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