Language object cache, for faster wfMsg() performance with unusual languages, and...
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * @defgroup Language Language
4 *
5 * @file
6 * @ingroup Language
7 */
8
9 if( !defined( 'MEDIAWIKI' ) ) {
10 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
11 exit( 1 );
12 }
13
14 # Read language names
15 global $wgLanguageNames;
16 require_once( dirname(__FILE__) . '/Names.php' ) ;
17
18 global $wgInputEncoding, $wgOutputEncoding;
19
20 /**
21 * These are always UTF-8, they exist only for backwards compatibility
22 */
23 $wgInputEncoding = "UTF-8";
24 $wgOutputEncoding = "UTF-8";
25
26 if( function_exists( 'mb_strtoupper' ) ) {
27 mb_internal_encoding('UTF-8');
28 }
29
30 /**
31 * a fake language converter
32 *
33 * @ingroup Language
34 */
35 class FakeConverter {
36 var $mLang;
37 function FakeConverter($langobj) {$this->mLang = $langobj;}
38 function convert($t, $i) {return $t;}
39 function parserConvert($t, $p) {return $t;}
40 function getVariants() { return array( $this->mLang->getCode() ); }
41 function getPreferredVariant() {return $this->mLang->getCode(); }
42 function findVariantLink(&$l, &$n, $forTemplate = false) {}
43 function getExtraHashOptions() {return '';}
44 function getParsedTitle() {return '';}
45 function markNoConversion($text, $noParse=false) {return $text;}
46 function convertCategoryKey( $key ) {return $key; }
47 function convertLinkToAllVariants($text){ return array( $this->mLang->getCode() => $text); }
48 function armourMath($text){ return $text; }
49 }
50
51 /**
52 * Internationalisation code
53 * @ingroup Language
54 */
55 class Language {
56 var $mConverter, $mVariants, $mCode, $mLoaded = false;
57 var $mMagicExtensions = array(), $mMagicHookDone = false;
58
59 static public $mLocalisationKeys = array( 'fallback', 'namespaceNames',
60 'skinNames', 'mathNames',
61 'bookstoreList', 'magicWords', 'messages', 'rtl', 'digitTransformTable',
62 'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
63 'defaultUserOptionOverrides', 'linkTrail', 'namespaceAliases',
64 'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
65 'defaultDateFormat', 'extraUserToggles', 'specialPageAliases',
66 'imageFiles'
67 );
68
69 static public $mMergeableMapKeys = array( 'messages', 'namespaceNames', 'mathNames',
70 'dateFormats', 'defaultUserOptionOverrides', 'magicWords', 'imageFiles' );
71
72 static public $mMergeableListKeys = array( 'extraUserToggles' );
73
74 static public $mMergeableAliasListKeys = array( 'specialPageAliases' );
75
76 static public $mLocalisationCache = array();
77 static public $mLangObjCache = array();
78
79 static public $mWeekdayMsgs = array(
80 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
81 'friday', 'saturday'
82 );
83
84 static public $mWeekdayAbbrevMsgs = array(
85 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
86 );
87
88 static public $mMonthMsgs = array(
89 'january', 'february', 'march', 'april', 'may_long', 'june',
90 'july', 'august', 'september', 'october', 'november',
91 'december'
92 );
93 static public $mMonthGenMsgs = array(
94 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
95 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
96 'december-gen'
97 );
98 static public $mMonthAbbrevMsgs = array(
99 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
100 'sep', 'oct', 'nov', 'dec'
101 );
102
103 static public $mIranianCalendarMonthMsgs = array(
104 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
105 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
106 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
107 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
108 );
109
110 static public $mHebrewCalendarMonthMsgs = array(
111 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
112 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
113 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
114 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
115 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
116 );
117
118 static public $mHebrewCalendarMonthGenMsgs = array(
119 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
120 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
121 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
122 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
123 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
124 );
125
126 static public $mHijriCalendarMonthMsgs = array(
127 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
128 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
129 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
130 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
131 );
132
133 /**
134 * Get a language object for a given language code
135 */
136 static function factory( $code ) {
137 global $IP;
138 static $recursionLevel = 0;
139
140 if ( isset( self::$mLangObjCache[$code] ) ) {
141 return self::$mLangObjCache[$code];
142 }
143
144 if ( $code == 'en' ) {
145 $class = 'Language';
146 } else {
147 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
148 // Preload base classes to work around APC/PHP5 bug
149 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
150 include_once("$IP/languages/classes/$class.deps.php");
151 }
152 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
153 include_once("$IP/languages/classes/$class.php");
154 }
155 }
156
157 if ( $recursionLevel > 5 ) {
158 throw new MWException( "Language fallback loop detected when creating class $class\n" );
159 }
160
161 if( ! class_exists( $class ) ) {
162 $fallback = Language::getFallbackFor( $code );
163 ++$recursionLevel;
164 $lang = Language::factory( $fallback );
165 --$recursionLevel;
166 $lang->setCode( $code );
167 } else {
168 $lang = new $class;
169 }
170
171 self::$mLangObjCache[$code] = $lang;
172 return $lang;
173 }
174
175 function __construct() {
176 $this->mConverter = new FakeConverter($this);
177 // Set the code to the name of the descendant
178 if ( get_class( $this ) == 'Language' ) {
179 $this->mCode = 'en';
180 } else {
181 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
182 }
183 }
184
185 /**
186 * Reduce memory usage
187 */
188 function __destruct() {
189 foreach ( $this as $name => $value ) {
190 unset( $this->$name );
191 }
192 }
193
194 /**
195 * Hook which will be called if this is the content language.
196 * Descendants can use this to register hook functions or modify globals
197 */
198 function initContLang() {}
199
200 /**
201 * @deprecated Use User::getDefaultOptions()
202 * @return array
203 */
204 function getDefaultUserOptions() {
205 wfDeprecated( __METHOD__ );
206 return User::getDefaultOptions();
207 }
208
209 function getFallbackLanguageCode() {
210 return self::getFallbackFor( $this->mCode );
211 }
212
213 /**
214 * Exports $wgBookstoreListEn
215 * @return array
216 */
217 function getBookstoreList() {
218 $this->load();
219 return $this->bookstoreList;
220 }
221
222 /**
223 * @return array
224 */
225 function getNamespaces() {
226 $this->load();
227 return $this->namespaceNames;
228 }
229
230 /**
231 * A convenience function that returns the same thing as
232 * getNamespaces() except with the array values changed to ' '
233 * where it found '_', useful for producing output to be displayed
234 * e.g. in <select> forms.
235 *
236 * @return array
237 */
238 function getFormattedNamespaces() {
239 $ns = $this->getNamespaces();
240 foreach($ns as $k => $v) {
241 $ns[$k] = strtr($v, '_', ' ');
242 }
243 return $ns;
244 }
245
246 /**
247 * Get a namespace value by key
248 * <code>
249 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
250 * echo $mw_ns; // prints 'MediaWiki'
251 * </code>
252 *
253 * @param $index Int: the array key of the namespace to return
254 * @return mixed, string if the namespace value exists, otherwise false
255 */
256 function getNsText( $index ) {
257 $ns = $this->getNamespaces();
258 return isset( $ns[$index] ) ? $ns[$index] : false;
259 }
260
261 /**
262 * A convenience function that returns the same thing as
263 * getNsText() except with '_' changed to ' ', useful for
264 * producing output.
265 *
266 * @return array
267 */
268 function getFormattedNsText( $index ) {
269 $ns = $this->getNsText( $index );
270 return strtr($ns, '_', ' ');
271 }
272
273 /**
274 * Get a namespace key by value, case insensitive.
275 * Only matches namespace names for the current language, not the
276 * canonical ones defined in Namespace.php.
277 *
278 * @param $text String
279 * @return mixed An integer if $text is a valid value otherwise false
280 */
281 function getLocalNsIndex( $text ) {
282 $this->load();
283 $lctext = $this->lc($text);
284 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
285 }
286
287 /**
288 * Get a namespace key by value, case insensitive. Canonical namespace
289 * names override custom ones defined for the current language.
290 *
291 * @param $text String
292 * @return mixed An integer if $text is a valid value otherwise false
293 */
294 function getNsIndex( $text ) {
295 $this->load();
296 $lctext = $this->lc($text);
297 if( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
298 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
299 }
300
301 /**
302 * short names for language variants used for language conversion links.
303 *
304 * @param $code String
305 * @return string
306 */
307 function getVariantname( $code ) {
308 return $this->getMessageFromDB( "variantname-$code" );
309 }
310
311 function specialPage( $name ) {
312 $aliases = $this->getSpecialPageAliases();
313 if ( isset( $aliases[$name][0] ) ) {
314 $name = $aliases[$name][0];
315 }
316 return $this->getNsText(NS_SPECIAL) . ':' . $name;
317 }
318
319 function getQuickbarSettings() {
320 return array(
321 $this->getMessage( 'qbsettings-none' ),
322 $this->getMessage( 'qbsettings-fixedleft' ),
323 $this->getMessage( 'qbsettings-fixedright' ),
324 $this->getMessage( 'qbsettings-floatingleft' ),
325 $this->getMessage( 'qbsettings-floatingright' )
326 );
327 }
328
329 function getSkinNames() {
330 $this->load();
331 return $this->skinNames;
332 }
333
334 function getMathNames() {
335 $this->load();
336 return $this->mathNames;
337 }
338
339 function getDatePreferences() {
340 $this->load();
341 return $this->datePreferences;
342 }
343
344 function getDateFormats() {
345 $this->load();
346 return $this->dateFormats;
347 }
348
349 function getDefaultDateFormat() {
350 $this->load();
351 return $this->defaultDateFormat;
352 }
353
354 function getDatePreferenceMigrationMap() {
355 $this->load();
356 return $this->datePreferenceMigrationMap;
357 }
358
359 function getImageFile( $image ) {
360 $this->load();
361 return $this->imageFiles[$image];
362 }
363
364 function getDefaultUserOptionOverrides() {
365 $this->load();
366 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
367 if (is_array($this->defaultUserOptionOverrides)) {
368 return $this->defaultUserOptionOverrides;
369 } else {
370 return array();
371 }
372 }
373
374 function getExtraUserToggles() {
375 $this->load();
376 return $this->extraUserToggles;
377 }
378
379 function getUserToggle( $tog ) {
380 return $this->getMessageFromDB( "tog-$tog" );
381 }
382
383 /**
384 * Get language names, indexed by code.
385 * If $customisedOnly is true, only returns codes with a messages file
386 */
387 public static function getLanguageNames( $customisedOnly = false ) {
388 global $wgLanguageNames, $wgExtraLanguageNames;
389 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
390 if ( !$customisedOnly ) {
391 return $allNames;
392 }
393
394 global $IP;
395 $names = array();
396 $dir = opendir( "$IP/languages/messages" );
397 while( false !== ( $file = readdir( $dir ) ) ) {
398 $m = array();
399 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
400 $code = str_replace( '_', '-', strtolower( $m[1] ) );
401 if ( isset( $allNames[$code] ) ) {
402 $names[$code] = $allNames[$code];
403 }
404 }
405 }
406 closedir( $dir );
407 return $names;
408 }
409
410 /**
411 * Get a message from the MediaWiki namespace.
412 *
413 * @param $msg String: message name
414 * @return string
415 */
416 function getMessageFromDB( $msg ) {
417 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
418 }
419
420 function getLanguageName( $code ) {
421 $names = self::getLanguageNames();
422 if ( !array_key_exists( $code, $names ) ) {
423 return '';
424 }
425 return $names[$code];
426 }
427
428 function getMonthName( $key ) {
429 return $this->getMessageFromDB( self::$mMonthMsgs[$key-1] );
430 }
431
432 function getMonthNameGen( $key ) {
433 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key-1] );
434 }
435
436 function getMonthAbbreviation( $key ) {
437 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key-1] );
438 }
439
440 function getWeekdayName( $key ) {
441 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key-1] );
442 }
443
444 function getWeekdayAbbreviation( $key ) {
445 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key-1] );
446 }
447
448 function getIranianCalendarMonthName( $key ) {
449 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key-1] );
450 }
451
452 function getHebrewCalendarMonthName( $key ) {
453 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key-1] );
454 }
455
456 function getHebrewCalendarMonthNameGen( $key ) {
457 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key-1] );
458 }
459
460 function getHijriCalendarMonthName( $key ) {
461 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key-1] );
462 }
463
464 /**
465 * Used by date() and time() to adjust the time output.
466 *
467 * @param $ts Int the time in date('YmdHis') format
468 * @param $tz Mixed: adjust the time by this amount (default false, mean we
469 * get user timecorrection setting)
470 * @return int
471 */
472 function userAdjust( $ts, $tz = false ) {
473 global $wgUser, $wgLocalTZoffset;
474
475 if (!$tz) {
476 $tz = $wgUser->getOption( 'timecorrection' );
477 }
478
479 # minutes and hours differences:
480 $minDiff = 0;
481 $hrDiff = 0;
482
483 if ( $tz === '' ) {
484 # Global offset in minutes.
485 if( isset($wgLocalTZoffset) ) {
486 if( $wgLocalTZoffset >= 0 ) {
487 $hrDiff = floor($wgLocalTZoffset / 60);
488 } else {
489 $hrDiff = ceil($wgLocalTZoffset / 60);
490 }
491 $minDiff = $wgLocalTZoffset % 60;
492 }
493 } elseif ( strpos( $tz, ':' ) !== false ) {
494 $tzArray = explode( ':', $tz );
495 $hrDiff = intval($tzArray[0]);
496 $minDiff = intval($hrDiff < 0 ? -$tzArray[1] : $tzArray[1]);
497 } else {
498 $hrDiff = intval( $tz );
499 }
500
501 # No difference ? Return time unchanged
502 if ( 0 == $hrDiff && 0 == $minDiff ) { return $ts; }
503
504 wfSuppressWarnings(); // E_STRICT system time bitching
505 # Generate an adjusted date
506 $t = mktime( (
507 (int)substr( $ts, 8, 2) ) + $hrDiff, # Hours
508 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
509 (int)substr( $ts, 12, 2 ), # Seconds
510 (int)substr( $ts, 4, 2 ), # Month
511 (int)substr( $ts, 6, 2 ), # Day
512 (int)substr( $ts, 0, 4 ) ); #Year
513
514 $date = date( 'YmdHis', $t );
515 wfRestoreWarnings();
516
517 return $date;
518 }
519
520 /**
521 * This is a workalike of PHP's date() function, but with better
522 * internationalisation, a reduced set of format characters, and a better
523 * escaping format.
524 *
525 * Supported format characters are dDjlNwzWFmMntLYyaAgGhHiscrU. See the
526 * PHP manual for definitions. There are a number of extensions, which
527 * start with "x":
528 *
529 * xn Do not translate digits of the next numeric format character
530 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
531 * xr Use roman numerals for the next numeric format character
532 * xh Use hebrew numerals for the next numeric format character
533 * xx Literal x
534 * xg Genitive month name
535 *
536 * xij j (day number) in Iranian calendar
537 * xiF F (month name) in Iranian calendar
538 * xin n (month number) in Iranian calendar
539 * xiY Y (full year) in Iranian calendar
540 *
541 * xjj j (day number) in Hebrew calendar
542 * xjF F (month name) in Hebrew calendar
543 * xjt t (days in month) in Hebrew calendar
544 * xjx xg (genitive month name) in Hebrew calendar
545 * xjn n (month number) in Hebrew calendar
546 * xjY Y (full year) in Hebrew calendar
547 *
548 * xmj j (day number) in Hijri calendar
549 * xmF F (month name) in Hijri calendar
550 * xmn n (month number) in Hijri calendar
551 * xmY Y (full year) in Hijri calendar
552 *
553 * xkY Y (full year) in Thai solar calendar. Months and days are
554 * identical to the Gregorian calendar
555 *
556 * Characters enclosed in double quotes will be considered literal (with
557 * the quotes themselves removed). Unmatched quotes will be considered
558 * literal quotes. Example:
559 *
560 * "The month is" F => The month is January
561 * i's" => 20'11"
562 *
563 * Backslash escaping is also supported.
564 *
565 * Input timestamp is assumed to be pre-normalized to the desired local
566 * time zone, if any.
567 *
568 * @param $format String
569 * @param $ts String: 14-character timestamp
570 * YYYYMMDDHHMMSS
571 * 01234567890123
572 */
573 function sprintfDate( $format, $ts ) {
574 $s = '';
575 $raw = false;
576 $roman = false;
577 $hebrewNum = false;
578 $unix = false;
579 $rawToggle = false;
580 $iranian = false;
581 $hebrew = false;
582 $hijri = false;
583 $thai = false;
584 for ( $p = 0; $p < strlen( $format ); $p++ ) {
585 $num = false;
586 $code = $format[$p];
587 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
588 $code .= $format[++$p];
589 }
590
591 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' ) && $p < strlen( $format ) - 1 ) {
592 $code .= $format[++$p];
593 }
594
595 switch ( $code ) {
596 case 'xx':
597 $s .= 'x';
598 break;
599 case 'xn':
600 $raw = true;
601 break;
602 case 'xN':
603 $rawToggle = !$rawToggle;
604 break;
605 case 'xr':
606 $roman = true;
607 break;
608 case 'xh':
609 $hebrewNum = true;
610 break;
611 case 'xg':
612 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
613 break;
614 case 'xjx':
615 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
616 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
617 break;
618 case 'd':
619 $num = substr( $ts, 6, 2 );
620 break;
621 case 'D':
622 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
623 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
624 break;
625 case 'j':
626 $num = intval( substr( $ts, 6, 2 ) );
627 break;
628 case 'xij':
629 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
630 $num = $iranian[2];
631 break;
632 case 'xmj':
633 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
634 $num = $hijri[2];
635 break;
636 case 'xjj':
637 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
638 $num = $hebrew[2];
639 break;
640 case 'l':
641 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
642 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
643 break;
644 case 'N':
645 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
646 $w = gmdate( 'w', $unix );
647 $num = $w ? $w : 7;
648 break;
649 case 'w':
650 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
651 $num = gmdate( 'w', $unix );
652 break;
653 case 'z':
654 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
655 $num = gmdate( 'z', $unix );
656 break;
657 case 'W':
658 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
659 $num = gmdate( 'W', $unix );
660 break;
661 case 'F':
662 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
663 break;
664 case 'xiF':
665 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
666 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
667 break;
668 case 'xmF':
669 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
670 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
671 break;
672 case 'xjF':
673 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
674 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
675 break;
676 case 'm':
677 $num = substr( $ts, 4, 2 );
678 break;
679 case 'M':
680 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
681 break;
682 case 'n':
683 $num = intval( substr( $ts, 4, 2 ) );
684 break;
685 case 'xin':
686 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
687 $num = $iranian[1];
688 break;
689 case 'xmn':
690 if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
691 $num = $hijri[1];
692 break;
693 case 'xjn':
694 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
695 $num = $hebrew[1];
696 break;
697 case 't':
698 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
699 $num = gmdate( 't', $unix );
700 break;
701 case 'xjt':
702 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
703 $num = $hebrew[3];
704 break;
705 case 'L':
706 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
707 $num = gmdate( 'L', $unix );
708 break;
709 case 'Y':
710 $num = substr( $ts, 0, 4 );
711 break;
712 case 'xiY':
713 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
714 $num = $iranian[0];
715 break;
716 case 'xmY':
717 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
718 $num = $hijri[0];
719 break;
720 case 'xjY':
721 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
722 $num = $hebrew[0];
723 break;
724 case 'xkY':
725 if ( !$thai ) $thai = self::tsToThai( $ts );
726 $num = $thai[0];
727 break;
728 case 'y':
729 $num = substr( $ts, 2, 2 );
730 break;
731 case 'a':
732 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
733 break;
734 case 'A':
735 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
736 break;
737 case 'g':
738 $h = substr( $ts, 8, 2 );
739 $num = $h % 12 ? $h % 12 : 12;
740 break;
741 case 'G':
742 $num = intval( substr( $ts, 8, 2 ) );
743 break;
744 case 'h':
745 $h = substr( $ts, 8, 2 );
746 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
747 break;
748 case 'H':
749 $num = substr( $ts, 8, 2 );
750 break;
751 case 'i':
752 $num = substr( $ts, 10, 2 );
753 break;
754 case 's':
755 $num = substr( $ts, 12, 2 );
756 break;
757 case 'c':
758 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
759 $s .= gmdate( 'c', $unix );
760 break;
761 case 'r':
762 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
763 $s .= gmdate( 'r', $unix );
764 break;
765 case 'U':
766 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
767 $num = $unix;
768 break;
769 case '\\':
770 # Backslash escaping
771 if ( $p < strlen( $format ) - 1 ) {
772 $s .= $format[++$p];
773 } else {
774 $s .= '\\';
775 }
776 break;
777 case '"':
778 # Quoted literal
779 if ( $p < strlen( $format ) - 1 ) {
780 $endQuote = strpos( $format, '"', $p + 1 );
781 if ( $endQuote === false ) {
782 # No terminating quote, assume literal "
783 $s .= '"';
784 } else {
785 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
786 $p = $endQuote;
787 }
788 } else {
789 # Quote at end of string, assume literal "
790 $s .= '"';
791 }
792 break;
793 default:
794 $s .= $format[$p];
795 }
796 if ( $num !== false ) {
797 if ( $rawToggle || $raw ) {
798 $s .= $num;
799 $raw = false;
800 } elseif ( $roman ) {
801 $s .= self::romanNumeral( $num );
802 $roman = false;
803 } elseif( $hebrewNum ) {
804 $s .= self::hebrewNumeral( $num );
805 $hebrewNum = false;
806 } else {
807 $s .= $this->formatNum( $num, true );
808 }
809 $num = false;
810 }
811 }
812 return $s;
813 }
814
815 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
816 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
817 /**
818 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
819 * Gregorian dates to Iranian dates. Originally written in C, it
820 * is released under the terms of GNU Lesser General Public
821 * License. Conversion to PHP was performed by Niklas Laxström.
822 *
823 * Link: http://www.farsiweb.info/jalali/jalali.c
824 */
825 private static function tsToIranian( $ts ) {
826 $gy = substr( $ts, 0, 4 ) -1600;
827 $gm = substr( $ts, 4, 2 ) -1;
828 $gd = substr( $ts, 6, 2 ) -1;
829
830 # Days passed from the beginning (including leap years)
831 $gDayNo = 365*$gy
832 + floor(($gy+3) / 4)
833 - floor(($gy+99) / 100)
834 + floor(($gy+399) / 400);
835
836
837 // Add days of the past months of this year
838 for( $i = 0; $i < $gm; $i++ ) {
839 $gDayNo += self::$GREG_DAYS[$i];
840 }
841
842 // Leap years
843 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
844 $gDayNo++;
845 }
846
847 // Days passed in current month
848 $gDayNo += $gd;
849
850 $jDayNo = $gDayNo - 79;
851
852 $jNp = floor($jDayNo / 12053);
853 $jDayNo %= 12053;
854
855 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
856 $jDayNo %= 1461;
857
858 if ( $jDayNo >= 366 ) {
859 $jy += floor(($jDayNo-1)/365);
860 $jDayNo = floor(($jDayNo-1)%365);
861 }
862
863 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
864 $jDayNo -= self::$IRANIAN_DAYS[$i];
865 }
866
867 $jm= $i+1;
868 $jd= $jDayNo+1;
869
870 return array($jy, $jm, $jd);
871 }
872 /**
873 * Converting Gregorian dates to Hijri dates.
874 *
875 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
876 *
877 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
878 */
879 private static function tsToHijri ( $ts ) {
880 $year = substr( $ts, 0, 4 );
881 $month = substr( $ts, 4, 2 );
882 $day = substr( $ts, 6, 2 );
883
884 $zyr = $year;
885 $zd=$day;
886 $zm=$month;
887 $zy=$zyr;
888
889
890
891 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
892 {
893
894
895 $zjd=(int)((1461*($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;
896 }
897 else
898 {
899 $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
900 }
901
902 $zl=$zjd-1948440+10632;
903 $zn=(int)(($zl-1)/10631);
904 $zl=$zl-10631*$zn+354;
905 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
906 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
907 $zm=(int)((24*$zl)/709);
908 $zd=$zl-(int)((709*$zm)/24);
909 $zy=30*$zn+$zj-30;
910
911 return array ($zy, $zm, $zd);
912 }
913
914 /**
915 * Converting Gregorian dates to Hebrew dates.
916 *
917 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
918 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
919 * to translate the relevant functions into PHP and release them under
920 * GNU GPL.
921 *
922 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
923 * and Adar II is 14. In a non-leap year, Adar is 6.
924 */
925 private static function tsToHebrew( $ts ) {
926 # Parse date
927 $year = substr( $ts, 0, 4 );
928 $month = substr( $ts, 4, 2 );
929 $day = substr( $ts, 6, 2 );
930
931 # Calculate Hebrew year
932 $hebrewYear = $year + 3760;
933
934 # Month number when September = 1, August = 12
935 $month += 4;
936 if( $month > 12 ) {
937 # Next year
938 $month -= 12;
939 $year++;
940 $hebrewYear++;
941 }
942
943 # Calculate day of year from 1 September
944 $dayOfYear = $day;
945 for( $i = 1; $i < $month; $i++ ) {
946 if( $i == 6 ) {
947 # February
948 $dayOfYear += 28;
949 # Check if the year is leap
950 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
951 $dayOfYear++;
952 }
953 } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
954 $dayOfYear += 30;
955 } else {
956 $dayOfYear += 31;
957 }
958 }
959
960 # Calculate the start of the Hebrew year
961 $start = self::hebrewYearStart( $hebrewYear );
962
963 # Calculate next year's start
964 if( $dayOfYear <= $start ) {
965 # Day is before the start of the year - it is the previous year
966 # Next year's start
967 $nextStart = $start;
968 # Previous year
969 $year--;
970 $hebrewYear--;
971 # Add days since previous year's 1 September
972 $dayOfYear += 365;
973 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
974 # Leap year
975 $dayOfYear++;
976 }
977 # Start of the new (previous) year
978 $start = self::hebrewYearStart( $hebrewYear );
979 } else {
980 # Next year's start
981 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
982 }
983
984 # Calculate Hebrew day of year
985 $hebrewDayOfYear = $dayOfYear - $start;
986
987 # Difference between year's days
988 $diff = $nextStart - $start;
989 # Add 12 (or 13 for leap years) days to ignore the difference between
990 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
991 # difference is only about the year type
992 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
993 $diff += 13;
994 } else {
995 $diff += 12;
996 }
997
998 # Check the year pattern, and is leap year
999 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1000 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1001 # and non-leap years
1002 $yearPattern = $diff % 30;
1003 # Check if leap year
1004 $isLeap = $diff >= 30;
1005
1006 # Calculate day in the month from number of day in the Hebrew year
1007 # Don't check Adar - if the day is not in Adar, we will stop before;
1008 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1009 $hebrewDay = $hebrewDayOfYear;
1010 $hebrewMonth = 1;
1011 $days = 0;
1012 while( $hebrewMonth <= 12 ) {
1013 # Calculate days in this month
1014 if( $isLeap && $hebrewMonth == 6 ) {
1015 # Adar in a leap year
1016 if( $isLeap ) {
1017 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1018 $days = 30;
1019 if( $hebrewDay <= $days ) {
1020 # Day in Adar I
1021 $hebrewMonth = 13;
1022 } else {
1023 # Subtract the days of Adar I
1024 $hebrewDay -= $days;
1025 # Try Adar II
1026 $days = 29;
1027 if( $hebrewDay <= $days ) {
1028 # Day in Adar II
1029 $hebrewMonth = 14;
1030 }
1031 }
1032 }
1033 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1034 # Cheshvan in a complete year (otherwise as the rule below)
1035 $days = 30;
1036 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1037 # Kislev in an incomplete year (otherwise as the rule below)
1038 $days = 29;
1039 } else {
1040 # Odd months have 30 days, even have 29
1041 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1042 }
1043 if( $hebrewDay <= $days ) {
1044 # In the current month
1045 break;
1046 } else {
1047 # Subtract the days of the current month
1048 $hebrewDay -= $days;
1049 # Try in the next month
1050 $hebrewMonth++;
1051 }
1052 }
1053
1054 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1055 }
1056
1057 /**
1058 * This calculates the Hebrew year start, as days since 1 September.
1059 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1060 * Used for Hebrew date.
1061 */
1062 private static function hebrewYearStart( $year ) {
1063 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1064 $b = intval( ( $year - 1 ) % 4 );
1065 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1066 if( $m < 0 ) {
1067 $m--;
1068 }
1069 $Mar = intval( $m );
1070 if( $m < 0 ) {
1071 $m++;
1072 }
1073 $m -= $Mar;
1074
1075 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1076 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1077 $Mar++;
1078 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1079 $Mar += 2;
1080 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1081 $Mar++;
1082 }
1083
1084 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1085 return $Mar;
1086 }
1087
1088 /**
1089 * Algorithm to convert Gregorian dates to Thai solar dates.
1090 *
1091 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1092 *
1093 * @param $ts String: 14-character timestamp
1094 * @return array converted year, month, day
1095 */
1096 private static function tsToThai( $ts ) {
1097 $gy = substr( $ts, 0, 4 );
1098 $gm = substr( $ts, 4, 2 );
1099 $gd = substr( $ts, 6, 2 );
1100
1101 # Add 543 years to the Gregorian calendar
1102 # Months and days are identical
1103 $gy_thai = $gy + 543;
1104
1105 return array( $gy_thai, $gm, $gd );
1106 }
1107
1108
1109 /**
1110 * Roman number formatting up to 3000
1111 */
1112 static function romanNumeral( $num ) {
1113 static $table = array(
1114 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1115 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1116 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1117 array( '', 'M', 'MM', 'MMM' )
1118 );
1119
1120 $num = intval( $num );
1121 if ( $num > 3000 || $num <= 0 ) {
1122 return $num;
1123 }
1124
1125 $s = '';
1126 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1127 if ( $num >= $pow10 ) {
1128 $s .= $table[$i][floor($num / $pow10)];
1129 }
1130 $num = $num % $pow10;
1131 }
1132 return $s;
1133 }
1134
1135 /**
1136 * Hebrew Gematria number formatting up to 9999
1137 */
1138 static function hebrewNumeral( $num ) {
1139 static $table = array(
1140 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1141 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1142 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1143 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1144 );
1145
1146 $num = intval( $num );
1147 if ( $num > 9999 || $num <= 0 ) {
1148 return $num;
1149 }
1150
1151 $s = '';
1152 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1153 if ( $num >= $pow10 ) {
1154 if ( $num == 15 || $num == 16 ) {
1155 $s .= $table[0][9] . $table[0][$num - 9];
1156 $num = 0;
1157 } else {
1158 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1159 if( $pow10 == 1000 ) {
1160 $s .= "'";
1161 }
1162 }
1163 }
1164 $num = $num % $pow10;
1165 }
1166 if( strlen( $s ) == 2 ) {
1167 $str = $s . "'";
1168 } else {
1169 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1170 $str .= substr( $s, strlen( $s ) - 2, 2 );
1171 }
1172 $start = substr( $str, 0, strlen( $str ) - 2 );
1173 $end = substr( $str, strlen( $str ) - 2 );
1174 switch( $end ) {
1175 case 'כ':
1176 $str = $start . 'ך';
1177 break;
1178 case 'מ':
1179 $str = $start . 'ם';
1180 break;
1181 case 'נ':
1182 $str = $start . 'ן';
1183 break;
1184 case 'פ':
1185 $str = $start . 'ף';
1186 break;
1187 case 'צ':
1188 $str = $start . 'ץ';
1189 break;
1190 }
1191 return $str;
1192 }
1193
1194 /**
1195 * This is meant to be used by time(), date(), and timeanddate() to get
1196 * the date preference they're supposed to use, it should be used in
1197 * all children.
1198 *
1199 *<code>
1200 * function timeanddate([...], $format = true) {
1201 * $datePreference = $this->dateFormat($format);
1202 * [...]
1203 * }
1204 *</code>
1205 *
1206 * @param $usePrefs Mixed: if true, the user's preference is used
1207 * if false, the site/language default is used
1208 * if int/string, assumed to be a format.
1209 * @return string
1210 */
1211 function dateFormat( $usePrefs = true ) {
1212 global $wgUser;
1213
1214 if( is_bool( $usePrefs ) ) {
1215 if( $usePrefs ) {
1216 $datePreference = $wgUser->getDatePreference();
1217 } else {
1218 $options = User::getDefaultOptions();
1219 $datePreference = (string)$options['date'];
1220 }
1221 } else {
1222 $datePreference = (string)$usePrefs;
1223 }
1224
1225 // return int
1226 if( $datePreference == '' ) {
1227 return 'default';
1228 }
1229
1230 return $datePreference;
1231 }
1232
1233 /**
1234 * @param $ts Mixed: the time format which needs to be turned into a
1235 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1236 * @param $adj Bool: whether to adjust the time output according to the
1237 * user configured offset ($timecorrection)
1238 * @param $format Mixed: true to use user's date format preference
1239 * @param $timecorrection String: the time offset as returned by
1240 * validateTimeZone() in Special:Preferences
1241 * @return string
1242 */
1243 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1244 $this->load();
1245 if ( $adj ) {
1246 $ts = $this->userAdjust( $ts, $timecorrection );
1247 }
1248
1249 $pref = $this->dateFormat( $format );
1250 if( $pref == 'default' || !isset( $this->dateFormats["$pref date"] ) ) {
1251 $pref = $this->defaultDateFormat;
1252 }
1253 return $this->sprintfDate( $this->dateFormats["$pref date"], $ts );
1254 }
1255
1256 /**
1257 * @param $ts Mixed: the time format which needs to be turned into a
1258 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1259 * @param $adj Bool: whether to adjust the time output according to the
1260 * user configured offset ($timecorrection)
1261 * @param $format Mixed: true to use user's date format preference
1262 * @param $timecorrection String: the time offset as returned by
1263 * validateTimeZone() in Special:Preferences
1264 * @return string
1265 */
1266 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1267 $this->load();
1268 if ( $adj ) {
1269 $ts = $this->userAdjust( $ts, $timecorrection );
1270 }
1271
1272 $pref = $this->dateFormat( $format );
1273 if( $pref == 'default' || !isset( $this->dateFormats["$pref time"] ) ) {
1274 $pref = $this->defaultDateFormat;
1275 }
1276 return $this->sprintfDate( $this->dateFormats["$pref time"], $ts );
1277 }
1278
1279 /**
1280 * @param $ts Mixed: the time format which needs to be turned into a
1281 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1282 * @param $adj Bool: whether to adjust the time output according to the
1283 * user configured offset ($timecorrection)
1284 * @param $format Mixed: what format to return, if it's false output the
1285 * default one (default true)
1286 * @param $timecorrection String: the time offset as returned by
1287 * validateTimeZone() in Special:Preferences
1288 * @return string
1289 */
1290 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1291 $this->load();
1292
1293 $ts = wfTimestamp( TS_MW, $ts );
1294
1295 if ( $adj ) {
1296 $ts = $this->userAdjust( $ts, $timecorrection );
1297 }
1298
1299 $pref = $this->dateFormat( $format );
1300 if( $pref == 'default' || !isset( $this->dateFormats["$pref both"] ) ) {
1301 $pref = $this->defaultDateFormat;
1302 }
1303
1304 return $this->sprintfDate( $this->dateFormats["$pref both"], $ts );
1305 }
1306
1307 function getMessage( $key ) {
1308 $this->load();
1309 return isset( $this->messages[$key] ) ? $this->messages[$key] : null;
1310 }
1311
1312 function getAllMessages() {
1313 $this->load();
1314 return $this->messages;
1315 }
1316
1317 function iconv( $in, $out, $string ) {
1318 # For most languages, this is a wrapper for iconv
1319 return iconv( $in, $out . '//IGNORE', $string );
1320 }
1321
1322 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1323 function ucwordbreaksCallbackAscii($matches){
1324 return $this->ucfirst($matches[1]);
1325 }
1326
1327 function ucwordbreaksCallbackMB($matches){
1328 return mb_strtoupper($matches[0]);
1329 }
1330
1331 function ucCallback($matches){
1332 list( $wikiUpperChars ) = self::getCaseMaps();
1333 return strtr( $matches[1], $wikiUpperChars );
1334 }
1335
1336 function lcCallback($matches){
1337 list( , $wikiLowerChars ) = self::getCaseMaps();
1338 return strtr( $matches[1], $wikiLowerChars );
1339 }
1340
1341 function ucwordsCallbackMB($matches){
1342 return mb_strtoupper($matches[0]);
1343 }
1344
1345 function ucwordsCallbackWiki($matches){
1346 list( $wikiUpperChars ) = self::getCaseMaps();
1347 return strtr( $matches[0], $wikiUpperChars );
1348 }
1349
1350 function ucfirst( $str ) {
1351 if ( empty($str) ) return $str;
1352 if ( ord($str[0]) < 128 ) return ucfirst($str);
1353 else return self::uc($str,true); // fall back to more complex logic in case of multibyte strings
1354 }
1355
1356 function uc( $str, $first = false ) {
1357 if ( function_exists( 'mb_strtoupper' ) ) {
1358 if ( $first ) {
1359 if ( self::isMultibyte( $str ) ) {
1360 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1361 } else {
1362 return ucfirst( $str );
1363 }
1364 } else {
1365 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1366 }
1367 } else {
1368 if ( self::isMultibyte( $str ) ) {
1369 list( $wikiUpperChars ) = $this->getCaseMaps();
1370 $x = $first ? '^' : '';
1371 return preg_replace_callback(
1372 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1373 array($this,"ucCallback"),
1374 $str
1375 );
1376 } else {
1377 return $first ? ucfirst( $str ) : strtoupper( $str );
1378 }
1379 }
1380 }
1381
1382 function lcfirst( $str ) {
1383 if ( empty($str) ) return $str;
1384 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1385 // editing string in place = cool
1386 $str[0]=strtolower($str[0]);
1387 return $str;
1388 }
1389 else return self::lc( $str, true );
1390 }
1391
1392 function lc( $str, $first = false ) {
1393 if ( function_exists( 'mb_strtolower' ) )
1394 if ( $first )
1395 if ( self::isMultibyte( $str ) )
1396 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1397 else
1398 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1399 else
1400 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1401 else
1402 if ( self::isMultibyte( $str ) ) {
1403 list( , $wikiLowerChars ) = self::getCaseMaps();
1404 $x = $first ? '^' : '';
1405 return preg_replace_callback(
1406 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1407 array($this,"lcCallback"),
1408 $str
1409 );
1410 } else
1411 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1412 }
1413
1414 function isMultibyte( $str ) {
1415 return (bool)preg_match( '/[\x80-\xff]/', $str );
1416 }
1417
1418 function ucwords($str) {
1419 if ( self::isMultibyte( $str ) ) {
1420 $str = self::lc($str);
1421
1422 // regexp to find first letter in each word (i.e. after each space)
1423 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1424
1425 // function to use to capitalize a single char
1426 if ( function_exists( 'mb_strtoupper' ) )
1427 return preg_replace_callback(
1428 $replaceRegexp,
1429 array($this,"ucwordsCallbackMB"),
1430 $str
1431 );
1432 else
1433 return preg_replace_callback(
1434 $replaceRegexp,
1435 array($this,"ucwordsCallbackWiki"),
1436 $str
1437 );
1438 }
1439 else
1440 return ucwords( strtolower( $str ) );
1441 }
1442
1443 # capitalize words at word breaks
1444 function ucwordbreaks($str){
1445 if (self::isMultibyte( $str ) ) {
1446 $str = self::lc($str);
1447
1448 // since \b doesn't work for UTF-8, we explicitely define word break chars
1449 $breaks= "[ \-\(\)\}\{\.,\?!]";
1450
1451 // find first letter after word break
1452 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1453
1454 if ( function_exists( 'mb_strtoupper' ) )
1455 return preg_replace_callback(
1456 $replaceRegexp,
1457 array($this,"ucwordbreaksCallbackMB"),
1458 $str
1459 );
1460 else
1461 return preg_replace_callback(
1462 $replaceRegexp,
1463 array($this,"ucwordsCallbackWiki"),
1464 $str
1465 );
1466 }
1467 else
1468 return preg_replace_callback(
1469 '/\b([\w\x80-\xff]+)\b/',
1470 array($this,"ucwordbreaksCallbackAscii"),
1471 $str );
1472 }
1473
1474 /**
1475 * Return a case-folded representation of $s
1476 *
1477 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1478 * and $s2 are the same except for the case of their characters. It is not
1479 * necessary for the value returned to make sense when displayed.
1480 *
1481 * Do *not* perform any other normalisation in this function. If a caller
1482 * uses this function when it should be using a more general normalisation
1483 * function, then fix the caller.
1484 */
1485 function caseFold( $s ) {
1486 return $this->uc( $s );
1487 }
1488
1489 function checkTitleEncoding( $s ) {
1490 if( is_array( $s ) ) {
1491 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1492 }
1493 # Check for non-UTF-8 URLs
1494 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1495 if(!$ishigh) return $s;
1496
1497 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1498 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1499 if( $isutf8 ) return $s;
1500
1501 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1502 }
1503
1504 function fallback8bitEncoding() {
1505 $this->load();
1506 return $this->fallback8bitEncoding;
1507 }
1508
1509 /**
1510 * Some languages have special punctuation to strip out
1511 * or characters which need to be converted for MySQL's
1512 * indexing to grok it correctly. Make such changes here.
1513 *
1514 * @param $string String
1515 * @return String
1516 */
1517 function stripForSearch( $string ) {
1518 global $wgDBtype;
1519 if ( $wgDBtype != 'mysql' ) {
1520 return $string;
1521 }
1522
1523 # MySQL fulltext index doesn't grok utf-8, so we
1524 # need to fold cases and convert to hex
1525
1526 wfProfileIn( __METHOD__ );
1527 if( function_exists( 'mb_strtolower' ) ) {
1528 $out = preg_replace(
1529 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1530 "'U8' . bin2hex( \"$1\" )",
1531 mb_strtolower( $string ) );
1532 } else {
1533 list( , $wikiLowerChars ) = self::getCaseMaps();
1534 $out = preg_replace(
1535 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1536 "'U8' . bin2hex( strtr( \"\$1\", \$wikiLowerChars ) )",
1537 $string );
1538 }
1539 wfProfileOut( __METHOD__ );
1540 return $out;
1541 }
1542
1543 function convertForSearchResult( $termsArray ) {
1544 # some languages, e.g. Chinese, need to do a conversion
1545 # in order for search results to be displayed correctly
1546 return $termsArray;
1547 }
1548
1549 /**
1550 * Get the first character of a string.
1551 *
1552 * @param $s string
1553 * @return string
1554 */
1555 function firstChar( $s ) {
1556 $matches = array();
1557 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1558 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1559
1560 if ( isset( $matches[1] ) ) {
1561 if ( strlen( $matches[1] ) != 3 ) {
1562 return $matches[1];
1563 }
1564
1565 // Break down Hangul syllables to grab the first jamo
1566 $code = utf8ToCodepoint( $matches[1] );
1567 if ( $code < 0xac00 || 0xd7a4 <= $code) {
1568 return $matches[1];
1569 } elseif ( $code < 0xb098 ) {
1570 return "\xe3\x84\xb1";
1571 } elseif ( $code < 0xb2e4 ) {
1572 return "\xe3\x84\xb4";
1573 } elseif ( $code < 0xb77c ) {
1574 return "\xe3\x84\xb7";
1575 } elseif ( $code < 0xb9c8 ) {
1576 return "\xe3\x84\xb9";
1577 } elseif ( $code < 0xbc14 ) {
1578 return "\xe3\x85\x81";
1579 } elseif ( $code < 0xc0ac ) {
1580 return "\xe3\x85\x82";
1581 } elseif ( $code < 0xc544 ) {
1582 return "\xe3\x85\x85";
1583 } elseif ( $code < 0xc790 ) {
1584 return "\xe3\x85\x87";
1585 } elseif ( $code < 0xcc28 ) {
1586 return "\xe3\x85\x88";
1587 } elseif ( $code < 0xce74 ) {
1588 return "\xe3\x85\x8a";
1589 } elseif ( $code < 0xd0c0 ) {
1590 return "\xe3\x85\x8b";
1591 } elseif ( $code < 0xd30c ) {
1592 return "\xe3\x85\x8c";
1593 } elseif ( $code < 0xd558 ) {
1594 return "\xe3\x85\x8d";
1595 } else {
1596 return "\xe3\x85\x8e";
1597 }
1598 } else {
1599 return "";
1600 }
1601 }
1602
1603 function initEncoding() {
1604 # Some languages may have an alternate char encoding option
1605 # (Esperanto X-coding, Japanese furigana conversion, etc)
1606 # If this language is used as the primary content language,
1607 # an override to the defaults can be set here on startup.
1608 }
1609
1610 function recodeForEdit( $s ) {
1611 # For some languages we'll want to explicitly specify
1612 # which characters make it into the edit box raw
1613 # or are converted in some way or another.
1614 # Note that if wgOutputEncoding is different from
1615 # wgInputEncoding, this text will be further converted
1616 # to wgOutputEncoding.
1617 global $wgEditEncoding;
1618 if( $wgEditEncoding == '' or
1619 $wgEditEncoding == 'UTF-8' ) {
1620 return $s;
1621 } else {
1622 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1623 }
1624 }
1625
1626 function recodeInput( $s ) {
1627 # Take the previous into account.
1628 global $wgEditEncoding;
1629 if($wgEditEncoding != "") {
1630 $enc = $wgEditEncoding;
1631 } else {
1632 $enc = 'UTF-8';
1633 }
1634 if( $enc == 'UTF-8' ) {
1635 return $s;
1636 } else {
1637 return $this->iconv( $enc, 'UTF-8', $s );
1638 }
1639 }
1640
1641 /**
1642 * For right-to-left language support
1643 *
1644 * @return bool
1645 */
1646 function isRTL() {
1647 $this->load();
1648 return $this->rtl;
1649 }
1650
1651 /**
1652 * A hidden direction mark (LRM or RLM), depending on the language direction
1653 *
1654 * @return string
1655 */
1656 function getDirMark() {
1657 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1658 }
1659
1660 /**
1661 * An arrow, depending on the language direction
1662 *
1663 * @return string
1664 */
1665 function getArrow() {
1666 return $this->isRTL() ? '←' : '→';
1667 }
1668
1669 /**
1670 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1671 *
1672 * @return bool
1673 */
1674 function linkPrefixExtension() {
1675 $this->load();
1676 return $this->linkPrefixExtension;
1677 }
1678
1679 function &getMagicWords() {
1680 $this->load();
1681 return $this->magicWords;
1682 }
1683
1684 # Fill a MagicWord object with data from here
1685 function getMagic( &$mw ) {
1686 if ( !$this->mMagicHookDone ) {
1687 $this->mMagicHookDone = true;
1688 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1689 }
1690 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1691 $rawEntry = $this->mMagicExtensions[$mw->mId];
1692 } else {
1693 $magicWords =& $this->getMagicWords();
1694 if ( isset( $magicWords[$mw->mId] ) ) {
1695 $rawEntry = $magicWords[$mw->mId];
1696 } else {
1697 # Fall back to English if local list is incomplete
1698 $magicWords =& Language::getMagicWords();
1699 if ( !isset($magicWords[$mw->mId]) ) {
1700 throw new MWException("Magic word '{$mw->mId}' not found" );
1701 }
1702 $rawEntry = $magicWords[$mw->mId];
1703 }
1704 }
1705
1706 if( !is_array( $rawEntry ) ) {
1707 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1708 } else {
1709 $mw->mCaseSensitive = $rawEntry[0];
1710 $mw->mSynonyms = array_slice( $rawEntry, 1 );
1711 }
1712 }
1713
1714 /**
1715 * Add magic words to the extension array
1716 */
1717 function addMagicWordsByLang( $newWords ) {
1718 $code = $this->getCode();
1719 $fallbackChain = array();
1720 while ( $code && !in_array( $code, $fallbackChain ) ) {
1721 $fallbackChain[] = $code;
1722 $code = self::getFallbackFor( $code );
1723 }
1724 if ( !in_array( 'en', $fallbackChain ) ) {
1725 $fallbackChain[] = 'en';
1726 }
1727 $fallbackChain = array_reverse( $fallbackChain );
1728 foreach ( $fallbackChain as $code ) {
1729 if ( isset( $newWords[$code] ) ) {
1730 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1731 }
1732 }
1733 }
1734
1735 /**
1736 * Get special page names, as an associative array
1737 * case folded alias => real name
1738 */
1739 function getSpecialPageAliases() {
1740 $this->load();
1741
1742 // Cache aliases because it may be slow to load them
1743 if ( !isset( $this->mExtendedSpecialPageAliases ) ) {
1744
1745 // Initialise array
1746 $this->mExtendedSpecialPageAliases = $this->specialPageAliases;
1747
1748 global $wgExtensionAliasesFiles;
1749 foreach ( $wgExtensionAliasesFiles as $file ) {
1750
1751 // Fail fast
1752 if ( !file_exists($file) )
1753 throw new MWException( "Aliases file does not exist: $file" );
1754
1755 $aliases = array();
1756 require($file);
1757
1758 // Check the availability of aliases
1759 if ( !isset($aliases['en']) )
1760 throw new MWException( "Malformed aliases file: $file" );
1761
1762 // Merge all aliases in fallback chain
1763 $code = $this->getCode();
1764 do {
1765 if ( !isset($aliases[$code]) ) continue;
1766
1767 $aliases[$code] = $this->fixSpecialPageAliases( $aliases[$code] );
1768 /* Merge the aliases, THIS will break if there is special page name
1769 * which looks like a numerical key, thanks to PHP...
1770 * See the array_merge_recursive manual entry */
1771 $this->mExtendedSpecialPageAliases = array_merge_recursive(
1772 $this->mExtendedSpecialPageAliases, $aliases[$code] );
1773
1774 } while ( $code = self::getFallbackFor( $code ) );
1775 }
1776
1777 wfRunHooks( 'LanguageGetSpecialPageAliases',
1778 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1779 }
1780
1781 return $this->mExtendedSpecialPageAliases;
1782 }
1783
1784 /**
1785 * Function to fix special page aliases. Will convert the first letter to
1786 * upper case and spaces to underscores. Can be given a full aliases array,
1787 * in which case it will recursively fix all aliases.
1788 */
1789 public function fixSpecialPageAliases( $mixed ) {
1790 // Work recursively until in string level
1791 if ( is_array($mixed) ) {
1792 $callback = array( $this, 'fixSpecialPageAliases' );
1793 return array_map( $callback, $mixed );
1794 }
1795 return str_replace( ' ', '_', $this->ucfirst( $mixed ) );
1796 }
1797
1798 /**
1799 * Italic is unsuitable for some languages
1800 *
1801 * @param $text String: the text to be emphasized.
1802 * @return string
1803 */
1804 function emphasize( $text ) {
1805 return "<em>$text</em>";
1806 }
1807
1808 /**
1809 * Normally we output all numbers in plain en_US style, that is
1810 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1811 * point twohundredthirtyfive. However this is not sutable for all
1812 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1813 * Icelandic just want to use commas instead of dots, and dots instead
1814 * of commas like "293.291,235".
1815 *
1816 * An example of this function being called:
1817 * <code>
1818 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1819 * </code>
1820 *
1821 * See LanguageGu.php for the Gujarati implementation and
1822 * $separatorTransformTable on MessageIs.php for
1823 * the , => . and . => , implementation.
1824 *
1825 * @todo check if it's viable to use localeconv() for the decimal
1826 * separator thing.
1827 * @param $number Mixed: the string to be formatted, should be an integer
1828 * or a floating point number.
1829 * @param $nocommafy Bool: set to true for special numbers like dates
1830 * @return string
1831 */
1832 function formatNum( $number, $nocommafy = false ) {
1833 global $wgTranslateNumerals;
1834 if (!$nocommafy) {
1835 $number = $this->commafy($number);
1836 $s = $this->separatorTransformTable();
1837 if (!is_null($s)) { $number = strtr($number, $s); }
1838 }
1839
1840 if ($wgTranslateNumerals) {
1841 $s = $this->digitTransformTable();
1842 if (!is_null($s)) { $number = strtr($number, $s); }
1843 }
1844
1845 return $number;
1846 }
1847
1848 function parseFormattedNumber( $number ) {
1849 $s = $this->digitTransformTable();
1850 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1851
1852 $s = $this->separatorTransformTable();
1853 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1854
1855 $number = strtr( $number, array (',' => '') );
1856 return $number;
1857 }
1858
1859 /**
1860 * Adds commas to a given number
1861 *
1862 * @param $_ mixed
1863 * @return string
1864 */
1865 function commafy($_) {
1866 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1867 }
1868
1869 function digitTransformTable() {
1870 $this->load();
1871 return $this->digitTransformTable;
1872 }
1873
1874 function separatorTransformTable() {
1875 $this->load();
1876 return $this->separatorTransformTable;
1877 }
1878
1879
1880 /**
1881 * For the credit list in includes/Credits.php (action=credits)
1882 *
1883 * @param $l Array
1884 * @return string
1885 */
1886 function listToText( $l ) {
1887 $s = '';
1888 $m = count($l) - 1;
1889 for ($i = $m; $i >= 0; $i--) {
1890 if ($i == $m) {
1891 $s = $l[$i];
1892 } else if ($i == $m - 1) {
1893 $s = $l[$i] . ' ' . $this->getMessageFromDB( 'and' ) . ' ' . $s;
1894 } else {
1895 $s = $l[$i] . ', ' . $s;
1896 }
1897 }
1898 return $s;
1899 }
1900
1901 /**
1902 * Take a list of strings and build a locale-friendly comma-separated
1903 * list, using the local comma-separator message.
1904 * @param $list array of strings to put in a comma list
1905 * @return string
1906 */
1907 function commaList( $list, $forContent = false ) {
1908 return implode(
1909 $list,
1910 wfMsgExt( 'comma-separator', array( 'escapenoentities', 'language' => $this ) ) );
1911 }
1912
1913 /**
1914 * Same as commaList, but separate it with the pipe instead.
1915 * @param $list array of strings to put in a pipe list
1916 * @return string
1917 */
1918 function pipeList( $list ) {
1919 return implode(
1920 $list,
1921 wfMsgExt( 'pipe-separator', array( 'escapenoentities', 'language' => $this ) ) );
1922 }
1923
1924 /**
1925 * Truncate a string to a specified length in bytes, appending an optional
1926 * string (e.g. for ellipses)
1927 *
1928 * The database offers limited byte lengths for some columns in the database;
1929 * multi-byte character sets mean we need to ensure that only whole characters
1930 * are included, otherwise broken characters can be passed to the user
1931 *
1932 * If $length is negative, the string will be truncated from the beginning
1933 *
1934 * @param $string String to truncate
1935 * @param $length Int: maximum length (excluding ellipses)
1936 * @param $ellipsis String to append to the truncated text
1937 * @return string
1938 */
1939 function truncate( $string, $length, $ellipsis = "" ) {
1940 if( $length == 0 ) {
1941 return $ellipsis;
1942 }
1943 if ( strlen( $string ) <= abs( $length ) ) {
1944 return $string;
1945 }
1946 if( $length > 0 ) {
1947 $string = substr( $string, 0, $length );
1948 $char = ord( $string[strlen( $string ) - 1] );
1949 $m = array();
1950 if ($char >= 0xc0) {
1951 # We got the first byte only of a multibyte char; remove it.
1952 $string = substr( $string, 0, -1 );
1953 } elseif( $char >= 0x80 &&
1954 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
1955 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
1956 # We chopped in the middle of a character; remove it
1957 $string = $m[1];
1958 }
1959 return $string . $ellipsis;
1960 } else {
1961 $string = substr( $string, $length );
1962 $char = ord( $string[0] );
1963 if( $char >= 0x80 && $char < 0xc0 ) {
1964 # We chopped in the middle of a character; remove the whole thing
1965 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
1966 }
1967 return $ellipsis . $string;
1968 }
1969 }
1970
1971 /**
1972 * Grammatical transformations, needed for inflected languages
1973 * Invoked by putting {{grammar:case|word}} in a message
1974 *
1975 * @param $word string
1976 * @param $case string
1977 * @return string
1978 */
1979 function convertGrammar( $word, $case ) {
1980 global $wgGrammarForms;
1981 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
1982 return $wgGrammarForms[$this->getCode()][$case][$word];
1983 }
1984 return $word;
1985 }
1986
1987 /**
1988 * Plural form transformations, needed for some languages.
1989 * For example, there are 3 form of plural in Russian and Polish,
1990 * depending on "count mod 10". See [[w:Plural]]
1991 * For English it is pretty simple.
1992 *
1993 * Invoked by putting {{plural:count|wordform1|wordform2}}
1994 * or {{plural:count|wordform1|wordform2|wordform3}}
1995 *
1996 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
1997 *
1998 * @param $count Integer: non-localized number
1999 * @param $forms Array: different plural forms
2000 * @return string Correct form of plural for $count in this language
2001 */
2002 function convertPlural( $count, $forms ) {
2003 if ( !count($forms) ) { return ''; }
2004 $forms = $this->preConvertPlural( $forms, 2 );
2005
2006 return ( $count == 1 ) ? $forms[0] : $forms[1];
2007 }
2008
2009 /**
2010 * Checks that convertPlural was given an array and pads it to requested
2011 * amound of forms by copying the last one.
2012 *
2013 * @param $count Integer: How many forms should there be at least
2014 * @param $forms Array of forms given to convertPlural
2015 * @return array Padded array of forms or an exception if not an array
2016 */
2017 protected function preConvertPlural( /* Array */ $forms, $count ) {
2018 while ( count($forms) < $count ) {
2019 $forms[] = $forms[count($forms)-1];
2020 }
2021 return $forms;
2022 }
2023
2024 /**
2025 * For translaing of expiry times
2026 * @param $str String: the validated block time in English
2027 * @return Somehow translated block time
2028 * @see LanguageFi.php for example implementation
2029 */
2030 function translateBlockExpiry( $str ) {
2031
2032 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2033
2034 if ( $scBlockExpiryOptions == '-') {
2035 return $str;
2036 }
2037
2038 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2039 if ( strpos($option, ":") === false )
2040 continue;
2041 list($show, $value) = explode(":", $option);
2042 if ( strcmp ( $str, $value) == 0 ) {
2043 return htmlspecialchars( trim( $show ) );
2044 }
2045 }
2046
2047 return $str;
2048 }
2049
2050 /**
2051 * languages like Chinese need to be segmented in order for the diff
2052 * to be of any use
2053 *
2054 * @param $text String
2055 * @return String
2056 */
2057 function segmentForDiff( $text ) {
2058 return $text;
2059 }
2060
2061 /**
2062 * and unsegment to show the result
2063 *
2064 * @param $text String
2065 * @return String
2066 */
2067 function unsegmentForDiff( $text ) {
2068 return $text;
2069 }
2070
2071 # convert text to different variants of a language.
2072 function convert( $text, $isTitle = false) {
2073 return $this->mConverter->convert($text, $isTitle);
2074 }
2075
2076 # Convert text from within Parser
2077 function parserConvert( $text, &$parser ) {
2078 return $this->mConverter->parserConvert( $text, $parser );
2079 }
2080
2081 # Check if this is a language with variants
2082 function hasVariants(){
2083 return sizeof($this->getVariants())>1;
2084 }
2085
2086 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2087 function armourMath($text){
2088 return $this->mConverter->armourMath($text);
2089 }
2090
2091
2092 /**
2093 * Perform output conversion on a string, and encode for safe HTML output.
2094 * @param $text String
2095 * @param $isTitle Bool -- wtf?
2096 * @return string
2097 * @todo this should get integrated somewhere sane
2098 */
2099 function convertHtml( $text, $isTitle = false ) {
2100 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2101 }
2102
2103 function convertCategoryKey( $key ) {
2104 return $this->mConverter->convertCategoryKey( $key );
2105 }
2106
2107 /**
2108 * get the list of variants supported by this langauge
2109 * see sample implementation in LanguageZh.php
2110 *
2111 * @return array an array of language codes
2112 */
2113 function getVariants() {
2114 return $this->mConverter->getVariants();
2115 }
2116
2117
2118 function getPreferredVariant( $fromUser = true ) {
2119 return $this->mConverter->getPreferredVariant( $fromUser );
2120 }
2121
2122 /**
2123 * if a language supports multiple variants, it is
2124 * possible that non-existing link in one variant
2125 * actually exists in another variant. this function
2126 * tries to find it. See e.g. LanguageZh.php
2127 *
2128 * @param $link String: the name of the link
2129 * @param $nt Mixed: the title object of the link
2130 * @return null the input parameters may be modified upon return
2131 */
2132 function findVariantLink( &$link, &$nt, $forTemplate = false ) {
2133 $this->mConverter->findVariantLink($link, $nt, $forTemplate );
2134 }
2135
2136 /**
2137 * If a language supports multiple variants, converts text
2138 * into an array of all possible variants of the text:
2139 * 'variant' => text in that variant
2140 */
2141
2142 function convertLinkToAllVariants($text){
2143 return $this->mConverter->convertLinkToAllVariants($text);
2144 }
2145
2146
2147 /**
2148 * returns language specific options used by User::getPageRenderHash()
2149 * for example, the preferred language variant
2150 *
2151 * @return string
2152 */
2153 function getExtraHashOptions() {
2154 return $this->mConverter->getExtraHashOptions();
2155 }
2156
2157 /**
2158 * for languages that support multiple variants, the title of an
2159 * article may be displayed differently in different variants. this
2160 * function returns the apporiate title defined in the body of the article.
2161 *
2162 * @return string
2163 */
2164 function getParsedTitle() {
2165 return $this->mConverter->getParsedTitle();
2166 }
2167
2168 /**
2169 * Enclose a string with the "no conversion" tag. This is used by
2170 * various functions in the Parser
2171 *
2172 * @param $text String: text to be tagged for no conversion
2173 * @param $noParse
2174 * @return string the tagged text
2175 */
2176 function markNoConversion( $text, $noParse=false ) {
2177 return $this->mConverter->markNoConversion( $text, $noParse );
2178 }
2179
2180 /**
2181 * A regular expression to match legal word-trailing characters
2182 * which should be merged onto a link of the form [[foo]]bar.
2183 *
2184 * @return string
2185 */
2186 function linkTrail() {
2187 $this->load();
2188 return $this->linkTrail;
2189 }
2190
2191 function getLangObj() {
2192 return $this;
2193 }
2194
2195 /**
2196 * Get the RFC 3066 code for this language object
2197 */
2198 function getCode() {
2199 return $this->mCode;
2200 }
2201
2202 function setCode( $code ) {
2203 $this->mCode = $code;
2204 }
2205
2206 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2207 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2208 }
2209
2210 static function getMessagesFileName( $code ) {
2211 global $IP;
2212 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2213 }
2214
2215 static function getClassFileName( $code ) {
2216 global $IP;
2217 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2218 }
2219
2220 static function getLocalisationArray( $code, $disableCache = false ) {
2221 self::loadLocalisation( $code, $disableCache );
2222 return self::$mLocalisationCache[$code];
2223 }
2224
2225 /**
2226 * Load localisation data for a given code into the static cache
2227 *
2228 * @return array Dependencies, map of filenames to mtimes
2229 */
2230 static function loadLocalisation( $code, $disableCache = false ) {
2231 static $recursionGuard = array();
2232 global $wgMemc, $wgCheckSerialized;
2233
2234 if ( !$code ) {
2235 throw new MWException( "Invalid language code requested" );
2236 }
2237
2238 if ( !$disableCache ) {
2239 # Try the per-process cache
2240 if ( isset( self::$mLocalisationCache[$code] ) ) {
2241 return self::$mLocalisationCache[$code]['deps'];
2242 }
2243
2244 wfProfileIn( __METHOD__ );
2245
2246 # Try the serialized directory
2247 $cache = wfGetPrecompiledData( self::getFileName( "Messages", $code, '.ser' ) );
2248 if ( $cache ) {
2249 if ( $wgCheckSerialized && self::isLocalisationOutOfDate( $cache ) ) {
2250 $cache = false;
2251 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2252 } else {
2253 self::$mLocalisationCache[$code] = $cache;
2254 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2255 wfProfileOut( __METHOD__ );
2256 return self::$mLocalisationCache[$code]['deps'];
2257 }
2258 }
2259
2260 # Try the global cache
2261 $memcKey = wfMemcKey('localisation', $code );
2262 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2263 $cache = $wgMemc->get( $memcKey );
2264 if ( $cache ) {
2265 if ( self::isLocalisationOutOfDate( $cache ) ) {
2266 $wgMemc->delete( $memcKey );
2267 $wgMemc->delete( $fbMemcKey );
2268 $cache = false;
2269 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2270 } else {
2271 self::$mLocalisationCache[$code] = $cache;
2272 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2273 wfProfileOut( __METHOD__ );
2274 return $cache['deps'];
2275 }
2276 }
2277 } else {
2278 wfProfileIn( __METHOD__ );
2279 }
2280
2281 # Default fallback, may be overridden when the messages file is included
2282 if ( $code != 'en' ) {
2283 $fallback = 'en';
2284 } else {
2285 $fallback = false;
2286 }
2287
2288 # Load the primary localisation from the source file
2289 $filename = self::getMessagesFileName( $code );
2290 if ( !file_exists( $filename ) ) {
2291 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2292 $cache = compact( self::$mLocalisationKeys ); // Set correct fallback
2293 $deps = array();
2294 } else {
2295 $deps = array( $filename => filemtime( $filename ) );
2296 require( $filename );
2297 $cache = compact( self::$mLocalisationKeys );
2298 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2299 }
2300
2301 if ( !empty( $fallback ) ) {
2302 # Load the fallback localisation, with a circular reference guard
2303 if ( isset( $recursionGuard[$code] ) ) {
2304 throw new MWException( "Error: Circular fallback reference in language code $code" );
2305 }
2306 $recursionGuard[$code] = true;
2307 $newDeps = self::loadLocalisation( $fallback, $disableCache );
2308 unset( $recursionGuard[$code] );
2309
2310 $secondary = self::$mLocalisationCache[$fallback];
2311 $deps = array_merge( $deps, $newDeps );
2312
2313 # Merge the fallback localisation with the current localisation
2314 foreach ( self::$mLocalisationKeys as $key ) {
2315 if ( isset( $cache[$key] ) ) {
2316 if ( isset( $secondary[$key] ) ) {
2317 if ( in_array( $key, self::$mMergeableMapKeys ) ) {
2318 $cache[$key] = $cache[$key] + $secondary[$key];
2319 } elseif ( in_array( $key, self::$mMergeableListKeys ) ) {
2320 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2321 } elseif ( in_array( $key, self::$mMergeableAliasListKeys ) ) {
2322 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2323 }
2324 }
2325 } else {
2326 $cache[$key] = $secondary[$key];
2327 }
2328 }
2329
2330 # Merge bookstore lists if requested
2331 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2332 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2333 }
2334 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2335 unset( $cache['bookstoreList']['inherit'] );
2336 }
2337 }
2338
2339 # Add dependencies to the cache entry
2340 $cache['deps'] = $deps;
2341
2342 # Replace spaces with underscores in namespace names
2343 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2344
2345 # And do the same for specialpage aliases. $page is an array.
2346 foreach ( $cache['specialPageAliases'] as &$page ) {
2347 $page = str_replace( ' ', '_', $page );
2348 }
2349 # Decouple the reference to prevent accidental damage
2350 unset($page);
2351
2352 # Save to both caches
2353 self::$mLocalisationCache[$code] = $cache;
2354 if ( !$disableCache ) {
2355 $wgMemc->set( $memcKey, $cache );
2356 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2357 }
2358
2359 wfProfileOut( __METHOD__ );
2360 return $deps;
2361 }
2362
2363 /**
2364 * Test if a given localisation cache is out of date with respect to the
2365 * source Messages files. This is done automatically for the global cache
2366 * in $wgMemc, but is only done on certain occasions for the serialized
2367 * data file.
2368 *
2369 * @param $cache mixed Either a language code or a cache array
2370 */
2371 static function isLocalisationOutOfDate( $cache ) {
2372 if ( !is_array( $cache ) ) {
2373 self::loadLocalisation( $cache );
2374 $cache = self::$mLocalisationCache[$cache];
2375 }
2376 $expired = false;
2377 foreach ( $cache['deps'] as $file => $mtime ) {
2378 if ( !file_exists( $file ) || filemtime( $file ) > $mtime ) {
2379 $expired = true;
2380 break;
2381 }
2382 }
2383 return $expired;
2384 }
2385
2386 /**
2387 * Get the fallback for a given language
2388 */
2389 static function getFallbackFor( $code ) {
2390 // Shortcut
2391 if ( $code === 'en' ) return false;
2392
2393 // Local cache
2394 static $cache = array();
2395 // Quick return
2396 if ( isset($cache[$code]) ) return $cache[$code];
2397
2398 // Try memcache
2399 global $wgMemc;
2400 $memcKey = wfMemcKey( 'fallback', $code );
2401 $fbcode = $wgMemc->get( $memcKey );
2402
2403 if ( is_string($fbcode) ) {
2404 // False is stored as a string to detect failures in memcache properly
2405 if ( $fbcode === '' ) $fbcode = false;
2406
2407 // Update local cache and return
2408 $cache[$code] = $fbcode;
2409 return $fbcode;
2410 }
2411
2412 // Nothing in caches, load and and update both caches
2413 self::loadLocalisation( $code );
2414 $fbcode = self::$mLocalisationCache[$code]['fallback'];
2415
2416 $cache[$code] = $fbcode;
2417 $wgMemc->set( $memcKey, (string) $fbcode );
2418
2419 return $fbcode;
2420 }
2421
2422 /**
2423 * Get all messages for a given language
2424 */
2425 static function getMessagesFor( $code ) {
2426 self::loadLocalisation( $code );
2427 return self::$mLocalisationCache[$code]['messages'];
2428 }
2429
2430 /**
2431 * Get a message for a given language
2432 */
2433 static function getMessageFor( $key, $code ) {
2434 self::loadLocalisation( $code );
2435 return isset( self::$mLocalisationCache[$code]['messages'][$key] ) ? self::$mLocalisationCache[$code]['messages'][$key] : null;
2436 }
2437
2438 /**
2439 * Load localisation data for this object
2440 */
2441 function load() {
2442 if ( !$this->mLoaded ) {
2443 self::loadLocalisation( $this->getCode() );
2444 $cache =& self::$mLocalisationCache[$this->getCode()];
2445 foreach ( self::$mLocalisationKeys as $key ) {
2446 $this->$key = $cache[$key];
2447 }
2448 $this->mLoaded = true;
2449
2450 $this->fixUpSettings();
2451 }
2452 }
2453
2454 /**
2455 * Do any necessary post-cache-load settings adjustment
2456 */
2457 function fixUpSettings() {
2458 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2459 $wgNamespaceAliases, $wgAmericanDates;
2460 wfProfileIn( __METHOD__ );
2461 if ( $wgExtraNamespaces ) {
2462 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
2463 }
2464
2465 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
2466 if ( $wgMetaNamespaceTalk ) {
2467 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
2468 } else {
2469 $talk = $this->namespaceNames[NS_PROJECT_TALK];
2470 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2471
2472 # Allow grammar transformations
2473 # Allowing full message-style parsing would make simple requests
2474 # such as action=raw much more expensive than they need to be.
2475 # This will hopefully cover most cases.
2476 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2477 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2478 $talk = str_replace( ' ', '_', $talk );
2479 $this->namespaceNames[NS_PROJECT_TALK] = $talk;
2480 }
2481
2482 # The above mixing may leave namespaces out of canonical order.
2483 # Re-order by namespace ID number...
2484 ksort( $this->namespaceNames );
2485
2486 # Put namespace names and aliases into a hashtable.
2487 # If this is too slow, then we should arrange it so that it is done
2488 # before caching. The catch is that at pre-cache time, the above
2489 # class-specific fixup hasn't been done.
2490 $this->mNamespaceIds = array();
2491 foreach ( $this->namespaceNames as $index => $name ) {
2492 $this->mNamespaceIds[$this->lc($name)] = $index;
2493 }
2494 if ( $this->namespaceAliases ) {
2495 foreach ( $this->namespaceAliases as $name => $index ) {
2496 $this->mNamespaceIds[$this->lc($name)] = $index;
2497 }
2498 }
2499 if ( $wgNamespaceAliases ) {
2500 foreach ( $wgNamespaceAliases as $name => $index ) {
2501 $this->mNamespaceIds[$this->lc($name)] = $index;
2502 }
2503 }
2504
2505 if ( $this->defaultDateFormat == 'dmy or mdy' ) {
2506 $this->defaultDateFormat = $wgAmericanDates ? 'mdy' : 'dmy';
2507 }
2508 wfProfileOut( __METHOD__ );
2509 }
2510
2511 function replaceGrammarInNamespace( $m ) {
2512 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2513 }
2514
2515 static function getCaseMaps() {
2516 static $wikiUpperChars, $wikiLowerChars;
2517 if ( isset( $wikiUpperChars ) ) {
2518 return array( $wikiUpperChars, $wikiLowerChars );
2519 }
2520
2521 wfProfileIn( __METHOD__ );
2522 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2523 if ( $arr === false ) {
2524 throw new MWException(
2525 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2526 }
2527 extract( $arr );
2528 wfProfileOut( __METHOD__ );
2529 return array( $wikiUpperChars, $wikiLowerChars );
2530 }
2531
2532 function formatTimePeriod( $seconds ) {
2533 if ( $seconds < 10 ) {
2534 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2535 } elseif ( $seconds < 60 ) {
2536 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2537 } elseif ( $seconds < 3600 ) {
2538 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2539 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2540 } else {
2541 $hours = floor( $seconds / 3600 );
2542 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2543 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2544 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2545 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2546 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2547 }
2548 }
2549
2550 function formatBitrate( $bps ) {
2551 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2552 if ( $bps <= 0 ) {
2553 return $this->formatNum( $bps ) . $units[0];
2554 }
2555 $unitIndex = floor( log10( $bps ) / 3 );
2556 $mantissa = $bps / pow( 1000, $unitIndex );
2557 if ( $mantissa < 10 ) {
2558 $mantissa = round( $mantissa, 1 );
2559 } else {
2560 $mantissa = round( $mantissa );
2561 }
2562 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2563 }
2564
2565 /**
2566 * Format a size in bytes for output, using an appropriate
2567 * unit (B, KB, MB or GB) according to the magnitude in question
2568 *
2569 * @param $size Size to format
2570 * @return string Plain text (not HTML)
2571 */
2572 function formatSize( $size ) {
2573 // For small sizes no decimal places necessary
2574 $round = 0;
2575 if( $size > 1024 ) {
2576 $size = $size / 1024;
2577 if( $size > 1024 ) {
2578 $size = $size / 1024;
2579 // For MB and bigger two decimal places are smarter
2580 $round = 2;
2581 if( $size > 1024 ) {
2582 $size = $size / 1024;
2583 $msg = 'size-gigabytes';
2584 } else {
2585 $msg = 'size-megabytes';
2586 }
2587 } else {
2588 $msg = 'size-kilobytes';
2589 }
2590 } else {
2591 $msg = 'size-bytes';
2592 }
2593 $size = round( $size, $round );
2594 $text = $this->getMessageFromDB( $msg );
2595 return str_replace( '$1', $this->formatNum( $size ), $text );
2596 }
2597 }