dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Apply hack from 1.16wmf4 in r71329 to trunk (follow-up to r71327). The version in...
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * @defgroup Language Language
4 *
5 * @file
6 * @ingroup Language
7 */
8
9 if ( !defined( 'MEDIAWIKI' ) ) {
10 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
11 exit( 1 );
12 }
13
14 # Read language names
15 global $wgLanguageNames;
16 require_once( dirname( __FILE__ ) . '/Names.php' );
17
18 global $wgInputEncoding, $wgOutputEncoding;
19
20 /**
21 * These are always UTF-8, they exist only for backwards compatibility
22 */
23 $wgInputEncoding = 'UTF-8';
24 $wgOutputEncoding = 'UTF-8';
25
26 if ( function_exists( 'mb_strtoupper' ) ) {
27 mb_internal_encoding( 'UTF-8' );
28 }
29
30 /**
31 * a fake language converter
32 *
33 * @ingroup Language
34 */
35 class FakeConverter {
36 var $mLang;
37 function FakeConverter( $langobj ) { $this->mLang = $langobj; }
38 function autoConvertToAllVariants( $text ) { return $text; }
39 function convert( $t ) { return $t; }
40 function convertTitle( $t ) { return $t->getPrefixedText(); }
41 function getVariants() { return array( $this->mLang->getCode() ); }
42 function getPreferredVariant() { return $this->mLang->getCode(); }
43 function getConvRuleTitle() { return false; }
44 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
45 function getExtraHashOptions() { return ''; }
46 function getParsedTitle() { return ''; }
47 function markNoConversion( $text, $noParse = false ) { return $text; }
48 function convertCategoryKey( $key ) { return $key; }
49 function convertLinkToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
50 function armourMath( $text ) { return $text; }
51 }
52
53 /**
54 * Internationalisation code
55 * @ingroup Language
56 */
57 class Language {
58 var $mConverter, $mVariants, $mCode, $mLoaded = false;
59 var $mMagicExtensions = array(), $mMagicHookDone = false;
60
61 var $mNamespaceIds, $namespaceNames, $namespaceAliases;
62 var $dateFormatStrings = array();
63 var $mExtendedSpecialPageAliases;
64
65 /**
66 * ReplacementArray object caches
67 */
68 var $transformData = array();
69
70 static public $dataCache;
71 static public $mLangObjCache = array();
72
73 static public $mWeekdayMsgs = array(
74 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
75 'friday', 'saturday'
76 );
77
78 static public $mWeekdayAbbrevMsgs = array(
79 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
80 );
81
82 static public $mMonthMsgs = array(
83 'january', 'february', 'march', 'april', 'may_long', 'june',
84 'july', 'august', 'september', 'october', 'november',
85 'december'
86 );
87 static public $mMonthGenMsgs = array(
88 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
89 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
90 'december-gen'
91 );
92 static public $mMonthAbbrevMsgs = array(
93 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
94 'sep', 'oct', 'nov', 'dec'
95 );
96
97 static public $mIranianCalendarMonthMsgs = array(
98 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
99 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
100 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
101 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
102 );
103
104 static public $mHebrewCalendarMonthMsgs = array(
105 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
106 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
107 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
108 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
109 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
110 );
111
112 static public $mHebrewCalendarMonthGenMsgs = array(
113 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
114 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
115 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
116 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
117 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
118 );
119
120 static public $mHijriCalendarMonthMsgs = array(
121 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
122 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
123 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
124 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
125 );
126
127 /**
128 * Get a cached language object for a given language code
129 */
130 static function factory( $code ) {
131 if ( !isset( self::$mLangObjCache[$code] ) ) {
132 if ( count( self::$mLangObjCache ) > 10 ) {
133 // Don't keep a billion objects around, that's stupid.
134 self::$mLangObjCache = array();
135 }
136 self::$mLangObjCache[$code] = self::newFromCode( $code );
137 }
138 return self::$mLangObjCache[$code];
139 }
140
141 /**
142 * Create a language object for a given language code
143 */
144 protected static function newFromCode( $code ) {
145 global $IP;
146 static $recursionLevel = 0;
147 if ( $code == 'en' ) {
148 $class = 'Language';
149 } else {
150 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
151 // Preload base classes to work around APC/PHP5 bug
152 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
153 include_once( "$IP/languages/classes/$class.deps.php" );
154 }
155 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
156 include_once( "$IP/languages/classes/$class.php" );
157 }
158 }
159
160 if ( $recursionLevel > 5 ) {
161 throw new MWException( "Language fallback loop detected when creating class $class\n" );
162 }
163
164 if ( !class_exists( $class ) ) {
165 $fallback = Language::getFallbackFor( $code );
166 ++$recursionLevel;
167 $lang = Language::newFromCode( $fallback );
168 --$recursionLevel;
169 $lang->setCode( $code );
170 } else {
171 $lang = new $class;
172 }
173 return $lang;
174 }
175
176 /**
177 * Get the LocalisationCache instance
178 */
179 public static function getLocalisationCache() {
180 if ( is_null( self::$dataCache ) ) {
181 global $wgLocalisationCacheConf;
182 $class = $wgLocalisationCacheConf['class'];
183 self::$dataCache = new $class( $wgLocalisationCacheConf );
184 }
185 return self::$dataCache;
186 }
187
188 function __construct() {
189 $this->mConverter = new FakeConverter( $this );
190 // Set the code to the name of the descendant
191 if ( get_class( $this ) == 'Language' ) {
192 $this->mCode = 'en';
193 } else {
194 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
195 }
196 self::getLocalisationCache();
197 }
198
199 /**
200 * Reduce memory usage
201 */
202 function __destruct() {
203 foreach ( $this as $name => $value ) {
204 unset( $this->$name );
205 }
206 }
207
208 /**
209 * Hook which will be called if this is the content language.
210 * Descendants can use this to register hook functions or modify globals
211 */
212 function initContLang() { }
213
214 /**
215 * @deprecated Use User::getDefaultOptions()
216 * @return array
217 */
218 function getDefaultUserOptions() {
219 wfDeprecated( __METHOD__ );
220 return User::getDefaultOptions();
221 }
222
223 function getFallbackLanguageCode() {
224 if ( $this->mCode === 'en' ) {
225 return false;
226 } else {
227 return self::$dataCache->getItem( $this->mCode, 'fallback' );
228 }
229 }
230
231 /**
232 * Exports $wgBookstoreListEn
233 * @return array
234 */
235 function getBookstoreList() {
236 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
237 }
238
239 /**
240 * @return array
241 */
242 function getNamespaces() {
243 if ( is_null( $this->namespaceNames ) ) {
244 global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk;
245
246 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
247 if ( $wgExtraNamespaces ) {
248 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
249 }
250
251 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
252 if ( $wgMetaNamespaceTalk ) {
253 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
254 } else {
255 $talk = $this->namespaceNames[NS_PROJECT_TALK];
256 $this->namespaceNames[NS_PROJECT_TALK] =
257 $this->fixVariableInNamespace( $talk );
258 }
259
260 # Sometimes a language will be localised but not actually exist on this wiki.
261 global $wgCanonicalNamespaceNames;
262 $validNamespaces = array_keys($wgCanonicalNamespaceNames);
263 $validNamespaces[] = NS_MAIN;
264 foreach( $this->namespaceNames as $key => $text ) {
265 if ( ! in_array( $key, $validNamespaces ) ) {
266 unset( $this->namespaceNames[$key] );
267 }
268 }
269
270 # The above mixing may leave namespaces out of canonical order.
271 # Re-order by namespace ID number...
272 ksort( $this->namespaceNames );
273 }
274 return $this->namespaceNames;
275 }
276
277 /**
278 * A convenience function that returns the same thing as
279 * getNamespaces() except with the array values changed to ' '
280 * where it found '_', useful for producing output to be displayed
281 * e.g. in <select> forms.
282 *
283 * @return array
284 */
285 function getFormattedNamespaces() {
286 $ns = $this->getNamespaces();
287 foreach ( $ns as $k => $v ) {
288 $ns[$k] = strtr( $v, '_', ' ' );
289 }
290 return $ns;
291 }
292
293 /**
294 * Get a namespace value by key
295 * <code>
296 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
297 * echo $mw_ns; // prints 'MediaWiki'
298 * </code>
299 *
300 * @param $index Int: the array key of the namespace to return
301 * @return mixed, string if the namespace value exists, otherwise false
302 */
303 function getNsText( $index ) {
304 $ns = $this->getNamespaces();
305 return isset( $ns[$index] ) ? $ns[$index] : false;
306 }
307
308 /**
309 * A convenience function that returns the same thing as
310 * getNsText() except with '_' changed to ' ', useful for
311 * producing output.
312 *
313 * @return array
314 */
315 function getFormattedNsText( $index ) {
316 $ns = $this->getNsText( $index );
317 return strtr( $ns, '_', ' ' );
318 }
319
320 /**
321 * Get a namespace key by value, case insensitive.
322 * Only matches namespace names for the current language, not the
323 * canonical ones defined in Namespace.php.
324 *
325 * @param $text String
326 * @return mixed An integer if $text is a valid value otherwise false
327 */
328 function getLocalNsIndex( $text ) {
329 $lctext = $this->lc( $text );
330 $ids = $this->getNamespaceIds();
331 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
332 }
333
334 function getNamespaceAliases() {
335 if ( is_null( $this->namespaceAliases ) ) {
336 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
337 if ( !$aliases ) {
338 $aliases = array();
339 } else {
340 foreach ( $aliases as $name => $index ) {
341 if ( $index === NS_PROJECT_TALK ) {
342 unset( $aliases[$name] );
343 $name = $this->fixVariableInNamespace( $name );
344 $aliases[$name] = $index;
345 }
346 }
347 }
348 $this->namespaceAliases = $aliases;
349 }
350 return $this->namespaceAliases;
351 }
352
353 function getNamespaceIds() {
354 if ( is_null( $this->mNamespaceIds ) ) {
355 global $wgNamespaceAliases;
356 # Put namespace names and aliases into a hashtable.
357 # If this is too slow, then we should arrange it so that it is done
358 # before caching. The catch is that at pre-cache time, the above
359 # class-specific fixup hasn't been done.
360 $this->mNamespaceIds = array();
361 foreach ( $this->getNamespaces() as $index => $name ) {
362 $this->mNamespaceIds[$this->lc( $name )] = $index;
363 }
364 foreach ( $this->getNamespaceAliases() as $name => $index ) {
365 $this->mNamespaceIds[$this->lc( $name )] = $index;
366 }
367 if ( $wgNamespaceAliases ) {
368 foreach ( $wgNamespaceAliases as $name => $index ) {
369 $this->mNamespaceIds[$this->lc( $name )] = $index;
370 }
371 }
372 }
373 return $this->mNamespaceIds;
374 }
375
376
377 /**
378 * Get a namespace key by value, case insensitive. Canonical namespace
379 * names override custom ones defined for the current language.
380 *
381 * @param $text String
382 * @return mixed An integer if $text is a valid value otherwise false
383 */
384 function getNsIndex( $text ) {
385 $lctext = $this->lc( $text );
386 if ( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) {
387 return $ns;
388 }
389 $ids = $this->getNamespaceIds();
390 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
391 }
392
393 /**
394 * short names for language variants used for language conversion links.
395 *
396 * @param $code String
397 * @return string
398 */
399 function getVariantname( $code ) {
400 return $this->getMessageFromDB( "variantname-$code" );
401 }
402
403 function specialPage( $name ) {
404 $aliases = $this->getSpecialPageAliases();
405 if ( isset( $aliases[$name][0] ) ) {
406 $name = $aliases[$name][0];
407 }
408 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
409 }
410
411 function getQuickbarSettings() {
412 return array(
413 $this->getMessage( 'qbsettings-none' ),
414 $this->getMessage( 'qbsettings-fixedleft' ),
415 $this->getMessage( 'qbsettings-fixedright' ),
416 $this->getMessage( 'qbsettings-floatingleft' ),
417 $this->getMessage( 'qbsettings-floatingright' )
418 );
419 }
420
421 function getMathNames() {
422 return self::$dataCache->getItem( $this->mCode, 'mathNames' );
423 }
424
425 function getDatePreferences() {
426 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
427 }
428
429 function getDateFormats() {
430 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
431 }
432
433 function getDefaultDateFormat() {
434 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
435 if ( $df === 'dmy or mdy' ) {
436 global $wgAmericanDates;
437 return $wgAmericanDates ? 'mdy' : 'dmy';
438 } else {
439 return $df;
440 }
441 }
442
443 function getDatePreferenceMigrationMap() {
444 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
445 }
446
447 function getImageFile( $image ) {
448 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
449 }
450
451 function getDefaultUserOptionOverrides() {
452 return self::$dataCache->getItem( $this->mCode, 'defaultUserOptionOverrides' );
453 }
454
455 function getExtraUserToggles() {
456 return self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
457 }
458
459 function getUserToggle( $tog ) {
460 return $this->getMessageFromDB( "tog-$tog" );
461 }
462
463 /**
464 * Get language names, indexed by code.
465 * If $customisedOnly is true, only returns codes with a messages file
466 */
467 public static function getLanguageNames( $customisedOnly = false ) {
468 global $wgLanguageNames, $wgExtraLanguageNames;
469 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
470 if ( !$customisedOnly ) {
471 return $allNames;
472 }
473
474 global $IP;
475 $names = array();
476 $dir = opendir( "$IP/languages/messages" );
477 while ( false !== ( $file = readdir( $dir ) ) ) {
478 $code = self::getCodeFromFileName( $file, 'Messages' );
479 if ( $code && isset( $allNames[$code] ) ) {
480 $names[$code] = $allNames[$code];
481 }
482 }
483 closedir( $dir );
484 return $names;
485 }
486
487 /**
488 * Get a message from the MediaWiki namespace.
489 *
490 * @param $msg String: message name
491 * @return string
492 */
493 function getMessageFromDB( $msg ) {
494 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
495 }
496
497 function getLanguageName( $code ) {
498 $names = self::getLanguageNames();
499 if ( !array_key_exists( $code, $names ) ) {
500 return '';
501 }
502 return $names[$code];
503 }
504
505 function getMonthName( $key ) {
506 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
507 }
508
509 function getMonthNameGen( $key ) {
510 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
511 }
512
513 function getMonthAbbreviation( $key ) {
514 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
515 }
516
517 function getWeekdayName( $key ) {
518 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
519 }
520
521 function getWeekdayAbbreviation( $key ) {
522 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
523 }
524
525 function getIranianCalendarMonthName( $key ) {
526 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
527 }
528
529 function getHebrewCalendarMonthName( $key ) {
530 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
531 }
532
533 function getHebrewCalendarMonthNameGen( $key ) {
534 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
535 }
536
537 function getHijriCalendarMonthName( $key ) {
538 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
539 }
540
541 /**
542 * Used by date() and time() to adjust the time output.
543 *
544 * @param $ts Int the time in date('YmdHis') format
545 * @param $tz Mixed: adjust the time by this amount (default false, mean we
546 * get user timecorrection setting)
547 * @return int
548 */
549 function userAdjust( $ts, $tz = false ) {
550 global $wgUser, $wgLocalTZoffset;
551
552 if ( $tz === false ) {
553 $tz = $wgUser->getOption( 'timecorrection' );
554 }
555
556 $data = explode( '|', $tz, 3 );
557
558 if ( $data[0] == 'ZoneInfo' ) {
559 if ( function_exists( 'timezone_open' ) && @timezone_open( $data[2] ) !== false ) {
560 $date = date_create( $ts, timezone_open( 'UTC' ) );
561 date_timezone_set( $date, timezone_open( $data[2] ) );
562 $date = date_format( $date, 'YmdHis' );
563 return $date;
564 }
565 # Unrecognized timezone, default to 'Offset' with the stored offset.
566 $data[0] = 'Offset';
567 }
568
569 $minDiff = 0;
570 if ( $data[0] == 'System' || $tz == '' ) {
571 #  Global offset in minutes.
572 if ( isset( $wgLocalTZoffset ) ) {
573 $minDiff = $wgLocalTZoffset;
574 }
575 } else if ( $data[0] == 'Offset' ) {
576 $minDiff = intval( $data[1] );
577 } else {
578 $data = explode( ':', $tz );
579 if ( count( $data ) == 2 ) {
580 $data[0] = intval( $data[0] );
581 $data[1] = intval( $data[1] );
582 $minDiff = abs( $data[0] ) * 60 + $data[1];
583 if ( $data[0] < 0 ) {
584 $minDiff = -$minDiff;
585 }
586 } else {
587 $minDiff = intval( $data[0] ) * 60;
588 }
589 }
590
591 # No difference ? Return time unchanged
592 if ( 0 == $minDiff ) {
593 return $ts;
594 }
595
596 wfSuppressWarnings(); // E_STRICT system time bitching
597 # Generate an adjusted date; take advantage of the fact that mktime
598 # will normalize out-of-range values so we don't have to split $minDiff
599 # into hours and minutes.
600 $t = mktime( (
601 (int)substr( $ts, 8, 2 ) ), # Hours
602 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
603 (int)substr( $ts, 12, 2 ), # Seconds
604 (int)substr( $ts, 4, 2 ), # Month
605 (int)substr( $ts, 6, 2 ), # Day
606 (int)substr( $ts, 0, 4 ) ); # Year
607
608 $date = date( 'YmdHis', $t );
609 wfRestoreWarnings();
610
611 return $date;
612 }
613
614 /**
615 * This is a workalike of PHP's date() function, but with better
616 * internationalisation, a reduced set of format characters, and a better
617 * escaping format.
618 *
619 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
620 * PHP manual for definitions. "o" format character is supported since
621 * PHP 5.1.0, previous versions return literal o.
622 * There are a number of extensions, which start with "x":
623 *
624 * xn Do not translate digits of the next numeric format character
625 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
626 * xr Use roman numerals for the next numeric format character
627 * xh Use hebrew numerals for the next numeric format character
628 * xx Literal x
629 * xg Genitive month name
630 *
631 * xij j (day number) in Iranian calendar
632 * xiF F (month name) in Iranian calendar
633 * xin n (month number) in Iranian calendar
634 * xiY Y (full year) in Iranian calendar
635 *
636 * xjj j (day number) in Hebrew calendar
637 * xjF F (month name) in Hebrew calendar
638 * xjt t (days in month) in Hebrew calendar
639 * xjx xg (genitive month name) in Hebrew calendar
640 * xjn n (month number) in Hebrew calendar
641 * xjY Y (full year) in Hebrew calendar
642 *
643 * xmj j (day number) in Hijri calendar
644 * xmF F (month name) in Hijri calendar
645 * xmn n (month number) in Hijri calendar
646 * xmY Y (full year) in Hijri calendar
647 *
648 * xkY Y (full year) in Thai solar calendar. Months and days are
649 * identical to the Gregorian calendar
650 * xoY Y (full year) in Minguo calendar or Juche year.
651 * Months and days are identical to the
652 * Gregorian calendar
653 * xtY Y (full year) in Japanese nengo. Months and days are
654 * identical to the Gregorian calendar
655 *
656 * Characters enclosed in double quotes will be considered literal (with
657 * the quotes themselves removed). Unmatched quotes will be considered
658 * literal quotes. Example:
659 *
660 * "The month is" F => The month is January
661 * i's" => 20'11"
662 *
663 * Backslash escaping is also supported.
664 *
665 * Input timestamp is assumed to be pre-normalized to the desired local
666 * time zone, if any.
667 *
668 * @param $format String
669 * @param $ts String: 14-character timestamp
670 * YYYYMMDDHHMMSS
671 * 01234567890123
672 * @todo emulation of "o" format character for PHP pre 5.1.0
673 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
674 */
675 function sprintfDate( $format, $ts ) {
676 $s = '';
677 $raw = false;
678 $roman = false;
679 $hebrewNum = false;
680 $unix = false;
681 $rawToggle = false;
682 $iranian = false;
683 $hebrew = false;
684 $hijri = false;
685 $thai = false;
686 $minguo = false;
687 $tenno = false;
688 for ( $p = 0; $p < strlen( $format ); $p++ ) {
689 $num = false;
690 $code = $format[$p];
691 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
692 $code .= $format[++$p];
693 }
694
695 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
696 $code .= $format[++$p];
697 }
698
699 switch ( $code ) {
700 case 'xx':
701 $s .= 'x';
702 break;
703 case 'xn':
704 $raw = true;
705 break;
706 case 'xN':
707 $rawToggle = !$rawToggle;
708 break;
709 case 'xr':
710 $roman = true;
711 break;
712 case 'xh':
713 $hebrewNum = true;
714 break;
715 case 'xg':
716 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
717 break;
718 case 'xjx':
719 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
720 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
721 break;
722 case 'd':
723 $num = substr( $ts, 6, 2 );
724 break;
725 case 'D':
726 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
727 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
728 break;
729 case 'j':
730 $num = intval( substr( $ts, 6, 2 ) );
731 break;
732 case 'xij':
733 if ( !$iranian ) {
734 $iranian = self::tsToIranian( $ts );
735 }
736 $num = $iranian[2];
737 break;
738 case 'xmj':
739 if ( !$hijri ) {
740 $hijri = self::tsToHijri( $ts );
741 }
742 $num = $hijri[2];
743 break;
744 case 'xjj':
745 if ( !$hebrew ) {
746 $hebrew = self::tsToHebrew( $ts );
747 }
748 $num = $hebrew[2];
749 break;
750 case 'l':
751 if ( !$unix ) {
752 $unix = wfTimestamp( TS_UNIX, $ts );
753 }
754 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
755 break;
756 case 'N':
757 if ( !$unix ) {
758 $unix = wfTimestamp( TS_UNIX, $ts );
759 }
760 $w = gmdate( 'w', $unix );
761 $num = $w ? $w : 7;
762 break;
763 case 'w':
764 if ( !$unix ) {
765 $unix = wfTimestamp( TS_UNIX, $ts );
766 }
767 $num = gmdate( 'w', $unix );
768 break;
769 case 'z':
770 if ( !$unix ) {
771 $unix = wfTimestamp( TS_UNIX, $ts );
772 }
773 $num = gmdate( 'z', $unix );
774 break;
775 case 'W':
776 if ( !$unix ) {
777 $unix = wfTimestamp( TS_UNIX, $ts );
778 }
779 $num = gmdate( 'W', $unix );
780 break;
781 case 'F':
782 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
783 break;
784 case 'xiF':
785 if ( !$iranian ) {
786 $iranian = self::tsToIranian( $ts );
787 }
788 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
789 break;
790 case 'xmF':
791 if ( !$hijri ) {
792 $hijri = self::tsToHijri( $ts );
793 }
794 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
795 break;
796 case 'xjF':
797 if ( !$hebrew ) {
798 $hebrew = self::tsToHebrew( $ts );
799 }
800 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
801 break;
802 case 'm':
803 $num = substr( $ts, 4, 2 );
804 break;
805 case 'M':
806 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
807 break;
808 case 'n':
809 $num = intval( substr( $ts, 4, 2 ) );
810 break;
811 case 'xin':
812 if ( !$iranian ) {
813 $iranian = self::tsToIranian( $ts );
814 }
815 $num = $iranian[1];
816 break;
817 case 'xmn':
818 if ( !$hijri ) {
819 $hijri = self::tsToHijri ( $ts );
820 }
821 $num = $hijri[1];
822 break;
823 case 'xjn':
824 if ( !$hebrew ) {
825 $hebrew = self::tsToHebrew( $ts );
826 }
827 $num = $hebrew[1];
828 break;
829 case 't':
830 if ( !$unix ) {
831 $unix = wfTimestamp( TS_UNIX, $ts );
832 }
833 $num = gmdate( 't', $unix );
834 break;
835 case 'xjt':
836 if ( !$hebrew ) {
837 $hebrew = self::tsToHebrew( $ts );
838 }
839 $num = $hebrew[3];
840 break;
841 case 'L':
842 if ( !$unix ) {
843 $unix = wfTimestamp( TS_UNIX, $ts );
844 }
845 $num = gmdate( 'L', $unix );
846 break;
847 # 'o' is supported since PHP 5.1.0
848 # return literal if not supported
849 # TODO: emulation for pre 5.1.0 versions
850 case 'o':
851 if ( !$unix ) {
852 $unix = wfTimestamp( TS_UNIX, $ts );
853 }
854 if ( version_compare( PHP_VERSION, '5.1.0' ) === 1 ) {
855 $num = date( 'o', $unix );
856 } else {
857 $s .= 'o';
858 }
859 break;
860 case 'Y':
861 $num = substr( $ts, 0, 4 );
862 break;
863 case 'xiY':
864 if ( !$iranian ) {
865 $iranian = self::tsToIranian( $ts );
866 }
867 $num = $iranian[0];
868 break;
869 case 'xmY':
870 if ( !$hijri ) {
871 $hijri = self::tsToHijri( $ts );
872 }
873 $num = $hijri[0];
874 break;
875 case 'xjY':
876 if ( !$hebrew ) {
877 $hebrew = self::tsToHebrew( $ts );
878 }
879 $num = $hebrew[0];
880 break;
881 case 'xkY':
882 if ( !$thai ) {
883 $thai = self::tsToYear( $ts, 'thai' );
884 }
885 $num = $thai[0];
886 break;
887 case 'xoY':
888 if ( !$minguo ) {
889 $minguo = self::tsToYear( $ts, 'minguo' );
890 }
891 $num = $minguo[0];
892 break;
893 case 'xtY':
894 if ( !$tenno ) {
895 $tenno = self::tsToYear( $ts, 'tenno' );
896 }
897 $num = $tenno[0];
898 break;
899 case 'y':
900 $num = substr( $ts, 2, 2 );
901 break;
902 case 'a':
903 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
904 break;
905 case 'A':
906 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
907 break;
908 case 'g':
909 $h = substr( $ts, 8, 2 );
910 $num = $h % 12 ? $h % 12 : 12;
911 break;
912 case 'G':
913 $num = intval( substr( $ts, 8, 2 ) );
914 break;
915 case 'h':
916 $h = substr( $ts, 8, 2 );
917 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
918 break;
919 case 'H':
920 $num = substr( $ts, 8, 2 );
921 break;
922 case 'i':
923 $num = substr( $ts, 10, 2 );
924 break;
925 case 's':
926 $num = substr( $ts, 12, 2 );
927 break;
928 case 'c':
929 if ( !$unix ) {
930 $unix = wfTimestamp( TS_UNIX, $ts );
931 }
932 $s .= gmdate( 'c', $unix );
933 break;
934 case 'r':
935 if ( !$unix ) {
936 $unix = wfTimestamp( TS_UNIX, $ts );
937 }
938 $s .= gmdate( 'r', $unix );
939 break;
940 case 'U':
941 if ( !$unix ) {
942 $unix = wfTimestamp( TS_UNIX, $ts );
943 }
944 $num = $unix;
945 break;
946 case '\\':
947 # Backslash escaping
948 if ( $p < strlen( $format ) - 1 ) {
949 $s .= $format[++$p];
950 } else {
951 $s .= '\\';
952 }
953 break;
954 case '"':
955 # Quoted literal
956 if ( $p < strlen( $format ) - 1 ) {
957 $endQuote = strpos( $format, '"', $p + 1 );
958 if ( $endQuote === false ) {
959 # No terminating quote, assume literal "
960 $s .= '"';
961 } else {
962 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
963 $p = $endQuote;
964 }
965 } else {
966 # Quote at end of string, assume literal "
967 $s .= '"';
968 }
969 break;
970 default:
971 $s .= $format[$p];
972 }
973 if ( $num !== false ) {
974 if ( $rawToggle || $raw ) {
975 $s .= $num;
976 $raw = false;
977 } elseif ( $roman ) {
978 $s .= self::romanNumeral( $num );
979 $roman = false;
980 } elseif ( $hebrewNum ) {
981 $s .= self::hebrewNumeral( $num );
982 $hebrewNum = false;
983 } else {
984 $s .= $this->formatNum( $num, true );
985 }
986 $num = false;
987 }
988 }
989 return $s;
990 }
991
992 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
993 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
994 /**
995 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
996 * Gregorian dates to Iranian dates. Originally written in C, it
997 * is released under the terms of GNU Lesser General Public
998 * License. Conversion to PHP was performed by Niklas Laxström.
999 *
1000 * Link: http://www.farsiweb.info/jalali/jalali.c
1001 */
1002 private static function tsToIranian( $ts ) {
1003 $gy = substr( $ts, 0, 4 ) -1600;
1004 $gm = substr( $ts, 4, 2 ) -1;
1005 $gd = substr( $ts, 6, 2 ) -1;
1006
1007 # Days passed from the beginning (including leap years)
1008 $gDayNo = 365 * $gy
1009 + floor( ( $gy + 3 ) / 4 )
1010 - floor( ( $gy + 99 ) / 100 )
1011 + floor( ( $gy + 399 ) / 400 );
1012
1013
1014 // Add days of the past months of this year
1015 for ( $i = 0; $i < $gm; $i++ ) {
1016 $gDayNo += self::$GREG_DAYS[$i];
1017 }
1018
1019 // Leap years
1020 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1021 $gDayNo++;
1022 }
1023
1024 // Days passed in current month
1025 $gDayNo += $gd;
1026
1027 $jDayNo = $gDayNo - 79;
1028
1029 $jNp = floor( $jDayNo / 12053 );
1030 $jDayNo %= 12053;
1031
1032 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1033 $jDayNo %= 1461;
1034
1035 if ( $jDayNo >= 366 ) {
1036 $jy += floor( ( $jDayNo - 1 ) / 365 );
1037 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1038 }
1039
1040 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1041 $jDayNo -= self::$IRANIAN_DAYS[$i];
1042 }
1043
1044 $jm = $i + 1;
1045 $jd = $jDayNo + 1;
1046
1047 return array( $jy, $jm, $jd );
1048 }
1049
1050 /**
1051 * Converting Gregorian dates to Hijri dates.
1052 *
1053 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1054 *
1055 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1056 */
1057 private static function tsToHijri( $ts ) {
1058 $year = substr( $ts, 0, 4 );
1059 $month = substr( $ts, 4, 2 );
1060 $day = substr( $ts, 6, 2 );
1061
1062 $zyr = $year;
1063 $zd = $day;
1064 $zm = $month;
1065 $zy = $zyr;
1066
1067 if (
1068 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1069 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1070 )
1071 {
1072 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1073 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1074 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1075 $zd - 32075;
1076 } else {
1077 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1078 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1079 }
1080
1081 $zl = $zjd -1948440 + 10632;
1082 $zn = (int)( ( $zl - 1 ) / 10631 );
1083 $zl = $zl - 10631 * $zn + 354;
1084 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1085 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1086 $zm = (int)( ( 24 * $zl ) / 709 );
1087 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1088 $zy = 30 * $zn + $zj - 30;
1089
1090 return array( $zy, $zm, $zd );
1091 }
1092
1093 /**
1094 * Converting Gregorian dates to Hebrew dates.
1095 *
1096 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1097 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1098 * to translate the relevant functions into PHP and release them under
1099 * GNU GPL.
1100 *
1101 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1102 * and Adar II is 14. In a non-leap year, Adar is 6.
1103 */
1104 private static function tsToHebrew( $ts ) {
1105 # Parse date
1106 $year = substr( $ts, 0, 4 );
1107 $month = substr( $ts, 4, 2 );
1108 $day = substr( $ts, 6, 2 );
1109
1110 # Calculate Hebrew year
1111 $hebrewYear = $year + 3760;
1112
1113 # Month number when September = 1, August = 12
1114 $month += 4;
1115 if ( $month > 12 ) {
1116 # Next year
1117 $month -= 12;
1118 $year++;
1119 $hebrewYear++;
1120 }
1121
1122 # Calculate day of year from 1 September
1123 $dayOfYear = $day;
1124 for ( $i = 1; $i < $month; $i++ ) {
1125 if ( $i == 6 ) {
1126 # February
1127 $dayOfYear += 28;
1128 # Check if the year is leap
1129 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1130 $dayOfYear++;
1131 }
1132 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1133 $dayOfYear += 30;
1134 } else {
1135 $dayOfYear += 31;
1136 }
1137 }
1138
1139 # Calculate the start of the Hebrew year
1140 $start = self::hebrewYearStart( $hebrewYear );
1141
1142 # Calculate next year's start
1143 if ( $dayOfYear <= $start ) {
1144 # Day is before the start of the year - it is the previous year
1145 # Next year's start
1146 $nextStart = $start;
1147 # Previous year
1148 $year--;
1149 $hebrewYear--;
1150 # Add days since previous year's 1 September
1151 $dayOfYear += 365;
1152 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1153 # Leap year
1154 $dayOfYear++;
1155 }
1156 # Start of the new (previous) year
1157 $start = self::hebrewYearStart( $hebrewYear );
1158 } else {
1159 # Next year's start
1160 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1161 }
1162
1163 # Calculate Hebrew day of year
1164 $hebrewDayOfYear = $dayOfYear - $start;
1165
1166 # Difference between year's days
1167 $diff = $nextStart - $start;
1168 # Add 12 (or 13 for leap years) days to ignore the difference between
1169 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1170 # difference is only about the year type
1171 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1172 $diff += 13;
1173 } else {
1174 $diff += 12;
1175 }
1176
1177 # Check the year pattern, and is leap year
1178 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1179 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1180 # and non-leap years
1181 $yearPattern = $diff % 30;
1182 # Check if leap year
1183 $isLeap = $diff >= 30;
1184
1185 # Calculate day in the month from number of day in the Hebrew year
1186 # Don't check Adar - if the day is not in Adar, we will stop before;
1187 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1188 $hebrewDay = $hebrewDayOfYear;
1189 $hebrewMonth = 1;
1190 $days = 0;
1191 while ( $hebrewMonth <= 12 ) {
1192 # Calculate days in this month
1193 if ( $isLeap && $hebrewMonth == 6 ) {
1194 # Adar in a leap year
1195 if ( $isLeap ) {
1196 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1197 $days = 30;
1198 if ( $hebrewDay <= $days ) {
1199 # Day in Adar I
1200 $hebrewMonth = 13;
1201 } else {
1202 # Subtract the days of Adar I
1203 $hebrewDay -= $days;
1204 # Try Adar II
1205 $days = 29;
1206 if ( $hebrewDay <= $days ) {
1207 # Day in Adar II
1208 $hebrewMonth = 14;
1209 }
1210 }
1211 }
1212 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1213 # Cheshvan in a complete year (otherwise as the rule below)
1214 $days = 30;
1215 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1216 # Kislev in an incomplete year (otherwise as the rule below)
1217 $days = 29;
1218 } else {
1219 # Odd months have 30 days, even have 29
1220 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1221 }
1222 if ( $hebrewDay <= $days ) {
1223 # In the current month
1224 break;
1225 } else {
1226 # Subtract the days of the current month
1227 $hebrewDay -= $days;
1228 # Try in the next month
1229 $hebrewMonth++;
1230 }
1231 }
1232
1233 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1234 }
1235
1236 /**
1237 * This calculates the Hebrew year start, as days since 1 September.
1238 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1239 * Used for Hebrew date.
1240 */
1241 private static function hebrewYearStart( $year ) {
1242 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1243 $b = intval( ( $year - 1 ) % 4 );
1244 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1245 if ( $m < 0 ) {
1246 $m--;
1247 }
1248 $Mar = intval( $m );
1249 if ( $m < 0 ) {
1250 $m++;
1251 }
1252 $m -= $Mar;
1253
1254 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1255 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1256 $Mar++;
1257 } else if ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1258 $Mar += 2;
1259 } else if ( $c == 2 || $c == 4 || $c == 6 ) {
1260 $Mar++;
1261 }
1262
1263 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1264 return $Mar;
1265 }
1266
1267 /**
1268 * Algorithm to convert Gregorian dates to Thai solar dates,
1269 * Minguo dates or Minguo dates.
1270 *
1271 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1272 * http://en.wikipedia.org/wiki/Minguo_calendar
1273 * http://en.wikipedia.org/wiki/Japanese_era_name
1274 *
1275 * @param $ts String: 14-character timestamp
1276 * @param $cName String: calender name
1277 * @return Array: converted year, month, day
1278 */
1279 private static function tsToYear( $ts, $cName ) {
1280 $gy = substr( $ts, 0, 4 );
1281 $gm = substr( $ts, 4, 2 );
1282 $gd = substr( $ts, 6, 2 );
1283
1284 if ( !strcmp( $cName, 'thai' ) ) {
1285 # Thai solar dates
1286 # Add 543 years to the Gregorian calendar
1287 # Months and days are identical
1288 $gy_offset = $gy + 543;
1289 } else if ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1290 # Minguo dates
1291 # Deduct 1911 years from the Gregorian calendar
1292 # Months and days are identical
1293 $gy_offset = $gy - 1911;
1294 } else if ( !strcmp( $cName, 'tenno' ) ) {
1295 # Nengō dates up to Meiji period
1296 # Deduct years from the Gregorian calendar
1297 # depending on the nengo periods
1298 # Months and days are identical
1299 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1300 # Meiji period
1301 $gy_gannen = $gy - 1868 + 1;
1302 $gy_offset = $gy_gannen;
1303 if ( $gy_gannen == 1 ) {
1304 $gy_offset = '元';
1305 }
1306 $gy_offset = '明治' . $gy_offset;
1307 } else if (
1308 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1309 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1310 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1311 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1312 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1313 )
1314 {
1315 # Taishō period
1316 $gy_gannen = $gy - 1912 + 1;
1317 $gy_offset = $gy_gannen;
1318 if ( $gy_gannen == 1 ) {
1319 $gy_offset = '元';
1320 }
1321 $gy_offset = '大正' . $gy_offset;
1322 } else if (
1323 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1324 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1325 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1326 )
1327 {
1328 # Shōwa period
1329 $gy_gannen = $gy - 1926 + 1;
1330 $gy_offset = $gy_gannen;
1331 if ( $gy_gannen == 1 ) {
1332 $gy_offset = '元';
1333 }
1334 $gy_offset = '昭和' . $gy_offset;
1335 } else {
1336 # Heisei period
1337 $gy_gannen = $gy - 1989 + 1;
1338 $gy_offset = $gy_gannen;
1339 if ( $gy_gannen == 1 ) {
1340 $gy_offset = '元';
1341 }
1342 $gy_offset = '平成' . $gy_offset;
1343 }
1344 } else {
1345 $gy_offset = $gy;
1346 }
1347
1348 return array( $gy_offset, $gm, $gd );
1349 }
1350
1351 /**
1352 * Roman number formatting up to 3000
1353 */
1354 static function romanNumeral( $num ) {
1355 static $table = array(
1356 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1357 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1358 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1359 array( '', 'M', 'MM', 'MMM' )
1360 );
1361
1362 $num = intval( $num );
1363 if ( $num > 3000 || $num <= 0 ) {
1364 return $num;
1365 }
1366
1367 $s = '';
1368 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1369 if ( $num >= $pow10 ) {
1370 $s .= $table[$i][floor( $num / $pow10 )];
1371 }
1372 $num = $num % $pow10;
1373 }
1374 return $s;
1375 }
1376
1377 /**
1378 * Hebrew Gematria number formatting up to 9999
1379 */
1380 static function hebrewNumeral( $num ) {
1381 static $table = array(
1382 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1383 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1384 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1385 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1386 );
1387
1388 $num = intval( $num );
1389 if ( $num > 9999 || $num <= 0 ) {
1390 return $num;
1391 }
1392
1393 $s = '';
1394 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1395 if ( $num >= $pow10 ) {
1396 if ( $num == 15 || $num == 16 ) {
1397 $s .= $table[0][9] . $table[0][$num - 9];
1398 $num = 0;
1399 } else {
1400 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1401 if ( $pow10 == 1000 ) {
1402 $s .= "'";
1403 }
1404 }
1405 }
1406 $num = $num % $pow10;
1407 }
1408 if ( strlen( $s ) == 2 ) {
1409 $str = $s . "'";
1410 } else {
1411 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1412 $str .= substr( $s, strlen( $s ) - 2, 2 );
1413 }
1414 $start = substr( $str, 0, strlen( $str ) - 2 );
1415 $end = substr( $str, strlen( $str ) - 2 );
1416 switch( $end ) {
1417 case 'כ':
1418 $str = $start . 'ך';
1419 break;
1420 case 'מ':
1421 $str = $start . 'ם';
1422 break;
1423 case 'נ':
1424 $str = $start . 'ן';
1425 break;
1426 case 'פ':
1427 $str = $start . 'ף';
1428 break;
1429 case 'צ':
1430 $str = $start . 'ץ';
1431 break;
1432 }
1433 return $str;
1434 }
1435
1436 /**
1437 * This is meant to be used by time(), date(), and timeanddate() to get
1438 * the date preference they're supposed to use, it should be used in
1439 * all children.
1440 *
1441 *<code>
1442 * function timeanddate([...], $format = true) {
1443 * $datePreference = $this->dateFormat($format);
1444 * [...]
1445 * }
1446 *</code>
1447 *
1448 * @param $usePrefs Mixed: if true, the user's preference is used
1449 * if false, the site/language default is used
1450 * if int/string, assumed to be a format.
1451 * @return string
1452 */
1453 function dateFormat( $usePrefs = true ) {
1454 global $wgUser;
1455
1456 if ( is_bool( $usePrefs ) ) {
1457 if ( $usePrefs ) {
1458 $datePreference = $wgUser->getDatePreference();
1459 } else {
1460 $datePreference = (string)User::getDefaultOption( 'date' );
1461 }
1462 } else {
1463 $datePreference = (string)$usePrefs;
1464 }
1465
1466 // return int
1467 if ( $datePreference == '' ) {
1468 return 'default';
1469 }
1470
1471 return $datePreference;
1472 }
1473
1474 /**
1475 * Get a format string for a given type and preference
1476 * @param $type May be date, time or both
1477 * @param $pref The format name as it appears in Messages*.php
1478 */
1479 function getDateFormatString( $type, $pref ) {
1480 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
1481 if ( $pref == 'default' ) {
1482 $pref = $this->getDefaultDateFormat();
1483 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1484 } else {
1485 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1486 if ( is_null( $df ) ) {
1487 $pref = $this->getDefaultDateFormat();
1488 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1489 }
1490 }
1491 $this->dateFormatStrings[$type][$pref] = $df;
1492 }
1493 return $this->dateFormatStrings[$type][$pref];
1494 }
1495
1496 /**
1497 * @param $ts Mixed: the time format which needs to be turned into a
1498 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1499 * @param $adj Bool: whether to adjust the time output according to the
1500 * user configured offset ($timecorrection)
1501 * @param $format Mixed: true to use user's date format preference
1502 * @param $timecorrection String: the time offset as returned by
1503 * validateTimeZone() in Special:Preferences
1504 * @return string
1505 */
1506 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1507 if ( $adj ) {
1508 $ts = $this->userAdjust( $ts, $timecorrection );
1509 }
1510 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1511 return $this->sprintfDate( $df, $ts );
1512 }
1513
1514 /**
1515 * @param $ts Mixed: the time format which needs to be turned into a
1516 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1517 * @param $adj Bool: whether to adjust the time output according to the
1518 * user configured offset ($timecorrection)
1519 * @param $format Mixed: true to use user's date format preference
1520 * @param $timecorrection String: the time offset as returned by
1521 * validateTimeZone() in Special:Preferences
1522 * @return string
1523 */
1524 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1525 if ( $adj ) {
1526 $ts = $this->userAdjust( $ts, $timecorrection );
1527 }
1528 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1529 return $this->sprintfDate( $df, $ts );
1530 }
1531
1532 /**
1533 * @param $ts Mixed: the time format which needs to be turned into a
1534 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1535 * @param $adj Bool: whether to adjust the time output according to the
1536 * user configured offset ($timecorrection)
1537 * @param $format Mixed: what format to return, if it's false output the
1538 * default one (default true)
1539 * @param $timecorrection String: the time offset as returned by
1540 * validateTimeZone() in Special:Preferences
1541 * @return string
1542 */
1543 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1544 $ts = wfTimestamp( TS_MW, $ts );
1545 if ( $adj ) {
1546 $ts = $this->userAdjust( $ts, $timecorrection );
1547 }
1548 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1549 return $this->sprintfDate( $df, $ts );
1550 }
1551
1552 function getMessage( $key ) {
1553 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
1554 }
1555
1556 function getAllMessages() {
1557 return self::$dataCache->getItem( $this->mCode, 'messages' );
1558 }
1559
1560 function iconv( $in, $out, $string ) {
1561 # This is a wrapper for iconv in all languages except esperanto,
1562 # which does some nasty x-conversions beforehand
1563
1564 # Even with //IGNORE iconv can whine about illegal characters in
1565 # *input* string. We just ignore those too.
1566 # REF: http://bugs.php.net/bug.php?id=37166
1567 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1568 wfSuppressWarnings();
1569 $text = iconv( $in, $out . '//IGNORE', $string );
1570 wfRestoreWarnings();
1571 return $text;
1572 }
1573
1574 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1575 function ucwordbreaksCallbackAscii( $matches ) {
1576 return $this->ucfirst( $matches[1] );
1577 }
1578
1579 function ucwordbreaksCallbackMB( $matches ) {
1580 return mb_strtoupper( $matches[0] );
1581 }
1582
1583 function ucCallback( $matches ) {
1584 list( $wikiUpperChars ) = self::getCaseMaps();
1585 return strtr( $matches[1], $wikiUpperChars );
1586 }
1587
1588 function lcCallback( $matches ) {
1589 list( , $wikiLowerChars ) = self::getCaseMaps();
1590 return strtr( $matches[1], $wikiLowerChars );
1591 }
1592
1593 function ucwordsCallbackMB( $matches ) {
1594 return mb_strtoupper( $matches[0] );
1595 }
1596
1597 function ucwordsCallbackWiki( $matches ) {
1598 list( $wikiUpperChars ) = self::getCaseMaps();
1599 return strtr( $matches[0], $wikiUpperChars );
1600 }
1601
1602 function ucfirst( $str ) {
1603 $o = ord( $str );
1604 if ( $o < 96 ) {
1605 return $str;
1606 } elseif ( $o < 128 ) {
1607 return ucfirst( $str );
1608 } else {
1609 // fall back to more complex logic in case of multibyte strings
1610 return $this->uc( $str, true );
1611 }
1612 }
1613
1614 function uc( $str, $first = false ) {
1615 if ( function_exists( 'mb_strtoupper' ) ) {
1616 if ( $first ) {
1617 if ( $this->isMultibyte( $str ) ) {
1618 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1619 } else {
1620 return ucfirst( $str );
1621 }
1622 } else {
1623 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1624 }
1625 } else {
1626 if ( $this->isMultibyte( $str ) ) {
1627 $x = $first ? '^' : '';
1628 return preg_replace_callback(
1629 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1630 array( $this, 'ucCallback' ),
1631 $str
1632 );
1633 } else {
1634 return $first ? ucfirst( $str ) : strtoupper( $str );
1635 }
1636 }
1637 }
1638
1639 function lcfirst( $str ) {
1640 $o = ord( $str );
1641 if ( !$o ) {
1642 return strval( $str );
1643 } elseif ( $o >= 128 ) {
1644 return $this->lc( $str, true );
1645 } elseif ( $o > 96 ) {
1646 return $str;
1647 } else {
1648 $str[0] = strtolower( $str[0] );
1649 return $str;
1650 }
1651 }
1652
1653 function lc( $str, $first = false ) {
1654 if ( function_exists( 'mb_strtolower' ) ) {
1655 if ( $first ) {
1656 if ( $this->isMultibyte( $str ) ) {
1657 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1658 } else {
1659 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1660 }
1661 } else {
1662 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1663 }
1664 } else {
1665 if ( $this->isMultibyte( $str ) ) {
1666 $x = $first ? '^' : '';
1667 return preg_replace_callback(
1668 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1669 array( $this, 'lcCallback' ),
1670 $str
1671 );
1672 } else {
1673 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1674 }
1675 }
1676 }
1677
1678 function isMultibyte( $str ) {
1679 return (bool)preg_match( '/[\x80-\xff]/', $str );
1680 }
1681
1682 function ucwords( $str ) {
1683 if ( $this->isMultibyte( $str ) ) {
1684 $str = $this->lc( $str );
1685
1686 // regexp to find first letter in each word (i.e. after each space)
1687 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1688
1689 // function to use to capitalize a single char
1690 if ( function_exists( 'mb_strtoupper' ) ) {
1691 return preg_replace_callback(
1692 $replaceRegexp,
1693 array( $this, 'ucwordsCallbackMB' ),
1694 $str
1695 );
1696 } else {
1697 return preg_replace_callback(
1698 $replaceRegexp,
1699 array( $this, 'ucwordsCallbackWiki' ),
1700 $str
1701 );
1702 }
1703 } else {
1704 return ucwords( strtolower( $str ) );
1705 }
1706 }
1707
1708 # capitalize words at word breaks
1709 function ucwordbreaks( $str ) {
1710 if ( $this->isMultibyte( $str ) ) {
1711 $str = $this->lc( $str );
1712
1713 // since \b doesn't work for UTF-8, we explicitely define word break chars
1714 $breaks = "[ \-\(\)\}\{\.,\?!]";
1715
1716 // find first letter after word break
1717 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1718
1719 if ( function_exists( 'mb_strtoupper' ) ) {
1720 return preg_replace_callback(
1721 $replaceRegexp,
1722 array( $this, 'ucwordbreaksCallbackMB' ),
1723 $str
1724 );
1725 } else {
1726 return preg_replace_callback(
1727 $replaceRegexp,
1728 array( $this, 'ucwordsCallbackWiki' ),
1729 $str
1730 );
1731 }
1732 } else {
1733 return preg_replace_callback(
1734 '/\b([\w\x80-\xff]+)\b/',
1735 array( $this, 'ucwordbreaksCallbackAscii' ),
1736 $str
1737 );
1738 }
1739 }
1740
1741 /**
1742 * Return a case-folded representation of $s
1743 *
1744 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1745 * and $s2 are the same except for the case of their characters. It is not
1746 * necessary for the value returned to make sense when displayed.
1747 *
1748 * Do *not* perform any other normalisation in this function. If a caller
1749 * uses this function when it should be using a more general normalisation
1750 * function, then fix the caller.
1751 */
1752 function caseFold( $s ) {
1753 return $this->uc( $s );
1754 }
1755
1756 function checkTitleEncoding( $s ) {
1757 if ( is_array( $s ) ) {
1758 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1759 }
1760 # Check for non-UTF-8 URLs
1761 $ishigh = preg_match( '/[\x80-\xff]/', $s );
1762 if ( !$ishigh ) {
1763 return $s;
1764 }
1765
1766 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1767 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1768 if ( $isutf8 ) {
1769 return $s;
1770 }
1771
1772 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
1773 }
1774
1775 function fallback8bitEncoding() {
1776 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
1777 }
1778
1779 /**
1780 * Most writing systems use whitespace to break up words.
1781 * Some languages such as Chinese don't conventionally do this,
1782 * which requires special handling when breaking up words for
1783 * searching etc.
1784 */
1785 function hasWordBreaks() {
1786 return true;
1787 }
1788
1789 /**
1790 * Some languages such as Chinese require word segmentation,
1791 * Specify such segmentation when overridden in derived class.
1792 *
1793 * @param $string String
1794 * @return String
1795 */
1796 function segmentByWord( $string ) {
1797 return $string;
1798 }
1799
1800 /**
1801 * Some languages have special punctuation need to be normalized.
1802 * Make such changes here.
1803 *
1804 * @param $string String
1805 * @return String
1806 */
1807 function normalizeForSearch( $string ) {
1808 return self::convertDoubleWidth( $string );
1809 }
1810
1811 /**
1812 * convert double-width roman characters to single-width.
1813 * range: ff00-ff5f ~= 0020-007f
1814 */
1815 protected static function convertDoubleWidth( $string ) {
1816 static $full = null;
1817 static $half = null;
1818
1819 if ( $full === null ) {
1820 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
1821 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1822 $full = str_split( $fullWidth, 3 );
1823 $half = str_split( $halfWidth );
1824 }
1825
1826 $string = str_replace( $full, $half, $string );
1827 return $string;
1828 }
1829
1830 protected static function insertSpace( $string, $pattern ) {
1831 $string = preg_replace( $pattern, " $1 ", $string );
1832 $string = preg_replace( '/ +/', ' ', $string );
1833 return $string;
1834 }
1835
1836 function convertForSearchResult( $termsArray ) {
1837 # some languages, e.g. Chinese, need to do a conversion
1838 # in order for search results to be displayed correctly
1839 return $termsArray;
1840 }
1841
1842 /**
1843 * Get the first character of a string.
1844 *
1845 * @param $s string
1846 * @return string
1847 */
1848 function firstChar( $s ) {
1849 $matches = array();
1850 preg_match(
1851 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1852 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
1853 $s,
1854 $matches
1855 );
1856
1857 if ( isset( $matches[1] ) ) {
1858 if ( strlen( $matches[1] ) != 3 ) {
1859 return $matches[1];
1860 }
1861
1862 // Break down Hangul syllables to grab the first jamo
1863 $code = utf8ToCodepoint( $matches[1] );
1864 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
1865 return $matches[1];
1866 } elseif ( $code < 0xb098 ) {
1867 return "\xe3\x84\xb1";
1868 } elseif ( $code < 0xb2e4 ) {
1869 return "\xe3\x84\xb4";
1870 } elseif ( $code < 0xb77c ) {
1871 return "\xe3\x84\xb7";
1872 } elseif ( $code < 0xb9c8 ) {
1873 return "\xe3\x84\xb9";
1874 } elseif ( $code < 0xbc14 ) {
1875 return "\xe3\x85\x81";
1876 } elseif ( $code < 0xc0ac ) {
1877 return "\xe3\x85\x82";
1878 } elseif ( $code < 0xc544 ) {
1879 return "\xe3\x85\x85";
1880 } elseif ( $code < 0xc790 ) {
1881 return "\xe3\x85\x87";
1882 } elseif ( $code < 0xcc28 ) {
1883 return "\xe3\x85\x88";
1884 } elseif ( $code < 0xce74 ) {
1885 return "\xe3\x85\x8a";
1886 } elseif ( $code < 0xd0c0 ) {
1887 return "\xe3\x85\x8b";
1888 } elseif ( $code < 0xd30c ) {
1889 return "\xe3\x85\x8c";
1890 } elseif ( $code < 0xd558 ) {
1891 return "\xe3\x85\x8d";
1892 } else {
1893 return "\xe3\x85\x8e";
1894 }
1895 } else {
1896 return '';
1897 }
1898 }
1899
1900 function initEncoding() {
1901 # Some languages may have an alternate char encoding option
1902 # (Esperanto X-coding, Japanese furigana conversion, etc)
1903 # If this language is used as the primary content language,
1904 # an override to the defaults can be set here on startup.
1905 }
1906
1907 function recodeForEdit( $s ) {
1908 # For some languages we'll want to explicitly specify
1909 # which characters make it into the edit box raw
1910 # or are converted in some way or another.
1911 # Note that if wgOutputEncoding is different from
1912 # wgInputEncoding, this text will be further converted
1913 # to wgOutputEncoding.
1914 global $wgEditEncoding;
1915 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
1916 return $s;
1917 } else {
1918 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1919 }
1920 }
1921
1922 function recodeInput( $s ) {
1923 # Take the previous into account.
1924 global $wgEditEncoding;
1925 if ( $wgEditEncoding != '' ) {
1926 $enc = $wgEditEncoding;
1927 } else {
1928 $enc = 'UTF-8';
1929 }
1930 if ( $enc == 'UTF-8' ) {
1931 return $s;
1932 } else {
1933 return $this->iconv( $enc, 'UTF-8', $s );
1934 }
1935 }
1936
1937 /**
1938 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
1939 * also cleans up certain backwards-compatible sequences, converting them
1940 * to the modern Unicode equivalent.
1941 *
1942 * This is language-specific for performance reasons only.
1943 */
1944 function normalize( $s ) {
1945 global $wgAllUnicodeFixes;
1946 $s = UtfNormal::cleanUp( $s );
1947 if ( $wgAllUnicodeFixes ) {
1948 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
1949 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
1950 }
1951
1952 return $s;
1953 }
1954
1955 /**
1956 * Transform a string using serialized data stored in the given file (which
1957 * must be in the serialized subdirectory of $IP). The file contains pairs
1958 * mapping source characters to destination characters.
1959 *
1960 * The data is cached in process memory. This will go faster if you have the
1961 * FastStringSearch extension.
1962 */
1963 function transformUsingPairFile( $file, $string ) {
1964 if ( !isset( $this->transformData[$file] ) ) {
1965 $data = wfGetPrecompiledData( $file );
1966 if ( $data === false ) {
1967 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
1968 }
1969 $this->transformData[$file] = new ReplacementArray( $data );
1970 }
1971 return $this->transformData[$file]->replace( $string );
1972 }
1973
1974 /**
1975 * For right-to-left language support
1976 *
1977 * @return bool
1978 */
1979 function isRTL() {
1980 return self::$dataCache->getItem( $this->mCode, 'rtl' );
1981 }
1982
1983 /**
1984 * Return the correct HTML 'dir' attribute value for this language.
1985 * @return String
1986 */
1987 function getDir() {
1988 return $this->isRTL() ? 'rtl' : 'ltr';
1989 }
1990
1991 /**
1992 * Return 'left' or 'right' as appropriate alignment for line-start
1993 * for this language's text direction.
1994 *
1995 * Should be equivalent to CSS3 'start' text-align value....
1996 *
1997 * @return String
1998 */
1999 function alignStart() {
2000 return $this->isRTL() ? 'right' : 'left';
2001 }
2002
2003 /**
2004 * Return 'right' or 'left' as appropriate alignment for line-end
2005 * for this language's text direction.
2006 *
2007 * Should be equivalent to CSS3 'end' text-align value....
2008 *
2009 * @return String
2010 */
2011 function alignEnd() {
2012 return $this->isRTL() ? 'left' : 'right';
2013 }
2014
2015 /**
2016 * A hidden direction mark (LRM or RLM), depending on the language direction
2017 *
2018 * @return string
2019 */
2020 function getDirMark() {
2021 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
2022 }
2023
2024 function capitalizeAllNouns() {
2025 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2026 }
2027
2028 /**
2029 * An arrow, depending on the language direction
2030 *
2031 * @return string
2032 */
2033 function getArrow() {
2034 return $this->isRTL() ? '←' : '→';
2035 }
2036
2037 /**
2038 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2039 *
2040 * @return bool
2041 */
2042 function linkPrefixExtension() {
2043 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2044 }
2045
2046 function getMagicWords() {
2047 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2048 }
2049
2050 # Fill a MagicWord object with data from here
2051 function getMagic( $mw ) {
2052 if ( !$this->mMagicHookDone ) {
2053 $this->mMagicHookDone = true;
2054 wfProfileIn( 'LanguageGetMagic' );
2055 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2056 wfProfileOut( 'LanguageGetMagic' );
2057 }
2058 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
2059 $rawEntry = $this->mMagicExtensions[$mw->mId];
2060 } else {
2061 $magicWords = $this->getMagicWords();
2062 if ( isset( $magicWords[$mw->mId] ) ) {
2063 $rawEntry = $magicWords[$mw->mId];
2064 } else {
2065 $rawEntry = false;
2066 }
2067 }
2068
2069 if ( !is_array( $rawEntry ) ) {
2070 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
2071 } else {
2072 $mw->mCaseSensitive = $rawEntry[0];
2073 $mw->mSynonyms = array_slice( $rawEntry, 1 );
2074 }
2075 }
2076
2077 /**
2078 * Add magic words to the extension array
2079 */
2080 function addMagicWordsByLang( $newWords ) {
2081 $code = $this->getCode();
2082 $fallbackChain = array();
2083 while ( $code && !in_array( $code, $fallbackChain ) ) {
2084 $fallbackChain[] = $code;
2085 $code = self::getFallbackFor( $code );
2086 }
2087 if ( !in_array( 'en', $fallbackChain ) ) {
2088 $fallbackChain[] = 'en';
2089 }
2090 $fallbackChain = array_reverse( $fallbackChain );
2091 foreach ( $fallbackChain as $code ) {
2092 if ( isset( $newWords[$code] ) ) {
2093 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
2094 }
2095 }
2096 }
2097
2098 /**
2099 * Get special page names, as an associative array
2100 * case folded alias => real name
2101 */
2102 function getSpecialPageAliases() {
2103 // Cache aliases because it may be slow to load them
2104 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
2105 // Initialise array
2106 $this->mExtendedSpecialPageAliases =
2107 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
2108 wfRunHooks( 'LanguageGetSpecialPageAliases',
2109 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
2110 }
2111
2112 return $this->mExtendedSpecialPageAliases;
2113 }
2114
2115 /**
2116 * Italic is unsuitable for some languages
2117 *
2118 * @param $text String: the text to be emphasized.
2119 * @return string
2120 */
2121 function emphasize( $text ) {
2122 return "<em>$text</em>";
2123 }
2124
2125 /**
2126 * Normally we output all numbers in plain en_US style, that is
2127 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2128 * point twohundredthirtyfive. However this is not sutable for all
2129 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2130 * Icelandic just want to use commas instead of dots, and dots instead
2131 * of commas like "293.291,235".
2132 *
2133 * An example of this function being called:
2134 * <code>
2135 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2136 * </code>
2137 *
2138 * See LanguageGu.php for the Gujarati implementation and
2139 * $separatorTransformTable on MessageIs.php for
2140 * the , => . and . => , implementation.
2141 *
2142 * @todo check if it's viable to use localeconv() for the decimal
2143 * separator thing.
2144 * @param $number Mixed: the string to be formatted, should be an integer
2145 * or a floating point number.
2146 * @param $nocommafy Bool: set to true for special numbers like dates
2147 * @return string
2148 */
2149 function formatNum( $number, $nocommafy = false ) {
2150 global $wgTranslateNumerals;
2151 if ( !$nocommafy ) {
2152 $number = $this->commafy( $number );
2153 $s = $this->separatorTransformTable();
2154 if ( $s ) {
2155 $number = strtr( $number, $s );
2156 }
2157 }
2158
2159 if ( $wgTranslateNumerals ) {
2160 $s = $this->digitTransformTable();
2161 if ( $s ) {
2162 $number = strtr( $number, $s );
2163 }
2164 }
2165
2166 return $number;
2167 }
2168
2169 function parseFormattedNumber( $number ) {
2170 $s = $this->digitTransformTable();
2171 if ( $s ) {
2172 $number = strtr( $number, array_flip( $s ) );
2173 }
2174
2175 $s = $this->separatorTransformTable();
2176 if ( $s ) {
2177 $number = strtr( $number, array_flip( $s ) );
2178 }
2179
2180 $number = strtr( $number, array( ',' => '' ) );
2181 return $number;
2182 }
2183
2184 /**
2185 * Adds commas to a given number
2186 *
2187 * @param $_ mixed
2188 * @return string
2189 */
2190 function commafy( $_ ) {
2191 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2192 }
2193
2194 function digitTransformTable() {
2195 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
2196 }
2197
2198 function separatorTransformTable() {
2199 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
2200 }
2201
2202 /**
2203 * Take a list of strings and build a locale-friendly comma-separated
2204 * list, using the local comma-separator message.
2205 * The last two strings are chained with an "and".
2206 *
2207 * @param $l Array
2208 * @return string
2209 */
2210 function listToText( $l ) {
2211 $s = '';
2212 $m = count( $l ) - 1;
2213 if ( $m == 1 ) {
2214 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2215 } else {
2216 for ( $i = $m; $i >= 0; $i-- ) {
2217 if ( $i == $m ) {
2218 $s = $l[$i];
2219 } else if ( $i == $m - 1 ) {
2220 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2221 } else {
2222 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2223 }
2224 }
2225 return $s;
2226 }
2227 }
2228
2229 /**
2230 * Take a list of strings and build a locale-friendly comma-separated
2231 * list, using the local comma-separator message.
2232 * @param $list array of strings to put in a comma list
2233 * @return string
2234 */
2235 function commaList( $list ) {
2236 return implode(
2237 $list,
2238 wfMsgExt(
2239 'comma-separator',
2240 array( 'parsemag', 'escapenoentities', 'language' => $this )
2241 )
2242 );
2243 }
2244
2245 /**
2246 * Take a list of strings and build a locale-friendly semicolon-separated
2247 * list, using the local semicolon-separator message.
2248 * @param $list array of strings to put in a semicolon list
2249 * @return string
2250 */
2251 function semicolonList( $list ) {
2252 return implode(
2253 $list,
2254 wfMsgExt(
2255 'semicolon-separator',
2256 array( 'parsemag', 'escapenoentities', 'language' => $this )
2257 )
2258 );
2259 }
2260
2261 /**
2262 * Same as commaList, but separate it with the pipe instead.
2263 * @param $list array of strings to put in a pipe list
2264 * @return string
2265 */
2266 function pipeList( $list ) {
2267 return implode(
2268 $list,
2269 wfMsgExt(
2270 'pipe-separator',
2271 array( 'escapenoentities', 'language' => $this )
2272 )
2273 );
2274 }
2275
2276 /**
2277 * Truncate a string to a specified length in bytes, appending an optional
2278 * string (e.g. for ellipses)
2279 *
2280 * The database offers limited byte lengths for some columns in the database;
2281 * multi-byte character sets mean we need to ensure that only whole characters
2282 * are included, otherwise broken characters can be passed to the user
2283 *
2284 * If $length is negative, the string will be truncated from the beginning
2285 *
2286 * @param $string String to truncate
2287 * @param $length Int: maximum length (excluding ellipses)
2288 * @param $ellipsis String to append to the truncated text
2289 * @return string
2290 */
2291 function truncate( $string, $length, $ellipsis = '...' ) {
2292 # Use the localized ellipsis character
2293 if ( $ellipsis == '...' ) {
2294 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2295 }
2296 # Check if there is no need to truncate
2297 if ( $length == 0 ) {
2298 return $ellipsis;
2299 } elseif ( strlen( $string ) <= abs( $length ) ) {
2300 return $string;
2301 }
2302 $stringOriginal = $string;
2303 if ( $length > 0 ) {
2304 $string = substr( $string, 0, $length ); // xyz...
2305 $string = $this->removeBadCharLast( $string );
2306 $string = $string . $ellipsis;
2307 } else {
2308 $string = substr( $string, $length ); // ...xyz
2309 $string = $this->removeBadCharFirst( $string );
2310 $string = $ellipsis . $string;
2311 }
2312 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181)
2313 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
2314 return $string;
2315 } else {
2316 return $stringOriginal;
2317 }
2318 }
2319
2320 /**
2321 * Remove bytes that represent an incomplete Unicode character
2322 * at the end of string (e.g. bytes of the char are missing)
2323 *
2324 * @param $string String
2325 * @return string
2326 */
2327 protected function removeBadCharLast( $string ) {
2328 $char = ord( $string[strlen( $string ) - 1] );
2329 $m = array();
2330 if ( $char >= 0xc0 ) {
2331 # We got the first byte only of a multibyte char; remove it.
2332 $string = substr( $string, 0, -1 );
2333 } elseif ( $char >= 0x80 &&
2334 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2335 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
2336 {
2337 # We chopped in the middle of a character; remove it
2338 $string = $m[1];
2339 }
2340 return $string;
2341 }
2342
2343 /**
2344 * Remove bytes that represent an incomplete Unicode character
2345 * at the start of string (e.g. bytes of the char are missing)
2346 *
2347 * @param $string String
2348 * @return string
2349 */
2350 protected function removeBadCharFirst( $string ) {
2351 $char = ord( $string[0] );
2352 if ( $char >= 0x80 && $char < 0xc0 ) {
2353 # We chopped in the middle of a character; remove the whole thing
2354 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2355 }
2356 return $string;
2357 }
2358
2359 /*
2360 * Truncate a string of valid HTML to a specified length in bytes,
2361 * appending an optional string (e.g. for ellipses), and return valid HTML
2362 *
2363 * This is only intended for styled/linked text, such as HTML with
2364 * tags like <span> and <a>, were the tags are self-contained (valid HTML)
2365 *
2366 * Note: tries to fix broken HTML with MWTidy
2367 *
2368 * @param string $text String to truncate
2369 * @param int $length (zero/positive) Maximum length (excluding ellipses)
2370 * @param string $ellipsis String to append to the truncated text
2371 * @returns string
2372 */
2373 function truncateHtml( $text, $length, $ellipsis = '...' ) {
2374 # Use the localized ellipsis character
2375 if ( $ellipsis == '...' ) {
2376 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2377 }
2378 # Check if there is no need to truncate
2379 if ( $length <= 0 ) {
2380 return $ellipsis; // no text shown, nothing to format
2381 } elseif ( strlen( $text ) <= $length ) {
2382 return $text; // string short enough even *with* HTML
2383 }
2384 $text = MWTidy::tidy( $text ); // fix tags
2385 $displayLen = 0; // innerHTML legth so far
2386 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
2387 $tagType = 0; // 0-open, 1-close
2388 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
2389 $entityState = 0; // 0-not entity, 1-entity
2390 $tag = $ret = $ch = '';
2391 $openTags = array();
2392 $textLen = strlen( $text );
2393 for ( $pos = 0; $pos < $textLen; ++$pos ) {
2394 $ch = $text[$pos];
2395 $lastCh = $pos ? $text[$pos - 1] : '';
2396 $ret .= $ch; // add to result string
2397 if ( $ch == '<' ) {
2398 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
2399 $entityState = 0; // for bad HTML
2400 $bracketState = 1; // tag started (checking for backslash)
2401 } elseif ( $ch == '>' ) {
2402 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
2403 $entityState = 0; // for bad HTML
2404 $bracketState = 0; // out of brackets
2405 } elseif ( $bracketState == 1 ) {
2406 if ( $ch == '/' ) {
2407 $tagType = 1; // close tag (e.g. "</span>")
2408 } else {
2409 $tagType = 0; // open tag (e.g. "<span>")
2410 $tag .= $ch;
2411 }
2412 $bracketState = 2; // building tag name
2413 } elseif ( $bracketState == 2 ) {
2414 if ( $ch != ' ' ) {
2415 $tag .= $ch;
2416 } else {
2417 // Name found (e.g. "<a href=..."), add on tag attributes...
2418 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
2419 }
2420 } elseif ( $bracketState == 0 ) {
2421 if ( $entityState ) {
2422 if ( $ch == ';' ) {
2423 $entityState = 0;
2424 $displayLen++; // entity is one displayed char
2425 }
2426 } else {
2427 if ( $ch == '&' ) {
2428 $entityState = 1; // entity found, (e.g. "&#160;")
2429 } else {
2430 $displayLen++; // this char is displayed
2431 // Add on the other display text after this...
2432 $skipped = $this->truncate_skip(
2433 $ret, $text, "<>&", $pos + 1, $length - $displayLen );
2434 $displayLen += $skipped;
2435 $pos += $skipped;
2436 }
2437 }
2438 }
2439 # Consider truncation once the display length has reached the maximim.
2440 # Double-check that we're not in the middle of a bracket/entity...
2441 if ( $displayLen >= $length && $bracketState == 0 && $entityState == 0 ) {
2442 if ( !$testingEllipsis ) {
2443 $testingEllipsis = true;
2444 # Save where we are; we will truncate here unless
2445 # the ellipsis actually makes the string longer.
2446 $pOpenTags = $openTags; // save state
2447 $pRet = $ret; // save state
2448 } elseif ( $displayLen > ( $length + strlen( $ellipsis ) ) ) {
2449 # Ellipsis won't make string longer/equal, the truncation point was OK.
2450 $openTags = $pOpenTags; // reload state
2451 $ret = $this->removeBadCharLast( $pRet ); // reload state, multi-byte char fix
2452 $ret .= $ellipsis; // add ellipsis
2453 break;
2454 }
2455 }
2456 }
2457 if ( $displayLen == 0 ) {
2458 return ''; // no text shown, nothing to format
2459 }
2460 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags ); // for bad HTML
2461 while ( count( $openTags ) > 0 ) {
2462 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
2463 }
2464 return $ret;
2465 }
2466
2467 // truncateHtml() helper function
2468 // like strcspn() but adds the skipped chars to $ret
2469 private function truncate_skip( &$ret, $text, $search, $start, $len = -1 ) {
2470 $skipCount = 0;
2471 if ( $start < strlen( $text ) ) {
2472 $skipCount = strcspn( $text, $search, $start, $len );
2473 $ret .= substr( $text, $start, $skipCount );
2474 }
2475 return $skipCount;
2476 }
2477
2478 // truncateHtml() helper function
2479 // (a) push or pop $tag from $openTags as needed
2480 // (b) clear $tag value
2481 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
2482 $tag = ltrim( $tag );
2483 if ( $tag != '' ) {
2484 if ( $tagType == 0 && $lastCh != '/' ) {
2485 $openTags[] = $tag; // tag opened (didn't close itself)
2486 } else if ( $tagType == 1 ) {
2487 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
2488 array_pop( $openTags ); // tag closed
2489 }
2490 }
2491 $tag = '';
2492 }
2493 }
2494
2495 /**
2496 * Grammatical transformations, needed for inflected languages
2497 * Invoked by putting {{grammar:case|word}} in a message
2498 *
2499 * @param $word string
2500 * @param $case string
2501 * @return string
2502 */
2503 function convertGrammar( $word, $case ) {
2504 global $wgGrammarForms;
2505 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
2506 return $wgGrammarForms[$this->getCode()][$case][$word];
2507 }
2508 return $word;
2509 }
2510
2511 /**
2512 * Provides an alternative text depending on specified gender.
2513 * Usage {{gender:username|masculine|feminine|neutral}}.
2514 * username is optional, in which case the gender of current user is used,
2515 * but only in (some) interface messages; otherwise default gender is used.
2516 * If second or third parameter are not specified, masculine is used.
2517 * These details may be overriden per language.
2518 */
2519 function gender( $gender, $forms ) {
2520 if ( !count( $forms ) ) {
2521 return '';
2522 }
2523 $forms = $this->preConvertPlural( $forms, 2 );
2524 if ( $gender === 'male' ) {
2525 return $forms[0];
2526 }
2527 if ( $gender === 'female' ) {
2528 return $forms[1];
2529 }
2530 return isset( $forms[2] ) ? $forms[2] : $forms[0];
2531 }
2532
2533 /**
2534 * Plural form transformations, needed for some languages.
2535 * For example, there are 3 form of plural in Russian and Polish,
2536 * depending on "count mod 10". See [[w:Plural]]
2537 * For English it is pretty simple.
2538 *
2539 * Invoked by putting {{plural:count|wordform1|wordform2}}
2540 * or {{plural:count|wordform1|wordform2|wordform3}}
2541 *
2542 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2543 *
2544 * @param $count Integer: non-localized number
2545 * @param $forms Array: different plural forms
2546 * @return string Correct form of plural for $count in this language
2547 */
2548 function convertPlural( $count, $forms ) {
2549 if ( !count( $forms ) ) {
2550 return '';
2551 }
2552 $forms = $this->preConvertPlural( $forms, 2 );
2553
2554 return ( $count == 1 ) ? $forms[0] : $forms[1];
2555 }
2556
2557 /**
2558 * Checks that convertPlural was given an array and pads it to requested
2559 * amound of forms by copying the last one.
2560 *
2561 * @param $count Integer: How many forms should there be at least
2562 * @param $forms Array of forms given to convertPlural
2563 * @return array Padded array of forms or an exception if not an array
2564 */
2565 protected function preConvertPlural( /* Array */ $forms, $count ) {
2566 while ( count( $forms ) < $count ) {
2567 $forms[] = $forms[count( $forms ) - 1];
2568 }
2569 return $forms;
2570 }
2571
2572 /**
2573 * For translating of expiry times
2574 * @param $str String: the validated block time in English
2575 * @return Somehow translated block time
2576 * @see LanguageFi.php for example implementation
2577 */
2578 function translateBlockExpiry( $str ) {
2579 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2580
2581 if ( $scBlockExpiryOptions == '-' ) {
2582 return $str;
2583 }
2584
2585 foreach ( explode( ',', $scBlockExpiryOptions ) as $option ) {
2586 if ( strpos( $option, ':' ) === false ) {
2587 continue;
2588 }
2589 list( $show, $value ) = explode( ':', $option );
2590 if ( strcmp( $str, $value ) == 0 ) {
2591 return htmlspecialchars( trim( $show ) );
2592 }
2593 }
2594
2595 return $str;
2596 }
2597
2598 /**
2599 * languages like Chinese need to be segmented in order for the diff
2600 * to be of any use
2601 *
2602 * @param $text String
2603 * @return String
2604 */
2605 function segmentForDiff( $text ) {
2606 return $text;
2607 }
2608
2609 /**
2610 * and unsegment to show the result
2611 *
2612 * @param $text String
2613 * @return String
2614 */
2615 function unsegmentForDiff( $text ) {
2616 return $text;
2617 }
2618
2619 # convert text to all supported variants
2620 function autoConvertToAllVariants( $text ) {
2621 return $this->mConverter->autoConvertToAllVariants( $text );
2622 }
2623
2624 # convert text to different variants of a language.
2625 function convert( $text ) {
2626 return $this->mConverter->convert( $text );
2627 }
2628
2629 # Convert a Title object to a string in the preferred variant
2630 function convertTitle( $title ) {
2631 return $this->mConverter->convertTitle( $title );
2632 }
2633
2634 # Check if this is a language with variants
2635 function hasVariants() {
2636 return sizeof( $this->getVariants() ) > 1;
2637 }
2638
2639 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2640 function armourMath( $text ) {
2641 return $this->mConverter->armourMath( $text );
2642 }
2643
2644 /**
2645 * Perform output conversion on a string, and encode for safe HTML output.
2646 * @param $text String text to be converted
2647 * @param $isTitle Bool whether this conversion is for the article title
2648 * @return string
2649 * @todo this should get integrated somewhere sane
2650 */
2651 function convertHtml( $text, $isTitle = false ) {
2652 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2653 }
2654
2655 function convertCategoryKey( $key ) {
2656 return $this->mConverter->convertCategoryKey( $key );
2657 }
2658
2659 /**
2660 * Get the list of variants supported by this langauge
2661 * see sample implementation in LanguageZh.php
2662 *
2663 * @return array an array of language codes
2664 */
2665 function getVariants() {
2666 return $this->mConverter->getVariants();
2667 }
2668
2669 function getPreferredVariant( $fromUser = true, $fromHeader = false ) {
2670 return $this->mConverter->getPreferredVariant( $fromUser, $fromHeader );
2671 }
2672
2673 /**
2674 * If a language supports multiple variants, it is
2675 * possible that non-existing link in one variant
2676 * actually exists in another variant. this function
2677 * tries to find it. See e.g. LanguageZh.php
2678 *
2679 * @param $link String: the name of the link
2680 * @param $nt Mixed: the title object of the link
2681 * @param $ignoreOtherCond Boolean: to disable other conditions when
2682 * we need to transclude a template or update a category's link
2683 * @return null the input parameters may be modified upon return
2684 */
2685 function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
2686 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
2687 }
2688
2689 /**
2690 * If a language supports multiple variants, converts text
2691 * into an array of all possible variants of the text:
2692 * 'variant' => text in that variant
2693 */
2694 function convertLinkToAllVariants( $text ) {
2695 return $this->mConverter->convertLinkToAllVariants( $text );
2696 }
2697
2698 /**
2699 * returns language specific options used by User::getPageRenderHash()
2700 * for example, the preferred language variant
2701 *
2702 * @return string
2703 */
2704 function getExtraHashOptions() {
2705 return $this->mConverter->getExtraHashOptions();
2706 }
2707
2708 /**
2709 * For languages that support multiple variants, the title of an
2710 * article may be displayed differently in different variants. this
2711 * function returns the apporiate title defined in the body of the article.
2712 *
2713 * @return string
2714 */
2715 function getParsedTitle() {
2716 return $this->mConverter->getParsedTitle();
2717 }
2718
2719 /**
2720 * Enclose a string with the "no conversion" tag. This is used by
2721 * various functions in the Parser
2722 *
2723 * @param $text String: text to be tagged for no conversion
2724 * @param $noParse
2725 * @return string the tagged text
2726 */
2727 function markNoConversion( $text, $noParse = false ) {
2728 return $this->mConverter->markNoConversion( $text, $noParse );
2729 }
2730
2731 /**
2732 * A regular expression to match legal word-trailing characters
2733 * which should be merged onto a link of the form [[foo]]bar.
2734 *
2735 * @return string
2736 */
2737 function linkTrail() {
2738 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
2739 }
2740
2741 function getLangObj() {
2742 return $this;
2743 }
2744
2745 /**
2746 * Get the RFC 3066 code for this language object
2747 */
2748 function getCode() {
2749 return $this->mCode;
2750 }
2751
2752 function setCode( $code ) {
2753 $this->mCode = $code;
2754 }
2755
2756 /**
2757 * Get the name of a file for a certain language code
2758 * @param $prefix string Prepend this to the filename
2759 * @param $code string Language code
2760 * @param $suffix string Append this to the filename
2761 * @return string $prefix . $mangledCode . $suffix
2762 */
2763 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2764 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2765 }
2766
2767 /**
2768 * Get the language code from a file name. Inverse of getFileName()
2769 * @param $filename string $prefix . $languageCode . $suffix
2770 * @param $prefix string Prefix before the language code
2771 * @param $suffix string Suffix after the language code
2772 * @return Language code, or false if $prefix or $suffix isn't found
2773 */
2774 static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
2775 $m = null;
2776 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
2777 preg_quote( $suffix, '/' ) . '/', $filename, $m );
2778 if ( !count( $m ) ) {
2779 return false;
2780 }
2781 return str_replace( '_', '-', strtolower( $m[1] ) );
2782 }
2783
2784 static function getMessagesFileName( $code ) {
2785 global $IP;
2786 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2787 }
2788
2789 static function getClassFileName( $code ) {
2790 global $IP;
2791 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2792 }
2793
2794 /**
2795 * Get the fallback for a given language
2796 */
2797 static function getFallbackFor( $code ) {
2798 if ( $code === 'en' ) {
2799 // Shortcut
2800 return false;
2801 } else {
2802 return self::getLocalisationCache()->getItem( $code, 'fallback' );
2803 }
2804 }
2805
2806 /**
2807 * Get all messages for a given language
2808 * WARNING: this may take a long time
2809 */
2810 static function getMessagesFor( $code ) {
2811 return self::getLocalisationCache()->getItem( $code, 'messages' );
2812 }
2813
2814 /**
2815 * Get a message for a given language
2816 */
2817 static function getMessageFor( $key, $code ) {
2818 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
2819 }
2820
2821 function fixVariableInNamespace( $talk ) {
2822 if ( strpos( $talk, '$1' ) === false ) {
2823 return $talk;
2824 }
2825
2826 global $wgMetaNamespace;
2827 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2828
2829 # Allow grammar transformations
2830 # Allowing full message-style parsing would make simple requests
2831 # such as action=raw much more expensive than they need to be.
2832 # This will hopefully cover most cases.
2833 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2834 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2835 return str_replace( ' ', '_', $talk );
2836 }
2837
2838 function replaceGrammarInNamespace( $m ) {
2839 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2840 }
2841
2842 static function getCaseMaps() {
2843 static $wikiUpperChars, $wikiLowerChars;
2844 if ( isset( $wikiUpperChars ) ) {
2845 return array( $wikiUpperChars, $wikiLowerChars );
2846 }
2847
2848 wfProfileIn( __METHOD__ );
2849 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2850 if ( $arr === false ) {
2851 throw new MWException(
2852 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2853 }
2854 extract( $arr );
2855 wfProfileOut( __METHOD__ );
2856 return array( $wikiUpperChars, $wikiLowerChars );
2857 }
2858
2859 function formatTimePeriod( $seconds ) {
2860 if ( $seconds < 10 ) {
2861 return $this->formatNum( sprintf( "%.1f", $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2862 } elseif ( $seconds < 60 ) {
2863 return $this->formatNum( round( $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2864 } elseif ( $seconds < 3600 ) {
2865 $minutes = floor( $seconds / 60 );
2866 $secondsPart = round( fmod( $seconds, 60 ) );
2867 if ( $secondsPart == 60 ) {
2868 $secondsPart = 0;
2869 $minutes++;
2870 }
2871 return $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2872 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2873 } else {
2874 $hours = floor( $seconds / 3600 );
2875 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2876 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2877 if ( $secondsPart == 60 ) {
2878 $secondsPart = 0;
2879 $minutes++;
2880 }
2881 if ( $minutes == 60 ) {
2882 $minutes = 0;
2883 $hours++;
2884 }
2885 return $this->formatNum( $hours ) . $this->getMessageFromDB( 'hours-abbrev' ) . ' ' .
2886 $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2887 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2888 }
2889 }
2890
2891 function formatBitrate( $bps ) {
2892 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2893 if ( $bps <= 0 ) {
2894 return $this->formatNum( $bps ) . $units[0];
2895 }
2896 $unitIndex = floor( log10( $bps ) / 3 );
2897 $mantissa = $bps / pow( 1000, $unitIndex );
2898 if ( $mantissa < 10 ) {
2899 $mantissa = round( $mantissa, 1 );
2900 } else {
2901 $mantissa = round( $mantissa );
2902 }
2903 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2904 }
2905
2906 /**
2907 * Format a size in bytes for output, using an appropriate
2908 * unit (B, KB, MB or GB) according to the magnitude in question
2909 *
2910 * @param $size Size to format
2911 * @return string Plain text (not HTML)
2912 */
2913 function formatSize( $size ) {
2914 // For small sizes no decimal places necessary
2915 $round = 0;
2916 if ( $size > 1024 ) {
2917 $size = $size / 1024;
2918 if ( $size > 1024 ) {
2919 $size = $size / 1024;
2920 // For MB and bigger two decimal places are smarter
2921 $round = 2;
2922 if ( $size > 1024 ) {
2923 $size = $size / 1024;
2924 $msg = 'size-gigabytes';
2925 } else {
2926 $msg = 'size-megabytes';
2927 }
2928 } else {
2929 $msg = 'size-kilobytes';
2930 }
2931 } else {
2932 $msg = 'size-bytes';
2933 }
2934 $size = round( $size, $round );
2935 $text = $this->getMessageFromDB( $msg );
2936 return str_replace( '$1', $this->formatNum( $size ), $text );
2937 }
2938
2939 /**
2940 * Get the conversion rule title, if any.
2941 */
2942 function getConvRuleTitle() {
2943 return $this->mConverter->getConvRuleTitle();
2944 }
2945
2946 /**
2947 * Given a string, convert it to a (hopefully short) key that can be used
2948 * for efficient sorting. A binary sort according to the sortkeys
2949 * corresponds to a logical sort of the corresponding strings. Current
2950 * code expects that a null character should sort before all others, but
2951 * has no other particular expectations (and that one can be changed if
2952 * necessary).
2953 *
2954 * @param string $string UTF-8 string
2955 * @return string Binary sortkey
2956 */
2957 public function convertToSortkey( $string ) {
2958 # Fake function for now
2959 return strtoupper( $string );
2960 }
2961
2962 /**
2963 * Does it make sense for lists to be split up into sections based on their
2964 * first letter? Logogram-based scripts probably want to return false.
2965 *
2966 * TODO: Use this in CategoryPage.php.
2967 *
2968 * @return boolean
2969 */
2970 public function usesFirstLettersInLists() {
2971 return true;
2972 }
2973
2974 /**
2975 * Given a string, return the logical "first letter" to be used for
2976 * grouping on category pages and so on. This has to be coordinated
2977 * carefully with convertToSortkey(), or else the sorted list might jump
2978 * back and forth between the same "initial letters" or other pathological
2979 * behavior. For instance, if you just return the first character, but "a"
2980 * sorts the same as "A" based on convertToSortkey(), then you might get a
2981 * list like
2982 *
2983 * == A ==
2984 * * [[Aardvark]]
2985 *
2986 * == a ==
2987 * * [[antelope]]
2988 *
2989 * == A ==
2990 * * [[Ape]]
2991 *
2992 * etc., assuming for the sake of argument that $wgCapitalLinks is false.
2993 * Obviously, this is ignored if usesFirstLettersInLists() is false.
2994 *
2995 * @param string $string UTF-8 string
2996 * @return string UTF-8 string corresponding to the first letter of input
2997 */
2998 public function firstLetterForLists( $string ) {
2999 if ( $string[0] == "\0" ) {
3000 $string = substr( $string, 1 );
3001 }
3002 return strtoupper( $this->firstChar( $string ) );
3003 }
3004 }
Wiklou, le Wiki du Wiklou