* $wgExtensionAliasesFiles option to simplify adding aliases for special pages provid...
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * @defgroup Language Language
4 *
5 * @file
6 * @ingroup Language
7 */
8
9 if( !defined( 'MEDIAWIKI' ) ) {
10 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
11 exit( 1 );
12 }
13
14 # Read language names
15 global $wgLanguageNames;
16 require_once( dirname(__FILE__) . '/Names.php' ) ;
17
18 global $wgInputEncoding, $wgOutputEncoding;
19
20 /**
21 * These are always UTF-8, they exist only for backwards compatibility
22 */
23 $wgInputEncoding = "UTF-8";
24 $wgOutputEncoding = "UTF-8";
25
26 if( function_exists( 'mb_strtoupper' ) ) {
27 mb_internal_encoding('UTF-8');
28 }
29
30 /**
31 * a fake language converter
32 *
33 * @ingroup Language
34 */
35 class FakeConverter {
36 var $mLang;
37 function FakeConverter($langobj) {$this->mLang = $langobj;}
38 function convert($t, $i) {return $t;}
39 function parserConvert($t, $p) {return $t;}
40 function getVariants() { return array( $this->mLang->getCode() ); }
41 function getPreferredVariant() {return $this->mLang->getCode(); }
42 function findVariantLink(&$l, &$n) {}
43 function getExtraHashOptions() {return '';}
44 function getParsedTitle() {return '';}
45 function markNoConversion($text, $noParse=false) {return $text;}
46 function convertCategoryKey( $key ) {return $key; }
47 function convertLinkToAllVariants($text){ return array( $this->mLang->getCode() => $text); }
48 function armourMath($text){ return $text; }
49 }
50
51 /**
52 * Internationalisation code
53 * @ingroup Language
54 */
55 class Language {
56 var $mConverter, $mVariants, $mCode, $mLoaded = false;
57 var $mMagicExtensions = array(), $mMagicHookDone = false;
58
59 static public $mLocalisationKeys = array( 'fallback', 'namespaceNames',
60 'skinNames', 'mathNames',
61 'bookstoreList', 'magicWords', 'messages', 'rtl', 'digitTransformTable',
62 'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
63 'defaultUserOptionOverrides', 'linkTrail', 'namespaceAliases',
64 'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
65 'defaultDateFormat', 'extraUserToggles', 'specialPageAliases' );
66
67 static public $mMergeableMapKeys = array( 'messages', 'namespaceNames', 'mathNames',
68 'dateFormats', 'defaultUserOptionOverrides', 'magicWords' );
69
70 static public $mMergeableListKeys = array( 'extraUserToggles' );
71
72 static public $mMergeableAliasListKeys = array( 'specialPageAliases' );
73
74 static public $mLocalisationCache = array();
75
76 static public $mWeekdayMsgs = array(
77 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
78 'friday', 'saturday'
79 );
80
81 static public $mWeekdayAbbrevMsgs = array(
82 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
83 );
84
85 static public $mMonthMsgs = array(
86 'january', 'february', 'march', 'april', 'may_long', 'june',
87 'july', 'august', 'september', 'october', 'november',
88 'december'
89 );
90 static public $mMonthGenMsgs = array(
91 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
92 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
93 'december-gen'
94 );
95 static public $mMonthAbbrevMsgs = array(
96 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
97 'sep', 'oct', 'nov', 'dec'
98 );
99
100 static public $mIranianCalendarMonthMsgs = array(
101 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
102 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
103 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
104 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
105 );
106
107 static public $mHebrewCalendarMonthMsgs = array(
108 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
109 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
110 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
111 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
112 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
113 );
114
115 static public $mHebrewCalendarMonthGenMsgs = array(
116 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
117 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
118 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
119 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
120 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
121 );
122
123 static public $mHijriCalendarMonthMsgs = array(
124 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
125 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
126 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
127 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
128 );
129
130 /**
131 * Create a language object for a given language code
132 */
133 static function factory( $code ) {
134 global $IP;
135 static $recursionLevel = 0;
136
137 if ( $code == 'en' ) {
138 $class = 'Language';
139 } else {
140 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
141 // Preload base classes to work around APC/PHP5 bug
142 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
143 include_once("$IP/languages/classes/$class.deps.php");
144 }
145 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
146 include_once("$IP/languages/classes/$class.php");
147 }
148 }
149
150 if ( $recursionLevel > 5 ) {
151 throw new MWException( "Language fallback loop detected when creating class $class\n" );
152 }
153
154 if( ! class_exists( $class ) ) {
155 $fallback = Language::getFallbackFor( $code );
156 ++$recursionLevel;
157 $lang = Language::factory( $fallback );
158 --$recursionLevel;
159 $lang->setCode( $code );
160 } else {
161 $lang = new $class;
162 }
163
164 return $lang;
165 }
166
167 function __construct() {
168 $this->mConverter = new FakeConverter($this);
169 // Set the code to the name of the descendant
170 if ( get_class( $this ) == 'Language' ) {
171 $this->mCode = 'en';
172 } else {
173 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
174 }
175 }
176
177 /**
178 * Hook which will be called if this is the content language.
179 * Descendants can use this to register hook functions or modify globals
180 */
181 function initContLang() {}
182
183 /**
184 * @deprecated Use User::getDefaultOptions()
185 * @return array
186 */
187 function getDefaultUserOptions() {
188 wfDeprecated( __METHOD__ );
189 return User::getDefaultOptions();
190 }
191
192 function getFallbackLanguageCode() {
193 return self::getFallbackFor( $this->mCode );
194 }
195
196 /**
197 * Exports $wgBookstoreListEn
198 * @return array
199 */
200 function getBookstoreList() {
201 $this->load();
202 return $this->bookstoreList;
203 }
204
205 /**
206 * @return array
207 */
208 function getNamespaces() {
209 $this->load();
210 return $this->namespaceNames;
211 }
212
213 /**
214 * A convenience function that returns the same thing as
215 * getNamespaces() except with the array values changed to ' '
216 * where it found '_', useful for producing output to be displayed
217 * e.g. in <select> forms.
218 *
219 * @return array
220 */
221 function getFormattedNamespaces() {
222 $ns = $this->getNamespaces();
223 foreach($ns as $k => $v) {
224 $ns[$k] = strtr($v, '_', ' ');
225 }
226 return $ns;
227 }
228
229 /**
230 * Get a namespace value by key
231 * <code>
232 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
233 * echo $mw_ns; // prints 'MediaWiki'
234 * </code>
235 *
236 * @param $index Int: the array key of the namespace to return
237 * @return mixed, string if the namespace value exists, otherwise false
238 */
239 function getNsText( $index ) {
240 $ns = $this->getNamespaces();
241 return isset( $ns[$index] ) ? $ns[$index] : false;
242 }
243
244 /**
245 * A convenience function that returns the same thing as
246 * getNsText() except with '_' changed to ' ', useful for
247 * producing output.
248 *
249 * @return array
250 */
251 function getFormattedNsText( $index ) {
252 $ns = $this->getNsText( $index );
253 return strtr($ns, '_', ' ');
254 }
255
256 /**
257 * Get a namespace key by value, case insensitive.
258 * Only matches namespace names for the current language, not the
259 * canonical ones defined in Namespace.php.
260 *
261 * @param $text String
262 * @return mixed An integer if $text is a valid value otherwise false
263 */
264 function getLocalNsIndex( $text ) {
265 $this->load();
266 $lctext = $this->lc($text);
267 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
268 }
269
270 /**
271 * Get a namespace key by value, case insensitive. Canonical namespace
272 * names override custom ones defined for the current language.
273 *
274 * @param $text String
275 * @return mixed An integer if $text is a valid value otherwise false
276 */
277 function getNsIndex( $text ) {
278 $this->load();
279 $lctext = $this->lc($text);
280 if( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
281 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
282 }
283
284 /**
285 * short names for language variants used for language conversion links.
286 *
287 * @param $code String
288 * @return string
289 */
290 function getVariantname( $code ) {
291 return $this->getMessageFromDB( "variantname-$code" );
292 }
293
294 function specialPage( $name ) {
295 $aliases = $this->getSpecialPageAliases();
296 if ( isset( $aliases[$name][0] ) ) {
297 $name = $aliases[$name][0];
298 }
299 return $this->getNsText(NS_SPECIAL) . ':' . $name;
300 }
301
302 function getQuickbarSettings() {
303 return array(
304 $this->getMessage( 'qbsettings-none' ),
305 $this->getMessage( 'qbsettings-fixedleft' ),
306 $this->getMessage( 'qbsettings-fixedright' ),
307 $this->getMessage( 'qbsettings-floatingleft' ),
308 $this->getMessage( 'qbsettings-floatingright' )
309 );
310 }
311
312 function getSkinNames() {
313 $this->load();
314 return $this->skinNames;
315 }
316
317 function getMathNames() {
318 $this->load();
319 return $this->mathNames;
320 }
321
322 function getDatePreferences() {
323 $this->load();
324 return $this->datePreferences;
325 }
326
327 function getDateFormats() {
328 $this->load();
329 return $this->dateFormats;
330 }
331
332 function getDefaultDateFormat() {
333 $this->load();
334 return $this->defaultDateFormat;
335 }
336
337 function getDatePreferenceMigrationMap() {
338 $this->load();
339 return $this->datePreferenceMigrationMap;
340 }
341
342 function getDefaultUserOptionOverrides() {
343 $this->load();
344 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
345 if (is_array($this->defaultUserOptionOverrides)) {
346 return $this->defaultUserOptionOverrides;
347 } else {
348 return array();
349 }
350 }
351
352 function getExtraUserToggles() {
353 $this->load();
354 return $this->extraUserToggles;
355 }
356
357 function getUserToggle( $tog ) {
358 return $this->getMessageFromDB( "tog-$tog" );
359 }
360
361 /**
362 * Get language names, indexed by code.
363 * If $customisedOnly is true, only returns codes with a messages file
364 */
365 public static function getLanguageNames( $customisedOnly = false ) {
366 global $wgLanguageNames, $wgExtraLanguageNames;
367 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
368 if ( !$customisedOnly ) {
369 return $allNames;
370 }
371
372 global $IP;
373 $names = array();
374 $dir = opendir( "$IP/languages/messages" );
375 while( false !== ( $file = readdir( $dir ) ) ) {
376 $m = array();
377 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
378 $code = str_replace( '_', '-', strtolower( $m[1] ) );
379 if ( isset( $allNames[$code] ) ) {
380 $names[$code] = $allNames[$code];
381 }
382 }
383 }
384 closedir( $dir );
385 return $names;
386 }
387
388 /**
389 * Ugly hack to get a message maybe from the MediaWiki namespace, if this
390 * language object is the content or user language.
391 */
392 function getMessageFromDB( $msg ) {
393 global $wgContLang, $wgLang;
394 if ( $wgContLang->getCode() == $this->getCode() ) {
395 # Content language
396 return wfMsgForContent( $msg );
397 } elseif ( $wgLang->getCode() == $this->getCode() ) {
398 # User language
399 return wfMsg( $msg );
400 } else {
401 # Neither, get from localisation
402 return $this->getMessage( $msg );
403 }
404 }
405
406 function getLanguageName( $code ) {
407 $names = self::getLanguageNames();
408 if ( !array_key_exists( $code, $names ) ) {
409 return '';
410 }
411 return $names[$code];
412 }
413
414 function getMonthName( $key ) {
415 return $this->getMessageFromDB( self::$mMonthMsgs[$key-1] );
416 }
417
418 function getMonthNameGen( $key ) {
419 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key-1] );
420 }
421
422 function getMonthAbbreviation( $key ) {
423 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key-1] );
424 }
425
426 function getWeekdayName( $key ) {
427 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key-1] );
428 }
429
430 function getWeekdayAbbreviation( $key ) {
431 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key-1] );
432 }
433
434 function getIranianCalendarMonthName( $key ) {
435 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key-1] );
436 }
437
438 function getHebrewCalendarMonthName( $key ) {
439 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key-1] );
440 }
441
442 function getHebrewCalendarMonthNameGen( $key ) {
443 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key-1] );
444 }
445
446 function getHijriCalendarMonthName( $key ) {
447 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key-1] );
448 }
449
450 /**
451 * Used by date() and time() to adjust the time output.
452 *
453 * @param $ts Int the time in date('YmdHis') format
454 * @param $tz Mixed: adjust the time by this amount (default false, mean we
455 * get user timecorrection setting)
456 * @return int
457 */
458 function userAdjust( $ts, $tz = false ) {
459 global $wgUser, $wgLocalTZoffset;
460
461 if (!$tz) {
462 $tz = $wgUser->getOption( 'timecorrection' );
463 }
464
465 # minutes and hours differences:
466 $minDiff = 0;
467 $hrDiff = 0;
468
469 if ( $tz === '' ) {
470 # Global offset in minutes.
471 if( isset($wgLocalTZoffset) ) {
472 if( $wgLocalTZoffset >= 0 ) {
473 $hrDiff = floor($wgLocalTZoffset / 60);
474 } else {
475 $hrDiff = ceil($wgLocalTZoffset / 60);
476 }
477 $minDiff = $wgLocalTZoffset % 60;
478 }
479 } elseif ( strpos( $tz, ':' ) !== false ) {
480 $tzArray = explode( ':', $tz );
481 $hrDiff = intval($tzArray[0]);
482 $minDiff = intval($hrDiff < 0 ? -$tzArray[1] : $tzArray[1]);
483 } else {
484 $hrDiff = intval( $tz );
485 }
486
487 # No difference ? Return time unchanged
488 if ( 0 == $hrDiff && 0 == $minDiff ) { return $ts; }
489
490 wfSuppressWarnings(); // E_STRICT system time bitching
491 # Generate an adjusted date
492 $t = mktime( (
493 (int)substr( $ts, 8, 2) ) + $hrDiff, # Hours
494 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
495 (int)substr( $ts, 12, 2 ), # Seconds
496 (int)substr( $ts, 4, 2 ), # Month
497 (int)substr( $ts, 6, 2 ), # Day
498 (int)substr( $ts, 0, 4 ) ); #Year
499
500 $date = date( 'YmdHis', $t );
501 wfRestoreWarnings();
502
503 return $date;
504 }
505
506 /**
507 * This is a workalike of PHP's date() function, but with better
508 * internationalisation, a reduced set of format characters, and a better
509 * escaping format.
510 *
511 * Supported format characters are dDjlNwzWFmMntLYyaAgGhHiscrU. See the
512 * PHP manual for definitions. There are a number of extensions, which
513 * start with "x":
514 *
515 * xn Do not translate digits of the next numeric format character
516 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
517 * xr Use roman numerals for the next numeric format character
518 * xh Use hebrew numerals for the next numeric format character
519 * xx Literal x
520 * xg Genitive month name
521 *
522 * xij j (day number) in Iranian calendar
523 * xiF F (month name) in Iranian calendar
524 * xin n (month number) in Iranian calendar
525 * xiY Y (full year) in Iranian calendar
526 *
527 * xjj j (day number) in Hebrew calendar
528 * xjF F (month name) in Hebrew calendar
529 * xjt t (days in month) in Hebrew calendar
530 * xjx xg (genitive month name) in Hebrew calendar
531 * xjn n (month number) in Hebrew calendar
532 * xjY Y (full year) in Hebrew calendar
533 *
534 * xmj j (day number) in Hijri calendar
535 * xmF F (month name) in Hijri calendar
536 * xmn n (month number) in Hijri calendar
537 * xmY Y (full year) in Hijri calendar
538 *
539 * xkY Y (full year) in Thai solar calendar. Months and days are
540 * identical to the Gregorian calendar
541 *
542 * Characters enclosed in double quotes will be considered literal (with
543 * the quotes themselves removed). Unmatched quotes will be considered
544 * literal quotes. Example:
545 *
546 * "The month is" F => The month is January
547 * i's" => 20'11"
548 *
549 * Backslash escaping is also supported.
550 *
551 * Input timestamp is assumed to be pre-normalized to the desired local
552 * time zone, if any.
553 *
554 * @param $format String
555 * @param $ts String: 14-character timestamp
556 * YYYYMMDDHHMMSS
557 * 01234567890123
558 */
559 function sprintfDate( $format, $ts ) {
560 $s = '';
561 $raw = false;
562 $roman = false;
563 $hebrewNum = false;
564 $unix = false;
565 $rawToggle = false;
566 $iranian = false;
567 $hebrew = false;
568 $hijri = false;
569 $thai = false;
570 for ( $p = 0; $p < strlen( $format ); $p++ ) {
571 $num = false;
572 $code = $format[$p];
573 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
574 $code .= $format[++$p];
575 }
576
577 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' ) && $p < strlen( $format ) - 1 ) {
578 $code .= $format[++$p];
579 }
580
581 switch ( $code ) {
582 case 'xx':
583 $s .= 'x';
584 break;
585 case 'xn':
586 $raw = true;
587 break;
588 case 'xN':
589 $rawToggle = !$rawToggle;
590 break;
591 case 'xr':
592 $roman = true;
593 break;
594 case 'xh':
595 $hebrewNum = true;
596 break;
597 case 'xg':
598 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
599 break;
600 case 'xjx':
601 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
602 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
603 break;
604 case 'd':
605 $num = substr( $ts, 6, 2 );
606 break;
607 case 'D':
608 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
609 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
610 break;
611 case 'j':
612 $num = intval( substr( $ts, 6, 2 ) );
613 break;
614 case 'xij':
615 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
616 $num = $iranian[2];
617 break;
618 case 'xmj':
619 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
620 $num = $hijri[2];
621 break;
622 case 'xjj':
623 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
624 $num = $hebrew[2];
625 break;
626 case 'l':
627 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
628 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
629 break;
630 case 'N':
631 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
632 $w = gmdate( 'w', $unix );
633 $num = $w ? $w : 7;
634 break;
635 case 'w':
636 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
637 $num = gmdate( 'w', $unix );
638 break;
639 case 'z':
640 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
641 $num = gmdate( 'z', $unix );
642 break;
643 case 'W':
644 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
645 $num = gmdate( 'W', $unix );
646 break;
647 case 'F':
648 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
649 break;
650 case 'xiF':
651 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
652 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
653 break;
654 case 'xmF':
655 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
656 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
657 break;
658 case 'xjF':
659 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
660 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
661 break;
662 case 'm':
663 $num = substr( $ts, 4, 2 );
664 break;
665 case 'M':
666 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
667 break;
668 case 'n':
669 $num = intval( substr( $ts, 4, 2 ) );
670 break;
671 case 'xin':
672 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
673 $num = $iranian[1];
674 break;
675 case 'xmn':
676 if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
677 $num = $hijri[1];
678 break;
679 case 'xjn':
680 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
681 $num = $hebrew[1];
682 break;
683 case 't':
684 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
685 $num = gmdate( 't', $unix );
686 break;
687 case 'xjt':
688 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
689 $num = $hebrew[3];
690 break;
691 case 'L':
692 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
693 $num = gmdate( 'L', $unix );
694 break;
695 case 'Y':
696 $num = substr( $ts, 0, 4 );
697 break;
698 case 'xiY':
699 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
700 $num = $iranian[0];
701 break;
702 case 'xmY':
703 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
704 $num = $hijri[0];
705 break;
706 case 'xjY':
707 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
708 $num = $hebrew[0];
709 break;
710 case 'xkY':
711 if ( !$thai ) $thai = self::tsToThai( $ts );
712 $num = $thai[0];
713 break;
714 case 'y':
715 $num = substr( $ts, 2, 2 );
716 break;
717 case 'a':
718 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
719 break;
720 case 'A':
721 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
722 break;
723 case 'g':
724 $h = substr( $ts, 8, 2 );
725 $num = $h % 12 ? $h % 12 : 12;
726 break;
727 case 'G':
728 $num = intval( substr( $ts, 8, 2 ) );
729 break;
730 case 'h':
731 $h = substr( $ts, 8, 2 );
732 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
733 break;
734 case 'H':
735 $num = substr( $ts, 8, 2 );
736 break;
737 case 'i':
738 $num = substr( $ts, 10, 2 );
739 break;
740 case 's':
741 $num = substr( $ts, 12, 2 );
742 break;
743 case 'c':
744 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
745 $s .= gmdate( 'c', $unix );
746 break;
747 case 'r':
748 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
749 $s .= gmdate( 'r', $unix );
750 break;
751 case 'U':
752 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
753 $num = $unix;
754 break;
755 case '\\':
756 # Backslash escaping
757 if ( $p < strlen( $format ) - 1 ) {
758 $s .= $format[++$p];
759 } else {
760 $s .= '\\';
761 }
762 break;
763 case '"':
764 # Quoted literal
765 if ( $p < strlen( $format ) - 1 ) {
766 $endQuote = strpos( $format, '"', $p + 1 );
767 if ( $endQuote === false ) {
768 # No terminating quote, assume literal "
769 $s .= '"';
770 } else {
771 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
772 $p = $endQuote;
773 }
774 } else {
775 # Quote at end of string, assume literal "
776 $s .= '"';
777 }
778 break;
779 default:
780 $s .= $format[$p];
781 }
782 if ( $num !== false ) {
783 if ( $rawToggle || $raw ) {
784 $s .= $num;
785 $raw = false;
786 } elseif ( $roman ) {
787 $s .= self::romanNumeral( $num );
788 $roman = false;
789 } elseif( $hebrewNum ) {
790 $s .= self::hebrewNumeral( $num );
791 $hebrewNum = false;
792 } else {
793 $s .= $this->formatNum( $num, true );
794 }
795 $num = false;
796 }
797 }
798 return $s;
799 }
800
801 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
802 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
803 /**
804 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
805 * Gregorian dates to Iranian dates. Originally written in C, it
806 * is released under the terms of GNU Lesser General Public
807 * License. Conversion to PHP was performed by Niklas Laxström.
808 *
809 * Link: http://www.farsiweb.info/jalali/jalali.c
810 */
811 private static function tsToIranian( $ts ) {
812 $gy = substr( $ts, 0, 4 ) -1600;
813 $gm = substr( $ts, 4, 2 ) -1;
814 $gd = substr( $ts, 6, 2 ) -1;
815
816 # Days passed from the beginning (including leap years)
817 $gDayNo = 365*$gy
818 + floor(($gy+3) / 4)
819 - floor(($gy+99) / 100)
820 + floor(($gy+399) / 400);
821
822
823 // Add days of the past months of this year
824 for( $i = 0; $i < $gm; $i++ ) {
825 $gDayNo += self::$GREG_DAYS[$i];
826 }
827
828 // Leap years
829 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
830 $gDayNo++;
831 }
832
833 // Days passed in current month
834 $gDayNo += $gd;
835
836 $jDayNo = $gDayNo - 79;
837
838 $jNp = floor($jDayNo / 12053);
839 $jDayNo %= 12053;
840
841 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
842 $jDayNo %= 1461;
843
844 if ( $jDayNo >= 366 ) {
845 $jy += floor(($jDayNo-1)/365);
846 $jDayNo = floor(($jDayNo-1)%365);
847 }
848
849 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
850 $jDayNo -= self::$IRANIAN_DAYS[$i];
851 }
852
853 $jm= $i+1;
854 $jd= $jDayNo+1;
855
856 return array($jy, $jm, $jd);
857 }
858 /**
859 * Converting Gregorian dates to Hijri dates.
860 *
861 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
862 *
863 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
864 */
865 private static function tsToHijri ( $ts ) {
866 $year = substr( $ts, 0, 4 );
867 $month = substr( $ts, 4, 2 );
868 $day = substr( $ts, 6, 2 );
869
870 $zyr = $year;
871 $zd=$day;
872 $zm=$month;
873 $zy=$zyr;
874
875
876
877 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
878 {
879
880
881 $zjd=(int)((1461*($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;
882 }
883 else
884 {
885 $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
886 }
887
888 $zl=$zjd-1948440+10632;
889 $zn=(int)(($zl-1)/10631);
890 $zl=$zl-10631*$zn+354;
891 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
892 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
893 $zm=(int)((24*$zl)/709);
894 $zd=$zl-(int)((709*$zm)/24);
895 $zy=30*$zn+$zj-30;
896
897 return array ($zy, $zm, $zd);
898 }
899
900 /**
901 * Converting Gregorian dates to Hebrew dates.
902 *
903 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
904 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
905 * to translate the relevant functions into PHP and release them under
906 * GNU GPL.
907 */
908 private static function tsToHebrew( $ts ) {
909 # Parse date
910 $year = substr( $ts, 0, 4 );
911 $month = substr( $ts, 4, 2 );
912 $day = substr( $ts, 6, 2 );
913
914 # Calculate Hebrew year
915 $hebrewYear = $year + 3760;
916
917 # Month number when September = 1, August = 12
918 $month += 4;
919 if( $month > 12 ) {
920 # Next year
921 $month -= 12;
922 $year++;
923 $hebrewYear++;
924 }
925
926 # Calculate day of year from 1 September
927 $dayOfYear = $day;
928 for( $i = 1; $i < $month; $i++ ) {
929 if( $i == 6 ) {
930 # February
931 $dayOfYear += 28;
932 # Check if the year is leap
933 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
934 $dayOfYear++;
935 }
936 } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
937 $dayOfYear += 30;
938 } else {
939 $dayOfYear += 31;
940 }
941 }
942
943 # Calculate the start of the Hebrew year
944 $start = self::hebrewYearStart( $hebrewYear );
945
946 # Calculate next year's start
947 if( $dayOfYear <= $start ) {
948 # Day is before the start of the year - it is the previous year
949 # Next year's start
950 $nextStart = $start;
951 # Previous year
952 $year--;
953 $hebrewYear--;
954 # Add days since previous year's 1 September
955 $dayOfYear += 365;
956 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
957 # Leap year
958 $dayOfYear++;
959 }
960 # Start of the new (previous) year
961 $start = self::hebrewYearStart( $hebrewYear );
962 } else {
963 # Next year's start
964 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
965 }
966
967 # Calculate Hebrew day of year
968 $hebrewDayOfYear = $dayOfYear - $start;
969
970 # Difference between year's days
971 $diff = $nextStart - $start;
972 # Add 12 (or 13 for leap years) days to ignore the difference between
973 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
974 # difference is only about the year type
975 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
976 $diff += 13;
977 } else {
978 $diff += 12;
979 }
980
981 # Check the year pattern, and is leap year
982 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
983 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
984 # and non-leap years
985 $yearPattern = $diff % 30;
986 # Check if leap year
987 $isLeap = $diff >= 30;
988
989 # Calculate day in the month from number of day in the Hebrew year
990 # Don't check Adar - if the day is not in Adar, we will stop before;
991 # if it is in Adar, we will use it to check if it is Adar I or Adar II
992 $hebrewDay = $hebrewDayOfYear;
993 $hebrewMonth = 1;
994 $days = 0;
995 while( $hebrewMonth <= 12 ) {
996 # Calculate days in this month
997 if( $isLeap && $hebrewMonth == 6 ) {
998 # Adar in a leap year
999 if( $isLeap ) {
1000 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1001 $days = 30;
1002 if( $hebrewDay <= $days ) {
1003 # Day in Adar I
1004 $hebrewMonth = 13;
1005 } else {
1006 # Subtract the days of Adar I
1007 $hebrewDay -= $days;
1008 # Try Adar II
1009 $days = 29;
1010 if( $hebrewDay <= $days ) {
1011 # Day in Adar II
1012 $hebrewMonth = 14;
1013 }
1014 }
1015 }
1016 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1017 # Cheshvan in a complete year (otherwise as the rule below)
1018 $days = 30;
1019 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1020 # Kislev in an incomplete year (otherwise as the rule below)
1021 $days = 29;
1022 } else {
1023 # Odd months have 30 days, even have 29
1024 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1025 }
1026 if( $hebrewDay <= $days ) {
1027 # In the current month
1028 break;
1029 } else {
1030 # Subtract the days of the current month
1031 $hebrewDay -= $days;
1032 # Try in the next month
1033 $hebrewMonth++;
1034 }
1035 }
1036
1037 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1038 }
1039
1040 /**
1041 * This calculates the Hebrew year start, as days since 1 September.
1042 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1043 * Used for Hebrew date.
1044 */
1045 private static function hebrewYearStart( $year ) {
1046 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1047 $b = intval( ( $year - 1 ) % 4 );
1048 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1049 if( $m < 0 ) {
1050 $m--;
1051 }
1052 $Mar = intval( $m );
1053 if( $m < 0 ) {
1054 $m++;
1055 }
1056 $m -= $Mar;
1057
1058 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1059 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1060 $Mar++;
1061 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1062 $Mar += 2;
1063 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1064 $Mar++;
1065 }
1066
1067 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1068 return $Mar;
1069 }
1070
1071 /**
1072 * Algorithm to convert Gregorian dates to Thai solar dates.
1073 *
1074 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1075 *
1076 * @param $ts String: 14-character timestamp
1077 * @return array converted year, month, day
1078 */
1079 private static function tsToThai( $ts ) {
1080 $gy = substr( $ts, 0, 4 );
1081 $gm = substr( $ts, 4, 2 );
1082 $gd = substr( $ts, 6, 2 );
1083
1084 # Add 543 years to the Gregorian calendar
1085 # Months and days are identical
1086 $gy_thai = $gy + 543;
1087
1088 return array( $gy_thai, $gm, $gd );
1089 }
1090
1091
1092 /**
1093 * Roman number formatting up to 3000
1094 */
1095 static function romanNumeral( $num ) {
1096 static $table = array(
1097 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1098 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1099 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1100 array( '', 'M', 'MM', 'MMM' )
1101 );
1102
1103 $num = intval( $num );
1104 if ( $num > 3000 || $num <= 0 ) {
1105 return $num;
1106 }
1107
1108 $s = '';
1109 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1110 if ( $num >= $pow10 ) {
1111 $s .= $table[$i][floor($num / $pow10)];
1112 }
1113 $num = $num % $pow10;
1114 }
1115 return $s;
1116 }
1117
1118 /**
1119 * Hebrew Gematria number formatting up to 9999
1120 */
1121 static function hebrewNumeral( $num ) {
1122 static $table = array(
1123 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1124 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1125 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1126 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1127 );
1128
1129 $num = intval( $num );
1130 if ( $num > 9999 || $num <= 0 ) {
1131 return $num;
1132 }
1133
1134 $s = '';
1135 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1136 if ( $num >= $pow10 ) {
1137 if ( $num == 15 || $num == 16 ) {
1138 $s .= $table[0][9] . $table[0][$num - 9];
1139 $num = 0;
1140 } else {
1141 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1142 if( $pow10 == 1000 ) {
1143 $s .= "'";
1144 }
1145 }
1146 }
1147 $num = $num % $pow10;
1148 }
1149 if( strlen( $s ) == 2 ) {
1150 $str = $s . "'";
1151 } else {
1152 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1153 $str .= substr( $s, strlen( $s ) - 2, 2 );
1154 }
1155 $start = substr( $str, 0, strlen( $str ) - 2 );
1156 $end = substr( $str, strlen( $str ) - 2 );
1157 switch( $end ) {
1158 case 'כ':
1159 $str = $start . 'ך';
1160 break;
1161 case 'מ':
1162 $str = $start . 'ם';
1163 break;
1164 case 'נ':
1165 $str = $start . 'ן';
1166 break;
1167 case 'פ':
1168 $str = $start . 'ף';
1169 break;
1170 case 'צ':
1171 $str = $start . 'ץ';
1172 break;
1173 }
1174 return $str;
1175 }
1176
1177 /**
1178 * This is meant to be used by time(), date(), and timeanddate() to get
1179 * the date preference they're supposed to use, it should be used in
1180 * all children.
1181 *
1182 *<code>
1183 * function timeanddate([...], $format = true) {
1184 * $datePreference = $this->dateFormat($format);
1185 * [...]
1186 * }
1187 *</code>
1188 *
1189 * @param $usePrefs Mixed: if true, the user's preference is used
1190 * if false, the site/language default is used
1191 * if int/string, assumed to be a format.
1192 * @return string
1193 */
1194 function dateFormat( $usePrefs = true ) {
1195 global $wgUser;
1196
1197 if( is_bool( $usePrefs ) ) {
1198 if( $usePrefs ) {
1199 $datePreference = $wgUser->getDatePreference();
1200 } else {
1201 $options = User::getDefaultOptions();
1202 $datePreference = (string)$options['date'];
1203 }
1204 } else {
1205 $datePreference = (string)$usePrefs;
1206 }
1207
1208 // return int
1209 if( $datePreference == '' ) {
1210 return 'default';
1211 }
1212
1213 return $datePreference;
1214 }
1215
1216 /**
1217 * @param $ts Mixed: the time format which needs to be turned into a
1218 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1219 * @param $adj Bool: whether to adjust the time output according to the
1220 * user configured offset ($timecorrection)
1221 * @param $format Mixed: true to use user's date format preference
1222 * @param $timecorrection String: the time offset as returned by
1223 * validateTimeZone() in Special:Preferences
1224 * @return string
1225 */
1226 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1227 $this->load();
1228 if ( $adj ) {
1229 $ts = $this->userAdjust( $ts, $timecorrection );
1230 }
1231
1232 $pref = $this->dateFormat( $format );
1233 if( $pref == 'default' || !isset( $this->dateFormats["$pref date"] ) ) {
1234 $pref = $this->defaultDateFormat;
1235 }
1236 return $this->sprintfDate( $this->dateFormats["$pref date"], $ts );
1237 }
1238
1239 /**
1240 * @param $ts Mixed: the time format which needs to be turned into a
1241 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1242 * @param $adj Bool: whether to adjust the time output according to the
1243 * user configured offset ($timecorrection)
1244 * @param $format Mixed: true to use user's date format preference
1245 * @param $timecorrection String: the time offset as returned by
1246 * validateTimeZone() in Special:Preferences
1247 * @return string
1248 */
1249 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1250 $this->load();
1251 if ( $adj ) {
1252 $ts = $this->userAdjust( $ts, $timecorrection );
1253 }
1254
1255 $pref = $this->dateFormat( $format );
1256 if( $pref == 'default' || !isset( $this->dateFormats["$pref time"] ) ) {
1257 $pref = $this->defaultDateFormat;
1258 }
1259 return $this->sprintfDate( $this->dateFormats["$pref time"], $ts );
1260 }
1261
1262 /**
1263 * @param $ts Mixed: the time format which needs to be turned into a
1264 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1265 * @param $adj Bool: whether to adjust the time output according to the
1266 * user configured offset ($timecorrection)
1267 * @param $format Mixed: what format to return, if it's false output the
1268 * default one (default true)
1269 * @param $timecorrection String: the time offset as returned by
1270 * validateTimeZone() in Special:Preferences
1271 * @return string
1272 */
1273 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1274 $this->load();
1275
1276 $ts = wfTimestamp( TS_MW, $ts );
1277
1278 if ( $adj ) {
1279 $ts = $this->userAdjust( $ts, $timecorrection );
1280 }
1281
1282 $pref = $this->dateFormat( $format );
1283 if( $pref == 'default' || !isset( $this->dateFormats["$pref both"] ) ) {
1284 $pref = $this->defaultDateFormat;
1285 }
1286
1287 return $this->sprintfDate( $this->dateFormats["$pref both"], $ts );
1288 }
1289
1290 function getMessage( $key ) {
1291 $this->load();
1292 return isset( $this->messages[$key] ) ? $this->messages[$key] : null;
1293 }
1294
1295 function getAllMessages() {
1296 $this->load();
1297 return $this->messages;
1298 }
1299
1300 function iconv( $in, $out, $string ) {
1301 # For most languages, this is a wrapper for iconv
1302 return iconv( $in, $out . '//IGNORE', $string );
1303 }
1304
1305 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1306 function ucwordbreaksCallbackAscii($matches){
1307 return $this->ucfirst($matches[1]);
1308 }
1309
1310 function ucwordbreaksCallbackMB($matches){
1311 return mb_strtoupper($matches[0]);
1312 }
1313
1314 function ucCallback($matches){
1315 list( $wikiUpperChars ) = self::getCaseMaps();
1316 return strtr( $matches[1], $wikiUpperChars );
1317 }
1318
1319 function lcCallback($matches){
1320 list( , $wikiLowerChars ) = self::getCaseMaps();
1321 return strtr( $matches[1], $wikiLowerChars );
1322 }
1323
1324 function ucwordsCallbackMB($matches){
1325 return mb_strtoupper($matches[0]);
1326 }
1327
1328 function ucwordsCallbackWiki($matches){
1329 list( $wikiUpperChars ) = self::getCaseMaps();
1330 return strtr( $matches[0], $wikiUpperChars );
1331 }
1332
1333 function ucfirst( $str ) {
1334 if ( empty($str) ) return $str;
1335 if ( ord($str[0]) < 128 ) return ucfirst($str);
1336 else return self::uc($str,true); // fall back to more complex logic in case of multibyte strings
1337 }
1338
1339 function uc( $str, $first = false ) {
1340 if ( function_exists( 'mb_strtoupper' ) ) {
1341 if ( $first ) {
1342 if ( self::isMultibyte( $str ) ) {
1343 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1344 } else {
1345 return ucfirst( $str );
1346 }
1347 } else {
1348 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1349 }
1350 } else {
1351 if ( self::isMultibyte( $str ) ) {
1352 list( $wikiUpperChars ) = $this->getCaseMaps();
1353 $x = $first ? '^' : '';
1354 return preg_replace_callback(
1355 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1356 array($this,"ucCallback"),
1357 $str
1358 );
1359 } else {
1360 return $first ? ucfirst( $str ) : strtoupper( $str );
1361 }
1362 }
1363 }
1364
1365 function lcfirst( $str ) {
1366 if ( empty($str) ) return $str;
1367 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1368 // editing string in place = cool
1369 $str[0]=strtolower($str[0]);
1370 return $str;
1371 }
1372 else return self::lc( $str, true );
1373 }
1374
1375 function lc( $str, $first = false ) {
1376 if ( function_exists( 'mb_strtolower' ) )
1377 if ( $first )
1378 if ( self::isMultibyte( $str ) )
1379 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1380 else
1381 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1382 else
1383 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1384 else
1385 if ( self::isMultibyte( $str ) ) {
1386 list( , $wikiLowerChars ) = self::getCaseMaps();
1387 $x = $first ? '^' : '';
1388 return preg_replace_callback(
1389 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1390 array($this,"lcCallback"),
1391 $str
1392 );
1393 } else
1394 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1395 }
1396
1397 function isMultibyte( $str ) {
1398 return (bool)preg_match( '/[\x80-\xff]/', $str );
1399 }
1400
1401 function ucwords($str) {
1402 if ( self::isMultibyte( $str ) ) {
1403 $str = self::lc($str);
1404
1405 // regexp to find first letter in each word (i.e. after each space)
1406 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1407
1408 // function to use to capitalize a single char
1409 if ( function_exists( 'mb_strtoupper' ) )
1410 return preg_replace_callback(
1411 $replaceRegexp,
1412 array($this,"ucwordsCallbackMB"),
1413 $str
1414 );
1415 else
1416 return preg_replace_callback(
1417 $replaceRegexp,
1418 array($this,"ucwordsCallbackWiki"),
1419 $str
1420 );
1421 }
1422 else
1423 return ucwords( strtolower( $str ) );
1424 }
1425
1426 # capitalize words at word breaks
1427 function ucwordbreaks($str){
1428 if (self::isMultibyte( $str ) ) {
1429 $str = self::lc($str);
1430
1431 // since \b doesn't work for UTF-8, we explicitely define word break chars
1432 $breaks= "[ \-\(\)\}\{\.,\?!]";
1433
1434 // find first letter after word break
1435 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1436
1437 if ( function_exists( 'mb_strtoupper' ) )
1438 return preg_replace_callback(
1439 $replaceRegexp,
1440 array($this,"ucwordbreaksCallbackMB"),
1441 $str
1442 );
1443 else
1444 return preg_replace_callback(
1445 $replaceRegexp,
1446 array($this,"ucwordsCallbackWiki"),
1447 $str
1448 );
1449 }
1450 else
1451 return preg_replace_callback(
1452 '/\b([\w\x80-\xff]+)\b/',
1453 array($this,"ucwordbreaksCallbackAscii"),
1454 $str );
1455 }
1456
1457 /**
1458 * Return a case-folded representation of $s
1459 *
1460 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1461 * and $s2 are the same except for the case of their characters. It is not
1462 * necessary for the value returned to make sense when displayed.
1463 *
1464 * Do *not* perform any other normalisation in this function. If a caller
1465 * uses this function when it should be using a more general normalisation
1466 * function, then fix the caller.
1467 */
1468 function caseFold( $s ) {
1469 return $this->uc( $s );
1470 }
1471
1472 function checkTitleEncoding( $s ) {
1473 if( is_array( $s ) ) {
1474 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1475 }
1476 # Check for non-UTF-8 URLs
1477 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1478 if(!$ishigh) return $s;
1479
1480 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1481 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1482 if( $isutf8 ) return $s;
1483
1484 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1485 }
1486
1487 function fallback8bitEncoding() {
1488 $this->load();
1489 return $this->fallback8bitEncoding;
1490 }
1491
1492 /**
1493 * Some languages have special punctuation to strip out
1494 * or characters which need to be converted for MySQL's
1495 * indexing to grok it correctly. Make such changes here.
1496 *
1497 * @param $string String
1498 * @return String
1499 */
1500 function stripForSearch( $string ) {
1501 global $wgDBtype;
1502 if ( $wgDBtype != 'mysql' ) {
1503 return $string;
1504 }
1505
1506 # MySQL fulltext index doesn't grok utf-8, so we
1507 # need to fold cases and convert to hex
1508
1509 wfProfileIn( __METHOD__ );
1510 if( function_exists( 'mb_strtolower' ) ) {
1511 $out = preg_replace(
1512 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1513 "'U8' . bin2hex( \"$1\" )",
1514 mb_strtolower( $string ) );
1515 } else {
1516 list( , $wikiLowerChars ) = self::getCaseMaps();
1517 $out = preg_replace(
1518 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1519 "'U8' . bin2hex( strtr( \"\$1\", \$wikiLowerChars ) )",
1520 $string );
1521 }
1522 wfProfileOut( __METHOD__ );
1523 return $out;
1524 }
1525
1526 function convertForSearchResult( $termsArray ) {
1527 # some languages, e.g. Chinese, need to do a conversion
1528 # in order for search results to be displayed correctly
1529 return $termsArray;
1530 }
1531
1532 /**
1533 * Get the first character of a string.
1534 *
1535 * @param $s string
1536 * @return string
1537 */
1538 function firstChar( $s ) {
1539 $matches = array();
1540 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1541 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1542
1543 if ( isset( $matches[1] ) ) {
1544 if ( strlen( $matches[1] ) != 3 ) {
1545 return $matches[1];
1546 }
1547
1548 // Break down Hangul syllables to grab the first jamo
1549 $code = utf8ToCodepoint( $matches[1] );
1550 if ( $code < 0xac00 || 0xd7a4 <= $code) {
1551 return $matches[1];
1552 } elseif ( $code < 0xb098 ) {
1553 return "\xe3\x84\xb1";
1554 } elseif ( $code < 0xb2e4 ) {
1555 return "\xe3\x84\xb4";
1556 } elseif ( $code < 0xb77c ) {
1557 return "\xe3\x84\xb7";
1558 } elseif ( $code < 0xb9c8 ) {
1559 return "\xe3\x84\xb9";
1560 } elseif ( $code < 0xbc14 ) {
1561 return "\xe3\x85\x81";
1562 } elseif ( $code < 0xc0ac ) {
1563 return "\xe3\x85\x82";
1564 } elseif ( $code < 0xc544 ) {
1565 return "\xe3\x85\x85";
1566 } elseif ( $code < 0xc790 ) {
1567 return "\xe3\x85\x87";
1568 } elseif ( $code < 0xcc28 ) {
1569 return "\xe3\x85\x88";
1570 } elseif ( $code < 0xce74 ) {
1571 return "\xe3\x85\x8a";
1572 } elseif ( $code < 0xd0c0 ) {
1573 return "\xe3\x85\x8b";
1574 } elseif ( $code < 0xd30c ) {
1575 return "\xe3\x85\x8c";
1576 } elseif ( $code < 0xd558 ) {
1577 return "\xe3\x85\x8d";
1578 } else {
1579 return "\xe3\x85\x8e";
1580 }
1581 } else {
1582 return "";
1583 }
1584 }
1585
1586 function initEncoding() {
1587 # Some languages may have an alternate char encoding option
1588 # (Esperanto X-coding, Japanese furigana conversion, etc)
1589 # If this language is used as the primary content language,
1590 # an override to the defaults can be set here on startup.
1591 }
1592
1593 function recodeForEdit( $s ) {
1594 # For some languages we'll want to explicitly specify
1595 # which characters make it into the edit box raw
1596 # or are converted in some way or another.
1597 # Note that if wgOutputEncoding is different from
1598 # wgInputEncoding, this text will be further converted
1599 # to wgOutputEncoding.
1600 global $wgEditEncoding;
1601 if( $wgEditEncoding == '' or
1602 $wgEditEncoding == 'UTF-8' ) {
1603 return $s;
1604 } else {
1605 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1606 }
1607 }
1608
1609 function recodeInput( $s ) {
1610 # Take the previous into account.
1611 global $wgEditEncoding;
1612 if($wgEditEncoding != "") {
1613 $enc = $wgEditEncoding;
1614 } else {
1615 $enc = 'UTF-8';
1616 }
1617 if( $enc == 'UTF-8' ) {
1618 return $s;
1619 } else {
1620 return $this->iconv( $enc, 'UTF-8', $s );
1621 }
1622 }
1623
1624 /**
1625 * For right-to-left language support
1626 *
1627 * @return bool
1628 */
1629 function isRTL() {
1630 $this->load();
1631 return $this->rtl;
1632 }
1633
1634 /**
1635 * A hidden direction mark (LRM or RLM), depending on the language direction
1636 *
1637 * @return string
1638 */
1639 function getDirMark() {
1640 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1641 }
1642
1643 /**
1644 * An arrow, depending on the language direction
1645 *
1646 * @return string
1647 */
1648 function getArrow() {
1649 return $this->isRTL() ? '←' : '→';
1650 }
1651
1652 /**
1653 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1654 *
1655 * @return bool
1656 */
1657 function linkPrefixExtension() {
1658 $this->load();
1659 return $this->linkPrefixExtension;
1660 }
1661
1662 function &getMagicWords() {
1663 $this->load();
1664 return $this->magicWords;
1665 }
1666
1667 # Fill a MagicWord object with data from here
1668 function getMagic( &$mw ) {
1669 if ( !$this->mMagicHookDone ) {
1670 $this->mMagicHookDone = true;
1671 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1672 }
1673 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1674 $rawEntry = $this->mMagicExtensions[$mw->mId];
1675 } else {
1676 $magicWords =& $this->getMagicWords();
1677 if ( isset( $magicWords[$mw->mId] ) ) {
1678 $rawEntry = $magicWords[$mw->mId];
1679 } else {
1680 # Fall back to English if local list is incomplete
1681 $magicWords =& Language::getMagicWords();
1682 if ( !isset($magicWords[$mw->mId]) ) { throw new MWException("Magic word not found" ); }
1683 $rawEntry = $magicWords[$mw->mId];
1684 }
1685 }
1686
1687 if( !is_array( $rawEntry ) ) {
1688 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1689 } else {
1690 $mw->mCaseSensitive = $rawEntry[0];
1691 $mw->mSynonyms = array_slice( $rawEntry, 1 );
1692 }
1693 }
1694
1695 /**
1696 * Add magic words to the extension array
1697 */
1698 function addMagicWordsByLang( $newWords ) {
1699 $code = $this->getCode();
1700 $fallbackChain = array();
1701 while ( $code && !in_array( $code, $fallbackChain ) ) {
1702 $fallbackChain[] = $code;
1703 $code = self::getFallbackFor( $code );
1704 }
1705 if ( !in_array( 'en', $fallbackChain ) ) {
1706 $fallbackChain[] = 'en';
1707 }
1708 $fallbackChain = array_reverse( $fallbackChain );
1709 foreach ( $fallbackChain as $code ) {
1710 if ( isset( $newWords[$code] ) ) {
1711 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1712 }
1713 }
1714 }
1715
1716 /**
1717 * Get special page names, as an associative array
1718 * case folded alias => real name
1719 */
1720 function getSpecialPageAliases() {
1721 $this->load();
1722
1723 // Cache aliases because it may be slow to load them
1724 if ( !isset( $this->mExtendedSpecialPageAliases ) ) {
1725
1726 // Initialise array
1727 $this->mExtendedSpecialPageAliases = $this->specialPageAliases;
1728
1729 global $wgExtensionAliasesFiles;
1730 foreach ( $wgExtensionAliasesFiles as $file ) {
1731
1732 // Fail fast
1733 if ( !file_exists($file) )
1734 throw new MWException( 'Aliases file does not exist' );
1735
1736 $aliases = array();
1737 require($file);
1738
1739 // Check the availability of aliases
1740 if ( !isset($aliases['en']) )
1741 throw new MWException( 'Malformed aliases file' );
1742
1743 $code = $this->getCode();
1744
1745 if ( isset($aliases[$code]) ) {
1746 $aliases[$code] = $this->fixSpecialPageAliases( $aliases[$code] );
1747 /* Merge the aliases, THIS will break if there is special page name
1748 * which looks like a numerical key, thanks to PHP...
1749 * See the comments for wfArrayMerge in GlobalSettings.php. */
1750 $this->mExtendedSpecialPageAliases = array_merge_recursive(
1751 $this->mExtendedSpecialPageAliases, $aliases[$code] );
1752 }
1753
1754 /* Add the English aliases to the end of list as aliases... unless we
1755 * already added them! */
1756 if ( $code !== 'en' ) {
1757 $aliases['en'] = $this->fixSpecialPageAliases( $aliases['en'] );
1758 $this->mExtendedSpecialPageAliases = array_merge_recursive(
1759 $this->mExtendedSpecialPageAliases, $aliases['en'] );
1760 }
1761
1762 }
1763
1764 wfRunHooks( 'LanguageGetSpecialPageAliases',
1765 array( &$this->mExtendedSpecialPageAliases, $code ) );
1766 }
1767
1768 return $this->mExtendedSpecialPageAliases;
1769 }
1770
1771 /**
1772 * Function to fix special page aliases. Will convert the first letter to
1773 * upper case and spaces to underscores. Can be given a full aliases array,
1774 * in which case it will recursively fix all aliases.
1775 */
1776 public function fixSpecialPageAliases( $mixed ) {
1777 // Work recursively until in string level
1778 if ( is_array($mixed) ) {
1779 $callback = array( $this, 'fixSpecialPageAliases' );
1780 return array_map( $callback, $mixed );
1781 }
1782 return str_replace( ' ', '_', $this->ucfirst( $mixed ) );
1783 }
1784
1785 /**
1786 * Italic is unsuitable for some languages
1787 *
1788 * @param $text String: the text to be emphasized.
1789 * @return string
1790 */
1791 function emphasize( $text ) {
1792 return "<em>$text</em>";
1793 }
1794
1795 /**
1796 * Normally we output all numbers in plain en_US style, that is
1797 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1798 * point twohundredthirtyfive. However this is not sutable for all
1799 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1800 * Icelandic just want to use commas instead of dots, and dots instead
1801 * of commas like "293.291,235".
1802 *
1803 * An example of this function being called:
1804 * <code>
1805 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1806 * </code>
1807 *
1808 * See LanguageGu.php for the Gujarati implementation and
1809 * LanguageIs.php for the , => . and . => , implementation.
1810 *
1811 * @todo check if it's viable to use localeconv() for the decimal
1812 * seperator thing.
1813 * @param $number Mixed: the string to be formatted, should be an integer
1814 * or a floating point number.
1815 * @param $nocommafy Bool: set to true for special numbers like dates
1816 * @return string
1817 */
1818 function formatNum( $number, $nocommafy = false ) {
1819 global $wgTranslateNumerals;
1820 if (!$nocommafy) {
1821 $number = $this->commafy($number);
1822 $s = $this->separatorTransformTable();
1823 if (!is_null($s)) { $number = strtr($number, $s); }
1824 }
1825
1826 if ($wgTranslateNumerals) {
1827 $s = $this->digitTransformTable();
1828 if (!is_null($s)) { $number = strtr($number, $s); }
1829 }
1830
1831 return $number;
1832 }
1833
1834 function parseFormattedNumber( $number ) {
1835 $s = $this->digitTransformTable();
1836 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1837
1838 $s = $this->separatorTransformTable();
1839 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1840
1841 $number = strtr( $number, array (',' => '') );
1842 return $number;
1843 }
1844
1845 /**
1846 * Adds commas to a given number
1847 *
1848 * @param $_ mixed
1849 * @return string
1850 */
1851 function commafy($_) {
1852 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1853 }
1854
1855 function digitTransformTable() {
1856 $this->load();
1857 return $this->digitTransformTable;
1858 }
1859
1860 function separatorTransformTable() {
1861 $this->load();
1862 return $this->separatorTransformTable;
1863 }
1864
1865
1866 /**
1867 * For the credit list in includes/Credits.php (action=credits)
1868 *
1869 * @param $l Array
1870 * @return string
1871 */
1872 function listToText( $l ) {
1873 $s = '';
1874 $m = count($l) - 1;
1875 for ($i = $m; $i >= 0; $i--) {
1876 if ($i == $m) {
1877 $s = $l[$i];
1878 } else if ($i == $m - 1) {
1879 $s = $l[$i] . ' ' . $this->getMessageFromDB( 'and' ) . ' ' . $s;
1880 } else {
1881 $s = $l[$i] . ', ' . $s;
1882 }
1883 }
1884 return $s;
1885 }
1886
1887 /**
1888 * Truncate a string to a specified length in bytes, appending an optional
1889 * string (e.g. for ellipses)
1890 *
1891 * The database offers limited byte lengths for some columns in the database;
1892 * multi-byte character sets mean we need to ensure that only whole characters
1893 * are included, otherwise broken characters can be passed to the user
1894 *
1895 * If $length is negative, the string will be truncated from the beginning
1896 *
1897 * @param $string String to truncate
1898 * @param $length Int: maximum length (excluding ellipses)
1899 * @param $ellipsis String to append to the truncated text
1900 * @return string
1901 */
1902 function truncate( $string, $length, $ellipsis = "" ) {
1903 if( $length == 0 ) {
1904 return $ellipsis;
1905 }
1906 if ( strlen( $string ) <= abs( $length ) ) {
1907 return $string;
1908 }
1909 if( $length > 0 ) {
1910 $string = substr( $string, 0, $length );
1911 $char = ord( $string[strlen( $string ) - 1] );
1912 $m = array();
1913 if ($char >= 0xc0) {
1914 # We got the first byte only of a multibyte char; remove it.
1915 $string = substr( $string, 0, -1 );
1916 } elseif( $char >= 0x80 &&
1917 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
1918 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
1919 # We chopped in the middle of a character; remove it
1920 $string = $m[1];
1921 }
1922 return $string . $ellipsis;
1923 } else {
1924 $string = substr( $string, $length );
1925 $char = ord( $string[0] );
1926 if( $char >= 0x80 && $char < 0xc0 ) {
1927 # We chopped in the middle of a character; remove the whole thing
1928 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
1929 }
1930 return $ellipsis . $string;
1931 }
1932 }
1933
1934 /**
1935 * Grammatical transformations, needed for inflected languages
1936 * Invoked by putting {{grammar:case|word}} in a message
1937 *
1938 * @param $word string
1939 * @param $case string
1940 * @return string
1941 */
1942 function convertGrammar( $word, $case ) {
1943 global $wgGrammarForms;
1944 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
1945 return $wgGrammarForms[$this->getCode()][$case][$word];
1946 }
1947 return $word;
1948 }
1949
1950 /**
1951 * Plural form transformations, needed for some languages.
1952 * For example, there are 3 form of plural in Russian and Polish,
1953 * depending on "count mod 10". See [[w:Plural]]
1954 * For English it is pretty simple.
1955 *
1956 * Invoked by putting {{plural:count|wordform1|wordform2}}
1957 * or {{plural:count|wordform1|wordform2|wordform3}}
1958 *
1959 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
1960 *
1961 * @param $count Integer: non-localized number
1962 * @param $forms Array: different plural forms
1963 * @return string Correct form of plural for $count in this language
1964 */
1965 function convertPlural( $count, $forms ) {
1966 if ( !count($forms) ) { return ''; }
1967 $forms = $this->preConvertPlural( $forms, 2 );
1968
1969 return ( abs($count) == 1 ) ? $forms[0] : $forms[1];
1970 }
1971
1972 /**
1973 * Checks that convertPlural was given an array and pads it to requested
1974 * amound of forms by copying the last one.
1975 *
1976 * @param $count Integer: How many forms should there be at least
1977 * @param $forms Array of forms given to convertPlural
1978 * @return array Padded array of forms or an exception if not an array
1979 */
1980 protected function preConvertPlural( /* Array */ $forms, $count ) {
1981 while ( count($forms) < $count ) {
1982 $forms[] = $forms[count($forms)-1];
1983 }
1984 return $forms;
1985 }
1986
1987 /**
1988 * For translaing of expiry times
1989 * @param $str String: the validated block time in English
1990 * @return Somehow translated block time
1991 * @see LanguageFi.php for example implementation
1992 */
1993 function translateBlockExpiry( $str ) {
1994
1995 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
1996
1997 if ( $scBlockExpiryOptions == '-') {
1998 return $str;
1999 }
2000
2001 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2002 if ( strpos($option, ":") === false )
2003 continue;
2004 list($show, $value) = explode(":", $option);
2005 if ( strcmp ( $str, $value) == 0 ) {
2006 return htmlspecialchars( trim( $show ) );
2007 }
2008 }
2009
2010 return $str;
2011 }
2012
2013 /**
2014 * languages like Chinese need to be segmented in order for the diff
2015 * to be of any use
2016 *
2017 * @param $text String
2018 * @return String
2019 */
2020 function segmentForDiff( $text ) {
2021 return $text;
2022 }
2023
2024 /**
2025 * and unsegment to show the result
2026 *
2027 * @param $text String
2028 * @return String
2029 */
2030 function unsegmentForDiff( $text ) {
2031 return $text;
2032 }
2033
2034 # convert text to different variants of a language.
2035 function convert( $text, $isTitle = false) {
2036 return $this->mConverter->convert($text, $isTitle);
2037 }
2038
2039 # Convert text from within Parser
2040 function parserConvert( $text, &$parser ) {
2041 return $this->mConverter->parserConvert( $text, $parser );
2042 }
2043
2044 # Check if this is a language with variants
2045 function hasVariants(){
2046 return sizeof($this->getVariants())>1;
2047 }
2048
2049 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2050 function armourMath($text){
2051 return $this->mConverter->armourMath($text);
2052 }
2053
2054
2055 /**
2056 * Perform output conversion on a string, and encode for safe HTML output.
2057 * @param $text String
2058 * @param $isTitle Bool -- wtf?
2059 * @return string
2060 * @todo this should get integrated somewhere sane
2061 */
2062 function convertHtml( $text, $isTitle = false ) {
2063 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2064 }
2065
2066 function convertCategoryKey( $key ) {
2067 return $this->mConverter->convertCategoryKey( $key );
2068 }
2069
2070 /**
2071 * get the list of variants supported by this langauge
2072 * see sample implementation in LanguageZh.php
2073 *
2074 * @return array an array of language codes
2075 */
2076 function getVariants() {
2077 return $this->mConverter->getVariants();
2078 }
2079
2080
2081 function getPreferredVariant( $fromUser = true ) {
2082 return $this->mConverter->getPreferredVariant( $fromUser );
2083 }
2084
2085 /**
2086 * if a language supports multiple variants, it is
2087 * possible that non-existing link in one variant
2088 * actually exists in another variant. this function
2089 * tries to find it. See e.g. LanguageZh.php
2090 *
2091 * @param $link String: the name of the link
2092 * @param $nt Mixed: the title object of the link
2093 * @return null the input parameters may be modified upon return
2094 */
2095 function findVariantLink( &$link, &$nt ) {
2096 $this->mConverter->findVariantLink($link, $nt);
2097 }
2098
2099 /**
2100 * If a language supports multiple variants, converts text
2101 * into an array of all possible variants of the text:
2102 * 'variant' => text in that variant
2103 */
2104
2105 function convertLinkToAllVariants($text){
2106 return $this->mConverter->convertLinkToAllVariants($text);
2107 }
2108
2109
2110 /**
2111 * returns language specific options used by User::getPageRenderHash()
2112 * for example, the preferred language variant
2113 *
2114 * @return string
2115 */
2116 function getExtraHashOptions() {
2117 return $this->mConverter->getExtraHashOptions();
2118 }
2119
2120 /**
2121 * for languages that support multiple variants, the title of an
2122 * article may be displayed differently in different variants. this
2123 * function returns the apporiate title defined in the body of the article.
2124 *
2125 * @return string
2126 */
2127 function getParsedTitle() {
2128 return $this->mConverter->getParsedTitle();
2129 }
2130
2131 /**
2132 * Enclose a string with the "no conversion" tag. This is used by
2133 * various functions in the Parser
2134 *
2135 * @param $text String: text to be tagged for no conversion
2136 * @param $noParse
2137 * @return string the tagged text
2138 */
2139 function markNoConversion( $text, $noParse=false ) {
2140 return $this->mConverter->markNoConversion( $text, $noParse );
2141 }
2142
2143 /**
2144 * A regular expression to match legal word-trailing characters
2145 * which should be merged onto a link of the form [[foo]]bar.
2146 *
2147 * @return string
2148 */
2149 function linkTrail() {
2150 $this->load();
2151 return $this->linkTrail;
2152 }
2153
2154 function getLangObj() {
2155 return $this;
2156 }
2157
2158 /**
2159 * Get the RFC 3066 code for this language object
2160 */
2161 function getCode() {
2162 return $this->mCode;
2163 }
2164
2165 function setCode( $code ) {
2166 $this->mCode = $code;
2167 }
2168
2169 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2170 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2171 }
2172
2173 static function getMessagesFileName( $code ) {
2174 global $IP;
2175 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2176 }
2177
2178 static function getClassFileName( $code ) {
2179 global $IP;
2180 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2181 }
2182
2183 static function getLocalisationArray( $code, $disableCache = false ) {
2184 self::loadLocalisation( $code, $disableCache );
2185 return self::$mLocalisationCache[$code];
2186 }
2187
2188 /**
2189 * Load localisation data for a given code into the static cache
2190 *
2191 * @return array Dependencies, map of filenames to mtimes
2192 */
2193 static function loadLocalisation( $code, $disableCache = false ) {
2194 static $recursionGuard = array();
2195 global $wgMemc, $wgCheckSerialized;
2196
2197 if ( !$code ) {
2198 throw new MWException( "Invalid language code requested" );
2199 }
2200
2201 if ( !$disableCache ) {
2202 # Try the per-process cache
2203 if ( isset( self::$mLocalisationCache[$code] ) ) {
2204 return self::$mLocalisationCache[$code]['deps'];
2205 }
2206
2207 wfProfileIn( __METHOD__ );
2208
2209 # Try the serialized directory
2210 $cache = wfGetPrecompiledData( self::getFileName( "Messages", $code, '.ser' ) );
2211 if ( $cache ) {
2212 if ( $wgCheckSerialized && self::isLocalisationOutOfDate( $cache ) ) {
2213 $cache = false;
2214 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2215 } else {
2216 self::$mLocalisationCache[$code] = $cache;
2217 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2218 wfProfileOut( __METHOD__ );
2219 return self::$mLocalisationCache[$code]['deps'];
2220 }
2221 }
2222
2223 # Try the global cache
2224 $memcKey = wfMemcKey('localisation', $code );
2225 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2226 $cache = $wgMemc->get( $memcKey );
2227 if ( $cache ) {
2228 if ( self::isLocalisationOutOfDate( $cache ) ) {
2229 $wgMemc->delete( $memcKey );
2230 $wgMemc->delete( $fbMemcKey );
2231 $cache = false;
2232 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2233 } else {
2234 self::$mLocalisationCache[$code] = $cache;
2235 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2236 wfProfileOut( __METHOD__ );
2237 return $cache['deps'];
2238 }
2239 }
2240 } else {
2241 wfProfileIn( __METHOD__ );
2242 }
2243
2244 # Default fallback, may be overridden when the messages file is included
2245 if ( $code != 'en' ) {
2246 $fallback = 'en';
2247 } else {
2248 $fallback = false;
2249 }
2250
2251 # Load the primary localisation from the source file
2252 $filename = self::getMessagesFileName( $code );
2253 if ( !file_exists( $filename ) ) {
2254 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2255 $cache = compact( self::$mLocalisationKeys ); // Set correct fallback
2256 $deps = array();
2257 } else {
2258 $deps = array( $filename => filemtime( $filename ) );
2259 require( $filename );
2260 $cache = compact( self::$mLocalisationKeys );
2261 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2262 }
2263
2264 if ( !empty( $fallback ) ) {
2265 # Load the fallback localisation, with a circular reference guard
2266 if ( isset( $recursionGuard[$code] ) ) {
2267 throw new MWException( "Error: Circular fallback reference in language code $code" );
2268 }
2269 $recursionGuard[$code] = true;
2270 $newDeps = self::loadLocalisation( $fallback, $disableCache );
2271 unset( $recursionGuard[$code] );
2272
2273 $secondary = self::$mLocalisationCache[$fallback];
2274 $deps = array_merge( $deps, $newDeps );
2275
2276 # Merge the fallback localisation with the current localisation
2277 foreach ( self::$mLocalisationKeys as $key ) {
2278 if ( isset( $cache[$key] ) ) {
2279 if ( isset( $secondary[$key] ) ) {
2280 if ( in_array( $key, self::$mMergeableMapKeys ) ) {
2281 $cache[$key] = $cache[$key] + $secondary[$key];
2282 } elseif ( in_array( $key, self::$mMergeableListKeys ) ) {
2283 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2284 } elseif ( in_array( $key, self::$mMergeableAliasListKeys ) ) {
2285 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2286 }
2287 }
2288 } else {
2289 $cache[$key] = $secondary[$key];
2290 }
2291 }
2292
2293 # Merge bookstore lists if requested
2294 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2295 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2296 }
2297 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2298 unset( $cache['bookstoreList']['inherit'] );
2299 }
2300 }
2301
2302 # Add dependencies to the cache entry
2303 $cache['deps'] = $deps;
2304
2305 # Replace spaces with underscores in namespace names
2306 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2307
2308 # And do the same for specialpage aliases.
2309 $cache['specialPageAliases'] =
2310 $this->fixSpecialPageAliases( $cache['specialPageAliases'] );
2311
2312 # Save to both caches
2313 self::$mLocalisationCache[$code] = $cache;
2314 if ( !$disableCache ) {
2315 $wgMemc->set( $memcKey, $cache );
2316 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2317 }
2318
2319 wfProfileOut( __METHOD__ );
2320 return $deps;
2321 }
2322
2323 /**
2324 * Test if a given localisation cache is out of date with respect to the
2325 * source Messages files. This is done automatically for the global cache
2326 * in $wgMemc, but is only done on certain occasions for the serialized
2327 * data file.
2328 *
2329 * @param $cache mixed Either a language code or a cache array
2330 */
2331 static function isLocalisationOutOfDate( $cache ) {
2332 if ( !is_array( $cache ) ) {
2333 self::loadLocalisation( $cache );
2334 $cache = self::$mLocalisationCache[$cache];
2335 }
2336 $expired = false;
2337 foreach ( $cache['deps'] as $file => $mtime ) {
2338 if ( !file_exists( $file ) || filemtime( $file ) > $mtime ) {
2339 $expired = true;
2340 break;
2341 }
2342 }
2343 return $expired;
2344 }
2345
2346 /**
2347 * Get the fallback for a given language
2348 */
2349 static function getFallbackFor( $code ) {
2350 // Shortcut
2351 if ( $code === 'en' ) return false;
2352
2353 // Local cache
2354 static $cache = array();
2355 // Quick return
2356 if ( isset($cache[$code]) ) return $cache[$code];
2357
2358 // Try memcache
2359 global $wgMemc;
2360 $memcKey = wfMemcKey( 'fallback', $code );
2361 $fbcode = $wgMemc->get( $memcKey );
2362
2363 if ( is_string($fbcode) ) {
2364 // False is stored as a string to detect failures in memcache properly
2365 if ( $fbcode === '' ) $fbcode = false;
2366
2367 // Update local cache and return
2368 $cache[$code] = $fbcode;
2369 return $fbcode;
2370 }
2371
2372 // Nothing in caches, load and and update both caches
2373 self::loadLocalisation( $code );
2374 $fbcode = self::$mLocalisationCache[$code]['fallback'];
2375
2376 $cache[$code] = $fbcode;
2377 $wgMemc->set( $memcKey, (string) $fbcode );
2378
2379 return $fbcode;
2380 }
2381
2382 /**
2383 * Get all messages for a given language
2384 */
2385 static function getMessagesFor( $code ) {
2386 self::loadLocalisation( $code );
2387 return self::$mLocalisationCache[$code]['messages'];
2388 }
2389
2390 /**
2391 * Get a message for a given language
2392 */
2393 static function getMessageFor( $key, $code ) {
2394 self::loadLocalisation( $code );
2395 return isset( self::$mLocalisationCache[$code]['messages'][$key] ) ? self::$mLocalisationCache[$code]['messages'][$key] : null;
2396 }
2397
2398 /**
2399 * Load localisation data for this object
2400 */
2401 function load() {
2402 if ( !$this->mLoaded ) {
2403 self::loadLocalisation( $this->getCode() );
2404 $cache =& self::$mLocalisationCache[$this->getCode()];
2405 foreach ( self::$mLocalisationKeys as $key ) {
2406 $this->$key = $cache[$key];
2407 }
2408 $this->mLoaded = true;
2409
2410 $this->fixUpSettings();
2411 }
2412 }
2413
2414 /**
2415 * Do any necessary post-cache-load settings adjustment
2416 */
2417 function fixUpSettings() {
2418 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2419 $wgNamespaceAliases, $wgAmericanDates;
2420 wfProfileIn( __METHOD__ );
2421 if ( $wgExtraNamespaces ) {
2422 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
2423 }
2424
2425 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
2426 if ( $wgMetaNamespaceTalk ) {
2427 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
2428 } else {
2429 $talk = $this->namespaceNames[NS_PROJECT_TALK];
2430 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2431
2432 # Allow grammar transformations
2433 # Allowing full message-style parsing would make simple requests
2434 # such as action=raw much more expensive than they need to be.
2435 # This will hopefully cover most cases.
2436 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2437 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2438 $talk = str_replace( ' ', '_', $talk );
2439 $this->namespaceNames[NS_PROJECT_TALK] = $talk;
2440 }
2441
2442 # The above mixing may leave namespaces out of canonical order.
2443 # Re-order by namespace ID number...
2444 ksort( $this->namespaceNames );
2445
2446 # Put namespace names and aliases into a hashtable.
2447 # If this is too slow, then we should arrange it so that it is done
2448 # before caching. The catch is that at pre-cache time, the above
2449 # class-specific fixup hasn't been done.
2450 $this->mNamespaceIds = array();
2451 foreach ( $this->namespaceNames as $index => $name ) {
2452 $this->mNamespaceIds[$this->lc($name)] = $index;
2453 }
2454 if ( $this->namespaceAliases ) {
2455 foreach ( $this->namespaceAliases as $name => $index ) {
2456 $this->mNamespaceIds[$this->lc($name)] = $index;
2457 }
2458 }
2459 if ( $wgNamespaceAliases ) {
2460 foreach ( $wgNamespaceAliases as $name => $index ) {
2461 $this->mNamespaceIds[$this->lc($name)] = $index;
2462 }
2463 }
2464
2465 if ( $this->defaultDateFormat == 'dmy or mdy' ) {
2466 $this->defaultDateFormat = $wgAmericanDates ? 'mdy' : 'dmy';
2467 }
2468 wfProfileOut( __METHOD__ );
2469 }
2470
2471 function replaceGrammarInNamespace( $m ) {
2472 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2473 }
2474
2475 static function getCaseMaps() {
2476 static $wikiUpperChars, $wikiLowerChars;
2477 if ( isset( $wikiUpperChars ) ) {
2478 return array( $wikiUpperChars, $wikiLowerChars );
2479 }
2480
2481 wfProfileIn( __METHOD__ );
2482 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2483 if ( $arr === false ) {
2484 throw new MWException(
2485 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2486 }
2487 extract( $arr );
2488 wfProfileOut( __METHOD__ );
2489 return array( $wikiUpperChars, $wikiLowerChars );
2490 }
2491
2492 function formatTimePeriod( $seconds ) {
2493 if ( $seconds < 10 ) {
2494 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2495 } elseif ( $seconds < 60 ) {
2496 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2497 } elseif ( $seconds < 3600 ) {
2498 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2499 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2500 } else {
2501 $hours = floor( $seconds / 3600 );
2502 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2503 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2504 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2505 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2506 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2507 }
2508 }
2509
2510 function formatBitrate( $bps ) {
2511 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2512 if ( $bps <= 0 ) {
2513 return $this->formatNum( $bps ) . $units[0];
2514 }
2515 $unitIndex = floor( log10( $bps ) / 3 );
2516 $mantissa = $bps / pow( 1000, $unitIndex );
2517 if ( $mantissa < 10 ) {
2518 $mantissa = round( $mantissa, 1 );
2519 } else {
2520 $mantissa = round( $mantissa );
2521 }
2522 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2523 }
2524
2525 /**
2526 * Format a size in bytes for output, using an appropriate
2527 * unit (B, KB, MB or GB) according to the magnitude in question
2528 *
2529 * @param $size Size to format
2530 * @return string Plain text (not HTML)
2531 */
2532 function formatSize( $size ) {
2533 // For small sizes no decimal places necessary
2534 $round = 0;
2535 if( $size > 1024 ) {
2536 $size = $size / 1024;
2537 if( $size > 1024 ) {
2538 $size = $size / 1024;
2539 // For MB and bigger two decimal places are smarter
2540 $round = 2;
2541 if( $size > 1024 ) {
2542 $size = $size / 1024;
2543 $msg = 'size-gigabytes';
2544 } else {
2545 $msg = 'size-megabytes';
2546 }
2547 } else {
2548 $msg = 'size-kilobytes';
2549 }
2550 } else {
2551 $msg = 'size-bytes';
2552 }
2553 $size = round( $size, $round );
2554 $text = $this->getMessageFromDB( $msg );
2555 return str_replace( '$1', $this->formatNum( $size ), $text );
2556 }
2557 }