dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Fixed syntax errors in message get function
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * Internationalisation code
4 *
5 * @file
6 * @ingroup Language
7 */
8
9 /**
10 * @defgroup Language Language
11 */
12
13 if ( !defined( 'MEDIAWIKI' ) ) {
14 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
15 exit( 1 );
16 }
17
18 # Read language names
19 global $wgLanguageNames;
20 require_once( dirname( __FILE__ ) . '/Names.php' );
21
22 global $wgInputEncoding, $wgOutputEncoding;
23
24 /**
25 * These are always UTF-8, they exist only for backwards compatibility
26 */
27 $wgInputEncoding = 'UTF-8';
28 $wgOutputEncoding = 'UTF-8';
29
30 if ( function_exists( 'mb_strtoupper' ) ) {
31 mb_internal_encoding( 'UTF-8' );
32 }
33
34 /**
35 * a fake language converter
36 *
37 * @ingroup Language
38 */
39 class FakeConverter {
40 var $mLang;
41 function __construct( $langobj ) { $this->mLang = $langobj; }
42 function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
43 function convert( $t ) { return $t; }
44 function convertTitle( $t ) { return $t->getPrefixedText(); }
45 function getVariants() { return array( $this->mLang->getCode() ); }
46 function getPreferredVariant() { return $this->mLang->getCode(); }
47 function getConvRuleTitle() { return false; }
48 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
49 function getExtraHashOptions() { return ''; }
50 function getParsedTitle() { return ''; }
51 function markNoConversion( $text, $noParse = false ) { return $text; }
52 function convertCategoryKey( $key ) { return $key; }
53 function convertLinkToAllVariants( $text ) { return autoConvertToAllVariants( $text ); }
54 function armourMath( $text ) { return $text; }
55 }
56
57 /**
58 * Internationalisation code
59 * @ingroup Language
60 */
61 class Language {
62 var $mConverter, $mVariants, $mCode, $mLoaded = false;
63 var $mMagicExtensions = array(), $mMagicHookDone = false;
64
65 var $mNamespaceIds, $namespaceNames, $namespaceAliases;
66 var $dateFormatStrings = array();
67 var $mExtendedSpecialPageAliases;
68
69 /**
70 * ReplacementArray object caches
71 */
72 var $transformData = array();
73
74 static public $dataCache;
75 static public $mLangObjCache = array();
76
77 static public $mWeekdayMsgs = array(
78 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
79 'friday', 'saturday'
80 );
81
82 static public $mWeekdayAbbrevMsgs = array(
83 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
84 );
85
86 static public $mMonthMsgs = array(
87 'january', 'february', 'march', 'april', 'may_long', 'june',
88 'july', 'august', 'september', 'october', 'november',
89 'december'
90 );
91 static public $mMonthGenMsgs = array(
92 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
93 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
94 'december-gen'
95 );
96 static public $mMonthAbbrevMsgs = array(
97 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
98 'sep', 'oct', 'nov', 'dec'
99 );
100
101 static public $mIranianCalendarMonthMsgs = array(
102 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
103 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
104 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
105 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
106 );
107
108 static public $mHebrewCalendarMonthMsgs = array(
109 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
110 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
111 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
112 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
113 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
114 );
115
116 static public $mHebrewCalendarMonthGenMsgs = array(
117 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
118 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
119 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
120 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
121 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
122 );
123
124 static public $mHijriCalendarMonthMsgs = array(
125 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
126 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
127 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
128 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
129 );
130
131 /**
132 * Get a cached language object for a given language code
133 */
134 static function factory( $code ) {
135 if ( !isset( self::$mLangObjCache[$code] ) ) {
136 if ( count( self::$mLangObjCache ) > 10 ) {
137 // Don't keep a billion objects around, that's stupid.
138 self::$mLangObjCache = array();
139 }
140 self::$mLangObjCache[$code] = self::newFromCode( $code );
141 }
142 return self::$mLangObjCache[$code];
143 }
144
145 /**
146 * Create a language object for a given language code
147 */
148 protected static function newFromCode( $code ) {
149 global $IP;
150 static $recursionLevel = 0;
151 if ( $code == 'en' ) {
152 $class = 'Language';
153 } else {
154 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
155 // Preload base classes to work around APC/PHP5 bug
156 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
157 include_once( "$IP/languages/classes/$class.deps.php" );
158 }
159 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
160 include_once( "$IP/languages/classes/$class.php" );
161 }
162 }
163
164 if ( $recursionLevel > 5 ) {
165 throw new MWException( "Language fallback loop detected when creating class $class\n" );
166 }
167
168 if ( !class_exists( $class ) ) {
169 $fallback = Language::getFallbackFor( $code );
170 ++$recursionLevel;
171 $lang = Language::newFromCode( $fallback );
172 --$recursionLevel;
173 $lang->setCode( $code );
174 } else {
175 $lang = new $class;
176 }
177 return $lang;
178 }
179
180 /**
181 * Get the LocalisationCache instance
182 */
183 public static function getLocalisationCache() {
184 if ( is_null( self::$dataCache ) ) {
185 global $wgLocalisationCacheConf;
186 $class = $wgLocalisationCacheConf['class'];
187 self::$dataCache = new $class( $wgLocalisationCacheConf );
188 }
189 return self::$dataCache;
190 }
191
192 function __construct() {
193 $this->mConverter = new FakeConverter( $this );
194 // Set the code to the name of the descendant
195 if ( get_class( $this ) == 'Language' ) {
196 $this->mCode = 'en';
197 } else {
198 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
199 }
200 self::getLocalisationCache();
201 }
202
203 /**
204 * Reduce memory usage
205 */
206 function __destruct() {
207 foreach ( $this as $name => $value ) {
208 unset( $this->$name );
209 }
210 }
211
212 /**
213 * Hook which will be called if this is the content language.
214 * Descendants can use this to register hook functions or modify globals
215 */
216 function initContLang() { }
217
218 /**
219 * @deprecated Use User::getDefaultOptions()
220 * @return array
221 */
222 function getDefaultUserOptions() {
223 wfDeprecated( __METHOD__ );
224 return User::getDefaultOptions();
225 }
226
227 function getFallbackLanguageCode() {
228 if ( $this->mCode === 'en' ) {
229 return false;
230 } else {
231 return self::$dataCache->getItem( $this->mCode, 'fallback' );
232 }
233 }
234
235 /**
236 * Exports $wgBookstoreListEn
237 * @return array
238 */
239 function getBookstoreList() {
240 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
241 }
242
243 /**
244 * @return array
245 */
246 function getNamespaces() {
247 if ( is_null( $this->namespaceNames ) ) {
248 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
249
250 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
251 $validNamespaces = MWNamespace::getCanonicalNamespaces();
252
253 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
254
255 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
256 if ( $wgMetaNamespaceTalk ) {
257 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
258 } else {
259 $talk = $this->namespaceNames[NS_PROJECT_TALK];
260 $this->namespaceNames[NS_PROJECT_TALK] =
261 $this->fixVariableInNamespace( $talk );
262 }
263
264 # Sometimes a language will be localised but not actually exist on this wiki.
265 foreach( $this->namespaceNames as $key => $text ) {
266 if ( !isset( $validNamespaces[$key] ) ) {
267 unset( $this->namespaceNames[$key] );
268 }
269 }
270
271 # The above mixing may leave namespaces out of canonical order.
272 # Re-order by namespace ID number...
273 ksort( $this->namespaceNames );
274 }
275 return $this->namespaceNames;
276 }
277
278 /**
279 * A convenience function that returns the same thing as
280 * getNamespaces() except with the array values changed to ' '
281 * where it found '_', useful for producing output to be displayed
282 * e.g. in <select> forms.
283 *
284 * @return array
285 */
286 function getFormattedNamespaces() {
287 $ns = $this->getNamespaces();
288 foreach ( $ns as $k => $v ) {
289 $ns[$k] = strtr( $v, '_', ' ' );
290 }
291 return $ns;
292 }
293
294 /**
295 * Get a namespace value by key
296 * <code>
297 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
298 * echo $mw_ns; // prints 'MediaWiki'
299 * </code>
300 *
301 * @param $index Int: the array key of the namespace to return
302 * @return mixed, string if the namespace value exists, otherwise false
303 */
304 function getNsText( $index ) {
305 $ns = $this->getNamespaces();
306 return isset( $ns[$index] ) ? $ns[$index] : false;
307 }
308
309 /**
310 * A convenience function that returns the same thing as
311 * getNsText() except with '_' changed to ' ', useful for
312 * producing output.
313 *
314 * @return array
315 */
316 function getFormattedNsText( $index ) {
317 $ns = $this->getNsText( $index );
318 return strtr( $ns, '_', ' ' );
319 }
320
321 /**
322 * Get a namespace key by value, case insensitive.
323 * Only matches namespace names for the current language, not the
324 * canonical ones defined in Namespace.php.
325 *
326 * @param $text String
327 * @return mixed An integer if $text is a valid value otherwise false
328 */
329 function getLocalNsIndex( $text ) {
330 $lctext = $this->lc( $text );
331 $ids = $this->getNamespaceIds();
332 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
333 }
334
335 function getNamespaceAliases() {
336 if ( is_null( $this->namespaceAliases ) ) {
337 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
338 if ( !$aliases ) {
339 $aliases = array();
340 } else {
341 foreach ( $aliases as $name => $index ) {
342 if ( $index === NS_PROJECT_TALK ) {
343 unset( $aliases[$name] );
344 $name = $this->fixVariableInNamespace( $name );
345 $aliases[$name] = $index;
346 }
347 }
348 }
349 $this->namespaceAliases = $aliases;
350 }
351 return $this->namespaceAliases;
352 }
353
354 function getNamespaceIds() {
355 if ( is_null( $this->mNamespaceIds ) ) {
356 global $wgNamespaceAliases;
357 # Put namespace names and aliases into a hashtable.
358 # If this is too slow, then we should arrange it so that it is done
359 # before caching. The catch is that at pre-cache time, the above
360 # class-specific fixup hasn't been done.
361 $this->mNamespaceIds = array();
362 foreach ( $this->getNamespaces() as $index => $name ) {
363 $this->mNamespaceIds[$this->lc( $name )] = $index;
364 }
365 foreach ( $this->getNamespaceAliases() as $name => $index ) {
366 $this->mNamespaceIds[$this->lc( $name )] = $index;
367 }
368 if ( $wgNamespaceAliases ) {
369 foreach ( $wgNamespaceAliases as $name => $index ) {
370 $this->mNamespaceIds[$this->lc( $name )] = $index;
371 }
372 }
373 }
374 return $this->mNamespaceIds;
375 }
376
377
378 /**
379 * Get a namespace key by value, case insensitive. Canonical namespace
380 * names override custom ones defined for the current language.
381 *
382 * @param $text String
383 * @return mixed An integer if $text is a valid value otherwise false
384 */
385 function getNsIndex( $text ) {
386 $lctext = $this->lc( $text );
387 if ( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) {
388 return $ns;
389 }
390 $ids = $this->getNamespaceIds();
391 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
392 }
393
394 /**
395 * short names for language variants used for language conversion links.
396 *
397 * @param $code String
398 * @return string
399 */
400 function getVariantname( $code ) {
401 return $this->getMessageFromDB( "variantname-$code" );
402 }
403
404 function specialPage( $name ) {
405 $aliases = $this->getSpecialPageAliases();
406 if ( isset( $aliases[$name][0] ) ) {
407 $name = $aliases[$name][0];
408 }
409 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
410 }
411
412 function getQuickbarSettings() {
413 return array(
414 $this->getMessage( 'qbsettings-none' ),
415 $this->getMessage( 'qbsettings-fixedleft' ),
416 $this->getMessage( 'qbsettings-fixedright' ),
417 $this->getMessage( 'qbsettings-floatingleft' ),
418 $this->getMessage( 'qbsettings-floatingright' )
419 );
420 }
421
422 function getMathNames() {
423 return self::$dataCache->getItem( $this->mCode, 'mathNames' );
424 }
425
426 function getDatePreferences() {
427 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
428 }
429
430 function getDateFormats() {
431 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
432 }
433
434 function getDefaultDateFormat() {
435 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
436 if ( $df === 'dmy or mdy' ) {
437 global $wgAmericanDates;
438 return $wgAmericanDates ? 'mdy' : 'dmy';
439 } else {
440 return $df;
441 }
442 }
443
444 function getDatePreferenceMigrationMap() {
445 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
446 }
447
448 function getImageFile( $image ) {
449 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
450 }
451
452 function getDefaultUserOptionOverrides() {
453 return self::$dataCache->getItem( $this->mCode, 'defaultUserOptionOverrides' );
454 }
455
456 function getExtraUserToggles() {
457 return self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
458 }
459
460 function getUserToggle( $tog ) {
461 return $this->getMessageFromDB( "tog-$tog" );
462 }
463
464 /**
465 * Get language names, indexed by code.
466 * If $customisedOnly is true, only returns codes with a messages file
467 */
468 public static function getLanguageNames( $customisedOnly = false ) {
469 global $wgLanguageNames, $wgExtraLanguageNames;
470 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
471 if ( !$customisedOnly ) {
472 return $allNames;
473 }
474
475 global $IP;
476 $names = array();
477 $dir = opendir( "$IP/languages/messages" );
478 while ( false !== ( $file = readdir( $dir ) ) ) {
479 $code = self::getCodeFromFileName( $file, 'Messages' );
480 if ( $code && isset( $allNames[$code] ) ) {
481 $names[$code] = $allNames[$code];
482 }
483 }
484 closedir( $dir );
485 return $names;
486 }
487
488 /**
489 * Get a message from the MediaWiki namespace.
490 *
491 * @param $msg String: message name
492 * @return string
493 */
494 function getMessageFromDB( $msg ) {
495 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
496 }
497
498 function getLanguageName( $code ) {
499 $names = self::getLanguageNames();
500 if ( !array_key_exists( $code, $names ) ) {
501 return '';
502 }
503 return $names[$code];
504 }
505
506 function getMonthName( $key ) {
507 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
508 }
509
510 function getMonthNameGen( $key ) {
511 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
512 }
513
514 function getMonthAbbreviation( $key ) {
515 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
516 }
517
518 function getWeekdayName( $key ) {
519 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
520 }
521
522 function getWeekdayAbbreviation( $key ) {
523 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
524 }
525
526 function getIranianCalendarMonthName( $key ) {
527 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
528 }
529
530 function getHebrewCalendarMonthName( $key ) {
531 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
532 }
533
534 function getHebrewCalendarMonthNameGen( $key ) {
535 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
536 }
537
538 function getHijriCalendarMonthName( $key ) {
539 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
540 }
541
542 /**
543 * Used by date() and time() to adjust the time output.
544 *
545 * @param $ts Int the time in date('YmdHis') format
546 * @param $tz Mixed: adjust the time by this amount (default false, mean we
547 * get user timecorrection setting)
548 * @return int
549 */
550 function userAdjust( $ts, $tz = false ) {
551 global $wgUser, $wgLocalTZoffset;
552
553 if ( $tz === false ) {
554 $tz = $wgUser->getOption( 'timecorrection' );
555 }
556
557 $data = explode( '|', $tz, 3 );
558
559 if ( $data[0] == 'ZoneInfo' ) {
560 if ( function_exists( 'timezone_open' ) && @timezone_open( $data[2] ) !== false ) {
561 $date = date_create( $ts, timezone_open( 'UTC' ) );
562 date_timezone_set( $date, timezone_open( $data[2] ) );
563 $date = date_format( $date, 'YmdHis' );
564 return $date;
565 }
566 # Unrecognized timezone, default to 'Offset' with the stored offset.
567 $data[0] = 'Offset';
568 }
569
570 $minDiff = 0;
571 if ( $data[0] == 'System' || $tz == '' ) {
572 #  Global offset in minutes.
573 if ( isset( $wgLocalTZoffset ) ) {
574 $minDiff = $wgLocalTZoffset;
575 }
576 } else if ( $data[0] == 'Offset' ) {
577 $minDiff = intval( $data[1] );
578 } else {
579 $data = explode( ':', $tz );
580 if ( count( $data ) == 2 ) {
581 $data[0] = intval( $data[0] );
582 $data[1] = intval( $data[1] );
583 $minDiff = abs( $data[0] ) * 60 + $data[1];
584 if ( $data[0] < 0 ) {
585 $minDiff = -$minDiff;
586 }
587 } else {
588 $minDiff = intval( $data[0] ) * 60;
589 }
590 }
591
592 # No difference ? Return time unchanged
593 if ( 0 == $minDiff ) {
594 return $ts;
595 }
596
597 wfSuppressWarnings(); // E_STRICT system time bitching
598 # Generate an adjusted date; take advantage of the fact that mktime
599 # will normalize out-of-range values so we don't have to split $minDiff
600 # into hours and minutes.
601 $t = mktime( (
602 (int)substr( $ts, 8, 2 ) ), # Hours
603 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
604 (int)substr( $ts, 12, 2 ), # Seconds
605 (int)substr( $ts, 4, 2 ), # Month
606 (int)substr( $ts, 6, 2 ), # Day
607 (int)substr( $ts, 0, 4 ) ); # Year
608
609 $date = date( 'YmdHis', $t );
610 wfRestoreWarnings();
611
612 return $date;
613 }
614
615 /**
616 * This is a workalike of PHP's date() function, but with better
617 * internationalisation, a reduced set of format characters, and a better
618 * escaping format.
619 *
620 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
621 * PHP manual for definitions. "o" format character is supported since
622 * PHP 5.1.0, previous versions return literal o.
623 * There are a number of extensions, which start with "x":
624 *
625 * xn Do not translate digits of the next numeric format character
626 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
627 * xr Use roman numerals for the next numeric format character
628 * xh Use hebrew numerals for the next numeric format character
629 * xx Literal x
630 * xg Genitive month name
631 *
632 * xij j (day number) in Iranian calendar
633 * xiF F (month name) in Iranian calendar
634 * xin n (month number) in Iranian calendar
635 * xiY Y (full year) in Iranian calendar
636 *
637 * xjj j (day number) in Hebrew calendar
638 * xjF F (month name) in Hebrew calendar
639 * xjt t (days in month) in Hebrew calendar
640 * xjx xg (genitive month name) in Hebrew calendar
641 * xjn n (month number) in Hebrew calendar
642 * xjY Y (full year) in Hebrew calendar
643 *
644 * xmj j (day number) in Hijri calendar
645 * xmF F (month name) in Hijri calendar
646 * xmn n (month number) in Hijri calendar
647 * xmY Y (full year) in Hijri calendar
648 *
649 * xkY Y (full year) in Thai solar calendar. Months and days are
650 * identical to the Gregorian calendar
651 * xoY Y (full year) in Minguo calendar or Juche year.
652 * Months and days are identical to the
653 * Gregorian calendar
654 * xtY Y (full year) in Japanese nengo. Months and days are
655 * identical to the Gregorian calendar
656 *
657 * Characters enclosed in double quotes will be considered literal (with
658 * the quotes themselves removed). Unmatched quotes will be considered
659 * literal quotes. Example:
660 *
661 * "The month is" F => The month is January
662 * i's" => 20'11"
663 *
664 * Backslash escaping is also supported.
665 *
666 * Input timestamp is assumed to be pre-normalized to the desired local
667 * time zone, if any.
668 *
669 * @param $format String
670 * @param $ts String: 14-character timestamp
671 * YYYYMMDDHHMMSS
672 * 01234567890123
673 * @todo emulation of "o" format character for PHP pre 5.1.0
674 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
675 */
676 function sprintfDate( $format, $ts ) {
677 $s = '';
678 $raw = false;
679 $roman = false;
680 $hebrewNum = false;
681 $unix = false;
682 $rawToggle = false;
683 $iranian = false;
684 $hebrew = false;
685 $hijri = false;
686 $thai = false;
687 $minguo = false;
688 $tenno = false;
689 for ( $p = 0; $p < strlen( $format ); $p++ ) {
690 $num = false;
691 $code = $format[$p];
692 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
693 $code .= $format[++$p];
694 }
695
696 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
697 $code .= $format[++$p];
698 }
699
700 switch ( $code ) {
701 case 'xx':
702 $s .= 'x';
703 break;
704 case 'xn':
705 $raw = true;
706 break;
707 case 'xN':
708 $rawToggle = !$rawToggle;
709 break;
710 case 'xr':
711 $roman = true;
712 break;
713 case 'xh':
714 $hebrewNum = true;
715 break;
716 case 'xg':
717 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
718 break;
719 case 'xjx':
720 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
721 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
722 break;
723 case 'd':
724 $num = substr( $ts, 6, 2 );
725 break;
726 case 'D':
727 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
728 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
729 break;
730 case 'j':
731 $num = intval( substr( $ts, 6, 2 ) );
732 break;
733 case 'xij':
734 if ( !$iranian ) {
735 $iranian = self::tsToIranian( $ts );
736 }
737 $num = $iranian[2];
738 break;
739 case 'xmj':
740 if ( !$hijri ) {
741 $hijri = self::tsToHijri( $ts );
742 }
743 $num = $hijri[2];
744 break;
745 case 'xjj':
746 if ( !$hebrew ) {
747 $hebrew = self::tsToHebrew( $ts );
748 }
749 $num = $hebrew[2];
750 break;
751 case 'l':
752 if ( !$unix ) {
753 $unix = wfTimestamp( TS_UNIX, $ts );
754 }
755 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
756 break;
757 case 'N':
758 if ( !$unix ) {
759 $unix = wfTimestamp( TS_UNIX, $ts );
760 }
761 $w = gmdate( 'w', $unix );
762 $num = $w ? $w : 7;
763 break;
764 case 'w':
765 if ( !$unix ) {
766 $unix = wfTimestamp( TS_UNIX, $ts );
767 }
768 $num = gmdate( 'w', $unix );
769 break;
770 case 'z':
771 if ( !$unix ) {
772 $unix = wfTimestamp( TS_UNIX, $ts );
773 }
774 $num = gmdate( 'z', $unix );
775 break;
776 case 'W':
777 if ( !$unix ) {
778 $unix = wfTimestamp( TS_UNIX, $ts );
779 }
780 $num = gmdate( 'W', $unix );
781 break;
782 case 'F':
783 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
784 break;
785 case 'xiF':
786 if ( !$iranian ) {
787 $iranian = self::tsToIranian( $ts );
788 }
789 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
790 break;
791 case 'xmF':
792 if ( !$hijri ) {
793 $hijri = self::tsToHijri( $ts );
794 }
795 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
796 break;
797 case 'xjF':
798 if ( !$hebrew ) {
799 $hebrew = self::tsToHebrew( $ts );
800 }
801 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
802 break;
803 case 'm':
804 $num = substr( $ts, 4, 2 );
805 break;
806 case 'M':
807 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
808 break;
809 case 'n':
810 $num = intval( substr( $ts, 4, 2 ) );
811 break;
812 case 'xin':
813 if ( !$iranian ) {
814 $iranian = self::tsToIranian( $ts );
815 }
816 $num = $iranian[1];
817 break;
818 case 'xmn':
819 if ( !$hijri ) {
820 $hijri = self::tsToHijri ( $ts );
821 }
822 $num = $hijri[1];
823 break;
824 case 'xjn':
825 if ( !$hebrew ) {
826 $hebrew = self::tsToHebrew( $ts );
827 }
828 $num = $hebrew[1];
829 break;
830 case 't':
831 if ( !$unix ) {
832 $unix = wfTimestamp( TS_UNIX, $ts );
833 }
834 $num = gmdate( 't', $unix );
835 break;
836 case 'xjt':
837 if ( !$hebrew ) {
838 $hebrew = self::tsToHebrew( $ts );
839 }
840 $num = $hebrew[3];
841 break;
842 case 'L':
843 if ( !$unix ) {
844 $unix = wfTimestamp( TS_UNIX, $ts );
845 }
846 $num = gmdate( 'L', $unix );
847 break;
848 # 'o' is supported since PHP 5.1.0
849 # return literal if not supported
850 # TODO: emulation for pre 5.1.0 versions
851 case 'o':
852 if ( !$unix ) {
853 $unix = wfTimestamp( TS_UNIX, $ts );
854 }
855 if ( version_compare( PHP_VERSION, '5.1.0' ) === 1 ) {
856 $num = date( 'o', $unix );
857 } else {
858 $s .= 'o';
859 }
860 break;
861 case 'Y':
862 $num = substr( $ts, 0, 4 );
863 break;
864 case 'xiY':
865 if ( !$iranian ) {
866 $iranian = self::tsToIranian( $ts );
867 }
868 $num = $iranian[0];
869 break;
870 case 'xmY':
871 if ( !$hijri ) {
872 $hijri = self::tsToHijri( $ts );
873 }
874 $num = $hijri[0];
875 break;
876 case 'xjY':
877 if ( !$hebrew ) {
878 $hebrew = self::tsToHebrew( $ts );
879 }
880 $num = $hebrew[0];
881 break;
882 case 'xkY':
883 if ( !$thai ) {
884 $thai = self::tsToYear( $ts, 'thai' );
885 }
886 $num = $thai[0];
887 break;
888 case 'xoY':
889 if ( !$minguo ) {
890 $minguo = self::tsToYear( $ts, 'minguo' );
891 }
892 $num = $minguo[0];
893 break;
894 case 'xtY':
895 if ( !$tenno ) {
896 $tenno = self::tsToYear( $ts, 'tenno' );
897 }
898 $num = $tenno[0];
899 break;
900 case 'y':
901 $num = substr( $ts, 2, 2 );
902 break;
903 case 'a':
904 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
905 break;
906 case 'A':
907 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
908 break;
909 case 'g':
910 $h = substr( $ts, 8, 2 );
911 $num = $h % 12 ? $h % 12 : 12;
912 break;
913 case 'G':
914 $num = intval( substr( $ts, 8, 2 ) );
915 break;
916 case 'h':
917 $h = substr( $ts, 8, 2 );
918 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
919 break;
920 case 'H':
921 $num = substr( $ts, 8, 2 );
922 break;
923 case 'i':
924 $num = substr( $ts, 10, 2 );
925 break;
926 case 's':
927 $num = substr( $ts, 12, 2 );
928 break;
929 case 'c':
930 if ( !$unix ) {
931 $unix = wfTimestamp( TS_UNIX, $ts );
932 }
933 $s .= gmdate( 'c', $unix );
934 break;
935 case 'r':
936 if ( !$unix ) {
937 $unix = wfTimestamp( TS_UNIX, $ts );
938 }
939 $s .= gmdate( 'r', $unix );
940 break;
941 case 'U':
942 if ( !$unix ) {
943 $unix = wfTimestamp( TS_UNIX, $ts );
944 }
945 $num = $unix;
946 break;
947 case '\\':
948 # Backslash escaping
949 if ( $p < strlen( $format ) - 1 ) {
950 $s .= $format[++$p];
951 } else {
952 $s .= '\\';
953 }
954 break;
955 case '"':
956 # Quoted literal
957 if ( $p < strlen( $format ) - 1 ) {
958 $endQuote = strpos( $format, '"', $p + 1 );
959 if ( $endQuote === false ) {
960 # No terminating quote, assume literal "
961 $s .= '"';
962 } else {
963 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
964 $p = $endQuote;
965 }
966 } else {
967 # Quote at end of string, assume literal "
968 $s .= '"';
969 }
970 break;
971 default:
972 $s .= $format[$p];
973 }
974 if ( $num !== false ) {
975 if ( $rawToggle || $raw ) {
976 $s .= $num;
977 $raw = false;
978 } elseif ( $roman ) {
979 $s .= self::romanNumeral( $num );
980 $roman = false;
981 } elseif ( $hebrewNum ) {
982 $s .= self::hebrewNumeral( $num );
983 $hebrewNum = false;
984 } else {
985 $s .= $this->formatNum( $num, true );
986 }
987 $num = false;
988 }
989 }
990 return $s;
991 }
992
993 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
994 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
995 /**
996 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
997 * Gregorian dates to Iranian dates. Originally written in C, it
998 * is released under the terms of GNU Lesser General Public
999 * License. Conversion to PHP was performed by Niklas Laxström.
1000 *
1001 * Link: http://www.farsiweb.info/jalali/jalali.c
1002 */
1003 private static function tsToIranian( $ts ) {
1004 $gy = substr( $ts, 0, 4 ) -1600;
1005 $gm = substr( $ts, 4, 2 ) -1;
1006 $gd = substr( $ts, 6, 2 ) -1;
1007
1008 # Days passed from the beginning (including leap years)
1009 $gDayNo = 365 * $gy
1010 + floor( ( $gy + 3 ) / 4 )
1011 - floor( ( $gy + 99 ) / 100 )
1012 + floor( ( $gy + 399 ) / 400 );
1013
1014
1015 // Add days of the past months of this year
1016 for ( $i = 0; $i < $gm; $i++ ) {
1017 $gDayNo += self::$GREG_DAYS[$i];
1018 }
1019
1020 // Leap years
1021 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1022 $gDayNo++;
1023 }
1024
1025 // Days passed in current month
1026 $gDayNo += $gd;
1027
1028 $jDayNo = $gDayNo - 79;
1029
1030 $jNp = floor( $jDayNo / 12053 );
1031 $jDayNo %= 12053;
1032
1033 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1034 $jDayNo %= 1461;
1035
1036 if ( $jDayNo >= 366 ) {
1037 $jy += floor( ( $jDayNo - 1 ) / 365 );
1038 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1039 }
1040
1041 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1042 $jDayNo -= self::$IRANIAN_DAYS[$i];
1043 }
1044
1045 $jm = $i + 1;
1046 $jd = $jDayNo + 1;
1047
1048 return array( $jy, $jm, $jd );
1049 }
1050
1051 /**
1052 * Converting Gregorian dates to Hijri dates.
1053 *
1054 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1055 *
1056 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1057 */
1058 private static function tsToHijri( $ts ) {
1059 $year = substr( $ts, 0, 4 );
1060 $month = substr( $ts, 4, 2 );
1061 $day = substr( $ts, 6, 2 );
1062
1063 $zyr = $year;
1064 $zd = $day;
1065 $zm = $month;
1066 $zy = $zyr;
1067
1068 if (
1069 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1070 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1071 )
1072 {
1073 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1074 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1075 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1076 $zd - 32075;
1077 } else {
1078 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1079 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1080 }
1081
1082 $zl = $zjd -1948440 + 10632;
1083 $zn = (int)( ( $zl - 1 ) / 10631 );
1084 $zl = $zl - 10631 * $zn + 354;
1085 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1086 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1087 $zm = (int)( ( 24 * $zl ) / 709 );
1088 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1089 $zy = 30 * $zn + $zj - 30;
1090
1091 return array( $zy, $zm, $zd );
1092 }
1093
1094 /**
1095 * Converting Gregorian dates to Hebrew dates.
1096 *
1097 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1098 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1099 * to translate the relevant functions into PHP and release them under
1100 * GNU GPL.
1101 *
1102 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1103 * and Adar II is 14. In a non-leap year, Adar is 6.
1104 */
1105 private static function tsToHebrew( $ts ) {
1106 # Parse date
1107 $year = substr( $ts, 0, 4 );
1108 $month = substr( $ts, 4, 2 );
1109 $day = substr( $ts, 6, 2 );
1110
1111 # Calculate Hebrew year
1112 $hebrewYear = $year + 3760;
1113
1114 # Month number when September = 1, August = 12
1115 $month += 4;
1116 if ( $month > 12 ) {
1117 # Next year
1118 $month -= 12;
1119 $year++;
1120 $hebrewYear++;
1121 }
1122
1123 # Calculate day of year from 1 September
1124 $dayOfYear = $day;
1125 for ( $i = 1; $i < $month; $i++ ) {
1126 if ( $i == 6 ) {
1127 # February
1128 $dayOfYear += 28;
1129 # Check if the year is leap
1130 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1131 $dayOfYear++;
1132 }
1133 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1134 $dayOfYear += 30;
1135 } else {
1136 $dayOfYear += 31;
1137 }
1138 }
1139
1140 # Calculate the start of the Hebrew year
1141 $start = self::hebrewYearStart( $hebrewYear );
1142
1143 # Calculate next year's start
1144 if ( $dayOfYear <= $start ) {
1145 # Day is before the start of the year - it is the previous year
1146 # Next year's start
1147 $nextStart = $start;
1148 # Previous year
1149 $year--;
1150 $hebrewYear--;
1151 # Add days since previous year's 1 September
1152 $dayOfYear += 365;
1153 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1154 # Leap year
1155 $dayOfYear++;
1156 }
1157 # Start of the new (previous) year
1158 $start = self::hebrewYearStart( $hebrewYear );
1159 } else {
1160 # Next year's start
1161 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1162 }
1163
1164 # Calculate Hebrew day of year
1165 $hebrewDayOfYear = $dayOfYear - $start;
1166
1167 # Difference between year's days
1168 $diff = $nextStart - $start;
1169 # Add 12 (or 13 for leap years) days to ignore the difference between
1170 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1171 # difference is only about the year type
1172 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1173 $diff += 13;
1174 } else {
1175 $diff += 12;
1176 }
1177
1178 # Check the year pattern, and is leap year
1179 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1180 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1181 # and non-leap years
1182 $yearPattern = $diff % 30;
1183 # Check if leap year
1184 $isLeap = $diff >= 30;
1185
1186 # Calculate day in the month from number of day in the Hebrew year
1187 # Don't check Adar - if the day is not in Adar, we will stop before;
1188 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1189 $hebrewDay = $hebrewDayOfYear;
1190 $hebrewMonth = 1;
1191 $days = 0;
1192 while ( $hebrewMonth <= 12 ) {
1193 # Calculate days in this month
1194 if ( $isLeap && $hebrewMonth == 6 ) {
1195 # Adar in a leap year
1196 if ( $isLeap ) {
1197 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1198 $days = 30;
1199 if ( $hebrewDay <= $days ) {
1200 # Day in Adar I
1201 $hebrewMonth = 13;
1202 } else {
1203 # Subtract the days of Adar I
1204 $hebrewDay -= $days;
1205 # Try Adar II
1206 $days = 29;
1207 if ( $hebrewDay <= $days ) {
1208 # Day in Adar II
1209 $hebrewMonth = 14;
1210 }
1211 }
1212 }
1213 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1214 # Cheshvan in a complete year (otherwise as the rule below)
1215 $days = 30;
1216 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1217 # Kislev in an incomplete year (otherwise as the rule below)
1218 $days = 29;
1219 } else {
1220 # Odd months have 30 days, even have 29
1221 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1222 }
1223 if ( $hebrewDay <= $days ) {
1224 # In the current month
1225 break;
1226 } else {
1227 # Subtract the days of the current month
1228 $hebrewDay -= $days;
1229 # Try in the next month
1230 $hebrewMonth++;
1231 }
1232 }
1233
1234 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1235 }
1236
1237 /**
1238 * This calculates the Hebrew year start, as days since 1 September.
1239 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1240 * Used for Hebrew date.
1241 */
1242 private static function hebrewYearStart( $year ) {
1243 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1244 $b = intval( ( $year - 1 ) % 4 );
1245 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1246 if ( $m < 0 ) {
1247 $m--;
1248 }
1249 $Mar = intval( $m );
1250 if ( $m < 0 ) {
1251 $m++;
1252 }
1253 $m -= $Mar;
1254
1255 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1256 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1257 $Mar++;
1258 } else if ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1259 $Mar += 2;
1260 } else if ( $c == 2 || $c == 4 || $c == 6 ) {
1261 $Mar++;
1262 }
1263
1264 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1265 return $Mar;
1266 }
1267
1268 /**
1269 * Algorithm to convert Gregorian dates to Thai solar dates,
1270 * Minguo dates or Minguo dates.
1271 *
1272 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1273 * http://en.wikipedia.org/wiki/Minguo_calendar
1274 * http://en.wikipedia.org/wiki/Japanese_era_name
1275 *
1276 * @param $ts String: 14-character timestamp
1277 * @param $cName String: calender name
1278 * @return Array: converted year, month, day
1279 */
1280 private static function tsToYear( $ts, $cName ) {
1281 $gy = substr( $ts, 0, 4 );
1282 $gm = substr( $ts, 4, 2 );
1283 $gd = substr( $ts, 6, 2 );
1284
1285 if ( !strcmp( $cName, 'thai' ) ) {
1286 # Thai solar dates
1287 # Add 543 years to the Gregorian calendar
1288 # Months and days are identical
1289 $gy_offset = $gy + 543;
1290 } else if ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1291 # Minguo dates
1292 # Deduct 1911 years from the Gregorian calendar
1293 # Months and days are identical
1294 $gy_offset = $gy - 1911;
1295 } else if ( !strcmp( $cName, 'tenno' ) ) {
1296 # Nengō dates up to Meiji period
1297 # Deduct years from the Gregorian calendar
1298 # depending on the nengo periods
1299 # Months and days are identical
1300 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1301 # Meiji period
1302 $gy_gannen = $gy - 1868 + 1;
1303 $gy_offset = $gy_gannen;
1304 if ( $gy_gannen == 1 ) {
1305 $gy_offset = '元';
1306 }
1307 $gy_offset = '明治' . $gy_offset;
1308 } else if (
1309 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1310 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1311 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1312 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1313 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1314 )
1315 {
1316 # Taishō period
1317 $gy_gannen = $gy - 1912 + 1;
1318 $gy_offset = $gy_gannen;
1319 if ( $gy_gannen == 1 ) {
1320 $gy_offset = '元';
1321 }
1322 $gy_offset = '大正' . $gy_offset;
1323 } else if (
1324 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1325 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1326 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1327 )
1328 {
1329 # Shōwa period
1330 $gy_gannen = $gy - 1926 + 1;
1331 $gy_offset = $gy_gannen;
1332 if ( $gy_gannen == 1 ) {
1333 $gy_offset = '元';
1334 }
1335 $gy_offset = '昭和' . $gy_offset;
1336 } else {
1337 # Heisei period
1338 $gy_gannen = $gy - 1989 + 1;
1339 $gy_offset = $gy_gannen;
1340 if ( $gy_gannen == 1 ) {
1341 $gy_offset = '元';
1342 }
1343 $gy_offset = '平成' . $gy_offset;
1344 }
1345 } else {
1346 $gy_offset = $gy;
1347 }
1348
1349 return array( $gy_offset, $gm, $gd );
1350 }
1351
1352 /**
1353 * Roman number formatting up to 3000
1354 */
1355 static function romanNumeral( $num ) {
1356 static $table = array(
1357 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1358 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1359 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1360 array( '', 'M', 'MM', 'MMM' )
1361 );
1362
1363 $num = intval( $num );
1364 if ( $num > 3000 || $num <= 0 ) {
1365 return $num;
1366 }
1367
1368 $s = '';
1369 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1370 if ( $num >= $pow10 ) {
1371 $s .= $table[$i][floor( $num / $pow10 )];
1372 }
1373 $num = $num % $pow10;
1374 }
1375 return $s;
1376 }
1377
1378 /**
1379 * Hebrew Gematria number formatting up to 9999
1380 */
1381 static function hebrewNumeral( $num ) {
1382 static $table = array(
1383 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1384 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1385 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1386 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1387 );
1388
1389 $num = intval( $num );
1390 if ( $num > 9999 || $num <= 0 ) {
1391 return $num;
1392 }
1393
1394 $s = '';
1395 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1396 if ( $num >= $pow10 ) {
1397 if ( $num == 15 || $num == 16 ) {
1398 $s .= $table[0][9] . $table[0][$num - 9];
1399 $num = 0;
1400 } else {
1401 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1402 if ( $pow10 == 1000 ) {
1403 $s .= "'";
1404 }
1405 }
1406 }
1407 $num = $num % $pow10;
1408 }
1409 if ( strlen( $s ) == 2 ) {
1410 $str = $s . "'";
1411 } else {
1412 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1413 $str .= substr( $s, strlen( $s ) - 2, 2 );
1414 }
1415 $start = substr( $str, 0, strlen( $str ) - 2 );
1416 $end = substr( $str, strlen( $str ) - 2 );
1417 switch( $end ) {
1418 case 'כ':
1419 $str = $start . 'ך';
1420 break;
1421 case 'מ':
1422 $str = $start . 'ם';
1423 break;
1424 case 'נ':
1425 $str = $start . 'ן';
1426 break;
1427 case 'פ':
1428 $str = $start . 'ף';
1429 break;
1430 case 'צ':
1431 $str = $start . 'ץ';
1432 break;
1433 }
1434 return $str;
1435 }
1436
1437 /**
1438 * This is meant to be used by time(), date(), and timeanddate() to get
1439 * the date preference they're supposed to use, it should be used in
1440 * all children.
1441 *
1442 *<code>
1443 * function timeanddate([...], $format = true) {
1444 * $datePreference = $this->dateFormat($format);
1445 * [...]
1446 * }
1447 *</code>
1448 *
1449 * @param $usePrefs Mixed: if true, the user's preference is used
1450 * if false, the site/language default is used
1451 * if int/string, assumed to be a format.
1452 * @return string
1453 */
1454 function dateFormat( $usePrefs = true ) {
1455 global $wgUser;
1456
1457 if ( is_bool( $usePrefs ) ) {
1458 if ( $usePrefs ) {
1459 $datePreference = $wgUser->getDatePreference();
1460 } else {
1461 $datePreference = (string)User::getDefaultOption( 'date' );
1462 }
1463 } else {
1464 $datePreference = (string)$usePrefs;
1465 }
1466
1467 // return int
1468 if ( $datePreference == '' ) {
1469 return 'default';
1470 }
1471
1472 return $datePreference;
1473 }
1474
1475 /**
1476 * Get a format string for a given type and preference
1477 * @param $type May be date, time or both
1478 * @param $pref The format name as it appears in Messages*.php
1479 */
1480 function getDateFormatString( $type, $pref ) {
1481 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
1482 if ( $pref == 'default' ) {
1483 $pref = $this->getDefaultDateFormat();
1484 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1485 } else {
1486 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1487 if ( is_null( $df ) ) {
1488 $pref = $this->getDefaultDateFormat();
1489 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1490 }
1491 }
1492 $this->dateFormatStrings[$type][$pref] = $df;
1493 }
1494 return $this->dateFormatStrings[$type][$pref];
1495 }
1496
1497 /**
1498 * @param $ts Mixed: the time format which needs to be turned into a
1499 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1500 * @param $adj Bool: whether to adjust the time output according to the
1501 * user configured offset ($timecorrection)
1502 * @param $format Mixed: true to use user's date format preference
1503 * @param $timecorrection String: the time offset as returned by
1504 * validateTimeZone() in Special:Preferences
1505 * @return string
1506 */
1507 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1508 if ( $adj ) {
1509 $ts = $this->userAdjust( $ts, $timecorrection );
1510 }
1511 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1512 return $this->sprintfDate( $df, $ts );
1513 }
1514
1515 /**
1516 * @param $ts Mixed: the time format which needs to be turned into a
1517 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1518 * @param $adj Bool: whether to adjust the time output according to the
1519 * user configured offset ($timecorrection)
1520 * @param $format Mixed: true to use user's date format preference
1521 * @param $timecorrection String: the time offset as returned by
1522 * validateTimeZone() in Special:Preferences
1523 * @return string
1524 */
1525 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1526 if ( $adj ) {
1527 $ts = $this->userAdjust( $ts, $timecorrection );
1528 }
1529 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1530 return $this->sprintfDate( $df, $ts );
1531 }
1532
1533 /**
1534 * @param $ts Mixed: the time format which needs to be turned into a
1535 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1536 * @param $adj Bool: whether to adjust the time output according to the
1537 * user configured offset ($timecorrection)
1538 * @param $format Mixed: what format to return, if it's false output the
1539 * default one (default true)
1540 * @param $timecorrection String: the time offset as returned by
1541 * validateTimeZone() in Special:Preferences
1542 * @return string
1543 */
1544 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1545 $ts = wfTimestamp( TS_MW, $ts );
1546 if ( $adj ) {
1547 $ts = $this->userAdjust( $ts, $timecorrection );
1548 }
1549 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1550 return $this->sprintfDate( $df, $ts );
1551 }
1552
1553 function getMessage( $key ) {
1554 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
1555 }
1556
1557 function getAllMessages() {
1558 return self::$dataCache->getItem( $this->mCode, 'messages' );
1559 }
1560
1561 function iconv( $in, $out, $string ) {
1562 # This is a wrapper for iconv in all languages except esperanto,
1563 # which does some nasty x-conversions beforehand
1564
1565 # Even with //IGNORE iconv can whine about illegal characters in
1566 # *input* string. We just ignore those too.
1567 # REF: http://bugs.php.net/bug.php?id=37166
1568 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1569 wfSuppressWarnings();
1570 $text = iconv( $in, $out . '//IGNORE', $string );
1571 wfRestoreWarnings();
1572 return $text;
1573 }
1574
1575 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1576 function ucwordbreaksCallbackAscii( $matches ) {
1577 return $this->ucfirst( $matches[1] );
1578 }
1579
1580 function ucwordbreaksCallbackMB( $matches ) {
1581 return mb_strtoupper( $matches[0] );
1582 }
1583
1584 function ucCallback( $matches ) {
1585 list( $wikiUpperChars ) = self::getCaseMaps();
1586 return strtr( $matches[1], $wikiUpperChars );
1587 }
1588
1589 function lcCallback( $matches ) {
1590 list( , $wikiLowerChars ) = self::getCaseMaps();
1591 return strtr( $matches[1], $wikiLowerChars );
1592 }
1593
1594 function ucwordsCallbackMB( $matches ) {
1595 return mb_strtoupper( $matches[0] );
1596 }
1597
1598 function ucwordsCallbackWiki( $matches ) {
1599 list( $wikiUpperChars ) = self::getCaseMaps();
1600 return strtr( $matches[0], $wikiUpperChars );
1601 }
1602
1603 /**
1604 * Make a string's first character uppercase
1605 */
1606 function ucfirst( $str ) {
1607 $o = ord( $str );
1608 if ( $o < 96 ) { // if already uppercase...
1609 return $str;
1610 } elseif ( $o < 128 ) {
1611 return ucfirst( $str ); // use PHP's ucfirst()
1612 } else {
1613 // fall back to more complex logic in case of multibyte strings
1614 return $this->uc( $str, true );
1615 }
1616 }
1617
1618 /**
1619 * Convert a string to uppercase
1620 */
1621 function uc( $str, $first = false ) {
1622 if ( function_exists( 'mb_strtoupper' ) ) {
1623 if ( $first ) {
1624 if ( $this->isMultibyte( $str ) ) {
1625 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1626 } else {
1627 return ucfirst( $str );
1628 }
1629 } else {
1630 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1631 }
1632 } else {
1633 if ( $this->isMultibyte( $str ) ) {
1634 $x = $first ? '^' : '';
1635 return preg_replace_callback(
1636 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1637 array( $this, 'ucCallback' ),
1638 $str
1639 );
1640 } else {
1641 return $first ? ucfirst( $str ) : strtoupper( $str );
1642 }
1643 }
1644 }
1645
1646 function lcfirst( $str ) {
1647 $o = ord( $str );
1648 if ( !$o ) {
1649 return strval( $str );
1650 } elseif ( $o >= 128 ) {
1651 return $this->lc( $str, true );
1652 } elseif ( $o > 96 ) {
1653 return $str;
1654 } else {
1655 $str[0] = strtolower( $str[0] );
1656 return $str;
1657 }
1658 }
1659
1660 function lc( $str, $first = false ) {
1661 if ( function_exists( 'mb_strtolower' ) ) {
1662 if ( $first ) {
1663 if ( $this->isMultibyte( $str ) ) {
1664 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1665 } else {
1666 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1667 }
1668 } else {
1669 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1670 }
1671 } else {
1672 if ( $this->isMultibyte( $str ) ) {
1673 $x = $first ? '^' : '';
1674 return preg_replace_callback(
1675 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1676 array( $this, 'lcCallback' ),
1677 $str
1678 );
1679 } else {
1680 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1681 }
1682 }
1683 }
1684
1685 function isMultibyte( $str ) {
1686 return (bool)preg_match( '/[\x80-\xff]/', $str );
1687 }
1688
1689 function ucwords( $str ) {
1690 if ( $this->isMultibyte( $str ) ) {
1691 $str = $this->lc( $str );
1692
1693 // regexp to find first letter in each word (i.e. after each space)
1694 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1695
1696 // function to use to capitalize a single char
1697 if ( function_exists( 'mb_strtoupper' ) ) {
1698 return preg_replace_callback(
1699 $replaceRegexp,
1700 array( $this, 'ucwordsCallbackMB' ),
1701 $str
1702 );
1703 } else {
1704 return preg_replace_callback(
1705 $replaceRegexp,
1706 array( $this, 'ucwordsCallbackWiki' ),
1707 $str
1708 );
1709 }
1710 } else {
1711 return ucwords( strtolower( $str ) );
1712 }
1713 }
1714
1715 # capitalize words at word breaks
1716 function ucwordbreaks( $str ) {
1717 if ( $this->isMultibyte( $str ) ) {
1718 $str = $this->lc( $str );
1719
1720 // since \b doesn't work for UTF-8, we explicitely define word break chars
1721 $breaks = "[ \-\(\)\}\{\.,\?!]";
1722
1723 // find first letter after word break
1724 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1725
1726 if ( function_exists( 'mb_strtoupper' ) ) {
1727 return preg_replace_callback(
1728 $replaceRegexp,
1729 array( $this, 'ucwordbreaksCallbackMB' ),
1730 $str
1731 );
1732 } else {
1733 return preg_replace_callback(
1734 $replaceRegexp,
1735 array( $this, 'ucwordsCallbackWiki' ),
1736 $str
1737 );
1738 }
1739 } else {
1740 return preg_replace_callback(
1741 '/\b([\w\x80-\xff]+)\b/',
1742 array( $this, 'ucwordbreaksCallbackAscii' ),
1743 $str
1744 );
1745 }
1746 }
1747
1748 /**
1749 * Return a case-folded representation of $s
1750 *
1751 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1752 * and $s2 are the same except for the case of their characters. It is not
1753 * necessary for the value returned to make sense when displayed.
1754 *
1755 * Do *not* perform any other normalisation in this function. If a caller
1756 * uses this function when it should be using a more general normalisation
1757 * function, then fix the caller.
1758 */
1759 function caseFold( $s ) {
1760 return $this->uc( $s );
1761 }
1762
1763 function checkTitleEncoding( $s ) {
1764 if ( is_array( $s ) ) {
1765 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1766 }
1767 # Check for non-UTF-8 URLs
1768 $ishigh = preg_match( '/[\x80-\xff]/', $s );
1769 if ( !$ishigh ) {
1770 return $s;
1771 }
1772
1773 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1774 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1775 if ( $isutf8 ) {
1776 return $s;
1777 }
1778
1779 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
1780 }
1781
1782 function fallback8bitEncoding() {
1783 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
1784 }
1785
1786 /**
1787 * Most writing systems use whitespace to break up words.
1788 * Some languages such as Chinese don't conventionally do this,
1789 * which requires special handling when breaking up words for
1790 * searching etc.
1791 */
1792 function hasWordBreaks() {
1793 return true;
1794 }
1795
1796 /**
1797 * Some languages such as Chinese require word segmentation,
1798 * Specify such segmentation when overridden in derived class.
1799 *
1800 * @param $string String
1801 * @return String
1802 */
1803 function segmentByWord( $string ) {
1804 return $string;
1805 }
1806
1807 /**
1808 * Some languages have special punctuation need to be normalized.
1809 * Make such changes here.
1810 *
1811 * @param $string String
1812 * @return String
1813 */
1814 function normalizeForSearch( $string ) {
1815 return self::convertDoubleWidth( $string );
1816 }
1817
1818 /**
1819 * convert double-width roman characters to single-width.
1820 * range: ff00-ff5f ~= 0020-007f
1821 */
1822 protected static function convertDoubleWidth( $string ) {
1823 static $full = null;
1824 static $half = null;
1825
1826 if ( $full === null ) {
1827 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
1828 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1829 $full = str_split( $fullWidth, 3 );
1830 $half = str_split( $halfWidth );
1831 }
1832
1833 $string = str_replace( $full, $half, $string );
1834 return $string;
1835 }
1836
1837 protected static function insertSpace( $string, $pattern ) {
1838 $string = preg_replace( $pattern, " $1 ", $string );
1839 $string = preg_replace( '/ +/', ' ', $string );
1840 return $string;
1841 }
1842
1843 function convertForSearchResult( $termsArray ) {
1844 # some languages, e.g. Chinese, need to do a conversion
1845 # in order for search results to be displayed correctly
1846 return $termsArray;
1847 }
1848
1849 /**
1850 * Get the first character of a string.
1851 *
1852 * @param $s string
1853 * @return string
1854 */
1855 function firstChar( $s ) {
1856 $matches = array();
1857 preg_match(
1858 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1859 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
1860 $s,
1861 $matches
1862 );
1863
1864 if ( isset( $matches[1] ) ) {
1865 if ( strlen( $matches[1] ) != 3 ) {
1866 return $matches[1];
1867 }
1868
1869 // Break down Hangul syllables to grab the first jamo
1870 $code = utf8ToCodepoint( $matches[1] );
1871 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
1872 return $matches[1];
1873 } elseif ( $code < 0xb098 ) {
1874 return "\xe3\x84\xb1";
1875 } elseif ( $code < 0xb2e4 ) {
1876 return "\xe3\x84\xb4";
1877 } elseif ( $code < 0xb77c ) {
1878 return "\xe3\x84\xb7";
1879 } elseif ( $code < 0xb9c8 ) {
1880 return "\xe3\x84\xb9";
1881 } elseif ( $code < 0xbc14 ) {
1882 return "\xe3\x85\x81";
1883 } elseif ( $code < 0xc0ac ) {
1884 return "\xe3\x85\x82";
1885 } elseif ( $code < 0xc544 ) {
1886 return "\xe3\x85\x85";
1887 } elseif ( $code < 0xc790 ) {
1888 return "\xe3\x85\x87";
1889 } elseif ( $code < 0xcc28 ) {
1890 return "\xe3\x85\x88";
1891 } elseif ( $code < 0xce74 ) {
1892 return "\xe3\x85\x8a";
1893 } elseif ( $code < 0xd0c0 ) {
1894 return "\xe3\x85\x8b";
1895 } elseif ( $code < 0xd30c ) {
1896 return "\xe3\x85\x8c";
1897 } elseif ( $code < 0xd558 ) {
1898 return "\xe3\x85\x8d";
1899 } else {
1900 return "\xe3\x85\x8e";
1901 }
1902 } else {
1903 return '';
1904 }
1905 }
1906
1907 function initEncoding() {
1908 # Some languages may have an alternate char encoding option
1909 # (Esperanto X-coding, Japanese furigana conversion, etc)
1910 # If this language is used as the primary content language,
1911 # an override to the defaults can be set here on startup.
1912 }
1913
1914 function recodeForEdit( $s ) {
1915 # For some languages we'll want to explicitly specify
1916 # which characters make it into the edit box raw
1917 # or are converted in some way or another.
1918 # Note that if wgOutputEncoding is different from
1919 # wgInputEncoding, this text will be further converted
1920 # to wgOutputEncoding.
1921 global $wgEditEncoding;
1922 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
1923 return $s;
1924 } else {
1925 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1926 }
1927 }
1928
1929 function recodeInput( $s ) {
1930 # Take the previous into account.
1931 global $wgEditEncoding;
1932 if ( $wgEditEncoding != '' ) {
1933 $enc = $wgEditEncoding;
1934 } else {
1935 $enc = 'UTF-8';
1936 }
1937 if ( $enc == 'UTF-8' ) {
1938 return $s;
1939 } else {
1940 return $this->iconv( $enc, 'UTF-8', $s );
1941 }
1942 }
1943
1944 /**
1945 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
1946 * also cleans up certain backwards-compatible sequences, converting them
1947 * to the modern Unicode equivalent.
1948 *
1949 * This is language-specific for performance reasons only.
1950 */
1951 function normalize( $s ) {
1952 global $wgAllUnicodeFixes;
1953 $s = UtfNormal::cleanUp( $s );
1954 if ( $wgAllUnicodeFixes ) {
1955 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
1956 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
1957 }
1958
1959 return $s;
1960 }
1961
1962 /**
1963 * Transform a string using serialized data stored in the given file (which
1964 * must be in the serialized subdirectory of $IP). The file contains pairs
1965 * mapping source characters to destination characters.
1966 *
1967 * The data is cached in process memory. This will go faster if you have the
1968 * FastStringSearch extension.
1969 */
1970 function transformUsingPairFile( $file, $string ) {
1971 if ( !isset( $this->transformData[$file] ) ) {
1972 $data = wfGetPrecompiledData( $file );
1973 if ( $data === false ) {
1974 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
1975 }
1976 $this->transformData[$file] = new ReplacementArray( $data );
1977 }
1978 return $this->transformData[$file]->replace( $string );
1979 }
1980
1981 /**
1982 * For right-to-left language support
1983 *
1984 * @return bool
1985 */
1986 function isRTL() {
1987 return self::$dataCache->getItem( $this->mCode, 'rtl' );
1988 }
1989
1990 /**
1991 * Return the correct HTML 'dir' attribute value for this language.
1992 * @return String
1993 */
1994 function getDir() {
1995 return $this->isRTL() ? 'rtl' : 'ltr';
1996 }
1997
1998 /**
1999 * Return 'left' or 'right' as appropriate alignment for line-start
2000 * for this language's text direction.
2001 *
2002 * Should be equivalent to CSS3 'start' text-align value....
2003 *
2004 * @return String
2005 */
2006 function alignStart() {
2007 return $this->isRTL() ? 'right' : 'left';
2008 }
2009
2010 /**
2011 * Return 'right' or 'left' as appropriate alignment for line-end
2012 * for this language's text direction.
2013 *
2014 * Should be equivalent to CSS3 'end' text-align value....
2015 *
2016 * @return String
2017 */
2018 function alignEnd() {
2019 return $this->isRTL() ? 'left' : 'right';
2020 }
2021
2022 /**
2023 * A hidden direction mark (LRM or RLM), depending on the language direction
2024 *
2025 * @return string
2026 */
2027 function getDirMark() {
2028 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
2029 }
2030
2031 function capitalizeAllNouns() {
2032 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2033 }
2034
2035 /**
2036 * An arrow, depending on the language direction
2037 *
2038 * @return string
2039 */
2040 function getArrow() {
2041 return $this->isRTL() ? '←' : '→';
2042 }
2043
2044 /**
2045 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2046 *
2047 * @return bool
2048 */
2049 function linkPrefixExtension() {
2050 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2051 }
2052
2053 function getMagicWords() {
2054 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2055 }
2056
2057 # Fill a MagicWord object with data from here
2058 function getMagic( $mw ) {
2059 if ( !$this->mMagicHookDone ) {
2060 $this->mMagicHookDone = true;
2061 wfProfileIn( 'LanguageGetMagic' );
2062 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2063 wfProfileOut( 'LanguageGetMagic' );
2064 }
2065 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
2066 $rawEntry = $this->mMagicExtensions[$mw->mId];
2067 } else {
2068 $magicWords = $this->getMagicWords();
2069 if ( isset( $magicWords[$mw->mId] ) ) {
2070 $rawEntry = $magicWords[$mw->mId];
2071 } else {
2072 $rawEntry = false;
2073 }
2074 }
2075
2076 if ( !is_array( $rawEntry ) ) {
2077 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
2078 } else {
2079 $mw->mCaseSensitive = $rawEntry[0];
2080 $mw->mSynonyms = array_slice( $rawEntry, 1 );
2081 }
2082 }
2083
2084 /**
2085 * Add magic words to the extension array
2086 */
2087 function addMagicWordsByLang( $newWords ) {
2088 $code = $this->getCode();
2089 $fallbackChain = array();
2090 while ( $code && !in_array( $code, $fallbackChain ) ) {
2091 $fallbackChain[] = $code;
2092 $code = self::getFallbackFor( $code );
2093 }
2094 if ( !in_array( 'en', $fallbackChain ) ) {
2095 $fallbackChain[] = 'en';
2096 }
2097 $fallbackChain = array_reverse( $fallbackChain );
2098 foreach ( $fallbackChain as $code ) {
2099 if ( isset( $newWords[$code] ) ) {
2100 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
2101 }
2102 }
2103 }
2104
2105 /**
2106 * Get special page names, as an associative array
2107 * case folded alias => real name
2108 */
2109 function getSpecialPageAliases() {
2110 // Cache aliases because it may be slow to load them
2111 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
2112 // Initialise array
2113 $this->mExtendedSpecialPageAliases =
2114 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
2115 wfRunHooks( 'LanguageGetSpecialPageAliases',
2116 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
2117 }
2118
2119 return $this->mExtendedSpecialPageAliases;
2120 }
2121
2122 /**
2123 * Italic is unsuitable for some languages
2124 *
2125 * @param $text String: the text to be emphasized.
2126 * @return string
2127 */
2128 function emphasize( $text ) {
2129 return "<em>$text</em>";
2130 }
2131
2132 /**
2133 * Normally we output all numbers in plain en_US style, that is
2134 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2135 * point twohundredthirtyfive. However this is not sutable for all
2136 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2137 * Icelandic just want to use commas instead of dots, and dots instead
2138 * of commas like "293.291,235".
2139 *
2140 * An example of this function being called:
2141 * <code>
2142 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2143 * </code>
2144 *
2145 * See LanguageGu.php for the Gujarati implementation and
2146 * $separatorTransformTable on MessageIs.php for
2147 * the , => . and . => , implementation.
2148 *
2149 * @todo check if it's viable to use localeconv() for the decimal
2150 * separator thing.
2151 * @param $number Mixed: the string to be formatted, should be an integer
2152 * or a floating point number.
2153 * @param $nocommafy Bool: set to true for special numbers like dates
2154 * @return string
2155 */
2156 function formatNum( $number, $nocommafy = false ) {
2157 global $wgTranslateNumerals;
2158 if ( !$nocommafy ) {
2159 $number = $this->commafy( $number );
2160 $s = $this->separatorTransformTable();
2161 if ( $s ) {
2162 $number = strtr( $number, $s );
2163 }
2164 }
2165
2166 if ( $wgTranslateNumerals ) {
2167 $s = $this->digitTransformTable();
2168 if ( $s ) {
2169 $number = strtr( $number, $s );
2170 }
2171 }
2172
2173 return $number;
2174 }
2175
2176 function parseFormattedNumber( $number ) {
2177 $s = $this->digitTransformTable();
2178 if ( $s ) {
2179 $number = strtr( $number, array_flip( $s ) );
2180 }
2181
2182 $s = $this->separatorTransformTable();
2183 if ( $s ) {
2184 $number = strtr( $number, array_flip( $s ) );
2185 }
2186
2187 $number = strtr( $number, array( ',' => '' ) );
2188 return $number;
2189 }
2190
2191 /**
2192 * Adds commas to a given number
2193 *
2194 * @param $_ mixed
2195 * @return string
2196 */
2197 function commafy( $_ ) {
2198 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2199 }
2200
2201 function digitTransformTable() {
2202 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
2203 }
2204
2205 function separatorTransformTable() {
2206 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
2207 }
2208
2209 /**
2210 * Take a list of strings and build a locale-friendly comma-separated
2211 * list, using the local comma-separator message.
2212 * The last two strings are chained with an "and".
2213 *
2214 * @param $l Array
2215 * @return string
2216 */
2217 function listToText( $l ) {
2218 $s = '';
2219 $m = count( $l ) - 1;
2220 if ( $m == 1 ) {
2221 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2222 } else {
2223 for ( $i = $m; $i >= 0; $i-- ) {
2224 if ( $i == $m ) {
2225 $s = $l[$i];
2226 } else if ( $i == $m - 1 ) {
2227 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2228 } else {
2229 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2230 }
2231 }
2232 return $s;
2233 }
2234 }
2235
2236 /**
2237 * Take a list of strings and build a locale-friendly comma-separated
2238 * list, using the local comma-separator message.
2239 * @param $list array of strings to put in a comma list
2240 * @return string
2241 */
2242 function commaList( $list ) {
2243 return implode(
2244 $list,
2245 wfMsgExt(
2246 'comma-separator',
2247 array( 'parsemag', 'escapenoentities', 'language' => $this )
2248 )
2249 );
2250 }
2251
2252 /**
2253 * Take a list of strings and build a locale-friendly semicolon-separated
2254 * list, using the local semicolon-separator message.
2255 * @param $list array of strings to put in a semicolon list
2256 * @return string
2257 */
2258 function semicolonList( $list ) {
2259 return implode(
2260 $list,
2261 wfMsgExt(
2262 'semicolon-separator',
2263 array( 'parsemag', 'escapenoentities', 'language' => $this )
2264 )
2265 );
2266 }
2267
2268 /**
2269 * Same as commaList, but separate it with the pipe instead.
2270 * @param $list array of strings to put in a pipe list
2271 * @return string
2272 */
2273 function pipeList( $list ) {
2274 return implode(
2275 $list,
2276 wfMsgExt(
2277 'pipe-separator',
2278 array( 'escapenoentities', 'language' => $this )
2279 )
2280 );
2281 }
2282
2283 /**
2284 * Truncate a string to a specified length in bytes, appending an optional
2285 * string (e.g. for ellipses)
2286 *
2287 * The database offers limited byte lengths for some columns in the database;
2288 * multi-byte character sets mean we need to ensure that only whole characters
2289 * are included, otherwise broken characters can be passed to the user
2290 *
2291 * If $length is negative, the string will be truncated from the beginning
2292 *
2293 * @param $string String to truncate
2294 * @param $length Int: maximum length (excluding ellipses)
2295 * @param $ellipsis String to append to the truncated text
2296 * @return string
2297 */
2298 function truncate( $string, $length, $ellipsis = '...' ) {
2299 # Use the localized ellipsis character
2300 if ( $ellipsis == '...' ) {
2301 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2302 }
2303 # Check if there is no need to truncate
2304 if ( $length == 0 ) {
2305 return $ellipsis;
2306 } elseif ( strlen( $string ) <= abs( $length ) ) {
2307 return $string;
2308 }
2309 $stringOriginal = $string;
2310 if ( $length > 0 ) {
2311 $string = substr( $string, 0, $length ); // xyz...
2312 $string = $this->removeBadCharLast( $string );
2313 $string = $string . $ellipsis;
2314 } else {
2315 $string = substr( $string, $length ); // ...xyz
2316 $string = $this->removeBadCharFirst( $string );
2317 $string = $ellipsis . $string;
2318 }
2319 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181)
2320 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
2321 return $string;
2322 } else {
2323 return $stringOriginal;
2324 }
2325 }
2326
2327 /**
2328 * Remove bytes that represent an incomplete Unicode character
2329 * at the end of string (e.g. bytes of the char are missing)
2330 *
2331 * @param $string String
2332 * @return string
2333 */
2334 protected function removeBadCharLast( $string ) {
2335 $char = ord( $string[strlen( $string ) - 1] );
2336 $m = array();
2337 if ( $char >= 0xc0 ) {
2338 # We got the first byte only of a multibyte char; remove it.
2339 $string = substr( $string, 0, -1 );
2340 } elseif ( $char >= 0x80 &&
2341 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2342 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
2343 {
2344 # We chopped in the middle of a character; remove it
2345 $string = $m[1];
2346 }
2347 return $string;
2348 }
2349
2350 /**
2351 * Remove bytes that represent an incomplete Unicode character
2352 * at the start of string (e.g. bytes of the char are missing)
2353 *
2354 * @param $string String
2355 * @return string
2356 */
2357 protected function removeBadCharFirst( $string ) {
2358 $char = ord( $string[0] );
2359 if ( $char >= 0x80 && $char < 0xc0 ) {
2360 # We chopped in the middle of a character; remove the whole thing
2361 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2362 }
2363 return $string;
2364 }
2365
2366 /*
2367 * Truncate a string of valid HTML to a specified length in bytes,
2368 * appending an optional string (e.g. for ellipses), and return valid HTML
2369 *
2370 * This is only intended for styled/linked text, such as HTML with
2371 * tags like <span> and <a>, were the tags are self-contained (valid HTML)
2372 *
2373 * Note: tries to fix broken HTML with MWTidy
2374 *
2375 * @param string $text String to truncate
2376 * @param int $length (zero/positive) Maximum length (excluding ellipses)
2377 * @param string $ellipsis String to append to the truncated text
2378 * @returns string
2379 */
2380 function truncateHtml( $text, $length, $ellipsis = '...' ) {
2381 # Use the localized ellipsis character
2382 if ( $ellipsis == '...' ) {
2383 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2384 }
2385 # Check if there is no need to truncate
2386 if ( $length <= 0 ) {
2387 return $ellipsis; // no text shown, nothing to format
2388 } elseif ( strlen( $text ) <= $length ) {
2389 return $text; // string short enough even *with* HTML
2390 }
2391 $text = MWTidy::tidy( $text ); // fix tags
2392 $displayLen = 0; // innerHTML legth so far
2393 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
2394 $tagType = 0; // 0-open, 1-close
2395 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
2396 $entityState = 0; // 0-not entity, 1-entity
2397 $tag = $ret = $ch = '';
2398 $openTags = array();
2399 $textLen = strlen( $text );
2400 for ( $pos = 0; $pos < $textLen; ++$pos ) {
2401 $ch = $text[$pos];
2402 $lastCh = $pos ? $text[$pos - 1] : '';
2403 $ret .= $ch; // add to result string
2404 if ( $ch == '<' ) {
2405 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
2406 $entityState = 0; // for bad HTML
2407 $bracketState = 1; // tag started (checking for backslash)
2408 } elseif ( $ch == '>' ) {
2409 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
2410 $entityState = 0; // for bad HTML
2411 $bracketState = 0; // out of brackets
2412 } elseif ( $bracketState == 1 ) {
2413 if ( $ch == '/' ) {
2414 $tagType = 1; // close tag (e.g. "</span>")
2415 } else {
2416 $tagType = 0; // open tag (e.g. "<span>")
2417 $tag .= $ch;
2418 }
2419 $bracketState = 2; // building tag name
2420 } elseif ( $bracketState == 2 ) {
2421 if ( $ch != ' ' ) {
2422 $tag .= $ch;
2423 } else {
2424 // Name found (e.g. "<a href=..."), add on tag attributes...
2425 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
2426 }
2427 } elseif ( $bracketState == 0 ) {
2428 if ( $entityState ) {
2429 if ( $ch == ';' ) {
2430 $entityState = 0;
2431 $displayLen++; // entity is one displayed char
2432 }
2433 } else {
2434 if ( $ch == '&' ) {
2435 $entityState = 1; // entity found, (e.g. "&#160;")
2436 } else {
2437 $displayLen++; // this char is displayed
2438 // Add on the other display text after this...
2439 $skipped = $this->truncate_skip(
2440 $ret, $text, "<>&", $pos + 1, $length - $displayLen );
2441 $displayLen += $skipped;
2442 $pos += $skipped;
2443 }
2444 }
2445 }
2446 # Consider truncation once the display length has reached the maximim.
2447 # Double-check that we're not in the middle of a bracket/entity...
2448 if ( $displayLen >= $length && $bracketState == 0 && $entityState == 0 ) {
2449 if ( !$testingEllipsis ) {
2450 $testingEllipsis = true;
2451 # Save where we are; we will truncate here unless
2452 # the ellipsis actually makes the string longer.
2453 $pOpenTags = $openTags; // save state
2454 $pRet = $ret; // save state
2455 } elseif ( $displayLen > ( $length + strlen( $ellipsis ) ) ) {
2456 # Ellipsis won't make string longer/equal, the truncation point was OK.
2457 $openTags = $pOpenTags; // reload state
2458 $ret = $this->removeBadCharLast( $pRet ); // reload state, multi-byte char fix
2459 $ret .= $ellipsis; // add ellipsis
2460 break;
2461 }
2462 }
2463 }
2464 if ( $displayLen == 0 ) {
2465 return ''; // no text shown, nothing to format
2466 }
2467 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags ); // for bad HTML
2468 while ( count( $openTags ) > 0 ) {
2469 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
2470 }
2471 return $ret;
2472 }
2473
2474 // truncateHtml() helper function
2475 // like strcspn() but adds the skipped chars to $ret
2476 private function truncate_skip( &$ret, $text, $search, $start, $len = -1 ) {
2477 $skipCount = 0;
2478 if ( $start < strlen( $text ) ) {
2479 $skipCount = strcspn( $text, $search, $start, $len );
2480 $ret .= substr( $text, $start, $skipCount );
2481 }
2482 return $skipCount;
2483 }
2484
2485 // truncateHtml() helper function
2486 // (a) push or pop $tag from $openTags as needed
2487 // (b) clear $tag value
2488 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
2489 $tag = ltrim( $tag );
2490 if ( $tag != '' ) {
2491 if ( $tagType == 0 && $lastCh != '/' ) {
2492 $openTags[] = $tag; // tag opened (didn't close itself)
2493 } else if ( $tagType == 1 ) {
2494 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
2495 array_pop( $openTags ); // tag closed
2496 }
2497 }
2498 $tag = '';
2499 }
2500 }
2501
2502 /**
2503 * Grammatical transformations, needed for inflected languages
2504 * Invoked by putting {{grammar:case|word}} in a message
2505 *
2506 * @param $word string
2507 * @param $case string
2508 * @return string
2509 */
2510 function convertGrammar( $word, $case ) {
2511 global $wgGrammarForms;
2512 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
2513 return $wgGrammarForms[$this->getCode()][$case][$word];
2514 }
2515 return $word;
2516 }
2517
2518 /**
2519 * Provides an alternative text depending on specified gender.
2520 * Usage {{gender:username|masculine|feminine|neutral}}.
2521 * username is optional, in which case the gender of current user is used,
2522 * but only in (some) interface messages; otherwise default gender is used.
2523 * If second or third parameter are not specified, masculine is used.
2524 * These details may be overriden per language.
2525 */
2526 function gender( $gender, $forms ) {
2527 if ( !count( $forms ) ) {
2528 return '';
2529 }
2530 $forms = $this->preConvertPlural( $forms, 2 );
2531 if ( $gender === 'male' ) {
2532 return $forms[0];
2533 }
2534 if ( $gender === 'female' ) {
2535 return $forms[1];
2536 }
2537 return isset( $forms[2] ) ? $forms[2] : $forms[0];
2538 }
2539
2540 /**
2541 * Plural form transformations, needed for some languages.
2542 * For example, there are 3 form of plural in Russian and Polish,
2543 * depending on "count mod 10". See [[w:Plural]]
2544 * For English it is pretty simple.
2545 *
2546 * Invoked by putting {{plural:count|wordform1|wordform2}}
2547 * or {{plural:count|wordform1|wordform2|wordform3}}
2548 *
2549 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2550 *
2551 * @param $count Integer: non-localized number
2552 * @param $forms Array: different plural forms
2553 * @return string Correct form of plural for $count in this language
2554 */
2555 function convertPlural( $count, $forms ) {
2556 if ( !count( $forms ) ) {
2557 return '';
2558 }
2559 $forms = $this->preConvertPlural( $forms, 2 );
2560
2561 return ( $count == 1 ) ? $forms[0] : $forms[1];
2562 }
2563
2564 /**
2565 * Checks that convertPlural was given an array and pads it to requested
2566 * amound of forms by copying the last one.
2567 *
2568 * @param $count Integer: How many forms should there be at least
2569 * @param $forms Array of forms given to convertPlural
2570 * @return array Padded array of forms or an exception if not an array
2571 */
2572 protected function preConvertPlural( /* Array */ $forms, $count ) {
2573 while ( count( $forms ) < $count ) {
2574 $forms[] = $forms[count( $forms ) - 1];
2575 }
2576 return $forms;
2577 }
2578
2579 /**
2580 * For translating of expiry times
2581 * @param $str String: the validated block time in English
2582 * @return Somehow translated block time
2583 * @see LanguageFi.php for example implementation
2584 */
2585 function translateBlockExpiry( $str ) {
2586 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2587
2588 if ( $scBlockExpiryOptions == '-' ) {
2589 return $str;
2590 }
2591
2592 foreach ( explode( ',', $scBlockExpiryOptions ) as $option ) {
2593 if ( strpos( $option, ':' ) === false ) {
2594 continue;
2595 }
2596 list( $show, $value ) = explode( ':', $option );
2597 if ( strcmp( $str, $value ) == 0 ) {
2598 return htmlspecialchars( trim( $show ) );
2599 }
2600 }
2601
2602 return $str;
2603 }
2604
2605 /**
2606 * languages like Chinese need to be segmented in order for the diff
2607 * to be of any use
2608 *
2609 * @param $text String
2610 * @return String
2611 */
2612 function segmentForDiff( $text ) {
2613 return $text;
2614 }
2615
2616 /**
2617 * and unsegment to show the result
2618 *
2619 * @param $text String
2620 * @return String
2621 */
2622 function unsegmentForDiff( $text ) {
2623 return $text;
2624 }
2625
2626 # convert text to all supported variants
2627 function autoConvertToAllVariants( $text ) {
2628 return $this->mConverter->autoConvertToAllVariants( $text );
2629 }
2630
2631 # convert text to different variants of a language.
2632 function convert( $text ) {
2633 return $this->mConverter->convert( $text );
2634 }
2635
2636 # Convert a Title object to a string in the preferred variant
2637 function convertTitle( $title ) {
2638 return $this->mConverter->convertTitle( $title );
2639 }
2640
2641 # Check if this is a language with variants
2642 function hasVariants() {
2643 return sizeof( $this->getVariants() ) > 1;
2644 }
2645
2646 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2647 function armourMath( $text ) {
2648 return $this->mConverter->armourMath( $text );
2649 }
2650
2651 /**
2652 * Perform output conversion on a string, and encode for safe HTML output.
2653 * @param $text String text to be converted
2654 * @param $isTitle Bool whether this conversion is for the article title
2655 * @return string
2656 * @todo this should get integrated somewhere sane
2657 */
2658 function convertHtml( $text, $isTitle = false ) {
2659 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2660 }
2661
2662 function convertCategoryKey( $key ) {
2663 return $this->mConverter->convertCategoryKey( $key );
2664 }
2665
2666 /**
2667 * Get the list of variants supported by this langauge
2668 * see sample implementation in LanguageZh.php
2669 *
2670 * @return array an array of language codes
2671 */
2672 function getVariants() {
2673 return $this->mConverter->getVariants();
2674 }
2675
2676 function getPreferredVariant( $fromUser = true, $fromHeader = false ) {
2677 return $this->mConverter->getPreferredVariant( $fromUser, $fromHeader );
2678 }
2679
2680 /**
2681 * If a language supports multiple variants, it is
2682 * possible that non-existing link in one variant
2683 * actually exists in another variant. this function
2684 * tries to find it. See e.g. LanguageZh.php
2685 *
2686 * @param $link String: the name of the link
2687 * @param $nt Mixed: the title object of the link
2688 * @param $ignoreOtherCond Boolean: to disable other conditions when
2689 * we need to transclude a template or update a category's link
2690 * @return null the input parameters may be modified upon return
2691 */
2692 function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
2693 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
2694 }
2695
2696 /**
2697 * If a language supports multiple variants, converts text
2698 * into an array of all possible variants of the text:
2699 * 'variant' => text in that variant
2700 */
2701 function convertLinkToAllVariants( $text ) {
2702 return $this->mConverter->convertLinkToAllVariants( $text );
2703 }
2704
2705 /**
2706 * returns language specific options used by User::getPageRenderHash()
2707 * for example, the preferred language variant
2708 *
2709 * @return string
2710 */
2711 function getExtraHashOptions() {
2712 return $this->mConverter->getExtraHashOptions();
2713 }
2714
2715 /**
2716 * For languages that support multiple variants, the title of an
2717 * article may be displayed differently in different variants. this
2718 * function returns the apporiate title defined in the body of the article.
2719 *
2720 * @return string
2721 */
2722 function getParsedTitle() {
2723 return $this->mConverter->getParsedTitle();
2724 }
2725
2726 /**
2727 * Enclose a string with the "no conversion" tag. This is used by
2728 * various functions in the Parser
2729 *
2730 * @param $text String: text to be tagged for no conversion
2731 * @param $noParse
2732 * @return string the tagged text
2733 */
2734 function markNoConversion( $text, $noParse = false ) {
2735 return $this->mConverter->markNoConversion( $text, $noParse );
2736 }
2737
2738 /**
2739 * A regular expression to match legal word-trailing characters
2740 * which should be merged onto a link of the form [[foo]]bar.
2741 *
2742 * @return string
2743 */
2744 function linkTrail() {
2745 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
2746 }
2747
2748 function getLangObj() {
2749 return $this;
2750 }
2751
2752 /**
2753 * Get the RFC 3066 code for this language object
2754 */
2755 function getCode() {
2756 return $this->mCode;
2757 }
2758
2759 function setCode( $code ) {
2760 $this->mCode = $code;
2761 }
2762
2763 /**
2764 * Get the name of a file for a certain language code
2765 * @param $prefix string Prepend this to the filename
2766 * @param $code string Language code
2767 * @param $suffix string Append this to the filename
2768 * @return string $prefix . $mangledCode . $suffix
2769 */
2770 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2771 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2772 }
2773
2774 /**
2775 * Get the language code from a file name. Inverse of getFileName()
2776 * @param $filename string $prefix . $languageCode . $suffix
2777 * @param $prefix string Prefix before the language code
2778 * @param $suffix string Suffix after the language code
2779 * @return Language code, or false if $prefix or $suffix isn't found
2780 */
2781 static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
2782 $m = null;
2783 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
2784 preg_quote( $suffix, '/' ) . '/', $filename, $m );
2785 if ( !count( $m ) ) {
2786 return false;
2787 }
2788 return str_replace( '_', '-', strtolower( $m[1] ) );
2789 }
2790
2791 static function getMessagesFileName( $code ) {
2792 global $IP;
2793 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2794 }
2795
2796 static function getClassFileName( $code ) {
2797 global $IP;
2798 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2799 }
2800
2801 /**
2802 * Get the fallback for a given language
2803 */
2804 static function getFallbackFor( $code ) {
2805 if ( $code === 'en' ) {
2806 // Shortcut
2807 return false;
2808 } else {
2809 return self::getLocalisationCache()->getItem( $code, 'fallback' );
2810 }
2811 }
2812
2813 /**
2814 * Get all messages for a given language
2815 * WARNING: this may take a long time
2816 */
2817 static function getMessagesFor( $code ) {
2818 return self::getLocalisationCache()->getItem( $code, 'messages' );
2819 }
2820
2821 /**
2822 * Get a message for a given language
2823 */
2824 static function getMessageFor( $key, $code ) {
2825 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
2826 }
2827
2828 function fixVariableInNamespace( $talk ) {
2829 if ( strpos( $talk, '$1' ) === false ) {
2830 return $talk;
2831 }
2832
2833 global $wgMetaNamespace;
2834 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2835
2836 # Allow grammar transformations
2837 # Allowing full message-style parsing would make simple requests
2838 # such as action=raw much more expensive than they need to be.
2839 # This will hopefully cover most cases.
2840 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2841 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2842 return str_replace( ' ', '_', $talk );
2843 }
2844
2845 function replaceGrammarInNamespace( $m ) {
2846 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2847 }
2848
2849 static function getCaseMaps() {
2850 static $wikiUpperChars, $wikiLowerChars;
2851 if ( isset( $wikiUpperChars ) ) {
2852 return array( $wikiUpperChars, $wikiLowerChars );
2853 }
2854
2855 wfProfileIn( __METHOD__ );
2856 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2857 if ( $arr === false ) {
2858 throw new MWException(
2859 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2860 }
2861 extract( $arr );
2862 wfProfileOut( __METHOD__ );
2863 return array( $wikiUpperChars, $wikiLowerChars );
2864 }
2865
2866 function formatTimePeriod( $seconds ) {
2867 if ( $seconds < 10 ) {
2868 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2869 } elseif ( $seconds < 60 ) {
2870 return $this->formatNum( round( $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2871 } elseif ( $seconds < 3600 ) {
2872 $minutes = floor( $seconds / 60 );
2873 $secondsPart = round( fmod( $seconds, 60 ) );
2874 if ( $secondsPart == 60 ) {
2875 $secondsPart = 0;
2876 $minutes++;
2877 }
2878 return $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2879 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2880 } else {
2881 $hours = floor( $seconds / 3600 );
2882 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2883 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2884 if ( $secondsPart == 60 ) {
2885 $secondsPart = 0;
2886 $minutes++;
2887 }
2888 if ( $minutes == 60 ) {
2889 $minutes = 0;
2890 $hours++;
2891 }
2892 return $this->formatNum( $hours ) . $this->getMessageFromDB( 'hours-abbrev' ) . ' ' .
2893 $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2894 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2895 }
2896 }
2897
2898 function formatBitrate( $bps ) {
2899 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2900 if ( $bps <= 0 ) {
2901 return $this->formatNum( $bps ) . $units[0];
2902 }
2903 $unitIndex = floor( log10( $bps ) / 3 );
2904 $mantissa = $bps / pow( 1000, $unitIndex );
2905 if ( $mantissa < 10 ) {
2906 $mantissa = round( $mantissa, 1 );
2907 } else {
2908 $mantissa = round( $mantissa );
2909 }
2910 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2911 }
2912
2913 /**
2914 * Format a size in bytes for output, using an appropriate
2915 * unit (B, KB, MB or GB) according to the magnitude in question
2916 *
2917 * @param $size Size to format
2918 * @return string Plain text (not HTML)
2919 */
2920 function formatSize( $size ) {
2921 // For small sizes no decimal places necessary
2922 $round = 0;
2923 if ( $size > 1024 ) {
2924 $size = $size / 1024;
2925 if ( $size > 1024 ) {
2926 $size = $size / 1024;
2927 // For MB and bigger two decimal places are smarter
2928 $round = 2;
2929 if ( $size > 1024 ) {
2930 $size = $size / 1024;
2931 $msg = 'size-gigabytes';
2932 } else {
2933 $msg = 'size-megabytes';
2934 }
2935 } else {
2936 $msg = 'size-kilobytes';
2937 }
2938 } else {
2939 $msg = 'size-bytes';
2940 }
2941 $size = round( $size, $round );
2942 $text = $this->getMessageFromDB( $msg );
2943 return str_replace( '$1', $this->formatNum( $size ), $text );
2944 }
2945
2946 /**
2947 * Get the conversion rule title, if any.
2948 */
2949 function getConvRuleTitle() {
2950 return $this->mConverter->getConvRuleTitle();
2951 }
2952
2953 /**
2954 * Given a string, convert it to a (hopefully short) key that can be used
2955 * for efficient sorting. A binary sort according to the sortkeys
2956 * corresponds to a logical sort of the corresponding strings. Current
2957 * code expects that a null character should sort before all others, but
2958 * has no other particular expectations (and that one can be changed if
2959 * necessary).
2960 *
2961 * @param string $string UTF-8 string
2962 * @return string Binary sortkey
2963 */
2964 public function convertToSortkey( $string ) {
2965 # Fake function for now
2966 return strtoupper( $string );
2967 }
2968
2969 /**
2970 * Given a string, return the logical "first letter" to be used for
2971 * grouping on category pages and so on. This has to be coordinated
2972 * carefully with convertToSortkey(), or else the sorted list might jump
2973 * back and forth between the same "initial letters" or other pathological
2974 * behavior. For instance, if you just return the first character, but "a"
2975 * sorts the same as "A" based on convertToSortkey(), then you might get a
2976 * list like
2977 *
2978 * == A ==
2979 * * [[Aardvark]]
2980 *
2981 * == a ==
2982 * * [[antelope]]
2983 *
2984 * == A ==
2985 * * [[Ape]]
2986 *
2987 * etc., assuming for the sake of argument that $wgCapitalLinks is false.
2988 *
2989 * @param string $string UTF-8 string
2990 * @return string UTF-8 string corresponding to the first letter of input
2991 */
2992 public function firstLetterForLists( $string ) {
2993 if ( $string[0] == "\0" ) {
2994 $string = substr( $string, 1 );
2995 }
2996 return strtoupper( $this->firstChar( $string ) );
2997 }
2998 }
Wiklou, le Wiki du Wiklou