dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Merge "Moved MessageBlobStore to /cache"
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * Internationalisation code.
4 *
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
19 *
20 * @file
21 * @ingroup Language
22 */
23
24 /**
25 * @defgroup Language Language
26 */
27
28 if ( !defined( 'MEDIAWIKI' ) ) {
29 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
30 exit( 1 );
31 }
32
33 if ( function_exists( 'mb_strtoupper' ) ) {
34 mb_internal_encoding( 'UTF-8' );
35 }
36
37 /**
38 * Internationalisation code
39 * @ingroup Language
40 */
41 class Language {
42 /**
43 * @var LanguageConverter
44 */
45 public $mConverter;
46
47 public $mVariants, $mCode, $mLoaded = false;
48 public $mMagicExtensions = array(), $mMagicHookDone = false;
49 private $mHtmlCode = null, $mParentLanguage = false;
50
51 public $dateFormatStrings = array();
52 public $mExtendedSpecialPageAliases;
53
54 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
55
56 /**
57 * ReplacementArray object caches
58 */
59 public $transformData = array();
60
61 /**
62 * @var LocalisationCache
63 */
64 static public $dataCache;
65
66 static public $mLangObjCache = array();
67
68 static public $mWeekdayMsgs = array(
69 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
70 'friday', 'saturday'
71 );
72
73 static public $mWeekdayAbbrevMsgs = array(
74 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
75 );
76
77 static public $mMonthMsgs = array(
78 'january', 'february', 'march', 'april', 'may_long', 'june',
79 'july', 'august', 'september', 'october', 'november',
80 'december'
81 );
82 static public $mMonthGenMsgs = array(
83 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
84 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
85 'december-gen'
86 );
87 static public $mMonthAbbrevMsgs = array(
88 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
89 'sep', 'oct', 'nov', 'dec'
90 );
91
92 static public $mIranianCalendarMonthMsgs = array(
93 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
94 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
95 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
96 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
97 );
98
99 static public $mHebrewCalendarMonthMsgs = array(
100 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
101 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
102 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
103 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
104 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
105 );
106
107 static public $mHebrewCalendarMonthGenMsgs = array(
108 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
109 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
110 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
111 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
112 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
113 );
114
115 static public $mHijriCalendarMonthMsgs = array(
116 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
117 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
118 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
119 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
120 );
121
122 /**
123 * @since 1.20
124 * @var array
125 */
126 static public $durationIntervals = array(
127 'millennia' => 31556952000,
128 'centuries' => 3155695200,
129 'decades' => 315569520,
130 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
131 'weeks' => 604800,
132 'days' => 86400,
133 'hours' => 3600,
134 'minutes' => 60,
135 'seconds' => 1,
136 );
137
138 /**
139 * Cache for language fallbacks.
140 * @see Language::getFallbacksIncludingSiteLanguage
141 * @since 1.21
142 * @var array
143 */
144 static private $fallbackLanguageCache = array();
145
146 /**
147 * Cache for language names
148 * @var MapCacheLRU|null
149 */
150 static private $languageNameCache;
151
152 /**
153 * Get a cached or new language object for a given language code
154 * @param string $code
155 * @return Language
156 */
157 static function factory( $code ) {
158 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
159
160 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
161 $code = $wgDummyLanguageCodes[$code];
162 }
163
164 // get the language object to process
165 $langObj = isset( self::$mLangObjCache[$code] )
166 ? self::$mLangObjCache[$code]
167 : self::newFromCode( $code );
168
169 // merge the language object in to get it up front in the cache
170 self::$mLangObjCache = array_merge( array( $code => $langObj ), self::$mLangObjCache );
171 // get rid of the oldest ones in case we have an overflow
172 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
173
174 return $langObj;
175 }
176
177 /**
178 * Create a language object for a given language code
179 * @param string $code
180 * @throws MWException
181 * @return Language
182 */
183 protected static function newFromCode( $code ) {
184 // Protect against path traversal below
185 if ( !Language::isValidCode( $code )
186 || strcspn( $code, ":/\\\000" ) !== strlen( $code )
187 ) {
188 throw new MWException( "Invalid language code \"$code\"" );
189 }
190
191 if ( !Language::isValidBuiltInCode( $code ) ) {
192 // It's not possible to customise this code with class files, so
193 // just return a Language object. This is to support uselang= hacks.
194 $lang = new Language;
195 $lang->setCode( $code );
196 return $lang;
197 }
198
199 // Check if there is a language class for the code
200 $class = self::classFromCode( $code );
201 self::preloadLanguageClass( $class );
202 if ( class_exists( $class ) ) {
203 $lang = new $class;
204 return $lang;
205 }
206
207 // Keep trying the fallback list until we find an existing class
208 $fallbacks = Language::getFallbacksFor( $code );
209 foreach ( $fallbacks as $fallbackCode ) {
210 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
211 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
212 }
213
214 $class = self::classFromCode( $fallbackCode );
215 self::preloadLanguageClass( $class );
216 if ( class_exists( $class ) ) {
217 $lang = Language::newFromCode( $fallbackCode );
218 $lang->setCode( $code );
219 return $lang;
220 }
221 }
222
223 throw new MWException( "Invalid fallback sequence for language '$code'" );
224 }
225
226 /**
227 * Checks whether any localisation is available for that language tag
228 * in MediaWiki (MessagesXx.php exists).
229 *
230 * @param string $code Language tag (in lower case)
231 * @return bool Whether language is supported
232 * @since 1.21
233 */
234 public static function isSupportedLanguage( $code ) {
235 return self::isValidBuiltInCode( $code )
236 && ( is_readable( self::getMessagesFileName( $code ) )
237 || is_readable( self::getJsonMessagesFileName( $code ) )
238 );
239 }
240
241 /**
242 * Returns true if a language code string is a well-formed language tag
243 * according to RFC 5646.
244 * This function only checks well-formedness; it doesn't check that
245 * language, script or variant codes actually exist in the repositories.
246 *
247 * Based on regexes by Mark Davis of the Unicode Consortium:
248 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
249 *
250 * @param string $code
251 * @param bool $lenient Whether to allow '_' as separator. The default is only '-'.
252 *
253 * @return bool
254 * @since 1.21
255 */
256 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
257 $alpha = '[a-z]';
258 $digit = '[0-9]';
259 $alphanum = '[a-z0-9]';
260 $x = 'x'; # private use singleton
261 $singleton = '[a-wy-z]'; # other singleton
262 $s = $lenient ? '[-_]' : '-';
263
264 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
265 $script = "$alpha{4}"; # ISO 15924
266 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
267 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
268 $extension = "$singleton(?:$s$alphanum{2,8})+";
269 $privateUse = "$x(?:$s$alphanum{1,8})+";
270
271 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
272 # Since these are limited, this is safe even later changes to the registry --
273 # the only oddity is that it might change the type of the tag, and thus
274 # the results from the capturing groups.
275 # http://www.iana.org/assignments/language-subtag-registry
276
277 $grandfathered = "en{$s}GB{$s}oed"
278 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
279 . "|no{$s}(?:bok|nyn)"
280 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
281 . "|zh{$s}min{$s}nan";
282
283 $variantList = "$variant(?:$s$variant)*";
284 $extensionList = "$extension(?:$s$extension)*";
285
286 $langtag = "(?:($language)"
287 . "(?:$s$script)?"
288 . "(?:$s$region)?"
289 . "(?:$s$variantList)?"
290 . "(?:$s$extensionList)?"
291 . "(?:$s$privateUse)?)";
292
293 # The final breakdown, with capturing groups for each of these components
294 # The variants, extensions, grandfathered, and private-use may have interior '-'
295
296 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
297
298 return (bool)preg_match( "/$root/", strtolower( $code ) );
299 }
300
301 /**
302 * Returns true if a language code string is of a valid form, whether or
303 * not it exists. This includes codes which are used solely for
304 * customisation via the MediaWiki namespace.
305 *
306 * @param string $code
307 *
308 * @return bool
309 */
310 public static function isValidCode( $code ) {
311 static $cache = array();
312 if ( isset( $cache[$code] ) ) {
313 return $cache[$code];
314 }
315 // People think language codes are html safe, so enforce it.
316 // Ideally we should only allow a-zA-Z0-9-
317 // but, .+ and other chars are often used for {{int:}} hacks
318 // see bugs 37564, 37587, 36938
319 $cache[$code] =
320 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
321 && !preg_match( MediaWikiTitleCodec::getTitleInvalidRegex(), $code );
322
323 return $cache[$code];
324 }
325
326 /**
327 * Returns true if a language code is of a valid form for the purposes of
328 * internal customisation of MediaWiki, via Messages*.php or *.json.
329 *
330 * @param string $code
331 *
332 * @throws MWException
333 * @since 1.18
334 * @return bool
335 */
336 public static function isValidBuiltInCode( $code ) {
337
338 if ( !is_string( $code ) ) {
339 if ( is_object( $code ) ) {
340 $addmsg = " of class " . get_class( $code );
341 } else {
342 $addmsg = '';
343 }
344 $type = gettype( $code );
345 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
346 }
347
348 return (bool)preg_match( '/^[a-z0-9-]{2,}$/', $code );
349 }
350
351 /**
352 * Returns true if a language code is an IETF tag known to MediaWiki.
353 *
354 * @param string $tag
355 *
356 * @since 1.21
357 * @return bool
358 */
359 public static function isKnownLanguageTag( $tag ) {
360 static $coreLanguageNames;
361
362 // Quick escape for invalid input to avoid exceptions down the line
363 // when code tries to process tags which are not valid at all.
364 if ( !self::isValidBuiltInCode( $tag ) ) {
365 return false;
366 }
367
368 if ( $coreLanguageNames === null ) {
369 global $IP;
370 include "$IP/languages/Names.php";
371 }
372
373 if ( isset( $coreLanguageNames[$tag] )
374 || self::fetchLanguageName( $tag, $tag ) !== ''
375 ) {
376 return true;
377 }
378
379 return false;
380 }
381
382 /**
383 * @param string $code
384 * @return string Name of the language class
385 */
386 public static function classFromCode( $code ) {
387 if ( $code == 'en' ) {
388 return 'Language';
389 } else {
390 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
391 }
392 }
393
394 /**
395 * Includes language class files
396 *
397 * @param string $class Name of the language class
398 */
399 public static function preloadLanguageClass( $class ) {
400 global $IP;
401
402 if ( $class === 'Language' ) {
403 return;
404 }
405
406 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
407 include_once "$IP/languages/classes/$class.php";
408 }
409 }
410
411 /**
412 * Get the LocalisationCache instance
413 *
414 * @return LocalisationCache
415 */
416 public static function getLocalisationCache() {
417 if ( is_null( self::$dataCache ) ) {
418 global $wgLocalisationCacheConf;
419 $class = $wgLocalisationCacheConf['class'];
420 self::$dataCache = new $class( $wgLocalisationCacheConf );
421 }
422 return self::$dataCache;
423 }
424
425 function __construct() {
426 $this->mConverter = new FakeConverter( $this );
427 // Set the code to the name of the descendant
428 if ( get_class( $this ) == 'Language' ) {
429 $this->mCode = 'en';
430 } else {
431 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
432 }
433 self::getLocalisationCache();
434 }
435
436 /**
437 * Reduce memory usage
438 */
439 function __destruct() {
440 foreach ( $this as $name => $value ) {
441 unset( $this->$name );
442 }
443 }
444
445 /**
446 * Hook which will be called if this is the content language.
447 * Descendants can use this to register hook functions or modify globals
448 */
449 function initContLang() {
450 }
451
452 /**
453 * @return array
454 * @since 1.19
455 */
456 function getFallbackLanguages() {
457 return self::getFallbacksFor( $this->mCode );
458 }
459
460 /**
461 * Exports $wgBookstoreListEn
462 * @return array
463 */
464 function getBookstoreList() {
465 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
466 }
467
468 /**
469 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
470 * available in localised form, it will be included in English.
471 *
472 * @return array
473 */
474 public function getNamespaces() {
475 if ( is_null( $this->namespaceNames ) ) {
476 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
477
478 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
479 $validNamespaces = MWNamespace::getCanonicalNamespaces();
480
481 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
482
483 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
484 if ( $wgMetaNamespaceTalk ) {
485 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
486 } else {
487 $talk = $this->namespaceNames[NS_PROJECT_TALK];
488 $this->namespaceNames[NS_PROJECT_TALK] =
489 $this->fixVariableInNamespace( $talk );
490 }
491
492 # Sometimes a language will be localised but not actually exist on this wiki.
493 foreach ( $this->namespaceNames as $key => $text ) {
494 if ( !isset( $validNamespaces[$key] ) ) {
495 unset( $this->namespaceNames[$key] );
496 }
497 }
498
499 # The above mixing may leave namespaces out of canonical order.
500 # Re-order by namespace ID number...
501 ksort( $this->namespaceNames );
502
503 Hooks::run( 'LanguageGetNamespaces', array( &$this->namespaceNames ) );
504 }
505
506 return $this->namespaceNames;
507 }
508
509 /**
510 * Arbitrarily set all of the namespace names at once. Mainly used for testing
511 * @param array $namespaces Array of namespaces (id => name)
512 */
513 public function setNamespaces( array $namespaces ) {
514 $this->namespaceNames = $namespaces;
515 $this->mNamespaceIds = null;
516 }
517
518 /**
519 * Resets all of the namespace caches. Mainly used for testing
520 */
521 public function resetNamespaces() {
522 $this->namespaceNames = null;
523 $this->mNamespaceIds = null;
524 $this->namespaceAliases = null;
525 }
526
527 /**
528 * A convenience function that returns the same thing as
529 * getNamespaces() except with the array values changed to ' '
530 * where it found '_', useful for producing output to be displayed
531 * e.g. in <select> forms.
532 *
533 * @return array
534 */
535 function getFormattedNamespaces() {
536 $ns = $this->getNamespaces();
537 foreach ( $ns as $k => $v ) {
538 $ns[$k] = strtr( $v, '_', ' ' );
539 }
540 return $ns;
541 }
542
543 /**
544 * Get a namespace value by key
545 * <code>
546 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
547 * echo $mw_ns; // prints 'MediaWiki'
548 * </code>
549 *
550 * @param int $index The array key of the namespace to return
551 * @return string|bool String if the namespace value exists, otherwise false
552 */
553 function getNsText( $index ) {
554 $ns = $this->getNamespaces();
555
556 return isset( $ns[$index] ) ? $ns[$index] : false;
557 }
558
559 /**
560 * A convenience function that returns the same thing as
561 * getNsText() except with '_' changed to ' ', useful for
562 * producing output.
563 *
564 * <code>
565 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
566 * echo $mw_ns; // prints 'MediaWiki talk'
567 * </code>
568 *
569 * @param int $index The array key of the namespace to return
570 * @return string Namespace name without underscores (empty string if namespace does not exist)
571 */
572 function getFormattedNsText( $index ) {
573 $ns = $this->getNsText( $index );
574
575 return strtr( $ns, '_', ' ' );
576 }
577
578 /**
579 * Returns gender-dependent namespace alias if available.
580 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
581 * @param int $index Namespace index
582 * @param string $gender Gender key (male, female... )
583 * @return string
584 * @since 1.18
585 */
586 function getGenderNsText( $index, $gender ) {
587 global $wgExtraGenderNamespaces;
588
589 $ns = $wgExtraGenderNamespaces +
590 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
591
592 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
593 }
594
595 /**
596 * Whether this language uses gender-dependent namespace aliases.
597 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
598 * @return bool
599 * @since 1.18
600 */
601 function needsGenderDistinction() {
602 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
603 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
604 // $wgExtraGenderNamespaces overrides everything
605 return true;
606 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
607 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
608 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
609 return false;
610 } else {
611 // Check what is in i18n files
612 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
613 return count( $aliases ) > 0;
614 }
615 }
616
617 /**
618 * Get a namespace key by value, case insensitive.
619 * Only matches namespace names for the current language, not the
620 * canonical ones defined in Namespace.php.
621 *
622 * @param string $text
623 * @return int|bool An integer if $text is a valid value otherwise false
624 */
625 function getLocalNsIndex( $text ) {
626 $lctext = $this->lc( $text );
627 $ids = $this->getNamespaceIds();
628 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
629 }
630
631 /**
632 * @return array
633 */
634 function getNamespaceAliases() {
635 if ( is_null( $this->namespaceAliases ) ) {
636 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
637 if ( !$aliases ) {
638 $aliases = array();
639 } else {
640 foreach ( $aliases as $name => $index ) {
641 if ( $index === NS_PROJECT_TALK ) {
642 unset( $aliases[$name] );
643 $name = $this->fixVariableInNamespace( $name );
644 $aliases[$name] = $index;
645 }
646 }
647 }
648
649 global $wgExtraGenderNamespaces;
650 $genders = $wgExtraGenderNamespaces +
651 (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
652 foreach ( $genders as $index => $forms ) {
653 foreach ( $forms as $alias ) {
654 $aliases[$alias] = $index;
655 }
656 }
657
658 # Also add converted namespace names as aliases, to avoid confusion.
659 $convertedNames = array();
660 foreach ( $this->getVariants() as $variant ) {
661 if ( $variant === $this->mCode ) {
662 continue;
663 }
664 foreach ( $this->getNamespaces() as $ns => $_ ) {
665 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
666 }
667 }
668
669 $this->namespaceAliases = $aliases + $convertedNames;
670 }
671
672 return $this->namespaceAliases;
673 }
674
675 /**
676 * @return array
677 */
678 function getNamespaceIds() {
679 if ( is_null( $this->mNamespaceIds ) ) {
680 global $wgNamespaceAliases;
681 # Put namespace names and aliases into a hashtable.
682 # If this is too slow, then we should arrange it so that it is done
683 # before caching. The catch is that at pre-cache time, the above
684 # class-specific fixup hasn't been done.
685 $this->mNamespaceIds = array();
686 foreach ( $this->getNamespaces() as $index => $name ) {
687 $this->mNamespaceIds[$this->lc( $name )] = $index;
688 }
689 foreach ( $this->getNamespaceAliases() as $name => $index ) {
690 $this->mNamespaceIds[$this->lc( $name )] = $index;
691 }
692 if ( $wgNamespaceAliases ) {
693 foreach ( $wgNamespaceAliases as $name => $index ) {
694 $this->mNamespaceIds[$this->lc( $name )] = $index;
695 }
696 }
697 }
698 return $this->mNamespaceIds;
699 }
700
701 /**
702 * Get a namespace key by value, case insensitive. Canonical namespace
703 * names override custom ones defined for the current language.
704 *
705 * @param string $text
706 * @return int|bool An integer if $text is a valid value otherwise false
707 */
708 function getNsIndex( $text ) {
709 $lctext = $this->lc( $text );
710 $ns = MWNamespace::getCanonicalIndex( $lctext );
711 if ( $ns !== null ) {
712 return $ns;
713 }
714 $ids = $this->getNamespaceIds();
715 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
716 }
717
718 /**
719 * short names for language variants used for language conversion links.
720 *
721 * @param string $code
722 * @param bool $usemsg Use the "variantname-xyz" message if it exists
723 * @return string
724 */
725 function getVariantname( $code, $usemsg = true ) {
726 $msg = "variantname-$code";
727 if ( $usemsg && wfMessage( $msg )->exists() ) {
728 return $this->getMessageFromDB( $msg );
729 }
730 $name = self::fetchLanguageName( $code );
731 if ( $name ) {
732 return $name; # if it's defined as a language name, show that
733 } else {
734 # otherwise, output the language code
735 return $code;
736 }
737 }
738
739 /**
740 * @deprecated since 1.24, doesn't handle conflicting aliases. Use
741 * SpecialPageFactory::getLocalNameFor instead.
742 * @param string $name
743 * @return string
744 */
745 function specialPage( $name ) {
746 $aliases = $this->getSpecialPageAliases();
747 if ( isset( $aliases[$name][0] ) ) {
748 $name = $aliases[$name][0];
749 }
750 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
751 }
752
753 /**
754 * @return array
755 */
756 function getDatePreferences() {
757 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
758 }
759
760 /**
761 * @return array
762 */
763 function getDateFormats() {
764 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
765 }
766
767 /**
768 * @return array|string
769 */
770 function getDefaultDateFormat() {
771 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
772 if ( $df === 'dmy or mdy' ) {
773 global $wgAmericanDates;
774 return $wgAmericanDates ? 'mdy' : 'dmy';
775 } else {
776 return $df;
777 }
778 }
779
780 /**
781 * @return array
782 */
783 function getDatePreferenceMigrationMap() {
784 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
785 }
786
787 /**
788 * @param string $image
789 * @return array|null
790 */
791 function getImageFile( $image ) {
792 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
793 }
794
795 /**
796 * @return array
797 * @since 1.24
798 */
799 function getImageFiles() {
800 return self::$dataCache->getItem( $this->mCode, 'imageFiles' );
801 }
802
803 /**
804 * @return array
805 */
806 function getExtraUserToggles() {
807 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
808 }
809
810 /**
811 * @param string $tog
812 * @return string
813 */
814 function getUserToggle( $tog ) {
815 return $this->getMessageFromDB( "tog-$tog" );
816 }
817
818 /**
819 * Get native language names, indexed by code.
820 * Only those defined in MediaWiki, no other data like CLDR.
821 * If $customisedOnly is true, only returns codes with a messages file
822 *
823 * @param bool $customisedOnly
824 *
825 * @return array
826 * @deprecated since 1.20, use fetchLanguageNames()
827 */
828 public static function getLanguageNames( $customisedOnly = false ) {
829 return self::fetchLanguageNames( null, $customisedOnly ? 'mwfile' : 'mw' );
830 }
831
832 /**
833 * Get translated language names. This is done on best effort and
834 * by default this is exactly the same as Language::getLanguageNames.
835 * The CLDR extension provides translated names.
836 * @param string $code Language code.
837 * @return array Language code => language name
838 * @since 1.18.0
839 * @deprecated since 1.20, use fetchLanguageNames()
840 */
841 public static function getTranslatedLanguageNames( $code ) {
842 return self::fetchLanguageNames( $code, 'all' );
843 }
844
845 /**
846 * Get an array of language names, indexed by code.
847 * @param null|string $inLanguage Code of language in which to return the names
848 * Use null for autonyms (native names)
849 * @param string $include One of:
850 * 'all' all available languages
851 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
852 * 'mwfile' only if the language is in 'mw' *and* has a message file
853 * @return array Language code => language name
854 * @since 1.20
855 */
856 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
857 $cacheKey = $inLanguage === null ? 'null' : $inLanguage;
858 $cacheKey .= ":$include";
859 if ( self::$languageNameCache === null ) {
860 self::$languageNameCache = new MapCacheLRU( 20 );
861 }
862 if ( self::$languageNameCache->has( $cacheKey ) ) {
863 $ret = self::$languageNameCache->get( $cacheKey );
864 } else {
865 $ret = self::fetchLanguageNamesUncached( $inLanguage, $include );
866 self::$languageNameCache->set( $cacheKey, $ret );
867 }
868 return $ret;
869 }
870
871 /**
872 * Uncached helper for fetchLanguageNames
873 * @param null|string $inLanguage Code of language in which to return the names
874 * Use null for autonyms (native names)
875 * @param string $include One of:
876 * 'all' all available languages
877 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
878 * 'mwfile' only if the language is in 'mw' *and* has a message file
879 * @return array Language code => language name
880 */
881 private static function fetchLanguageNamesUncached( $inLanguage = null, $include = 'mw' ) {
882 global $wgExtraLanguageNames;
883 static $coreLanguageNames;
884
885 if ( $coreLanguageNames === null ) {
886 global $IP;
887 include "$IP/languages/Names.php";
888 }
889
890 // If passed an invalid language code to use, fallback to en
891 if ( $inLanguage !== null && !Language::isValidCode( $inLanguage ) ) {
892 $inLanguage = 'en';
893 }
894
895 $names = array();
896
897 if ( $inLanguage ) {
898 # TODO: also include when $inLanguage is null, when this code is more efficient
899 Hooks::run( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
900 }
901
902 $mwNames = $wgExtraLanguageNames + $coreLanguageNames;
903 foreach ( $mwNames as $mwCode => $mwName ) {
904 # - Prefer own MediaWiki native name when not using the hook
905 # - For other names just add if not added through the hook
906 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
907 $names[$mwCode] = $mwName;
908 }
909 }
910
911 if ( $include === 'all' ) {
912 ksort( $names );
913 return $names;
914 }
915
916 $returnMw = array();
917 $coreCodes = array_keys( $mwNames );
918 foreach ( $coreCodes as $coreCode ) {
919 $returnMw[$coreCode] = $names[$coreCode];
920 }
921
922 if ( $include === 'mwfile' ) {
923 $namesMwFile = array();
924 # We do this using a foreach over the codes instead of a directory
925 # loop so that messages files in extensions will work correctly.
926 foreach ( $returnMw as $code => $value ) {
927 if ( is_readable( self::getMessagesFileName( $code ) )
928 || is_readable( self::getJsonMessagesFileName( $code ) )
929 ) {
930 $namesMwFile[$code] = $names[$code];
931 }
932 }
933
934 ksort( $namesMwFile );
935 return $namesMwFile;
936 }
937
938 ksort( $returnMw );
939 # 'mw' option; default if it's not one of the other two options (all/mwfile)
940 return $returnMw;
941 }
942
943 /**
944 * @param string $code The code of the language for which to get the name
945 * @param null|string $inLanguage Code of language in which to return the name (null for autonyms)
946 * @param string $include 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
947 * @return string Language name or empty
948 * @since 1.20
949 */
950 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
951 $code = strtolower( $code );
952 $array = self::fetchLanguageNames( $inLanguage, $include );
953 return !array_key_exists( $code, $array ) ? '' : $array[$code];
954 }
955
956 /**
957 * Get a message from the MediaWiki namespace.
958 *
959 * @param string $msg Message name
960 * @return string
961 */
962 function getMessageFromDB( $msg ) {
963 return $this->msg( $msg )->text();
964 }
965
966 /**
967 * Get message object in this language. Only for use inside this class.
968 *
969 * @param string $msg Message name
970 * @return Message
971 */
972 protected function msg( $msg ) {
973 return wfMessage( $msg )->inLanguage( $this );
974 }
975
976 /**
977 * Get the native language name of $code.
978 * Only if defined in MediaWiki, no other data like CLDR.
979 * @param string $code
980 * @return string
981 * @deprecated since 1.20, use fetchLanguageName()
982 */
983 function getLanguageName( $code ) {
984 return self::fetchLanguageName( $code );
985 }
986
987 /**
988 * @param string $key
989 * @return string
990 */
991 function getMonthName( $key ) {
992 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
993 }
994
995 /**
996 * @return array
997 */
998 function getMonthNamesArray() {
999 $monthNames = array( '' );
1000 for ( $i = 1; $i < 13; $i++ ) {
1001 $monthNames[] = $this->getMonthName( $i );
1002 }
1003 return $monthNames;
1004 }
1005
1006 /**
1007 * @param string $key
1008 * @return string
1009 */
1010 function getMonthNameGen( $key ) {
1011 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
1012 }
1013
1014 /**
1015 * @param string $key
1016 * @return string
1017 */
1018 function getMonthAbbreviation( $key ) {
1019 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
1020 }
1021
1022 /**
1023 * @return array
1024 */
1025 function getMonthAbbreviationsArray() {
1026 $monthNames = array( '' );
1027 for ( $i = 1; $i < 13; $i++ ) {
1028 $monthNames[] = $this->getMonthAbbreviation( $i );
1029 }
1030 return $monthNames;
1031 }
1032
1033 /**
1034 * @param string $key
1035 * @return string
1036 */
1037 function getWeekdayName( $key ) {
1038 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
1039 }
1040
1041 /**
1042 * @param string $key
1043 * @return string
1044 */
1045 function getWeekdayAbbreviation( $key ) {
1046 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
1047 }
1048
1049 /**
1050 * @param string $key
1051 * @return string
1052 */
1053 function getIranianCalendarMonthName( $key ) {
1054 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
1055 }
1056
1057 /**
1058 * @param string $key
1059 * @return string
1060 */
1061 function getHebrewCalendarMonthName( $key ) {
1062 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
1063 }
1064
1065 /**
1066 * @param string $key
1067 * @return string
1068 */
1069 function getHebrewCalendarMonthNameGen( $key ) {
1070 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1071 }
1072
1073 /**
1074 * @param string $key
1075 * @return string
1076 */
1077 function getHijriCalendarMonthName( $key ) {
1078 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1079 }
1080
1081 /**
1082 * Pass through result from $dateTimeObj->format()
1083 * @param DateTime|bool|null &$dateTimeObj
1084 * @param string $ts
1085 * @param DateTimeZone|bool|null $zone
1086 * @param string $code
1087 * @return string
1088 */
1089 private static function dateTimeObjFormat( &$dateTimeObj, $ts, $zone, $code ) {
1090 if ( !$dateTimeObj ) {
1091 $dateTimeObj = DateTime::createFromFormat(
1092 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1093 );
1094 }
1095 return $dateTimeObj->format( $code );
1096 }
1097
1098 /**
1099 * This is a workalike of PHP's date() function, but with better
1100 * internationalisation, a reduced set of format characters, and a better
1101 * escaping format.
1102 *
1103 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1104 * the PHP manual for definitions. There are a number of extensions, which
1105 * start with "x":
1106 *
1107 * xn Do not translate digits of the next numeric format character
1108 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1109 * xr Use roman numerals for the next numeric format character
1110 * xh Use hebrew numerals for the next numeric format character
1111 * xx Literal x
1112 * xg Genitive month name
1113 *
1114 * xij j (day number) in Iranian calendar
1115 * xiF F (month name) in Iranian calendar
1116 * xin n (month number) in Iranian calendar
1117 * xiy y (two digit year) in Iranian calendar
1118 * xiY Y (full year) in Iranian calendar
1119 *
1120 * xjj j (day number) in Hebrew calendar
1121 * xjF F (month name) in Hebrew calendar
1122 * xjt t (days in month) in Hebrew calendar
1123 * xjx xg (genitive month name) in Hebrew calendar
1124 * xjn n (month number) in Hebrew calendar
1125 * xjY Y (full year) in Hebrew calendar
1126 *
1127 * xmj j (day number) in Hijri calendar
1128 * xmF F (month name) in Hijri calendar
1129 * xmn n (month number) in Hijri calendar
1130 * xmY Y (full year) in Hijri calendar
1131 *
1132 * xkY Y (full year) in Thai solar calendar. Months and days are
1133 * identical to the Gregorian calendar
1134 * xoY Y (full year) in Minguo calendar or Juche year.
1135 * Months and days are identical to the
1136 * Gregorian calendar
1137 * xtY Y (full year) in Japanese nengo. Months and days are
1138 * identical to the Gregorian calendar
1139 *
1140 * Characters enclosed in double quotes will be considered literal (with
1141 * the quotes themselves removed). Unmatched quotes will be considered
1142 * literal quotes. Example:
1143 *
1144 * "The month is" F => The month is January
1145 * i's" => 20'11"
1146 *
1147 * Backslash escaping is also supported.
1148 *
1149 * Input timestamp is assumed to be pre-normalized to the desired local
1150 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1151 * $ts is UTC if $zone is not given.
1152 *
1153 * @param string $format
1154 * @param string $ts 14-character timestamp
1155 * YYYYMMDDHHMMSS
1156 * 01234567890123
1157 * @param DateTimeZone $zone Timezone of $ts
1158 * @param[out] int $ttl The amount of time (in seconds) the output may be cached for.
1159 * Only makes sense if $ts is the current time.
1160 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1161 *
1162 * @throws MWException
1163 * @return string
1164 */
1165 function sprintfDate( $format, $ts, DateTimeZone $zone = null, &$ttl = null ) {
1166 $s = '';
1167 $raw = false;
1168 $roman = false;
1169 $hebrewNum = false;
1170 $dateTimeObj = false;
1171 $rawToggle = false;
1172 $iranian = false;
1173 $hebrew = false;
1174 $hijri = false;
1175 $thai = false;
1176 $minguo = false;
1177 $tenno = false;
1178
1179 $usedSecond = false;
1180 $usedMinute = false;
1181 $usedHour = false;
1182 $usedAMPM = false;
1183 $usedDay = false;
1184 $usedWeek = false;
1185 $usedMonth = false;
1186 $usedYear = false;
1187 $usedISOYear = false;
1188 $usedIsLeapYear = false;
1189
1190 $usedHebrewMonth = false;
1191 $usedIranianMonth = false;
1192 $usedHijriMonth = false;
1193 $usedHebrewYear = false;
1194 $usedIranianYear = false;
1195 $usedHijriYear = false;
1196 $usedTennoYear = false;
1197
1198 if ( strlen( $ts ) !== 14 ) {
1199 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1200 }
1201
1202 if ( !ctype_digit( $ts ) ) {
1203 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1204 }
1205
1206 $formatLength = strlen( $format );
1207 for ( $p = 0; $p < $formatLength; $p++ ) {
1208 $num = false;
1209 $code = $format[$p];
1210 if ( $code == 'x' && $p < $formatLength - 1 ) {
1211 $code .= $format[++$p];
1212 }
1213
1214 if ( ( $code === 'xi'
1215 || $code === 'xj'
1216 || $code === 'xk'
1217 || $code === 'xm'
1218 || $code === 'xo'
1219 || $code === 'xt' )
1220 && $p < $formatLength - 1 ) {
1221 $code .= $format[++$p];
1222 }
1223
1224 switch ( $code ) {
1225 case 'xx':
1226 $s .= 'x';
1227 break;
1228 case 'xn':
1229 $raw = true;
1230 break;
1231 case 'xN':
1232 $rawToggle = !$rawToggle;
1233 break;
1234 case 'xr':
1235 $roman = true;
1236 break;
1237 case 'xh':
1238 $hebrewNum = true;
1239 break;
1240 case 'xg':
1241 $usedMonth = true;
1242 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1243 break;
1244 case 'xjx':
1245 $usedHebrewMonth = true;
1246 if ( !$hebrew ) {
1247 $hebrew = self::tsToHebrew( $ts );
1248 }
1249 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1250 break;
1251 case 'd':
1252 $usedDay = true;
1253 $num = substr( $ts, 6, 2 );
1254 break;
1255 case 'D':
1256 $usedDay = true;
1257 $s .= $this->getWeekdayAbbreviation(
1258 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1259 );
1260 break;
1261 case 'j':
1262 $usedDay = true;
1263 $num = intval( substr( $ts, 6, 2 ) );
1264 break;
1265 case 'xij':
1266 $usedDay = true;
1267 if ( !$iranian ) {
1268 $iranian = self::tsToIranian( $ts );
1269 }
1270 $num = $iranian[2];
1271 break;
1272 case 'xmj':
1273 $usedDay = true;
1274 if ( !$hijri ) {
1275 $hijri = self::tsToHijri( $ts );
1276 }
1277 $num = $hijri[2];
1278 break;
1279 case 'xjj':
1280 $usedDay = true;
1281 if ( !$hebrew ) {
1282 $hebrew = self::tsToHebrew( $ts );
1283 }
1284 $num = $hebrew[2];
1285 break;
1286 case 'l':
1287 $usedDay = true;
1288 $s .= $this->getWeekdayName(
1289 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'w' ) + 1
1290 );
1291 break;
1292 case 'F':
1293 $usedMonth = true;
1294 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1295 break;
1296 case 'xiF':
1297 $usedIranianMonth = true;
1298 if ( !$iranian ) {
1299 $iranian = self::tsToIranian( $ts );
1300 }
1301 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1302 break;
1303 case 'xmF':
1304 $usedHijriMonth = true;
1305 if ( !$hijri ) {
1306 $hijri = self::tsToHijri( $ts );
1307 }
1308 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1309 break;
1310 case 'xjF':
1311 $usedHebrewMonth = true;
1312 if ( !$hebrew ) {
1313 $hebrew = self::tsToHebrew( $ts );
1314 }
1315 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1316 break;
1317 case 'm':
1318 $usedMonth = true;
1319 $num = substr( $ts, 4, 2 );
1320 break;
1321 case 'M':
1322 $usedMonth = true;
1323 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1324 break;
1325 case 'n':
1326 $usedMonth = true;
1327 $num = intval( substr( $ts, 4, 2 ) );
1328 break;
1329 case 'xin':
1330 $usedIranianMonth = true;
1331 if ( !$iranian ) {
1332 $iranian = self::tsToIranian( $ts );
1333 }
1334 $num = $iranian[1];
1335 break;
1336 case 'xmn':
1337 $usedHijriMonth = true;
1338 if ( !$hijri ) {
1339 $hijri = self::tsToHijri ( $ts );
1340 }
1341 $num = $hijri[1];
1342 break;
1343 case 'xjn':
1344 $usedHebrewMonth = true;
1345 if ( !$hebrew ) {
1346 $hebrew = self::tsToHebrew( $ts );
1347 }
1348 $num = $hebrew[1];
1349 break;
1350 case 'xjt':
1351 $usedHebrewMonth = true;
1352 if ( !$hebrew ) {
1353 $hebrew = self::tsToHebrew( $ts );
1354 }
1355 $num = $hebrew[3];
1356 break;
1357 case 'Y':
1358 $usedYear = true;
1359 $num = substr( $ts, 0, 4 );
1360 break;
1361 case 'xiY':
1362 $usedIranianYear = true;
1363 if ( !$iranian ) {
1364 $iranian = self::tsToIranian( $ts );
1365 }
1366 $num = $iranian[0];
1367 break;
1368 case 'xmY':
1369 $usedHijriYear = true;
1370 if ( !$hijri ) {
1371 $hijri = self::tsToHijri( $ts );
1372 }
1373 $num = $hijri[0];
1374 break;
1375 case 'xjY':
1376 $usedHebrewYear = true;
1377 if ( !$hebrew ) {
1378 $hebrew = self::tsToHebrew( $ts );
1379 }
1380 $num = $hebrew[0];
1381 break;
1382 case 'xkY':
1383 $usedYear = true;
1384 if ( !$thai ) {
1385 $thai = self::tsToYear( $ts, 'thai' );
1386 }
1387 $num = $thai[0];
1388 break;
1389 case 'xoY':
1390 $usedYear = true;
1391 if ( !$minguo ) {
1392 $minguo = self::tsToYear( $ts, 'minguo' );
1393 }
1394 $num = $minguo[0];
1395 break;
1396 case 'xtY':
1397 $usedTennoYear = true;
1398 if ( !$tenno ) {
1399 $tenno = self::tsToYear( $ts, 'tenno' );
1400 }
1401 $num = $tenno[0];
1402 break;
1403 case 'y':
1404 $usedYear = true;
1405 $num = substr( $ts, 2, 2 );
1406 break;
1407 case 'xiy':
1408 $usedIranianYear = true;
1409 if ( !$iranian ) {
1410 $iranian = self::tsToIranian( $ts );
1411 }
1412 $num = substr( $iranian[0], -2 );
1413 break;
1414 case 'a':
1415 $usedAMPM = true;
1416 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1417 break;
1418 case 'A':
1419 $usedAMPM = true;
1420 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1421 break;
1422 case 'g':
1423 $usedHour = true;
1424 $h = substr( $ts, 8, 2 );
1425 $num = $h % 12 ? $h % 12 : 12;
1426 break;
1427 case 'G':
1428 $usedHour = true;
1429 $num = intval( substr( $ts, 8, 2 ) );
1430 break;
1431 case 'h':
1432 $usedHour = true;
1433 $h = substr( $ts, 8, 2 );
1434 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1435 break;
1436 case 'H':
1437 $usedHour = true;
1438 $num = substr( $ts, 8, 2 );
1439 break;
1440 case 'i':
1441 $usedMinute = true;
1442 $num = substr( $ts, 10, 2 );
1443 break;
1444 case 's':
1445 $usedSecond = true;
1446 $num = substr( $ts, 12, 2 );
1447 break;
1448 case 'c':
1449 case 'r':
1450 $usedSecond = true;
1451 // fall through
1452 case 'e':
1453 case 'O':
1454 case 'P':
1455 case 'T':
1456 $s .= Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1457 break;
1458 case 'w':
1459 case 'N':
1460 case 'z':
1461 $usedDay = true;
1462 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1463 break;
1464 case 'W':
1465 $usedWeek = true;
1466 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1467 break;
1468 case 't':
1469 $usedMonth = true;
1470 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1471 break;
1472 case 'L':
1473 $usedIsLeapYear = true;
1474 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1475 break;
1476 case 'o':
1477 $usedISOYear = true;
1478 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1479 break;
1480 case 'U':
1481 $usedSecond = true;
1482 // fall through
1483 case 'I':
1484 case 'Z':
1485 $num = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, $code );
1486 break;
1487 case '\\':
1488 # Backslash escaping
1489 if ( $p < $formatLength - 1 ) {
1490 $s .= $format[++$p];
1491 } else {
1492 $s .= '\\';
1493 }
1494 break;
1495 case '"':
1496 # Quoted literal
1497 if ( $p < $formatLength - 1 ) {
1498 $endQuote = strpos( $format, '"', $p + 1 );
1499 if ( $endQuote === false ) {
1500 # No terminating quote, assume literal "
1501 $s .= '"';
1502 } else {
1503 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1504 $p = $endQuote;
1505 }
1506 } else {
1507 # Quote at end of string, assume literal "
1508 $s .= '"';
1509 }
1510 break;
1511 default:
1512 $s .= $format[$p];
1513 }
1514 if ( $num !== false ) {
1515 if ( $rawToggle || $raw ) {
1516 $s .= $num;
1517 $raw = false;
1518 } elseif ( $roman ) {
1519 $s .= Language::romanNumeral( $num );
1520 $roman = false;
1521 } elseif ( $hebrewNum ) {
1522 $s .= self::hebrewNumeral( $num );
1523 $hebrewNum = false;
1524 } else {
1525 $s .= $this->formatNum( $num, true );
1526 }
1527 }
1528 }
1529
1530 if ( $usedSecond ) {
1531 $ttl = 1;
1532 } elseif ( $usedMinute ) {
1533 $ttl = 60 - substr( $ts, 12, 2 );
1534 } elseif ( $usedHour ) {
1535 $ttl = 3600 - substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1536 } elseif ( $usedAMPM ) {
1537 $ttl = 43200 - ( substr( $ts, 8, 2 ) % 12 ) * 3600 -
1538 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1539 } elseif (
1540 $usedDay ||
1541 $usedHebrewMonth ||
1542 $usedIranianMonth ||
1543 $usedHijriMonth ||
1544 $usedHebrewYear ||
1545 $usedIranianYear ||
1546 $usedHijriYear ||
1547 $usedTennoYear
1548 ) {
1549 // @todo Someone who understands the non-Gregorian calendars
1550 // should write proper logic for them so that they don't need purged every day.
1551 $ttl = 86400 - substr( $ts, 8, 2 ) * 3600 -
1552 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1553 } else {
1554 $possibleTtls = array();
1555 $timeRemainingInDay = 86400 - substr( $ts, 8, 2 ) * 3600 -
1556 substr( $ts, 10, 2 ) * 60 - substr( $ts, 12, 2 );
1557 if ( $usedWeek ) {
1558 $possibleTtls[] =
1559 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400 +
1560 $timeRemainingInDay;
1561 } elseif ( $usedISOYear ) {
1562 // December 28th falls on the last ISO week of the year, every year.
1563 // The last ISO week of a year can be 52 or 53.
1564 $lastWeekOfISOYear = DateTime::createFromFormat(
1565 'Ymd',
1566 substr( $ts, 0, 4 ) . '1228',
1567 $zone ?: new DateTimeZone( 'UTC' )
1568 )->format( 'W' );
1569 $currentISOWeek = Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'W' );
1570 $weeksRemaining = $lastWeekOfISOYear - $currentISOWeek;
1571 $timeRemainingInWeek =
1572 ( 7 - Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'N' ) ) * 86400
1573 + $timeRemainingInDay;
1574 $possibleTtls[] = $weeksRemaining * 604800 + $timeRemainingInWeek;
1575 }
1576
1577 if ( $usedMonth ) {
1578 $possibleTtls[] =
1579 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 't' ) -
1580 substr( $ts, 6, 2 ) ) * 86400
1581 + $timeRemainingInDay;
1582 } elseif ( $usedYear ) {
1583 $possibleTtls[] =
1584 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1585 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1586 + $timeRemainingInDay;
1587 } elseif ( $usedIsLeapYear ) {
1588 $year = substr( $ts, 0, 4 );
1589 $timeRemainingInYear =
1590 ( Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'L' ) + 364 -
1591 Language::dateTimeObjFormat( $dateTimeObj, $ts, $zone, 'z' ) ) * 86400
1592 + $timeRemainingInDay;
1593 $mod = $year % 4;
1594 if ( $mod || ( !( $year % 100 ) && $year % 400 ) ) {
1595 // this isn't a leap year. see when the next one starts
1596 $nextCandidate = $year - $mod + 4;
1597 if ( $nextCandidate % 100 || !( $nextCandidate % 400 ) ) {
1598 $possibleTtls[] = ( $nextCandidate - $year - 1 ) * 365 * 86400 +
1599 $timeRemainingInYear;
1600 } else {
1601 $possibleTtls[] = ( $nextCandidate - $year + 3 ) * 365 * 86400 +
1602 $timeRemainingInYear;
1603 }
1604 } else {
1605 // this is a leap year, so the next year isn't
1606 $possibleTtls[] = $timeRemainingInYear;
1607 }
1608 }
1609
1610 if ( $possibleTtls ) {
1611 $ttl = min( $possibleTtls );
1612 }
1613 }
1614
1615 return $s;
1616 }
1617
1618 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1619 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1620
1621 /**
1622 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1623 * Gregorian dates to Iranian dates. Originally written in C, it
1624 * is released under the terms of GNU Lesser General Public
1625 * License. Conversion to PHP was performed by Niklas Laxström.
1626 *
1627 * Link: http://www.farsiweb.info/jalali/jalali.c
1628 *
1629 * @param string $ts
1630 *
1631 * @return string
1632 */
1633 private static function tsToIranian( $ts ) {
1634 $gy = substr( $ts, 0, 4 ) -1600;
1635 $gm = substr( $ts, 4, 2 ) -1;
1636 $gd = substr( $ts, 6, 2 ) -1;
1637
1638 # Days passed from the beginning (including leap years)
1639 $gDayNo = 365 * $gy
1640 + floor( ( $gy + 3 ) / 4 )
1641 - floor( ( $gy + 99 ) / 100 )
1642 + floor( ( $gy + 399 ) / 400 );
1643
1644 // Add days of the past months of this year
1645 for ( $i = 0; $i < $gm; $i++ ) {
1646 $gDayNo += self::$GREG_DAYS[$i];
1647 }
1648
1649 // Leap years
1650 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1651 $gDayNo++;
1652 }
1653
1654 // Days passed in current month
1655 $gDayNo += (int)$gd;
1656
1657 $jDayNo = $gDayNo - 79;
1658
1659 $jNp = floor( $jDayNo / 12053 );
1660 $jDayNo %= 12053;
1661
1662 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1663 $jDayNo %= 1461;
1664
1665 if ( $jDayNo >= 366 ) {
1666 $jy += floor( ( $jDayNo - 1 ) / 365 );
1667 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1668 }
1669
1670 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1671 $jDayNo -= self::$IRANIAN_DAYS[$i];
1672 }
1673
1674 $jm = $i + 1;
1675 $jd = $jDayNo + 1;
1676
1677 return array( $jy, $jm, $jd );
1678 }
1679
1680 /**
1681 * Converting Gregorian dates to Hijri dates.
1682 *
1683 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1684 *
1685 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1686 *
1687 * @param string $ts
1688 *
1689 * @return string
1690 */
1691 private static function tsToHijri( $ts ) {
1692 $year = substr( $ts, 0, 4 );
1693 $month = substr( $ts, 4, 2 );
1694 $day = substr( $ts, 6, 2 );
1695
1696 $zyr = $year;
1697 $zd = $day;
1698 $zm = $month;
1699 $zy = $zyr;
1700
1701 if (
1702 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1703 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1704 ) {
1705 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1706 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1707 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1708 $zd - 32075;
1709 } else {
1710 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1711 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1712 }
1713
1714 $zl = $zjd -1948440 + 10632;
1715 $zn = (int)( ( $zl - 1 ) / 10631 );
1716 $zl = $zl - 10631 * $zn + 354;
1717 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) +
1718 ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1719 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) -
1720 ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1721 $zm = (int)( ( 24 * $zl ) / 709 );
1722 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1723 $zy = 30 * $zn + $zj - 30;
1724
1725 return array( $zy, $zm, $zd );
1726 }
1727
1728 /**
1729 * Converting Gregorian dates to Hebrew dates.
1730 *
1731 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1732 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1733 * to translate the relevant functions into PHP and release them under
1734 * GNU GPL.
1735 *
1736 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1737 * and Adar II is 14. In a non-leap year, Adar is 6.
1738 *
1739 * @param string $ts
1740 *
1741 * @return string
1742 */
1743 private static function tsToHebrew( $ts ) {
1744 # Parse date
1745 $year = substr( $ts, 0, 4 );
1746 $month = substr( $ts, 4, 2 );
1747 $day = substr( $ts, 6, 2 );
1748
1749 # Calculate Hebrew year
1750 $hebrewYear = $year + 3760;
1751
1752 # Month number when September = 1, August = 12
1753 $month += 4;
1754 if ( $month > 12 ) {
1755 # Next year
1756 $month -= 12;
1757 $year++;
1758 $hebrewYear++;
1759 }
1760
1761 # Calculate day of year from 1 September
1762 $dayOfYear = $day;
1763 for ( $i = 1; $i < $month; $i++ ) {
1764 if ( $i == 6 ) {
1765 # February
1766 $dayOfYear += 28;
1767 # Check if the year is leap
1768 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1769 $dayOfYear++;
1770 }
1771 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1772 $dayOfYear += 30;
1773 } else {
1774 $dayOfYear += 31;
1775 }
1776 }
1777
1778 # Calculate the start of the Hebrew year
1779 $start = self::hebrewYearStart( $hebrewYear );
1780
1781 # Calculate next year's start
1782 if ( $dayOfYear <= $start ) {
1783 # Day is before the start of the year - it is the previous year
1784 # Next year's start
1785 $nextStart = $start;
1786 # Previous year
1787 $year--;
1788 $hebrewYear--;
1789 # Add days since previous year's 1 September
1790 $dayOfYear += 365;
1791 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1792 # Leap year
1793 $dayOfYear++;
1794 }
1795 # Start of the new (previous) year
1796 $start = self::hebrewYearStart( $hebrewYear );
1797 } else {
1798 # Next year's start
1799 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1800 }
1801
1802 # Calculate Hebrew day of year
1803 $hebrewDayOfYear = $dayOfYear - $start;
1804
1805 # Difference between year's days
1806 $diff = $nextStart - $start;
1807 # Add 12 (or 13 for leap years) days to ignore the difference between
1808 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1809 # difference is only about the year type
1810 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1811 $diff += 13;
1812 } else {
1813 $diff += 12;
1814 }
1815
1816 # Check the year pattern, and is leap year
1817 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1818 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1819 # and non-leap years
1820 $yearPattern = $diff % 30;
1821 # Check if leap year
1822 $isLeap = $diff >= 30;
1823
1824 # Calculate day in the month from number of day in the Hebrew year
1825 # Don't check Adar - if the day is not in Adar, we will stop before;
1826 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1827 $hebrewDay = $hebrewDayOfYear;
1828 $hebrewMonth = 1;
1829 $days = 0;
1830 while ( $hebrewMonth <= 12 ) {
1831 # Calculate days in this month
1832 if ( $isLeap && $hebrewMonth == 6 ) {
1833 # Adar in a leap year
1834 if ( $isLeap ) {
1835 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1836 $days = 30;
1837 if ( $hebrewDay <= $days ) {
1838 # Day in Adar I
1839 $hebrewMonth = 13;
1840 } else {
1841 # Subtract the days of Adar I
1842 $hebrewDay -= $days;
1843 # Try Adar II
1844 $days = 29;
1845 if ( $hebrewDay <= $days ) {
1846 # Day in Adar II
1847 $hebrewMonth = 14;
1848 }
1849 }
1850 }
1851 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1852 # Cheshvan in a complete year (otherwise as the rule below)
1853 $days = 30;
1854 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1855 # Kislev in an incomplete year (otherwise as the rule below)
1856 $days = 29;
1857 } else {
1858 # Odd months have 30 days, even have 29
1859 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1860 }
1861 if ( $hebrewDay <= $days ) {
1862 # In the current month
1863 break;
1864 } else {
1865 # Subtract the days of the current month
1866 $hebrewDay -= $days;
1867 # Try in the next month
1868 $hebrewMonth++;
1869 }
1870 }
1871
1872 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1873 }
1874
1875 /**
1876 * This calculates the Hebrew year start, as days since 1 September.
1877 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1878 * Used for Hebrew date.
1879 *
1880 * @param int $year
1881 *
1882 * @return string
1883 */
1884 private static function hebrewYearStart( $year ) {
1885 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1886 $b = intval( ( $year - 1 ) % 4 );
1887 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1888 if ( $m < 0 ) {
1889 $m--;
1890 }
1891 $Mar = intval( $m );
1892 if ( $m < 0 ) {
1893 $m++;
1894 }
1895 $m -= $Mar;
1896
1897 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1898 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1899 $Mar++;
1900 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1901 $Mar += 2;
1902 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1903 $Mar++;
1904 }
1905
1906 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1907 return $Mar;
1908 }
1909
1910 /**
1911 * Algorithm to convert Gregorian dates to Thai solar dates,
1912 * Minguo dates or Minguo dates.
1913 *
1914 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1915 * http://en.wikipedia.org/wiki/Minguo_calendar
1916 * http://en.wikipedia.org/wiki/Japanese_era_name
1917 *
1918 * @param string $ts 14-character timestamp
1919 * @param string $cName Calender name
1920 * @return array Converted year, month, day
1921 */
1922 private static function tsToYear( $ts, $cName ) {
1923 $gy = substr( $ts, 0, 4 );
1924 $gm = substr( $ts, 4, 2 );
1925 $gd = substr( $ts, 6, 2 );
1926
1927 if ( !strcmp( $cName, 'thai' ) ) {
1928 # Thai solar dates
1929 # Add 543 years to the Gregorian calendar
1930 # Months and days are identical
1931 $gy_offset = $gy + 543;
1932 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1933 # Minguo dates
1934 # Deduct 1911 years from the Gregorian calendar
1935 # Months and days are identical
1936 $gy_offset = $gy - 1911;
1937 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1938 # Nengō dates up to Meiji period
1939 # Deduct years from the Gregorian calendar
1940 # depending on the nengo periods
1941 # Months and days are identical
1942 if ( ( $gy < 1912 )
1943 || ( ( $gy == 1912 ) && ( $gm < 7 ) )
1944 || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) )
1945 ) {
1946 # Meiji period
1947 $gy_gannen = $gy - 1868 + 1;
1948 $gy_offset = $gy_gannen;
1949 if ( $gy_gannen == 1 ) {
1950 $gy_offset = '元';
1951 }
1952 $gy_offset = '明治' . $gy_offset;
1953 } elseif (
1954 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1955 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1956 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1957 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1958 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1959 ) {
1960 # Taishō period
1961 $gy_gannen = $gy - 1912 + 1;
1962 $gy_offset = $gy_gannen;
1963 if ( $gy_gannen == 1 ) {
1964 $gy_offset = '元';
1965 }
1966 $gy_offset = '大正' . $gy_offset;
1967 } elseif (
1968 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1969 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1970 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1971 ) {
1972 # Shōwa period
1973 $gy_gannen = $gy - 1926 + 1;
1974 $gy_offset = $gy_gannen;
1975 if ( $gy_gannen == 1 ) {
1976 $gy_offset = '元';
1977 }
1978 $gy_offset = '昭和' . $gy_offset;
1979 } else {
1980 # Heisei period
1981 $gy_gannen = $gy - 1989 + 1;
1982 $gy_offset = $gy_gannen;
1983 if ( $gy_gannen == 1 ) {
1984 $gy_offset = '元';
1985 }
1986 $gy_offset = '平成' . $gy_offset;
1987 }
1988 } else {
1989 $gy_offset = $gy;
1990 }
1991
1992 return array( $gy_offset, $gm, $gd );
1993 }
1994
1995 /**
1996 * Roman number formatting up to 10000
1997 *
1998 * @param int $num
1999 *
2000 * @return string
2001 */
2002 static function romanNumeral( $num ) {
2003 static $table = array(
2004 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
2005 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
2006 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
2007 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM',
2008 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
2009 );
2010
2011 $num = intval( $num );
2012 if ( $num > 10000 || $num <= 0 ) {
2013 return $num;
2014 }
2015
2016 $s = '';
2017 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2018 if ( $num >= $pow10 ) {
2019 $s .= $table[$i][(int)floor( $num / $pow10 )];
2020 }
2021 $num = $num % $pow10;
2022 }
2023 return $s;
2024 }
2025
2026 /**
2027 * Hebrew Gematria number formatting up to 9999
2028 *
2029 * @param int $num
2030 *
2031 * @return string
2032 */
2033 static function hebrewNumeral( $num ) {
2034 static $table = array(
2035 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
2036 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
2037 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
2038 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
2039 );
2040
2041 $num = intval( $num );
2042 if ( $num > 9999 || $num <= 0 ) {
2043 return $num;
2044 }
2045
2046 $s = '';
2047 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
2048 if ( $num >= $pow10 ) {
2049 if ( $num == 15 || $num == 16 ) {
2050 $s .= $table[0][9] . $table[0][$num - 9];
2051 $num = 0;
2052 } else {
2053 $s .= $table[$i][intval( ( $num / $pow10 ) )];
2054 if ( $pow10 == 1000 ) {
2055 $s .= "'";
2056 }
2057 }
2058 }
2059 $num = $num % $pow10;
2060 }
2061 if ( strlen( $s ) == 2 ) {
2062 $str = $s . "'";
2063 } else {
2064 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
2065 $str .= substr( $s, strlen( $s ) - 2, 2 );
2066 }
2067 $start = substr( $str, 0, strlen( $str ) - 2 );
2068 $end = substr( $str, strlen( $str ) - 2 );
2069 switch ( $end ) {
2070 case 'כ':
2071 $str = $start . 'ך';
2072 break;
2073 case 'מ':
2074 $str = $start . 'ם';
2075 break;
2076 case 'נ':
2077 $str = $start . 'ן';
2078 break;
2079 case 'פ':
2080 $str = $start . 'ף';
2081 break;
2082 case 'צ':
2083 $str = $start . 'ץ';
2084 break;
2085 }
2086 return $str;
2087 }
2088
2089 /**
2090 * Used by date() and time() to adjust the time output.
2091 *
2092 * @param string $ts The time in date('YmdHis') format
2093 * @param mixed $tz Adjust the time by this amount (default false, mean we
2094 * get user timecorrection setting)
2095 * @return int
2096 */
2097 function userAdjust( $ts, $tz = false ) {
2098 global $wgUser, $wgLocalTZoffset;
2099
2100 if ( $tz === false ) {
2101 $tz = $wgUser->getOption( 'timecorrection' );
2102 }
2103
2104 $data = explode( '|', $tz, 3 );
2105
2106 if ( $data[0] == 'ZoneInfo' ) {
2107 wfSuppressWarnings();
2108 $userTZ = timezone_open( $data[2] );
2109 wfRestoreWarnings();
2110 if ( $userTZ !== false ) {
2111 $date = date_create( $ts, timezone_open( 'UTC' ) );
2112 date_timezone_set( $date, $userTZ );
2113 $date = date_format( $date, 'YmdHis' );
2114 return $date;
2115 }
2116 # Unrecognized timezone, default to 'Offset' with the stored offset.
2117 $data[0] = 'Offset';
2118 }
2119
2120 if ( $data[0] == 'System' || $tz == '' ) {
2121 # Global offset in minutes.
2122 $minDiff = $wgLocalTZoffset;
2123 } elseif ( $data[0] == 'Offset' ) {
2124 $minDiff = intval( $data[1] );
2125 } else {
2126 $data = explode( ':', $tz );
2127 if ( count( $data ) == 2 ) {
2128 $data[0] = intval( $data[0] );
2129 $data[1] = intval( $data[1] );
2130 $minDiff = abs( $data[0] ) * 60 + $data[1];
2131 if ( $data[0] < 0 ) {
2132 $minDiff = -$minDiff;
2133 }
2134 } else {
2135 $minDiff = intval( $data[0] ) * 60;
2136 }
2137 }
2138
2139 # No difference ? Return time unchanged
2140 if ( 0 == $minDiff ) {
2141 return $ts;
2142 }
2143
2144 wfSuppressWarnings(); // E_STRICT system time bitching
2145 # Generate an adjusted date; take advantage of the fact that mktime
2146 # will normalize out-of-range values so we don't have to split $minDiff
2147 # into hours and minutes.
2148 $t = mktime( (
2149 (int)substr( $ts, 8, 2 ) ), # Hours
2150 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
2151 (int)substr( $ts, 12, 2 ), # Seconds
2152 (int)substr( $ts, 4, 2 ), # Month
2153 (int)substr( $ts, 6, 2 ), # Day
2154 (int)substr( $ts, 0, 4 ) ); # Year
2155
2156 $date = date( 'YmdHis', $t );
2157 wfRestoreWarnings();
2158
2159 return $date;
2160 }
2161
2162 /**
2163 * This is meant to be used by time(), date(), and timeanddate() to get
2164 * the date preference they're supposed to use, it should be used in
2165 * all children.
2166 *
2167 *<code>
2168 * function timeanddate([...], $format = true) {
2169 * $datePreference = $this->dateFormat($format);
2170 * [...]
2171 * }
2172 *</code>
2173 *
2174 * @param int|string|bool $usePrefs If true, the user's preference is used
2175 * if false, the site/language default is used
2176 * if int/string, assumed to be a format.
2177 * @return string
2178 */
2179 function dateFormat( $usePrefs = true ) {
2180 global $wgUser;
2181
2182 if ( is_bool( $usePrefs ) ) {
2183 if ( $usePrefs ) {
2184 $datePreference = $wgUser->getDatePreference();
2185 } else {
2186 $datePreference = (string)User::getDefaultOption( 'date' );
2187 }
2188 } else {
2189 $datePreference = (string)$usePrefs;
2190 }
2191
2192 // return int
2193 if ( $datePreference == '' ) {
2194 return 'default';
2195 }
2196
2197 return $datePreference;
2198 }
2199
2200 /**
2201 * Get a format string for a given type and preference
2202 * @param string $type May be date, time or both
2203 * @param string $pref The format name as it appears in Messages*.php
2204 *
2205 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2206 *
2207 * @return string
2208 */
2209 function getDateFormatString( $type, $pref ) {
2210 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2211 if ( $pref == 'default' ) {
2212 $pref = $this->getDefaultDateFormat();
2213 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2214 } else {
2215 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2216
2217 if ( $type === 'pretty' && $df === null ) {
2218 $df = $this->getDateFormatString( 'date', $pref );
2219 }
2220
2221 if ( $df === null ) {
2222 $pref = $this->getDefaultDateFormat();
2223 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2224 }
2225 }
2226 $this->dateFormatStrings[$type][$pref] = $df;
2227 }
2228 return $this->dateFormatStrings[$type][$pref];
2229 }
2230
2231 /**
2232 * @param string $ts The time format which needs to be turned into a
2233 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2234 * @param bool $adj Whether to adjust the time output according to the
2235 * user configured offset ($timecorrection)
2236 * @param mixed $format True to use user's date format preference
2237 * @param string|bool $timecorrection The time offset as returned by
2238 * validateTimeZone() in Special:Preferences
2239 * @return string
2240 */
2241 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2242 $ts = wfTimestamp( TS_MW, $ts );
2243 if ( $adj ) {
2244 $ts = $this->userAdjust( $ts, $timecorrection );
2245 }
2246 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2247 return $this->sprintfDate( $df, $ts );
2248 }
2249
2250 /**
2251 * @param string $ts The time format which needs to be turned into a
2252 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2253 * @param bool $adj Whether to adjust the time output according to the
2254 * user configured offset ($timecorrection)
2255 * @param mixed $format True to use user's date format preference
2256 * @param string|bool $timecorrection The time offset as returned by
2257 * validateTimeZone() in Special:Preferences
2258 * @return string
2259 */
2260 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2261 $ts = wfTimestamp( TS_MW, $ts );
2262 if ( $adj ) {
2263 $ts = $this->userAdjust( $ts, $timecorrection );
2264 }
2265 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2266 return $this->sprintfDate( $df, $ts );
2267 }
2268
2269 /**
2270 * @param string $ts The time format which needs to be turned into a
2271 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2272 * @param bool $adj Whether to adjust the time output according to the
2273 * user configured offset ($timecorrection)
2274 * @param mixed $format What format to return, if it's false output the
2275 * default one (default true)
2276 * @param string|bool $timecorrection The time offset as returned by
2277 * validateTimeZone() in Special:Preferences
2278 * @return string
2279 */
2280 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2281 $ts = wfTimestamp( TS_MW, $ts );
2282 if ( $adj ) {
2283 $ts = $this->userAdjust( $ts, $timecorrection );
2284 }
2285 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2286 return $this->sprintfDate( $df, $ts );
2287 }
2288
2289 /**
2290 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2291 *
2292 * @since 1.20
2293 *
2294 * @param int $seconds The amount of seconds.
2295 * @param array $chosenIntervals The intervals to enable.
2296 *
2297 * @return string
2298 */
2299 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2300 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2301
2302 $segments = array();
2303
2304 foreach ( $intervals as $intervalName => $intervalValue ) {
2305 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2306 // duration-years, duration-decades, duration-centuries, duration-millennia
2307 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2308 $segments[] = $message->inLanguage( $this )->escaped();
2309 }
2310
2311 return $this->listToText( $segments );
2312 }
2313
2314 /**
2315 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2316 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2317 *
2318 * @since 1.20
2319 *
2320 * @param int $seconds The amount of seconds.
2321 * @param array $chosenIntervals The intervals to enable.
2322 *
2323 * @return array
2324 */
2325 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2326 if ( empty( $chosenIntervals ) ) {
2327 $chosenIntervals = array(
2328 'millennia',
2329 'centuries',
2330 'decades',
2331 'years',
2332 'days',
2333 'hours',
2334 'minutes',
2335 'seconds'
2336 );
2337 }
2338
2339 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2340 $sortedNames = array_keys( $intervals );
2341 $smallestInterval = array_pop( $sortedNames );
2342
2343 $segments = array();
2344
2345 foreach ( $intervals as $name => $length ) {
2346 $value = floor( $seconds / $length );
2347
2348 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2349 $seconds -= $value * $length;
2350 $segments[$name] = $value;
2351 }
2352 }
2353
2354 return $segments;
2355 }
2356
2357 /**
2358 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2359 *
2360 * @param string $type Can be 'date', 'time' or 'both'
2361 * @param string $ts The time format which needs to be turned into a
2362 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2363 * @param User $user User object used to get preferences for timezone and format
2364 * @param array $options Array, can contain the following keys:
2365 * - 'timecorrection': time correction, can have the following values:
2366 * - true: use user's preference
2367 * - false: don't use time correction
2368 * - int: value of time correction in minutes
2369 * - 'format': format to use, can have the following values:
2370 * - true: use user's preference
2371 * - false: use default preference
2372 * - string: format to use
2373 * @since 1.19
2374 * @return string
2375 */
2376 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2377 $ts = wfTimestamp( TS_MW, $ts );
2378 $options += array( 'timecorrection' => true, 'format' => true );
2379 if ( $options['timecorrection'] !== false ) {
2380 if ( $options['timecorrection'] === true ) {
2381 $offset = $user->getOption( 'timecorrection' );
2382 } else {
2383 $offset = $options['timecorrection'];
2384 }
2385 $ts = $this->userAdjust( $ts, $offset );
2386 }
2387 if ( $options['format'] === true ) {
2388 $format = $user->getDatePreference();
2389 } else {
2390 $format = $options['format'];
2391 }
2392 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2393 return $this->sprintfDate( $df, $ts );
2394 }
2395
2396 /**
2397 * Get the formatted date for the given timestamp and formatted for
2398 * the given user.
2399 *
2400 * @param mixed $ts Mixed: the time format which needs to be turned into a
2401 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2402 * @param User $user User object used to get preferences for timezone and format
2403 * @param array $options Array, can contain the following keys:
2404 * - 'timecorrection': time correction, can have the following values:
2405 * - true: use user's preference
2406 * - false: don't use time correction
2407 * - int: value of time correction in minutes
2408 * - 'format': format to use, can have the following values:
2409 * - true: use user's preference
2410 * - false: use default preference
2411 * - string: format to use
2412 * @since 1.19
2413 * @return string
2414 */
2415 public function userDate( $ts, User $user, array $options = array() ) {
2416 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2417 }
2418
2419 /**
2420 * Get the formatted time for the given timestamp and formatted for
2421 * the given user.
2422 *
2423 * @param mixed $ts The time format which needs to be turned into a
2424 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2425 * @param User $user User object used to get preferences for timezone and format
2426 * @param array $options Array, can contain the following keys:
2427 * - 'timecorrection': time correction, can have the following values:
2428 * - true: use user's preference
2429 * - false: don't use time correction
2430 * - int: value of time correction in minutes
2431 * - 'format': format to use, can have the following values:
2432 * - true: use user's preference
2433 * - false: use default preference
2434 * - string: format to use
2435 * @since 1.19
2436 * @return string
2437 */
2438 public function userTime( $ts, User $user, array $options = array() ) {
2439 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2440 }
2441
2442 /**
2443 * Get the formatted date and time for the given timestamp and formatted for
2444 * the given user.
2445 *
2446 * @param mixed $ts The time format which needs to be turned into a
2447 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2448 * @param User $user User object used to get preferences for timezone and format
2449 * @param array $options Array, can contain the following keys:
2450 * - 'timecorrection': time correction, can have the following values:
2451 * - true: use user's preference
2452 * - false: don't use time correction
2453 * - int: value of time correction in minutes
2454 * - 'format': format to use, can have the following values:
2455 * - true: use user's preference
2456 * - false: use default preference
2457 * - string: format to use
2458 * @since 1.19
2459 * @return string
2460 */
2461 public function userTimeAndDate( $ts, User $user, array $options = array() ) {
2462 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2463 }
2464
2465 /**
2466 * Convert an MWTimestamp into a pretty human-readable timestamp using
2467 * the given user preferences and relative base time.
2468 *
2469 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2470 * on your timestamp object, which will then call this function. Calling
2471 * this function directly will cause hooks to be skipped over.
2472 *
2473 * @see MWTimestamp::getHumanTimestamp
2474 * @param MWTimestamp $ts Timestamp to prettify
2475 * @param MWTimestamp $relativeTo Base timestamp
2476 * @param User $user User preferences to use
2477 * @return string Human timestamp
2478 * @since 1.22
2479 */
2480 public function getHumanTimestamp( MWTimestamp $ts, MWTimestamp $relativeTo, User $user ) {
2481 $diff = $ts->diff( $relativeTo );
2482 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) -
2483 (int)$relativeTo->timestamp->format( 'w' ) );
2484 $days = $diff->days ?: (int)$diffDay;
2485 if ( $diff->invert || $days > 5
2486 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' )
2487 ) {
2488 // Timestamps are in different years: use full timestamp
2489 // Also do full timestamp for future dates
2490 /**
2491 * @todo FIXME: Add better handling of future timestamps.
2492 */
2493 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2494 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2495 } elseif ( $days > 5 ) {
2496 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2497 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2498 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2499 } elseif ( $days > 1 ) {
2500 // Timestamp within the past week: show the day of the week and time
2501 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2502 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2503 // Messages:
2504 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2505 $ts = wfMessage( "$weekday-at" )
2506 ->inLanguage( $this )
2507 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2508 ->text();
2509 } elseif ( $days == 1 ) {
2510 // Timestamp was yesterday: say 'yesterday' and the time.
2511 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2512 $ts = wfMessage( 'yesterday-at' )
2513 ->inLanguage( $this )
2514 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2515 ->text();
2516 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2517 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2518 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2519 $ts = wfMessage( 'today-at' )
2520 ->inLanguage( $this )
2521 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2522 ->text();
2523
2524 // From here on in, the timestamp was soon enough ago so that we can simply say
2525 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2526 } elseif ( $diff->h == 1 ) {
2527 // Less than 90 minutes, but more than an hour ago.
2528 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2529 } elseif ( $diff->i >= 1 ) {
2530 // A few minutes ago.
2531 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2532 } elseif ( $diff->s >= 30 ) {
2533 // Less than a minute, but more than 30 sec ago.
2534 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2535 } else {
2536 // Less than 30 seconds ago.
2537 $ts = wfMessage( 'just-now' )->text();
2538 }
2539
2540 return $ts;
2541 }
2542
2543 /**
2544 * @param string $key
2545 * @return array|null
2546 */
2547 function getMessage( $key ) {
2548 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2549 }
2550
2551 /**
2552 * @return array
2553 */
2554 function getAllMessages() {
2555 return self::$dataCache->getItem( $this->mCode, 'messages' );
2556 }
2557
2558 /**
2559 * @param string $in
2560 * @param string $out
2561 * @param string $string
2562 * @return string
2563 */
2564 function iconv( $in, $out, $string ) {
2565 # This is a wrapper for iconv in all languages except esperanto,
2566 # which does some nasty x-conversions beforehand
2567
2568 # Even with //IGNORE iconv can whine about illegal characters in
2569 # *input* string. We just ignore those too.
2570 # REF: http://bugs.php.net/bug.php?id=37166
2571 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2572 wfSuppressWarnings();
2573 $text = iconv( $in, $out . '//IGNORE', $string );
2574 wfRestoreWarnings();
2575 return $text;
2576 }
2577
2578 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2579
2580 /**
2581 * @param array $matches
2582 * @return mixed|string
2583 */
2584 function ucwordbreaksCallbackAscii( $matches ) {
2585 return $this->ucfirst( $matches[1] );
2586 }
2587
2588 /**
2589 * @param array $matches
2590 * @return string
2591 */
2592 function ucwordbreaksCallbackMB( $matches ) {
2593 return mb_strtoupper( $matches[0] );
2594 }
2595
2596 /**
2597 * @param array $matches
2598 * @return string
2599 */
2600 function ucCallback( $matches ) {
2601 list( $wikiUpperChars ) = self::getCaseMaps();
2602 return strtr( $matches[1], $wikiUpperChars );
2603 }
2604
2605 /**
2606 * @param array $matches
2607 * @return string
2608 */
2609 function lcCallback( $matches ) {
2610 list( , $wikiLowerChars ) = self::getCaseMaps();
2611 return strtr( $matches[1], $wikiLowerChars );
2612 }
2613
2614 /**
2615 * @param array $matches
2616 * @return string
2617 */
2618 function ucwordsCallbackMB( $matches ) {
2619 return mb_strtoupper( $matches[0] );
2620 }
2621
2622 /**
2623 * @param array $matches
2624 * @return string
2625 */
2626 function ucwordsCallbackWiki( $matches ) {
2627 list( $wikiUpperChars ) = self::getCaseMaps();
2628 return strtr( $matches[0], $wikiUpperChars );
2629 }
2630
2631 /**
2632 * Make a string's first character uppercase
2633 *
2634 * @param string $str
2635 *
2636 * @return string
2637 */
2638 function ucfirst( $str ) {
2639 $o = ord( $str );
2640 if ( $o < 96 ) { // if already uppercase...
2641 return $str;
2642 } elseif ( $o < 128 ) {
2643 return ucfirst( $str ); // use PHP's ucfirst()
2644 } else {
2645 // fall back to more complex logic in case of multibyte strings
2646 return $this->uc( $str, true );
2647 }
2648 }
2649
2650 /**
2651 * Convert a string to uppercase
2652 *
2653 * @param string $str
2654 * @param bool $first
2655 *
2656 * @return string
2657 */
2658 function uc( $str, $first = false ) {
2659 if ( function_exists( 'mb_strtoupper' ) ) {
2660 if ( $first ) {
2661 if ( $this->isMultibyte( $str ) ) {
2662 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2663 } else {
2664 return ucfirst( $str );
2665 }
2666 } else {
2667 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2668 }
2669 } else {
2670 if ( $this->isMultibyte( $str ) ) {
2671 $x = $first ? '^' : '';
2672 return preg_replace_callback(
2673 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2674 array( $this, 'ucCallback' ),
2675 $str
2676 );
2677 } else {
2678 return $first ? ucfirst( $str ) : strtoupper( $str );
2679 }
2680 }
2681 }
2682
2683 /**
2684 * @param string $str
2685 * @return mixed|string
2686 */
2687 function lcfirst( $str ) {
2688 $o = ord( $str );
2689 if ( !$o ) {
2690 return strval( $str );
2691 } elseif ( $o >= 128 ) {
2692 return $this->lc( $str, true );
2693 } elseif ( $o > 96 ) {
2694 return $str;
2695 } else {
2696 $str[0] = strtolower( $str[0] );
2697 return $str;
2698 }
2699 }
2700
2701 /**
2702 * @param string $str
2703 * @param bool $first
2704 * @return mixed|string
2705 */
2706 function lc( $str, $first = false ) {
2707 if ( function_exists( 'mb_strtolower' ) ) {
2708 if ( $first ) {
2709 if ( $this->isMultibyte( $str ) ) {
2710 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2711 } else {
2712 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2713 }
2714 } else {
2715 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2716 }
2717 } else {
2718 if ( $this->isMultibyte( $str ) ) {
2719 $x = $first ? '^' : '';
2720 return preg_replace_callback(
2721 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2722 array( $this, 'lcCallback' ),
2723 $str
2724 );
2725 } else {
2726 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2727 }
2728 }
2729 }
2730
2731 /**
2732 * @param string $str
2733 * @return bool
2734 */
2735 function isMultibyte( $str ) {
2736 return (bool)preg_match( '/[\x80-\xff]/', $str );
2737 }
2738
2739 /**
2740 * @param string $str
2741 * @return mixed|string
2742 */
2743 function ucwords( $str ) {
2744 if ( $this->isMultibyte( $str ) ) {
2745 $str = $this->lc( $str );
2746
2747 // regexp to find first letter in each word (i.e. after each space)
2748 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2749
2750 // function to use to capitalize a single char
2751 if ( function_exists( 'mb_strtoupper' ) ) {
2752 return preg_replace_callback(
2753 $replaceRegexp,
2754 array( $this, 'ucwordsCallbackMB' ),
2755 $str
2756 );
2757 } else {
2758 return preg_replace_callback(
2759 $replaceRegexp,
2760 array( $this, 'ucwordsCallbackWiki' ),
2761 $str
2762 );
2763 }
2764 } else {
2765 return ucwords( strtolower( $str ) );
2766 }
2767 }
2768
2769 /**
2770 * capitalize words at word breaks
2771 *
2772 * @param string $str
2773 * @return mixed
2774 */
2775 function ucwordbreaks( $str ) {
2776 if ( $this->isMultibyte( $str ) ) {
2777 $str = $this->lc( $str );
2778
2779 // since \b doesn't work for UTF-8, we explicitely define word break chars
2780 $breaks = "[ \-\(\)\}\{\.,\?!]";
2781
2782 // find first letter after word break
2783 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|" .
2784 "$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2785
2786 if ( function_exists( 'mb_strtoupper' ) ) {
2787 return preg_replace_callback(
2788 $replaceRegexp,
2789 array( $this, 'ucwordbreaksCallbackMB' ),
2790 $str
2791 );
2792 } else {
2793 return preg_replace_callback(
2794 $replaceRegexp,
2795 array( $this, 'ucwordsCallbackWiki' ),
2796 $str
2797 );
2798 }
2799 } else {
2800 return preg_replace_callback(
2801 '/\b([\w\x80-\xff]+)\b/',
2802 array( $this, 'ucwordbreaksCallbackAscii' ),
2803 $str
2804 );
2805 }
2806 }
2807
2808 /**
2809 * Return a case-folded representation of $s
2810 *
2811 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2812 * and $s2 are the same except for the case of their characters. It is not
2813 * necessary for the value returned to make sense when displayed.
2814 *
2815 * Do *not* perform any other normalisation in this function. If a caller
2816 * uses this function when it should be using a more general normalisation
2817 * function, then fix the caller.
2818 *
2819 * @param string $s
2820 *
2821 * @return string
2822 */
2823 function caseFold( $s ) {
2824 return $this->uc( $s );
2825 }
2826
2827 /**
2828 * @param string $s
2829 * @return string
2830 */
2831 function checkTitleEncoding( $s ) {
2832 if ( is_array( $s ) ) {
2833 throw new MWException( 'Given array to checkTitleEncoding.' );
2834 }
2835 if ( StringUtils::isUtf8( $s ) ) {
2836 return $s;
2837 }
2838
2839 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2840 }
2841
2842 /**
2843 * @return array
2844 */
2845 function fallback8bitEncoding() {
2846 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2847 }
2848
2849 /**
2850 * Most writing systems use whitespace to break up words.
2851 * Some languages such as Chinese don't conventionally do this,
2852 * which requires special handling when breaking up words for
2853 * searching etc.
2854 *
2855 * @return bool
2856 */
2857 function hasWordBreaks() {
2858 return true;
2859 }
2860
2861 /**
2862 * Some languages such as Chinese require word segmentation,
2863 * Specify such segmentation when overridden in derived class.
2864 *
2865 * @param string $string
2866 * @return string
2867 */
2868 function segmentByWord( $string ) {
2869 return $string;
2870 }
2871
2872 /**
2873 * Some languages have special punctuation need to be normalized.
2874 * Make such changes here.
2875 *
2876 * @param string $string
2877 * @return string
2878 */
2879 function normalizeForSearch( $string ) {
2880 return self::convertDoubleWidth( $string );
2881 }
2882
2883 /**
2884 * convert double-width roman characters to single-width.
2885 * range: ff00-ff5f ~= 0020-007f
2886 *
2887 * @param string $string
2888 *
2889 * @return string
2890 */
2891 protected static function convertDoubleWidth( $string ) {
2892 static $full = null;
2893 static $half = null;
2894
2895 if ( $full === null ) {
2896 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2897 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2898 $full = str_split( $fullWidth, 3 );
2899 $half = str_split( $halfWidth );
2900 }
2901
2902 $string = str_replace( $full, $half, $string );
2903 return $string;
2904 }
2905
2906 /**
2907 * @param string $string
2908 * @param string $pattern
2909 * @return string
2910 */
2911 protected static function insertSpace( $string, $pattern ) {
2912 $string = preg_replace( $pattern, " $1 ", $string );
2913 $string = preg_replace( '/ +/', ' ', $string );
2914 return $string;
2915 }
2916
2917 /**
2918 * @param array $termsArray
2919 * @return array
2920 */
2921 function convertForSearchResult( $termsArray ) {
2922 # some languages, e.g. Chinese, need to do a conversion
2923 # in order for search results to be displayed correctly
2924 return $termsArray;
2925 }
2926
2927 /**
2928 * Get the first character of a string.
2929 *
2930 * @param string $s
2931 * @return string
2932 */
2933 function firstChar( $s ) {
2934 $matches = array();
2935 preg_match(
2936 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2937 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2938 $s,
2939 $matches
2940 );
2941
2942 if ( isset( $matches[1] ) ) {
2943 if ( strlen( $matches[1] ) != 3 ) {
2944 return $matches[1];
2945 }
2946
2947 // Break down Hangul syllables to grab the first jamo
2948 $code = UtfNormal\Utils::utf8ToCodepoint( $matches[1] );
2949 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2950 return $matches[1];
2951 } elseif ( $code < 0xb098 ) {
2952 return "\xe3\x84\xb1";
2953 } elseif ( $code < 0xb2e4 ) {
2954 return "\xe3\x84\xb4";
2955 } elseif ( $code < 0xb77c ) {
2956 return "\xe3\x84\xb7";
2957 } elseif ( $code < 0xb9c8 ) {
2958 return "\xe3\x84\xb9";
2959 } elseif ( $code < 0xbc14 ) {
2960 return "\xe3\x85\x81";
2961 } elseif ( $code < 0xc0ac ) {
2962 return "\xe3\x85\x82";
2963 } elseif ( $code < 0xc544 ) {
2964 return "\xe3\x85\x85";
2965 } elseif ( $code < 0xc790 ) {
2966 return "\xe3\x85\x87";
2967 } elseif ( $code < 0xcc28 ) {
2968 return "\xe3\x85\x88";
2969 } elseif ( $code < 0xce74 ) {
2970 return "\xe3\x85\x8a";
2971 } elseif ( $code < 0xd0c0 ) {
2972 return "\xe3\x85\x8b";
2973 } elseif ( $code < 0xd30c ) {
2974 return "\xe3\x85\x8c";
2975 } elseif ( $code < 0xd558 ) {
2976 return "\xe3\x85\x8d";
2977 } else {
2978 return "\xe3\x85\x8e";
2979 }
2980 } else {
2981 return '';
2982 }
2983 }
2984
2985 function initEncoding() {
2986 # Some languages may have an alternate char encoding option
2987 # (Esperanto X-coding, Japanese furigana conversion, etc)
2988 # If this language is used as the primary content language,
2989 # an override to the defaults can be set here on startup.
2990 }
2991
2992 /**
2993 * @param string $s
2994 * @return string
2995 */
2996 function recodeForEdit( $s ) {
2997 # For some languages we'll want to explicitly specify
2998 # which characters make it into the edit box raw
2999 # or are converted in some way or another.
3000 global $wgEditEncoding;
3001 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
3002 return $s;
3003 } else {
3004 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
3005 }
3006 }
3007
3008 /**
3009 * @param string $s
3010 * @return string
3011 */
3012 function recodeInput( $s ) {
3013 # Take the previous into account.
3014 global $wgEditEncoding;
3015 if ( $wgEditEncoding != '' ) {
3016 $enc = $wgEditEncoding;
3017 } else {
3018 $enc = 'UTF-8';
3019 }
3020 if ( $enc == 'UTF-8' ) {
3021 return $s;
3022 } else {
3023 return $this->iconv( $enc, 'UTF-8', $s );
3024 }
3025 }
3026
3027 /**
3028 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
3029 * also cleans up certain backwards-compatible sequences, converting them
3030 * to the modern Unicode equivalent.
3031 *
3032 * This is language-specific for performance reasons only.
3033 *
3034 * @param string $s
3035 *
3036 * @return string
3037 */
3038 function normalize( $s ) {
3039 global $wgAllUnicodeFixes;
3040 $s = UtfNormal\Validator::cleanUp( $s );
3041 if ( $wgAllUnicodeFixes ) {
3042 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
3043 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
3044 }
3045
3046 return $s;
3047 }
3048
3049 /**
3050 * Transform a string using serialized data stored in the given file (which
3051 * must be in the serialized subdirectory of $IP). The file contains pairs
3052 * mapping source characters to destination characters.
3053 *
3054 * The data is cached in process memory. This will go faster if you have the
3055 * FastStringSearch extension.
3056 *
3057 * @param string $file
3058 * @param string $string
3059 *
3060 * @throws MWException
3061 * @return string
3062 */
3063 function transformUsingPairFile( $file, $string ) {
3064 if ( !isset( $this->transformData[$file] ) ) {
3065 $data = wfGetPrecompiledData( $file );
3066 if ( $data === false ) {
3067 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
3068 }
3069 $this->transformData[$file] = new ReplacementArray( $data );
3070 }
3071 return $this->transformData[$file]->replace( $string );
3072 }
3073
3074 /**
3075 * For right-to-left language support
3076 *
3077 * @return bool
3078 */
3079 function isRTL() {
3080 return self::$dataCache->getItem( $this->mCode, 'rtl' );
3081 }
3082
3083 /**
3084 * Return the correct HTML 'dir' attribute value for this language.
3085 * @return string
3086 */
3087 function getDir() {
3088 return $this->isRTL() ? 'rtl' : 'ltr';
3089 }
3090
3091 /**
3092 * Return 'left' or 'right' as appropriate alignment for line-start
3093 * for this language's text direction.
3094 *
3095 * Should be equivalent to CSS3 'start' text-align value....
3096 *
3097 * @return string
3098 */
3099 function alignStart() {
3100 return $this->isRTL() ? 'right' : 'left';
3101 }
3102
3103 /**
3104 * Return 'right' or 'left' as appropriate alignment for line-end
3105 * for this language's text direction.
3106 *
3107 * Should be equivalent to CSS3 'end' text-align value....
3108 *
3109 * @return string
3110 */
3111 function alignEnd() {
3112 return $this->isRTL() ? 'left' : 'right';
3113 }
3114
3115 /**
3116 * A hidden direction mark (LRM or RLM), depending on the language direction.
3117 * Unlike getDirMark(), this function returns the character as an HTML entity.
3118 * This function should be used when the output is guaranteed to be HTML,
3119 * because it makes the output HTML source code more readable. When
3120 * the output is plain text or can be escaped, getDirMark() should be used.
3121 *
3122 * @param bool $opposite Get the direction mark opposite to your language
3123 * @return string
3124 * @since 1.20
3125 */
3126 function getDirMarkEntity( $opposite = false ) {
3127 if ( $opposite ) {
3128 return $this->isRTL() ? '&lrm;' : '&rlm;';
3129 }
3130 return $this->isRTL() ? '&rlm;' : '&lrm;';
3131 }
3132
3133 /**
3134 * A hidden direction mark (LRM or RLM), depending on the language direction.
3135 * This function produces them as invisible Unicode characters and
3136 * the output may be hard to read and debug, so it should only be used
3137 * when the output is plain text or can be escaped. When the output is
3138 * HTML, use getDirMarkEntity() instead.
3139 *
3140 * @param bool $opposite Get the direction mark opposite to your language
3141 * @return string
3142 */
3143 function getDirMark( $opposite = false ) {
3144 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
3145 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
3146 if ( $opposite ) {
3147 return $this->isRTL() ? $lrm : $rlm;
3148 }
3149 return $this->isRTL() ? $rlm : $lrm;
3150 }
3151
3152 /**
3153 * @return array
3154 */
3155 function capitalizeAllNouns() {
3156 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
3157 }
3158
3159 /**
3160 * An arrow, depending on the language direction.
3161 *
3162 * @param string $direction The direction of the arrow: forwards (default),
3163 * backwards, left, right, up, down.
3164 * @return string
3165 */
3166 function getArrow( $direction = 'forwards' ) {
3167 switch ( $direction ) {
3168 case 'forwards':
3169 return $this->isRTL() ? '←' : '→';
3170 case 'backwards':
3171 return $this->isRTL() ? '→' : '←';
3172 case 'left':
3173 return '←';
3174 case 'right':
3175 return '→';
3176 case 'up':
3177 return '↑';
3178 case 'down':
3179 return '↓';
3180 }
3181 }
3182
3183 /**
3184 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
3185 *
3186 * @return bool
3187 */
3188 function linkPrefixExtension() {
3189 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
3190 }
3191
3192 /**
3193 * Get all magic words from cache.
3194 * @return array
3195 */
3196 function getMagicWords() {
3197 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
3198 }
3199
3200 /**
3201 * Run the LanguageGetMagic hook once.
3202 */
3203 protected function doMagicHook() {
3204 if ( $this->mMagicHookDone ) {
3205 return;
3206 }
3207 $this->mMagicHookDone = true;
3208 Hooks::run( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
3209 }
3210
3211 /**
3212 * Fill a MagicWord object with data from here
3213 *
3214 * @param MagicWord $mw
3215 */
3216 function getMagic( $mw ) {
3217 // Saves a function call
3218 if ( !$this->mMagicHookDone ) {
3219 $this->doMagicHook();
3220 }
3221
3222 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3223 $rawEntry = $this->mMagicExtensions[$mw->mId];
3224 } else {
3225 $rawEntry = self::$dataCache->getSubitem(
3226 $this->mCode, 'magicWords', $mw->mId );
3227 }
3228
3229 if ( !is_array( $rawEntry ) ) {
3230 wfWarn( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3231 } else {
3232 $mw->mCaseSensitive = $rawEntry[0];
3233 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3234 }
3235 }
3236
3237 /**
3238 * Add magic words to the extension array
3239 *
3240 * @param array $newWords
3241 */
3242 function addMagicWordsByLang( $newWords ) {
3243 $fallbackChain = $this->getFallbackLanguages();
3244 $fallbackChain = array_reverse( $fallbackChain );
3245 foreach ( $fallbackChain as $code ) {
3246 if ( isset( $newWords[$code] ) ) {
3247 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3248 }
3249 }
3250 }
3251
3252 /**
3253 * Get special page names, as an associative array
3254 * canonical name => array of valid names, including aliases
3255 * @return array
3256 */
3257 function getSpecialPageAliases() {
3258 // Cache aliases because it may be slow to load them
3259 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3260 // Initialise array
3261 $this->mExtendedSpecialPageAliases =
3262 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3263 Hooks::run( 'LanguageGetSpecialPageAliases',
3264 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
3265 }
3266
3267 return $this->mExtendedSpecialPageAliases;
3268 }
3269
3270 /**
3271 * Italic is unsuitable for some languages
3272 *
3273 * @param string $text The text to be emphasized.
3274 * @return string
3275 */
3276 function emphasize( $text ) {
3277 return "<em>$text</em>";
3278 }
3279
3280 /**
3281 * Normally we output all numbers in plain en_US style, that is
3282 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3283 * point twohundredthirtyfive. However this is not suitable for all
3284 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3285 * Icelandic just want to use commas instead of dots, and dots instead
3286 * of commas like "293.291,235".
3287 *
3288 * An example of this function being called:
3289 * <code>
3290 * wfMessage( 'message' )->numParams( $num )->text()
3291 * </code>
3292 *
3293 * See $separatorTransformTable on MessageIs.php for
3294 * the , => . and . => , implementation.
3295 *
3296 * @todo check if it's viable to use localeconv() for the decimal separator thing.
3297 * @param int|float $number The string to be formatted, should be an integer
3298 * or a floating point number.
3299 * @param bool $nocommafy Set to true for special numbers like dates
3300 * @return string
3301 */
3302 public function formatNum( $number, $nocommafy = false ) {
3303 global $wgTranslateNumerals;
3304 if ( !$nocommafy ) {
3305 $number = $this->commafy( $number );
3306 $s = $this->separatorTransformTable();
3307 if ( $s ) {
3308 $number = strtr( $number, $s );
3309 }
3310 }
3311
3312 if ( $wgTranslateNumerals ) {
3313 $s = $this->digitTransformTable();
3314 if ( $s ) {
3315 $number = strtr( $number, $s );
3316 }
3317 }
3318
3319 return $number;
3320 }
3321
3322 /**
3323 * Front-end for non-commafied formatNum
3324 *
3325 * @param int|float $number The string to be formatted, should be an integer
3326 * or a floating point number.
3327 * @since 1.21
3328 * @return string
3329 */
3330 public function formatNumNoSeparators( $number ) {
3331 return $this->formatNum( $number, true );
3332 }
3333
3334 /**
3335 * @param string $number
3336 * @return string
3337 */
3338 public function parseFormattedNumber( $number ) {
3339 $s = $this->digitTransformTable();
3340 if ( $s ) {
3341 // eliminate empty array values such as ''. (bug 64347)
3342 $s = array_filter( $s );
3343 $number = strtr( $number, array_flip( $s ) );
3344 }
3345
3346 $s = $this->separatorTransformTable();
3347 if ( $s ) {
3348 // eliminate empty array values such as ''. (bug 64347)
3349 $s = array_filter( $s );
3350 $number = strtr( $number, array_flip( $s ) );
3351 }
3352
3353 $number = strtr( $number, array( ',' => '' ) );
3354 return $number;
3355 }
3356
3357 /**
3358 * Adds commas to a given number
3359 * @since 1.19
3360 * @param mixed $number
3361 * @return string
3362 */
3363 function commafy( $number ) {
3364 $digitGroupingPattern = $this->digitGroupingPattern();
3365 if ( $number === null ) {
3366 return '';
3367 }
3368
3369 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3370 // default grouping is at thousands, use the same for ###,###,### pattern too.
3371 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3372 } else {
3373 // Ref: http://cldr.unicode.org/translation/number-patterns
3374 $sign = "";
3375 if ( intval( $number ) < 0 ) {
3376 // For negative numbers apply the algorithm like positive number and add sign.
3377 $sign = "-";
3378 $number = substr( $number, 1 );
3379 }
3380 $integerPart = array();
3381 $decimalPart = array();
3382 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3383 preg_match( "/\d+/", $number, $integerPart );
3384 preg_match( "/\.\d*/", $number, $decimalPart );
3385 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3386 if ( $groupedNumber === $number ) {
3387 // the string does not have any number part. Eg: .12345
3388 return $sign . $groupedNumber;
3389 }
3390 $start = $end = ($integerPart) ? strlen( $integerPart[0] ) : 0;
3391 while ( $start > 0 ) {
3392 $match = $matches[0][$numMatches - 1];
3393 $matchLen = strlen( $match );
3394 $start = $end - $matchLen;
3395 if ( $start < 0 ) {
3396 $start = 0;
3397 }
3398 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3399 $end = $start;
3400 if ( $numMatches > 1 ) {
3401 // use the last pattern for the rest of the number
3402 $numMatches--;
3403 }
3404 if ( $start > 0 ) {
3405 $groupedNumber = "," . $groupedNumber;
3406 }
3407 }
3408 return $sign . $groupedNumber;
3409 }
3410 }
3411
3412 /**
3413 * @return string
3414 */
3415 function digitGroupingPattern() {
3416 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3417 }
3418
3419 /**
3420 * @return array
3421 */
3422 function digitTransformTable() {
3423 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3424 }
3425
3426 /**
3427 * @return array
3428 */
3429 function separatorTransformTable() {
3430 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3431 }
3432
3433 /**
3434 * Take a list of strings and build a locale-friendly comma-separated
3435 * list, using the local comma-separator message.
3436 * The last two strings are chained with an "and".
3437 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3438 *
3439 * @param string[] $l
3440 * @return string
3441 */
3442 function listToText( array $l ) {
3443 $m = count( $l ) - 1;
3444 if ( $m < 0 ) {
3445 return '';
3446 }
3447 if ( $m > 0 ) {
3448 $and = $this->msg( 'and' )->escaped();
3449 $space = $this->msg( 'word-separator' )->escaped();
3450 if ( $m > 1 ) {
3451 $comma = $this->msg( 'comma-separator' )->escaped();
3452 }
3453 }
3454 $s = $l[$m];
3455 for ( $i = $m - 1; $i >= 0; $i-- ) {
3456 if ( $i == $m - 1 ) {
3457 $s = $l[$i] . $and . $space . $s;
3458 } else {
3459 $s = $l[$i] . $comma . $s;
3460 }
3461 }
3462 return $s;
3463 }
3464
3465 /**
3466 * Take a list of strings and build a locale-friendly comma-separated
3467 * list, using the local comma-separator message.
3468 * @param string[] $list Array of strings to put in a comma list
3469 * @return string
3470 */
3471 function commaList( array $list ) {
3472 return implode(
3473 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3474 $list
3475 );
3476 }
3477
3478 /**
3479 * Take a list of strings and build a locale-friendly semicolon-separated
3480 * list, using the local semicolon-separator message.
3481 * @param string[] $list Array of strings to put in a semicolon list
3482 * @return string
3483 */
3484 function semicolonList( array $list ) {
3485 return implode(
3486 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3487 $list
3488 );
3489 }
3490
3491 /**
3492 * Same as commaList, but separate it with the pipe instead.
3493 * @param string[] $list Array of strings to put in a pipe list
3494 * @return string
3495 */
3496 function pipeList( array $list ) {
3497 return implode(
3498 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3499 $list
3500 );
3501 }
3502
3503 /**
3504 * Truncate a string to a specified length in bytes, appending an optional
3505 * string (e.g. for ellipses)
3506 *
3507 * The database offers limited byte lengths for some columns in the database;
3508 * multi-byte character sets mean we need to ensure that only whole characters
3509 * are included, otherwise broken characters can be passed to the user
3510 *
3511 * If $length is negative, the string will be truncated from the beginning
3512 *
3513 * @param string $string String to truncate
3514 * @param int $length Maximum length (including ellipses)
3515 * @param string $ellipsis String to append to the truncated text
3516 * @param bool $adjustLength Subtract length of ellipsis from $length.
3517 * $adjustLength was introduced in 1.18, before that behaved as if false.
3518 * @return string
3519 */
3520 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3521 # Use the localized ellipsis character
3522 if ( $ellipsis == '...' ) {
3523 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3524 }
3525 # Check if there is no need to truncate
3526 if ( $length == 0 ) {
3527 return $ellipsis; // convention
3528 } elseif ( strlen( $string ) <= abs( $length ) ) {
3529 return $string; // no need to truncate
3530 }
3531 $stringOriginal = $string;
3532 # If ellipsis length is >= $length then we can't apply $adjustLength
3533 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3534 $string = $ellipsis; // this can be slightly unexpected
3535 # Otherwise, truncate and add ellipsis...
3536 } else {
3537 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3538 if ( $length > 0 ) {
3539 $length -= $eLength;
3540 $string = substr( $string, 0, $length ); // xyz...
3541 $string = $this->removeBadCharLast( $string );
3542 $string = rtrim( $string );
3543 $string = $string . $ellipsis;
3544 } else {
3545 $length += $eLength;
3546 $string = substr( $string, $length ); // ...xyz
3547 $string = $this->removeBadCharFirst( $string );
3548 $string = ltrim( $string );
3549 $string = $ellipsis . $string;
3550 }
3551 }
3552 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3553 # This check is *not* redundant if $adjustLength, due to the single case where
3554 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3555 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3556 return $string;
3557 } else {
3558 return $stringOriginal;
3559 }
3560 }
3561
3562 /**
3563 * Remove bytes that represent an incomplete Unicode character
3564 * at the end of string (e.g. bytes of the char are missing)
3565 *
3566 * @param string $string
3567 * @return string
3568 */
3569 protected function removeBadCharLast( $string ) {
3570 if ( $string != '' ) {
3571 $char = ord( $string[strlen( $string ) - 1] );
3572 $m = array();
3573 if ( $char >= 0xc0 ) {
3574 # We got the first byte only of a multibyte char; remove it.
3575 $string = substr( $string, 0, -1 );
3576 } elseif ( $char >= 0x80 &&
3577 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3578 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3579 ) {
3580 # We chopped in the middle of a character; remove it
3581 $string = $m[1];
3582 }
3583 }
3584 return $string;
3585 }
3586
3587 /**
3588 * Remove bytes that represent an incomplete Unicode character
3589 * at the start of string (e.g. bytes of the char are missing)
3590 *
3591 * @param string $string
3592 * @return string
3593 */
3594 protected function removeBadCharFirst( $string ) {
3595 if ( $string != '' ) {
3596 $char = ord( $string[0] );
3597 if ( $char >= 0x80 && $char < 0xc0 ) {
3598 # We chopped in the middle of a character; remove the whole thing
3599 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3600 }
3601 }
3602 return $string;
3603 }
3604
3605 /**
3606 * Truncate a string of valid HTML to a specified length in bytes,
3607 * appending an optional string (e.g. for ellipses), and return valid HTML
3608 *
3609 * This is only intended for styled/linked text, such as HTML with
3610 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3611 * Also, this will not detect things like "display:none" CSS.
3612 *
3613 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3614 *
3615 * @param string $text HTML string to truncate
3616 * @param int $length (zero/positive) Maximum length (including ellipses)
3617 * @param string $ellipsis String to append to the truncated text
3618 * @return string
3619 */
3620 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3621 # Use the localized ellipsis character
3622 if ( $ellipsis == '...' ) {
3623 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3624 }
3625 # Check if there is clearly no need to truncate
3626 if ( $length <= 0 ) {
3627 return $ellipsis; // no text shown, nothing to format (convention)
3628 } elseif ( strlen( $text ) <= $length ) {
3629 return $text; // string short enough even *with* HTML (short-circuit)
3630 }
3631
3632 $dispLen = 0; // innerHTML legth so far
3633 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3634 $tagType = 0; // 0-open, 1-close
3635 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3636 $entityState = 0; // 0-not entity, 1-entity
3637 $tag = $ret = ''; // accumulated tag name, accumulated result string
3638 $openTags = array(); // open tag stack
3639 $maybeState = null; // possible truncation state
3640
3641 $textLen = strlen( $text );
3642 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3643 for ( $pos = 0; true; ++$pos ) {
3644 # Consider truncation once the display length has reached the maximim.
3645 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3646 # Check that we're not in the middle of a bracket/entity...
3647 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3648 if ( !$testingEllipsis ) {
3649 $testingEllipsis = true;
3650 # Save where we are; we will truncate here unless there turn out to
3651 # be so few remaining characters that truncation is not necessary.
3652 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3653 $maybeState = array( $ret, $openTags ); // save state
3654 }
3655 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3656 # String in fact does need truncation, the truncation point was OK.
3657 list( $ret, $openTags ) = $maybeState; // reload state
3658 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3659 $ret .= $ellipsis; // add ellipsis
3660 break;
3661 }
3662 }
3663 if ( $pos >= $textLen ) {
3664 break; // extra iteration just for above checks
3665 }
3666
3667 # Read the next char...
3668 $ch = $text[$pos];
3669 $lastCh = $pos ? $text[$pos - 1] : '';
3670 $ret .= $ch; // add to result string
3671 if ( $ch == '<' ) {
3672 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3673 $entityState = 0; // for bad HTML
3674 $bracketState = 1; // tag started (checking for backslash)
3675 } elseif ( $ch == '>' ) {
3676 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3677 $entityState = 0; // for bad HTML
3678 $bracketState = 0; // out of brackets
3679 } elseif ( $bracketState == 1 ) {
3680 if ( $ch == '/' ) {
3681 $tagType = 1; // close tag (e.g. "</span>")
3682 } else {
3683 $tagType = 0; // open tag (e.g. "<span>")
3684 $tag .= $ch;
3685 }
3686 $bracketState = 2; // building tag name
3687 } elseif ( $bracketState == 2 ) {
3688 if ( $ch != ' ' ) {
3689 $tag .= $ch;
3690 } else {
3691 // Name found (e.g. "<a href=..."), add on tag attributes...
3692 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3693 }
3694 } elseif ( $bracketState == 0 ) {
3695 if ( $entityState ) {
3696 if ( $ch == ';' ) {
3697 $entityState = 0;
3698 $dispLen++; // entity is one displayed char
3699 }
3700 } else {
3701 if ( $neLength == 0 && !$maybeState ) {
3702 // Save state without $ch. We want to *hit* the first
3703 // display char (to get tags) but not *use* it if truncating.
3704 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3705 }
3706 if ( $ch == '&' ) {
3707 $entityState = 1; // entity found, (e.g. "&#160;")
3708 } else {
3709 $dispLen++; // this char is displayed
3710 // Add the next $max display text chars after this in one swoop...
3711 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3712 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3713 $dispLen += $skipped;
3714 $pos += $skipped;
3715 }
3716 }
3717 }
3718 }
3719 // Close the last tag if left unclosed by bad HTML
3720 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3721 while ( count( $openTags ) > 0 ) {
3722 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3723 }
3724 return $ret;
3725 }
3726
3727 /**
3728 * truncateHtml() helper function
3729 * like strcspn() but adds the skipped chars to $ret
3730 *
3731 * @param string $ret
3732 * @param string $text
3733 * @param string $search
3734 * @param int $start
3735 * @param null|int $len
3736 * @return int
3737 */
3738 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3739 if ( $len === null ) {
3740 $len = -1; // -1 means "no limit" for strcspn
3741 } elseif ( $len < 0 ) {
3742 $len = 0; // sanity
3743 }
3744 $skipCount = 0;
3745 if ( $start < strlen( $text ) ) {
3746 $skipCount = strcspn( $text, $search, $start, $len );
3747 $ret .= substr( $text, $start, $skipCount );
3748 }
3749 return $skipCount;
3750 }
3751
3752 /**
3753 * truncateHtml() helper function
3754 * (a) push or pop $tag from $openTags as needed
3755 * (b) clear $tag value
3756 * @param string &$tag Current HTML tag name we are looking at
3757 * @param int $tagType (0-open tag, 1-close tag)
3758 * @param string $lastCh Character before the '>' that ended this tag
3759 * @param array &$openTags Open tag stack (not accounting for $tag)
3760 */
3761 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3762 $tag = ltrim( $tag );
3763 if ( $tag != '' ) {
3764 if ( $tagType == 0 && $lastCh != '/' ) {
3765 $openTags[] = $tag; // tag opened (didn't close itself)
3766 } elseif ( $tagType == 1 ) {
3767 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3768 array_pop( $openTags ); // tag closed
3769 }
3770 }
3771 $tag = '';
3772 }
3773 }
3774
3775 /**
3776 * Grammatical transformations, needed for inflected languages
3777 * Invoked by putting {{grammar:case|word}} in a message
3778 *
3779 * @param string $word
3780 * @param string $case
3781 * @return string
3782 */
3783 function convertGrammar( $word, $case ) {
3784 global $wgGrammarForms;
3785 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3786 return $wgGrammarForms[$this->getCode()][$case][$word];
3787 }
3788
3789 return $word;
3790 }
3791 /**
3792 * Get the grammar forms for the content language
3793 * @return array Array of grammar forms
3794 * @since 1.20
3795 */
3796 function getGrammarForms() {
3797 global $wgGrammarForms;
3798 if ( isset( $wgGrammarForms[$this->getCode()] )
3799 && is_array( $wgGrammarForms[$this->getCode()] )
3800 ) {
3801 return $wgGrammarForms[$this->getCode()];
3802 }
3803
3804 return array();
3805 }
3806 /**
3807 * Provides an alternative text depending on specified gender.
3808 * Usage {{gender:username|masculine|feminine|unknown}}.
3809 * username is optional, in which case the gender of current user is used,
3810 * but only in (some) interface messages; otherwise default gender is used.
3811 *
3812 * If no forms are given, an empty string is returned. If only one form is
3813 * given, it will be returned unconditionally. These details are implied by
3814 * the caller and cannot be overridden in subclasses.
3815 *
3816 * If three forms are given, the default is to use the third (unknown) form.
3817 * If fewer than three forms are given, the default is to use the first (masculine) form.
3818 * These details can be overridden in subclasses.
3819 *
3820 * @param string $gender
3821 * @param array $forms
3822 *
3823 * @return string
3824 */
3825 function gender( $gender, $forms ) {
3826 if ( !count( $forms ) ) {
3827 return '';
3828 }
3829 $forms = $this->preConvertPlural( $forms, 2 );
3830 if ( $gender === 'male' ) {
3831 return $forms[0];
3832 }
3833 if ( $gender === 'female' ) {
3834 return $forms[1];
3835 }
3836 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3837 }
3838
3839 /**
3840 * Plural form transformations, needed for some languages.
3841 * For example, there are 3 form of plural in Russian and Polish,
3842 * depending on "count mod 10". See [[w:Plural]]
3843 * For English it is pretty simple.
3844 *
3845 * Invoked by putting {{plural:count|wordform1|wordform2}}
3846 * or {{plural:count|wordform1|wordform2|wordform3}}
3847 *
3848 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3849 *
3850 * @param int $count Non-localized number
3851 * @param array $forms Different plural forms
3852 * @return string Correct form of plural for $count in this language
3853 */
3854 function convertPlural( $count, $forms ) {
3855 // Handle explicit n=pluralform cases
3856 $forms = $this->handleExplicitPluralForms( $count, $forms );
3857 if ( is_string( $forms ) ) {
3858 return $forms;
3859 }
3860 if ( !count( $forms ) ) {
3861 return '';
3862 }
3863
3864 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3865 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3866 return $forms[$pluralForm];
3867 }
3868
3869 /**
3870 * Handles explicit plural forms for Language::convertPlural()
3871 *
3872 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3873 * If an explicitly defined plural form matches the $count, then
3874 * string value returned, otherwise array returned for further consideration
3875 * by CLDR rules or overridden convertPlural().
3876 *
3877 * @since 1.23
3878 *
3879 * @param int $count Non-localized number
3880 * @param array $forms Different plural forms
3881 *
3882 * @return array|string
3883 */
3884 protected function handleExplicitPluralForms( $count, array $forms ) {
3885 foreach ( $forms as $index => $form ) {
3886 if ( preg_match( '/\d+=/i', $form ) ) {
3887 $pos = strpos( $form, '=' );
3888 if ( substr( $form, 0, $pos ) === (string)$count ) {
3889 return substr( $form, $pos + 1 );
3890 }
3891 unset( $forms[$index] );
3892 }
3893 }
3894 return array_values( $forms );
3895 }
3896
3897 /**
3898 * Checks that convertPlural was given an array and pads it to requested
3899 * amount of forms by copying the last one.
3900 *
3901 * @param array $forms Array of forms given to convertPlural
3902 * @param int $count How many forms should there be at least
3903 * @return array Padded array of forms or an exception if not an array
3904 */
3905 protected function preConvertPlural( /* Array */ $forms, $count ) {
3906 while ( count( $forms ) < $count ) {
3907 $forms[] = $forms[count( $forms ) - 1];
3908 }
3909 return $forms;
3910 }
3911
3912 /**
3913 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3914 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3915 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3916 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3917 * match up with it.
3918 *
3919 * @param string $str The validated block duration in English
3920 * @return string Somehow translated block duration
3921 * @see LanguageFi.php for example implementation
3922 */
3923 function translateBlockExpiry( $str ) {
3924 $duration = SpecialBlock::getSuggestedDurations( $this );
3925 foreach ( $duration as $show => $value ) {
3926 if ( strcmp( $str, $value ) == 0 ) {
3927 return htmlspecialchars( trim( $show ) );
3928 }
3929 }
3930
3931 if ( wfIsInfinity( $str ) ) {
3932 foreach ( $duration as $show => $value ) {
3933 if ( wfIsInfinity( $value ) ) {
3934 return htmlspecialchars( trim( $show ) );
3935 }
3936 }
3937 }
3938
3939 // If all else fails, return a standard duration or timestamp description.
3940 $time = strtotime( $str, 0 );
3941 if ( $time === false ) { // Unknown format. Return it as-is in case.
3942 return $str;
3943 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3944 // $time is relative to 0 so it's a duration length.
3945 return $this->formatDuration( $time );
3946 } else { // It's an absolute timestamp.
3947 if ( $time === 0 ) {
3948 // wfTimestamp() handles 0 as current time instead of epoch.
3949 return $this->timeanddate( '19700101000000' );
3950 } else {
3951 return $this->timeanddate( $time );
3952 }
3953 }
3954 }
3955
3956 /**
3957 * languages like Chinese need to be segmented in order for the diff
3958 * to be of any use
3959 *
3960 * @param string $text
3961 * @return string
3962 */
3963 public function segmentForDiff( $text ) {
3964 return $text;
3965 }
3966
3967 /**
3968 * and unsegment to show the result
3969 *
3970 * @param string $text
3971 * @return string
3972 */
3973 public function unsegmentForDiff( $text ) {
3974 return $text;
3975 }
3976
3977 /**
3978 * Return the LanguageConverter used in the Language
3979 *
3980 * @since 1.19
3981 * @return LanguageConverter
3982 */
3983 public function getConverter() {
3984 return $this->mConverter;
3985 }
3986
3987 /**
3988 * convert text to all supported variants
3989 *
3990 * @param string $text
3991 * @return array
3992 */
3993 public function autoConvertToAllVariants( $text ) {
3994 return $this->mConverter->autoConvertToAllVariants( $text );
3995 }
3996
3997 /**
3998 * convert text to different variants of a language.
3999 *
4000 * @param string $text
4001 * @return string
4002 */
4003 public function convert( $text ) {
4004 return $this->mConverter->convert( $text );
4005 }
4006
4007 /**
4008 * Convert a Title object to a string in the preferred variant
4009 *
4010 * @param Title $title
4011 * @return string
4012 */
4013 public function convertTitle( $title ) {
4014 return $this->mConverter->convertTitle( $title );
4015 }
4016
4017 /**
4018 * Convert a namespace index to a string in the preferred variant
4019 *
4020 * @param int $ns
4021 * @return string
4022 */
4023 public function convertNamespace( $ns ) {
4024 return $this->mConverter->convertNamespace( $ns );
4025 }
4026
4027 /**
4028 * Check if this is a language with variants
4029 *
4030 * @return bool
4031 */
4032 public function hasVariants() {
4033 return count( $this->getVariants() ) > 1;
4034 }
4035
4036 /**
4037 * Check if the language has the specific variant
4038 *
4039 * @since 1.19
4040 * @param string $variant
4041 * @return bool
4042 */
4043 public function hasVariant( $variant ) {
4044 return (bool)$this->mConverter->validateVariant( $variant );
4045 }
4046
4047 /**
4048 * Put custom tags (e.g. -{ }-) around math to prevent conversion
4049 *
4050 * @param string $text
4051 * @return string
4052 * @deprecated since 1.22 is no longer used
4053 */
4054 public function armourMath( $text ) {
4055 return $this->mConverter->armourMath( $text );
4056 }
4057
4058 /**
4059 * Perform output conversion on a string, and encode for safe HTML output.
4060 * @param string $text Text to be converted
4061 * @param bool $isTitle Whether this conversion is for the article title
4062 * @return string
4063 * @todo this should get integrated somewhere sane
4064 */
4065 public function convertHtml( $text, $isTitle = false ) {
4066 return htmlspecialchars( $this->convert( $text, $isTitle ) );
4067 }
4068
4069 /**
4070 * @param string $key
4071 * @return string
4072 */
4073 public function convertCategoryKey( $key ) {
4074 return $this->mConverter->convertCategoryKey( $key );
4075 }
4076
4077 /**
4078 * Get the list of variants supported by this language
4079 * see sample implementation in LanguageZh.php
4080 *
4081 * @return array An array of language codes
4082 */
4083 public function getVariants() {
4084 return $this->mConverter->getVariants();
4085 }
4086
4087 /**
4088 * @return string
4089 */
4090 public function getPreferredVariant() {
4091 return $this->mConverter->getPreferredVariant();
4092 }
4093
4094 /**
4095 * @return string
4096 */
4097 public function getDefaultVariant() {
4098 return $this->mConverter->getDefaultVariant();
4099 }
4100
4101 /**
4102 * @return string
4103 */
4104 public function getURLVariant() {
4105 return $this->mConverter->getURLVariant();
4106 }
4107
4108 /**
4109 * If a language supports multiple variants, it is
4110 * possible that non-existing link in one variant
4111 * actually exists in another variant. this function
4112 * tries to find it. See e.g. LanguageZh.php
4113 * The input parameters may be modified upon return
4114 *
4115 * @param string &$link The name of the link
4116 * @param Title &$nt The title object of the link
4117 * @param bool $ignoreOtherCond To disable other conditions when
4118 * we need to transclude a template or update a category's link
4119 */
4120 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
4121 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
4122 }
4123
4124 /**
4125 * returns language specific options used by User::getPageRenderHash()
4126 * for example, the preferred language variant
4127 *
4128 * @return string
4129 */
4130 function getExtraHashOptions() {
4131 return $this->mConverter->getExtraHashOptions();
4132 }
4133
4134 /**
4135 * For languages that support multiple variants, the title of an
4136 * article may be displayed differently in different variants. this
4137 * function returns the apporiate title defined in the body of the article.
4138 *
4139 * @return string
4140 */
4141 public function getParsedTitle() {
4142 return $this->mConverter->getParsedTitle();
4143 }
4144
4145 /**
4146 * Prepare external link text for conversion. When the text is
4147 * a URL, it shouldn't be converted, and it'll be wrapped in
4148 * the "raw" tag (-{R| }-) to prevent conversion.
4149 *
4150 * This function is called "markNoConversion" for historical
4151 * reasons.
4152 *
4153 * @param string $text Text to be used for external link
4154 * @param bool $noParse Wrap it without confirming it's a real URL first
4155 * @return string The tagged text
4156 */
4157 public function markNoConversion( $text, $noParse = false ) {
4158 // Excluding protocal-relative URLs may avoid many false positives.
4159 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
4160 return $this->mConverter->markNoConversion( $text );
4161 } else {
4162 return $text;
4163 }
4164 }
4165
4166 /**
4167 * A regular expression to match legal word-trailing characters
4168 * which should be merged onto a link of the form [[foo]]bar.
4169 *
4170 * @return string
4171 */
4172 public function linkTrail() {
4173 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
4174 }
4175
4176 /**
4177 * A regular expression character set to match legal word-prefixing
4178 * characters which should be merged onto a link of the form foo[[bar]].
4179 *
4180 * @return string
4181 */
4182 public function linkPrefixCharset() {
4183 return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
4184 }
4185
4186 /**
4187 * @deprecated since 1.24, will be removed in 1.25
4188 * @return Language
4189 */
4190 function getLangObj() {
4191 wfDeprecated( __METHOD__, '1.24' );
4192 return $this;
4193 }
4194
4195 /**
4196 * Get the "parent" language which has a converter to convert a "compatible" language
4197 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
4198 *
4199 * @return Language|null
4200 * @since 1.22
4201 */
4202 public function getParentLanguage() {
4203 if ( $this->mParentLanguage !== false ) {
4204 return $this->mParentLanguage;
4205 }
4206
4207 $pieces = explode( '-', $this->getCode() );
4208 $code = $pieces[0];
4209 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
4210 $this->mParentLanguage = null;
4211 return null;
4212 }
4213 $lang = Language::factory( $code );
4214 if ( !$lang->hasVariant( $this->getCode() ) ) {
4215 $this->mParentLanguage = null;
4216 return null;
4217 }
4218
4219 $this->mParentLanguage = $lang;
4220 return $lang;
4221 }
4222
4223 /**
4224 * Get the RFC 3066 code for this language object
4225 *
4226 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4227 * htmlspecialchars() or similar
4228 *
4229 * @return string
4230 */
4231 public function getCode() {
4232 return $this->mCode;
4233 }
4234
4235 /**
4236 * Get the code in Bcp47 format which we can use
4237 * inside of html lang="" tags.
4238 *
4239 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4240 * htmlspecialchars() or similar.
4241 *
4242 * @since 1.19
4243 * @return string
4244 */
4245 public function getHtmlCode() {
4246 if ( is_null( $this->mHtmlCode ) ) {
4247 $this->mHtmlCode = wfBCP47( $this->getCode() );
4248 }
4249 return $this->mHtmlCode;
4250 }
4251
4252 /**
4253 * @param string $code
4254 */
4255 public function setCode( $code ) {
4256 $this->mCode = $code;
4257 // Ensure we don't leave incorrect cached data lying around
4258 $this->mHtmlCode = null;
4259 $this->mParentLanguage = false;
4260 }
4261
4262 /**
4263 * Get the name of a file for a certain language code
4264 * @param string $prefix Prepend this to the filename
4265 * @param string $code Language code
4266 * @param string $suffix Append this to the filename
4267 * @throws MWException
4268 * @return string $prefix . $mangledCode . $suffix
4269 */
4270 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4271 if ( !self::isValidBuiltInCode( $code ) ) {
4272 throw new MWException( "Invalid language code \"$code\"" );
4273 }
4274
4275 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4276 }
4277
4278 /**
4279 * Get the language code from a file name. Inverse of getFileName()
4280 * @param string $filename $prefix . $languageCode . $suffix
4281 * @param string $prefix Prefix before the language code
4282 * @param string $suffix Suffix after the language code
4283 * @return string Language code, or false if $prefix or $suffix isn't found
4284 */
4285 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4286 $m = null;
4287 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4288 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4289 if ( !count( $m ) ) {
4290 return false;
4291 }
4292 return str_replace( '_', '-', strtolower( $m[1] ) );
4293 }
4294
4295 /**
4296 * @param string $code
4297 * @return string
4298 */
4299 public static function getMessagesFileName( $code ) {
4300 global $IP;
4301 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4302 Hooks::run( 'Language::getMessagesFileName', array( $code, &$file ) );
4303 return $file;
4304 }
4305
4306 /**
4307 * @param string $code
4308 * @return string
4309 * @since 1.23
4310 */
4311 public static function getJsonMessagesFileName( $code ) {
4312 global $IP;
4313
4314 if ( !self::isValidBuiltInCode( $code ) ) {
4315 throw new MWException( "Invalid language code \"$code\"" );
4316 }
4317
4318 return "$IP/languages/i18n/$code.json";
4319 }
4320
4321 /**
4322 * @param string $code
4323 * @return string
4324 */
4325 public static function getClassFileName( $code ) {
4326 global $IP;
4327 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4328 }
4329
4330 /**
4331 * Get the first fallback for a given language.
4332 *
4333 * @param string $code
4334 *
4335 * @return bool|string
4336 */
4337 public static function getFallbackFor( $code ) {
4338 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4339 return false;
4340 } else {
4341 $fallbacks = self::getFallbacksFor( $code );
4342 $first = array_shift( $fallbacks );
4343 return $first;
4344 }
4345 }
4346
4347 /**
4348 * Get the ordered list of fallback languages.
4349 *
4350 * @since 1.19
4351 * @param string $code Language code
4352 * @return array
4353 */
4354 public static function getFallbacksFor( $code ) {
4355 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4356 return array();
4357 } else {
4358 $v = self::getLocalisationCache()->getItem( $code, 'fallback' );
4359 $v = array_map( 'trim', explode( ',', $v ) );
4360 if ( $v[count( $v ) - 1] !== 'en' ) {
4361 $v[] = 'en';
4362 }
4363 return $v;
4364 }
4365 }
4366
4367 /**
4368 * Get the ordered list of fallback languages, ending with the fallback
4369 * language chain for the site language.
4370 *
4371 * @since 1.22
4372 * @param string $code Language code
4373 * @return array Array( fallbacks, site fallbacks )
4374 */
4375 public static function getFallbacksIncludingSiteLanguage( $code ) {
4376 global $wgLanguageCode;
4377
4378 // Usually, we will only store a tiny number of fallback chains, so we
4379 // keep them in static memory.
4380 $cacheKey = "{$code}-{$wgLanguageCode}";
4381
4382 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4383 $fallbacks = self::getFallbacksFor( $code );
4384
4385 // Append the site's fallback chain, including the site language itself
4386 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4387 array_unshift( $siteFallbacks, $wgLanguageCode );
4388
4389 // Eliminate any languages already included in the chain
4390 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4391
4392 self::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4393 }
4394 return self::$fallbackLanguageCache[$cacheKey];
4395 }
4396
4397 /**
4398 * Get all messages for a given language
4399 * WARNING: this may take a long time. If you just need all message *keys*
4400 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4401 *
4402 * @param string $code
4403 *
4404 * @return array
4405 */
4406 public static function getMessagesFor( $code ) {
4407 return self::getLocalisationCache()->getItem( $code, 'messages' );
4408 }
4409
4410 /**
4411 * Get a message for a given language
4412 *
4413 * @param string $key
4414 * @param string $code
4415 *
4416 * @return string
4417 */
4418 public static function getMessageFor( $key, $code ) {
4419 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4420 }
4421
4422 /**
4423 * Get all message keys for a given language. This is a faster alternative to
4424 * array_keys( Language::getMessagesFor( $code ) )
4425 *
4426 * @since 1.19
4427 * @param string $code Language code
4428 * @return array Array of message keys (strings)
4429 */
4430 public static function getMessageKeysFor( $code ) {
4431 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
4432 }
4433
4434 /**
4435 * @param string $talk
4436 * @return mixed
4437 */
4438 function fixVariableInNamespace( $talk ) {
4439 if ( strpos( $talk, '$1' ) === false ) {
4440 return $talk;
4441 }
4442
4443 global $wgMetaNamespace;
4444 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4445
4446 # Allow grammar transformations
4447 # Allowing full message-style parsing would make simple requests
4448 # such as action=raw much more expensive than they need to be.
4449 # This will hopefully cover most cases.
4450 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4451 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4452 return str_replace( ' ', '_', $talk );
4453 }
4454
4455 /**
4456 * @param string $m
4457 * @return string
4458 */
4459 function replaceGrammarInNamespace( $m ) {
4460 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4461 }
4462
4463 /**
4464 * @throws MWException
4465 * @return array
4466 */
4467 static function getCaseMaps() {
4468 static $wikiUpperChars, $wikiLowerChars;
4469 if ( isset( $wikiUpperChars ) ) {
4470 return array( $wikiUpperChars, $wikiLowerChars );
4471 }
4472
4473 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4474 if ( $arr === false ) {
4475 throw new MWException(
4476 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4477 }
4478 $wikiUpperChars = $arr['wikiUpperChars'];
4479 $wikiLowerChars = $arr['wikiLowerChars'];
4480 return array( $wikiUpperChars, $wikiLowerChars );
4481 }
4482
4483 /**
4484 * Decode an expiry (block, protection, etc) which has come from the DB
4485 *
4486 * @param string $expiry Database expiry String
4487 * @param bool|int $format True to process using language functions, or TS_ constant
4488 * to return the expiry in a given timestamp
4489 * @param string $inifinity If $format is not true, use this string for infinite expiry
4490 * @return string
4491 * @since 1.18
4492 */
4493 public function formatExpiry( $expiry, $format = true, $infinity = 'infinity' ) {
4494 static $dbInfinity;
4495 if ( $dbInfinity === null ) {
4496 $dbInfinity = wfGetDB( DB_SLAVE )->getInfinity();
4497 }
4498
4499 if ( $expiry == '' || $expiry === 'infinity' || $expiry == $dbInfinity ) {
4500 return $format === true
4501 ? $this->getMessageFromDB( 'infiniteblock' )
4502 : $infinity;
4503 } else {
4504 return $format === true
4505 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4506 : wfTimestamp( $format, $expiry );
4507 }
4508 }
4509
4510 /**
4511 * @todo Document
4512 * @param int|float $seconds
4513 * @param array $format Optional
4514 * If $format['avoid'] === 'avoidseconds': don't mention seconds if $seconds >= 1 hour.
4515 * If $format['avoid'] === 'avoidminutes': don't mention seconds/minutes if $seconds > 48 hours.
4516 * If $format['noabbrevs'] is true: use 'seconds' and friends instead of 'seconds-abbrev'
4517 * and friends.
4518 * For backwards compatibility, $format may also be one of the strings 'avoidseconds'
4519 * or 'avoidminutes'.
4520 * @return string
4521 */
4522 function formatTimePeriod( $seconds, $format = array() ) {
4523 if ( !is_array( $format ) ) {
4524 $format = array( 'avoid' => $format ); // For backwards compatibility
4525 }
4526 if ( !isset( $format['avoid'] ) ) {
4527 $format['avoid'] = false;
4528 }
4529 if ( !isset( $format['noabbrevs'] ) ) {
4530 $format['noabbrevs'] = false;
4531 }
4532 $secondsMsg = wfMessage(
4533 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4534 $minutesMsg = wfMessage(
4535 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4536 $hoursMsg = wfMessage(
4537 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4538 $daysMsg = wfMessage(
4539 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4540
4541 if ( round( $seconds * 10 ) < 100 ) {
4542 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4543 $s = $secondsMsg->params( $s )->text();
4544 } elseif ( round( $seconds ) < 60 ) {
4545 $s = $this->formatNum( round( $seconds ) );
4546 $s = $secondsMsg->params( $s )->text();
4547 } elseif ( round( $seconds ) < 3600 ) {
4548 $minutes = floor( $seconds / 60 );
4549 $secondsPart = round( fmod( $seconds, 60 ) );
4550 if ( $secondsPart == 60 ) {
4551 $secondsPart = 0;
4552 $minutes++;
4553 }
4554 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4555 $s .= ' ';
4556 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4557 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4558 $hours = floor( $seconds / 3600 );
4559 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4560 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4561 if ( $secondsPart == 60 ) {
4562 $secondsPart = 0;
4563 $minutes++;
4564 }
4565 if ( $minutes == 60 ) {
4566 $minutes = 0;
4567 $hours++;
4568 }
4569 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4570 $s .= ' ';
4571 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4572 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4573 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4574 }
4575 } else {
4576 $days = floor( $seconds / 86400 );
4577 if ( $format['avoid'] === 'avoidminutes' ) {
4578 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4579 if ( $hours == 24 ) {
4580 $hours = 0;
4581 $days++;
4582 }
4583 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4584 $s .= ' ';
4585 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4586 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4587 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4588 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4589 if ( $minutes == 60 ) {
4590 $minutes = 0;
4591 $hours++;
4592 }
4593 if ( $hours == 24 ) {
4594 $hours = 0;
4595 $days++;
4596 }
4597 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4598 $s .= ' ';
4599 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4600 $s .= ' ';
4601 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4602 } else {
4603 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4604 $s .= ' ';
4605 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4606 }
4607 }
4608 return $s;
4609 }
4610
4611 /**
4612 * Format a bitrate for output, using an appropriate
4613 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to
4614 * the magnitude in question.
4615 *
4616 * This use base 1000. For base 1024 use formatSize(), for another base
4617 * see formatComputingNumbers().
4618 *
4619 * @param int $bps
4620 * @return string
4621 */
4622 function formatBitrate( $bps ) {
4623 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4624 }
4625
4626 /**
4627 * @param int $size Size of the unit
4628 * @param int $boundary Size boundary (1000, or 1024 in most cases)
4629 * @param string $messageKey Message key to be uesd
4630 * @return string
4631 */
4632 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4633 if ( $size <= 0 ) {
4634 return str_replace( '$1', $this->formatNum( $size ),
4635 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4636 );
4637 }
4638 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4639 $index = 0;
4640
4641 $maxIndex = count( $sizes ) - 1;
4642 while ( $size >= $boundary && $index < $maxIndex ) {
4643 $index++;
4644 $size /= $boundary;
4645 }
4646
4647 // For small sizes no decimal places necessary
4648 $round = 0;
4649 if ( $index > 1 ) {
4650 // For MB and bigger two decimal places are smarter
4651 $round = 2;
4652 }
4653 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4654
4655 $size = round( $size, $round );
4656 $text = $this->getMessageFromDB( $msg );
4657 return str_replace( '$1', $this->formatNum( $size ), $text );
4658 }
4659
4660 /**
4661 * Format a size in bytes for output, using an appropriate
4662 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4663 *
4664 * This method use base 1024. For base 1000 use formatBitrate(), for
4665 * another base see formatComputingNumbers()
4666 *
4667 * @param int $size Size to format
4668 * @return string Plain text (not HTML)
4669 */
4670 function formatSize( $size ) {
4671 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4672 }
4673
4674 /**
4675 * Make a list item, used by various special pages
4676 *
4677 * @param string $page Page link
4678 * @param string $details HTML safe text between brackets
4679 * @param bool $oppositedm Add the direction mark opposite to your
4680 * language, to display text properly
4681 * @return HTML escaped string
4682 */
4683 function specialList( $page, $details, $oppositedm = true ) {
4684 if ( !$details ) {
4685 return $page;
4686 }
4687
4688 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) . $this->getDirMark();
4689 return
4690 $page .
4691 $dirmark .
4692 $this->msg( 'word-separator' )->escaped() .
4693 $this->msg( 'parentheses' )->rawParams( $details )->escaped();
4694 }
4695
4696 /**
4697 * Generate (prev x| next x) (20|50|100...) type links for paging
4698 *
4699 * @param Title $title Title object to link
4700 * @param int $offset
4701 * @param int $limit
4702 * @param array $query Optional URL query parameter string
4703 * @param bool $atend Optional param for specified if this is the last page
4704 * @return string
4705 */
4706 public function viewPrevNext( Title $title, $offset, $limit,
4707 array $query = array(), $atend = false
4708 ) {
4709 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4710
4711 # Make 'previous' link
4712 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4713 if ( $offset > 0 ) {
4714 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4715 $query, $prev, 'prevn-title', 'mw-prevlink' );
4716 } else {
4717 $plink = htmlspecialchars( $prev );
4718 }
4719
4720 # Make 'next' link
4721 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4722 if ( $atend ) {
4723 $nlink = htmlspecialchars( $next );
4724 } else {
4725 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4726 $query, $next, 'nextn-title', 'mw-nextlink' );
4727 }
4728
4729 # Make links to set number of items per page
4730 $numLinks = array();
4731 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4732 $numLinks[] = $this->numLink( $title, $offset, $num,
4733 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4734 }
4735
4736 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4737 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4738 }
4739
4740 /**
4741 * Helper function for viewPrevNext() that generates links
4742 *
4743 * @param Title $title Title object to link
4744 * @param int $offset
4745 * @param int $limit
4746 * @param array $query Extra query parameters
4747 * @param string $link Text to use for the link; will be escaped
4748 * @param string $tooltipMsg Name of the message to use as tooltip
4749 * @param string $class Value of the "class" attribute of the link
4750 * @return string HTML fragment
4751 */
4752 private function numLink( Title $title, $offset, $limit, array $query, $link,
4753 $tooltipMsg, $class
4754 ) {
4755 $query = array( 'limit' => $limit, 'offset' => $offset ) + $query;
4756 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )
4757 ->numParams( $limit )->text();
4758
4759 return Html::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4760 'title' => $tooltip, 'class' => $class ), $link );
4761 }
4762
4763 /**
4764 * Get the conversion rule title, if any.
4765 *
4766 * @return string
4767 */
4768 public function getConvRuleTitle() {
4769 return $this->mConverter->getConvRuleTitle();
4770 }
4771
4772 /**
4773 * Get the compiled plural rules for the language
4774 * @since 1.20
4775 * @return array Associative array with plural form, and plural rule as key-value pairs
4776 */
4777 public function getCompiledPluralRules() {
4778 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4779 $fallbacks = Language::getFallbacksFor( $this->mCode );
4780 if ( !$pluralRules ) {
4781 foreach ( $fallbacks as $fallbackCode ) {
4782 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4783 if ( $pluralRules ) {
4784 break;
4785 }
4786 }
4787 }
4788 return $pluralRules;
4789 }
4790
4791 /**
4792 * Get the plural rules for the language
4793 * @since 1.20
4794 * @return array Associative array with plural form number and plural rule as key-value pairs
4795 */
4796 public function getPluralRules() {
4797 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4798 $fallbacks = Language::getFallbacksFor( $this->mCode );
4799 if ( !$pluralRules ) {
4800 foreach ( $fallbacks as $fallbackCode ) {
4801 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4802 if ( $pluralRules ) {
4803 break;
4804 }
4805 }
4806 }
4807 return $pluralRules;
4808 }
4809
4810 /**
4811 * Get the plural rule types for the language
4812 * @since 1.22
4813 * @return array Associative array with plural form number and plural rule type as key-value pairs
4814 */
4815 public function getPluralRuleTypes() {
4816 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4817 $fallbacks = Language::getFallbacksFor( $this->mCode );
4818 if ( !$pluralRuleTypes ) {
4819 foreach ( $fallbacks as $fallbackCode ) {
4820 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4821 if ( $pluralRuleTypes ) {
4822 break;
4823 }
4824 }
4825 }
4826 return $pluralRuleTypes;
4827 }
4828
4829 /**
4830 * Find the index number of the plural rule appropriate for the given number
4831 * @param int $number
4832 * @return int The index number of the plural rule
4833 */
4834 public function getPluralRuleIndexNumber( $number ) {
4835 $pluralRules = $this->getCompiledPluralRules();
4836 $form = CLDRPluralRuleEvaluator::evaluateCompiled( $number, $pluralRules );
4837 return $form;
4838 }
4839
4840 /**
4841 * Find the plural rule type appropriate for the given number
4842 * For example, if the language is set to Arabic, getPluralType(5) should
4843 * return 'few'.
4844 * @since 1.22
4845 * @param int $number
4846 * @return string The name of the plural rule type, e.g. one, two, few, many
4847 */
4848 public function getPluralRuleType( $number ) {
4849 $index = $this->getPluralRuleIndexNumber( $number );
4850 $pluralRuleTypes = $this->getPluralRuleTypes();
4851 if ( isset( $pluralRuleTypes[$index] ) ) {
4852 return $pluralRuleTypes[$index];
4853 } else {
4854 return 'other';
4855 }
4856 }
4857 }
Wiklou, le Wiki du Wiklou