* Rewrote wfArrayMerge() in terms of array plus
[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, $forTemplate = false) {}
43 function getExtraHashOptions() {return '';}
44 function getParsedTitle() {return '';}
45 function markNoConversion($text, $noParse=false) {return $text;}
46 function convertCategoryKey( $key ) {return $key; }
47 function convertLinkToAllVariants($text){ return array( $this->mLang->getCode() => $text); }
48 function armourMath($text){ return $text; }
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 'imageFiles'
67 );
68
69 static public $mMergeableMapKeys = array( 'messages', 'namespaceNames', 'mathNames',
70 'dateFormats', 'defaultUserOptionOverrides', 'magicWords', 'imageFiles' );
71
72 static public $mMergeableListKeys = array( 'extraUserToggles' );
73
74 static public $mMergeableAliasListKeys = array( 'specialPageAliases' );
75
76 static public $mLocalisationCache = array();
77
78 static public $mWeekdayMsgs = array(
79 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
80 'friday', 'saturday'
81 );
82
83 static public $mWeekdayAbbrevMsgs = array(
84 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
85 );
86
87 static public $mMonthMsgs = array(
88 'january', 'february', 'march', 'april', 'may_long', 'june',
89 'july', 'august', 'september', 'october', 'november',
90 'december'
91 );
92 static public $mMonthGenMsgs = array(
93 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
94 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
95 'december-gen'
96 );
97 static public $mMonthAbbrevMsgs = array(
98 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
99 'sep', 'oct', 'nov', 'dec'
100 );
101
102 static public $mIranianCalendarMonthMsgs = array(
103 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
104 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
105 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
106 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
107 );
108
109 static public $mHebrewCalendarMonthMsgs = array(
110 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
111 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
112 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
113 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
114 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
115 );
116
117 static public $mHebrewCalendarMonthGenMsgs = array(
118 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
119 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
120 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
121 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
122 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
123 );
124
125 static public $mHijriCalendarMonthMsgs = array(
126 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
127 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
128 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
129 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
130 );
131
132 /**
133 * Create a language object for a given language code
134 */
135 static function factory( $code ) {
136 global $IP;
137 static $recursionLevel = 0;
138
139 if ( $code == 'en' ) {
140 $class = 'Language';
141 } else {
142 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
143 // Preload base classes to work around APC/PHP5 bug
144 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
145 include_once("$IP/languages/classes/$class.deps.php");
146 }
147 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
148 include_once("$IP/languages/classes/$class.php");
149 }
150 }
151
152 if ( $recursionLevel > 5 ) {
153 throw new MWException( "Language fallback loop detected when creating class $class\n" );
154 }
155
156 if( ! class_exists( $class ) ) {
157 $fallback = Language::getFallbackFor( $code );
158 ++$recursionLevel;
159 $lang = Language::factory( $fallback );
160 --$recursionLevel;
161 $lang->setCode( $code );
162 } else {
163 $lang = new $class;
164 }
165
166 return $lang;
167 }
168
169 function __construct() {
170 $this->mConverter = new FakeConverter($this);
171 // Set the code to the name of the descendant
172 if ( get_class( $this ) == 'Language' ) {
173 $this->mCode = 'en';
174 } else {
175 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
176 }
177 }
178
179 /**
180 * Reduce memory usage
181 */
182 function __destruct() {
183 foreach ( $this as $name => $value ) {
184 unset( $this->$name );
185 }
186 }
187
188 /**
189 * Hook which will be called if this is the content language.
190 * Descendants can use this to register hook functions or modify globals
191 */
192 function initContLang() {}
193
194 /**
195 * @deprecated Use User::getDefaultOptions()
196 * @return array
197 */
198 function getDefaultUserOptions() {
199 wfDeprecated( __METHOD__ );
200 return User::getDefaultOptions();
201 }
202
203 function getFallbackLanguageCode() {
204 return self::getFallbackFor( $this->mCode );
205 }
206
207 /**
208 * Exports $wgBookstoreListEn
209 * @return array
210 */
211 function getBookstoreList() {
212 $this->load();
213 return $this->bookstoreList;
214 }
215
216 /**
217 * @return array
218 */
219 function getNamespaces() {
220 $this->load();
221 return $this->namespaceNames;
222 }
223
224 /**
225 * A convenience function that returns the same thing as
226 * getNamespaces() except with the array values changed to ' '
227 * where it found '_', useful for producing output to be displayed
228 * e.g. in <select> forms.
229 *
230 * @return array
231 */
232 function getFormattedNamespaces() {
233 $ns = $this->getNamespaces();
234 foreach($ns as $k => $v) {
235 $ns[$k] = strtr($v, '_', ' ');
236 }
237 return $ns;
238 }
239
240 /**
241 * Get a namespace value by key
242 * <code>
243 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
244 * echo $mw_ns; // prints 'MediaWiki'
245 * </code>
246 *
247 * @param $index Int: the array key of the namespace to return
248 * @return mixed, string if the namespace value exists, otherwise false
249 */
250 function getNsText( $index ) {
251 $ns = $this->getNamespaces();
252 return isset( $ns[$index] ) ? $ns[$index] : false;
253 }
254
255 /**
256 * A convenience function that returns the same thing as
257 * getNsText() except with '_' changed to ' ', useful for
258 * producing output.
259 *
260 * @return array
261 */
262 function getFormattedNsText( $index ) {
263 $ns = $this->getNsText( $index );
264 return strtr($ns, '_', ' ');
265 }
266
267 /**
268 * Get a namespace key by value, case insensitive.
269 * Only matches namespace names for the current language, not the
270 * canonical ones defined in Namespace.php.
271 *
272 * @param $text String
273 * @return mixed An integer if $text is a valid value otherwise false
274 */
275 function getLocalNsIndex( $text ) {
276 $this->load();
277 $lctext = $this->lc($text);
278 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
279 }
280
281 /**
282 * Get a namespace key by value, case insensitive. Canonical namespace
283 * names override custom ones defined for the current language.
284 *
285 * @param $text String
286 * @return mixed An integer if $text is a valid value otherwise false
287 */
288 function getNsIndex( $text ) {
289 $this->load();
290 $lctext = $this->lc($text);
291 if( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
292 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
293 }
294
295 /**
296 * short names for language variants used for language conversion links.
297 *
298 * @param $code String
299 * @return string
300 */
301 function getVariantname( $code ) {
302 return $this->getMessageFromDB( "variantname-$code" );
303 }
304
305 function specialPage( $name ) {
306 $aliases = $this->getSpecialPageAliases();
307 if ( isset( $aliases[$name][0] ) ) {
308 $name = $aliases[$name][0];
309 }
310 return $this->getNsText(NS_SPECIAL) . ':' . $name;
311 }
312
313 function getQuickbarSettings() {
314 return array(
315 $this->getMessage( 'qbsettings-none' ),
316 $this->getMessage( 'qbsettings-fixedleft' ),
317 $this->getMessage( 'qbsettings-fixedright' ),
318 $this->getMessage( 'qbsettings-floatingleft' ),
319 $this->getMessage( 'qbsettings-floatingright' )
320 );
321 }
322
323 function getSkinNames() {
324 $this->load();
325 return $this->skinNames;
326 }
327
328 function getMathNames() {
329 $this->load();
330 return $this->mathNames;
331 }
332
333 function getDatePreferences() {
334 $this->load();
335 return $this->datePreferences;
336 }
337
338 function getDateFormats() {
339 $this->load();
340 return $this->dateFormats;
341 }
342
343 function getDefaultDateFormat() {
344 $this->load();
345 return $this->defaultDateFormat;
346 }
347
348 function getDatePreferenceMigrationMap() {
349 $this->load();
350 return $this->datePreferenceMigrationMap;
351 }
352
353 function getImageFile( $image ) {
354 $this->load();
355 return $this->imageFiles[$image];
356 }
357
358 function getDefaultUserOptionOverrides() {
359 $this->load();
360 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
361 if (is_array($this->defaultUserOptionOverrides)) {
362 return $this->defaultUserOptionOverrides;
363 } else {
364 return array();
365 }
366 }
367
368 function getExtraUserToggles() {
369 $this->load();
370 return $this->extraUserToggles;
371 }
372
373 function getUserToggle( $tog ) {
374 return $this->getMessageFromDB( "tog-$tog" );
375 }
376
377 /**
378 * Get language names, indexed by code.
379 * If $customisedOnly is true, only returns codes with a messages file
380 */
381 public static function getLanguageNames( $customisedOnly = false ) {
382 global $wgLanguageNames, $wgExtraLanguageNames;
383 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
384 if ( !$customisedOnly ) {
385 return $allNames;
386 }
387
388 global $IP;
389 $names = array();
390 $dir = opendir( "$IP/languages/messages" );
391 while( false !== ( $file = readdir( $dir ) ) ) {
392 $m = array();
393 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
394 $code = str_replace( '_', '-', strtolower( $m[1] ) );
395 if ( isset( $allNames[$code] ) ) {
396 $names[$code] = $allNames[$code];
397 }
398 }
399 }
400 closedir( $dir );
401 return $names;
402 }
403
404 /**
405 * Get a message from the MediaWiki namespace.
406 *
407 * @param $msg String: message name
408 * @return string
409 */
410 function getMessageFromDB( $msg ) {
411 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
412 }
413
414 function getLanguageName( $code ) {
415 $names = self::getLanguageNames();
416 if ( !array_key_exists( $code, $names ) ) {
417 return '';
418 }
419 return $names[$code];
420 }
421
422 function getMonthName( $key ) {
423 return $this->getMessageFromDB( self::$mMonthMsgs[$key-1] );
424 }
425
426 function getMonthNameGen( $key ) {
427 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key-1] );
428 }
429
430 function getMonthAbbreviation( $key ) {
431 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key-1] );
432 }
433
434 function getWeekdayName( $key ) {
435 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key-1] );
436 }
437
438 function getWeekdayAbbreviation( $key ) {
439 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key-1] );
440 }
441
442 function getIranianCalendarMonthName( $key ) {
443 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key-1] );
444 }
445
446 function getHebrewCalendarMonthName( $key ) {
447 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key-1] );
448 }
449
450 function getHebrewCalendarMonthNameGen( $key ) {
451 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key-1] );
452 }
453
454 function getHijriCalendarMonthName( $key ) {
455 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key-1] );
456 }
457
458 /**
459 * Used by date() and time() to adjust the time output.
460 *
461 * @param $ts Int the time in date('YmdHis') format
462 * @param $tz Mixed: adjust the time by this amount (default false, mean we
463 * get user timecorrection setting)
464 * @return int
465 */
466 function userAdjust( $ts, $tz = false ) {
467 global $wgUser, $wgLocalTZoffset;
468
469 if (!$tz) {
470 $tz = $wgUser->getOption( 'timecorrection' );
471 }
472
473 # minutes and hours differences:
474 $minDiff = 0;
475 $hrDiff = 0;
476
477 if ( $tz === '' ) {
478 # Global offset in minutes.
479 if( isset($wgLocalTZoffset) ) {
480 if( $wgLocalTZoffset >= 0 ) {
481 $hrDiff = floor($wgLocalTZoffset / 60);
482 } else {
483 $hrDiff = ceil($wgLocalTZoffset / 60);
484 }
485 $minDiff = $wgLocalTZoffset % 60;
486 }
487 } elseif ( strpos( $tz, ':' ) !== false ) {
488 $tzArray = explode( ':', $tz );
489 $hrDiff = intval($tzArray[0]);
490 $minDiff = intval($hrDiff < 0 ? -$tzArray[1] : $tzArray[1]);
491 } else {
492 $hrDiff = intval( $tz );
493 }
494
495 # No difference ? Return time unchanged
496 if ( 0 == $hrDiff && 0 == $minDiff ) { return $ts; }
497
498 wfSuppressWarnings(); // E_STRICT system time bitching
499 # Generate an adjusted date
500 $t = mktime( (
501 (int)substr( $ts, 8, 2) ) + $hrDiff, # Hours
502 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
503 (int)substr( $ts, 12, 2 ), # Seconds
504 (int)substr( $ts, 4, 2 ), # Month
505 (int)substr( $ts, 6, 2 ), # Day
506 (int)substr( $ts, 0, 4 ) ); #Year
507
508 $date = date( 'YmdHis', $t );
509 wfRestoreWarnings();
510
511 return $date;
512 }
513
514 /**
515 * This is a workalike of PHP's date() function, but with better
516 * internationalisation, a reduced set of format characters, and a better
517 * escaping format.
518 *
519 * Supported format characters are dDjlNwzWFmMntLYyaAgGhHiscrU. See the
520 * PHP manual for definitions. There are a number of extensions, which
521 * start with "x":
522 *
523 * xn Do not translate digits of the next numeric format character
524 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
525 * xr Use roman numerals for the next numeric format character
526 * xh Use hebrew numerals for the next numeric format character
527 * xx Literal x
528 * xg Genitive month name
529 *
530 * xij j (day number) in Iranian calendar
531 * xiF F (month name) in Iranian calendar
532 * xin n (month number) in Iranian calendar
533 * xiY Y (full year) in Iranian calendar
534 *
535 * xjj j (day number) in Hebrew calendar
536 * xjF F (month name) in Hebrew calendar
537 * xjt t (days in month) in Hebrew calendar
538 * xjx xg (genitive month name) in Hebrew calendar
539 * xjn n (month number) in Hebrew calendar
540 * xjY Y (full year) in Hebrew calendar
541 *
542 * xmj j (day number) in Hijri calendar
543 * xmF F (month name) in Hijri calendar
544 * xmn n (month number) in Hijri calendar
545 * xmY Y (full year) in Hijri calendar
546 *
547 * xkY Y (full year) in Thai solar calendar. Months and days are
548 * identical to the Gregorian calendar
549 *
550 * Characters enclosed in double quotes will be considered literal (with
551 * the quotes themselves removed). Unmatched quotes will be considered
552 * literal quotes. Example:
553 *
554 * "The month is" F => The month is January
555 * i's" => 20'11"
556 *
557 * Backslash escaping is also supported.
558 *
559 * Input timestamp is assumed to be pre-normalized to the desired local
560 * time zone, if any.
561 *
562 * @param $format String
563 * @param $ts String: 14-character timestamp
564 * YYYYMMDDHHMMSS
565 * 01234567890123
566 */
567 function sprintfDate( $format, $ts ) {
568 $s = '';
569 $raw = false;
570 $roman = false;
571 $hebrewNum = false;
572 $unix = false;
573 $rawToggle = false;
574 $iranian = false;
575 $hebrew = false;
576 $hijri = false;
577 $thai = false;
578 for ( $p = 0; $p < strlen( $format ); $p++ ) {
579 $num = false;
580 $code = $format[$p];
581 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
582 $code .= $format[++$p];
583 }
584
585 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' ) && $p < strlen( $format ) - 1 ) {
586 $code .= $format[++$p];
587 }
588
589 switch ( $code ) {
590 case 'xx':
591 $s .= 'x';
592 break;
593 case 'xn':
594 $raw = true;
595 break;
596 case 'xN':
597 $rawToggle = !$rawToggle;
598 break;
599 case 'xr':
600 $roman = true;
601 break;
602 case 'xh':
603 $hebrewNum = true;
604 break;
605 case 'xg':
606 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
607 break;
608 case 'xjx':
609 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
610 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
611 break;
612 case 'd':
613 $num = substr( $ts, 6, 2 );
614 break;
615 case 'D':
616 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
617 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
618 break;
619 case 'j':
620 $num = intval( substr( $ts, 6, 2 ) );
621 break;
622 case 'xij':
623 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
624 $num = $iranian[2];
625 break;
626 case 'xmj':
627 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
628 $num = $hijri[2];
629 break;
630 case 'xjj':
631 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
632 $num = $hebrew[2];
633 break;
634 case 'l':
635 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
636 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
637 break;
638 case 'N':
639 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
640 $w = gmdate( 'w', $unix );
641 $num = $w ? $w : 7;
642 break;
643 case 'w':
644 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
645 $num = gmdate( 'w', $unix );
646 break;
647 case 'z':
648 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
649 $num = gmdate( 'z', $unix );
650 break;
651 case 'W':
652 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
653 $num = gmdate( 'W', $unix );
654 break;
655 case 'F':
656 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
657 break;
658 case 'xiF':
659 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
660 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
661 break;
662 case 'xmF':
663 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
664 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
665 break;
666 case 'xjF':
667 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
668 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
669 break;
670 case 'm':
671 $num = substr( $ts, 4, 2 );
672 break;
673 case 'M':
674 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
675 break;
676 case 'n':
677 $num = intval( substr( $ts, 4, 2 ) );
678 break;
679 case 'xin':
680 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
681 $num = $iranian[1];
682 break;
683 case 'xmn':
684 if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
685 $num = $hijri[1];
686 break;
687 case 'xjn':
688 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
689 $num = $hebrew[1];
690 break;
691 case 't':
692 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
693 $num = gmdate( 't', $unix );
694 break;
695 case 'xjt':
696 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
697 $num = $hebrew[3];
698 break;
699 case 'L':
700 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
701 $num = gmdate( 'L', $unix );
702 break;
703 case 'Y':
704 $num = substr( $ts, 0, 4 );
705 break;
706 case 'xiY':
707 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
708 $num = $iranian[0];
709 break;
710 case 'xmY':
711 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
712 $num = $hijri[0];
713 break;
714 case 'xjY':
715 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
716 $num = $hebrew[0];
717 break;
718 case 'xkY':
719 if ( !$thai ) $thai = self::tsToThai( $ts );
720 $num = $thai[0];
721 break;
722 case 'y':
723 $num = substr( $ts, 2, 2 );
724 break;
725 case 'a':
726 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
727 break;
728 case 'A':
729 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
730 break;
731 case 'g':
732 $h = substr( $ts, 8, 2 );
733 $num = $h % 12 ? $h % 12 : 12;
734 break;
735 case 'G':
736 $num = intval( substr( $ts, 8, 2 ) );
737 break;
738 case 'h':
739 $h = substr( $ts, 8, 2 );
740 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
741 break;
742 case 'H':
743 $num = substr( $ts, 8, 2 );
744 break;
745 case 'i':
746 $num = substr( $ts, 10, 2 );
747 break;
748 case 's':
749 $num = substr( $ts, 12, 2 );
750 break;
751 case 'c':
752 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
753 $s .= gmdate( 'c', $unix );
754 break;
755 case 'r':
756 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
757 $s .= gmdate( 'r', $unix );
758 break;
759 case 'U':
760 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
761 $num = $unix;
762 break;
763 case '\\':
764 # Backslash escaping
765 if ( $p < strlen( $format ) - 1 ) {
766 $s .= $format[++$p];
767 } else {
768 $s .= '\\';
769 }
770 break;
771 case '"':
772 # Quoted literal
773 if ( $p < strlen( $format ) - 1 ) {
774 $endQuote = strpos( $format, '"', $p + 1 );
775 if ( $endQuote === false ) {
776 # No terminating quote, assume literal "
777 $s .= '"';
778 } else {
779 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
780 $p = $endQuote;
781 }
782 } else {
783 # Quote at end of string, assume literal "
784 $s .= '"';
785 }
786 break;
787 default:
788 $s .= $format[$p];
789 }
790 if ( $num !== false ) {
791 if ( $rawToggle || $raw ) {
792 $s .= $num;
793 $raw = false;
794 } elseif ( $roman ) {
795 $s .= self::romanNumeral( $num );
796 $roman = false;
797 } elseif( $hebrewNum ) {
798 $s .= self::hebrewNumeral( $num );
799 $hebrewNum = false;
800 } else {
801 $s .= $this->formatNum( $num, true );
802 }
803 $num = false;
804 }
805 }
806 return $s;
807 }
808
809 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
810 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
811 /**
812 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
813 * Gregorian dates to Iranian dates. Originally written in C, it
814 * is released under the terms of GNU Lesser General Public
815 * License. Conversion to PHP was performed by Niklas Laxström.
816 *
817 * Link: http://www.farsiweb.info/jalali/jalali.c
818 */
819 private static function tsToIranian( $ts ) {
820 $gy = substr( $ts, 0, 4 ) -1600;
821 $gm = substr( $ts, 4, 2 ) -1;
822 $gd = substr( $ts, 6, 2 ) -1;
823
824 # Days passed from the beginning (including leap years)
825 $gDayNo = 365*$gy
826 + floor(($gy+3) / 4)
827 - floor(($gy+99) / 100)
828 + floor(($gy+399) / 400);
829
830
831 // Add days of the past months of this year
832 for( $i = 0; $i < $gm; $i++ ) {
833 $gDayNo += self::$GREG_DAYS[$i];
834 }
835
836 // Leap years
837 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
838 $gDayNo++;
839 }
840
841 // Days passed in current month
842 $gDayNo += $gd;
843
844 $jDayNo = $gDayNo - 79;
845
846 $jNp = floor($jDayNo / 12053);
847 $jDayNo %= 12053;
848
849 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
850 $jDayNo %= 1461;
851
852 if ( $jDayNo >= 366 ) {
853 $jy += floor(($jDayNo-1)/365);
854 $jDayNo = floor(($jDayNo-1)%365);
855 }
856
857 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
858 $jDayNo -= self::$IRANIAN_DAYS[$i];
859 }
860
861 $jm= $i+1;
862 $jd= $jDayNo+1;
863
864 return array($jy, $jm, $jd);
865 }
866 /**
867 * Converting Gregorian dates to Hijri dates.
868 *
869 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
870 *
871 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
872 */
873 private static function tsToHijri ( $ts ) {
874 $year = substr( $ts, 0, 4 );
875 $month = substr( $ts, 4, 2 );
876 $day = substr( $ts, 6, 2 );
877
878 $zyr = $year;
879 $zd=$day;
880 $zm=$month;
881 $zy=$zyr;
882
883
884
885 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
886 {
887
888
889 $zjd=(int)((1461*($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;
890 }
891 else
892 {
893 $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
894 }
895
896 $zl=$zjd-1948440+10632;
897 $zn=(int)(($zl-1)/10631);
898 $zl=$zl-10631*$zn+354;
899 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
900 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
901 $zm=(int)((24*$zl)/709);
902 $zd=$zl-(int)((709*$zm)/24);
903 $zy=30*$zn+$zj-30;
904
905 return array ($zy, $zm, $zd);
906 }
907
908 /**
909 * Converting Gregorian dates to Hebrew dates.
910 *
911 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
912 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
913 * to translate the relevant functions into PHP and release them under
914 * GNU GPL.
915 *
916 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
917 * and Adar II is 14. In a non-leap year, Adar is 6.
918 */
919 private static function tsToHebrew( $ts ) {
920 # Parse date
921 $year = substr( $ts, 0, 4 );
922 $month = substr( $ts, 4, 2 );
923 $day = substr( $ts, 6, 2 );
924
925 # Calculate Hebrew year
926 $hebrewYear = $year + 3760;
927
928 # Month number when September = 1, August = 12
929 $month += 4;
930 if( $month > 12 ) {
931 # Next year
932 $month -= 12;
933 $year++;
934 $hebrewYear++;
935 }
936
937 # Calculate day of year from 1 September
938 $dayOfYear = $day;
939 for( $i = 1; $i < $month; $i++ ) {
940 if( $i == 6 ) {
941 # February
942 $dayOfYear += 28;
943 # Check if the year is leap
944 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
945 $dayOfYear++;
946 }
947 } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
948 $dayOfYear += 30;
949 } else {
950 $dayOfYear += 31;
951 }
952 }
953
954 # Calculate the start of the Hebrew year
955 $start = self::hebrewYearStart( $hebrewYear );
956
957 # Calculate next year's start
958 if( $dayOfYear <= $start ) {
959 # Day is before the start of the year - it is the previous year
960 # Next year's start
961 $nextStart = $start;
962 # Previous year
963 $year--;
964 $hebrewYear--;
965 # Add days since previous year's 1 September
966 $dayOfYear += 365;
967 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
968 # Leap year
969 $dayOfYear++;
970 }
971 # Start of the new (previous) year
972 $start = self::hebrewYearStart( $hebrewYear );
973 } else {
974 # Next year's start
975 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
976 }
977
978 # Calculate Hebrew day of year
979 $hebrewDayOfYear = $dayOfYear - $start;
980
981 # Difference between year's days
982 $diff = $nextStart - $start;
983 # Add 12 (or 13 for leap years) days to ignore the difference between
984 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
985 # difference is only about the year type
986 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
987 $diff += 13;
988 } else {
989 $diff += 12;
990 }
991
992 # Check the year pattern, and is leap year
993 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
994 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
995 # and non-leap years
996 $yearPattern = $diff % 30;
997 # Check if leap year
998 $isLeap = $diff >= 30;
999
1000 # Calculate day in the month from number of day in the Hebrew year
1001 # Don't check Adar - if the day is not in Adar, we will stop before;
1002 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1003 $hebrewDay = $hebrewDayOfYear;
1004 $hebrewMonth = 1;
1005 $days = 0;
1006 while( $hebrewMonth <= 12 ) {
1007 # Calculate days in this month
1008 if( $isLeap && $hebrewMonth == 6 ) {
1009 # Adar in a leap year
1010 if( $isLeap ) {
1011 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1012 $days = 30;
1013 if( $hebrewDay <= $days ) {
1014 # Day in Adar I
1015 $hebrewMonth = 13;
1016 } else {
1017 # Subtract the days of Adar I
1018 $hebrewDay -= $days;
1019 # Try Adar II
1020 $days = 29;
1021 if( $hebrewDay <= $days ) {
1022 # Day in Adar II
1023 $hebrewMonth = 14;
1024 }
1025 }
1026 }
1027 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1028 # Cheshvan in a complete year (otherwise as the rule below)
1029 $days = 30;
1030 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1031 # Kislev in an incomplete year (otherwise as the rule below)
1032 $days = 29;
1033 } else {
1034 # Odd months have 30 days, even have 29
1035 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1036 }
1037 if( $hebrewDay <= $days ) {
1038 # In the current month
1039 break;
1040 } else {
1041 # Subtract the days of the current month
1042 $hebrewDay -= $days;
1043 # Try in the next month
1044 $hebrewMonth++;
1045 }
1046 }
1047
1048 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1049 }
1050
1051 /**
1052 * This calculates the Hebrew year start, as days since 1 September.
1053 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1054 * Used for Hebrew date.
1055 */
1056 private static function hebrewYearStart( $year ) {
1057 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1058 $b = intval( ( $year - 1 ) % 4 );
1059 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1060 if( $m < 0 ) {
1061 $m--;
1062 }
1063 $Mar = intval( $m );
1064 if( $m < 0 ) {
1065 $m++;
1066 }
1067 $m -= $Mar;
1068
1069 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1070 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1071 $Mar++;
1072 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1073 $Mar += 2;
1074 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1075 $Mar++;
1076 }
1077
1078 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1079 return $Mar;
1080 }
1081
1082 /**
1083 * Algorithm to convert Gregorian dates to Thai solar dates.
1084 *
1085 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1086 *
1087 * @param $ts String: 14-character timestamp
1088 * @return array converted year, month, day
1089 */
1090 private static function tsToThai( $ts ) {
1091 $gy = substr( $ts, 0, 4 );
1092 $gm = substr( $ts, 4, 2 );
1093 $gd = substr( $ts, 6, 2 );
1094
1095 # Add 543 years to the Gregorian calendar
1096 # Months and days are identical
1097 $gy_thai = $gy + 543;
1098
1099 return array( $gy_thai, $gm, $gd );
1100 }
1101
1102
1103 /**
1104 * Roman number formatting up to 3000
1105 */
1106 static function romanNumeral( $num ) {
1107 static $table = array(
1108 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1109 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1110 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1111 array( '', 'M', 'MM', 'MMM' )
1112 );
1113
1114 $num = intval( $num );
1115 if ( $num > 3000 || $num <= 0 ) {
1116 return $num;
1117 }
1118
1119 $s = '';
1120 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1121 if ( $num >= $pow10 ) {
1122 $s .= $table[$i][floor($num / $pow10)];
1123 }
1124 $num = $num % $pow10;
1125 }
1126 return $s;
1127 }
1128
1129 /**
1130 * Hebrew Gematria number formatting up to 9999
1131 */
1132 static function hebrewNumeral( $num ) {
1133 static $table = array(
1134 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1135 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1136 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1137 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1138 );
1139
1140 $num = intval( $num );
1141 if ( $num > 9999 || $num <= 0 ) {
1142 return $num;
1143 }
1144
1145 $s = '';
1146 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1147 if ( $num >= $pow10 ) {
1148 if ( $num == 15 || $num == 16 ) {
1149 $s .= $table[0][9] . $table[0][$num - 9];
1150 $num = 0;
1151 } else {
1152 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1153 if( $pow10 == 1000 ) {
1154 $s .= "'";
1155 }
1156 }
1157 }
1158 $num = $num % $pow10;
1159 }
1160 if( strlen( $s ) == 2 ) {
1161 $str = $s . "'";
1162 } else {
1163 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1164 $str .= substr( $s, strlen( $s ) - 2, 2 );
1165 }
1166 $start = substr( $str, 0, strlen( $str ) - 2 );
1167 $end = substr( $str, strlen( $str ) - 2 );
1168 switch( $end ) {
1169 case 'כ':
1170 $str = $start . 'ך';
1171 break;
1172 case 'מ':
1173 $str = $start . 'ם';
1174 break;
1175 case 'נ':
1176 $str = $start . 'ן';
1177 break;
1178 case 'פ':
1179 $str = $start . 'ף';
1180 break;
1181 case 'צ':
1182 $str = $start . 'ץ';
1183 break;
1184 }
1185 return $str;
1186 }
1187
1188 /**
1189 * This is meant to be used by time(), date(), and timeanddate() to get
1190 * the date preference they're supposed to use, it should be used in
1191 * all children.
1192 *
1193 *<code>
1194 * function timeanddate([...], $format = true) {
1195 * $datePreference = $this->dateFormat($format);
1196 * [...]
1197 * }
1198 *</code>
1199 *
1200 * @param $usePrefs Mixed: if true, the user's preference is used
1201 * if false, the site/language default is used
1202 * if int/string, assumed to be a format.
1203 * @return string
1204 */
1205 function dateFormat( $usePrefs = true ) {
1206 global $wgUser;
1207
1208 if( is_bool( $usePrefs ) ) {
1209 if( $usePrefs ) {
1210 $datePreference = $wgUser->getDatePreference();
1211 } else {
1212 $options = User::getDefaultOptions();
1213 $datePreference = (string)$options['date'];
1214 }
1215 } else {
1216 $datePreference = (string)$usePrefs;
1217 }
1218
1219 // return int
1220 if( $datePreference == '' ) {
1221 return 'default';
1222 }
1223
1224 return $datePreference;
1225 }
1226
1227 /**
1228 * @param $ts Mixed: the time format which needs to be turned into a
1229 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1230 * @param $adj Bool: whether to adjust the time output according to the
1231 * user configured offset ($timecorrection)
1232 * @param $format Mixed: true to use user's date format preference
1233 * @param $timecorrection String: the time offset as returned by
1234 * validateTimeZone() in Special:Preferences
1235 * @return string
1236 */
1237 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1238 $this->load();
1239 if ( $adj ) {
1240 $ts = $this->userAdjust( $ts, $timecorrection );
1241 }
1242
1243 $pref = $this->dateFormat( $format );
1244 if( $pref == 'default' || !isset( $this->dateFormats["$pref date"] ) ) {
1245 $pref = $this->defaultDateFormat;
1246 }
1247 return $this->sprintfDate( $this->dateFormats["$pref date"], $ts );
1248 }
1249
1250 /**
1251 * @param $ts Mixed: the time format which needs to be turned into a
1252 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1253 * @param $adj Bool: whether to adjust the time output according to the
1254 * user configured offset ($timecorrection)
1255 * @param $format Mixed: true to use user's date format preference
1256 * @param $timecorrection String: the time offset as returned by
1257 * validateTimeZone() in Special:Preferences
1258 * @return string
1259 */
1260 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1261 $this->load();
1262 if ( $adj ) {
1263 $ts = $this->userAdjust( $ts, $timecorrection );
1264 }
1265
1266 $pref = $this->dateFormat( $format );
1267 if( $pref == 'default' || !isset( $this->dateFormats["$pref time"] ) ) {
1268 $pref = $this->defaultDateFormat;
1269 }
1270 return $this->sprintfDate( $this->dateFormats["$pref time"], $ts );
1271 }
1272
1273 /**
1274 * @param $ts Mixed: the time format which needs to be turned into a
1275 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1276 * @param $adj Bool: whether to adjust the time output according to the
1277 * user configured offset ($timecorrection)
1278 * @param $format Mixed: what format to return, if it's false output the
1279 * default one (default true)
1280 * @param $timecorrection String: the time offset as returned by
1281 * validateTimeZone() in Special:Preferences
1282 * @return string
1283 */
1284 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1285 $this->load();
1286
1287 $ts = wfTimestamp( TS_MW, $ts );
1288
1289 if ( $adj ) {
1290 $ts = $this->userAdjust( $ts, $timecorrection );
1291 }
1292
1293 $pref = $this->dateFormat( $format );
1294 if( $pref == 'default' || !isset( $this->dateFormats["$pref both"] ) ) {
1295 $pref = $this->defaultDateFormat;
1296 }
1297
1298 return $this->sprintfDate( $this->dateFormats["$pref both"], $ts );
1299 }
1300
1301 function getMessage( $key ) {
1302 $this->load();
1303 return isset( $this->messages[$key] ) ? $this->messages[$key] : null;
1304 }
1305
1306 function getAllMessages() {
1307 $this->load();
1308 return $this->messages;
1309 }
1310
1311 function iconv( $in, $out, $string ) {
1312 # For most languages, this is a wrapper for iconv
1313 return iconv( $in, $out . '//IGNORE', $string );
1314 }
1315
1316 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1317 function ucwordbreaksCallbackAscii($matches){
1318 return $this->ucfirst($matches[1]);
1319 }
1320
1321 function ucwordbreaksCallbackMB($matches){
1322 return mb_strtoupper($matches[0]);
1323 }
1324
1325 function ucCallback($matches){
1326 list( $wikiUpperChars ) = self::getCaseMaps();
1327 return strtr( $matches[1], $wikiUpperChars );
1328 }
1329
1330 function lcCallback($matches){
1331 list( , $wikiLowerChars ) = self::getCaseMaps();
1332 return strtr( $matches[1], $wikiLowerChars );
1333 }
1334
1335 function ucwordsCallbackMB($matches){
1336 return mb_strtoupper($matches[0]);
1337 }
1338
1339 function ucwordsCallbackWiki($matches){
1340 list( $wikiUpperChars ) = self::getCaseMaps();
1341 return strtr( $matches[0], $wikiUpperChars );
1342 }
1343
1344 function ucfirst( $str ) {
1345 if ( empty($str) ) return $str;
1346 if ( ord($str[0]) < 128 ) return ucfirst($str);
1347 else return self::uc($str,true); // fall back to more complex logic in case of multibyte strings
1348 }
1349
1350 function uc( $str, $first = false ) {
1351 if ( function_exists( 'mb_strtoupper' ) ) {
1352 if ( $first ) {
1353 if ( self::isMultibyte( $str ) ) {
1354 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1355 } else {
1356 return ucfirst( $str );
1357 }
1358 } else {
1359 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1360 }
1361 } else {
1362 if ( self::isMultibyte( $str ) ) {
1363 list( $wikiUpperChars ) = $this->getCaseMaps();
1364 $x = $first ? '^' : '';
1365 return preg_replace_callback(
1366 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1367 array($this,"ucCallback"),
1368 $str
1369 );
1370 } else {
1371 return $first ? ucfirst( $str ) : strtoupper( $str );
1372 }
1373 }
1374 }
1375
1376 function lcfirst( $str ) {
1377 if ( empty($str) ) return $str;
1378 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1379 // editing string in place = cool
1380 $str[0]=strtolower($str[0]);
1381 return $str;
1382 }
1383 else return self::lc( $str, true );
1384 }
1385
1386 function lc( $str, $first = false ) {
1387 if ( function_exists( 'mb_strtolower' ) )
1388 if ( $first )
1389 if ( self::isMultibyte( $str ) )
1390 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1391 else
1392 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1393 else
1394 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1395 else
1396 if ( self::isMultibyte( $str ) ) {
1397 list( , $wikiLowerChars ) = self::getCaseMaps();
1398 $x = $first ? '^' : '';
1399 return preg_replace_callback(
1400 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1401 array($this,"lcCallback"),
1402 $str
1403 );
1404 } else
1405 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1406 }
1407
1408 function isMultibyte( $str ) {
1409 return (bool)preg_match( '/[\x80-\xff]/', $str );
1410 }
1411
1412 function ucwords($str) {
1413 if ( self::isMultibyte( $str ) ) {
1414 $str = self::lc($str);
1415
1416 // regexp to find first letter in each word (i.e. after each space)
1417 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1418
1419 // function to use to capitalize a single char
1420 if ( function_exists( 'mb_strtoupper' ) )
1421 return preg_replace_callback(
1422 $replaceRegexp,
1423 array($this,"ucwordsCallbackMB"),
1424 $str
1425 );
1426 else
1427 return preg_replace_callback(
1428 $replaceRegexp,
1429 array($this,"ucwordsCallbackWiki"),
1430 $str
1431 );
1432 }
1433 else
1434 return ucwords( strtolower( $str ) );
1435 }
1436
1437 # capitalize words at word breaks
1438 function ucwordbreaks($str){
1439 if (self::isMultibyte( $str ) ) {
1440 $str = self::lc($str);
1441
1442 // since \b doesn't work for UTF-8, we explicitely define word break chars
1443 $breaks= "[ \-\(\)\}\{\.,\?!]";
1444
1445 // find first letter after word break
1446 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1447
1448 if ( function_exists( 'mb_strtoupper' ) )
1449 return preg_replace_callback(
1450 $replaceRegexp,
1451 array($this,"ucwordbreaksCallbackMB"),
1452 $str
1453 );
1454 else
1455 return preg_replace_callback(
1456 $replaceRegexp,
1457 array($this,"ucwordsCallbackWiki"),
1458 $str
1459 );
1460 }
1461 else
1462 return preg_replace_callback(
1463 '/\b([\w\x80-\xff]+)\b/',
1464 array($this,"ucwordbreaksCallbackAscii"),
1465 $str );
1466 }
1467
1468 /**
1469 * Return a case-folded representation of $s
1470 *
1471 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1472 * and $s2 are the same except for the case of their characters. It is not
1473 * necessary for the value returned to make sense when displayed.
1474 *
1475 * Do *not* perform any other normalisation in this function. If a caller
1476 * uses this function when it should be using a more general normalisation
1477 * function, then fix the caller.
1478 */
1479 function caseFold( $s ) {
1480 return $this->uc( $s );
1481 }
1482
1483 function checkTitleEncoding( $s ) {
1484 if( is_array( $s ) ) {
1485 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1486 }
1487 # Check for non-UTF-8 URLs
1488 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1489 if(!$ishigh) return $s;
1490
1491 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1492 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1493 if( $isutf8 ) return $s;
1494
1495 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1496 }
1497
1498 function fallback8bitEncoding() {
1499 $this->load();
1500 return $this->fallback8bitEncoding;
1501 }
1502
1503 /**
1504 * Some languages have special punctuation to strip out
1505 * or characters which need to be converted for MySQL's
1506 * indexing to grok it correctly. Make such changes here.
1507 *
1508 * @param $string String
1509 * @return String
1510 */
1511 function stripForSearch( $string ) {
1512 global $wgDBtype;
1513 if ( $wgDBtype != 'mysql' ) {
1514 return $string;
1515 }
1516
1517 # MySQL fulltext index doesn't grok utf-8, so we
1518 # need to fold cases and convert to hex
1519
1520 wfProfileIn( __METHOD__ );
1521 if( function_exists( 'mb_strtolower' ) ) {
1522 $out = preg_replace(
1523 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1524 "'U8' . bin2hex( \"$1\" )",
1525 mb_strtolower( $string ) );
1526 } else {
1527 list( , $wikiLowerChars ) = self::getCaseMaps();
1528 $out = preg_replace(
1529 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1530 "'U8' . bin2hex( strtr( \"\$1\", \$wikiLowerChars ) )",
1531 $string );
1532 }
1533 wfProfileOut( __METHOD__ );
1534 return $out;
1535 }
1536
1537 function convertForSearchResult( $termsArray ) {
1538 # some languages, e.g. Chinese, need to do a conversion
1539 # in order for search results to be displayed correctly
1540 return $termsArray;
1541 }
1542
1543 /**
1544 * Get the first character of a string.
1545 *
1546 * @param $s string
1547 * @return string
1548 */
1549 function firstChar( $s ) {
1550 $matches = array();
1551 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1552 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1553
1554 if ( isset( $matches[1] ) ) {
1555 if ( strlen( $matches[1] ) != 3 ) {
1556 return $matches[1];
1557 }
1558
1559 // Break down Hangul syllables to grab the first jamo
1560 $code = utf8ToCodepoint( $matches[1] );
1561 if ( $code < 0xac00 || 0xd7a4 <= $code) {
1562 return $matches[1];
1563 } elseif ( $code < 0xb098 ) {
1564 return "\xe3\x84\xb1";
1565 } elseif ( $code < 0xb2e4 ) {
1566 return "\xe3\x84\xb4";
1567 } elseif ( $code < 0xb77c ) {
1568 return "\xe3\x84\xb7";
1569 } elseif ( $code < 0xb9c8 ) {
1570 return "\xe3\x84\xb9";
1571 } elseif ( $code < 0xbc14 ) {
1572 return "\xe3\x85\x81";
1573 } elseif ( $code < 0xc0ac ) {
1574 return "\xe3\x85\x82";
1575 } elseif ( $code < 0xc544 ) {
1576 return "\xe3\x85\x85";
1577 } elseif ( $code < 0xc790 ) {
1578 return "\xe3\x85\x87";
1579 } elseif ( $code < 0xcc28 ) {
1580 return "\xe3\x85\x88";
1581 } elseif ( $code < 0xce74 ) {
1582 return "\xe3\x85\x8a";
1583 } elseif ( $code < 0xd0c0 ) {
1584 return "\xe3\x85\x8b";
1585 } elseif ( $code < 0xd30c ) {
1586 return "\xe3\x85\x8c";
1587 } elseif ( $code < 0xd558 ) {
1588 return "\xe3\x85\x8d";
1589 } else {
1590 return "\xe3\x85\x8e";
1591 }
1592 } else {
1593 return "";
1594 }
1595 }
1596
1597 function initEncoding() {
1598 # Some languages may have an alternate char encoding option
1599 # (Esperanto X-coding, Japanese furigana conversion, etc)
1600 # If this language is used as the primary content language,
1601 # an override to the defaults can be set here on startup.
1602 }
1603
1604 function recodeForEdit( $s ) {
1605 # For some languages we'll want to explicitly specify
1606 # which characters make it into the edit box raw
1607 # or are converted in some way or another.
1608 # Note that if wgOutputEncoding is different from
1609 # wgInputEncoding, this text will be further converted
1610 # to wgOutputEncoding.
1611 global $wgEditEncoding;
1612 if( $wgEditEncoding == '' or
1613 $wgEditEncoding == 'UTF-8' ) {
1614 return $s;
1615 } else {
1616 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1617 }
1618 }
1619
1620 function recodeInput( $s ) {
1621 # Take the previous into account.
1622 global $wgEditEncoding;
1623 if($wgEditEncoding != "") {
1624 $enc = $wgEditEncoding;
1625 } else {
1626 $enc = 'UTF-8';
1627 }
1628 if( $enc == 'UTF-8' ) {
1629 return $s;
1630 } else {
1631 return $this->iconv( $enc, 'UTF-8', $s );
1632 }
1633 }
1634
1635 /**
1636 * For right-to-left language support
1637 *
1638 * @return bool
1639 */
1640 function isRTL() {
1641 $this->load();
1642 return $this->rtl;
1643 }
1644
1645 /**
1646 * A hidden direction mark (LRM or RLM), depending on the language direction
1647 *
1648 * @return string
1649 */
1650 function getDirMark() {
1651 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1652 }
1653
1654 /**
1655 * An arrow, depending on the language direction
1656 *
1657 * @return string
1658 */
1659 function getArrow() {
1660 return $this->isRTL() ? '←' : '→';
1661 }
1662
1663 /**
1664 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1665 *
1666 * @return bool
1667 */
1668 function linkPrefixExtension() {
1669 $this->load();
1670 return $this->linkPrefixExtension;
1671 }
1672
1673 function &getMagicWords() {
1674 $this->load();
1675 return $this->magicWords;
1676 }
1677
1678 # Fill a MagicWord object with data from here
1679 function getMagic( &$mw ) {
1680 if ( !$this->mMagicHookDone ) {
1681 $this->mMagicHookDone = true;
1682 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1683 }
1684 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1685 $rawEntry = $this->mMagicExtensions[$mw->mId];
1686 } else {
1687 $magicWords =& $this->getMagicWords();
1688 if ( isset( $magicWords[$mw->mId] ) ) {
1689 $rawEntry = $magicWords[$mw->mId];
1690 } else {
1691 # Fall back to English if local list is incomplete
1692 $magicWords =& Language::getMagicWords();
1693 if ( !isset($magicWords[$mw->mId]) ) {
1694 throw new MWException("Magic word '{$mw->mId}' not found" );
1695 }
1696 $rawEntry = $magicWords[$mw->mId];
1697 }
1698 }
1699
1700 if( !is_array( $rawEntry ) ) {
1701 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1702 } else {
1703 $mw->mCaseSensitive = $rawEntry[0];
1704 $mw->mSynonyms = array_slice( $rawEntry, 1 );
1705 }
1706 }
1707
1708 /**
1709 * Add magic words to the extension array
1710 */
1711 function addMagicWordsByLang( $newWords ) {
1712 $code = $this->getCode();
1713 $fallbackChain = array();
1714 while ( $code && !in_array( $code, $fallbackChain ) ) {
1715 $fallbackChain[] = $code;
1716 $code = self::getFallbackFor( $code );
1717 }
1718 if ( !in_array( 'en', $fallbackChain ) ) {
1719 $fallbackChain[] = 'en';
1720 }
1721 $fallbackChain = array_reverse( $fallbackChain );
1722 foreach ( $fallbackChain as $code ) {
1723 if ( isset( $newWords[$code] ) ) {
1724 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1725 }
1726 }
1727 }
1728
1729 /**
1730 * Get special page names, as an associative array
1731 * case folded alias => real name
1732 */
1733 function getSpecialPageAliases() {
1734 $this->load();
1735
1736 // Cache aliases because it may be slow to load them
1737 if ( !isset( $this->mExtendedSpecialPageAliases ) ) {
1738
1739 // Initialise array
1740 $this->mExtendedSpecialPageAliases = $this->specialPageAliases;
1741
1742 global $wgExtensionAliasesFiles;
1743 foreach ( $wgExtensionAliasesFiles as $file ) {
1744
1745 // Fail fast
1746 if ( !file_exists($file) )
1747 throw new MWException( "Aliases file does not exist: $file" );
1748
1749 $aliases = array();
1750 require($file);
1751
1752 // Check the availability of aliases
1753 if ( !isset($aliases['en']) )
1754 throw new MWException( "Malformed aliases file: $file" );
1755
1756 // Merge all aliases in fallback chain
1757 $code = $this->getCode();
1758 do {
1759 if ( !isset($aliases[$code]) ) continue;
1760
1761 $aliases[$code] = $this->fixSpecialPageAliases( $aliases[$code] );
1762 /* Merge the aliases, THIS will break if there is special page name
1763 * which looks like a numerical key, thanks to PHP...
1764 * See the array_merge_recursive manual entry */
1765 $this->mExtendedSpecialPageAliases = array_merge_recursive(
1766 $this->mExtendedSpecialPageAliases, $aliases[$code] );
1767
1768 } while ( $code = self::getFallbackFor( $code ) );
1769 }
1770
1771 wfRunHooks( 'LanguageGetSpecialPageAliases',
1772 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1773 }
1774
1775 return $this->mExtendedSpecialPageAliases;
1776 }
1777
1778 /**
1779 * Function to fix special page aliases. Will convert the first letter to
1780 * upper case and spaces to underscores. Can be given a full aliases array,
1781 * in which case it will recursively fix all aliases.
1782 */
1783 public function fixSpecialPageAliases( $mixed ) {
1784 // Work recursively until in string level
1785 if ( is_array($mixed) ) {
1786 $callback = array( $this, 'fixSpecialPageAliases' );
1787 return array_map( $callback, $mixed );
1788 }
1789 return str_replace( ' ', '_', $this->ucfirst( $mixed ) );
1790 }
1791
1792 /**
1793 * Italic is unsuitable for some languages
1794 *
1795 * @param $text String: the text to be emphasized.
1796 * @return string
1797 */
1798 function emphasize( $text ) {
1799 return "<em>$text</em>";
1800 }
1801
1802 /**
1803 * Normally we output all numbers in plain en_US style, that is
1804 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1805 * point twohundredthirtyfive. However this is not sutable for all
1806 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1807 * Icelandic just want to use commas instead of dots, and dots instead
1808 * of commas like "293.291,235".
1809 *
1810 * An example of this function being called:
1811 * <code>
1812 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1813 * </code>
1814 *
1815 * See LanguageGu.php for the Gujarati implementation and
1816 * $separatorTransformTable on MessageIs.php for
1817 * the , => . and . => , implementation.
1818 *
1819 * @todo check if it's viable to use localeconv() for the decimal
1820 * separator thing.
1821 * @param $number Mixed: the string to be formatted, should be an integer
1822 * or a floating point number.
1823 * @param $nocommafy Bool: set to true for special numbers like dates
1824 * @return string
1825 */
1826 function formatNum( $number, $nocommafy = false ) {
1827 global $wgTranslateNumerals;
1828 if (!$nocommafy) {
1829 $number = $this->commafy($number);
1830 $s = $this->separatorTransformTable();
1831 if (!is_null($s)) { $number = strtr($number, $s); }
1832 }
1833
1834 if ($wgTranslateNumerals) {
1835 $s = $this->digitTransformTable();
1836 if (!is_null($s)) { $number = strtr($number, $s); }
1837 }
1838
1839 return $number;
1840 }
1841
1842 function parseFormattedNumber( $number ) {
1843 $s = $this->digitTransformTable();
1844 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1845
1846 $s = $this->separatorTransformTable();
1847 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1848
1849 $number = strtr( $number, array (',' => '') );
1850 return $number;
1851 }
1852
1853 /**
1854 * Adds commas to a given number
1855 *
1856 * @param $_ mixed
1857 * @return string
1858 */
1859 function commafy($_) {
1860 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1861 }
1862
1863 function digitTransformTable() {
1864 $this->load();
1865 return $this->digitTransformTable;
1866 }
1867
1868 function separatorTransformTable() {
1869 $this->load();
1870 return $this->separatorTransformTable;
1871 }
1872
1873
1874 /**
1875 * For the credit list in includes/Credits.php (action=credits)
1876 *
1877 * @param $l Array
1878 * @return string
1879 */
1880 function listToText( $l ) {
1881 $s = '';
1882 $m = count($l) - 1;
1883 for ($i = $m; $i >= 0; $i--) {
1884 if ($i == $m) {
1885 $s = $l[$i];
1886 } else if ($i == $m - 1) {
1887 $s = $l[$i] . ' ' . $this->getMessageFromDB( 'and' ) . ' ' . $s;
1888 } else {
1889 $s = $l[$i] . ', ' . $s;
1890 }
1891 }
1892 return $s;
1893 }
1894
1895 /**
1896 * Take a list of strings and build a locale-friendly comma-separated
1897 * list, using the local comma-separator message.
1898 * @param $list array of strings to put in a comma list
1899 * @return string
1900 */
1901 function commaList( $list, $forContent = false ) {
1902 return implode(
1903 $list,
1904 wfMsgExt( 'comma-separator', array( 'escapenoentities', 'language' => $this ) ) );
1905 }
1906
1907 /**
1908 * Same as commaList, but separate it with the pipe instead.
1909 * @param $list array of strings to put in a pipe list
1910 * @return string
1911 */
1912 function pipeList( $list ) {
1913 return implode(
1914 $list,
1915 wfMsgExt( 'pipe-separator', array( 'escapenoentities', 'language' => $this ) ) );
1916 }
1917
1918 /**
1919 * Truncate a string to a specified length in bytes, appending an optional
1920 * string (e.g. for ellipses)
1921 *
1922 * The database offers limited byte lengths for some columns in the database;
1923 * multi-byte character sets mean we need to ensure that only whole characters
1924 * are included, otherwise broken characters can be passed to the user
1925 *
1926 * If $length is negative, the string will be truncated from the beginning
1927 *
1928 * @param $string String to truncate
1929 * @param $length Int: maximum length (excluding ellipses)
1930 * @param $ellipsis String to append to the truncated text
1931 * @return string
1932 */
1933 function truncate( $string, $length, $ellipsis = "" ) {
1934 if( $length == 0 ) {
1935 return $ellipsis;
1936 }
1937 if ( strlen( $string ) <= abs( $length ) ) {
1938 return $string;
1939 }
1940 if( $length > 0 ) {
1941 $string = substr( $string, 0, $length );
1942 $char = ord( $string[strlen( $string ) - 1] );
1943 $m = array();
1944 if ($char >= 0xc0) {
1945 # We got the first byte only of a multibyte char; remove it.
1946 $string = substr( $string, 0, -1 );
1947 } elseif( $char >= 0x80 &&
1948 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
1949 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
1950 # We chopped in the middle of a character; remove it
1951 $string = $m[1];
1952 }
1953 return $string . $ellipsis;
1954 } else {
1955 $string = substr( $string, $length );
1956 $char = ord( $string[0] );
1957 if( $char >= 0x80 && $char < 0xc0 ) {
1958 # We chopped in the middle of a character; remove the whole thing
1959 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
1960 }
1961 return $ellipsis . $string;
1962 }
1963 }
1964
1965 /**
1966 * Grammatical transformations, needed for inflected languages
1967 * Invoked by putting {{grammar:case|word}} in a message
1968 *
1969 * @param $word string
1970 * @param $case string
1971 * @return string
1972 */
1973 function convertGrammar( $word, $case ) {
1974 global $wgGrammarForms;
1975 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
1976 return $wgGrammarForms[$this->getCode()][$case][$word];
1977 }
1978 return $word;
1979 }
1980
1981 /**
1982 * Plural form transformations, needed for some languages.
1983 * For example, there are 3 form of plural in Russian and Polish,
1984 * depending on "count mod 10". See [[w:Plural]]
1985 * For English it is pretty simple.
1986 *
1987 * Invoked by putting {{plural:count|wordform1|wordform2}}
1988 * or {{plural:count|wordform1|wordform2|wordform3}}
1989 *
1990 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
1991 *
1992 * @param $count Integer: non-localized number
1993 * @param $forms Array: different plural forms
1994 * @return string Correct form of plural for $count in this language
1995 */
1996 function convertPlural( $count, $forms ) {
1997 if ( !count($forms) ) { return ''; }
1998 $forms = $this->preConvertPlural( $forms, 2 );
1999
2000 return ( $count == 1 ) ? $forms[0] : $forms[1];
2001 }
2002
2003 /**
2004 * Checks that convertPlural was given an array and pads it to requested
2005 * amound of forms by copying the last one.
2006 *
2007 * @param $count Integer: How many forms should there be at least
2008 * @param $forms Array of forms given to convertPlural
2009 * @return array Padded array of forms or an exception if not an array
2010 */
2011 protected function preConvertPlural( /* Array */ $forms, $count ) {
2012 while ( count($forms) < $count ) {
2013 $forms[] = $forms[count($forms)-1];
2014 }
2015 return $forms;
2016 }
2017
2018 /**
2019 * For translaing of expiry times
2020 * @param $str String: the validated block time in English
2021 * @return Somehow translated block time
2022 * @see LanguageFi.php for example implementation
2023 */
2024 function translateBlockExpiry( $str ) {
2025
2026 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2027
2028 if ( $scBlockExpiryOptions == '-') {
2029 return $str;
2030 }
2031
2032 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2033 if ( strpos($option, ":") === false )
2034 continue;
2035 list($show, $value) = explode(":", $option);
2036 if ( strcmp ( $str, $value) == 0 ) {
2037 return htmlspecialchars( trim( $show ) );
2038 }
2039 }
2040
2041 return $str;
2042 }
2043
2044 /**
2045 * languages like Chinese need to be segmented in order for the diff
2046 * to be of any use
2047 *
2048 * @param $text String
2049 * @return String
2050 */
2051 function segmentForDiff( $text ) {
2052 return $text;
2053 }
2054
2055 /**
2056 * and unsegment to show the result
2057 *
2058 * @param $text String
2059 * @return String
2060 */
2061 function unsegmentForDiff( $text ) {
2062 return $text;
2063 }
2064
2065 # convert text to different variants of a language.
2066 function convert( $text, $isTitle = false) {
2067 return $this->mConverter->convert($text, $isTitle);
2068 }
2069
2070 # Convert text from within Parser
2071 function parserConvert( $text, &$parser ) {
2072 return $this->mConverter->parserConvert( $text, $parser );
2073 }
2074
2075 # Check if this is a language with variants
2076 function hasVariants(){
2077 return sizeof($this->getVariants())>1;
2078 }
2079
2080 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2081 function armourMath($text){
2082 return $this->mConverter->armourMath($text);
2083 }
2084
2085
2086 /**
2087 * Perform output conversion on a string, and encode for safe HTML output.
2088 * @param $text String
2089 * @param $isTitle Bool -- wtf?
2090 * @return string
2091 * @todo this should get integrated somewhere sane
2092 */
2093 function convertHtml( $text, $isTitle = false ) {
2094 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2095 }
2096
2097 function convertCategoryKey( $key ) {
2098 return $this->mConverter->convertCategoryKey( $key );
2099 }
2100
2101 /**
2102 * get the list of variants supported by this langauge
2103 * see sample implementation in LanguageZh.php
2104 *
2105 * @return array an array of language codes
2106 */
2107 function getVariants() {
2108 return $this->mConverter->getVariants();
2109 }
2110
2111
2112 function getPreferredVariant( $fromUser = true ) {
2113 return $this->mConverter->getPreferredVariant( $fromUser );
2114 }
2115
2116 /**
2117 * if a language supports multiple variants, it is
2118 * possible that non-existing link in one variant
2119 * actually exists in another variant. this function
2120 * tries to find it. See e.g. LanguageZh.php
2121 *
2122 * @param $link String: the name of the link
2123 * @param $nt Mixed: the title object of the link
2124 * @return null the input parameters may be modified upon return
2125 */
2126 function findVariantLink( &$link, &$nt, $forTemplate = false ) {
2127 $this->mConverter->findVariantLink($link, $nt, $forTemplate );
2128 }
2129
2130 /**
2131 * If a language supports multiple variants, converts text
2132 * into an array of all possible variants of the text:
2133 * 'variant' => text in that variant
2134 */
2135
2136 function convertLinkToAllVariants($text){
2137 return $this->mConverter->convertLinkToAllVariants($text);
2138 }
2139
2140
2141 /**
2142 * returns language specific options used by User::getPageRenderHash()
2143 * for example, the preferred language variant
2144 *
2145 * @return string
2146 */
2147 function getExtraHashOptions() {
2148 return $this->mConverter->getExtraHashOptions();
2149 }
2150
2151 /**
2152 * for languages that support multiple variants, the title of an
2153 * article may be displayed differently in different variants. this
2154 * function returns the apporiate title defined in the body of the article.
2155 *
2156 * @return string
2157 */
2158 function getParsedTitle() {
2159 return $this->mConverter->getParsedTitle();
2160 }
2161
2162 /**
2163 * Enclose a string with the "no conversion" tag. This is used by
2164 * various functions in the Parser
2165 *
2166 * @param $text String: text to be tagged for no conversion
2167 * @param $noParse
2168 * @return string the tagged text
2169 */
2170 function markNoConversion( $text, $noParse=false ) {
2171 return $this->mConverter->markNoConversion( $text, $noParse );
2172 }
2173
2174 /**
2175 * A regular expression to match legal word-trailing characters
2176 * which should be merged onto a link of the form [[foo]]bar.
2177 *
2178 * @return string
2179 */
2180 function linkTrail() {
2181 $this->load();
2182 return $this->linkTrail;
2183 }
2184
2185 function getLangObj() {
2186 return $this;
2187 }
2188
2189 /**
2190 * Get the RFC 3066 code for this language object
2191 */
2192 function getCode() {
2193 return $this->mCode;
2194 }
2195
2196 function setCode( $code ) {
2197 $this->mCode = $code;
2198 }
2199
2200 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2201 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2202 }
2203
2204 static function getMessagesFileName( $code ) {
2205 global $IP;
2206 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2207 }
2208
2209 static function getClassFileName( $code ) {
2210 global $IP;
2211 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2212 }
2213
2214 static function getLocalisationArray( $code, $disableCache = false ) {
2215 self::loadLocalisation( $code, $disableCache );
2216 return self::$mLocalisationCache[$code];
2217 }
2218
2219 /**
2220 * Load localisation data for a given code into the static cache
2221 *
2222 * @return array Dependencies, map of filenames to mtimes
2223 */
2224 static function loadLocalisation( $code, $disableCache = false ) {
2225 static $recursionGuard = array();
2226 global $wgMemc, $wgCheckSerialized;
2227
2228 if ( !$code ) {
2229 throw new MWException( "Invalid language code requested" );
2230 }
2231
2232 if ( !$disableCache ) {
2233 # Try the per-process cache
2234 if ( isset( self::$mLocalisationCache[$code] ) ) {
2235 return self::$mLocalisationCache[$code]['deps'];
2236 }
2237
2238 wfProfileIn( __METHOD__ );
2239
2240 # Try the serialized directory
2241 $cache = wfGetPrecompiledData( self::getFileName( "Messages", $code, '.ser' ) );
2242 if ( $cache ) {
2243 if ( $wgCheckSerialized && self::isLocalisationOutOfDate( $cache ) ) {
2244 $cache = false;
2245 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2246 } else {
2247 self::$mLocalisationCache[$code] = $cache;
2248 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2249 wfProfileOut( __METHOD__ );
2250 return self::$mLocalisationCache[$code]['deps'];
2251 }
2252 }
2253
2254 # Try the global cache
2255 $memcKey = wfMemcKey('localisation', $code );
2256 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2257 $cache = $wgMemc->get( $memcKey );
2258 if ( $cache ) {
2259 if ( self::isLocalisationOutOfDate( $cache ) ) {
2260 $wgMemc->delete( $memcKey );
2261 $wgMemc->delete( $fbMemcKey );
2262 $cache = false;
2263 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2264 } else {
2265 self::$mLocalisationCache[$code] = $cache;
2266 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2267 wfProfileOut( __METHOD__ );
2268 return $cache['deps'];
2269 }
2270 }
2271 } else {
2272 wfProfileIn( __METHOD__ );
2273 }
2274
2275 # Default fallback, may be overridden when the messages file is included
2276 if ( $code != 'en' ) {
2277 $fallback = 'en';
2278 } else {
2279 $fallback = false;
2280 }
2281
2282 # Load the primary localisation from the source file
2283 $filename = self::getMessagesFileName( $code );
2284 if ( !file_exists( $filename ) ) {
2285 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2286 $cache = compact( self::$mLocalisationKeys ); // Set correct fallback
2287 $deps = array();
2288 } else {
2289 $deps = array( $filename => filemtime( $filename ) );
2290 require( $filename );
2291 $cache = compact( self::$mLocalisationKeys );
2292 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2293 }
2294
2295 if ( !empty( $fallback ) ) {
2296 # Load the fallback localisation, with a circular reference guard
2297 if ( isset( $recursionGuard[$code] ) ) {
2298 throw new MWException( "Error: Circular fallback reference in language code $code" );
2299 }
2300 $recursionGuard[$code] = true;
2301 $newDeps = self::loadLocalisation( $fallback, $disableCache );
2302 unset( $recursionGuard[$code] );
2303
2304 $secondary = self::$mLocalisationCache[$fallback];
2305 $deps = array_merge( $deps, $newDeps );
2306
2307 # Merge the fallback localisation with the current localisation
2308 foreach ( self::$mLocalisationKeys as $key ) {
2309 if ( isset( $cache[$key] ) ) {
2310 if ( isset( $secondary[$key] ) ) {
2311 if ( in_array( $key, self::$mMergeableMapKeys ) ) {
2312 $cache[$key] = $cache[$key] + $secondary[$key];
2313 } elseif ( in_array( $key, self::$mMergeableListKeys ) ) {
2314 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2315 } elseif ( in_array( $key, self::$mMergeableAliasListKeys ) ) {
2316 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2317 }
2318 }
2319 } else {
2320 $cache[$key] = $secondary[$key];
2321 }
2322 }
2323
2324 # Merge bookstore lists if requested
2325 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2326 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2327 }
2328 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2329 unset( $cache['bookstoreList']['inherit'] );
2330 }
2331 }
2332
2333 # Add dependencies to the cache entry
2334 $cache['deps'] = $deps;
2335
2336 # Replace spaces with underscores in namespace names
2337 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2338
2339 # And do the same for specialpage aliases. $page is an array.
2340 foreach ( $cache['specialPageAliases'] as &$page ) {
2341 $page = str_replace( ' ', '_', $page );
2342 }
2343 # Decouple the reference to prevent accidental damage
2344 unset($page);
2345
2346 # Save to both caches
2347 self::$mLocalisationCache[$code] = $cache;
2348 if ( !$disableCache ) {
2349 $wgMemc->set( $memcKey, $cache );
2350 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2351 }
2352
2353 wfProfileOut( __METHOD__ );
2354 return $deps;
2355 }
2356
2357 /**
2358 * Test if a given localisation cache is out of date with respect to the
2359 * source Messages files. This is done automatically for the global cache
2360 * in $wgMemc, but is only done on certain occasions for the serialized
2361 * data file.
2362 *
2363 * @param $cache mixed Either a language code or a cache array
2364 */
2365 static function isLocalisationOutOfDate( $cache ) {
2366 if ( !is_array( $cache ) ) {
2367 self::loadLocalisation( $cache );
2368 $cache = self::$mLocalisationCache[$cache];
2369 }
2370 $expired = false;
2371 foreach ( $cache['deps'] as $file => $mtime ) {
2372 if ( !file_exists( $file ) || filemtime( $file ) > $mtime ) {
2373 $expired = true;
2374 break;
2375 }
2376 }
2377 return $expired;
2378 }
2379
2380 /**
2381 * Get the fallback for a given language
2382 */
2383 static function getFallbackFor( $code ) {
2384 // Shortcut
2385 if ( $code === 'en' ) return false;
2386
2387 // Local cache
2388 static $cache = array();
2389 // Quick return
2390 if ( isset($cache[$code]) ) return $cache[$code];
2391
2392 // Try memcache
2393 global $wgMemc;
2394 $memcKey = wfMemcKey( 'fallback', $code );
2395 $fbcode = $wgMemc->get( $memcKey );
2396
2397 if ( is_string($fbcode) ) {
2398 // False is stored as a string to detect failures in memcache properly
2399 if ( $fbcode === '' ) $fbcode = false;
2400
2401 // Update local cache and return
2402 $cache[$code] = $fbcode;
2403 return $fbcode;
2404 }
2405
2406 // Nothing in caches, load and and update both caches
2407 self::loadLocalisation( $code );
2408 $fbcode = self::$mLocalisationCache[$code]['fallback'];
2409
2410 $cache[$code] = $fbcode;
2411 $wgMemc->set( $memcKey, (string) $fbcode );
2412
2413 return $fbcode;
2414 }
2415
2416 /**
2417 * Get all messages for a given language
2418 */
2419 static function getMessagesFor( $code ) {
2420 self::loadLocalisation( $code );
2421 return self::$mLocalisationCache[$code]['messages'];
2422 }
2423
2424 /**
2425 * Get a message for a given language
2426 */
2427 static function getMessageFor( $key, $code ) {
2428 self::loadLocalisation( $code );
2429 return isset( self::$mLocalisationCache[$code]['messages'][$key] ) ? self::$mLocalisationCache[$code]['messages'][$key] : null;
2430 }
2431
2432 /**
2433 * Load localisation data for this object
2434 */
2435 function load() {
2436 if ( !$this->mLoaded ) {
2437 self::loadLocalisation( $this->getCode() );
2438 $cache =& self::$mLocalisationCache[$this->getCode()];
2439 foreach ( self::$mLocalisationKeys as $key ) {
2440 $this->$key = $cache[$key];
2441 }
2442 $this->mLoaded = true;
2443
2444 $this->fixUpSettings();
2445 }
2446 }
2447
2448 /**
2449 * Do any necessary post-cache-load settings adjustment
2450 */
2451 function fixUpSettings() {
2452 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2453 $wgNamespaceAliases, $wgAmericanDates;
2454 wfProfileIn( __METHOD__ );
2455 if ( $wgExtraNamespaces ) {
2456 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
2457 }
2458
2459 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
2460 if ( $wgMetaNamespaceTalk ) {
2461 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
2462 } else {
2463 $talk = $this->namespaceNames[NS_PROJECT_TALK];
2464 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2465
2466 # Allow grammar transformations
2467 # Allowing full message-style parsing would make simple requests
2468 # such as action=raw much more expensive than they need to be.
2469 # This will hopefully cover most cases.
2470 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2471 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2472 $talk = str_replace( ' ', '_', $talk );
2473 $this->namespaceNames[NS_PROJECT_TALK] = $talk;
2474 }
2475
2476 # The above mixing may leave namespaces out of canonical order.
2477 # Re-order by namespace ID number...
2478 ksort( $this->namespaceNames );
2479
2480 # Put namespace names and aliases into a hashtable.
2481 # If this is too slow, then we should arrange it so that it is done
2482 # before caching. The catch is that at pre-cache time, the above
2483 # class-specific fixup hasn't been done.
2484 $this->mNamespaceIds = array();
2485 foreach ( $this->namespaceNames as $index => $name ) {
2486 $this->mNamespaceIds[$this->lc($name)] = $index;
2487 }
2488 if ( $this->namespaceAliases ) {
2489 foreach ( $this->namespaceAliases as $name => $index ) {
2490 $this->mNamespaceIds[$this->lc($name)] = $index;
2491 }
2492 }
2493 if ( $wgNamespaceAliases ) {
2494 foreach ( $wgNamespaceAliases as $name => $index ) {
2495 $this->mNamespaceIds[$this->lc($name)] = $index;
2496 }
2497 }
2498
2499 if ( $this->defaultDateFormat == 'dmy or mdy' ) {
2500 $this->defaultDateFormat = $wgAmericanDates ? 'mdy' : 'dmy';
2501 }
2502 wfProfileOut( __METHOD__ );
2503 }
2504
2505 function replaceGrammarInNamespace( $m ) {
2506 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2507 }
2508
2509 static function getCaseMaps() {
2510 static $wikiUpperChars, $wikiLowerChars;
2511 if ( isset( $wikiUpperChars ) ) {
2512 return array( $wikiUpperChars, $wikiLowerChars );
2513 }
2514
2515 wfProfileIn( __METHOD__ );
2516 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2517 if ( $arr === false ) {
2518 throw new MWException(
2519 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2520 }
2521 extract( $arr );
2522 wfProfileOut( __METHOD__ );
2523 return array( $wikiUpperChars, $wikiLowerChars );
2524 }
2525
2526 function formatTimePeriod( $seconds ) {
2527 if ( $seconds < 10 ) {
2528 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2529 } elseif ( $seconds < 60 ) {
2530 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2531 } elseif ( $seconds < 3600 ) {
2532 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2533 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2534 } else {
2535 $hours = floor( $seconds / 3600 );
2536 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2537 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2538 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2539 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2540 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2541 }
2542 }
2543
2544 function formatBitrate( $bps ) {
2545 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2546 if ( $bps <= 0 ) {
2547 return $this->formatNum( $bps ) . $units[0];
2548 }
2549 $unitIndex = floor( log10( $bps ) / 3 );
2550 $mantissa = $bps / pow( 1000, $unitIndex );
2551 if ( $mantissa < 10 ) {
2552 $mantissa = round( $mantissa, 1 );
2553 } else {
2554 $mantissa = round( $mantissa );
2555 }
2556 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2557 }
2558
2559 /**
2560 * Format a size in bytes for output, using an appropriate
2561 * unit (B, KB, MB or GB) according to the magnitude in question
2562 *
2563 * @param $size Size to format
2564 * @return string Plain text (not HTML)
2565 */
2566 function formatSize( $size ) {
2567 // For small sizes no decimal places necessary
2568 $round = 0;
2569 if( $size > 1024 ) {
2570 $size = $size / 1024;
2571 if( $size > 1024 ) {
2572 $size = $size / 1024;
2573 // For MB and bigger two decimal places are smarter
2574 $round = 2;
2575 if( $size > 1024 ) {
2576 $size = $size / 1024;
2577 $msg = 'size-gigabytes';
2578 } else {
2579 $msg = 'size-megabytes';
2580 }
2581 } else {
2582 $msg = 'size-kilobytes';
2583 }
2584 } else {
2585 $msg = 'size-bytes';
2586 }
2587 $size = round( $size, $round );
2588 $text = $this->getMessageFromDB( $msg );
2589 return str_replace( '$1', $this->formatNum( $size ), $text );
2590 }
2591 }