Followup r114067 rename to formatDuration to match other methods
[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
33 /**
34 * @var Language
35 */
36 var $mLang;
37 function __construct( $langobj ) { $this->mLang = $langobj; }
38 function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
39 function convert( $t ) { return $t; }
40 function convertTo( $text, $variant ) { return $text; }
41 function convertTitle( $t ) { return $t->getPrefixedText(); }
42 function getVariants() { return array( $this->mLang->getCode() ); }
43 function getPreferredVariant() { return $this->mLang->getCode(); }
44 function getDefaultVariant() { return $this->mLang->getCode(); }
45 function getURLVariant() { return ''; }
46 function getConvRuleTitle() { return false; }
47 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
48 function getExtraHashOptions() { return ''; }
49 function getParsedTitle() { return ''; }
50 function markNoConversion( $text, $noParse = false ) { return $text; }
51 function convertCategoryKey( $key ) { return $key; }
52 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
53 function armourMath( $text ) { return $text; }
54 }
55
56 /**
57 * Internationalisation code
58 * @ingroup Language
59 */
60 class Language {
61
62 /**
63 * @var LanguageConverter
64 */
65 var $mConverter;
66
67 var $mVariants, $mCode, $mLoaded = false;
68 var $mMagicExtensions = array(), $mMagicHookDone = false;
69 private $mHtmlCode = null;
70
71 var $dateFormatStrings = array();
72 var $mExtendedSpecialPageAliases;
73
74 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
75
76 /**
77 * ReplacementArray object caches
78 */
79 var $transformData = array();
80
81 /**
82 * @var LocalisationCache
83 */
84 static public $dataCache;
85
86 static public $mLangObjCache = array();
87
88 static public $mWeekdayMsgs = array(
89 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
90 'friday', 'saturday'
91 );
92
93 static public $mWeekdayAbbrevMsgs = array(
94 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
95 );
96
97 static public $mMonthMsgs = array(
98 'january', 'february', 'march', 'april', 'may_long', 'june',
99 'july', 'august', 'september', 'october', 'november',
100 'december'
101 );
102 static public $mMonthGenMsgs = array(
103 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
104 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
105 'december-gen'
106 );
107 static public $mMonthAbbrevMsgs = array(
108 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
109 'sep', 'oct', 'nov', 'dec'
110 );
111
112 static public $mIranianCalendarMonthMsgs = array(
113 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
114 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
115 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
116 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
117 );
118
119 static public $mHebrewCalendarMonthMsgs = array(
120 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
121 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
122 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
123 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
124 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
125 );
126
127 static public $mHebrewCalendarMonthGenMsgs = array(
128 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
129 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
130 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
131 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
132 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
133 );
134
135 static public $mHijriCalendarMonthMsgs = array(
136 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
137 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
138 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
139 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
140 );
141
142 /**
143 * Get a cached language object for a given language code
144 * @param $code String
145 * @return Language
146 */
147 static function factory( $code ) {
148 if ( !isset( self::$mLangObjCache[$code] ) ) {
149 if ( count( self::$mLangObjCache ) > 10 ) {
150 // Don't keep a billion objects around, that's stupid.
151 self::$mLangObjCache = array();
152 }
153 self::$mLangObjCache[$code] = self::newFromCode( $code );
154 }
155 return self::$mLangObjCache[$code];
156 }
157
158 /**
159 * Create a language object for a given language code
160 * @param $code String
161 * @throws MWException
162 * @return Language
163 */
164 protected static function newFromCode( $code ) {
165 // Protect against path traversal below
166 if ( !Language::isValidCode( $code )
167 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
168 {
169 throw new MWException( "Invalid language code \"$code\"" );
170 }
171
172 if ( !Language::isValidBuiltInCode( $code ) ) {
173 // It's not possible to customise this code with class files, so
174 // just return a Language object. This is to support uselang= hacks.
175 $lang = new Language;
176 $lang->setCode( $code );
177 return $lang;
178 }
179
180 // Check if there is a language class for the code
181 $class = self::classFromCode( $code );
182 self::preloadLanguageClass( $class );
183 if ( MWInit::classExists( $class ) ) {
184 $lang = new $class;
185 return $lang;
186 }
187
188 // Keep trying the fallback list until we find an existing class
189 $fallbacks = Language::getFallbacksFor( $code );
190 foreach ( $fallbacks as $fallbackCode ) {
191 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
192 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
193 }
194
195 $class = self::classFromCode( $fallbackCode );
196 self::preloadLanguageClass( $class );
197 if ( MWInit::classExists( $class ) ) {
198 $lang = Language::newFromCode( $fallbackCode );
199 $lang->setCode( $code );
200 return $lang;
201 }
202 }
203
204 throw new MWException( "Invalid fallback sequence for language '$code'" );
205 }
206
207 /**
208 * Returns true if a language code string is of a valid form, whether or
209 * not it exists. This includes codes which are used solely for
210 * customisation via the MediaWiki namespace.
211 *
212 * @param $code string
213 *
214 * @return bool
215 */
216 public static function isValidCode( $code ) {
217 return
218 strcspn( $code, ":/\\\000" ) === strlen( $code )
219 && !preg_match( Title::getTitleInvalidRegex(), $code );
220 }
221
222 /**
223 * Returns true if a language code is of a valid form for the purposes of
224 * internal customisation of MediaWiki, via Messages*.php.
225 *
226 * @param $code string
227 *
228 * @since 1.18
229 * @return bool
230 */
231 public static function isValidBuiltInCode( $code ) {
232 return preg_match( '/^[a-z0-9-]+$/i', $code );
233 }
234
235 /**
236 * @param $code
237 * @return String Name of the language class
238 */
239 public static function classFromCode( $code ) {
240 if ( $code == 'en' ) {
241 return 'Language';
242 } else {
243 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
244 }
245 }
246
247 /**
248 * Includes language class files
249 *
250 * @param $class string Name of the language class
251 */
252 public static function preloadLanguageClass( $class ) {
253 global $IP;
254
255 if ( $class === 'Language' ) {
256 return;
257 }
258
259 if ( !defined( 'MW_COMPILED' ) ) {
260 // Preload base classes to work around APC/PHP5 bug
261 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
262 include_once( "$IP/languages/classes/$class.deps.php" );
263 }
264 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
265 include_once( "$IP/languages/classes/$class.php" );
266 }
267 }
268 }
269
270 /**
271 * Get the LocalisationCache instance
272 *
273 * @return LocalisationCache
274 */
275 public static function getLocalisationCache() {
276 if ( is_null( self::$dataCache ) ) {
277 global $wgLocalisationCacheConf;
278 $class = $wgLocalisationCacheConf['class'];
279 self::$dataCache = new $class( $wgLocalisationCacheConf );
280 }
281 return self::$dataCache;
282 }
283
284 function __construct() {
285 $this->mConverter = new FakeConverter( $this );
286 // Set the code to the name of the descendant
287 if ( get_class( $this ) == 'Language' ) {
288 $this->mCode = 'en';
289 } else {
290 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
291 }
292 self::getLocalisationCache();
293 }
294
295 /**
296 * Reduce memory usage
297 */
298 function __destruct() {
299 foreach ( $this as $name => $value ) {
300 unset( $this->$name );
301 }
302 }
303
304 /**
305 * Hook which will be called if this is the content language.
306 * Descendants can use this to register hook functions or modify globals
307 */
308 function initContLang() { }
309
310 /**
311 * Same as getFallbacksFor for current language.
312 * @return array|bool
313 * @deprecated in 1.19
314 */
315 function getFallbackLanguageCode() {
316 wfDeprecated( __METHOD__ );
317 return self::getFallbackFor( $this->mCode );
318 }
319
320 /**
321 * @return array
322 * @since 1.19
323 */
324 function getFallbackLanguages() {
325 return self::getFallbacksFor( $this->mCode );
326 }
327
328 /**
329 * Exports $wgBookstoreListEn
330 * @return array
331 */
332 function getBookstoreList() {
333 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
334 }
335
336 /**
337 * @return array
338 */
339 public function getNamespaces() {
340 if ( is_null( $this->namespaceNames ) ) {
341 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
342
343 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
344 $validNamespaces = MWNamespace::getCanonicalNamespaces();
345
346 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
347
348 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
349 if ( $wgMetaNamespaceTalk ) {
350 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
351 } else {
352 $talk = $this->namespaceNames[NS_PROJECT_TALK];
353 $this->namespaceNames[NS_PROJECT_TALK] =
354 $this->fixVariableInNamespace( $talk );
355 }
356
357 # Sometimes a language will be localised but not actually exist on this wiki.
358 foreach ( $this->namespaceNames as $key => $text ) {
359 if ( !isset( $validNamespaces[$key] ) ) {
360 unset( $this->namespaceNames[$key] );
361 }
362 }
363
364 # The above mixing may leave namespaces out of canonical order.
365 # Re-order by namespace ID number...
366 ksort( $this->namespaceNames );
367
368 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames ) );
369 }
370 return $this->namespaceNames;
371 }
372
373 /**
374 * Arbitrarily set all of the namespace names at once. Mainly used for testing
375 * @param $namespaces Array of namespaces (id => name)
376 */
377 public function setNamespaces( array $namespaces ) {
378 $this->namespaceNames = $namespaces;
379 }
380
381 /**
382 * A convenience function that returns the same thing as
383 * getNamespaces() except with the array values changed to ' '
384 * where it found '_', useful for producing output to be displayed
385 * e.g. in <select> forms.
386 *
387 * @return array
388 */
389 function getFormattedNamespaces() {
390 $ns = $this->getNamespaces();
391 foreach ( $ns as $k => $v ) {
392 $ns[$k] = strtr( $v, '_', ' ' );
393 }
394 return $ns;
395 }
396
397 /**
398 * Get a namespace value by key
399 * <code>
400 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
401 * echo $mw_ns; // prints 'MediaWiki'
402 * </code>
403 *
404 * @param $index Int: the array key of the namespace to return
405 * @return mixed, string if the namespace value exists, otherwise false
406 */
407 function getNsText( $index ) {
408 $ns = $this->getNamespaces();
409 return isset( $ns[$index] ) ? $ns[$index] : false;
410 }
411
412 /**
413 * A convenience function that returns the same thing as
414 * getNsText() except with '_' changed to ' ', useful for
415 * producing output.
416 *
417 * @param $index string
418 *
419 * @return array
420 */
421 function getFormattedNsText( $index ) {
422 $ns = $this->getNsText( $index );
423 return strtr( $ns, '_', ' ' );
424 }
425
426 /**
427 * Returns gender-dependent namespace alias if available.
428 * @param $index Int: namespace index
429 * @param $gender String: gender key (male, female... )
430 * @return String
431 * @since 1.18
432 */
433 function getGenderNsText( $index, $gender ) {
434 global $wgExtraGenderNamespaces;
435
436 $ns = $wgExtraGenderNamespaces + self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
437 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
438 }
439
440 /**
441 * Whether this language makes distinguishes genders for example in
442 * namespaces.
443 * @return bool
444 * @since 1.18
445 */
446 function needsGenderDistinction() {
447 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
448 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
449 // $wgExtraGenderNamespaces overrides everything
450 return true;
451 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
452 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
453 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
454 return false;
455 } else {
456 // Check what is in i18n files
457 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
458 return count( $aliases ) > 0;
459 }
460 }
461
462 /**
463 * Get a namespace key by value, case insensitive.
464 * Only matches namespace names for the current language, not the
465 * canonical ones defined in Namespace.php.
466 *
467 * @param $text String
468 * @return mixed An integer if $text is a valid value otherwise false
469 */
470 function getLocalNsIndex( $text ) {
471 $lctext = $this->lc( $text );
472 $ids = $this->getNamespaceIds();
473 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
474 }
475
476 /**
477 * @return array
478 */
479 function getNamespaceAliases() {
480 if ( is_null( $this->namespaceAliases ) ) {
481 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
482 if ( !$aliases ) {
483 $aliases = array();
484 } else {
485 foreach ( $aliases as $name => $index ) {
486 if ( $index === NS_PROJECT_TALK ) {
487 unset( $aliases[$name] );
488 $name = $this->fixVariableInNamespace( $name );
489 $aliases[$name] = $index;
490 }
491 }
492 }
493
494 global $wgExtraGenderNamespaces;
495 $genders = $wgExtraGenderNamespaces + (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
496 foreach ( $genders as $index => $forms ) {
497 foreach ( $forms as $alias ) {
498 $aliases[$alias] = $index;
499 }
500 }
501
502 $this->namespaceAliases = $aliases;
503 }
504 return $this->namespaceAliases;
505 }
506
507 /**
508 * @return array
509 */
510 function getNamespaceIds() {
511 if ( is_null( $this->mNamespaceIds ) ) {
512 global $wgNamespaceAliases;
513 # Put namespace names and aliases into a hashtable.
514 # If this is too slow, then we should arrange it so that it is done
515 # before caching. The catch is that at pre-cache time, the above
516 # class-specific fixup hasn't been done.
517 $this->mNamespaceIds = array();
518 foreach ( $this->getNamespaces() as $index => $name ) {
519 $this->mNamespaceIds[$this->lc( $name )] = $index;
520 }
521 foreach ( $this->getNamespaceAliases() as $name => $index ) {
522 $this->mNamespaceIds[$this->lc( $name )] = $index;
523 }
524 if ( $wgNamespaceAliases ) {
525 foreach ( $wgNamespaceAliases as $name => $index ) {
526 $this->mNamespaceIds[$this->lc( $name )] = $index;
527 }
528 }
529 }
530 return $this->mNamespaceIds;
531 }
532
533 /**
534 * Get a namespace key by value, case insensitive. Canonical namespace
535 * names override custom ones defined for the current language.
536 *
537 * @param $text String
538 * @return mixed An integer if $text is a valid value otherwise false
539 */
540 function getNsIndex( $text ) {
541 $lctext = $this->lc( $text );
542 $ns = MWNamespace::getCanonicalIndex( $lctext );
543 if ( $ns !== null ) {
544 return $ns;
545 }
546 $ids = $this->getNamespaceIds();
547 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
548 }
549
550 /**
551 * short names for language variants used for language conversion links.
552 *
553 * @param $code String
554 * @param $usemsg bool Use the "variantname-xyz" message if it exists
555 * @return string
556 */
557 function getVariantname( $code, $usemsg = true ) {
558 $msg = "variantname-$code";
559 list( $rootCode ) = explode( '-', $code );
560 if ( $usemsg && wfMessage( $msg )->exists() ) {
561 return $this->getMessageFromDB( $msg );
562 }
563 $name = self::fetchLanguageName( $code );
564 if ( $name ) {
565 return $name; # if it's defined as a language name, show that
566 } else {
567 # otherwise, output the language code
568 return $code;
569 }
570 }
571
572 /**
573 * @param $name string
574 * @return string
575 */
576 function specialPage( $name ) {
577 $aliases = $this->getSpecialPageAliases();
578 if ( isset( $aliases[$name][0] ) ) {
579 $name = $aliases[$name][0];
580 }
581 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
582 }
583
584 /**
585 * @return array
586 */
587 function getQuickbarSettings() {
588 return array(
589 $this->getMessage( 'qbsettings-none' ),
590 $this->getMessage( 'qbsettings-fixedleft' ),
591 $this->getMessage( 'qbsettings-fixedright' ),
592 $this->getMessage( 'qbsettings-floatingleft' ),
593 $this->getMessage( 'qbsettings-floatingright' ),
594 $this->getMessage( 'qbsettings-directionality' )
595 );
596 }
597
598 /**
599 * @return array
600 */
601 function getDatePreferences() {
602 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
603 }
604
605 /**
606 * @return array
607 */
608 function getDateFormats() {
609 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
610 }
611
612 /**
613 * @return array|string
614 */
615 function getDefaultDateFormat() {
616 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
617 if ( $df === 'dmy or mdy' ) {
618 global $wgAmericanDates;
619 return $wgAmericanDates ? 'mdy' : 'dmy';
620 } else {
621 return $df;
622 }
623 }
624
625 /**
626 * @return array
627 */
628 function getDatePreferenceMigrationMap() {
629 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
630 }
631
632 /**
633 * @param $image
634 * @return array|null
635 */
636 function getImageFile( $image ) {
637 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
638 }
639
640 /**
641 * @return array
642 */
643 function getExtraUserToggles() {
644 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
645 }
646
647 /**
648 * @param $tog
649 * @return string
650 */
651 function getUserToggle( $tog ) {
652 return $this->getMessageFromDB( "tog-$tog" );
653 }
654
655 /**
656 * Get native language names, indexed by code.
657 * Only those defined in MediaWiki, no other data like CLDR.
658 * If $customisedOnly is true, only returns codes with a messages file
659 *
660 * @param $customisedOnly bool
661 *
662 * @return array
663 * @deprecated in 1.20, use fetchLanguageNames()
664 */
665 public static function getLanguageNames( $customisedOnly = false ) {
666 return self::fetchLanguageNames( null, $customisedOnly ? 'mwfile' : 'mw' );
667 }
668
669 /**
670 * Get translated language names. This is done on best effort and
671 * by default this is exactly the same as Language::getLanguageNames.
672 * The CLDR extension provides translated names.
673 * @param $code String Language code.
674 * @return Array language code => language name
675 * @since 1.18.0
676 * @deprecated in 1.20, use fetchLanguageNames()
677 */
678 public static function getTranslatedLanguageNames( $code ) {
679 return self::fetchLanguageNames( $code, 'all' );
680 }
681
682 /**
683 * Get an array of language names, indexed by code.
684 * @param $inLanguage null|string: Code of language in which to return the names
685 * Use null for autonyms (native names)
686 * @param $include string:
687 * 'all' all available languages
688 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames
689 * 'mwfile' only if the language is in 'mw' *and* has a message file
690 * @return array|bool: language code => language name, false if $include is wrong
691 * @since 1.20
692 */
693 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
694 global $wgExtraLanguageNames;
695 static $coreLanguageNames;
696
697 if ( $coreLanguageNames === null ) {
698 include( MWInit::compiledPath( 'languages/Names.php' ) );
699 }
700
701 $names = array();
702
703 if( $inLanguage ) {
704 # TODO: also include when $inLanguage is null, when this code is more efficient
705 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
706 }
707
708 $mwNames = $wgExtraLanguageNames + $coreLanguageNames;
709 foreach ( $mwNames as $mwCode => $mwName ) {
710 # - Prefer own MediaWiki native name when not using the hook
711 # TODO: prefer it always to make it consistent, but casing is different in CLDR
712 # - For other names just add if not added through the hook
713 if ( ( $mwCode === $inLanguage && !$inLanguage ) || !isset( $names[$mwCode] ) ) {
714 $names[$mwCode] = $mwName;
715 }
716 }
717
718 if ( $include === 'all' ) {
719 return $names;
720 }
721
722 $returnMw = array();
723 $coreCodes = array_keys( $mwNames );
724 foreach( $coreCodes as $coreCode ) {
725 $returnMw[$coreCode] = $names[$coreCode];
726 }
727
728 if( $include === 'mw' ) {
729 return $returnMw;
730 } elseif( $include === 'mwfile' ) {
731 $namesMwFile = array();
732 # We do this using a foreach over the codes instead of a directory
733 # loop so that messages files in extensions will work correctly.
734 foreach ( $returnMw as $code => $value ) {
735 if ( is_readable( self::getMessagesFileName( $code ) ) ) {
736 $namesMwFile[$code] = $names[$code];
737 }
738 }
739 return $namesMwFile;
740 }
741 return false;
742 }
743
744 /**
745 * @param $code string: The code of the language for which to get the name
746 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
747 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
748 * @return string: Language name or empty
749 * @since 1.20
750 */
751 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
752 $array = self::fetchLanguageNames( $inLanguage, $include );
753 return !array_key_exists( $code, $array ) ? '' : $array[$code];
754 }
755
756 /**
757 * Get a message from the MediaWiki namespace.
758 *
759 * @param $msg String: message name
760 * @return string
761 */
762 function getMessageFromDB( $msg ) {
763 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
764 }
765
766 /**
767 * Get the native language name of $code.
768 * Only if defined in MediaWiki, no other data like CLDR.
769 * @param $code string
770 * @return string
771 * @deprecated in 1.20, use fetchLanguageName()
772 */
773 function getLanguageName( $code ) {
774 return self::fetchLanguageName( $code );
775 }
776
777 /**
778 * @param $key string
779 * @return string
780 */
781 function getMonthName( $key ) {
782 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
783 }
784
785 /**
786 * @return array
787 */
788 function getMonthNamesArray() {
789 $monthNames = array( '' );
790 for ( $i = 1; $i < 13; $i++ ) {
791 $monthNames[] = $this->getMonthName( $i );
792 }
793 return $monthNames;
794 }
795
796 /**
797 * @param $key string
798 * @return string
799 */
800 function getMonthNameGen( $key ) {
801 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
802 }
803
804 /**
805 * @param $key string
806 * @return string
807 */
808 function getMonthAbbreviation( $key ) {
809 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
810 }
811
812 /**
813 * @return array
814 */
815 function getMonthAbbreviationsArray() {
816 $monthNames = array( '' );
817 for ( $i = 1; $i < 13; $i++ ) {
818 $monthNames[] = $this->getMonthAbbreviation( $i );
819 }
820 return $monthNames;
821 }
822
823 /**
824 * @param $key string
825 * @return string
826 */
827 function getWeekdayName( $key ) {
828 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
829 }
830
831 /**
832 * @param $key string
833 * @return string
834 */
835 function getWeekdayAbbreviation( $key ) {
836 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
837 }
838
839 /**
840 * @param $key string
841 * @return string
842 */
843 function getIranianCalendarMonthName( $key ) {
844 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
845 }
846
847 /**
848 * @param $key string
849 * @return string
850 */
851 function getHebrewCalendarMonthName( $key ) {
852 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
853 }
854
855 /**
856 * @param $key string
857 * @return string
858 */
859 function getHebrewCalendarMonthNameGen( $key ) {
860 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
861 }
862
863 /**
864 * @param $key string
865 * @return string
866 */
867 function getHijriCalendarMonthName( $key ) {
868 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
869 }
870
871 /**
872 * This is a workalike of PHP's date() function, but with better
873 * internationalisation, a reduced set of format characters, and a better
874 * escaping format.
875 *
876 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
877 * PHP manual for definitions. There are a number of extensions, which
878 * start with "x":
879 *
880 * xn Do not translate digits of the next numeric format character
881 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
882 * xr Use roman numerals for the next numeric format character
883 * xh Use hebrew numerals for the next numeric format character
884 * xx Literal x
885 * xg Genitive month name
886 *
887 * xij j (day number) in Iranian calendar
888 * xiF F (month name) in Iranian calendar
889 * xin n (month number) in Iranian calendar
890 * xiy y (two digit year) in Iranian calendar
891 * xiY Y (full year) in Iranian calendar
892 *
893 * xjj j (day number) in Hebrew calendar
894 * xjF F (month name) in Hebrew calendar
895 * xjt t (days in month) in Hebrew calendar
896 * xjx xg (genitive month name) in Hebrew calendar
897 * xjn n (month number) in Hebrew calendar
898 * xjY Y (full year) in Hebrew calendar
899 *
900 * xmj j (day number) in Hijri calendar
901 * xmF F (month name) in Hijri calendar
902 * xmn n (month number) in Hijri calendar
903 * xmY Y (full year) in Hijri calendar
904 *
905 * xkY Y (full year) in Thai solar calendar. Months and days are
906 * identical to the Gregorian calendar
907 * xoY Y (full year) in Minguo calendar or Juche year.
908 * Months and days are identical to the
909 * Gregorian calendar
910 * xtY Y (full year) in Japanese nengo. Months and days are
911 * identical to the Gregorian calendar
912 *
913 * Characters enclosed in double quotes will be considered literal (with
914 * the quotes themselves removed). Unmatched quotes will be considered
915 * literal quotes. Example:
916 *
917 * "The month is" F => The month is January
918 * i's" => 20'11"
919 *
920 * Backslash escaping is also supported.
921 *
922 * Input timestamp is assumed to be pre-normalized to the desired local
923 * time zone, if any.
924 *
925 * @param $format String
926 * @param $ts String: 14-character timestamp
927 * YYYYMMDDHHMMSS
928 * 01234567890123
929 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
930 *
931 * @return string
932 */
933 function sprintfDate( $format, $ts ) {
934 $s = '';
935 $raw = false;
936 $roman = false;
937 $hebrewNum = false;
938 $unix = false;
939 $rawToggle = false;
940 $iranian = false;
941 $hebrew = false;
942 $hijri = false;
943 $thai = false;
944 $minguo = false;
945 $tenno = false;
946 for ( $p = 0; $p < strlen( $format ); $p++ ) {
947 $num = false;
948 $code = $format[$p];
949 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
950 $code .= $format[++$p];
951 }
952
953 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
954 $code .= $format[++$p];
955 }
956
957 switch ( $code ) {
958 case 'xx':
959 $s .= 'x';
960 break;
961 case 'xn':
962 $raw = true;
963 break;
964 case 'xN':
965 $rawToggle = !$rawToggle;
966 break;
967 case 'xr':
968 $roman = true;
969 break;
970 case 'xh':
971 $hebrewNum = true;
972 break;
973 case 'xg':
974 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
975 break;
976 case 'xjx':
977 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
978 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
979 break;
980 case 'd':
981 $num = substr( $ts, 6, 2 );
982 break;
983 case 'D':
984 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
985 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
986 break;
987 case 'j':
988 $num = intval( substr( $ts, 6, 2 ) );
989 break;
990 case 'xij':
991 if ( !$iranian ) {
992 $iranian = self::tsToIranian( $ts );
993 }
994 $num = $iranian[2];
995 break;
996 case 'xmj':
997 if ( !$hijri ) {
998 $hijri = self::tsToHijri( $ts );
999 }
1000 $num = $hijri[2];
1001 break;
1002 case 'xjj':
1003 if ( !$hebrew ) {
1004 $hebrew = self::tsToHebrew( $ts );
1005 }
1006 $num = $hebrew[2];
1007 break;
1008 case 'l':
1009 if ( !$unix ) {
1010 $unix = wfTimestamp( TS_UNIX, $ts );
1011 }
1012 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
1013 break;
1014 case 'N':
1015 if ( !$unix ) {
1016 $unix = wfTimestamp( TS_UNIX, $ts );
1017 }
1018 $w = gmdate( 'w', $unix );
1019 $num = $w ? $w : 7;
1020 break;
1021 case 'w':
1022 if ( !$unix ) {
1023 $unix = wfTimestamp( TS_UNIX, $ts );
1024 }
1025 $num = gmdate( 'w', $unix );
1026 break;
1027 case 'z':
1028 if ( !$unix ) {
1029 $unix = wfTimestamp( TS_UNIX, $ts );
1030 }
1031 $num = gmdate( 'z', $unix );
1032 break;
1033 case 'W':
1034 if ( !$unix ) {
1035 $unix = wfTimestamp( TS_UNIX, $ts );
1036 }
1037 $num = gmdate( 'W', $unix );
1038 break;
1039 case 'F':
1040 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1041 break;
1042 case 'xiF':
1043 if ( !$iranian ) {
1044 $iranian = self::tsToIranian( $ts );
1045 }
1046 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1047 break;
1048 case 'xmF':
1049 if ( !$hijri ) {
1050 $hijri = self::tsToHijri( $ts );
1051 }
1052 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1053 break;
1054 case 'xjF':
1055 if ( !$hebrew ) {
1056 $hebrew = self::tsToHebrew( $ts );
1057 }
1058 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1059 break;
1060 case 'm':
1061 $num = substr( $ts, 4, 2 );
1062 break;
1063 case 'M':
1064 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1065 break;
1066 case 'n':
1067 $num = intval( substr( $ts, 4, 2 ) );
1068 break;
1069 case 'xin':
1070 if ( !$iranian ) {
1071 $iranian = self::tsToIranian( $ts );
1072 }
1073 $num = $iranian[1];
1074 break;
1075 case 'xmn':
1076 if ( !$hijri ) {
1077 $hijri = self::tsToHijri ( $ts );
1078 }
1079 $num = $hijri[1];
1080 break;
1081 case 'xjn':
1082 if ( !$hebrew ) {
1083 $hebrew = self::tsToHebrew( $ts );
1084 }
1085 $num = $hebrew[1];
1086 break;
1087 case 't':
1088 if ( !$unix ) {
1089 $unix = wfTimestamp( TS_UNIX, $ts );
1090 }
1091 $num = gmdate( 't', $unix );
1092 break;
1093 case 'xjt':
1094 if ( !$hebrew ) {
1095 $hebrew = self::tsToHebrew( $ts );
1096 }
1097 $num = $hebrew[3];
1098 break;
1099 case 'L':
1100 if ( !$unix ) {
1101 $unix = wfTimestamp( TS_UNIX, $ts );
1102 }
1103 $num = gmdate( 'L', $unix );
1104 break;
1105 case 'o':
1106 if ( !$unix ) {
1107 $unix = wfTimestamp( TS_UNIX, $ts );
1108 }
1109 $num = gmdate( 'o', $unix );
1110 break;
1111 case 'Y':
1112 $num = substr( $ts, 0, 4 );
1113 break;
1114 case 'xiY':
1115 if ( !$iranian ) {
1116 $iranian = self::tsToIranian( $ts );
1117 }
1118 $num = $iranian[0];
1119 break;
1120 case 'xmY':
1121 if ( !$hijri ) {
1122 $hijri = self::tsToHijri( $ts );
1123 }
1124 $num = $hijri[0];
1125 break;
1126 case 'xjY':
1127 if ( !$hebrew ) {
1128 $hebrew = self::tsToHebrew( $ts );
1129 }
1130 $num = $hebrew[0];
1131 break;
1132 case 'xkY':
1133 if ( !$thai ) {
1134 $thai = self::tsToYear( $ts, 'thai' );
1135 }
1136 $num = $thai[0];
1137 break;
1138 case 'xoY':
1139 if ( !$minguo ) {
1140 $minguo = self::tsToYear( $ts, 'minguo' );
1141 }
1142 $num = $minguo[0];
1143 break;
1144 case 'xtY':
1145 if ( !$tenno ) {
1146 $tenno = self::tsToYear( $ts, 'tenno' );
1147 }
1148 $num = $tenno[0];
1149 break;
1150 case 'y':
1151 $num = substr( $ts, 2, 2 );
1152 break;
1153 case 'xiy':
1154 if ( !$iranian ) {
1155 $iranian = self::tsToIranian( $ts );
1156 }
1157 $num = substr( $iranian[0], -2 );
1158 break;
1159 case 'a':
1160 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1161 break;
1162 case 'A':
1163 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1164 break;
1165 case 'g':
1166 $h = substr( $ts, 8, 2 );
1167 $num = $h % 12 ? $h % 12 : 12;
1168 break;
1169 case 'G':
1170 $num = intval( substr( $ts, 8, 2 ) );
1171 break;
1172 case 'h':
1173 $h = substr( $ts, 8, 2 );
1174 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1175 break;
1176 case 'H':
1177 $num = substr( $ts, 8, 2 );
1178 break;
1179 case 'i':
1180 $num = substr( $ts, 10, 2 );
1181 break;
1182 case 's':
1183 $num = substr( $ts, 12, 2 );
1184 break;
1185 case 'c':
1186 if ( !$unix ) {
1187 $unix = wfTimestamp( TS_UNIX, $ts );
1188 }
1189 $s .= gmdate( 'c', $unix );
1190 break;
1191 case 'r':
1192 if ( !$unix ) {
1193 $unix = wfTimestamp( TS_UNIX, $ts );
1194 }
1195 $s .= gmdate( 'r', $unix );
1196 break;
1197 case 'U':
1198 if ( !$unix ) {
1199 $unix = wfTimestamp( TS_UNIX, $ts );
1200 }
1201 $num = $unix;
1202 break;
1203 case '\\':
1204 # Backslash escaping
1205 if ( $p < strlen( $format ) - 1 ) {
1206 $s .= $format[++$p];
1207 } else {
1208 $s .= '\\';
1209 }
1210 break;
1211 case '"':
1212 # Quoted literal
1213 if ( $p < strlen( $format ) - 1 ) {
1214 $endQuote = strpos( $format, '"', $p + 1 );
1215 if ( $endQuote === false ) {
1216 # No terminating quote, assume literal "
1217 $s .= '"';
1218 } else {
1219 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1220 $p = $endQuote;
1221 }
1222 } else {
1223 # Quote at end of string, assume literal "
1224 $s .= '"';
1225 }
1226 break;
1227 default:
1228 $s .= $format[$p];
1229 }
1230 if ( $num !== false ) {
1231 if ( $rawToggle || $raw ) {
1232 $s .= $num;
1233 $raw = false;
1234 } elseif ( $roman ) {
1235 $s .= self::romanNumeral( $num );
1236 $roman = false;
1237 } elseif ( $hebrewNum ) {
1238 $s .= self::hebrewNumeral( $num );
1239 $hebrewNum = false;
1240 } else {
1241 $s .= $this->formatNum( $num, true );
1242 }
1243 }
1244 }
1245 return $s;
1246 }
1247
1248 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1249 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1250
1251 /**
1252 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1253 * Gregorian dates to Iranian dates. Originally written in C, it
1254 * is released under the terms of GNU Lesser General Public
1255 * License. Conversion to PHP was performed by Niklas Laxström.
1256 *
1257 * Link: http://www.farsiweb.info/jalali/jalali.c
1258 *
1259 * @param $ts string
1260 *
1261 * @return string
1262 */
1263 private static function tsToIranian( $ts ) {
1264 $gy = substr( $ts, 0, 4 ) -1600;
1265 $gm = substr( $ts, 4, 2 ) -1;
1266 $gd = substr( $ts, 6, 2 ) -1;
1267
1268 # Days passed from the beginning (including leap years)
1269 $gDayNo = 365 * $gy
1270 + floor( ( $gy + 3 ) / 4 )
1271 - floor( ( $gy + 99 ) / 100 )
1272 + floor( ( $gy + 399 ) / 400 );
1273
1274 // Add days of the past months of this year
1275 for ( $i = 0; $i < $gm; $i++ ) {
1276 $gDayNo += self::$GREG_DAYS[$i];
1277 }
1278
1279 // Leap years
1280 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1281 $gDayNo++;
1282 }
1283
1284 // Days passed in current month
1285 $gDayNo += (int)$gd;
1286
1287 $jDayNo = $gDayNo - 79;
1288
1289 $jNp = floor( $jDayNo / 12053 );
1290 $jDayNo %= 12053;
1291
1292 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1293 $jDayNo %= 1461;
1294
1295 if ( $jDayNo >= 366 ) {
1296 $jy += floor( ( $jDayNo - 1 ) / 365 );
1297 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1298 }
1299
1300 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1301 $jDayNo -= self::$IRANIAN_DAYS[$i];
1302 }
1303
1304 $jm = $i + 1;
1305 $jd = $jDayNo + 1;
1306
1307 return array( $jy, $jm, $jd );
1308 }
1309
1310 /**
1311 * Converting Gregorian dates to Hijri dates.
1312 *
1313 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1314 *
1315 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1316 *
1317 * @param $ts string
1318 *
1319 * @return string
1320 */
1321 private static function tsToHijri( $ts ) {
1322 $year = substr( $ts, 0, 4 );
1323 $month = substr( $ts, 4, 2 );
1324 $day = substr( $ts, 6, 2 );
1325
1326 $zyr = $year;
1327 $zd = $day;
1328 $zm = $month;
1329 $zy = $zyr;
1330
1331 if (
1332 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1333 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1334 )
1335 {
1336 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1337 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1338 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1339 $zd - 32075;
1340 } else {
1341 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1342 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1343 }
1344
1345 $zl = $zjd -1948440 + 10632;
1346 $zn = (int)( ( $zl - 1 ) / 10631 );
1347 $zl = $zl - 10631 * $zn + 354;
1348 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1349 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1350 $zm = (int)( ( 24 * $zl ) / 709 );
1351 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1352 $zy = 30 * $zn + $zj - 30;
1353
1354 return array( $zy, $zm, $zd );
1355 }
1356
1357 /**
1358 * Converting Gregorian dates to Hebrew dates.
1359 *
1360 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1361 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1362 * to translate the relevant functions into PHP and release them under
1363 * GNU GPL.
1364 *
1365 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1366 * and Adar II is 14. In a non-leap year, Adar is 6.
1367 *
1368 * @param $ts string
1369 *
1370 * @return string
1371 */
1372 private static function tsToHebrew( $ts ) {
1373 # Parse date
1374 $year = substr( $ts, 0, 4 );
1375 $month = substr( $ts, 4, 2 );
1376 $day = substr( $ts, 6, 2 );
1377
1378 # Calculate Hebrew year
1379 $hebrewYear = $year + 3760;
1380
1381 # Month number when September = 1, August = 12
1382 $month += 4;
1383 if ( $month > 12 ) {
1384 # Next year
1385 $month -= 12;
1386 $year++;
1387 $hebrewYear++;
1388 }
1389
1390 # Calculate day of year from 1 September
1391 $dayOfYear = $day;
1392 for ( $i = 1; $i < $month; $i++ ) {
1393 if ( $i == 6 ) {
1394 # February
1395 $dayOfYear += 28;
1396 # Check if the year is leap
1397 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1398 $dayOfYear++;
1399 }
1400 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1401 $dayOfYear += 30;
1402 } else {
1403 $dayOfYear += 31;
1404 }
1405 }
1406
1407 # Calculate the start of the Hebrew year
1408 $start = self::hebrewYearStart( $hebrewYear );
1409
1410 # Calculate next year's start
1411 if ( $dayOfYear <= $start ) {
1412 # Day is before the start of the year - it is the previous year
1413 # Next year's start
1414 $nextStart = $start;
1415 # Previous year
1416 $year--;
1417 $hebrewYear--;
1418 # Add days since previous year's 1 September
1419 $dayOfYear += 365;
1420 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1421 # Leap year
1422 $dayOfYear++;
1423 }
1424 # Start of the new (previous) year
1425 $start = self::hebrewYearStart( $hebrewYear );
1426 } else {
1427 # Next year's start
1428 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1429 }
1430
1431 # Calculate Hebrew day of year
1432 $hebrewDayOfYear = $dayOfYear - $start;
1433
1434 # Difference between year's days
1435 $diff = $nextStart - $start;
1436 # Add 12 (or 13 for leap years) days to ignore the difference between
1437 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1438 # difference is only about the year type
1439 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1440 $diff += 13;
1441 } else {
1442 $diff += 12;
1443 }
1444
1445 # Check the year pattern, and is leap year
1446 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1447 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1448 # and non-leap years
1449 $yearPattern = $diff % 30;
1450 # Check if leap year
1451 $isLeap = $diff >= 30;
1452
1453 # Calculate day in the month from number of day in the Hebrew year
1454 # Don't check Adar - if the day is not in Adar, we will stop before;
1455 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1456 $hebrewDay = $hebrewDayOfYear;
1457 $hebrewMonth = 1;
1458 $days = 0;
1459 while ( $hebrewMonth <= 12 ) {
1460 # Calculate days in this month
1461 if ( $isLeap && $hebrewMonth == 6 ) {
1462 # Adar in a leap year
1463 if ( $isLeap ) {
1464 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1465 $days = 30;
1466 if ( $hebrewDay <= $days ) {
1467 # Day in Adar I
1468 $hebrewMonth = 13;
1469 } else {
1470 # Subtract the days of Adar I
1471 $hebrewDay -= $days;
1472 # Try Adar II
1473 $days = 29;
1474 if ( $hebrewDay <= $days ) {
1475 # Day in Adar II
1476 $hebrewMonth = 14;
1477 }
1478 }
1479 }
1480 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1481 # Cheshvan in a complete year (otherwise as the rule below)
1482 $days = 30;
1483 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1484 # Kislev in an incomplete year (otherwise as the rule below)
1485 $days = 29;
1486 } else {
1487 # Odd months have 30 days, even have 29
1488 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1489 }
1490 if ( $hebrewDay <= $days ) {
1491 # In the current month
1492 break;
1493 } else {
1494 # Subtract the days of the current month
1495 $hebrewDay -= $days;
1496 # Try in the next month
1497 $hebrewMonth++;
1498 }
1499 }
1500
1501 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1502 }
1503
1504 /**
1505 * This calculates the Hebrew year start, as days since 1 September.
1506 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1507 * Used for Hebrew date.
1508 *
1509 * @param $year int
1510 *
1511 * @return string
1512 */
1513 private static function hebrewYearStart( $year ) {
1514 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1515 $b = intval( ( $year - 1 ) % 4 );
1516 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1517 if ( $m < 0 ) {
1518 $m--;
1519 }
1520 $Mar = intval( $m );
1521 if ( $m < 0 ) {
1522 $m++;
1523 }
1524 $m -= $Mar;
1525
1526 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1527 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1528 $Mar++;
1529 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1530 $Mar += 2;
1531 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1532 $Mar++;
1533 }
1534
1535 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1536 return $Mar;
1537 }
1538
1539 /**
1540 * Algorithm to convert Gregorian dates to Thai solar dates,
1541 * Minguo dates or Minguo dates.
1542 *
1543 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1544 * http://en.wikipedia.org/wiki/Minguo_calendar
1545 * http://en.wikipedia.org/wiki/Japanese_era_name
1546 *
1547 * @param $ts String: 14-character timestamp
1548 * @param $cName String: calender name
1549 * @return Array: converted year, month, day
1550 */
1551 private static function tsToYear( $ts, $cName ) {
1552 $gy = substr( $ts, 0, 4 );
1553 $gm = substr( $ts, 4, 2 );
1554 $gd = substr( $ts, 6, 2 );
1555
1556 if ( !strcmp( $cName, 'thai' ) ) {
1557 # Thai solar dates
1558 # Add 543 years to the Gregorian calendar
1559 # Months and days are identical
1560 $gy_offset = $gy + 543;
1561 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1562 # Minguo dates
1563 # Deduct 1911 years from the Gregorian calendar
1564 # Months and days are identical
1565 $gy_offset = $gy - 1911;
1566 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1567 # Nengō dates up to Meiji period
1568 # Deduct years from the Gregorian calendar
1569 # depending on the nengo periods
1570 # Months and days are identical
1571 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1572 # Meiji period
1573 $gy_gannen = $gy - 1868 + 1;
1574 $gy_offset = $gy_gannen;
1575 if ( $gy_gannen == 1 ) {
1576 $gy_offset = '元';
1577 }
1578 $gy_offset = '明治' . $gy_offset;
1579 } elseif (
1580 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1581 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1582 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1583 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1584 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1585 )
1586 {
1587 # Taishō period
1588 $gy_gannen = $gy - 1912 + 1;
1589 $gy_offset = $gy_gannen;
1590 if ( $gy_gannen == 1 ) {
1591 $gy_offset = '元';
1592 }
1593 $gy_offset = '大正' . $gy_offset;
1594 } elseif (
1595 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1596 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1597 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1598 )
1599 {
1600 # Shōwa period
1601 $gy_gannen = $gy - 1926 + 1;
1602 $gy_offset = $gy_gannen;
1603 if ( $gy_gannen == 1 ) {
1604 $gy_offset = '元';
1605 }
1606 $gy_offset = '昭和' . $gy_offset;
1607 } else {
1608 # Heisei period
1609 $gy_gannen = $gy - 1989 + 1;
1610 $gy_offset = $gy_gannen;
1611 if ( $gy_gannen == 1 ) {
1612 $gy_offset = '元';
1613 }
1614 $gy_offset = '平成' . $gy_offset;
1615 }
1616 } else {
1617 $gy_offset = $gy;
1618 }
1619
1620 return array( $gy_offset, $gm, $gd );
1621 }
1622
1623 /**
1624 * Roman number formatting up to 3000
1625 *
1626 * @param $num int
1627 *
1628 * @return string
1629 */
1630 static function romanNumeral( $num ) {
1631 static $table = array(
1632 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1633 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1634 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1635 array( '', 'M', 'MM', 'MMM' )
1636 );
1637
1638 $num = intval( $num );
1639 if ( $num > 3000 || $num <= 0 ) {
1640 return $num;
1641 }
1642
1643 $s = '';
1644 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1645 if ( $num >= $pow10 ) {
1646 $s .= $table[$i][(int)floor( $num / $pow10 )];
1647 }
1648 $num = $num % $pow10;
1649 }
1650 return $s;
1651 }
1652
1653 /**
1654 * Hebrew Gematria number formatting up to 9999
1655 *
1656 * @param $num int
1657 *
1658 * @return string
1659 */
1660 static function hebrewNumeral( $num ) {
1661 static $table = array(
1662 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1663 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1664 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1665 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1666 );
1667
1668 $num = intval( $num );
1669 if ( $num > 9999 || $num <= 0 ) {
1670 return $num;
1671 }
1672
1673 $s = '';
1674 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1675 if ( $num >= $pow10 ) {
1676 if ( $num == 15 || $num == 16 ) {
1677 $s .= $table[0][9] . $table[0][$num - 9];
1678 $num = 0;
1679 } else {
1680 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1681 if ( $pow10 == 1000 ) {
1682 $s .= "'";
1683 }
1684 }
1685 }
1686 $num = $num % $pow10;
1687 }
1688 if ( strlen( $s ) == 2 ) {
1689 $str = $s . "'";
1690 } else {
1691 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1692 $str .= substr( $s, strlen( $s ) - 2, 2 );
1693 }
1694 $start = substr( $str, 0, strlen( $str ) - 2 );
1695 $end = substr( $str, strlen( $str ) - 2 );
1696 switch( $end ) {
1697 case 'כ':
1698 $str = $start . 'ך';
1699 break;
1700 case 'מ':
1701 $str = $start . 'ם';
1702 break;
1703 case 'נ':
1704 $str = $start . 'ן';
1705 break;
1706 case 'פ':
1707 $str = $start . 'ף';
1708 break;
1709 case 'צ':
1710 $str = $start . 'ץ';
1711 break;
1712 }
1713 return $str;
1714 }
1715
1716 /**
1717 * Used by date() and time() to adjust the time output.
1718 *
1719 * @param $ts Int the time in date('YmdHis') format
1720 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1721 * get user timecorrection setting)
1722 * @return int
1723 */
1724 function userAdjust( $ts, $tz = false ) {
1725 global $wgUser, $wgLocalTZoffset;
1726
1727 if ( $tz === false ) {
1728 $tz = $wgUser->getOption( 'timecorrection' );
1729 }
1730
1731 $data = explode( '|', $tz, 3 );
1732
1733 if ( $data[0] == 'ZoneInfo' ) {
1734 wfSuppressWarnings();
1735 $userTZ = timezone_open( $data[2] );
1736 wfRestoreWarnings();
1737 if ( $userTZ !== false ) {
1738 $date = date_create( $ts, timezone_open( 'UTC' ) );
1739 date_timezone_set( $date, $userTZ );
1740 $date = date_format( $date, 'YmdHis' );
1741 return $date;
1742 }
1743 # Unrecognized timezone, default to 'Offset' with the stored offset.
1744 $data[0] = 'Offset';
1745 }
1746
1747 $minDiff = 0;
1748 if ( $data[0] == 'System' || $tz == '' ) {
1749 #  Global offset in minutes.
1750 if ( isset( $wgLocalTZoffset ) ) {
1751 $minDiff = $wgLocalTZoffset;
1752 }
1753 } elseif ( $data[0] == 'Offset' ) {
1754 $minDiff = intval( $data[1] );
1755 } else {
1756 $data = explode( ':', $tz );
1757 if ( count( $data ) == 2 ) {
1758 $data[0] = intval( $data[0] );
1759 $data[1] = intval( $data[1] );
1760 $minDiff = abs( $data[0] ) * 60 + $data[1];
1761 if ( $data[0] < 0 ) {
1762 $minDiff = -$minDiff;
1763 }
1764 } else {
1765 $minDiff = intval( $data[0] ) * 60;
1766 }
1767 }
1768
1769 # No difference ? Return time unchanged
1770 if ( 0 == $minDiff ) {
1771 return $ts;
1772 }
1773
1774 wfSuppressWarnings(); // E_STRICT system time bitching
1775 # Generate an adjusted date; take advantage of the fact that mktime
1776 # will normalize out-of-range values so we don't have to split $minDiff
1777 # into hours and minutes.
1778 $t = mktime( (
1779 (int)substr( $ts, 8, 2 ) ), # Hours
1780 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
1781 (int)substr( $ts, 12, 2 ), # Seconds
1782 (int)substr( $ts, 4, 2 ), # Month
1783 (int)substr( $ts, 6, 2 ), # Day
1784 (int)substr( $ts, 0, 4 ) ); # Year
1785
1786 $date = date( 'YmdHis', $t );
1787 wfRestoreWarnings();
1788
1789 return $date;
1790 }
1791
1792 /**
1793 * This is meant to be used by time(), date(), and timeanddate() to get
1794 * the date preference they're supposed to use, it should be used in
1795 * all children.
1796 *
1797 *<code>
1798 * function timeanddate([...], $format = true) {
1799 * $datePreference = $this->dateFormat($format);
1800 * [...]
1801 * }
1802 *</code>
1803 *
1804 * @param $usePrefs Mixed: if true, the user's preference is used
1805 * if false, the site/language default is used
1806 * if int/string, assumed to be a format.
1807 * @return string
1808 */
1809 function dateFormat( $usePrefs = true ) {
1810 global $wgUser;
1811
1812 if ( is_bool( $usePrefs ) ) {
1813 if ( $usePrefs ) {
1814 $datePreference = $wgUser->getDatePreference();
1815 } else {
1816 $datePreference = (string)User::getDefaultOption( 'date' );
1817 }
1818 } else {
1819 $datePreference = (string)$usePrefs;
1820 }
1821
1822 // return int
1823 if ( $datePreference == '' ) {
1824 return 'default';
1825 }
1826
1827 return $datePreference;
1828 }
1829
1830 /**
1831 * Get a format string for a given type and preference
1832 * @param $type string May be date, time or both
1833 * @param $pref string The format name as it appears in Messages*.php
1834 *
1835 * @return string
1836 */
1837 function getDateFormatString( $type, $pref ) {
1838 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
1839 if ( $pref == 'default' ) {
1840 $pref = $this->getDefaultDateFormat();
1841 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1842 } else {
1843 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1844 if ( is_null( $df ) ) {
1845 $pref = $this->getDefaultDateFormat();
1846 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1847 }
1848 }
1849 $this->dateFormatStrings[$type][$pref] = $df;
1850 }
1851 return $this->dateFormatStrings[$type][$pref];
1852 }
1853
1854 /**
1855 * @param $ts Mixed: the time format which needs to be turned into a
1856 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1857 * @param $adj Bool: whether to adjust the time output according to the
1858 * user configured offset ($timecorrection)
1859 * @param $format Mixed: true to use user's date format preference
1860 * @param $timecorrection String|bool the time offset as returned by
1861 * validateTimeZone() in Special:Preferences
1862 * @return string
1863 */
1864 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1865 $ts = wfTimestamp( TS_MW, $ts );
1866 if ( $adj ) {
1867 $ts = $this->userAdjust( $ts, $timecorrection );
1868 }
1869 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1870 return $this->sprintfDate( $df, $ts );
1871 }
1872
1873 /**
1874 * @param $ts Mixed: the time format which needs to be turned into a
1875 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1876 * @param $adj Bool: whether to adjust the time output according to the
1877 * user configured offset ($timecorrection)
1878 * @param $format Mixed: true to use user's date format preference
1879 * @param $timecorrection String|bool the time offset as returned by
1880 * validateTimeZone() in Special:Preferences
1881 * @return string
1882 */
1883 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1884 $ts = wfTimestamp( TS_MW, $ts );
1885 if ( $adj ) {
1886 $ts = $this->userAdjust( $ts, $timecorrection );
1887 }
1888 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1889 return $this->sprintfDate( $df, $ts );
1890 }
1891
1892 /**
1893 * @param $ts Mixed: the time format which needs to be turned into a
1894 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1895 * @param $adj Bool: whether to adjust the time output according to the
1896 * user configured offset ($timecorrection)
1897 * @param $format Mixed: what format to return, if it's false output the
1898 * default one (default true)
1899 * @param $timecorrection String|bool the time offset as returned by
1900 * validateTimeZone() in Special:Preferences
1901 * @return string
1902 */
1903 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1904 $ts = wfTimestamp( TS_MW, $ts );
1905 if ( $adj ) {
1906 $ts = $this->userAdjust( $ts, $timecorrection );
1907 }
1908 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1909 return $this->sprintfDate( $df, $ts );
1910 }
1911
1912 /**
1913 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
1914 *
1915 * @since 1.20
1916 *
1917 * @param integer $seconds The amount of seconds.
1918 * @param array $chosenIntervals The intervals to enable.
1919 *
1920 * @return string
1921 */
1922 public function formatDuration( $seconds, array $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' ) ) {
1923 $intervals = array(
1924 'millennia' => 1000 * 31557600,
1925 'centuries' => 100 * 31557600,
1926 'decades' => 10 * 31557600,
1927 'years' => 31557600, // 86400 * 365.25
1928 'weeks' => 604800,
1929 'days' => 86400,
1930 'hours' => 3600,
1931 'minutes' => 60,
1932 'seconds' => 1,
1933 );
1934
1935 if ( !empty( $chosenIntervals ) ) {
1936 $intervals = array_intersect_key( $intervals, array_flip( $chosenIntervals ) );
1937 }
1938
1939 $segments = array();
1940
1941 foreach ( $intervals as $name => $length ) {
1942 $value = floor( $seconds / $length );
1943
1944 if ( $value > 0 || ( $name == 'seconds' && empty( $segments ) ) ) {
1945 $seconds -= $value * $length;
1946 $message = new Message( 'duration-' . $name, array( $value ) );
1947 $segments[] = $message->inLanguage( $this )->escaped();
1948 }
1949 }
1950
1951 return $this->listToText( $segments );
1952 }
1953
1954 /**
1955 * Internal helper function for userDate(), userTime() and userTimeAndDate()
1956 *
1957 * @param $type String: can be 'date', 'time' or 'both'
1958 * @param $ts Mixed: the time format which needs to be turned into a
1959 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1960 * @param $user User object used to get preferences for timezone and format
1961 * @param $options Array, can contain the following keys:
1962 * - 'timecorrection': time correction, can have the following values:
1963 * - true: use user's preference
1964 * - false: don't use time correction
1965 * - integer: value of time correction in minutes
1966 * - 'format': format to use, can have the following values:
1967 * - true: use user's preference
1968 * - false: use default preference
1969 * - string: format to use
1970 * @since 1.19
1971 * @return String
1972 */
1973 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
1974 $ts = wfTimestamp( TS_MW, $ts );
1975 $options += array( 'timecorrection' => true, 'format' => true );
1976 if ( $options['timecorrection'] !== false ) {
1977 if ( $options['timecorrection'] === true ) {
1978 $offset = $user->getOption( 'timecorrection' );
1979 } else {
1980 $offset = $options['timecorrection'];
1981 }
1982 $ts = $this->userAdjust( $ts, $offset );
1983 }
1984 if ( $options['format'] === true ) {
1985 $format = $user->getDatePreference();
1986 } else {
1987 $format = $options['format'];
1988 }
1989 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
1990 return $this->sprintfDate( $df, $ts );
1991 }
1992
1993 /**
1994 * Get the formatted date for the given timestamp and formatted for
1995 * the given user.
1996 *
1997 * @param $ts Mixed: the time format which needs to be turned into a
1998 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1999 * @param $user User object used to get preferences for timezone and format
2000 * @param $options Array, can contain the following keys:
2001 * - 'timecorrection': time correction, can have the following values:
2002 * - true: use user's preference
2003 * - false: don't use time correction
2004 * - integer: value of time correction in minutes
2005 * - 'format': format to use, can have the following values:
2006 * - true: use user's preference
2007 * - false: use default preference
2008 * - string: format to use
2009 * @since 1.19
2010 * @return String
2011 */
2012 public function userDate( $ts, User $user, array $options = array() ) {
2013 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2014 }
2015
2016 /**
2017 * Get the formatted time for the given timestamp and formatted for
2018 * the given user.
2019 *
2020 * @param $ts Mixed: the time format which needs to be turned into a
2021 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2022 * @param $user User object used to get preferences for timezone and format
2023 * @param $options Array, can contain the following keys:
2024 * - 'timecorrection': time correction, can have the following values:
2025 * - true: use user's preference
2026 * - false: don't use time correction
2027 * - integer: value of time correction in minutes
2028 * - 'format': format to use, can have the following values:
2029 * - true: use user's preference
2030 * - false: use default preference
2031 * - string: format to use
2032 * @since 1.19
2033 * @return String
2034 */
2035 public function userTime( $ts, User $user, array $options = array() ) {
2036 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2037 }
2038
2039 /**
2040 * Get the formatted date and time for the given timestamp and formatted for
2041 * the given user.
2042 *
2043 * @param $ts Mixed: the time format which needs to be turned into a
2044 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2045 * @param $user User object used to get preferences for timezone and format
2046 * @param $options Array, can contain the following keys:
2047 * - 'timecorrection': time correction, can have the following values:
2048 * - true: use user's preference
2049 * - false: don't use time correction
2050 * - integer: value of time correction in minutes
2051 * - 'format': format to use, can have the following values:
2052 * - true: use user's preference
2053 * - false: use default preference
2054 * - string: format to use
2055 * @since 1.19
2056 * @return String
2057 */
2058 public function userTimeAndDate( $ts, User $user, array $options = array() ) {
2059 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2060 }
2061
2062 /**
2063 * @param $key string
2064 * @return array|null
2065 */
2066 function getMessage( $key ) {
2067 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2068 }
2069
2070 /**
2071 * @return array
2072 */
2073 function getAllMessages() {
2074 return self::$dataCache->getItem( $this->mCode, 'messages' );
2075 }
2076
2077 /**
2078 * @param $in
2079 * @param $out
2080 * @param $string
2081 * @return string
2082 */
2083 function iconv( $in, $out, $string ) {
2084 # This is a wrapper for iconv in all languages except esperanto,
2085 # which does some nasty x-conversions beforehand
2086
2087 # Even with //IGNORE iconv can whine about illegal characters in
2088 # *input* string. We just ignore those too.
2089 # REF: http://bugs.php.net/bug.php?id=37166
2090 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2091 wfSuppressWarnings();
2092 $text = iconv( $in, $out . '//IGNORE', $string );
2093 wfRestoreWarnings();
2094 return $text;
2095 }
2096
2097 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2098
2099 /**
2100 * @param $matches array
2101 * @return mixed|string
2102 */
2103 function ucwordbreaksCallbackAscii( $matches ) {
2104 return $this->ucfirst( $matches[1] );
2105 }
2106
2107 /**
2108 * @param $matches array
2109 * @return string
2110 */
2111 function ucwordbreaksCallbackMB( $matches ) {
2112 return mb_strtoupper( $matches[0] );
2113 }
2114
2115 /**
2116 * @param $matches array
2117 * @return string
2118 */
2119 function ucCallback( $matches ) {
2120 list( $wikiUpperChars ) = self::getCaseMaps();
2121 return strtr( $matches[1], $wikiUpperChars );
2122 }
2123
2124 /**
2125 * @param $matches array
2126 * @return string
2127 */
2128 function lcCallback( $matches ) {
2129 list( , $wikiLowerChars ) = self::getCaseMaps();
2130 return strtr( $matches[1], $wikiLowerChars );
2131 }
2132
2133 /**
2134 * @param $matches array
2135 * @return string
2136 */
2137 function ucwordsCallbackMB( $matches ) {
2138 return mb_strtoupper( $matches[0] );
2139 }
2140
2141 /**
2142 * @param $matches array
2143 * @return string
2144 */
2145 function ucwordsCallbackWiki( $matches ) {
2146 list( $wikiUpperChars ) = self::getCaseMaps();
2147 return strtr( $matches[0], $wikiUpperChars );
2148 }
2149
2150 /**
2151 * Make a string's first character uppercase
2152 *
2153 * @param $str string
2154 *
2155 * @return string
2156 */
2157 function ucfirst( $str ) {
2158 $o = ord( $str );
2159 if ( $o < 96 ) { // if already uppercase...
2160 return $str;
2161 } elseif ( $o < 128 ) {
2162 return ucfirst( $str ); // use PHP's ucfirst()
2163 } else {
2164 // fall back to more complex logic in case of multibyte strings
2165 return $this->uc( $str, true );
2166 }
2167 }
2168
2169 /**
2170 * Convert a string to uppercase
2171 *
2172 * @param $str string
2173 * @param $first bool
2174 *
2175 * @return string
2176 */
2177 function uc( $str, $first = false ) {
2178 if ( function_exists( 'mb_strtoupper' ) ) {
2179 if ( $first ) {
2180 if ( $this->isMultibyte( $str ) ) {
2181 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2182 } else {
2183 return ucfirst( $str );
2184 }
2185 } else {
2186 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2187 }
2188 } else {
2189 if ( $this->isMultibyte( $str ) ) {
2190 $x = $first ? '^' : '';
2191 return preg_replace_callback(
2192 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2193 array( $this, 'ucCallback' ),
2194 $str
2195 );
2196 } else {
2197 return $first ? ucfirst( $str ) : strtoupper( $str );
2198 }
2199 }
2200 }
2201
2202 /**
2203 * @param $str string
2204 * @return mixed|string
2205 */
2206 function lcfirst( $str ) {
2207 $o = ord( $str );
2208 if ( !$o ) {
2209 return strval( $str );
2210 } elseif ( $o >= 128 ) {
2211 return $this->lc( $str, true );
2212 } elseif ( $o > 96 ) {
2213 return $str;
2214 } else {
2215 $str[0] = strtolower( $str[0] );
2216 return $str;
2217 }
2218 }
2219
2220 /**
2221 * @param $str string
2222 * @param $first bool
2223 * @return mixed|string
2224 */
2225 function lc( $str, $first = false ) {
2226 if ( function_exists( 'mb_strtolower' ) ) {
2227 if ( $first ) {
2228 if ( $this->isMultibyte( $str ) ) {
2229 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2230 } else {
2231 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2232 }
2233 } else {
2234 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2235 }
2236 } else {
2237 if ( $this->isMultibyte( $str ) ) {
2238 $x = $first ? '^' : '';
2239 return preg_replace_callback(
2240 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2241 array( $this, 'lcCallback' ),
2242 $str
2243 );
2244 } else {
2245 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2246 }
2247 }
2248 }
2249
2250 /**
2251 * @param $str string
2252 * @return bool
2253 */
2254 function isMultibyte( $str ) {
2255 return (bool)preg_match( '/[\x80-\xff]/', $str );
2256 }
2257
2258 /**
2259 * @param $str string
2260 * @return mixed|string
2261 */
2262 function ucwords( $str ) {
2263 if ( $this->isMultibyte( $str ) ) {
2264 $str = $this->lc( $str );
2265
2266 // regexp to find first letter in each word (i.e. after each space)
2267 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2268
2269 // function to use to capitalize a single char
2270 if ( function_exists( 'mb_strtoupper' ) ) {
2271 return preg_replace_callback(
2272 $replaceRegexp,
2273 array( $this, 'ucwordsCallbackMB' ),
2274 $str
2275 );
2276 } else {
2277 return preg_replace_callback(
2278 $replaceRegexp,
2279 array( $this, 'ucwordsCallbackWiki' ),
2280 $str
2281 );
2282 }
2283 } else {
2284 return ucwords( strtolower( $str ) );
2285 }
2286 }
2287
2288 /**
2289 * capitalize words at word breaks
2290 *
2291 * @param $str string
2292 * @return mixed
2293 */
2294 function ucwordbreaks( $str ) {
2295 if ( $this->isMultibyte( $str ) ) {
2296 $str = $this->lc( $str );
2297
2298 // since \b doesn't work for UTF-8, we explicitely define word break chars
2299 $breaks = "[ \-\(\)\}\{\.,\?!]";
2300
2301 // find first letter after word break
2302 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2303
2304 if ( function_exists( 'mb_strtoupper' ) ) {
2305 return preg_replace_callback(
2306 $replaceRegexp,
2307 array( $this, 'ucwordbreaksCallbackMB' ),
2308 $str
2309 );
2310 } else {
2311 return preg_replace_callback(
2312 $replaceRegexp,
2313 array( $this, 'ucwordsCallbackWiki' ),
2314 $str
2315 );
2316 }
2317 } else {
2318 return preg_replace_callback(
2319 '/\b([\w\x80-\xff]+)\b/',
2320 array( $this, 'ucwordbreaksCallbackAscii' ),
2321 $str
2322 );
2323 }
2324 }
2325
2326 /**
2327 * Return a case-folded representation of $s
2328 *
2329 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2330 * and $s2 are the same except for the case of their characters. It is not
2331 * necessary for the value returned to make sense when displayed.
2332 *
2333 * Do *not* perform any other normalisation in this function. If a caller
2334 * uses this function when it should be using a more general normalisation
2335 * function, then fix the caller.
2336 *
2337 * @param $s string
2338 *
2339 * @return string
2340 */
2341 function caseFold( $s ) {
2342 return $this->uc( $s );
2343 }
2344
2345 /**
2346 * @param $s string
2347 * @return string
2348 */
2349 function checkTitleEncoding( $s ) {
2350 if ( is_array( $s ) ) {
2351 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2352 }
2353 # Check for non-UTF-8 URLs
2354 $ishigh = preg_match( '/[\x80-\xff]/', $s );
2355 if ( !$ishigh ) {
2356 return $s;
2357 }
2358
2359 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2360 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
2361 if ( $isutf8 ) {
2362 return $s;
2363 }
2364
2365 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2366 }
2367
2368 /**
2369 * @return array
2370 */
2371 function fallback8bitEncoding() {
2372 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2373 }
2374
2375 /**
2376 * Most writing systems use whitespace to break up words.
2377 * Some languages such as Chinese don't conventionally do this,
2378 * which requires special handling when breaking up words for
2379 * searching etc.
2380 *
2381 * @return bool
2382 */
2383 function hasWordBreaks() {
2384 return true;
2385 }
2386
2387 /**
2388 * Some languages such as Chinese require word segmentation,
2389 * Specify such segmentation when overridden in derived class.
2390 *
2391 * @param $string String
2392 * @return String
2393 */
2394 function segmentByWord( $string ) {
2395 return $string;
2396 }
2397
2398 /**
2399 * Some languages have special punctuation need to be normalized.
2400 * Make such changes here.
2401 *
2402 * @param $string String
2403 * @return String
2404 */
2405 function normalizeForSearch( $string ) {
2406 return self::convertDoubleWidth( $string );
2407 }
2408
2409 /**
2410 * convert double-width roman characters to single-width.
2411 * range: ff00-ff5f ~= 0020-007f
2412 *
2413 * @param $string string
2414 *
2415 * @return string
2416 */
2417 protected static function convertDoubleWidth( $string ) {
2418 static $full = null;
2419 static $half = null;
2420
2421 if ( $full === null ) {
2422 $fullWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2423 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2424 $full = str_split( $fullWidth, 3 );
2425 $half = str_split( $halfWidth );
2426 }
2427
2428 $string = str_replace( $full, $half, $string );
2429 return $string;
2430 }
2431
2432 /**
2433 * @param $string string
2434 * @param $pattern string
2435 * @return string
2436 */
2437 protected static function insertSpace( $string, $pattern ) {
2438 $string = preg_replace( $pattern, " $1 ", $string );
2439 $string = preg_replace( '/ +/', ' ', $string );
2440 return $string;
2441 }
2442
2443 /**
2444 * @param $termsArray array
2445 * @return array
2446 */
2447 function convertForSearchResult( $termsArray ) {
2448 # some languages, e.g. Chinese, need to do a conversion
2449 # in order for search results to be displayed correctly
2450 return $termsArray;
2451 }
2452
2453 /**
2454 * Get the first character of a string.
2455 *
2456 * @param $s string
2457 * @return string
2458 */
2459 function firstChar( $s ) {
2460 $matches = array();
2461 preg_match(
2462 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2463 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2464 $s,
2465 $matches
2466 );
2467
2468 if ( isset( $matches[1] ) ) {
2469 if ( strlen( $matches[1] ) != 3 ) {
2470 return $matches[1];
2471 }
2472
2473 // Break down Hangul syllables to grab the first jamo
2474 $code = utf8ToCodepoint( $matches[1] );
2475 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2476 return $matches[1];
2477 } elseif ( $code < 0xb098 ) {
2478 return "\xe3\x84\xb1";
2479 } elseif ( $code < 0xb2e4 ) {
2480 return "\xe3\x84\xb4";
2481 } elseif ( $code < 0xb77c ) {
2482 return "\xe3\x84\xb7";
2483 } elseif ( $code < 0xb9c8 ) {
2484 return "\xe3\x84\xb9";
2485 } elseif ( $code < 0xbc14 ) {
2486 return "\xe3\x85\x81";
2487 } elseif ( $code < 0xc0ac ) {
2488 return "\xe3\x85\x82";
2489 } elseif ( $code < 0xc544 ) {
2490 return "\xe3\x85\x85";
2491 } elseif ( $code < 0xc790 ) {
2492 return "\xe3\x85\x87";
2493 } elseif ( $code < 0xcc28 ) {
2494 return "\xe3\x85\x88";
2495 } elseif ( $code < 0xce74 ) {
2496 return "\xe3\x85\x8a";
2497 } elseif ( $code < 0xd0c0 ) {
2498 return "\xe3\x85\x8b";
2499 } elseif ( $code < 0xd30c ) {
2500 return "\xe3\x85\x8c";
2501 } elseif ( $code < 0xd558 ) {
2502 return "\xe3\x85\x8d";
2503 } else {
2504 return "\xe3\x85\x8e";
2505 }
2506 } else {
2507 return '';
2508 }
2509 }
2510
2511 function initEncoding() {
2512 # Some languages may have an alternate char encoding option
2513 # (Esperanto X-coding, Japanese furigana conversion, etc)
2514 # If this language is used as the primary content language,
2515 # an override to the defaults can be set here on startup.
2516 }
2517
2518 /**
2519 * @param $s string
2520 * @return string
2521 */
2522 function recodeForEdit( $s ) {
2523 # For some languages we'll want to explicitly specify
2524 # which characters make it into the edit box raw
2525 # or are converted in some way or another.
2526 global $wgEditEncoding;
2527 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
2528 return $s;
2529 } else {
2530 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2531 }
2532 }
2533
2534 /**
2535 * @param $s string
2536 * @return string
2537 */
2538 function recodeInput( $s ) {
2539 # Take the previous into account.
2540 global $wgEditEncoding;
2541 if ( $wgEditEncoding != '' ) {
2542 $enc = $wgEditEncoding;
2543 } else {
2544 $enc = 'UTF-8';
2545 }
2546 if ( $enc == 'UTF-8' ) {
2547 return $s;
2548 } else {
2549 return $this->iconv( $enc, 'UTF-8', $s );
2550 }
2551 }
2552
2553 /**
2554 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2555 * also cleans up certain backwards-compatible sequences, converting them
2556 * to the modern Unicode equivalent.
2557 *
2558 * This is language-specific for performance reasons only.
2559 *
2560 * @param $s string
2561 *
2562 * @return string
2563 */
2564 function normalize( $s ) {
2565 global $wgAllUnicodeFixes;
2566 $s = UtfNormal::cleanUp( $s );
2567 if ( $wgAllUnicodeFixes ) {
2568 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2569 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2570 }
2571
2572 return $s;
2573 }
2574
2575 /**
2576 * Transform a string using serialized data stored in the given file (which
2577 * must be in the serialized subdirectory of $IP). The file contains pairs
2578 * mapping source characters to destination characters.
2579 *
2580 * The data is cached in process memory. This will go faster if you have the
2581 * FastStringSearch extension.
2582 *
2583 * @param $file string
2584 * @param $string string
2585 *
2586 * @throws MWException
2587 * @return string
2588 */
2589 function transformUsingPairFile( $file, $string ) {
2590 if ( !isset( $this->transformData[$file] ) ) {
2591 $data = wfGetPrecompiledData( $file );
2592 if ( $data === false ) {
2593 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2594 }
2595 $this->transformData[$file] = new ReplacementArray( $data );
2596 }
2597 return $this->transformData[$file]->replace( $string );
2598 }
2599
2600 /**
2601 * For right-to-left language support
2602 *
2603 * @return bool
2604 */
2605 function isRTL() {
2606 return self::$dataCache->getItem( $this->mCode, 'rtl' );
2607 }
2608
2609 /**
2610 * Return the correct HTML 'dir' attribute value for this language.
2611 * @return String
2612 */
2613 function getDir() {
2614 return $this->isRTL() ? 'rtl' : 'ltr';
2615 }
2616
2617 /**
2618 * Return 'left' or 'right' as appropriate alignment for line-start
2619 * for this language's text direction.
2620 *
2621 * Should be equivalent to CSS3 'start' text-align value....
2622 *
2623 * @return String
2624 */
2625 function alignStart() {
2626 return $this->isRTL() ? 'right' : 'left';
2627 }
2628
2629 /**
2630 * Return 'right' or 'left' as appropriate alignment for line-end
2631 * for this language's text direction.
2632 *
2633 * Should be equivalent to CSS3 'end' text-align value....
2634 *
2635 * @return String
2636 */
2637 function alignEnd() {
2638 return $this->isRTL() ? 'left' : 'right';
2639 }
2640
2641 /**
2642 * A hidden direction mark (LRM or RLM), depending on the language direction
2643 *
2644 * @param $opposite Boolean Get the direction mark opposite to your language
2645 * @return string
2646 */
2647 function getDirMark( $opposite = false ) {
2648 $rtl = "\xE2\x80\x8F";
2649 $ltr = "\xE2\x80\x8E";
2650 if ( $opposite ) { return $this->isRTL() ? $ltr : $rtl; }
2651 return $this->isRTL() ? $rtl : $ltr;
2652 }
2653
2654 /**
2655 * @return array
2656 */
2657 function capitalizeAllNouns() {
2658 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2659 }
2660
2661 /**
2662 * An arrow, depending on the language direction
2663 *
2664 * @return string
2665 */
2666 function getArrow() {
2667 return $this->isRTL() ? '←' : '→';
2668 }
2669
2670 /**
2671 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2672 *
2673 * @return bool
2674 */
2675 function linkPrefixExtension() {
2676 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2677 }
2678
2679 /**
2680 * @return array
2681 */
2682 function getMagicWords() {
2683 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2684 }
2685
2686 protected function doMagicHook() {
2687 if ( $this->mMagicHookDone ) {
2688 return;
2689 }
2690 $this->mMagicHookDone = true;
2691 wfProfileIn( 'LanguageGetMagic' );
2692 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2693 wfProfileOut( 'LanguageGetMagic' );
2694 }
2695
2696 /**
2697 * Fill a MagicWord object with data from here
2698 *
2699 * @param $mw
2700 */
2701 function getMagic( $mw ) {
2702 $this->doMagicHook();
2703
2704 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
2705 $rawEntry = $this->mMagicExtensions[$mw->mId];
2706 } else {
2707 $magicWords = $this->getMagicWords();
2708 if ( isset( $magicWords[$mw->mId] ) ) {
2709 $rawEntry = $magicWords[$mw->mId];
2710 } else {
2711 $rawEntry = false;
2712 }
2713 }
2714
2715 if ( !is_array( $rawEntry ) ) {
2716 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
2717 } else {
2718 $mw->mCaseSensitive = $rawEntry[0];
2719 $mw->mSynonyms = array_slice( $rawEntry, 1 );
2720 }
2721 }
2722
2723 /**
2724 * Add magic words to the extension array
2725 *
2726 * @param $newWords array
2727 */
2728 function addMagicWordsByLang( $newWords ) {
2729 $fallbackChain = $this->getFallbackLanguages();
2730 $fallbackChain = array_reverse( $fallbackChain );
2731 foreach ( $fallbackChain as $code ) {
2732 if ( isset( $newWords[$code] ) ) {
2733 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
2734 }
2735 }
2736 }
2737
2738 /**
2739 * Get special page names, as an associative array
2740 * case folded alias => real name
2741 */
2742 function getSpecialPageAliases() {
2743 // Cache aliases because it may be slow to load them
2744 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
2745 // Initialise array
2746 $this->mExtendedSpecialPageAliases =
2747 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
2748 wfRunHooks( 'LanguageGetSpecialPageAliases',
2749 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
2750 }
2751
2752 return $this->mExtendedSpecialPageAliases;
2753 }
2754
2755 /**
2756 * Italic is unsuitable for some languages
2757 *
2758 * @param $text String: the text to be emphasized.
2759 * @return string
2760 */
2761 function emphasize( $text ) {
2762 return "<em>$text</em>";
2763 }
2764
2765 /**
2766 * Normally we output all numbers in plain en_US style, that is
2767 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2768 * point twohundredthirtyfive. However this is not suitable for all
2769 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2770 * Icelandic just want to use commas instead of dots, and dots instead
2771 * of commas like "293.291,235".
2772 *
2773 * An example of this function being called:
2774 * <code>
2775 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2776 * </code>
2777 *
2778 * See LanguageGu.php for the Gujarati implementation and
2779 * $separatorTransformTable on MessageIs.php for
2780 * the , => . and . => , implementation.
2781 *
2782 * @todo check if it's viable to use localeconv() for the decimal
2783 * separator thing.
2784 * @param $number Mixed: the string to be formatted, should be an integer
2785 * or a floating point number.
2786 * @param $nocommafy Bool: set to true for special numbers like dates
2787 * @return string
2788 */
2789 public function formatNum( $number, $nocommafy = false ) {
2790 global $wgTranslateNumerals;
2791 if ( !$nocommafy ) {
2792 $number = $this->commafy( $number );
2793 $s = $this->separatorTransformTable();
2794 if ( $s ) {
2795 $number = strtr( $number, $s );
2796 }
2797 }
2798
2799 if ( $wgTranslateNumerals ) {
2800 $s = $this->digitTransformTable();
2801 if ( $s ) {
2802 $number = strtr( $number, $s );
2803 }
2804 }
2805
2806 return $number;
2807 }
2808
2809 /**
2810 * @param $number string
2811 * @return string
2812 */
2813 function parseFormattedNumber( $number ) {
2814 $s = $this->digitTransformTable();
2815 if ( $s ) {
2816 $number = strtr( $number, array_flip( $s ) );
2817 }
2818
2819 $s = $this->separatorTransformTable();
2820 if ( $s ) {
2821 $number = strtr( $number, array_flip( $s ) );
2822 }
2823
2824 $number = strtr( $number, array( ',' => '' ) );
2825 return $number;
2826 }
2827
2828 /**
2829 * Adds commas to a given number
2830 * @since 1.19
2831 * @param $_ mixed
2832 * @return string
2833 */
2834 function commafy( $_ ) {
2835 $digitGroupingPattern = $this->digitGroupingPattern();
2836 if ( $_ === null ) {
2837 return '';
2838 }
2839
2840 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
2841 // default grouping is at thousands, use the same for ###,###,### pattern too.
2842 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2843 } else {
2844 // Ref: http://cldr.unicode.org/translation/number-patterns
2845 $sign = "";
2846 if ( intval( $_ ) < 0 ) {
2847 // For negative numbers apply the algorithm like positive number and add sign.
2848 $sign = "-";
2849 $_ = substr( $_, 1 );
2850 }
2851 $numberpart = array();
2852 $decimalpart = array();
2853 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
2854 preg_match( "/\d+/", $_, $numberpart );
2855 preg_match( "/\.\d*/", $_, $decimalpart );
2856 $groupedNumber = ( count( $decimalpart ) > 0 ) ? $decimalpart[0]:"";
2857 if ( $groupedNumber === $_ ) {
2858 // the string does not have any number part. Eg: .12345
2859 return $sign . $groupedNumber;
2860 }
2861 $start = $end = strlen( $numberpart[0] );
2862 while ( $start > 0 ) {
2863 $match = $matches[0][$numMatches -1] ;
2864 $matchLen = strlen( $match );
2865 $start = $end - $matchLen;
2866 if ( $start < 0 ) {
2867 $start = 0;
2868 }
2869 $groupedNumber = substr( $_ , $start, $end -$start ) . $groupedNumber ;
2870 $end = $start;
2871 if ( $numMatches > 1 ) {
2872 // use the last pattern for the rest of the number
2873 $numMatches--;
2874 }
2875 if ( $start > 0 ) {
2876 $groupedNumber = "," . $groupedNumber;
2877 }
2878 }
2879 return $sign . $groupedNumber;
2880 }
2881 }
2882 /**
2883 * @return String
2884 */
2885 function digitGroupingPattern() {
2886 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
2887 }
2888
2889 /**
2890 * @return array
2891 */
2892 function digitTransformTable() {
2893 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
2894 }
2895
2896 /**
2897 * @return array
2898 */
2899 function separatorTransformTable() {
2900 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
2901 }
2902
2903 /**
2904 * Take a list of strings and build a locale-friendly comma-separated
2905 * list, using the local comma-separator message.
2906 * The last two strings are chained with an "and".
2907 *
2908 * @param $l Array
2909 * @return string
2910 */
2911 function listToText( array $l ) {
2912 $s = '';
2913 $m = count( $l ) - 1;
2914 if ( $m == 1 ) {
2915 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2916 } else {
2917 for ( $i = $m; $i >= 0; $i-- ) {
2918 if ( $i == $m ) {
2919 $s = $l[$i];
2920 } elseif ( $i == $m - 1 ) {
2921 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2922 } else {
2923 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2924 }
2925 }
2926 return $s;
2927 }
2928 }
2929
2930 /**
2931 * Take a list of strings and build a locale-friendly comma-separated
2932 * list, using the local comma-separator message.
2933 * @param $list array of strings to put in a comma list
2934 * @return string
2935 */
2936 function commaList( array $list ) {
2937 return implode(
2938 wfMsgExt(
2939 'comma-separator',
2940 array( 'parsemag', 'escapenoentities', 'language' => $this )
2941 ),
2942 $list
2943 );
2944 }
2945
2946 /**
2947 * Take a list of strings and build a locale-friendly semicolon-separated
2948 * list, using the local semicolon-separator message.
2949 * @param $list array of strings to put in a semicolon list
2950 * @return string
2951 */
2952 function semicolonList( array $list ) {
2953 return implode(
2954 wfMsgExt(
2955 'semicolon-separator',
2956 array( 'parsemag', 'escapenoentities', 'language' => $this )
2957 ),
2958 $list
2959 );
2960 }
2961
2962 /**
2963 * Same as commaList, but separate it with the pipe instead.
2964 * @param $list array of strings to put in a pipe list
2965 * @return string
2966 */
2967 function pipeList( array $list ) {
2968 return implode(
2969 wfMsgExt(
2970 'pipe-separator',
2971 array( 'escapenoentities', 'language' => $this )
2972 ),
2973 $list
2974 );
2975 }
2976
2977 /**
2978 * Truncate a string to a specified length in bytes, appending an optional
2979 * string (e.g. for ellipses)
2980 *
2981 * The database offers limited byte lengths for some columns in the database;
2982 * multi-byte character sets mean we need to ensure that only whole characters
2983 * are included, otherwise broken characters can be passed to the user
2984 *
2985 * If $length is negative, the string will be truncated from the beginning
2986 *
2987 * @param $string String to truncate
2988 * @param $length Int: maximum length (including ellipses)
2989 * @param $ellipsis String to append to the truncated text
2990 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
2991 * $adjustLength was introduced in 1.18, before that behaved as if false.
2992 * @return string
2993 */
2994 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
2995 # Use the localized ellipsis character
2996 if ( $ellipsis == '...' ) {
2997 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2998 }
2999 # Check if there is no need to truncate
3000 if ( $length == 0 ) {
3001 return $ellipsis; // convention
3002 } elseif ( strlen( $string ) <= abs( $length ) ) {
3003 return $string; // no need to truncate
3004 }
3005 $stringOriginal = $string;
3006 # If ellipsis length is >= $length then we can't apply $adjustLength
3007 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3008 $string = $ellipsis; // this can be slightly unexpected
3009 # Otherwise, truncate and add ellipsis...
3010 } else {
3011 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3012 if ( $length > 0 ) {
3013 $length -= $eLength;
3014 $string = substr( $string, 0, $length ); // xyz...
3015 $string = $this->removeBadCharLast( $string );
3016 $string = $string . $ellipsis;
3017 } else {
3018 $length += $eLength;
3019 $string = substr( $string, $length ); // ...xyz
3020 $string = $this->removeBadCharFirst( $string );
3021 $string = $ellipsis . $string;
3022 }
3023 }
3024 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3025 # This check is *not* redundant if $adjustLength, due to the single case where
3026 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3027 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3028 return $string;
3029 } else {
3030 return $stringOriginal;
3031 }
3032 }
3033
3034 /**
3035 * Remove bytes that represent an incomplete Unicode character
3036 * at the end of string (e.g. bytes of the char are missing)
3037 *
3038 * @param $string String
3039 * @return string
3040 */
3041 protected function removeBadCharLast( $string ) {
3042 if ( $string != '' ) {
3043 $char = ord( $string[strlen( $string ) - 1] );
3044 $m = array();
3045 if ( $char >= 0xc0 ) {
3046 # We got the first byte only of a multibyte char; remove it.
3047 $string = substr( $string, 0, -1 );
3048 } elseif ( $char >= 0x80 &&
3049 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3050 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
3051 {
3052 # We chopped in the middle of a character; remove it
3053 $string = $m[1];
3054 }
3055 }
3056 return $string;
3057 }
3058
3059 /**
3060 * Remove bytes that represent an incomplete Unicode character
3061 * at the start of string (e.g. bytes of the char are missing)
3062 *
3063 * @param $string String
3064 * @return string
3065 */
3066 protected function removeBadCharFirst( $string ) {
3067 if ( $string != '' ) {
3068 $char = ord( $string[0] );
3069 if ( $char >= 0x80 && $char < 0xc0 ) {
3070 # We chopped in the middle of a character; remove the whole thing
3071 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3072 }
3073 }
3074 return $string;
3075 }
3076
3077 /**
3078 * Truncate a string of valid HTML to a specified length in bytes,
3079 * appending an optional string (e.g. for ellipses), and return valid HTML
3080 *
3081 * This is only intended for styled/linked text, such as HTML with
3082 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3083 * Also, this will not detect things like "display:none" CSS.
3084 *
3085 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3086 *
3087 * @param string $text HTML string to truncate
3088 * @param int $length (zero/positive) Maximum length (including ellipses)
3089 * @param string $ellipsis String to append to the truncated text
3090 * @return string
3091 */
3092 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3093 # Use the localized ellipsis character
3094 if ( $ellipsis == '...' ) {
3095 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
3096 }
3097 # Check if there is clearly no need to truncate
3098 if ( $length <= 0 ) {
3099 return $ellipsis; // no text shown, nothing to format (convention)
3100 } elseif ( strlen( $text ) <= $length ) {
3101 return $text; // string short enough even *with* HTML (short-circuit)
3102 }
3103
3104 $dispLen = 0; // innerHTML legth so far
3105 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3106 $tagType = 0; // 0-open, 1-close
3107 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3108 $entityState = 0; // 0-not entity, 1-entity
3109 $tag = $ret = ''; // accumulated tag name, accumulated result string
3110 $openTags = array(); // open tag stack
3111 $maybeState = null; // possible truncation state
3112
3113 $textLen = strlen( $text );
3114 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3115 for ( $pos = 0; true; ++$pos ) {
3116 # Consider truncation once the display length has reached the maximim.
3117 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3118 # Check that we're not in the middle of a bracket/entity...
3119 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3120 if ( !$testingEllipsis ) {
3121 $testingEllipsis = true;
3122 # Save where we are; we will truncate here unless there turn out to
3123 # be so few remaining characters that truncation is not necessary.
3124 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3125 $maybeState = array( $ret, $openTags ); // save state
3126 }
3127 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3128 # String in fact does need truncation, the truncation point was OK.
3129 list( $ret, $openTags ) = $maybeState; // reload state
3130 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3131 $ret .= $ellipsis; // add ellipsis
3132 break;
3133 }
3134 }
3135 if ( $pos >= $textLen ) break; // extra iteration just for above checks
3136
3137 # Read the next char...
3138 $ch = $text[$pos];
3139 $lastCh = $pos ? $text[$pos - 1] : '';
3140 $ret .= $ch; // add to result string
3141 if ( $ch == '<' ) {
3142 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3143 $entityState = 0; // for bad HTML
3144 $bracketState = 1; // tag started (checking for backslash)
3145 } elseif ( $ch == '>' ) {
3146 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3147 $entityState = 0; // for bad HTML
3148 $bracketState = 0; // out of brackets
3149 } elseif ( $bracketState == 1 ) {
3150 if ( $ch == '/' ) {
3151 $tagType = 1; // close tag (e.g. "</span>")
3152 } else {
3153 $tagType = 0; // open tag (e.g. "<span>")
3154 $tag .= $ch;
3155 }
3156 $bracketState = 2; // building tag name
3157 } elseif ( $bracketState == 2 ) {
3158 if ( $ch != ' ' ) {
3159 $tag .= $ch;
3160 } else {
3161 // Name found (e.g. "<a href=..."), add on tag attributes...
3162 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3163 }
3164 } elseif ( $bracketState == 0 ) {
3165 if ( $entityState ) {
3166 if ( $ch == ';' ) {
3167 $entityState = 0;
3168 $dispLen++; // entity is one displayed char
3169 }
3170 } else {
3171 if ( $neLength == 0 && !$maybeState ) {
3172 // Save state without $ch. We want to *hit* the first
3173 // display char (to get tags) but not *use* it if truncating.
3174 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3175 }
3176 if ( $ch == '&' ) {
3177 $entityState = 1; // entity found, (e.g. "&#160;")
3178 } else {
3179 $dispLen++; // this char is displayed
3180 // Add the next $max display text chars after this in one swoop...
3181 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3182 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3183 $dispLen += $skipped;
3184 $pos += $skipped;
3185 }
3186 }
3187 }
3188 }
3189 // Close the last tag if left unclosed by bad HTML
3190 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3191 while ( count( $openTags ) > 0 ) {
3192 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3193 }
3194 return $ret;
3195 }
3196
3197 /**
3198 * truncateHtml() helper function
3199 * like strcspn() but adds the skipped chars to $ret
3200 *
3201 * @param $ret
3202 * @param $text
3203 * @param $search
3204 * @param $start
3205 * @param $len
3206 * @return int
3207 */
3208 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3209 if ( $len === null ) {
3210 $len = -1; // -1 means "no limit" for strcspn
3211 } elseif ( $len < 0 ) {
3212 $len = 0; // sanity
3213 }
3214 $skipCount = 0;
3215 if ( $start < strlen( $text ) ) {
3216 $skipCount = strcspn( $text, $search, $start, $len );
3217 $ret .= substr( $text, $start, $skipCount );
3218 }
3219 return $skipCount;
3220 }
3221
3222 /**
3223 * truncateHtml() helper function
3224 * (a) push or pop $tag from $openTags as needed
3225 * (b) clear $tag value
3226 * @param &$tag string Current HTML tag name we are looking at
3227 * @param $tagType int (0-open tag, 1-close tag)
3228 * @param $lastCh string Character before the '>' that ended this tag
3229 * @param &$openTags array Open tag stack (not accounting for $tag)
3230 */
3231 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3232 $tag = ltrim( $tag );
3233 if ( $tag != '' ) {
3234 if ( $tagType == 0 && $lastCh != '/' ) {
3235 $openTags[] = $tag; // tag opened (didn't close itself)
3236 } elseif ( $tagType == 1 ) {
3237 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3238 array_pop( $openTags ); // tag closed
3239 }
3240 }
3241 $tag = '';
3242 }
3243 }
3244
3245 /**
3246 * Grammatical transformations, needed for inflected languages
3247 * Invoked by putting {{grammar:case|word}} in a message
3248 *
3249 * @param $word string
3250 * @param $case string
3251 * @return string
3252 */
3253 function convertGrammar( $word, $case ) {
3254 global $wgGrammarForms;
3255 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3256 return $wgGrammarForms[$this->getCode()][$case][$word];
3257 }
3258 return $word;
3259 }
3260
3261 /**
3262 * Provides an alternative text depending on specified gender.
3263 * Usage {{gender:username|masculine|feminine|neutral}}.
3264 * username is optional, in which case the gender of current user is used,
3265 * but only in (some) interface messages; otherwise default gender is used.
3266 *
3267 * If no forms are given, an empty string is returned. If only one form is
3268 * given, it will be returned unconditionally. These details are implied by
3269 * the caller and cannot be overridden in subclasses.
3270 *
3271 * If more than one form is given, the default is to use the neutral one
3272 * if it is specified, and to use the masculine one otherwise. These
3273 * details can be overridden in subclasses.
3274 *
3275 * @param $gender string
3276 * @param $forms array
3277 *
3278 * @return string
3279 */
3280 function gender( $gender, $forms ) {
3281 if ( !count( $forms ) ) {
3282 return '';
3283 }
3284 $forms = $this->preConvertPlural( $forms, 2 );
3285 if ( $gender === 'male' ) {
3286 return $forms[0];
3287 }
3288 if ( $gender === 'female' ) {
3289 return $forms[1];
3290 }
3291 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3292 }
3293
3294 /**
3295 * Plural form transformations, needed for some languages.
3296 * For example, there are 3 form of plural in Russian and Polish,
3297 * depending on "count mod 10". See [[w:Plural]]
3298 * For English it is pretty simple.
3299 *
3300 * Invoked by putting {{plural:count|wordform1|wordform2}}
3301 * or {{plural:count|wordform1|wordform2|wordform3}}
3302 *
3303 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3304 *
3305 * @param $count Integer: non-localized number
3306 * @param $forms Array: different plural forms
3307 * @return string Correct form of plural for $count in this language
3308 */
3309 function convertPlural( $count, $forms ) {
3310 if ( !count( $forms ) ) {
3311 return '';
3312 }
3313 $forms = $this->preConvertPlural( $forms, 2 );
3314
3315 return ( $count == 1 ) ? $forms[0] : $forms[1];
3316 }
3317
3318 /**
3319 * Checks that convertPlural was given an array and pads it to requested
3320 * amount of forms by copying the last one.
3321 *
3322 * @param $count Integer: How many forms should there be at least
3323 * @param $forms Array of forms given to convertPlural
3324 * @return array Padded array of forms or an exception if not an array
3325 */
3326 protected function preConvertPlural( /* Array */ $forms, $count ) {
3327 while ( count( $forms ) < $count ) {
3328 $forms[] = $forms[count( $forms ) - 1];
3329 }
3330 return $forms;
3331 }
3332
3333 /**
3334 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3335 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3336 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3337 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3338 * match up with it.
3339 *
3340 * @param $str String: the validated block duration in English
3341 * @return string Somehow translated block duration
3342 * @see LanguageFi.php for example implementation
3343 */
3344 function translateBlockExpiry( $str ) {
3345 $duration = SpecialBlock::getSuggestedDurations( $this );
3346 foreach ( $duration as $show => $value ) {
3347 if ( strcmp( $str, $value ) == 0 ) {
3348 return htmlspecialchars( trim( $show ) );
3349 }
3350 }
3351
3352 // Since usually only infinite or indefinite is only on list, so try
3353 // equivalents if still here.
3354 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3355 if ( in_array( $str, $indefs ) ) {
3356 foreach ( $indefs as $val ) {
3357 $show = array_search( $val, $duration, true );
3358 if ( $show !== false ) {
3359 return htmlspecialchars( trim( $show ) );
3360 }
3361 }
3362 }
3363 // If all else fails, return the original string.
3364 return $str;
3365 }
3366
3367 /**
3368 * languages like Chinese need to be segmented in order for the diff
3369 * to be of any use
3370 *
3371 * @param $text String
3372 * @return String
3373 */
3374 public function segmentForDiff( $text ) {
3375 return $text;
3376 }
3377
3378 /**
3379 * and unsegment to show the result
3380 *
3381 * @param $text String
3382 * @return String
3383 */
3384 public function unsegmentForDiff( $text ) {
3385 return $text;
3386 }
3387
3388 /**
3389 * Return the LanguageConverter used in the Language
3390 *
3391 * @since 1.19
3392 * @return LanguageConverter
3393 */
3394 public function getConverter() {
3395 return $this->mConverter;
3396 }
3397
3398 /**
3399 * convert text to all supported variants
3400 *
3401 * @param $text string
3402 * @return array
3403 */
3404 public function autoConvertToAllVariants( $text ) {
3405 return $this->mConverter->autoConvertToAllVariants( $text );
3406 }
3407
3408 /**
3409 * convert text to different variants of a language.
3410 *
3411 * @param $text string
3412 * @return string
3413 */
3414 public function convert( $text ) {
3415 return $this->mConverter->convert( $text );
3416 }
3417
3418 /**
3419 * Convert a Title object to a string in the preferred variant
3420 *
3421 * @param $title Title
3422 * @return string
3423 */
3424 public function convertTitle( $title ) {
3425 return $this->mConverter->convertTitle( $title );
3426 }
3427
3428 /**
3429 * Check if this is a language with variants
3430 *
3431 * @return bool
3432 */
3433 public function hasVariants() {
3434 return sizeof( $this->getVariants() ) > 1;
3435 }
3436
3437 /**
3438 * Check if the language has the specific variant
3439 *
3440 * @since 1.19
3441 * @param $variant string
3442 * @return bool
3443 */
3444 public function hasVariant( $variant ) {
3445 return (bool)$this->mConverter->validateVariant( $variant );
3446 }
3447
3448 /**
3449 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3450 *
3451 * @param $text string
3452 * @return string
3453 */
3454 public function armourMath( $text ) {
3455 return $this->mConverter->armourMath( $text );
3456 }
3457
3458 /**
3459 * Perform output conversion on a string, and encode for safe HTML output.
3460 * @param $text String text to be converted
3461 * @param $isTitle Bool whether this conversion is for the article title
3462 * @return string
3463 * @todo this should get integrated somewhere sane
3464 */
3465 public function convertHtml( $text, $isTitle = false ) {
3466 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3467 }
3468
3469 /**
3470 * @param $key string
3471 * @return string
3472 */
3473 public function convertCategoryKey( $key ) {
3474 return $this->mConverter->convertCategoryKey( $key );
3475 }
3476
3477 /**
3478 * Get the list of variants supported by this language
3479 * see sample implementation in LanguageZh.php
3480 *
3481 * @return array an array of language codes
3482 */
3483 public function getVariants() {
3484 return $this->mConverter->getVariants();
3485 }
3486
3487 /**
3488 * @return string
3489 */
3490 public function getPreferredVariant() {
3491 return $this->mConverter->getPreferredVariant();
3492 }
3493
3494 /**
3495 * @return string
3496 */
3497 public function getDefaultVariant() {
3498 return $this->mConverter->getDefaultVariant();
3499 }
3500
3501 /**
3502 * @return string
3503 */
3504 public function getURLVariant() {
3505 return $this->mConverter->getURLVariant();
3506 }
3507
3508 /**
3509 * If a language supports multiple variants, it is
3510 * possible that non-existing link in one variant
3511 * actually exists in another variant. this function
3512 * tries to find it. See e.g. LanguageZh.php
3513 *
3514 * @param $link String: the name of the link
3515 * @param $nt Mixed: the title object of the link
3516 * @param $ignoreOtherCond Boolean: to disable other conditions when
3517 * we need to transclude a template or update a category's link
3518 * @return null the input parameters may be modified upon return
3519 */
3520 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3521 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
3522 }
3523
3524 /**
3525 * If a language supports multiple variants, converts text
3526 * into an array of all possible variants of the text:
3527 * 'variant' => text in that variant
3528 *
3529 * @deprecated since 1.17 Use autoConvertToAllVariants()
3530 *
3531 * @param $text string
3532 *
3533 * @return string
3534 */
3535 public function convertLinkToAllVariants( $text ) {
3536 return $this->mConverter->convertLinkToAllVariants( $text );
3537 }
3538
3539 /**
3540 * returns language specific options used by User::getPageRenderHash()
3541 * for example, the preferred language variant
3542 *
3543 * @return string
3544 */
3545 function getExtraHashOptions() {
3546 return $this->mConverter->getExtraHashOptions();
3547 }
3548
3549 /**
3550 * For languages that support multiple variants, the title of an
3551 * article may be displayed differently in different variants. this
3552 * function returns the apporiate title defined in the body of the article.
3553 *
3554 * @return string
3555 */
3556 public function getParsedTitle() {
3557 return $this->mConverter->getParsedTitle();
3558 }
3559
3560 /**
3561 * Enclose a string with the "no conversion" tag. This is used by
3562 * various functions in the Parser
3563 *
3564 * @param $text String: text to be tagged for no conversion
3565 * @param $noParse bool
3566 * @return string the tagged text
3567 */
3568 public function markNoConversion( $text, $noParse = false ) {
3569 return $this->mConverter->markNoConversion( $text, $noParse );
3570 }
3571
3572 /**
3573 * A regular expression to match legal word-trailing characters
3574 * which should be merged onto a link of the form [[foo]]bar.
3575 *
3576 * @return string
3577 */
3578 public function linkTrail() {
3579 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
3580 }
3581
3582 /**
3583 * @return Language
3584 */
3585 function getLangObj() {
3586 return $this;
3587 }
3588
3589 /**
3590 * Get the RFC 3066 code for this language object
3591 *
3592 * @return string
3593 */
3594 public function getCode() {
3595 return $this->mCode;
3596 }
3597
3598 /**
3599 * Get the code in Bcp47 format which we can use
3600 * inside of html lang="" tags.
3601 * @since 1.19
3602 * @return string
3603 */
3604 public function getHtmlCode() {
3605 if ( is_null( $this->mHtmlCode ) ) {
3606 $this->mHtmlCode = wfBCP47( $this->getCode() );
3607 }
3608 return $this->mHtmlCode;
3609 }
3610
3611 /**
3612 * @param $code string
3613 */
3614 public function setCode( $code ) {
3615 $this->mCode = $code;
3616 // Ensure we don't leave an incorrect html code lying around
3617 $this->mHtmlCode = null;
3618 }
3619
3620 /**
3621 * Get the name of a file for a certain language code
3622 * @param $prefix string Prepend this to the filename
3623 * @param $code string Language code
3624 * @param $suffix string Append this to the filename
3625 * @throws MWException
3626 * @return string $prefix . $mangledCode . $suffix
3627 */
3628 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
3629 // Protect against path traversal
3630 if ( !Language::isValidCode( $code )
3631 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
3632 {
3633 throw new MWException( "Invalid language code \"$code\"" );
3634 }
3635
3636 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
3637 }
3638
3639 /**
3640 * Get the language code from a file name. Inverse of getFileName()
3641 * @param $filename string $prefix . $languageCode . $suffix
3642 * @param $prefix string Prefix before the language code
3643 * @param $suffix string Suffix after the language code
3644 * @return string Language code, or false if $prefix or $suffix isn't found
3645 */
3646 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
3647 $m = null;
3648 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
3649 preg_quote( $suffix, '/' ) . '/', $filename, $m );
3650 if ( !count( $m ) ) {
3651 return false;
3652 }
3653 return str_replace( '_', '-', strtolower( $m[1] ) );
3654 }
3655
3656 /**
3657 * @param $code string
3658 * @return string
3659 */
3660 public static function getMessagesFileName( $code ) {
3661 global $IP;
3662 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
3663 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
3664 return $file;
3665 }
3666
3667 /**
3668 * @param $code string
3669 * @return string
3670 */
3671 public static function getClassFileName( $code ) {
3672 global $IP;
3673 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
3674 }
3675
3676 /**
3677 * Get the first fallback for a given language.
3678 *
3679 * @param $code string
3680 *
3681 * @return bool|string
3682 */
3683 public static function getFallbackFor( $code ) {
3684 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
3685 return false;
3686 } else {
3687 $fallbacks = self::getFallbacksFor( $code );
3688 $first = array_shift( $fallbacks );
3689 return $first;
3690 }
3691 }
3692
3693 /**
3694 * Get the ordered list of fallback languages.
3695 *
3696 * @since 1.19
3697 * @param $code string Language code
3698 * @return array
3699 */
3700 public static function getFallbacksFor( $code ) {
3701 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
3702 return array();
3703 } else {
3704 $v = self::getLocalisationCache()->getItem( $code, 'fallback' );
3705 $v = array_map( 'trim', explode( ',', $v ) );
3706 if ( $v[count( $v ) - 1] !== 'en' ) {
3707 $v[] = 'en';
3708 }
3709 return $v;
3710 }
3711 }
3712
3713 /**
3714 * Get all messages for a given language
3715 * WARNING: this may take a long time. If you just need all message *keys*
3716 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
3717 *
3718 * @param $code string
3719 *
3720 * @return array
3721 */
3722 public static function getMessagesFor( $code ) {
3723 return self::getLocalisationCache()->getItem( $code, 'messages' );
3724 }
3725
3726 /**
3727 * Get a message for a given language
3728 *
3729 * @param $key string
3730 * @param $code string
3731 *
3732 * @return string
3733 */
3734 public static function getMessageFor( $key, $code ) {
3735 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
3736 }
3737
3738 /**
3739 * Get all message keys for a given language. This is a faster alternative to
3740 * array_keys( Language::getMessagesFor( $code ) )
3741 *
3742 * @since 1.19
3743 * @param $code string Language code
3744 * @return array of message keys (strings)
3745 */
3746 public static function getMessageKeysFor( $code ) {
3747 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
3748 }
3749
3750 /**
3751 * @param $talk
3752 * @return mixed
3753 */
3754 function fixVariableInNamespace( $talk ) {
3755 if ( strpos( $talk, '$1' ) === false ) {
3756 return $talk;
3757 }
3758
3759 global $wgMetaNamespace;
3760 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
3761
3762 # Allow grammar transformations
3763 # Allowing full message-style parsing would make simple requests
3764 # such as action=raw much more expensive than they need to be.
3765 # This will hopefully cover most cases.
3766 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
3767 array( &$this, 'replaceGrammarInNamespace' ), $talk );
3768 return str_replace( ' ', '_', $talk );
3769 }
3770
3771 /**
3772 * @param $m string
3773 * @return string
3774 */
3775 function replaceGrammarInNamespace( $m ) {
3776 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
3777 }
3778
3779 /**
3780 * @throws MWException
3781 * @return array
3782 */
3783 static function getCaseMaps() {
3784 static $wikiUpperChars, $wikiLowerChars;
3785 if ( isset( $wikiUpperChars ) ) {
3786 return array( $wikiUpperChars, $wikiLowerChars );
3787 }
3788
3789 wfProfileIn( __METHOD__ );
3790 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
3791 if ( $arr === false ) {
3792 throw new MWException(
3793 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
3794 }
3795 $wikiUpperChars = $arr['wikiUpperChars'];
3796 $wikiLowerChars = $arr['wikiLowerChars'];
3797 wfProfileOut( __METHOD__ );
3798 return array( $wikiUpperChars, $wikiLowerChars );
3799 }
3800
3801 /**
3802 * Decode an expiry (block, protection, etc) which has come from the DB
3803 *
3804 * @FIXME: why are we returnings DBMS-dependent strings???
3805 *
3806 * @param $expiry String: Database expiry String
3807 * @param $format Bool|Int true to process using language functions, or TS_ constant
3808 * to return the expiry in a given timestamp
3809 * @return String
3810 */
3811 public function formatExpiry( $expiry, $format = true ) {
3812 static $infinity, $infinityMsg;
3813 if ( $infinity === null ) {
3814 $infinityMsg = wfMessage( 'infiniteblock' );
3815 $infinity = wfGetDB( DB_SLAVE )->getInfinity();
3816 }
3817
3818 if ( $expiry == '' || $expiry == $infinity ) {
3819 return $format === true
3820 ? $infinityMsg
3821 : $infinity;
3822 } else {
3823 return $format === true
3824 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
3825 : wfTimestamp( $format, $expiry );
3826 }
3827 }
3828
3829 /**
3830 * @todo Document
3831 * @param $seconds int|float
3832 * @param $format Array Optional
3833 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
3834 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
3835 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
3836 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
3837 * @return string
3838 */
3839 function formatTimePeriod( $seconds, $format = array() ) {
3840 if ( !is_array( $format ) ) {
3841 $format = array( 'avoid' => $format ); // For backwards compatibility
3842 }
3843 if ( !isset( $format['avoid'] ) ) {
3844 $format['avoid'] = false;
3845 }
3846 if ( !isset( $format['noabbrevs' ] ) ) {
3847 $format['noabbrevs'] = false;
3848 }
3849 $secondsMsg = wfMessage(
3850 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
3851 $minutesMsg = wfMessage(
3852 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
3853 $hoursMsg = wfMessage(
3854 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
3855 $daysMsg = wfMessage(
3856 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
3857
3858 if ( round( $seconds * 10 ) < 100 ) {
3859 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
3860 $s = $secondsMsg->params( $s )->text();
3861 } elseif ( round( $seconds ) < 60 ) {
3862 $s = $this->formatNum( round( $seconds ) );
3863 $s = $secondsMsg->params( $s )->text();
3864 } elseif ( round( $seconds ) < 3600 ) {
3865 $minutes = floor( $seconds / 60 );
3866 $secondsPart = round( fmod( $seconds, 60 ) );
3867 if ( $secondsPart == 60 ) {
3868 $secondsPart = 0;
3869 $minutes++;
3870 }
3871 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3872 $s .= ' ';
3873 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3874 } elseif ( round( $seconds ) <= 2 * 86400 ) {
3875 $hours = floor( $seconds / 3600 );
3876 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
3877 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
3878 if ( $secondsPart == 60 ) {
3879 $secondsPart = 0;
3880 $minutes++;
3881 }
3882 if ( $minutes == 60 ) {
3883 $minutes = 0;
3884 $hours++;
3885 }
3886 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
3887 $s .= ' ';
3888 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3889 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
3890 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
3891 }
3892 } else {
3893 $days = floor( $seconds / 86400 );
3894 if ( $format['avoid'] === 'avoidminutes' ) {
3895 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
3896 if ( $hours == 24 ) {
3897 $hours = 0;
3898 $days++;
3899 }
3900 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3901 $s .= ' ';
3902 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
3903 } elseif ( $format['avoid'] === 'avoidseconds' ) {
3904 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
3905 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
3906 if ( $minutes == 60 ) {
3907 $minutes = 0;
3908 $hours++;
3909 }
3910 if ( $hours == 24 ) {
3911 $hours = 0;
3912 $days++;
3913 }
3914 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3915 $s .= ' ';
3916 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
3917 $s .= ' ';
3918 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
3919 } else {
3920 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
3921 $s .= ' ';
3922 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
3923 }
3924 }
3925 return $s;
3926 }
3927
3928 /**
3929 * Format a bitrate for output, using an appropriate
3930 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
3931 *
3932 * This use base 1000. For base 1024 use formatSize(), for another base
3933 * see formatComputingNumbers()
3934 *
3935 * @param $bps int
3936 * @return string
3937 */
3938 function formatBitrate( $bps ) {
3939 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
3940 }
3941
3942 /**
3943 * @param $size int Size of the unit
3944 * @param $boundary int Size boundary (1000, or 1024 in most cases)
3945 * @param $messageKey string Message key to be uesd
3946 * @return string
3947 */
3948 function formatComputingNumbers( $size, $boundary, $messageKey ) {
3949 if ( $size <= 0 ) {
3950 return str_replace( '$1', $this->formatNum( $size ),
3951 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
3952 );
3953 }
3954 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
3955 $index = 0;
3956
3957 $maxIndex = count( $sizes ) - 1;
3958 while ( $size >= $boundary && $index < $maxIndex ) {
3959 $index++;
3960 $size /= $boundary;
3961 }
3962
3963 // For small sizes no decimal places necessary
3964 $round = 0;
3965 if ( $index > 1 ) {
3966 // For MB and bigger two decimal places are smarter
3967 $round = 2;
3968 }
3969 $msg = str_replace( '$1', $sizes[$index], $messageKey );
3970
3971 $size = round( $size, $round );
3972 $text = $this->getMessageFromDB( $msg );
3973 return str_replace( '$1', $this->formatNum( $size ), $text );
3974 }
3975
3976 /**
3977 * Format a size in bytes for output, using an appropriate
3978 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
3979 *
3980 * This method use base 1024. For base 1000 use formatBitrate(), for
3981 * another base see formatComputingNumbers()
3982 *
3983 * @param $size int Size to format
3984 * @return string Plain text (not HTML)
3985 */
3986 function formatSize( $size ) {
3987 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
3988 }
3989
3990 /**
3991 * Make a list item, used by various special pages
3992 *
3993 * @param $page String Page link
3994 * @param $details String Text between brackets
3995 * @param $oppositedm Boolean Add the direction mark opposite to your
3996 * language, to display text properly
3997 * @return String
3998 */
3999 function specialList( $page, $details, $oppositedm = true ) {
4000 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) .
4001 $this->getDirMark();
4002 $details = $details ? $dirmark . $this->getMessageFromDB( 'word-separator' ) .
4003 wfMsgExt( 'parentheses', array( 'escape', 'replaceafter', 'language' => $this ), $details ) : '';
4004 return $page . $details;
4005 }
4006
4007 /**
4008 * Generate (prev x| next x) (20|50|100...) type links for paging
4009 *
4010 * @param $title Title object to link
4011 * @param $offset Integer offset parameter
4012 * @param $limit Integer limit parameter
4013 * @param $query String optional URL query parameter string
4014 * @param $atend Bool optional param for specified if this is the last page
4015 * @return String
4016 */
4017 public function viewPrevNext( Title $title, $offset, $limit, array $query = array(), $atend = false ) {
4018 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4019
4020 # Make 'previous' link
4021 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4022 if ( $offset > 0 ) {
4023 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4024 $query, $prev, 'prevn-title', 'mw-prevlink' );
4025 } else {
4026 $plink = htmlspecialchars( $prev );
4027 }
4028
4029 # Make 'next' link
4030 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4031 if ( $atend ) {
4032 $nlink = htmlspecialchars( $next );
4033 } else {
4034 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4035 $query, $next, 'prevn-title', 'mw-nextlink' );
4036 }
4037
4038 # Make links to set number of items per page
4039 $numLinks = array();
4040 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4041 $numLinks[] = $this->numLink( $title, $offset, $num,
4042 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4043 }
4044
4045 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4046 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4047 }
4048
4049 /**
4050 * Helper function for viewPrevNext() that generates links
4051 *
4052 * @param $title Title object to link
4053 * @param $offset Integer offset parameter
4054 * @param $limit Integer limit parameter
4055 * @param $query Array extra query parameters
4056 * @param $link String text to use for the link; will be escaped
4057 * @param $tooltipMsg String name of the message to use as tooltip
4058 * @param $class String value of the "class" attribute of the link
4059 * @return String HTML fragment
4060 */
4061 private function numLink( Title $title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4062 $query = array( 'limit' => $limit, 'offset' => $offset ) + $query;
4063 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4064 return Html::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4065 'title' => $tooltip, 'class' => $class ), $link );
4066 }
4067
4068 /**
4069 * Get the conversion rule title, if any.
4070 *
4071 * @return string
4072 */
4073 public function getConvRuleTitle() {
4074 return $this->mConverter->getConvRuleTitle();
4075 }
4076 }