* Bug 16484 Remove an unneeded extra comma in user rights log (and other lists)
[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(
60 'fallback', 'namespaceNames', 'mathNames', 'bookstoreList',
61 'magicWords', 'messages', 'rtl', 'digitTransformTable',
62 'separatorTransformTable', 'fallback8bitEncoding', 'linkPrefixExtension',
63 'defaultUserOptionOverrides', 'linkTrail', 'namespaceAliases',
64 'dateFormats', 'datePreferences', 'datePreferenceMigrationMap',
65 'defaultDateFormat', 'extraUserToggles', 'specialPageAliases',
66 'imageFiles'
67 );
68
69 static public $mMergeableMapKeys = array( 'messages', 'namespaceNames', 'mathNames',
70 'dateFormats', 'defaultUserOptionOverrides', 'magicWords', 'imageFiles' );
71
72 static public $mMergeableListKeys = array( 'extraUserToggles' );
73
74 static public $mMergeableAliasListKeys = array( 'specialPageAliases' );
75
76 static public $mLocalisationCache = array();
77 static public $mLangObjCache = array();
78
79 static public $mWeekdayMsgs = array(
80 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
81 'friday', 'saturday'
82 );
83
84 static public $mWeekdayAbbrevMsgs = array(
85 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
86 );
87
88 static public $mMonthMsgs = array(
89 'january', 'february', 'march', 'april', 'may_long', 'june',
90 'july', 'august', 'september', 'october', 'november',
91 'december'
92 );
93 static public $mMonthGenMsgs = array(
94 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
95 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
96 'december-gen'
97 );
98 static public $mMonthAbbrevMsgs = array(
99 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
100 'sep', 'oct', 'nov', 'dec'
101 );
102
103 static public $mIranianCalendarMonthMsgs = array(
104 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
105 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
106 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
107 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
108 );
109
110 static public $mHebrewCalendarMonthMsgs = array(
111 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
112 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
113 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
114 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
115 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
116 );
117
118 static public $mHebrewCalendarMonthGenMsgs = array(
119 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
120 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
121 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
122 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
123 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
124 );
125
126 static public $mHijriCalendarMonthMsgs = array(
127 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
128 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
129 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
130 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
131 );
132
133 /**
134 * Get a cached language object for a given language code
135 */
136 static function factory( $code ) {
137 if ( !isset( self::$mLangObjCache[$code] ) ) {
138 if( count( self::$mLangObjCache ) > 10 ) {
139 // Don't keep a billion objects around, that's stupid.
140 self::$mLangObjCache = array();
141 }
142 self::$mLangObjCache[$code] = self::newFromCode( $code );
143 }
144 return self::$mLangObjCache[$code];
145 }
146
147 /**
148 * Create a language object for a given language code
149 */
150 protected static function newFromCode( $code ) {
151 global $IP;
152 static $recursionLevel = 0;
153 if ( $code == 'en' ) {
154 $class = 'Language';
155 } else {
156 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
157 // Preload base classes to work around APC/PHP5 bug
158 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
159 include_once("$IP/languages/classes/$class.deps.php");
160 }
161 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
162 include_once("$IP/languages/classes/$class.php");
163 }
164 }
165
166 if ( $recursionLevel > 5 ) {
167 throw new MWException( "Language fallback loop detected when creating class $class\n" );
168 }
169
170 if( ! class_exists( $class ) ) {
171 $fallback = Language::getFallbackFor( $code );
172 ++$recursionLevel;
173 $lang = Language::newFromCode( $fallback );
174 --$recursionLevel;
175 $lang->setCode( $code );
176 } else {
177 $lang = new $class;
178 }
179 return $lang;
180 }
181
182 function __construct() {
183 $this->mConverter = new FakeConverter($this);
184 // Set the code to the name of the descendant
185 if ( get_class( $this ) == 'Language' ) {
186 $this->mCode = 'en';
187 } else {
188 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
189 }
190 }
191
192 /**
193 * Reduce memory usage
194 */
195 function __destruct() {
196 foreach ( $this as $name => $value ) {
197 unset( $this->$name );
198 }
199 }
200
201 /**
202 * Hook which will be called if this is the content language.
203 * Descendants can use this to register hook functions or modify globals
204 */
205 function initContLang() {}
206
207 /**
208 * @deprecated Use User::getDefaultOptions()
209 * @return array
210 */
211 function getDefaultUserOptions() {
212 wfDeprecated( __METHOD__ );
213 return User::getDefaultOptions();
214 }
215
216 function getFallbackLanguageCode() {
217 return self::getFallbackFor( $this->mCode );
218 }
219
220 /**
221 * Exports $wgBookstoreListEn
222 * @return array
223 */
224 function getBookstoreList() {
225 $this->load();
226 return $this->bookstoreList;
227 }
228
229 /**
230 * @return array
231 */
232 function getNamespaces() {
233 $this->load();
234 return $this->namespaceNames;
235 }
236
237 /**
238 * A convenience function that returns the same thing as
239 * getNamespaces() except with the array values changed to ' '
240 * where it found '_', useful for producing output to be displayed
241 * e.g. in <select> forms.
242 *
243 * @return array
244 */
245 function getFormattedNamespaces() {
246 $ns = $this->getNamespaces();
247 foreach($ns as $k => $v) {
248 $ns[$k] = strtr($v, '_', ' ');
249 }
250 return $ns;
251 }
252
253 /**
254 * Get a namespace value by key
255 * <code>
256 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
257 * echo $mw_ns; // prints 'MediaWiki'
258 * </code>
259 *
260 * @param $index Int: the array key of the namespace to return
261 * @return mixed, string if the namespace value exists, otherwise false
262 */
263 function getNsText( $index ) {
264 $ns = $this->getNamespaces();
265 return isset( $ns[$index] ) ? $ns[$index] : false;
266 }
267
268 /**
269 * A convenience function that returns the same thing as
270 * getNsText() except with '_' changed to ' ', useful for
271 * producing output.
272 *
273 * @return array
274 */
275 function getFormattedNsText( $index ) {
276 $ns = $this->getNsText( $index );
277 return strtr($ns, '_', ' ');
278 }
279
280 /**
281 * Get a namespace key by value, case insensitive.
282 * Only matches namespace names for the current language, not the
283 * canonical ones defined in Namespace.php.
284 *
285 * @param $text String
286 * @return mixed An integer if $text is a valid value otherwise false
287 */
288 function getLocalNsIndex( $text ) {
289 $this->load();
290 $lctext = $this->lc($text);
291 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
292 }
293
294 /**
295 * Get a namespace key by value, case insensitive. Canonical namespace
296 * names override custom ones defined for the current language.
297 *
298 * @param $text String
299 * @return mixed An integer if $text is a valid value otherwise false
300 */
301 function getNsIndex( $text ) {
302 $this->load();
303 $lctext = $this->lc($text);
304 if( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) return $ns;
305 return isset( $this->mNamespaceIds[$lctext] ) ? $this->mNamespaceIds[$lctext] : false;
306 }
307
308 /**
309 * short names for language variants used for language conversion links.
310 *
311 * @param $code String
312 * @return string
313 */
314 function getVariantname( $code ) {
315 return $this->getMessageFromDB( "variantname-$code" );
316 }
317
318 function specialPage( $name ) {
319 $aliases = $this->getSpecialPageAliases();
320 if ( isset( $aliases[$name][0] ) ) {
321 $name = $aliases[$name][0];
322 }
323 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
324 }
325
326 function getQuickbarSettings() {
327 return array(
328 $this->getMessage( 'qbsettings-none' ),
329 $this->getMessage( 'qbsettings-fixedleft' ),
330 $this->getMessage( 'qbsettings-fixedright' ),
331 $this->getMessage( 'qbsettings-floatingleft' ),
332 $this->getMessage( 'qbsettings-floatingright' )
333 );
334 }
335
336 function getMathNames() {
337 $this->load();
338 return $this->mathNames;
339 }
340
341 function getDatePreferences() {
342 $this->load();
343 return $this->datePreferences;
344 }
345
346 function getDateFormats() {
347 $this->load();
348 return $this->dateFormats;
349 }
350
351 function getDefaultDateFormat() {
352 $this->load();
353 return $this->defaultDateFormat;
354 }
355
356 function getDatePreferenceMigrationMap() {
357 $this->load();
358 return $this->datePreferenceMigrationMap;
359 }
360
361 function getImageFile( $image ) {
362 $this->load();
363 return $this->imageFiles[$image];
364 }
365
366 function getDefaultUserOptionOverrides() {
367 $this->load();
368 # XXX - apparently some languageas get empty arrays, didn't get to it yet -- midom
369 if (is_array($this->defaultUserOptionOverrides)) {
370 return $this->defaultUserOptionOverrides;
371 } else {
372 return array();
373 }
374 }
375
376 function getExtraUserToggles() {
377 $this->load();
378 return $this->extraUserToggles;
379 }
380
381 function getUserToggle( $tog ) {
382 return $this->getMessageFromDB( "tog-$tog" );
383 }
384
385 /**
386 * Get language names, indexed by code.
387 * If $customisedOnly is true, only returns codes with a messages file
388 */
389 public static function getLanguageNames( $customisedOnly = false ) {
390 global $wgLanguageNames, $wgExtraLanguageNames;
391 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
392 if ( !$customisedOnly ) {
393 return $allNames;
394 }
395
396 global $IP;
397 $names = array();
398 $dir = opendir( "$IP/languages/messages" );
399 while( false !== ( $file = readdir( $dir ) ) ) {
400 $m = array();
401 if( preg_match( '/Messages([A-Z][a-z_]+)\.php$/', $file, $m ) ) {
402 $code = str_replace( '_', '-', strtolower( $m[1] ) );
403 if ( isset( $allNames[$code] ) ) {
404 $names[$code] = $allNames[$code];
405 }
406 }
407 }
408 closedir( $dir );
409 return $names;
410 }
411
412 /**
413 * Get a message from the MediaWiki namespace.
414 *
415 * @param $msg String: message name
416 * @return string
417 */
418 function getMessageFromDB( $msg ) {
419 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
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 === false ) {
478 $tz = $wgUser->getOption( 'timecorrection' );
479 }
480
481 $data = explode( '|', $tz, 3 );
482
483 if ( $data[0] == 'ZoneInfo' ) {
484 if ( function_exists( 'timezone_open' ) && @timezone_open( $data[2] ) !== false ) {
485 $date = date_create( $ts, timezone_open( 'UTC' ) );
486 date_timezone_set( $date, timezone_open( $data[2] ) );
487 $date = date_format( $date, 'YmdHis' );
488 return $date;
489 }
490 # Unrecognized timezone, default to 'Offset' with the stored offset.
491 $data[0] = 'Offset';
492 }
493
494 $minDiff = 0;
495 if ( $data[0] == 'System' || $tz == '' ) {
496 # Global offset in minutes.
497 if( isset($wgLocalTZoffset) ) $minDiff = $wgLocalTZoffset;
498 } else if ( $data[0] == 'Offset' ) {
499 $minDiff = intval( $data[1] );
500 } else {
501 $data = explode( ':', $tz );
502 if( count( $data ) == 2 ) {
503 $data[0] = intval( $data[0] );
504 $data[1] = intval( $data[1] );
505 $minDiff = abs( $data[0] ) * 60 + $data[1];
506 if ( $data[0] < 0 ) $minDiff = -$minDiff;
507 } else {
508 $minDiff = intval( $data[0] ) * 60;
509 }
510 }
511
512 # No difference ? Return time unchanged
513 if ( 0 == $minDiff ) return $ts;
514
515 wfSuppressWarnings(); // E_STRICT system time bitching
516 # Generate an adjusted date; take advantage of the fact that mktime
517 # will normalize out-of-range values so we don't have to split $minDiff
518 # into hours and minutes.
519 $t = mktime( (
520 (int)substr( $ts, 8, 2) ), # Hours
521 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
522 (int)substr( $ts, 12, 2 ), # Seconds
523 (int)substr( $ts, 4, 2 ), # Month
524 (int)substr( $ts, 6, 2 ), # Day
525 (int)substr( $ts, 0, 4 ) ); #Year
526
527 $date = date( 'YmdHis', $t );
528 wfRestoreWarnings();
529
530 return $date;
531 }
532
533 /**
534 * This is a workalike of PHP's date() function, but with better
535 * internationalisation, a reduced set of format characters, and a better
536 * escaping format.
537 *
538 * Supported format characters are dDjlNwzWFmMntLYyaAgGhHiscrU. See the
539 * PHP manual for definitions. There are a number of extensions, which
540 * start with "x":
541 *
542 * xn Do not translate digits of the next numeric format character
543 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
544 * xr Use roman numerals for the next numeric format character
545 * xh Use hebrew numerals for the next numeric format character
546 * xx Literal x
547 * xg Genitive month name
548 *
549 * xij j (day number) in Iranian calendar
550 * xiF F (month name) in Iranian calendar
551 * xin n (month number) in Iranian calendar
552 * xiY Y (full year) in Iranian calendar
553 *
554 * xjj j (day number) in Hebrew calendar
555 * xjF F (month name) in Hebrew calendar
556 * xjt t (days in month) in Hebrew calendar
557 * xjx xg (genitive month name) in Hebrew calendar
558 * xjn n (month number) in Hebrew calendar
559 * xjY Y (full year) in Hebrew calendar
560 *
561 * xmj j (day number) in Hijri calendar
562 * xmF F (month name) in Hijri calendar
563 * xmn n (month number) in Hijri calendar
564 * xmY Y (full year) in Hijri calendar
565 *
566 * xkY Y (full year) in Thai solar calendar. Months and days are
567 * identical to the Gregorian calendar
568 *
569 * Characters enclosed in double quotes will be considered literal (with
570 * the quotes themselves removed). Unmatched quotes will be considered
571 * literal quotes. Example:
572 *
573 * "The month is" F => The month is January
574 * i's" => 20'11"
575 *
576 * Backslash escaping is also supported.
577 *
578 * Input timestamp is assumed to be pre-normalized to the desired local
579 * time zone, if any.
580 *
581 * @param $format String
582 * @param $ts String: 14-character timestamp
583 * YYYYMMDDHHMMSS
584 * 01234567890123
585 */
586 function sprintfDate( $format, $ts ) {
587 $s = '';
588 $raw = false;
589 $roman = false;
590 $hebrewNum = false;
591 $unix = false;
592 $rawToggle = false;
593 $iranian = false;
594 $hebrew = false;
595 $hijri = false;
596 $thai = false;
597 for ( $p = 0; $p < strlen( $format ); $p++ ) {
598 $num = false;
599 $code = $format[$p];
600 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
601 $code .= $format[++$p];
602 }
603
604 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' ) && $p < strlen( $format ) - 1 ) {
605 $code .= $format[++$p];
606 }
607
608 switch ( $code ) {
609 case 'xx':
610 $s .= 'x';
611 break;
612 case 'xn':
613 $raw = true;
614 break;
615 case 'xN':
616 $rawToggle = !$rawToggle;
617 break;
618 case 'xr':
619 $roman = true;
620 break;
621 case 'xh':
622 $hebrewNum = true;
623 break;
624 case 'xg':
625 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
626 break;
627 case 'xjx':
628 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
629 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
630 break;
631 case 'd':
632 $num = substr( $ts, 6, 2 );
633 break;
634 case 'D':
635 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
636 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
637 break;
638 case 'j':
639 $num = intval( substr( $ts, 6, 2 ) );
640 break;
641 case 'xij':
642 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
643 $num = $iranian[2];
644 break;
645 case 'xmj':
646 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
647 $num = $hijri[2];
648 break;
649 case 'xjj':
650 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
651 $num = $hebrew[2];
652 break;
653 case 'l':
654 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
655 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
656 break;
657 case 'N':
658 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
659 $w = gmdate( 'w', $unix );
660 $num = $w ? $w : 7;
661 break;
662 case 'w':
663 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
664 $num = gmdate( 'w', $unix );
665 break;
666 case 'z':
667 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
668 $num = gmdate( 'z', $unix );
669 break;
670 case 'W':
671 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
672 $num = gmdate( 'W', $unix );
673 break;
674 case 'F':
675 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
676 break;
677 case 'xiF':
678 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
679 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
680 break;
681 case 'xmF':
682 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
683 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
684 break;
685 case 'xjF':
686 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
687 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
688 break;
689 case 'm':
690 $num = substr( $ts, 4, 2 );
691 break;
692 case 'M':
693 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
694 break;
695 case 'n':
696 $num = intval( substr( $ts, 4, 2 ) );
697 break;
698 case 'xin':
699 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
700 $num = $iranian[1];
701 break;
702 case 'xmn':
703 if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
704 $num = $hijri[1];
705 break;
706 case 'xjn':
707 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
708 $num = $hebrew[1];
709 break;
710 case 't':
711 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
712 $num = gmdate( 't', $unix );
713 break;
714 case 'xjt':
715 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
716 $num = $hebrew[3];
717 break;
718 case 'L':
719 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
720 $num = gmdate( 'L', $unix );
721 break;
722 case 'Y':
723 $num = substr( $ts, 0, 4 );
724 break;
725 case 'xiY':
726 if ( !$iranian ) $iranian = self::tsToIranian( $ts );
727 $num = $iranian[0];
728 break;
729 case 'xmY':
730 if ( !$hijri ) $hijri = self::tsToHijri( $ts );
731 $num = $hijri[0];
732 break;
733 case 'xjY':
734 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
735 $num = $hebrew[0];
736 break;
737 case 'xkY':
738 if ( !$thai ) $thai = self::tsToThai( $ts );
739 $num = $thai[0];
740 break;
741 case 'y':
742 $num = substr( $ts, 2, 2 );
743 break;
744 case 'a':
745 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
746 break;
747 case 'A':
748 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
749 break;
750 case 'g':
751 $h = substr( $ts, 8, 2 );
752 $num = $h % 12 ? $h % 12 : 12;
753 break;
754 case 'G':
755 $num = intval( substr( $ts, 8, 2 ) );
756 break;
757 case 'h':
758 $h = substr( $ts, 8, 2 );
759 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
760 break;
761 case 'H':
762 $num = substr( $ts, 8, 2 );
763 break;
764 case 'i':
765 $num = substr( $ts, 10, 2 );
766 break;
767 case 's':
768 $num = substr( $ts, 12, 2 );
769 break;
770 case 'c':
771 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
772 $s .= gmdate( 'c', $unix );
773 break;
774 case 'r':
775 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
776 $s .= gmdate( 'r', $unix );
777 break;
778 case 'U':
779 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
780 $num = $unix;
781 break;
782 case '\\':
783 # Backslash escaping
784 if ( $p < strlen( $format ) - 1 ) {
785 $s .= $format[++$p];
786 } else {
787 $s .= '\\';
788 }
789 break;
790 case '"':
791 # Quoted literal
792 if ( $p < strlen( $format ) - 1 ) {
793 $endQuote = strpos( $format, '"', $p + 1 );
794 if ( $endQuote === false ) {
795 # No terminating quote, assume literal "
796 $s .= '"';
797 } else {
798 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
799 $p = $endQuote;
800 }
801 } else {
802 # Quote at end of string, assume literal "
803 $s .= '"';
804 }
805 break;
806 default:
807 $s .= $format[$p];
808 }
809 if ( $num !== false ) {
810 if ( $rawToggle || $raw ) {
811 $s .= $num;
812 $raw = false;
813 } elseif ( $roman ) {
814 $s .= self::romanNumeral( $num );
815 $roman = false;
816 } elseif( $hebrewNum ) {
817 $s .= self::hebrewNumeral( $num );
818 $hebrewNum = false;
819 } else {
820 $s .= $this->formatNum( $num, true );
821 }
822 $num = false;
823 }
824 }
825 return $s;
826 }
827
828 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
829 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
830 /**
831 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
832 * Gregorian dates to Iranian dates. Originally written in C, it
833 * is released under the terms of GNU Lesser General Public
834 * License. Conversion to PHP was performed by Niklas Laxström.
835 *
836 * Link: http://www.farsiweb.info/jalali/jalali.c
837 */
838 private static function tsToIranian( $ts ) {
839 $gy = substr( $ts, 0, 4 ) -1600;
840 $gm = substr( $ts, 4, 2 ) -1;
841 $gd = substr( $ts, 6, 2 ) -1;
842
843 # Days passed from the beginning (including leap years)
844 $gDayNo = 365*$gy
845 + floor(($gy+3) / 4)
846 - floor(($gy+99) / 100)
847 + floor(($gy+399) / 400);
848
849
850 // Add days of the past months of this year
851 for( $i = 0; $i < $gm; $i++ ) {
852 $gDayNo += self::$GREG_DAYS[$i];
853 }
854
855 // Leap years
856 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
857 $gDayNo++;
858 }
859
860 // Days passed in current month
861 $gDayNo += $gd;
862
863 $jDayNo = $gDayNo - 79;
864
865 $jNp = floor($jDayNo / 12053);
866 $jDayNo %= 12053;
867
868 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
869 $jDayNo %= 1461;
870
871 if ( $jDayNo >= 366 ) {
872 $jy += floor(($jDayNo-1)/365);
873 $jDayNo = floor(($jDayNo-1)%365);
874 }
875
876 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
877 $jDayNo -= self::$IRANIAN_DAYS[$i];
878 }
879
880 $jm= $i+1;
881 $jd= $jDayNo+1;
882
883 return array($jy, $jm, $jd);
884 }
885 /**
886 * Converting Gregorian dates to Hijri dates.
887 *
888 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
889 *
890 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
891 */
892 private static function tsToHijri ( $ts ) {
893 $year = substr( $ts, 0, 4 );
894 $month = substr( $ts, 4, 2 );
895 $day = substr( $ts, 6, 2 );
896
897 $zyr = $year;
898 $zd=$day;
899 $zm=$month;
900 $zy=$zyr;
901
902
903
904 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
905 {
906
907
908 $zjd=(int)((1461*($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;
909 }
910 else
911 {
912 $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
913 }
914
915 $zl=$zjd-1948440+10632;
916 $zn=(int)(($zl-1)/10631);
917 $zl=$zl-10631*$zn+354;
918 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
919 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
920 $zm=(int)((24*$zl)/709);
921 $zd=$zl-(int)((709*$zm)/24);
922 $zy=30*$zn+$zj-30;
923
924 return array ($zy, $zm, $zd);
925 }
926
927 /**
928 * Converting Gregorian dates to Hebrew dates.
929 *
930 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
931 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
932 * to translate the relevant functions into PHP and release them under
933 * GNU GPL.
934 *
935 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
936 * and Adar II is 14. In a non-leap year, Adar is 6.
937 */
938 private static function tsToHebrew( $ts ) {
939 # Parse date
940 $year = substr( $ts, 0, 4 );
941 $month = substr( $ts, 4, 2 );
942 $day = substr( $ts, 6, 2 );
943
944 # Calculate Hebrew year
945 $hebrewYear = $year + 3760;
946
947 # Month number when September = 1, August = 12
948 $month += 4;
949 if( $month > 12 ) {
950 # Next year
951 $month -= 12;
952 $year++;
953 $hebrewYear++;
954 }
955
956 # Calculate day of year from 1 September
957 $dayOfYear = $day;
958 for( $i = 1; $i < $month; $i++ ) {
959 if( $i == 6 ) {
960 # February
961 $dayOfYear += 28;
962 # Check if the year is leap
963 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
964 $dayOfYear++;
965 }
966 } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
967 $dayOfYear += 30;
968 } else {
969 $dayOfYear += 31;
970 }
971 }
972
973 # Calculate the start of the Hebrew year
974 $start = self::hebrewYearStart( $hebrewYear );
975
976 # Calculate next year's start
977 if( $dayOfYear <= $start ) {
978 # Day is before the start of the year - it is the previous year
979 # Next year's start
980 $nextStart = $start;
981 # Previous year
982 $year--;
983 $hebrewYear--;
984 # Add days since previous year's 1 September
985 $dayOfYear += 365;
986 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
987 # Leap year
988 $dayOfYear++;
989 }
990 # Start of the new (previous) year
991 $start = self::hebrewYearStart( $hebrewYear );
992 } else {
993 # Next year's start
994 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
995 }
996
997 # Calculate Hebrew day of year
998 $hebrewDayOfYear = $dayOfYear - $start;
999
1000 # Difference between year's days
1001 $diff = $nextStart - $start;
1002 # Add 12 (or 13 for leap years) days to ignore the difference between
1003 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1004 # difference is only about the year type
1005 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1006 $diff += 13;
1007 } else {
1008 $diff += 12;
1009 }
1010
1011 # Check the year pattern, and is leap year
1012 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1013 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1014 # and non-leap years
1015 $yearPattern = $diff % 30;
1016 # Check if leap year
1017 $isLeap = $diff >= 30;
1018
1019 # Calculate day in the month from number of day in the Hebrew year
1020 # Don't check Adar - if the day is not in Adar, we will stop before;
1021 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1022 $hebrewDay = $hebrewDayOfYear;
1023 $hebrewMonth = 1;
1024 $days = 0;
1025 while( $hebrewMonth <= 12 ) {
1026 # Calculate days in this month
1027 if( $isLeap && $hebrewMonth == 6 ) {
1028 # Adar in a leap year
1029 if( $isLeap ) {
1030 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1031 $days = 30;
1032 if( $hebrewDay <= $days ) {
1033 # Day in Adar I
1034 $hebrewMonth = 13;
1035 } else {
1036 # Subtract the days of Adar I
1037 $hebrewDay -= $days;
1038 # Try Adar II
1039 $days = 29;
1040 if( $hebrewDay <= $days ) {
1041 # Day in Adar II
1042 $hebrewMonth = 14;
1043 }
1044 }
1045 }
1046 } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1047 # Cheshvan in a complete year (otherwise as the rule below)
1048 $days = 30;
1049 } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1050 # Kislev in an incomplete year (otherwise as the rule below)
1051 $days = 29;
1052 } else {
1053 # Odd months have 30 days, even have 29
1054 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1055 }
1056 if( $hebrewDay <= $days ) {
1057 # In the current month
1058 break;
1059 } else {
1060 # Subtract the days of the current month
1061 $hebrewDay -= $days;
1062 # Try in the next month
1063 $hebrewMonth++;
1064 }
1065 }
1066
1067 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1068 }
1069
1070 /**
1071 * This calculates the Hebrew year start, as days since 1 September.
1072 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1073 * Used for Hebrew date.
1074 */
1075 private static function hebrewYearStart( $year ) {
1076 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1077 $b = intval( ( $year - 1 ) % 4 );
1078 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1079 if( $m < 0 ) {
1080 $m--;
1081 }
1082 $Mar = intval( $m );
1083 if( $m < 0 ) {
1084 $m++;
1085 }
1086 $m -= $Mar;
1087
1088 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1089 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1090 $Mar++;
1091 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1092 $Mar += 2;
1093 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1094 $Mar++;
1095 }
1096
1097 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1098 return $Mar;
1099 }
1100
1101 /**
1102 * Algorithm to convert Gregorian dates to Thai solar dates.
1103 *
1104 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1105 *
1106 * @param $ts String: 14-character timestamp
1107 * @return array converted year, month, day
1108 */
1109 private static function tsToThai( $ts ) {
1110 $gy = substr( $ts, 0, 4 );
1111 $gm = substr( $ts, 4, 2 );
1112 $gd = substr( $ts, 6, 2 );
1113
1114 # Add 543 years to the Gregorian calendar
1115 # Months and days are identical
1116 $gy_thai = $gy + 543;
1117
1118 return array( $gy_thai, $gm, $gd );
1119 }
1120
1121
1122 /**
1123 * Roman number formatting up to 3000
1124 */
1125 static function romanNumeral( $num ) {
1126 static $table = array(
1127 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1128 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1129 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1130 array( '', 'M', 'MM', 'MMM' )
1131 );
1132
1133 $num = intval( $num );
1134 if ( $num > 3000 || $num <= 0 ) {
1135 return $num;
1136 }
1137
1138 $s = '';
1139 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1140 if ( $num >= $pow10 ) {
1141 $s .= $table[$i][floor($num / $pow10)];
1142 }
1143 $num = $num % $pow10;
1144 }
1145 return $s;
1146 }
1147
1148 /**
1149 * Hebrew Gematria number formatting up to 9999
1150 */
1151 static function hebrewNumeral( $num ) {
1152 static $table = array(
1153 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1154 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1155 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1156 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1157 );
1158
1159 $num = intval( $num );
1160 if ( $num > 9999 || $num <= 0 ) {
1161 return $num;
1162 }
1163
1164 $s = '';
1165 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1166 if ( $num >= $pow10 ) {
1167 if ( $num == 15 || $num == 16 ) {
1168 $s .= $table[0][9] . $table[0][$num - 9];
1169 $num = 0;
1170 } else {
1171 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1172 if( $pow10 == 1000 ) {
1173 $s .= "'";
1174 }
1175 }
1176 }
1177 $num = $num % $pow10;
1178 }
1179 if( strlen( $s ) == 2 ) {
1180 $str = $s . "'";
1181 } else {
1182 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1183 $str .= substr( $s, strlen( $s ) - 2, 2 );
1184 }
1185 $start = substr( $str, 0, strlen( $str ) - 2 );
1186 $end = substr( $str, strlen( $str ) - 2 );
1187 switch( $end ) {
1188 case 'כ':
1189 $str = $start . 'ך';
1190 break;
1191 case 'מ':
1192 $str = $start . 'ם';
1193 break;
1194 case 'נ':
1195 $str = $start . 'ן';
1196 break;
1197 case 'פ':
1198 $str = $start . 'ף';
1199 break;
1200 case 'צ':
1201 $str = $start . 'ץ';
1202 break;
1203 }
1204 return $str;
1205 }
1206
1207 /**
1208 * This is meant to be used by time(), date(), and timeanddate() to get
1209 * the date preference they're supposed to use, it should be used in
1210 * all children.
1211 *
1212 *<code>
1213 * function timeanddate([...], $format = true) {
1214 * $datePreference = $this->dateFormat($format);
1215 * [...]
1216 * }
1217 *</code>
1218 *
1219 * @param $usePrefs Mixed: if true, the user's preference is used
1220 * if false, the site/language default is used
1221 * if int/string, assumed to be a format.
1222 * @return string
1223 */
1224 function dateFormat( $usePrefs = true ) {
1225 global $wgUser;
1226
1227 if( is_bool( $usePrefs ) ) {
1228 if( $usePrefs ) {
1229 $datePreference = $wgUser->getDatePreference();
1230 } else {
1231 $options = User::getDefaultOptions();
1232 $datePreference = (string)$options['date'];
1233 }
1234 } else {
1235 $datePreference = (string)$usePrefs;
1236 }
1237
1238 // return int
1239 if( $datePreference == '' ) {
1240 return 'default';
1241 }
1242
1243 return $datePreference;
1244 }
1245
1246 /**
1247 * @param $ts Mixed: the time format which needs to be turned into a
1248 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1249 * @param $adj Bool: whether to adjust the time output according to the
1250 * user configured offset ($timecorrection)
1251 * @param $format Mixed: true to use user's date format preference
1252 * @param $timecorrection String: the time offset as returned by
1253 * validateTimeZone() in Special:Preferences
1254 * @return string
1255 */
1256 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1257 $this->load();
1258 if ( $adj ) {
1259 $ts = $this->userAdjust( $ts, $timecorrection );
1260 }
1261
1262 $pref = $this->dateFormat( $format );
1263 if( $pref == 'default' || !isset( $this->dateFormats["$pref date"] ) ) {
1264 $pref = $this->defaultDateFormat;
1265 }
1266 return $this->sprintfDate( $this->dateFormats["$pref date"], $ts );
1267 }
1268
1269 /**
1270 * @param $ts Mixed: the time format which needs to be turned into a
1271 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1272 * @param $adj Bool: whether to adjust the time output according to the
1273 * user configured offset ($timecorrection)
1274 * @param $format Mixed: true to use user's date format preference
1275 * @param $timecorrection String: the time offset as returned by
1276 * validateTimeZone() in Special:Preferences
1277 * @return string
1278 */
1279 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1280 $this->load();
1281 if ( $adj ) {
1282 $ts = $this->userAdjust( $ts, $timecorrection );
1283 }
1284
1285 $pref = $this->dateFormat( $format );
1286 if( $pref == 'default' || !isset( $this->dateFormats["$pref time"] ) ) {
1287 $pref = $this->defaultDateFormat;
1288 }
1289 return $this->sprintfDate( $this->dateFormats["$pref time"], $ts );
1290 }
1291
1292 /**
1293 * @param $ts Mixed: the time format which needs to be turned into a
1294 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1295 * @param $adj Bool: whether to adjust the time output according to the
1296 * user configured offset ($timecorrection)
1297 * @param $format Mixed: what format to return, if it's false output the
1298 * default one (default true)
1299 * @param $timecorrection String: the time offset as returned by
1300 * validateTimeZone() in Special:Preferences
1301 * @return string
1302 */
1303 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1304 $this->load();
1305
1306 $ts = wfTimestamp( TS_MW, $ts );
1307
1308 if ( $adj ) {
1309 $ts = $this->userAdjust( $ts, $timecorrection );
1310 }
1311
1312 $pref = $this->dateFormat( $format );
1313 if( $pref == 'default' || !isset( $this->dateFormats["$pref both"] ) ) {
1314 $pref = $this->defaultDateFormat;
1315 }
1316
1317 return $this->sprintfDate( $this->dateFormats["$pref both"], $ts );
1318 }
1319
1320 function getMessage( $key ) {
1321 $this->load();
1322 return isset( $this->messages[$key] ) ? $this->messages[$key] : null;
1323 }
1324
1325 function getAllMessages() {
1326 $this->load();
1327 return $this->messages;
1328 }
1329
1330 function iconv( $in, $out, $string ) {
1331 # For most languages, this is a wrapper for iconv
1332 return iconv( $in, $out . '//IGNORE', $string );
1333 }
1334
1335 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1336 function ucwordbreaksCallbackAscii($matches){
1337 return $this->ucfirst($matches[1]);
1338 }
1339
1340 function ucwordbreaksCallbackMB($matches){
1341 return mb_strtoupper($matches[0]);
1342 }
1343
1344 function ucCallback($matches){
1345 list( $wikiUpperChars ) = self::getCaseMaps();
1346 return strtr( $matches[1], $wikiUpperChars );
1347 }
1348
1349 function lcCallback($matches){
1350 list( , $wikiLowerChars ) = self::getCaseMaps();
1351 return strtr( $matches[1], $wikiLowerChars );
1352 }
1353
1354 function ucwordsCallbackMB($matches){
1355 return mb_strtoupper($matches[0]);
1356 }
1357
1358 function ucwordsCallbackWiki($matches){
1359 list( $wikiUpperChars ) = self::getCaseMaps();
1360 return strtr( $matches[0], $wikiUpperChars );
1361 }
1362
1363 function ucfirst( $str ) {
1364 if ( empty($str) ) return $str;
1365 if ( ord($str[0]) < 128 ) return ucfirst($str);
1366 else return self::uc($str,true); // fall back to more complex logic in case of multibyte strings
1367 }
1368
1369 function uc( $str, $first = false ) {
1370 if ( function_exists( 'mb_strtoupper' ) ) {
1371 if ( $first ) {
1372 if ( self::isMultibyte( $str ) ) {
1373 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1374 } else {
1375 return ucfirst( $str );
1376 }
1377 } else {
1378 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1379 }
1380 } else {
1381 if ( self::isMultibyte( $str ) ) {
1382 list( $wikiUpperChars ) = $this->getCaseMaps();
1383 $x = $first ? '^' : '';
1384 return preg_replace_callback(
1385 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1386 array($this,"ucCallback"),
1387 $str
1388 );
1389 } else {
1390 return $first ? ucfirst( $str ) : strtoupper( $str );
1391 }
1392 }
1393 }
1394
1395 function lcfirst( $str ) {
1396 if ( empty($str) ) return $str;
1397 if ( is_string( $str ) && ord($str[0]) < 128 ) {
1398 // editing string in place = cool
1399 $str[0]=strtolower($str[0]);
1400 return $str;
1401 }
1402 else return self::lc( $str, true );
1403 }
1404
1405 function lc( $str, $first = false ) {
1406 if ( function_exists( 'mb_strtolower' ) )
1407 if ( $first )
1408 if ( self::isMultibyte( $str ) )
1409 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1410 else
1411 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1412 else
1413 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1414 else
1415 if ( self::isMultibyte( $str ) ) {
1416 list( , $wikiLowerChars ) = self::getCaseMaps();
1417 $x = $first ? '^' : '';
1418 return preg_replace_callback(
1419 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1420 array($this,"lcCallback"),
1421 $str
1422 );
1423 } else
1424 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1425 }
1426
1427 function isMultibyte( $str ) {
1428 return (bool)preg_match( '/[\x80-\xff]/', $str );
1429 }
1430
1431 function ucwords($str) {
1432 if ( self::isMultibyte( $str ) ) {
1433 $str = self::lc($str);
1434
1435 // regexp to find first letter in each word (i.e. after each space)
1436 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1437
1438 // function to use to capitalize a single char
1439 if ( function_exists( 'mb_strtoupper' ) )
1440 return preg_replace_callback(
1441 $replaceRegexp,
1442 array($this,"ucwordsCallbackMB"),
1443 $str
1444 );
1445 else
1446 return preg_replace_callback(
1447 $replaceRegexp,
1448 array($this,"ucwordsCallbackWiki"),
1449 $str
1450 );
1451 }
1452 else
1453 return ucwords( strtolower( $str ) );
1454 }
1455
1456 # capitalize words at word breaks
1457 function ucwordbreaks($str){
1458 if (self::isMultibyte( $str ) ) {
1459 $str = self::lc($str);
1460
1461 // since \b doesn't work for UTF-8, we explicitely define word break chars
1462 $breaks= "[ \-\(\)\}\{\.,\?!]";
1463
1464 // find first letter after word break
1465 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1466
1467 if ( function_exists( 'mb_strtoupper' ) )
1468 return preg_replace_callback(
1469 $replaceRegexp,
1470 array($this,"ucwordbreaksCallbackMB"),
1471 $str
1472 );
1473 else
1474 return preg_replace_callback(
1475 $replaceRegexp,
1476 array($this,"ucwordsCallbackWiki"),
1477 $str
1478 );
1479 }
1480 else
1481 return preg_replace_callback(
1482 '/\b([\w\x80-\xff]+)\b/',
1483 array($this,"ucwordbreaksCallbackAscii"),
1484 $str );
1485 }
1486
1487 /**
1488 * Return a case-folded representation of $s
1489 *
1490 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1491 * and $s2 are the same except for the case of their characters. It is not
1492 * necessary for the value returned to make sense when displayed.
1493 *
1494 * Do *not* perform any other normalisation in this function. If a caller
1495 * uses this function when it should be using a more general normalisation
1496 * function, then fix the caller.
1497 */
1498 function caseFold( $s ) {
1499 return $this->uc( $s );
1500 }
1501
1502 function checkTitleEncoding( $s ) {
1503 if( is_array( $s ) ) {
1504 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1505 }
1506 # Check for non-UTF-8 URLs
1507 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1508 if(!$ishigh) return $s;
1509
1510 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1511 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1512 if( $isutf8 ) return $s;
1513
1514 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1515 }
1516
1517 function fallback8bitEncoding() {
1518 $this->load();
1519 return $this->fallback8bitEncoding;
1520 }
1521
1522 /**
1523 * Some languages have special punctuation to strip out
1524 * or characters which need to be converted for MySQL's
1525 * indexing to grok it correctly. Make such changes here.
1526 *
1527 * @param $string String
1528 * @return String
1529 */
1530 function stripForSearch( $string ) {
1531 global $wgDBtype;
1532 if ( $wgDBtype != 'mysql' ) {
1533 return $string;
1534 }
1535
1536
1537 wfProfileIn( __METHOD__ );
1538
1539 // MySQL fulltext index doesn't grok utf-8, so we
1540 // need to fold cases and convert to hex
1541 $out = preg_replace_callback(
1542 "/([\\xc0-\\xff][\\x80-\\xbf]*)/",
1543 array( $this, 'stripForSearchCallback' ),
1544 $this->lc( $string ) );
1545
1546 // And to add insult to injury, the default indexing
1547 // ignores short words... Pad them so we can pass them
1548 // through without reconfiguring the server...
1549 $minLength = $this->minSearchLength();
1550 if( $minLength > 1 ) {
1551 $n = $minLength-1;
1552 $out = preg_replace(
1553 "/\b(\w{1,$n})\b/",
1554 "$1U800",
1555 $out );
1556 }
1557
1558 // Periods within things like hostnames and IP addresses
1559 // are also important -- we want a search for "example.com"
1560 // or "192.168.1.1" to work sanely.
1561 //
1562 // MySQL's search seems to ignore them, so you'd match on
1563 // "example.wikipedia.com" and "192.168.83.1" as well.
1564 $out = preg_replace(
1565 "/(\w)\.(\w|\*)/u",
1566 "$1U82e$2",
1567 $out );
1568
1569 wfProfileOut( __METHOD__ );
1570 return $out;
1571 }
1572
1573 /**
1574 * Armor a case-folded UTF-8 string to get through MySQL's
1575 * fulltext search without being mucked up by funny charset
1576 * settings or anything else of the sort.
1577 */
1578 protected function stripForSearchCallback( $matches ) {
1579 return 'U8' . bin2hex( $matches[1] );
1580 }
1581
1582 /**
1583 * Check MySQL server's ft_min_word_len setting so we know
1584 * if we need to pad short words...
1585 */
1586 protected function minSearchLength() {
1587 if( !isset( $this->minSearchLength ) ) {
1588 $sql = "show global variables like 'ft\\_min\\_word\\_len'";
1589 $dbr = wfGetDB( DB_SLAVE );
1590 $result = $dbr->query( $sql );
1591 $row = $result->fetchObject();
1592 $result->free();
1593
1594 if( $row && $row->Variable_name == 'ft_min_word_len' ) {
1595 $this->minSearchLength = intval( $row->Value );
1596 } else {
1597 $this->minSearchLength = 0;
1598 }
1599 }
1600 return $this->minSearchLength;
1601 }
1602
1603 function convertForSearchResult( $termsArray ) {
1604 # some languages, e.g. Chinese, need to do a conversion
1605 # in order for search results to be displayed correctly
1606 return $termsArray;
1607 }
1608
1609 /**
1610 * Get the first character of a string.
1611 *
1612 * @param $s string
1613 * @return string
1614 */
1615 function firstChar( $s ) {
1616 $matches = array();
1617 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1618 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1619
1620 if ( isset( $matches[1] ) ) {
1621 if ( strlen( $matches[1] ) != 3 ) {
1622 return $matches[1];
1623 }
1624
1625 // Break down Hangul syllables to grab the first jamo
1626 $code = utf8ToCodepoint( $matches[1] );
1627 if ( $code < 0xac00 || 0xd7a4 <= $code) {
1628 return $matches[1];
1629 } elseif ( $code < 0xb098 ) {
1630 return "\xe3\x84\xb1";
1631 } elseif ( $code < 0xb2e4 ) {
1632 return "\xe3\x84\xb4";
1633 } elseif ( $code < 0xb77c ) {
1634 return "\xe3\x84\xb7";
1635 } elseif ( $code < 0xb9c8 ) {
1636 return "\xe3\x84\xb9";
1637 } elseif ( $code < 0xbc14 ) {
1638 return "\xe3\x85\x81";
1639 } elseif ( $code < 0xc0ac ) {
1640 return "\xe3\x85\x82";
1641 } elseif ( $code < 0xc544 ) {
1642 return "\xe3\x85\x85";
1643 } elseif ( $code < 0xc790 ) {
1644 return "\xe3\x85\x87";
1645 } elseif ( $code < 0xcc28 ) {
1646 return "\xe3\x85\x88";
1647 } elseif ( $code < 0xce74 ) {
1648 return "\xe3\x85\x8a";
1649 } elseif ( $code < 0xd0c0 ) {
1650 return "\xe3\x85\x8b";
1651 } elseif ( $code < 0xd30c ) {
1652 return "\xe3\x85\x8c";
1653 } elseif ( $code < 0xd558 ) {
1654 return "\xe3\x85\x8d";
1655 } else {
1656 return "\xe3\x85\x8e";
1657 }
1658 } else {
1659 return "";
1660 }
1661 }
1662
1663 function initEncoding() {
1664 # Some languages may have an alternate char encoding option
1665 # (Esperanto X-coding, Japanese furigana conversion, etc)
1666 # If this language is used as the primary content language,
1667 # an override to the defaults can be set here on startup.
1668 }
1669
1670 function recodeForEdit( $s ) {
1671 # For some languages we'll want to explicitly specify
1672 # which characters make it into the edit box raw
1673 # or are converted in some way or another.
1674 # Note that if wgOutputEncoding is different from
1675 # wgInputEncoding, this text will be further converted
1676 # to wgOutputEncoding.
1677 global $wgEditEncoding;
1678 if( $wgEditEncoding == '' or
1679 $wgEditEncoding == 'UTF-8' ) {
1680 return $s;
1681 } else {
1682 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1683 }
1684 }
1685
1686 function recodeInput( $s ) {
1687 # Take the previous into account.
1688 global $wgEditEncoding;
1689 if($wgEditEncoding != "") {
1690 $enc = $wgEditEncoding;
1691 } else {
1692 $enc = 'UTF-8';
1693 }
1694 if( $enc == 'UTF-8' ) {
1695 return $s;
1696 } else {
1697 return $this->iconv( $enc, 'UTF-8', $s );
1698 }
1699 }
1700
1701 /**
1702 * For right-to-left language support
1703 *
1704 * @return bool
1705 */
1706 function isRTL() {
1707 $this->load();
1708 return $this->rtl;
1709 }
1710
1711 /**
1712 * A hidden direction mark (LRM or RLM), depending on the language direction
1713 *
1714 * @return string
1715 */
1716 function getDirMark() {
1717 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1718 }
1719
1720 /**
1721 * An arrow, depending on the language direction
1722 *
1723 * @return string
1724 */
1725 function getArrow() {
1726 return $this->isRTL() ? '←' : '→';
1727 }
1728
1729 /**
1730 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1731 *
1732 * @return bool
1733 */
1734 function linkPrefixExtension() {
1735 $this->load();
1736 return $this->linkPrefixExtension;
1737 }
1738
1739 function &getMagicWords() {
1740 $this->load();
1741 return $this->magicWords;
1742 }
1743
1744 # Fill a MagicWord object with data from here
1745 function getMagic( &$mw ) {
1746 if ( !$this->mMagicHookDone ) {
1747 $this->mMagicHookDone = true;
1748 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1749 }
1750 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1751 $rawEntry = $this->mMagicExtensions[$mw->mId];
1752 } else {
1753 $magicWords =& $this->getMagicWords();
1754 if ( isset( $magicWords[$mw->mId] ) ) {
1755 $rawEntry = $magicWords[$mw->mId];
1756 } else {
1757 # Fall back to English if local list is incomplete
1758 $magicWords =& Language::getMagicWords();
1759 if ( !isset($magicWords[$mw->mId]) ) {
1760 throw new MWException("Magic word '{$mw->mId}' not found" );
1761 }
1762 $rawEntry = $magicWords[$mw->mId];
1763 }
1764 }
1765
1766 if( !is_array( $rawEntry ) ) {
1767 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1768 } else {
1769 $mw->mCaseSensitive = $rawEntry[0];
1770 $mw->mSynonyms = array_slice( $rawEntry, 1 );
1771 }
1772 }
1773
1774 /**
1775 * Add magic words to the extension array
1776 */
1777 function addMagicWordsByLang( $newWords ) {
1778 $code = $this->getCode();
1779 $fallbackChain = array();
1780 while ( $code && !in_array( $code, $fallbackChain ) ) {
1781 $fallbackChain[] = $code;
1782 $code = self::getFallbackFor( $code );
1783 }
1784 if ( !in_array( 'en', $fallbackChain ) ) {
1785 $fallbackChain[] = 'en';
1786 }
1787 $fallbackChain = array_reverse( $fallbackChain );
1788 foreach ( $fallbackChain as $code ) {
1789 if ( isset( $newWords[$code] ) ) {
1790 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1791 }
1792 }
1793 }
1794
1795 /**
1796 * Get special page names, as an associative array
1797 * case folded alias => real name
1798 */
1799 function getSpecialPageAliases() {
1800 $this->load();
1801
1802 // Cache aliases because it may be slow to load them
1803 if ( !isset( $this->mExtendedSpecialPageAliases ) ) {
1804
1805 // Initialise array
1806 $this->mExtendedSpecialPageAliases = $this->specialPageAliases;
1807
1808 global $wgExtensionAliasesFiles;
1809 foreach ( $wgExtensionAliasesFiles as $file ) {
1810
1811 // Fail fast
1812 if ( !file_exists($file) )
1813 throw new MWException( "Aliases file does not exist: $file" );
1814
1815 $aliases = array();
1816 require($file);
1817
1818 // Check the availability of aliases
1819 if ( !isset($aliases['en']) )
1820 throw new MWException( "Malformed aliases file: $file" );
1821
1822 // Merge all aliases in fallback chain
1823 $code = $this->getCode();
1824 do {
1825 if ( !isset($aliases[$code]) ) continue;
1826
1827 $aliases[$code] = $this->fixSpecialPageAliases( $aliases[$code] );
1828 /* Merge the aliases, THIS will break if there is special page name
1829 * which looks like a numerical key, thanks to PHP...
1830 * See the array_merge_recursive manual entry */
1831 $this->mExtendedSpecialPageAliases = array_merge_recursive(
1832 $this->mExtendedSpecialPageAliases, $aliases[$code] );
1833
1834 } while ( $code = self::getFallbackFor( $code ) );
1835 }
1836
1837 wfRunHooks( 'LanguageGetSpecialPageAliases',
1838 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1839 }
1840
1841 return $this->mExtendedSpecialPageAliases;
1842 }
1843
1844 /**
1845 * Function to fix special page aliases. Will convert the first letter to
1846 * upper case and spaces to underscores. Can be given a full aliases array,
1847 * in which case it will recursively fix all aliases.
1848 */
1849 public function fixSpecialPageAliases( $mixed ) {
1850 // Work recursively until in string level
1851 if ( is_array($mixed) ) {
1852 $callback = array( $this, 'fixSpecialPageAliases' );
1853 return array_map( $callback, $mixed );
1854 }
1855 return str_replace( ' ', '_', $this->ucfirst( $mixed ) );
1856 }
1857
1858 /**
1859 * Italic is unsuitable for some languages
1860 *
1861 * @param $text String: the text to be emphasized.
1862 * @return string
1863 */
1864 function emphasize( $text ) {
1865 return "<em>$text</em>";
1866 }
1867
1868 /**
1869 * Normally we output all numbers in plain en_US style, that is
1870 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1871 * point twohundredthirtyfive. However this is not sutable for all
1872 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
1873 * Icelandic just want to use commas instead of dots, and dots instead
1874 * of commas like "293.291,235".
1875 *
1876 * An example of this function being called:
1877 * <code>
1878 * wfMsg( 'message', $wgLang->formatNum( $num ) )
1879 * </code>
1880 *
1881 * See LanguageGu.php for the Gujarati implementation and
1882 * $separatorTransformTable on MessageIs.php for
1883 * the , => . and . => , implementation.
1884 *
1885 * @todo check if it's viable to use localeconv() for the decimal
1886 * separator thing.
1887 * @param $number Mixed: the string to be formatted, should be an integer
1888 * or a floating point number.
1889 * @param $nocommafy Bool: set to true for special numbers like dates
1890 * @return string
1891 */
1892 function formatNum( $number, $nocommafy = false ) {
1893 global $wgTranslateNumerals;
1894 if (!$nocommafy) {
1895 $number = $this->commafy($number);
1896 $s = $this->separatorTransformTable();
1897 if ($s) { $number = strtr($number, $s); }
1898 }
1899
1900 if ($wgTranslateNumerals) {
1901 $s = $this->digitTransformTable();
1902 if ($s) { $number = strtr($number, $s); }
1903 }
1904
1905 return $number;
1906 }
1907
1908 function parseFormattedNumber( $number ) {
1909 $s = $this->digitTransformTable();
1910 if ($s) { $number = strtr($number, array_flip($s)); }
1911
1912 $s = $this->separatorTransformTable();
1913 if ($s) { $number = strtr($number, array_flip($s)); }
1914
1915 $number = strtr( $number, array (',' => '') );
1916 return $number;
1917 }
1918
1919 /**
1920 * Adds commas to a given number
1921 *
1922 * @param $_ mixed
1923 * @return string
1924 */
1925 function commafy($_) {
1926 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
1927 }
1928
1929 function digitTransformTable() {
1930 $this->load();
1931 return $this->digitTransformTable;
1932 }
1933
1934 function separatorTransformTable() {
1935 $this->load();
1936 return $this->separatorTransformTable;
1937 }
1938
1939
1940 /**
1941 * Take a list of strings and build a locale-friendly comma-separated
1942 * list, using the local comma-separator message.
1943 * The last two strings are chained with an "and".
1944 *
1945 * @param $l Array
1946 * @return string
1947 */
1948 function listToText( $l ) {
1949 $s = '';
1950 $m = count( $l ) - 1;
1951 if( $m == 1 ) {
1952 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
1953 }
1954 else {
1955 for ( $i = $m; $i >= 0; $i-- ) {
1956 if ( $i == $m ) {
1957 $s = $l[$i];
1958 } else if( $i == $m - 1 ) {
1959 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
1960 } else {
1961 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
1962 }
1963 }
1964 return $s;
1965 }
1966 }
1967
1968 /**
1969 * Take a list of strings and build a locale-friendly comma-separated
1970 * list, using the local comma-separator message.
1971 * @param $list array of strings to put in a comma list
1972 * @return string
1973 */
1974 function commaList( $list, $forContent = false ) {
1975 return implode(
1976 $list,
1977 wfMsgExt( 'comma-separator', array( 'escapenoentities', 'language' => $this ) ) );
1978 }
1979
1980 /**
1981 * Same as commaList, but separate it with the pipe instead.
1982 * @param $list array of strings to put in a pipe list
1983 * @return string
1984 */
1985 function pipeList( $list ) {
1986 return implode(
1987 $list,
1988 wfMsgExt( 'pipe-separator', array( 'escapenoentities', 'language' => $this ) ) );
1989 }
1990
1991 /**
1992 * Truncate a string to a specified length in bytes, appending an optional
1993 * string (e.g. for ellipses)
1994 *
1995 * The database offers limited byte lengths for some columns in the database;
1996 * multi-byte character sets mean we need to ensure that only whole characters
1997 * are included, otherwise broken characters can be passed to the user
1998 *
1999 * If $length is negative, the string will be truncated from the beginning
2000 *
2001 * @param $string String to truncate
2002 * @param $length Int: maximum length (excluding ellipses)
2003 * @param $ellipsis String to append to the truncated text
2004 * @return string
2005 */
2006 function truncate( $string, $length, $ellipsis = "" ) {
2007 if( $length == 0 ) {
2008 return $ellipsis;
2009 }
2010 if ( strlen( $string ) <= abs( $length ) ) {
2011 return $string;
2012 }
2013 if( $length > 0 ) {
2014 $string = substr( $string, 0, $length );
2015 $char = ord( $string[strlen( $string ) - 1] );
2016 $m = array();
2017 if ($char >= 0xc0) {
2018 # We got the first byte only of a multibyte char; remove it.
2019 $string = substr( $string, 0, -1 );
2020 } elseif( $char >= 0x80 &&
2021 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2022 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
2023 # We chopped in the middle of a character; remove it
2024 $string = $m[1];
2025 }
2026 return $string . $ellipsis;
2027 } else {
2028 $string = substr( $string, $length );
2029 $char = ord( $string[0] );
2030 if( $char >= 0x80 && $char < 0xc0 ) {
2031 # We chopped in the middle of a character; remove the whole thing
2032 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2033 }
2034 return $ellipsis . $string;
2035 }
2036 }
2037
2038 /**
2039 * Grammatical transformations, needed for inflected languages
2040 * Invoked by putting {{grammar:case|word}} in a message
2041 *
2042 * @param $word string
2043 * @param $case string
2044 * @return string
2045 */
2046 function convertGrammar( $word, $case ) {
2047 global $wgGrammarForms;
2048 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
2049 return $wgGrammarForms[$this->getCode()][$case][$word];
2050 }
2051 return $word;
2052 }
2053
2054 /**
2055 * Plural form transformations, needed for some languages.
2056 * For example, there are 3 form of plural in Russian and Polish,
2057 * depending on "count mod 10". See [[w:Plural]]
2058 * For English it is pretty simple.
2059 *
2060 * Invoked by putting {{plural:count|wordform1|wordform2}}
2061 * or {{plural:count|wordform1|wordform2|wordform3}}
2062 *
2063 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2064 *
2065 * @param $count Integer: non-localized number
2066 * @param $forms Array: different plural forms
2067 * @return string Correct form of plural for $count in this language
2068 */
2069 function convertPlural( $count, $forms ) {
2070 if ( !count($forms) ) { return ''; }
2071 $forms = $this->preConvertPlural( $forms, 2 );
2072
2073 return ( $count == 1 ) ? $forms[0] : $forms[1];
2074 }
2075
2076 /**
2077 * Checks that convertPlural was given an array and pads it to requested
2078 * amound of forms by copying the last one.
2079 *
2080 * @param $count Integer: How many forms should there be at least
2081 * @param $forms Array of forms given to convertPlural
2082 * @return array Padded array of forms or an exception if not an array
2083 */
2084 protected function preConvertPlural( /* Array */ $forms, $count ) {
2085 while ( count($forms) < $count ) {
2086 $forms[] = $forms[count($forms)-1];
2087 }
2088 return $forms;
2089 }
2090
2091 /**
2092 * For translaing of expiry times
2093 * @param $str String: the validated block time in English
2094 * @return Somehow translated block time
2095 * @see LanguageFi.php for example implementation
2096 */
2097 function translateBlockExpiry( $str ) {
2098
2099 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2100
2101 if ( $scBlockExpiryOptions == '-') {
2102 return $str;
2103 }
2104
2105 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2106 if ( strpos($option, ":") === false )
2107 continue;
2108 list($show, $value) = explode(":", $option);
2109 if ( strcmp ( $str, $value) == 0 ) {
2110 return htmlspecialchars( trim( $show ) );
2111 }
2112 }
2113
2114 return $str;
2115 }
2116
2117 /**
2118 * languages like Chinese need to be segmented in order for the diff
2119 * to be of any use
2120 *
2121 * @param $text String
2122 * @return String
2123 */
2124 function segmentForDiff( $text ) {
2125 return $text;
2126 }
2127
2128 /**
2129 * and unsegment to show the result
2130 *
2131 * @param $text String
2132 * @return String
2133 */
2134 function unsegmentForDiff( $text ) {
2135 return $text;
2136 }
2137
2138 # convert text to different variants of a language.
2139 function convert( $text, $isTitle = false) {
2140 return $this->mConverter->convert($text, $isTitle);
2141 }
2142
2143 # Convert text from within Parser
2144 function parserConvert( $text, &$parser ) {
2145 return $this->mConverter->parserConvert( $text, $parser );
2146 }
2147
2148 # Check if this is a language with variants
2149 function hasVariants(){
2150 return sizeof($this->getVariants())>1;
2151 }
2152
2153 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2154 function armourMath($text){
2155 return $this->mConverter->armourMath($text);
2156 }
2157
2158
2159 /**
2160 * Perform output conversion on a string, and encode for safe HTML output.
2161 * @param $text String
2162 * @param $isTitle Bool -- wtf?
2163 * @return string
2164 * @todo this should get integrated somewhere sane
2165 */
2166 function convertHtml( $text, $isTitle = false ) {
2167 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2168 }
2169
2170 function convertCategoryKey( $key ) {
2171 return $this->mConverter->convertCategoryKey( $key );
2172 }
2173
2174 /**
2175 * get the list of variants supported by this langauge
2176 * see sample implementation in LanguageZh.php
2177 *
2178 * @return array an array of language codes
2179 */
2180 function getVariants() {
2181 return $this->mConverter->getVariants();
2182 }
2183
2184
2185 function getPreferredVariant( $fromUser = true ) {
2186 return $this->mConverter->getPreferredVariant( $fromUser );
2187 }
2188
2189 /**
2190 * if a language supports multiple variants, it is
2191 * possible that non-existing link in one variant
2192 * actually exists in another variant. this function
2193 * tries to find it. See e.g. LanguageZh.php
2194 *
2195 * @param $link String: the name of the link
2196 * @param $nt Mixed: the title object of the link
2197 * @return null the input parameters may be modified upon return
2198 */
2199 function findVariantLink( &$link, &$nt, $forTemplate = false ) {
2200 $this->mConverter->findVariantLink($link, $nt, $forTemplate );
2201 }
2202
2203 /**
2204 * If a language supports multiple variants, converts text
2205 * into an array of all possible variants of the text:
2206 * 'variant' => text in that variant
2207 */
2208
2209 function convertLinkToAllVariants($text){
2210 return $this->mConverter->convertLinkToAllVariants($text);
2211 }
2212
2213
2214 /**
2215 * returns language specific options used by User::getPageRenderHash()
2216 * for example, the preferred language variant
2217 *
2218 * @return string
2219 */
2220 function getExtraHashOptions() {
2221 return $this->mConverter->getExtraHashOptions();
2222 }
2223
2224 /**
2225 * for languages that support multiple variants, the title of an
2226 * article may be displayed differently in different variants. this
2227 * function returns the apporiate title defined in the body of the article.
2228 *
2229 * @return string
2230 */
2231 function getParsedTitle() {
2232 return $this->mConverter->getParsedTitle();
2233 }
2234
2235 /**
2236 * Enclose a string with the "no conversion" tag. This is used by
2237 * various functions in the Parser
2238 *
2239 * @param $text String: text to be tagged for no conversion
2240 * @param $noParse
2241 * @return string the tagged text
2242 */
2243 function markNoConversion( $text, $noParse=false ) {
2244 return $this->mConverter->markNoConversion( $text, $noParse );
2245 }
2246
2247 /**
2248 * A regular expression to match legal word-trailing characters
2249 * which should be merged onto a link of the form [[foo]]bar.
2250 *
2251 * @return string
2252 */
2253 function linkTrail() {
2254 $this->load();
2255 return $this->linkTrail;
2256 }
2257
2258 function getLangObj() {
2259 return $this;
2260 }
2261
2262 /**
2263 * Get the RFC 3066 code for this language object
2264 */
2265 function getCode() {
2266 return $this->mCode;
2267 }
2268
2269 function setCode( $code ) {
2270 $this->mCode = $code;
2271 }
2272
2273 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2274 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2275 }
2276
2277 static function getMessagesFileName( $code ) {
2278 global $IP;
2279 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2280 }
2281
2282 static function getClassFileName( $code ) {
2283 global $IP;
2284 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2285 }
2286
2287 static function getLocalisationArray( $code, $disableCache = false ) {
2288 self::loadLocalisation( $code, $disableCache );
2289 return self::$mLocalisationCache[$code];
2290 }
2291
2292 /**
2293 * Load localisation data for a given code into the static cache
2294 *
2295 * @return array Dependencies, map of filenames to mtimes
2296 */
2297 static function loadLocalisation( $code, $disableCache = false ) {
2298 static $recursionGuard = array();
2299 global $wgMemc, $wgEnableSerializedMessages, $wgCheckSerialized;
2300
2301 if ( !$code ) {
2302 throw new MWException( "Invalid language code requested" );
2303 }
2304
2305 if ( !$disableCache ) {
2306 # Try the per-process cache
2307 if ( isset( self::$mLocalisationCache[$code] ) ) {
2308 return self::$mLocalisationCache[$code]['deps'];
2309 }
2310
2311 wfProfileIn( __METHOD__ );
2312
2313 # Try the serialized directory
2314 if( $wgEnableSerializedMessages ) {
2315 $cache = wfGetPrecompiledData( self::getFileName( "Messages", $code, '.ser' ) );
2316 if ( $cache ) {
2317 if ( $wgCheckSerialized && self::isLocalisationOutOfDate( $cache ) ) {
2318 $cache = false;
2319 wfDebug( "Language::loadLocalisation(): precompiled data file for $code is out of date\n" );
2320 } else {
2321 self::$mLocalisationCache[$code] = $cache;
2322 wfDebug( "Language::loadLocalisation(): got localisation for $code from precompiled data file\n" );
2323 wfProfileOut( __METHOD__ );
2324 return self::$mLocalisationCache[$code]['deps'];
2325 }
2326 }
2327 }
2328
2329 # Try the global cache
2330 $memcKey = wfMemcKey('localisation', $code );
2331 $fbMemcKey = wfMemcKey('fallback', $cache['fallback'] );
2332 $cache = $wgMemc->get( $memcKey );
2333 if ( $cache ) {
2334 if ( self::isLocalisationOutOfDate( $cache ) ) {
2335 $wgMemc->delete( $memcKey );
2336 $wgMemc->delete( $fbMemcKey );
2337 $cache = false;
2338 wfDebug( "Language::loadLocalisation(): localisation cache for $code had expired\n" );
2339 } else {
2340 self::$mLocalisationCache[$code] = $cache;
2341 wfDebug( "Language::loadLocalisation(): got localisation for $code from cache\n" );
2342 wfProfileOut( __METHOD__ );
2343 return $cache['deps'];
2344 }
2345 }
2346 } else {
2347 wfProfileIn( __METHOD__ );
2348 }
2349
2350 # Default fallback, may be overridden when the messages file is included
2351 if ( $code != 'en' ) {
2352 $fallback = 'en';
2353 } else {
2354 $fallback = false;
2355 }
2356
2357 # Load the primary localisation from the source file
2358 $filename = self::getMessagesFileName( $code );
2359 if ( !file_exists( $filename ) ) {
2360 wfDebug( "Language::loadLocalisation(): no localisation file for $code, using implicit fallback to en\n" );
2361 $cache = compact( self::$mLocalisationKeys ); // Set correct fallback
2362 $deps = array();
2363 } else {
2364 $deps = array( $filename => filemtime( $filename ) );
2365 require( $filename );
2366 $cache = compact( self::$mLocalisationKeys );
2367 wfDebug( "Language::loadLocalisation(): got localisation for $code from source\n" );
2368 }
2369
2370 if ( !empty( $fallback ) ) {
2371 # Load the fallback localisation, with a circular reference guard
2372 if ( isset( $recursionGuard[$code] ) ) {
2373 throw new MWException( "Error: Circular fallback reference in language code $code" );
2374 }
2375 $recursionGuard[$code] = true;
2376 $newDeps = self::loadLocalisation( $fallback, $disableCache );
2377 unset( $recursionGuard[$code] );
2378
2379 $secondary = self::$mLocalisationCache[$fallback];
2380 $deps = array_merge( $deps, $newDeps );
2381
2382 # Merge the fallback localisation with the current localisation
2383 foreach ( self::$mLocalisationKeys as $key ) {
2384 if ( isset( $cache[$key] ) ) {
2385 if ( isset( $secondary[$key] ) ) {
2386 if ( in_array( $key, self::$mMergeableMapKeys ) ) {
2387 $cache[$key] = $cache[$key] + $secondary[$key];
2388 } elseif ( in_array( $key, self::$mMergeableListKeys ) ) {
2389 $cache[$key] = array_merge( $secondary[$key], $cache[$key] );
2390 } elseif ( in_array( $key, self::$mMergeableAliasListKeys ) ) {
2391 $cache[$key] = array_merge_recursive( $cache[$key], $secondary[$key] );
2392 }
2393 }
2394 } else {
2395 $cache[$key] = $secondary[$key];
2396 }
2397 }
2398
2399 # Merge bookstore lists if requested
2400 if ( !empty( $cache['bookstoreList']['inherit'] ) ) {
2401 $cache['bookstoreList'] = array_merge( $cache['bookstoreList'], $secondary['bookstoreList'] );
2402 }
2403 if ( isset( $cache['bookstoreList']['inherit'] ) ) {
2404 unset( $cache['bookstoreList']['inherit'] );
2405 }
2406 }
2407
2408 # Add dependencies to the cache entry
2409 $cache['deps'] = $deps;
2410
2411 # Replace spaces with underscores in namespace names
2412 $cache['namespaceNames'] = str_replace( ' ', '_', $cache['namespaceNames'] );
2413
2414 # And do the same for specialpage aliases. $page is an array.
2415 foreach ( $cache['specialPageAliases'] as &$page ) {
2416 $page = str_replace( ' ', '_', $page );
2417 }
2418 # Decouple the reference to prevent accidental damage
2419 unset($page);
2420
2421 # Save to both caches
2422 self::$mLocalisationCache[$code] = $cache;
2423 if ( !$disableCache ) {
2424 $wgMemc->set( $memcKey, $cache );
2425 $wgMemc->set( $fbMemcKey, (string) $cache['fallback'] );
2426 }
2427
2428 wfProfileOut( __METHOD__ );
2429 return $deps;
2430 }
2431
2432 /**
2433 * Test if a given localisation cache is out of date with respect to the
2434 * source Messages files. This is done automatically for the global cache
2435 * in $wgMemc, but is only done on certain occasions for the serialized
2436 * data file.
2437 *
2438 * @param $cache mixed Either a language code or a cache array
2439 */
2440 static function isLocalisationOutOfDate( $cache ) {
2441 if ( !is_array( $cache ) ) {
2442 self::loadLocalisation( $cache );
2443 $cache = self::$mLocalisationCache[$cache];
2444 }
2445 $expired = false;
2446 foreach ( $cache['deps'] as $file => $mtime ) {
2447 if ( !file_exists( $file ) || filemtime( $file ) > $mtime ) {
2448 $expired = true;
2449 break;
2450 }
2451 }
2452 return $expired;
2453 }
2454
2455 /**
2456 * Get the fallback for a given language
2457 */
2458 static function getFallbackFor( $code ) {
2459 // Shortcut
2460 if ( $code === 'en' ) return false;
2461
2462 // Local cache
2463 static $cache = array();
2464 // Quick return
2465 if ( isset($cache[$code]) ) return $cache[$code];
2466
2467 // Try memcache
2468 global $wgMemc;
2469 $memcKey = wfMemcKey( 'fallback', $code );
2470 $fbcode = $wgMemc->get( $memcKey );
2471
2472 if ( is_string($fbcode) ) {
2473 // False is stored as a string to detect failures in memcache properly
2474 if ( $fbcode === '' ) $fbcode = false;
2475
2476 // Update local cache and return
2477 $cache[$code] = $fbcode;
2478 return $fbcode;
2479 }
2480
2481 // Nothing in caches, load and and update both caches
2482 self::loadLocalisation( $code );
2483 $fbcode = self::$mLocalisationCache[$code]['fallback'];
2484
2485 $cache[$code] = $fbcode;
2486 $wgMemc->set( $memcKey, (string) $fbcode );
2487
2488 return $fbcode;
2489 }
2490
2491 /**
2492 * Get all messages for a given language
2493 */
2494 static function getMessagesFor( $code ) {
2495 self::loadLocalisation( $code );
2496 return self::$mLocalisationCache[$code]['messages'];
2497 }
2498
2499 /**
2500 * Get a message for a given language
2501 */
2502 static function getMessageFor( $key, $code ) {
2503 self::loadLocalisation( $code );
2504 return isset( self::$mLocalisationCache[$code]['messages'][$key] ) ? self::$mLocalisationCache[$code]['messages'][$key] : null;
2505 }
2506
2507 /**
2508 * Load localisation data for this object
2509 */
2510 function load() {
2511 if ( !$this->mLoaded ) {
2512 self::loadLocalisation( $this->getCode() );
2513 $cache =& self::$mLocalisationCache[$this->getCode()];
2514 foreach ( self::$mLocalisationKeys as $key ) {
2515 $this->$key = $cache[$key];
2516 }
2517 $this->mLoaded = true;
2518
2519 $this->fixUpSettings();
2520 }
2521 }
2522
2523 /**
2524 * Do any necessary post-cache-load settings adjustment
2525 */
2526 function fixUpSettings() {
2527 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk,
2528 $wgNamespaceAliases, $wgAmericanDates;
2529 wfProfileIn( __METHOD__ );
2530 if ( $wgExtraNamespaces ) {
2531 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
2532 }
2533
2534 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
2535 if ( $wgMetaNamespaceTalk ) {
2536 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
2537 } else {
2538 $talk = $this->namespaceNames[NS_PROJECT_TALK];
2539 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2540
2541 # Allow grammar transformations
2542 # Allowing full message-style parsing would make simple requests
2543 # such as action=raw much more expensive than they need to be.
2544 # This will hopefully cover most cases.
2545 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2546 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2547 $talk = str_replace( ' ', '_', $talk );
2548 $this->namespaceNames[NS_PROJECT_TALK] = $talk;
2549 }
2550
2551 # The above mixing may leave namespaces out of canonical order.
2552 # Re-order by namespace ID number...
2553 ksort( $this->namespaceNames );
2554
2555 # Put namespace names and aliases into a hashtable.
2556 # If this is too slow, then we should arrange it so that it is done
2557 # before caching. The catch is that at pre-cache time, the above
2558 # class-specific fixup hasn't been done.
2559 $this->mNamespaceIds = array();
2560 foreach ( $this->namespaceNames as $index => $name ) {
2561 $this->mNamespaceIds[$this->lc($name)] = $index;
2562 }
2563 if ( $this->namespaceAliases ) {
2564 foreach ( $this->namespaceAliases as $name => $index ) {
2565 $this->mNamespaceIds[$this->lc($name)] = $index;
2566 }
2567 }
2568 if ( $wgNamespaceAliases ) {
2569 foreach ( $wgNamespaceAliases as $name => $index ) {
2570 $this->mNamespaceIds[$this->lc($name)] = $index;
2571 }
2572 }
2573
2574 if ( $this->defaultDateFormat == 'dmy or mdy' ) {
2575 $this->defaultDateFormat = $wgAmericanDates ? 'mdy' : 'dmy';
2576 }
2577 wfProfileOut( __METHOD__ );
2578 }
2579
2580 function replaceGrammarInNamespace( $m ) {
2581 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2582 }
2583
2584 static function getCaseMaps() {
2585 static $wikiUpperChars, $wikiLowerChars;
2586 if ( isset( $wikiUpperChars ) ) {
2587 return array( $wikiUpperChars, $wikiLowerChars );
2588 }
2589
2590 wfProfileIn( __METHOD__ );
2591 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2592 if ( $arr === false ) {
2593 throw new MWException(
2594 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2595 }
2596 extract( $arr );
2597 wfProfileOut( __METHOD__ );
2598 return array( $wikiUpperChars, $wikiLowerChars );
2599 }
2600
2601 function formatTimePeriod( $seconds ) {
2602 if ( $seconds < 10 ) {
2603 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2604 } elseif ( $seconds < 60 ) {
2605 return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2606 } elseif ( $seconds < 3600 ) {
2607 return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2608 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2609 } else {
2610 $hours = floor( $seconds / 3600 );
2611 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2612 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2613 return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2614 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2615 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2616 }
2617 }
2618
2619 function formatBitrate( $bps ) {
2620 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2621 if ( $bps <= 0 ) {
2622 return $this->formatNum( $bps ) . $units[0];
2623 }
2624 $unitIndex = floor( log10( $bps ) / 3 );
2625 $mantissa = $bps / pow( 1000, $unitIndex );
2626 if ( $mantissa < 10 ) {
2627 $mantissa = round( $mantissa, 1 );
2628 } else {
2629 $mantissa = round( $mantissa );
2630 }
2631 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2632 }
2633
2634 /**
2635 * Format a size in bytes for output, using an appropriate
2636 * unit (B, KB, MB or GB) according to the magnitude in question
2637 *
2638 * @param $size Size to format
2639 * @return string Plain text (not HTML)
2640 */
2641 function formatSize( $size ) {
2642 // For small sizes no decimal places necessary
2643 $round = 0;
2644 if( $size > 1024 ) {
2645 $size = $size / 1024;
2646 if( $size > 1024 ) {
2647 $size = $size / 1024;
2648 // For MB and bigger two decimal places are smarter
2649 $round = 2;
2650 if( $size > 1024 ) {
2651 $size = $size / 1024;
2652 $msg = 'size-gigabytes';
2653 } else {
2654 $msg = 'size-megabytes';
2655 }
2656 } else {
2657 $msg = 'size-kilobytes';
2658 }
2659 } else {
2660 $msg = 'size-bytes';
2661 }
2662 $size = round( $size, $round );
2663 $text = $this->getMessageFromDB( $msg );
2664 return str_replace( '$1', $this->formatNum( $size ), $text );
2665 }
2666 }