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