dépôts / lhc / web / wiklou.git / blob
? search:


49a37ea1a1f069050db9b339e6ec7cc6d142e1a9
[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 static public $mLangObjCache = array();
78
79 static public $mWeekdayMsgs = array(
80 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
81 'friday', 'saturday'
82 );
83
84 static public $mWeekdayAbbrevMsgs = array(
85 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
86 );
87
88 static public $mMonthMsgs = array(
89 'january', 'february', 'march', 'april', 'may_long', 'june',
90 'july', 'august', 'september', 'october', 'november',
91 'december'
92 );
93 static public $mMonthGenMsgs = array(
94 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
95 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
96 'december-gen'
97 );
98 static public $mMonthAbbrevMsgs = array(
99 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
100 'sep', 'oct', 'nov', 'dec'
101 );
102
103 static public $mIranianCalendarMonthMsgs = array(
104 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
105 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
106 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
107 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
108 );
109
110 static public $mHebrewCalendarMonthMsgs = array(
111 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
112 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
113 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
114 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
115 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
116 );
117
118 static public $mHebrewCalendarMonthGenMsgs = array(
119 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
120 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
121 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
122 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
123 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
124 );
125
126 static public $mHijriCalendarMonthMsgs = array(
127 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
128 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
129 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
130 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
131 );
132
133 /**
134 * Get a cached language object for a given language code
135 */
136 static function factory( $code ) {
137 if ( !isset( self::$mLangObjCache[$code] ) ) {
138 if( count( self::$mLangObjCache ) > 10 ) {
139 // Don't keep a billion objects around, that's stupid.
140 self::$mLangObjCache = array();
141 }
142 self::$mLangObjCache[$code] = self::newFromCode( $code );
143 }
144 return self::$mLangObjCache[$code];
145 }
146
147 /**
148 * Create a language object for a given language code
149 */
150 protected static function newFromCode( $code ) {
151 global $IP;
152 static $recursionLevel = 0;
153 if ( $code == 'en' ) {
154 $class = 'Language';
155 } else {
156 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
157 // Preload base classes to work around APC/PHP5 bug
158 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
159 include_once("$IP/languages/classes/$class.deps.php");
160 }
161 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
162 include_once("$IP/languages/classes/$class.php");
163 }
164 }
165
166 if ( $recursionLevel > 5 ) {
167 throw new MWException( "Language fallback loop detected when creating class $class\n" );
168 }
169
170 if( ! class_exists( $class ) ) {
171 $fallback = Language::getFallbackFor( $code );
172 ++$recursionLevel;
173 $lang = Language::newFromCode( $fallback );
174 --$recursionLevel;
175 $lang->setCode( $code );
176 } else {
177 $lang = new $class;
178 }
179 return $lang;
180 }
181
182 function __construct() {
183 $this->mConverter = new FakeConverter($this);
184 // Set the code to the name of the descendant
185 if ( get_class( $this ) == 'Language' ) {
186 $this->mCode = 'en';
187 } else {
188 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
189 }
190 }
191
192 /**
193 * Reduce memory usage
194 */
195 function __destruct() {
196 foreach ( $this as $name => $value ) {
197 unset( $this->$name );
198 }
199 }
200
201 /**
202 * Hook which will be called if this is the content language.
203 * Descendants can use this to register hook functions or modify globals
204 */
205 function initContLang() {}
206
207 /**
208 * @deprecated Use User::getDefaultOptions()
209 * @return array
210 */
211 function getDefaultUserOptions() {
212 wfDeprecated( __METHOD__ );
213 return User::getDefaultOptions();
214 }
215
216 function getFallbackLanguageCode() {
217 return self::getFallbackFor( $this->mCode );
218 }
219
220 /**
221 * Exports $wgBookstoreListEn
222 * @return array
223 */
224 function getBookstoreList() {
225 $this->load();
226 return $this->bookstoreList;
227 }
228
229 /**
230 * @return array
231 */
232 function getNamespaces() {
233 $this->load();
234 return $this->namespaceNames;
235 }
236
237 /**
238 * A convenience function that returns the same thing as
239 * getNamespaces() except with the array values changed to ' '
240 * where it found '_', useful for producing output to be displayed
241 * e.g. in <select> forms.
242 *
243 * @return array
244 */
245 function getFormattedNamespaces() {
246 $ns = $this->getNamespaces();
247 foreach($ns as $k => $v) {
248 $ns[$k] = strtr($v, '_', ' ');
249 }
250 return $ns;
251 }
252
253 /**
254 * Get a namespace value by key
255 * <code>
256 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
257 * echo $mw_ns; // prints 'MediaWiki'
258 * </code>
259 *
260 * @param $index Int: the array key of the namespace to return
261 * @return mixed, string if the namespace value exists, otherwise false
262 */
263 function getNsText( $index ) {
264 $ns = $this->getNamespaces();
265 return isset( $ns[$index] ) ? $ns[$index] : false;
266 }
267
268 /**
269 * A convenience function that returns the same thing as
270 * getNsText() except with '_' changed to ' ', useful for
271 * producing output.
272 *
273 * @return array
274 */
275 function getFormattedNsText( $index ) {
276 $ns = $this->getNsText( $index );
277 return strtr($ns, '_', ' ');
278 }
279
280 /**
281 * Get a namespace key by value, case insensitive.
282 * Only matches namespace names for the current language, not the
283 * canonical ones defined in Namespace.php.
284 *
285 * @param $text String
286 * @return mixed An integer if $text is a valid value otherwise false
287 */
288 function getLocalNsIndex( $text ) {
289 $this->load();
290 $lctext = $this->lc($text);
291 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
292 }
293
294 /**
295 * Get a namespace key by value, case insensitive. Canonical namespace
296 * names override custom ones defined for the current language.
297 *
298 * @param $text String
299 * @return mixed An integer if $text is a valid value otherwise false
300 */
301 function getNsIndex( $text ) {
302 $this->load();
303 $lctext = $this->lc($text);
304 if( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
305 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
306 }
307
308 /**
309 * short names for language variants used for language conversion links.
310 *
311 * @param $code String
312 * @return string
313 */
314 function getVariantname( $code ) {
315 return $this->getMessageFromDB( "variantname-$code" );
316 }
317
318 function specialPage( $name ) {
319 $aliases = $this->getSpecialPageAliases();
320 if ( isset( $aliases[$name][0] ) ) {
321 $name = $aliases[$name][0];
322 }
323 return $this->getNsText(NS_SPECIAL) . ':' . $name;
324 }
325
326 function getQuickbarSettings() {
327 return array(
328 $this->getMessage( 'qbsettings-none' ),
329 $this->getMessage( 'qbsettings-fixedleft' ),
330 $this->getMessage( 'qbsettings-fixedright' ),
331 $this->getMessage( 'qbsettings-floatingleft' ),
332 $this->getMessage( 'qbsettings-floatingright' )
333 );
334 }
335
336 function getSkinNames() {
337 $this->load();
338 return $this->skinNames;
339 }
340
341 function getMathNames() {
342 $this->load();
343 return $this->mathNames;
344 }
345
346 function getDatePreferences() {
347 $this->load();
348 return $this->datePreferences;
349 }
350
351 function getDateFormats() {
352 $this->load();
353 return $this->dateFormats;
354 }
355
356 function getDefaultDateFormat() {
357 $this->load();
358 return $this->defaultDateFormat;
359 }
360
361 function getDatePreferenceMigrationMap() {
362 $this->load();
363 return $this->datePreferenceMigrationMap;
364 }
365
366 function getImageFile( $image ) {
367 $this->load();
368 return $this->imageFiles[$image];
369 }
370
371 function getDefaultUserOptionOverrides() {
372 $this->load();
373 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
374 if (is_array($this->defaultUserOptionOverrides)) {
375 return $this->defaultUserOptionOverrides;
376 } else {
377 return array();
378 }
379 }
380
381 function getExtraUserToggles() {
382 $this->load();
383 return $this->extraUserToggles;
384 }
385
386 function getUserToggle( $tog ) {
387 return $this->getMessageFromDB( "tog-$tog" );
388 }
389
390 /**
391 * Get language names, indexed by code.
392 * If $customisedOnly is true, only returns codes with a messages file
393 */
394 public static function getLanguageNames( $customisedOnly = false ) {
395 global $wgLanguageNames, $wgExtraLanguageNames;
396 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
397 if ( !$customisedOnly ) {
398 return $allNames;
399 }
400
401 global $IP;
402 $names = array();
403 $dir = opendir( "$IP/languages/messages" );
404 while( false !== ( $file = readdir( $dir ) ) ) {
405 $m = array();
406 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
407 $code = str_replace( '_', '-', strtolower( $m[1] ) );
408 if ( isset( $allNames[$code] ) ) {
409 $names[$code] = $allNames[$code];
410 }
411 }
412 }
413 closedir( $dir );
414 return $names;
415 }
416
417 /**
418 * Get a message from the MediaWiki namespace.
419 *
420 * @param $msg String: message name
421 * @return string
422 */
423 function getMessageFromDB( $msg ) {
424 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
425 }
426
427 function getLanguageName( $code ) {
428 $names = self::getLanguageNames();
429 if ( !array_key_exists( $code, $names ) ) {
430 return '';
431 }
432 return $names[$code];
433 }
434
435 function getMonthName( $key ) {
436 return $this->getMessageFromDB( self::$mMonthMsgs[$key-1] );
437 }
438
439 function getMonthNameGen( $key ) {
440 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key-1] );
441 }
442
443 function getMonthAbbreviation( $key ) {
444 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key-1] );
445 }
446
447 function getWeekdayName( $key ) {
448 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key-1] );
449 }
450
451 function getWeekdayAbbreviation( $key ) {
452 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key-1] );
453 }
454
455 function getIranianCalendarMonthName( $key ) {
456 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key-1] );
457 }
458
459 function getHebrewCalendarMonthName( $key ) {
460 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key-1] );
461 }
462
463 function getHebrewCalendarMonthNameGen( $key ) {
464 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key-1] );
465 }
466
467 function getHijriCalendarMonthName( $key ) {
468 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key-1] );
469 }
470
471 /**
472 * Used by date() and time() to adjust the time output.
473 *
474 * @param $ts Int the time in date('YmdHis') format
475 * @param $tz Mixed: adjust the time by this amount (default false, mean we
476 * get user timecorrection setting)
477 * @return int
478 */
479 function userAdjust( $ts, $tz = false ) {
480 global $wgUser, $wgLocalTZoffset;
481
482 if (!$tz) {
483 $tz = $wgUser->getOption( 'timecorrection' );
484 }
485
486 # minutes and hours differences:
487 $minDiff = 0;
488 $hrDiff = 0;
489
490 if ( $tz === '' ) {
491 # Global offset in minutes.
492 if( isset($wgLocalTZoffset) ) {
493 if( $wgLocalTZoffset >= 0 ) {
494 $hrDiff = floor($wgLocalTZoffset / 60);
495 } else {
496 $hrDiff = ceil($wgLocalTZoffset / 60);
497 }
498 $minDiff = $wgLocalTZoffset % 60;
499 }
500 } elseif ( strpos( $tz, ':' ) !== false ) {
501 $tzArray = explode( ':', $tz );
502 $hrDiff = intval($tzArray[0]);
503 $minDiff = intval($hrDiff < 0 ? -$tzArray[1] : $tzArray[1]);
504 } else {
505 $hrDiff = intval( $tz );
506 }
507
508 # No difference ? Return time unchanged
509 if ( 0 == $hrDiff && 0 == $minDiff ) { return $ts; }
510
511 wfSuppressWarnings(); // E_STRICT system time bitching
512 # Generate an adjusted date
513 $t = mktime( (
514 (int)substr( $ts, 8, 2) ) + $hrDiff, # Hours
515 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
516 (int)substr( $ts, 12, 2 ), # Seconds
517 (int)substr( $ts, 4, 2 ), # Month
518 (int)substr( $ts, 6, 2 ), # Day
519 (int)substr( $ts, 0, 4 ) ); #Year
520
521 $date = date( 'YmdHis', $t );
522 wfRestoreWarnings();
523
524 return $date;
525 }
526
527 /**
528 * This is a workalike of PHP's date() function, but with better
529 * internationalisation, a reduced set of format characters, and a better
530 * escaping format.
531 *
532 * Supported format characters are dDjlNwzWFmMntLYyaAgGhHiscrU. See the
533 * PHP manual for definitions. There are a number of extensions, which
534 * start with "x":
535 *
536 * xn Do not translate digits of the next numeric format character
537 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
538 * xr Use roman numerals for the next numeric format character
539 * xh Use hebrew numerals for the next numeric format character
540 * xx Literal x
541 * xg Genitive month name
542 *
543 * xij j (day number) in Iranian calendar
544 * xiF F (month name) in Iranian calendar
545 * xin n (month number) in Iranian calendar
546 * xiY Y (full year) in Iranian calendar
547 *
548 * xjj j (day number) in Hebrew calendar
549 * xjF F (month name) in Hebrew calendar
550 * xjt t (days in month) in Hebrew calendar
551 * xjx xg (genitive month name) in Hebrew calendar
552 * xjn n (month number) in Hebrew calendar
553 * xjY Y (full year) in Hebrew calendar
554 *
555 * xmj j (day number) in Hijri calendar
556 * xmF F (month name) in Hijri calendar
557 * xmn n (month number) in Hijri calendar
558 * xmY Y (full year) in Hijri calendar
559 *
560 * xkY Y (full year) in Thai solar calendar. Months and days are
561 * identical to the Gregorian calendar
562 *
563 * Characters enclosed in double quotes will be considered literal (with
564 * the quotes themselves removed). Unmatched quotes will be considered
565 * literal quotes. Example:
566 *
567 * "The month is" F => The month is January
568 * i's" => 20'11"
569 *
570 * Backslash escaping is also supported.
571 *
572 * Input timestamp is assumed to be pre-normalized to the desired local
573 * time zone, if any.
574 *
575 * @param $format String
576 * @param $ts String: 14-character timestamp
577 * YYYYMMDDHHMMSS
578 * 01234567890123
579 */
580 function sprintfDate( $format, $ts ) {
581 $s = '';
582 $raw = false;
583 $roman = false;
584 $hebrewNum = false;
585 $unix = false;
586 $rawToggle = false;
587 $iranian = false;
588 $hebrew = false;
589 $hijri = false;
590 $thai = false;
591 for ( $p = 0; $p < strlen( $format ); $p++ ) {
592 $num = false;
593 $code = $format[$p];
594 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
595 $code .= $format[++$p];
596 }
597
598 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' ) && $p < strlen( $format ) - 1 ) {
599 $code .= $format[++$p];
600 }
601
602 switch ( $code ) {
603 case 'xx':
604 $s .= 'x';
605 break;
606 case 'xn':
607 $raw = true;
608 break;
609 case 'xN':
610 $rawToggle = !$rawToggle;
611 break;
612 case 'xr':
613 $roman = true;
614 break;
615 case 'xh':
616 $hebrewNum = true;
617 break;
618 case 'xg':
619 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
620 break;
621 case 'xjx':
622 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
623 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
624 break;
625 case 'd':
626 $num = substr( $ts, 6, 2 );
627 break;
628 case 'D':
629 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
630 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
631 break;
632 case 'j':
633 $num = intval( substr( $ts, 6, 2 ) );
634 break;
635 case 'xij':
636 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
637 $num = $iranian[2];
638 break;
639 case 'xmj':
640 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
641 $num = $hijri[2];
642 break;
643 case 'xjj':
644 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
645 $num = $hebrew[2];
646 break;
647 case 'l':
648 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
649 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
650 break;
651 case 'N':
652 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
653 $w = gmdate( 'w', $unix );
654 $num = $w ? $w : 7;
655 break;
656 case 'w':
657 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
658 $num = gmdate( 'w', $unix );
659 break;
660 case 'z':
661 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
662 $num = gmdate( 'z', $unix );
663 break;
664 case 'W':
665 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
666 $num = gmdate( 'W', $unix );
667 break;
668 case 'F':
669 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
670 break;
671 case 'xiF':
672 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
673 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
674 break;
675 case 'xmF':
676 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
677 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
678 break;
679 case 'xjF':
680 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
681 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
682 break;
683 case 'm':
684 $num = substr( $ts, 4, 2 );
685 break;
686 case 'M':
687 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
688 break;
689 case 'n':
690 $num = intval( substr( $ts, 4, 2 ) );
691 break;
692 case 'xin':
693 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
694 $num = $iranian[1];
695 break;
696 case 'xmn':
697 if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
698 $num = $hijri[1];
699 break;
700 case 'xjn':
701 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
702 $num = $hebrew[1];
703 break;
704 case 't':
705 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
706 $num = gmdate( 't', $unix );
707 break;
708 case 'xjt':
709 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
710 $num = $hebrew[3];
711 break;
712 case 'L':
713 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
714 $num = gmdate( 'L', $unix );
715 break;
716 case 'Y':
717 $num = substr( $ts, 0, 4 );
718 break;
719 case 'xiY':
720 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
721 $num = $iranian[0];
722 break;
723 case 'xmY':
724 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
725 $num = $hijri[0];
726 break;
727 case 'xjY':
728 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
729 $num = $hebrew[0];
730 break;
731 case 'xkY':
732 if ( !$thai ) $thai = self::tsToThai( $ts );
733 $num = $thai[0];
734 break;
735 case 'y':
736 $num = substr( $ts, 2, 2 );
737 break;
738 case 'a':
739 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
740 break;
741 case 'A':
742 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
743 break;
744 case 'g':
745 $h = substr( $ts, 8, 2 );
746 $num = $h % 12 ? $h % 12 : 12;
747 break;
748 case 'G':
749 $num = intval( substr( $ts, 8, 2 ) );
750 break;
751 case 'h':
752 $h = substr( $ts, 8, 2 );
753 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
754 break;
755 case 'H':
756 $num = substr( $ts, 8, 2 );
757 break;
758 case 'i':
759 $num = substr( $ts, 10, 2 );
760 break;
761 case 's':
762 $num = substr( $ts, 12, 2 );
763 break;
764 case 'c':
765 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
766 $s .= gmdate( 'c', $unix );
767 break;
768 case 'r':
769 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
770 $s .= gmdate( 'r', $unix );
771 break;
772 case 'U':
773 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
774 $num = $unix;
775 break;
776 case '\\':
777 # Backslash escaping
778 if ( $p < strlen( $format ) - 1 ) {
779 $s .= $format[++$p];
780 } else {
781 $s .= '\\';
782 }
783 break;
784 case '"':
785 # Quoted literal
786 if ( $p < strlen( $format ) - 1 ) {
787 $endQuote = strpos( $format, '"', $p + 1 );
788 if ( $endQuote === false ) {
789 # No terminating quote, assume literal "
790 $s .= '"';
791 } else {
792 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
793 $p = $endQuote;
794 }
795 } else {
796 # Quote at end of string, assume literal "
797 $s .= '"';
798 }
799 break;
800 default:
801 $s .= $format[$p];
802 }
803 if ( $num !== false ) {
804 if ( $rawToggle || $raw ) {
805 $s .= $num;
806 $raw = false;
807 } elseif ( $roman ) {
808 $s .= self::romanNumeral( $num );
809 $roman = false;
810 } elseif( $hebrewNum ) {
811 $s .= self::hebrewNumeral( $num );
812 $hebrewNum = false;
813 } else {
814 $s .= $this->formatNum( $num, true );
815 }
816 $num = false;
817 }
818 }
819 return $s;
820 }
821
822 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
823 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
824 /**
825 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
826 * Gregorian dates to Iranian dates. Originally written in C, it
827 * is released under the terms of GNU Lesser General Public
828 * License. Conversion to PHP was performed by Niklas Laxström.
829 *
830 * Link: http://www.farsiweb.info/jalali/jalali.c
831 */
832 private static function tsToIranian( $ts ) {
833 $gy = substr( $ts, 0, 4 ) -1600;
834 $gm = substr( $ts, 4, 2 ) -1;
835 $gd = substr( $ts, 6, 2 ) -1;
836
837 # Days passed from the beginning (including leap years)
838 $gDayNo = 365*$gy
839 + floor(($gy+3) / 4)
840 - floor(($gy+99) / 100)
841 + floor(($gy+399) / 400);
842
843
844 // Add days of the past months of this year
845 for( $i = 0; $i < $gm; $i++ ) {
846 $gDayNo += self::$GREG_DAYS[$i];
847 }
848
849 // Leap years
850 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
851 $gDayNo++;
852 }
853
854 // Days passed in current month
855 $gDayNo += $gd;
856
857 $jDayNo = $gDayNo - 79;
858
859 $jNp = floor($jDayNo / 12053);
860 $jDayNo %= 12053;
861
862 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
863 $jDayNo %= 1461;
864
865 if ( $jDayNo >= 366 ) {
866 $jy += floor(($jDayNo-1)/365);
867 $jDayNo = floor(($jDayNo-1)%365);
868 }
869
870 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
871 $jDayNo -= self::$IRANIAN_DAYS[$i];
872 }
873
874 $jm= $i+1;
875 $jd= $jDayNo+1;
876
877 return array($jy, $jm, $jd);
878 }
879 /**
880 * Converting Gregorian dates to Hijri dates.
881 *
882 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
883 *
884 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
885 */
886 private static function tsToHijri ( $ts ) {
887 $year = substr( $ts, 0, 4 );
888 $month = substr( $ts, 4, 2 );
889 $day = substr( $ts, 6, 2 );
890
891 $zyr = $year;
892 $zd=$day;
893 $zm=$month;
894 $zy=$zyr;
895
896
897
898 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
899 {
900
901
902 $zjd=(int)((1461*($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;
903 }
904 else
905 {
906 $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
907 }
908
909 $zl=$zjd-1948440+10632;
910 $zn=(int)(($zl-1)/10631);
911 $zl=$zl-10631*$zn+354;
912 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
913 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
914 $zm=(int)((24*$zl)/709);
915 $zd=$zl-(int)((709*$zm)/24);
916 $zy=30*$zn+$zj-30;
917
918 return array ($zy, $zm, $zd);
919 }
920
921 /**
922 * Converting Gregorian dates to Hebrew dates.
923 *
924 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
925 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
926 * to translate the relevant functions into PHP and release them under
927 * GNU GPL.
928 *
929 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
930 * and Adar II is 14. In a non-leap year, Adar is 6.
931 */
932 private static function tsToHebrew( $ts ) {
933 # Parse date
934 $year = substr( $ts, 0, 4 );
935 $month = substr( $ts, 4, 2 );
936 $day = substr( $ts, 6, 2 );
937
938 # Calculate Hebrew year
939 $hebrewYear = $year + 3760;
940
941 # Month number when September = 1, August = 12
942 $month += 4;
943 if( $month > 12 ) {
944 # Next year
945 $month -= 12;
946 $year++;
947 $hebrewYear++;
948 }
949
950 # Calculate day of year from 1 September
951 $dayOfYear = $day;
952 for( $i = 1; $i < $month; $i++ ) {
953 if( $i == 6 ) {
954 # February
955 $dayOfYear += 28;
956 # Check if the year is leap
957 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
958 $dayOfYear++;
959 }
960 } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
961 $dayOfYear += 30;
962 } else {
963 $dayOfYear += 31;
964 }
965 }
966
967 # Calculate the start of the Hebrew year
968 $start = self::hebrewYearStart( $hebrewYear );
969
970 # Calculate next year's start
971 if( $dayOfYear <= $start ) {
972 # Day is before the start of the year - it is the previous year
973 # Next year's start
974 $nextStart = $start;
975 # Previous year
976 $year--;
977 $hebrewYear--;
978 # Add days since previous year's 1 September
979 $dayOfYear += 365;
980 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
981 # Leap year
982 $dayOfYear++;
983 }
984 # Start of the new (previous) year
985 $start = self::hebrewYearStart( $hebrewYear );
986 } else {
987 # Next year's start
988 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
989 }
990
991 # Calculate Hebrew day of year
992 $hebrewDayOfYear = $dayOfYear - $start;
993
994 # Difference between year's days
995 $diff = $nextStart - $start;
996 # Add 12 (or 13 for leap years) days to ignore the difference between
997 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
998 # difference is only about the year type
999 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1000 $diff += 13;
1001 } else {
1002 $diff += 12;
1003 }
1004
1005 # Check the year pattern, and is leap year
1006 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1007 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1008 # and non-leap years
1009 $yearPattern = $diff % 30;
1010 # Check if leap year
1011 $isLeap = $diff >= 30;
1012
1013 # Calculate day in the month from number of day in the Hebrew year
1014 # Don't check Adar - if the day is not in Adar, we will stop before;
1015 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1016 $hebrewDay = $hebrewDayOfYear;
1017 $hebrewMonth = 1;
1018 $days = 0;
1019 while( $hebrewMonth <= 12 ) {
1020 # Calculate days in this month
1021 if( $isLeap && $hebrewMonth == 6 ) {
1022 # Adar in a leap year
1023 if( $isLeap ) {
1024 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1025 $days = 30;
1026 if( $hebrewDay <= $days ) {
1027 # Day in Adar I
1028 $hebrewMonth = 13;
1029 } else {
1030 # Subtract the days of Adar I
1031 $hebrewDay -= $days;
1032 # Try Adar II
1033 $days = 29;
1034 if( $hebrewDay <= $days ) {
1035 # Day in Adar II
1036 $hebrewMonth = 14;
1037 }
1038 }
1039 }
1040 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1041 # Cheshvan in a complete year (otherwise as the rule below)
1042 $days = 30;
1043 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1044 # Kislev in an incomplete year (otherwise as the rule below)
1045 $days = 29;
1046 } else {
1047 # Odd months have 30 days, even have 29
1048 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1049 }
1050 if( $hebrewDay <= $days ) {
1051 # In the current month
1052 break;
1053 } else {
1054 # Subtract the days of the current month
1055 $hebrewDay -= $days;
1056 # Try in the next month
1057 $hebrewMonth++;
1058 }
1059 }
1060
1061 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1062 }
1063
1064 /**
1065 * This calculates the Hebrew year start, as days since 1 September.
1066 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1067 * Used for Hebrew date.
1068 */
1069 private static function hebrewYearStart( $year ) {
1070 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1071 $b = intval( ( $year - 1 ) % 4 );
1072 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1073 if( $m < 0 ) {
1074 $m--;
1075 }
1076 $Mar = intval( $m );
1077 if( $m < 0 ) {
1078 $m++;
1079 }
1080 $m -= $Mar;
1081
1082 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1083 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1084 $Mar++;
1085 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1086 $Mar += 2;
1087 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1088 $Mar++;
1089 }
1090
1091 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1092 return $Mar;
1093 }
1094
1095 /**
1096 * Algorithm to convert Gregorian dates to Thai solar dates.
1097 *
1098 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1099 *
1100 * @param $ts String: 14-character timestamp
1101 * @return array converted year, month, day
1102 */
1103 private static function tsToThai( $ts ) {
1104 $gy = substr( $ts, 0, 4 );
1105 $gm = substr( $ts, 4, 2 );
1106 $gd = substr( $ts, 6, 2 );
1107
1108 # Add 543 years to the Gregorian calendar
1109 # Months and days are identical
1110 $gy_thai = $gy + 543;
1111
1112 return array( $gy_thai, $gm, $gd );
1113 }
1114
1115
1116 /**
1117 * Roman number formatting up to 3000
1118 */
1119 static function romanNumeral( $num ) {
1120 static $table = array(
1121 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1122 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1123 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1124 array( '', 'M', 'MM', 'MMM' )
1125 );
1126
1127 $num = intval( $num );
1128 if ( $num > 3000 || $num <= 0 ) {
1129 return $num;
1130 }
1131
1132 $s = '';
1133 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1134 if ( $num >= $pow10 ) {
1135 $s .= $table[$i][floor($num / $pow10)];
1136 }
1137 $num = $num % $pow10;
1138 }
1139 return $s;
1140 }
1141
1142 /**
1143 * Hebrew Gematria number formatting up to 9999
1144 */
1145 static function hebrewNumeral( $num ) {
1146 static $table = array(
1147 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1148 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1149 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1150 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1151 );
1152
1153 $num = intval( $num );
1154 if ( $num > 9999 || $num <= 0 ) {
1155 return $num;
1156 }
1157
1158 $s = '';
1159 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1160 if ( $num >= $pow10 ) {
1161 if ( $num == 15 || $num == 16 ) {
1162 $s .= $table[0][9] . $table[0][$num - 9];
1163 $num = 0;
1164 } else {
1165 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1166 if( $pow10 == 1000 ) {
1167 $s .= "'";
1168 }
1169 }
1170 }
1171 $num = $num % $pow10;
1172 }
1173 if( strlen( $s ) == 2 ) {
1174 $str = $s . "'";
1175 } else {
1176 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1177 $str .= substr( $s, strlen( $s ) - 2, 2 );
1178 }
1179 $start = substr( $str, 0, strlen( $str ) - 2 );
1180 $end = substr( $str, strlen( $str ) - 2 );
1181 switch( $end ) {
1182 case 'כ':
1183 $str = $start . 'ך';
1184 break;
1185 case 'מ':
1186 $str = $start . 'ם';
1187 break;
1188 case 'נ':
1189 $str = $start . 'ן';
1190 break;
1191 case 'פ':
1192 $str = $start . 'ף';
1193 break;
1194 case 'צ':
1195 $str = $start . 'ץ';
1196 break;
1197 }
1198 return $str;
1199 }
1200
1201 /**
1202 * This is meant to be used by time(), date(), and timeanddate() to get
1203 * the date preference they're supposed to use, it should be used in
1204 * all children.
1205 *
1206 *<code>
1207 * function timeanddate([...], $format = true) {
1208 * $datePreference = $this->dateFormat($format);
1209 * [...]
1210 * }
1211 *</code>
1212 *
1213 * @param $usePrefs Mixed: if true, the user's preference is used
1214 * if false, the site/language default is used
1215 * if int/string, assumed to be a format.
1216 * @return string
1217 */
1218 function dateFormat( $usePrefs = true ) {
1219 global $wgUser;
1220
1221 if( is_bool( $usePrefs ) ) {
1222 if( $usePrefs ) {
1223 $datePreference = $wgUser->getDatePreference();
1224 } else {
1225 $options = User::getDefaultOptions();
1226 $datePreference = (string)$options['date'];
1227 }
1228 } else {
1229 $datePreference = (string)$usePrefs;
1230 }
1231
1232 // return int
1233 if( $datePreference == '' ) {
1234 return 'default';
1235 }
1236
1237 return $datePreference;
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 date( $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 date"] ) ) {
1258 $pref = $this->defaultDateFormat;
1259 }
1260 return $this->sprintfDate( $this->dateFormats["$pref date"], $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: true to use user's date format preference
1269 * @param $timecorrection String: the time offset as returned by
1270 * validateTimeZone() in Special:Preferences
1271 * @return string
1272 */
1273 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1274 $this->load();
1275 if ( $adj ) {
1276 $ts = $this->userAdjust( $ts, $timecorrection );
1277 }
1278
1279 $pref = $this->dateFormat( $format );
1280 if( $pref == 'default' || !isset( $this->dateFormats["$pref time"] ) ) {
1281 $pref = $this->defaultDateFormat;
1282 }
1283 return $this->sprintfDate( $this->dateFormats["$pref time"], $ts );
1284 }
1285
1286 /**
1287 * @param $ts Mixed: the time format which needs to be turned into a
1288 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1289 * @param $adj Bool: whether to adjust the time output according to the
1290 * user configured offset ($timecorrection)
1291 * @param $format Mixed: what format to return, if it's false output the
1292 * default one (default true)
1293 * @param $timecorrection String: the time offset as returned by
1294 * validateTimeZone() in Special:Preferences
1295 * @return string
1296 */
1297 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1298 $this->load();
1299
1300 $ts = wfTimestamp( TS_MW, $ts );
1301
1302 if ( $adj ) {
1303 $ts = $this->userAdjust( $ts, $timecorrection );
1304 }
1305
1306 $pref = $this->dateFormat( $format );
1307 if( $pref == 'default' || !isset( $this->dateFormats["$pref both"] ) ) {
1308 $pref = $this->defaultDateFormat;
1309 }
1310
1311 return $this->sprintfDate( $this->dateFormats["$pref both"], $ts );
1312 }
1313
1314 function getMessage( $key ) {
1315 $this->load();
1316 return isset( $this->messages[$key] ) ? $this->messages[$key] : null;
1317 }
1318
1319 function getAllMessages() {
1320 $this->load();
1321 return $this->messages;
1322 }
1323
1324 function iconv( $in, $out, $string ) {
1325 # For most languages, this is a wrapper for iconv
1326 return iconv( $in, $out . '//IGNORE', $string );
1327 }
1328
1329 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1330 function ucwordbreaksCallbackAscii($matches){
1331 return $this->ucfirst($matches[1]);
1332 }
1333
1334 function ucwordbreaksCallbackMB($matches){
1335 return mb_strtoupper($matches[0]);
1336 }
1337
1338 function ucCallback($matches){
1339 list( $wikiUpperChars ) = self::getCaseMaps();
1340 return strtr( $matches[1], $wikiUpperChars );
1341 }
1342
1343 function lcCallback($matches){
1344 list( , $wikiLowerChars ) = self::getCaseMaps();
1345 return strtr( $matches[1], $wikiLowerChars );
1346 }
1347
1348 function ucwordsCallbackMB($matches){
1349 return mb_strtoupper($matches[0]);
1350 }
1351
1352 function ucwordsCallbackWiki($matches){
1353 list( $wikiUpperChars ) = self::getCaseMaps();
1354 return strtr( $matches[0], $wikiUpperChars );
1355 }
1356
1357 function ucfirst( $str ) {
1358 if ( empty($str) ) return $str;
1359 if ( ord($str[0]) < 128 ) return ucfirst($str);
1360 else return self::uc($str,true); // fall back to more complex logic in case of multibyte strings
1361 }
1362
1363 function uc( $str, $first = false ) {
1364 if ( function_exists( 'mb_strtoupper' ) ) {
1365 if ( $first ) {
1366 if ( self::isMultibyte( $str ) ) {
1367 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1368 } else {
1369 return ucfirst( $str );
1370 }
1371 } else {
1372 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1373 }
1374 } else {
1375 if ( self::isMultibyte( $str ) ) {
1376 list( $wikiUpperChars ) = $this->getCaseMaps();
1377 $x = $first ? '^' : '';
1378 return preg_replace_callback(
1379 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1380 array($this,"ucCallback"),
1381 $str
1382 );
1383 } else {
1384 return $first ? ucfirst( $str ) : strtoupper( $str );
1385 }
1386 }
1387 }
1388
1389 function lcfirst( $str ) {
1390 if ( empty($str) ) return $str;
1391 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1392 // editing string in place = cool
1393 $str[0]=strtolower($str[0]);
1394 return $str;
1395 }
1396 else return self::lc( $str, true );
1397 }
1398
1399 function lc( $str, $first = false ) {
1400 if ( function_exists( 'mb_strtolower' ) )
1401 if ( $first )
1402 if ( self::isMultibyte( $str ) )
1403 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1404 else
1405 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1406 else
1407 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1408 else
1409 if ( self::isMultibyte( $str ) ) {
1410 list( , $wikiLowerChars ) = self::getCaseMaps();
1411 $x = $first ? '^' : '';
1412 return preg_replace_callback(
1413 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1414 array($this,"lcCallback"),
1415 $str
1416 );
1417 } else
1418 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1419 }
1420
1421 function isMultibyte( $str ) {
1422 return (bool)preg_match( '/[\x80-\xff]/', $str );
1423 }
1424
1425 function ucwords($str) {
1426 if ( self::isMultibyte( $str ) ) {
1427 $str = self::lc($str);
1428
1429 // regexp to find first letter in each word (i.e. after each space)
1430 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1431
1432 // function to use to capitalize a single char
1433 if ( function_exists( 'mb_strtoupper' ) )
1434 return preg_replace_callback(
1435 $replaceRegexp,
1436 array($this,"ucwordsCallbackMB"),
1437 $str
1438 );
1439 else
1440 return preg_replace_callback(
1441 $replaceRegexp,
1442 array($this,"ucwordsCallbackWiki"),
1443 $str
1444 );
1445 }
1446 else
1447 return ucwords( strtolower( $str ) );
1448 }
1449
1450 # capitalize words at word breaks
1451 function ucwordbreaks($str){
1452 if (self::isMultibyte( $str ) ) {
1453 $str = self::lc($str);
1454
1455 // since \b doesn't work for UTF-8, we explicitely define word break chars
1456 $breaks= "[ \-\(\)\}\{\.,\?!]";
1457
1458 // find first letter after word break
1459 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1460
1461 if ( function_exists( 'mb_strtoupper' ) )
1462 return preg_replace_callback(
1463 $replaceRegexp,
1464 array($this,"ucwordbreaksCallbackMB"),
1465 $str
1466 );
1467 else
1468 return preg_replace_callback(
1469 $replaceRegexp,
1470 array($this,"ucwordsCallbackWiki"),
1471 $str
1472 );
1473 }
1474 else
1475 return preg_replace_callback(
1476 '/\b([\w\x80-\xff]+)\b/',
1477 array($this,"ucwordbreaksCallbackAscii"),
1478 $str );
1479 }
1480
1481 /**
1482 * Return a case-folded representation of $s
1483 *
1484 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1485 * and $s2 are the same except for the case of their characters. It is not
1486 * necessary for the value returned to make sense when displayed.
1487 *
1488 * Do *not* perform any other normalisation in this function. If a caller
1489 * uses this function when it should be using a more general normalisation
1490 * function, then fix the caller.
1491 */
1492 function caseFold( $s ) {
1493 return $this->uc( $s );
1494 }
1495
1496 function checkTitleEncoding( $s ) {
1497 if( is_array( $s ) ) {
1498 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1499 }
1500 # Check for non-UTF-8 URLs
1501 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1502 if(!$ishigh) return $s;
1503
1504 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1505 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1506 if( $isutf8 ) return $s;
1507
1508 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1509 }
1510
1511 function fallback8bitEncoding() {
1512 $this->load();
1513 return $this->fallback8bitEncoding;
1514 }
1515
1516 /**
1517 * Some languages have special punctuation to strip out
1518 * or characters which need to be converted for MySQL's
1519 * indexing to grok it correctly. Make such changes here.
1520 *
1521 * @param $string String
1522 * @return String
1523 */
1524 function stripForSearch( $string ) {
1525 global $wgDBtype;
1526 if ( $wgDBtype != 'mysql' ) {
1527 return $string;
1528 }
1529
1530
1531 wfProfileIn( __METHOD__ );
1532
1533 // MySQL fulltext index doesn't grok utf-8, so we
1534 // need to fold cases and convert to hex
1535 $out = preg_replace_callback(
1536 "/([\\xc0-\\xff][\\x80-\\xbf]*)/",
1537 array( $this, 'stripForSearchCallback' ),
1538 $this->lc( $string ) );
1539
1540 // And to add insult to injury, the default indexing
1541 // ignores short words... Pad them so we can pass them
1542 // through without reconfiguring the server...
1543 $minLength = $this->minSearchLength();
1544 if( $minLength > 1 ) {
1545 $n = $minLength-1;
1546 $out = preg_replace(
1547 "/\b(\w{1,$n})\b/",
1548 "$1U800",
1549 $out );
1550 }
1551
1552 wfProfileOut( __METHOD__ );
1553 return $out;
1554 }
1555
1556 /**
1557 * Armor a case-folded UTF-8 string to get through MySQL's
1558 * fulltext search without being mucked up by funny charset
1559 * settings or anything else of the sort.
1560 */
1561 protected function stripForSearchCallback( $matches ) {
1562 return 'U8' . bin2hex( $matches[1] );
1563 }
1564
1565 /**
1566 * Check MySQL server's ft_min_word_len setting so we know
1567 * if we need to pad short words...
1568 */
1569 protected function minSearchLength() {
1570 if( !isset( $this->minSearchLength ) ) {
1571 $sql = "show global variables like 'ft\\_min\\_word\\_len'";
1572 $dbr = wfGetDB( DB_SLAVE );
1573 $result = $dbr->query( $sql );
1574 $row = $result->fetchObject();
1575 $result->free();
1576
1577 if( $row && $row->Variable_name == 'ft_min_word_len' ) {
1578 $this->minSearchLength = intval( $row->Value );
1579 } else {
1580 $this->minSearchLength = 0;
1581 }
1582 }
1583 return $this->minSearchLength;
1584 }
1585
1586 function convertForSearchResult( $termsArray ) {
1587 # some languages, e.g. Chinese, need to do a conversion
1588 # in order for search results to be displayed correctly
1589 return $termsArray;
1590 }
1591
1592 /**
1593 * Get the first character of a string.
1594 *
1595 * @param $s string
1596 * @return string
1597 */
1598 function firstChar( $s ) {
1599 $matches = array();
1600 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1601 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1602
1603 if ( isset( $matches[1] ) ) {
1604 if ( strlen( $matches[1] ) != 3 ) {
1605 return $matches[1];
1606 }
1607
1608 // Break down Hangul syllables to grab the first jamo
1609 $code = utf8ToCodepoint( $matches[1] );
1610 if ( $code < 0xac00 || 0xd7a4 <= $code) {
1611 return $matches[1];
1612 } elseif ( $code < 0xb098 ) {
1613 return "\xe3\x84\xb1";
1614 } elseif ( $code < 0xb2e4 ) {
1615 return "\xe3\x84\xb4";
1616 } elseif ( $code < 0xb77c ) {
1617 return "\xe3\x84\xb7";
1618 } elseif ( $code < 0xb9c8 ) {
1619 return "\xe3\x84\xb9";
1620 } elseif ( $code < 0xbc14 ) {
1621 return "\xe3\x85\x81";
1622 } elseif ( $code < 0xc0ac ) {
1623 return "\xe3\x85\x82";
1624 } elseif ( $code < 0xc544 ) {
1625 return "\xe3\x85\x85";
1626 } elseif ( $code < 0xc790 ) {
1627 return "\xe3\x85\x87";
1628 } elseif ( $code < 0xcc28 ) {
1629 return "\xe3\x85\x88";
1630 } elseif ( $code < 0xce74 ) {
1631 return "\xe3\x85\x8a";
1632 } elseif ( $code < 0xd0c0 ) {
1633 return "\xe3\x85\x8b";
1634 } elseif ( $code < 0xd30c ) {
1635 return "\xe3\x85\x8c";
1636 } elseif ( $code < 0xd558 ) {
1637 return "\xe3\x85\x8d";
1638 } else {
1639 return "\xe3\x85\x8e";
1640 }
1641 } else {
1642 return "";
1643 }
1644 }
1645
1646 function initEncoding() {
1647 # Some languages may have an alternate char encoding option
1648 # (Esperanto X-coding, Japanese furigana conversion, etc)
1649 # If this language is used as the primary content language,
1650 # an override to the defaults can be set here on startup.
1651 }
1652
1653 function recodeForEdit( $s ) {
1654 # For some languages we'll want to explicitly specify
1655 # which characters make it into the edit box raw
1656 # or are converted in some way or another.
1657 # Note that if wgOutputEncoding is different from
1658 # wgInputEncoding, this text will be further converted
1659 # to wgOutputEncoding.
1660 global $wgEditEncoding;
1661 if( $wgEditEncoding == '' or
1662 $wgEditEncoding == 'UTF-8' ) {
1663 return $s;
1664 } else {
1665 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1666 }
1667 }
1668
1669 function recodeInput( $s ) {
1670 # Take the previous into account.
1671 global $wgEditEncoding;
1672 if($wgEditEncoding != "") {
1673 $enc = $wgEditEncoding;
1674 } else {
1675 $enc = 'UTF-8';
1676 }
1677 if( $enc == 'UTF-8' ) {
1678 return $s;
1679 } else {
1680 return $this->iconv( $enc, 'UTF-8', $s );
1681 }
1682 }
1683
1684 /**
1685 * For right-to-left language support
1686 *
1687 * @return bool
1688 */
1689 function isRTL() {
1690 $this->load();
1691 return $this->rtl;
1692 }
1693
1694 /**
1695 * A hidden direction mark (LRM or RLM), depending on the language direction
1696 *
1697 * @return string
1698 */
1699 function getDirMark() {
1700 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1701 }
1702
1703 /**
1704 * An arrow, depending on the language direction
1705 *
1706 * @return string
1707 */
1708 function getArrow() {
1709 return $this->isRTL() ? '←' : '→';
1710 }
1711
1712 /**
1713 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1714 *
1715 * @return bool
1716 */
1717 function linkPrefixExtension() {
1718 $this->load();
1719 return $this->linkPrefixExtension;
1720 }
1721
1722 function &getMagicWords() {
1723 $this->load();
1724 return $this->magicWords;
1725 }
1726
1727 # Fill a MagicWord object with data from here
1728 function getMagic( &$mw ) {
1729 if ( !$this->mMagicHookDone ) {
1730 $this->mMagicHookDone = true;
1731 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1732 }
1733 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1734 $rawEntry = $this->mMagicExtensions[$mw->mId];
1735 } else {
1736 $magicWords =& $this->getMagicWords();
1737 if ( isset( $magicWords[$mw->mId] ) ) {
1738 $rawEntry = $magicWords[$mw->mId];
1739 } else {
1740 # Fall back to English if local list is incomplete
1741 $magicWords =& Language::getMagicWords();
1742 if ( !isset($magicWords[$mw->mId]) ) {
1743 throw new MWException("Magic word '{$mw->mId}' not found" );
1744 }
1745 $rawEntry = $magicWords[$mw->mId];
1746 }
1747 }
1748
1749 if( !is_array( $rawEntry ) ) {
1750 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1751 } else {
1752 $mw->mCaseSensitive = $rawEntry[0];
1753 $mw->mSynonyms = array_slice( $rawEntry, 1 );
1754 }
1755 }
1756
1757 /**
1758 * Add magic words to the extension array
1759 */
1760 function addMagicWordsByLang( $newWords ) {
1761 $code = $this->getCode();
1762 $fallbackChain = array();
1763 while ( $code && !in_array( $code, $fallbackChain ) ) {
1764 $fallbackChain[] = $code;
1765 $code = self::getFallbackFor( $code );
1766 }
1767 if ( !in_array( 'en', $fallbackChain ) ) {
1768 $fallbackChain[] = 'en';
1769 }
1770 $fallbackChain = array_reverse( $fallbackChain );
1771 foreach ( $fallbackChain as $code ) {
1772 if ( isset( $newWords[$code] ) ) {
1773 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1774 }
1775 }
1776 }
1777
1778 /**
1779 * Get special page names, as an associative array
1780 * case folded alias => real name
1781 */
1782 function getSpecialPageAliases() {
1783 $this->load();
1784
1785 // Cache aliases because it may be slow to load them
1786 if ( !isset( $this->mExtendedSpecialPageAliases ) ) {
1787
1788 // Initialise array
1789 $this->mExtendedSpecialPageAliases = $this->specialPageAliases;
1790
1791 global $wgExtensionAliasesFiles;
1792 foreach ( $wgExtensionAliasesFiles as $file ) {
1793
1794 // Fail fast
1795 if ( !file_exists($file) )
1796 throw new MWException( "Aliases file does not exist: $file" );
1797
1798 $aliases = array();
1799 require($file);
1800
1801 // Check the availability of aliases
1802 if ( !isset($aliases['en']) )
1803 throw new MWException( "Malformed aliases file: $file" );
1804
1805 // Merge all aliases in fallback chain
1806 $code = $this->getCode();
1807 do {
1808 if ( !isset($aliases[$code]) ) continue;
1809
1810 $aliases[$code] = $this->fixSpecialPageAliases( $aliases[$code] );
1811 /* Merge the aliases, THIS will break if there is special page name
1812 * which looks like a numerical key, thanks to PHP...
1813 * See the array_merge_recursive manual entry */
1814 $this->mExtendedSpecialPageAliases = array_merge_recursive(
1815 $this->mExtendedSpecialPageAliases, $aliases[$code] );
1816
1817 } while ( $code = self::getFallbackFor( $code ) );
1818 }
1819
1820 wfRunHooks( 'LanguageGetSpecialPageAliases',
1821 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1822 }
1823
1824 return $this->mExtendedSpecialPageAliases;
1825 }
1826
1827 /**
1828 * Function to fix special page aliases. Will convert the first letter to
1829 * upper case and spaces to underscores. Can be given a full aliases array,
1830 * in which case it will recursively fix all aliases.
1831 */
1832 public function fixSpecialPageAliases( $mixed ) {
1833 // Work recursively until in string level
1834 if ( is_array($mixed) ) {
1835 $callback = array( $this, 'fixSpecialPageAliases' );
1836 return array_map( $callback, $mixed );
1837 }
1838 return str_replace( ' ', '_', $this->ucfirst( $mixed ) );
1839 }
1840
1841 /**
1842 * Italic is unsuitable for some languages
1843 *
1844 * @param $text String: the text to be emphasized.
1845 * @return string
1846 */
1847 function emphasize( $text ) {
1848 return "<em>$text</em>";
1849 }
1850
1851 /**
1852 * Normally we output all numbers in plain en_US style, that is
1853 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1854 * point twohundredthirtyfive. However this is not sutable for all
1855 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1856 * Icelandic just want to use commas instead of dots, and dots instead
1857 * of commas like "293.291,235".
1858 *
1859 * An example of this function being called:
1860 * <code>
1861 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1862 * </code>
1863 *
1864 * See LanguageGu.php for the Gujarati implementation and
1865 * $separatorTransformTable on MessageIs.php for
1866 * the , => . and . => , implementation.
1867 *
1868 * @todo check if it's viable to use localeconv() for the decimal
1869 * separator thing.
1870 * @param $number Mixed: the string to be formatted, should be an integer
1871 * or a floating point number.
1872 * @param $nocommafy Bool: set to true for special numbers like dates
1873 * @return string
1874 */
1875 function formatNum( $number, $nocommafy = false ) {
1876 global $wgTranslateNumerals;
1877 if (!$nocommafy) {
1878 $number = $this->commafy($number);
1879 $s = $this->separatorTransformTable();
1880 if (!is_null($s)) { $number = strtr($number, $s); }
1881 }
1882
1883 if ($wgTranslateNumerals) {
1884 $s = $this->digitTransformTable();
1885 if (!is_null($s)) { $number = strtr($number, $s); }
1886 }
1887
1888 return $number;
1889 }
1890
1891 function parseFormattedNumber( $number ) {
1892 $s = $this->digitTransformTable();
1893 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1894
1895 $s = $this->separatorTransformTable();
1896 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1897
1898 $number = strtr( $number, array (',' => '') );
1899 return $number;
1900 }
1901
1902 /**
1903 * Adds commas to a given number
1904 *
1905 * @param $_ mixed
1906 * @return string
1907 */
1908 function commafy($_) {
1909 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1910 }
1911
1912 function digitTransformTable() {
1913 $this->load();
1914 return $this->digitTransformTable;
1915 }
1916
1917 function separatorTransformTable() {
1918 $this->load();
1919 return $this->separatorTransformTable;
1920 }
1921
1922
1923 /**
1924 * For the credit list in includes/Credits.php (action=credits)
1925 *
1926 * @param $l Array
1927 * @return string
1928 */
1929 function listToText( $l ) {
1930 $s = '';
1931 $m = count($l) - 1;
1932 for ($i = $m; $i >= 0; $i--) {
1933 if ($i == $m) {
1934 $s = $l[$i];
1935 } else if ($i == $m - 1) {
1936 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
1937 } else {
1938 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
1939 }
1940 }
1941 return $s;
1942 }
1943
1944 /**
1945 * Take a list of strings and build a locale-friendly comma-separated
1946 * list, using the local comma-separator message.
1947 * @param $list array of strings to put in a comma list
1948 * @return string
1949 */
1950 function commaList( $list, $forContent = false ) {
1951 return implode(
1952 $list,
1953 wfMsgExt( 'comma-separator', array( 'escapenoentities', 'language' => $this ) ) );
1954 }
1955
1956 /**
1957 * Same as commaList, but separate it with the pipe instead.
1958 * @param $list array of strings to put in a pipe list
1959 * @return string
1960 */
1961 function pipeList( $list ) {
1962 return implode(
1963 $list,
1964 wfMsgExt( 'pipe-separator', array( 'escapenoentities', 'language' => $this ) ) );
1965 }
1966
1967 /**
1968 * Truncate a string to a specified length in bytes, appending an optional
1969 * string (e.g. for ellipses)
1970 *
1971 * The database offers limited byte lengths for some columns in the database;
1972 * multi-byte character sets mean we need to ensure that only whole characters
1973 * are included, otherwise broken characters can be passed to the user
1974 *
1975 * If $length is negative, the string will be truncated from the beginning
1976 *
1977 * @param $string String to truncate
1978 * @param $length Int: maximum length (excluding ellipses)
1979 * @param $ellipsis String to append to the truncated text
1980 * @return string
1981 */
1982 function truncate( $string, $length, $ellipsis = "" ) {
1983 if( $length == 0 ) {
1984 return $ellipsis;
1985 }
1986 if ( strlen( $string ) <= abs( $length ) ) {
1987 return $string;
1988 }
1989 if( $length > 0 ) {
1990 $string = substr( $string, 0, $length );
1991 $char = ord( $string[strlen( $string ) - 1] );
1992 $m = array();
1993 if ($char >= 0xc0) {
1994 # We got the first byte only of a multibyte char; remove it.
1995 $string = substr( $string, 0, -1 );
1996 } elseif( $char >= 0x80 &&
1997 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
1998 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
1999 # We chopped in the middle of a character; remove it
2000 $string = $m[1];
2001 }
2002 return $string . $ellipsis;
2003 } else {
2004 $string = substr( $string, $length );
2005 $char = ord( $string[0] );
2006 if( $char >= 0x80 && $char < 0xc0 ) {
2007 # We chopped in the middle of a character; remove the whole thing
2008 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2009 }
2010 return $ellipsis . $string;
2011 }
2012 }
2013
2014 /**
2015 * Grammatical transformations, needed for inflected languages
2016 * Invoked by putting {{grammar:case|word}} in a message
2017 *
2018 * @param $word string
2019 * @param $case string
2020 * @return string
2021 */
2022 function convertGrammar( $word, $case ) {
2023 global $wgGrammarForms;
2024 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
2025 return $wgGrammarForms[$this->getCode()][$case][$word];
2026 }
2027 return $word;
2028 }
2029
2030 /**
2031 * Plural form transformations, needed for some languages.
2032 * For example, there are 3 form of plural in Russian and Polish,
2033 * depending on "count mod 10". See [[w:Plural]]
2034 * For English it is pretty simple.
2035 *
2036 * Invoked by putting {{plural:count|wordform1|wordform2}}
2037 * or {{plural:count|wordform1|wordform2|wordform3}}
2038 *
2039 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2040 *
2041 * @param $count Integer: non-localized number
2042 * @param $forms Array: different plural forms
2043 * @return string Correct form of plural for $count in this language
2044 */
2045 function convertPlural( $count, $forms ) {
2046 if ( !count($forms) ) { return ''; }
2047 $forms = $this->preConvertPlural( $forms, 2 );
2048
2049 return ( $count == 1 ) ? $forms[0] : $forms[1];
2050 }
2051
2052 /**
2053 * Checks that convertPlural was given an array and pads it to requested
2054 * amound of forms by copying the last one.
2055 *
2056 * @param $count Integer: How many forms should there be at least
2057 * @param $forms Array of forms given to convertPlural
2058 * @return array Padded array of forms or an exception if not an array
2059 */
2060 protected function preConvertPlural( /* Array */ $forms, $count ) {
2061 while ( count($forms) < $count ) {
2062 $forms[] = $forms[count($forms)-1];
2063 }
2064 return $forms;
2065 }
2066
2067 /**
2068 * For translaing of expiry times
2069 * @param $str String: the validated block time in English
2070 * @return Somehow translated block time
2071 * @see LanguageFi.php for example implementation
2072 */
2073 function translateBlockExpiry( $str ) {
2074
2075 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2076
2077 if ( $scBlockExpiryOptions == '-') {
2078 return $str;
2079 }
2080
2081 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2082 if ( strpos($option, ":") === false )
2083 continue;
2084 list($show, $value) = explode(":", $option);
2085 if ( strcmp ( $str, $value) == 0 ) {
2086 return htmlspecialchars( trim( $show ) );
2087 }
2088 }
2089
2090 return $str;
2091 }
2092
2093 /**
2094 * languages like Chinese need to be segmented in order for the diff
2095 * to be of any use
2096 *
2097 * @param $text String
2098 * @return String
2099 */
2100 function segmentForDiff( $text ) {
2101 return $text;
2102 }
2103
2104 /**
2105 * and unsegment to show the result
2106 *
2107 * @param $text String
2108 * @return String
2109 */
2110 function unsegmentForDiff( $text ) {
2111 return $text;
2112 }
2113
2114 # convert text to different variants of a language.
2115 function convert( $text, $isTitle = false) {
2116 return $this->mConverter->convert($text, $isTitle);
2117 }
2118
2119 # Convert text from within Parser
2120 function parserConvert( $text, &$parser ) {
2121 return $this->mConverter->parserConvert( $text, $parser );
2122 }
2123
2124 # Check if this is a language with variants
2125 function hasVariants(){
2126 return sizeof($this->getVariants())>1;
2127 }
2128
2129 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2130 function armourMath($text){
2131 return $this->mConverter->armourMath($text);
2132 }
2133
2134
2135 /**
2136 * Perform output conversion on a string, and encode for safe HTML output.
2137 * @param $text String
2138 * @param $isTitle Bool -- wtf?
2139 * @return string
2140 * @todo this should get integrated somewhere sane
2141 */
2142 function convertHtml( $text, $isTitle = false ) {
2143 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2144 }
2145
2146 function convertCategoryKey( $key ) {
2147 return $this->mConverter->convertCategoryKey( $key );
2148 }
2149
2150 /**
2151 * get the list of variants supported by this langauge
2152 * see sample implementation in LanguageZh.php
2153 *
2154 * @return array an array of language codes
2155 */
2156 function getVariants() {
2157 return $this->mConverter->getVariants();
2158 }
2159
2160
2161 function getPreferredVariant( $fromUser = true ) {
2162 return $this->mConverter->getPreferredVariant( $fromUser );
2163 }
2164
2165 /**
2166 * if a language supports multiple variants, it is
2167 * possible that non-existing link in one variant
2168 * actually exists in another variant. this function
2169 * tries to find it. See e.g. LanguageZh.php
2170 *
2171 * @param $link String: the name of the link
2172 * @param $nt Mixed: the title object of the link
2173 * @return null the input parameters may be modified upon return
2174 */
2175 function findVariantLink( &$link, &$nt, $forTemplate = false ) {
2176 $this->mConverter->findVariantLink($link, $nt, $forTemplate );
2177 }
2178
2179 /**
2180 * If a language supports multiple variants, converts text
2181 * into an array of all possible variants of the text:
2182 * 'variant' => text in that variant
2183 */
2184
2185 function convertLinkToAllVariants($text){
2186 return $this->mConverter->convertLinkToAllVariants($text);
2187 }
2188
2189
2190 /**
2191 * returns language specific options used by User::getPageRenderHash()
2192 * for example, the preferred language variant
2193 *
2194 * @return string
2195 */
2196 function getExtraHashOptions() {
2197 return $this->mConverter->getExtraHashOptions();
2198 }
2199
2200 /**
2201 * for languages that support multiple variants, the title of an
2202 * article may be displayed differently in different variants. this
2203 * function returns the apporiate title defined in the body of the article.
2204 *
2205 * @return string
2206 */
2207 function getParsedTitle() {
2208 return $this->mConverter->getParsedTitle();
2209 }
2210
2211 /**
2212 * Enclose a string with the "no conversion" tag. This is used by
2213 * various functions in the Parser
2214 *
2215 * @param $text String: text to be tagged for no conversion
2216 * @param $noParse
2217 * @return string the tagged text
2218 */
2219 function markNoConversion( $text, $noParse=false ) {
2220 return $this->mConverter->markNoConversion( $text, $noParse );
2221 }
2222
2223 /**
2224 * A regular expression to match legal word-trailing characters
2225 * which should be merged onto a link of the form [[foo]]bar.
2226 *
2227 * @return string
2228 */
2229 function linkTrail() {
2230 $this->load();
2231 return $this->linkTrail;
2232 }
2233
2234 function getLangObj() {
2235 return $this;
2236 }
2237
2238 /**
2239 * Get the RFC 3066 code for this language object
2240 */
2241 function getCode() {
2242 return $this->mCode;
2243 }
2244
2245 function setCode( $code ) {
2246 $this->mCode = $code;
2247 }
2248
2249 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2250 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2251 }
2252
2253 static function getMessagesFileName( $code ) {
2254 global $IP;
2255 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2256 }
2257
2258 static function getClassFileName( $code ) {
2259 global $IP;
2260 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2261 }
2262
2263 static function getLocalisationArray( $code, $disableCache = false ) {
2264 self::loadLocalisation( $code, $disableCache );
2265 return self::$mLocalisationCache[$code];
2266 }
2267
2268 /**
2269 * Load localisation data for a given code into the static cache
2270 *
2271 * @return array Dependencies, map of filenames to mtimes
2272 */
2273 static function loadLocalisation( $code, $disableCache = false ) {
2274 static $recursionGuard = array();
2275 global $wgMemc, $wgCheckSerialized;
2276
2277 if ( !$code ) {
2278 throw new MWException( "Invalid language code requested" );
2279 }
2280
2281 if ( !$disableCache ) {
2282 # Try the per-process cache
2283 if ( isset( self::$mLocalisationCache[$code] ) ) {
2284 return self::$mLocalisationCache[$code]['deps'];
2285 }
2286
2287 wfProfileIn( __METHOD__ );
2288
2289 # Try the serialized directory
2290 $cache = wfGetPrecompiledData( self::getFileName( "Messages", $code, '.ser' ) );
2291 if ( $cache ) {
2292 if ( $wgCheckSerialized && self::isLocalisationOutOfDate( $cache ) ) {
2293 $cache = false;
2294 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2295 } else {
2296 self::$mLocalisationCache[$code] = $cache;
2297 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2298 wfProfileOut( __METHOD__ );
2299 return self::$mLocalisationCache[$code]['deps'];
2300 }
2301 }
2302
2303 # Try the global cache
2304 $memcKey = wfMemcKey('localisation', $code );
2305 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2306 $cache = $wgMemc->get( $memcKey );
2307 if ( $cache ) {
2308 if ( self::isLocalisationOutOfDate( $cache ) ) {
2309 $wgMemc->delete( $memcKey );
2310 $wgMemc->delete( $fbMemcKey );
2311 $cache = false;
2312 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2313 } else {
2314 self::$mLocalisationCache[$code] = $cache;
2315 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2316 wfProfileOut( __METHOD__ );
2317 return $cache['deps'];
2318 }
2319 }
2320 } else {
2321 wfProfileIn( __METHOD__ );
2322 }
2323
2324 # Default fallback, may be overridden when the messages file is included
2325 if ( $code != 'en' ) {
2326 $fallback = 'en';
2327 } else {
2328 $fallback = false;
2329 }
2330
2331 # Load the primary localisation from the source file
2332 $filename = self::getMessagesFileName( $code );
2333 if ( !file_exists( $filename ) ) {
2334 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2335 $cache = compact( self::$mLocalisationKeys ); // Set correct fallback
2336 $deps = array();
2337 } else {
2338 $deps = array( $filename => filemtime( $filename ) );
2339 require( $filename );
2340 $cache = compact( self::$mLocalisationKeys );
2341 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2342 }
2343
2344 if ( !empty( $fallback ) ) {
2345 # Load the fallback localisation, with a circular reference guard
2346 if ( isset( $recursionGuard[$code] ) ) {
2347 throw new MWException( "Error: Circular fallback reference in language code $code" );
2348 }
2349 $recursionGuard[$code] = true;
2350 $newDeps = self::loadLocalisation( $fallback, $disableCache );
2351 unset( $recursionGuard[$code] );
2352
2353 $secondary = self::$mLocalisationCache[$fallback];
2354 $deps = array_merge( $deps, $newDeps );
2355
2356 # Merge the fallback localisation with the current localisation
2357 foreach ( self::$mLocalisationKeys as $key ) {
2358 if ( isset( $cache[$key] ) ) {
2359 if ( isset( $secondary[$key] ) ) {
2360 if ( in_array( $key, self::$mMergeableMapKeys ) ) {
2361 $cache[$key] = $cache[$key] + $secondary[$key];
2362 } elseif ( in_array( $key, self::$mMergeableListKeys ) ) {
2363 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2364 } elseif ( in_array( $key, self::$mMergeableAliasListKeys ) ) {
2365 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2366 }
2367 }
2368 } else {
2369 $cache[$key] = $secondary[$key];
2370 }
2371 }
2372
2373 # Merge bookstore lists if requested
2374 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2375 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2376 }
2377 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2378 unset( $cache['bookstoreList']['inherit'] );
2379 }
2380 }
2381
2382 # Add dependencies to the cache entry
2383 $cache['deps'] = $deps;
2384
2385 # Replace spaces with underscores in namespace names
2386 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2387
2388 # And do the same for specialpage aliases. $page is an array.
2389 foreach ( $cache['specialPageAliases'] as &$page ) {
2390 $page = str_replace( ' ', '_', $page );
2391 }
2392 # Decouple the reference to prevent accidental damage
2393 unset($page);
2394
2395 # Save to both caches
2396 self::$mLocalisationCache[$code] = $cache;
2397 if ( !$disableCache ) {
2398 $wgMemc->set( $memcKey, $cache );
2399 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2400 }
2401
2402 wfProfileOut( __METHOD__ );
2403 return $deps;
2404 }
2405
2406 /**
2407 * Test if a given localisation cache is out of date with respect to the
2408 * source Messages files. This is done automatically for the global cache
2409 * in $wgMemc, but is only done on certain occasions for the serialized
2410 * data file.
2411 *
2412 * @param $cache mixed Either a language code or a cache array
2413 */
2414 static function isLocalisationOutOfDate( $cache ) {
2415 if ( !is_array( $cache ) ) {
2416 self::loadLocalisation( $cache );
2417 $cache = self::$mLocalisationCache[$cache];
2418 }
2419 $expired = false;
2420 foreach ( $cache['deps'] as $file => $mtime ) {
2421 if ( !file_exists( $file ) || filemtime( $file ) > $mtime ) {
2422 $expired = true;
2423 break;
2424 }
2425 }
2426 return $expired;
2427 }
2428
2429 /**
2430 * Get the fallback for a given language
2431 */
2432 static function getFallbackFor( $code ) {
2433 // Shortcut
2434 if ( $code === 'en' ) return false;
2435
2436 // Local cache
2437 static $cache = array();
2438 // Quick return
2439 if ( isset($cache[$code]) ) return $cache[$code];
2440
2441 // Try memcache
2442 global $wgMemc;
2443 $memcKey = wfMemcKey( 'fallback', $code );
2444 $fbcode = $wgMemc->get( $memcKey );
2445
2446 if ( is_string($fbcode) ) {
2447 // False is stored as a string to detect failures in memcache properly
2448 if ( $fbcode === '' ) $fbcode = false;
2449
2450 // Update local cache and return
2451 $cache[$code] = $fbcode;
2452 return $fbcode;
2453 }
2454
2455 // Nothing in caches, load and and update both caches
2456 self::loadLocalisation( $code );
2457 $fbcode = self::$mLocalisationCache[$code]['fallback'];
2458
2459 $cache[$code] = $fbcode;
2460 $wgMemc->set( $memcKey, (string) $fbcode );
2461
2462 return $fbcode;
2463 }
2464
2465 /**
2466 * Get all messages for a given language
2467 */
2468 static function getMessagesFor( $code ) {
2469 self::loadLocalisation( $code );
2470 return self::$mLocalisationCache[$code]['messages'];
2471 }
2472
2473 /**
2474 * Get a message for a given language
2475 */
2476 static function getMessageFor( $key, $code ) {
2477 self::loadLocalisation( $code );
2478 return isset( self::$mLocalisationCache[$code]['messages'][$key] ) ? self::$mLocalisationCache[$code]['messages'][$key] : null;
2479 }
2480
2481 /**
2482 * Load localisation data for this object
2483 */
2484 function load() {
2485 if ( !$this->mLoaded ) {
2486 self::loadLocalisation( $this->getCode() );
2487 $cache =& self::$mLocalisationCache[$this->getCode()];
2488 foreach ( self::$mLocalisationKeys as $key ) {
2489 $this->$key = $cache[$key];
2490 }
2491 $this->mLoaded = true;
2492
2493 $this->fixUpSettings();
2494 }
2495 }
2496
2497 /**
2498 * Do any necessary post-cache-load settings adjustment
2499 */
2500 function fixUpSettings() {
2501 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2502 $wgNamespaceAliases, $wgAmericanDates;
2503 wfProfileIn( __METHOD__ );
2504 if ( $wgExtraNamespaces ) {
2505 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
2506 }
2507
2508 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
2509 if ( $wgMetaNamespaceTalk ) {
2510 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
2511 } else {
2512 $talk = $this->namespaceNames[NS_PROJECT_TALK];
2513 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2514
2515 # Allow grammar transformations
2516 # Allowing full message-style parsing would make simple requests
2517 # such as action=raw much more expensive than they need to be.
2518 # This will hopefully cover most cases.
2519 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2520 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2521 $talk = str_replace( ' ', '_', $talk );
2522 $this->namespaceNames[NS_PROJECT_TALK] = $talk;
2523 }
2524
2525 # The above mixing may leave namespaces out of canonical order.
2526 # Re-order by namespace ID number...
2527 ksort( $this->namespaceNames );
2528
2529 # Put namespace names and aliases into a hashtable.
2530 # If this is too slow, then we should arrange it so that it is done
2531 # before caching. The catch is that at pre-cache time, the above
2532 # class-specific fixup hasn't been done.
2533 $this->mNamespaceIds = array();
2534 foreach ( $this->namespaceNames as $index => $name ) {
2535 $this->mNamespaceIds[$this->lc($name)] = $index;
2536 }
2537 if ( $this->namespaceAliases ) {
2538 foreach ( $this->namespaceAliases as $name => $index ) {
2539 $this->mNamespaceIds[$this->lc($name)] = $index;
2540 }
2541 }
2542 if ( $wgNamespaceAliases ) {
2543 foreach ( $wgNamespaceAliases as $name => $index ) {
2544 $this->mNamespaceIds[$this->lc($name)] = $index;
2545 }
2546 }
2547
2548 if ( $this->defaultDateFormat == 'dmy or mdy' ) {
2549 $this->defaultDateFormat = $wgAmericanDates ? 'mdy' : 'dmy';
2550 }
2551 wfProfileOut( __METHOD__ );
2552 }
2553
2554 function replaceGrammarInNamespace( $m ) {
2555 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2556 }
2557
2558 static function getCaseMaps() {
2559 static $wikiUpperChars, $wikiLowerChars;
2560 if ( isset( $wikiUpperChars ) ) {
2561 return array( $wikiUpperChars, $wikiLowerChars );
2562 }
2563
2564 wfProfileIn( __METHOD__ );
2565 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2566 if ( $arr === false ) {
2567 throw new MWException(
2568 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2569 }
2570 extract( $arr );
2571 wfProfileOut( __METHOD__ );
2572 return array( $wikiUpperChars, $wikiLowerChars );
2573 }
2574
2575 function formatTimePeriod( $seconds ) {
2576 if ( $seconds < 10 ) {
2577 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2578 } elseif ( $seconds < 60 ) {
2579 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2580 } elseif ( $seconds < 3600 ) {
2581 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2582 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2583 } else {
2584 $hours = floor( $seconds / 3600 );
2585 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2586 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2587 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2588 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2589 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2590 }
2591 }
2592
2593 function formatBitrate( $bps ) {
2594 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2595 if ( $bps <= 0 ) {
2596 return $this->formatNum( $bps ) . $units[0];
2597 }
2598 $unitIndex = floor( log10( $bps ) / 3 );
2599 $mantissa = $bps / pow( 1000, $unitIndex );
2600 if ( $mantissa < 10 ) {
2601 $mantissa = round( $mantissa, 1 );
2602 } else {
2603 $mantissa = round( $mantissa );
2604 }
2605 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2606 }
2607
2608 /**
2609 * Format a size in bytes for output, using an appropriate
2610 * unit (B, KB, MB or GB) according to the magnitude in question
2611 *
2612 * @param $size Size to format
2613 * @return string Plain text (not HTML)
2614 */
2615 function formatSize( $size ) {
2616 // For small sizes no decimal places necessary
2617 $round = 0;
2618 if( $size > 1024 ) {
2619 $size = $size / 1024;
2620 if( $size > 1024 ) {
2621 $size = $size / 1024;
2622 // For MB and bigger two decimal places are smarter
2623 $round = 2;
2624 if( $size > 1024 ) {
2625 $size = $size / 1024;
2626 $msg = 'size-gigabytes';
2627 } else {
2628 $msg = 'size-megabytes';
2629 }
2630 } else {
2631 $msg = 'size-kilobytes';
2632 }
2633 } else {
2634 $msg = 'size-bytes';
2635 }
2636 $size = round( $size, $round );
2637 $text = $this->getMessageFromDB( $msg );
2638 return str_replace( '$1', $this->formatNum( $size ), $text );
2639 }
2640 }
Wiklou, le Wiki du Wiklou