* Cleanup for r37375. Split duplicated code into function and really fix it
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * @defgroup Language Language
4 *
5 * @file
6 * @ingroup Language
7 */
8
9 if( !defined( 'MEDIAWIKI' ) ) {
10 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
11 exit( 1 );
12 }
13
14 # Read language names
15 global $wgLanguageNames;
16 require_once( dirname(__FILE__) . '/Names.php' ) ;
17
18 global $wgInputEncoding, $wgOutputEncoding;
19
20 /**
21 * These are always UTF-8, they exist only for backwards compatibility
22 */
23 $wgInputEncoding = "UTF-8";
24 $wgOutputEncoding = "UTF-8";
25
26 if( function_exists( 'mb_strtoupper' ) ) {
27 mb_internal_encoding('UTF-8');
28 }
29
30 /**
31 * a fake language converter
32 *
33 * @ingroup Language
34 */
35 class FakeConverter {
36 var $mLang;
37 function FakeConverter($langobj) {$this->mLang = $langobj;}
38 function convert($t, $i) {return $t;}
39 function parserConvert($t, $p) {return $t;}
40 function getVariants() { return array( $this->mLang->getCode() ); }
41 function getPreferredVariant() {return $this->mLang->getCode(); }
42 function findVariantLink(&$l, &$n) {}
43 function getExtraHashOptions() {return '';}
44 function getParsedTitle() {return '';}
45 function markNoConversion($text, $noParse=false) {return $text;}
46 function convertCategoryKey( $key ) {return $key; }
47 function convertLinkToAllVariants($text){ return array( $this->mLang->getCode() => $text); }
48 function armourMath($text){ return $text; }
49 }
50
51 /**
52 * Internationalisation code
53 * @ingroup Language
54 */
55 class Language {
56 var $mConverter, $mVariants, $mCode, $mLoaded = false;
57 var $mMagicExtensions = array(), $mMagicHookDone = false;
58
59 static public $mLocalisationKeys = array( 'fallback', 'namespaceNames',
60 'skinNames', 'mathNames',
61 'bookstoreList', 'magicWords', 'messages', 'rtl', 'digitTransformTable',
62 'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
63 'defaultUserOptionOverrides', 'linkTrail', 'namespaceAliases',
64 'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
65 'defaultDateFormat', 'extraUserToggles', 'specialPageAliases',
66 '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 * Hook which will be called if this is the content language.
181 * Descendants can use this to register hook functions or modify globals
182 */
183 function initContLang() {}
184
185 /**
186 * @deprecated Use User::getDefaultOptions()
187 * @return array
188 */
189 function getDefaultUserOptions() {
190 wfDeprecated( __METHOD__ );
191 return User::getDefaultOptions();
192 }
193
194 function getFallbackLanguageCode() {
195 return self::getFallbackFor( $this->mCode );
196 }
197
198 /**
199 * Exports $wgBookstoreListEn
200 * @return array
201 */
202 function getBookstoreList() {
203 $this->load();
204 return $this->bookstoreList;
205 }
206
207 /**
208 * @return array
209 */
210 function getNamespaces() {
211 $this->load();
212 return $this->namespaceNames;
213 }
214
215 /**
216 * A convenience function that returns the same thing as
217 * getNamespaces() except with the array values changed to ' '
218 * where it found '_', useful for producing output to be displayed
219 * e.g. in <select> forms.
220 *
221 * @return array
222 */
223 function getFormattedNamespaces() {
224 $ns = $this->getNamespaces();
225 foreach($ns as $k => $v) {
226 $ns[$k] = strtr($v, '_', ' ');
227 }
228 return $ns;
229 }
230
231 /**
232 * Get a namespace value by key
233 * <code>
234 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
235 * echo $mw_ns; // prints 'MediaWiki'
236 * </code>
237 *
238 * @param $index Int: the array key of the namespace to return
239 * @return mixed, string if the namespace value exists, otherwise false
240 */
241 function getNsText( $index ) {
242 $ns = $this->getNamespaces();
243 return isset( $ns[$index] ) ? $ns[$index] : false;
244 }
245
246 /**
247 * A convenience function that returns the same thing as
248 * getNsText() except with '_' changed to ' ', useful for
249 * producing output.
250 *
251 * @return array
252 */
253 function getFormattedNsText( $index ) {
254 $ns = $this->getNsText( $index );
255 return strtr($ns, '_', ' ');
256 }
257
258 /**
259 * Get a namespace key by value, case insensitive.
260 * Only matches namespace names for the current language, not the
261 * canonical ones defined in Namespace.php.
262 *
263 * @param $text String
264 * @return mixed An integer if $text is a valid value otherwise false
265 */
266 function getLocalNsIndex( $text ) {
267 $this->load();
268 $lctext = $this->lc($text);
269 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
270 }
271
272 /**
273 * Get a namespace key by value, case insensitive. Canonical namespace
274 * names override custom ones defined for the current language.
275 *
276 * @param $text String
277 * @return mixed An integer if $text is a valid value otherwise false
278 */
279 function getNsIndex( $text ) {
280 $this->load();
281 $lctext = $this->lc($text);
282 if( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
283 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
284 }
285
286 /**
287 * short names for language variants used for language conversion links.
288 *
289 * @param $code String
290 * @return string
291 */
292 function getVariantname( $code ) {
293 return $this->getMessageFromDB( "variantname-$code" );
294 }
295
296 function specialPage( $name ) {
297 $aliases = $this->getSpecialPageAliases();
298 if ( isset( $aliases[$name][0] ) ) {
299 $name = $aliases[$name][0];
300 }
301 return $this->getNsText(NS_SPECIAL) . ':' . $name;
302 }
303
304 function getQuickbarSettings() {
305 return array(
306 $this->getMessage( 'qbsettings-none' ),
307 $this->getMessage( 'qbsettings-fixedleft' ),
308 $this->getMessage( 'qbsettings-fixedright' ),
309 $this->getMessage( 'qbsettings-floatingleft' ),
310 $this->getMessage( 'qbsettings-floatingright' )
311 );
312 }
313
314 function getSkinNames() {
315 $this->load();
316 return $this->skinNames;
317 }
318
319 function getMathNames() {
320 $this->load();
321 return $this->mathNames;
322 }
323
324 function getDatePreferences() {
325 $this->load();
326 return $this->datePreferences;
327 }
328
329 function getDateFormats() {
330 $this->load();
331 return $this->dateFormats;
332 }
333
334 function getDefaultDateFormat() {
335 $this->load();
336 return $this->defaultDateFormat;
337 }
338
339 function getDatePreferenceMigrationMap() {
340 $this->load();
341 return $this->datePreferenceMigrationMap;
342 }
343
344 function getImageFile( $image ) {
345 $this->load();
346 return $this->imageFiles[$image];
347 }
348
349 function getDefaultUserOptionOverrides() {
350 $this->load();
351 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
352 if (is_array($this->defaultUserOptionOverrides)) {
353 return $this->defaultUserOptionOverrides;
354 } else {
355 return array();
356 }
357 }
358
359 function getExtraUserToggles() {
360 $this->load();
361 return $this->extraUserToggles;
362 }
363
364 function getUserToggle( $tog ) {
365 return $this->getMessageFromDB( "tog-$tog" );
366 }
367
368 /**
369 * Get language names, indexed by code.
370 * If $customisedOnly is true, only returns codes with a messages file
371 */
372 public static function getLanguageNames( $customisedOnly = false ) {
373 global $wgLanguageNames, $wgExtraLanguageNames;
374 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
375 if ( !$customisedOnly ) {
376 return $allNames;
377 }
378
379 global $IP;
380 $names = array();
381 $dir = opendir( "$IP/languages/messages" );
382 while( false !== ( $file = readdir( $dir ) ) ) {
383 $m = array();
384 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
385 $code = str_replace( '_', '-', strtolower( $m[1] ) );
386 if ( isset( $allNames[$code] ) ) {
387 $names[$code] = $allNames[$code];
388 }
389 }
390 }
391 closedir( $dir );
392 return $names;
393 }
394
395 /**
396 * Ugly hack to get a message maybe from the MediaWiki namespace, if this
397 * language object is the content or user language.
398 */
399 function getMessageFromDB( $msg ) {
400 global $wgContLang, $wgLang;
401 if ( $wgContLang->getCode() == $this->getCode() ) {
402 # Content language
403 return wfMsgForContent( $msg );
404 } elseif ( $wgLang->getCode() == $this->getCode() ) {
405 # User language
406 return wfMsg( $msg );
407 } else {
408 # Neither, get from localisation
409 return $this->getMessage( $msg );
410 }
411 }
412
413 function getLanguageName( $code ) {
414 $names = self::getLanguageNames();
415 if ( !array_key_exists( $code, $names ) ) {
416 return '';
417 }
418 return $names[$code];
419 }
420
421 function getMonthName( $key ) {
422 return $this->getMessageFromDB( self::$mMonthMsgs[$key-1] );
423 }
424
425 function getMonthNameGen( $key ) {
426 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key-1] );
427 }
428
429 function getMonthAbbreviation( $key ) {
430 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key-1] );
431 }
432
433 function getWeekdayName( $key ) {
434 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key-1] );
435 }
436
437 function getWeekdayAbbreviation( $key ) {
438 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key-1] );
439 }
440
441 function getIranianCalendarMonthName( $key ) {
442 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key-1] );
443 }
444
445 function getHebrewCalendarMonthName( $key ) {
446 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key-1] );
447 }
448
449 function getHebrewCalendarMonthNameGen( $key ) {
450 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key-1] );
451 }
452
453 function getHijriCalendarMonthName( $key ) {
454 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key-1] );
455 }
456
457 /**
458 * Used by date() and time() to adjust the time output.
459 *
460 * @param $ts Int the time in date('YmdHis') format
461 * @param $tz Mixed: adjust the time by this amount (default false, mean we
462 * get user timecorrection setting)
463 * @return int
464 */
465 function userAdjust( $ts, $tz = false ) {
466 global $wgUser, $wgLocalTZoffset;
467
468 if (!$tz) {
469 $tz = $wgUser->getOption( 'timecorrection' );
470 }
471
472 # minutes and hours differences:
473 $minDiff = 0;
474 $hrDiff = 0;
475
476 if ( $tz === '' ) {
477 # Global offset in minutes.
478 if( isset($wgLocalTZoffset) ) {
479 if( $wgLocalTZoffset >= 0 ) {
480 $hrDiff = floor($wgLocalTZoffset / 60);
481 } else {
482 $hrDiff = ceil($wgLocalTZoffset / 60);
483 }
484 $minDiff = $wgLocalTZoffset % 60;
485 }
486 } elseif ( strpos( $tz, ':' ) !== false ) {
487 $tzArray = explode( ':', $tz );
488 $hrDiff = intval($tzArray[0]);
489 $minDiff = intval($hrDiff < 0 ? -$tzArray[1] : $tzArray[1]);
490 } else {
491 $hrDiff = intval( $tz );
492 }
493
494 # No difference ? Return time unchanged
495 if ( 0 == $hrDiff && 0 == $minDiff ) { return $ts; }
496
497 wfSuppressWarnings(); // E_STRICT system time bitching
498 # Generate an adjusted date
499 $t = mktime( (
500 (int)substr( $ts, 8, 2) ) + $hrDiff, # Hours
501 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
502 (int)substr( $ts, 12, 2 ), # Seconds
503 (int)substr( $ts, 4, 2 ), # Month
504 (int)substr( $ts, 6, 2 ), # Day
505 (int)substr( $ts, 0, 4 ) ); #Year
506
507 $date = date( 'YmdHis', $t );
508 wfRestoreWarnings();
509
510 return $date;
511 }
512
513 /**
514 * This is a workalike of PHP's date() function, but with better
515 * internationalisation, a reduced set of format characters, and a better
516 * escaping format.
517 *
518 * Supported format characters are dDjlNwzWFmMntLYyaAgGhHiscrU. See the
519 * PHP manual for definitions. There are a number of extensions, which
520 * start with "x":
521 *
522 * xn Do not translate digits of the next numeric format character
523 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
524 * xr Use roman numerals for the next numeric format character
525 * xh Use hebrew numerals for the next numeric format character
526 * xx Literal x
527 * xg Genitive month name
528 *
529 * xij j (day number) in Iranian calendar
530 * xiF F (month name) in Iranian calendar
531 * xin n (month number) in Iranian calendar
532 * xiY Y (full year) in Iranian calendar
533 *
534 * xjj j (day number) in Hebrew calendar
535 * xjF F (month name) in Hebrew calendar
536 * xjt t (days in month) in Hebrew calendar
537 * xjx xg (genitive month name) in Hebrew calendar
538 * xjn n (month number) in Hebrew calendar
539 * xjY Y (full year) in Hebrew calendar
540 *
541 * xmj j (day number) in Hijri calendar
542 * xmF F (month name) in Hijri calendar
543 * xmn n (month number) in Hijri calendar
544 * xmY Y (full year) in Hijri calendar
545 *
546 * xkY Y (full year) in Thai solar calendar. Months and days are
547 * identical to the Gregorian calendar
548 *
549 * Characters enclosed in double quotes will be considered literal (with
550 * the quotes themselves removed). Unmatched quotes will be considered
551 * literal quotes. Example:
552 *
553 * "The month is" F => The month is January
554 * i's" => 20'11"
555 *
556 * Backslash escaping is also supported.
557 *
558 * Input timestamp is assumed to be pre-normalized to the desired local
559 * time zone, if any.
560 *
561 * @param $format String
562 * @param $ts String: 14-character timestamp
563 * YYYYMMDDHHMMSS
564 * 01234567890123
565 */
566 function sprintfDate( $format, $ts ) {
567 $s = '';
568 $raw = false;
569 $roman = false;
570 $hebrewNum = false;
571 $unix = false;
572 $rawToggle = false;
573 $iranian = false;
574 $hebrew = false;
575 $hijri = false;
576 $thai = false;
577 for ( $p = 0; $p < strlen( $format ); $p++ ) {
578 $num = false;
579 $code = $format[$p];
580 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
581 $code .= $format[++$p];
582 }
583
584 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' ) && $p < strlen( $format ) - 1 ) {
585 $code .= $format[++$p];
586 }
587
588 switch ( $code ) {
589 case 'xx':
590 $s .= 'x';
591 break;
592 case 'xn':
593 $raw = true;
594 break;
595 case 'xN':
596 $rawToggle = !$rawToggle;
597 break;
598 case 'xr':
599 $roman = true;
600 break;
601 case 'xh':
602 $hebrewNum = true;
603 break;
604 case 'xg':
605 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
606 break;
607 case 'xjx':
608 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
609 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
610 break;
611 case 'd':
612 $num = substr( $ts, 6, 2 );
613 break;
614 case 'D':
615 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
616 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
617 break;
618 case 'j':
619 $num = intval( substr( $ts, 6, 2 ) );
620 break;
621 case 'xij':
622 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
623 $num = $iranian[2];
624 break;
625 case 'xmj':
626 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
627 $num = $hijri[2];
628 break;
629 case 'xjj':
630 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
631 $num = $hebrew[2];
632 break;
633 case 'l':
634 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
635 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
636 break;
637 case 'N':
638 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
639 $w = gmdate( 'w', $unix );
640 $num = $w ? $w : 7;
641 break;
642 case 'w':
643 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
644 $num = gmdate( 'w', $unix );
645 break;
646 case 'z':
647 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
648 $num = gmdate( 'z', $unix );
649 break;
650 case 'W':
651 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
652 $num = gmdate( 'W', $unix );
653 break;
654 case 'F':
655 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
656 break;
657 case 'xiF':
658 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
659 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
660 break;
661 case 'xmF':
662 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
663 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
664 break;
665 case 'xjF':
666 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
667 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
668 break;
669 case 'm':
670 $num = substr( $ts, 4, 2 );
671 break;
672 case 'M':
673 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
674 break;
675 case 'n':
676 $num = intval( substr( $ts, 4, 2 ) );
677 break;
678 case 'xin':
679 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
680 $num = $iranian[1];
681 break;
682 case 'xmn':
683 if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
684 $num = $hijri[1];
685 break;
686 case 'xjn':
687 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
688 $num = $hebrew[1];
689 break;
690 case 't':
691 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
692 $num = gmdate( 't', $unix );
693 break;
694 case 'xjt':
695 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
696 $num = $hebrew[3];
697 break;
698 case 'L':
699 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
700 $num = gmdate( 'L', $unix );
701 break;
702 case 'Y':
703 $num = substr( $ts, 0, 4 );
704 break;
705 case 'xiY':
706 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
707 $num = $iranian[0];
708 break;
709 case 'xmY':
710 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
711 $num = $hijri[0];
712 break;
713 case 'xjY':
714 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
715 $num = $hebrew[0];
716 break;
717 case 'xkY':
718 if ( !$thai ) $thai = self::tsToThai( $ts );
719 $num = $thai[0];
720 break;
721 case 'y':
722 $num = substr( $ts, 2, 2 );
723 break;
724 case 'a':
725 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
726 break;
727 case 'A':
728 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
729 break;
730 case 'g':
731 $h = substr( $ts, 8, 2 );
732 $num = $h % 12 ? $h % 12 : 12;
733 break;
734 case 'G':
735 $num = intval( substr( $ts, 8, 2 ) );
736 break;
737 case 'h':
738 $h = substr( $ts, 8, 2 );
739 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
740 break;
741 case 'H':
742 $num = substr( $ts, 8, 2 );
743 break;
744 case 'i':
745 $num = substr( $ts, 10, 2 );
746 break;
747 case 's':
748 $num = substr( $ts, 12, 2 );
749 break;
750 case 'c':
751 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
752 $s .= gmdate( 'c', $unix );
753 break;
754 case 'r':
755 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
756 $s .= gmdate( 'r', $unix );
757 break;
758 case 'U':
759 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
760 $num = $unix;
761 break;
762 case '\\':
763 # Backslash escaping
764 if ( $p < strlen( $format ) - 1 ) {
765 $s .= $format[++$p];
766 } else {
767 $s .= '\\';
768 }
769 break;
770 case '"':
771 # Quoted literal
772 if ( $p < strlen( $format ) - 1 ) {
773 $endQuote = strpos( $format, '"', $p + 1 );
774 if ( $endQuote === false ) {
775 # No terminating quote, assume literal "
776 $s .= '"';
777 } else {
778 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
779 $p = $endQuote;
780 }
781 } else {
782 # Quote at end of string, assume literal "
783 $s .= '"';
784 }
785 break;
786 default:
787 $s .= $format[$p];
788 }
789 if ( $num !== false ) {
790 if ( $rawToggle || $raw ) {
791 $s .= $num;
792 $raw = false;
793 } elseif ( $roman ) {
794 $s .= self::romanNumeral( $num );
795 $roman = false;
796 } elseif( $hebrewNum ) {
797 $s .= self::hebrewNumeral( $num );
798 $hebrewNum = false;
799 } else {
800 $s .= $this->formatNum( $num, true );
801 }
802 $num = false;
803 }
804 }
805 return $s;
806 }
807
808 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
809 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
810 /**
811 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
812 * Gregorian dates to Iranian dates. Originally written in C, it
813 * is released under the terms of GNU Lesser General Public
814 * License. Conversion to PHP was performed by Niklas Laxström.
815 *
816 * Link: http://www.farsiweb.info/jalali/jalali.c
817 */
818 private static function tsToIranian( $ts ) {
819 $gy = substr( $ts, 0, 4 ) -1600;
820 $gm = substr( $ts, 4, 2 ) -1;
821 $gd = substr( $ts, 6, 2 ) -1;
822
823 # Days passed from the beginning (including leap years)
824 $gDayNo = 365*$gy
825 + floor(($gy+3) / 4)
826 - floor(($gy+99) / 100)
827 + floor(($gy+399) / 400);
828
829
830 // Add days of the past months of this year
831 for( $i = 0; $i < $gm; $i++ ) {
832 $gDayNo += self::$GREG_DAYS[$i];
833 }
834
835 // Leap years
836 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
837 $gDayNo++;
838 }
839
840 // Days passed in current month
841 $gDayNo += $gd;
842
843 $jDayNo = $gDayNo - 79;
844
845 $jNp = floor($jDayNo / 12053);
846 $jDayNo %= 12053;
847
848 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
849 $jDayNo %= 1461;
850
851 if ( $jDayNo >= 366 ) {
852 $jy += floor(($jDayNo-1)/365);
853 $jDayNo = floor(($jDayNo-1)%365);
854 }
855
856 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
857 $jDayNo -= self::$IRANIAN_DAYS[$i];
858 }
859
860 $jm= $i+1;
861 $jd= $jDayNo+1;
862
863 return array($jy, $jm, $jd);
864 }
865 /**
866 * Converting Gregorian dates to Hijri dates.
867 *
868 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
869 *
870 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
871 */
872 private static function tsToHijri ( $ts ) {
873 $year = substr( $ts, 0, 4 );
874 $month = substr( $ts, 4, 2 );
875 $day = substr( $ts, 6, 2 );
876
877 $zyr = $year;
878 $zd=$day;
879 $zm=$month;
880 $zy=$zyr;
881
882
883
884 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
885 {
886
887
888 $zjd=(int)((1461*($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;
889 }
890 else
891 {
892 $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
893 }
894
895 $zl=$zjd-1948440+10632;
896 $zn=(int)(($zl-1)/10631);
897 $zl=$zl-10631*$zn+354;
898 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
899 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
900 $zm=(int)((24*$zl)/709);
901 $zd=$zl-(int)((709*$zm)/24);
902 $zy=30*$zn+$zj-30;
903
904 return array ($zy, $zm, $zd);
905 }
906
907 /**
908 * Converting Gregorian dates to Hebrew dates.
909 *
910 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
911 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
912 * to translate the relevant functions into PHP and release them under
913 * GNU GPL.
914 */
915 private static function tsToHebrew( $ts ) {
916 # Parse date
917 $year = substr( $ts, 0, 4 );
918 $month = substr( $ts, 4, 2 );
919 $day = substr( $ts, 6, 2 );
920
921 # Calculate Hebrew year
922 $hebrewYear = $year + 3760;
923
924 # Month number when September = 1, August = 12
925 $month += 4;
926 if( $month > 12 ) {
927 # Next year
928 $month -= 12;
929 $year++;
930 $hebrewYear++;
931 }
932
933 # Calculate day of year from 1 September
934 $dayOfYear = $day;
935 for( $i = 1; $i < $month; $i++ ) {
936 if( $i == 6 ) {
937 # February
938 $dayOfYear += 28;
939 # Check if the year is leap
940 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
941 $dayOfYear++;
942 }
943 } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
944 $dayOfYear += 30;
945 } else {
946 $dayOfYear += 31;
947 }
948 }
949
950 # Calculate the start of the Hebrew year
951 $start = self::hebrewYearStart( $hebrewYear );
952
953 # Calculate next year's start
954 if( $dayOfYear <= $start ) {
955 # Day is before the start of the year - it is the previous year
956 # Next year's start
957 $nextStart = $start;
958 # Previous year
959 $year--;
960 $hebrewYear--;
961 # Add days since previous year's 1 September
962 $dayOfYear += 365;
963 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
964 # Leap year
965 $dayOfYear++;
966 }
967 # Start of the new (previous) year
968 $start = self::hebrewYearStart( $hebrewYear );
969 } else {
970 # Next year's start
971 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
972 }
973
974 # Calculate Hebrew day of year
975 $hebrewDayOfYear = $dayOfYear - $start;
976
977 # Difference between year's days
978 $diff = $nextStart - $start;
979 # Add 12 (or 13 for leap years) days to ignore the difference between
980 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
981 # difference is only about the year type
982 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
983 $diff += 13;
984 } else {
985 $diff += 12;
986 }
987
988 # Check the year pattern, and is leap year
989 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
990 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
991 # and non-leap years
992 $yearPattern = $diff % 30;
993 # Check if leap year
994 $isLeap = $diff >= 30;
995
996 # Calculate day in the month from number of day in the Hebrew year
997 # Don't check Adar - if the day is not in Adar, we will stop before;
998 # if it is in Adar, we will use it to check if it is Adar I or Adar II
999 $hebrewDay = $hebrewDayOfYear;
1000 $hebrewMonth = 1;
1001 $days = 0;
1002 while( $hebrewMonth <= 12 ) {
1003 # Calculate days in this month
1004 if( $isLeap && $hebrewMonth == 6 ) {
1005 # Adar in a leap year
1006 if( $isLeap ) {
1007 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1008 $days = 30;
1009 if( $hebrewDay <= $days ) {
1010 # Day in Adar I
1011 $hebrewMonth = 13;
1012 } else {
1013 # Subtract the days of Adar I
1014 $hebrewDay -= $days;
1015 # Try Adar II
1016 $days = 29;
1017 if( $hebrewDay <= $days ) {
1018 # Day in Adar II
1019 $hebrewMonth = 14;
1020 }
1021 }
1022 }
1023 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1024 # Cheshvan in a complete year (otherwise as the rule below)
1025 $days = 30;
1026 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1027 # Kislev in an incomplete year (otherwise as the rule below)
1028 $days = 29;
1029 } else {
1030 # Odd months have 30 days, even have 29
1031 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1032 }
1033 if( $hebrewDay <= $days ) {
1034 # In the current month
1035 break;
1036 } else {
1037 # Subtract the days of the current month
1038 $hebrewDay -= $days;
1039 # Try in the next month
1040 $hebrewMonth++;
1041 }
1042 }
1043
1044 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1045 }
1046
1047 /**
1048 * This calculates the Hebrew year start, as days since 1 September.
1049 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1050 * Used for Hebrew date.
1051 */
1052 private static function hebrewYearStart( $year ) {
1053 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1054 $b = intval( ( $year - 1 ) % 4 );
1055 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1056 if( $m < 0 ) {
1057 $m--;
1058 }
1059 $Mar = intval( $m );
1060 if( $m < 0 ) {
1061 $m++;
1062 }
1063 $m -= $Mar;
1064
1065 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1066 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1067 $Mar++;
1068 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1069 $Mar += 2;
1070 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1071 $Mar++;
1072 }
1073
1074 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1075 return $Mar;
1076 }
1077
1078 /**
1079 * Algorithm to convert Gregorian dates to Thai solar dates.
1080 *
1081 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1082 *
1083 * @param $ts String: 14-character timestamp
1084 * @return array converted year, month, day
1085 */
1086 private static function tsToThai( $ts ) {
1087 $gy = substr( $ts, 0, 4 );
1088 $gm = substr( $ts, 4, 2 );
1089 $gd = substr( $ts, 6, 2 );
1090
1091 # Add 543 years to the Gregorian calendar
1092 # Months and days are identical
1093 $gy_thai = $gy + 543;
1094
1095 return array( $gy_thai, $gm, $gd );
1096 }
1097
1098
1099 /**
1100 * Roman number formatting up to 3000
1101 */
1102 static function romanNumeral( $num ) {
1103 static $table = array(
1104 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1105 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1106 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1107 array( '', 'M', 'MM', 'MMM' )
1108 );
1109
1110 $num = intval( $num );
1111 if ( $num > 3000 || $num <= 0 ) {
1112 return $num;
1113 }
1114
1115 $s = '';
1116 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1117 if ( $num >= $pow10 ) {
1118 $s .= $table[$i][floor($num / $pow10)];
1119 }
1120 $num = $num % $pow10;
1121 }
1122 return $s;
1123 }
1124
1125 /**
1126 * Hebrew Gematria number formatting up to 9999
1127 */
1128 static function hebrewNumeral( $num ) {
1129 static $table = array(
1130 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1131 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1132 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1133 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1134 );
1135
1136 $num = intval( $num );
1137 if ( $num > 9999 || $num <= 0 ) {
1138 return $num;
1139 }
1140
1141 $s = '';
1142 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1143 if ( $num >= $pow10 ) {
1144 if ( $num == 15 || $num == 16 ) {
1145 $s .= $table[0][9] . $table[0][$num - 9];
1146 $num = 0;
1147 } else {
1148 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1149 if( $pow10 == 1000 ) {
1150 $s .= "'";
1151 }
1152 }
1153 }
1154 $num = $num % $pow10;
1155 }
1156 if( strlen( $s ) == 2 ) {
1157 $str = $s . "'";
1158 } else {
1159 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1160 $str .= substr( $s, strlen( $s ) - 2, 2 );
1161 }
1162 $start = substr( $str, 0, strlen( $str ) - 2 );
1163 $end = substr( $str, strlen( $str ) - 2 );
1164 switch( $end ) {
1165 case 'כ':
1166 $str = $start . 'ך';
1167 break;
1168 case 'מ':
1169 $str = $start . 'ם';
1170 break;
1171 case 'נ':
1172 $str = $start . 'ן';
1173 break;
1174 case 'פ':
1175 $str = $start . 'ף';
1176 break;
1177 case 'צ':
1178 $str = $start . 'ץ';
1179 break;
1180 }
1181 return $str;
1182 }
1183
1184 /**
1185 * This is meant to be used by time(), date(), and timeanddate() to get
1186 * the date preference they're supposed to use, it should be used in
1187 * all children.
1188 *
1189 *<code>
1190 * function timeanddate([...], $format = true) {
1191 * $datePreference = $this->dateFormat($format);
1192 * [...]
1193 * }
1194 *</code>
1195 *
1196 * @param $usePrefs Mixed: if true, the user's preference is used
1197 * if false, the site/language default is used
1198 * if int/string, assumed to be a format.
1199 * @return string
1200 */
1201 function dateFormat( $usePrefs = true ) {
1202 global $wgUser;
1203
1204 if( is_bool( $usePrefs ) ) {
1205 if( $usePrefs ) {
1206 $datePreference = $wgUser->getDatePreference();
1207 } else {
1208 $options = User::getDefaultOptions();
1209 $datePreference = (string)$options['date'];
1210 }
1211 } else {
1212 $datePreference = (string)$usePrefs;
1213 }
1214
1215 // return int
1216 if( $datePreference == '' ) {
1217 return 'default';
1218 }
1219
1220 return $datePreference;
1221 }
1222
1223 /**
1224 * @param $ts Mixed: the time format which needs to be turned into a
1225 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1226 * @param $adj Bool: whether to adjust the time output according to the
1227 * user configured offset ($timecorrection)
1228 * @param $format Mixed: true to use user's date format preference
1229 * @param $timecorrection String: the time offset as returned by
1230 * validateTimeZone() in Special:Preferences
1231 * @return string
1232 */
1233 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1234 $this->load();
1235 if ( $adj ) {
1236 $ts = $this->userAdjust( $ts, $timecorrection );
1237 }
1238
1239 $pref = $this->dateFormat( $format );
1240 if( $pref == 'default' || !isset( $this->dateFormats["$pref date"] ) ) {
1241 $pref = $this->defaultDateFormat;
1242 }
1243 return $this->sprintfDate( $this->dateFormats["$pref date"], $ts );
1244 }
1245
1246 /**
1247 * @param $ts Mixed: the time format which needs to be turned into a
1248 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1249 * @param $adj Bool: whether to adjust the time output according to the
1250 * user configured offset ($timecorrection)
1251 * @param $format Mixed: true to use user's date format preference
1252 * @param $timecorrection String: the time offset as returned by
1253 * validateTimeZone() in Special:Preferences
1254 * @return string
1255 */
1256 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1257 $this->load();
1258 if ( $adj ) {
1259 $ts = $this->userAdjust( $ts, $timecorrection );
1260 }
1261
1262 $pref = $this->dateFormat( $format );
1263 if( $pref == 'default' || !isset( $this->dateFormats["$pref time"] ) ) {
1264 $pref = $this->defaultDateFormat;
1265 }
1266 return $this->sprintfDate( $this->dateFormats["$pref time"], $ts );
1267 }
1268
1269 /**
1270 * @param $ts Mixed: the time format which needs to be turned into a
1271 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1272 * @param $adj Bool: whether to adjust the time output according to the
1273 * user configured offset ($timecorrection)
1274 * @param $format Mixed: what format to return, if it's false output the
1275 * default one (default true)
1276 * @param $timecorrection String: the time offset as returned by
1277 * validateTimeZone() in Special:Preferences
1278 * @return string
1279 */
1280 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1281 $this->load();
1282
1283 $ts = wfTimestamp( TS_MW, $ts );
1284
1285 if ( $adj ) {
1286 $ts = $this->userAdjust( $ts, $timecorrection );
1287 }
1288
1289 $pref = $this->dateFormat( $format );
1290 if( $pref == 'default' || !isset( $this->dateFormats["$pref both"] ) ) {
1291 $pref = $this->defaultDateFormat;
1292 }
1293
1294 return $this->sprintfDate( $this->dateFormats["$pref both"], $ts );
1295 }
1296
1297 function getMessage( $key ) {
1298 $this->load();
1299 return isset( $this->messages[$key] ) ? $this->messages[$key] : null;
1300 }
1301
1302 function getAllMessages() {
1303 $this->load();
1304 return $this->messages;
1305 }
1306
1307 function iconv( $in, $out, $string ) {
1308 # For most languages, this is a wrapper for iconv
1309 return iconv( $in, $out . '//IGNORE', $string );
1310 }
1311
1312 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1313 function ucwordbreaksCallbackAscii($matches){
1314 return $this->ucfirst($matches[1]);
1315 }
1316
1317 function ucwordbreaksCallbackMB($matches){
1318 return mb_strtoupper($matches[0]);
1319 }
1320
1321 function ucCallback($matches){
1322 list( $wikiUpperChars ) = self::getCaseMaps();
1323 return strtr( $matches[1], $wikiUpperChars );
1324 }
1325
1326 function lcCallback($matches){
1327 list( , $wikiLowerChars ) = self::getCaseMaps();
1328 return strtr( $matches[1], $wikiLowerChars );
1329 }
1330
1331 function ucwordsCallbackMB($matches){
1332 return mb_strtoupper($matches[0]);
1333 }
1334
1335 function ucwordsCallbackWiki($matches){
1336 list( $wikiUpperChars ) = self::getCaseMaps();
1337 return strtr( $matches[0], $wikiUpperChars );
1338 }
1339
1340 function ucfirst( $str ) {
1341 if ( empty($str) ) return $str;
1342 if ( ord($str[0]) < 128 ) return ucfirst($str);
1343 else return self::uc($str,true); // fall back to more complex logic in case of multibyte strings
1344 }
1345
1346 function uc( $str, $first = false ) {
1347 if ( function_exists( 'mb_strtoupper' ) ) {
1348 if ( $first ) {
1349 if ( self::isMultibyte( $str ) ) {
1350 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1351 } else {
1352 return ucfirst( $str );
1353 }
1354 } else {
1355 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1356 }
1357 } else {
1358 if ( self::isMultibyte( $str ) ) {
1359 list( $wikiUpperChars ) = $this->getCaseMaps();
1360 $x = $first ? '^' : '';
1361 return preg_replace_callback(
1362 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1363 array($this,"ucCallback"),
1364 $str
1365 );
1366 } else {
1367 return $first ? ucfirst( $str ) : strtoupper( $str );
1368 }
1369 }
1370 }
1371
1372 function lcfirst( $str ) {
1373 if ( empty($str) ) return $str;
1374 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1375 // editing string in place = cool
1376 $str[0]=strtolower($str[0]);
1377 return $str;
1378 }
1379 else return self::lc( $str, true );
1380 }
1381
1382 function lc( $str, $first = false ) {
1383 if ( function_exists( 'mb_strtolower' ) )
1384 if ( $first )
1385 if ( self::isMultibyte( $str ) )
1386 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1387 else
1388 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1389 else
1390 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1391 else
1392 if ( self::isMultibyte( $str ) ) {
1393 list( , $wikiLowerChars ) = self::getCaseMaps();
1394 $x = $first ? '^' : '';
1395 return preg_replace_callback(
1396 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1397 array($this,"lcCallback"),
1398 $str
1399 );
1400 } else
1401 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1402 }
1403
1404 function isMultibyte( $str ) {
1405 return (bool)preg_match( '/[\x80-\xff]/', $str );
1406 }
1407
1408 function ucwords($str) {
1409 if ( self::isMultibyte( $str ) ) {
1410 $str = self::lc($str);
1411
1412 // regexp to find first letter in each word (i.e. after each space)
1413 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1414
1415 // function to use to capitalize a single char
1416 if ( function_exists( 'mb_strtoupper' ) )
1417 return preg_replace_callback(
1418 $replaceRegexp,
1419 array($this,"ucwordsCallbackMB"),
1420 $str
1421 );
1422 else
1423 return preg_replace_callback(
1424 $replaceRegexp,
1425 array($this,"ucwordsCallbackWiki"),
1426 $str
1427 );
1428 }
1429 else
1430 return ucwords( strtolower( $str ) );
1431 }
1432
1433 # capitalize words at word breaks
1434 function ucwordbreaks($str){
1435 if (self::isMultibyte( $str ) ) {
1436 $str = self::lc($str);
1437
1438 // since \b doesn't work for UTF-8, we explicitely define word break chars
1439 $breaks= "[ \-\(\)\}\{\.,\?!]";
1440
1441 // find first letter after word break
1442 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1443
1444 if ( function_exists( 'mb_strtoupper' ) )
1445 return preg_replace_callback(
1446 $replaceRegexp,
1447 array($this,"ucwordbreaksCallbackMB"),
1448 $str
1449 );
1450 else
1451 return preg_replace_callback(
1452 $replaceRegexp,
1453 array($this,"ucwordsCallbackWiki"),
1454 $str
1455 );
1456 }
1457 else
1458 return preg_replace_callback(
1459 '/\b([\w\x80-\xff]+)\b/',
1460 array($this,"ucwordbreaksCallbackAscii"),
1461 $str );
1462 }
1463
1464 /**
1465 * Return a case-folded representation of $s
1466 *
1467 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1468 * and $s2 are the same except for the case of their characters. It is not
1469 * necessary for the value returned to make sense when displayed.
1470 *
1471 * Do *not* perform any other normalisation in this function. If a caller
1472 * uses this function when it should be using a more general normalisation
1473 * function, then fix the caller.
1474 */
1475 function caseFold( $s ) {
1476 return $this->uc( $s );
1477 }
1478
1479 function checkTitleEncoding( $s ) {
1480 if( is_array( $s ) ) {
1481 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1482 }
1483 # Check for non-UTF-8 URLs
1484 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1485 if(!$ishigh) return $s;
1486
1487 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1488 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1489 if( $isutf8 ) return $s;
1490
1491 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1492 }
1493
1494 function fallback8bitEncoding() {
1495 $this->load();
1496 return $this->fallback8bitEncoding;
1497 }
1498
1499 /**
1500 * Some languages have special punctuation to strip out
1501 * or characters which need to be converted for MySQL's
1502 * indexing to grok it correctly. Make such changes here.
1503 *
1504 * @param $string String
1505 * @return String
1506 */
1507 function stripForSearch( $string ) {
1508 global $wgDBtype;
1509 if ( $wgDBtype != 'mysql' ) {
1510 return $string;
1511 }
1512
1513 # MySQL fulltext index doesn't grok utf-8, so we
1514 # need to fold cases and convert to hex
1515
1516 wfProfileIn( __METHOD__ );
1517 if( function_exists( 'mb_strtolower' ) ) {
1518 $out = preg_replace(
1519 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1520 "'U8' . bin2hex( \"$1\" )",
1521 mb_strtolower( $string ) );
1522 } else {
1523 list( , $wikiLowerChars ) = self::getCaseMaps();
1524 $out = preg_replace(
1525 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1526 "'U8' . bin2hex( strtr( \"\$1\", \$wikiLowerChars ) )",
1527 $string );
1528 }
1529 wfProfileOut( __METHOD__ );
1530 return $out;
1531 }
1532
1533 function convertForSearchResult( $termsArray ) {
1534 # some languages, e.g. Chinese, need to do a conversion
1535 # in order for search results to be displayed correctly
1536 return $termsArray;
1537 }
1538
1539 /**
1540 * Get the first character of a string.
1541 *
1542 * @param $s string
1543 * @return string
1544 */
1545 function firstChar( $s ) {
1546 $matches = array();
1547 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1548 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1549
1550 if ( isset( $matches[1] ) ) {
1551 if ( strlen( $matches[1] ) != 3 ) {
1552 return $matches[1];
1553 }
1554
1555 // Break down Hangul syllables to grab the first jamo
1556 $code = utf8ToCodepoint( $matches[1] );
1557 if ( $code < 0xac00 || 0xd7a4 <= $code) {
1558 return $matches[1];
1559 } elseif ( $code < 0xb098 ) {
1560 return "\xe3\x84\xb1";
1561 } elseif ( $code < 0xb2e4 ) {
1562 return "\xe3\x84\xb4";
1563 } elseif ( $code < 0xb77c ) {
1564 return "\xe3\x84\xb7";
1565 } elseif ( $code < 0xb9c8 ) {
1566 return "\xe3\x84\xb9";
1567 } elseif ( $code < 0xbc14 ) {
1568 return "\xe3\x85\x81";
1569 } elseif ( $code < 0xc0ac ) {
1570 return "\xe3\x85\x82";
1571 } elseif ( $code < 0xc544 ) {
1572 return "\xe3\x85\x85";
1573 } elseif ( $code < 0xc790 ) {
1574 return "\xe3\x85\x87";
1575 } elseif ( $code < 0xcc28 ) {
1576 return "\xe3\x85\x88";
1577 } elseif ( $code < 0xce74 ) {
1578 return "\xe3\x85\x8a";
1579 } elseif ( $code < 0xd0c0 ) {
1580 return "\xe3\x85\x8b";
1581 } elseif ( $code < 0xd30c ) {
1582 return "\xe3\x85\x8c";
1583 } elseif ( $code < 0xd558 ) {
1584 return "\xe3\x85\x8d";
1585 } else {
1586 return "\xe3\x85\x8e";
1587 }
1588 } else {
1589 return "";
1590 }
1591 }
1592
1593 function initEncoding() {
1594 # Some languages may have an alternate char encoding option
1595 # (Esperanto X-coding, Japanese furigana conversion, etc)
1596 # If this language is used as the primary content language,
1597 # an override to the defaults can be set here on startup.
1598 }
1599
1600 function recodeForEdit( $s ) {
1601 # For some languages we'll want to explicitly specify
1602 # which characters make it into the edit box raw
1603 # or are converted in some way or another.
1604 # Note that if wgOutputEncoding is different from
1605 # wgInputEncoding, this text will be further converted
1606 # to wgOutputEncoding.
1607 global $wgEditEncoding;
1608 if( $wgEditEncoding == '' or
1609 $wgEditEncoding == 'UTF-8' ) {
1610 return $s;
1611 } else {
1612 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1613 }
1614 }
1615
1616 function recodeInput( $s ) {
1617 # Take the previous into account.
1618 global $wgEditEncoding;
1619 if($wgEditEncoding != "") {
1620 $enc = $wgEditEncoding;
1621 } else {
1622 $enc = 'UTF-8';
1623 }
1624 if( $enc == 'UTF-8' ) {
1625 return $s;
1626 } else {
1627 return $this->iconv( $enc, 'UTF-8', $s );
1628 }
1629 }
1630
1631 /**
1632 * For right-to-left language support
1633 *
1634 * @return bool
1635 */
1636 function isRTL() {
1637 $this->load();
1638 return $this->rtl;
1639 }
1640
1641 /**
1642 * A hidden direction mark (LRM or RLM), depending on the language direction
1643 *
1644 * @return string
1645 */
1646 function getDirMark() {
1647 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1648 }
1649
1650 /**
1651 * An arrow, depending on the language direction
1652 *
1653 * @return string
1654 */
1655 function getArrow() {
1656 return $this->isRTL() ? '←' : '→';
1657 }
1658
1659 /**
1660 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1661 *
1662 * @return bool
1663 */
1664 function linkPrefixExtension() {
1665 $this->load();
1666 return $this->linkPrefixExtension;
1667 }
1668
1669 function &getMagicWords() {
1670 $this->load();
1671 return $this->magicWords;
1672 }
1673
1674 # Fill a MagicWord object with data from here
1675 function getMagic( &$mw ) {
1676 if ( !$this->mMagicHookDone ) {
1677 $this->mMagicHookDone = true;
1678 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1679 }
1680 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1681 $rawEntry = $this->mMagicExtensions[$mw->mId];
1682 } else {
1683 $magicWords =& $this->getMagicWords();
1684 if ( isset( $magicWords[$mw->mId] ) ) {
1685 $rawEntry = $magicWords[$mw->mId];
1686 } else {
1687 # Fall back to English if local list is incomplete
1688 $magicWords =& Language::getMagicWords();
1689 if ( !isset($magicWords[$mw->mId]) ) { throw new MWException("Magic word not found" ); }
1690 $rawEntry = $magicWords[$mw->mId];
1691 }
1692 }
1693
1694 if( !is_array( $rawEntry ) ) {
1695 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1696 } else {
1697 $mw->mCaseSensitive = $rawEntry[0];
1698 $mw->mSynonyms = array_slice( $rawEntry, 1 );
1699 }
1700 }
1701
1702 /**
1703 * Add magic words to the extension array
1704 */
1705 function addMagicWordsByLang( $newWords ) {
1706 $code = $this->getCode();
1707 $fallbackChain = array();
1708 while ( $code && !in_array( $code, $fallbackChain ) ) {
1709 $fallbackChain[] = $code;
1710 $code = self::getFallbackFor( $code );
1711 }
1712 if ( !in_array( 'en', $fallbackChain ) ) {
1713 $fallbackChain[] = 'en';
1714 }
1715 $fallbackChain = array_reverse( $fallbackChain );
1716 foreach ( $fallbackChain as $code ) {
1717 if ( isset( $newWords[$code] ) ) {
1718 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1719 }
1720 }
1721 }
1722
1723 /**
1724 * Get special page names, as an associative array
1725 * case folded alias => real name
1726 */
1727 function getSpecialPageAliases() {
1728 $this->load();
1729
1730 // Cache aliases because it may be slow to load them
1731 if ( !isset( $this->mExtendedSpecialPageAliases ) ) {
1732
1733 // Initialise array
1734 $this->mExtendedSpecialPageAliases = $this->specialPageAliases;
1735
1736 global $wgExtensionAliasesFiles;
1737 foreach ( $wgExtensionAliasesFiles as $file ) {
1738
1739 // Fail fast
1740 if ( !file_exists($file) )
1741 throw new MWException( 'Aliases file does not exist' );
1742
1743 $aliases = array();
1744 require($file);
1745
1746 // Check the availability of aliases
1747 if ( !isset($aliases['en']) )
1748 throw new MWException( 'Malformed aliases file' );
1749
1750 // Merge all aliases in fallback chain
1751 $code = $this->getCode();
1752 do {
1753 if ( !isset($aliases[$code]) ) continue;
1754
1755 $aliases[$code] = $this->fixSpecialPageAliases( $aliases[$code] );
1756 /* Merge the aliases, THIS will break if there is special page name
1757 * which looks like a numerical key, thanks to PHP...
1758 * See the comments for wfArrayMerge in GlobalSettings.php. */
1759 $this->mExtendedSpecialPageAliases = array_merge_recursive(
1760 $this->mExtendedSpecialPageAliases, $aliases[$code] );
1761
1762 } while ( $code = self::getFallbackFor( $code ) );
1763 }
1764
1765 wfRunHooks( 'LanguageGetSpecialPageAliases',
1766 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1767 }
1768
1769 return $this->mExtendedSpecialPageAliases;
1770 }
1771
1772 /**
1773 * Function to fix special page aliases. Will convert the first letter to
1774 * upper case and spaces to underscores. Can be given a full aliases array,
1775 * in which case it will recursively fix all aliases.
1776 */
1777 public function fixSpecialPageAliases( $mixed ) {
1778 // Work recursively until in string level
1779 if ( is_array($mixed) ) {
1780 $callback = array( $this, 'fixSpecialPageAliases' );
1781 return array_map( $callback, $mixed );
1782 }
1783 return str_replace( ' ', '_', $this->ucfirst( $mixed ) );
1784 }
1785
1786 /**
1787 * Italic is unsuitable for some languages
1788 *
1789 * @param $text String: the text to be emphasized.
1790 * @return string
1791 */
1792 function emphasize( $text ) {
1793 return "<em>$text</em>";
1794 }
1795
1796 /**
1797 * Normally we output all numbers in plain en_US style, that is
1798 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1799 * point twohundredthirtyfive. However this is not sutable for all
1800 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1801 * Icelandic just want to use commas instead of dots, and dots instead
1802 * of commas like "293.291,235".
1803 *
1804 * An example of this function being called:
1805 * <code>
1806 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1807 * </code>
1808 *
1809 * See LanguageGu.php for the Gujarati implementation and
1810 * LanguageIs.php for the , => . and . => , implementation.
1811 *
1812 * @todo check if it's viable to use localeconv() for the decimal
1813 * seperator thing.
1814 * @param $number Mixed: the string to be formatted, should be an integer
1815 * or a floating point number.
1816 * @param $nocommafy Bool: set to true for special numbers like dates
1817 * @return string
1818 */
1819 function formatNum( $number, $nocommafy = false ) {
1820 global $wgTranslateNumerals;
1821 if (!$nocommafy) {
1822 $number = $this->commafy($number);
1823 $s = $this->separatorTransformTable();
1824 if (!is_null($s)) { $number = strtr($number, $s); }
1825 }
1826
1827 if ($wgTranslateNumerals) {
1828 $s = $this->digitTransformTable();
1829 if (!is_null($s)) { $number = strtr($number, $s); }
1830 }
1831
1832 return $number;
1833 }
1834
1835 function parseFormattedNumber( $number ) {
1836 $s = $this->digitTransformTable();
1837 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1838
1839 $s = $this->separatorTransformTable();
1840 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1841
1842 $number = strtr( $number, array (',' => '') );
1843 return $number;
1844 }
1845
1846 /**
1847 * Adds commas to a given number
1848 *
1849 * @param $_ mixed
1850 * @return string
1851 */
1852 function commafy($_) {
1853 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1854 }
1855
1856 function digitTransformTable() {
1857 $this->load();
1858 return $this->digitTransformTable;
1859 }
1860
1861 function separatorTransformTable() {
1862 $this->load();
1863 return $this->separatorTransformTable;
1864 }
1865
1866
1867 /**
1868 * For the credit list in includes/Credits.php (action=credits)
1869 *
1870 * @param $l Array
1871 * @return string
1872 */
1873 function listToText( $l ) {
1874 $s = '';
1875 $m = count($l) - 1;
1876 for ($i = $m; $i >= 0; $i--) {
1877 if ($i == $m) {
1878 $s = $l[$i];
1879 } else if ($i == $m - 1) {
1880 $s = $l[$i] . ' ' . $this->getMessageFromDB( 'and' ) . ' ' . $s;
1881 } else {
1882 $s = $l[$i] . ', ' . $s;
1883 }
1884 }
1885 return $s;
1886 }
1887
1888 /**
1889 * Truncate a string to a specified length in bytes, appending an optional
1890 * string (e.g. for ellipses)
1891 *
1892 * The database offers limited byte lengths for some columns in the database;
1893 * multi-byte character sets mean we need to ensure that only whole characters
1894 * are included, otherwise broken characters can be passed to the user
1895 *
1896 * If $length is negative, the string will be truncated from the beginning
1897 *
1898 * @param $string String to truncate
1899 * @param $length Int: maximum length (excluding ellipses)
1900 * @param $ellipsis String to append to the truncated text
1901 * @return string
1902 */
1903 function truncate( $string, $length, $ellipsis = "" ) {
1904 if( $length == 0 ) {
1905 return $ellipsis;
1906 }
1907 if ( strlen( $string ) <= abs( $length ) ) {
1908 return $string;
1909 }
1910 if( $length > 0 ) {
1911 $string = substr( $string, 0, $length );
1912 $char = ord( $string[strlen( $string ) - 1] );
1913 $m = array();
1914 if ($char >= 0xc0) {
1915 # We got the first byte only of a multibyte char; remove it.
1916 $string = substr( $string, 0, -1 );
1917 } elseif( $char >= 0x80 &&
1918 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
1919 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
1920 # We chopped in the middle of a character; remove it
1921 $string = $m[1];
1922 }
1923 return $string . $ellipsis;
1924 } else {
1925 $string = substr( $string, $length );
1926 $char = ord( $string[0] );
1927 if( $char >= 0x80 && $char < 0xc0 ) {
1928 # We chopped in the middle of a character; remove the whole thing
1929 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
1930 }
1931 return $ellipsis . $string;
1932 }
1933 }
1934
1935 /**
1936 * Grammatical transformations, needed for inflected languages
1937 * Invoked by putting {{grammar:case|word}} in a message
1938 *
1939 * @param $word string
1940 * @param $case string
1941 * @return string
1942 */
1943 function convertGrammar( $word, $case ) {
1944 global $wgGrammarForms;
1945 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
1946 return $wgGrammarForms[$this->getCode()][$case][$word];
1947 }
1948 return $word;
1949 }
1950
1951 /**
1952 * Plural form transformations, needed for some languages.
1953 * For example, there are 3 form of plural in Russian and Polish,
1954 * depending on "count mod 10". See [[w:Plural]]
1955 * For English it is pretty simple.
1956 *
1957 * Invoked by putting {{plural:count|wordform1|wordform2}}
1958 * or {{plural:count|wordform1|wordform2|wordform3}}
1959 *
1960 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
1961 *
1962 * @param $count Integer: non-localized number
1963 * @param $forms Array: different plural forms
1964 * @return string Correct form of plural for $count in this language
1965 */
1966 function convertPlural( $count, $forms ) {
1967 if ( !count($forms) ) { return ''; }
1968 $forms = $this->preConvertPlural( $forms, 2 );
1969
1970 return ( abs($count) == 1 ) ? $forms[0] : $forms[1];
1971 }
1972
1973 /**
1974 * Checks that convertPlural was given an array and pads it to requested
1975 * amound of forms by copying the last one.
1976 *
1977 * @param $count Integer: How many forms should there be at least
1978 * @param $forms Array of forms given to convertPlural
1979 * @return array Padded array of forms or an exception if not an array
1980 */
1981 protected function preConvertPlural( /* Array */ $forms, $count ) {
1982 while ( count($forms) < $count ) {
1983 $forms[] = $forms[count($forms)-1];
1984 }
1985 return $forms;
1986 }
1987
1988 /**
1989 * For translaing of expiry times
1990 * @param $str String: the validated block time in English
1991 * @return Somehow translated block time
1992 * @see LanguageFi.php for example implementation
1993 */
1994 function translateBlockExpiry( $str ) {
1995
1996 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
1997
1998 if ( $scBlockExpiryOptions == '-') {
1999 return $str;
2000 }
2001
2002 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2003 if ( strpos($option, ":") === false )
2004 continue;
2005 list($show, $value) = explode(":", $option);
2006 if ( strcmp ( $str, $value) == 0 ) {
2007 return htmlspecialchars( trim( $show ) );
2008 }
2009 }
2010
2011 return $str;
2012 }
2013
2014 /**
2015 * languages like Chinese need to be segmented in order for the diff
2016 * to be of any use
2017 *
2018 * @param $text String
2019 * @return String
2020 */
2021 function segmentForDiff( $text ) {
2022 return $text;
2023 }
2024
2025 /**
2026 * and unsegment to show the result
2027 *
2028 * @param $text String
2029 * @return String
2030 */
2031 function unsegmentForDiff( $text ) {
2032 return $text;
2033 }
2034
2035 # convert text to different variants of a language.
2036 function convert( $text, $isTitle = false) {
2037 return $this->mConverter->convert($text, $isTitle);
2038 }
2039
2040 # Convert text from within Parser
2041 function parserConvert( $text, &$parser ) {
2042 return $this->mConverter->parserConvert( $text, $parser );
2043 }
2044
2045 # Check if this is a language with variants
2046 function hasVariants(){
2047 return sizeof($this->getVariants())>1;
2048 }
2049
2050 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2051 function armourMath($text){
2052 return $this->mConverter->armourMath($text);
2053 }
2054
2055
2056 /**
2057 * Perform output conversion on a string, and encode for safe HTML output.
2058 * @param $text String
2059 * @param $isTitle Bool -- wtf?
2060 * @return string
2061 * @todo this should get integrated somewhere sane
2062 */
2063 function convertHtml( $text, $isTitle = false ) {
2064 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2065 }
2066
2067 function convertCategoryKey( $key ) {
2068 return $this->mConverter->convertCategoryKey( $key );
2069 }
2070
2071 /**
2072 * get the list of variants supported by this langauge
2073 * see sample implementation in LanguageZh.php
2074 *
2075 * @return array an array of language codes
2076 */
2077 function getVariants() {
2078 return $this->mConverter->getVariants();
2079 }
2080
2081
2082 function getPreferredVariant( $fromUser = true ) {
2083 return $this->mConverter->getPreferredVariant( $fromUser );
2084 }
2085
2086 /**
2087 * if a language supports multiple variants, it is
2088 * possible that non-existing link in one variant
2089 * actually exists in another variant. this function
2090 * tries to find it. See e.g. LanguageZh.php
2091 *
2092 * @param $link String: the name of the link
2093 * @param $nt Mixed: the title object of the link
2094 * @return null the input parameters may be modified upon return
2095 */
2096 function findVariantLink( &$link, &$nt ) {
2097 $this->mConverter->findVariantLink($link, $nt);
2098 }
2099
2100 /**
2101 * If a language supports multiple variants, converts text
2102 * into an array of all possible variants of the text:
2103 * 'variant' => text in that variant
2104 */
2105
2106 function convertLinkToAllVariants($text){
2107 return $this->mConverter->convertLinkToAllVariants($text);
2108 }
2109
2110
2111 /**
2112 * returns language specific options used by User::getPageRenderHash()
2113 * for example, the preferred language variant
2114 *
2115 * @return string
2116 */
2117 function getExtraHashOptions() {
2118 return $this->mConverter->getExtraHashOptions();
2119 }
2120
2121 /**
2122 * for languages that support multiple variants, the title of an
2123 * article may be displayed differently in different variants. this
2124 * function returns the apporiate title defined in the body of the article.
2125 *
2126 * @return string
2127 */
2128 function getParsedTitle() {
2129 return $this->mConverter->getParsedTitle();
2130 }
2131
2132 /**
2133 * Enclose a string with the "no conversion" tag. This is used by
2134 * various functions in the Parser
2135 *
2136 * @param $text String: text to be tagged for no conversion
2137 * @param $noParse
2138 * @return string the tagged text
2139 */
2140 function markNoConversion( $text, $noParse=false ) {
2141 return $this->mConverter->markNoConversion( $text, $noParse );
2142 }
2143
2144 /**
2145 * A regular expression to match legal word-trailing characters
2146 * which should be merged onto a link of the form [[foo]]bar.
2147 *
2148 * @return string
2149 */
2150 function linkTrail() {
2151 $this->load();
2152 return $this->linkTrail;
2153 }
2154
2155 function getLangObj() {
2156 return $this;
2157 }
2158
2159 /**
2160 * Get the RFC 3066 code for this language object
2161 */
2162 function getCode() {
2163 return $this->mCode;
2164 }
2165
2166 function setCode( $code ) {
2167 $this->mCode = $code;
2168 }
2169
2170 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2171 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2172 }
2173
2174 static function getMessagesFileName( $code ) {
2175 global $IP;
2176 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2177 }
2178
2179 static function getClassFileName( $code ) {
2180 global $IP;
2181 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2182 }
2183
2184 static function getLocalisationArray( $code, $disableCache = false ) {
2185 self::loadLocalisation( $code, $disableCache );
2186 return self::$mLocalisationCache[$code];
2187 }
2188
2189 /**
2190 * Load localisation data for a given code into the static cache
2191 *
2192 * @return array Dependencies, map of filenames to mtimes
2193 */
2194 static function loadLocalisation( $code, $disableCache = false ) {
2195 static $recursionGuard = array();
2196 global $wgMemc, $wgCheckSerialized;
2197
2198 if ( !$code ) {
2199 throw new MWException( "Invalid language code requested" );
2200 }
2201
2202 if ( !$disableCache ) {
2203 # Try the per-process cache
2204 if ( isset( self::$mLocalisationCache[$code] ) ) {
2205 return self::$mLocalisationCache[$code]['deps'];
2206 }
2207
2208 wfProfileIn( __METHOD__ );
2209
2210 # Try the serialized directory
2211 $cache = wfGetPrecompiledData( self::getFileName( "Messages", $code, '.ser' ) );
2212 if ( $cache ) {
2213 if ( $wgCheckSerialized && self::isLocalisationOutOfDate( $cache ) ) {
2214 $cache = false;
2215 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2216 } else {
2217 self::$mLocalisationCache[$code] = $cache;
2218 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2219 wfProfileOut( __METHOD__ );
2220 return self::$mLocalisationCache[$code]['deps'];
2221 }
2222 }
2223
2224 # Try the global cache
2225 $memcKey = wfMemcKey('localisation', $code );
2226 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2227 $cache = $wgMemc->get( $memcKey );
2228 if ( $cache ) {
2229 if ( self::isLocalisationOutOfDate( $cache ) ) {
2230 $wgMemc->delete( $memcKey );
2231 $wgMemc->delete( $fbMemcKey );
2232 $cache = false;
2233 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2234 } else {
2235 self::$mLocalisationCache[$code] = $cache;
2236 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2237 wfProfileOut( __METHOD__ );
2238 return $cache['deps'];
2239 }
2240 }
2241 } else {
2242 wfProfileIn( __METHOD__ );
2243 }
2244
2245 # Default fallback, may be overridden when the messages file is included
2246 if ( $code != 'en' ) {
2247 $fallback = 'en';
2248 } else {
2249 $fallback = false;
2250 }
2251
2252 # Load the primary localisation from the source file
2253 $filename = self::getMessagesFileName( $code );
2254 if ( !file_exists( $filename ) ) {
2255 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2256 $cache = compact( self::$mLocalisationKeys ); // Set correct fallback
2257 $deps = array();
2258 } else {
2259 $deps = array( $filename => filemtime( $filename ) );
2260 require( $filename );
2261 $cache = compact( self::$mLocalisationKeys );
2262 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2263 }
2264
2265 if ( !empty( $fallback ) ) {
2266 # Load the fallback localisation, with a circular reference guard
2267 if ( isset( $recursionGuard[$code] ) ) {
2268 throw new MWException( "Error: Circular fallback reference in language code $code" );
2269 }
2270 $recursionGuard[$code] = true;
2271 $newDeps = self::loadLocalisation( $fallback, $disableCache );
2272 unset( $recursionGuard[$code] );
2273
2274 $secondary = self::$mLocalisationCache[$fallback];
2275 $deps = array_merge( $deps, $newDeps );
2276
2277 # Merge the fallback localisation with the current localisation
2278 foreach ( self::$mLocalisationKeys as $key ) {
2279 if ( isset( $cache[$key] ) ) {
2280 if ( isset( $secondary[$key] ) ) {
2281 if ( in_array( $key, self::$mMergeableMapKeys ) ) {
2282 $cache[$key] = $cache[$key] + $secondary[$key];
2283 } elseif ( in_array( $key, self::$mMergeableListKeys ) ) {
2284 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2285 } elseif ( in_array( $key, self::$mMergeableAliasListKeys ) ) {
2286 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2287 }
2288 }
2289 } else {
2290 $cache[$key] = $secondary[$key];
2291 }
2292 }
2293
2294 # Merge bookstore lists if requested
2295 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2296 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2297 }
2298 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2299 unset( $cache['bookstoreList']['inherit'] );
2300 }
2301 }
2302
2303 # Add dependencies to the cache entry
2304 $cache['deps'] = $deps;
2305
2306 # Replace spaces with underscores in namespace names
2307 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2308
2309 # And do the same for specialpage aliases. $page is an array.
2310 foreach ( $cache['specialPageAliases'] as &$page ) {
2311 $page = str_replace( ' ', '_', $page );
2312 }
2313 # Decouple the reference to prevent accidental damage
2314 unset($page);
2315
2316 # Save to both caches
2317 self::$mLocalisationCache[$code] = $cache;
2318 if ( !$disableCache ) {
2319 $wgMemc->set( $memcKey, $cache );
2320 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2321 }
2322
2323 wfProfileOut( __METHOD__ );
2324 return $deps;
2325 }
2326
2327 /**
2328 * Test if a given localisation cache is out of date with respect to the
2329 * source Messages files. This is done automatically for the global cache
2330 * in $wgMemc, but is only done on certain occasions for the serialized
2331 * data file.
2332 *
2333 * @param $cache mixed Either a language code or a cache array
2334 */
2335 static function isLocalisationOutOfDate( $cache ) {
2336 if ( !is_array( $cache ) ) {
2337 self::loadLocalisation( $cache );
2338 $cache = self::$mLocalisationCache[$cache];
2339 }
2340 $expired = false;
2341 foreach ( $cache['deps'] as $file => $mtime ) {
2342 if ( !file_exists( $file ) || filemtime( $file ) > $mtime ) {
2343 $expired = true;
2344 break;
2345 }
2346 }
2347 return $expired;
2348 }
2349
2350 /**
2351 * Get the fallback for a given language
2352 */
2353 static function getFallbackFor( $code ) {
2354 // Shortcut
2355 if ( $code === 'en' ) return false;
2356
2357 // Local cache
2358 static $cache = array();
2359 // Quick return
2360 if ( isset($cache[$code]) ) return $cache[$code];
2361
2362 // Try memcache
2363 global $wgMemc;
2364 $memcKey = wfMemcKey( 'fallback', $code );
2365 $fbcode = $wgMemc->get( $memcKey );
2366
2367 if ( is_string($fbcode) ) {
2368 // False is stored as a string to detect failures in memcache properly
2369 if ( $fbcode === '' ) $fbcode = false;
2370
2371 // Update local cache and return
2372 $cache[$code] = $fbcode;
2373 return $fbcode;
2374 }
2375
2376 // Nothing in caches, load and and update both caches
2377 self::loadLocalisation( $code );
2378 $fbcode = self::$mLocalisationCache[$code]['fallback'];
2379
2380 $cache[$code] = $fbcode;
2381 $wgMemc->set( $memcKey, (string) $fbcode );
2382
2383 return $fbcode;
2384 }
2385
2386 /**
2387 * Get all messages for a given language
2388 */
2389 static function getMessagesFor( $code ) {
2390 self::loadLocalisation( $code );
2391 return self::$mLocalisationCache[$code]['messages'];
2392 }
2393
2394 /**
2395 * Get a message for a given language
2396 */
2397 static function getMessageFor( $key, $code ) {
2398 self::loadLocalisation( $code );
2399 return isset( self::$mLocalisationCache[$code]['messages'][$key] ) ? self::$mLocalisationCache[$code]['messages'][$key] : null;
2400 }
2401
2402 /**
2403 * Load localisation data for this object
2404 */
2405 function load() {
2406 if ( !$this->mLoaded ) {
2407 self::loadLocalisation( $this->getCode() );
2408 $cache =& self::$mLocalisationCache[$this->getCode()];
2409 foreach ( self::$mLocalisationKeys as $key ) {
2410 $this->$key = $cache[$key];
2411 }
2412 $this->mLoaded = true;
2413
2414 $this->fixUpSettings();
2415 }
2416 }
2417
2418 /**
2419 * Do any necessary post-cache-load settings adjustment
2420 */
2421 function fixUpSettings() {
2422 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2423 $wgNamespaceAliases, $wgAmericanDates;
2424 wfProfileIn( __METHOD__ );
2425 if ( $wgExtraNamespaces ) {
2426 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
2427 }
2428
2429 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
2430 if ( $wgMetaNamespaceTalk ) {
2431 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
2432 } else {
2433 $talk = $this->namespaceNames[NS_PROJECT_TALK];
2434 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2435
2436 # Allow grammar transformations
2437 # Allowing full message-style parsing would make simple requests
2438 # such as action=raw much more expensive than they need to be.
2439 # This will hopefully cover most cases.
2440 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2441 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2442 $talk = str_replace( ' ', '_', $talk );
2443 $this->namespaceNames[NS_PROJECT_TALK] = $talk;
2444 }
2445
2446 # The above mixing may leave namespaces out of canonical order.
2447 # Re-order by namespace ID number...
2448 ksort( $this->namespaceNames );
2449
2450 # Put namespace names and aliases into a hashtable.
2451 # If this is too slow, then we should arrange it so that it is done
2452 # before caching. The catch is that at pre-cache time, the above
2453 # class-specific fixup hasn't been done.
2454 $this->mNamespaceIds = array();
2455 foreach ( $this->namespaceNames as $index => $name ) {
2456 $this->mNamespaceIds[$this->lc($name)] = $index;
2457 }
2458 if ( $this->namespaceAliases ) {
2459 foreach ( $this->namespaceAliases as $name => $index ) {
2460 $this->mNamespaceIds[$this->lc($name)] = $index;
2461 }
2462 }
2463 if ( $wgNamespaceAliases ) {
2464 foreach ( $wgNamespaceAliases as $name => $index ) {
2465 $this->mNamespaceIds[$this->lc($name)] = $index;
2466 }
2467 }
2468
2469 if ( $this->defaultDateFormat == 'dmy or mdy' ) {
2470 $this->defaultDateFormat = $wgAmericanDates ? 'mdy' : 'dmy';
2471 }
2472 wfProfileOut( __METHOD__ );
2473 }
2474
2475 function replaceGrammarInNamespace( $m ) {
2476 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2477 }
2478
2479 static function getCaseMaps() {
2480 static $wikiUpperChars, $wikiLowerChars;
2481 if ( isset( $wikiUpperChars ) ) {
2482 return array( $wikiUpperChars, $wikiLowerChars );
2483 }
2484
2485 wfProfileIn( __METHOD__ );
2486 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2487 if ( $arr === false ) {
2488 throw new MWException(
2489 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2490 }
2491 extract( $arr );
2492 wfProfileOut( __METHOD__ );
2493 return array( $wikiUpperChars, $wikiLowerChars );
2494 }
2495
2496 function formatTimePeriod( $seconds ) {
2497 if ( $seconds < 10 ) {
2498 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2499 } elseif ( $seconds < 60 ) {
2500 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2501 } elseif ( $seconds < 3600 ) {
2502 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2503 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2504 } else {
2505 $hours = floor( $seconds / 3600 );
2506 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2507 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2508 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2509 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2510 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2511 }
2512 }
2513
2514 function formatBitrate( $bps ) {
2515 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2516 if ( $bps <= 0 ) {
2517 return $this->formatNum( $bps ) . $units[0];
2518 }
2519 $unitIndex = floor( log10( $bps ) / 3 );
2520 $mantissa = $bps / pow( 1000, $unitIndex );
2521 if ( $mantissa < 10 ) {
2522 $mantissa = round( $mantissa, 1 );
2523 } else {
2524 $mantissa = round( $mantissa );
2525 }
2526 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2527 }
2528
2529 /**
2530 * Format a size in bytes for output, using an appropriate
2531 * unit (B, KB, MB or GB) according to the magnitude in question
2532 *
2533 * @param $size Size to format
2534 * @return string Plain text (not HTML)
2535 */
2536 function formatSize( $size ) {
2537 // For small sizes no decimal places necessary
2538 $round = 0;
2539 if( $size > 1024 ) {
2540 $size = $size / 1024;
2541 if( $size > 1024 ) {
2542 $size = $size / 1024;
2543 // For MB and bigger two decimal places are smarter
2544 $round = 2;
2545 if( $size > 1024 ) {
2546 $size = $size / 1024;
2547 $msg = 'size-gigabytes';
2548 } else {
2549 $msg = 'size-megabytes';
2550 }
2551 } else {
2552 $msg = 'size-kilobytes';
2553 }
2554 } else {
2555 $msg = 'size-bytes';
2556 }
2557 $size = round( $size, $round );
2558 $text = $this->getMessageFromDB( $msg );
2559 return str_replace( '$1', $this->formatNum( $size ), $text );
2560 }
2561 }