Localisation updates for core and extension messages from translatewiki.net
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * Internationalisation code
4 *
5 * @file
6 * @ingroup Language
7 */
8
9 /**
10 * @defgroup Language Language
11 */
12
13 if ( !defined( 'MEDIAWIKI' ) ) {
14 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
15 exit( 1 );
16 }
17
18 # Read language names
19 global $wgLanguageNames;
20 require_once( dirname( __FILE__ ) . '/Names.php' );
21
22 if ( function_exists( 'mb_strtoupper' ) ) {
23 mb_internal_encoding( 'UTF-8' );
24 }
25
26 /**
27 * a fake language converter
28 *
29 * @ingroup Language
30 */
31 class FakeConverter {
32 var $mLang;
33 function __construct( $langobj ) { $this->mLang = $langobj; }
34 function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
35 function convert( $t ) { return $t; }
36 function convertTitle( $t ) { return $t->getPrefixedText(); }
37 function getVariants() { return array( $this->mLang->getCode() ); }
38 function getPreferredVariant() { return $this->mLang->getCode(); }
39 function getDefaultVariant() { return $this->mLang->getCode(); }
40 function getURLVariant() { return ''; }
41 function getConvRuleTitle() { return false; }
42 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
43 function getExtraHashOptions() { return ''; }
44 function getParsedTitle() { return ''; }
45 function markNoConversion( $text, $noParse = false ) { return $text; }
46 function convertCategoryKey( $key ) { return $key; }
47 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
48 function armourMath( $text ) { return $text; }
49 }
50
51 /**
52 * Internationalisation code
53 * @ingroup Language
54 */
55 class Language {
56
57 /**
58 * @var LanguageConverter
59 */
60 var $mConverter;
61
62 var $mVariants, $mCode, $mLoaded = false;
63 var $mMagicExtensions = array(), $mMagicHookDone = false;
64
65 var $mNamespaceIds, $namespaceNames, $namespaceAliases;
66 var $dateFormatStrings = array();
67 var $mExtendedSpecialPageAliases;
68
69 /**
70 * ReplacementArray object caches
71 */
72 var $transformData = array();
73
74 /**
75 * @var LocalisationCache
76 */
77 static public $dataCache;
78
79 static public $mLangObjCache = array();
80
81 static public $mWeekdayMsgs = array(
82 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
83 'friday', 'saturday'
84 );
85
86 static public $mWeekdayAbbrevMsgs = array(
87 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
88 );
89
90 static public $mMonthMsgs = array(
91 'january', 'february', 'march', 'april', 'may_long', 'june',
92 'july', 'august', 'september', 'october', 'november',
93 'december'
94 );
95 static public $mMonthGenMsgs = array(
96 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
97 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
98 'december-gen'
99 );
100 static public $mMonthAbbrevMsgs = array(
101 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
102 'sep', 'oct', 'nov', 'dec'
103 );
104
105 static public $mIranianCalendarMonthMsgs = array(
106 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
107 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
108 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
109 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
110 );
111
112 static public $mHebrewCalendarMonthMsgs = array(
113 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
114 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
115 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
116 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
117 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
118 );
119
120 static public $mHebrewCalendarMonthGenMsgs = array(
121 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
122 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
123 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
124 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
125 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
126 );
127
128 static public $mHijriCalendarMonthMsgs = array(
129 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
130 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
131 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
132 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
133 );
134
135 /**
136 * Get a cached language object for a given language code
137 * @param $code String
138 * @return Language
139 */
140 static function factory( $code ) {
141 if ( !isset( self::$mLangObjCache[$code] ) ) {
142 if ( count( self::$mLangObjCache ) > 10 ) {
143 // Don't keep a billion objects around, that's stupid.
144 self::$mLangObjCache = array();
145 }
146 self::$mLangObjCache[$code] = self::newFromCode( $code );
147 }
148 return self::$mLangObjCache[$code];
149 }
150
151 /**
152 * Create a language object for a given language code
153 * @param $code String
154 * @return Language
155 */
156 protected static function newFromCode( $code ) {
157 // Protect against path traversal below
158 if ( !Language::isValidCode( $code )
159 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
160 {
161 throw new MWException( "Invalid language code \"$code\"" );
162 }
163
164 if ( !Language::isValidBuiltInCode( $code ) ) {
165 // It's not possible to customise this code with class files, so
166 // just return a Language object. This is to support uselang= hacks.
167 $lang = new Language;
168 $lang->setCode( $code );
169 return $lang;
170 }
171
172 // Check if there is a language class for the code
173 $class = self::classFromCode( $code );
174 self::preloadLanguageClass( $class );
175 if ( MWInit::classExists( $class ) ) {
176 $lang = new $class;
177 return $lang;
178 }
179
180 // Keep trying the fallback list until we find an existing class
181 $fallbacks = Language::getFallbacksFor( $code );
182 foreach ( $fallbacks as $fallbackCode ) {
183 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
184 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
185 }
186
187 $class = self::classFromCode( $fallbackCode );
188 self::preloadLanguageClass( $class );
189 if ( MWInit::classExists( $class ) ) {
190 $lang = Language::newFromCode( $fallbackCode );
191 $lang->setCode( $code );
192 return $lang;
193 }
194 }
195
196 throw new MWException( "Invalid fallback sequence for language '$code'" );
197 }
198
199 /**
200 * Returns true if a language code string is of a valid form, whether or
201 * not it exists. This includes codes which are used solely for
202 * customisation via the MediaWiki namespace.
203 *
204 * @param $code string
205 *
206 * @return bool
207 */
208 public static function isValidCode( $code ) {
209 return
210 strcspn( $code, ":/\\\000" ) === strlen( $code )
211 && !preg_match( Title::getTitleInvalidRegex(), $code );
212 }
213
214 /**
215 * Returns true if a language code is of a valid form for the purposes of
216 * internal customisation of MediaWiki, via Messages*.php.
217 *
218 * @param $code string
219 *
220 * @since 1.18
221 * @return bool
222 */
223 public static function isValidBuiltInCode( $code ) {
224 return preg_match( '/^[a-z0-9-]+$/i', $code );
225 }
226
227 /**
228 * @param $code
229 * @return String Name of the language class
230 */
231 public static function classFromCode( $code ) {
232 if ( $code == 'en' ) {
233 return 'Language';
234 } else {
235 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
236 }
237 }
238
239 /**
240 * Includes language class files
241 *
242 * @param $class Name of the language class
243 */
244 public static function preloadLanguageClass( $class ) {
245 global $IP;
246
247 if ( $class === 'Language' ) {
248 return;
249 }
250
251 if ( !defined( 'MW_COMPILED' ) ) {
252 // Preload base classes to work around APC/PHP5 bug
253 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
254 include_once( "$IP/languages/classes/$class.deps.php" );
255 }
256 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
257 include_once( "$IP/languages/classes/$class.php" );
258 }
259 }
260 }
261
262 /**
263 * Get the LocalisationCache instance
264 *
265 * @return LocalisationCache
266 */
267 public static function getLocalisationCache() {
268 if ( is_null( self::$dataCache ) ) {
269 global $wgLocalisationCacheConf;
270 $class = $wgLocalisationCacheConf['class'];
271 self::$dataCache = new $class( $wgLocalisationCacheConf );
272 }
273 return self::$dataCache;
274 }
275
276 function __construct() {
277 $this->mConverter = new FakeConverter( $this );
278 // Set the code to the name of the descendant
279 if ( get_class( $this ) == 'Language' ) {
280 $this->mCode = 'en';
281 } else {
282 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
283 }
284 self::getLocalisationCache();
285 }
286
287 /**
288 * Reduce memory usage
289 */
290 function __destruct() {
291 foreach ( $this as $name => $value ) {
292 unset( $this->$name );
293 }
294 }
295
296 /**
297 * Hook which will be called if this is the content language.
298 * Descendants can use this to register hook functions or modify globals
299 */
300 function initContLang() { }
301
302 /**
303 * Same as getFallbacksFor for current language.
304 * @return array|bool
305 * @deprecated in 1.19
306 */
307 function getFallbackLanguageCode() {
308 wfDeprecated( __METHOD__ );
309 return self::getFallbackFor( $this->mCode );
310 }
311
312 /**
313 * @return array
314 * @since 1.19
315 */
316 function getFallbackLanguages() {
317 return self::getFallbacksFor( $this->mCode );
318 }
319
320 /**
321 * Exports $wgBookstoreListEn
322 * @return array
323 */
324 function getBookstoreList() {
325 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
326 }
327
328 /**
329 * @return array
330 */
331 function getNamespaces() {
332 if ( is_null( $this->namespaceNames ) ) {
333 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
334
335 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
336 $validNamespaces = MWNamespace::getCanonicalNamespaces();
337
338 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
339
340 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
341 if ( $wgMetaNamespaceTalk ) {
342 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
343 } else {
344 $talk = $this->namespaceNames[NS_PROJECT_TALK];
345 $this->namespaceNames[NS_PROJECT_TALK] =
346 $this->fixVariableInNamespace( $talk );
347 }
348
349 # Sometimes a language will be localised but not actually exist on this wiki.
350 foreach ( $this->namespaceNames as $key => $text ) {
351 if ( !isset( $validNamespaces[$key] ) ) {
352 unset( $this->namespaceNames[$key] );
353 }
354 }
355
356 # The above mixing may leave namespaces out of canonical order.
357 # Re-order by namespace ID number...
358 ksort( $this->namespaceNames );
359
360 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames ) );
361 }
362 return $this->namespaceNames;
363 }
364
365 /**
366 * A convenience function that returns the same thing as
367 * getNamespaces() except with the array values changed to ' '
368 * where it found '_', useful for producing output to be displayed
369 * e.g. in <select> forms.
370 *
371 * @return array
372 */
373 function getFormattedNamespaces() {
374 $ns = $this->getNamespaces();
375 foreach ( $ns as $k => $v ) {
376 $ns[$k] = strtr( $v, '_', ' ' );
377 }
378 return $ns;
379 }
380
381 /**
382 * Get a namespace value by key
383 * <code>
384 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
385 * echo $mw_ns; // prints 'MediaWiki'
386 * </code>
387 *
388 * @param $index Int: the array key of the namespace to return
389 * @return mixed, string if the namespace value exists, otherwise false
390 */
391 function getNsText( $index ) {
392 $ns = $this->getNamespaces();
393 return isset( $ns[$index] ) ? $ns[$index] : false;
394 }
395
396 /**
397 * A convenience function that returns the same thing as
398 * getNsText() except with '_' changed to ' ', useful for
399 * producing output.
400 *
401 * @param $index string
402 *
403 * @return array
404 */
405 function getFormattedNsText( $index ) {
406 $ns = $this->getNsText( $index );
407 return strtr( $ns, '_', ' ' );
408 }
409
410 /**
411 * Returns gender-dependent namespace alias if available.
412 * @param $index Int: namespace index
413 * @param $gender String: gender key (male, female... )
414 * @return String
415 * @since 1.18
416 */
417 function getGenderNsText( $index, $gender ) {
418 global $wgExtraGenderNamespaces;
419
420 $ns = $wgExtraGenderNamespaces + self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
421 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
422 }
423
424 /**
425 * Whether this language makes distinguishes genders for example in
426 * namespaces.
427 * @return bool
428 * @since 1.18
429 */
430 function needsGenderDistinction() {
431 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
432 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
433 // $wgExtraGenderNamespaces overrides everything
434 return true;
435 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
436 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
437 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
438 return false;
439 } else {
440 // Check what is in i18n files
441 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
442 return count( $aliases ) > 0;
443 }
444 }
445
446 /**
447 * Get a namespace key by value, case insensitive.
448 * Only matches namespace names for the current language, not the
449 * canonical ones defined in Namespace.php.
450 *
451 * @param $text String
452 * @return mixed An integer if $text is a valid value otherwise false
453 */
454 function getLocalNsIndex( $text ) {
455 $lctext = $this->lc( $text );
456 $ids = $this->getNamespaceIds();
457 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
458 }
459
460 /**
461 * @return array
462 */
463 function getNamespaceAliases() {
464 if ( is_null( $this->namespaceAliases ) ) {
465 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
466 if ( !$aliases ) {
467 $aliases = array();
468 } else {
469 foreach ( $aliases as $name => $index ) {
470 if ( $index === NS_PROJECT_TALK ) {
471 unset( $aliases[$name] );
472 $name = $this->fixVariableInNamespace( $name );
473 $aliases[$name] = $index;
474 }
475 }
476 }
477
478 global $wgExtraGenderNamespaces;
479 $genders = $wgExtraGenderNamespaces + self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
480 foreach ( $genders as $index => $forms ) {
481 foreach ( $forms as $alias ) {
482 $aliases[$alias] = $index;
483 }
484 }
485
486 $this->namespaceAliases = $aliases;
487 }
488 return $this->namespaceAliases;
489 }
490
491 /**
492 * @return array
493 */
494 function getNamespaceIds() {
495 if ( is_null( $this->mNamespaceIds ) ) {
496 global $wgNamespaceAliases;
497 # Put namespace names and aliases into a hashtable.
498 # If this is too slow, then we should arrange it so that it is done
499 # before caching. The catch is that at pre-cache time, the above
500 # class-specific fixup hasn't been done.
501 $this->mNamespaceIds = array();
502 foreach ( $this->getNamespaces() as $index => $name ) {
503 $this->mNamespaceIds[$this->lc( $name )] = $index;
504 }
505 foreach ( $this->getNamespaceAliases() as $name => $index ) {
506 $this->mNamespaceIds[$this->lc( $name )] = $index;
507 }
508 if ( $wgNamespaceAliases ) {
509 foreach ( $wgNamespaceAliases as $name => $index ) {
510 $this->mNamespaceIds[$this->lc( $name )] = $index;
511 }
512 }
513 }
514 return $this->mNamespaceIds;
515 }
516
517 /**
518 * Get a namespace key by value, case insensitive. Canonical namespace
519 * names override custom ones defined for the current language.
520 *
521 * @param $text String
522 * @return mixed An integer if $text is a valid value otherwise false
523 */
524 function getNsIndex( $text ) {
525 $lctext = $this->lc( $text );
526 $ns = MWNamespace::getCanonicalIndex( $lctext );
527 if ( $ns !== null ) {
528 return $ns;
529 }
530 $ids = $this->getNamespaceIds();
531 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
532 }
533
534 /**
535 * short names for language variants used for language conversion links.
536 *
537 * @param $code String
538 * @param $usemsg bool Use the "variantname-xyz" message if it exists
539 * @return string
540 */
541 function getVariantname( $code, $usemsg = true ) {
542 $msg = "variantname-$code";
543 list( $rootCode ) = explode( '-', $code );
544 if ( $usemsg && wfMessage( $msg )->exists() ) {
545 return $this->getMessageFromDB( $msg );
546 }
547 $name = self::getLanguageName( $code );
548 if ( $name ) {
549 return $name; # if it's defined as a language name, show that
550 } else {
551 # otherwise, output the language code
552 return $code;
553 }
554 }
555
556 /**
557 * @param $name string
558 * @return string
559 */
560 function specialPage( $name ) {
561 $aliases = $this->getSpecialPageAliases();
562 if ( isset( $aliases[$name][0] ) ) {
563 $name = $aliases[$name][0];
564 }
565 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
566 }
567
568 /**
569 * @return array
570 */
571 function getQuickbarSettings() {
572 return array(
573 $this->getMessage( 'qbsettings-none' ),
574 $this->getMessage( 'qbsettings-fixedleft' ),
575 $this->getMessage( 'qbsettings-fixedright' ),
576 $this->getMessage( 'qbsettings-floatingleft' ),
577 $this->getMessage( 'qbsettings-floatingright' ),
578 $this->getMessage( 'qbsettings-directionality' )
579 );
580 }
581
582 /**
583 * @return array
584 */
585 function getDatePreferences() {
586 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
587 }
588
589 /**
590 * @return array
591 */
592 function getDateFormats() {
593 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
594 }
595
596 /**
597 * @return array|string
598 */
599 function getDefaultDateFormat() {
600 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
601 if ( $df === 'dmy or mdy' ) {
602 global $wgAmericanDates;
603 return $wgAmericanDates ? 'mdy' : 'dmy';
604 } else {
605 return $df;
606 }
607 }
608
609 /**
610 * @return array
611 */
612 function getDatePreferenceMigrationMap() {
613 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
614 }
615
616 /**
617 * @param $image
618 * @return array|null
619 */
620 function getImageFile( $image ) {
621 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
622 }
623
624 /**
625 * @return array
626 */
627 function getExtraUserToggles() {
628 return self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
629 }
630
631 /**
632 * @param $tog
633 * @return string
634 */
635 function getUserToggle( $tog ) {
636 return $this->getMessageFromDB( "tog-$tog" );
637 }
638
639 /**
640 * Get language names, indexed by code.
641 * If $customisedOnly is true, only returns codes with a messages file
642 *
643 * @param $customisedOnly bool
644 *
645 * @return array
646 */
647 public static function getLanguageNames( $customisedOnly = false ) {
648 global $wgExtraLanguageNames;
649 static $coreLanguageNames;
650
651 if ( $coreLanguageNames === null ) {
652 include( MWInit::compiledPath( 'languages/Names.php' ) );
653 }
654
655 $allNames = $wgExtraLanguageNames + $coreLanguageNames;
656 if ( !$customisedOnly ) {
657 return $allNames;
658 }
659
660 global $IP;
661 $names = array();
662 $dir = opendir( "$IP/languages/messages" );
663 while ( false !== ( $file = readdir( $dir ) ) ) {
664 $code = self::getCodeFromFileName( $file, 'Messages' );
665 if ( $code && isset( $allNames[$code] ) ) {
666 $names[$code] = $allNames[$code];
667 }
668 }
669 closedir( $dir );
670 return $names;
671 }
672
673 /**
674 * Get translated language names. This is done on best effort and
675 * by default this is exactly the same as Language::getLanguageNames.
676 * The CLDR extension provides translated names.
677 * @param $code String Language code.
678 * @return Array language code => language name
679 * @since 1.18.0
680 */
681 public static function getTranslatedLanguageNames( $code ) {
682 $names = array();
683 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $code ) );
684
685 foreach ( self::getLanguageNames() as $code => $name ) {
686 if ( !isset( $names[$code] ) ) $names[$code] = $name;
687 }
688
689 return $names;
690 }
691
692 /**
693 * Get a message from the MediaWiki namespace.
694 *
695 * @param $msg String: message name
696 * @return string
697 */
698 function getMessageFromDB( $msg ) {
699 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
700 }
701
702 /**
703 * @param $code string
704 * @return string
705 */
706 function getLanguageName( $code ) {
707 $names = self::getLanguageNames();
708 if ( !array_key_exists( $code, $names ) ) {
709 return '';
710 }
711 return $names[$code];
712 }
713
714 /**
715 * @param $key string
716 * @return string
717 */
718 function getMonthName( $key ) {
719 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
720 }
721
722 /**
723 * @return array
724 */
725 function getMonthNamesArray() {
726 $monthNames = array( '' );
727 for ( $i = 1; $i < 13; $i++ ) {
728 $monthNames[] = $this->getMonthName( $i );
729 }
730 return $monthNames;
731 }
732
733 /**
734 * @param $key string
735 * @return string
736 */
737 function getMonthNameGen( $key ) {
738 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
739 }
740
741 /**
742 * @param $key string
743 * @return string
744 */
745 function getMonthAbbreviation( $key ) {
746 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
747 }
748
749 /**
750 * @return array
751 */
752 function getMonthAbbreviationsArray() {
753 $monthNames = array( '' );
754 for ( $i = 1; $i < 13; $i++ ) {
755 $monthNames[] = $this->getMonthAbbreviation( $i );
756 }
757 return $monthNames;
758 }
759
760 /**
761 * @param $key string
762 * @return string
763 */
764 function getWeekdayName( $key ) {
765 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
766 }
767
768 /**
769 * @param $key string
770 * @return string
771 */
772 function getWeekdayAbbreviation( $key ) {
773 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
774 }
775
776 /**
777 * @param $key string
778 * @return string
779 */
780 function getIranianCalendarMonthName( $key ) {
781 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
782 }
783
784 /**
785 * @param $key string
786 * @return string
787 */
788 function getHebrewCalendarMonthName( $key ) {
789 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
790 }
791
792 /**
793 * @param $key string
794 * @return string
795 */
796 function getHebrewCalendarMonthNameGen( $key ) {
797 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
798 }
799
800 /**
801 * @param $key string
802 * @return string
803 */
804 function getHijriCalendarMonthName( $key ) {
805 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
806 }
807
808 /**
809 * This is a workalike of PHP's date() function, but with better
810 * internationalisation, a reduced set of format characters, and a better
811 * escaping format.
812 *
813 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
814 * PHP manual for definitions. There are a number of extensions, which
815 * start with "x":
816 *
817 * xn Do not translate digits of the next numeric format character
818 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
819 * xr Use roman numerals for the next numeric format character
820 * xh Use hebrew numerals for the next numeric format character
821 * xx Literal x
822 * xg Genitive month name
823 *
824 * xij j (day number) in Iranian calendar
825 * xiF F (month name) in Iranian calendar
826 * xin n (month number) in Iranian calendar
827 * xiY Y (full year) in Iranian calendar
828 *
829 * xjj j (day number) in Hebrew calendar
830 * xjF F (month name) in Hebrew calendar
831 * xjt t (days in month) in Hebrew calendar
832 * xjx xg (genitive month name) in Hebrew calendar
833 * xjn n (month number) in Hebrew calendar
834 * xjY Y (full year) in Hebrew calendar
835 *
836 * xmj j (day number) in Hijri calendar
837 * xmF F (month name) in Hijri calendar
838 * xmn n (month number) in Hijri calendar
839 * xmY Y (full year) in Hijri calendar
840 *
841 * xkY Y (full year) in Thai solar calendar. Months and days are
842 * identical to the Gregorian calendar
843 * xoY Y (full year) in Minguo calendar or Juche year.
844 * Months and days are identical to the
845 * Gregorian calendar
846 * xtY Y (full year) in Japanese nengo. Months and days are
847 * identical to the Gregorian calendar
848 *
849 * Characters enclosed in double quotes will be considered literal (with
850 * the quotes themselves removed). Unmatched quotes will be considered
851 * literal quotes. Example:
852 *
853 * "The month is" F => The month is January
854 * i's" => 20'11"
855 *
856 * Backslash escaping is also supported.
857 *
858 * Input timestamp is assumed to be pre-normalized to the desired local
859 * time zone, if any.
860 *
861 * @param $format String
862 * @param $ts String: 14-character timestamp
863 * YYYYMMDDHHMMSS
864 * 01234567890123
865 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
866 *
867 * @return string
868 */
869 function sprintfDate( $format, $ts ) {
870 $s = '';
871 $raw = false;
872 $roman = false;
873 $hebrewNum = false;
874 $unix = false;
875 $rawToggle = false;
876 $iranian = false;
877 $hebrew = false;
878 $hijri = false;
879 $thai = false;
880 $minguo = false;
881 $tenno = false;
882 for ( $p = 0; $p < strlen( $format ); $p++ ) {
883 $num = false;
884 $code = $format[$p];
885 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
886 $code .= $format[++$p];
887 }
888
889 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
890 $code .= $format[++$p];
891 }
892
893 switch ( $code ) {
894 case 'xx':
895 $s .= 'x';
896 break;
897 case 'xn':
898 $raw = true;
899 break;
900 case 'xN':
901 $rawToggle = !$rawToggle;
902 break;
903 case 'xr':
904 $roman = true;
905 break;
906 case 'xh':
907 $hebrewNum = true;
908 break;
909 case 'xg':
910 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
911 break;
912 case 'xjx':
913 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
914 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
915 break;
916 case 'd':
917 $num = substr( $ts, 6, 2 );
918 break;
919 case 'D':
920 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
921 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
922 break;
923 case 'j':
924 $num = intval( substr( $ts, 6, 2 ) );
925 break;
926 case 'xij':
927 if ( !$iranian ) {
928 $iranian = self::tsToIranian( $ts );
929 }
930 $num = $iranian[2];
931 break;
932 case 'xmj':
933 if ( !$hijri ) {
934 $hijri = self::tsToHijri( $ts );
935 }
936 $num = $hijri[2];
937 break;
938 case 'xjj':
939 if ( !$hebrew ) {
940 $hebrew = self::tsToHebrew( $ts );
941 }
942 $num = $hebrew[2];
943 break;
944 case 'l':
945 if ( !$unix ) {
946 $unix = wfTimestamp( TS_UNIX, $ts );
947 }
948 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
949 break;
950 case 'N':
951 if ( !$unix ) {
952 $unix = wfTimestamp( TS_UNIX, $ts );
953 }
954 $w = gmdate( 'w', $unix );
955 $num = $w ? $w : 7;
956 break;
957 case 'w':
958 if ( !$unix ) {
959 $unix = wfTimestamp( TS_UNIX, $ts );
960 }
961 $num = gmdate( 'w', $unix );
962 break;
963 case 'z':
964 if ( !$unix ) {
965 $unix = wfTimestamp( TS_UNIX, $ts );
966 }
967 $num = gmdate( 'z', $unix );
968 break;
969 case 'W':
970 if ( !$unix ) {
971 $unix = wfTimestamp( TS_UNIX, $ts );
972 }
973 $num = gmdate( 'W', $unix );
974 break;
975 case 'F':
976 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
977 break;
978 case 'xiF':
979 if ( !$iranian ) {
980 $iranian = self::tsToIranian( $ts );
981 }
982 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
983 break;
984 case 'xmF':
985 if ( !$hijri ) {
986 $hijri = self::tsToHijri( $ts );
987 }
988 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
989 break;
990 case 'xjF':
991 if ( !$hebrew ) {
992 $hebrew = self::tsToHebrew( $ts );
993 }
994 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
995 break;
996 case 'm':
997 $num = substr( $ts, 4, 2 );
998 break;
999 case 'M':
1000 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1001 break;
1002 case 'n':
1003 $num = intval( substr( $ts, 4, 2 ) );
1004 break;
1005 case 'xin':
1006 if ( !$iranian ) {
1007 $iranian = self::tsToIranian( $ts );
1008 }
1009 $num = $iranian[1];
1010 break;
1011 case 'xmn':
1012 if ( !$hijri ) {
1013 $hijri = self::tsToHijri ( $ts );
1014 }
1015 $num = $hijri[1];
1016 break;
1017 case 'xjn':
1018 if ( !$hebrew ) {
1019 $hebrew = self::tsToHebrew( $ts );
1020 }
1021 $num = $hebrew[1];
1022 break;
1023 case 't':
1024 if ( !$unix ) {
1025 $unix = wfTimestamp( TS_UNIX, $ts );
1026 }
1027 $num = gmdate( 't', $unix );
1028 break;
1029 case 'xjt':
1030 if ( !$hebrew ) {
1031 $hebrew = self::tsToHebrew( $ts );
1032 }
1033 $num = $hebrew[3];
1034 break;
1035 case 'L':
1036 if ( !$unix ) {
1037 $unix = wfTimestamp( TS_UNIX, $ts );
1038 }
1039 $num = gmdate( 'L', $unix );
1040 break;
1041 case 'o':
1042 if ( !$unix ) {
1043 $unix = wfTimestamp( TS_UNIX, $ts );
1044 }
1045 $num = date( 'o', $unix );
1046 break;
1047 case 'Y':
1048 $num = substr( $ts, 0, 4 );
1049 break;
1050 case 'xiY':
1051 if ( !$iranian ) {
1052 $iranian = self::tsToIranian( $ts );
1053 }
1054 $num = $iranian[0];
1055 break;
1056 case 'xmY':
1057 if ( !$hijri ) {
1058 $hijri = self::tsToHijri( $ts );
1059 }
1060 $num = $hijri[0];
1061 break;
1062 case 'xjY':
1063 if ( !$hebrew ) {
1064 $hebrew = self::tsToHebrew( $ts );
1065 }
1066 $num = $hebrew[0];
1067 break;
1068 case 'xkY':
1069 if ( !$thai ) {
1070 $thai = self::tsToYear( $ts, 'thai' );
1071 }
1072 $num = $thai[0];
1073 break;
1074 case 'xoY':
1075 if ( !$minguo ) {
1076 $minguo = self::tsToYear( $ts, 'minguo' );
1077 }
1078 $num = $minguo[0];
1079 break;
1080 case 'xtY':
1081 if ( !$tenno ) {
1082 $tenno = self::tsToYear( $ts, 'tenno' );
1083 }
1084 $num = $tenno[0];
1085 break;
1086 case 'y':
1087 $num = substr( $ts, 2, 2 );
1088 break;
1089 case 'a':
1090 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1091 break;
1092 case 'A':
1093 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1094 break;
1095 case 'g':
1096 $h = substr( $ts, 8, 2 );
1097 $num = $h % 12 ? $h % 12 : 12;
1098 break;
1099 case 'G':
1100 $num = intval( substr( $ts, 8, 2 ) );
1101 break;
1102 case 'h':
1103 $h = substr( $ts, 8, 2 );
1104 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1105 break;
1106 case 'H':
1107 $num = substr( $ts, 8, 2 );
1108 break;
1109 case 'i':
1110 $num = substr( $ts, 10, 2 );
1111 break;
1112 case 's':
1113 $num = substr( $ts, 12, 2 );
1114 break;
1115 case 'c':
1116 if ( !$unix ) {
1117 $unix = wfTimestamp( TS_UNIX, $ts );
1118 }
1119 $s .= gmdate( 'c', $unix );
1120 break;
1121 case 'r':
1122 if ( !$unix ) {
1123 $unix = wfTimestamp( TS_UNIX, $ts );
1124 }
1125 $s .= gmdate( 'r', $unix );
1126 break;
1127 case 'U':
1128 if ( !$unix ) {
1129 $unix = wfTimestamp( TS_UNIX, $ts );
1130 }
1131 $num = $unix;
1132 break;
1133 case '\\':
1134 # Backslash escaping
1135 if ( $p < strlen( $format ) - 1 ) {
1136 $s .= $format[++$p];
1137 } else {
1138 $s .= '\\';
1139 }
1140 break;
1141 case '"':
1142 # Quoted literal
1143 if ( $p < strlen( $format ) - 1 ) {
1144 $endQuote = strpos( $format, '"', $p + 1 );
1145 if ( $endQuote === false ) {
1146 # No terminating quote, assume literal "
1147 $s .= '"';
1148 } else {
1149 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1150 $p = $endQuote;
1151 }
1152 } else {
1153 # Quote at end of string, assume literal "
1154 $s .= '"';
1155 }
1156 break;
1157 default:
1158 $s .= $format[$p];
1159 }
1160 if ( $num !== false ) {
1161 if ( $rawToggle || $raw ) {
1162 $s .= $num;
1163 $raw = false;
1164 } elseif ( $roman ) {
1165 $s .= self::romanNumeral( $num );
1166 $roman = false;
1167 } elseif ( $hebrewNum ) {
1168 $s .= self::hebrewNumeral( $num );
1169 $hebrewNum = false;
1170 } else {
1171 $s .= $this->formatNum( $num, true );
1172 }
1173 }
1174 }
1175 return $s;
1176 }
1177
1178 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1179 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1180
1181 /**
1182 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1183 * Gregorian dates to Iranian dates. Originally written in C, it
1184 * is released under the terms of GNU Lesser General Public
1185 * License. Conversion to PHP was performed by Niklas Laxström.
1186 *
1187 * Link: http://www.farsiweb.info/jalali/jalali.c
1188 *
1189 * @param $ts string
1190 *
1191 * @return string
1192 */
1193 private static function tsToIranian( $ts ) {
1194 $gy = substr( $ts, 0, 4 ) -1600;
1195 $gm = substr( $ts, 4, 2 ) -1;
1196 $gd = substr( $ts, 6, 2 ) -1;
1197
1198 # Days passed from the beginning (including leap years)
1199 $gDayNo = 365 * $gy
1200 + floor( ( $gy + 3 ) / 4 )
1201 - floor( ( $gy + 99 ) / 100 )
1202 + floor( ( $gy + 399 ) / 400 );
1203
1204 // Add days of the past months of this year
1205 for ( $i = 0; $i < $gm; $i++ ) {
1206 $gDayNo += self::$GREG_DAYS[$i];
1207 }
1208
1209 // Leap years
1210 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1211 $gDayNo++;
1212 }
1213
1214 // Days passed in current month
1215 $gDayNo += $gd;
1216
1217 $jDayNo = $gDayNo - 79;
1218
1219 $jNp = floor( $jDayNo / 12053 );
1220 $jDayNo %= 12053;
1221
1222 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1223 $jDayNo %= 1461;
1224
1225 if ( $jDayNo >= 366 ) {
1226 $jy += floor( ( $jDayNo - 1 ) / 365 );
1227 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1228 }
1229
1230 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1231 $jDayNo -= self::$IRANIAN_DAYS[$i];
1232 }
1233
1234 $jm = $i + 1;
1235 $jd = $jDayNo + 1;
1236
1237 return array( $jy, $jm, $jd );
1238 }
1239
1240 /**
1241 * Converting Gregorian dates to Hijri dates.
1242 *
1243 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1244 *
1245 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1246 *
1247 * @param $ts string
1248 *
1249 * @return string
1250 */
1251 private static function tsToHijri( $ts ) {
1252 $year = substr( $ts, 0, 4 );
1253 $month = substr( $ts, 4, 2 );
1254 $day = substr( $ts, 6, 2 );
1255
1256 $zyr = $year;
1257 $zd = $day;
1258 $zm = $month;
1259 $zy = $zyr;
1260
1261 if (
1262 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1263 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1264 )
1265 {
1266 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1267 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1268 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1269 $zd - 32075;
1270 } else {
1271 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1272 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1273 }
1274
1275 $zl = $zjd -1948440 + 10632;
1276 $zn = (int)( ( $zl - 1 ) / 10631 );
1277 $zl = $zl - 10631 * $zn + 354;
1278 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1279 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1280 $zm = (int)( ( 24 * $zl ) / 709 );
1281 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1282 $zy = 30 * $zn + $zj - 30;
1283
1284 return array( $zy, $zm, $zd );
1285 }
1286
1287 /**
1288 * Converting Gregorian dates to Hebrew dates.
1289 *
1290 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1291 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1292 * to translate the relevant functions into PHP and release them under
1293 * GNU GPL.
1294 *
1295 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1296 * and Adar II is 14. In a non-leap year, Adar is 6.
1297 *
1298 * @param $ts string
1299 *
1300 * @return string
1301 */
1302 private static function tsToHebrew( $ts ) {
1303 # Parse date
1304 $year = substr( $ts, 0, 4 );
1305 $month = substr( $ts, 4, 2 );
1306 $day = substr( $ts, 6, 2 );
1307
1308 # Calculate Hebrew year
1309 $hebrewYear = $year + 3760;
1310
1311 # Month number when September = 1, August = 12
1312 $month += 4;
1313 if ( $month > 12 ) {
1314 # Next year
1315 $month -= 12;
1316 $year++;
1317 $hebrewYear++;
1318 }
1319
1320 # Calculate day of year from 1 September
1321 $dayOfYear = $day;
1322 for ( $i = 1; $i < $month; $i++ ) {
1323 if ( $i == 6 ) {
1324 # February
1325 $dayOfYear += 28;
1326 # Check if the year is leap
1327 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1328 $dayOfYear++;
1329 }
1330 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1331 $dayOfYear += 30;
1332 } else {
1333 $dayOfYear += 31;
1334 }
1335 }
1336
1337 # Calculate the start of the Hebrew year
1338 $start = self::hebrewYearStart( $hebrewYear );
1339
1340 # Calculate next year's start
1341 if ( $dayOfYear <= $start ) {
1342 # Day is before the start of the year - it is the previous year
1343 # Next year's start
1344 $nextStart = $start;
1345 # Previous year
1346 $year--;
1347 $hebrewYear--;
1348 # Add days since previous year's 1 September
1349 $dayOfYear += 365;
1350 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1351 # Leap year
1352 $dayOfYear++;
1353 }
1354 # Start of the new (previous) year
1355 $start = self::hebrewYearStart( $hebrewYear );
1356 } else {
1357 # Next year's start
1358 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1359 }
1360
1361 # Calculate Hebrew day of year
1362 $hebrewDayOfYear = $dayOfYear - $start;
1363
1364 # Difference between year's days
1365 $diff = $nextStart - $start;
1366 # Add 12 (or 13 for leap years) days to ignore the difference between
1367 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1368 # difference is only about the year type
1369 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1370 $diff += 13;
1371 } else {
1372 $diff += 12;
1373 }
1374
1375 # Check the year pattern, and is leap year
1376 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1377 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1378 # and non-leap years
1379 $yearPattern = $diff % 30;
1380 # Check if leap year
1381 $isLeap = $diff >= 30;
1382
1383 # Calculate day in the month from number of day in the Hebrew year
1384 # Don't check Adar - if the day is not in Adar, we will stop before;
1385 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1386 $hebrewDay = $hebrewDayOfYear;
1387 $hebrewMonth = 1;
1388 $days = 0;
1389 while ( $hebrewMonth <= 12 ) {
1390 # Calculate days in this month
1391 if ( $isLeap && $hebrewMonth == 6 ) {
1392 # Adar in a leap year
1393 if ( $isLeap ) {
1394 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1395 $days = 30;
1396 if ( $hebrewDay <= $days ) {
1397 # Day in Adar I
1398 $hebrewMonth = 13;
1399 } else {
1400 # Subtract the days of Adar I
1401 $hebrewDay -= $days;
1402 # Try Adar II
1403 $days = 29;
1404 if ( $hebrewDay <= $days ) {
1405 # Day in Adar II
1406 $hebrewMonth = 14;
1407 }
1408 }
1409 }
1410 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1411 # Cheshvan in a complete year (otherwise as the rule below)
1412 $days = 30;
1413 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1414 # Kislev in an incomplete year (otherwise as the rule below)
1415 $days = 29;
1416 } else {
1417 # Odd months have 30 days, even have 29
1418 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1419 }
1420 if ( $hebrewDay <= $days ) {
1421 # In the current month
1422 break;
1423 } else {
1424 # Subtract the days of the current month
1425 $hebrewDay -= $days;
1426 # Try in the next month
1427 $hebrewMonth++;
1428 }
1429 }
1430
1431 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1432 }
1433
1434 /**
1435 * This calculates the Hebrew year start, as days since 1 September.
1436 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1437 * Used for Hebrew date.
1438 *
1439 * @param $year int
1440 *
1441 * @return string
1442 */
1443 private static function hebrewYearStart( $year ) {
1444 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1445 $b = intval( ( $year - 1 ) % 4 );
1446 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1447 if ( $m < 0 ) {
1448 $m--;
1449 }
1450 $Mar = intval( $m );
1451 if ( $m < 0 ) {
1452 $m++;
1453 }
1454 $m -= $Mar;
1455
1456 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1457 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1458 $Mar++;
1459 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1460 $Mar += 2;
1461 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1462 $Mar++;
1463 }
1464
1465 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1466 return $Mar;
1467 }
1468
1469 /**
1470 * Algorithm to convert Gregorian dates to Thai solar dates,
1471 * Minguo dates or Minguo dates.
1472 *
1473 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1474 * http://en.wikipedia.org/wiki/Minguo_calendar
1475 * http://en.wikipedia.org/wiki/Japanese_era_name
1476 *
1477 * @param $ts String: 14-character timestamp
1478 * @param $cName String: calender name
1479 * @return Array: converted year, month, day
1480 */
1481 private static function tsToYear( $ts, $cName ) {
1482 $gy = substr( $ts, 0, 4 );
1483 $gm = substr( $ts, 4, 2 );
1484 $gd = substr( $ts, 6, 2 );
1485
1486 if ( !strcmp( $cName, 'thai' ) ) {
1487 # Thai solar dates
1488 # Add 543 years to the Gregorian calendar
1489 # Months and days are identical
1490 $gy_offset = $gy + 543;
1491 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1492 # Minguo dates
1493 # Deduct 1911 years from the Gregorian calendar
1494 # Months and days are identical
1495 $gy_offset = $gy - 1911;
1496 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1497 # Nengō dates up to Meiji period
1498 # Deduct years from the Gregorian calendar
1499 # depending on the nengo periods
1500 # Months and days are identical
1501 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1502 # Meiji period
1503 $gy_gannen = $gy - 1868 + 1;
1504 $gy_offset = $gy_gannen;
1505 if ( $gy_gannen == 1 ) {
1506 $gy_offset = '元';
1507 }
1508 $gy_offset = '明治' . $gy_offset;
1509 } elseif (
1510 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1511 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1512 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1513 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1514 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1515 )
1516 {
1517 # Taishō period
1518 $gy_gannen = $gy - 1912 + 1;
1519 $gy_offset = $gy_gannen;
1520 if ( $gy_gannen == 1 ) {
1521 $gy_offset = '元';
1522 }
1523 $gy_offset = '大正' . $gy_offset;
1524 } elseif (
1525 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1526 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1527 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1528 )
1529 {
1530 # Shōwa period
1531 $gy_gannen = $gy - 1926 + 1;
1532 $gy_offset = $gy_gannen;
1533 if ( $gy_gannen == 1 ) {
1534 $gy_offset = '元';
1535 }
1536 $gy_offset = '昭和' . $gy_offset;
1537 } else {
1538 # Heisei period
1539 $gy_gannen = $gy - 1989 + 1;
1540 $gy_offset = $gy_gannen;
1541 if ( $gy_gannen == 1 ) {
1542 $gy_offset = '元';
1543 }
1544 $gy_offset = '平成' . $gy_offset;
1545 }
1546 } else {
1547 $gy_offset = $gy;
1548 }
1549
1550 return array( $gy_offset, $gm, $gd );
1551 }
1552
1553 /**
1554 * Roman number formatting up to 3000
1555 *
1556 * @param $num int
1557 *
1558 * @return string
1559 */
1560 static function romanNumeral( $num ) {
1561 static $table = array(
1562 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1563 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1564 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1565 array( '', 'M', 'MM', 'MMM' )
1566 );
1567
1568 $num = intval( $num );
1569 if ( $num > 3000 || $num <= 0 ) {
1570 return $num;
1571 }
1572
1573 $s = '';
1574 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1575 if ( $num >= $pow10 ) {
1576 $s .= $table[$i][floor( $num / $pow10 )];
1577 }
1578 $num = $num % $pow10;
1579 }
1580 return $s;
1581 }
1582
1583 /**
1584 * Hebrew Gematria number formatting up to 9999
1585 *
1586 * @param $num int
1587 *
1588 * @return string
1589 */
1590 static function hebrewNumeral( $num ) {
1591 static $table = array(
1592 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1593 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1594 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1595 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1596 );
1597
1598 $num = intval( $num );
1599 if ( $num > 9999 || $num <= 0 ) {
1600 return $num;
1601 }
1602
1603 $s = '';
1604 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1605 if ( $num >= $pow10 ) {
1606 if ( $num == 15 || $num == 16 ) {
1607 $s .= $table[0][9] . $table[0][$num - 9];
1608 $num = 0;
1609 } else {
1610 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1611 if ( $pow10 == 1000 ) {
1612 $s .= "'";
1613 }
1614 }
1615 }
1616 $num = $num % $pow10;
1617 }
1618 if ( strlen( $s ) == 2 ) {
1619 $str = $s . "'";
1620 } else {
1621 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1622 $str .= substr( $s, strlen( $s ) - 2, 2 );
1623 }
1624 $start = substr( $str, 0, strlen( $str ) - 2 );
1625 $end = substr( $str, strlen( $str ) - 2 );
1626 switch( $end ) {
1627 case 'כ':
1628 $str = $start . 'ך';
1629 break;
1630 case 'מ':
1631 $str = $start . 'ם';
1632 break;
1633 case 'נ':
1634 $str = $start . 'ן';
1635 break;
1636 case 'פ':
1637 $str = $start . 'ף';
1638 break;
1639 case 'צ':
1640 $str = $start . 'ץ';
1641 break;
1642 }
1643 return $str;
1644 }
1645
1646 /**
1647 * Used by date() and time() to adjust the time output.
1648 *
1649 * @param $ts Int the time in date('YmdHis') format
1650 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1651 * get user timecorrection setting)
1652 * @return int
1653 */
1654 function userAdjust( $ts, $tz = false ) {
1655 global $wgUser, $wgLocalTZoffset;
1656
1657 if ( $tz === false ) {
1658 $tz = $wgUser->getOption( 'timecorrection' );
1659 }
1660
1661 $data = explode( '|', $tz, 3 );
1662
1663 if ( $data[0] == 'ZoneInfo' ) {
1664 wfSuppressWarnings();
1665 $userTZ = timezone_open( $data[2] );
1666 wfRestoreWarnings();
1667 if ( $userTZ !== false ) {
1668 $date = date_create( $ts, timezone_open( 'UTC' ) );
1669 date_timezone_set( $date, $userTZ );
1670 $date = date_format( $date, 'YmdHis' );
1671 return $date;
1672 }
1673 # Unrecognized timezone, default to 'Offset' with the stored offset.
1674 $data[0] = 'Offset';
1675 }
1676
1677 $minDiff = 0;
1678 if ( $data[0] == 'System' || $tz == '' ) {
1679 #  Global offset in minutes.
1680 if ( isset( $wgLocalTZoffset ) ) {
1681 $minDiff = $wgLocalTZoffset;
1682 }
1683 } elseif ( $data[0] == 'Offset' ) {
1684 $minDiff = intval( $data[1] );
1685 } else {
1686 $data = explode( ':', $tz );
1687 if ( count( $data ) == 2 ) {
1688 $data[0] = intval( $data[0] );
1689 $data[1] = intval( $data[1] );
1690 $minDiff = abs( $data[0] ) * 60 + $data[1];
1691 if ( $data[0] < 0 ) {
1692 $minDiff = -$minDiff;
1693 }
1694 } else {
1695 $minDiff = intval( $data[0] ) * 60;
1696 }
1697 }
1698
1699 # No difference ? Return time unchanged
1700 if ( 0 == $minDiff ) {
1701 return $ts;
1702 }
1703
1704 wfSuppressWarnings(); // E_STRICT system time bitching
1705 # Generate an adjusted date; take advantage of the fact that mktime
1706 # will normalize out-of-range values so we don't have to split $minDiff
1707 # into hours and minutes.
1708 $t = mktime( (
1709 (int)substr( $ts, 8, 2 ) ), # Hours
1710 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
1711 (int)substr( $ts, 12, 2 ), # Seconds
1712 (int)substr( $ts, 4, 2 ), # Month
1713 (int)substr( $ts, 6, 2 ), # Day
1714 (int)substr( $ts, 0, 4 ) ); # Year
1715
1716 $date = date( 'YmdHis', $t );
1717 wfRestoreWarnings();
1718
1719 return $date;
1720 }
1721
1722 /**
1723 * This is meant to be used by time(), date(), and timeanddate() to get
1724 * the date preference they're supposed to use, it should be used in
1725 * all children.
1726 *
1727 *<code>
1728 * function timeanddate([...], $format = true) {
1729 * $datePreference = $this->dateFormat($format);
1730 * [...]
1731 * }
1732 *</code>
1733 *
1734 * @param $usePrefs Mixed: if true, the user's preference is used
1735 * if false, the site/language default is used
1736 * if int/string, assumed to be a format.
1737 * @return string
1738 */
1739 function dateFormat( $usePrefs = true ) {
1740 global $wgUser;
1741
1742 if ( is_bool( $usePrefs ) ) {
1743 if ( $usePrefs ) {
1744 $datePreference = $wgUser->getDatePreference();
1745 } else {
1746 $datePreference = (string)User::getDefaultOption( 'date' );
1747 }
1748 } else {
1749 $datePreference = (string)$usePrefs;
1750 }
1751
1752 // return int
1753 if ( $datePreference == '' ) {
1754 return 'default';
1755 }
1756
1757 return $datePreference;
1758 }
1759
1760 /**
1761 * Get a format string for a given type and preference
1762 * @param $type string May be date, time or both
1763 * @param $pref string The format name as it appears in Messages*.php
1764 *
1765 * @return string
1766 */
1767 function getDateFormatString( $type, $pref ) {
1768 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
1769 if ( $pref == 'default' ) {
1770 $pref = $this->getDefaultDateFormat();
1771 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1772 } else {
1773 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1774 if ( is_null( $df ) ) {
1775 $pref = $this->getDefaultDateFormat();
1776 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1777 }
1778 }
1779 $this->dateFormatStrings[$type][$pref] = $df;
1780 }
1781 return $this->dateFormatStrings[$type][$pref];
1782 }
1783
1784 /**
1785 * @param $ts Mixed: the time format which needs to be turned into a
1786 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1787 * @param $adj Bool: whether to adjust the time output according to the
1788 * user configured offset ($timecorrection)
1789 * @param $format Mixed: true to use user's date format preference
1790 * @param $timecorrection String|bool the time offset as returned by
1791 * validateTimeZone() in Special:Preferences
1792 * @return string
1793 */
1794 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1795 $ts = wfTimestamp( TS_MW, $ts );
1796 if ( $adj ) {
1797 $ts = $this->userAdjust( $ts, $timecorrection );
1798 }
1799 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1800 return $this->sprintfDate( $df, $ts );
1801 }
1802
1803 /**
1804 * @param $ts Mixed: the time format which needs to be turned into a
1805 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1806 * @param $adj Bool: whether to adjust the time output according to the
1807 * user configured offset ($timecorrection)
1808 * @param $format Mixed: true to use user's date format preference
1809 * @param $timecorrection String|bool the time offset as returned by
1810 * validateTimeZone() in Special:Preferences
1811 * @return string
1812 */
1813 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1814 $ts = wfTimestamp( TS_MW, $ts );
1815 if ( $adj ) {
1816 $ts = $this->userAdjust( $ts, $timecorrection );
1817 }
1818 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1819 return $this->sprintfDate( $df, $ts );
1820 }
1821
1822 /**
1823 * @param $ts Mixed: the time format which needs to be turned into a
1824 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1825 * @param $adj Bool: whether to adjust the time output according to the
1826 * user configured offset ($timecorrection)
1827 * @param $format Mixed: what format to return, if it's false output the
1828 * default one (default true)
1829 * @param $timecorrection String|bool the time offset as returned by
1830 * validateTimeZone() in Special:Preferences
1831 * @return string
1832 */
1833 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1834 $ts = wfTimestamp( TS_MW, $ts );
1835 if ( $adj ) {
1836 $ts = $this->userAdjust( $ts, $timecorrection );
1837 }
1838 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1839 return $this->sprintfDate( $df, $ts );
1840 }
1841
1842 /**
1843 * Internal helper function for userDate(), userTime() and userTimeAndDate()
1844 *
1845 * @param $type String: can be 'date', 'time' or 'both'
1846 * @param $ts Mixed: the time format which needs to be turned into a
1847 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1848 * @param $user User object used to get preferences for timezone and format
1849 * @param $options Array, can contain the following keys:
1850 * - 'timecorrection': time correction, can have the following values:
1851 * - true: use user's preference
1852 * - false: don't use time correction
1853 * - integer: value of time correction in minutes
1854 * - 'format': format to use, can have the following values:
1855 * - true: use user's preference
1856 * - false: use default preference
1857 * - string: format to use
1858 * @return String
1859 */
1860 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
1861 $ts = wfTimestamp( TS_MW, $ts );
1862 $options += array( 'timecorrection' => true, 'format' => true );
1863 if ( $options['timecorrection'] !== false ) {
1864 if ( $options['timecorrection'] === true ) {
1865 $offset = $user->getOption( 'timecorrection' );
1866 } else {
1867 $offset = $options['timecorrection'];
1868 }
1869 $ts = $this->userAdjust( $ts, $offset );
1870 }
1871 if ( $options['format'] === true ) {
1872 $format = $user->getDatePreference();
1873 } else {
1874 $format = $options['format'];
1875 }
1876 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
1877 return $this->sprintfDate( $df, $ts );
1878 }
1879
1880 /**
1881 * Get the formatted date for the given timestamp and formatted for
1882 * the given user.
1883 *
1884 * @param $type String: can be 'date', 'time' or 'both'
1885 * @param $ts Mixed: the time format which needs to be turned into a
1886 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1887 * @param $user User object used to get preferences for timezone and format
1888 * @param $options Array, can contain the following keys:
1889 * - 'timecorrection': time correction, can have the following values:
1890 * - true: use user's preference
1891 * - false: don't use time correction
1892 * - integer: value of time correction in minutes
1893 * - 'format': format to use, can have the following values:
1894 * - true: use user's preference
1895 * - false: use default preference
1896 * - string: format to use
1897 * @return String
1898 */
1899 public function userDate( $ts, User $user, array $options = array() ) {
1900 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
1901 }
1902
1903 /**
1904 * Get the formatted time for the given timestamp and formatted for
1905 * the given user.
1906 *
1907 * @param $type String: can be 'date', 'time' or 'both'
1908 * @param $ts Mixed: the time format which needs to be turned into a
1909 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1910 * @param $user User object used to get preferences for timezone and format
1911 * @param $options Array, can contain the following keys:
1912 * - 'timecorrection': time correction, can have the following values:
1913 * - true: use user's preference
1914 * - false: don't use time correction
1915 * - integer: value of time correction in minutes
1916 * - 'format': format to use, can have the following values:
1917 * - true: use user's preference
1918 * - false: use default preference
1919 * - string: format to use
1920 * @return String
1921 */
1922 public function userTime( $ts, User $user, array $options = array() ) {
1923 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
1924 }
1925
1926 /**
1927 * Get the formatted date and time for the given timestamp and formatted for
1928 * the given user.
1929 *
1930 * @param $type String: can be 'date', 'time' or 'both'
1931 * @param $ts Mixed: the time format which needs to be turned into a
1932 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1933 * @param $user User object used to get preferences for timezone and format
1934 * @param $options Array, can contain the following keys:
1935 * - 'timecorrection': time correction, can have the following values:
1936 * - true: use user's preference
1937 * - false: don't use time correction
1938 * - integer: value of time correction in minutes
1939 * - 'format': format to use, can have the following values:
1940 * - true: use user's preference
1941 * - false: use default preference
1942 * - string: format to use
1943 * @return String
1944 */
1945 public function userTimeAndDate( $ts, User $user, array $options = array() ) {
1946 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
1947 }
1948
1949 /**
1950 * @param $key string
1951 * @return array|null
1952 */
1953 function getMessage( $key ) {
1954 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
1955 }
1956
1957 /**
1958 * @return array
1959 */
1960 function getAllMessages() {
1961 return self::$dataCache->getItem( $this->mCode, 'messages' );
1962 }
1963
1964 /**
1965 * @param $in
1966 * @param $out
1967 * @param $string
1968 * @return string
1969 */
1970 function iconv( $in, $out, $string ) {
1971 # This is a wrapper for iconv in all languages except esperanto,
1972 # which does some nasty x-conversions beforehand
1973
1974 # Even with //IGNORE iconv can whine about illegal characters in
1975 # *input* string. We just ignore those too.
1976 # REF: http://bugs.php.net/bug.php?id=37166
1977 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1978 wfSuppressWarnings();
1979 $text = iconv( $in, $out . '//IGNORE', $string );
1980 wfRestoreWarnings();
1981 return $text;
1982 }
1983
1984 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1985
1986 /**
1987 * @param $matches array
1988 * @return mixed|string
1989 */
1990 function ucwordbreaksCallbackAscii( $matches ) {
1991 return $this->ucfirst( $matches[1] );
1992 }
1993
1994 /**
1995 * @param $matches array
1996 * @return string
1997 */
1998 function ucwordbreaksCallbackMB( $matches ) {
1999 return mb_strtoupper( $matches[0] );
2000 }
2001
2002 /**
2003 * @param $matches array
2004 * @return string
2005 */
2006 function ucCallback( $matches ) {
2007 list( $wikiUpperChars ) = self::getCaseMaps();
2008 return strtr( $matches[1], $wikiUpperChars );
2009 }
2010
2011 /**
2012 * @param $matches array
2013 * @return string
2014 */
2015 function lcCallback( $matches ) {
2016 list( , $wikiLowerChars ) = self::getCaseMaps();
2017 return strtr( $matches[1], $wikiLowerChars );
2018 }
2019
2020 /**
2021 * @param $matches array
2022 * @return string
2023 */
2024 function ucwordsCallbackMB( $matches ) {
2025 return mb_strtoupper( $matches[0] );
2026 }
2027
2028 /**
2029 * @param $matches array
2030 * @return string
2031 */
2032 function ucwordsCallbackWiki( $matches ) {
2033 list( $wikiUpperChars ) = self::getCaseMaps();
2034 return strtr( $matches[0], $wikiUpperChars );
2035 }
2036
2037 /**
2038 * Make a string's first character uppercase
2039 *
2040 * @param $str string
2041 *
2042 * @return string
2043 */
2044 function ucfirst( $str ) {
2045 $o = ord( $str );
2046 if ( $o < 96 ) { // if already uppercase...
2047 return $str;
2048 } elseif ( $o < 128 ) {
2049 return ucfirst( $str ); // use PHP's ucfirst()
2050 } else {
2051 // fall back to more complex logic in case of multibyte strings
2052 return $this->uc( $str, true );
2053 }
2054 }
2055
2056 /**
2057 * Convert a string to uppercase
2058 *
2059 * @param $str string
2060 * @param $first bool
2061 *
2062 * @return string
2063 */
2064 function uc( $str, $first = false ) {
2065 if ( function_exists( 'mb_strtoupper' ) ) {
2066 if ( $first ) {
2067 if ( $this->isMultibyte( $str ) ) {
2068 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2069 } else {
2070 return ucfirst( $str );
2071 }
2072 } else {
2073 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2074 }
2075 } else {
2076 if ( $this->isMultibyte( $str ) ) {
2077 $x = $first ? '^' : '';
2078 return preg_replace_callback(
2079 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2080 array( $this, 'ucCallback' ),
2081 $str
2082 );
2083 } else {
2084 return $first ? ucfirst( $str ) : strtoupper( $str );
2085 }
2086 }
2087 }
2088
2089 /**
2090 * @param $str string
2091 * @return mixed|string
2092 */
2093 function lcfirst( $str ) {
2094 $o = ord( $str );
2095 if ( !$o ) {
2096 return strval( $str );
2097 } elseif ( $o >= 128 ) {
2098 return $this->lc( $str, true );
2099 } elseif ( $o > 96 ) {
2100 return $str;
2101 } else {
2102 $str[0] = strtolower( $str[0] );
2103 return $str;
2104 }
2105 }
2106
2107 /**
2108 * @param $str string
2109 * @param $first bool
2110 * @return mixed|string
2111 */
2112 function lc( $str, $first = false ) {
2113 if ( function_exists( 'mb_strtolower' ) ) {
2114 if ( $first ) {
2115 if ( $this->isMultibyte( $str ) ) {
2116 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2117 } else {
2118 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2119 }
2120 } else {
2121 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2122 }
2123 } else {
2124 if ( $this->isMultibyte( $str ) ) {
2125 $x = $first ? '^' : '';
2126 return preg_replace_callback(
2127 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2128 array( $this, 'lcCallback' ),
2129 $str
2130 );
2131 } else {
2132 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2133 }
2134 }
2135 }
2136
2137 /**
2138 * @param $str string
2139 * @return bool
2140 */
2141 function isMultibyte( $str ) {
2142 return (bool)preg_match( '/[\x80-\xff]/', $str );
2143 }
2144
2145 /**
2146 * @param $str string
2147 * @return mixed|string
2148 */
2149 function ucwords( $str ) {
2150 if ( $this->isMultibyte( $str ) ) {
2151 $str = $this->lc( $str );
2152
2153 // regexp to find first letter in each word (i.e. after each space)
2154 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2155
2156 // function to use to capitalize a single char
2157 if ( function_exists( 'mb_strtoupper' ) ) {
2158 return preg_replace_callback(
2159 $replaceRegexp,
2160 array( $this, 'ucwordsCallbackMB' ),
2161 $str
2162 );
2163 } else {
2164 return preg_replace_callback(
2165 $replaceRegexp,
2166 array( $this, 'ucwordsCallbackWiki' ),
2167 $str
2168 );
2169 }
2170 } else {
2171 return ucwords( strtolower( $str ) );
2172 }
2173 }
2174
2175 /**
2176 * capitalize words at word breaks
2177 *
2178 * @param $str string
2179 * @return mixed
2180 */
2181 function ucwordbreaks( $str ) {
2182 if ( $this->isMultibyte( $str ) ) {
2183 $str = $this->lc( $str );
2184
2185 // since \b doesn't work for UTF-8, we explicitely define word break chars
2186 $breaks = "[ \-\(\)\}\{\.,\?!]";
2187
2188 // find first letter after word break
2189 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2190
2191 if ( function_exists( 'mb_strtoupper' ) ) {
2192 return preg_replace_callback(
2193 $replaceRegexp,
2194 array( $this, 'ucwordbreaksCallbackMB' ),
2195 $str
2196 );
2197 } else {
2198 return preg_replace_callback(
2199 $replaceRegexp,
2200 array( $this, 'ucwordsCallbackWiki' ),
2201 $str
2202 );
2203 }
2204 } else {
2205 return preg_replace_callback(
2206 '/\b([\w\x80-\xff]+)\b/',
2207 array( $this, 'ucwordbreaksCallbackAscii' ),
2208 $str
2209 );
2210 }
2211 }
2212
2213 /**
2214 * Return a case-folded representation of $s
2215 *
2216 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2217 * and $s2 are the same except for the case of their characters. It is not
2218 * necessary for the value returned to make sense when displayed.
2219 *
2220 * Do *not* perform any other normalisation in this function. If a caller
2221 * uses this function when it should be using a more general normalisation
2222 * function, then fix the caller.
2223 *
2224 * @param $s string
2225 *
2226 * @return string
2227 */
2228 function caseFold( $s ) {
2229 return $this->uc( $s );
2230 }
2231
2232 /**
2233 * @param $s string
2234 * @return string
2235 */
2236 function checkTitleEncoding( $s ) {
2237 if ( is_array( $s ) ) {
2238 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2239 }
2240 # Check for non-UTF-8 URLs
2241 $ishigh = preg_match( '/[\x80-\xff]/', $s );
2242 if ( !$ishigh ) {
2243 return $s;
2244 }
2245
2246 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2247 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
2248 if ( $isutf8 ) {
2249 return $s;
2250 }
2251
2252 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2253 }
2254
2255 /**
2256 * @return array
2257 */
2258 function fallback8bitEncoding() {
2259 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2260 }
2261
2262 /**
2263 * Most writing systems use whitespace to break up words.
2264 * Some languages such as Chinese don't conventionally do this,
2265 * which requires special handling when breaking up words for
2266 * searching etc.
2267 *
2268 * @return bool
2269 */
2270 function hasWordBreaks() {
2271 return true;
2272 }
2273
2274 /**
2275 * Some languages such as Chinese require word segmentation,
2276 * Specify such segmentation when overridden in derived class.
2277 *
2278 * @param $string String
2279 * @return String
2280 */
2281 function segmentByWord( $string ) {
2282 return $string;
2283 }
2284
2285 /**
2286 * Some languages have special punctuation need to be normalized.
2287 * Make such changes here.
2288 *
2289 * @param $string String
2290 * @return String
2291 */
2292 function normalizeForSearch( $string ) {
2293 return self::convertDoubleWidth( $string );
2294 }
2295
2296 /**
2297 * convert double-width roman characters to single-width.
2298 * range: ff00-ff5f ~= 0020-007f
2299 *
2300 * @param $string string
2301 *
2302 * @return string
2303 */
2304 protected static function convertDoubleWidth( $string ) {
2305 static $full = null;
2306 static $half = null;
2307
2308 if ( $full === null ) {
2309 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2310 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2311 $full = str_split( $fullWidth, 3 );
2312 $half = str_split( $halfWidth );
2313 }
2314
2315 $string = str_replace( $full, $half, $string );
2316 return $string;
2317 }
2318
2319 /**
2320 * @param $string string
2321 * @param $pattern string
2322 * @return string
2323 */
2324 protected static function insertSpace( $string, $pattern ) {
2325 $string = preg_replace( $pattern, " $1 ", $string );
2326 $string = preg_replace( '/ +/', ' ', $string );
2327 return $string;
2328 }
2329
2330 /**
2331 * @param $termsArray array
2332 * @return array
2333 */
2334 function convertForSearchResult( $termsArray ) {
2335 # some languages, e.g. Chinese, need to do a conversion
2336 # in order for search results to be displayed correctly
2337 return $termsArray;
2338 }
2339
2340 /**
2341 * Get the first character of a string.
2342 *
2343 * @param $s string
2344 * @return string
2345 */
2346 function firstChar( $s ) {
2347 $matches = array();
2348 preg_match(
2349 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2350 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2351 $s,
2352 $matches
2353 );
2354
2355 if ( isset( $matches[1] ) ) {
2356 if ( strlen( $matches[1] ) != 3 ) {
2357 return $matches[1];
2358 }
2359
2360 // Break down Hangul syllables to grab the first jamo
2361 $code = utf8ToCodepoint( $matches[1] );
2362 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2363 return $matches[1];
2364 } elseif ( $code < 0xb098 ) {
2365 return "\xe3\x84\xb1";
2366 } elseif ( $code < 0xb2e4 ) {
2367 return "\xe3\x84\xb4";
2368 } elseif ( $code < 0xb77c ) {
2369 return "\xe3\x84\xb7";
2370 } elseif ( $code < 0xb9c8 ) {
2371 return "\xe3\x84\xb9";
2372 } elseif ( $code < 0xbc14 ) {
2373 return "\xe3\x85\x81";
2374 } elseif ( $code < 0xc0ac ) {
2375 return "\xe3\x85\x82";
2376 } elseif ( $code < 0xc544 ) {
2377 return "\xe3\x85\x85";
2378 } elseif ( $code < 0xc790 ) {
2379 return "\xe3\x85\x87";
2380 } elseif ( $code < 0xcc28 ) {
2381 return "\xe3\x85\x88";
2382 } elseif ( $code < 0xce74 ) {
2383 return "\xe3\x85\x8a";
2384 } elseif ( $code < 0xd0c0 ) {
2385 return "\xe3\x85\x8b";
2386 } elseif ( $code < 0xd30c ) {
2387 return "\xe3\x85\x8c";
2388 } elseif ( $code < 0xd558 ) {
2389 return "\xe3\x85\x8d";
2390 } else {
2391 return "\xe3\x85\x8e";
2392 }
2393 } else {
2394 return '';
2395 }
2396 }
2397
2398 function initEncoding() {
2399 # Some languages may have an alternate char encoding option
2400 # (Esperanto X-coding, Japanese furigana conversion, etc)
2401 # If this language is used as the primary content language,
2402 # an override to the defaults can be set here on startup.
2403 }
2404
2405 /**
2406 * @param $s string
2407 * @return string
2408 */
2409 function recodeForEdit( $s ) {
2410 # For some languages we'll want to explicitly specify
2411 # which characters make it into the edit box raw
2412 # or are converted in some way or another.
2413 global $wgEditEncoding;
2414 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
2415 return $s;
2416 } else {
2417 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2418 }
2419 }
2420
2421 /**
2422 * @param $s string
2423 * @return string
2424 */
2425 function recodeInput( $s ) {
2426 # Take the previous into account.
2427 global $wgEditEncoding;
2428 if ( $wgEditEncoding != '' ) {
2429 $enc = $wgEditEncoding;
2430 } else {
2431 $enc = 'UTF-8';
2432 }
2433 if ( $enc == 'UTF-8' ) {
2434 return $s;
2435 } else {
2436 return $this->iconv( $enc, 'UTF-8', $s );
2437 }
2438 }
2439
2440 /**
2441 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2442 * also cleans up certain backwards-compatible sequences, converting them
2443 * to the modern Unicode equivalent.
2444 *
2445 * This is language-specific for performance reasons only.
2446 *
2447 * @param $s string
2448 *
2449 * @return string
2450 */
2451 function normalize( $s ) {
2452 global $wgAllUnicodeFixes;
2453 $s = UtfNormal::cleanUp( $s );
2454 if ( $wgAllUnicodeFixes ) {
2455 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2456 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2457 }
2458
2459 return $s;
2460 }
2461
2462 /**
2463 * Transform a string using serialized data stored in the given file (which
2464 * must be in the serialized subdirectory of $IP). The file contains pairs
2465 * mapping source characters to destination characters.
2466 *
2467 * The data is cached in process memory. This will go faster if you have the
2468 * FastStringSearch extension.
2469 *
2470 * @param $file string
2471 * @param $string string
2472 *
2473 * @return string
2474 */
2475 function transformUsingPairFile( $file, $string ) {
2476 if ( !isset( $this->transformData[$file] ) ) {
2477 $data = wfGetPrecompiledData( $file );
2478 if ( $data === false ) {
2479 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2480 }
2481 $this->transformData[$file] = new ReplacementArray( $data );
2482 }
2483 return $this->transformData[$file]->replace( $string );
2484 }
2485
2486 /**
2487 * For right-to-left language support
2488 *
2489 * @return bool
2490 */
2491 function isRTL() {
2492 return self::$dataCache->getItem( $this->mCode, 'rtl' );
2493 }
2494
2495 /**
2496 * Return the correct HTML 'dir' attribute value for this language.
2497 * @return String
2498 */
2499 function getDir() {
2500 return $this->isRTL() ? 'rtl' : 'ltr';
2501 }
2502
2503 /**
2504 * Return 'left' or 'right' as appropriate alignment for line-start
2505 * for this language's text direction.
2506 *
2507 * Should be equivalent to CSS3 'start' text-align value....
2508 *
2509 * @return String
2510 */
2511 function alignStart() {
2512 return $this->isRTL() ? 'right' : 'left';
2513 }
2514
2515 /**
2516 * Return 'right' or 'left' as appropriate alignment for line-end
2517 * for this language's text direction.
2518 *
2519 * Should be equivalent to CSS3 'end' text-align value....
2520 *
2521 * @return String
2522 */
2523 function alignEnd() {
2524 return $this->isRTL() ? 'left' : 'right';
2525 }
2526
2527 /**
2528 * A hidden direction mark (LRM or RLM), depending on the language direction
2529 *
2530 * @param $opposite Boolean Get the direction mark opposite to your language
2531 * @return string
2532 */
2533 function getDirMark( $opposite = false ) {
2534 $rtl = "\xE2\x80\x8F";
2535 $ltr = "\xE2\x80\x8E";
2536 if ( $opposite ) { return $this->isRTL() ? $ltr : $rtl; }
2537 return $this->isRTL() ? $rtl : $ltr;
2538 }
2539
2540 /**
2541 * @return array
2542 */
2543 function capitalizeAllNouns() {
2544 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2545 }
2546
2547 /**
2548 * An arrow, depending on the language direction
2549 *
2550 * @return string
2551 */
2552 function getArrow() {
2553 return $this->isRTL() ? '←' : '→';
2554 }
2555
2556 /**
2557 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2558 *
2559 * @return bool
2560 */
2561 function linkPrefixExtension() {
2562 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2563 }
2564
2565 /**
2566 * @return array
2567 */
2568 function getMagicWords() {
2569 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2570 }
2571
2572 protected function doMagicHook() {
2573 if ( $this->mMagicHookDone ) {
2574 return;
2575 }
2576 $this->mMagicHookDone = true;
2577 wfProfileIn( 'LanguageGetMagic' );
2578 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2579 wfProfileOut( 'LanguageGetMagic' );
2580 }
2581
2582 /**
2583 * Fill a MagicWord object with data from here
2584 *
2585 * @param $mw
2586 */
2587 function getMagic( $mw ) {
2588 $this->doMagicHook();
2589
2590 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
2591 $rawEntry = $this->mMagicExtensions[$mw->mId];
2592 } else {
2593 $magicWords = $this->getMagicWords();
2594 if ( isset( $magicWords[$mw->mId] ) ) {
2595 $rawEntry = $magicWords[$mw->mId];
2596 } else {
2597 $rawEntry = false;
2598 }
2599 }
2600
2601 if ( !is_array( $rawEntry ) ) {
2602 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
2603 } else {
2604 $mw->mCaseSensitive = $rawEntry[0];
2605 $mw->mSynonyms = array_slice( $rawEntry, 1 );
2606 }
2607 }
2608
2609 /**
2610 * Add magic words to the extension array
2611 *
2612 * @param $newWords array
2613 */
2614 function addMagicWordsByLang( $newWords ) {
2615 $fallbackChain = $this->getFallbackLanguages();
2616 $fallbackChain = array_reverse( $fallbackChain );
2617 foreach ( $fallbackChain as $code ) {
2618 if ( isset( $newWords[$code] ) ) {
2619 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
2620 }
2621 }
2622 }
2623
2624 /**
2625 * Get special page names, as an associative array
2626 * case folded alias => real name
2627 */
2628 function getSpecialPageAliases() {
2629 // Cache aliases because it may be slow to load them
2630 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
2631 // Initialise array
2632 $this->mExtendedSpecialPageAliases =
2633 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
2634 wfRunHooks( 'LanguageGetSpecialPageAliases',
2635 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
2636 }
2637
2638 return $this->mExtendedSpecialPageAliases;
2639 }
2640
2641 /**
2642 * Italic is unsuitable for some languages
2643 *
2644 * @param $text String: the text to be emphasized.
2645 * @return string
2646 */
2647 function emphasize( $text ) {
2648 return "<em>$text</em>";
2649 }
2650
2651 /**
2652 * Normally we output all numbers in plain en_US style, that is
2653 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2654 * point twohundredthirtyfive. However this is not suitable for all
2655 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2656 * Icelandic just want to use commas instead of dots, and dots instead
2657 * of commas like "293.291,235".
2658 *
2659 * An example of this function being called:
2660 * <code>
2661 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2662 * </code>
2663 *
2664 * See LanguageGu.php for the Gujarati implementation and
2665 * $separatorTransformTable on MessageIs.php for
2666 * the , => . and . => , implementation.
2667 *
2668 * @todo check if it's viable to use localeconv() for the decimal
2669 * separator thing.
2670 * @param $number Mixed: the string to be formatted, should be an integer
2671 * or a floating point number.
2672 * @param $nocommafy Bool: set to true for special numbers like dates
2673 * @return string
2674 */
2675 function formatNum( $number, $nocommafy = false ) {
2676 global $wgTranslateNumerals;
2677 if ( !$nocommafy ) {
2678 $number = $this->commafy( $number );
2679 $s = $this->separatorTransformTable();
2680 if ( $s ) {
2681 $number = strtr( $number, $s );
2682 }
2683 }
2684
2685 if ( $wgTranslateNumerals ) {
2686 $s = $this->digitTransformTable();
2687 if ( $s ) {
2688 $number = strtr( $number, $s );
2689 }
2690 }
2691
2692 return $number;
2693 }
2694
2695 /**
2696 * @param $number string
2697 * @return string
2698 */
2699 function parseFormattedNumber( $number ) {
2700 $s = $this->digitTransformTable();
2701 if ( $s ) {
2702 $number = strtr( $number, array_flip( $s ) );
2703 }
2704
2705 $s = $this->separatorTransformTable();
2706 if ( $s ) {
2707 $number = strtr( $number, array_flip( $s ) );
2708 }
2709
2710 $number = strtr( $number, array( ',' => '' ) );
2711 return $number;
2712 }
2713
2714 /**
2715 * Adds commas to a given number
2716 * @since 1.19
2717 * @param $_ mixed
2718 * @return string
2719 */
2720 function commafy( $_ ) {
2721 $digitGroupingPattern = $this->digitGroupingPattern();
2722
2723 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
2724 // default grouping is at thousands, use the same for ###,###,### pattern too.
2725 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2726 } else {
2727 // Ref: http://cldr.unicode.org/translation/number-patterns
2728 $numberpart = array();
2729 $decimalpart = array();
2730 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
2731 preg_match( "/\d+/", $_, $numberpart );
2732 preg_match( "/\.\d*/", $_, $decimalpart );
2733 $groupedNumber = ( count( $decimalpart ) > 0 ) ? $decimalpart[0]:"";
2734 if ( $groupedNumber === $_ ) {
2735 // the string does not have any number part. Eg: .12345
2736 return $groupedNumber;
2737 }
2738 $start = $end = strlen( $numberpart[0] );
2739 while ( $start > 0 ) {
2740 $match = $matches[0][$numMatches -1] ;
2741 $matchLen = strlen( $match );
2742 $start = $end - $matchLen;
2743 if ( $start < 0 ) {
2744 $start = 0;
2745 }
2746 $groupedNumber = substr( $_ , $start, $end -$start ) . $groupedNumber ;
2747 $end = $start;
2748 if ( $numMatches > 1 ) {
2749 // use the last pattern for the rest of the number
2750 $numMatches--;
2751 }
2752 if ( $start > 0 ) {
2753 $groupedNumber = "," . $groupedNumber;
2754 }
2755 }
2756 return $groupedNumber;
2757 }
2758 }
2759 /**
2760 * @return String
2761 */
2762 function digitGroupingPattern() {
2763 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
2764 }
2765
2766 /**
2767 * @return array
2768 */
2769 function digitTransformTable() {
2770 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
2771 }
2772
2773 /**
2774 * @return array
2775 */
2776 function separatorTransformTable() {
2777 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
2778 }
2779
2780 /**
2781 * Take a list of strings and build a locale-friendly comma-separated
2782 * list, using the local comma-separator message.
2783 * The last two strings are chained with an "and".
2784 *
2785 * @param $l Array
2786 * @return string
2787 */
2788 function listToText( $l ) {
2789 $s = '';
2790 $m = count( $l ) - 1;
2791 if ( $m == 1 ) {
2792 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2793 } else {
2794 for ( $i = $m; $i >= 0; $i-- ) {
2795 if ( $i == $m ) {
2796 $s = $l[$i];
2797 } elseif ( $i == $m - 1 ) {
2798 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2799 } else {
2800 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2801 }
2802 }
2803 return $s;
2804 }
2805 }
2806
2807 /**
2808 * Take a list of strings and build a locale-friendly comma-separated
2809 * list, using the local comma-separator message.
2810 * @param $list array of strings to put in a comma list
2811 * @return string
2812 */
2813 function commaList( $list ) {
2814 return implode(
2815 $list,
2816 wfMsgExt(
2817 'comma-separator',
2818 array( 'parsemag', 'escapenoentities', 'language' => $this )
2819 )
2820 );
2821 }
2822
2823 /**
2824 * Take a list of strings and build a locale-friendly semicolon-separated
2825 * list, using the local semicolon-separator message.
2826 * @param $list array of strings to put in a semicolon list
2827 * @return string
2828 */
2829 function semicolonList( $list ) {
2830 return implode(
2831 $list,
2832 wfMsgExt(
2833 'semicolon-separator',
2834 array( 'parsemag', 'escapenoentities', 'language' => $this )
2835 )
2836 );
2837 }
2838
2839 /**
2840 * Same as commaList, but separate it with the pipe instead.
2841 * @param $list array of strings to put in a pipe list
2842 * @return string
2843 */
2844 function pipeList( $list ) {
2845 return implode(
2846 $list,
2847 wfMsgExt(
2848 'pipe-separator',
2849 array( 'escapenoentities', 'language' => $this )
2850 )
2851 );
2852 }
2853
2854 /**
2855 * Truncate a string to a specified length in bytes, appending an optional
2856 * string (e.g. for ellipses)
2857 *
2858 * The database offers limited byte lengths for some columns in the database;
2859 * multi-byte character sets mean we need to ensure that only whole characters
2860 * are included, otherwise broken characters can be passed to the user
2861 *
2862 * If $length is negative, the string will be truncated from the beginning
2863 *
2864 * @param $string String to truncate
2865 * @param $length Int: maximum length (including ellipses)
2866 * @param $ellipsis String to append to the truncated text
2867 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
2868 * $adjustLength was introduced in 1.18, before that behaved as if false.
2869 * @return string
2870 */
2871 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
2872 # Use the localized ellipsis character
2873 if ( $ellipsis == '...' ) {
2874 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2875 }
2876 # Check if there is no need to truncate
2877 if ( $length == 0 ) {
2878 return $ellipsis; // convention
2879 } elseif ( strlen( $string ) <= abs( $length ) ) {
2880 return $string; // no need to truncate
2881 }
2882 $stringOriginal = $string;
2883 # If ellipsis length is >= $length then we can't apply $adjustLength
2884 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
2885 $string = $ellipsis; // this can be slightly unexpected
2886 # Otherwise, truncate and add ellipsis...
2887 } else {
2888 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
2889 if ( $length > 0 ) {
2890 $length -= $eLength;
2891 $string = substr( $string, 0, $length ); // xyz...
2892 $string = $this->removeBadCharLast( $string );
2893 $string = $string . $ellipsis;
2894 } else {
2895 $length += $eLength;
2896 $string = substr( $string, $length ); // ...xyz
2897 $string = $this->removeBadCharFirst( $string );
2898 $string = $ellipsis . $string;
2899 }
2900 }
2901 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
2902 # This check is *not* redundant if $adjustLength, due to the single case where
2903 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
2904 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
2905 return $string;
2906 } else {
2907 return $stringOriginal;
2908 }
2909 }
2910
2911 /**
2912 * Remove bytes that represent an incomplete Unicode character
2913 * at the end of string (e.g. bytes of the char are missing)
2914 *
2915 * @param $string String
2916 * @return string
2917 */
2918 protected function removeBadCharLast( $string ) {
2919 if ( $string != '' ) {
2920 $char = ord( $string[strlen( $string ) - 1] );
2921 $m = array();
2922 if ( $char >= 0xc0 ) {
2923 # We got the first byte only of a multibyte char; remove it.
2924 $string = substr( $string, 0, -1 );
2925 } elseif ( $char >= 0x80 &&
2926 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2927 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
2928 {
2929 # We chopped in the middle of a character; remove it
2930 $string = $m[1];
2931 }
2932 }
2933 return $string;
2934 }
2935
2936 /**
2937 * Remove bytes that represent an incomplete Unicode character
2938 * at the start of string (e.g. bytes of the char are missing)
2939 *
2940 * @param $string String
2941 * @return string
2942 */
2943 protected function removeBadCharFirst( $string ) {
2944 if ( $string != '' ) {
2945 $char = ord( $string[0] );
2946 if ( $char >= 0x80 && $char < 0xc0 ) {
2947 # We chopped in the middle of a character; remove the whole thing
2948 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2949 }
2950 }
2951 return $string;
2952 }
2953
2954 /**
2955 * Truncate a string of valid HTML to a specified length in bytes,
2956 * appending an optional string (e.g. for ellipses), and return valid HTML
2957 *
2958 * This is only intended for styled/linked text, such as HTML with
2959 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
2960 * Also, this will not detect things like "display:none" CSS.
2961 *
2962 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
2963 *
2964 * @param string $text HTML string to truncate
2965 * @param int $length (zero/positive) Maximum length (including ellipses)
2966 * @param string $ellipsis String to append to the truncated text
2967 * @return string
2968 */
2969 function truncateHtml( $text, $length, $ellipsis = '...' ) {
2970 # Use the localized ellipsis character
2971 if ( $ellipsis == '...' ) {
2972 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2973 }
2974 # Check if there is clearly no need to truncate
2975 if ( $length <= 0 ) {
2976 return $ellipsis; // no text shown, nothing to format (convention)
2977 } elseif ( strlen( $text ) <= $length ) {
2978 return $text; // string short enough even *with* HTML (short-circuit)
2979 }
2980
2981 $dispLen = 0; // innerHTML legth so far
2982 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
2983 $tagType = 0; // 0-open, 1-close
2984 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
2985 $entityState = 0; // 0-not entity, 1-entity
2986 $tag = $ret = ''; // accumulated tag name, accumulated result string
2987 $openTags = array(); // open tag stack
2988 $maybeState = null; // possible truncation state
2989
2990 $textLen = strlen( $text );
2991 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
2992 for ( $pos = 0; true; ++$pos ) {
2993 # Consider truncation once the display length has reached the maximim.
2994 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
2995 # Check that we're not in the middle of a bracket/entity...
2996 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
2997 if ( !$testingEllipsis ) {
2998 $testingEllipsis = true;
2999 # Save where we are; we will truncate here unless there turn out to
3000 # be so few remaining characters that truncation is not necessary.
3001 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3002 $maybeState = array( $ret, $openTags ); // save state
3003 }
3004 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3005 # String in fact does need truncation, the truncation point was OK.
3006 list( $ret, $openTags ) = $maybeState; // reload state
3007 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3008 $ret .= $ellipsis; // add ellipsis
3009 break;
3010 }
3011 }
3012 if ( $pos >= $textLen ) break; // extra iteration just for above checks
3013
3014 # Read the next char...
3015 $ch = $text[$pos];
3016 $lastCh = $pos ? $text[$pos - 1] : '';
3017 $ret .= $ch; // add to result string
3018 if ( $ch == '<' ) {
3019 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3020 $entityState = 0; // for bad HTML
3021 $bracketState = 1; // tag started (checking for backslash)
3022 } elseif ( $ch == '>' ) {
3023 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3024 $entityState = 0; // for bad HTML
3025 $bracketState = 0; // out of brackets
3026 } elseif ( $bracketState == 1 ) {
3027 if ( $ch == '/' ) {
3028 $tagType = 1; // close tag (e.g. "</span>")
3029 } else {
3030 $tagType = 0; // open tag (e.g. "<span>")
3031 $tag .= $ch;
3032 }
3033 $bracketState = 2; // building tag name
3034 } elseif ( $bracketState == 2 ) {
3035 if ( $ch != ' ' ) {
3036 $tag .= $ch;
3037 } else {
3038 // Name found (e.g. "<a href=..."), add on tag attributes...
3039 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3040 }
3041 } elseif ( $bracketState == 0 ) {
3042 if ( $entityState ) {
3043 if ( $ch == ';' ) {
3044 $entityState = 0;
3045 $dispLen++; // entity is one displayed char
3046 }
3047 } else {
3048 if ( $neLength == 0 && !$maybeState ) {
3049 // Save state without $ch. We want to *hit* the first
3050 // display char (to get tags) but not *use* it if truncating.
3051 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3052 }
3053 if ( $ch == '&' ) {
3054 $entityState = 1; // entity found, (e.g. "&#160;")
3055 } else {
3056 $dispLen++; // this char is displayed
3057 // Add the next $max display text chars after this in one swoop...
3058 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3059 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3060 $dispLen += $skipped;
3061 $pos += $skipped;
3062 }
3063 }
3064 }
3065 }
3066 // Close the last tag if left unclosed by bad HTML
3067 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3068 while ( count( $openTags ) > 0 ) {
3069 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3070 }
3071 return $ret;
3072 }
3073
3074 /**
3075 * truncateHtml() helper function
3076 * like strcspn() but adds the skipped chars to $ret
3077 *
3078 * @param $ret
3079 * @param $text
3080 * @param $search
3081 * @param $start
3082 * @param $len
3083 * @return int
3084 */
3085 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3086 if ( $len === null ) {
3087 $len = -1; // -1 means "no limit" for strcspn
3088 } elseif ( $len < 0 ) {
3089 $len = 0; // sanity
3090 }
3091 $skipCount = 0;
3092 if ( $start < strlen( $text ) ) {
3093 $skipCount = strcspn( $text, $search, $start, $len );
3094 $ret .= substr( $text, $start, $skipCount );
3095 }
3096 return $skipCount;
3097 }
3098
3099 /**
3100 * truncateHtml() helper function
3101 * (a) push or pop $tag from $openTags as needed
3102 * (b) clear $tag value
3103 * @param String &$tag Current HTML tag name we are looking at
3104 * @param int $tagType (0-open tag, 1-close tag)
3105 * @param char $lastCh Character before the '>' that ended this tag
3106 * @param array &$openTags Open tag stack (not accounting for $tag)
3107 */
3108 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3109 $tag = ltrim( $tag );
3110 if ( $tag != '' ) {
3111 if ( $tagType == 0 && $lastCh != '/' ) {
3112 $openTags[] = $tag; // tag opened (didn't close itself)
3113 } elseif ( $tagType == 1 ) {
3114 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3115 array_pop( $openTags ); // tag closed
3116 }
3117 }
3118 $tag = '';
3119 }
3120 }
3121
3122 /**
3123 * Grammatical transformations, needed for inflected languages
3124 * Invoked by putting {{grammar:case|word}} in a message
3125 *
3126 * @param $word string
3127 * @param $case string
3128 * @return string
3129 */
3130 function convertGrammar( $word, $case ) {
3131 global $wgGrammarForms;
3132 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3133 return $wgGrammarForms[$this->getCode()][$case][$word];
3134 }
3135 return $word;
3136 }
3137
3138 /**
3139 * Provides an alternative text depending on specified gender.
3140 * Usage {{gender:username|masculine|feminine|neutral}}.
3141 * username is optional, in which case the gender of current user is used,
3142 * but only in (some) interface messages; otherwise default gender is used.
3143 * If second or third parameter are not specified, masculine is used.
3144 * These details may be overriden per language.
3145 *
3146 * @param $gender string
3147 * @param $forms array
3148 *
3149 * @return string
3150 */
3151 function gender( $gender, $forms ) {
3152 if ( !count( $forms ) ) {
3153 return '';
3154 }
3155 $forms = $this->preConvertPlural( $forms, 2 );
3156 if ( $gender === 'male' ) {
3157 return $forms[0];
3158 }
3159 if ( $gender === 'female' ) {
3160 return $forms[1];
3161 }
3162 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3163 }
3164
3165 /**
3166 * Plural form transformations, needed for some languages.
3167 * For example, there are 3 form of plural in Russian and Polish,
3168 * depending on "count mod 10". See [[w:Plural]]
3169 * For English it is pretty simple.
3170 *
3171 * Invoked by putting {{plural:count|wordform1|wordform2}}
3172 * or {{plural:count|wordform1|wordform2|wordform3}}
3173 *
3174 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3175 *
3176 * @param $count Integer: non-localized number
3177 * @param $forms Array: different plural forms
3178 * @return string Correct form of plural for $count in this language
3179 */
3180 function convertPlural( $count, $forms ) {
3181 if ( !count( $forms ) ) {
3182 return '';
3183 }
3184 $forms = $this->preConvertPlural( $forms, 2 );
3185
3186 return ( $count == 1 ) ? $forms[0] : $forms[1];
3187 }
3188
3189 /**
3190 * Checks that convertPlural was given an array and pads it to requested
3191 * amount of forms by copying the last one.
3192 *
3193 * @param $count Integer: How many forms should there be at least
3194 * @param $forms Array of forms given to convertPlural
3195 * @return array Padded array of forms or an exception if not an array
3196 */
3197 protected function preConvertPlural( /* Array */ $forms, $count ) {
3198 while ( count( $forms ) < $count ) {
3199 $forms[] = $forms[count( $forms ) - 1];
3200 }
3201 return $forms;
3202 }
3203
3204 /**
3205 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3206 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3207 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3208 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3209 * match up with it.
3210 *
3211 * @param $str String: the validated block duration in English
3212 * @return Somehow translated block duration
3213 * @see LanguageFi.php for example implementation
3214 */
3215 function translateBlockExpiry( $str ) {
3216 $duration = SpecialBlock::getSuggestedDurations( $this );
3217 foreach ( $duration as $show => $value ) {
3218 if ( strcmp( $str, $value ) == 0 ) {
3219 return htmlspecialchars( trim( $show ) );
3220 }
3221 }
3222
3223 // Since usually only infinite or indefinite is only on list, so try
3224 // equivalents if still here.
3225 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3226 if ( in_array( $str, $indefs ) ) {
3227 foreach ( $indefs as $val ) {
3228 $show = array_search( $val, $duration, true );
3229 if ( $show !== false ) {
3230 return htmlspecialchars( trim( $show ) );
3231 }
3232 }
3233 }
3234 // If all else fails, return the original string.
3235 return $str;
3236 }
3237
3238 /**
3239 * languages like Chinese need to be segmented in order for the diff
3240 * to be of any use
3241 *
3242 * @param $text String
3243 * @return String
3244 */
3245 function segmentForDiff( $text ) {
3246 return $text;
3247 }
3248
3249 /**
3250 * and unsegment to show the result
3251 *
3252 * @param $text String
3253 * @return String
3254 */
3255 function unsegmentForDiff( $text ) {
3256 return $text;
3257 }
3258
3259 /**
3260 * convert text to all supported variants
3261 *
3262 * @param $text string
3263 * @return array
3264 */
3265 function autoConvertToAllVariants( $text ) {
3266 return $this->mConverter->autoConvertToAllVariants( $text );
3267 }
3268
3269 /**
3270 * convert text to different variants of a language.
3271 *
3272 * @param $text string
3273 * @return string
3274 */
3275 function convert( $text ) {
3276 return $this->mConverter->convert( $text );
3277 }
3278
3279
3280 /**
3281 * Convert a Title object to a string in the preferred variant
3282 *
3283 * @param $title Title
3284 * @return string
3285 */
3286 function convertTitle( $title ) {
3287 return $this->mConverter->convertTitle( $title );
3288 }
3289
3290 /**
3291 * Check if this is a language with variants
3292 *
3293 * @return bool
3294 */
3295 function hasVariants() {
3296 return sizeof( $this->getVariants() ) > 1;
3297 }
3298
3299 /**
3300 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3301 *
3302 * @param $text string
3303 * @return string
3304 */
3305 function armourMath( $text ) {
3306 return $this->mConverter->armourMath( $text );
3307 }
3308
3309 /**
3310 * Perform output conversion on a string, and encode for safe HTML output.
3311 * @param $text String text to be converted
3312 * @param $isTitle Bool whether this conversion is for the article title
3313 * @return string
3314 * @todo this should get integrated somewhere sane
3315 */
3316 function convertHtml( $text, $isTitle = false ) {
3317 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3318 }
3319
3320 /**
3321 * @param $key string
3322 * @return string
3323 */
3324 function convertCategoryKey( $key ) {
3325 return $this->mConverter->convertCategoryKey( $key );
3326 }
3327
3328 /**
3329 * Get the list of variants supported by this language
3330 * see sample implementation in LanguageZh.php
3331 *
3332 * @return array an array of language codes
3333 */
3334 function getVariants() {
3335 return $this->mConverter->getVariants();
3336 }
3337
3338 /**
3339 * @return string
3340 */
3341 function getPreferredVariant() {
3342 return $this->mConverter->getPreferredVariant();
3343 }
3344
3345 /**
3346 * @return string
3347 */
3348 function getDefaultVariant() {
3349 return $this->mConverter->getDefaultVariant();
3350 }
3351
3352 /**
3353 * @return string
3354 */
3355 function getURLVariant() {
3356 return $this->mConverter->getURLVariant();
3357 }
3358
3359 /**
3360 * If a language supports multiple variants, it is
3361 * possible that non-existing link in one variant
3362 * actually exists in another variant. this function
3363 * tries to find it. See e.g. LanguageZh.php
3364 *
3365 * @param $link String: the name of the link
3366 * @param $nt Mixed: the title object of the link
3367 * @param $ignoreOtherCond Boolean: to disable other conditions when
3368 * we need to transclude a template or update a category's link
3369 * @return null the input parameters may be modified upon return
3370 */
3371 function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3372 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
3373 }
3374
3375 /**
3376 * If a language supports multiple variants, converts text
3377 * into an array of all possible variants of the text:
3378 * 'variant' => text in that variant
3379 *
3380 * @deprecated since 1.17 Use autoConvertToAllVariants()
3381 *
3382 * @param $text string
3383 *
3384 * @return string
3385 */
3386 function convertLinkToAllVariants( $text ) {
3387 return $this->mConverter->convertLinkToAllVariants( $text );
3388 }
3389
3390 /**
3391 * returns language specific options used by User::getPageRenderHash()
3392 * for example, the preferred language variant
3393 *
3394 * @return string
3395 */
3396 function getExtraHashOptions() {
3397 return $this->mConverter->getExtraHashOptions();
3398 }
3399
3400 /**
3401 * For languages that support multiple variants, the title of an
3402 * article may be displayed differently in different variants. this
3403 * function returns the apporiate title defined in the body of the article.
3404 *
3405 * @return string
3406 */
3407 function getParsedTitle() {
3408 return $this->mConverter->getParsedTitle();
3409 }
3410
3411 /**
3412 * Enclose a string with the "no conversion" tag. This is used by
3413 * various functions in the Parser
3414 *
3415 * @param $text String: text to be tagged for no conversion
3416 * @param $noParse bool
3417 * @return string the tagged text
3418 */
3419 function markNoConversion( $text, $noParse = false ) {
3420 return $this->mConverter->markNoConversion( $text, $noParse );
3421 }
3422
3423 /**
3424 * A regular expression to match legal word-trailing characters
3425 * which should be merged onto a link of the form [[foo]]bar.
3426 *
3427 * @return string
3428 */
3429 function linkTrail() {
3430 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
3431 }
3432
3433 /**
3434 * @return Language
3435 */
3436 function getLangObj() {
3437 return $this;
3438 }
3439
3440 /**
3441 * Get the RFC 3066 code for this language object
3442 *
3443 * @return string
3444 */
3445 function getCode() {
3446 return $this->mCode;
3447 }
3448
3449 /**
3450 * @param $code string
3451 */
3452 function setCode( $code ) {
3453 $this->mCode = $code;
3454 }
3455
3456 /**
3457 * Get the name of a file for a certain language code
3458 * @param $prefix string Prepend this to the filename
3459 * @param $code string Language code
3460 * @param $suffix string Append this to the filename
3461 * @return string $prefix . $mangledCode . $suffix
3462 */
3463 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
3464 // Protect against path traversal
3465 if ( !Language::isValidCode( $code )
3466 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
3467 {
3468 throw new MWException( "Invalid language code \"$code\"" );
3469 }
3470
3471 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
3472 }
3473
3474 /**
3475 * Get the language code from a file name. Inverse of getFileName()
3476 * @param $filename string $prefix . $languageCode . $suffix
3477 * @param $prefix string Prefix before the language code
3478 * @param $suffix string Suffix after the language code
3479 * @return string Language code, or false if $prefix or $suffix isn't found
3480 */
3481 static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
3482 $m = null;
3483 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
3484 preg_quote( $suffix, '/' ) . '/', $filename, $m );
3485 if ( !count( $m ) ) {
3486 return false;
3487 }
3488 return str_replace( '_', '-', strtolower( $m[1] ) );
3489 }
3490
3491 /**
3492 * @param $code string
3493 * @return string
3494 */
3495 static function getMessagesFileName( $code ) {
3496 global $IP;
3497 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
3498 }
3499
3500 /**
3501 * @param $code string
3502 * @return string
3503 */
3504 static function getClassFileName( $code ) {
3505 global $IP;
3506 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
3507 }
3508
3509 /**
3510 * Get the first fallback for a given language.
3511 *
3512 * @param $code string
3513 *
3514 * @return false|string
3515 */
3516 static function getFallbackFor( $code ) {
3517 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
3518 return false;
3519 } else {
3520 $fallbacks = self::getFallbacksFor( $code );
3521 $first = array_shift( $fallbacks );
3522 return $first;
3523 }
3524 }
3525
3526 /**
3527 * Get the ordered list of fallback languages.
3528 *
3529 * @since 1.19
3530 * @param $code string Language code
3531 * @return array
3532 */
3533 static function getFallbacksFor( $code ) {
3534 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
3535 return array();
3536 } else {
3537 $v = self::getLocalisationCache()->getItem( $code, 'fallback' );
3538 $v = array_map( 'trim', explode( ',', $v ) );
3539 if ( $v[count( $v ) - 1] !== 'en' ) {
3540 $v[] = 'en';
3541 }
3542 return $v;
3543 }
3544 }
3545
3546 /**
3547 * Get all messages for a given language
3548 * WARNING: this may take a long time. If you just need all message *keys*
3549 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
3550 *
3551 * @param $code string
3552 *
3553 * @return array
3554 */
3555 static function getMessagesFor( $code ) {
3556 return self::getLocalisationCache()->getItem( $code, 'messages' );
3557 }
3558
3559 /**
3560 * Get a message for a given language
3561 *
3562 * @param $key string
3563 * @param $code string
3564 *
3565 * @return string
3566 */
3567 static function getMessageFor( $key, $code ) {
3568 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
3569 }
3570
3571 /**
3572 * Get all message keys for a given language. This is a faster alternative to
3573 * array_keys( Language::getMessagesFor( $code ) )
3574 *
3575 * @since 1.19
3576 * @param $code string Language code
3577 * @return array of message keys (strings)
3578 */
3579 static function getMessageKeysFor( $code ) {
3580 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
3581 }
3582
3583 /**
3584 * @param $talk
3585 * @return mixed
3586 */
3587 function fixVariableInNamespace( $talk ) {
3588 if ( strpos( $talk, '$1' ) === false ) {
3589 return $talk;
3590 }
3591
3592 global $wgMetaNamespace;
3593 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
3594
3595 # Allow grammar transformations
3596 # Allowing full message-style parsing would make simple requests
3597 # such as action=raw much more expensive than they need to be.
3598 # This will hopefully cover most cases.
3599 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
3600 array( &$this, 'replaceGrammarInNamespace' ), $talk );
3601 return str_replace( ' ', '_', $talk );
3602 }
3603
3604 /**
3605 * @param $m string
3606 * @return string
3607 */
3608 function replaceGrammarInNamespace( $m ) {
3609 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
3610 }
3611
3612 /**
3613 * @throws MWException
3614 * @return array
3615 */
3616 static function getCaseMaps() {
3617 static $wikiUpperChars, $wikiLowerChars;
3618 if ( isset( $wikiUpperChars ) ) {
3619 return array( $wikiUpperChars, $wikiLowerChars );
3620 }
3621
3622 wfProfileIn( __METHOD__ );
3623 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
3624 if ( $arr === false ) {
3625 throw new MWException(
3626 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
3627 }
3628 $wikiUpperChars = $arr['wikiUpperChars'];
3629 $wikiLowerChars = $arr['wikiLowerChars'];
3630 wfProfileOut( __METHOD__ );
3631 return array( $wikiUpperChars, $wikiLowerChars );
3632 }
3633
3634 /**
3635 * Decode an expiry (block, protection, etc) which has come from the DB
3636 *
3637 * @param $expiry String: Database expiry String
3638 * @param $format Bool|Int true to process using language functions, or TS_ constant
3639 * to return the expiry in a given timestamp
3640 * @return String
3641 */
3642 public function formatExpiry( $expiry, $format = true ) {
3643 static $infinity, $infinityMsg;
3644 if ( $infinity === null ) {
3645 $infinityMsg = wfMessage( 'infiniteblock' );
3646 $infinity = wfGetDB( DB_SLAVE )->getInfinity();
3647 }
3648
3649 if ( $expiry == '' || $expiry == $infinity ) {
3650 return $format === true
3651 ? $infinityMsg
3652 : $infinity;
3653 } else {
3654 return $format === true
3655 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
3656 : wfTimestamp( $format, $expiry );
3657 }
3658 }
3659
3660 /**
3661 * @todo Document
3662 * @param $seconds int|float
3663 * @param $format Array Optional
3664 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
3665 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
3666 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
3667 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
3668 * @return string
3669 */
3670 function formatTimePeriod( $seconds, $format = array() ) {
3671 if ( !is_array( $format ) ) {
3672 $format = array( 'avoid' => $format ); // For backwards compatibility
3673 }
3674 if ( !isset( $format['avoid'] ) ) {
3675 $format['avoid'] = false;
3676 }
3677 if ( !isset( $format['noabbrevs' ] ) ) {
3678 $format['noabbrevs'] = false;
3679 }
3680 $secondsMsg = wfMessage(
3681 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
3682 $minutesMsg = wfMessage(
3683 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
3684 $hoursMsg = wfMessage(
3685 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
3686 $daysMsg = wfMessage(
3687 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
3688
3689 if ( round( $seconds * 10 ) < 100 ) {
3690 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
3691 $s = $secondsMsg->params( $s )->text();
3692 } elseif ( round( $seconds ) < 60 ) {
3693 $s = $this->formatNum( round( $seconds ) );
3694 $s = $secondsMsg->params( $s )->text();
3695 } elseif ( round( $seconds ) < 3600 ) {
3696 $minutes = floor( $seconds / 60 );
3697 $secondsPart = round( fmod( $seconds, 60 ) );
3698 if ( $secondsPart == 60 ) {
3699 $secondsPart = 0;
3700 $minutes++;
3701 }
3702 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3703 $s .= ' ';
3704 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3705 } elseif ( round( $seconds ) <= 2 * 86400 ) {
3706 $hours = floor( $seconds / 3600 );
3707 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
3708 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
3709 if ( $secondsPart == 60 ) {
3710 $secondsPart = 0;
3711 $minutes++;
3712 }
3713 if ( $minutes == 60 ) {
3714 $minutes = 0;
3715 $hours++;
3716 }
3717 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
3718 $s .= ' ';
3719 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3720 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
3721 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3722 }
3723 } else {
3724 $days = floor( $seconds / 86400 );
3725 if ( $format['avoid'] === 'avoidminutes' ) {
3726 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
3727 if ( $hours == 24 ) {
3728 $hours = 0;
3729 $days++;
3730 }
3731 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3732 $s .= ' ';
3733 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
3734 } elseif ( $format['avoid'] === 'avoidseconds' ) {
3735 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
3736 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
3737 if ( $minutes == 60 ) {
3738 $minutes = 0;
3739 $hours++;
3740 }
3741 if ( $hours == 24 ) {
3742 $hours = 0;
3743 $days++;
3744 }
3745 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3746 $s .= ' ';
3747 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
3748 $s .= ' ';
3749 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3750 } else {
3751 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3752 $s .= ' ';
3753 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
3754 }
3755 }
3756 return $s;
3757 }
3758
3759 /**
3760 * @param $bps int
3761 * @return string
3762 */
3763 function formatBitrate( $bps ) {
3764 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
3765 if ( $bps <= 0 ) {
3766 return $this->formatNum( $bps ) . $units[0];
3767 }
3768 $unitIndex = floor( log10( $bps ) / 3 );
3769 $mantissa = $bps / pow( 1000, $unitIndex );
3770 if ( $mantissa < 10 ) {
3771 $mantissa = round( $mantissa, 1 );
3772 } else {
3773 $mantissa = round( $mantissa );
3774 }
3775 return $this->formatNum( $mantissa ) . $units[$unitIndex];
3776 }
3777
3778 /**
3779 * Format a size in bytes for output, using an appropriate
3780 * unit (B, KB, MB or GB) according to the magnitude in question
3781 *
3782 * @param $size int Size to format
3783 * @return string Plain text (not HTML)
3784 */
3785 function formatSize( $size ) {
3786 // For small sizes no decimal places necessary
3787 $round = 0;
3788 if ( $size > 1024 ) {
3789 $size = $size / 1024;
3790 if ( $size > 1024 ) {
3791 $size = $size / 1024;
3792 // For MB and bigger two decimal places are smarter
3793 $round = 2;
3794 if ( $size > 1024 ) {
3795 $size = $size / 1024;
3796 $msg = 'size-gigabytes';
3797 } else {
3798 $msg = 'size-megabytes';
3799 }
3800 } else {
3801 $msg = 'size-kilobytes';
3802 }
3803 } else {
3804 $msg = 'size-bytes';
3805 }
3806 $size = round( $size, $round );
3807 $text = $this->getMessageFromDB( $msg );
3808 return str_replace( '$1', $this->formatNum( $size ), $text );
3809 }
3810
3811 /**
3812 * Make a list item, used by various special pages
3813 *
3814 * @param $page String Page link
3815 * @param $details String Text between brackets
3816 * @param $oppositedm Boolean Add the direction mark opposite to your
3817 * language, to display text properly
3818 * @return String
3819 */
3820 function specialList( $page, $details, $oppositedm = true ) {
3821 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) .
3822 $this->getDirMark();
3823 $details = $details ? $dirmark . $this->getMessageFromDB( 'word-separator' ) .
3824 wfMsgExt( 'parentheses', array( 'escape', 'replaceafter', 'language' => $this ), $details ) : '';
3825 return $page . $details;
3826 }
3827
3828 /**
3829 * Generate (prev x| next x) (20|50|100...) type links for paging
3830 *
3831 * @param $title Title object to link
3832 * @param $offset Integer offset parameter
3833 * @param $limit Integer limit parameter
3834 * @param $query String optional URL query parameter string
3835 * @param $atend Bool optional param for specified if this is the last page
3836 * @return String
3837 */
3838 public function viewPrevNext( Title $title, $offset, $limit, array $query = array(), $atend = false ) {
3839 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
3840
3841 # Make 'previous' link
3842 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
3843 if( $offset > 0 ) {
3844 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
3845 $query, $prev, 'prevn-title', 'mw-prevlink' );
3846 } else {
3847 $plink = htmlspecialchars( $prev );
3848 }
3849
3850 # Make 'next' link
3851 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
3852 if( $atend ) {
3853 $nlink = htmlspecialchars( $next );
3854 } else {
3855 $nlink = $this->numLink( $title, $offset + $limit, $limit,
3856 $query, $next, 'prevn-title', 'mw-nextlink' );
3857 }
3858
3859 # Make links to set number of items per page
3860 $numLinks = array();
3861 foreach( array( 20, 50, 100, 250, 500 ) as $num ) {
3862 $numLinks[] = $this->numLink( $title, $offset, $num,
3863 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
3864 }
3865
3866 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
3867 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
3868 }
3869
3870 /**
3871 * Helper function for viewPrevNext() that generates links
3872 *
3873 * @param $title Title object to link
3874 * @param $offset Integer offset parameter
3875 * @param $limit Integer limit parameter
3876 * @param $query Array extra query parameters
3877 * @param $link String text to use for the link; will be escaped
3878 * @param $tooltipMsg String name of the message to use as tooltip
3879 * @param $class String value of the "class" attribute of the link
3880 * @return String HTML fragment
3881 */
3882 private function numLink( Title $title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
3883 $query = array( 'limit' => $limit, 'offset' => $offset ) + $query;
3884 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
3885 return Html::element( 'a', array( 'href' => $title->getLocalURL( $query ),
3886 'title' => $tooltip, 'class' => $class ), $link );
3887 }
3888
3889 /**
3890 * Get the conversion rule title, if any.
3891 *
3892 * @return string
3893 */
3894 function getConvRuleTitle() {
3895 return $this->mConverter->getConvRuleTitle();
3896 }
3897 }