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