dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Merge "New function Language::getParentLanguage()."
[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 # Read language names
34 global $wgLanguageNames;
35 require_once __DIR__ . '/Names.php';
36
37 if ( function_exists( 'mb_strtoupper' ) ) {
38 mb_internal_encoding( 'UTF-8' );
39 }
40
41 /**
42 * a fake language converter
43 *
44 * @ingroup Language
45 */
46 class FakeConverter {
47 /**
48 * @var Language
49 */
50 public $mLang;
51 function __construct( $langobj ) { $this->mLang = $langobj; }
52 function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
53 function convert( $t ) { return $t; }
54 function convertTo( $text, $variant ) { return $text; }
55 function convertTitle( $t ) { return $t->getPrefixedText(); }
56 function convertNamespace( $ns ) { return $this->mLang->getFormattedNsText( $ns ); }
57 function getVariants() { return array( $this->mLang->getCode() ); }
58 function getPreferredVariant() { return $this->mLang->getCode(); }
59 function getDefaultVariant() { return $this->mLang->getCode(); }
60 function getURLVariant() { return ''; }
61 function getConvRuleTitle() { return false; }
62 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
63 function getExtraHashOptions() { return ''; }
64 function getParsedTitle() { return ''; }
65 function markNoConversion( $text, $noParse = false ) { return $text; }
66 function convertCategoryKey( $key ) { return $key; }
67 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
68 function armourMath( $text ) { 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 ( MWInit::classExists( $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 ( MWInit::classExists( $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 include MWInit::compiledPath( 'languages/Names.php' );
396 }
397
398 if ( isset( $coreLanguageNames[$tag] )
399 || self::fetchLanguageName( $tag, $tag ) !== ''
400 ) {
401 return true;
402 }
403
404 return false;
405 }
406
407 /**
408 * @param $code
409 * @return String Name of the language class
410 */
411 public static function classFromCode( $code ) {
412 if ( $code == 'en' ) {
413 return 'Language';
414 } else {
415 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
416 }
417 }
418
419 /**
420 * Includes language class files
421 *
422 * @param $class string Name of the language class
423 */
424 public static function preloadLanguageClass( $class ) {
425 global $IP;
426
427 if ( $class === 'Language' ) {
428 return;
429 }
430
431 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
432 include_once "$IP/languages/classes/$class.php";
433 }
434 }
435
436 /**
437 * Get the LocalisationCache instance
438 *
439 * @return LocalisationCache
440 */
441 public static function getLocalisationCache() {
442 if ( is_null( self::$dataCache ) ) {
443 global $wgLocalisationCacheConf;
444 $class = $wgLocalisationCacheConf['class'];
445 self::$dataCache = new $class( $wgLocalisationCacheConf );
446 }
447 return self::$dataCache;
448 }
449
450 function __construct() {
451 $this->mConverter = new FakeConverter( $this );
452 // Set the code to the name of the descendant
453 if ( get_class( $this ) == 'Language' ) {
454 $this->mCode = 'en';
455 } else {
456 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
457 }
458 self::getLocalisationCache();
459 }
460
461 /**
462 * Reduce memory usage
463 */
464 function __destruct() {
465 foreach ( $this as $name => $value ) {
466 unset( $this->$name );
467 }
468 }
469
470 /**
471 * Hook which will be called if this is the content language.
472 * Descendants can use this to register hook functions or modify globals
473 */
474 function initContLang() { }
475
476 /**
477 * Same as getFallbacksFor for current language.
478 * @return array|bool
479 * @deprecated in 1.19
480 */
481 function getFallbackLanguageCode() {
482 wfDeprecated( __METHOD__, '1.19' );
483 return self::getFallbackFor( $this->mCode );
484 }
485
486 /**
487 * @return array
488 * @since 1.19
489 */
490 function getFallbackLanguages() {
491 return self::getFallbacksFor( $this->mCode );
492 }
493
494 /**
495 * Exports $wgBookstoreListEn
496 * @return array
497 */
498 function getBookstoreList() {
499 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
500 }
501
502 /**
503 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
504 * available in localised form, it will be included in English.
505 *
506 * @return array
507 */
508 public function getNamespaces() {
509 if ( is_null( $this->namespaceNames ) ) {
510 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
511
512 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
513 $validNamespaces = MWNamespace::getCanonicalNamespaces();
514
515 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
516
517 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
518 if ( $wgMetaNamespaceTalk ) {
519 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
520 } else {
521 $talk = $this->namespaceNames[NS_PROJECT_TALK];
522 $this->namespaceNames[NS_PROJECT_TALK] =
523 $this->fixVariableInNamespace( $talk );
524 }
525
526 # Sometimes a language will be localised but not actually exist on this wiki.
527 foreach ( $this->namespaceNames as $key => $text ) {
528 if ( !isset( $validNamespaces[$key] ) ) {
529 unset( $this->namespaceNames[$key] );
530 }
531 }
532
533 # The above mixing may leave namespaces out of canonical order.
534 # Re-order by namespace ID number...
535 ksort( $this->namespaceNames );
536
537 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames ) );
538 }
539 return $this->namespaceNames;
540 }
541
542 /**
543 * Arbitrarily set all of the namespace names at once. Mainly used for testing
544 * @param $namespaces Array of namespaces (id => name)
545 */
546 public function setNamespaces( array $namespaces ) {
547 $this->namespaceNames = $namespaces;
548 $this->mNamespaceIds = null;
549 }
550
551 /**
552 * Resets all of the namespace caches. Mainly used for testing
553 */
554 public function resetNamespaces() {
555 $this->namespaceNames = null;
556 $this->mNamespaceIds = null;
557 $this->namespaceAliases = null;
558 }
559
560 /**
561 * A convenience function that returns the same thing as
562 * getNamespaces() except with the array values changed to ' '
563 * where it found '_', useful for producing output to be displayed
564 * e.g. in <select> forms.
565 *
566 * @return array
567 */
568 function getFormattedNamespaces() {
569 $ns = $this->getNamespaces();
570 foreach ( $ns as $k => $v ) {
571 $ns[$k] = strtr( $v, '_', ' ' );
572 }
573 return $ns;
574 }
575
576 /**
577 * Get a namespace value by key
578 * <code>
579 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
580 * echo $mw_ns; // prints 'MediaWiki'
581 * </code>
582 *
583 * @param $index Int: the array key of the namespace to return
584 * @return mixed, string if the namespace value exists, otherwise false
585 */
586 function getNsText( $index ) {
587 $ns = $this->getNamespaces();
588 return isset( $ns[$index] ) ? $ns[$index] : false;
589 }
590
591 /**
592 * A convenience function that returns the same thing as
593 * getNsText() except with '_' changed to ' ', useful for
594 * producing output.
595 *
596 * <code>
597 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
598 * echo $mw_ns; // prints 'MediaWiki talk'
599 * </code>
600 *
601 * @param int $index The array key of the namespace to return
602 * @return string Namespace name without underscores (empty string if namespace does not exist)
603 */
604 function getFormattedNsText( $index ) {
605 $ns = $this->getNsText( $index );
606 return strtr( $ns, '_', ' ' );
607 }
608
609 /**
610 * Returns gender-dependent namespace alias if available.
611 * @param $index Int: namespace index
612 * @param $gender String: gender key (male, female... )
613 * @return String
614 * @since 1.18
615 */
616 function getGenderNsText( $index, $gender ) {
617 global $wgExtraGenderNamespaces;
618
619 $ns = $wgExtraGenderNamespaces + self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
620 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
621 }
622
623 /**
624 * Whether this language makes distinguishes genders for example in
625 * namespaces.
626 * @return bool
627 * @since 1.18
628 */
629 function needsGenderDistinction() {
630 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
631 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
632 // $wgExtraGenderNamespaces overrides everything
633 return true;
634 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
635 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
636 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
637 return false;
638 } else {
639 // Check what is in i18n files
640 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
641 return count( $aliases ) > 0;
642 }
643 }
644
645 /**
646 * Get a namespace key by value, case insensitive.
647 * Only matches namespace names for the current language, not the
648 * canonical ones defined in Namespace.php.
649 *
650 * @param $text String
651 * @return mixed An integer if $text is a valid value otherwise false
652 */
653 function getLocalNsIndex( $text ) {
654 $lctext = $this->lc( $text );
655 $ids = $this->getNamespaceIds();
656 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
657 }
658
659 /**
660 * @return array
661 */
662 function getNamespaceAliases() {
663 if ( is_null( $this->namespaceAliases ) ) {
664 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
665 if ( !$aliases ) {
666 $aliases = array();
667 } else {
668 foreach ( $aliases as $name => $index ) {
669 if ( $index === NS_PROJECT_TALK ) {
670 unset( $aliases[$name] );
671 $name = $this->fixVariableInNamespace( $name );
672 $aliases[$name] = $index;
673 }
674 }
675 }
676
677 global $wgExtraGenderNamespaces;
678 $genders = $wgExtraGenderNamespaces + (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
679 foreach ( $genders as $index => $forms ) {
680 foreach ( $forms as $alias ) {
681 $aliases[$alias] = $index;
682 }
683 }
684
685 $this->namespaceAliases = $aliases;
686 }
687 return $this->namespaceAliases;
688 }
689
690 /**
691 * @return array
692 */
693 function getNamespaceIds() {
694 if ( is_null( $this->mNamespaceIds ) ) {
695 global $wgNamespaceAliases;
696 # Put namespace names and aliases into a hashtable.
697 # If this is too slow, then we should arrange it so that it is done
698 # before caching. The catch is that at pre-cache time, the above
699 # class-specific fixup hasn't been done.
700 $this->mNamespaceIds = array();
701 foreach ( $this->getNamespaces() as $index => $name ) {
702 $this->mNamespaceIds[$this->lc( $name )] = $index;
703 }
704 foreach ( $this->getNamespaceAliases() as $name => $index ) {
705 $this->mNamespaceIds[$this->lc( $name )] = $index;
706 }
707 if ( $wgNamespaceAliases ) {
708 foreach ( $wgNamespaceAliases as $name => $index ) {
709 $this->mNamespaceIds[$this->lc( $name )] = $index;
710 }
711 }
712 }
713 return $this->mNamespaceIds;
714 }
715
716 /**
717 * Get a namespace key by value, case insensitive. Canonical namespace
718 * names override custom ones defined for the current language.
719 *
720 * @param $text String
721 * @return mixed An integer if $text is a valid value otherwise false
722 */
723 function getNsIndex( $text ) {
724 $lctext = $this->lc( $text );
725 $ns = MWNamespace::getCanonicalIndex( $lctext );
726 if ( $ns !== null ) {
727 return $ns;
728 }
729 $ids = $this->getNamespaceIds();
730 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
731 }
732
733 /**
734 * short names for language variants used for language conversion links.
735 *
736 * @param $code String
737 * @param $usemsg bool Use the "variantname-xyz" message if it exists
738 * @return string
739 */
740 function getVariantname( $code, $usemsg = true ) {
741 $msg = "variantname-$code";
742 if ( $usemsg && wfMessage( $msg )->exists() ) {
743 return $this->getMessageFromDB( $msg );
744 }
745 $name = self::fetchLanguageName( $code );
746 if ( $name ) {
747 return $name; # if it's defined as a language name, show that
748 } else {
749 # otherwise, output the language code
750 return $code;
751 }
752 }
753
754 /**
755 * @param $name string
756 * @return string
757 */
758 function specialPage( $name ) {
759 $aliases = $this->getSpecialPageAliases();
760 if ( isset( $aliases[$name][0] ) ) {
761 $name = $aliases[$name][0];
762 }
763 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
764 }
765
766 /**
767 * @return array
768 */
769 function getDatePreferences() {
770 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
771 }
772
773 /**
774 * @return array
775 */
776 function getDateFormats() {
777 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
778 }
779
780 /**
781 * @return array|string
782 */
783 function getDefaultDateFormat() {
784 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
785 if ( $df === 'dmy or mdy' ) {
786 global $wgAmericanDates;
787 return $wgAmericanDates ? 'mdy' : 'dmy';
788 } else {
789 return $df;
790 }
791 }
792
793 /**
794 * @return array
795 */
796 function getDatePreferenceMigrationMap() {
797 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
798 }
799
800 /**
801 * @param $image
802 * @return array|null
803 */
804 function getImageFile( $image ) {
805 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
806 }
807
808 /**
809 * @return array
810 */
811 function getExtraUserToggles() {
812 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
813 }
814
815 /**
816 * @param $tog
817 * @return string
818 */
819 function getUserToggle( $tog ) {
820 return $this->getMessageFromDB( "tog-$tog" );
821 }
822
823 /**
824 * Get native language names, indexed by code.
825 * Only those defined in MediaWiki, no other data like CLDR.
826 * If $customisedOnly is true, only returns codes with a messages file
827 *
828 * @param $customisedOnly bool
829 *
830 * @return array
831 * @deprecated in 1.20, use fetchLanguageNames()
832 */
833 public static function getLanguageNames( $customisedOnly = false ) {
834 return self::fetchLanguageNames( null, $customisedOnly ? 'mwfile' : 'mw' );
835 }
836
837 /**
838 * Get translated language names. This is done on best effort and
839 * by default this is exactly the same as Language::getLanguageNames.
840 * The CLDR extension provides translated names.
841 * @param $code String Language code.
842 * @return Array language code => language name
843 * @since 1.18.0
844 * @deprecated in 1.20, use fetchLanguageNames()
845 */
846 public static function getTranslatedLanguageNames( $code ) {
847 return self::fetchLanguageNames( $code, 'all' );
848 }
849
850 /**
851 * Get an array of language names, indexed by code.
852 * @param $inLanguage null|string: Code of language in which to return the names
853 * Use null for autonyms (native names)
854 * @param $include string:
855 * 'all' all available languages
856 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
857 * 'mwfile' only if the language is in 'mw' *and* has a message file
858 * @return array: language code => language name
859 * @since 1.20
860 */
861 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
862 global $wgExtraLanguageNames;
863 static $coreLanguageNames;
864
865 if ( $coreLanguageNames === null ) {
866 include MWInit::compiledPath( 'languages/Names.php' );
867 }
868
869 $names = array();
870
871 if ( $inLanguage ) {
872 # TODO: also include when $inLanguage is null, when this code is more efficient
873 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
874 }
875
876 $mwNames = $wgExtraLanguageNames + $coreLanguageNames;
877 foreach ( $mwNames as $mwCode => $mwName ) {
878 # - Prefer own MediaWiki native name when not using the hook
879 # - For other names just add if not added through the hook
880 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
881 $names[$mwCode] = $mwName;
882 }
883 }
884
885 if ( $include === 'all' ) {
886 return $names;
887 }
888
889 $returnMw = array();
890 $coreCodes = array_keys( $mwNames );
891 foreach ( $coreCodes as $coreCode ) {
892 $returnMw[$coreCode] = $names[$coreCode];
893 }
894
895 if ( $include === 'mwfile' ) {
896 $namesMwFile = array();
897 # We do this using a foreach over the codes instead of a directory
898 # loop so that messages files in extensions will work correctly.
899 foreach ( $returnMw as $code => $value ) {
900 if ( is_readable( self::getMessagesFileName( $code ) ) ) {
901 $namesMwFile[$code] = $names[$code];
902 }
903 }
904 return $namesMwFile;
905 }
906 # 'mw' option; default if it's not one of the other two options (all/mwfile)
907 return $returnMw;
908 }
909
910 /**
911 * @param $code string: The code of the language for which to get the name
912 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
913 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
914 * @return string: Language name or empty
915 * @since 1.20
916 */
917 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
918 $array = self::fetchLanguageNames( $inLanguage, $include );
919 return !array_key_exists( $code, $array ) ? '' : $array[$code];
920 }
921
922 /**
923 * Get a message from the MediaWiki namespace.
924 *
925 * @param $msg String: message name
926 * @return string
927 */
928 function getMessageFromDB( $msg ) {
929 return wfMessage( $msg )->inLanguage( $this )->text();
930 }
931
932 /**
933 * Get the native language name of $code.
934 * Only if defined in MediaWiki, no other data like CLDR.
935 * @param $code string
936 * @return string
937 * @deprecated in 1.20, use fetchLanguageName()
938 */
939 function getLanguageName( $code ) {
940 return self::fetchLanguageName( $code );
941 }
942
943 /**
944 * @param $key string
945 * @return string
946 */
947 function getMonthName( $key ) {
948 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
949 }
950
951 /**
952 * @return array
953 */
954 function getMonthNamesArray() {
955 $monthNames = array( '' );
956 for ( $i = 1; $i < 13; $i++ ) {
957 $monthNames[] = $this->getMonthName( $i );
958 }
959 return $monthNames;
960 }
961
962 /**
963 * @param $key string
964 * @return string
965 */
966 function getMonthNameGen( $key ) {
967 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
968 }
969
970 /**
971 * @param $key string
972 * @return string
973 */
974 function getMonthAbbreviation( $key ) {
975 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
976 }
977
978 /**
979 * @return array
980 */
981 function getMonthAbbreviationsArray() {
982 $monthNames = array( '' );
983 for ( $i = 1; $i < 13; $i++ ) {
984 $monthNames[] = $this->getMonthAbbreviation( $i );
985 }
986 return $monthNames;
987 }
988
989 /**
990 * @param $key string
991 * @return string
992 */
993 function getWeekdayName( $key ) {
994 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
995 }
996
997 /**
998 * @param $key string
999 * @return string
1000 */
1001 function getWeekdayAbbreviation( $key ) {
1002 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
1003 }
1004
1005 /**
1006 * @param $key string
1007 * @return string
1008 */
1009 function getIranianCalendarMonthName( $key ) {
1010 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
1011 }
1012
1013 /**
1014 * @param $key string
1015 * @return string
1016 */
1017 function getHebrewCalendarMonthName( $key ) {
1018 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
1019 }
1020
1021 /**
1022 * @param $key string
1023 * @return string
1024 */
1025 function getHebrewCalendarMonthNameGen( $key ) {
1026 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1027 }
1028
1029 /**
1030 * @param $key string
1031 * @return string
1032 */
1033 function getHijriCalendarMonthName( $key ) {
1034 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1035 }
1036
1037 /**
1038 * This is a workalike of PHP's date() function, but with better
1039 * internationalisation, a reduced set of format characters, and a better
1040 * escaping format.
1041 *
1042 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1043 * the PHP manual for definitions. There are a number of extensions, which
1044 * start with "x":
1045 *
1046 * xn Do not translate digits of the next numeric format character
1047 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1048 * xr Use roman numerals for the next numeric format character
1049 * xh Use hebrew numerals for the next numeric format character
1050 * xx Literal x
1051 * xg Genitive month name
1052 *
1053 * xij j (day number) in Iranian calendar
1054 * xiF F (month name) in Iranian calendar
1055 * xin n (month number) in Iranian calendar
1056 * xiy y (two digit year) in Iranian calendar
1057 * xiY Y (full year) in Iranian calendar
1058 *
1059 * xjj j (day number) in Hebrew calendar
1060 * xjF F (month name) in Hebrew calendar
1061 * xjt t (days in month) in Hebrew calendar
1062 * xjx xg (genitive month name) in Hebrew calendar
1063 * xjn n (month number) in Hebrew calendar
1064 * xjY Y (full year) in Hebrew calendar
1065 *
1066 * xmj j (day number) in Hijri calendar
1067 * xmF F (month name) in Hijri calendar
1068 * xmn n (month number) in Hijri calendar
1069 * xmY Y (full year) in Hijri calendar
1070 *
1071 * xkY Y (full year) in Thai solar calendar. Months and days are
1072 * identical to the Gregorian calendar
1073 * xoY Y (full year) in Minguo calendar or Juche year.
1074 * Months and days are identical to the
1075 * Gregorian calendar
1076 * xtY Y (full year) in Japanese nengo. Months and days are
1077 * identical to the Gregorian calendar
1078 *
1079 * Characters enclosed in double quotes will be considered literal (with
1080 * the quotes themselves removed). Unmatched quotes will be considered
1081 * literal quotes. Example:
1082 *
1083 * "The month is" F => The month is January
1084 * i's" => 20'11"
1085 *
1086 * Backslash escaping is also supported.
1087 *
1088 * Input timestamp is assumed to be pre-normalized to the desired local
1089 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1090 * $ts is UTC if $zone is not given.
1091 *
1092 * @param $format String
1093 * @param $ts String: 14-character timestamp
1094 * YYYYMMDDHHMMSS
1095 * 01234567890123
1096 * @param $zone DateTimeZone: Timezone of $ts
1097 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1098 *
1099 * @throws MWException
1100 * @return string
1101 */
1102 function sprintfDate( $format, $ts, DateTimeZone $zone = null ) {
1103 $s = '';
1104 $raw = false;
1105 $roman = false;
1106 $hebrewNum = false;
1107 $dateTimeObj = false;
1108 $rawToggle = false;
1109 $iranian = false;
1110 $hebrew = false;
1111 $hijri = false;
1112 $thai = false;
1113 $minguo = false;
1114 $tenno = false;
1115
1116 if ( strlen( $ts ) !== 14 ) {
1117 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1118 }
1119
1120 if ( !ctype_digit( $ts ) ) {
1121 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1122 }
1123
1124 for ( $p = 0; $p < strlen( $format ); $p++ ) {
1125 $num = false;
1126 $code = $format[$p];
1127 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
1128 $code .= $format[++$p];
1129 }
1130
1131 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
1132 $code .= $format[++$p];
1133 }
1134
1135 switch ( $code ) {
1136 case 'xx':
1137 $s .= 'x';
1138 break;
1139 case 'xn':
1140 $raw = true;
1141 break;
1142 case 'xN':
1143 $rawToggle = !$rawToggle;
1144 break;
1145 case 'xr':
1146 $roman = true;
1147 break;
1148 case 'xh':
1149 $hebrewNum = true;
1150 break;
1151 case 'xg':
1152 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1153 break;
1154 case 'xjx':
1155 if ( !$hebrew ) {
1156 $hebrew = self::tsToHebrew( $ts );
1157 }
1158 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1159 break;
1160 case 'd':
1161 $num = substr( $ts, 6, 2 );
1162 break;
1163 case 'D':
1164 if ( !$dateTimeObj ) {
1165 $dateTimeObj = DateTime::createFromFormat(
1166 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1167 );
1168 }
1169 $s .= $this->getWeekdayAbbreviation( $dateTimeObj->format( 'w' ) + 1 );
1170 break;
1171 case 'j':
1172 $num = intval( substr( $ts, 6, 2 ) );
1173 break;
1174 case 'xij':
1175 if ( !$iranian ) {
1176 $iranian = self::tsToIranian( $ts );
1177 }
1178 $num = $iranian[2];
1179 break;
1180 case 'xmj':
1181 if ( !$hijri ) {
1182 $hijri = self::tsToHijri( $ts );
1183 }
1184 $num = $hijri[2];
1185 break;
1186 case 'xjj':
1187 if ( !$hebrew ) {
1188 $hebrew = self::tsToHebrew( $ts );
1189 }
1190 $num = $hebrew[2];
1191 break;
1192 case 'l':
1193 if ( !$dateTimeObj ) {
1194 $dateTimeObj = DateTime::createFromFormat(
1195 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1196 );
1197 }
1198 $s .= $this->getWeekdayName( $dateTimeObj->format( 'w' ) + 1 );
1199 break;
1200 case 'F':
1201 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1202 break;
1203 case 'xiF':
1204 if ( !$iranian ) {
1205 $iranian = self::tsToIranian( $ts );
1206 }
1207 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1208 break;
1209 case 'xmF':
1210 if ( !$hijri ) {
1211 $hijri = self::tsToHijri( $ts );
1212 }
1213 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1214 break;
1215 case 'xjF':
1216 if ( !$hebrew ) {
1217 $hebrew = self::tsToHebrew( $ts );
1218 }
1219 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1220 break;
1221 case 'm':
1222 $num = substr( $ts, 4, 2 );
1223 break;
1224 case 'M':
1225 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1226 break;
1227 case 'n':
1228 $num = intval( substr( $ts, 4, 2 ) );
1229 break;
1230 case 'xin':
1231 if ( !$iranian ) {
1232 $iranian = self::tsToIranian( $ts );
1233 }
1234 $num = $iranian[1];
1235 break;
1236 case 'xmn':
1237 if ( !$hijri ) {
1238 $hijri = self::tsToHijri ( $ts );
1239 }
1240 $num = $hijri[1];
1241 break;
1242 case 'xjn':
1243 if ( !$hebrew ) {
1244 $hebrew = self::tsToHebrew( $ts );
1245 }
1246 $num = $hebrew[1];
1247 break;
1248 case 'xjt':
1249 if ( !$hebrew ) {
1250 $hebrew = self::tsToHebrew( $ts );
1251 }
1252 $num = $hebrew[3];
1253 break;
1254 case 'Y':
1255 $num = substr( $ts, 0, 4 );
1256 break;
1257 case 'xiY':
1258 if ( !$iranian ) {
1259 $iranian = self::tsToIranian( $ts );
1260 }
1261 $num = $iranian[0];
1262 break;
1263 case 'xmY':
1264 if ( !$hijri ) {
1265 $hijri = self::tsToHijri( $ts );
1266 }
1267 $num = $hijri[0];
1268 break;
1269 case 'xjY':
1270 if ( !$hebrew ) {
1271 $hebrew = self::tsToHebrew( $ts );
1272 }
1273 $num = $hebrew[0];
1274 break;
1275 case 'xkY':
1276 if ( !$thai ) {
1277 $thai = self::tsToYear( $ts, 'thai' );
1278 }
1279 $num = $thai[0];
1280 break;
1281 case 'xoY':
1282 if ( !$minguo ) {
1283 $minguo = self::tsToYear( $ts, 'minguo' );
1284 }
1285 $num = $minguo[0];
1286 break;
1287 case 'xtY':
1288 if ( !$tenno ) {
1289 $tenno = self::tsToYear( $ts, 'tenno' );
1290 }
1291 $num = $tenno[0];
1292 break;
1293 case 'y':
1294 $num = substr( $ts, 2, 2 );
1295 break;
1296 case 'xiy':
1297 if ( !$iranian ) {
1298 $iranian = self::tsToIranian( $ts );
1299 }
1300 $num = substr( $iranian[0], -2 );
1301 break;
1302 case 'a':
1303 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1304 break;
1305 case 'A':
1306 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1307 break;
1308 case 'g':
1309 $h = substr( $ts, 8, 2 );
1310 $num = $h % 12 ? $h % 12 : 12;
1311 break;
1312 case 'G':
1313 $num = intval( substr( $ts, 8, 2 ) );
1314 break;
1315 case 'h':
1316 $h = substr( $ts, 8, 2 );
1317 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1318 break;
1319 case 'H':
1320 $num = substr( $ts, 8, 2 );
1321 break;
1322 case 'i':
1323 $num = substr( $ts, 10, 2 );
1324 break;
1325 case 's':
1326 $num = substr( $ts, 12, 2 );
1327 break;
1328 case 'c':
1329 case 'r':
1330 case 'e':
1331 case 'O':
1332 case 'P':
1333 case 'T':
1334 // Pass through string from $dateTimeObj->format()
1335 if ( !$dateTimeObj ) {
1336 $dateTimeObj = DateTime::createFromFormat(
1337 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1338 );
1339 }
1340 $s .= $dateTimeObj->format( $code );
1341 break;
1342 case 'w':
1343 case 'N':
1344 case 'z':
1345 case 'W':
1346 case 't':
1347 case 'L':
1348 case 'o':
1349 case 'U':
1350 case 'I':
1351 case 'Z':
1352 // Pass through number from $dateTimeObj->format()
1353 if ( !$dateTimeObj ) {
1354 $dateTimeObj = DateTime::createFromFormat(
1355 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1356 );
1357 }
1358 $num = $dateTimeObj->format( $code );
1359 break;
1360 case '\\':
1361 # Backslash escaping
1362 if ( $p < strlen( $format ) - 1 ) {
1363 $s .= $format[++$p];
1364 } else {
1365 $s .= '\\';
1366 }
1367 break;
1368 case '"':
1369 # Quoted literal
1370 if ( $p < strlen( $format ) - 1 ) {
1371 $endQuote = strpos( $format, '"', $p + 1 );
1372 if ( $endQuote === false ) {
1373 # No terminating quote, assume literal "
1374 $s .= '"';
1375 } else {
1376 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1377 $p = $endQuote;
1378 }
1379 } else {
1380 # Quote at end of string, assume literal "
1381 $s .= '"';
1382 }
1383 break;
1384 default:
1385 $s .= $format[$p];
1386 }
1387 if ( $num !== false ) {
1388 if ( $rawToggle || $raw ) {
1389 $s .= $num;
1390 $raw = false;
1391 } elseif ( $roman ) {
1392 $s .= Language::romanNumeral( $num );
1393 $roman = false;
1394 } elseif ( $hebrewNum ) {
1395 $s .= self::hebrewNumeral( $num );
1396 $hebrewNum = false;
1397 } else {
1398 $s .= $this->formatNum( $num, true );
1399 }
1400 }
1401 }
1402 return $s;
1403 }
1404
1405 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1406 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1407
1408 /**
1409 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1410 * Gregorian dates to Iranian dates. Originally written in C, it
1411 * is released under the terms of GNU Lesser General Public
1412 * License. Conversion to PHP was performed by Niklas Laxström.
1413 *
1414 * Link: http://www.farsiweb.info/jalali/jalali.c
1415 *
1416 * @param $ts string
1417 *
1418 * @return string
1419 */
1420 private static function tsToIranian( $ts ) {
1421 $gy = substr( $ts, 0, 4 ) -1600;
1422 $gm = substr( $ts, 4, 2 ) -1;
1423 $gd = substr( $ts, 6, 2 ) -1;
1424
1425 # Days passed from the beginning (including leap years)
1426 $gDayNo = 365 * $gy
1427 + floor( ( $gy + 3 ) / 4 )
1428 - floor( ( $gy + 99 ) / 100 )
1429 + floor( ( $gy + 399 ) / 400 );
1430
1431 // Add days of the past months of this year
1432 for ( $i = 0; $i < $gm; $i++ ) {
1433 $gDayNo += self::$GREG_DAYS[$i];
1434 }
1435
1436 // Leap years
1437 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1438 $gDayNo++;
1439 }
1440
1441 // Days passed in current month
1442 $gDayNo += (int)$gd;
1443
1444 $jDayNo = $gDayNo - 79;
1445
1446 $jNp = floor( $jDayNo / 12053 );
1447 $jDayNo %= 12053;
1448
1449 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1450 $jDayNo %= 1461;
1451
1452 if ( $jDayNo >= 366 ) {
1453 $jy += floor( ( $jDayNo - 1 ) / 365 );
1454 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1455 }
1456
1457 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1458 $jDayNo -= self::$IRANIAN_DAYS[$i];
1459 }
1460
1461 $jm = $i + 1;
1462 $jd = $jDayNo + 1;
1463
1464 return array( $jy, $jm, $jd );
1465 }
1466
1467 /**
1468 * Converting Gregorian dates to Hijri dates.
1469 *
1470 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1471 *
1472 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1473 *
1474 * @param $ts string
1475 *
1476 * @return string
1477 */
1478 private static function tsToHijri( $ts ) {
1479 $year = substr( $ts, 0, 4 );
1480 $month = substr( $ts, 4, 2 );
1481 $day = substr( $ts, 6, 2 );
1482
1483 $zyr = $year;
1484 $zd = $day;
1485 $zm = $month;
1486 $zy = $zyr;
1487
1488 if (
1489 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1490 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1491 )
1492 {
1493 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1494 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1495 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1496 $zd - 32075;
1497 } else {
1498 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1499 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1500 }
1501
1502 $zl = $zjd -1948440 + 10632;
1503 $zn = (int)( ( $zl - 1 ) / 10631 );
1504 $zl = $zl - 10631 * $zn + 354;
1505 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1506 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1507 $zm = (int)( ( 24 * $zl ) / 709 );
1508 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1509 $zy = 30 * $zn + $zj - 30;
1510
1511 return array( $zy, $zm, $zd );
1512 }
1513
1514 /**
1515 * Converting Gregorian dates to Hebrew dates.
1516 *
1517 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1518 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1519 * to translate the relevant functions into PHP and release them under
1520 * GNU GPL.
1521 *
1522 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1523 * and Adar II is 14. In a non-leap year, Adar is 6.
1524 *
1525 * @param $ts string
1526 *
1527 * @return string
1528 */
1529 private static function tsToHebrew( $ts ) {
1530 # Parse date
1531 $year = substr( $ts, 0, 4 );
1532 $month = substr( $ts, 4, 2 );
1533 $day = substr( $ts, 6, 2 );
1534
1535 # Calculate Hebrew year
1536 $hebrewYear = $year + 3760;
1537
1538 # Month number when September = 1, August = 12
1539 $month += 4;
1540 if ( $month > 12 ) {
1541 # Next year
1542 $month -= 12;
1543 $year++;
1544 $hebrewYear++;
1545 }
1546
1547 # Calculate day of year from 1 September
1548 $dayOfYear = $day;
1549 for ( $i = 1; $i < $month; $i++ ) {
1550 if ( $i == 6 ) {
1551 # February
1552 $dayOfYear += 28;
1553 # Check if the year is leap
1554 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1555 $dayOfYear++;
1556 }
1557 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1558 $dayOfYear += 30;
1559 } else {
1560 $dayOfYear += 31;
1561 }
1562 }
1563
1564 # Calculate the start of the Hebrew year
1565 $start = self::hebrewYearStart( $hebrewYear );
1566
1567 # Calculate next year's start
1568 if ( $dayOfYear <= $start ) {
1569 # Day is before the start of the year - it is the previous year
1570 # Next year's start
1571 $nextStart = $start;
1572 # Previous year
1573 $year--;
1574 $hebrewYear--;
1575 # Add days since previous year's 1 September
1576 $dayOfYear += 365;
1577 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1578 # Leap year
1579 $dayOfYear++;
1580 }
1581 # Start of the new (previous) year
1582 $start = self::hebrewYearStart( $hebrewYear );
1583 } else {
1584 # Next year's start
1585 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1586 }
1587
1588 # Calculate Hebrew day of year
1589 $hebrewDayOfYear = $dayOfYear - $start;
1590
1591 # Difference between year's days
1592 $diff = $nextStart - $start;
1593 # Add 12 (or 13 for leap years) days to ignore the difference between
1594 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1595 # difference is only about the year type
1596 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1597 $diff += 13;
1598 } else {
1599 $diff += 12;
1600 }
1601
1602 # Check the year pattern, and is leap year
1603 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1604 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1605 # and non-leap years
1606 $yearPattern = $diff % 30;
1607 # Check if leap year
1608 $isLeap = $diff >= 30;
1609
1610 # Calculate day in the month from number of day in the Hebrew year
1611 # Don't check Adar - if the day is not in Adar, we will stop before;
1612 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1613 $hebrewDay = $hebrewDayOfYear;
1614 $hebrewMonth = 1;
1615 $days = 0;
1616 while ( $hebrewMonth <= 12 ) {
1617 # Calculate days in this month
1618 if ( $isLeap && $hebrewMonth == 6 ) {
1619 # Adar in a leap year
1620 if ( $isLeap ) {
1621 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1622 $days = 30;
1623 if ( $hebrewDay <= $days ) {
1624 # Day in Adar I
1625 $hebrewMonth = 13;
1626 } else {
1627 # Subtract the days of Adar I
1628 $hebrewDay -= $days;
1629 # Try Adar II
1630 $days = 29;
1631 if ( $hebrewDay <= $days ) {
1632 # Day in Adar II
1633 $hebrewMonth = 14;
1634 }
1635 }
1636 }
1637 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1638 # Cheshvan in a complete year (otherwise as the rule below)
1639 $days = 30;
1640 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1641 # Kislev in an incomplete year (otherwise as the rule below)
1642 $days = 29;
1643 } else {
1644 # Odd months have 30 days, even have 29
1645 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1646 }
1647 if ( $hebrewDay <= $days ) {
1648 # In the current month
1649 break;
1650 } else {
1651 # Subtract the days of the current month
1652 $hebrewDay -= $days;
1653 # Try in the next month
1654 $hebrewMonth++;
1655 }
1656 }
1657
1658 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1659 }
1660
1661 /**
1662 * This calculates the Hebrew year start, as days since 1 September.
1663 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1664 * Used for Hebrew date.
1665 *
1666 * @param $year int
1667 *
1668 * @return string
1669 */
1670 private static function hebrewYearStart( $year ) {
1671 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1672 $b = intval( ( $year - 1 ) % 4 );
1673 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1674 if ( $m < 0 ) {
1675 $m--;
1676 }
1677 $Mar = intval( $m );
1678 if ( $m < 0 ) {
1679 $m++;
1680 }
1681 $m -= $Mar;
1682
1683 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1684 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1685 $Mar++;
1686 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1687 $Mar += 2;
1688 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1689 $Mar++;
1690 }
1691
1692 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1693 return $Mar;
1694 }
1695
1696 /**
1697 * Algorithm to convert Gregorian dates to Thai solar dates,
1698 * Minguo dates or Minguo dates.
1699 *
1700 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1701 * http://en.wikipedia.org/wiki/Minguo_calendar
1702 * http://en.wikipedia.org/wiki/Japanese_era_name
1703 *
1704 * @param $ts String: 14-character timestamp
1705 * @param $cName String: calender name
1706 * @return Array: converted year, month, day
1707 */
1708 private static function tsToYear( $ts, $cName ) {
1709 $gy = substr( $ts, 0, 4 );
1710 $gm = substr( $ts, 4, 2 );
1711 $gd = substr( $ts, 6, 2 );
1712
1713 if ( !strcmp( $cName, 'thai' ) ) {
1714 # Thai solar dates
1715 # Add 543 years to the Gregorian calendar
1716 # Months and days are identical
1717 $gy_offset = $gy + 543;
1718 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1719 # Minguo dates
1720 # Deduct 1911 years from the Gregorian calendar
1721 # Months and days are identical
1722 $gy_offset = $gy - 1911;
1723 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1724 # Nengō dates up to Meiji period
1725 # Deduct years from the Gregorian calendar
1726 # depending on the nengo periods
1727 # Months and days are identical
1728 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1729 # Meiji period
1730 $gy_gannen = $gy - 1868 + 1;
1731 $gy_offset = $gy_gannen;
1732 if ( $gy_gannen == 1 ) {
1733 $gy_offset = '元';
1734 }
1735 $gy_offset = '明治' . $gy_offset;
1736 } elseif (
1737 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1738 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1739 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1740 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1741 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1742 )
1743 {
1744 # Taishō period
1745 $gy_gannen = $gy - 1912 + 1;
1746 $gy_offset = $gy_gannen;
1747 if ( $gy_gannen == 1 ) {
1748 $gy_offset = '元';
1749 }
1750 $gy_offset = '大正' . $gy_offset;
1751 } elseif (
1752 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1753 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1754 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1755 )
1756 {
1757 # Shōwa period
1758 $gy_gannen = $gy - 1926 + 1;
1759 $gy_offset = $gy_gannen;
1760 if ( $gy_gannen == 1 ) {
1761 $gy_offset = '元';
1762 }
1763 $gy_offset = '昭和' . $gy_offset;
1764 } else {
1765 # Heisei period
1766 $gy_gannen = $gy - 1989 + 1;
1767 $gy_offset = $gy_gannen;
1768 if ( $gy_gannen == 1 ) {
1769 $gy_offset = '元';
1770 }
1771 $gy_offset = '平成' . $gy_offset;
1772 }
1773 } else {
1774 $gy_offset = $gy;
1775 }
1776
1777 return array( $gy_offset, $gm, $gd );
1778 }
1779
1780 /**
1781 * Roman number formatting up to 10000
1782 *
1783 * @param $num int
1784 *
1785 * @return string
1786 */
1787 static function romanNumeral( $num ) {
1788 static $table = array(
1789 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1790 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1791 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1792 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM', 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1793 );
1794
1795 $num = intval( $num );
1796 if ( $num > 10000 || $num <= 0 ) {
1797 return $num;
1798 }
1799
1800 $s = '';
1801 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1802 if ( $num >= $pow10 ) {
1803 $s .= $table[$i][(int)floor( $num / $pow10 )];
1804 }
1805 $num = $num % $pow10;
1806 }
1807 return $s;
1808 }
1809
1810 /**
1811 * Hebrew Gematria number formatting up to 9999
1812 *
1813 * @param $num int
1814 *
1815 * @return string
1816 */
1817 static function hebrewNumeral( $num ) {
1818 static $table = array(
1819 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1820 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1821 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1822 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1823 );
1824
1825 $num = intval( $num );
1826 if ( $num > 9999 || $num <= 0 ) {
1827 return $num;
1828 }
1829
1830 $s = '';
1831 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1832 if ( $num >= $pow10 ) {
1833 if ( $num == 15 || $num == 16 ) {
1834 $s .= $table[0][9] . $table[0][$num - 9];
1835 $num = 0;
1836 } else {
1837 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1838 if ( $pow10 == 1000 ) {
1839 $s .= "'";
1840 }
1841 }
1842 }
1843 $num = $num % $pow10;
1844 }
1845 if ( strlen( $s ) == 2 ) {
1846 $str = $s . "'";
1847 } else {
1848 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1849 $str .= substr( $s, strlen( $s ) - 2, 2 );
1850 }
1851 $start = substr( $str, 0, strlen( $str ) - 2 );
1852 $end = substr( $str, strlen( $str ) - 2 );
1853 switch ( $end ) {
1854 case 'כ':
1855 $str = $start . 'ך';
1856 break;
1857 case 'מ':
1858 $str = $start . 'ם';
1859 break;
1860 case 'נ':
1861 $str = $start . 'ן';
1862 break;
1863 case 'פ':
1864 $str = $start . 'ף';
1865 break;
1866 case 'צ':
1867 $str = $start . 'ץ';
1868 break;
1869 }
1870 return $str;
1871 }
1872
1873 /**
1874 * Used by date() and time() to adjust the time output.
1875 *
1876 * @param $ts Int the time in date('YmdHis') format
1877 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1878 * get user timecorrection setting)
1879 * @return int
1880 */
1881 function userAdjust( $ts, $tz = false ) {
1882 global $wgUser, $wgLocalTZoffset;
1883
1884 if ( $tz === false ) {
1885 $tz = $wgUser->getOption( 'timecorrection' );
1886 }
1887
1888 $data = explode( '|', $tz, 3 );
1889
1890 if ( $data[0] == 'ZoneInfo' ) {
1891 wfSuppressWarnings();
1892 $userTZ = timezone_open( $data[2] );
1893 wfRestoreWarnings();
1894 if ( $userTZ !== false ) {
1895 $date = date_create( $ts, timezone_open( 'UTC' ) );
1896 date_timezone_set( $date, $userTZ );
1897 $date = date_format( $date, 'YmdHis' );
1898 return $date;
1899 }
1900 # Unrecognized timezone, default to 'Offset' with the stored offset.
1901 $data[0] = 'Offset';
1902 }
1903
1904 $minDiff = 0;
1905 if ( $data[0] == 'System' || $tz == '' ) {
1906 #  Global offset in minutes.
1907 if ( isset( $wgLocalTZoffset ) ) {
1908 $minDiff = $wgLocalTZoffset;
1909 }
1910 } elseif ( $data[0] == 'Offset' ) {
1911 $minDiff = intval( $data[1] );
1912 } else {
1913 $data = explode( ':', $tz );
1914 if ( count( $data ) == 2 ) {
1915 $data[0] = intval( $data[0] );
1916 $data[1] = intval( $data[1] );
1917 $minDiff = abs( $data[0] ) * 60 + $data[1];
1918 if ( $data[0] < 0 ) {
1919 $minDiff = -$minDiff;
1920 }
1921 } else {
1922 $minDiff = intval( $data[0] ) * 60;
1923 }
1924 }
1925
1926 # No difference ? Return time unchanged
1927 if ( 0 == $minDiff ) {
1928 return $ts;
1929 }
1930
1931 wfSuppressWarnings(); // E_STRICT system time bitching
1932 # Generate an adjusted date; take advantage of the fact that mktime
1933 # will normalize out-of-range values so we don't have to split $minDiff
1934 # into hours and minutes.
1935 $t = mktime( (
1936 (int)substr( $ts, 8, 2 ) ), # Hours
1937 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
1938 (int)substr( $ts, 12, 2 ), # Seconds
1939 (int)substr( $ts, 4, 2 ), # Month
1940 (int)substr( $ts, 6, 2 ), # Day
1941 (int)substr( $ts, 0, 4 ) ); # Year
1942
1943 $date = date( 'YmdHis', $t );
1944 wfRestoreWarnings();
1945
1946 return $date;
1947 }
1948
1949 /**
1950 * This is meant to be used by time(), date(), and timeanddate() to get
1951 * the date preference they're supposed to use, it should be used in
1952 * all children.
1953 *
1954 *<code>
1955 * function timeanddate([...], $format = true) {
1956 * $datePreference = $this->dateFormat($format);
1957 * [...]
1958 * }
1959 *</code>
1960 *
1961 * @param $usePrefs Mixed: if true, the user's preference is used
1962 * if false, the site/language default is used
1963 * if int/string, assumed to be a format.
1964 * @return string
1965 */
1966 function dateFormat( $usePrefs = true ) {
1967 global $wgUser;
1968
1969 if ( is_bool( $usePrefs ) ) {
1970 if ( $usePrefs ) {
1971 $datePreference = $wgUser->getDatePreference();
1972 } else {
1973 $datePreference = (string)User::getDefaultOption( 'date' );
1974 }
1975 } else {
1976 $datePreference = (string)$usePrefs;
1977 }
1978
1979 // return int
1980 if ( $datePreference == '' ) {
1981 return 'default';
1982 }
1983
1984 return $datePreference;
1985 }
1986
1987 /**
1988 * Get a format string for a given type and preference
1989 * @param $type string May be date, time or both
1990 * @param $pref string The format name as it appears in Messages*.php
1991 *
1992 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
1993 *
1994 * @return string
1995 */
1996 function getDateFormatString( $type, $pref ) {
1997 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
1998 if ( $pref == 'default' ) {
1999 $pref = $this->getDefaultDateFormat();
2000 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2001 } else {
2002 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2003
2004 if ( $type === 'pretty' && $df === null ) {
2005 $df = $this->getDateFormatString( 'date', $pref );
2006 }
2007
2008 if ( $df === null ) {
2009 $pref = $this->getDefaultDateFormat();
2010 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2011 }
2012 }
2013 $this->dateFormatStrings[$type][$pref] = $df;
2014 }
2015 return $this->dateFormatStrings[$type][$pref];
2016 }
2017
2018 /**
2019 * @param $ts Mixed: the time format which needs to be turned into a
2020 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2021 * @param $adj Bool: whether to adjust the time output according to the
2022 * user configured offset ($timecorrection)
2023 * @param $format Mixed: true to use user's date format preference
2024 * @param $timecorrection String|bool the time offset as returned by
2025 * validateTimeZone() in Special:Preferences
2026 * @return string
2027 */
2028 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2029 $ts = wfTimestamp( TS_MW, $ts );
2030 if ( $adj ) {
2031 $ts = $this->userAdjust( $ts, $timecorrection );
2032 }
2033 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2034 return $this->sprintfDate( $df, $ts );
2035 }
2036
2037 /**
2038 * @param $ts Mixed: the time format which needs to be turned into a
2039 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2040 * @param $adj Bool: whether to adjust the time output according to the
2041 * user configured offset ($timecorrection)
2042 * @param $format Mixed: true to use user's date format preference
2043 * @param $timecorrection String|bool the time offset as returned by
2044 * validateTimeZone() in Special:Preferences
2045 * @return string
2046 */
2047 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2048 $ts = wfTimestamp( TS_MW, $ts );
2049 if ( $adj ) {
2050 $ts = $this->userAdjust( $ts, $timecorrection );
2051 }
2052 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2053 return $this->sprintfDate( $df, $ts );
2054 }
2055
2056 /**
2057 * @param $ts Mixed: the time format which needs to be turned into a
2058 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2059 * @param $adj Bool: whether to adjust the time output according to the
2060 * user configured offset ($timecorrection)
2061 * @param $format Mixed: what format to return, if it's false output the
2062 * default one (default true)
2063 * @param $timecorrection String|bool the time offset as returned by
2064 * validateTimeZone() in Special:Preferences
2065 * @return string
2066 */
2067 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2068 $ts = wfTimestamp( TS_MW, $ts );
2069 if ( $adj ) {
2070 $ts = $this->userAdjust( $ts, $timecorrection );
2071 }
2072 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2073 return $this->sprintfDate( $df, $ts );
2074 }
2075
2076 /**
2077 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2078 *
2079 * @since 1.20
2080 *
2081 * @param integer $seconds The amount of seconds.
2082 * @param array $chosenIntervals The intervals to enable.
2083 *
2084 * @return string
2085 */
2086 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2087 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2088
2089 $segments = array();
2090
2091 foreach ( $intervals as $intervalName => $intervalValue ) {
2092 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2093 $segments[] = $message->inLanguage( $this )->escaped();
2094 }
2095
2096 return $this->listToText( $segments );
2097 }
2098
2099 /**
2100 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2101 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2102 *
2103 * @since 1.20
2104 *
2105 * @param integer $seconds The amount of seconds.
2106 * @param array $chosenIntervals The intervals to enable.
2107 *
2108 * @return array
2109 */
2110 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2111 if ( empty( $chosenIntervals ) ) {
2112 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
2113 }
2114
2115 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2116 $sortedNames = array_keys( $intervals );
2117 $smallestInterval = array_pop( $sortedNames );
2118
2119 $segments = array();
2120
2121 foreach ( $intervals as $name => $length ) {
2122 $value = floor( $seconds / $length );
2123
2124 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2125 $seconds -= $value * $length;
2126 $segments[$name] = $value;
2127 }
2128 }
2129
2130 return $segments;
2131 }
2132
2133 /**
2134 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2135 *
2136 * @param $type String: can be 'date', 'time' or 'both'
2137 * @param $ts Mixed: the time format which needs to be turned into a
2138 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2139 * @param $user User object used to get preferences for timezone and format
2140 * @param $options Array, can contain the following keys:
2141 * - 'timecorrection': time correction, can have the following values:
2142 * - true: use user's preference
2143 * - false: don't use time correction
2144 * - integer: value of time correction in minutes
2145 * - 'format': format to use, can have the following values:
2146 * - true: use user's preference
2147 * - false: use default preference
2148 * - string: format to use
2149 * @since 1.19
2150 * @return String
2151 */
2152 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2153 $ts = wfTimestamp( TS_MW, $ts );
2154 $options += array( 'timecorrection' => true, 'format' => true );
2155 if ( $options['timecorrection'] !== false ) {
2156 if ( $options['timecorrection'] === true ) {
2157 $offset = $user->getOption( 'timecorrection' );
2158 } else {
2159 $offset = $options['timecorrection'];
2160 }
2161 $ts = $this->userAdjust( $ts, $offset );
2162 }
2163 if ( $options['format'] === true ) {
2164 $format = $user->getDatePreference();
2165 } else {
2166 $format = $options['format'];
2167 }
2168 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2169 return $this->sprintfDate( $df, $ts );
2170 }
2171
2172 /**
2173 * Get the formatted date for the given timestamp and formatted for
2174 * the given user.
2175 *
2176 * @param $ts Mixed: the time format which needs to be turned into a
2177 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2178 * @param $user User object used to get preferences for timezone and format
2179 * @param $options Array, can contain the following keys:
2180 * - 'timecorrection': time correction, can have the following values:
2181 * - true: use user's preference
2182 * - false: don't use time correction
2183 * - integer: value of time correction in minutes
2184 * - 'format': format to use, can have the following values:
2185 * - true: use user's preference
2186 * - false: use default preference
2187 * - string: format to use
2188 * @since 1.19
2189 * @return String
2190 */
2191 public function userDate( $ts, User $user, array $options = array() ) {
2192 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2193 }
2194
2195 /**
2196 * Get the formatted time for the given timestamp and formatted for
2197 * the given user.
2198 *
2199 * @param $ts Mixed: the time format which needs to be turned into a
2200 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2201 * @param $user User object used to get preferences for timezone and format
2202 * @param $options Array, can contain the following keys:
2203 * - 'timecorrection': time correction, can have the following values:
2204 * - true: use user's preference
2205 * - false: don't use time correction
2206 * - integer: value of time correction in minutes
2207 * - 'format': format to use, can have the following values:
2208 * - true: use user's preference
2209 * - false: use default preference
2210 * - string: format to use
2211 * @since 1.19
2212 * @return String
2213 */
2214 public function userTime( $ts, User $user, array $options = array() ) {
2215 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2216 }
2217
2218 /**
2219 * Get the formatted date and time for the given timestamp and formatted for
2220 * the given user.
2221 *
2222 * @param $ts Mixed: the time format which needs to be turned into a
2223 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2224 * @param $user User object used to get preferences for timezone and format
2225 * @param $options Array, can contain the following keys:
2226 * - 'timecorrection': time correction, can have the following values:
2227 * - true: use user's preference
2228 * - false: don't use time correction
2229 * - integer: value of time correction in minutes
2230 * - 'format': format to use, can have the following values:
2231 * - true: use user's preference
2232 * - false: use default preference
2233 * - string: format to use
2234 * @since 1.19
2235 * @return String
2236 */
2237 public function userTimeAndDate( $ts, User $user, array $options = array() ) {
2238 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2239 }
2240
2241 /**
2242 * Convert an MWTimestamp into a pretty human-readable timestamp using
2243 * the given user preferences and relative base time.
2244 *
2245 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2246 * on your timestamp object, which will then call this function. Calling
2247 * this function directly will cause hooks to be skipped over.
2248 *
2249 * @see MWTimestamp::getHumanTimestamp
2250 * @param MWTimestamp $ts Timestamp to prettify
2251 * @param MWTimestamp $relativeTo Base timestamp
2252 * @param User $user User preferences to use
2253 * @return string Human timestamp
2254 * @since 1.22
2255 */
2256 public function getHumanTimestamp( MWTimestamp $ts, MWTimestamp $relativeTo, User $user ) {
2257 $diff = $ts->diff( $relativeTo );
2258 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) - (int)$relativeTo->timestamp->format( 'w' ) );
2259 $days = $diff->days ?: (int)$diffDay;
2260 if ( $diff->invert || $days > 5 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' ) ) {
2261 // Timestamps are in different years: use full timestamp
2262 // Also do full timestamp for future dates
2263 /**
2264 * @FIXME Add better handling of future timestamps.
2265 */
2266 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2267 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2268 } elseif ( $days > 5 ) {
2269 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2270 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2271 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2272 } elseif ( $days > 1 ) {
2273 // Timestamp within the past week: show the day of the week and time
2274 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2275 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2276 $ts = wfMessage( "$weekday-at" )
2277 ->inLanguage( $this )
2278 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2279 ->text();
2280 } elseif ( $days == 1 ) {
2281 // Timestamp was yesterday: say 'yesterday' and the time.
2282 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2283 $ts = wfMessage( 'yesterday-at' )
2284 ->inLanguage( $this )
2285 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2286 ->text();
2287 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2288 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2289 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2290 $ts = wfMessage( 'today-at' )
2291 ->inLanguage( $this )
2292 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2293 ->text();
2294
2295 // From here on in, the timestamp was soon enough ago so that we can simply say
2296 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2297 } elseif ( $diff->h == 1 ) {
2298 // Less than 90 minutes, but more than an hour ago.
2299 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2300 } elseif ( $diff->i >= 1 ) {
2301 // A few minutes ago.
2302 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2303 } elseif ( $diff->s >= 30 ) {
2304 // Less than a minute, but more than 30 sec ago.
2305 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2306 } else {
2307 // Less than 30 seconds ago.
2308 $ts = wfMessage( 'just-now' )->text();
2309 }
2310
2311 return $ts;
2312 }
2313
2314 /**
2315 * @param $key string
2316 * @return array|null
2317 */
2318 function getMessage( $key ) {
2319 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2320 }
2321
2322 /**
2323 * @return array
2324 */
2325 function getAllMessages() {
2326 return self::$dataCache->getItem( $this->mCode, 'messages' );
2327 }
2328
2329 /**
2330 * @param $in
2331 * @param $out
2332 * @param $string
2333 * @return string
2334 */
2335 function iconv( $in, $out, $string ) {
2336 # This is a wrapper for iconv in all languages except esperanto,
2337 # which does some nasty x-conversions beforehand
2338
2339 # Even with //IGNORE iconv can whine about illegal characters in
2340 # *input* string. We just ignore those too.
2341 # REF: http://bugs.php.net/bug.php?id=37166
2342 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2343 wfSuppressWarnings();
2344 $text = iconv( $in, $out . '//IGNORE', $string );
2345 wfRestoreWarnings();
2346 return $text;
2347 }
2348
2349 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2350
2351 /**
2352 * @param $matches array
2353 * @return mixed|string
2354 */
2355 function ucwordbreaksCallbackAscii( $matches ) {
2356 return $this->ucfirst( $matches[1] );
2357 }
2358
2359 /**
2360 * @param $matches array
2361 * @return string
2362 */
2363 function ucwordbreaksCallbackMB( $matches ) {
2364 return mb_strtoupper( $matches[0] );
2365 }
2366
2367 /**
2368 * @param $matches array
2369 * @return string
2370 */
2371 function ucCallback( $matches ) {
2372 list( $wikiUpperChars ) = self::getCaseMaps();
2373 return strtr( $matches[1], $wikiUpperChars );
2374 }
2375
2376 /**
2377 * @param $matches array
2378 * @return string
2379 */
2380 function lcCallback( $matches ) {
2381 list( , $wikiLowerChars ) = self::getCaseMaps();
2382 return strtr( $matches[1], $wikiLowerChars );
2383 }
2384
2385 /**
2386 * @param $matches array
2387 * @return string
2388 */
2389 function ucwordsCallbackMB( $matches ) {
2390 return mb_strtoupper( $matches[0] );
2391 }
2392
2393 /**
2394 * @param $matches array
2395 * @return string
2396 */
2397 function ucwordsCallbackWiki( $matches ) {
2398 list( $wikiUpperChars ) = self::getCaseMaps();
2399 return strtr( $matches[0], $wikiUpperChars );
2400 }
2401
2402 /**
2403 * Make a string's first character uppercase
2404 *
2405 * @param $str string
2406 *
2407 * @return string
2408 */
2409 function ucfirst( $str ) {
2410 $o = ord( $str );
2411 if ( $o < 96 ) { // if already uppercase...
2412 return $str;
2413 } elseif ( $o < 128 ) {
2414 return ucfirst( $str ); // use PHP's ucfirst()
2415 } else {
2416 // fall back to more complex logic in case of multibyte strings
2417 return $this->uc( $str, true );
2418 }
2419 }
2420
2421 /**
2422 * Convert a string to uppercase
2423 *
2424 * @param $str string
2425 * @param $first bool
2426 *
2427 * @return string
2428 */
2429 function uc( $str, $first = false ) {
2430 if ( function_exists( 'mb_strtoupper' ) ) {
2431 if ( $first ) {
2432 if ( $this->isMultibyte( $str ) ) {
2433 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2434 } else {
2435 return ucfirst( $str );
2436 }
2437 } else {
2438 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2439 }
2440 } else {
2441 if ( $this->isMultibyte( $str ) ) {
2442 $x = $first ? '^' : '';
2443 return preg_replace_callback(
2444 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2445 array( $this, 'ucCallback' ),
2446 $str
2447 );
2448 } else {
2449 return $first ? ucfirst( $str ) : strtoupper( $str );
2450 }
2451 }
2452 }
2453
2454 /**
2455 * @param $str string
2456 * @return mixed|string
2457 */
2458 function lcfirst( $str ) {
2459 $o = ord( $str );
2460 if ( !$o ) {
2461 return strval( $str );
2462 } elseif ( $o >= 128 ) {
2463 return $this->lc( $str, true );
2464 } elseif ( $o > 96 ) {
2465 return $str;
2466 } else {
2467 $str[0] = strtolower( $str[0] );
2468 return $str;
2469 }
2470 }
2471
2472 /**
2473 * @param $str string
2474 * @param $first bool
2475 * @return mixed|string
2476 */
2477 function lc( $str, $first = false ) {
2478 if ( function_exists( 'mb_strtolower' ) ) {
2479 if ( $first ) {
2480 if ( $this->isMultibyte( $str ) ) {
2481 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2482 } else {
2483 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2484 }
2485 } else {
2486 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2487 }
2488 } else {
2489 if ( $this->isMultibyte( $str ) ) {
2490 $x = $first ? '^' : '';
2491 return preg_replace_callback(
2492 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2493 array( $this, 'lcCallback' ),
2494 $str
2495 );
2496 } else {
2497 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2498 }
2499 }
2500 }
2501
2502 /**
2503 * @param $str string
2504 * @return bool
2505 */
2506 function isMultibyte( $str ) {
2507 return (bool)preg_match( '/[\x80-\xff]/', $str );
2508 }
2509
2510 /**
2511 * @param $str string
2512 * @return mixed|string
2513 */
2514 function ucwords( $str ) {
2515 if ( $this->isMultibyte( $str ) ) {
2516 $str = $this->lc( $str );
2517
2518 // regexp to find first letter in each word (i.e. after each space)
2519 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2520
2521 // function to use to capitalize a single char
2522 if ( function_exists( 'mb_strtoupper' ) ) {
2523 return preg_replace_callback(
2524 $replaceRegexp,
2525 array( $this, 'ucwordsCallbackMB' ),
2526 $str
2527 );
2528 } else {
2529 return preg_replace_callback(
2530 $replaceRegexp,
2531 array( $this, 'ucwordsCallbackWiki' ),
2532 $str
2533 );
2534 }
2535 } else {
2536 return ucwords( strtolower( $str ) );
2537 }
2538 }
2539
2540 /**
2541 * capitalize words at word breaks
2542 *
2543 * @param $str string
2544 * @return mixed
2545 */
2546 function ucwordbreaks( $str ) {
2547 if ( $this->isMultibyte( $str ) ) {
2548 $str = $this->lc( $str );
2549
2550 // since \b doesn't work for UTF-8, we explicitely define word break chars
2551 $breaks = "[ \-\(\)\}\{\.,\?!]";
2552
2553 // find first letter after word break
2554 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2555
2556 if ( function_exists( 'mb_strtoupper' ) ) {
2557 return preg_replace_callback(
2558 $replaceRegexp,
2559 array( $this, 'ucwordbreaksCallbackMB' ),
2560 $str
2561 );
2562 } else {
2563 return preg_replace_callback(
2564 $replaceRegexp,
2565 array( $this, 'ucwordsCallbackWiki' ),
2566 $str
2567 );
2568 }
2569 } else {
2570 return preg_replace_callback(
2571 '/\b([\w\x80-\xff]+)\b/',
2572 array( $this, 'ucwordbreaksCallbackAscii' ),
2573 $str
2574 );
2575 }
2576 }
2577
2578 /**
2579 * Return a case-folded representation of $s
2580 *
2581 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2582 * and $s2 are the same except for the case of their characters. It is not
2583 * necessary for the value returned to make sense when displayed.
2584 *
2585 * Do *not* perform any other normalisation in this function. If a caller
2586 * uses this function when it should be using a more general normalisation
2587 * function, then fix the caller.
2588 *
2589 * @param $s string
2590 *
2591 * @return string
2592 */
2593 function caseFold( $s ) {
2594 return $this->uc( $s );
2595 }
2596
2597 /**
2598 * @param $s string
2599 * @return string
2600 */
2601 function checkTitleEncoding( $s ) {
2602 if ( is_array( $s ) ) {
2603 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2604 }
2605 if ( StringUtils::isUtf8( $s ) ) {
2606 return $s;
2607 }
2608
2609 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2610 }
2611
2612 /**
2613 * @return array
2614 */
2615 function fallback8bitEncoding() {
2616 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2617 }
2618
2619 /**
2620 * Most writing systems use whitespace to break up words.
2621 * Some languages such as Chinese don't conventionally do this,
2622 * which requires special handling when breaking up words for
2623 * searching etc.
2624 *
2625 * @return bool
2626 */
2627 function hasWordBreaks() {
2628 return true;
2629 }
2630
2631 /**
2632 * Some languages such as Chinese require word segmentation,
2633 * Specify such segmentation when overridden in derived class.
2634 *
2635 * @param $string String
2636 * @return String
2637 */
2638 function segmentByWord( $string ) {
2639 return $string;
2640 }
2641
2642 /**
2643 * Some languages have special punctuation need to be normalized.
2644 * Make such changes here.
2645 *
2646 * @param $string String
2647 * @return String
2648 */
2649 function normalizeForSearch( $string ) {
2650 return self::convertDoubleWidth( $string );
2651 }
2652
2653 /**
2654 * convert double-width roman characters to single-width.
2655 * range: ff00-ff5f ~= 0020-007f
2656 *
2657 * @param $string string
2658 *
2659 * @return string
2660 */
2661 protected static function convertDoubleWidth( $string ) {
2662 static $full = null;
2663 static $half = null;
2664
2665 if ( $full === null ) {
2666 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2667 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2668 $full = str_split( $fullWidth, 3 );
2669 $half = str_split( $halfWidth );
2670 }
2671
2672 $string = str_replace( $full, $half, $string );
2673 return $string;
2674 }
2675
2676 /**
2677 * @param $string string
2678 * @param $pattern string
2679 * @return string
2680 */
2681 protected static function insertSpace( $string, $pattern ) {
2682 $string = preg_replace( $pattern, " $1 ", $string );
2683 $string = preg_replace( '/ +/', ' ', $string );
2684 return $string;
2685 }
2686
2687 /**
2688 * @param $termsArray array
2689 * @return array
2690 */
2691 function convertForSearchResult( $termsArray ) {
2692 # some languages, e.g. Chinese, need to do a conversion
2693 # in order for search results to be displayed correctly
2694 return $termsArray;
2695 }
2696
2697 /**
2698 * Get the first character of a string.
2699 *
2700 * @param $s string
2701 * @return string
2702 */
2703 function firstChar( $s ) {
2704 $matches = array();
2705 preg_match(
2706 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2707 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2708 $s,
2709 $matches
2710 );
2711
2712 if ( isset( $matches[1] ) ) {
2713 if ( strlen( $matches[1] ) != 3 ) {
2714 return $matches[1];
2715 }
2716
2717 // Break down Hangul syllables to grab the first jamo
2718 $code = utf8ToCodepoint( $matches[1] );
2719 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2720 return $matches[1];
2721 } elseif ( $code < 0xb098 ) {
2722 return "\xe3\x84\xb1";
2723 } elseif ( $code < 0xb2e4 ) {
2724 return "\xe3\x84\xb4";
2725 } elseif ( $code < 0xb77c ) {
2726 return "\xe3\x84\xb7";
2727 } elseif ( $code < 0xb9c8 ) {
2728 return "\xe3\x84\xb9";
2729 } elseif ( $code < 0xbc14 ) {
2730 return "\xe3\x85\x81";
2731 } elseif ( $code < 0xc0ac ) {
2732 return "\xe3\x85\x82";
2733 } elseif ( $code < 0xc544 ) {
2734 return "\xe3\x85\x85";
2735 } elseif ( $code < 0xc790 ) {
2736 return "\xe3\x85\x87";
2737 } elseif ( $code < 0xcc28 ) {
2738 return "\xe3\x85\x88";
2739 } elseif ( $code < 0xce74 ) {
2740 return "\xe3\x85\x8a";
2741 } elseif ( $code < 0xd0c0 ) {
2742 return "\xe3\x85\x8b";
2743 } elseif ( $code < 0xd30c ) {
2744 return "\xe3\x85\x8c";
2745 } elseif ( $code < 0xd558 ) {
2746 return "\xe3\x85\x8d";
2747 } else {
2748 return "\xe3\x85\x8e";
2749 }
2750 } else {
2751 return '';
2752 }
2753 }
2754
2755 function initEncoding() {
2756 # Some languages may have an alternate char encoding option
2757 # (Esperanto X-coding, Japanese furigana conversion, etc)
2758 # If this language is used as the primary content language,
2759 # an override to the defaults can be set here on startup.
2760 }
2761
2762 /**
2763 * @param $s string
2764 * @return string
2765 */
2766 function recodeForEdit( $s ) {
2767 # For some languages we'll want to explicitly specify
2768 # which characters make it into the edit box raw
2769 # or are converted in some way or another.
2770 global $wgEditEncoding;
2771 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
2772 return $s;
2773 } else {
2774 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2775 }
2776 }
2777
2778 /**
2779 * @param $s string
2780 * @return string
2781 */
2782 function recodeInput( $s ) {
2783 # Take the previous into account.
2784 global $wgEditEncoding;
2785 if ( $wgEditEncoding != '' ) {
2786 $enc = $wgEditEncoding;
2787 } else {
2788 $enc = 'UTF-8';
2789 }
2790 if ( $enc == 'UTF-8' ) {
2791 return $s;
2792 } else {
2793 return $this->iconv( $enc, 'UTF-8', $s );
2794 }
2795 }
2796
2797 /**
2798 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2799 * also cleans up certain backwards-compatible sequences, converting them
2800 * to the modern Unicode equivalent.
2801 *
2802 * This is language-specific for performance reasons only.
2803 *
2804 * @param $s string
2805 *
2806 * @return string
2807 */
2808 function normalize( $s ) {
2809 global $wgAllUnicodeFixes;
2810 $s = UtfNormal::cleanUp( $s );
2811 if ( $wgAllUnicodeFixes ) {
2812 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2813 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2814 }
2815
2816 return $s;
2817 }
2818
2819 /**
2820 * Transform a string using serialized data stored in the given file (which
2821 * must be in the serialized subdirectory of $IP). The file contains pairs
2822 * mapping source characters to destination characters.
2823 *
2824 * The data is cached in process memory. This will go faster if you have the
2825 * FastStringSearch extension.
2826 *
2827 * @param $file string
2828 * @param $string string
2829 *
2830 * @throws MWException
2831 * @return string
2832 */
2833 function transformUsingPairFile( $file, $string ) {
2834 if ( !isset( $this->transformData[$file] ) ) {
2835 $data = wfGetPrecompiledData( $file );
2836 if ( $data === false ) {
2837 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2838 }
2839 $this->transformData[$file] = new ReplacementArray( $data );
2840 }
2841 return $this->transformData[$file]->replace( $string );
2842 }
2843
2844 /**
2845 * For right-to-left language support
2846 *
2847 * @return bool
2848 */
2849 function isRTL() {
2850 return self::$dataCache->getItem( $this->mCode, 'rtl' );
2851 }
2852
2853 /**
2854 * Return the correct HTML 'dir' attribute value for this language.
2855 * @return String
2856 */
2857 function getDir() {
2858 return $this->isRTL() ? 'rtl' : 'ltr';
2859 }
2860
2861 /**
2862 * Return 'left' or 'right' as appropriate alignment for line-start
2863 * for this language's text direction.
2864 *
2865 * Should be equivalent to CSS3 'start' text-align value....
2866 *
2867 * @return String
2868 */
2869 function alignStart() {
2870 return $this->isRTL() ? 'right' : 'left';
2871 }
2872
2873 /**
2874 * Return 'right' or 'left' as appropriate alignment for line-end
2875 * for this language's text direction.
2876 *
2877 * Should be equivalent to CSS3 'end' text-align value....
2878 *
2879 * @return String
2880 */
2881 function alignEnd() {
2882 return $this->isRTL() ? 'left' : 'right';
2883 }
2884
2885 /**
2886 * A hidden direction mark (LRM or RLM), depending on the language direction.
2887 * Unlike getDirMark(), this function returns the character as an HTML entity.
2888 * This function should be used when the output is guaranteed to be HTML,
2889 * because it makes the output HTML source code more readable. When
2890 * the output is plain text or can be escaped, getDirMark() should be used.
2891 *
2892 * @param $opposite Boolean Get the direction mark opposite to your language
2893 * @return string
2894 * @since 1.20
2895 */
2896 function getDirMarkEntity( $opposite = false ) {
2897 if ( $opposite ) {
2898 return $this->isRTL() ? '&lrm;' : '&rlm;';
2899 }
2900 return $this->isRTL() ? '&rlm;' : '&lrm;';
2901 }
2902
2903 /**
2904 * A hidden direction mark (LRM or RLM), depending on the language direction.
2905 * This function produces them as invisible Unicode characters and
2906 * the output may be hard to read and debug, so it should only be used
2907 * when the output is plain text or can be escaped. When the output is
2908 * HTML, use getDirMarkEntity() instead.
2909 *
2910 * @param $opposite Boolean Get the direction mark opposite to your language
2911 * @return string
2912 */
2913 function getDirMark( $opposite = false ) {
2914 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2915 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2916 if ( $opposite ) {
2917 return $this->isRTL() ? $lrm : $rlm;
2918 }
2919 return $this->isRTL() ? $rlm : $lrm;
2920 }
2921
2922 /**
2923 * @return array
2924 */
2925 function capitalizeAllNouns() {
2926 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2927 }
2928
2929 /**
2930 * An arrow, depending on the language direction.
2931 *
2932 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2933 * @return string
2934 */
2935 function getArrow( $direction = 'forwards' ) {
2936 switch ( $direction ) {
2937 case 'forwards':
2938 return $this->isRTL() ? '←' : '→';
2939 case 'backwards':
2940 return $this->isRTL() ? '→' : '←';
2941 case 'left':
2942 return '←';
2943 case 'right':
2944 return '→';
2945 case 'up':
2946 return '↑';
2947 case 'down':
2948 return '↓';
2949 }
2950 }
2951
2952 /**
2953 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2954 *
2955 * @return bool
2956 */
2957 function linkPrefixExtension() {
2958 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2959 }
2960
2961 /**
2962 * Get all magic words from cache.
2963 * @return array
2964 */
2965 function getMagicWords() {
2966 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2967 }
2968
2969 /**
2970 * Run the LanguageGetMagic hook once.
2971 */
2972 protected function doMagicHook() {
2973 if ( $this->mMagicHookDone ) {
2974 return;
2975 }
2976 $this->mMagicHookDone = true;
2977 wfProfileIn( 'LanguageGetMagic' );
2978 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2979 wfProfileOut( 'LanguageGetMagic' );
2980 }
2981
2982 /**
2983 * Fill a MagicWord object with data from here
2984 *
2985 * @param $mw
2986 */
2987 function getMagic( $mw ) {
2988 // Saves a function call
2989 if ( ! $this->mMagicHookDone ) {
2990 $this->doMagicHook();
2991 }
2992
2993 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
2994 $rawEntry = $this->mMagicExtensions[$mw->mId];
2995 } else {
2996 $rawEntry = self::$dataCache->getSubitem(
2997 $this->mCode, 'magicWords', $mw->mId );
2998 }
2999
3000 if ( !is_array( $rawEntry ) ) {
3001 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3002 } else {
3003 $mw->mCaseSensitive = $rawEntry[0];
3004 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3005 }
3006 }
3007
3008 /**
3009 * Add magic words to the extension array
3010 *
3011 * @param $newWords array
3012 */
3013 function addMagicWordsByLang( $newWords ) {
3014 $fallbackChain = $this->getFallbackLanguages();
3015 $fallbackChain = array_reverse( $fallbackChain );
3016 foreach ( $fallbackChain as $code ) {
3017 if ( isset( $newWords[$code] ) ) {
3018 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3019 }
3020 }
3021 }
3022
3023 /**
3024 * Get special page names, as an associative array
3025 * case folded alias => real name
3026 */
3027 function getSpecialPageAliases() {
3028 // Cache aliases because it may be slow to load them
3029 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3030 // Initialise array
3031 $this->mExtendedSpecialPageAliases =
3032 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3033 wfRunHooks( 'LanguageGetSpecialPageAliases',
3034 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
3035 }
3036
3037 return $this->mExtendedSpecialPageAliases;
3038 }
3039
3040 /**
3041 * Italic is unsuitable for some languages
3042 *
3043 * @param $text String: the text to be emphasized.
3044 * @return string
3045 */
3046 function emphasize( $text ) {
3047 return "<em>$text</em>";
3048 }
3049
3050 /**
3051 * Normally we output all numbers in plain en_US style, that is
3052 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3053 * point twohundredthirtyfive. However this is not suitable for all
3054 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
3055 * Icelandic just want to use commas instead of dots, and dots instead
3056 * of commas like "293.291,235".
3057 *
3058 * An example of this function being called:
3059 * <code>
3060 * wfMessage( 'message' )->numParams( $num )->text()
3061 * </code>
3062 *
3063 * See LanguageGu.php for the Gujarati implementation and
3064 * $separatorTransformTable on MessageIs.php for
3065 * the , => . and . => , implementation.
3066 *
3067 * @todo check if it's viable to use localeconv() for the decimal
3068 * separator thing.
3069 * @param $number Mixed: the string to be formatted, should be an integer
3070 * or a floating point number.
3071 * @param $nocommafy Bool: set to true for special numbers like dates
3072 * @return string
3073 */
3074 public function formatNum( $number, $nocommafy = false ) {
3075 global $wgTranslateNumerals;
3076 if ( !$nocommafy ) {
3077 $number = $this->commafy( $number );
3078 $s = $this->separatorTransformTable();
3079 if ( $s ) {
3080 $number = strtr( $number, $s );
3081 }
3082 }
3083
3084 if ( $wgTranslateNumerals ) {
3085 $s = $this->digitTransformTable();
3086 if ( $s ) {
3087 $number = strtr( $number, $s );
3088 }
3089 }
3090
3091 return $number;
3092 }
3093
3094 /**
3095 * Front-end for non-commafied formatNum
3096 *
3097 * @param mixed $number the string to be formatted, should be an integer
3098 * or a floating point number.
3099 * @since 1.21
3100 * @return string
3101 */
3102 public function formatNumNoSeparators( $number ) {
3103 return $this->formatNum( $number, true );
3104 }
3105
3106 /**
3107 * @param $number string
3108 * @return string
3109 */
3110 function parseFormattedNumber( $number ) {
3111 $s = $this->digitTransformTable();
3112 if ( $s ) {
3113 $number = strtr( $number, array_flip( $s ) );
3114 }
3115
3116 $s = $this->separatorTransformTable();
3117 if ( $s ) {
3118 $number = strtr( $number, array_flip( $s ) );
3119 }
3120
3121 $number = strtr( $number, array( ',' => '' ) );
3122 return $number;
3123 }
3124
3125 /**
3126 * Adds commas to a given number
3127 * @since 1.19
3128 * @param $number mixed
3129 * @return string
3130 */
3131 function commafy( $number ) {
3132 $digitGroupingPattern = $this->digitGroupingPattern();
3133 if ( $number === null ) {
3134 return '';
3135 }
3136
3137 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3138 // default grouping is at thousands, use the same for ###,###,### pattern too.
3139 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3140 } else {
3141 // Ref: http://cldr.unicode.org/translation/number-patterns
3142 $sign = "";
3143 if ( intval( $number ) < 0 ) {
3144 // For negative numbers apply the algorithm like positive number and add sign.
3145 $sign = "-";
3146 $number = substr( $number, 1 );
3147 }
3148 $integerPart = array();
3149 $decimalPart = array();
3150 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3151 preg_match( "/\d+/", $number, $integerPart );
3152 preg_match( "/\.\d*/", $number, $decimalPart );
3153 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3154 if ( $groupedNumber === $number ) {
3155 // the string does not have any number part. Eg: .12345
3156 return $sign . $groupedNumber;
3157 }
3158 $start = $end = strlen( $integerPart[0] );
3159 while ( $start > 0 ) {
3160 $match = $matches[0][$numMatches - 1];
3161 $matchLen = strlen( $match );
3162 $start = $end - $matchLen;
3163 if ( $start < 0 ) {
3164 $start = 0;
3165 }
3166 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3167 $end = $start;
3168 if ( $numMatches > 1 ) {
3169 // use the last pattern for the rest of the number
3170 $numMatches--;
3171 }
3172 if ( $start > 0 ) {
3173 $groupedNumber = "," . $groupedNumber;
3174 }
3175 }
3176 return $sign . $groupedNumber;
3177 }
3178 }
3179
3180 /**
3181 * @return String
3182 */
3183 function digitGroupingPattern() {
3184 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3185 }
3186
3187 /**
3188 * @return array
3189 */
3190 function digitTransformTable() {
3191 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3192 }
3193
3194 /**
3195 * @return array
3196 */
3197 function separatorTransformTable() {
3198 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3199 }
3200
3201 /**
3202 * Take a list of strings and build a locale-friendly comma-separated
3203 * list, using the local comma-separator message.
3204 * The last two strings are chained with an "and".
3205 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3206 *
3207 * @param $l Array
3208 * @return string
3209 */
3210 function listToText( array $l ) {
3211 $m = count( $l ) - 1;
3212 if ( $m < 0 ) {
3213 return '';
3214 }
3215 if ( $m > 0 ) {
3216 $and = $this->getMessageFromDB( 'and' );
3217 $space = $this->getMessageFromDB( 'word-separator' );
3218 if ( $m > 1 ) {
3219 $comma = $this->getMessageFromDB( 'comma-separator' );
3220 }
3221 }
3222 $s = $l[$m];
3223 for ( $i = $m - 1; $i >= 0; $i-- ) {
3224 if ( $i == $m - 1 ) {
3225 $s = $l[$i] . $and . $space . $s;
3226 } else {
3227 $s = $l[$i] . $comma . $s;
3228 }
3229 }
3230 return $s;
3231 }
3232
3233 /**
3234 * Take a list of strings and build a locale-friendly comma-separated
3235 * list, using the local comma-separator message.
3236 * @param $list array of strings to put in a comma list
3237 * @return string
3238 */
3239 function commaList( array $list ) {
3240 return implode(
3241 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3242 $list
3243 );
3244 }
3245
3246 /**
3247 * Take a list of strings and build a locale-friendly semicolon-separated
3248 * list, using the local semicolon-separator message.
3249 * @param $list array of strings to put in a semicolon list
3250 * @return string
3251 */
3252 function semicolonList( array $list ) {
3253 return implode(
3254 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3255 $list
3256 );
3257 }
3258
3259 /**
3260 * Same as commaList, but separate it with the pipe instead.
3261 * @param $list array of strings to put in a pipe list
3262 * @return string
3263 */
3264 function pipeList( array $list ) {
3265 return implode(
3266 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3267 $list
3268 );
3269 }
3270
3271 /**
3272 * Truncate a string to a specified length in bytes, appending an optional
3273 * string (e.g. for ellipses)
3274 *
3275 * The database offers limited byte lengths for some columns in the database;
3276 * multi-byte character sets mean we need to ensure that only whole characters
3277 * are included, otherwise broken characters can be passed to the user
3278 *
3279 * If $length is negative, the string will be truncated from the beginning
3280 *
3281 * @param $string String to truncate
3282 * @param $length Int: maximum length (including ellipses)
3283 * @param $ellipsis String to append to the truncated text
3284 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3285 * $adjustLength was introduced in 1.18, before that behaved as if false.
3286 * @return string
3287 */
3288 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3289 # Use the localized ellipsis character
3290 if ( $ellipsis == '...' ) {
3291 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3292 }
3293 # Check if there is no need to truncate
3294 if ( $length == 0 ) {
3295 return $ellipsis; // convention
3296 } elseif ( strlen( $string ) <= abs( $length ) ) {
3297 return $string; // no need to truncate
3298 }
3299 $stringOriginal = $string;
3300 # If ellipsis length is >= $length then we can't apply $adjustLength
3301 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3302 $string = $ellipsis; // this can be slightly unexpected
3303 # Otherwise, truncate and add ellipsis...
3304 } else {
3305 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3306 if ( $length > 0 ) {
3307 $length -= $eLength;
3308 $string = substr( $string, 0, $length ); // xyz...
3309 $string = $this->removeBadCharLast( $string );
3310 $string = $string . $ellipsis;
3311 } else {
3312 $length += $eLength;
3313 $string = substr( $string, $length ); // ...xyz
3314 $string = $this->removeBadCharFirst( $string );
3315 $string = $ellipsis . $string;
3316 }
3317 }
3318 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3319 # This check is *not* redundant if $adjustLength, due to the single case where
3320 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3321 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3322 return $string;
3323 } else {
3324 return $stringOriginal;
3325 }
3326 }
3327
3328 /**
3329 * Remove bytes that represent an incomplete Unicode character
3330 * at the end of string (e.g. bytes of the char are missing)
3331 *
3332 * @param $string String
3333 * @return string
3334 */
3335 protected function removeBadCharLast( $string ) {
3336 if ( $string != '' ) {
3337 $char = ord( $string[strlen( $string ) - 1] );
3338 $m = array();
3339 if ( $char >= 0xc0 ) {
3340 # We got the first byte only of a multibyte char; remove it.
3341 $string = substr( $string, 0, -1 );
3342 } elseif ( $char >= 0x80 &&
3343 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3344 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3345 ) {
3346 # We chopped in the middle of a character; remove it
3347 $string = $m[1];
3348 }
3349 }
3350 return $string;
3351 }
3352
3353 /**
3354 * Remove bytes that represent an incomplete Unicode character
3355 * at the start of string (e.g. bytes of the char are missing)
3356 *
3357 * @param $string String
3358 * @return string
3359 */
3360 protected function removeBadCharFirst( $string ) {
3361 if ( $string != '' ) {
3362 $char = ord( $string[0] );
3363 if ( $char >= 0x80 && $char < 0xc0 ) {
3364 # We chopped in the middle of a character; remove the whole thing
3365 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3366 }
3367 }
3368 return $string;
3369 }
3370
3371 /**
3372 * Truncate a string of valid HTML to a specified length in bytes,
3373 * appending an optional string (e.g. for ellipses), and return valid HTML
3374 *
3375 * This is only intended for styled/linked text, such as HTML with
3376 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3377 * Also, this will not detect things like "display:none" CSS.
3378 *
3379 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3380 *
3381 * @param string $text HTML string to truncate
3382 * @param int $length (zero/positive) Maximum length (including ellipses)
3383 * @param string $ellipsis String to append to the truncated text
3384 * @return string
3385 */
3386 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3387 # Use the localized ellipsis character
3388 if ( $ellipsis == '...' ) {
3389 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3390 }
3391 # Check if there is clearly no need to truncate
3392 if ( $length <= 0 ) {
3393 return $ellipsis; // no text shown, nothing to format (convention)
3394 } elseif ( strlen( $text ) <= $length ) {
3395 return $text; // string short enough even *with* HTML (short-circuit)
3396 }
3397
3398 $dispLen = 0; // innerHTML legth so far
3399 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3400 $tagType = 0; // 0-open, 1-close
3401 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3402 $entityState = 0; // 0-not entity, 1-entity
3403 $tag = $ret = ''; // accumulated tag name, accumulated result string
3404 $openTags = array(); // open tag stack
3405 $maybeState = null; // possible truncation state
3406
3407 $textLen = strlen( $text );
3408 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3409 for ( $pos = 0; true; ++$pos ) {
3410 # Consider truncation once the display length has reached the maximim.
3411 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3412 # Check that we're not in the middle of a bracket/entity...
3413 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3414 if ( !$testingEllipsis ) {
3415 $testingEllipsis = true;
3416 # Save where we are; we will truncate here unless there turn out to
3417 # be so few remaining characters that truncation is not necessary.
3418 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3419 $maybeState = array( $ret, $openTags ); // save state
3420 }
3421 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3422 # String in fact does need truncation, the truncation point was OK.
3423 list( $ret, $openTags ) = $maybeState; // reload state
3424 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3425 $ret .= $ellipsis; // add ellipsis
3426 break;
3427 }
3428 }
3429 if ( $pos >= $textLen ) {
3430 break; // extra iteration just for above checks
3431 }
3432
3433 # Read the next char...
3434 $ch = $text[$pos];
3435 $lastCh = $pos ? $text[$pos - 1] : '';
3436 $ret .= $ch; // add to result string
3437 if ( $ch == '<' ) {
3438 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3439 $entityState = 0; // for bad HTML
3440 $bracketState = 1; // tag started (checking for backslash)
3441 } elseif ( $ch == '>' ) {
3442 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3443 $entityState = 0; // for bad HTML
3444 $bracketState = 0; // out of brackets
3445 } elseif ( $bracketState == 1 ) {
3446 if ( $ch == '/' ) {
3447 $tagType = 1; // close tag (e.g. "</span>")
3448 } else {
3449 $tagType = 0; // open tag (e.g. "<span>")
3450 $tag .= $ch;
3451 }
3452 $bracketState = 2; // building tag name
3453 } elseif ( $bracketState == 2 ) {
3454 if ( $ch != ' ' ) {
3455 $tag .= $ch;
3456 } else {
3457 // Name found (e.g. "<a href=..."), add on tag attributes...
3458 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3459 }
3460 } elseif ( $bracketState == 0 ) {
3461 if ( $entityState ) {
3462 if ( $ch == ';' ) {
3463 $entityState = 0;
3464 $dispLen++; // entity is one displayed char
3465 }
3466 } else {
3467 if ( $neLength == 0 && !$maybeState ) {
3468 // Save state without $ch. We want to *hit* the first
3469 // display char (to get tags) but not *use* it if truncating.
3470 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3471 }
3472 if ( $ch == '&' ) {
3473 $entityState = 1; // entity found, (e.g. "&#160;")
3474 } else {
3475 $dispLen++; // this char is displayed
3476 // Add the next $max display text chars after this in one swoop...
3477 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3478 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3479 $dispLen += $skipped;
3480 $pos += $skipped;
3481 }
3482 }
3483 }
3484 }
3485 // Close the last tag if left unclosed by bad HTML
3486 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3487 while ( count( $openTags ) > 0 ) {
3488 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3489 }
3490 return $ret;
3491 }
3492
3493 /**
3494 * truncateHtml() helper function
3495 * like strcspn() but adds the skipped chars to $ret
3496 *
3497 * @param $ret
3498 * @param $text
3499 * @param $search
3500 * @param $start
3501 * @param $len
3502 * @return int
3503 */
3504 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3505 if ( $len === null ) {
3506 $len = -1; // -1 means "no limit" for strcspn
3507 } elseif ( $len < 0 ) {
3508 $len = 0; // sanity
3509 }
3510 $skipCount = 0;
3511 if ( $start < strlen( $text ) ) {
3512 $skipCount = strcspn( $text, $search, $start, $len );
3513 $ret .= substr( $text, $start, $skipCount );
3514 }
3515 return $skipCount;
3516 }
3517
3518 /**
3519 * truncateHtml() helper function
3520 * (a) push or pop $tag from $openTags as needed
3521 * (b) clear $tag value
3522 * @param &$tag string Current HTML tag name we are looking at
3523 * @param $tagType int (0-open tag, 1-close tag)
3524 * @param $lastCh string Character before the '>' that ended this tag
3525 * @param &$openTags array Open tag stack (not accounting for $tag)
3526 */
3527 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3528 $tag = ltrim( $tag );
3529 if ( $tag != '' ) {
3530 if ( $tagType == 0 && $lastCh != '/' ) {
3531 $openTags[] = $tag; // tag opened (didn't close itself)
3532 } elseif ( $tagType == 1 ) {
3533 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3534 array_pop( $openTags ); // tag closed
3535 }
3536 }
3537 $tag = '';
3538 }
3539 }
3540
3541 /**
3542 * Grammatical transformations, needed for inflected languages
3543 * Invoked by putting {{grammar:case|word}} in a message
3544 *
3545 * @param $word string
3546 * @param $case string
3547 * @return string
3548 */
3549 function convertGrammar( $word, $case ) {
3550 global $wgGrammarForms;
3551 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3552 return $wgGrammarForms[$this->getCode()][$case][$word];
3553 }
3554 return $word;
3555 }
3556 /**
3557 * Get the grammar forms for the content language
3558 * @return array of grammar forms
3559 * @since 1.20
3560 */
3561 function getGrammarForms() {
3562 global $wgGrammarForms;
3563 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3564 return $wgGrammarForms[$this->getCode()];
3565 }
3566 return array();
3567 }
3568 /**
3569 * Provides an alternative text depending on specified gender.
3570 * Usage {{gender:username|masculine|feminine|neutral}}.
3571 * username is optional, in which case the gender of current user is used,
3572 * but only in (some) interface messages; otherwise default gender is used.
3573 *
3574 * If no forms are given, an empty string is returned. If only one form is
3575 * given, it will be returned unconditionally. These details are implied by
3576 * the caller and cannot be overridden in subclasses.
3577 *
3578 * If more than one form is given, the default is to use the neutral one
3579 * if it is specified, and to use the masculine one otherwise. These
3580 * details can be overridden in subclasses.
3581 *
3582 * @param $gender string
3583 * @param $forms array
3584 *
3585 * @return string
3586 */
3587 function gender( $gender, $forms ) {
3588 if ( !count( $forms ) ) {
3589 return '';
3590 }
3591 $forms = $this->preConvertPlural( $forms, 2 );
3592 if ( $gender === 'male' ) {
3593 return $forms[0];
3594 }
3595 if ( $gender === 'female' ) {
3596 return $forms[1];
3597 }
3598 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3599 }
3600
3601 /**
3602 * Plural form transformations, needed for some languages.
3603 * For example, there are 3 form of plural in Russian and Polish,
3604 * depending on "count mod 10". See [[w:Plural]]
3605 * For English it is pretty simple.
3606 *
3607 * Invoked by putting {{plural:count|wordform1|wordform2}}
3608 * or {{plural:count|wordform1|wordform2|wordform3}}
3609 *
3610 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3611 *
3612 * @param $count Integer: non-localized number
3613 * @param $forms Array: different plural forms
3614 * @return string Correct form of plural for $count in this language
3615 */
3616 function convertPlural( $count, $forms ) {
3617 // Handle explicit n=pluralform cases
3618 foreach ( $forms as $index => $form ) {
3619 if ( preg_match( '/\d+=/i', $form ) ) {
3620 $pos = strpos( $form, '=' );
3621 if ( substr( $form, 0, $pos ) === (string) $count ) {
3622 return substr( $form, $pos + 1 );
3623 }
3624 unset( $forms[$index] );
3625 }
3626 }
3627
3628 $forms = array_values( $forms );
3629 if ( !count( $forms ) ) {
3630 return '';
3631 }
3632
3633 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3634 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3635 return $forms[$pluralForm];
3636 }
3637
3638 /**
3639 * Checks that convertPlural was given an array and pads it to requested
3640 * amount of forms by copying the last one.
3641 *
3642 * @param $count Integer: How many forms should there be at least
3643 * @param $forms Array of forms given to convertPlural
3644 * @return array Padded array of forms or an exception if not an array
3645 */
3646 protected function preConvertPlural( /* Array */ $forms, $count ) {
3647 while ( count( $forms ) < $count ) {
3648 $forms[] = $forms[count( $forms ) - 1];
3649 }
3650 return $forms;
3651 }
3652
3653 /**
3654 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3655 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3656 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3657 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3658 * match up with it.
3659 *
3660 * @param $str String: the validated block duration in English
3661 * @return string Somehow translated block duration
3662 * @see LanguageFi.php for example implementation
3663 */
3664 function translateBlockExpiry( $str ) {
3665 $duration = SpecialBlock::getSuggestedDurations( $this );
3666 foreach ( $duration as $show => $value ) {
3667 if ( strcmp( $str, $value ) == 0 ) {
3668 return htmlspecialchars( trim( $show ) );
3669 }
3670 }
3671
3672 // Since usually only infinite or indefinite is only on list, so try
3673 // equivalents if still here.
3674 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3675 if ( in_array( $str, $indefs ) ) {
3676 foreach ( $indefs as $val ) {
3677 $show = array_search( $val, $duration, true );
3678 if ( $show !== false ) {
3679 return htmlspecialchars( trim( $show ) );
3680 }
3681 }
3682 }
3683
3684 // If all else fails, return a standard duration or timestamp description.
3685 $time = strtotime( $str, 0 );
3686 if ( $time === false ) { // Unknown format. Return it as-is in case.
3687 return $str;
3688 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3689 // $time is relative to 0 so it's a duration length.
3690 return $this->formatDuration( $time );
3691 } else { // It's an absolute timestamp.
3692 if ( $time === 0 ) {
3693 // wfTimestamp() handles 0 as current time instead of epoch.
3694 return $this->timeanddate( '19700101000000' );
3695 } else {
3696 return $this->timeanddate( $time );
3697 }
3698 }
3699 }
3700
3701 /**
3702 * languages like Chinese need to be segmented in order for the diff
3703 * to be of any use
3704 *
3705 * @param $text String
3706 * @return String
3707 */
3708 public function segmentForDiff( $text ) {
3709 return $text;
3710 }
3711
3712 /**
3713 * and unsegment to show the result
3714 *
3715 * @param $text String
3716 * @return String
3717 */
3718 public function unsegmentForDiff( $text ) {
3719 return $text;
3720 }
3721
3722 /**
3723 * Return the LanguageConverter used in the Language
3724 *
3725 * @since 1.19
3726 * @return LanguageConverter
3727 */
3728 public function getConverter() {
3729 return $this->mConverter;
3730 }
3731
3732 /**
3733 * convert text to all supported variants
3734 *
3735 * @param $text string
3736 * @return array
3737 */
3738 public function autoConvertToAllVariants( $text ) {
3739 return $this->mConverter->autoConvertToAllVariants( $text );
3740 }
3741
3742 /**
3743 * convert text to different variants of a language.
3744 *
3745 * @param $text string
3746 * @return string
3747 */
3748 public function convert( $text ) {
3749 return $this->mConverter->convert( $text );
3750 }
3751
3752 /**
3753 * Convert a Title object to a string in the preferred variant
3754 *
3755 * @param $title Title
3756 * @return string
3757 */
3758 public function convertTitle( $title ) {
3759 return $this->mConverter->convertTitle( $title );
3760 }
3761
3762 /**
3763 * Convert a namespace index to a string in the preferred variant
3764 *
3765 * @param $ns int
3766 * @return string
3767 */
3768 public function convertNamespace( $ns ) {
3769 return $this->mConverter->convertNamespace( $ns );
3770 }
3771
3772 /**
3773 * Check if this is a language with variants
3774 *
3775 * @return bool
3776 */
3777 public function hasVariants() {
3778 return count( $this->getVariants() ) > 1;
3779 }
3780
3781 /**
3782 * Check if the language has the specific variant
3783 *
3784 * @since 1.19
3785 * @param $variant string
3786 * @return bool
3787 */
3788 public function hasVariant( $variant ) {
3789 return (bool)$this->mConverter->validateVariant( $variant );
3790 }
3791
3792 /**
3793 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3794 *
3795 * @param $text string
3796 * @return string
3797 */
3798 public function armourMath( $text ) {
3799 return $this->mConverter->armourMath( $text );
3800 }
3801
3802 /**
3803 * Perform output conversion on a string, and encode for safe HTML output.
3804 * @param $text String text to be converted
3805 * @param $isTitle Bool whether this conversion is for the article title
3806 * @return string
3807 * @todo this should get integrated somewhere sane
3808 */
3809 public function convertHtml( $text, $isTitle = false ) {
3810 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3811 }
3812
3813 /**
3814 * @param $key string
3815 * @return string
3816 */
3817 public function convertCategoryKey( $key ) {
3818 return $this->mConverter->convertCategoryKey( $key );
3819 }
3820
3821 /**
3822 * Get the list of variants supported by this language
3823 * see sample implementation in LanguageZh.php
3824 *
3825 * @return array an array of language codes
3826 */
3827 public function getVariants() {
3828 return $this->mConverter->getVariants();
3829 }
3830
3831 /**
3832 * @return string
3833 */
3834 public function getPreferredVariant() {
3835 return $this->mConverter->getPreferredVariant();
3836 }
3837
3838 /**
3839 * @return string
3840 */
3841 public function getDefaultVariant() {
3842 return $this->mConverter->getDefaultVariant();
3843 }
3844
3845 /**
3846 * @return string
3847 */
3848 public function getURLVariant() {
3849 return $this->mConverter->getURLVariant();
3850 }
3851
3852 /**
3853 * If a language supports multiple variants, it is
3854 * possible that non-existing link in one variant
3855 * actually exists in another variant. this function
3856 * tries to find it. See e.g. LanguageZh.php
3857 *
3858 * @param $link String: the name of the link
3859 * @param $nt Mixed: the title object of the link
3860 * @param $ignoreOtherCond Boolean: to disable other conditions when
3861 * we need to transclude a template or update a category's link
3862 * @return null the input parameters may be modified upon return
3863 */
3864 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3865 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
3866 }
3867
3868 /**
3869 * If a language supports multiple variants, converts text
3870 * into an array of all possible variants of the text:
3871 * 'variant' => text in that variant
3872 *
3873 * @deprecated since 1.17 Use autoConvertToAllVariants()
3874 *
3875 * @param $text string
3876 *
3877 * @return string
3878 */
3879 public function convertLinkToAllVariants( $text ) {
3880 return $this->mConverter->convertLinkToAllVariants( $text );
3881 }
3882
3883 /**
3884 * returns language specific options used by User::getPageRenderHash()
3885 * for example, the preferred language variant
3886 *
3887 * @return string
3888 */
3889 function getExtraHashOptions() {
3890 return $this->mConverter->getExtraHashOptions();
3891 }
3892
3893 /**
3894 * For languages that support multiple variants, the title of an
3895 * article may be displayed differently in different variants. this
3896 * function returns the apporiate title defined in the body of the article.
3897 *
3898 * @return string
3899 */
3900 public function getParsedTitle() {
3901 return $this->mConverter->getParsedTitle();
3902 }
3903
3904 /**
3905 * Prepare external link text for conversion. When the text is
3906 * a URL, it shouldn't be converted, and it'll be wrapped in
3907 * the "raw" tag (-{R| }-) to prevent conversion.
3908 *
3909 * This function is called "markNoConversion" for historical
3910 * reasons.
3911 *
3912 * @param $text String: text to be used for external link
3913 * @param $noParse bool: wrap it without confirming it's a real URL first
3914 * @return string the tagged text
3915 */
3916 public function markNoConversion( $text, $noParse = false ) {
3917 // Excluding protocal-relative URLs may avoid many false positives.
3918 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
3919 return $this->mConverter->markNoConversion( $text );
3920 } else {
3921 return $text;
3922 }
3923 }
3924
3925 /**
3926 * A regular expression to match legal word-trailing characters
3927 * which should be merged onto a link of the form [[foo]]bar.
3928 *
3929 * @return string
3930 */
3931 public function linkTrail() {
3932 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
3933 }
3934
3935 /**
3936 * @return Language
3937 */
3938 function getLangObj() {
3939 return $this;
3940 }
3941
3942 /**
3943 * Get the "parent" language which has a converter to convert a "compatible" language
3944 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
3945 *
3946 * @return Language|null
3947 * @since 1.22
3948 */
3949 public function getParentLanguage() {
3950 if ( $this->mParentLanguage !== false ) {
3951 return $this->mParentLanguage;
3952 }
3953
3954 $pieces = explode( '-', $this->getCode() );
3955 $code = $pieces[0];
3956 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
3957 $this->mParentLanguage = null;
3958 return null;
3959 }
3960 $lang = Language::factory( $code );
3961 if ( !$lang->hasVariant( $this->getCode() ) ) {
3962 $this->mParentLanguage = null;
3963 return null;
3964 }
3965
3966 $this->mParentLanguage = $lang;
3967 return $lang;
3968 }
3969
3970 /**
3971 * Get the RFC 3066 code for this language object
3972 *
3973 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3974 * htmlspecialchars() or similar
3975 *
3976 * @return string
3977 */
3978 public function getCode() {
3979 return $this->mCode;
3980 }
3981
3982 /**
3983 * Get the code in Bcp47 format which we can use
3984 * inside of html lang="" tags.
3985 *
3986 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3987 * htmlspecialchars() or similar.
3988 *
3989 * @since 1.19
3990 * @return string
3991 */
3992 public function getHtmlCode() {
3993 if ( is_null( $this->mHtmlCode ) ) {
3994 $this->mHtmlCode = wfBCP47( $this->getCode() );
3995 }
3996 return $this->mHtmlCode;
3997 }
3998
3999 /**
4000 * @param $code string
4001 */
4002 public function setCode( $code ) {
4003 $this->mCode = $code;
4004 // Ensure we don't leave incorrect cached data lying around
4005 $this->mHtmlCode = null;
4006 $this->mParentLanguage = false;
4007 }
4008
4009 /**
4010 * Get the name of a file for a certain language code
4011 * @param $prefix string Prepend this to the filename
4012 * @param $code string Language code
4013 * @param $suffix string Append this to the filename
4014 * @throws MWException
4015 * @return string $prefix . $mangledCode . $suffix
4016 */
4017 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4018 // Protect against path traversal
4019 if ( !Language::isValidCode( $code )
4020 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
4021 {
4022 throw new MWException( "Invalid language code \"$code\"" );
4023 }
4024
4025 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4026 }
4027
4028 /**
4029 * Get the language code from a file name. Inverse of getFileName()
4030 * @param $filename string $prefix . $languageCode . $suffix
4031 * @param $prefix string Prefix before the language code
4032 * @param $suffix string Suffix after the language code
4033 * @return string Language code, or false if $prefix or $suffix isn't found
4034 */
4035 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4036 $m = null;
4037 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4038 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4039 if ( !count( $m ) ) {
4040 return false;
4041 }
4042 return str_replace( '_', '-', strtolower( $m[1] ) );
4043 }
4044
4045 /**
4046 * @param $code string
4047 * @return string
4048 */
4049 public static function getMessagesFileName( $code ) {
4050 global $IP;
4051 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4052 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
4053 return $file;
4054 }
4055
4056 /**
4057 * @param $code string
4058 * @return string
4059 */
4060 public static function getClassFileName( $code ) {
4061 global $IP;
4062 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4063 }
4064
4065 /**
4066 * Get the first fallback for a given language.
4067 *
4068 * @param $code string
4069 *
4070 * @return bool|string
4071 */
4072 public static function getFallbackFor( $code ) {
4073 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4074 return false;
4075 } else {
4076 $fallbacks = self::getFallbacksFor( $code );
4077 $first = array_shift( $fallbacks );
4078 return $first;
4079 }
4080 }
4081
4082 /**
4083 * Get the ordered list of fallback languages.
4084 *
4085 * @since 1.19
4086 * @param $code string Language code
4087 * @return array
4088 */
4089 public static function getFallbacksFor( $code ) {
4090 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4091 return array();
4092 } else {
4093 $v = self::getLocalisationCache()->getItem( $code, 'fallback' );
4094 $v = array_map( 'trim', explode( ',', $v ) );
4095 if ( $v[count( $v ) - 1] !== 'en' ) {
4096 $v[] = 'en';
4097 }
4098 return $v;
4099 }
4100 }
4101
4102 /**
4103 * Get the ordered list of fallback languages, ending with the fallback
4104 * language chain for the site language.
4105 *
4106 * @since 1.22
4107 * @param string $code Language code
4108 * @return array array( fallbacks, site fallbacks )
4109 */
4110 public static function getFallbacksIncludingSiteLanguage( $code ) {
4111 global $wgLanguageCode;
4112
4113 // Usually, we will only store a tiny number of fallback chains, so we
4114 // keep them in static memory.
4115 $cacheKey = "{$code}-{$wgLanguageCode}";
4116
4117 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4118 $fallbacks = self::getFallbacksFor( $code );
4119
4120 // Append the site's fallback chain, including the site language itself
4121 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4122 array_unshift( $siteFallbacks, $wgLanguageCode );
4123
4124 // Eliminate any languages already included in the chain
4125 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4126
4127 self::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4128 }
4129 return self::$fallbackLanguageCache[$cacheKey];
4130 }
4131
4132 /**
4133 * Get all messages for a given language
4134 * WARNING: this may take a long time. If you just need all message *keys*
4135 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4136 *
4137 * @param $code string
4138 *
4139 * @return array
4140 */
4141 public static function getMessagesFor( $code ) {
4142 return self::getLocalisationCache()->getItem( $code, 'messages' );
4143 }
4144
4145 /**
4146 * Get a message for a given language
4147 *
4148 * @param $key string
4149 * @param $code string
4150 *
4151 * @return string
4152 */
4153 public static function getMessageFor( $key, $code ) {
4154 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4155 }
4156
4157 /**
4158 * Get all message keys for a given language. This is a faster alternative to
4159 * array_keys( Language::getMessagesFor( $code ) )
4160 *
4161 * @since 1.19
4162 * @param $code string Language code
4163 * @return array of message keys (strings)
4164 */
4165 public static function getMessageKeysFor( $code ) {
4166 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
4167 }
4168
4169 /**
4170 * @param $talk
4171 * @return mixed
4172 */
4173 function fixVariableInNamespace( $talk ) {
4174 if ( strpos( $talk, '$1' ) === false ) {
4175 return $talk;
4176 }
4177
4178 global $wgMetaNamespace;
4179 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4180
4181 # Allow grammar transformations
4182 # Allowing full message-style parsing would make simple requests
4183 # such as action=raw much more expensive than they need to be.
4184 # This will hopefully cover most cases.
4185 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4186 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4187 return str_replace( ' ', '_', $talk );
4188 }
4189
4190 /**
4191 * @param $m string
4192 * @return string
4193 */
4194 function replaceGrammarInNamespace( $m ) {
4195 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4196 }
4197
4198 /**
4199 * @throws MWException
4200 * @return array
4201 */
4202 static function getCaseMaps() {
4203 static $wikiUpperChars, $wikiLowerChars;
4204 if ( isset( $wikiUpperChars ) ) {
4205 return array( $wikiUpperChars, $wikiLowerChars );
4206 }
4207
4208 wfProfileIn( __METHOD__ );
4209 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4210 if ( $arr === false ) {
4211 throw new MWException(
4212 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4213 }
4214 $wikiUpperChars = $arr['wikiUpperChars'];
4215 $wikiLowerChars = $arr['wikiLowerChars'];
4216 wfProfileOut( __METHOD__ );
4217 return array( $wikiUpperChars, $wikiLowerChars );
4218 }
4219
4220 /**
4221 * Decode an expiry (block, protection, etc) which has come from the DB
4222 *
4223 * @todo FIXME: why are we returnings DBMS-dependent strings???
4224 *
4225 * @param $expiry String: Database expiry String
4226 * @param $format Bool|Int true to process using language functions, or TS_ constant
4227 * to return the expiry in a given timestamp
4228 * @return String
4229 * @since 1.18
4230 */
4231 public function formatExpiry( $expiry, $format = true ) {
4232 static $infinity;
4233 if ( $infinity === null ) {
4234 $infinity = wfGetDB( DB_SLAVE )->getInfinity();
4235 }
4236
4237 if ( $expiry == '' || $expiry == $infinity ) {
4238 return $format === true
4239 ? $this->getMessageFromDB( 'infiniteblock' )
4240 : $infinity;
4241 } else {
4242 return $format === true
4243 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4244 : wfTimestamp( $format, $expiry );
4245 }
4246 }
4247
4248 /**
4249 * @todo Document
4250 * @param $seconds int|float
4251 * @param $format Array Optional
4252 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
4253 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
4254 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
4255 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
4256 * @return string
4257 */
4258 function formatTimePeriod( $seconds, $format = array() ) {
4259 if ( !is_array( $format ) ) {
4260 $format = array( 'avoid' => $format ); // For backwards compatibility
4261 }
4262 if ( !isset( $format['avoid'] ) ) {
4263 $format['avoid'] = false;
4264 }
4265 if ( !isset( $format['noabbrevs' ] ) ) {
4266 $format['noabbrevs'] = false;
4267 }
4268 $secondsMsg = wfMessage(
4269 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4270 $minutesMsg = wfMessage(
4271 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4272 $hoursMsg = wfMessage(
4273 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4274 $daysMsg = wfMessage(
4275 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4276
4277 if ( round( $seconds * 10 ) < 100 ) {
4278 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4279 $s = $secondsMsg->params( $s )->text();
4280 } elseif ( round( $seconds ) < 60 ) {
4281 $s = $this->formatNum( round( $seconds ) );
4282 $s = $secondsMsg->params( $s )->text();
4283 } elseif ( round( $seconds ) < 3600 ) {
4284 $minutes = floor( $seconds / 60 );
4285 $secondsPart = round( fmod( $seconds, 60 ) );
4286 if ( $secondsPart == 60 ) {
4287 $secondsPart = 0;
4288 $minutes++;
4289 }
4290 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4291 $s .= ' ';
4292 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4293 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4294 $hours = floor( $seconds / 3600 );
4295 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4296 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4297 if ( $secondsPart == 60 ) {
4298 $secondsPart = 0;
4299 $minutes++;
4300 }
4301 if ( $minutes == 60 ) {
4302 $minutes = 0;
4303 $hours++;
4304 }
4305 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4306 $s .= ' ';
4307 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4308 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4309 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4310 }
4311 } else {
4312 $days = floor( $seconds / 86400 );
4313 if ( $format['avoid'] === 'avoidminutes' ) {
4314 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4315 if ( $hours == 24 ) {
4316 $hours = 0;
4317 $days++;
4318 }
4319 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4320 $s .= ' ';
4321 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4322 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4323 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4324 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4325 if ( $minutes == 60 ) {
4326 $minutes = 0;
4327 $hours++;
4328 }
4329 if ( $hours == 24 ) {
4330 $hours = 0;
4331 $days++;
4332 }
4333 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4334 $s .= ' ';
4335 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4336 $s .= ' ';
4337 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4338 } else {
4339 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4340 $s .= ' ';
4341 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4342 }
4343 }
4344 return $s;
4345 }
4346
4347 /**
4348 * Format a bitrate for output, using an appropriate
4349 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4350 *
4351 * This use base 1000. For base 1024 use formatSize(), for another base
4352 * see formatComputingNumbers()
4353 *
4354 * @param $bps int
4355 * @return string
4356 */
4357 function formatBitrate( $bps ) {
4358 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4359 }
4360
4361 /**
4362 * @param $size int Size of the unit
4363 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4364 * @param $messageKey string Message key to be uesd
4365 * @return string
4366 */
4367 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4368 if ( $size <= 0 ) {
4369 return str_replace( '$1', $this->formatNum( $size ),
4370 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4371 );
4372 }
4373 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4374 $index = 0;
4375
4376 $maxIndex = count( $sizes ) - 1;
4377 while ( $size >= $boundary && $index < $maxIndex ) {
4378 $index++;
4379 $size /= $boundary;
4380 }
4381
4382 // For small sizes no decimal places necessary
4383 $round = 0;
4384 if ( $index > 1 ) {
4385 // For MB and bigger two decimal places are smarter
4386 $round = 2;
4387 }
4388 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4389
4390 $size = round( $size, $round );
4391 $text = $this->getMessageFromDB( $msg );
4392 return str_replace( '$1', $this->formatNum( $size ), $text );
4393 }
4394
4395 /**
4396 * Format a size in bytes for output, using an appropriate
4397 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4398 *
4399 * This method use base 1024. For base 1000 use formatBitrate(), for
4400 * another base see formatComputingNumbers()
4401 *
4402 * @param $size int Size to format
4403 * @return string Plain text (not HTML)
4404 */
4405 function formatSize( $size ) {
4406 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4407 }
4408
4409 /**
4410 * Make a list item, used by various special pages
4411 *
4412 * @param $page String Page link
4413 * @param $details String Text between brackets
4414 * @param $oppositedm Boolean Add the direction mark opposite to your
4415 * language, to display text properly
4416 * @return String
4417 */
4418 function specialList( $page, $details, $oppositedm = true ) {
4419 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) .
4420 $this->getDirMark();
4421 $details = $details ? $dirmark . $this->getMessageFromDB( 'word-separator' ) .
4422 wfMessage( 'parentheses' )->rawParams( $details )->inLanguage( $this )->escaped() : '';
4423 return $page . $details;
4424 }
4425
4426 /**
4427 * Generate (prev x| next x) (20|50|100...) type links for paging
4428 *
4429 * @param $title Title object to link
4430 * @param $offset Integer offset parameter
4431 * @param $limit Integer limit parameter
4432 * @param $query array|String optional URL query parameter string
4433 * @param $atend Bool optional param for specified if this is the last page
4434 * @return String
4435 */
4436 public function viewPrevNext( Title $title, $offset, $limit, array $query = array(), $atend = false ) {
4437 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4438
4439 # Make 'previous' link
4440 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4441 if ( $offset > 0 ) {
4442 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4443 $query, $prev, 'prevn-title', 'mw-prevlink' );
4444 } else {
4445 $plink = htmlspecialchars( $prev );
4446 }
4447
4448 # Make 'next' link
4449 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4450 if ( $atend ) {
4451 $nlink = htmlspecialchars( $next );
4452 } else {
4453 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4454 $query, $next, 'prevn-title', 'mw-nextlink' );
4455 }
4456
4457 # Make links to set number of items per page
4458 $numLinks = array();
4459 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4460 $numLinks[] = $this->numLink( $title, $offset, $num,
4461 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4462 }
4463
4464 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4465 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4466 }
4467
4468 /**
4469 * Helper function for viewPrevNext() that generates links
4470 *
4471 * @param $title Title object to link
4472 * @param $offset Integer offset parameter
4473 * @param $limit Integer limit parameter
4474 * @param $query Array extra query parameters
4475 * @param $link String text to use for the link; will be escaped
4476 * @param $tooltipMsg String name of the message to use as tooltip
4477 * @param $class String value of the "class" attribute of the link
4478 * @return String HTML fragment
4479 */
4480 private function numLink( Title $title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4481 $query = array( 'limit' => $limit, 'offset' => $offset ) + $query;
4482 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4483 return Html::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4484 'title' => $tooltip, 'class' => $class ), $link );
4485 }
4486
4487 /**
4488 * Get the conversion rule title, if any.
4489 *
4490 * @return string
4491 */
4492 public function getConvRuleTitle() {
4493 return $this->mConverter->getConvRuleTitle();
4494 }
4495
4496 /**
4497 * Get the compiled plural rules for the language
4498 * @since 1.20
4499 * @return array Associative array with plural form, and plural rule as key-value pairs
4500 */
4501 public function getCompiledPluralRules() {
4502 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4503 $fallbacks = Language::getFallbacksFor( $this->mCode );
4504 if ( !$pluralRules ) {
4505 foreach ( $fallbacks as $fallbackCode ) {
4506 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4507 if ( $pluralRules ) {
4508 break;
4509 }
4510 }
4511 }
4512 return $pluralRules;
4513 }
4514
4515 /**
4516 * Get the plural rules for the language
4517 * @since 1.20
4518 * @return array Associative array with plural form number and plural rule as key-value pairs
4519 */
4520 public function getPluralRules() {
4521 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4522 $fallbacks = Language::getFallbacksFor( $this->mCode );
4523 if ( !$pluralRules ) {
4524 foreach ( $fallbacks as $fallbackCode ) {
4525 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4526 if ( $pluralRules ) {
4527 break;
4528 }
4529 }
4530 }
4531 return $pluralRules;
4532 }
4533
4534 /**
4535 * Get the plural rule types for the language
4536 * @since 1.22
4537 * @return array Associative array with plural form number and plural rule type as key-value pairs
4538 */
4539 public function getPluralRuleTypes() {
4540 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4541 $fallbacks = Language::getFallbacksFor( $this->mCode );
4542 if ( !$pluralRuleTypes ) {
4543 foreach ( $fallbacks as $fallbackCode ) {
4544 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4545 if ( $pluralRuleTypes ) {
4546 break;
4547 }
4548 }
4549 }
4550 return $pluralRuleTypes;
4551 }
4552
4553 /**
4554 * Find the index number of the plural rule appropriate for the given number
4555 * @return int The index number of the plural rule
4556 */
4557 public function getPluralRuleIndexNumber( $number ) {
4558 $pluralRules = $this->getCompiledPluralRules();
4559 $form = CLDRPluralRuleEvaluator::evaluateCompiled( $number, $pluralRules );
4560 return $form;
4561 }
4562
4563 /**
4564 * Find the plural rule type appropriate for the given number
4565 * For example, if the language is set to Arabic, getPluralType(5) should
4566 * return 'few'.
4567 * @since 1.22
4568 * @return string The name of the plural rule type, e.g. one, two, few, many
4569 */
4570 public function getPluralRuleType( $number ) {
4571 $index = $this->getPluralRuleIndexNumber( $number );
4572 $pluralRuleTypes = $this->getPluralRuleTypes();
4573 if ( isset( $pluralRuleTypes[$index] ) ) {
4574 return $pluralRuleTypes[$index];
4575 } else {
4576 return 'other';
4577 }
4578 }
4579 }
Wiklou, le Wiki du Wiklou