Revert r39949 "* Revert revert r39662 of my parser changes."
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * @defgroup Language Language
4 *
5 * @file
6 * @ingroup Language
7 */
8
9 if( !defined( 'MEDIAWIKI' ) ) {
10 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
11 exit( 1 );
12 }
13
14 # Read language names
15 global $wgLanguageNames;
16 require_once( dirname(__FILE__) . '/Names.php' ) ;
17
18 global $wgInputEncoding, $wgOutputEncoding;
19
20 /**
21 * These are always UTF-8, they exist only for backwards compatibility
22 */
23 $wgInputEncoding = "UTF-8";
24 $wgOutputEncoding = "UTF-8";
25
26 if( function_exists( 'mb_strtoupper' ) ) {
27 mb_internal_encoding('UTF-8');
28 }
29
30 /**
31 * a fake language converter
32 *
33 * @ingroup Language
34 */
35 class FakeConverter {
36 var $mLang;
37 function FakeConverter($langobj) {$this->mLang = $langobj;}
38 function convert($t, $i) {return $t;}
39 function parserConvert($t, $p) {return $t;}
40 function getVariants() { return array( $this->mLang->getCode() ); }
41 function getPreferredVariant() {return $this->mLang->getCode(); }
42 function findVariantLink(&$l, &$n, $forTemplate = false) {}
43 function getExtraHashOptions() {return '';}
44 function getParsedTitle() {return '';}
45 function markNoConversion($text, $noParse=false) {return $text;}
46 function convertCategoryKey( $key ) {return $key; }
47 function convertLinkToAllVariants($text){ return array( $this->mLang->getCode() => $text); }
48 function armourMath($text){ return $text; }
49 }
50
51 /**
52 * Internationalisation code
53 * @ingroup Language
54 */
55 class Language {
56 var $mConverter, $mVariants, $mCode, $mLoaded = false;
57 var $mMagicExtensions = array(), $mMagicHookDone = false;
58
59 static public $mLocalisationKeys = array( 'fallback', 'namespaceNames',
60 'skinNames', 'mathNames',
61 'bookstoreList', 'magicWords', 'messages', 'rtl', 'digitTransformTable',
62 'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
63 'defaultUserOptionOverrides', 'linkTrail', 'namespaceAliases',
64 'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
65 'defaultDateFormat', 'extraUserToggles', 'specialPageAliases',
66 'imageFiles'
67 );
68
69 static public $mMergeableMapKeys = array( 'messages', 'namespaceNames', 'mathNames',
70 'dateFormats', 'defaultUserOptionOverrides', 'magicWords', 'imageFiles' );
71
72 static public $mMergeableListKeys = array( 'extraUserToggles' );
73
74 static public $mMergeableAliasListKeys = array( 'specialPageAliases' );
75
76 static public $mLocalisationCache = array();
77
78 static public $mWeekdayMsgs = array(
79 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
80 'friday', 'saturday'
81 );
82
83 static public $mWeekdayAbbrevMsgs = array(
84 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
85 );
86
87 static public $mMonthMsgs = array(
88 'january', 'february', 'march', 'april', 'may_long', 'june',
89 'july', 'august', 'september', 'october', 'november',
90 'december'
91 );
92 static public $mMonthGenMsgs = array(
93 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
94 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
95 'december-gen'
96 );
97 static public $mMonthAbbrevMsgs = array(
98 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
99 'sep', 'oct', 'nov', 'dec'
100 );
101
102 static public $mIranianCalendarMonthMsgs = array(
103 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
104 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
105 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
106 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
107 );
108
109 static public $mHebrewCalendarMonthMsgs = array(
110 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
111 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
112 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
113 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
114 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
115 );
116
117 static public $mHebrewCalendarMonthGenMsgs = array(
118 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
119 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
120 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
121 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
122 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
123 );
124
125 static public $mHijriCalendarMonthMsgs = array(
126 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
127 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
128 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
129 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
130 );
131
132 /**
133 * Create a language object for a given language code
134 */
135 static function factory( $code ) {
136 global $IP;
137 static $recursionLevel = 0;
138
139 if ( $code == 'en' ) {
140 $class = 'Language';
141 } else {
142 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
143 // Preload base classes to work around APC/PHP5 bug
144 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
145 include_once("$IP/languages/classes/$class.deps.php");
146 }
147 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
148 include_once("$IP/languages/classes/$class.php");
149 }
150 }
151
152 if ( $recursionLevel > 5 ) {
153 throw new MWException( "Language fallback loop detected when creating class $class\n" );
154 }
155
156 if( ! class_exists( $class ) ) {
157 $fallback = Language::getFallbackFor( $code );
158 ++$recursionLevel;
159 $lang = Language::factory( $fallback );
160 --$recursionLevel;
161 $lang->setCode( $code );
162 } else {
163 $lang = new $class;
164 }
165
166 return $lang;
167 }
168
169 function __construct() {
170 $this->mConverter = new FakeConverter($this);
171 // Set the code to the name of the descendant
172 if ( get_class( $this ) == 'Language' ) {
173 $this->mCode = 'en';
174 } else {
175 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
176 }
177 }
178
179 /**
180 * Hook which will be called if this is the content language.
181 * Descendants can use this to register hook functions or modify globals
182 */
183 function initContLang() {}
184
185 /**
186 * @deprecated Use User::getDefaultOptions()
187 * @return array
188 */
189 function getDefaultUserOptions() {
190 wfDeprecated( __METHOD__ );
191 return User::getDefaultOptions();
192 }
193
194 function getFallbackLanguageCode() {
195 return self::getFallbackFor( $this->mCode );
196 }
197
198 /**
199 * Exports $wgBookstoreListEn
200 * @return array
201 */
202 function getBookstoreList() {
203 $this->load();
204 return $this->bookstoreList;
205 }
206
207 /**
208 * @return array
209 */
210 function getNamespaces() {
211 $this->load();
212 return $this->namespaceNames;
213 }
214
215 /**
216 * A convenience function that returns the same thing as
217 * getNamespaces() except with the array values changed to ' '
218 * where it found '_', useful for producing output to be displayed
219 * e.g. in <select> forms.
220 *
221 * @return array
222 */
223 function getFormattedNamespaces() {
224 $ns = $this->getNamespaces();
225 foreach($ns as $k => $v) {
226 $ns[$k] = strtr($v, '_', ' ');
227 }
228 return $ns;
229 }
230
231 /**
232 * Get a namespace value by key
233 * <code>
234 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
235 * echo $mw_ns; // prints 'MediaWiki'
236 * </code>
237 *
238 * @param $index Int: the array key of the namespace to return
239 * @return mixed, string if the namespace value exists, otherwise false
240 */
241 function getNsText( $index ) {
242 $ns = $this->getNamespaces();
243 return isset( $ns[$index] ) ? $ns[$index] : false;
244 }
245
246 /**
247 * A convenience function that returns the same thing as
248 * getNsText() except with '_' changed to ' ', useful for
249 * producing output.
250 *
251 * @return array
252 */
253 function getFormattedNsText( $index ) {
254 $ns = $this->getNsText( $index );
255 return strtr($ns, '_', ' ');
256 }
257
258 /**
259 * Get a namespace key by value, case insensitive.
260 * Only matches namespace names for the current language, not the
261 * canonical ones defined in Namespace.php.
262 *
263 * @param $text String
264 * @return mixed An integer if $text is a valid value otherwise false
265 */
266 function getLocalNsIndex( $text ) {
267 $this->load();
268 $lctext = $this->lc($text);
269 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
270 }
271
272 /**
273 * Get a namespace key by value, case insensitive. Canonical namespace
274 * names override custom ones defined for the current language.
275 *
276 * @param $text String
277 * @return mixed An integer if $text is a valid value otherwise false
278 */
279 function getNsIndex( $text ) {
280 $this->load();
281 $lctext = $this->lc($text);
282 if( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
283 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
284 }
285
286 /**
287 * short names for language variants used for language conversion links.
288 *
289 * @param $code String
290 * @return string
291 */
292 function getVariantname( $code ) {
293 return $this->getMessageFromDB( "variantname-$code" );
294 }
295
296 function specialPage( $name ) {
297 $aliases = $this->getSpecialPageAliases();
298 if ( isset( $aliases[$name][0] ) ) {
299 $name = $aliases[$name][0];
300 }
301 return $this->getNsText(NS_SPECIAL) . ':' . $name;
302 }
303
304 function getQuickbarSettings() {
305 return array(
306 $this->getMessage( 'qbsettings-none' ),
307 $this->getMessage( 'qbsettings-fixedleft' ),
308 $this->getMessage( 'qbsettings-fixedright' ),
309 $this->getMessage( 'qbsettings-floatingleft' ),
310 $this->getMessage( 'qbsettings-floatingright' )
311 );
312 }
313
314 function getSkinNames() {
315 $this->load();
316 return $this->skinNames;
317 }
318
319 function getMathNames() {
320 $this->load();
321 return $this->mathNames;
322 }
323
324 function getDatePreferences() {
325 $this->load();
326 return $this->datePreferences;
327 }
328
329 function getDateFormats() {
330 $this->load();
331 return $this->dateFormats;
332 }
333
334 function getDefaultDateFormat() {
335 $this->load();
336 return $this->defaultDateFormat;
337 }
338
339 function getDatePreferenceMigrationMap() {
340 $this->load();
341 return $this->datePreferenceMigrationMap;
342 }
343
344 function getImageFile( $image ) {
345 $this->load();
346 return $this->imageFiles[$image];
347 }
348
349 function getDefaultUserOptionOverrides() {
350 $this->load();
351 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
352 if (is_array($this->defaultUserOptionOverrides)) {
353 return $this->defaultUserOptionOverrides;
354 } else {
355 return array();
356 }
357 }
358
359 function getExtraUserToggles() {
360 $this->load();
361 return $this->extraUserToggles;
362 }
363
364 function getUserToggle( $tog ) {
365 return $this->getMessageFromDB( "tog-$tog" );
366 }
367
368 /**
369 * Get language names, indexed by code.
370 * If $customisedOnly is true, only returns codes with a messages file
371 */
372 public static function getLanguageNames( $customisedOnly = false ) {
373 global $wgLanguageNames, $wgExtraLanguageNames;
374 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
375 if ( !$customisedOnly ) {
376 return $allNames;
377 }
378
379 global $IP;
380 $names = array();
381 $dir = opendir( "$IP/languages/messages" );
382 while( false !== ( $file = readdir( $dir ) ) ) {
383 $m = array();
384 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
385 $code = str_replace( '_', '-', strtolower( $m[1] ) );
386 if ( isset( $allNames[$code] ) ) {
387 $names[$code] = $allNames[$code];
388 }
389 }
390 }
391 closedir( $dir );
392 return $names;
393 }
394
395 /**
396 * Ugly hack to get a message maybe from the MediaWiki namespace, if this
397 * language object is the content or user language.
398 */
399 function getMessageFromDB( $msg ) {
400 global $wgContLang, $wgLang;
401 if ( $wgContLang->getCode() == $this->getCode() ) {
402 # Content language
403 return wfMsgForContent( $msg );
404 } elseif ( $wgLang->getCode() == $this->getCode() ) {
405 # User language
406 return wfMsg( $msg );
407 } else {
408 # Neither, get from localisation
409 return $this->getMessage( $msg );
410 }
411 }
412
413 function getLanguageName( $code ) {
414 $names = self::getLanguageNames();
415 if ( !array_key_exists( $code, $names ) ) {
416 return '';
417 }
418 return $names[$code];
419 }
420
421 function getMonthName( $key ) {
422 return $this->getMessageFromDB( self::$mMonthMsgs[$key-1] );
423 }
424
425 function getMonthNameGen( $key ) {
426 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key-1] );
427 }
428
429 function getMonthAbbreviation( $key ) {
430 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key-1] );
431 }
432
433 function getWeekdayName( $key ) {
434 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key-1] );
435 }
436
437 function getWeekdayAbbreviation( $key ) {
438 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key-1] );
439 }
440
441 function getIranianCalendarMonthName( $key ) {
442 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key-1] );
443 }
444
445 function getHebrewCalendarMonthName( $key ) {
446 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key-1] );
447 }
448
449 function getHebrewCalendarMonthNameGen( $key ) {
450 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key-1] );
451 }
452
453 function getHijriCalendarMonthName( $key ) {
454 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key-1] );
455 }
456
457 /**
458 * Used by date() and time() to adjust the time output.
459 *
460 * @param $ts Int the time in date('YmdHis') format
461 * @param $tz Mixed: adjust the time by this amount (default false, mean we
462 * get user timecorrection setting)
463 * @return int
464 */
465 function userAdjust( $ts, $tz = false ) {
466 global $wgUser, $wgLocalTZoffset;
467
468 if (!$tz) {
469 $tz = $wgUser->getOption( 'timecorrection' );
470 }
471
472 # minutes and hours differences:
473 $minDiff = 0;
474 $hrDiff = 0;
475
476 if ( $tz === '' ) {
477 # Global offset in minutes.
478 if( isset($wgLocalTZoffset) ) {
479 if( $wgLocalTZoffset >= 0 ) {
480 $hrDiff = floor($wgLocalTZoffset / 60);
481 } else {
482 $hrDiff = ceil($wgLocalTZoffset / 60);
483 }
484 $minDiff = $wgLocalTZoffset % 60;
485 }
486 } elseif ( strpos( $tz, ':' ) !== false ) {
487 $tzArray = explode( ':', $tz );
488 $hrDiff = intval($tzArray[0]);
489 $minDiff = intval($hrDiff < 0 ? -$tzArray[1] : $tzArray[1]);
490 } else {
491 $hrDiff = intval( $tz );
492 }
493
494 # No difference ? Return time unchanged
495 if ( 0 == $hrDiff && 0 == $minDiff ) { return $ts; }
496
497 wfSuppressWarnings(); // E_STRICT system time bitching
498 # Generate an adjusted date
499 $t = mktime( (
500 (int)substr( $ts, 8, 2) ) + $hrDiff, # Hours
501 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
502 (int)substr( $ts, 12, 2 ), # Seconds
503 (int)substr( $ts, 4, 2 ), # Month
504 (int)substr( $ts, 6, 2 ), # Day
505 (int)substr( $ts, 0, 4 ) ); #Year
506
507 $date = date( 'YmdHis', $t );
508 wfRestoreWarnings();
509
510 return $date;
511 }
512
513 /**
514 * This is a workalike of PHP's date() function, but with better
515 * internationalisation, a reduced set of format characters, and a better
516 * escaping format.
517 *
518 * Supported format characters are dDjlNwzWFmMntLYyaAgGhHiscrU. See the
519 * PHP manual for definitions. There are a number of extensions, which
520 * start with "x":
521 *
522 * xn Do not translate digits of the next numeric format character
523 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
524 * xr Use roman numerals for the next numeric format character
525 * xh Use hebrew numerals for the next numeric format character
526 * xx Literal x
527 * xg Genitive month name
528 *
529 * xij j (day number) in Iranian calendar
530 * xiF F (month name) in Iranian calendar
531 * xin n (month number) in Iranian calendar
532 * xiY Y (full year) in Iranian calendar
533 *
534 * xjj j (day number) in Hebrew calendar
535 * xjF F (month name) in Hebrew calendar
536 * xjt t (days in month) in Hebrew calendar
537 * xjx xg (genitive month name) in Hebrew calendar
538 * xjn n (month number) in Hebrew calendar
539 * xjY Y (full year) in Hebrew calendar
540 *
541 * xmj j (day number) in Hijri calendar
542 * xmF F (month name) in Hijri calendar
543 * xmn n (month number) in Hijri calendar
544 * xmY Y (full year) in Hijri calendar
545 *
546 * xkY Y (full year) in Thai solar calendar. Months and days are
547 * identical to the Gregorian calendar
548 *
549 * Characters enclosed in double quotes will be considered literal (with
550 * the quotes themselves removed). Unmatched quotes will be considered
551 * literal quotes. Example:
552 *
553 * "The month is" F => The month is January
554 * i's" => 20'11"
555 *
556 * Backslash escaping is also supported.
557 *
558 * Input timestamp is assumed to be pre-normalized to the desired local
559 * time zone, if any.
560 *
561 * @param $format String
562 * @param $ts String: 14-character timestamp
563 * YYYYMMDDHHMMSS
564 * 01234567890123
565 */
566 function sprintfDate( $format, $ts ) {
567 $s = '';
568 $raw = false;
569 $roman = false;
570 $hebrewNum = false;
571 $unix = false;
572 $rawToggle = false;
573 $iranian = false;
574 $hebrew = false;
575 $hijri = false;
576 $thai = false;
577 for ( $p = 0; $p < strlen( $format ); $p++ ) {
578 $num = false;
579 $code = $format[$p];
580 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
581 $code .= $format[++$p];
582 }
583
584 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' ) && $p < strlen( $format ) - 1 ) {
585 $code .= $format[++$p];
586 }
587
588 switch ( $code ) {
589 case 'xx':
590 $s .= 'x';
591 break;
592 case 'xn':
593 $raw = true;
594 break;
595 case 'xN':
596 $rawToggle = !$rawToggle;
597 break;
598 case 'xr':
599 $roman = true;
600 break;
601 case 'xh':
602 $hebrewNum = true;
603 break;
604 case 'xg':
605 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
606 break;
607 case 'xjx':
608 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
609 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
610 break;
611 case 'd':
612 $num = substr( $ts, 6, 2 );
613 break;
614 case 'D':
615 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
616 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
617 break;
618 case 'j':
619 $num = intval( substr( $ts, 6, 2 ) );
620 break;
621 case 'xij':
622 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
623 $num = $iranian[2];
624 break;
625 case 'xmj':
626 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
627 $num = $hijri[2];
628 break;
629 case 'xjj':
630 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
631 $num = $hebrew[2];
632 break;
633 case 'l':
634 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
635 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
636 break;
637 case 'N':
638 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
639 $w = gmdate( 'w', $unix );
640 $num = $w ? $w : 7;
641 break;
642 case 'w':
643 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
644 $num = gmdate( 'w', $unix );
645 break;
646 case 'z':
647 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
648 $num = gmdate( 'z', $unix );
649 break;
650 case 'W':
651 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
652 $num = gmdate( 'W', $unix );
653 break;
654 case 'F':
655 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
656 break;
657 case 'xiF':
658 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
659 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
660 break;
661 case 'xmF':
662 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
663 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
664 break;
665 case 'xjF':
666 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
667 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
668 break;
669 case 'm':
670 $num = substr( $ts, 4, 2 );
671 break;
672 case 'M':
673 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
674 break;
675 case 'n':
676 $num = intval( substr( $ts, 4, 2 ) );
677 break;
678 case 'xin':
679 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
680 $num = $iranian[1];
681 break;
682 case 'xmn':
683 if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
684 $num = $hijri[1];
685 break;
686 case 'xjn':
687 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
688 $num = $hebrew[1];
689 break;
690 case 't':
691 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
692 $num = gmdate( 't', $unix );
693 break;
694 case 'xjt':
695 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
696 $num = $hebrew[3];
697 break;
698 case 'L':
699 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
700 $num = gmdate( 'L', $unix );
701 break;
702 case 'Y':
703 $num = substr( $ts, 0, 4 );
704 break;
705 case 'xiY':
706 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
707 $num = $iranian[0];
708 break;
709 case 'xmY':
710 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
711 $num = $hijri[0];
712 break;
713 case 'xjY':
714 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
715 $num = $hebrew[0];
716 break;
717 case 'xkY':
718 if ( !$thai ) $thai = self::tsToThai( $ts );
719 $num = $thai[0];
720 break;
721 case 'y':
722 $num = substr( $ts, 2, 2 );
723 break;
724 case 'a':
725 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
726 break;
727 case 'A':
728 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
729 break;
730 case 'g':
731 $h = substr( $ts, 8, 2 );
732 $num = $h % 12 ? $h % 12 : 12;
733 break;
734 case 'G':
735 $num = intval( substr( $ts, 8, 2 ) );
736 break;
737 case 'h':
738 $h = substr( $ts, 8, 2 );
739 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
740 break;
741 case 'H':
742 $num = substr( $ts, 8, 2 );
743 break;
744 case 'i':
745 $num = substr( $ts, 10, 2 );
746 break;
747 case 's':
748 $num = substr( $ts, 12, 2 );
749 break;
750 case 'c':
751 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
752 $s .= gmdate( 'c', $unix );
753 break;
754 case 'r':
755 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
756 $s .= gmdate( 'r', $unix );
757 break;
758 case 'U':
759 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
760 $num = $unix;
761 break;
762 case '\\':
763 # Backslash escaping
764 if ( $p < strlen( $format ) - 1 ) {
765 $s .= $format[++$p];
766 } else {
767 $s .= '\\';
768 }
769 break;
770 case '"':
771 # Quoted literal
772 if ( $p < strlen( $format ) - 1 ) {
773 $endQuote = strpos( $format, '"', $p + 1 );
774 if ( $endQuote === false ) {
775 # No terminating quote, assume literal "
776 $s .= '"';
777 } else {
778 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
779 $p = $endQuote;
780 }
781 } else {
782 # Quote at end of string, assume literal "
783 $s .= '"';
784 }
785 break;
786 default:
787 $s .= $format[$p];
788 }
789 if ( $num !== false ) {
790 if ( $rawToggle || $raw ) {
791 $s .= $num;
792 $raw = false;
793 } elseif ( $roman ) {
794 $s .= self::romanNumeral( $num );
795 $roman = false;
796 } elseif( $hebrewNum ) {
797 $s .= self::hebrewNumeral( $num );
798 $hebrewNum = false;
799 } else {
800 $s .= $this->formatNum( $num, true );
801 }
802 $num = false;
803 }
804 }
805 return $s;
806 }
807
808 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
809 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
810 /**
811 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
812 * Gregorian dates to Iranian dates. Originally written in C, it
813 * is released under the terms of GNU Lesser General Public
814 * License. Conversion to PHP was performed by Niklas Laxström.
815 *
816 * Link: http://www.farsiweb.info/jalali/jalali.c
817 */
818 private static function tsToIranian( $ts ) {
819 $gy = substr( $ts, 0, 4 ) -1600;
820 $gm = substr( $ts, 4, 2 ) -1;
821 $gd = substr( $ts, 6, 2 ) -1;
822
823 # Days passed from the beginning (including leap years)
824 $gDayNo = 365*$gy
825 + floor(($gy+3) / 4)
826 - floor(($gy+99) / 100)
827 + floor(($gy+399) / 400);
828
829
830 // Add days of the past months of this year
831 for( $i = 0; $i < $gm; $i++ ) {
832 $gDayNo += self::$GREG_DAYS[$i];
833 }
834
835 // Leap years
836 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
837 $gDayNo++;
838 }
839
840 // Days passed in current month
841 $gDayNo += $gd;
842
843 $jDayNo = $gDayNo - 79;
844
845 $jNp = floor($jDayNo / 12053);
846 $jDayNo %= 12053;
847
848 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
849 $jDayNo %= 1461;
850
851 if ( $jDayNo >= 366 ) {
852 $jy += floor(($jDayNo-1)/365);
853 $jDayNo = floor(($jDayNo-1)%365);
854 }
855
856 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
857 $jDayNo -= self::$IRANIAN_DAYS[$i];
858 }
859
860 $jm= $i+1;
861 $jd= $jDayNo+1;
862
863 return array($jy, $jm, $jd);
864 }
865 /**
866 * Converting Gregorian dates to Hijri dates.
867 *
868 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
869 *
870 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
871 */
872 private static function tsToHijri ( $ts ) {
873 $year = substr( $ts, 0, 4 );
874 $month = substr( $ts, 4, 2 );
875 $day = substr( $ts, 6, 2 );
876
877 $zyr = $year;
878 $zd=$day;
879 $zm=$month;
880 $zy=$zyr;
881
882
883
884 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
885 {
886
887
888 $zjd=(int)((1461*($zy + 4800 + (int)( ($zm-14) /12) ))/4) + (int)((367*($zm-2-12*((int)(($zm-14)/12))))/12)-(int)((3*(int)(( ($zy+4900+(int)(($zm-14)/12))/100)))/4)+$zd-32075;
889 }
890 else
891 {
892 $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
893 }
894
895 $zl=$zjd-1948440+10632;
896 $zn=(int)(($zl-1)/10631);
897 $zl=$zl-10631*$zn+354;
898 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
899 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
900 $zm=(int)((24*$zl)/709);
901 $zd=$zl-(int)((709*$zm)/24);
902 $zy=30*$zn+$zj-30;
903
904 return array ($zy, $zm, $zd);
905 }
906
907 /**
908 * Converting Gregorian dates to Hebrew dates.
909 *
910 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
911 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
912 * to translate the relevant functions into PHP and release them under
913 * GNU GPL.
914 *
915 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
916 * and Adar II is 14. In a non-leap year, Adar is 6.
917 */
918 private static function tsToHebrew( $ts ) {
919 # Parse date
920 $year = substr( $ts, 0, 4 );
921 $month = substr( $ts, 4, 2 );
922 $day = substr( $ts, 6, 2 );
923
924 # Calculate Hebrew year
925 $hebrewYear = $year + 3760;
926
927 # Month number when September = 1, August = 12
928 $month += 4;
929 if( $month > 12 ) {
930 # Next year
931 $month -= 12;
932 $year++;
933 $hebrewYear++;
934 }
935
936 # Calculate day of year from 1 September
937 $dayOfYear = $day;
938 for( $i = 1; $i < $month; $i++ ) {
939 if( $i == 6 ) {
940 # February
941 $dayOfYear += 28;
942 # Check if the year is leap
943 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
944 $dayOfYear++;
945 }
946 } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
947 $dayOfYear += 30;
948 } else {
949 $dayOfYear += 31;
950 }
951 }
952
953 # Calculate the start of the Hebrew year
954 $start = self::hebrewYearStart( $hebrewYear );
955
956 # Calculate next year's start
957 if( $dayOfYear <= $start ) {
958 # Day is before the start of the year - it is the previous year
959 # Next year's start
960 $nextStart = $start;
961 # Previous year
962 $year--;
963 $hebrewYear--;
964 # Add days since previous year's 1 September
965 $dayOfYear += 365;
966 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
967 # Leap year
968 $dayOfYear++;
969 }
970 # Start of the new (previous) year
971 $start = self::hebrewYearStart( $hebrewYear );
972 } else {
973 # Next year's start
974 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
975 }
976
977 # Calculate Hebrew day of year
978 $hebrewDayOfYear = $dayOfYear - $start;
979
980 # Difference between year's days
981 $diff = $nextStart - $start;
982 # Add 12 (or 13 for leap years) days to ignore the difference between
983 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
984 # difference is only about the year type
985 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
986 $diff += 13;
987 } else {
988 $diff += 12;
989 }
990
991 # Check the year pattern, and is leap year
992 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
993 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
994 # and non-leap years
995 $yearPattern = $diff % 30;
996 # Check if leap year
997 $isLeap = $diff >= 30;
998
999 # Calculate day in the month from number of day in the Hebrew year
1000 # Don't check Adar - if the day is not in Adar, we will stop before;
1001 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1002 $hebrewDay = $hebrewDayOfYear;
1003 $hebrewMonth = 1;
1004 $days = 0;
1005 while( $hebrewMonth <= 12 ) {
1006 # Calculate days in this month
1007 if( $isLeap && $hebrewMonth == 6 ) {
1008 # Adar in a leap year
1009 if( $isLeap ) {
1010 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1011 $days = 30;
1012 if( $hebrewDay <= $days ) {
1013 # Day in Adar I
1014 $hebrewMonth = 13;
1015 } else {
1016 # Subtract the days of Adar I
1017 $hebrewDay -= $days;
1018 # Try Adar II
1019 $days = 29;
1020 if( $hebrewDay <= $days ) {
1021 # Day in Adar II
1022 $hebrewMonth = 14;
1023 }
1024 }
1025 }
1026 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1027 # Cheshvan in a complete year (otherwise as the rule below)
1028 $days = 30;
1029 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1030 # Kislev in an incomplete year (otherwise as the rule below)
1031 $days = 29;
1032 } else {
1033 # Odd months have 30 days, even have 29
1034 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1035 }
1036 if( $hebrewDay <= $days ) {
1037 # In the current month
1038 break;
1039 } else {
1040 # Subtract the days of the current month
1041 $hebrewDay -= $days;
1042 # Try in the next month
1043 $hebrewMonth++;
1044 }
1045 }
1046
1047 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1048 }
1049
1050 /**
1051 * This calculates the Hebrew year start, as days since 1 September.
1052 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1053 * Used for Hebrew date.
1054 */
1055 private static function hebrewYearStart( $year ) {
1056 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1057 $b = intval( ( $year - 1 ) % 4 );
1058 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1059 if( $m < 0 ) {
1060 $m--;
1061 }
1062 $Mar = intval( $m );
1063 if( $m < 0 ) {
1064 $m++;
1065 }
1066 $m -= $Mar;
1067
1068 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1069 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1070 $Mar++;
1071 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1072 $Mar += 2;
1073 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1074 $Mar++;
1075 }
1076
1077 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1078 return $Mar;
1079 }
1080
1081 /**
1082 * Algorithm to convert Gregorian dates to Thai solar dates.
1083 *
1084 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1085 *
1086 * @param $ts String: 14-character timestamp
1087 * @return array converted year, month, day
1088 */
1089 private static function tsToThai( $ts ) {
1090 $gy = substr( $ts, 0, 4 );
1091 $gm = substr( $ts, 4, 2 );
1092 $gd = substr( $ts, 6, 2 );
1093
1094 # Add 543 years to the Gregorian calendar
1095 # Months and days are identical
1096 $gy_thai = $gy + 543;
1097
1098 return array( $gy_thai, $gm, $gd );
1099 }
1100
1101
1102 /**
1103 * Roman number formatting up to 3000
1104 */
1105 static function romanNumeral( $num ) {
1106 static $table = array(
1107 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1108 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1109 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1110 array( '', 'M', 'MM', 'MMM' )
1111 );
1112
1113 $num = intval( $num );
1114 if ( $num > 3000 || $num <= 0 ) {
1115 return $num;
1116 }
1117
1118 $s = '';
1119 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1120 if ( $num >= $pow10 ) {
1121 $s .= $table[$i][floor($num / $pow10)];
1122 }
1123 $num = $num % $pow10;
1124 }
1125 return $s;
1126 }
1127
1128 /**
1129 * Hebrew Gematria number formatting up to 9999
1130 */
1131 static function hebrewNumeral( $num ) {
1132 static $table = array(
1133 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1134 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1135 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1136 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1137 );
1138
1139 $num = intval( $num );
1140 if ( $num > 9999 || $num <= 0 ) {
1141 return $num;
1142 }
1143
1144 $s = '';
1145 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1146 if ( $num >= $pow10 ) {
1147 if ( $num == 15 || $num == 16 ) {
1148 $s .= $table[0][9] . $table[0][$num - 9];
1149 $num = 0;
1150 } else {
1151 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1152 if( $pow10 == 1000 ) {
1153 $s .= "'";
1154 }
1155 }
1156 }
1157 $num = $num % $pow10;
1158 }
1159 if( strlen( $s ) == 2 ) {
1160 $str = $s . "'";
1161 } else {
1162 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1163 $str .= substr( $s, strlen( $s ) - 2, 2 );
1164 }
1165 $start = substr( $str, 0, strlen( $str ) - 2 );
1166 $end = substr( $str, strlen( $str ) - 2 );
1167 switch( $end ) {
1168 case 'כ':
1169 $str = $start . 'ך';
1170 break;
1171 case 'מ':
1172 $str = $start . 'ם';
1173 break;
1174 case 'נ':
1175 $str = $start . 'ן';
1176 break;
1177 case 'פ':
1178 $str = $start . 'ף';
1179 break;
1180 case 'צ':
1181 $str = $start . 'ץ';
1182 break;
1183 }
1184 return $str;
1185 }
1186
1187 /**
1188 * This is meant to be used by time(), date(), and timeanddate() to get
1189 * the date preference they're supposed to use, it should be used in
1190 * all children.
1191 *
1192 *<code>
1193 * function timeanddate([...], $format = true) {
1194 * $datePreference = $this->dateFormat($format);
1195 * [...]
1196 * }
1197 *</code>
1198 *
1199 * @param $usePrefs Mixed: if true, the user's preference is used
1200 * if false, the site/language default is used
1201 * if int/string, assumed to be a format.
1202 * @return string
1203 */
1204 function dateFormat( $usePrefs = true ) {
1205 global $wgUser;
1206
1207 if( is_bool( $usePrefs ) ) {
1208 if( $usePrefs ) {
1209 $datePreference = $wgUser->getDatePreference();
1210 } else {
1211 $options = User::getDefaultOptions();
1212 $datePreference = (string)$options['date'];
1213 }
1214 } else {
1215 $datePreference = (string)$usePrefs;
1216 }
1217
1218 // return int
1219 if( $datePreference == '' ) {
1220 return 'default';
1221 }
1222
1223 return $datePreference;
1224 }
1225
1226 /**
1227 * @param $ts Mixed: the time format which needs to be turned into a
1228 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1229 * @param $adj Bool: whether to adjust the time output according to the
1230 * user configured offset ($timecorrection)
1231 * @param $format Mixed: true to use user's date format preference
1232 * @param $timecorrection String: the time offset as returned by
1233 * validateTimeZone() in Special:Preferences
1234 * @return string
1235 */
1236 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1237 $this->load();
1238 if ( $adj ) {
1239 $ts = $this->userAdjust( $ts, $timecorrection );
1240 }
1241
1242 $pref = $this->dateFormat( $format );
1243 if( $pref == 'default' || !isset( $this->dateFormats["$pref date"] ) ) {
1244 $pref = $this->defaultDateFormat;
1245 }
1246 return $this->sprintfDate( $this->dateFormats["$pref date"], $ts );
1247 }
1248
1249 /**
1250 * @param $ts Mixed: the time format which needs to be turned into a
1251 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1252 * @param $adj Bool: whether to adjust the time output according to the
1253 * user configured offset ($timecorrection)
1254 * @param $format Mixed: true to use user's date format preference
1255 * @param $timecorrection String: the time offset as returned by
1256 * validateTimeZone() in Special:Preferences
1257 * @return string
1258 */
1259 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1260 $this->load();
1261 if ( $adj ) {
1262 $ts = $this->userAdjust( $ts, $timecorrection );
1263 }
1264
1265 $pref = $this->dateFormat( $format );
1266 if( $pref == 'default' || !isset( $this->dateFormats["$pref time"] ) ) {
1267 $pref = $this->defaultDateFormat;
1268 }
1269 return $this->sprintfDate( $this->dateFormats["$pref time"], $ts );
1270 }
1271
1272 /**
1273 * @param $ts Mixed: the time format which needs to be turned into a
1274 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1275 * @param $adj Bool: whether to adjust the time output according to the
1276 * user configured offset ($timecorrection)
1277 * @param $format Mixed: what format to return, if it's false output the
1278 * default one (default true)
1279 * @param $timecorrection String: the time offset as returned by
1280 * validateTimeZone() in Special:Preferences
1281 * @return string
1282 */
1283 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1284 $this->load();
1285
1286 $ts = wfTimestamp( TS_MW, $ts );
1287
1288 if ( $adj ) {
1289 $ts = $this->userAdjust( $ts, $timecorrection );
1290 }
1291
1292 $pref = $this->dateFormat( $format );
1293 if( $pref == 'default' || !isset( $this->dateFormats["$pref both"] ) ) {
1294 $pref = $this->defaultDateFormat;
1295 }
1296
1297 return $this->sprintfDate( $this->dateFormats["$pref both"], $ts );
1298 }
1299
1300 function getMessage( $key ) {
1301 $this->load();
1302 return isset( $this->messages[$key] ) ? $this->messages[$key] : null;
1303 }
1304
1305 function getAllMessages() {
1306 $this->load();
1307 return $this->messages;
1308 }
1309
1310 function iconv( $in, $out, $string ) {
1311 # For most languages, this is a wrapper for iconv
1312 return iconv( $in, $out . '//IGNORE', $string );
1313 }
1314
1315 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1316 function ucwordbreaksCallbackAscii($matches){
1317 return $this->ucfirst($matches[1]);
1318 }
1319
1320 function ucwordbreaksCallbackMB($matches){
1321 return mb_strtoupper($matches[0]);
1322 }
1323
1324 function ucCallback($matches){
1325 list( $wikiUpperChars ) = self::getCaseMaps();
1326 return strtr( $matches[1], $wikiUpperChars );
1327 }
1328
1329 function lcCallback($matches){
1330 list( , $wikiLowerChars ) = self::getCaseMaps();
1331 return strtr( $matches[1], $wikiLowerChars );
1332 }
1333
1334 function ucwordsCallbackMB($matches){
1335 return mb_strtoupper($matches[0]);
1336 }
1337
1338 function ucwordsCallbackWiki($matches){
1339 list( $wikiUpperChars ) = self::getCaseMaps();
1340 return strtr( $matches[0], $wikiUpperChars );
1341 }
1342
1343 function ucfirst( $str ) {
1344 if ( empty($str) ) return $str;
1345 if ( ord($str[0]) < 128 ) return ucfirst($str);
1346 else return self::uc($str,true); // fall back to more complex logic in case of multibyte strings
1347 }
1348
1349 function uc( $str, $first = false ) {
1350 if ( function_exists( 'mb_strtoupper' ) ) {
1351 if ( $first ) {
1352 if ( self::isMultibyte( $str ) ) {
1353 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1354 } else {
1355 return ucfirst( $str );
1356 }
1357 } else {
1358 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1359 }
1360 } else {
1361 if ( self::isMultibyte( $str ) ) {
1362 list( $wikiUpperChars ) = $this->getCaseMaps();
1363 $x = $first ? '^' : '';
1364 return preg_replace_callback(
1365 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1366 array($this,"ucCallback"),
1367 $str
1368 );
1369 } else {
1370 return $first ? ucfirst( $str ) : strtoupper( $str );
1371 }
1372 }
1373 }
1374
1375 function lcfirst( $str ) {
1376 if ( empty($str) ) return $str;
1377 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1378 // editing string in place = cool
1379 $str[0]=strtolower($str[0]);
1380 return $str;
1381 }
1382 else return self::lc( $str, true );
1383 }
1384
1385 function lc( $str, $first = false ) {
1386 if ( function_exists( 'mb_strtolower' ) )
1387 if ( $first )
1388 if ( self::isMultibyte( $str ) )
1389 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1390 else
1391 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1392 else
1393 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1394 else
1395 if ( self::isMultibyte( $str ) ) {
1396 list( , $wikiLowerChars ) = self::getCaseMaps();
1397 $x = $first ? '^' : '';
1398 return preg_replace_callback(
1399 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1400 array($this,"lcCallback"),
1401 $str
1402 );
1403 } else
1404 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1405 }
1406
1407 function isMultibyte( $str ) {
1408 return (bool)preg_match( '/[\x80-\xff]/', $str );
1409 }
1410
1411 function ucwords($str) {
1412 if ( self::isMultibyte( $str ) ) {
1413 $str = self::lc($str);
1414
1415 // regexp to find first letter in each word (i.e. after each space)
1416 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1417
1418 // function to use to capitalize a single char
1419 if ( function_exists( 'mb_strtoupper' ) )
1420 return preg_replace_callback(
1421 $replaceRegexp,
1422 array($this,"ucwordsCallbackMB"),
1423 $str
1424 );
1425 else
1426 return preg_replace_callback(
1427 $replaceRegexp,
1428 array($this,"ucwordsCallbackWiki"),
1429 $str
1430 );
1431 }
1432 else
1433 return ucwords( strtolower( $str ) );
1434 }
1435
1436 # capitalize words at word breaks
1437 function ucwordbreaks($str){
1438 if (self::isMultibyte( $str ) ) {
1439 $str = self::lc($str);
1440
1441 // since \b doesn't work for UTF-8, we explicitely define word break chars
1442 $breaks= "[ \-\(\)\}\{\.,\?!]";
1443
1444 // find first letter after word break
1445 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1446
1447 if ( function_exists( 'mb_strtoupper' ) )
1448 return preg_replace_callback(
1449 $replaceRegexp,
1450 array($this,"ucwordbreaksCallbackMB"),
1451 $str
1452 );
1453 else
1454 return preg_replace_callback(
1455 $replaceRegexp,
1456 array($this,"ucwordsCallbackWiki"),
1457 $str
1458 );
1459 }
1460 else
1461 return preg_replace_callback(
1462 '/\b([\w\x80-\xff]+)\b/',
1463 array($this,"ucwordbreaksCallbackAscii"),
1464 $str );
1465 }
1466
1467 /**
1468 * Return a case-folded representation of $s
1469 *
1470 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1471 * and $s2 are the same except for the case of their characters. It is not
1472 * necessary for the value returned to make sense when displayed.
1473 *
1474 * Do *not* perform any other normalisation in this function. If a caller
1475 * uses this function when it should be using a more general normalisation
1476 * function, then fix the caller.
1477 */
1478 function caseFold( $s ) {
1479 return $this->uc( $s );
1480 }
1481
1482 function checkTitleEncoding( $s ) {
1483 if( is_array( $s ) ) {
1484 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1485 }
1486 # Check for non-UTF-8 URLs
1487 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1488 if(!$ishigh) return $s;
1489
1490 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1491 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1492 if( $isutf8 ) return $s;
1493
1494 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1495 }
1496
1497 function fallback8bitEncoding() {
1498 $this->load();
1499 return $this->fallback8bitEncoding;
1500 }
1501
1502 /**
1503 * Some languages have special punctuation to strip out
1504 * or characters which need to be converted for MySQL's
1505 * indexing to grok it correctly. Make such changes here.
1506 *
1507 * @param $string String
1508 * @return String
1509 */
1510 function stripForSearch( $string ) {
1511 global $wgDBtype;
1512 if ( $wgDBtype != 'mysql' ) {
1513 return $string;
1514 }
1515
1516 # MySQL fulltext index doesn't grok utf-8, so we
1517 # need to fold cases and convert to hex
1518
1519 wfProfileIn( __METHOD__ );
1520 if( function_exists( 'mb_strtolower' ) ) {
1521 $out = preg_replace(
1522 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1523 "'U8' . bin2hex( \"$1\" )",
1524 mb_strtolower( $string ) );
1525 } else {
1526 list( , $wikiLowerChars ) = self::getCaseMaps();
1527 $out = preg_replace(
1528 "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
1529 "'U8' . bin2hex( strtr( \"\$1\", \$wikiLowerChars ) )",
1530 $string );
1531 }
1532 wfProfileOut( __METHOD__ );
1533 return $out;
1534 }
1535
1536 function convertForSearchResult( $termsArray ) {
1537 # some languages, e.g. Chinese, need to do a conversion
1538 # in order for search results to be displayed correctly
1539 return $termsArray;
1540 }
1541
1542 /**
1543 * Get the first character of a string.
1544 *
1545 * @param $s string
1546 * @return string
1547 */
1548 function firstChar( $s ) {
1549 $matches = array();
1550 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1551 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1552
1553 if ( isset( $matches[1] ) ) {
1554 if ( strlen( $matches[1] ) != 3 ) {
1555 return $matches[1];
1556 }
1557
1558 // Break down Hangul syllables to grab the first jamo
1559 $code = utf8ToCodepoint( $matches[1] );
1560 if ( $code < 0xac00 || 0xd7a4 <= $code) {
1561 return $matches[1];
1562 } elseif ( $code < 0xb098 ) {
1563 return "\xe3\x84\xb1";
1564 } elseif ( $code < 0xb2e4 ) {
1565 return "\xe3\x84\xb4";
1566 } elseif ( $code < 0xb77c ) {
1567 return "\xe3\x84\xb7";
1568 } elseif ( $code < 0xb9c8 ) {
1569 return "\xe3\x84\xb9";
1570 } elseif ( $code < 0xbc14 ) {
1571 return "\xe3\x85\x81";
1572 } elseif ( $code < 0xc0ac ) {
1573 return "\xe3\x85\x82";
1574 } elseif ( $code < 0xc544 ) {
1575 return "\xe3\x85\x85";
1576 } elseif ( $code < 0xc790 ) {
1577 return "\xe3\x85\x87";
1578 } elseif ( $code < 0xcc28 ) {
1579 return "\xe3\x85\x88";
1580 } elseif ( $code < 0xce74 ) {
1581 return "\xe3\x85\x8a";
1582 } elseif ( $code < 0xd0c0 ) {
1583 return "\xe3\x85\x8b";
1584 } elseif ( $code < 0xd30c ) {
1585 return "\xe3\x85\x8c";
1586 } elseif ( $code < 0xd558 ) {
1587 return "\xe3\x85\x8d";
1588 } else {
1589 return "\xe3\x85\x8e";
1590 }
1591 } else {
1592 return "";
1593 }
1594 }
1595
1596 function initEncoding() {
1597 # Some languages may have an alternate char encoding option
1598 # (Esperanto X-coding, Japanese furigana conversion, etc)
1599 # If this language is used as the primary content language,
1600 # an override to the defaults can be set here on startup.
1601 }
1602
1603 function recodeForEdit( $s ) {
1604 # For some languages we'll want to explicitly specify
1605 # which characters make it into the edit box raw
1606 # or are converted in some way or another.
1607 # Note that if wgOutputEncoding is different from
1608 # wgInputEncoding, this text will be further converted
1609 # to wgOutputEncoding.
1610 global $wgEditEncoding;
1611 if( $wgEditEncoding == '' or
1612 $wgEditEncoding == 'UTF-8' ) {
1613 return $s;
1614 } else {
1615 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1616 }
1617 }
1618
1619 function recodeInput( $s ) {
1620 # Take the previous into account.
1621 global $wgEditEncoding;
1622 if($wgEditEncoding != "") {
1623 $enc = $wgEditEncoding;
1624 } else {
1625 $enc = 'UTF-8';
1626 }
1627 if( $enc == 'UTF-8' ) {
1628 return $s;
1629 } else {
1630 return $this->iconv( $enc, 'UTF-8', $s );
1631 }
1632 }
1633
1634 /**
1635 * For right-to-left language support
1636 *
1637 * @return bool
1638 */
1639 function isRTL() {
1640 $this->load();
1641 return $this->rtl;
1642 }
1643
1644 /**
1645 * A hidden direction mark (LRM or RLM), depending on the language direction
1646 *
1647 * @return string
1648 */
1649 function getDirMark() {
1650 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1651 }
1652
1653 /**
1654 * An arrow, depending on the language direction
1655 *
1656 * @return string
1657 */
1658 function getArrow() {
1659 return $this->isRTL() ? '←' : '→';
1660 }
1661
1662 /**
1663 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1664 *
1665 * @return bool
1666 */
1667 function linkPrefixExtension() {
1668 $this->load();
1669 return $this->linkPrefixExtension;
1670 }
1671
1672 function &getMagicWords() {
1673 $this->load();
1674 return $this->magicWords;
1675 }
1676
1677 # Fill a MagicWord object with data from here
1678 function getMagic( &$mw ) {
1679 if ( !$this->mMagicHookDone ) {
1680 $this->mMagicHookDone = true;
1681 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1682 }
1683 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1684 $rawEntry = $this->mMagicExtensions[$mw->mId];
1685 } else {
1686 $magicWords =& $this->getMagicWords();
1687 if ( isset( $magicWords[$mw->mId] ) ) {
1688 $rawEntry = $magicWords[$mw->mId];
1689 } else {
1690 # Fall back to English if local list is incomplete
1691 $magicWords =& Language::getMagicWords();
1692 if ( !isset($magicWords[$mw->mId]) ) {
1693 throw new MWException("Magic word '{$mw->mId}' not found" );
1694 }
1695 $rawEntry = $magicWords[$mw->mId];
1696 }
1697 }
1698
1699 if( !is_array( $rawEntry ) ) {
1700 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1701 } else {
1702 $mw->mCaseSensitive = $rawEntry[0];
1703 $mw->mSynonyms = array_slice( $rawEntry, 1 );
1704 }
1705 }
1706
1707 /**
1708 * Add magic words to the extension array
1709 */
1710 function addMagicWordsByLang( $newWords ) {
1711 $code = $this->getCode();
1712 $fallbackChain = array();
1713 while ( $code && !in_array( $code, $fallbackChain ) ) {
1714 $fallbackChain[] = $code;
1715 $code = self::getFallbackFor( $code );
1716 }
1717 if ( !in_array( 'en', $fallbackChain ) ) {
1718 $fallbackChain[] = 'en';
1719 }
1720 $fallbackChain = array_reverse( $fallbackChain );
1721 foreach ( $fallbackChain as $code ) {
1722 if ( isset( $newWords[$code] ) ) {
1723 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1724 }
1725 }
1726 }
1727
1728 /**
1729 * Get special page names, as an associative array
1730 * case folded alias => real name
1731 */
1732 function getSpecialPageAliases() {
1733 $this->load();
1734
1735 // Cache aliases because it may be slow to load them
1736 if ( !isset( $this->mExtendedSpecialPageAliases ) ) {
1737
1738 // Initialise array
1739 $this->mExtendedSpecialPageAliases = $this->specialPageAliases;
1740
1741 global $wgExtensionAliasesFiles;
1742 foreach ( $wgExtensionAliasesFiles as $file ) {
1743
1744 // Fail fast
1745 if ( !file_exists($file) )
1746 throw new MWException( "Aliases file does not exist: $file" );
1747
1748 $aliases = array();
1749 require($file);
1750
1751 // Check the availability of aliases
1752 if ( !isset($aliases['en']) )
1753 throw new MWException( "Malformed aliases file: $file" );
1754
1755 // Merge all aliases in fallback chain
1756 $code = $this->getCode();
1757 do {
1758 if ( !isset($aliases[$code]) ) continue;
1759
1760 $aliases[$code] = $this->fixSpecialPageAliases( $aliases[$code] );
1761 /* Merge the aliases, THIS will break if there is special page name
1762 * which looks like a numerical key, thanks to PHP...
1763 * See the comments for wfArrayMerge in GlobalSettings.php. */
1764 $this->mExtendedSpecialPageAliases = array_merge_recursive(
1765 $this->mExtendedSpecialPageAliases, $aliases[$code] );
1766
1767 } while ( $code = self::getFallbackFor( $code ) );
1768 }
1769
1770 wfRunHooks( 'LanguageGetSpecialPageAliases',
1771 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1772 }
1773
1774 return $this->mExtendedSpecialPageAliases;
1775 }
1776
1777 /**
1778 * Function to fix special page aliases. Will convert the first letter to
1779 * upper case and spaces to underscores. Can be given a full aliases array,
1780 * in which case it will recursively fix all aliases.
1781 */
1782 public function fixSpecialPageAliases( $mixed ) {
1783 // Work recursively until in string level
1784 if ( is_array($mixed) ) {
1785 $callback = array( $this, 'fixSpecialPageAliases' );
1786 return array_map( $callback, $mixed );
1787 }
1788 return str_replace( ' ', '_', $this->ucfirst( $mixed ) );
1789 }
1790
1791 /**
1792 * Italic is unsuitable for some languages
1793 *
1794 * @param $text String: the text to be emphasized.
1795 * @return string
1796 */
1797 function emphasize( $text ) {
1798 return "<em>$text</em>";
1799 }
1800
1801 /**
1802 * Normally we output all numbers in plain en_US style, that is
1803 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1804 * point twohundredthirtyfive. However this is not sutable for all
1805 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1806 * Icelandic just want to use commas instead of dots, and dots instead
1807 * of commas like "293.291,235".
1808 *
1809 * An example of this function being called:
1810 * <code>
1811 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1812 * </code>
1813 *
1814 * See LanguageGu.php for the Gujarati implementation and
1815 * LanguageIs.php for the , => . and . => , implementation.
1816 *
1817 * @todo check if it's viable to use localeconv() for the decimal
1818 * seperator thing.
1819 * @param $number Mixed: the string to be formatted, should be an integer
1820 * or a floating point number.
1821 * @param $nocommafy Bool: set to true for special numbers like dates
1822 * @return string
1823 */
1824 function formatNum( $number, $nocommafy = false ) {
1825 global $wgTranslateNumerals;
1826 if (!$nocommafy) {
1827 $number = $this->commafy($number);
1828 $s = $this->separatorTransformTable();
1829 if (!is_null($s)) { $number = strtr($number, $s); }
1830 }
1831
1832 if ($wgTranslateNumerals) {
1833 $s = $this->digitTransformTable();
1834 if (!is_null($s)) { $number = strtr($number, $s); }
1835 }
1836
1837 return $number;
1838 }
1839
1840 function parseFormattedNumber( $number ) {
1841 $s = $this->digitTransformTable();
1842 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1843
1844 $s = $this->separatorTransformTable();
1845 if (!is_null($s)) { $number = strtr($number, array_flip($s)); }
1846
1847 $number = strtr( $number, array (',' => '') );
1848 return $number;
1849 }
1850
1851 /**
1852 * Adds commas to a given number
1853 *
1854 * @param $_ mixed
1855 * @return string
1856 */
1857 function commafy($_) {
1858 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1859 }
1860
1861 function digitTransformTable() {
1862 $this->load();
1863 return $this->digitTransformTable;
1864 }
1865
1866 function separatorTransformTable() {
1867 $this->load();
1868 return $this->separatorTransformTable;
1869 }
1870
1871
1872 /**
1873 * For the credit list in includes/Credits.php (action=credits)
1874 *
1875 * @param $l Array
1876 * @return string
1877 */
1878 function listToText( $l ) {
1879 $s = '';
1880 $m = count($l) - 1;
1881 for ($i = $m; $i >= 0; $i--) {
1882 if ($i == $m) {
1883 $s = $l[$i];
1884 } else if ($i == $m - 1) {
1885 $s = $l[$i] . ' ' . $this->getMessageFromDB( 'and' ) . ' ' . $s;
1886 } else {
1887 $s = $l[$i] . ', ' . $s;
1888 }
1889 }
1890 return $s;
1891 }
1892
1893 /**
1894 * Truncate a string to a specified length in bytes, appending an optional
1895 * string (e.g. for ellipses)
1896 *
1897 * The database offers limited byte lengths for some columns in the database;
1898 * multi-byte character sets mean we need to ensure that only whole characters
1899 * are included, otherwise broken characters can be passed to the user
1900 *
1901 * If $length is negative, the string will be truncated from the beginning
1902 *
1903 * @param $string String to truncate
1904 * @param $length Int: maximum length (excluding ellipses)
1905 * @param $ellipsis String to append to the truncated text
1906 * @return string
1907 */
1908 function truncate( $string, $length, $ellipsis = "" ) {
1909 if( $length == 0 ) {
1910 return $ellipsis;
1911 }
1912 if ( strlen( $string ) <= abs( $length ) ) {
1913 return $string;
1914 }
1915 if( $length > 0 ) {
1916 $string = substr( $string, 0, $length );
1917 $char = ord( $string[strlen( $string ) - 1] );
1918 $m = array();
1919 if ($char >= 0xc0) {
1920 # We got the first byte only of a multibyte char; remove it.
1921 $string = substr( $string, 0, -1 );
1922 } elseif( $char >= 0x80 &&
1923 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
1924 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
1925 # We chopped in the middle of a character; remove it
1926 $string = $m[1];
1927 }
1928 return $string . $ellipsis;
1929 } else {
1930 $string = substr( $string, $length );
1931 $char = ord( $string[0] );
1932 if( $char >= 0x80 && $char < 0xc0 ) {
1933 # We chopped in the middle of a character; remove the whole thing
1934 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
1935 }
1936 return $ellipsis . $string;
1937 }
1938 }
1939
1940 /**
1941 * Grammatical transformations, needed for inflected languages
1942 * Invoked by putting {{grammar:case|word}} in a message
1943 *
1944 * @param $word string
1945 * @param $case string
1946 * @return string
1947 */
1948 function convertGrammar( $word, $case ) {
1949 global $wgGrammarForms;
1950 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
1951 return $wgGrammarForms[$this->getCode()][$case][$word];
1952 }
1953 return $word;
1954 }
1955
1956 /**
1957 * Plural form transformations, needed for some languages.
1958 * For example, there are 3 form of plural in Russian and Polish,
1959 * depending on "count mod 10". See [[w:Plural]]
1960 * For English it is pretty simple.
1961 *
1962 * Invoked by putting {{plural:count|wordform1|wordform2}}
1963 * or {{plural:count|wordform1|wordform2|wordform3}}
1964 *
1965 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
1966 *
1967 * @param $count Integer: non-localized number
1968 * @param $forms Array: different plural forms
1969 * @return string Correct form of plural for $count in this language
1970 */
1971 function convertPlural( $count, $forms ) {
1972 if ( !count($forms) ) { return ''; }
1973 $forms = $this->preConvertPlural( $forms, 2 );
1974
1975 return ( abs($count) == 1 ) ? $forms[0] : $forms[1];
1976 }
1977
1978 /**
1979 * Checks that convertPlural was given an array and pads it to requested
1980 * amound of forms by copying the last one.
1981 *
1982 * @param $count Integer: How many forms should there be at least
1983 * @param $forms Array of forms given to convertPlural
1984 * @return array Padded array of forms or an exception if not an array
1985 */
1986 protected function preConvertPlural( /* Array */ $forms, $count ) {
1987 while ( count($forms) < $count ) {
1988 $forms[] = $forms[count($forms)-1];
1989 }
1990 return $forms;
1991 }
1992
1993 /**
1994 * For translaing of expiry times
1995 * @param $str String: the validated block time in English
1996 * @return Somehow translated block time
1997 * @see LanguageFi.php for example implementation
1998 */
1999 function translateBlockExpiry( $str ) {
2000
2001 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2002
2003 if ( $scBlockExpiryOptions == '-') {
2004 return $str;
2005 }
2006
2007 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2008 if ( strpos($option, ":") === false )
2009 continue;
2010 list($show, $value) = explode(":", $option);
2011 if ( strcmp ( $str, $value) == 0 ) {
2012 return htmlspecialchars( trim( $show ) );
2013 }
2014 }
2015
2016 return $str;
2017 }
2018
2019 /**
2020 * languages like Chinese need to be segmented in order for the diff
2021 * to be of any use
2022 *
2023 * @param $text String
2024 * @return String
2025 */
2026 function segmentForDiff( $text ) {
2027 return $text;
2028 }
2029
2030 /**
2031 * and unsegment to show the result
2032 *
2033 * @param $text String
2034 * @return String
2035 */
2036 function unsegmentForDiff( $text ) {
2037 return $text;
2038 }
2039
2040 # convert text to different variants of a language.
2041 function convert( $text, $isTitle = false) {
2042 return $this->mConverter->convert($text, $isTitle);
2043 }
2044
2045 # Convert text from within Parser
2046 function parserConvert( $text, &$parser ) {
2047 return $this->mConverter->parserConvert( $text, $parser );
2048 }
2049
2050 # Check if this is a language with variants
2051 function hasVariants(){
2052 return sizeof($this->getVariants())>1;
2053 }
2054
2055 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2056 function armourMath($text){
2057 return $this->mConverter->armourMath($text);
2058 }
2059
2060
2061 /**
2062 * Perform output conversion on a string, and encode for safe HTML output.
2063 * @param $text String
2064 * @param $isTitle Bool -- wtf?
2065 * @return string
2066 * @todo this should get integrated somewhere sane
2067 */
2068 function convertHtml( $text, $isTitle = false ) {
2069 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2070 }
2071
2072 function convertCategoryKey( $key ) {
2073 return $this->mConverter->convertCategoryKey( $key );
2074 }
2075
2076 /**
2077 * get the list of variants supported by this langauge
2078 * see sample implementation in LanguageZh.php
2079 *
2080 * @return array an array of language codes
2081 */
2082 function getVariants() {
2083 return $this->mConverter->getVariants();
2084 }
2085
2086
2087 function getPreferredVariant( $fromUser = true ) {
2088 return $this->mConverter->getPreferredVariant( $fromUser );
2089 }
2090
2091 /**
2092 * if a language supports multiple variants, it is
2093 * possible that non-existing link in one variant
2094 * actually exists in another variant. this function
2095 * tries to find it. See e.g. LanguageZh.php
2096 *
2097 * @param $link String: the name of the link
2098 * @param $nt Mixed: the title object of the link
2099 * @return null the input parameters may be modified upon return
2100 */
2101 function findVariantLink( &$link, &$nt, $forTemplate = false ) {
2102 $this->mConverter->findVariantLink($link, $nt, $forTemplate );
2103 }
2104
2105 /**
2106 * If a language supports multiple variants, converts text
2107 * into an array of all possible variants of the text:
2108 * 'variant' => text in that variant
2109 */
2110
2111 function convertLinkToAllVariants($text){
2112 return $this->mConverter->convertLinkToAllVariants($text);
2113 }
2114
2115
2116 /**
2117 * returns language specific options used by User::getPageRenderHash()
2118 * for example, the preferred language variant
2119 *
2120 * @return string
2121 */
2122 function getExtraHashOptions() {
2123 return $this->mConverter->getExtraHashOptions();
2124 }
2125
2126 /**
2127 * for languages that support multiple variants, the title of an
2128 * article may be displayed differently in different variants. this
2129 * function returns the apporiate title defined in the body of the article.
2130 *
2131 * @return string
2132 */
2133 function getParsedTitle() {
2134 return $this->mConverter->getParsedTitle();
2135 }
2136
2137 /**
2138 * Enclose a string with the "no conversion" tag. This is used by
2139 * various functions in the Parser
2140 *
2141 * @param $text String: text to be tagged for no conversion
2142 * @param $noParse
2143 * @return string the tagged text
2144 */
2145 function markNoConversion( $text, $noParse=false ) {
2146 return $this->mConverter->markNoConversion( $text, $noParse );
2147 }
2148
2149 /**
2150 * A regular expression to match legal word-trailing characters
2151 * which should be merged onto a link of the form [[foo]]bar.
2152 *
2153 * @return string
2154 */
2155 function linkTrail() {
2156 $this->load();
2157 return $this->linkTrail;
2158 }
2159
2160 function getLangObj() {
2161 return $this;
2162 }
2163
2164 /**
2165 * Get the RFC 3066 code for this language object
2166 */
2167 function getCode() {
2168 return $this->mCode;
2169 }
2170
2171 function setCode( $code ) {
2172 $this->mCode = $code;
2173 }
2174
2175 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2176 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2177 }
2178
2179 static function getMessagesFileName( $code ) {
2180 global $IP;
2181 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2182 }
2183
2184 static function getClassFileName( $code ) {
2185 global $IP;
2186 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2187 }
2188
2189 static function getLocalisationArray( $code, $disableCache = false ) {
2190 self::loadLocalisation( $code, $disableCache );
2191 return self::$mLocalisationCache[$code];
2192 }
2193
2194 /**
2195 * Load localisation data for a given code into the static cache
2196 *
2197 * @return array Dependencies, map of filenames to mtimes
2198 */
2199 static function loadLocalisation( $code, $disableCache = false ) {
2200 static $recursionGuard = array();
2201 global $wgMemc, $wgCheckSerialized;
2202
2203 if ( !$code ) {
2204 throw new MWException( "Invalid language code requested" );
2205 }
2206
2207 if ( !$disableCache ) {
2208 # Try the per-process cache
2209 if ( isset( self::$mLocalisationCache[$code] ) ) {
2210 return self::$mLocalisationCache[$code]['deps'];
2211 }
2212
2213 wfProfileIn( __METHOD__ );
2214
2215 # Try the serialized directory
2216 $cache = wfGetPrecompiledData( self::getFileName( "Messages", $code, '.ser' ) );
2217 if ( $cache ) {
2218 if ( $wgCheckSerialized && self::isLocalisationOutOfDate( $cache ) ) {
2219 $cache = false;
2220 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2221 } else {
2222 self::$mLocalisationCache[$code] = $cache;
2223 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2224 wfProfileOut( __METHOD__ );
2225 return self::$mLocalisationCache[$code]['deps'];
2226 }
2227 }
2228
2229 # Try the global cache
2230 $memcKey = wfMemcKey('localisation', $code );
2231 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2232 $cache = $wgMemc->get( $memcKey );
2233 if ( $cache ) {
2234 if ( self::isLocalisationOutOfDate( $cache ) ) {
2235 $wgMemc->delete( $memcKey );
2236 $wgMemc->delete( $fbMemcKey );
2237 $cache = false;
2238 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2239 } else {
2240 self::$mLocalisationCache[$code] = $cache;
2241 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2242 wfProfileOut( __METHOD__ );
2243 return $cache['deps'];
2244 }
2245 }
2246 } else {
2247 wfProfileIn( __METHOD__ );
2248 }
2249
2250 # Default fallback, may be overridden when the messages file is included
2251 if ( $code != 'en' ) {
2252 $fallback = 'en';
2253 } else {
2254 $fallback = false;
2255 }
2256
2257 # Load the primary localisation from the source file
2258 $filename = self::getMessagesFileName( $code );
2259 if ( !file_exists( $filename ) ) {
2260 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2261 $cache = compact( self::$mLocalisationKeys ); // Set correct fallback
2262 $deps = array();
2263 } else {
2264 $deps = array( $filename => filemtime( $filename ) );
2265 require( $filename );
2266 $cache = compact( self::$mLocalisationKeys );
2267 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2268 }
2269
2270 if ( !empty( $fallback ) ) {
2271 # Load the fallback localisation, with a circular reference guard
2272 if ( isset( $recursionGuard[$code] ) ) {
2273 throw new MWException( "Error: Circular fallback reference in language code $code" );
2274 }
2275 $recursionGuard[$code] = true;
2276 $newDeps = self::loadLocalisation( $fallback, $disableCache );
2277 unset( $recursionGuard[$code] );
2278
2279 $secondary = self::$mLocalisationCache[$fallback];
2280 $deps = array_merge( $deps, $newDeps );
2281
2282 # Merge the fallback localisation with the current localisation
2283 foreach ( self::$mLocalisationKeys as $key ) {
2284 if ( isset( $cache[$key] ) ) {
2285 if ( isset( $secondary[$key] ) ) {
2286 if ( in_array( $key, self::$mMergeableMapKeys ) ) {
2287 $cache[$key] = $cache[$key] + $secondary[$key];
2288 } elseif ( in_array( $key, self::$mMergeableListKeys ) ) {
2289 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2290 } elseif ( in_array( $key, self::$mMergeableAliasListKeys ) ) {
2291 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2292 }
2293 }
2294 } else {
2295 $cache[$key] = $secondary[$key];
2296 }
2297 }
2298
2299 # Merge bookstore lists if requested
2300 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2301 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2302 }
2303 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2304 unset( $cache['bookstoreList']['inherit'] );
2305 }
2306 }
2307
2308 # Add dependencies to the cache entry
2309 $cache['deps'] = $deps;
2310
2311 # Replace spaces with underscores in namespace names
2312 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2313
2314 # And do the same for specialpage aliases. $page is an array.
2315 foreach ( $cache['specialPageAliases'] as &$page ) {
2316 $page = str_replace( ' ', '_', $page );
2317 }
2318 # Decouple the reference to prevent accidental damage
2319 unset($page);
2320
2321 # Save to both caches
2322 self::$mLocalisationCache[$code] = $cache;
2323 if ( !$disableCache ) {
2324 $wgMemc->set( $memcKey, $cache );
2325 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2326 }
2327
2328 wfProfileOut( __METHOD__ );
2329 return $deps;
2330 }
2331
2332 /**
2333 * Test if a given localisation cache is out of date with respect to the
2334 * source Messages files. This is done automatically for the global cache
2335 * in $wgMemc, but is only done on certain occasions for the serialized
2336 * data file.
2337 *
2338 * @param $cache mixed Either a language code or a cache array
2339 */
2340 static function isLocalisationOutOfDate( $cache ) {
2341 if ( !is_array( $cache ) ) {
2342 self::loadLocalisation( $cache );
2343 $cache = self::$mLocalisationCache[$cache];
2344 }
2345 $expired = false;
2346 foreach ( $cache['deps'] as $file => $mtime ) {
2347 if ( !file_exists( $file ) || filemtime( $file ) > $mtime ) {
2348 $expired = true;
2349 break;
2350 }
2351 }
2352 return $expired;
2353 }
2354
2355 /**
2356 * Get the fallback for a given language
2357 */
2358 static function getFallbackFor( $code ) {
2359 // Shortcut
2360 if ( $code === 'en' ) return false;
2361
2362 // Local cache
2363 static $cache = array();
2364 // Quick return
2365 if ( isset($cache[$code]) ) return $cache[$code];
2366
2367 // Try memcache
2368 global $wgMemc;
2369 $memcKey = wfMemcKey( 'fallback', $code );
2370 $fbcode = $wgMemc->get( $memcKey );
2371
2372 if ( is_string($fbcode) ) {
2373 // False is stored as a string to detect failures in memcache properly
2374 if ( $fbcode === '' ) $fbcode = false;
2375
2376 // Update local cache and return
2377 $cache[$code] = $fbcode;
2378 return $fbcode;
2379 }
2380
2381 // Nothing in caches, load and and update both caches
2382 self::loadLocalisation( $code );
2383 $fbcode = self::$mLocalisationCache[$code]['fallback'];
2384
2385 $cache[$code] = $fbcode;
2386 $wgMemc->set( $memcKey, (string) $fbcode );
2387
2388 return $fbcode;
2389 }
2390
2391 /**
2392 * Get all messages for a given language
2393 */
2394 static function getMessagesFor( $code ) {
2395 self::loadLocalisation( $code );
2396 return self::$mLocalisationCache[$code]['messages'];
2397 }
2398
2399 /**
2400 * Get a message for a given language
2401 */
2402 static function getMessageFor( $key, $code ) {
2403 self::loadLocalisation( $code );
2404 return isset( self::$mLocalisationCache[$code]['messages'][$key] ) ? self::$mLocalisationCache[$code]['messages'][$key] : null;
2405 }
2406
2407 /**
2408 * Load localisation data for this object
2409 */
2410 function load() {
2411 if ( !$this->mLoaded ) {
2412 self::loadLocalisation( $this->getCode() );
2413 $cache =& self::$mLocalisationCache[$this->getCode()];
2414 foreach ( self::$mLocalisationKeys as $key ) {
2415 $this->$key = $cache[$key];
2416 }
2417 $this->mLoaded = true;
2418
2419 $this->fixUpSettings();
2420 }
2421 }
2422
2423 /**
2424 * Do any necessary post-cache-load settings adjustment
2425 */
2426 function fixUpSettings() {
2427 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2428 $wgNamespaceAliases, $wgAmericanDates;
2429 wfProfileIn( __METHOD__ );
2430 if ( $wgExtraNamespaces ) {
2431 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
2432 }
2433
2434 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
2435 if ( $wgMetaNamespaceTalk ) {
2436 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
2437 } else {
2438 $talk = $this->namespaceNames[NS_PROJECT_TALK];
2439 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2440
2441 # Allow grammar transformations
2442 # Allowing full message-style parsing would make simple requests
2443 # such as action=raw much more expensive than they need to be.
2444 # This will hopefully cover most cases.
2445 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2446 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2447 $talk = str_replace( ' ', '_', $talk );
2448 $this->namespaceNames[NS_PROJECT_TALK] = $talk;
2449 }
2450
2451 # The above mixing may leave namespaces out of canonical order.
2452 # Re-order by namespace ID number...
2453 ksort( $this->namespaceNames );
2454
2455 # Put namespace names and aliases into a hashtable.
2456 # If this is too slow, then we should arrange it so that it is done
2457 # before caching. The catch is that at pre-cache time, the above
2458 # class-specific fixup hasn't been done.
2459 $this->mNamespaceIds = array();
2460 foreach ( $this->namespaceNames as $index => $name ) {
2461 $this->mNamespaceIds[$this->lc($name)] = $index;
2462 }
2463 if ( $this->namespaceAliases ) {
2464 foreach ( $this->namespaceAliases as $name => $index ) {
2465 $this->mNamespaceIds[$this->lc($name)] = $index;
2466 }
2467 }
2468 if ( $wgNamespaceAliases ) {
2469 foreach ( $wgNamespaceAliases as $name => $index ) {
2470 $this->mNamespaceIds[$this->lc($name)] = $index;
2471 }
2472 }
2473
2474 if ( $this->defaultDateFormat == 'dmy or mdy' ) {
2475 $this->defaultDateFormat = $wgAmericanDates ? 'mdy' : 'dmy';
2476 }
2477 wfProfileOut( __METHOD__ );
2478 }
2479
2480 function replaceGrammarInNamespace( $m ) {
2481 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2482 }
2483
2484 static function getCaseMaps() {
2485 static $wikiUpperChars, $wikiLowerChars;
2486 if ( isset( $wikiUpperChars ) ) {
2487 return array( $wikiUpperChars, $wikiLowerChars );
2488 }
2489
2490 wfProfileIn( __METHOD__ );
2491 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2492 if ( $arr === false ) {
2493 throw new MWException(
2494 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2495 }
2496 extract( $arr );
2497 wfProfileOut( __METHOD__ );
2498 return array( $wikiUpperChars, $wikiLowerChars );
2499 }
2500
2501 function formatTimePeriod( $seconds ) {
2502 if ( $seconds < 10 ) {
2503 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2504 } elseif ( $seconds < 60 ) {
2505 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2506 } elseif ( $seconds < 3600 ) {
2507 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2508 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2509 } else {
2510 $hours = floor( $seconds / 3600 );
2511 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2512 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2513 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2514 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2515 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2516 }
2517 }
2518
2519 function formatBitrate( $bps ) {
2520 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2521 if ( $bps <= 0 ) {
2522 return $this->formatNum( $bps ) . $units[0];
2523 }
2524 $unitIndex = floor( log10( $bps ) / 3 );
2525 $mantissa = $bps / pow( 1000, $unitIndex );
2526 if ( $mantissa < 10 ) {
2527 $mantissa = round( $mantissa, 1 );
2528 } else {
2529 $mantissa = round( $mantissa );
2530 }
2531 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2532 }
2533
2534 /**
2535 * Format a size in bytes for output, using an appropriate
2536 * unit (B, KB, MB or GB) according to the magnitude in question
2537 *
2538 * @param $size Size to format
2539 * @return string Plain text (not HTML)
2540 */
2541 function formatSize( $size ) {
2542 // For small sizes no decimal places necessary
2543 $round = 0;
2544 if( $size > 1024 ) {
2545 $size = $size / 1024;
2546 if( $size > 1024 ) {
2547 $size = $size / 1024;
2548 // For MB and bigger two decimal places are smarter
2549 $round = 2;
2550 if( $size > 1024 ) {
2551 $size = $size / 1024;
2552 $msg = 'size-gigabytes';
2553 } else {
2554 $msg = 'size-megabytes';
2555 }
2556 } else {
2557 $msg = 'size-kilobytes';
2558 }
2559 } else {
2560 $msg = 'size-bytes';
2561 }
2562 $size = round( $size, $round );
2563 $text = $this->getMessageFromDB( $msg );
2564 return str_replace( '$1', $this->formatNum( $size ), $text );
2565 }
2566 }