timecorrection)
1268 * @param $format Mixed: what format to return, if it's false output the
1269 * default one (default true)
1270 * @param $timecorrection String: the time offset as returned by
1271 * validateTimeZone() in Special:Preferences
1272 * @return string
1273 */
1274 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1275 $this->load();
1276
1277 $ts = wfTimestamp( TS_MW, $ts );
1278
1279 if ( $adj ) {
1280 $ts = $this->userAdjust( $ts, $timecorrection );
1281 }
1282
1283 $pref = $this->dateFormat( $format );
1284 if( $pref == 'default' || !isset( $this->dateFormats["$pref both"] ) ) {
1285 $pref = $this->defaultDateFormat;
1286 }
1287
1288 return $this->sprintfDate( $this->dateFormats["$pref both"], $ts );
1289 }
1290
1291 function getMessage( $key ) {
1292 $this->load();
1293 return isset( $this->messages[$key] ) ? $this->messages[$key] : null;
1294 }
1295
1296 function getAllMessages() {
1297 $this->load();
1298 return $this->messages;
1299 }
1300
1301 function iconv( $in, $out, $string ) {
1302 # For most languages, this is a wrapper for iconv
1303 return iconv( $in, $out . '//IGNORE', $string );
1304 }
1305
1306 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1307 function ucwordbreaksCallbackAscii($matches){
1308 return $this->ucfirst($matches[1]);
1309 }
1310
1311 function ucwordbreaksCallbackMB($matches){
1312 return mb_strtoupper($matches[0]);
1313 }
1314
1315 function ucCallback($matches){
1316 list( $wikiUpperChars ) = self::getCaseMaps();
1317 return strtr( $matches[1], $wikiUpperChars );
1318 }
1319
1320 function lcCallback($matches){
1321 list( , $wikiLowerChars ) = self::getCaseMaps();
1322 return strtr( $matches[1], $wikiLowerChars );
1323 }
1324
1325 function ucwordsCallbackMB($matches){
1326 return mb_strtoupper($matches[0]);
1327 }
1328
1329 function ucwordsCallbackWiki($matches){
1330 list( $wikiUpperChars ) = self::getCaseMaps();
1331 return strtr( $matches[0], $wikiUpperChars );
1332 }
1333
1334 function ucfirst( $str ) {
1335 if ( empty($str) ) return $str;
1336 if ( ord($str[0]) < 128 ) return ucfirst($str);
1337 else return self::uc($str,true); // fall back to more complex logic in case of multibyte strings
1338 }
1339
1340 function uc( $str, $first = false ) {
1341 if ( function_exists( 'mb_strtoupper' ) ) {
1342 if ( $first ) {
1343 if ( self::isMultibyte( $str ) ) {
1344 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1345 } else {
1346 return ucfirst( $str );
1347 }
1348 } else {
1349 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1350 }
1351 } else {
1352 if ( self::isMultibyte( $str ) ) {
1353 list( $wikiUpperChars ) = $this->getCaseMaps();
1354 $x = $first ? '^' : '';
1355 return preg_replace_callback(
1356 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1357 array($this,"ucCallback"),
1358 $str
1359 );
1360 } else {
1361 return $first ? ucfirst( $str ) : strtoupper( $str );
1362 }
1363 }
1364 }
1365
1366 function lcfirst( $str ) {
1367 if ( empty($str) ) return $str;
1368 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1369 // editing string in place = cool
1370 $str[0]=strtolower($str[0]);
1371 return $str;
1372 }
1373 else return self::lc( $str, true );
1374 }
1375
1376 function lc( $str, $first = false ) {
1377 if ( function_exists( 'mb_strtolower' ) )
1378 if ( $first )
1379 if ( self::isMultibyte( $str ) )
1380 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1381 else
1382 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1383 else
1384 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1385 else
1386 if ( self::isMultibyte( $str ) ) {
1387 list( , $wikiLowerChars ) = self::getCaseMaps();
1388 $x = $first ? '^' : '';
1389 return preg_replace_callback(
1390 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1391 array($this,"lcCallback"),
1392 $str
1393 );
1394 } else
1395 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1396 }
1397
1398 function isMultibyte( $str ) {
1399 return (bool)preg_match( '/[\x80-\xff]/', $str );
1400 }
1401
1402 function ucwords($str) {
1403 if ( self::isMultibyte( $str ) ) {
1404 $str = self::lc($str);
1405
1406 // regexp to find first letter in each word (i.e. after each space)
1407 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1408
1409 // function to use to capitalize a single char
1410 if ( function_exists( 'mb_strtoupper' ) )
1411 return preg_replace_callback(
1412 $replaceRegexp,
1413 array($this,"ucwordsCallbackMB"),
1414 $str
1415 );
1416 else
1417 return preg_replace_callback(
1418 $replaceRegexp,
1419 array($this,"ucwordsCallbackWiki"),
1420 $str
1421 );
1422 }
1423 else
1424 return ucwords( strtolower( $str ) );
1425 }
1426
1427 # capitalize words at word breaks
1428 function ucwordbreaks($str){
1429 if (self::isMultibyte( $str ) ) {
1430 $str = self::lc($str);
1431
1432 // since \b doesn't work for UTF-8, we explicitely define word break chars
1433 $breaks= "[ \-\(\)\}\{\.,\?!]";
1434
1435 // find first letter after word break
1436 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1437
1438 if ( function_exists( 'mb_strtoupper' ) )
1439 return preg_replace_callback(
1440 $replaceRegexp,
1441 array($this,"ucwordbreaksCallbackMB"),
1442 $str
1443 );
1444 else
1445 return preg_replace_callback(
1446 $replaceRegexp,
1447 array($this,"ucwordsCallbackWiki"),
1448 $str
1449 );
1450 }
1451 else
1452 return preg_replace_callback(
1453 '/\b([\w\x80-\xff]+)\b/',
1454 array($this,"ucwordbreaksCallbackAscii"),
1455 $str );
1456 }
1457
1458 /**
1459 * Return a case-folded representation of $s
1460 *
1461 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1462 * and $s2 are the same except for the case of their characters. It is not
1463 * necessary for the value returned to make sense when displayed.
1464 *
1465 * Do *not* perform any other normalisation in this function. If a caller
1466 * uses this function when it should be using a more general normalisation
1467 * function, then fix the caller.
1468 */
1469 function caseFold( $s ) {
1470 return $this->uc( $s );
1471 }
1472
1473 function checkTitleEncoding( $s ) {
1474 if( is_array( $s ) ) {
1475 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1476 }
1477 # Check for non-UTF-8 URLs
1478 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1479 if(!$ishigh) return $s;
1480
1481 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1482 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1483 if( $isutf8 ) return $s;
1484
1485 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1486 }
1487
1488 function fallback8bitEncoding() {
1489 $this->load();
1490 return $this->fallback8bitEncoding;
1491 }
1492
1493 /**
1494 * Some languages have special punctuation to strip out
1495 * or characters which need to be converted for MySQL's
1496 * indexing to grok it correctly. Make such changes here.
1497 *
1498 * @param $string String
1499 * @return String
1500 */
1501 function stripForSearch( $string ) {
1502 global $wgDBtype;
1503 if ( $wgDBtype != 'mysql' ) {
1504 return $string;
1505 }
1506
1507 # MySQL fulltext index doesn't grok utf-8, so we
1508 # need to fold cases and convert to hex
1509
1510 wfProfileIn( __METHOD__ );
1511 if( function_exists( 'mb_strtolower' ) ) {
1512 $out = preg_replace(
1513 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1514 "'U8' . bin2hex( \"$1\" )",
1515 mb_strtolower( $string ) );
1516 } else {
1517 list( , $wikiLowerChars ) = self::getCaseMaps();
1518 $out = preg_replace(
1519 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1520 "'U8' . bin2hex( strtr( \"\$1\", \$wikiLowerChars ) )",
1521 $string );
1522 }
1523 wfProfileOut( __METHOD__ );
1524 return $out;
1525 }
1526
1527 function convertForSearchResult( $termsArray ) {
1528 # some languages, e.g. Chinese, need to do a conversion
1529 # in order for search results to be displayed correctly
1530 return $termsArray;
1531 }
1532
1533 /**
1534 * Get the first character of a string.
1535 *
1536 * @param $s string
1537 * @return string
1538 */
1539 function firstChar( $s ) {
1540 $matches = array();
1541 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1542 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1543
1544 if ( isset( $matches[1] ) ) {
1545 if ( strlen( $matches[1] ) != 3 ) {
1546 return $matches[1];
1547 }
1548
1549 // Break down Hangul syllables to grab the first jamo
1550 $code = utf8ToCodepoint( $matches[1] );
1551 if ( $code < 0xac00 || 0xd7a4 <= $code) {
1552 return $matches[1];
1553 } elseif ( $code < 0xb098 ) {
1554 return "\xe3\x84\xb1";
1555 } elseif ( $code < 0xb2e4 ) {
1556 return "\xe3\x84\xb4";
1557 } elseif ( $code < 0xb77c ) {
1558 return "\xe3\x84\xb7";
1559 } elseif ( $code < 0xb9c8 ) {
1560 return "\xe3\x84\xb9";
1561 } elseif ( $code < 0xbc14 ) {
1562 return "\xe3\x85\x81";
1563 } elseif ( $code < 0xc0ac ) {
1564 return "\xe3\x85\x82";
1565 } elseif ( $code < 0xc544 ) {
1566 return "\xe3\x85\x85";
1567 } elseif ( $code < 0xc790 ) {
1568 return "\xe3\x85\x87";
1569 } elseif ( $code < 0xcc28 ) {
1570 return "\xe3\x85\x88";
1571 } elseif ( $code < 0xce74 ) {
1572 return "\xe3\x85\x8a";
1573 } elseif ( $code < 0xd0c0 ) {
1574 return "\xe3\x85\x8b";
1575 } elseif ( $code < 0xd30c ) {
1576 return "\xe3\x85\x8c";
1577 } elseif ( $code < 0xd558 ) {
1578 return "\xe3\x85\x8d";
1579 } else {
1580 return "\xe3\x85\x8e";
1581 }
1582 } else {
1583 return "";
1584 }
1585 }
1586
1587 function initEncoding() {
1588 # Some languages may have an alternate char encoding option
1589 # (Esperanto X-coding, Japanese furigana conversion, etc)
1590 # If this language is used as the primary content language,
1591 # an override to the defaults can be set here on startup.
1592 }
1593
1594 function recodeForEdit( $s ) {
1595 # For some languages we'll want to explicitly specify
1596 # which characters make it into the edit box raw
1597 # or are converted in some way or another.
1598 # Note that if wgOutputEncoding is different from
1599 # wgInputEncoding, this text will be further converted
1600 # to wgOutputEncoding.
1601 global $wgEditEncoding;
1602 if( $wgEditEncoding == '' or
1603 $wgEditEncoding == 'UTF-8' ) {
1604 return $s;
1605 } else {
1606 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1607 }
1608 }
1609
1610 function recodeInput( $s ) {
1611 # Take the previous into account.
1612 global $wgEditEncoding;
1613 if($wgEditEncoding != "") {
1614 $enc = $wgEditEncoding;
1615 } else {
1616 $enc = 'UTF-8';
1617 }
1618 if( $enc == 'UTF-8' ) {
1619 return $s;
1620 } else {
1621 return $this->iconv( $enc, 'UTF-8', $s );
1622 }
1623 }
1624
1625 /**
1626 * For right-to-left language support
1627 *
1628 * @return bool
1629 */
1630 function isRTL() {
1631 $this->load();
1632 return $this->rtl;
1633 }
1634
1635 /**
1636 * A hidden direction mark (LRM or RLM), depending on the language direction
1637 *
1638 * @return string
1639 */
1640 function getDirMark() {
1641 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1642 }
1643
1644 /**
1645 * An arrow, depending on the language direction
1646 *
1647 * @return string
1648 */
1649 function getArrow() {
1650 return $this->isRTL() ? '←' : '→';
1651 }
1652
1653 /**
1654 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1655 *
1656 * @return bool
1657 */
1658 function linkPrefixExtension() {
1659 $this->load();
1660 return $this->linkPrefixExtension;
1661 }
1662
1663 function &getMagicWords() {
1664 $this->load();
1665 return $this->magicWords;
1666 }
1667
1668 # Fill a MagicWord object with data from here
1669 function getMagic( &$mw ) {
1670 if ( !$this->mMagicHookDone ) {
1671 $this->mMagicHookDone = true;
1672 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1673 }
1674 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1675 $rawEntry = $this->mMagicExtensions[$mw->mId];
1676 } else {
1677 $magicWords =& $this->getMagicWords();
1678 if ( isset( $magicWords[$mw->mId] ) ) {
1679 $rawEntry = $magicWords[$mw->mId];
1680 } else {
1681 # Fall back to English if local list is incomplete
1682 $magicWords =& Language::getMagicWords();
1683 if ( !isset($magicWords[$mw->mId]) ) { throw new MWException("Magic word not found" ); }
1684 $rawEntry = $magicWords[$mw->mId];
1685 }
1686 }
1687
1688 if( !is_array( $rawEntry ) ) {
1689 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1690 } else {
1691 $mw->mCaseSensitive = $rawEntry[0];
1692 $mw->mSynonyms = array_slice( $rawEntry, 1 );
1693 }
1694 }
1695
1696 /**
1697 * Add magic words to the extension array
1698 */
1699 function addMagicWordsByLang( $newWords ) {
1700 $code = $this->getCode();
1701 $fallbackChain = array();
1702 while ( $code && !in_array( $code, $fallbackChain ) ) {
1703 $fallbackChain[] = $code;
1704 $code = self::getFallbackFor( $code );
1705 }
1706 if ( !in_array( 'en', $fallbackChain ) ) {
1707 $fallbackChain[] = 'en';
1708 }
1709 $fallbackChain = array_reverse( $fallbackChain );
1710 foreach ( $fallbackChain as $code ) {
1711 if ( isset( $newWords[$code] ) ) {
1712 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1713 }
1714 }
1715 }
1716
1717 /**
1718 * Get special page names, as an associative array
1719 * case folded alias => real name
1720 */
1721 function getSpecialPageAliases() {
1722 $this->load();
1723 if ( !isset( $this->mExtendedSpecialPageAliases ) ) {
1724 $this->mExtendedSpecialPageAliases = $this->specialPageAliases;
1725 wfRunHooks( 'LanguageGetSpecialPageAliases',
1726 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1727 }
1728 return $this->mExtendedSpecialPageAliases;
1729 }
1730
1731 /**
1732 * Italic is unsuitable for some languages
1733 *
1734 * @param $text String: the text to be emphasized.
1735 * @return string
1736 */
1737 function emphasize( $text ) {
1738 return "<em>$text</em>";
1739 }
1740
1741 /**
1742 * Normally we output all numbers in plain en_US style, that is
1743 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1744 * point twohundredthirtyfive. However this is not sutable for all
1745 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1746 * Icelandic just want to use commas instead of dots, and dots instead
1747 * of commas like "293.291,235".
1748 *
1749 * An example of this function being called:
1750 * <code>
1751 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1752 * </code>
1753 *
1754 * See LanguageGu.php for the Gujarati implementation and
1755 * LanguageIs.php for the , => . and . => , implementation.
1756 *
1757 * @todo check if it's viable to use localeconv() for the decimal
1758 * seperator thing.
1759 * @param $number Mixed: the string to be formatted, should be an integer
1760 * or a floating point number.
1761 * @param $nocommafy Bool: set to true for special numbers like dates
1762 * @return string
1763 */
1764 function formatNum( $number, $nocommafy = false ) {
1765 global $wgTranslateNumerals;
1766 if (!$nocommafy) {
1767 $number = $this->commafy($number);
1768 $s = $this->separatorTransformTable();
1769 if (!is_null($s)) { $number = strtr($number, $s); }
1770 }
1771
1772 if ($wgTranslateNumerals) {
1773 $s = $this->digitTransformTable();
1774 if (!is_null($s)) { $number = strtr($number, $s); }
1775 }
1776
1777 return $number;
1778 }
1779
1780 function parseFormattedNumber( $number ) {
1781 $s = $this->digitTransformTable();
1782 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1783
1784 $s = $this->separatorTransformTable();
1785 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1786
1787 $number = strtr( $number, array (',' => '') );
1788 return $number;
1789 }
1790
1791 /**
1792 * Adds commas to a given number
1793 *
1794 * @param $_ mixed
1795 * @return string
1796 */
1797 function commafy($_) {
1798 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1799 }
1800
1801 function digitTransformTable() {
1802 $this->load();
1803 return $this->digitTransformTable;
1804 }
1805
1806 function separatorTransformTable() {
1807 $this->load();
1808 return $this->separatorTransformTable;
1809 }
1810
1811
1812 /**
1813 * For the credit list in includes/Credits.php (action=credits)
1814 *
1815 * @param $l Array
1816 * @return string
1817 */
1818 function listToText( $l ) {
1819 $s = '';
1820 $m = count($l) - 1;
1821 for ($i = $m; $i >= 0; $i--) {
1822 if ($i == $m) {
1823 $s = $l[$i];
1824 } else if ($i == $m - 1) {
1825 $s = $l[$i] . ' ' . $this->getMessageFromDB( 'and' ) . ' ' . $s;
1826 } else {
1827 $s = $l[$i] . ', ' . $s;
1828 }
1829 }
1830 return $s;
1831 }
1832
1833 /**
1834 * Truncate a string to a specified length in bytes, appending an optional
1835 * string (e.g. for ellipses)
1836 *
1837 * The database offers limited byte lengths for some columns in the database;
1838 * multi-byte character sets mean we need to ensure that only whole characters
1839 * are included, otherwise broken characters can be passed to the user
1840 *
1841 * If $length is negative, the string will be truncated from the beginning
1842 *
1843 * @param $string String to truncate
1844 * @param $length Int: maximum length (excluding ellipses)
1845 * @param $ellipsis String to append to the truncated text
1846 * @return string
1847 */
1848 function truncate( $string, $length, $ellipsis = "" ) {
1849 if( $length == 0 ) {
1850 return $ellipsis;
1851 }
1852 if ( strlen( $string ) <= abs( $length ) ) {
1853 return $string;
1854 }
1855 if( $length > 0 ) {
1856 $string = substr( $string, 0, $length );
1857 $char = ord( $string[strlen( $string ) - 1] );
1858 $m = array();
1859 if ($char >= 0xc0) {
1860 # We got the first byte only of a multibyte char; remove it.
1861 $string = substr( $string, 0, -1 );
1862 } elseif( $char >= 0x80 &&
1863 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
1864 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
1865 # We chopped in the middle of a character; remove it
1866 $string = $m[1];
1867 }
1868 return $string . $ellipsis;
1869 } else {
1870 $string = substr( $string, $length );
1871 $char = ord( $string[0] );
1872 if( $char >= 0x80 && $char < 0xc0 ) {
1873 # We chopped in the middle of a character; remove the whole thing
1874 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
1875 }
1876 return $ellipsis . $string;
1877 }
1878 }
1879
1880 /**
1881 * Grammatical transformations, needed for inflected languages
1882 * Invoked by putting {{grammar:case|word}} in a message
1883 *
1884 * @param $word string
1885 * @param $case string
1886 * @return string
1887 */
1888 function convertGrammar( $word, $case ) {
1889 global $wgGrammarForms;
1890 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
1891 return $wgGrammarForms[$this->getCode()][$case][$word];
1892 }
1893 return $word;
1894 }
1895
1896 /**
1897 * Plural form transformations, needed for some languages.
1898 * For example, there are 3 form of plural in Russian and Polish,
1899 * depending on "count mod 10". See [[w:Plural]]
1900 * For English it is pretty simple.
1901 *
1902 * Invoked by putting {{plural:count|wordform1|wordform2}}
1903 * or {{plural:count|wordform1|wordform2|wordform3}}
1904 *
1905 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
1906 *
1907 * @param $count Integer: non-localized number
1908 * @param $forms Array: different plural forms
1909 * @return string Correct form of plural for $count in this language
1910 */
1911 function convertPlural( $count, $forms ) {
1912 if ( !count($forms) ) { return ''; }
1913 $forms = $this->preConvertPlural( $forms, 2 );
1914
1915 return ( abs($count) == 1 ) ? $forms[0] : $forms[1];
1916 }
1917
1918 /**
1919 * Checks that convertPlural was given an array and pads it to requested
1920 * amound of forms by copying the last one.
1921 *
1922 * @param $count Integer: How many forms should there be at least
1923 * @param $forms Array of forms given to convertPlural
1924 * @return array Padded array of forms or an exception if not an array
1925 */
1926 protected function preConvertPlural( /* Array */ $forms, $count ) {
1927 while ( count($forms) < $count ) {
1928 $forms[] = $forms[count($forms)-1];
1929 }
1930 return $forms;
1931 }
1932
1933 /**
1934 * For translaing of expiry times
1935 * @param $str String: the validated block time in English
1936 * @return Somehow translated block time
1937 * @see LanguageFi.php for example implementation
1938 */
1939 function translateBlockExpiry( $str ) {
1940
1941 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
1942
1943 if ( $scBlockExpiryOptions == '-') {
1944 return $str;
1945 }
1946
1947 foreach (explode(',', $scBlockExpiryOptions) as $option) {
1948 if ( strpos($option, ":") === false )
1949 continue;
1950 list($show, $value) = explode(":", $option);
1951 if ( strcmp ( $str, $value) == 0 ) {
1952 return htmlspecialchars( trim( $show ) );
1953 }
1954 }
1955
1956 return $str;
1957 }
1958
1959 /**
1960 * languages like Chinese need to be segmented in order for the diff
1961 * to be of any use
1962 *
1963 * @param $text String
1964 * @return String
1965 */
1966 function segmentForDiff( $text ) {
1967 return $text;
1968 }
1969
1970 /**
1971 * and unsegment to show the result
1972 *
1973 * @param $text String
1974 * @return String
1975 */
1976 function unsegmentForDiff( $text ) {
1977 return $text;
1978 }
1979
1980 # convert text to different variants of a language.
1981 function convert( $text, $isTitle = false) {
1982 return $this->mConverter->convert($text, $isTitle);
1983 }
1984
1985 # Convert text from within Parser
1986 function parserConvert( $text, &$parser ) {
1987 return $this->mConverter->parserConvert( $text, $parser );
1988 }
1989
1990 # Check if this is a language with variants
1991 function hasVariants(){
1992 return sizeof($this->getVariants())>1;
1993 }
1994
1995 # Put custom tags (e.g. -{ }-) around math to prevent conversion
1996 function armourMath($text){
1997 return $this->mConverter->armourMath($text);
1998 }
1999
2000
2001 /**
2002 * Perform output conversion on a string, and encode for safe HTML output.
2003 * @param $text String
2004 * @param $isTitle Bool -- wtf?
2005 * @return string
2006 * @todo this should get integrated somewhere sane
2007 */
2008 function convertHtml( $text, $isTitle = false ) {
2009 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2010 }
2011
2012 function convertCategoryKey( $key ) {
2013 return $this->mConverter->convertCategoryKey( $key );
2014 }
2015
2016 /**
2017 * get the list of variants supported by this langauge
2018 * see sample implementation in LanguageZh.php
2019 *
2020 * @return array an array of language codes
2021 */
2022 function getVariants() {
2023 return $this->mConverter->getVariants();
2024 }
2025
2026
2027 function getPreferredVariant( $fromUser = true ) {
2028 return $this->mConverter->getPreferredVariant( $fromUser );
2029 }
2030
2031 /**
2032 * if a language supports multiple variants, it is
2033 * possible that non-existing link in one variant
2034 * actually exists in another variant. this function
2035 * tries to find it. See e.g. LanguageZh.php
2036 *
2037 * @param $link String: the name of the link
2038 * @param $nt Mixed: the title object of the link
2039 * @return null the input parameters may be modified upon return
2040 */
2041 function findVariantLink( &$link, &$nt ) {
2042 $this->mConverter->findVariantLink($link, $nt);
2043 }
2044
2045 /**
2046 * If a language supports multiple variants, converts text
2047 * into an array of all possible variants of the text:
2048 * 'variant' => text in that variant
2049 */
2050
2051 function convertLinkToAllVariants($text){
2052 return $this->mConverter->convertLinkToAllVariants($text);
2053 }
2054
2055
2056 /**
2057 * returns language specific options used by User::getPageRenderHash()
2058 * for example, the preferred language variant
2059 *
2060 * @return string
2061 */
2062 function getExtraHashOptions() {
2063 return $this->mConverter->getExtraHashOptions();
2064 }
2065
2066 /**
2067 * for languages that support multiple variants, the title of an
2068 * article may be displayed differently in different variants. this
2069 * function returns the apporiate title defined in the body of the article.
2070 *
2071 * @return string
2072 */
2073 function getParsedTitle() {
2074 return $this->mConverter->getParsedTitle();
2075 }
2076
2077 /**
2078 * Enclose a string with the "no conversion" tag. This is used by
2079 * various functions in the Parser
2080 *
2081 * @param $text String: text to be tagged for no conversion
2082 * @param $noParse
2083 * @return string the tagged text
2084 */
2085 function markNoConversion( $text, $noParse=false ) {
2086 return $this->mConverter->markNoConversion( $text, $noParse );
2087 }
2088
2089 /**
2090 * A regular expression to match legal word-trailing characters
2091 * which should be merged onto a link of the form [[foo]]bar.
2092 *
2093 * @return string
2094 */
2095 function linkTrail() {
2096 $this->load();
2097 return $this->linkTrail;
2098 }
2099
2100 function getLangObj() {
2101 return $this;
2102 }
2103
2104 /**
2105 * Get the RFC 3066 code for this language object
2106 */
2107 function getCode() {
2108 return $this->mCode;
2109 }
2110
2111 function setCode( $code ) {
2112 $this->mCode = $code;
2113 }
2114
2115 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2116 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2117 }
2118
2119 static function getMessagesFileName( $code ) {
2120 global $IP;
2121 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2122 }
2123
2124 static function getClassFileName( $code ) {
2125 global $IP;
2126 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2127 }
2128
2129 static function getLocalisationArray( $code, $disableCache = false ) {
2130 self::loadLocalisation( $code, $disableCache );
2131 return self::$mLocalisationCache[$code];
2132 }
2133
2134 /**
2135 * Load localisation data for a given code into the static cache
2136 *
2137 * @return array Dependencies, map of filenames to mtimes
2138 */
2139 static function loadLocalisation( $code, $disableCache = false ) {
2140 static $recursionGuard = array();
2141 global $wgMemc, $wgCheckSerialized;
2142
2143 if ( !$code ) {
2144 throw new MWException( "Invalid language code requested" );
2145 }
2146
2147 if ( !$disableCache ) {
2148 # Try the per-process cache
2149 if ( isset( self::$mLocalisationCache[$code] ) ) {
2150 return self::$mLocalisationCache[$code]['deps'];
2151 }
2152
2153 wfProfileIn( __METHOD__ );
2154
2155 # Try the serialized directory
2156 $cache = wfGetPrecompiledData( self::getFileName( "Messages", $code, '.ser' ) );
2157 if ( $cache ) {
2158 if ( $wgCheckSerialized && self::isLocalisationOutOfDate( $cache ) ) {
2159 $cache = false;
2160 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2161 } else {
2162 self::$mLocalisationCache[$code] = $cache;
2163 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2164 wfProfileOut( __METHOD__ );
2165 return self::$mLocalisationCache[$code]['deps'];
2166 }
2167 }
2168
2169 # Try the global cache
2170 $memcKey = wfMemcKey('localisation', $code );
2171 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2172 $cache = $wgMemc->get( $memcKey );
2173 if ( $cache ) {
2174 if ( self::isLocalisationOutOfDate( $cache ) ) {
2175 $wgMemc->delete( $memcKey );
2176 $wgMemc->delete( $fbMemcKey );
2177 $cache = false;
2178 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2179 } else {
2180 self::$mLocalisationCache[$code] = $cache;
2181 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2182 wfProfileOut( __METHOD__ );
2183 return $cache['deps'];
2184 }
2185 }
2186 } else {
2187 wfProfileIn( __METHOD__ );
2188 }
2189
2190 # Default fallback, may be overridden when the messages file is included
2191 if ( $code != 'en' ) {
2192 $fallback = 'en';
2193 } else {
2194 $fallback = false;
2195 }
2196
2197 # Load the primary localisation from the source file
2198 $filename = self::getMessagesFileName( $code );
2199 if ( !file_exists( $filename ) ) {
2200 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2201 $cache = compact( self::$mLocalisationKeys ); // Set correct fallback
2202 $deps = array();
2203 } else {
2204 $deps = array( $filename => filemtime( $filename ) );
2205 require( $filename );
2206 $cache = compact( self::$mLocalisationKeys );
2207 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2208 }
2209
2210 if ( !empty( $fallback ) ) {
2211 # Load the fallback localisation, with a circular reference guard
2212 if ( isset( $recursionGuard[$code] ) ) {
2213 throw new MWException( "Error: Circular fallback reference in language code $code" );
2214 }
2215 $recursionGuard[$code] = true;
2216 $newDeps = self::loadLocalisation( $fallback, $disableCache );
2217 unset( $recursionGuard[$code] );
2218
2219 $secondary = self::$mLocalisationCache[$fallback];
2220 $deps = array_merge( $deps, $newDeps );
2221
2222 # Merge the fallback localisation with the current localisation
2223 foreach ( self::$mLocalisationKeys as $key ) {
2224 if ( isset( $cache[$key] ) ) {
2225 if ( isset( $secondary[$key] ) ) {
2226 if ( in_array( $key, self::$mMergeableMapKeys ) ) {
2227 $cache[$key] = $cache[$key] + $secondary[$key];
2228 } elseif ( in_array( $key, self::$mMergeableListKeys ) ) {
2229 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2230 } elseif ( in_array( $key, self::$mMergeableAliasListKeys ) ) {
2231 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2232 }
2233 }
2234 } else {
2235 $cache[$key] = $secondary[$key];
2236 }
2237 }
2238
2239 # Merge bookstore lists if requested
2240 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2241 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2242 }
2243 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2244 unset( $cache['bookstoreList']['inherit'] );
2245 }
2246 }
2247
2248 # Add dependencies to the cache entry
2249 $cache['deps'] = $deps;
2250
2251 # Replace spaces with underscores in namespace names
2252 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2253
2254 # And do the same for specialpage aliases. $page is an array.
2255 foreach ( $cache['specialPageAliases'] as &$page ) {
2256 $page = str_replace( ' ', '_', $page );
2257 }
2258 # Decouple the reference to prevent accidental damage
2259 unset($page);
2260
2261 # Save to both caches
2262 self::$mLocalisationCache[$code] = $cache;
2263 if ( !$disableCache ) {
2264 $wgMemc->set( $memcKey, $cache );
2265 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2266 }
2267
2268 wfProfileOut( __METHOD__ );
2269 return $deps;
2270 }
2271
2272 /**
2273 * Test if a given localisation cache is out of date with respect to the
2274 * source Messages files. This is done automatically for the global cache
2275 * in $wgMemc, but is only done on certain occasions for the serialized
2276 * data file.
2277 *
2278 * @param $cache mixed Either a language code or a cache array
2279 */
2280 static function isLocalisationOutOfDate( $cache ) {
2281 if ( !is_array( $cache ) ) {
2282 self::loadLocalisation( $cache );
2283 $cache = self::$mLocalisationCache[$cache];
2284 }
2285 $expired = false;
2286 foreach ( $cache['deps'] as $file => $mtime ) {
2287 if ( !file_exists( $file ) || filemtime( $file ) > $mtime ) {
2288 $expired = true;
2289 break;
2290 }
2291 }
2292 return $expired;
2293 }
2294
2295 /**
2296 * Get the fallback for a given language
2297 */
2298 static function getFallbackFor( $code ) {
2299 // Shortcut
2300 if ( $code === 'en' ) return false;
2301
2302 // Local cache
2303 static $cache = array();
2304 // Quick return
2305 if ( isset($cache[$code]) ) return $cache[$code];
2306
2307 // Try memcache
2308 global $wgMemc;
2309 $memcKey = wfMemcKey( 'fallback', $code );
2310 $fbcode = $wgMemc->get( $memcKey );
2311
2312 if ( is_string($fbcode) ) {
2313 // False is stored as a string to detect failures in memcache properly
2314 if ( $fbcode === '' ) $fbcode = false;
2315
2316 // Update local cache and return
2317 $cache[$code] = $fbcode;
2318 return $fbcode;
2319 }
2320
2321 // Nothing in caches, load and and update both caches
2322 self::loadLocalisation( $code );
2323 $fbcode = self::$mLocalisationCache[$code]['fallback'];
2324
2325 $cache[$code] = $fbcode;
2326 $wgMemc->set( $memcKey, (string) $fbcode );
2327
2328 return $fbcode;
2329 }
2330
2331 /**
2332 * Get all messages for a given language
2333 */
2334 static function getMessagesFor( $code ) {
2335 self::loadLocalisation( $code );
2336 return self::$mLocalisationCache[$code]['messages'];
2337 }
2338
2339 /**
2340 * Get a message for a given language
2341 */
2342 static function getMessageFor( $key, $code ) {
2343 self::loadLocalisation( $code );
2344 return isset( self::$mLocalisationCache[$code]['messages'][$key] ) ? self::$mLocalisationCache[$code]['messages'][$key] : null;
2345 }
2346
2347 /**
2348 * Load localisation data for this object
2349 */
2350 function load() {
2351 if ( !$this->mLoaded ) {
2352 self::loadLocalisation( $this->getCode() );
2353 $cache =& self::$mLocalisationCache[$this->getCode()];
2354 foreach ( self::$mLocalisationKeys as $key ) {
2355 $this->$key = $cache[$key];
2356 }
2357 $this->mLoaded = true;
2358
2359 $this->fixUpSettings();
2360 }
2361 }
2362
2363 /**
2364 * Do any necessary post-cache-load settings adjustment
2365 */
2366 function fixUpSettings() {
2367 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2368 $wgNamespaceAliases, $wgAmericanDates;
2369 wfProfileIn( __METHOD__ );
2370 if ( $wgExtraNamespaces ) {
2371 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
2372 }
2373
2374 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
2375 if ( $wgMetaNamespaceTalk ) {
2376 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
2377 } else {
2378 $talk = $this->namespaceNames[NS_PROJECT_TALK];
2379 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2380
2381 # Allow grammar transformations
2382 # Allowing full message-style parsing would make simple requests
2383 # such as action=raw much more expensive than they need to be.
2384 # This will hopefully cover most cases.
2385 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2386 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2387 $talk = str_replace( ' ', '_', $talk );
2388 $this->namespaceNames[NS_PROJECT_TALK] = $talk;
2389 }
2390
2391 # The above mixing may leave namespaces out of canonical order.
2392 # Re-order by namespace ID number...
2393 ksort( $this->namespaceNames );
2394
2395 # Put namespace names and aliases into a hashtable.
2396 # If this is too slow, then we should arrange it so that it is done
2397 # before caching. The catch is that at pre-cache time, the above
2398 # class-specific fixup hasn't been done.
2399 $this->mNamespaceIds = array();
2400 foreach ( $this->namespaceNames as $index => $name ) {
2401 $this->mNamespaceIds[$this->lc($name)] = $index;
2402 }
2403 if ( $this->namespaceAliases ) {
2404 foreach ( $this->namespaceAliases as $name => $index ) {
2405 $this->mNamespaceIds[$this->lc($name)] = $index;
2406 }
2407 }
2408 if ( $wgNamespaceAliases ) {
2409 foreach ( $wgNamespaceAliases as $name => $index ) {
2410 $this->mNamespaceIds[$this->lc($name)] = $index;
2411 }
2412 }
2413
2414 if ( $this->defaultDateFormat == 'dmy or mdy' ) {
2415 $this->defaultDateFormat = $wgAmericanDates ? 'mdy' : 'dmy';
2416 }
2417 wfProfileOut( __METHOD__ );
2418 }
2419
2420 function replaceGrammarInNamespace( $m ) {
2421 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2422 }
2423
2424 static function getCaseMaps() {
2425 static $wikiUpperChars, $wikiLowerChars;
2426 if ( isset( $wikiUpperChars ) ) {
2427 return array( $wikiUpperChars, $wikiLowerChars );
2428 }
2429
2430 wfProfileIn( __METHOD__ );
2431 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2432 if ( $arr === false ) {
2433 throw new MWException(
2434 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2435 }
2436 extract( $arr );
2437 wfProfileOut( __METHOD__ );
2438 return array( $wikiUpperChars, $wikiLowerChars );
2439 }
2440
2441 function formatTimePeriod( $seconds ) {
2442 if ( $seconds < 10 ) {
2443 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2444 } elseif ( $seconds < 60 ) {
2445 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2446 } elseif ( $seconds < 3600 ) {
2447 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2448 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2449 } else {
2450 $hours = floor( $seconds / 3600 );
2451 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2452 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2453 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2454 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2455 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2456 }
2457 }
2458
2459 function formatBitrate( $bps ) {
2460 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2461 if ( $bps <= 0 ) {
2462 return $this->formatNum( $bps ) . $units[0];
2463 }
2464 $unitIndex = floor( log10( $bps ) / 3 );
2465 $mantissa = $bps / pow( 1000, $unitIndex );
2466 if ( $mantissa < 10 ) {
2467 $mantissa = round( $mantissa, 1 );
2468 } else {
2469 $mantissa = round( $mantissa );
2470 }
2471 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2472 }
2473
2474 /**
2475 * Format a size in bytes for output, using an appropriate
2476 * unit (B, KB, MB or GB) according to the magnitude in question
2477 *
2478 * @param $size Size to format
2479 * @return string Plain text (not HTML)
2480 */
2481 function formatSize( $size ) {
2482 // For small sizes no decimal places necessary
2483 $round = 0;
2484 if( $size > 1024 ) {
2485 $size = $size / 1024;
2486 if( $size > 1024 ) {
2487 $size = $size / 1024;
2488 // For MB and bigger two decimal places are smarter
2489 $round = 2;
2490 if( $size > 1024 ) {
2491 $size = $size / 1024;
2492 $msg = 'size-gigabytes';
2493 } else {
2494 $msg = 'size-megabytes';
2495 }
2496 } else {
2497 $msg = 'size-kilobytes';
2498 }
2499 } else {
2500 $msg = 'size-bytes';
2501 }
2502 $size = round( $size, $round );
2503 $text = $this->getMessageFromDB( $msg );
2504 return str_replace( '$1', $this->formatNum( $size ), $text );
2505 }
2506 }