dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Merge "Add optional message to define default description for Upload"
[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 * a fake language converter
39 *
40 * @ingroup Language
41 */
42 class FakeConverter {
43 /**
44 * @var Language
45 */
46 public $mLang;
47 function __construct( $langobj ) { $this->mLang = $langobj; }
48 function autoConvert( $text, $variant = false ) { return $text; }
49 function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
50 function convert( $t ) { return $t; }
51 function convertTo( $text, $variant ) { return $text; }
52 function convertTitle( $t ) { return $t->getPrefixedText(); }
53 function convertNamespace( $ns ) { return $this->mLang->getFormattedNsText( $ns ); }
54 function getVariants() { return array( $this->mLang->getCode() ); }
55 function getVariantFallbacks( $variant ) { return $this->mLang->getCode(); }
56 function getPreferredVariant() { return $this->mLang->getCode(); }
57 function getDefaultVariant() { return $this->mLang->getCode(); }
58 function getURLVariant() { return ''; }
59 function getConvRuleTitle() { return false; }
60 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
61 function getExtraHashOptions() { return ''; }
62 function getParsedTitle() { return ''; }
63 function markNoConversion( $text, $noParse = false ) { return $text; }
64 function convertCategoryKey( $key ) { return $key; }
65 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
66 function armourMath( $text ) { return $text; }
67 function validateVariant( $variant = null ) { return $variant === $this->mLang->getCode() ? $variant : null; }
68 function translate( $text, $variant ) { return $text; }
69 }
70
71 /**
72 * Internationalisation code
73 * @ingroup Language
74 */
75 class Language {
76
77 /**
78 * @var LanguageConverter
79 */
80 public $mConverter;
81
82 public $mVariants, $mCode, $mLoaded = false;
83 public $mMagicExtensions = array(), $mMagicHookDone = false;
84 private $mHtmlCode = null, $mParentLanguage = false;
85
86 public $dateFormatStrings = array();
87 public $mExtendedSpecialPageAliases;
88
89 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
90
91 /**
92 * ReplacementArray object caches
93 */
94 public $transformData = array();
95
96 /**
97 * @var LocalisationCache
98 */
99 static public $dataCache;
100
101 static public $mLangObjCache = array();
102
103 static public $mWeekdayMsgs = array(
104 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
105 'friday', 'saturday'
106 );
107
108 static public $mWeekdayAbbrevMsgs = array(
109 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
110 );
111
112 static public $mMonthMsgs = array(
113 'january', 'february', 'march', 'april', 'may_long', 'june',
114 'july', 'august', 'september', 'october', 'november',
115 'december'
116 );
117 static public $mMonthGenMsgs = array(
118 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
119 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
120 'december-gen'
121 );
122 static public $mMonthAbbrevMsgs = array(
123 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
124 'sep', 'oct', 'nov', 'dec'
125 );
126
127 static public $mIranianCalendarMonthMsgs = array(
128 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
129 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
130 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
131 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
132 );
133
134 static public $mHebrewCalendarMonthMsgs = array(
135 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
136 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
137 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
138 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
139 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
140 );
141
142 static public $mHebrewCalendarMonthGenMsgs = array(
143 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
144 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
145 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
146 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
147 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
148 );
149
150 static public $mHijriCalendarMonthMsgs = array(
151 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
152 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
153 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
154 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
155 );
156
157 /**
158 * @since 1.20
159 * @var array
160 */
161 static public $durationIntervals = array(
162 'millennia' => 31556952000,
163 'centuries' => 3155695200,
164 'decades' => 315569520,
165 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
166 'weeks' => 604800,
167 'days' => 86400,
168 'hours' => 3600,
169 'minutes' => 60,
170 'seconds' => 1,
171 );
172
173 /**
174 * Cache for language fallbacks.
175 * @see Language::getFallbacksIncludingSiteLanguage
176 * @since 1.21
177 * @var array
178 */
179 static private $fallbackLanguageCache = array();
180
181 /**
182 * Get a cached or new language object for a given language code
183 * @param $code String
184 * @return Language
185 */
186 static function factory( $code ) {
187 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
188
189 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
190 $code = $wgDummyLanguageCodes[$code];
191 }
192
193 // get the language object to process
194 $langObj = isset( self::$mLangObjCache[$code] )
195 ? self::$mLangObjCache[$code]
196 : self::newFromCode( $code );
197
198 // merge the language object in to get it up front in the cache
199 self::$mLangObjCache = array_merge( array( $code => $langObj ), self::$mLangObjCache );
200 // get rid of the oldest ones in case we have an overflow
201 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
202
203 return $langObj;
204 }
205
206 /**
207 * Create a language object for a given language code
208 * @param $code String
209 * @throws MWException
210 * @return Language
211 */
212 protected static function newFromCode( $code ) {
213 // Protect against path traversal below
214 if ( !Language::isValidCode( $code )
215 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
216 {
217 throw new MWException( "Invalid language code \"$code\"" );
218 }
219
220 if ( !Language::isValidBuiltInCode( $code ) ) {
221 // It's not possible to customise this code with class files, so
222 // just return a Language object. This is to support uselang= hacks.
223 $lang = new Language;
224 $lang->setCode( $code );
225 return $lang;
226 }
227
228 // Check if there is a language class for the code
229 $class = self::classFromCode( $code );
230 self::preloadLanguageClass( $class );
231 if ( class_exists( $class ) ) {
232 $lang = new $class;
233 return $lang;
234 }
235
236 // Keep trying the fallback list until we find an existing class
237 $fallbacks = Language::getFallbacksFor( $code );
238 foreach ( $fallbacks as $fallbackCode ) {
239 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
240 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
241 }
242
243 $class = self::classFromCode( $fallbackCode );
244 self::preloadLanguageClass( $class );
245 if ( class_exists( $class ) ) {
246 $lang = Language::newFromCode( $fallbackCode );
247 $lang->setCode( $code );
248 return $lang;
249 }
250 }
251
252 throw new MWException( "Invalid fallback sequence for language '$code'" );
253 }
254
255 /**
256 * Checks whether any localisation is available for that language tag
257 * in MediaWiki (MessagesXx.php exists).
258 *
259 * @param string $code Language tag (in lower case)
260 * @return bool Whether language is supported
261 * @since 1.21
262 */
263 public static function isSupportedLanguage( $code ) {
264 return $code === strtolower( $code ) && is_readable( self::getMessagesFileName( $code ) );
265 }
266
267 /**
268 * Returns true if a language code string is a well-formed language tag
269 * according to RFC 5646.
270 * This function only checks well-formedness; it doesn't check that
271 * language, script or variant codes actually exist in the repositories.
272 *
273 * Based on regexes by Mark Davis of the Unicode Consortium:
274 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
275 *
276 * @param $code string
277 * @param $lenient boolean Whether to allow '_' as separator. The default is only '-'.
278 *
279 * @return bool
280 * @since 1.21
281 */
282 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
283 $alpha = '[a-z]';
284 $digit = '[0-9]';
285 $alphanum = '[a-z0-9]';
286 $x = 'x'; # private use singleton
287 $singleton = '[a-wy-z]'; # other singleton
288 $s = $lenient ? '[-_]' : '-';
289
290 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
291 $script = "$alpha{4}"; # ISO 15924
292 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
293 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
294 $extension = "$singleton(?:$s$alphanum{2,8})+";
295 $privateUse = "$x(?:$s$alphanum{1,8})+";
296
297 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
298 # Since these are limited, this is safe even later changes to the registry --
299 # the only oddity is that it might change the type of the tag, and thus
300 # the results from the capturing groups.
301 # http://www.iana.org/assignments/language-subtag-registry
302
303 $grandfathered = "en{$s}GB{$s}oed"
304 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
305 . "|no{$s}(?:bok|nyn)"
306 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
307 . "|zh{$s}min{$s}nan";
308
309 $variantList = "$variant(?:$s$variant)*";
310 $extensionList = "$extension(?:$s$extension)*";
311
312 $langtag = "(?:($language)"
313 . "(?:$s$script)?"
314 . "(?:$s$region)?"
315 . "(?:$s$variantList)?"
316 . "(?:$s$extensionList)?"
317 . "(?:$s$privateUse)?)";
318
319 # The final breakdown, with capturing groups for each of these components
320 # The variants, extensions, grandfathered, and private-use may have interior '-'
321
322 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
323
324 return (bool)preg_match( "/$root/", strtolower( $code ) );
325 }
326
327 /**
328 * Returns true if a language code string is of a valid form, whether or
329 * not it exists. This includes codes which are used solely for
330 * customisation via the MediaWiki namespace.
331 *
332 * @param $code string
333 *
334 * @return bool
335 */
336 public static function isValidCode( $code ) {
337 static $cache = array();
338 if ( isset( $cache[$code] ) ) {
339 return $cache[$code];
340 }
341 // People think language codes are html safe, so enforce it.
342 // Ideally we should only allow a-zA-Z0-9-
343 // but, .+ and other chars are often used for {{int:}} hacks
344 // see bugs 37564, 37587, 36938
345 $cache[$code] =
346 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
347 && !preg_match( Title::getTitleInvalidRegex(), $code );
348
349 return $cache[$code];
350 }
351
352 /**
353 * Returns true if a language code is of a valid form for the purposes of
354 * internal customisation of MediaWiki, via Messages*.php.
355 *
356 * @param $code string
357 *
358 * @throws MWException
359 * @since 1.18
360 * @return bool
361 */
362 public static function isValidBuiltInCode( $code ) {
363
364 if ( !is_string( $code ) ) {
365 if ( is_object( $code ) ) {
366 $addmsg = " of class " . get_class( $code );
367 } else {
368 $addmsg = '';
369 }
370 $type = gettype( $code );
371 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
372 }
373
374 return (bool)preg_match( '/^[a-z0-9-]{2,}$/i', $code );
375 }
376
377 /**
378 * Returns true if a language code is an IETF tag known to MediaWiki.
379 *
380 * @param $code string
381 *
382 * @since 1.21
383 * @return bool
384 */
385 public static function isKnownLanguageTag( $tag ) {
386 static $coreLanguageNames;
387
388 // Quick escape for invalid input to avoid exceptions down the line
389 // when code tries to process tags which are not valid at all.
390 if ( !self::isValidBuiltInCode( $tag ) ) {
391 return false;
392 }
393
394 if ( $coreLanguageNames === null ) {
395 global $IP;
396 include "$IP/languages/Names.php";
397 }
398
399 if ( isset( $coreLanguageNames[$tag] )
400 || self::fetchLanguageName( $tag, $tag ) !== ''
401 ) {
402 return true;
403 }
404
405 return false;
406 }
407
408 /**
409 * @param $code
410 * @return String Name of the language class
411 */
412 public static function classFromCode( $code ) {
413 if ( $code == 'en' ) {
414 return 'Language';
415 } else {
416 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
417 }
418 }
419
420 /**
421 * Includes language class files
422 *
423 * @param $class string Name of the language class
424 */
425 public static function preloadLanguageClass( $class ) {
426 global $IP;
427
428 if ( $class === 'Language' ) {
429 return;
430 }
431
432 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
433 include_once "$IP/languages/classes/$class.php";
434 }
435 }
436
437 /**
438 * Get the LocalisationCache instance
439 *
440 * @return LocalisationCache
441 */
442 public static function getLocalisationCache() {
443 if ( is_null( self::$dataCache ) ) {
444 global $wgLocalisationCacheConf;
445 $class = $wgLocalisationCacheConf['class'];
446 self::$dataCache = new $class( $wgLocalisationCacheConf );
447 }
448 return self::$dataCache;
449 }
450
451 function __construct() {
452 $this->mConverter = new FakeConverter( $this );
453 // Set the code to the name of the descendant
454 if ( get_class( $this ) == 'Language' ) {
455 $this->mCode = 'en';
456 } else {
457 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
458 }
459 self::getLocalisationCache();
460 }
461
462 /**
463 * Reduce memory usage
464 */
465 function __destruct() {
466 foreach ( $this as $name => $value ) {
467 unset( $this->$name );
468 }
469 }
470
471 /**
472 * Hook which will be called if this is the content language.
473 * Descendants can use this to register hook functions or modify globals
474 */
475 function initContLang() { }
476
477 /**
478 * Same as getFallbacksFor for current language.
479 * @return array|bool
480 * @deprecated in 1.19
481 */
482 function getFallbackLanguageCode() {
483 wfDeprecated( __METHOD__, '1.19' );
484 return self::getFallbackFor( $this->mCode );
485 }
486
487 /**
488 * @return array
489 * @since 1.19
490 */
491 function getFallbackLanguages() {
492 return self::getFallbacksFor( $this->mCode );
493 }
494
495 /**
496 * Exports $wgBookstoreListEn
497 * @return array
498 */
499 function getBookstoreList() {
500 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
501 }
502
503 /**
504 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
505 * available in localised form, it will be included in English.
506 *
507 * @return array
508 */
509 public function getNamespaces() {
510 if ( is_null( $this->namespaceNames ) ) {
511 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
512
513 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
514 $validNamespaces = MWNamespace::getCanonicalNamespaces();
515
516 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
517
518 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
519 if ( $wgMetaNamespaceTalk ) {
520 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
521 } else {
522 $talk = $this->namespaceNames[NS_PROJECT_TALK];
523 $this->namespaceNames[NS_PROJECT_TALK] =
524 $this->fixVariableInNamespace( $talk );
525 }
526
527 # Sometimes a language will be localised but not actually exist on this wiki.
528 foreach ( $this->namespaceNames as $key => $text ) {
529 if ( !isset( $validNamespaces[$key] ) ) {
530 unset( $this->namespaceNames[$key] );
531 }
532 }
533
534 # The above mixing may leave namespaces out of canonical order.
535 # Re-order by namespace ID number...
536 ksort( $this->namespaceNames );
537
538 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames ) );
539 }
540 return $this->namespaceNames;
541 }
542
543 /**
544 * Arbitrarily set all of the namespace names at once. Mainly used for testing
545 * @param $namespaces Array of namespaces (id => name)
546 */
547 public function setNamespaces( array $namespaces ) {
548 $this->namespaceNames = $namespaces;
549 $this->mNamespaceIds = null;
550 }
551
552 /**
553 * Resets all of the namespace caches. Mainly used for testing
554 */
555 public function resetNamespaces() {
556 $this->namespaceNames = null;
557 $this->mNamespaceIds = null;
558 $this->namespaceAliases = null;
559 }
560
561 /**
562 * A convenience function that returns the same thing as
563 * getNamespaces() except with the array values changed to ' '
564 * where it found '_', useful for producing output to be displayed
565 * e.g. in <select> forms.
566 *
567 * @return array
568 */
569 function getFormattedNamespaces() {
570 $ns = $this->getNamespaces();
571 foreach ( $ns as $k => $v ) {
572 $ns[$k] = strtr( $v, '_', ' ' );
573 }
574 return $ns;
575 }
576
577 /**
578 * Get a namespace value by key
579 * <code>
580 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
581 * echo $mw_ns; // prints 'MediaWiki'
582 * </code>
583 *
584 * @param $index Int: the array key of the namespace to return
585 * @return mixed, string if the namespace value exists, otherwise false
586 */
587 function getNsText( $index ) {
588 $ns = $this->getNamespaces();
589 return isset( $ns[$index] ) ? $ns[$index] : false;
590 }
591
592 /**
593 * A convenience function that returns the same thing as
594 * getNsText() except with '_' changed to ' ', useful for
595 * producing output.
596 *
597 * <code>
598 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
599 * echo $mw_ns; // prints 'MediaWiki talk'
600 * </code>
601 *
602 * @param int $index The array key of the namespace to return
603 * @return string Namespace name without underscores (empty string if namespace does not exist)
604 */
605 function getFormattedNsText( $index ) {
606 $ns = $this->getNsText( $index );
607 return strtr( $ns, '_', ' ' );
608 }
609
610 /**
611 * Returns gender-dependent namespace alias if available.
612 * @param $index Int: namespace index
613 * @param $gender String: gender key (male, female... )
614 * @return String
615 * @since 1.18
616 */
617 function getGenderNsText( $index, $gender ) {
618 global $wgExtraGenderNamespaces;
619
620 $ns = $wgExtraGenderNamespaces + self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
621 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
622 }
623
624 /**
625 * Whether this language makes distinguishes genders for example in
626 * namespaces.
627 * @return bool
628 * @since 1.18
629 */
630 function needsGenderDistinction() {
631 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
632 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
633 // $wgExtraGenderNamespaces overrides everything
634 return true;
635 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
636 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
637 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
638 return false;
639 } else {
640 // Check what is in i18n files
641 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
642 return count( $aliases ) > 0;
643 }
644 }
645
646 /**
647 * Get a namespace key by value, case insensitive.
648 * Only matches namespace names for the current language, not the
649 * canonical ones defined in Namespace.php.
650 *
651 * @param $text String
652 * @return mixed An integer if $text is a valid value otherwise false
653 */
654 function getLocalNsIndex( $text ) {
655 $lctext = $this->lc( $text );
656 $ids = $this->getNamespaceIds();
657 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
658 }
659
660 /**
661 * @return array
662 */
663 function getNamespaceAliases() {
664 if ( is_null( $this->namespaceAliases ) ) {
665 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
666 if ( !$aliases ) {
667 $aliases = array();
668 } else {
669 foreach ( $aliases as $name => $index ) {
670 if ( $index === NS_PROJECT_TALK ) {
671 unset( $aliases[$name] );
672 $name = $this->fixVariableInNamespace( $name );
673 $aliases[$name] = $index;
674 }
675 }
676 }
677
678 global $wgExtraGenderNamespaces;
679 $genders = $wgExtraGenderNamespaces + (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
680 foreach ( $genders as $index => $forms ) {
681 foreach ( $forms as $alias ) {
682 $aliases[$alias] = $index;
683 }
684 }
685
686 # Also add converted namespace names as aliases, to avoid confusion.
687 $convertedNames = array();
688 foreach ( $this->getVariants() as $variant ) {
689 if ( $variant === $this->mCode ) {
690 continue;
691 }
692 foreach ( $this->getNamespaces() as $ns => $_ ) {
693 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
694 }
695 }
696
697 $this->namespaceAliases = $aliases + $convertedNames;
698 }
699 return $this->namespaceAliases;
700 }
701
702 /**
703 * @return array
704 */
705 function getNamespaceIds() {
706 if ( is_null( $this->mNamespaceIds ) ) {
707 global $wgNamespaceAliases;
708 # Put namespace names and aliases into a hashtable.
709 # If this is too slow, then we should arrange it so that it is done
710 # before caching. The catch is that at pre-cache time, the above
711 # class-specific fixup hasn't been done.
712 $this->mNamespaceIds = array();
713 foreach ( $this->getNamespaces() as $index => $name ) {
714 $this->mNamespaceIds[$this->lc( $name )] = $index;
715 }
716 foreach ( $this->getNamespaceAliases() as $name => $index ) {
717 $this->mNamespaceIds[$this->lc( $name )] = $index;
718 }
719 if ( $wgNamespaceAliases ) {
720 foreach ( $wgNamespaceAliases as $name => $index ) {
721 $this->mNamespaceIds[$this->lc( $name )] = $index;
722 }
723 }
724 }
725 return $this->mNamespaceIds;
726 }
727
728 /**
729 * Get a namespace key by value, case insensitive. Canonical namespace
730 * names override custom ones defined for the current language.
731 *
732 * @param $text String
733 * @return mixed An integer if $text is a valid value otherwise false
734 */
735 function getNsIndex( $text ) {
736 $lctext = $this->lc( $text );
737 $ns = MWNamespace::getCanonicalIndex( $lctext );
738 if ( $ns !== null ) {
739 return $ns;
740 }
741 $ids = $this->getNamespaceIds();
742 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
743 }
744
745 /**
746 * short names for language variants used for language conversion links.
747 *
748 * @param $code String
749 * @param $usemsg bool Use the "variantname-xyz" message if it exists
750 * @return string
751 */
752 function getVariantname( $code, $usemsg = true ) {
753 $msg = "variantname-$code";
754 if ( $usemsg && wfMessage( $msg )->exists() ) {
755 return $this->getMessageFromDB( $msg );
756 }
757 $name = self::fetchLanguageName( $code );
758 if ( $name ) {
759 return $name; # if it's defined as a language name, show that
760 } else {
761 # otherwise, output the language code
762 return $code;
763 }
764 }
765
766 /**
767 * @param $name string
768 * @return string
769 */
770 function specialPage( $name ) {
771 $aliases = $this->getSpecialPageAliases();
772 if ( isset( $aliases[$name][0] ) ) {
773 $name = $aliases[$name][0];
774 }
775 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
776 }
777
778 /**
779 * @return array
780 */
781 function getDatePreferences() {
782 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
783 }
784
785 /**
786 * @return array
787 */
788 function getDateFormats() {
789 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
790 }
791
792 /**
793 * @return array|string
794 */
795 function getDefaultDateFormat() {
796 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
797 if ( $df === 'dmy or mdy' ) {
798 global $wgAmericanDates;
799 return $wgAmericanDates ? 'mdy' : 'dmy';
800 } else {
801 return $df;
802 }
803 }
804
805 /**
806 * @return array
807 */
808 function getDatePreferenceMigrationMap() {
809 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
810 }
811
812 /**
813 * @param $image
814 * @return array|null
815 */
816 function getImageFile( $image ) {
817 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
818 }
819
820 /**
821 * @return array
822 */
823 function getExtraUserToggles() {
824 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
825 }
826
827 /**
828 * @param $tog
829 * @return string
830 */
831 function getUserToggle( $tog ) {
832 return $this->getMessageFromDB( "tog-$tog" );
833 }
834
835 /**
836 * Get native language names, indexed by code.
837 * Only those defined in MediaWiki, no other data like CLDR.
838 * If $customisedOnly is true, only returns codes with a messages file
839 *
840 * @param $customisedOnly bool
841 *
842 * @return array
843 * @deprecated in 1.20, use fetchLanguageNames()
844 */
845 public static function getLanguageNames( $customisedOnly = false ) {
846 return self::fetchLanguageNames( null, $customisedOnly ? 'mwfile' : 'mw' );
847 }
848
849 /**
850 * Get translated language names. This is done on best effort and
851 * by default this is exactly the same as Language::getLanguageNames.
852 * The CLDR extension provides translated names.
853 * @param $code String Language code.
854 * @return Array language code => language name
855 * @since 1.18.0
856 * @deprecated in 1.20, use fetchLanguageNames()
857 */
858 public static function getTranslatedLanguageNames( $code ) {
859 return self::fetchLanguageNames( $code, 'all' );
860 }
861
862 /**
863 * Get an array of language names, indexed by code.
864 * @param $inLanguage null|string: Code of language in which to return the names
865 * Use null for autonyms (native names)
866 * @param $include string:
867 * 'all' all available languages
868 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
869 * 'mwfile' only if the language is in 'mw' *and* has a message file
870 * @return array: language code => language name
871 * @since 1.20
872 */
873 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
874 global $wgExtraLanguageNames;
875 static $coreLanguageNames;
876
877 if ( $coreLanguageNames === null ) {
878 global $IP;
879 include "$IP/languages/Names.php";
880 }
881
882 $names = array();
883
884 if ( $inLanguage ) {
885 # TODO: also include when $inLanguage is null, when this code is more efficient
886 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
887 }
888
889 $mwNames = $wgExtraLanguageNames + $coreLanguageNames;
890 foreach ( $mwNames as $mwCode => $mwName ) {
891 # - Prefer own MediaWiki native name when not using the hook
892 # - For other names just add if not added through the hook
893 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
894 $names[$mwCode] = $mwName;
895 }
896 }
897
898 if ( $include === 'all' ) {
899 return $names;
900 }
901
902 $returnMw = array();
903 $coreCodes = array_keys( $mwNames );
904 foreach ( $coreCodes as $coreCode ) {
905 $returnMw[$coreCode] = $names[$coreCode];
906 }
907
908 if ( $include === 'mwfile' ) {
909 $namesMwFile = array();
910 # We do this using a foreach over the codes instead of a directory
911 # loop so that messages files in extensions will work correctly.
912 foreach ( $returnMw as $code => $value ) {
913 if ( is_readable( self::getMessagesFileName( $code ) ) ) {
914 $namesMwFile[$code] = $names[$code];
915 }
916 }
917 return $namesMwFile;
918 }
919 # 'mw' option; default if it's not one of the other two options (all/mwfile)
920 return $returnMw;
921 }
922
923 /**
924 * @param $code string: The code of the language for which to get the name
925 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
926 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
927 * @return string: Language name or empty
928 * @since 1.20
929 */
930 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
931 $array = self::fetchLanguageNames( $inLanguage, $include );
932 return !array_key_exists( $code, $array ) ? '' : $array[$code];
933 }
934
935 /**
936 * Get a message from the MediaWiki namespace.
937 *
938 * @param $msg String: message name
939 * @return string
940 */
941 function getMessageFromDB( $msg ) {
942 return wfMessage( $msg )->inLanguage( $this )->text();
943 }
944
945 /**
946 * Get the native language name of $code.
947 * Only if defined in MediaWiki, no other data like CLDR.
948 * @param $code string
949 * @return string
950 * @deprecated in 1.20, use fetchLanguageName()
951 */
952 function getLanguageName( $code ) {
953 return self::fetchLanguageName( $code );
954 }
955
956 /**
957 * @param $key string
958 * @return string
959 */
960 function getMonthName( $key ) {
961 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
962 }
963
964 /**
965 * @return array
966 */
967 function getMonthNamesArray() {
968 $monthNames = array( '' );
969 for ( $i = 1; $i < 13; $i++ ) {
970 $monthNames[] = $this->getMonthName( $i );
971 }
972 return $monthNames;
973 }
974
975 /**
976 * @param $key string
977 * @return string
978 */
979 function getMonthNameGen( $key ) {
980 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
981 }
982
983 /**
984 * @param $key string
985 * @return string
986 */
987 function getMonthAbbreviation( $key ) {
988 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
989 }
990
991 /**
992 * @return array
993 */
994 function getMonthAbbreviationsArray() {
995 $monthNames = array( '' );
996 for ( $i = 1; $i < 13; $i++ ) {
997 $monthNames[] = $this->getMonthAbbreviation( $i );
998 }
999 return $monthNames;
1000 }
1001
1002 /**
1003 * @param $key string
1004 * @return string
1005 */
1006 function getWeekdayName( $key ) {
1007 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
1008 }
1009
1010 /**
1011 * @param $key string
1012 * @return string
1013 */
1014 function getWeekdayAbbreviation( $key ) {
1015 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
1016 }
1017
1018 /**
1019 * @param $key string
1020 * @return string
1021 */
1022 function getIranianCalendarMonthName( $key ) {
1023 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
1024 }
1025
1026 /**
1027 * @param $key string
1028 * @return string
1029 */
1030 function getHebrewCalendarMonthName( $key ) {
1031 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
1032 }
1033
1034 /**
1035 * @param $key string
1036 * @return string
1037 */
1038 function getHebrewCalendarMonthNameGen( $key ) {
1039 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1040 }
1041
1042 /**
1043 * @param $key string
1044 * @return string
1045 */
1046 function getHijriCalendarMonthName( $key ) {
1047 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1048 }
1049
1050 /**
1051 * This is a workalike of PHP's date() function, but with better
1052 * internationalisation, a reduced set of format characters, and a better
1053 * escaping format.
1054 *
1055 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1056 * the PHP manual for definitions. There are a number of extensions, which
1057 * start with "x":
1058 *
1059 * xn Do not translate digits of the next numeric format character
1060 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1061 * xr Use roman numerals for the next numeric format character
1062 * xh Use hebrew numerals for the next numeric format character
1063 * xx Literal x
1064 * xg Genitive month name
1065 *
1066 * xij j (day number) in Iranian calendar
1067 * xiF F (month name) in Iranian calendar
1068 * xin n (month number) in Iranian calendar
1069 * xiy y (two digit year) in Iranian calendar
1070 * xiY Y (full year) in Iranian calendar
1071 *
1072 * xjj j (day number) in Hebrew calendar
1073 * xjF F (month name) in Hebrew calendar
1074 * xjt t (days in month) in Hebrew calendar
1075 * xjx xg (genitive month name) in Hebrew calendar
1076 * xjn n (month number) in Hebrew calendar
1077 * xjY Y (full year) in Hebrew calendar
1078 *
1079 * xmj j (day number) in Hijri calendar
1080 * xmF F (month name) in Hijri calendar
1081 * xmn n (month number) in Hijri calendar
1082 * xmY Y (full year) in Hijri calendar
1083 *
1084 * xkY Y (full year) in Thai solar calendar. Months and days are
1085 * identical to the Gregorian calendar
1086 * xoY Y (full year) in Minguo calendar or Juche year.
1087 * Months and days are identical to the
1088 * Gregorian calendar
1089 * xtY Y (full year) in Japanese nengo. Months and days are
1090 * identical to the Gregorian calendar
1091 *
1092 * Characters enclosed in double quotes will be considered literal (with
1093 * the quotes themselves removed). Unmatched quotes will be considered
1094 * literal quotes. Example:
1095 *
1096 * "The month is" F => The month is January
1097 * i's" => 20'11"
1098 *
1099 * Backslash escaping is also supported.
1100 *
1101 * Input timestamp is assumed to be pre-normalized to the desired local
1102 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1103 * $ts is UTC if $zone is not given.
1104 *
1105 * @param $format String
1106 * @param $ts String: 14-character timestamp
1107 * YYYYMMDDHHMMSS
1108 * 01234567890123
1109 * @param $zone DateTimeZone: Timezone of $ts
1110 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1111 *
1112 * @throws MWException
1113 * @return string
1114 */
1115 function sprintfDate( $format, $ts, DateTimeZone $zone = null ) {
1116 $s = '';
1117 $raw = false;
1118 $roman = false;
1119 $hebrewNum = false;
1120 $dateTimeObj = false;
1121 $rawToggle = false;
1122 $iranian = false;
1123 $hebrew = false;
1124 $hijri = false;
1125 $thai = false;
1126 $minguo = false;
1127 $tenno = false;
1128
1129 if ( strlen( $ts ) !== 14 ) {
1130 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1131 }
1132
1133 if ( !ctype_digit( $ts ) ) {
1134 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1135 }
1136
1137 for ( $p = 0; $p < strlen( $format ); $p++ ) {
1138 $num = false;
1139 $code = $format[$p];
1140 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
1141 $code .= $format[++$p];
1142 }
1143
1144 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
1145 $code .= $format[++$p];
1146 }
1147
1148 switch ( $code ) {
1149 case 'xx':
1150 $s .= 'x';
1151 break;
1152 case 'xn':
1153 $raw = true;
1154 break;
1155 case 'xN':
1156 $rawToggle = !$rawToggle;
1157 break;
1158 case 'xr':
1159 $roman = true;
1160 break;
1161 case 'xh':
1162 $hebrewNum = true;
1163 break;
1164 case 'xg':
1165 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1166 break;
1167 case 'xjx':
1168 if ( !$hebrew ) {
1169 $hebrew = self::tsToHebrew( $ts );
1170 }
1171 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1172 break;
1173 case 'd':
1174 $num = substr( $ts, 6, 2 );
1175 break;
1176 case 'D':
1177 if ( !$dateTimeObj ) {
1178 $dateTimeObj = DateTime::createFromFormat(
1179 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1180 );
1181 }
1182 $s .= $this->getWeekdayAbbreviation( $dateTimeObj->format( 'w' ) + 1 );
1183 break;
1184 case 'j':
1185 $num = intval( substr( $ts, 6, 2 ) );
1186 break;
1187 case 'xij':
1188 if ( !$iranian ) {
1189 $iranian = self::tsToIranian( $ts );
1190 }
1191 $num = $iranian[2];
1192 break;
1193 case 'xmj':
1194 if ( !$hijri ) {
1195 $hijri = self::tsToHijri( $ts );
1196 }
1197 $num = $hijri[2];
1198 break;
1199 case 'xjj':
1200 if ( !$hebrew ) {
1201 $hebrew = self::tsToHebrew( $ts );
1202 }
1203 $num = $hebrew[2];
1204 break;
1205 case 'l':
1206 if ( !$dateTimeObj ) {
1207 $dateTimeObj = DateTime::createFromFormat(
1208 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1209 );
1210 }
1211 $s .= $this->getWeekdayName( $dateTimeObj->format( 'w' ) + 1 );
1212 break;
1213 case 'F':
1214 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1215 break;
1216 case 'xiF':
1217 if ( !$iranian ) {
1218 $iranian = self::tsToIranian( $ts );
1219 }
1220 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1221 break;
1222 case 'xmF':
1223 if ( !$hijri ) {
1224 $hijri = self::tsToHijri( $ts );
1225 }
1226 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1227 break;
1228 case 'xjF':
1229 if ( !$hebrew ) {
1230 $hebrew = self::tsToHebrew( $ts );
1231 }
1232 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1233 break;
1234 case 'm':
1235 $num = substr( $ts, 4, 2 );
1236 break;
1237 case 'M':
1238 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1239 break;
1240 case 'n':
1241 $num = intval( substr( $ts, 4, 2 ) );
1242 break;
1243 case 'xin':
1244 if ( !$iranian ) {
1245 $iranian = self::tsToIranian( $ts );
1246 }
1247 $num = $iranian[1];
1248 break;
1249 case 'xmn':
1250 if ( !$hijri ) {
1251 $hijri = self::tsToHijri ( $ts );
1252 }
1253 $num = $hijri[1];
1254 break;
1255 case 'xjn':
1256 if ( !$hebrew ) {
1257 $hebrew = self::tsToHebrew( $ts );
1258 }
1259 $num = $hebrew[1];
1260 break;
1261 case 'xjt':
1262 if ( !$hebrew ) {
1263 $hebrew = self::tsToHebrew( $ts );
1264 }
1265 $num = $hebrew[3];
1266 break;
1267 case 'Y':
1268 $num = substr( $ts, 0, 4 );
1269 break;
1270 case 'xiY':
1271 if ( !$iranian ) {
1272 $iranian = self::tsToIranian( $ts );
1273 }
1274 $num = $iranian[0];
1275 break;
1276 case 'xmY':
1277 if ( !$hijri ) {
1278 $hijri = self::tsToHijri( $ts );
1279 }
1280 $num = $hijri[0];
1281 break;
1282 case 'xjY':
1283 if ( !$hebrew ) {
1284 $hebrew = self::tsToHebrew( $ts );
1285 }
1286 $num = $hebrew[0];
1287 break;
1288 case 'xkY':
1289 if ( !$thai ) {
1290 $thai = self::tsToYear( $ts, 'thai' );
1291 }
1292 $num = $thai[0];
1293 break;
1294 case 'xoY':
1295 if ( !$minguo ) {
1296 $minguo = self::tsToYear( $ts, 'minguo' );
1297 }
1298 $num = $minguo[0];
1299 break;
1300 case 'xtY':
1301 if ( !$tenno ) {
1302 $tenno = self::tsToYear( $ts, 'tenno' );
1303 }
1304 $num = $tenno[0];
1305 break;
1306 case 'y':
1307 $num = substr( $ts, 2, 2 );
1308 break;
1309 case 'xiy':
1310 if ( !$iranian ) {
1311 $iranian = self::tsToIranian( $ts );
1312 }
1313 $num = substr( $iranian[0], -2 );
1314 break;
1315 case 'a':
1316 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1317 break;
1318 case 'A':
1319 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1320 break;
1321 case 'g':
1322 $h = substr( $ts, 8, 2 );
1323 $num = $h % 12 ? $h % 12 : 12;
1324 break;
1325 case 'G':
1326 $num = intval( substr( $ts, 8, 2 ) );
1327 break;
1328 case 'h':
1329 $h = substr( $ts, 8, 2 );
1330 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1331 break;
1332 case 'H':
1333 $num = substr( $ts, 8, 2 );
1334 break;
1335 case 'i':
1336 $num = substr( $ts, 10, 2 );
1337 break;
1338 case 's':
1339 $num = substr( $ts, 12, 2 );
1340 break;
1341 case 'c':
1342 case 'r':
1343 case 'e':
1344 case 'O':
1345 case 'P':
1346 case 'T':
1347 // Pass through string from $dateTimeObj->format()
1348 if ( !$dateTimeObj ) {
1349 $dateTimeObj = DateTime::createFromFormat(
1350 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1351 );
1352 }
1353 $s .= $dateTimeObj->format( $code );
1354 break;
1355 case 'w':
1356 case 'N':
1357 case 'z':
1358 case 'W':
1359 case 't':
1360 case 'L':
1361 case 'o':
1362 case 'U':
1363 case 'I':
1364 case 'Z':
1365 // Pass through number from $dateTimeObj->format()
1366 if ( !$dateTimeObj ) {
1367 $dateTimeObj = DateTime::createFromFormat(
1368 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1369 );
1370 }
1371 $num = $dateTimeObj->format( $code );
1372 break;
1373 case '\\':
1374 # Backslash escaping
1375 if ( $p < strlen( $format ) - 1 ) {
1376 $s .= $format[++$p];
1377 } else {
1378 $s .= '\\';
1379 }
1380 break;
1381 case '"':
1382 # Quoted literal
1383 if ( $p < strlen( $format ) - 1 ) {
1384 $endQuote = strpos( $format, '"', $p + 1 );
1385 if ( $endQuote === false ) {
1386 # No terminating quote, assume literal "
1387 $s .= '"';
1388 } else {
1389 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1390 $p = $endQuote;
1391 }
1392 } else {
1393 # Quote at end of string, assume literal "
1394 $s .= '"';
1395 }
1396 break;
1397 default:
1398 $s .= $format[$p];
1399 }
1400 if ( $num !== false ) {
1401 if ( $rawToggle || $raw ) {
1402 $s .= $num;
1403 $raw = false;
1404 } elseif ( $roman ) {
1405 $s .= Language::romanNumeral( $num );
1406 $roman = false;
1407 } elseif ( $hebrewNum ) {
1408 $s .= self::hebrewNumeral( $num );
1409 $hebrewNum = false;
1410 } else {
1411 $s .= $this->formatNum( $num, true );
1412 }
1413 }
1414 }
1415 return $s;
1416 }
1417
1418 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1419 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1420
1421 /**
1422 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1423 * Gregorian dates to Iranian dates. Originally written in C, it
1424 * is released under the terms of GNU Lesser General Public
1425 * License. Conversion to PHP was performed by Niklas Laxström.
1426 *
1427 * Link: http://www.farsiweb.info/jalali/jalali.c
1428 *
1429 * @param $ts string
1430 *
1431 * @return string
1432 */
1433 private static function tsToIranian( $ts ) {
1434 $gy = substr( $ts, 0, 4 ) -1600;
1435 $gm = substr( $ts, 4, 2 ) -1;
1436 $gd = substr( $ts, 6, 2 ) -1;
1437
1438 # Days passed from the beginning (including leap years)
1439 $gDayNo = 365 * $gy
1440 + floor( ( $gy + 3 ) / 4 )
1441 - floor( ( $gy + 99 ) / 100 )
1442 + floor( ( $gy + 399 ) / 400 );
1443
1444 // Add days of the past months of this year
1445 for ( $i = 0; $i < $gm; $i++ ) {
1446 $gDayNo += self::$GREG_DAYS[$i];
1447 }
1448
1449 // Leap years
1450 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1451 $gDayNo++;
1452 }
1453
1454 // Days passed in current month
1455 $gDayNo += (int)$gd;
1456
1457 $jDayNo = $gDayNo - 79;
1458
1459 $jNp = floor( $jDayNo / 12053 );
1460 $jDayNo %= 12053;
1461
1462 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1463 $jDayNo %= 1461;
1464
1465 if ( $jDayNo >= 366 ) {
1466 $jy += floor( ( $jDayNo - 1 ) / 365 );
1467 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1468 }
1469
1470 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1471 $jDayNo -= self::$IRANIAN_DAYS[$i];
1472 }
1473
1474 $jm = $i + 1;
1475 $jd = $jDayNo + 1;
1476
1477 return array( $jy, $jm, $jd );
1478 }
1479
1480 /**
1481 * Converting Gregorian dates to Hijri dates.
1482 *
1483 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1484 *
1485 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1486 *
1487 * @param $ts string
1488 *
1489 * @return string
1490 */
1491 private static function tsToHijri( $ts ) {
1492 $year = substr( $ts, 0, 4 );
1493 $month = substr( $ts, 4, 2 );
1494 $day = substr( $ts, 6, 2 );
1495
1496 $zyr = $year;
1497 $zd = $day;
1498 $zm = $month;
1499 $zy = $zyr;
1500
1501 if (
1502 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1503 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1504 )
1505 {
1506 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1507 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1508 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1509 $zd - 32075;
1510 } else {
1511 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1512 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1513 }
1514
1515 $zl = $zjd -1948440 + 10632;
1516 $zn = (int)( ( $zl - 1 ) / 10631 );
1517 $zl = $zl - 10631 * $zn + 354;
1518 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1519 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1520 $zm = (int)( ( 24 * $zl ) / 709 );
1521 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1522 $zy = 30 * $zn + $zj - 30;
1523
1524 return array( $zy, $zm, $zd );
1525 }
1526
1527 /**
1528 * Converting Gregorian dates to Hebrew dates.
1529 *
1530 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1531 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1532 * to translate the relevant functions into PHP and release them under
1533 * GNU GPL.
1534 *
1535 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1536 * and Adar II is 14. In a non-leap year, Adar is 6.
1537 *
1538 * @param $ts string
1539 *
1540 * @return string
1541 */
1542 private static function tsToHebrew( $ts ) {
1543 # Parse date
1544 $year = substr( $ts, 0, 4 );
1545 $month = substr( $ts, 4, 2 );
1546 $day = substr( $ts, 6, 2 );
1547
1548 # Calculate Hebrew year
1549 $hebrewYear = $year + 3760;
1550
1551 # Month number when September = 1, August = 12
1552 $month += 4;
1553 if ( $month > 12 ) {
1554 # Next year
1555 $month -= 12;
1556 $year++;
1557 $hebrewYear++;
1558 }
1559
1560 # Calculate day of year from 1 September
1561 $dayOfYear = $day;
1562 for ( $i = 1; $i < $month; $i++ ) {
1563 if ( $i == 6 ) {
1564 # February
1565 $dayOfYear += 28;
1566 # Check if the year is leap
1567 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1568 $dayOfYear++;
1569 }
1570 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1571 $dayOfYear += 30;
1572 } else {
1573 $dayOfYear += 31;
1574 }
1575 }
1576
1577 # Calculate the start of the Hebrew year
1578 $start = self::hebrewYearStart( $hebrewYear );
1579
1580 # Calculate next year's start
1581 if ( $dayOfYear <= $start ) {
1582 # Day is before the start of the year - it is the previous year
1583 # Next year's start
1584 $nextStart = $start;
1585 # Previous year
1586 $year--;
1587 $hebrewYear--;
1588 # Add days since previous year's 1 September
1589 $dayOfYear += 365;
1590 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1591 # Leap year
1592 $dayOfYear++;
1593 }
1594 # Start of the new (previous) year
1595 $start = self::hebrewYearStart( $hebrewYear );
1596 } else {
1597 # Next year's start
1598 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1599 }
1600
1601 # Calculate Hebrew day of year
1602 $hebrewDayOfYear = $dayOfYear - $start;
1603
1604 # Difference between year's days
1605 $diff = $nextStart - $start;
1606 # Add 12 (or 13 for leap years) days to ignore the difference between
1607 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1608 # difference is only about the year type
1609 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1610 $diff += 13;
1611 } else {
1612 $diff += 12;
1613 }
1614
1615 # Check the year pattern, and is leap year
1616 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1617 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1618 # and non-leap years
1619 $yearPattern = $diff % 30;
1620 # Check if leap year
1621 $isLeap = $diff >= 30;
1622
1623 # Calculate day in the month from number of day in the Hebrew year
1624 # Don't check Adar - if the day is not in Adar, we will stop before;
1625 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1626 $hebrewDay = $hebrewDayOfYear;
1627 $hebrewMonth = 1;
1628 $days = 0;
1629 while ( $hebrewMonth <= 12 ) {
1630 # Calculate days in this month
1631 if ( $isLeap && $hebrewMonth == 6 ) {
1632 # Adar in a leap year
1633 if ( $isLeap ) {
1634 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1635 $days = 30;
1636 if ( $hebrewDay <= $days ) {
1637 # Day in Adar I
1638 $hebrewMonth = 13;
1639 } else {
1640 # Subtract the days of Adar I
1641 $hebrewDay -= $days;
1642 # Try Adar II
1643 $days = 29;
1644 if ( $hebrewDay <= $days ) {
1645 # Day in Adar II
1646 $hebrewMonth = 14;
1647 }
1648 }
1649 }
1650 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1651 # Cheshvan in a complete year (otherwise as the rule below)
1652 $days = 30;
1653 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1654 # Kislev in an incomplete year (otherwise as the rule below)
1655 $days = 29;
1656 } else {
1657 # Odd months have 30 days, even have 29
1658 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1659 }
1660 if ( $hebrewDay <= $days ) {
1661 # In the current month
1662 break;
1663 } else {
1664 # Subtract the days of the current month
1665 $hebrewDay -= $days;
1666 # Try in the next month
1667 $hebrewMonth++;
1668 }
1669 }
1670
1671 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1672 }
1673
1674 /**
1675 * This calculates the Hebrew year start, as days since 1 September.
1676 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1677 * Used for Hebrew date.
1678 *
1679 * @param $year int
1680 *
1681 * @return string
1682 */
1683 private static function hebrewYearStart( $year ) {
1684 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1685 $b = intval( ( $year - 1 ) % 4 );
1686 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1687 if ( $m < 0 ) {
1688 $m--;
1689 }
1690 $Mar = intval( $m );
1691 if ( $m < 0 ) {
1692 $m++;
1693 }
1694 $m -= $Mar;
1695
1696 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1697 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1698 $Mar++;
1699 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1700 $Mar += 2;
1701 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1702 $Mar++;
1703 }
1704
1705 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1706 return $Mar;
1707 }
1708
1709 /**
1710 * Algorithm to convert Gregorian dates to Thai solar dates,
1711 * Minguo dates or Minguo dates.
1712 *
1713 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1714 * http://en.wikipedia.org/wiki/Minguo_calendar
1715 * http://en.wikipedia.org/wiki/Japanese_era_name
1716 *
1717 * @param $ts String: 14-character timestamp
1718 * @param $cName String: calender name
1719 * @return Array: converted year, month, day
1720 */
1721 private static function tsToYear( $ts, $cName ) {
1722 $gy = substr( $ts, 0, 4 );
1723 $gm = substr( $ts, 4, 2 );
1724 $gd = substr( $ts, 6, 2 );
1725
1726 if ( !strcmp( $cName, 'thai' ) ) {
1727 # Thai solar dates
1728 # Add 543 years to the Gregorian calendar
1729 # Months and days are identical
1730 $gy_offset = $gy + 543;
1731 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1732 # Minguo dates
1733 # Deduct 1911 years from the Gregorian calendar
1734 # Months and days are identical
1735 $gy_offset = $gy - 1911;
1736 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1737 # Nengō dates up to Meiji period
1738 # Deduct years from the Gregorian calendar
1739 # depending on the nengo periods
1740 # Months and days are identical
1741 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1742 # Meiji period
1743 $gy_gannen = $gy - 1868 + 1;
1744 $gy_offset = $gy_gannen;
1745 if ( $gy_gannen == 1 ) {
1746 $gy_offset = '元';
1747 }
1748 $gy_offset = '明治' . $gy_offset;
1749 } elseif (
1750 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1751 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1752 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1753 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1754 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1755 )
1756 {
1757 # Taishō period
1758 $gy_gannen = $gy - 1912 + 1;
1759 $gy_offset = $gy_gannen;
1760 if ( $gy_gannen == 1 ) {
1761 $gy_offset = '元';
1762 }
1763 $gy_offset = '大正' . $gy_offset;
1764 } elseif (
1765 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1766 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1767 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1768 )
1769 {
1770 # Shōwa period
1771 $gy_gannen = $gy - 1926 + 1;
1772 $gy_offset = $gy_gannen;
1773 if ( $gy_gannen == 1 ) {
1774 $gy_offset = '元';
1775 }
1776 $gy_offset = '昭和' . $gy_offset;
1777 } else {
1778 # Heisei period
1779 $gy_gannen = $gy - 1989 + 1;
1780 $gy_offset = $gy_gannen;
1781 if ( $gy_gannen == 1 ) {
1782 $gy_offset = '元';
1783 }
1784 $gy_offset = '平成' . $gy_offset;
1785 }
1786 } else {
1787 $gy_offset = $gy;
1788 }
1789
1790 return array( $gy_offset, $gm, $gd );
1791 }
1792
1793 /**
1794 * Roman number formatting up to 10000
1795 *
1796 * @param $num int
1797 *
1798 * @return string
1799 */
1800 static function romanNumeral( $num ) {
1801 static $table = array(
1802 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1803 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1804 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1805 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM', 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1806 );
1807
1808 $num = intval( $num );
1809 if ( $num > 10000 || $num <= 0 ) {
1810 return $num;
1811 }
1812
1813 $s = '';
1814 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1815 if ( $num >= $pow10 ) {
1816 $s .= $table[$i][(int)floor( $num / $pow10 )];
1817 }
1818 $num = $num % $pow10;
1819 }
1820 return $s;
1821 }
1822
1823 /**
1824 * Hebrew Gematria number formatting up to 9999
1825 *
1826 * @param $num int
1827 *
1828 * @return string
1829 */
1830 static function hebrewNumeral( $num ) {
1831 static $table = array(
1832 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1833 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1834 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1835 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1836 );
1837
1838 $num = intval( $num );
1839 if ( $num > 9999 || $num <= 0 ) {
1840 return $num;
1841 }
1842
1843 $s = '';
1844 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1845 if ( $num >= $pow10 ) {
1846 if ( $num == 15 || $num == 16 ) {
1847 $s .= $table[0][9] . $table[0][$num - 9];
1848 $num = 0;
1849 } else {
1850 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1851 if ( $pow10 == 1000 ) {
1852 $s .= "'";
1853 }
1854 }
1855 }
1856 $num = $num % $pow10;
1857 }
1858 if ( strlen( $s ) == 2 ) {
1859 $str = $s . "'";
1860 } else {
1861 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1862 $str .= substr( $s, strlen( $s ) - 2, 2 );
1863 }
1864 $start = substr( $str, 0, strlen( $str ) - 2 );
1865 $end = substr( $str, strlen( $str ) - 2 );
1866 switch ( $end ) {
1867 case 'כ':
1868 $str = $start . 'ך';
1869 break;
1870 case 'מ':
1871 $str = $start . 'ם';
1872 break;
1873 case 'נ':
1874 $str = $start . 'ן';
1875 break;
1876 case 'פ':
1877 $str = $start . 'ף';
1878 break;
1879 case 'צ':
1880 $str = $start . 'ץ';
1881 break;
1882 }
1883 return $str;
1884 }
1885
1886 /**
1887 * Used by date() and time() to adjust the time output.
1888 *
1889 * @param $ts Int the time in date('YmdHis') format
1890 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1891 * get user timecorrection setting)
1892 * @return int
1893 */
1894 function userAdjust( $ts, $tz = false ) {
1895 global $wgUser, $wgLocalTZoffset;
1896
1897 if ( $tz === false ) {
1898 $tz = $wgUser->getOption( 'timecorrection' );
1899 }
1900
1901 $data = explode( '|', $tz, 3 );
1902
1903 if ( $data[0] == 'ZoneInfo' ) {
1904 wfSuppressWarnings();
1905 $userTZ = timezone_open( $data[2] );
1906 wfRestoreWarnings();
1907 if ( $userTZ !== false ) {
1908 $date = date_create( $ts, timezone_open( 'UTC' ) );
1909 date_timezone_set( $date, $userTZ );
1910 $date = date_format( $date, 'YmdHis' );
1911 return $date;
1912 }
1913 # Unrecognized timezone, default to 'Offset' with the stored offset.
1914 $data[0] = 'Offset';
1915 }
1916
1917 $minDiff = 0;
1918 if ( $data[0] == 'System' || $tz == '' ) {
1919 #  Global offset in minutes.
1920 if ( isset( $wgLocalTZoffset ) ) {
1921 $minDiff = $wgLocalTZoffset;
1922 }
1923 } elseif ( $data[0] == 'Offset' ) {
1924 $minDiff = intval( $data[1] );
1925 } else {
1926 $data = explode( ':', $tz );
1927 if ( count( $data ) == 2 ) {
1928 $data[0] = intval( $data[0] );
1929 $data[1] = intval( $data[1] );
1930 $minDiff = abs( $data[0] ) * 60 + $data[1];
1931 if ( $data[0] < 0 ) {
1932 $minDiff = -$minDiff;
1933 }
1934 } else {
1935 $minDiff = intval( $data[0] ) * 60;
1936 }
1937 }
1938
1939 # No difference ? Return time unchanged
1940 if ( 0 == $minDiff ) {
1941 return $ts;
1942 }
1943
1944 wfSuppressWarnings(); // E_STRICT system time bitching
1945 # Generate an adjusted date; take advantage of the fact that mktime
1946 # will normalize out-of-range values so we don't have to split $minDiff
1947 # into hours and minutes.
1948 $t = mktime( (
1949 (int)substr( $ts, 8, 2 ) ), # Hours
1950 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
1951 (int)substr( $ts, 12, 2 ), # Seconds
1952 (int)substr( $ts, 4, 2 ), # Month
1953 (int)substr( $ts, 6, 2 ), # Day
1954 (int)substr( $ts, 0, 4 ) ); # Year
1955
1956 $date = date( 'YmdHis', $t );
1957 wfRestoreWarnings();
1958
1959 return $date;
1960 }
1961
1962 /**
1963 * This is meant to be used by time(), date(), and timeanddate() to get
1964 * the date preference they're supposed to use, it should be used in
1965 * all children.
1966 *
1967 *<code>
1968 * function timeanddate([...], $format = true) {
1969 * $datePreference = $this->dateFormat($format);
1970 * [...]
1971 * }
1972 *</code>
1973 *
1974 * @param $usePrefs Mixed: if true, the user's preference is used
1975 * if false, the site/language default is used
1976 * if int/string, assumed to be a format.
1977 * @return string
1978 */
1979 function dateFormat( $usePrefs = true ) {
1980 global $wgUser;
1981
1982 if ( is_bool( $usePrefs ) ) {
1983 if ( $usePrefs ) {
1984 $datePreference = $wgUser->getDatePreference();
1985 } else {
1986 $datePreference = (string)User::getDefaultOption( 'date' );
1987 }
1988 } else {
1989 $datePreference = (string)$usePrefs;
1990 }
1991
1992 // return int
1993 if ( $datePreference == '' ) {
1994 return 'default';
1995 }
1996
1997 return $datePreference;
1998 }
1999
2000 /**
2001 * Get a format string for a given type and preference
2002 * @param $type string May be date, time or both
2003 * @param $pref string The format name as it appears in Messages*.php
2004 *
2005 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2006 *
2007 * @return string
2008 */
2009 function getDateFormatString( $type, $pref ) {
2010 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2011 if ( $pref == 'default' ) {
2012 $pref = $this->getDefaultDateFormat();
2013 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2014 } else {
2015 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2016
2017 if ( $type === 'pretty' && $df === null ) {
2018 $df = $this->getDateFormatString( 'date', $pref );
2019 }
2020
2021 if ( $df === null ) {
2022 $pref = $this->getDefaultDateFormat();
2023 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2024 }
2025 }
2026 $this->dateFormatStrings[$type][$pref] = $df;
2027 }
2028 return $this->dateFormatStrings[$type][$pref];
2029 }
2030
2031 /**
2032 * @param $ts Mixed: the time format which needs to be turned into a
2033 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2034 * @param $adj Bool: whether to adjust the time output according to the
2035 * user configured offset ($timecorrection)
2036 * @param $format Mixed: true to use user's date format preference
2037 * @param $timecorrection String|bool the time offset as returned by
2038 * validateTimeZone() in Special:Preferences
2039 * @return string
2040 */
2041 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2042 $ts = wfTimestamp( TS_MW, $ts );
2043 if ( $adj ) {
2044 $ts = $this->userAdjust( $ts, $timecorrection );
2045 }
2046 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2047 return $this->sprintfDate( $df, $ts );
2048 }
2049
2050 /**
2051 * @param $ts Mixed: the time format which needs to be turned into a
2052 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2053 * @param $adj Bool: whether to adjust the time output according to the
2054 * user configured offset ($timecorrection)
2055 * @param $format Mixed: true to use user's date format preference
2056 * @param $timecorrection String|bool the time offset as returned by
2057 * validateTimeZone() in Special:Preferences
2058 * @return string
2059 */
2060 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2061 $ts = wfTimestamp( TS_MW, $ts );
2062 if ( $adj ) {
2063 $ts = $this->userAdjust( $ts, $timecorrection );
2064 }
2065 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2066 return $this->sprintfDate( $df, $ts );
2067 }
2068
2069 /**
2070 * @param $ts Mixed: the time format which needs to be turned into a
2071 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2072 * @param $adj Bool: whether to adjust the time output according to the
2073 * user configured offset ($timecorrection)
2074 * @param $format Mixed: what format to return, if it's false output the
2075 * default one (default true)
2076 * @param $timecorrection String|bool the time offset as returned by
2077 * validateTimeZone() in Special:Preferences
2078 * @return string
2079 */
2080 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2081 $ts = wfTimestamp( TS_MW, $ts );
2082 if ( $adj ) {
2083 $ts = $this->userAdjust( $ts, $timecorrection );
2084 }
2085 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2086 return $this->sprintfDate( $df, $ts );
2087 }
2088
2089 /**
2090 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2091 *
2092 * @since 1.20
2093 *
2094 * @param integer $seconds The amount of seconds.
2095 * @param array $chosenIntervals The intervals to enable.
2096 *
2097 * @return string
2098 */
2099 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2100 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2101
2102 $segments = array();
2103
2104 foreach ( $intervals as $intervalName => $intervalValue ) {
2105 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2106 // duration-years, duration-decades, duration-centuries, duration-millennia
2107 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2108 $segments[] = $message->inLanguage( $this )->escaped();
2109 }
2110
2111 return $this->listToText( $segments );
2112 }
2113
2114 /**
2115 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2116 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2117 *
2118 * @since 1.20
2119 *
2120 * @param integer $seconds The amount of seconds.
2121 * @param array $chosenIntervals The intervals to enable.
2122 *
2123 * @return array
2124 */
2125 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2126 if ( empty( $chosenIntervals ) ) {
2127 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
2128 }
2129
2130 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2131 $sortedNames = array_keys( $intervals );
2132 $smallestInterval = array_pop( $sortedNames );
2133
2134 $segments = array();
2135
2136 foreach ( $intervals as $name => $length ) {
2137 $value = floor( $seconds / $length );
2138
2139 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2140 $seconds -= $value * $length;
2141 $segments[$name] = $value;
2142 }
2143 }
2144
2145 return $segments;
2146 }
2147
2148 /**
2149 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2150 *
2151 * @param $type String: can be 'date', 'time' or 'both'
2152 * @param $ts Mixed: the time format which needs to be turned into a
2153 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2154 * @param $user User object used to get preferences for timezone and format
2155 * @param $options Array, can contain the following keys:
2156 * - 'timecorrection': time correction, can have the following values:
2157 * - true: use user's preference
2158 * - false: don't use time correction
2159 * - integer: value of time correction in minutes
2160 * - 'format': format to use, can have the following values:
2161 * - true: use user's preference
2162 * - false: use default preference
2163 * - string: format to use
2164 * @since 1.19
2165 * @return String
2166 */
2167 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2168 $ts = wfTimestamp( TS_MW, $ts );
2169 $options += array( 'timecorrection' => true, 'format' => true );
2170 if ( $options['timecorrection'] !== false ) {
2171 if ( $options['timecorrection'] === true ) {
2172 $offset = $user->getOption( 'timecorrection' );
2173 } else {
2174 $offset = $options['timecorrection'];
2175 }
2176 $ts = $this->userAdjust( $ts, $offset );
2177 }
2178 if ( $options['format'] === true ) {
2179 $format = $user->getDatePreference();
2180 } else {
2181 $format = $options['format'];
2182 }
2183 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2184 return $this->sprintfDate( $df, $ts );
2185 }
2186
2187 /**
2188 * Get the formatted date for the given timestamp and formatted for
2189 * the given user.
2190 *
2191 * @param $ts Mixed: the time format which needs to be turned into a
2192 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2193 * @param $user User object used to get preferences for timezone and format
2194 * @param $options Array, can contain the following keys:
2195 * - 'timecorrection': time correction, can have the following values:
2196 * - true: use user's preference
2197 * - false: don't use time correction
2198 * - integer: value of time correction in minutes
2199 * - 'format': format to use, can have the following values:
2200 * - true: use user's preference
2201 * - false: use default preference
2202 * - string: format to use
2203 * @since 1.19
2204 * @return String
2205 */
2206 public function userDate( $ts, User $user, array $options = array() ) {
2207 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2208 }
2209
2210 /**
2211 * Get the formatted time for the given timestamp and formatted for
2212 * the given user.
2213 *
2214 * @param $ts Mixed: the time format which needs to be turned into a
2215 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2216 * @param $user User object used to get preferences for timezone and format
2217 * @param $options Array, can contain the following keys:
2218 * - 'timecorrection': time correction, can have the following values:
2219 * - true: use user's preference
2220 * - false: don't use time correction
2221 * - integer: value of time correction in minutes
2222 * - 'format': format to use, can have the following values:
2223 * - true: use user's preference
2224 * - false: use default preference
2225 * - string: format to use
2226 * @since 1.19
2227 * @return String
2228 */
2229 public function userTime( $ts, User $user, array $options = array() ) {
2230 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2231 }
2232
2233 /**
2234 * Get the formatted date and time for the given timestamp and formatted for
2235 * the given user.
2236 *
2237 * @param $ts Mixed: the time format which needs to be turned into a
2238 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2239 * @param $user User object used to get preferences for timezone and format
2240 * @param $options Array, can contain the following keys:
2241 * - 'timecorrection': time correction, can have the following values:
2242 * - true: use user's preference
2243 * - false: don't use time correction
2244 * - integer: value of time correction in minutes
2245 * - 'format': format to use, can have the following values:
2246 * - true: use user's preference
2247 * - false: use default preference
2248 * - string: format to use
2249 * @since 1.19
2250 * @return String
2251 */
2252 public function userTimeAndDate( $ts, User $user, array $options = array() ) {
2253 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2254 }
2255
2256 /**
2257 * Convert an MWTimestamp into a pretty human-readable timestamp using
2258 * the given user preferences and relative base time.
2259 *
2260 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2261 * on your timestamp object, which will then call this function. Calling
2262 * this function directly will cause hooks to be skipped over.
2263 *
2264 * @see MWTimestamp::getHumanTimestamp
2265 * @param MWTimestamp $ts Timestamp to prettify
2266 * @param MWTimestamp $relativeTo Base timestamp
2267 * @param User $user User preferences to use
2268 * @return string Human timestamp
2269 * @since 1.22
2270 */
2271 public function getHumanTimestamp( MWTimestamp $ts, MWTimestamp $relativeTo, User $user ) {
2272 $diff = $ts->diff( $relativeTo );
2273 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) - (int)$relativeTo->timestamp->format( 'w' ) );
2274 $days = $diff->days ?: (int)$diffDay;
2275 if ( $diff->invert || $days > 5 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' ) ) {
2276 // Timestamps are in different years: use full timestamp
2277 // Also do full timestamp for future dates
2278 /**
2279 * @FIXME Add better handling of future timestamps.
2280 */
2281 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2282 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2283 } elseif ( $days > 5 ) {
2284 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2285 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2286 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2287 } elseif ( $days > 1 ) {
2288 // Timestamp within the past week: show the day of the week and time
2289 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2290 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2291 $ts = wfMessage( "$weekday-at" )
2292 ->inLanguage( $this )
2293 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2294 ->text();
2295 } elseif ( $days == 1 ) {
2296 // Timestamp was yesterday: say 'yesterday' and the time.
2297 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2298 $ts = wfMessage( 'yesterday-at' )
2299 ->inLanguage( $this )
2300 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2301 ->text();
2302 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2303 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2304 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2305 $ts = wfMessage( 'today-at' )
2306 ->inLanguage( $this )
2307 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2308 ->text();
2309
2310 // From here on in, the timestamp was soon enough ago so that we can simply say
2311 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2312 } elseif ( $diff->h == 1 ) {
2313 // Less than 90 minutes, but more than an hour ago.
2314 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2315 } elseif ( $diff->i >= 1 ) {
2316 // A few minutes ago.
2317 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2318 } elseif ( $diff->s >= 30 ) {
2319 // Less than a minute, but more than 30 sec ago.
2320 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2321 } else {
2322 // Less than 30 seconds ago.
2323 $ts = wfMessage( 'just-now' )->text();
2324 }
2325
2326 return $ts;
2327 }
2328
2329 /**
2330 * @param $key string
2331 * @return array|null
2332 */
2333 function getMessage( $key ) {
2334 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2335 }
2336
2337 /**
2338 * @return array
2339 */
2340 function getAllMessages() {
2341 return self::$dataCache->getItem( $this->mCode, 'messages' );
2342 }
2343
2344 /**
2345 * @param $in
2346 * @param $out
2347 * @param $string
2348 * @return string
2349 */
2350 function iconv( $in, $out, $string ) {
2351 # This is a wrapper for iconv in all languages except esperanto,
2352 # which does some nasty x-conversions beforehand
2353
2354 # Even with //IGNORE iconv can whine about illegal characters in
2355 # *input* string. We just ignore those too.
2356 # REF: http://bugs.php.net/bug.php?id=37166
2357 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2358 wfSuppressWarnings();
2359 $text = iconv( $in, $out . '//IGNORE', $string );
2360 wfRestoreWarnings();
2361 return $text;
2362 }
2363
2364 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2365
2366 /**
2367 * @param $matches array
2368 * @return mixed|string
2369 */
2370 function ucwordbreaksCallbackAscii( $matches ) {
2371 return $this->ucfirst( $matches[1] );
2372 }
2373
2374 /**
2375 * @param $matches array
2376 * @return string
2377 */
2378 function ucwordbreaksCallbackMB( $matches ) {
2379 return mb_strtoupper( $matches[0] );
2380 }
2381
2382 /**
2383 * @param $matches array
2384 * @return string
2385 */
2386 function ucCallback( $matches ) {
2387 list( $wikiUpperChars ) = self::getCaseMaps();
2388 return strtr( $matches[1], $wikiUpperChars );
2389 }
2390
2391 /**
2392 * @param $matches array
2393 * @return string
2394 */
2395 function lcCallback( $matches ) {
2396 list( , $wikiLowerChars ) = self::getCaseMaps();
2397 return strtr( $matches[1], $wikiLowerChars );
2398 }
2399
2400 /**
2401 * @param $matches array
2402 * @return string
2403 */
2404 function ucwordsCallbackMB( $matches ) {
2405 return mb_strtoupper( $matches[0] );
2406 }
2407
2408 /**
2409 * @param $matches array
2410 * @return string
2411 */
2412 function ucwordsCallbackWiki( $matches ) {
2413 list( $wikiUpperChars ) = self::getCaseMaps();
2414 return strtr( $matches[0], $wikiUpperChars );
2415 }
2416
2417 /**
2418 * Make a string's first character uppercase
2419 *
2420 * @param $str string
2421 *
2422 * @return string
2423 */
2424 function ucfirst( $str ) {
2425 $o = ord( $str );
2426 if ( $o < 96 ) { // if already uppercase...
2427 return $str;
2428 } elseif ( $o < 128 ) {
2429 return ucfirst( $str ); // use PHP's ucfirst()
2430 } else {
2431 // fall back to more complex logic in case of multibyte strings
2432 return $this->uc( $str, true );
2433 }
2434 }
2435
2436 /**
2437 * Convert a string to uppercase
2438 *
2439 * @param $str string
2440 * @param $first bool
2441 *
2442 * @return string
2443 */
2444 function uc( $str, $first = false ) {
2445 if ( function_exists( 'mb_strtoupper' ) ) {
2446 if ( $first ) {
2447 if ( $this->isMultibyte( $str ) ) {
2448 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2449 } else {
2450 return ucfirst( $str );
2451 }
2452 } else {
2453 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2454 }
2455 } else {
2456 if ( $this->isMultibyte( $str ) ) {
2457 $x = $first ? '^' : '';
2458 return preg_replace_callback(
2459 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2460 array( $this, 'ucCallback' ),
2461 $str
2462 );
2463 } else {
2464 return $first ? ucfirst( $str ) : strtoupper( $str );
2465 }
2466 }
2467 }
2468
2469 /**
2470 * @param $str string
2471 * @return mixed|string
2472 */
2473 function lcfirst( $str ) {
2474 $o = ord( $str );
2475 if ( !$o ) {
2476 return strval( $str );
2477 } elseif ( $o >= 128 ) {
2478 return $this->lc( $str, true );
2479 } elseif ( $o > 96 ) {
2480 return $str;
2481 } else {
2482 $str[0] = strtolower( $str[0] );
2483 return $str;
2484 }
2485 }
2486
2487 /**
2488 * @param $str string
2489 * @param $first bool
2490 * @return mixed|string
2491 */
2492 function lc( $str, $first = false ) {
2493 if ( function_exists( 'mb_strtolower' ) ) {
2494 if ( $first ) {
2495 if ( $this->isMultibyte( $str ) ) {
2496 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2497 } else {
2498 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2499 }
2500 } else {
2501 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2502 }
2503 } else {
2504 if ( $this->isMultibyte( $str ) ) {
2505 $x = $first ? '^' : '';
2506 return preg_replace_callback(
2507 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2508 array( $this, 'lcCallback' ),
2509 $str
2510 );
2511 } else {
2512 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2513 }
2514 }
2515 }
2516
2517 /**
2518 * @param $str string
2519 * @return bool
2520 */
2521 function isMultibyte( $str ) {
2522 return (bool)preg_match( '/[\x80-\xff]/', $str );
2523 }
2524
2525 /**
2526 * @param $str string
2527 * @return mixed|string
2528 */
2529 function ucwords( $str ) {
2530 if ( $this->isMultibyte( $str ) ) {
2531 $str = $this->lc( $str );
2532
2533 // regexp to find first letter in each word (i.e. after each space)
2534 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2535
2536 // function to use to capitalize a single char
2537 if ( function_exists( 'mb_strtoupper' ) ) {
2538 return preg_replace_callback(
2539 $replaceRegexp,
2540 array( $this, 'ucwordsCallbackMB' ),
2541 $str
2542 );
2543 } else {
2544 return preg_replace_callback(
2545 $replaceRegexp,
2546 array( $this, 'ucwordsCallbackWiki' ),
2547 $str
2548 );
2549 }
2550 } else {
2551 return ucwords( strtolower( $str ) );
2552 }
2553 }
2554
2555 /**
2556 * capitalize words at word breaks
2557 *
2558 * @param $str string
2559 * @return mixed
2560 */
2561 function ucwordbreaks( $str ) {
2562 if ( $this->isMultibyte( $str ) ) {
2563 $str = $this->lc( $str );
2564
2565 // since \b doesn't work for UTF-8, we explicitely define word break chars
2566 $breaks = "[ \-\(\)\}\{\.,\?!]";
2567
2568 // find first letter after word break
2569 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2570
2571 if ( function_exists( 'mb_strtoupper' ) ) {
2572 return preg_replace_callback(
2573 $replaceRegexp,
2574 array( $this, 'ucwordbreaksCallbackMB' ),
2575 $str
2576 );
2577 } else {
2578 return preg_replace_callback(
2579 $replaceRegexp,
2580 array( $this, 'ucwordsCallbackWiki' ),
2581 $str
2582 );
2583 }
2584 } else {
2585 return preg_replace_callback(
2586 '/\b([\w\x80-\xff]+)\b/',
2587 array( $this, 'ucwordbreaksCallbackAscii' ),
2588 $str
2589 );
2590 }
2591 }
2592
2593 /**
2594 * Return a case-folded representation of $s
2595 *
2596 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2597 * and $s2 are the same except for the case of their characters. It is not
2598 * necessary for the value returned to make sense when displayed.
2599 *
2600 * Do *not* perform any other normalisation in this function. If a caller
2601 * uses this function when it should be using a more general normalisation
2602 * function, then fix the caller.
2603 *
2604 * @param $s string
2605 *
2606 * @return string
2607 */
2608 function caseFold( $s ) {
2609 return $this->uc( $s );
2610 }
2611
2612 /**
2613 * @param $s string
2614 * @return string
2615 */
2616 function checkTitleEncoding( $s ) {
2617 if ( is_array( $s ) ) {
2618 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2619 }
2620 if ( StringUtils::isUtf8( $s ) ) {
2621 return $s;
2622 }
2623
2624 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2625 }
2626
2627 /**
2628 * @return array
2629 */
2630 function fallback8bitEncoding() {
2631 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2632 }
2633
2634 /**
2635 * Most writing systems use whitespace to break up words.
2636 * Some languages such as Chinese don't conventionally do this,
2637 * which requires special handling when breaking up words for
2638 * searching etc.
2639 *
2640 * @return bool
2641 */
2642 function hasWordBreaks() {
2643 return true;
2644 }
2645
2646 /**
2647 * Some languages such as Chinese require word segmentation,
2648 * Specify such segmentation when overridden in derived class.
2649 *
2650 * @param $string String
2651 * @return String
2652 */
2653 function segmentByWord( $string ) {
2654 return $string;
2655 }
2656
2657 /**
2658 * Some languages have special punctuation need to be normalized.
2659 * Make such changes here.
2660 *
2661 * @param $string String
2662 * @return String
2663 */
2664 function normalizeForSearch( $string ) {
2665 return self::convertDoubleWidth( $string );
2666 }
2667
2668 /**
2669 * convert double-width roman characters to single-width.
2670 * range: ff00-ff5f ~= 0020-007f
2671 *
2672 * @param $string string
2673 *
2674 * @return string
2675 */
2676 protected static function convertDoubleWidth( $string ) {
2677 static $full = null;
2678 static $half = null;
2679
2680 if ( $full === null ) {
2681 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2682 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2683 $full = str_split( $fullWidth, 3 );
2684 $half = str_split( $halfWidth );
2685 }
2686
2687 $string = str_replace( $full, $half, $string );
2688 return $string;
2689 }
2690
2691 /**
2692 * @param $string string
2693 * @param $pattern string
2694 * @return string
2695 */
2696 protected static function insertSpace( $string, $pattern ) {
2697 $string = preg_replace( $pattern, " $1 ", $string );
2698 $string = preg_replace( '/ +/', ' ', $string );
2699 return $string;
2700 }
2701
2702 /**
2703 * @param $termsArray array
2704 * @return array
2705 */
2706 function convertForSearchResult( $termsArray ) {
2707 # some languages, e.g. Chinese, need to do a conversion
2708 # in order for search results to be displayed correctly
2709 return $termsArray;
2710 }
2711
2712 /**
2713 * Get the first character of a string.
2714 *
2715 * @param $s string
2716 * @return string
2717 */
2718 function firstChar( $s ) {
2719 $matches = array();
2720 preg_match(
2721 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2722 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2723 $s,
2724 $matches
2725 );
2726
2727 if ( isset( $matches[1] ) ) {
2728 if ( strlen( $matches[1] ) != 3 ) {
2729 return $matches[1];
2730 }
2731
2732 // Break down Hangul syllables to grab the first jamo
2733 $code = utf8ToCodepoint( $matches[1] );
2734 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2735 return $matches[1];
2736 } elseif ( $code < 0xb098 ) {
2737 return "\xe3\x84\xb1";
2738 } elseif ( $code < 0xb2e4 ) {
2739 return "\xe3\x84\xb4";
2740 } elseif ( $code < 0xb77c ) {
2741 return "\xe3\x84\xb7";
2742 } elseif ( $code < 0xb9c8 ) {
2743 return "\xe3\x84\xb9";
2744 } elseif ( $code < 0xbc14 ) {
2745 return "\xe3\x85\x81";
2746 } elseif ( $code < 0xc0ac ) {
2747 return "\xe3\x85\x82";
2748 } elseif ( $code < 0xc544 ) {
2749 return "\xe3\x85\x85";
2750 } elseif ( $code < 0xc790 ) {
2751 return "\xe3\x85\x87";
2752 } elseif ( $code < 0xcc28 ) {
2753 return "\xe3\x85\x88";
2754 } elseif ( $code < 0xce74 ) {
2755 return "\xe3\x85\x8a";
2756 } elseif ( $code < 0xd0c0 ) {
2757 return "\xe3\x85\x8b";
2758 } elseif ( $code < 0xd30c ) {
2759 return "\xe3\x85\x8c";
2760 } elseif ( $code < 0xd558 ) {
2761 return "\xe3\x85\x8d";
2762 } else {
2763 return "\xe3\x85\x8e";
2764 }
2765 } else {
2766 return '';
2767 }
2768 }
2769
2770 function initEncoding() {
2771 # Some languages may have an alternate char encoding option
2772 # (Esperanto X-coding, Japanese furigana conversion, etc)
2773 # If this language is used as the primary content language,
2774 # an override to the defaults can be set here on startup.
2775 }
2776
2777 /**
2778 * @param $s string
2779 * @return string
2780 */
2781 function recodeForEdit( $s ) {
2782 # For some languages we'll want to explicitly specify
2783 # which characters make it into the edit box raw
2784 # or are converted in some way or another.
2785 global $wgEditEncoding;
2786 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
2787 return $s;
2788 } else {
2789 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2790 }
2791 }
2792
2793 /**
2794 * @param $s string
2795 * @return string
2796 */
2797 function recodeInput( $s ) {
2798 # Take the previous into account.
2799 global $wgEditEncoding;
2800 if ( $wgEditEncoding != '' ) {
2801 $enc = $wgEditEncoding;
2802 } else {
2803 $enc = 'UTF-8';
2804 }
2805 if ( $enc == 'UTF-8' ) {
2806 return $s;
2807 } else {
2808 return $this->iconv( $enc, 'UTF-8', $s );
2809 }
2810 }
2811
2812 /**
2813 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2814 * also cleans up certain backwards-compatible sequences, converting them
2815 * to the modern Unicode equivalent.
2816 *
2817 * This is language-specific for performance reasons only.
2818 *
2819 * @param $s string
2820 *
2821 * @return string
2822 */
2823 function normalize( $s ) {
2824 global $wgAllUnicodeFixes;
2825 $s = UtfNormal::cleanUp( $s );
2826 if ( $wgAllUnicodeFixes ) {
2827 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2828 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2829 }
2830
2831 return $s;
2832 }
2833
2834 /**
2835 * Transform a string using serialized data stored in the given file (which
2836 * must be in the serialized subdirectory of $IP). The file contains pairs
2837 * mapping source characters to destination characters.
2838 *
2839 * The data is cached in process memory. This will go faster if you have the
2840 * FastStringSearch extension.
2841 *
2842 * @param $file string
2843 * @param $string string
2844 *
2845 * @throws MWException
2846 * @return string
2847 */
2848 function transformUsingPairFile( $file, $string ) {
2849 if ( !isset( $this->transformData[$file] ) ) {
2850 $data = wfGetPrecompiledData( $file );
2851 if ( $data === false ) {
2852 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2853 }
2854 $this->transformData[$file] = new ReplacementArray( $data );
2855 }
2856 return $this->transformData[$file]->replace( $string );
2857 }
2858
2859 /**
2860 * For right-to-left language support
2861 *
2862 * @return bool
2863 */
2864 function isRTL() {
2865 return self::$dataCache->getItem( $this->mCode, 'rtl' );
2866 }
2867
2868 /**
2869 * Return the correct HTML 'dir' attribute value for this language.
2870 * @return String
2871 */
2872 function getDir() {
2873 return $this->isRTL() ? 'rtl' : 'ltr';
2874 }
2875
2876 /**
2877 * Return 'left' or 'right' as appropriate alignment for line-start
2878 * for this language's text direction.
2879 *
2880 * Should be equivalent to CSS3 'start' text-align value....
2881 *
2882 * @return String
2883 */
2884 function alignStart() {
2885 return $this->isRTL() ? 'right' : 'left';
2886 }
2887
2888 /**
2889 * Return 'right' or 'left' as appropriate alignment for line-end
2890 * for this language's text direction.
2891 *
2892 * Should be equivalent to CSS3 'end' text-align value....
2893 *
2894 * @return String
2895 */
2896 function alignEnd() {
2897 return $this->isRTL() ? 'left' : 'right';
2898 }
2899
2900 /**
2901 * A hidden direction mark (LRM or RLM), depending on the language direction.
2902 * Unlike getDirMark(), this function returns the character as an HTML entity.
2903 * This function should be used when the output is guaranteed to be HTML,
2904 * because it makes the output HTML source code more readable. When
2905 * the output is plain text or can be escaped, getDirMark() should be used.
2906 *
2907 * @param $opposite Boolean Get the direction mark opposite to your language
2908 * @return string
2909 * @since 1.20
2910 */
2911 function getDirMarkEntity( $opposite = false ) {
2912 if ( $opposite ) {
2913 return $this->isRTL() ? '&lrm;' : '&rlm;';
2914 }
2915 return $this->isRTL() ? '&rlm;' : '&lrm;';
2916 }
2917
2918 /**
2919 * A hidden direction mark (LRM or RLM), depending on the language direction.
2920 * This function produces them as invisible Unicode characters and
2921 * the output may be hard to read and debug, so it should only be used
2922 * when the output is plain text or can be escaped. When the output is
2923 * HTML, use getDirMarkEntity() instead.
2924 *
2925 * @param $opposite Boolean Get the direction mark opposite to your language
2926 * @return string
2927 */
2928 function getDirMark( $opposite = false ) {
2929 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2930 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2931 if ( $opposite ) {
2932 return $this->isRTL() ? $lrm : $rlm;
2933 }
2934 return $this->isRTL() ? $rlm : $lrm;
2935 }
2936
2937 /**
2938 * @return array
2939 */
2940 function capitalizeAllNouns() {
2941 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2942 }
2943
2944 /**
2945 * An arrow, depending on the language direction.
2946 *
2947 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2948 * @return string
2949 */
2950 function getArrow( $direction = 'forwards' ) {
2951 switch ( $direction ) {
2952 case 'forwards':
2953 return $this->isRTL() ? '←' : '→';
2954 case 'backwards':
2955 return $this->isRTL() ? '→' : '←';
2956 case 'left':
2957 return '←';
2958 case 'right':
2959 return '→';
2960 case 'up':
2961 return '↑';
2962 case 'down':
2963 return '↓';
2964 }
2965 }
2966
2967 /**
2968 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2969 *
2970 * @return bool
2971 */
2972 function linkPrefixExtension() {
2973 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2974 }
2975
2976 /**
2977 * Get all magic words from cache.
2978 * @return array
2979 */
2980 function getMagicWords() {
2981 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2982 }
2983
2984 /**
2985 * Run the LanguageGetMagic hook once.
2986 */
2987 protected function doMagicHook() {
2988 if ( $this->mMagicHookDone ) {
2989 return;
2990 }
2991 $this->mMagicHookDone = true;
2992 wfProfileIn( 'LanguageGetMagic' );
2993 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2994 wfProfileOut( 'LanguageGetMagic' );
2995 }
2996
2997 /**
2998 * Fill a MagicWord object with data from here
2999 *
3000 * @param $mw
3001 */
3002 function getMagic( $mw ) {
3003 // Saves a function call
3004 if ( ! $this->mMagicHookDone ) {
3005 $this->doMagicHook();
3006 }
3007
3008 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3009 $rawEntry = $this->mMagicExtensions[$mw->mId];
3010 } else {
3011 $rawEntry = self::$dataCache->getSubitem(
3012 $this->mCode, 'magicWords', $mw->mId );
3013 }
3014
3015 if ( !is_array( $rawEntry ) ) {
3016 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3017 } else {
3018 $mw->mCaseSensitive = $rawEntry[0];
3019 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3020 }
3021 }
3022
3023 /**
3024 * Add magic words to the extension array
3025 *
3026 * @param $newWords array
3027 */
3028 function addMagicWordsByLang( $newWords ) {
3029 $fallbackChain = $this->getFallbackLanguages();
3030 $fallbackChain = array_reverse( $fallbackChain );
3031 foreach ( $fallbackChain as $code ) {
3032 if ( isset( $newWords[$code] ) ) {
3033 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3034 }
3035 }
3036 }
3037
3038 /**
3039 * Get special page names, as an associative array
3040 * case folded alias => real name
3041 */
3042 function getSpecialPageAliases() {
3043 // Cache aliases because it may be slow to load them
3044 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3045 // Initialise array
3046 $this->mExtendedSpecialPageAliases =
3047 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3048 wfRunHooks( 'LanguageGetSpecialPageAliases',
3049 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
3050 }
3051
3052 return $this->mExtendedSpecialPageAliases;
3053 }
3054
3055 /**
3056 * Italic is unsuitable for some languages
3057 *
3058 * @param $text String: the text to be emphasized.
3059 * @return string
3060 */
3061 function emphasize( $text ) {
3062 return "<em>$text</em>";
3063 }
3064
3065 /**
3066 * Normally we output all numbers in plain en_US style, that is
3067 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3068 * point twohundredthirtyfive. However this is not suitable for all
3069 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
3070 * Icelandic just want to use commas instead of dots, and dots instead
3071 * of commas like "293.291,235".
3072 *
3073 * An example of this function being called:
3074 * <code>
3075 * wfMessage( 'message' )->numParams( $num )->text()
3076 * </code>
3077 *
3078 * See LanguageGu.php for the Gujarati implementation and
3079 * $separatorTransformTable on MessageIs.php for
3080 * the , => . and . => , implementation.
3081 *
3082 * @todo check if it's viable to use localeconv() for the decimal
3083 * separator thing.
3084 * @param $number Mixed: the string to be formatted, should be an integer
3085 * or a floating point number.
3086 * @param $nocommafy Bool: set to true for special numbers like dates
3087 * @return string
3088 */
3089 public function formatNum( $number, $nocommafy = false ) {
3090 global $wgTranslateNumerals;
3091 if ( !$nocommafy ) {
3092 $number = $this->commafy( $number );
3093 $s = $this->separatorTransformTable();
3094 if ( $s ) {
3095 $number = strtr( $number, $s );
3096 }
3097 }
3098
3099 if ( $wgTranslateNumerals ) {
3100 $s = $this->digitTransformTable();
3101 if ( $s ) {
3102 $number = strtr( $number, $s );
3103 }
3104 }
3105
3106 return $number;
3107 }
3108
3109 /**
3110 * Front-end for non-commafied formatNum
3111 *
3112 * @param mixed $number the string to be formatted, should be an integer
3113 * or a floating point number.
3114 * @since 1.21
3115 * @return string
3116 */
3117 public function formatNumNoSeparators( $number ) {
3118 return $this->formatNum( $number, true );
3119 }
3120
3121 /**
3122 * @param $number string
3123 * @return string
3124 */
3125 function parseFormattedNumber( $number ) {
3126 $s = $this->digitTransformTable();
3127 if ( $s ) {
3128 $number = strtr( $number, array_flip( $s ) );
3129 }
3130
3131 $s = $this->separatorTransformTable();
3132 if ( $s ) {
3133 $number = strtr( $number, array_flip( $s ) );
3134 }
3135
3136 $number = strtr( $number, array( ',' => '' ) );
3137 return $number;
3138 }
3139
3140 /**
3141 * Adds commas to a given number
3142 * @since 1.19
3143 * @param $number mixed
3144 * @return string
3145 */
3146 function commafy( $number ) {
3147 $digitGroupingPattern = $this->digitGroupingPattern();
3148 if ( $number === null ) {
3149 return '';
3150 }
3151
3152 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3153 // default grouping is at thousands, use the same for ###,###,### pattern too.
3154 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3155 } else {
3156 // Ref: http://cldr.unicode.org/translation/number-patterns
3157 $sign = "";
3158 if ( intval( $number ) < 0 ) {
3159 // For negative numbers apply the algorithm like positive number and add sign.
3160 $sign = "-";
3161 $number = substr( $number, 1 );
3162 }
3163 $integerPart = array();
3164 $decimalPart = array();
3165 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3166 preg_match( "/\d+/", $number, $integerPart );
3167 preg_match( "/\.\d*/", $number, $decimalPart );
3168 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3169 if ( $groupedNumber === $number ) {
3170 // the string does not have any number part. Eg: .12345
3171 return $sign . $groupedNumber;
3172 }
3173 $start = $end = strlen( $integerPart[0] );
3174 while ( $start > 0 ) {
3175 $match = $matches[0][$numMatches - 1];
3176 $matchLen = strlen( $match );
3177 $start = $end - $matchLen;
3178 if ( $start < 0 ) {
3179 $start = 0;
3180 }
3181 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3182 $end = $start;
3183 if ( $numMatches > 1 ) {
3184 // use the last pattern for the rest of the number
3185 $numMatches--;
3186 }
3187 if ( $start > 0 ) {
3188 $groupedNumber = "," . $groupedNumber;
3189 }
3190 }
3191 return $sign . $groupedNumber;
3192 }
3193 }
3194
3195 /**
3196 * @return String
3197 */
3198 function digitGroupingPattern() {
3199 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3200 }
3201
3202 /**
3203 * @return array
3204 */
3205 function digitTransformTable() {
3206 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3207 }
3208
3209 /**
3210 * @return array
3211 */
3212 function separatorTransformTable() {
3213 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3214 }
3215
3216 /**
3217 * Take a list of strings and build a locale-friendly comma-separated
3218 * list, using the local comma-separator message.
3219 * The last two strings are chained with an "and".
3220 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3221 *
3222 * @param $l Array
3223 * @return string
3224 */
3225 function listToText( array $l ) {
3226 $m = count( $l ) - 1;
3227 if ( $m < 0 ) {
3228 return '';
3229 }
3230 if ( $m > 0 ) {
3231 $and = $this->getMessageFromDB( 'and' );
3232 $space = $this->getMessageFromDB( 'word-separator' );
3233 if ( $m > 1 ) {
3234 $comma = $this->getMessageFromDB( 'comma-separator' );
3235 }
3236 }
3237 $s = $l[$m];
3238 for ( $i = $m - 1; $i >= 0; $i-- ) {
3239 if ( $i == $m - 1 ) {
3240 $s = $l[$i] . $and . $space . $s;
3241 } else {
3242 $s = $l[$i] . $comma . $s;
3243 }
3244 }
3245 return $s;
3246 }
3247
3248 /**
3249 * Take a list of strings and build a locale-friendly comma-separated
3250 * list, using the local comma-separator message.
3251 * @param $list array of strings to put in a comma list
3252 * @return string
3253 */
3254 function commaList( array $list ) {
3255 return implode(
3256 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3257 $list
3258 );
3259 }
3260
3261 /**
3262 * Take a list of strings and build a locale-friendly semicolon-separated
3263 * list, using the local semicolon-separator message.
3264 * @param $list array of strings to put in a semicolon list
3265 * @return string
3266 */
3267 function semicolonList( array $list ) {
3268 return implode(
3269 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3270 $list
3271 );
3272 }
3273
3274 /**
3275 * Same as commaList, but separate it with the pipe instead.
3276 * @param $list array of strings to put in a pipe list
3277 * @return string
3278 */
3279 function pipeList( array $list ) {
3280 return implode(
3281 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3282 $list
3283 );
3284 }
3285
3286 /**
3287 * Truncate a string to a specified length in bytes, appending an optional
3288 * string (e.g. for ellipses)
3289 *
3290 * The database offers limited byte lengths for some columns in the database;
3291 * multi-byte character sets mean we need to ensure that only whole characters
3292 * are included, otherwise broken characters can be passed to the user
3293 *
3294 * If $length is negative, the string will be truncated from the beginning
3295 *
3296 * @param $string String to truncate
3297 * @param $length Int: maximum length (including ellipses)
3298 * @param $ellipsis String to append to the truncated text
3299 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3300 * $adjustLength was introduced in 1.18, before that behaved as if false.
3301 * @return string
3302 */
3303 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3304 # Use the localized ellipsis character
3305 if ( $ellipsis == '...' ) {
3306 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3307 }
3308 # Check if there is no need to truncate
3309 if ( $length == 0 ) {
3310 return $ellipsis; // convention
3311 } elseif ( strlen( $string ) <= abs( $length ) ) {
3312 return $string; // no need to truncate
3313 }
3314 $stringOriginal = $string;
3315 # If ellipsis length is >= $length then we can't apply $adjustLength
3316 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3317 $string = $ellipsis; // this can be slightly unexpected
3318 # Otherwise, truncate and add ellipsis...
3319 } else {
3320 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3321 if ( $length > 0 ) {
3322 $length -= $eLength;
3323 $string = substr( $string, 0, $length ); // xyz...
3324 $string = $this->removeBadCharLast( $string );
3325 $string = $string . $ellipsis;
3326 } else {
3327 $length += $eLength;
3328 $string = substr( $string, $length ); // ...xyz
3329 $string = $this->removeBadCharFirst( $string );
3330 $string = $ellipsis . $string;
3331 }
3332 }
3333 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3334 # This check is *not* redundant if $adjustLength, due to the single case where
3335 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3336 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3337 return $string;
3338 } else {
3339 return $stringOriginal;
3340 }
3341 }
3342
3343 /**
3344 * Remove bytes that represent an incomplete Unicode character
3345 * at the end of string (e.g. bytes of the char are missing)
3346 *
3347 * @param $string String
3348 * @return string
3349 */
3350 protected function removeBadCharLast( $string ) {
3351 if ( $string != '' ) {
3352 $char = ord( $string[strlen( $string ) - 1] );
3353 $m = array();
3354 if ( $char >= 0xc0 ) {
3355 # We got the first byte only of a multibyte char; remove it.
3356 $string = substr( $string, 0, -1 );
3357 } elseif ( $char >= 0x80 &&
3358 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3359 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3360 ) {
3361 # We chopped in the middle of a character; remove it
3362 $string = $m[1];
3363 }
3364 }
3365 return $string;
3366 }
3367
3368 /**
3369 * Remove bytes that represent an incomplete Unicode character
3370 * at the start of string (e.g. bytes of the char are missing)
3371 *
3372 * @param $string String
3373 * @return string
3374 */
3375 protected function removeBadCharFirst( $string ) {
3376 if ( $string != '' ) {
3377 $char = ord( $string[0] );
3378 if ( $char >= 0x80 && $char < 0xc0 ) {
3379 # We chopped in the middle of a character; remove the whole thing
3380 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3381 }
3382 }
3383 return $string;
3384 }
3385
3386 /**
3387 * Truncate a string of valid HTML to a specified length in bytes,
3388 * appending an optional string (e.g. for ellipses), and return valid HTML
3389 *
3390 * This is only intended for styled/linked text, such as HTML with
3391 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3392 * Also, this will not detect things like "display:none" CSS.
3393 *
3394 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3395 *
3396 * @param string $text HTML string to truncate
3397 * @param int $length (zero/positive) Maximum length (including ellipses)
3398 * @param string $ellipsis String to append to the truncated text
3399 * @return string
3400 */
3401 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3402 # Use the localized ellipsis character
3403 if ( $ellipsis == '...' ) {
3404 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3405 }
3406 # Check if there is clearly no need to truncate
3407 if ( $length <= 0 ) {
3408 return $ellipsis; // no text shown, nothing to format (convention)
3409 } elseif ( strlen( $text ) <= $length ) {
3410 return $text; // string short enough even *with* HTML (short-circuit)
3411 }
3412
3413 $dispLen = 0; // innerHTML legth so far
3414 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3415 $tagType = 0; // 0-open, 1-close
3416 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3417 $entityState = 0; // 0-not entity, 1-entity
3418 $tag = $ret = ''; // accumulated tag name, accumulated result string
3419 $openTags = array(); // open tag stack
3420 $maybeState = null; // possible truncation state
3421
3422 $textLen = strlen( $text );
3423 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3424 for ( $pos = 0; true; ++$pos ) {
3425 # Consider truncation once the display length has reached the maximim.
3426 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3427 # Check that we're not in the middle of a bracket/entity...
3428 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3429 if ( !$testingEllipsis ) {
3430 $testingEllipsis = true;
3431 # Save where we are; we will truncate here unless there turn out to
3432 # be so few remaining characters that truncation is not necessary.
3433 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3434 $maybeState = array( $ret, $openTags ); // save state
3435 }
3436 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3437 # String in fact does need truncation, the truncation point was OK.
3438 list( $ret, $openTags ) = $maybeState; // reload state
3439 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3440 $ret .= $ellipsis; // add ellipsis
3441 break;
3442 }
3443 }
3444 if ( $pos >= $textLen ) {
3445 break; // extra iteration just for above checks
3446 }
3447
3448 # Read the next char...
3449 $ch = $text[$pos];
3450 $lastCh = $pos ? $text[$pos - 1] : '';
3451 $ret .= $ch; // add to result string
3452 if ( $ch == '<' ) {
3453 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3454 $entityState = 0; // for bad HTML
3455 $bracketState = 1; // tag started (checking for backslash)
3456 } elseif ( $ch == '>' ) {
3457 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3458 $entityState = 0; // for bad HTML
3459 $bracketState = 0; // out of brackets
3460 } elseif ( $bracketState == 1 ) {
3461 if ( $ch == '/' ) {
3462 $tagType = 1; // close tag (e.g. "</span>")
3463 } else {
3464 $tagType = 0; // open tag (e.g. "<span>")
3465 $tag .= $ch;
3466 }
3467 $bracketState = 2; // building tag name
3468 } elseif ( $bracketState == 2 ) {
3469 if ( $ch != ' ' ) {
3470 $tag .= $ch;
3471 } else {
3472 // Name found (e.g. "<a href=..."), add on tag attributes...
3473 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3474 }
3475 } elseif ( $bracketState == 0 ) {
3476 if ( $entityState ) {
3477 if ( $ch == ';' ) {
3478 $entityState = 0;
3479 $dispLen++; // entity is one displayed char
3480 }
3481 } else {
3482 if ( $neLength == 0 && !$maybeState ) {
3483 // Save state without $ch. We want to *hit* the first
3484 // display char (to get tags) but not *use* it if truncating.
3485 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3486 }
3487 if ( $ch == '&' ) {
3488 $entityState = 1; // entity found, (e.g. "&#160;")
3489 } else {
3490 $dispLen++; // this char is displayed
3491 // Add the next $max display text chars after this in one swoop...
3492 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3493 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3494 $dispLen += $skipped;
3495 $pos += $skipped;
3496 }
3497 }
3498 }
3499 }
3500 // Close the last tag if left unclosed by bad HTML
3501 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3502 while ( count( $openTags ) > 0 ) {
3503 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3504 }
3505 return $ret;
3506 }
3507
3508 /**
3509 * truncateHtml() helper function
3510 * like strcspn() but adds the skipped chars to $ret
3511 *
3512 * @param $ret
3513 * @param $text
3514 * @param $search
3515 * @param $start
3516 * @param $len
3517 * @return int
3518 */
3519 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3520 if ( $len === null ) {
3521 $len = -1; // -1 means "no limit" for strcspn
3522 } elseif ( $len < 0 ) {
3523 $len = 0; // sanity
3524 }
3525 $skipCount = 0;
3526 if ( $start < strlen( $text ) ) {
3527 $skipCount = strcspn( $text, $search, $start, $len );
3528 $ret .= substr( $text, $start, $skipCount );
3529 }
3530 return $skipCount;
3531 }
3532
3533 /**
3534 * truncateHtml() helper function
3535 * (a) push or pop $tag from $openTags as needed
3536 * (b) clear $tag value
3537 * @param &$tag string Current HTML tag name we are looking at
3538 * @param $tagType int (0-open tag, 1-close tag)
3539 * @param $lastCh string Character before the '>' that ended this tag
3540 * @param &$openTags array Open tag stack (not accounting for $tag)
3541 */
3542 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3543 $tag = ltrim( $tag );
3544 if ( $tag != '' ) {
3545 if ( $tagType == 0 && $lastCh != '/' ) {
3546 $openTags[] = $tag; // tag opened (didn't close itself)
3547 } elseif ( $tagType == 1 ) {
3548 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3549 array_pop( $openTags ); // tag closed
3550 }
3551 }
3552 $tag = '';
3553 }
3554 }
3555
3556 /**
3557 * Grammatical transformations, needed for inflected languages
3558 * Invoked by putting {{grammar:case|word}} in a message
3559 *
3560 * @param $word string
3561 * @param $case string
3562 * @return string
3563 */
3564 function convertGrammar( $word, $case ) {
3565 global $wgGrammarForms;
3566 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3567 return $wgGrammarForms[$this->getCode()][$case][$word];
3568 }
3569 return $word;
3570 }
3571 /**
3572 * Get the grammar forms for the content language
3573 * @return array of grammar forms
3574 * @since 1.20
3575 */
3576 function getGrammarForms() {
3577 global $wgGrammarForms;
3578 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3579 return $wgGrammarForms[$this->getCode()];
3580 }
3581 return array();
3582 }
3583 /**
3584 * Provides an alternative text depending on specified gender.
3585 * Usage {{gender:username|masculine|feminine|neutral}}.
3586 * username is optional, in which case the gender of current user is used,
3587 * but only in (some) interface messages; otherwise default gender is used.
3588 *
3589 * If no forms are given, an empty string is returned. If only one form is
3590 * given, it will be returned unconditionally. These details are implied by
3591 * the caller and cannot be overridden in subclasses.
3592 *
3593 * If more than one form is given, the default is to use the neutral one
3594 * if it is specified, and to use the masculine one otherwise. These
3595 * details can be overridden in subclasses.
3596 *
3597 * @param $gender string
3598 * @param $forms array
3599 *
3600 * @return string
3601 */
3602 function gender( $gender, $forms ) {
3603 if ( !count( $forms ) ) {
3604 return '';
3605 }
3606 $forms = $this->preConvertPlural( $forms, 2 );
3607 if ( $gender === 'male' ) {
3608 return $forms[0];
3609 }
3610 if ( $gender === 'female' ) {
3611 return $forms[1];
3612 }
3613 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3614 }
3615
3616 /**
3617 * Plural form transformations, needed for some languages.
3618 * For example, there are 3 form of plural in Russian and Polish,
3619 * depending on "count mod 10". See [[w:Plural]]
3620 * For English it is pretty simple.
3621 *
3622 * Invoked by putting {{plural:count|wordform1|wordform2}}
3623 * or {{plural:count|wordform1|wordform2|wordform3}}
3624 *
3625 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3626 *
3627 * @param $count Integer: non-localized number
3628 * @param $forms Array: different plural forms
3629 * @return string Correct form of plural for $count in this language
3630 */
3631 function convertPlural( $count, $forms ) {
3632 // Handle explicit n=pluralform cases
3633 foreach ( $forms as $index => $form ) {
3634 if ( preg_match( '/^\d+=/i', $form ) ) {
3635 $pos = strpos( $form, '=' );
3636 if ( substr( $form, 0, $pos ) === (string)$count ) {
3637 return substr( $form, $pos + 1 );
3638 }
3639 unset( $forms[$index] );
3640 }
3641 }
3642
3643 $forms = array_values( $forms );
3644 if ( !count( $forms ) ) {
3645 return '';
3646 }
3647
3648 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3649 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3650 return $forms[$pluralForm];
3651 }
3652
3653 /**
3654 * Checks that convertPlural was given an array and pads it to requested
3655 * amount of forms by copying the last one.
3656 *
3657 * @param $count Integer: How many forms should there be at least
3658 * @param $forms Array of forms given to convertPlural
3659 * @return array Padded array of forms or an exception if not an array
3660 */
3661 protected function preConvertPlural( /* Array */ $forms, $count ) {
3662 while ( count( $forms ) < $count ) {
3663 $forms[] = $forms[count( $forms ) - 1];
3664 }
3665 return $forms;
3666 }
3667
3668 /**
3669 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3670 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3671 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3672 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3673 * match up with it.
3674 *
3675 * @param $str String: the validated block duration in English
3676 * @return string Somehow translated block duration
3677 * @see LanguageFi.php for example implementation
3678 */
3679 function translateBlockExpiry( $str ) {
3680 $duration = SpecialBlock::getSuggestedDurations( $this );
3681 foreach ( $duration as $show => $value ) {
3682 if ( strcmp( $str, $value ) == 0 ) {
3683 return htmlspecialchars( trim( $show ) );
3684 }
3685 }
3686
3687 // Since usually only infinite or indefinite is only on list, so try
3688 // equivalents if still here.
3689 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3690 if ( in_array( $str, $indefs ) ) {
3691 foreach ( $indefs as $val ) {
3692 $show = array_search( $val, $duration, true );
3693 if ( $show !== false ) {
3694 return htmlspecialchars( trim( $show ) );
3695 }
3696 }
3697 }
3698
3699 // If all else fails, return a standard duration or timestamp description.
3700 $time = strtotime( $str, 0 );
3701 if ( $time === false ) { // Unknown format. Return it as-is in case.
3702 return $str;
3703 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3704 // $time is relative to 0 so it's a duration length.
3705 return $this->formatDuration( $time );
3706 } else { // It's an absolute timestamp.
3707 if ( $time === 0 ) {
3708 // wfTimestamp() handles 0 as current time instead of epoch.
3709 return $this->timeanddate( '19700101000000' );
3710 } else {
3711 return $this->timeanddate( $time );
3712 }
3713 }
3714 }
3715
3716 /**
3717 * languages like Chinese need to be segmented in order for the diff
3718 * to be of any use
3719 *
3720 * @param $text String
3721 * @return String
3722 */
3723 public function segmentForDiff( $text ) {
3724 return $text;
3725 }
3726
3727 /**
3728 * and unsegment to show the result
3729 *
3730 * @param $text String
3731 * @return String
3732 */
3733 public function unsegmentForDiff( $text ) {
3734 return $text;
3735 }
3736
3737 /**
3738 * Return the LanguageConverter used in the Language
3739 *
3740 * @since 1.19
3741 * @return LanguageConverter
3742 */
3743 public function getConverter() {
3744 return $this->mConverter;
3745 }
3746
3747 /**
3748 * convert text to all supported variants
3749 *
3750 * @param $text string
3751 * @return array
3752 */
3753 public function autoConvertToAllVariants( $text ) {
3754 return $this->mConverter->autoConvertToAllVariants( $text );
3755 }
3756
3757 /**
3758 * convert text to different variants of a language.
3759 *
3760 * @param $text string
3761 * @return string
3762 */
3763 public function convert( $text ) {
3764 return $this->mConverter->convert( $text );
3765 }
3766
3767 /**
3768 * Convert a Title object to a string in the preferred variant
3769 *
3770 * @param $title Title
3771 * @return string
3772 */
3773 public function convertTitle( $title ) {
3774 return $this->mConverter->convertTitle( $title );
3775 }
3776
3777 /**
3778 * Convert a namespace index to a string in the preferred variant
3779 *
3780 * @param $ns int
3781 * @return string
3782 */
3783 public function convertNamespace( $ns ) {
3784 return $this->mConverter->convertNamespace( $ns );
3785 }
3786
3787 /**
3788 * Check if this is a language with variants
3789 *
3790 * @return bool
3791 */
3792 public function hasVariants() {
3793 return count( $this->getVariants() ) > 1;
3794 }
3795
3796 /**
3797 * Check if the language has the specific variant
3798 *
3799 * @since 1.19
3800 * @param $variant string
3801 * @return bool
3802 */
3803 public function hasVariant( $variant ) {
3804 return (bool)$this->mConverter->validateVariant( $variant );
3805 }
3806
3807 /**
3808 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3809 *
3810 * @param $text string
3811 * @return string
3812 */
3813 public function armourMath( $text ) {
3814 return $this->mConverter->armourMath( $text );
3815 }
3816
3817 /**
3818 * Perform output conversion on a string, and encode for safe HTML output.
3819 * @param $text String text to be converted
3820 * @param $isTitle Bool whether this conversion is for the article title
3821 * @return string
3822 * @todo this should get integrated somewhere sane
3823 */
3824 public function convertHtml( $text, $isTitle = false ) {
3825 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3826 }
3827
3828 /**
3829 * @param $key string
3830 * @return string
3831 */
3832 public function convertCategoryKey( $key ) {
3833 return $this->mConverter->convertCategoryKey( $key );
3834 }
3835
3836 /**
3837 * Get the list of variants supported by this language
3838 * see sample implementation in LanguageZh.php
3839 *
3840 * @return array an array of language codes
3841 */
3842 public function getVariants() {
3843 return $this->mConverter->getVariants();
3844 }
3845
3846 /**
3847 * @return string
3848 */
3849 public function getPreferredVariant() {
3850 return $this->mConverter->getPreferredVariant();
3851 }
3852
3853 /**
3854 * @return string
3855 */
3856 public function getDefaultVariant() {
3857 return $this->mConverter->getDefaultVariant();
3858 }
3859
3860 /**
3861 * @return string
3862 */
3863 public function getURLVariant() {
3864 return $this->mConverter->getURLVariant();
3865 }
3866
3867 /**
3868 * If a language supports multiple variants, it is
3869 * possible that non-existing link in one variant
3870 * actually exists in another variant. this function
3871 * tries to find it. See e.g. LanguageZh.php
3872 *
3873 * @param $link String: the name of the link
3874 * @param $nt Mixed: the title object of the link
3875 * @param $ignoreOtherCond Boolean: to disable other conditions when
3876 * we need to transclude a template or update a category's link
3877 * @return null the input parameters may be modified upon return
3878 */
3879 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3880 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
3881 }
3882
3883 /**
3884 * If a language supports multiple variants, converts text
3885 * into an array of all possible variants of the text:
3886 * 'variant' => text in that variant
3887 *
3888 * @deprecated since 1.17 Use autoConvertToAllVariants()
3889 *
3890 * @param $text string
3891 *
3892 * @return string
3893 */
3894 public function convertLinkToAllVariants( $text ) {
3895 return $this->mConverter->convertLinkToAllVariants( $text );
3896 }
3897
3898 /**
3899 * returns language specific options used by User::getPageRenderHash()
3900 * for example, the preferred language variant
3901 *
3902 * @return string
3903 */
3904 function getExtraHashOptions() {
3905 return $this->mConverter->getExtraHashOptions();
3906 }
3907
3908 /**
3909 * For languages that support multiple variants, the title of an
3910 * article may be displayed differently in different variants. this
3911 * function returns the apporiate title defined in the body of the article.
3912 *
3913 * @return string
3914 */
3915 public function getParsedTitle() {
3916 return $this->mConverter->getParsedTitle();
3917 }
3918
3919 /**
3920 * Prepare external link text for conversion. When the text is
3921 * a URL, it shouldn't be converted, and it'll be wrapped in
3922 * the "raw" tag (-{R| }-) to prevent conversion.
3923 *
3924 * This function is called "markNoConversion" for historical
3925 * reasons.
3926 *
3927 * @param $text String: text to be used for external link
3928 * @param $noParse bool: wrap it without confirming it's a real URL first
3929 * @return string the tagged text
3930 */
3931 public function markNoConversion( $text, $noParse = false ) {
3932 // Excluding protocal-relative URLs may avoid many false positives.
3933 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
3934 return $this->mConverter->markNoConversion( $text );
3935 } else {
3936 return $text;
3937 }
3938 }
3939
3940 /**
3941 * A regular expression to match legal word-trailing characters
3942 * which should be merged onto a link of the form [[foo]]bar.
3943 *
3944 * @return string
3945 */
3946 public function linkTrail() {
3947 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
3948 }
3949
3950 /**
3951 * @return Language
3952 */
3953 function getLangObj() {
3954 return $this;
3955 }
3956
3957 /**
3958 * Get the "parent" language which has a converter to convert a "compatible" language
3959 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
3960 *
3961 * @return Language|null
3962 * @since 1.22
3963 */
3964 public function getParentLanguage() {
3965 if ( $this->mParentLanguage !== false ) {
3966 return $this->mParentLanguage;
3967 }
3968
3969 $pieces = explode( '-', $this->getCode() );
3970 $code = $pieces[0];
3971 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
3972 $this->mParentLanguage = null;
3973 return null;
3974 }
3975 $lang = Language::factory( $code );
3976 if ( !$lang->hasVariant( $this->getCode() ) ) {
3977 $this->mParentLanguage = null;
3978 return null;
3979 }
3980
3981 $this->mParentLanguage = $lang;
3982 return $lang;
3983 }
3984
3985 /**
3986 * Get the RFC 3066 code for this language object
3987 *
3988 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3989 * htmlspecialchars() or similar
3990 *
3991 * @return string
3992 */
3993 public function getCode() {
3994 return $this->mCode;
3995 }
3996
3997 /**
3998 * Get the code in Bcp47 format which we can use
3999 * inside of html lang="" tags.
4000 *
4001 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4002 * htmlspecialchars() or similar.
4003 *
4004 * @since 1.19
4005 * @return string
4006 */
4007 public function getHtmlCode() {
4008 if ( is_null( $this->mHtmlCode ) ) {
4009 $this->mHtmlCode = wfBCP47( $this->getCode() );
4010 }
4011 return $this->mHtmlCode;
4012 }
4013
4014 /**
4015 * @param $code string
4016 */
4017 public function setCode( $code ) {
4018 $this->mCode = $code;
4019 // Ensure we don't leave incorrect cached data lying around
4020 $this->mHtmlCode = null;
4021 $this->mParentLanguage = false;
4022 }
4023
4024 /**
4025 * Get the name of a file for a certain language code
4026 * @param $prefix string Prepend this to the filename
4027 * @param $code string Language code
4028 * @param $suffix string Append this to the filename
4029 * @throws MWException
4030 * @return string $prefix . $mangledCode . $suffix
4031 */
4032 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4033 // Protect against path traversal
4034 if ( !Language::isValidCode( $code )
4035 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
4036 {
4037 throw new MWException( "Invalid language code \"$code\"" );
4038 }
4039
4040 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4041 }
4042
4043 /**
4044 * Get the language code from a file name. Inverse of getFileName()
4045 * @param $filename string $prefix . $languageCode . $suffix
4046 * @param $prefix string Prefix before the language code
4047 * @param $suffix string Suffix after the language code
4048 * @return string Language code, or false if $prefix or $suffix isn't found
4049 */
4050 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4051 $m = null;
4052 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4053 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4054 if ( !count( $m ) ) {
4055 return false;
4056 }
4057 return str_replace( '_', '-', strtolower( $m[1] ) );
4058 }
4059
4060 /**
4061 * @param $code string
4062 * @return string
4063 */
4064 public static function getMessagesFileName( $code ) {
4065 global $IP;
4066 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4067 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
4068 return $file;
4069 }
4070
4071 /**
4072 * @param $code string
4073 * @return string
4074 */
4075 public static function getClassFileName( $code ) {
4076 global $IP;
4077 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4078 }
4079
4080 /**
4081 * Get the first fallback for a given language.
4082 *
4083 * @param $code string
4084 *
4085 * @return bool|string
4086 */
4087 public static function getFallbackFor( $code ) {
4088 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4089 return false;
4090 } else {
4091 $fallbacks = self::getFallbacksFor( $code );
4092 $first = array_shift( $fallbacks );
4093 return $first;
4094 }
4095 }
4096
4097 /**
4098 * Get the ordered list of fallback languages.
4099 *
4100 * @since 1.19
4101 * @param $code string Language code
4102 * @return array
4103 */
4104 public static function getFallbacksFor( $code ) {
4105 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4106 return array();
4107 } else {
4108 $v = self::getLocalisationCache()->getItem( $code, 'fallback' );
4109 $v = array_map( 'trim', explode( ',', $v ) );
4110 if ( $v[count( $v ) - 1] !== 'en' ) {
4111 $v[] = 'en';
4112 }
4113 return $v;
4114 }
4115 }
4116
4117 /**
4118 * Get the ordered list of fallback languages, ending with the fallback
4119 * language chain for the site language.
4120 *
4121 * @since 1.22
4122 * @param string $code Language code
4123 * @return array array( fallbacks, site fallbacks )
4124 */
4125 public static function getFallbacksIncludingSiteLanguage( $code ) {
4126 global $wgLanguageCode;
4127
4128 // Usually, we will only store a tiny number of fallback chains, so we
4129 // keep them in static memory.
4130 $cacheKey = "{$code}-{$wgLanguageCode}";
4131
4132 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4133 $fallbacks = self::getFallbacksFor( $code );
4134
4135 // Append the site's fallback chain, including the site language itself
4136 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4137 array_unshift( $siteFallbacks, $wgLanguageCode );
4138
4139 // Eliminate any languages already included in the chain
4140 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4141
4142 self::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4143 }
4144 return self::$fallbackLanguageCache[$cacheKey];
4145 }
4146
4147 /**
4148 * Get all messages for a given language
4149 * WARNING: this may take a long time. If you just need all message *keys*
4150 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4151 *
4152 * @param $code string
4153 *
4154 * @return array
4155 */
4156 public static function getMessagesFor( $code ) {
4157 return self::getLocalisationCache()->getItem( $code, 'messages' );
4158 }
4159
4160 /**
4161 * Get a message for a given language
4162 *
4163 * @param $key string
4164 * @param $code string
4165 *
4166 * @return string
4167 */
4168 public static function getMessageFor( $key, $code ) {
4169 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4170 }
4171
4172 /**
4173 * Get all message keys for a given language. This is a faster alternative to
4174 * array_keys( Language::getMessagesFor( $code ) )
4175 *
4176 * @since 1.19
4177 * @param $code string Language code
4178 * @return array of message keys (strings)
4179 */
4180 public static function getMessageKeysFor( $code ) {
4181 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
4182 }
4183
4184 /**
4185 * @param $talk
4186 * @return mixed
4187 */
4188 function fixVariableInNamespace( $talk ) {
4189 if ( strpos( $talk, '$1' ) === false ) {
4190 return $talk;
4191 }
4192
4193 global $wgMetaNamespace;
4194 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4195
4196 # Allow grammar transformations
4197 # Allowing full message-style parsing would make simple requests
4198 # such as action=raw much more expensive than they need to be.
4199 # This will hopefully cover most cases.
4200 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4201 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4202 return str_replace( ' ', '_', $talk );
4203 }
4204
4205 /**
4206 * @param $m string
4207 * @return string
4208 */
4209 function replaceGrammarInNamespace( $m ) {
4210 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4211 }
4212
4213 /**
4214 * @throws MWException
4215 * @return array
4216 */
4217 static function getCaseMaps() {
4218 static $wikiUpperChars, $wikiLowerChars;
4219 if ( isset( $wikiUpperChars ) ) {
4220 return array( $wikiUpperChars, $wikiLowerChars );
4221 }
4222
4223 wfProfileIn( __METHOD__ );
4224 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4225 if ( $arr === false ) {
4226 throw new MWException(
4227 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4228 }
4229 $wikiUpperChars = $arr['wikiUpperChars'];
4230 $wikiLowerChars = $arr['wikiLowerChars'];
4231 wfProfileOut( __METHOD__ );
4232 return array( $wikiUpperChars, $wikiLowerChars );
4233 }
4234
4235 /**
4236 * Decode an expiry (block, protection, etc) which has come from the DB
4237 *
4238 * @todo FIXME: why are we returnings DBMS-dependent strings???
4239 *
4240 * @param $expiry String: Database expiry String
4241 * @param $format Bool|Int true to process using language functions, or TS_ constant
4242 * to return the expiry in a given timestamp
4243 * @return String
4244 * @since 1.18
4245 */
4246 public function formatExpiry( $expiry, $format = true ) {
4247 static $infinity;
4248 if ( $infinity === null ) {
4249 $infinity = wfGetDB( DB_SLAVE )->getInfinity();
4250 }
4251
4252 if ( $expiry == '' || $expiry == $infinity ) {
4253 return $format === true
4254 ? $this->getMessageFromDB( 'infiniteblock' )
4255 : $infinity;
4256 } else {
4257 return $format === true
4258 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4259 : wfTimestamp( $format, $expiry );
4260 }
4261 }
4262
4263 /**
4264 * @todo Document
4265 * @param $seconds int|float
4266 * @param $format Array Optional
4267 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
4268 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
4269 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
4270 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
4271 * @return string
4272 */
4273 function formatTimePeriod( $seconds, $format = array() ) {
4274 if ( !is_array( $format ) ) {
4275 $format = array( 'avoid' => $format ); // For backwards compatibility
4276 }
4277 if ( !isset( $format['avoid'] ) ) {
4278 $format['avoid'] = false;
4279 }
4280 if ( !isset( $format['noabbrevs' ] ) ) {
4281 $format['noabbrevs'] = false;
4282 }
4283 $secondsMsg = wfMessage(
4284 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4285 $minutesMsg = wfMessage(
4286 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4287 $hoursMsg = wfMessage(
4288 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4289 $daysMsg = wfMessage(
4290 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4291
4292 if ( round( $seconds * 10 ) < 100 ) {
4293 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4294 $s = $secondsMsg->params( $s )->text();
4295 } elseif ( round( $seconds ) < 60 ) {
4296 $s = $this->formatNum( round( $seconds ) );
4297 $s = $secondsMsg->params( $s )->text();
4298 } elseif ( round( $seconds ) < 3600 ) {
4299 $minutes = floor( $seconds / 60 );
4300 $secondsPart = round( fmod( $seconds, 60 ) );
4301 if ( $secondsPart == 60 ) {
4302 $secondsPart = 0;
4303 $minutes++;
4304 }
4305 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4306 $s .= ' ';
4307 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4308 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4309 $hours = floor( $seconds / 3600 );
4310 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4311 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4312 if ( $secondsPart == 60 ) {
4313 $secondsPart = 0;
4314 $minutes++;
4315 }
4316 if ( $minutes == 60 ) {
4317 $minutes = 0;
4318 $hours++;
4319 }
4320 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4321 $s .= ' ';
4322 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4323 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4324 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4325 }
4326 } else {
4327 $days = floor( $seconds / 86400 );
4328 if ( $format['avoid'] === 'avoidminutes' ) {
4329 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4330 if ( $hours == 24 ) {
4331 $hours = 0;
4332 $days++;
4333 }
4334 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4335 $s .= ' ';
4336 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4337 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4338 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4339 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4340 if ( $minutes == 60 ) {
4341 $minutes = 0;
4342 $hours++;
4343 }
4344 if ( $hours == 24 ) {
4345 $hours = 0;
4346 $days++;
4347 }
4348 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4349 $s .= ' ';
4350 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4351 $s .= ' ';
4352 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4353 } else {
4354 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4355 $s .= ' ';
4356 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4357 }
4358 }
4359 return $s;
4360 }
4361
4362 /**
4363 * Format a bitrate for output, using an appropriate
4364 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4365 *
4366 * This use base 1000. For base 1024 use formatSize(), for another base
4367 * see formatComputingNumbers()
4368 *
4369 * @param $bps int
4370 * @return string
4371 */
4372 function formatBitrate( $bps ) {
4373 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4374 }
4375
4376 /**
4377 * @param $size int Size of the unit
4378 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4379 * @param $messageKey string Message key to be uesd
4380 * @return string
4381 */
4382 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4383 if ( $size <= 0 ) {
4384 return str_replace( '$1', $this->formatNum( $size ),
4385 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4386 );
4387 }
4388 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4389 $index = 0;
4390
4391 $maxIndex = count( $sizes ) - 1;
4392 while ( $size >= $boundary && $index < $maxIndex ) {
4393 $index++;
4394 $size /= $boundary;
4395 }
4396
4397 // For small sizes no decimal places necessary
4398 $round = 0;
4399 if ( $index > 1 ) {
4400 // For MB and bigger two decimal places are smarter
4401 $round = 2;
4402 }
4403 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4404
4405 $size = round( $size, $round );
4406 $text = $this->getMessageFromDB( $msg );
4407 return str_replace( '$1', $this->formatNum( $size ), $text );
4408 }
4409
4410 /**
4411 * Format a size in bytes for output, using an appropriate
4412 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4413 *
4414 * This method use base 1024. For base 1000 use formatBitrate(), for
4415 * another base see formatComputingNumbers()
4416 *
4417 * @param $size int Size to format
4418 * @return string Plain text (not HTML)
4419 */
4420 function formatSize( $size ) {
4421 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4422 }
4423
4424 /**
4425 * Make a list item, used by various special pages
4426 *
4427 * @param $page String Page link
4428 * @param $details String Text between brackets
4429 * @param $oppositedm Boolean Add the direction mark opposite to your
4430 * language, to display text properly
4431 * @return String
4432 */
4433 function specialList( $page, $details, $oppositedm = true ) {
4434 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) .
4435 $this->getDirMark();
4436 $details = $details ? $dirmark . $this->getMessageFromDB( 'word-separator' ) .
4437 wfMessage( 'parentheses' )->rawParams( $details )->inLanguage( $this )->escaped() : '';
4438 return $page . $details;
4439 }
4440
4441 /**
4442 * Generate (prev x| next x) (20|50|100...) type links for paging
4443 *
4444 * @param $title Title object to link
4445 * @param $offset Integer offset parameter
4446 * @param $limit Integer limit parameter
4447 * @param $query array|String optional URL query parameter string
4448 * @param $atend Bool optional param for specified if this is the last page
4449 * @return String
4450 */
4451 public function viewPrevNext( Title $title, $offset, $limit, array $query = array(), $atend = false ) {
4452 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4453
4454 # Make 'previous' link
4455 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4456 if ( $offset > 0 ) {
4457 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4458 $query, $prev, 'prevn-title', 'mw-prevlink' );
4459 } else {
4460 $plink = htmlspecialchars( $prev );
4461 }
4462
4463 # Make 'next' link
4464 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4465 if ( $atend ) {
4466 $nlink = htmlspecialchars( $next );
4467 } else {
4468 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4469 $query, $next, 'prevn-title', 'mw-nextlink' );
4470 }
4471
4472 # Make links to set number of items per page
4473 $numLinks = array();
4474 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4475 $numLinks[] = $this->numLink( $title, $offset, $num,
4476 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4477 }
4478
4479 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4480 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4481 }
4482
4483 /**
4484 * Helper function for viewPrevNext() that generates links
4485 *
4486 * @param $title Title object to link
4487 * @param $offset Integer offset parameter
4488 * @param $limit Integer limit parameter
4489 * @param $query Array extra query parameters
4490 * @param $link String text to use for the link; will be escaped
4491 * @param $tooltipMsg String name of the message to use as tooltip
4492 * @param $class String value of the "class" attribute of the link
4493 * @return String HTML fragment
4494 */
4495 private function numLink( Title $title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4496 $query = array( 'limit' => $limit, 'offset' => $offset ) + $query;
4497 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4498 return Html::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4499 'title' => $tooltip, 'class' => $class ), $link );
4500 }
4501
4502 /**
4503 * Get the conversion rule title, if any.
4504 *
4505 * @return string
4506 */
4507 public function getConvRuleTitle() {
4508 return $this->mConverter->getConvRuleTitle();
4509 }
4510
4511 /**
4512 * Get the compiled plural rules for the language
4513 * @since 1.20
4514 * @return array Associative array with plural form, and plural rule as key-value pairs
4515 */
4516 public function getCompiledPluralRules() {
4517 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4518 $fallbacks = Language::getFallbacksFor( $this->mCode );
4519 if ( !$pluralRules ) {
4520 foreach ( $fallbacks as $fallbackCode ) {
4521 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4522 if ( $pluralRules ) {
4523 break;
4524 }
4525 }
4526 }
4527 return $pluralRules;
4528 }
4529
4530 /**
4531 * Get the plural rules for the language
4532 * @since 1.20
4533 * @return array Associative array with plural form number and plural rule as key-value pairs
4534 */
4535 public function getPluralRules() {
4536 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4537 $fallbacks = Language::getFallbacksFor( $this->mCode );
4538 if ( !$pluralRules ) {
4539 foreach ( $fallbacks as $fallbackCode ) {
4540 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4541 if ( $pluralRules ) {
4542 break;
4543 }
4544 }
4545 }
4546 return $pluralRules;
4547 }
4548
4549 /**
4550 * Get the plural rule types for the language
4551 * @since 1.22
4552 * @return array Associative array with plural form number and plural rule type as key-value pairs
4553 */
4554 public function getPluralRuleTypes() {
4555 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4556 $fallbacks = Language::getFallbacksFor( $this->mCode );
4557 if ( !$pluralRuleTypes ) {
4558 foreach ( $fallbacks as $fallbackCode ) {
4559 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4560 if ( $pluralRuleTypes ) {
4561 break;
4562 }
4563 }
4564 }
4565 return $pluralRuleTypes;
4566 }
4567
4568 /**
4569 * Find the index number of the plural rule appropriate for the given number
4570 * @return int The index number of the plural rule
4571 */
4572 public function getPluralRuleIndexNumber( $number ) {
4573 $pluralRules = $this->getCompiledPluralRules();
4574 $form = CLDRPluralRuleEvaluator::evaluateCompiled( $number, $pluralRules );
4575 return $form;
4576 }
4577
4578 /**
4579 * Find the plural rule type appropriate for the given number
4580 * For example, if the language is set to Arabic, getPluralType(5) should
4581 * return 'few'.
4582 * @since 1.22
4583 * @return string The name of the plural rule type, e.g. one, two, few, many
4584 */
4585 public function getPluralRuleType( $number ) {
4586 $index = $this->getPluralRuleIndexNumber( $number );
4587 $pluralRuleTypes = $this->getPluralRuleTypes();
4588 if ( isset( $pluralRuleTypes[$index] ) ) {
4589 return $pluralRuleTypes[$index];
4590 } else {
4591 return 'other';
4592 }
4593 }
4594 }
Wiklou, le Wiki du Wiklou