dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Merge "action=edit API with section=new would fail if page didn't exist"
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * Internationalisation code.
4 *
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
19 *
20 * @file
21 * @ingroup Language
22 */
23
24 /**
25 * @defgroup Language Language
26 */
27
28 if ( !defined( 'MEDIAWIKI' ) ) {
29 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
30 exit( 1 );
31 }
32
33 # Read language names
34 global $wgLanguageNames;
35 require_once __DIR__ . '/Names.php';
36
37 if ( function_exists( 'mb_strtoupper' ) ) {
38 mb_internal_encoding( 'UTF-8' );
39 }
40
41 /**
42 * a fake language converter
43 *
44 * @ingroup Language
45 */
46 class FakeConverter {
47 /**
48 * @var Language
49 */
50 public $mLang;
51 function __construct( $langobj ) { $this->mLang = $langobj; }
52 function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
53 function convert( $t ) { return $t; }
54 function convertTo( $text, $variant ) { return $text; }
55 function convertTitle( $t ) { return $t->getPrefixedText(); }
56 function convertNamespace( $ns ) { return $this->mLang->getFormattedNsText( $ns ); }
57 function getVariants() { return array( $this->mLang->getCode() ); }
58 function getPreferredVariant() { return $this->mLang->getCode(); }
59 function getDefaultVariant() { return $this->mLang->getCode(); }
60 function getURLVariant() { return ''; }
61 function getConvRuleTitle() { return false; }
62 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
63 function getExtraHashOptions() { return ''; }
64 function getParsedTitle() { return ''; }
65 function markNoConversion( $text, $noParse = false ) { return $text; }
66 function convertCategoryKey( $key ) { return $key; }
67 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
68 function armourMath( $text ) { return $text; }
69 }
70
71 /**
72 * Internationalisation code
73 * @ingroup Language
74 */
75 class Language {
76
77 /**
78 * @var LanguageConverter
79 */
80 public $mConverter;
81
82 public $mVariants, $mCode, $mLoaded = false;
83 public $mMagicExtensions = array(), $mMagicHookDone = false;
84 private $mHtmlCode = null, $mParentLanguage = false;
85
86 public $dateFormatStrings = array();
87 public $mExtendedSpecialPageAliases;
88
89 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
90
91 /**
92 * ReplacementArray object caches
93 */
94 public $transformData = array();
95
96 /**
97 * @var LocalisationCache
98 */
99 static public $dataCache;
100
101 static public $mLangObjCache = array();
102
103 static public $mWeekdayMsgs = array(
104 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
105 'friday', 'saturday'
106 );
107
108 static public $mWeekdayAbbrevMsgs = array(
109 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
110 );
111
112 static public $mMonthMsgs = array(
113 'january', 'february', 'march', 'april', 'may_long', 'june',
114 'july', 'august', 'september', 'october', 'november',
115 'december'
116 );
117 static public $mMonthGenMsgs = array(
118 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
119 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
120 'december-gen'
121 );
122 static public $mMonthAbbrevMsgs = array(
123 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
124 'sep', 'oct', 'nov', 'dec'
125 );
126
127 static public $mIranianCalendarMonthMsgs = array(
128 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
129 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
130 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
131 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
132 );
133
134 static public $mHebrewCalendarMonthMsgs = array(
135 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
136 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
137 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
138 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
139 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
140 );
141
142 static public $mHebrewCalendarMonthGenMsgs = array(
143 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
144 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
145 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
146 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
147 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
148 );
149
150 static public $mHijriCalendarMonthMsgs = array(
151 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
152 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
153 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
154 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
155 );
156
157 /**
158 * @since 1.20
159 * @var array
160 */
161 static public $durationIntervals = array(
162 'millennia' => 31556952000,
163 'centuries' => 3155695200,
164 'decades' => 315569520,
165 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
166 'weeks' => 604800,
167 'days' => 86400,
168 'hours' => 3600,
169 'minutes' => 60,
170 'seconds' => 1,
171 );
172
173 /**
174 * Cache for language fallbacks.
175 * @see Language::getFallbacksIncludingSiteLanguage
176 * @since 1.21
177 * @var array
178 */
179 static private $fallbackLanguageCache = array();
180
181 /**
182 * Get a cached or new language object for a given language code
183 * @param $code String
184 * @return Language
185 */
186 static function factory( $code ) {
187 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
188
189 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
190 $code = $wgDummyLanguageCodes[$code];
191 }
192
193 // get the language object to process
194 $langObj = isset( self::$mLangObjCache[$code] )
195 ? self::$mLangObjCache[$code]
196 : self::newFromCode( $code );
197
198 // merge the language object in to get it up front in the cache
199 self::$mLangObjCache = array_merge( array( $code => $langObj ), self::$mLangObjCache );
200 // get rid of the oldest ones in case we have an overflow
201 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
202
203 return $langObj;
204 }
205
206 /**
207 * Create a language object for a given language code
208 * @param $code String
209 * @throws MWException
210 * @return Language
211 */
212 protected static function newFromCode( $code ) {
213 // Protect against path traversal below
214 if ( !Language::isValidCode( $code )
215 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
216 {
217 throw new MWException( "Invalid language code \"$code\"" );
218 }
219
220 if ( !Language::isValidBuiltInCode( $code ) ) {
221 // It's not possible to customise this code with class files, so
222 // just return a Language object. This is to support uselang= hacks.
223 $lang = new Language;
224 $lang->setCode( $code );
225 return $lang;
226 }
227
228 // Check if there is a language class for the code
229 $class = self::classFromCode( $code );
230 self::preloadLanguageClass( $class );
231 if ( MWInit::classExists( $class ) ) {
232 $lang = new $class;
233 return $lang;
234 }
235
236 // Keep trying the fallback list until we find an existing class
237 $fallbacks = Language::getFallbacksFor( $code );
238 foreach ( $fallbacks as $fallbackCode ) {
239 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
240 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
241 }
242
243 $class = self::classFromCode( $fallbackCode );
244 self::preloadLanguageClass( $class );
245 if ( MWInit::classExists( $class ) ) {
246 $lang = Language::newFromCode( $fallbackCode );
247 $lang->setCode( $code );
248 return $lang;
249 }
250 }
251
252 throw new MWException( "Invalid fallback sequence for language '$code'" );
253 }
254
255 /**
256 * Checks whether any localisation is available for that language tag
257 * in MediaWiki (MessagesXx.php exists).
258 *
259 * @param string $code Language tag (in lower case)
260 * @return bool Whether language is supported
261 * @since 1.21
262 */
263 public static function isSupportedLanguage( $code ) {
264 return $code === strtolower( $code ) && is_readable( self::getMessagesFileName( $code ) );
265 }
266
267 /**
268 * Returns true if a language code string is a well-formed language tag
269 * according to RFC 5646.
270 * This function only checks well-formedness; it doesn't check that
271 * language, script or variant codes actually exist in the repositories.
272 *
273 * Based on regexes by Mark Davis of the Unicode Consortium:
274 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
275 *
276 * @param $code string
277 * @param $lenient boolean Whether to allow '_' as separator. The default is only '-'.
278 *
279 * @return bool
280 * @since 1.21
281 */
282 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
283 $alpha = '[a-z]';
284 $digit = '[0-9]';
285 $alphanum = '[a-z0-9]';
286 $x = 'x'; # private use singleton
287 $singleton = '[a-wy-z]'; # other singleton
288 $s = $lenient ? '[-_]' : '-';
289
290 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
291 $script = "$alpha{4}"; # ISO 15924
292 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
293 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
294 $extension = "$singleton(?:$s$alphanum{2,8})+";
295 $privateUse = "$x(?:$s$alphanum{1,8})+";
296
297 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
298 # Since these are limited, this is safe even later changes to the registry --
299 # the only oddity is that it might change the type of the tag, and thus
300 # the results from the capturing groups.
301 # http://www.iana.org/assignments/language-subtag-registry
302
303 $grandfathered = "en{$s}GB{$s}oed"
304 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
305 . "|no{$s}(?:bok|nyn)"
306 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
307 . "|zh{$s}min{$s}nan";
308
309 $variantList = "$variant(?:$s$variant)*";
310 $extensionList = "$extension(?:$s$extension)*";
311
312 $langtag = "(?:($language)"
313 . "(?:$s$script)?"
314 . "(?:$s$region)?"
315 . "(?:$s$variantList)?"
316 . "(?:$s$extensionList)?"
317 . "(?:$s$privateUse)?)";
318
319 # The final breakdown, with capturing groups for each of these components
320 # The variants, extensions, grandfathered, and private-use may have interior '-'
321
322 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
323
324 return (bool)preg_match( "/$root/", strtolower( $code ) );
325 }
326
327 /**
328 * Returns true if a language code string is of a valid form, whether or
329 * not it exists. This includes codes which are used solely for
330 * customisation via the MediaWiki namespace.
331 *
332 * @param $code string
333 *
334 * @return bool
335 */
336 public static function isValidCode( $code ) {
337 static $cache = array();
338 if ( isset( $cache[$code] ) ) {
339 return $cache[$code];
340 }
341 // People think language codes are html safe, so enforce it.
342 // Ideally we should only allow a-zA-Z0-9-
343 // but, .+ and other chars are often used for {{int:}} hacks
344 // see bugs 37564, 37587, 36938
345 $cache[$code] =
346 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
347 && !preg_match( Title::getTitleInvalidRegex(), $code );
348
349 return $cache[$code];
350 }
351
352 /**
353 * Returns true if a language code is of a valid form for the purposes of
354 * internal customisation of MediaWiki, via Messages*.php.
355 *
356 * @param $code string
357 *
358 * @throws MWException
359 * @since 1.18
360 * @return bool
361 */
362 public static function isValidBuiltInCode( $code ) {
363
364 if ( !is_string( $code ) ) {
365 if ( is_object( $code ) ) {
366 $addmsg = " of class " . get_class( $code );
367 } else {
368 $addmsg = '';
369 }
370 $type = gettype( $code );
371 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
372 }
373
374 return (bool)preg_match( '/^[a-z0-9-]{2,}$/i', $code );
375 }
376
377 /**
378 * Returns true if a language code is an IETF tag known to MediaWiki.
379 *
380 * @param $code string
381 *
382 * @since 1.21
383 * @return bool
384 */
385 public static function isKnownLanguageTag( $tag ) {
386 static $coreLanguageNames;
387
388 // Quick escape for invalid input to avoid exceptions down the line
389 // when code tries to process tags which are not valid at all.
390 if ( !self::isValidBuiltInCode( $tag ) ) {
391 return false;
392 }
393
394 if ( $coreLanguageNames === null ) {
395 include MWInit::compiledPath( 'languages/Names.php' );
396 }
397
398 if ( isset( $coreLanguageNames[$tag] )
399 || self::fetchLanguageName( $tag, $tag ) !== ''
400 ) {
401 return true;
402 }
403
404 return false;
405 }
406
407 /**
408 * @param $code
409 * @return String Name of the language class
410 */
411 public static function classFromCode( $code ) {
412 if ( $code == 'en' ) {
413 return 'Language';
414 } else {
415 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
416 }
417 }
418
419 /**
420 * Includes language class files
421 *
422 * @param $class string Name of the language class
423 */
424 public static function preloadLanguageClass( $class ) {
425 global $IP;
426
427 if ( $class === 'Language' ) {
428 return;
429 }
430
431 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
432 include_once "$IP/languages/classes/$class.php";
433 }
434 }
435
436 /**
437 * Get the LocalisationCache instance
438 *
439 * @return LocalisationCache
440 */
441 public static function getLocalisationCache() {
442 if ( is_null( self::$dataCache ) ) {
443 global $wgLocalisationCacheConf;
444 $class = $wgLocalisationCacheConf['class'];
445 self::$dataCache = new $class( $wgLocalisationCacheConf );
446 }
447 return self::$dataCache;
448 }
449
450 function __construct() {
451 $this->mConverter = new FakeConverter( $this );
452 // Set the code to the name of the descendant
453 if ( get_class( $this ) == 'Language' ) {
454 $this->mCode = 'en';
455 } else {
456 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
457 }
458 self::getLocalisationCache();
459 }
460
461 /**
462 * Reduce memory usage
463 */
464 function __destruct() {
465 foreach ( $this as $name => $value ) {
466 unset( $this->$name );
467 }
468 }
469
470 /**
471 * Hook which will be called if this is the content language.
472 * Descendants can use this to register hook functions or modify globals
473 */
474 function initContLang() { }
475
476 /**
477 * Same as getFallbacksFor for current language.
478 * @return array|bool
479 * @deprecated in 1.19
480 */
481 function getFallbackLanguageCode() {
482 wfDeprecated( __METHOD__, '1.19' );
483 return self::getFallbackFor( $this->mCode );
484 }
485
486 /**
487 * @return array
488 * @since 1.19
489 */
490 function getFallbackLanguages() {
491 return self::getFallbacksFor( $this->mCode );
492 }
493
494 /**
495 * Exports $wgBookstoreListEn
496 * @return array
497 */
498 function getBookstoreList() {
499 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
500 }
501
502 /**
503 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
504 * available in localised form, it will be included in English.
505 *
506 * @return array
507 */
508 public function getNamespaces() {
509 if ( is_null( $this->namespaceNames ) ) {
510 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
511
512 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
513 $validNamespaces = MWNamespace::getCanonicalNamespaces();
514
515 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
516
517 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
518 if ( $wgMetaNamespaceTalk ) {
519 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
520 } else {
521 $talk = $this->namespaceNames[NS_PROJECT_TALK];
522 $this->namespaceNames[NS_PROJECT_TALK] =
523 $this->fixVariableInNamespace( $talk );
524 }
525
526 # Sometimes a language will be localised but not actually exist on this wiki.
527 foreach ( $this->namespaceNames as $key => $text ) {
528 if ( !isset( $validNamespaces[$key] ) ) {
529 unset( $this->namespaceNames[$key] );
530 }
531 }
532
533 # The above mixing may leave namespaces out of canonical order.
534 # Re-order by namespace ID number...
535 ksort( $this->namespaceNames );
536
537 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames ) );
538 }
539 return $this->namespaceNames;
540 }
541
542 /**
543 * Arbitrarily set all of the namespace names at once. Mainly used for testing
544 * @param $namespaces Array of namespaces (id => name)
545 */
546 public function setNamespaces( array $namespaces ) {
547 $this->namespaceNames = $namespaces;
548 $this->mNamespaceIds = null;
549 }
550
551 /**
552 * Resets all of the namespace caches. Mainly used for testing
553 */
554 public function resetNamespaces() {
555 $this->namespaceNames = null;
556 $this->mNamespaceIds = null;
557 $this->namespaceAliases = null;
558 }
559
560 /**
561 * A convenience function that returns the same thing as
562 * getNamespaces() except with the array values changed to ' '
563 * where it found '_', useful for producing output to be displayed
564 * e.g. in <select> forms.
565 *
566 * @return array
567 */
568 function getFormattedNamespaces() {
569 $ns = $this->getNamespaces();
570 foreach ( $ns as $k => $v ) {
571 $ns[$k] = strtr( $v, '_', ' ' );
572 }
573 return $ns;
574 }
575
576 /**
577 * Get a namespace value by key
578 * <code>
579 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
580 * echo $mw_ns; // prints 'MediaWiki'
581 * </code>
582 *
583 * @param $index Int: the array key of the namespace to return
584 * @return mixed, string if the namespace value exists, otherwise false
585 */
586 function getNsText( $index ) {
587 $ns = $this->getNamespaces();
588 return isset( $ns[$index] ) ? $ns[$index] : false;
589 }
590
591 /**
592 * A convenience function that returns the same thing as
593 * getNsText() except with '_' changed to ' ', useful for
594 * producing output.
595 *
596 * <code>
597 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
598 * echo $mw_ns; // prints 'MediaWiki talk'
599 * </code>
600 *
601 * @param int $index The array key of the namespace to return
602 * @return string Namespace name without underscores (empty string if namespace does not exist)
603 */
604 function getFormattedNsText( $index ) {
605 $ns = $this->getNsText( $index );
606 return strtr( $ns, '_', ' ' );
607 }
608
609 /**
610 * Returns gender-dependent namespace alias if available.
611 * @param $index Int: namespace index
612 * @param $gender String: gender key (male, female... )
613 * @return String
614 * @since 1.18
615 */
616 function getGenderNsText( $index, $gender ) {
617 global $wgExtraGenderNamespaces;
618
619 $ns = $wgExtraGenderNamespaces + self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
620 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
621 }
622
623 /**
624 * Whether this language makes distinguishes genders for example in
625 * namespaces.
626 * @return bool
627 * @since 1.18
628 */
629 function needsGenderDistinction() {
630 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
631 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
632 // $wgExtraGenderNamespaces overrides everything
633 return true;
634 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
635 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
636 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
637 return false;
638 } else {
639 // Check what is in i18n files
640 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
641 return count( $aliases ) > 0;
642 }
643 }
644
645 /**
646 * Get a namespace key by value, case insensitive.
647 * Only matches namespace names for the current language, not the
648 * canonical ones defined in Namespace.php.
649 *
650 * @param $text String
651 * @return mixed An integer if $text is a valid value otherwise false
652 */
653 function getLocalNsIndex( $text ) {
654 $lctext = $this->lc( $text );
655 $ids = $this->getNamespaceIds();
656 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
657 }
658
659 /**
660 * @return array
661 */
662 function getNamespaceAliases() {
663 if ( is_null( $this->namespaceAliases ) ) {
664 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
665 if ( !$aliases ) {
666 $aliases = array();
667 } else {
668 foreach ( $aliases as $name => $index ) {
669 if ( $index === NS_PROJECT_TALK ) {
670 unset( $aliases[$name] );
671 $name = $this->fixVariableInNamespace( $name );
672 $aliases[$name] = $index;
673 }
674 }
675 }
676
677 global $wgExtraGenderNamespaces;
678 $genders = $wgExtraGenderNamespaces + (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
679 foreach ( $genders as $index => $forms ) {
680 foreach ( $forms as $alias ) {
681 $aliases[$alias] = $index;
682 }
683 }
684
685 $this->namespaceAliases = $aliases;
686 }
687 return $this->namespaceAliases;
688 }
689
690 /**
691 * @return array
692 */
693 function getNamespaceIds() {
694 if ( is_null( $this->mNamespaceIds ) ) {
695 global $wgNamespaceAliases;
696 # Put namespace names and aliases into a hashtable.
697 # If this is too slow, then we should arrange it so that it is done
698 # before caching. The catch is that at pre-cache time, the above
699 # class-specific fixup hasn't been done.
700 $this->mNamespaceIds = array();
701 foreach ( $this->getNamespaces() as $index => $name ) {
702 $this->mNamespaceIds[$this->lc( $name )] = $index;
703 }
704 foreach ( $this->getNamespaceAliases() as $name => $index ) {
705 $this->mNamespaceIds[$this->lc( $name )] = $index;
706 }
707 if ( $wgNamespaceAliases ) {
708 foreach ( $wgNamespaceAliases as $name => $index ) {
709 $this->mNamespaceIds[$this->lc( $name )] = $index;
710 }
711 }
712 }
713 return $this->mNamespaceIds;
714 }
715
716 /**
717 * Get a namespace key by value, case insensitive. Canonical namespace
718 * names override custom ones defined for the current language.
719 *
720 * @param $text String
721 * @return mixed An integer if $text is a valid value otherwise false
722 */
723 function getNsIndex( $text ) {
724 $lctext = $this->lc( $text );
725 $ns = MWNamespace::getCanonicalIndex( $lctext );
726 if ( $ns !== null ) {
727 return $ns;
728 }
729 $ids = $this->getNamespaceIds();
730 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
731 }
732
733 /**
734 * short names for language variants used for language conversion links.
735 *
736 * @param $code String
737 * @param $usemsg bool Use the "variantname-xyz" message if it exists
738 * @return string
739 */
740 function getVariantname( $code, $usemsg = true ) {
741 $msg = "variantname-$code";
742 if ( $usemsg && wfMessage( $msg )->exists() ) {
743 return $this->getMessageFromDB( $msg );
744 }
745 $name = self::fetchLanguageName( $code );
746 if ( $name ) {
747 return $name; # if it's defined as a language name, show that
748 } else {
749 # otherwise, output the language code
750 return $code;
751 }
752 }
753
754 /**
755 * @param $name string
756 * @return string
757 */
758 function specialPage( $name ) {
759 $aliases = $this->getSpecialPageAliases();
760 if ( isset( $aliases[$name][0] ) ) {
761 $name = $aliases[$name][0];
762 }
763 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
764 }
765
766 /**
767 * @return array
768 */
769 function getDatePreferences() {
770 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
771 }
772
773 /**
774 * @return array
775 */
776 function getDateFormats() {
777 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
778 }
779
780 /**
781 * @return array|string
782 */
783 function getDefaultDateFormat() {
784 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
785 if ( $df === 'dmy or mdy' ) {
786 global $wgAmericanDates;
787 return $wgAmericanDates ? 'mdy' : 'dmy';
788 } else {
789 return $df;
790 }
791 }
792
793 /**
794 * @return array
795 */
796 function getDatePreferenceMigrationMap() {
797 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
798 }
799
800 /**
801 * @param $image
802 * @return array|null
803 */
804 function getImageFile( $image ) {
805 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
806 }
807
808 /**
809 * @return array
810 */
811 function getExtraUserToggles() {
812 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
813 }
814
815 /**
816 * @param $tog
817 * @return string
818 */
819 function getUserToggle( $tog ) {
820 return $this->getMessageFromDB( "tog-$tog" );
821 }
822
823 /**
824 * Get native language names, indexed by code.
825 * Only those defined in MediaWiki, no other data like CLDR.
826 * If $customisedOnly is true, only returns codes with a messages file
827 *
828 * @param $customisedOnly bool
829 *
830 * @return array
831 * @deprecated in 1.20, use fetchLanguageNames()
832 */
833 public static function getLanguageNames( $customisedOnly = false ) {
834 return self::fetchLanguageNames( null, $customisedOnly ? 'mwfile' : 'mw' );
835 }
836
837 /**
838 * Get translated language names. This is done on best effort and
839 * by default this is exactly the same as Language::getLanguageNames.
840 * The CLDR extension provides translated names.
841 * @param $code String Language code.
842 * @return Array language code => language name
843 * @since 1.18.0
844 * @deprecated in 1.20, use fetchLanguageNames()
845 */
846 public static function getTranslatedLanguageNames( $code ) {
847 return self::fetchLanguageNames( $code, 'all' );
848 }
849
850 /**
851 * Get an array of language names, indexed by code.
852 * @param $inLanguage null|string: Code of language in which to return the names
853 * Use null for autonyms (native names)
854 * @param $include string:
855 * 'all' all available languages
856 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
857 * 'mwfile' only if the language is in 'mw' *and* has a message file
858 * @return array: language code => language name
859 * @since 1.20
860 */
861 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
862 global $wgExtraLanguageNames;
863 static $coreLanguageNames;
864
865 if ( $coreLanguageNames === null ) {
866 include MWInit::compiledPath( 'languages/Names.php' );
867 }
868
869 $names = array();
870
871 if ( $inLanguage ) {
872 # TODO: also include when $inLanguage is null, when this code is more efficient
873 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
874 }
875
876 $mwNames = $wgExtraLanguageNames + $coreLanguageNames;
877 foreach ( $mwNames as $mwCode => $mwName ) {
878 # - Prefer own MediaWiki native name when not using the hook
879 # - For other names just add if not added through the hook
880 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
881 $names[$mwCode] = $mwName;
882 }
883 }
884
885 if ( $include === 'all' ) {
886 return $names;
887 }
888
889 $returnMw = array();
890 $coreCodes = array_keys( $mwNames );
891 foreach ( $coreCodes as $coreCode ) {
892 $returnMw[$coreCode] = $names[$coreCode];
893 }
894
895 if ( $include === 'mwfile' ) {
896 $namesMwFile = array();
897 # We do this using a foreach over the codes instead of a directory
898 # loop so that messages files in extensions will work correctly.
899 foreach ( $returnMw as $code => $value ) {
900 if ( is_readable( self::getMessagesFileName( $code ) ) ) {
901 $namesMwFile[$code] = $names[$code];
902 }
903 }
904 return $namesMwFile;
905 }
906 # 'mw' option; default if it's not one of the other two options (all/mwfile)
907 return $returnMw;
908 }
909
910 /**
911 * @param $code string: The code of the language for which to get the name
912 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
913 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
914 * @return string: Language name or empty
915 * @since 1.20
916 */
917 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
918 $array = self::fetchLanguageNames( $inLanguage, $include );
919 return !array_key_exists( $code, $array ) ? '' : $array[$code];
920 }
921
922 /**
923 * Get a message from the MediaWiki namespace.
924 *
925 * @param $msg String: message name
926 * @return string
927 */
928 function getMessageFromDB( $msg ) {
929 return wfMessage( $msg )->inLanguage( $this )->text();
930 }
931
932 /**
933 * Get the native language name of $code.
934 * Only if defined in MediaWiki, no other data like CLDR.
935 * @param $code string
936 * @return string
937 * @deprecated in 1.20, use fetchLanguageName()
938 */
939 function getLanguageName( $code ) {
940 return self::fetchLanguageName( $code );
941 }
942
943 /**
944 * @param $key string
945 * @return string
946 */
947 function getMonthName( $key ) {
948 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
949 }
950
951 /**
952 * @return array
953 */
954 function getMonthNamesArray() {
955 $monthNames = array( '' );
956 for ( $i = 1; $i < 13; $i++ ) {
957 $monthNames[] = $this->getMonthName( $i );
958 }
959 return $monthNames;
960 }
961
962 /**
963 * @param $key string
964 * @return string
965 */
966 function getMonthNameGen( $key ) {
967 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
968 }
969
970 /**
971 * @param $key string
972 * @return string
973 */
974 function getMonthAbbreviation( $key ) {
975 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
976 }
977
978 /**
979 * @return array
980 */
981 function getMonthAbbreviationsArray() {
982 $monthNames = array( '' );
983 for ( $i = 1; $i < 13; $i++ ) {
984 $monthNames[] = $this->getMonthAbbreviation( $i );
985 }
986 return $monthNames;
987 }
988
989 /**
990 * @param $key string
991 * @return string
992 */
993 function getWeekdayName( $key ) {
994 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
995 }
996
997 /**
998 * @param $key string
999 * @return string
1000 */
1001 function getWeekdayAbbreviation( $key ) {
1002 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
1003 }
1004
1005 /**
1006 * @param $key string
1007 * @return string
1008 */
1009 function getIranianCalendarMonthName( $key ) {
1010 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
1011 }
1012
1013 /**
1014 * @param $key string
1015 * @return string
1016 */
1017 function getHebrewCalendarMonthName( $key ) {
1018 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
1019 }
1020
1021 /**
1022 * @param $key string
1023 * @return string
1024 */
1025 function getHebrewCalendarMonthNameGen( $key ) {
1026 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1027 }
1028
1029 /**
1030 * @param $key string
1031 * @return string
1032 */
1033 function getHijriCalendarMonthName( $key ) {
1034 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1035 }
1036
1037 /**
1038 * This is a workalike of PHP's date() function, but with better
1039 * internationalisation, a reduced set of format characters, and a better
1040 * escaping format.
1041 *
1042 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1043 * the PHP manual for definitions. There are a number of extensions, which
1044 * start with "x":
1045 *
1046 * xn Do not translate digits of the next numeric format character
1047 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1048 * xr Use roman numerals for the next numeric format character
1049 * xh Use hebrew numerals for the next numeric format character
1050 * xx Literal x
1051 * xg Genitive month name
1052 *
1053 * xij j (day number) in Iranian calendar
1054 * xiF F (month name) in Iranian calendar
1055 * xin n (month number) in Iranian calendar
1056 * xiy y (two digit year) in Iranian calendar
1057 * xiY Y (full year) in Iranian calendar
1058 *
1059 * xjj j (day number) in Hebrew calendar
1060 * xjF F (month name) in Hebrew calendar
1061 * xjt t (days in month) in Hebrew calendar
1062 * xjx xg (genitive month name) in Hebrew calendar
1063 * xjn n (month number) in Hebrew calendar
1064 * xjY Y (full year) in Hebrew calendar
1065 *
1066 * xmj j (day number) in Hijri calendar
1067 * xmF F (month name) in Hijri calendar
1068 * xmn n (month number) in Hijri calendar
1069 * xmY Y (full year) in Hijri calendar
1070 *
1071 * xkY Y (full year) in Thai solar calendar. Months and days are
1072 * identical to the Gregorian calendar
1073 * xoY Y (full year) in Minguo calendar or Juche year.
1074 * Months and days are identical to the
1075 * Gregorian calendar
1076 * xtY Y (full year) in Japanese nengo. Months and days are
1077 * identical to the Gregorian calendar
1078 *
1079 * Characters enclosed in double quotes will be considered literal (with
1080 * the quotes themselves removed). Unmatched quotes will be considered
1081 * literal quotes. Example:
1082 *
1083 * "The month is" F => The month is January
1084 * i's" => 20'11"
1085 *
1086 * Backslash escaping is also supported.
1087 *
1088 * Input timestamp is assumed to be pre-normalized to the desired local
1089 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1090 * $ts is UTC if $zone is not given.
1091 *
1092 * @param $format String
1093 * @param $ts String: 14-character timestamp
1094 * YYYYMMDDHHMMSS
1095 * 01234567890123
1096 * @param $zone DateTimeZone: Timezone of $ts
1097 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1098 *
1099 * @throws MWException
1100 * @return string
1101 */
1102 function sprintfDate( $format, $ts, DateTimeZone $zone = null ) {
1103 $s = '';
1104 $raw = false;
1105 $roman = false;
1106 $hebrewNum = false;
1107 $dateTimeObj = false;
1108 $rawToggle = false;
1109 $iranian = false;
1110 $hebrew = false;
1111 $hijri = false;
1112 $thai = false;
1113 $minguo = false;
1114 $tenno = false;
1115
1116 if ( strlen( $ts ) !== 14 ) {
1117 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1118 }
1119
1120 if ( !ctype_digit( $ts ) ) {
1121 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1122 }
1123
1124 for ( $p = 0; $p < strlen( $format ); $p++ ) {
1125 $num = false;
1126 $code = $format[$p];
1127 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
1128 $code .= $format[++$p];
1129 }
1130
1131 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
1132 $code .= $format[++$p];
1133 }
1134
1135 switch ( $code ) {
1136 case 'xx':
1137 $s .= 'x';
1138 break;
1139 case 'xn':
1140 $raw = true;
1141 break;
1142 case 'xN':
1143 $rawToggle = !$rawToggle;
1144 break;
1145 case 'xr':
1146 $roman = true;
1147 break;
1148 case 'xh':
1149 $hebrewNum = true;
1150 break;
1151 case 'xg':
1152 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1153 break;
1154 case 'xjx':
1155 if ( !$hebrew ) {
1156 $hebrew = self::tsToHebrew( $ts );
1157 }
1158 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1159 break;
1160 case 'd':
1161 $num = substr( $ts, 6, 2 );
1162 break;
1163 case 'D':
1164 if ( !$dateTimeObj ) {
1165 $dateTimeObj = DateTime::createFromFormat(
1166 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1167 );
1168 }
1169 $s .= $this->getWeekdayAbbreviation( $dateTimeObj->format( 'w' ) + 1 );
1170 break;
1171 case 'j':
1172 $num = intval( substr( $ts, 6, 2 ) );
1173 break;
1174 case 'xij':
1175 if ( !$iranian ) {
1176 $iranian = self::tsToIranian( $ts );
1177 }
1178 $num = $iranian[2];
1179 break;
1180 case 'xmj':
1181 if ( !$hijri ) {
1182 $hijri = self::tsToHijri( $ts );
1183 }
1184 $num = $hijri[2];
1185 break;
1186 case 'xjj':
1187 if ( !$hebrew ) {
1188 $hebrew = self::tsToHebrew( $ts );
1189 }
1190 $num = $hebrew[2];
1191 break;
1192 case 'l':
1193 if ( !$dateTimeObj ) {
1194 $dateTimeObj = DateTime::createFromFormat(
1195 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1196 );
1197 }
1198 $s .= $this->getWeekdayName( $dateTimeObj->format( 'w' ) + 1 );
1199 break;
1200 case 'F':
1201 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1202 break;
1203 case 'xiF':
1204 if ( !$iranian ) {
1205 $iranian = self::tsToIranian( $ts );
1206 }
1207 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1208 break;
1209 case 'xmF':
1210 if ( !$hijri ) {
1211 $hijri = self::tsToHijri( $ts );
1212 }
1213 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1214 break;
1215 case 'xjF':
1216 if ( !$hebrew ) {
1217 $hebrew = self::tsToHebrew( $ts );
1218 }
1219 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1220 break;
1221 case 'm':
1222 $num = substr( $ts, 4, 2 );
1223 break;
1224 case 'M':
1225 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1226 break;
1227 case 'n':
1228 $num = intval( substr( $ts, 4, 2 ) );
1229 break;
1230 case 'xin':
1231 if ( !$iranian ) {
1232 $iranian = self::tsToIranian( $ts );
1233 }
1234 $num = $iranian[1];
1235 break;
1236 case 'xmn':
1237 if ( !$hijri ) {
1238 $hijri = self::tsToHijri ( $ts );
1239 }
1240 $num = $hijri[1];
1241 break;
1242 case 'xjn':
1243 if ( !$hebrew ) {
1244 $hebrew = self::tsToHebrew( $ts );
1245 }
1246 $num = $hebrew[1];
1247 break;
1248 case 'xjt':
1249 if ( !$hebrew ) {
1250 $hebrew = self::tsToHebrew( $ts );
1251 }
1252 $num = $hebrew[3];
1253 break;
1254 case 'Y':
1255 $num = substr( $ts, 0, 4 );
1256 break;
1257 case 'xiY':
1258 if ( !$iranian ) {
1259 $iranian = self::tsToIranian( $ts );
1260 }
1261 $num = $iranian[0];
1262 break;
1263 case 'xmY':
1264 if ( !$hijri ) {
1265 $hijri = self::tsToHijri( $ts );
1266 }
1267 $num = $hijri[0];
1268 break;
1269 case 'xjY':
1270 if ( !$hebrew ) {
1271 $hebrew = self::tsToHebrew( $ts );
1272 }
1273 $num = $hebrew[0];
1274 break;
1275 case 'xkY':
1276 if ( !$thai ) {
1277 $thai = self::tsToYear( $ts, 'thai' );
1278 }
1279 $num = $thai[0];
1280 break;
1281 case 'xoY':
1282 if ( !$minguo ) {
1283 $minguo = self::tsToYear( $ts, 'minguo' );
1284 }
1285 $num = $minguo[0];
1286 break;
1287 case 'xtY':
1288 if ( !$tenno ) {
1289 $tenno = self::tsToYear( $ts, 'tenno' );
1290 }
1291 $num = $tenno[0];
1292 break;
1293 case 'y':
1294 $num = substr( $ts, 2, 2 );
1295 break;
1296 case 'xiy':
1297 if ( !$iranian ) {
1298 $iranian = self::tsToIranian( $ts );
1299 }
1300 $num = substr( $iranian[0], -2 );
1301 break;
1302 case 'a':
1303 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1304 break;
1305 case 'A':
1306 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1307 break;
1308 case 'g':
1309 $h = substr( $ts, 8, 2 );
1310 $num = $h % 12 ? $h % 12 : 12;
1311 break;
1312 case 'G':
1313 $num = intval( substr( $ts, 8, 2 ) );
1314 break;
1315 case 'h':
1316 $h = substr( $ts, 8, 2 );
1317 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1318 break;
1319 case 'H':
1320 $num = substr( $ts, 8, 2 );
1321 break;
1322 case 'i':
1323 $num = substr( $ts, 10, 2 );
1324 break;
1325 case 's':
1326 $num = substr( $ts, 12, 2 );
1327 break;
1328 case 'c':
1329 case 'r':
1330 case 'e':
1331 case 'O':
1332 case 'P':
1333 case 'T':
1334 // Pass through string from $dateTimeObj->format()
1335 if ( !$dateTimeObj ) {
1336 $dateTimeObj = DateTime::createFromFormat(
1337 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1338 );
1339 }
1340 $s .= $dateTimeObj->format( $code );
1341 break;
1342 case 'w':
1343 case 'N':
1344 case 'z':
1345 case 'W':
1346 case 't':
1347 case 'L':
1348 case 'o':
1349 case 'U':
1350 case 'I':
1351 case 'Z':
1352 // Pass through number from $dateTimeObj->format()
1353 if ( !$dateTimeObj ) {
1354 $dateTimeObj = DateTime::createFromFormat(
1355 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1356 );
1357 }
1358 $num = $dateTimeObj->format( $code );
1359 break;
1360 case '\\':
1361 # Backslash escaping
1362 if ( $p < strlen( $format ) - 1 ) {
1363 $s .= $format[++$p];
1364 } else {
1365 $s .= '\\';
1366 }
1367 break;
1368 case '"':
1369 # Quoted literal
1370 if ( $p < strlen( $format ) - 1 ) {
1371 $endQuote = strpos( $format, '"', $p + 1 );
1372 if ( $endQuote === false ) {
1373 # No terminating quote, assume literal "
1374 $s .= '"';
1375 } else {
1376 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1377 $p = $endQuote;
1378 }
1379 } else {
1380 # Quote at end of string, assume literal "
1381 $s .= '"';
1382 }
1383 break;
1384 default:
1385 $s .= $format[$p];
1386 }
1387 if ( $num !== false ) {
1388 if ( $rawToggle || $raw ) {
1389 $s .= $num;
1390 $raw = false;
1391 } elseif ( $roman ) {
1392 $s .= Language::romanNumeral( $num );
1393 $roman = false;
1394 } elseif ( $hebrewNum ) {
1395 $s .= self::hebrewNumeral( $num );
1396 $hebrewNum = false;
1397 } else {
1398 $s .= $this->formatNum( $num, true );
1399 }
1400 }
1401 }
1402 return $s;
1403 }
1404
1405 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1406 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1407
1408 /**
1409 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1410 * Gregorian dates to Iranian dates. Originally written in C, it
1411 * is released under the terms of GNU Lesser General Public
1412 * License. Conversion to PHP was performed by Niklas Laxström.
1413 *
1414 * Link: http://www.farsiweb.info/jalali/jalali.c
1415 *
1416 * @param $ts string
1417 *
1418 * @return string
1419 */
1420 private static function tsToIranian( $ts ) {
1421 $gy = substr( $ts, 0, 4 ) -1600;
1422 $gm = substr( $ts, 4, 2 ) -1;
1423 $gd = substr( $ts, 6, 2 ) -1;
1424
1425 # Days passed from the beginning (including leap years)
1426 $gDayNo = 365 * $gy
1427 + floor( ( $gy + 3 ) / 4 )
1428 - floor( ( $gy + 99 ) / 100 )
1429 + floor( ( $gy + 399 ) / 400 );
1430
1431 // Add days of the past months of this year
1432 for ( $i = 0; $i < $gm; $i++ ) {
1433 $gDayNo += self::$GREG_DAYS[$i];
1434 }
1435
1436 // Leap years
1437 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1438 $gDayNo++;
1439 }
1440
1441 // Days passed in current month
1442 $gDayNo += (int)$gd;
1443
1444 $jDayNo = $gDayNo - 79;
1445
1446 $jNp = floor( $jDayNo / 12053 );
1447 $jDayNo %= 12053;
1448
1449 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1450 $jDayNo %= 1461;
1451
1452 if ( $jDayNo >= 366 ) {
1453 $jy += floor( ( $jDayNo - 1 ) / 365 );
1454 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1455 }
1456
1457 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1458 $jDayNo -= self::$IRANIAN_DAYS[$i];
1459 }
1460
1461 $jm = $i + 1;
1462 $jd = $jDayNo + 1;
1463
1464 return array( $jy, $jm, $jd );
1465 }
1466
1467 /**
1468 * Converting Gregorian dates to Hijri dates.
1469 *
1470 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1471 *
1472 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1473 *
1474 * @param $ts string
1475 *
1476 * @return string
1477 */
1478 private static function tsToHijri( $ts ) {
1479 $year = substr( $ts, 0, 4 );
1480 $month = substr( $ts, 4, 2 );
1481 $day = substr( $ts, 6, 2 );
1482
1483 $zyr = $year;
1484 $zd = $day;
1485 $zm = $month;
1486 $zy = $zyr;
1487
1488 if (
1489 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1490 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1491 )
1492 {
1493 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1494 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1495 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1496 $zd - 32075;
1497 } else {
1498 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1499 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1500 }
1501
1502 $zl = $zjd -1948440 + 10632;
1503 $zn = (int)( ( $zl - 1 ) / 10631 );
1504 $zl = $zl - 10631 * $zn + 354;
1505 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1506 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1507 $zm = (int)( ( 24 * $zl ) / 709 );
1508 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1509 $zy = 30 * $zn + $zj - 30;
1510
1511 return array( $zy, $zm, $zd );
1512 }
1513
1514 /**
1515 * Converting Gregorian dates to Hebrew dates.
1516 *
1517 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1518 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1519 * to translate the relevant functions into PHP and release them under
1520 * GNU GPL.
1521 *
1522 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1523 * and Adar II is 14. In a non-leap year, Adar is 6.
1524 *
1525 * @param $ts string
1526 *
1527 * @return string
1528 */
1529 private static function tsToHebrew( $ts ) {
1530 # Parse date
1531 $year = substr( $ts, 0, 4 );
1532 $month = substr( $ts, 4, 2 );
1533 $day = substr( $ts, 6, 2 );
1534
1535 # Calculate Hebrew year
1536 $hebrewYear = $year + 3760;
1537
1538 # Month number when September = 1, August = 12
1539 $month += 4;
1540 if ( $month > 12 ) {
1541 # Next year
1542 $month -= 12;
1543 $year++;
1544 $hebrewYear++;
1545 }
1546
1547 # Calculate day of year from 1 September
1548 $dayOfYear = $day;
1549 for ( $i = 1; $i < $month; $i++ ) {
1550 if ( $i == 6 ) {
1551 # February
1552 $dayOfYear += 28;
1553 # Check if the year is leap
1554 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1555 $dayOfYear++;
1556 }
1557 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1558 $dayOfYear += 30;
1559 } else {
1560 $dayOfYear += 31;
1561 }
1562 }
1563
1564 # Calculate the start of the Hebrew year
1565 $start = self::hebrewYearStart( $hebrewYear );
1566
1567 # Calculate next year's start
1568 if ( $dayOfYear <= $start ) {
1569 # Day is before the start of the year - it is the previous year
1570 # Next year's start
1571 $nextStart = $start;
1572 # Previous year
1573 $year--;
1574 $hebrewYear--;
1575 # Add days since previous year's 1 September
1576 $dayOfYear += 365;
1577 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1578 # Leap year
1579 $dayOfYear++;
1580 }
1581 # Start of the new (previous) year
1582 $start = self::hebrewYearStart( $hebrewYear );
1583 } else {
1584 # Next year's start
1585 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1586 }
1587
1588 # Calculate Hebrew day of year
1589 $hebrewDayOfYear = $dayOfYear - $start;
1590
1591 # Difference between year's days
1592 $diff = $nextStart - $start;
1593 # Add 12 (or 13 for leap years) days to ignore the difference between
1594 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1595 # difference is only about the year type
1596 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1597 $diff += 13;
1598 } else {
1599 $diff += 12;
1600 }
1601
1602 # Check the year pattern, and is leap year
1603 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1604 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1605 # and non-leap years
1606 $yearPattern = $diff % 30;
1607 # Check if leap year
1608 $isLeap = $diff >= 30;
1609
1610 # Calculate day in the month from number of day in the Hebrew year
1611 # Don't check Adar - if the day is not in Adar, we will stop before;
1612 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1613 $hebrewDay = $hebrewDayOfYear;
1614 $hebrewMonth = 1;
1615 $days = 0;
1616 while ( $hebrewMonth <= 12 ) {
1617 # Calculate days in this month
1618 if ( $isLeap && $hebrewMonth == 6 ) {
1619 # Adar in a leap year
1620 if ( $isLeap ) {
1621 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1622 $days = 30;
1623 if ( $hebrewDay <= $days ) {
1624 # Day in Adar I
1625 $hebrewMonth = 13;
1626 } else {
1627 # Subtract the days of Adar I
1628 $hebrewDay -= $days;
1629 # Try Adar II
1630 $days = 29;
1631 if ( $hebrewDay <= $days ) {
1632 # Day in Adar II
1633 $hebrewMonth = 14;
1634 }
1635 }
1636 }
1637 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1638 # Cheshvan in a complete year (otherwise as the rule below)
1639 $days = 30;
1640 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1641 # Kislev in an incomplete year (otherwise as the rule below)
1642 $days = 29;
1643 } else {
1644 # Odd months have 30 days, even have 29
1645 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1646 }
1647 if ( $hebrewDay <= $days ) {
1648 # In the current month
1649 break;
1650 } else {
1651 # Subtract the days of the current month
1652 $hebrewDay -= $days;
1653 # Try in the next month
1654 $hebrewMonth++;
1655 }
1656 }
1657
1658 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1659 }
1660
1661 /**
1662 * This calculates the Hebrew year start, as days since 1 September.
1663 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1664 * Used for Hebrew date.
1665 *
1666 * @param $year int
1667 *
1668 * @return string
1669 */
1670 private static function hebrewYearStart( $year ) {
1671 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1672 $b = intval( ( $year - 1 ) % 4 );
1673 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1674 if ( $m < 0 ) {
1675 $m--;
1676 }
1677 $Mar = intval( $m );
1678 if ( $m < 0 ) {
1679 $m++;
1680 }
1681 $m -= $Mar;
1682
1683 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1684 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1685 $Mar++;
1686 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1687 $Mar += 2;
1688 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1689 $Mar++;
1690 }
1691
1692 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1693 return $Mar;
1694 }
1695
1696 /**
1697 * Algorithm to convert Gregorian dates to Thai solar dates,
1698 * Minguo dates or Minguo dates.
1699 *
1700 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1701 * http://en.wikipedia.org/wiki/Minguo_calendar
1702 * http://en.wikipedia.org/wiki/Japanese_era_name
1703 *
1704 * @param $ts String: 14-character timestamp
1705 * @param $cName String: calender name
1706 * @return Array: converted year, month, day
1707 */
1708 private static function tsToYear( $ts, $cName ) {
1709 $gy = substr( $ts, 0, 4 );
1710 $gm = substr( $ts, 4, 2 );
1711 $gd = substr( $ts, 6, 2 );
1712
1713 if ( !strcmp( $cName, 'thai' ) ) {
1714 # Thai solar dates
1715 # Add 543 years to the Gregorian calendar
1716 # Months and days are identical
1717 $gy_offset = $gy + 543;
1718 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1719 # Minguo dates
1720 # Deduct 1911 years from the Gregorian calendar
1721 # Months and days are identical
1722 $gy_offset = $gy - 1911;
1723 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1724 # Nengō dates up to Meiji period
1725 # Deduct years from the Gregorian calendar
1726 # depending on the nengo periods
1727 # Months and days are identical
1728 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1729 # Meiji period
1730 $gy_gannen = $gy - 1868 + 1;
1731 $gy_offset = $gy_gannen;
1732 if ( $gy_gannen == 1 ) {
1733 $gy_offset = '元';
1734 }
1735 $gy_offset = '明治' . $gy_offset;
1736 } elseif (
1737 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1738 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1739 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1740 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1741 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1742 )
1743 {
1744 # Taishō period
1745 $gy_gannen = $gy - 1912 + 1;
1746 $gy_offset = $gy_gannen;
1747 if ( $gy_gannen == 1 ) {
1748 $gy_offset = '元';
1749 }
1750 $gy_offset = '大正' . $gy_offset;
1751 } elseif (
1752 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1753 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1754 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1755 )
1756 {
1757 # Shōwa period
1758 $gy_gannen = $gy - 1926 + 1;
1759 $gy_offset = $gy_gannen;
1760 if ( $gy_gannen == 1 ) {
1761 $gy_offset = '元';
1762 }
1763 $gy_offset = '昭和' . $gy_offset;
1764 } else {
1765 # Heisei period
1766 $gy_gannen = $gy - 1989 + 1;
1767 $gy_offset = $gy_gannen;
1768 if ( $gy_gannen == 1 ) {
1769 $gy_offset = '元';
1770 }
1771 $gy_offset = '平成' . $gy_offset;
1772 }
1773 } else {
1774 $gy_offset = $gy;
1775 }
1776
1777 return array( $gy_offset, $gm, $gd );
1778 }
1779
1780 /**
1781 * Roman number formatting up to 10000
1782 *
1783 * @param $num int
1784 *
1785 * @return string
1786 */
1787 static function romanNumeral( $num ) {
1788 static $table = array(
1789 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1790 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1791 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1792 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM', 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1793 );
1794
1795 $num = intval( $num );
1796 if ( $num > 10000 || $num <= 0 ) {
1797 return $num;
1798 }
1799
1800 $s = '';
1801 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1802 if ( $num >= $pow10 ) {
1803 $s .= $table[$i][(int)floor( $num / $pow10 )];
1804 }
1805 $num = $num % $pow10;
1806 }
1807 return $s;
1808 }
1809
1810 /**
1811 * Hebrew Gematria number formatting up to 9999
1812 *
1813 * @param $num int
1814 *
1815 * @return string
1816 */
1817 static function hebrewNumeral( $num ) {
1818 static $table = array(
1819 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1820 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1821 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1822 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1823 );
1824
1825 $num = intval( $num );
1826 if ( $num > 9999 || $num <= 0 ) {
1827 return $num;
1828 }
1829
1830 $s = '';
1831 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1832 if ( $num >= $pow10 ) {
1833 if ( $num == 15 || $num == 16 ) {
1834 $s .= $table[0][9] . $table[0][$num - 9];
1835 $num = 0;
1836 } else {
1837 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1838 if ( $pow10 == 1000 ) {
1839 $s .= "'";
1840 }
1841 }
1842 }
1843 $num = $num % $pow10;
1844 }
1845 if ( strlen( $s ) == 2 ) {
1846 $str = $s . "'";
1847 } else {
1848 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1849 $str .= substr( $s, strlen( $s ) - 2, 2 );
1850 }
1851 $start = substr( $str, 0, strlen( $str ) - 2 );
1852 $end = substr( $str, strlen( $str ) - 2 );
1853 switch ( $end ) {
1854 case 'כ':
1855 $str = $start . 'ך';
1856 break;
1857 case 'מ':
1858 $str = $start . 'ם';
1859 break;
1860 case 'נ':
1861 $str = $start . 'ן';
1862 break;
1863 case 'פ':
1864 $str = $start . 'ף';
1865 break;
1866 case 'צ':
1867 $str = $start . 'ץ';
1868 break;
1869 }
1870 return $str;
1871 }
1872
1873 /**
1874 * Used by date() and time() to adjust the time output.
1875 *
1876 * @param $ts Int the time in date('YmdHis') format
1877 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1878 * get user timecorrection setting)
1879 * @return int
1880 */
1881 function userAdjust( $ts, $tz = false ) {
1882 global $wgUser, $wgLocalTZoffset;
1883
1884 if ( $tz === false ) {
1885 $tz = $wgUser->getOption( 'timecorrection' );
1886 }
1887
1888 $data = explode( '|', $tz, 3 );
1889
1890 if ( $data[0] == 'ZoneInfo' ) {
1891 wfSuppressWarnings();
1892 $userTZ = timezone_open( $data[2] );
1893 wfRestoreWarnings();
1894 if ( $userTZ !== false ) {
1895 $date = date_create( $ts, timezone_open( 'UTC' ) );
1896 date_timezone_set( $date, $userTZ );
1897 $date = date_format( $date, 'YmdHis' );
1898 return $date;
1899 }
1900 # Unrecognized timezone, default to 'Offset' with the stored offset.
1901 $data[0] = 'Offset';
1902 }
1903
1904 $minDiff = 0;
1905 if ( $data[0] == 'System' || $tz == '' ) {
1906 #  Global offset in minutes.
1907 if ( isset( $wgLocalTZoffset ) ) {
1908 $minDiff = $wgLocalTZoffset;
1909 }
1910 } elseif ( $data[0] == 'Offset' ) {
1911 $minDiff = intval( $data[1] );
1912 } else {
1913 $data = explode( ':', $tz );
1914 if ( count( $data ) == 2 ) {
1915 $data[0] = intval( $data[0] );
1916 $data[1] = intval( $data[1] );
1917 $minDiff = abs( $data[0] ) * 60 + $data[1];
1918 if ( $data[0] < 0 ) {
1919 $minDiff = -$minDiff;
1920 }
1921 } else {
1922 $minDiff = intval( $data[0] ) * 60;
1923 }
1924 }
1925
1926 # No difference ? Return time unchanged
1927 if ( 0 == $minDiff ) {
1928 return $ts;
1929 }
1930
1931 wfSuppressWarnings(); // E_STRICT system time bitching
1932 # Generate an adjusted date; take advantage of the fact that mktime
1933 # will normalize out-of-range values so we don't have to split $minDiff
1934 # into hours and minutes.
1935 $t = mktime( (
1936 (int)substr( $ts, 8, 2 ) ), # Hours
1937 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
1938 (int)substr( $ts, 12, 2 ), # Seconds
1939 (int)substr( $ts, 4, 2 ), # Month
1940 (int)substr( $ts, 6, 2 ), # Day
1941 (int)substr( $ts, 0, 4 ) ); # Year
1942
1943 $date = date( 'YmdHis', $t );
1944 wfRestoreWarnings();
1945
1946 return $date;
1947 }
1948
1949 /**
1950 * This is meant to be used by time(), date(), and timeanddate() to get
1951 * the date preference they're supposed to use, it should be used in
1952 * all children.
1953 *
1954 *<code>
1955 * function timeanddate([...], $format = true) {
1956 * $datePreference = $this->dateFormat($format);
1957 * [...]
1958 * }
1959 *</code>
1960 *
1961 * @param $usePrefs Mixed: if true, the user's preference is used
1962 * if false, the site/language default is used
1963 * if int/string, assumed to be a format.
1964 * @return string
1965 */
1966 function dateFormat( $usePrefs = true ) {
1967 global $wgUser;
1968
1969 if ( is_bool( $usePrefs ) ) {
1970 if ( $usePrefs ) {
1971 $datePreference = $wgUser->getDatePreference();
1972 } else {
1973 $datePreference = (string)User::getDefaultOption( 'date' );
1974 }
1975 } else {
1976 $datePreference = (string)$usePrefs;
1977 }
1978
1979 // return int
1980 if ( $datePreference == '' ) {
1981 return 'default';
1982 }
1983
1984 return $datePreference;
1985 }
1986
1987 /**
1988 * Get a format string for a given type and preference
1989 * @param $type string May be date, time or both
1990 * @param $pref string The format name as it appears in Messages*.php
1991 *
1992 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
1993 *
1994 * @return string
1995 */
1996 function getDateFormatString( $type, $pref ) {
1997 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
1998 if ( $pref == 'default' ) {
1999 $pref = $this->getDefaultDateFormat();
2000 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2001 } else {
2002 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2003
2004 if ( $type === 'pretty' && $df === null ) {
2005 $df = $this->getDateFormatString( 'date', $pref );
2006 }
2007
2008 if ( $df === null ) {
2009 $pref = $this->getDefaultDateFormat();
2010 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2011 }
2012 }
2013 $this->dateFormatStrings[$type][$pref] = $df;
2014 }
2015 return $this->dateFormatStrings[$type][$pref];
2016 }
2017
2018 /**
2019 * @param $ts Mixed: the time format which needs to be turned into a
2020 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2021 * @param $adj Bool: whether to adjust the time output according to the
2022 * user configured offset ($timecorrection)
2023 * @param $format Mixed: true to use user's date format preference
2024 * @param $timecorrection String|bool the time offset as returned by
2025 * validateTimeZone() in Special:Preferences
2026 * @return string
2027 */
2028 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2029 $ts = wfTimestamp( TS_MW, $ts );
2030 if ( $adj ) {
2031 $ts = $this->userAdjust( $ts, $timecorrection );
2032 }
2033 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2034 return $this->sprintfDate( $df, $ts );
2035 }
2036
2037 /**
2038 * @param $ts Mixed: the time format which needs to be turned into a
2039 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2040 * @param $adj Bool: whether to adjust the time output according to the
2041 * user configured offset ($timecorrection)
2042 * @param $format Mixed: true to use user's date format preference
2043 * @param $timecorrection String|bool the time offset as returned by
2044 * validateTimeZone() in Special:Preferences
2045 * @return string
2046 */
2047 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2048 $ts = wfTimestamp( TS_MW, $ts );
2049 if ( $adj ) {
2050 $ts = $this->userAdjust( $ts, $timecorrection );
2051 }
2052 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2053 return $this->sprintfDate( $df, $ts );
2054 }
2055
2056 /**
2057 * @param $ts Mixed: the time format which needs to be turned into a
2058 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2059 * @param $adj Bool: whether to adjust the time output according to the
2060 * user configured offset ($timecorrection)
2061 * @param $format Mixed: what format to return, if it's false output the
2062 * default one (default true)
2063 * @param $timecorrection String|bool the time offset as returned by
2064 * validateTimeZone() in Special:Preferences
2065 * @return string
2066 */
2067 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2068 $ts = wfTimestamp( TS_MW, $ts );
2069 if ( $adj ) {
2070 $ts = $this->userAdjust( $ts, $timecorrection );
2071 }
2072 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2073 return $this->sprintfDate( $df, $ts );
2074 }
2075
2076 /**
2077 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2078 *
2079 * @since 1.20
2080 *
2081 * @param integer $seconds The amount of seconds.
2082 * @param array $chosenIntervals The intervals to enable.
2083 *
2084 * @return string
2085 */
2086 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2087 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2088
2089 $segments = array();
2090
2091 foreach ( $intervals as $intervalName => $intervalValue ) {
2092 // Give grep a chance to find the usages:
2093 // duration-centuries, duration-decades, duration-years, duration-days,
2094 // duration-hours, duration-minutes, duration-seconds
2095 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2096 $segments[] = $message->inLanguage( $this )->escaped();
2097 }
2098
2099 return $this->listToText( $segments );
2100 }
2101
2102 /**
2103 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2104 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2105 *
2106 * @since 1.20
2107 *
2108 * @param integer $seconds The amount of seconds.
2109 * @param array $chosenIntervals The intervals to enable.
2110 *
2111 * @return array
2112 */
2113 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2114 if ( empty( $chosenIntervals ) ) {
2115 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
2116 }
2117
2118 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2119 $sortedNames = array_keys( $intervals );
2120 $smallestInterval = array_pop( $sortedNames );
2121
2122 $segments = array();
2123
2124 foreach ( $intervals as $name => $length ) {
2125 $value = floor( $seconds / $length );
2126
2127 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2128 $seconds -= $value * $length;
2129 $segments[$name] = $value;
2130 }
2131 }
2132
2133 return $segments;
2134 }
2135
2136 /**
2137 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2138 *
2139 * @param $type String: can be 'date', 'time' or 'both'
2140 * @param $ts Mixed: the time format which needs to be turned into a
2141 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2142 * @param $user User object used to get preferences for timezone and format
2143 * @param $options Array, can contain the following keys:
2144 * - 'timecorrection': time correction, can have the following values:
2145 * - true: use user's preference
2146 * - false: don't use time correction
2147 * - integer: value of time correction in minutes
2148 * - 'format': format to use, can have the following values:
2149 * - true: use user's preference
2150 * - false: use default preference
2151 * - string: format to use
2152 * @since 1.19
2153 * @return String
2154 */
2155 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2156 $ts = wfTimestamp( TS_MW, $ts );
2157 $options += array( 'timecorrection' => true, 'format' => true );
2158 if ( $options['timecorrection'] !== false ) {
2159 if ( $options['timecorrection'] === true ) {
2160 $offset = $user->getOption( 'timecorrection' );
2161 } else {
2162 $offset = $options['timecorrection'];
2163 }
2164 $ts = $this->userAdjust( $ts, $offset );
2165 }
2166 if ( $options['format'] === true ) {
2167 $format = $user->getDatePreference();
2168 } else {
2169 $format = $options['format'];
2170 }
2171 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2172 return $this->sprintfDate( $df, $ts );
2173 }
2174
2175 /**
2176 * Get the formatted date for the given timestamp and formatted for
2177 * the given user.
2178 *
2179 * @param $ts Mixed: the time format which needs to be turned into a
2180 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2181 * @param $user User object used to get preferences for timezone and format
2182 * @param $options Array, can contain the following keys:
2183 * - 'timecorrection': time correction, can have the following values:
2184 * - true: use user's preference
2185 * - false: don't use time correction
2186 * - integer: value of time correction in minutes
2187 * - 'format': format to use, can have the following values:
2188 * - true: use user's preference
2189 * - false: use default preference
2190 * - string: format to use
2191 * @since 1.19
2192 * @return String
2193 */
2194 public function userDate( $ts, User $user, array $options = array() ) {
2195 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2196 }
2197
2198 /**
2199 * Get the formatted time for the given timestamp and formatted for
2200 * the given user.
2201 *
2202 * @param $ts Mixed: the time format which needs to be turned into a
2203 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2204 * @param $user User object used to get preferences for timezone and format
2205 * @param $options Array, can contain the following keys:
2206 * - 'timecorrection': time correction, can have the following values:
2207 * - true: use user's preference
2208 * - false: don't use time correction
2209 * - integer: value of time correction in minutes
2210 * - 'format': format to use, can have the following values:
2211 * - true: use user's preference
2212 * - false: use default preference
2213 * - string: format to use
2214 * @since 1.19
2215 * @return String
2216 */
2217 public function userTime( $ts, User $user, array $options = array() ) {
2218 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2219 }
2220
2221 /**
2222 * Get the formatted date and time for the given timestamp and formatted for
2223 * the given user.
2224 *
2225 * @param $ts Mixed: the time format which needs to be turned into a
2226 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2227 * @param $user User object used to get preferences for timezone and format
2228 * @param $options Array, can contain the following keys:
2229 * - 'timecorrection': time correction, can have the following values:
2230 * - true: use user's preference
2231 * - false: don't use time correction
2232 * - integer: value of time correction in minutes
2233 * - 'format': format to use, can have the following values:
2234 * - true: use user's preference
2235 * - false: use default preference
2236 * - string: format to use
2237 * @since 1.19
2238 * @return String
2239 */
2240 public function userTimeAndDate( $ts, User $user, array $options = array() ) {
2241 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2242 }
2243
2244 /**
2245 * Convert an MWTimestamp into a pretty human-readable timestamp using
2246 * the given user preferences and relative base time.
2247 *
2248 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2249 * on your timestamp object, which will then call this function. Calling
2250 * this function directly will cause hooks to be skipped over.
2251 *
2252 * @see MWTimestamp::getHumanTimestamp
2253 * @param MWTimestamp $ts Timestamp to prettify
2254 * @param MWTimestamp $relativeTo Base timestamp
2255 * @param User $user User preferences to use
2256 * @return string Human timestamp
2257 * @since 1.22
2258 */
2259 public function getHumanTimestamp( MWTimestamp $ts, MWTimestamp $relativeTo, User $user ) {
2260 $diff = $ts->diff( $relativeTo );
2261 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) - (int)$relativeTo->timestamp->format( 'w' ) );
2262 $days = $diff->days ?: (int)$diffDay;
2263 if ( $diff->invert || $days > 5 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' ) ) {
2264 // Timestamps are in different years: use full timestamp
2265 // Also do full timestamp for future dates
2266 /**
2267 * @FIXME Add better handling of future timestamps.
2268 */
2269 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2270 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2271 } elseif ( $days > 5 ) {
2272 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2273 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2274 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2275 } elseif ( $days > 1 ) {
2276 // Timestamp within the past week: show the day of the week and time
2277 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2278 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2279 $ts = wfMessage( "$weekday-at" )
2280 ->inLanguage( $this )
2281 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2282 ->text();
2283 } elseif ( $days == 1 ) {
2284 // Timestamp was yesterday: say 'yesterday' and the time.
2285 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2286 $ts = wfMessage( 'yesterday-at' )
2287 ->inLanguage( $this )
2288 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2289 ->text();
2290 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2291 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2292 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2293 $ts = wfMessage( 'today-at' )
2294 ->inLanguage( $this )
2295 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2296 ->text();
2297
2298 // From here on in, the timestamp was soon enough ago so that we can simply say
2299 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2300 } elseif ( $diff->h == 1 ) {
2301 // Less than 90 minutes, but more than an hour ago.
2302 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2303 } elseif ( $diff->i >= 1 ) {
2304 // A few minutes ago.
2305 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2306 } elseif ( $diff->s >= 30 ) {
2307 // Less than a minute, but more than 30 sec ago.
2308 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2309 } else {
2310 // Less than 30 seconds ago.
2311 $ts = wfMessage( 'just-now' )->text();
2312 }
2313
2314 return $ts;
2315 }
2316
2317 /**
2318 * @param $key string
2319 * @return array|null
2320 */
2321 function getMessage( $key ) {
2322 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2323 }
2324
2325 /**
2326 * @return array
2327 */
2328 function getAllMessages() {
2329 return self::$dataCache->getItem( $this->mCode, 'messages' );
2330 }
2331
2332 /**
2333 * @param $in
2334 * @param $out
2335 * @param $string
2336 * @return string
2337 */
2338 function iconv( $in, $out, $string ) {
2339 # This is a wrapper for iconv in all languages except esperanto,
2340 # which does some nasty x-conversions beforehand
2341
2342 # Even with //IGNORE iconv can whine about illegal characters in
2343 # *input* string. We just ignore those too.
2344 # REF: http://bugs.php.net/bug.php?id=37166
2345 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2346 wfSuppressWarnings();
2347 $text = iconv( $in, $out . '//IGNORE', $string );
2348 wfRestoreWarnings();
2349 return $text;
2350 }
2351
2352 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2353
2354 /**
2355 * @param $matches array
2356 * @return mixed|string
2357 */
2358 function ucwordbreaksCallbackAscii( $matches ) {
2359 return $this->ucfirst( $matches[1] );
2360 }
2361
2362 /**
2363 * @param $matches array
2364 * @return string
2365 */
2366 function ucwordbreaksCallbackMB( $matches ) {
2367 return mb_strtoupper( $matches[0] );
2368 }
2369
2370 /**
2371 * @param $matches array
2372 * @return string
2373 */
2374 function ucCallback( $matches ) {
2375 list( $wikiUpperChars ) = self::getCaseMaps();
2376 return strtr( $matches[1], $wikiUpperChars );
2377 }
2378
2379 /**
2380 * @param $matches array
2381 * @return string
2382 */
2383 function lcCallback( $matches ) {
2384 list( , $wikiLowerChars ) = self::getCaseMaps();
2385 return strtr( $matches[1], $wikiLowerChars );
2386 }
2387
2388 /**
2389 * @param $matches array
2390 * @return string
2391 */
2392 function ucwordsCallbackMB( $matches ) {
2393 return mb_strtoupper( $matches[0] );
2394 }
2395
2396 /**
2397 * @param $matches array
2398 * @return string
2399 */
2400 function ucwordsCallbackWiki( $matches ) {
2401 list( $wikiUpperChars ) = self::getCaseMaps();
2402 return strtr( $matches[0], $wikiUpperChars );
2403 }
2404
2405 /**
2406 * Make a string's first character uppercase
2407 *
2408 * @param $str string
2409 *
2410 * @return string
2411 */
2412 function ucfirst( $str ) {
2413 $o = ord( $str );
2414 if ( $o < 96 ) { // if already uppercase...
2415 return $str;
2416 } elseif ( $o < 128 ) {
2417 return ucfirst( $str ); // use PHP's ucfirst()
2418 } else {
2419 // fall back to more complex logic in case of multibyte strings
2420 return $this->uc( $str, true );
2421 }
2422 }
2423
2424 /**
2425 * Convert a string to uppercase
2426 *
2427 * @param $str string
2428 * @param $first bool
2429 *
2430 * @return string
2431 */
2432 function uc( $str, $first = false ) {
2433 if ( function_exists( 'mb_strtoupper' ) ) {
2434 if ( $first ) {
2435 if ( $this->isMultibyte( $str ) ) {
2436 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2437 } else {
2438 return ucfirst( $str );
2439 }
2440 } else {
2441 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2442 }
2443 } else {
2444 if ( $this->isMultibyte( $str ) ) {
2445 $x = $first ? '^' : '';
2446 return preg_replace_callback(
2447 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2448 array( $this, 'ucCallback' ),
2449 $str
2450 );
2451 } else {
2452 return $first ? ucfirst( $str ) : strtoupper( $str );
2453 }
2454 }
2455 }
2456
2457 /**
2458 * @param $str string
2459 * @return mixed|string
2460 */
2461 function lcfirst( $str ) {
2462 $o = ord( $str );
2463 if ( !$o ) {
2464 return strval( $str );
2465 } elseif ( $o >= 128 ) {
2466 return $this->lc( $str, true );
2467 } elseif ( $o > 96 ) {
2468 return $str;
2469 } else {
2470 $str[0] = strtolower( $str[0] );
2471 return $str;
2472 }
2473 }
2474
2475 /**
2476 * @param $str string
2477 * @param $first bool
2478 * @return mixed|string
2479 */
2480 function lc( $str, $first = false ) {
2481 if ( function_exists( 'mb_strtolower' ) ) {
2482 if ( $first ) {
2483 if ( $this->isMultibyte( $str ) ) {
2484 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2485 } else {
2486 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2487 }
2488 } else {
2489 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2490 }
2491 } else {
2492 if ( $this->isMultibyte( $str ) ) {
2493 $x = $first ? '^' : '';
2494 return preg_replace_callback(
2495 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2496 array( $this, 'lcCallback' ),
2497 $str
2498 );
2499 } else {
2500 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2501 }
2502 }
2503 }
2504
2505 /**
2506 * @param $str string
2507 * @return bool
2508 */
2509 function isMultibyte( $str ) {
2510 return (bool)preg_match( '/[\x80-\xff]/', $str );
2511 }
2512
2513 /**
2514 * @param $str string
2515 * @return mixed|string
2516 */
2517 function ucwords( $str ) {
2518 if ( $this->isMultibyte( $str ) ) {
2519 $str = $this->lc( $str );
2520
2521 // regexp to find first letter in each word (i.e. after each space)
2522 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2523
2524 // function to use to capitalize a single char
2525 if ( function_exists( 'mb_strtoupper' ) ) {
2526 return preg_replace_callback(
2527 $replaceRegexp,
2528 array( $this, 'ucwordsCallbackMB' ),
2529 $str
2530 );
2531 } else {
2532 return preg_replace_callback(
2533 $replaceRegexp,
2534 array( $this, 'ucwordsCallbackWiki' ),
2535 $str
2536 );
2537 }
2538 } else {
2539 return ucwords( strtolower( $str ) );
2540 }
2541 }
2542
2543 /**
2544 * capitalize words at word breaks
2545 *
2546 * @param $str string
2547 * @return mixed
2548 */
2549 function ucwordbreaks( $str ) {
2550 if ( $this->isMultibyte( $str ) ) {
2551 $str = $this->lc( $str );
2552
2553 // since \b doesn't work for UTF-8, we explicitely define word break chars
2554 $breaks = "[ \-\(\)\}\{\.,\?!]";
2555
2556 // find first letter after word break
2557 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2558
2559 if ( function_exists( 'mb_strtoupper' ) ) {
2560 return preg_replace_callback(
2561 $replaceRegexp,
2562 array( $this, 'ucwordbreaksCallbackMB' ),
2563 $str
2564 );
2565 } else {
2566 return preg_replace_callback(
2567 $replaceRegexp,
2568 array( $this, 'ucwordsCallbackWiki' ),
2569 $str
2570 );
2571 }
2572 } else {
2573 return preg_replace_callback(
2574 '/\b([\w\x80-\xff]+)\b/',
2575 array( $this, 'ucwordbreaksCallbackAscii' ),
2576 $str
2577 );
2578 }
2579 }
2580
2581 /**
2582 * Return a case-folded representation of $s
2583 *
2584 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2585 * and $s2 are the same except for the case of their characters. It is not
2586 * necessary for the value returned to make sense when displayed.
2587 *
2588 * Do *not* perform any other normalisation in this function. If a caller
2589 * uses this function when it should be using a more general normalisation
2590 * function, then fix the caller.
2591 *
2592 * @param $s string
2593 *
2594 * @return string
2595 */
2596 function caseFold( $s ) {
2597 return $this->uc( $s );
2598 }
2599
2600 /**
2601 * @param $s string
2602 * @return string
2603 */
2604 function checkTitleEncoding( $s ) {
2605 if ( is_array( $s ) ) {
2606 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2607 }
2608 if ( StringUtils::isUtf8( $s ) ) {
2609 return $s;
2610 }
2611
2612 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2613 }
2614
2615 /**
2616 * @return array
2617 */
2618 function fallback8bitEncoding() {
2619 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2620 }
2621
2622 /**
2623 * Most writing systems use whitespace to break up words.
2624 * Some languages such as Chinese don't conventionally do this,
2625 * which requires special handling when breaking up words for
2626 * searching etc.
2627 *
2628 * @return bool
2629 */
2630 function hasWordBreaks() {
2631 return true;
2632 }
2633
2634 /**
2635 * Some languages such as Chinese require word segmentation,
2636 * Specify such segmentation when overridden in derived class.
2637 *
2638 * @param $string String
2639 * @return String
2640 */
2641 function segmentByWord( $string ) {
2642 return $string;
2643 }
2644
2645 /**
2646 * Some languages have special punctuation need to be normalized.
2647 * Make such changes here.
2648 *
2649 * @param $string String
2650 * @return String
2651 */
2652 function normalizeForSearch( $string ) {
2653 return self::convertDoubleWidth( $string );
2654 }
2655
2656 /**
2657 * convert double-width roman characters to single-width.
2658 * range: ff00-ff5f ~= 0020-007f
2659 *
2660 * @param $string string
2661 *
2662 * @return string
2663 */
2664 protected static function convertDoubleWidth( $string ) {
2665 static $full = null;
2666 static $half = null;
2667
2668 if ( $full === null ) {
2669 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2670 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2671 $full = str_split( $fullWidth, 3 );
2672 $half = str_split( $halfWidth );
2673 }
2674
2675 $string = str_replace( $full, $half, $string );
2676 return $string;
2677 }
2678
2679 /**
2680 * @param $string string
2681 * @param $pattern string
2682 * @return string
2683 */
2684 protected static function insertSpace( $string, $pattern ) {
2685 $string = preg_replace( $pattern, " $1 ", $string );
2686 $string = preg_replace( '/ +/', ' ', $string );
2687 return $string;
2688 }
2689
2690 /**
2691 * @param $termsArray array
2692 * @return array
2693 */
2694 function convertForSearchResult( $termsArray ) {
2695 # some languages, e.g. Chinese, need to do a conversion
2696 # in order for search results to be displayed correctly
2697 return $termsArray;
2698 }
2699
2700 /**
2701 * Get the first character of a string.
2702 *
2703 * @param $s string
2704 * @return string
2705 */
2706 function firstChar( $s ) {
2707 $matches = array();
2708 preg_match(
2709 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2710 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2711 $s,
2712 $matches
2713 );
2714
2715 if ( isset( $matches[1] ) ) {
2716 if ( strlen( $matches[1] ) != 3 ) {
2717 return $matches[1];
2718 }
2719
2720 // Break down Hangul syllables to grab the first jamo
2721 $code = utf8ToCodepoint( $matches[1] );
2722 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2723 return $matches[1];
2724 } elseif ( $code < 0xb098 ) {
2725 return "\xe3\x84\xb1";
2726 } elseif ( $code < 0xb2e4 ) {
2727 return "\xe3\x84\xb4";
2728 } elseif ( $code < 0xb77c ) {
2729 return "\xe3\x84\xb7";
2730 } elseif ( $code < 0xb9c8 ) {
2731 return "\xe3\x84\xb9";
2732 } elseif ( $code < 0xbc14 ) {
2733 return "\xe3\x85\x81";
2734 } elseif ( $code < 0xc0ac ) {
2735 return "\xe3\x85\x82";
2736 } elseif ( $code < 0xc544 ) {
2737 return "\xe3\x85\x85";
2738 } elseif ( $code < 0xc790 ) {
2739 return "\xe3\x85\x87";
2740 } elseif ( $code < 0xcc28 ) {
2741 return "\xe3\x85\x88";
2742 } elseif ( $code < 0xce74 ) {
2743 return "\xe3\x85\x8a";
2744 } elseif ( $code < 0xd0c0 ) {
2745 return "\xe3\x85\x8b";
2746 } elseif ( $code < 0xd30c ) {
2747 return "\xe3\x85\x8c";
2748 } elseif ( $code < 0xd558 ) {
2749 return "\xe3\x85\x8d";
2750 } else {
2751 return "\xe3\x85\x8e";
2752 }
2753 } else {
2754 return '';
2755 }
2756 }
2757
2758 function initEncoding() {
2759 # Some languages may have an alternate char encoding option
2760 # (Esperanto X-coding, Japanese furigana conversion, etc)
2761 # If this language is used as the primary content language,
2762 # an override to the defaults can be set here on startup.
2763 }
2764
2765 /**
2766 * @param $s string
2767 * @return string
2768 */
2769 function recodeForEdit( $s ) {
2770 # For some languages we'll want to explicitly specify
2771 # which characters make it into the edit box raw
2772 # or are converted in some way or another.
2773 global $wgEditEncoding;
2774 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
2775 return $s;
2776 } else {
2777 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2778 }
2779 }
2780
2781 /**
2782 * @param $s string
2783 * @return string
2784 */
2785 function recodeInput( $s ) {
2786 # Take the previous into account.
2787 global $wgEditEncoding;
2788 if ( $wgEditEncoding != '' ) {
2789 $enc = $wgEditEncoding;
2790 } else {
2791 $enc = 'UTF-8';
2792 }
2793 if ( $enc == 'UTF-8' ) {
2794 return $s;
2795 } else {
2796 return $this->iconv( $enc, 'UTF-8', $s );
2797 }
2798 }
2799
2800 /**
2801 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2802 * also cleans up certain backwards-compatible sequences, converting them
2803 * to the modern Unicode equivalent.
2804 *
2805 * This is language-specific for performance reasons only.
2806 *
2807 * @param $s string
2808 *
2809 * @return string
2810 */
2811 function normalize( $s ) {
2812 global $wgAllUnicodeFixes;
2813 $s = UtfNormal::cleanUp( $s );
2814 if ( $wgAllUnicodeFixes ) {
2815 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2816 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2817 }
2818
2819 return $s;
2820 }
2821
2822 /**
2823 * Transform a string using serialized data stored in the given file (which
2824 * must be in the serialized subdirectory of $IP). The file contains pairs
2825 * mapping source characters to destination characters.
2826 *
2827 * The data is cached in process memory. This will go faster if you have the
2828 * FastStringSearch extension.
2829 *
2830 * @param $file string
2831 * @param $string string
2832 *
2833 * @throws MWException
2834 * @return string
2835 */
2836 function transformUsingPairFile( $file, $string ) {
2837 if ( !isset( $this->transformData[$file] ) ) {
2838 $data = wfGetPrecompiledData( $file );
2839 if ( $data === false ) {
2840 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2841 }
2842 $this->transformData[$file] = new ReplacementArray( $data );
2843 }
2844 return $this->transformData[$file]->replace( $string );
2845 }
2846
2847 /**
2848 * For right-to-left language support
2849 *
2850 * @return bool
2851 */
2852 function isRTL() {
2853 return self::$dataCache->getItem( $this->mCode, 'rtl' );
2854 }
2855
2856 /**
2857 * Return the correct HTML 'dir' attribute value for this language.
2858 * @return String
2859 */
2860 function getDir() {
2861 return $this->isRTL() ? 'rtl' : 'ltr';
2862 }
2863
2864 /**
2865 * Return 'left' or 'right' as appropriate alignment for line-start
2866 * for this language's text direction.
2867 *
2868 * Should be equivalent to CSS3 'start' text-align value....
2869 *
2870 * @return String
2871 */
2872 function alignStart() {
2873 return $this->isRTL() ? 'right' : 'left';
2874 }
2875
2876 /**
2877 * Return 'right' or 'left' as appropriate alignment for line-end
2878 * for this language's text direction.
2879 *
2880 * Should be equivalent to CSS3 'end' text-align value....
2881 *
2882 * @return String
2883 */
2884 function alignEnd() {
2885 return $this->isRTL() ? 'left' : 'right';
2886 }
2887
2888 /**
2889 * A hidden direction mark (LRM or RLM), depending on the language direction.
2890 * Unlike getDirMark(), this function returns the character as an HTML entity.
2891 * This function should be used when the output is guaranteed to be HTML,
2892 * because it makes the output HTML source code more readable. When
2893 * the output is plain text or can be escaped, getDirMark() should be used.
2894 *
2895 * @param $opposite Boolean Get the direction mark opposite to your language
2896 * @return string
2897 * @since 1.20
2898 */
2899 function getDirMarkEntity( $opposite = false ) {
2900 if ( $opposite ) {
2901 return $this->isRTL() ? '&lrm;' : '&rlm;';
2902 }
2903 return $this->isRTL() ? '&rlm;' : '&lrm;';
2904 }
2905
2906 /**
2907 * A hidden direction mark (LRM or RLM), depending on the language direction.
2908 * This function produces them as invisible Unicode characters and
2909 * the output may be hard to read and debug, so it should only be used
2910 * when the output is plain text or can be escaped. When the output is
2911 * HTML, use getDirMarkEntity() instead.
2912 *
2913 * @param $opposite Boolean Get the direction mark opposite to your language
2914 * @return string
2915 */
2916 function getDirMark( $opposite = false ) {
2917 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2918 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2919 if ( $opposite ) {
2920 return $this->isRTL() ? $lrm : $rlm;
2921 }
2922 return $this->isRTL() ? $rlm : $lrm;
2923 }
2924
2925 /**
2926 * @return array
2927 */
2928 function capitalizeAllNouns() {
2929 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2930 }
2931
2932 /**
2933 * An arrow, depending on the language direction.
2934 *
2935 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2936 * @return string
2937 */
2938 function getArrow( $direction = 'forwards' ) {
2939 switch ( $direction ) {
2940 case 'forwards':
2941 return $this->isRTL() ? '←' : '→';
2942 case 'backwards':
2943 return $this->isRTL() ? '→' : '←';
2944 case 'left':
2945 return '←';
2946 case 'right':
2947 return '→';
2948 case 'up':
2949 return '↑';
2950 case 'down':
2951 return '↓';
2952 }
2953 }
2954
2955 /**
2956 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2957 *
2958 * @return bool
2959 */
2960 function linkPrefixExtension() {
2961 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2962 }
2963
2964 /**
2965 * Get all magic words from cache.
2966 * @return array
2967 */
2968 function getMagicWords() {
2969 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2970 }
2971
2972 /**
2973 * Run the LanguageGetMagic hook once.
2974 */
2975 protected function doMagicHook() {
2976 if ( $this->mMagicHookDone ) {
2977 return;
2978 }
2979 $this->mMagicHookDone = true;
2980 wfProfileIn( 'LanguageGetMagic' );
2981 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2982 wfProfileOut( 'LanguageGetMagic' );
2983 }
2984
2985 /**
2986 * Fill a MagicWord object with data from here
2987 *
2988 * @param $mw
2989 */
2990 function getMagic( $mw ) {
2991 // Saves a function call
2992 if ( ! $this->mMagicHookDone ) {
2993 $this->doMagicHook();
2994 }
2995
2996 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
2997 $rawEntry = $this->mMagicExtensions[$mw->mId];
2998 } else {
2999 $rawEntry = self::$dataCache->getSubitem(
3000 $this->mCode, 'magicWords', $mw->mId );
3001 }
3002
3003 if ( !is_array( $rawEntry ) ) {
3004 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3005 } else {
3006 $mw->mCaseSensitive = $rawEntry[0];
3007 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3008 }
3009 }
3010
3011 /**
3012 * Add magic words to the extension array
3013 *
3014 * @param $newWords array
3015 */
3016 function addMagicWordsByLang( $newWords ) {
3017 $fallbackChain = $this->getFallbackLanguages();
3018 $fallbackChain = array_reverse( $fallbackChain );
3019 foreach ( $fallbackChain as $code ) {
3020 if ( isset( $newWords[$code] ) ) {
3021 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3022 }
3023 }
3024 }
3025
3026 /**
3027 * Get special page names, as an associative array
3028 * case folded alias => real name
3029 */
3030 function getSpecialPageAliases() {
3031 // Cache aliases because it may be slow to load them
3032 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3033 // Initialise array
3034 $this->mExtendedSpecialPageAliases =
3035 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3036 wfRunHooks( 'LanguageGetSpecialPageAliases',
3037 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
3038 }
3039
3040 return $this->mExtendedSpecialPageAliases;
3041 }
3042
3043 /**
3044 * Italic is unsuitable for some languages
3045 *
3046 * @param $text String: the text to be emphasized.
3047 * @return string
3048 */
3049 function emphasize( $text ) {
3050 return "<em>$text</em>";
3051 }
3052
3053 /**
3054 * Normally we output all numbers in plain en_US style, that is
3055 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3056 * point twohundredthirtyfive. However this is not suitable for all
3057 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
3058 * Icelandic just want to use commas instead of dots, and dots instead
3059 * of commas like "293.291,235".
3060 *
3061 * An example of this function being called:
3062 * <code>
3063 * wfMessage( 'message' )->numParams( $num )->text()
3064 * </code>
3065 *
3066 * See LanguageGu.php for the Gujarati implementation and
3067 * $separatorTransformTable on MessageIs.php for
3068 * the , => . and . => , implementation.
3069 *
3070 * @todo check if it's viable to use localeconv() for the decimal
3071 * separator thing.
3072 * @param $number Mixed: the string to be formatted, should be an integer
3073 * or a floating point number.
3074 * @param $nocommafy Bool: set to true for special numbers like dates
3075 * @return string
3076 */
3077 public function formatNum( $number, $nocommafy = false ) {
3078 global $wgTranslateNumerals;
3079 if ( !$nocommafy ) {
3080 $number = $this->commafy( $number );
3081 $s = $this->separatorTransformTable();
3082 if ( $s ) {
3083 $number = strtr( $number, $s );
3084 }
3085 }
3086
3087 if ( $wgTranslateNumerals ) {
3088 $s = $this->digitTransformTable();
3089 if ( $s ) {
3090 $number = strtr( $number, $s );
3091 }
3092 }
3093
3094 return $number;
3095 }
3096
3097 /**
3098 * Front-end for non-commafied formatNum
3099 *
3100 * @param mixed $number the string to be formatted, should be an integer
3101 * or a floating point number.
3102 * @since 1.21
3103 * @return string
3104 */
3105 public function formatNumNoSeparators( $number ) {
3106 return $this->formatNum( $number, true );
3107 }
3108
3109 /**
3110 * @param $number string
3111 * @return string
3112 */
3113 function parseFormattedNumber( $number ) {
3114 $s = $this->digitTransformTable();
3115 if ( $s ) {
3116 $number = strtr( $number, array_flip( $s ) );
3117 }
3118
3119 $s = $this->separatorTransformTable();
3120 if ( $s ) {
3121 $number = strtr( $number, array_flip( $s ) );
3122 }
3123
3124 $number = strtr( $number, array( ',' => '' ) );
3125 return $number;
3126 }
3127
3128 /**
3129 * Adds commas to a given number
3130 * @since 1.19
3131 * @param $number mixed
3132 * @return string
3133 */
3134 function commafy( $number ) {
3135 $digitGroupingPattern = $this->digitGroupingPattern();
3136 if ( $number === null ) {
3137 return '';
3138 }
3139
3140 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3141 // default grouping is at thousands, use the same for ###,###,### pattern too.
3142 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3143 } else {
3144 // Ref: http://cldr.unicode.org/translation/number-patterns
3145 $sign = "";
3146 if ( intval( $number ) < 0 ) {
3147 // For negative numbers apply the algorithm like positive number and add sign.
3148 $sign = "-";
3149 $number = substr( $number, 1 );
3150 }
3151 $integerPart = array();
3152 $decimalPart = array();
3153 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3154 preg_match( "/\d+/", $number, $integerPart );
3155 preg_match( "/\.\d*/", $number, $decimalPart );
3156 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3157 if ( $groupedNumber === $number ) {
3158 // the string does not have any number part. Eg: .12345
3159 return $sign . $groupedNumber;
3160 }
3161 $start = $end = strlen( $integerPart[0] );
3162 while ( $start > 0 ) {
3163 $match = $matches[0][$numMatches - 1];
3164 $matchLen = strlen( $match );
3165 $start = $end - $matchLen;
3166 if ( $start < 0 ) {
3167 $start = 0;
3168 }
3169 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3170 $end = $start;
3171 if ( $numMatches > 1 ) {
3172 // use the last pattern for the rest of the number
3173 $numMatches--;
3174 }
3175 if ( $start > 0 ) {
3176 $groupedNumber = "," . $groupedNumber;
3177 }
3178 }
3179 return $sign . $groupedNumber;
3180 }
3181 }
3182
3183 /**
3184 * @return String
3185 */
3186 function digitGroupingPattern() {
3187 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3188 }
3189
3190 /**
3191 * @return array
3192 */
3193 function digitTransformTable() {
3194 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3195 }
3196
3197 /**
3198 * @return array
3199 */
3200 function separatorTransformTable() {
3201 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3202 }
3203
3204 /**
3205 * Take a list of strings and build a locale-friendly comma-separated
3206 * list, using the local comma-separator message.
3207 * The last two strings are chained with an "and".
3208 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3209 *
3210 * @param $l Array
3211 * @return string
3212 */
3213 function listToText( array $l ) {
3214 $m = count( $l ) - 1;
3215 if ( $m < 0 ) {
3216 return '';
3217 }
3218 if ( $m > 0 ) {
3219 $and = $this->getMessageFromDB( 'and' );
3220 $space = $this->getMessageFromDB( 'word-separator' );
3221 if ( $m > 1 ) {
3222 $comma = $this->getMessageFromDB( 'comma-separator' );
3223 }
3224 }
3225 $s = $l[$m];
3226 for ( $i = $m - 1; $i >= 0; $i-- ) {
3227 if ( $i == $m - 1 ) {
3228 $s = $l[$i] . $and . $space . $s;
3229 } else {
3230 $s = $l[$i] . $comma . $s;
3231 }
3232 }
3233 return $s;
3234 }
3235
3236 /**
3237 * Take a list of strings and build a locale-friendly comma-separated
3238 * list, using the local comma-separator message.
3239 * @param $list array of strings to put in a comma list
3240 * @return string
3241 */
3242 function commaList( array $list ) {
3243 return implode(
3244 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3245 $list
3246 );
3247 }
3248
3249 /**
3250 * Take a list of strings and build a locale-friendly semicolon-separated
3251 * list, using the local semicolon-separator message.
3252 * @param $list array of strings to put in a semicolon list
3253 * @return string
3254 */
3255 function semicolonList( array $list ) {
3256 return implode(
3257 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3258 $list
3259 );
3260 }
3261
3262 /**
3263 * Same as commaList, but separate it with the pipe instead.
3264 * @param $list array of strings to put in a pipe list
3265 * @return string
3266 */
3267 function pipeList( array $list ) {
3268 return implode(
3269 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3270 $list
3271 );
3272 }
3273
3274 /**
3275 * Truncate a string to a specified length in bytes, appending an optional
3276 * string (e.g. for ellipses)
3277 *
3278 * The database offers limited byte lengths for some columns in the database;
3279 * multi-byte character sets mean we need to ensure that only whole characters
3280 * are included, otherwise broken characters can be passed to the user
3281 *
3282 * If $length is negative, the string will be truncated from the beginning
3283 *
3284 * @param $string String to truncate
3285 * @param $length Int: maximum length (including ellipses)
3286 * @param $ellipsis String to append to the truncated text
3287 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3288 * $adjustLength was introduced in 1.18, before that behaved as if false.
3289 * @return string
3290 */
3291 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3292 # Use the localized ellipsis character
3293 if ( $ellipsis == '...' ) {
3294 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3295 }
3296 # Check if there is no need to truncate
3297 if ( $length == 0 ) {
3298 return $ellipsis; // convention
3299 } elseif ( strlen( $string ) <= abs( $length ) ) {
3300 return $string; // no need to truncate
3301 }
3302 $stringOriginal = $string;
3303 # If ellipsis length is >= $length then we can't apply $adjustLength
3304 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3305 $string = $ellipsis; // this can be slightly unexpected
3306 # Otherwise, truncate and add ellipsis...
3307 } else {
3308 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3309 if ( $length > 0 ) {
3310 $length -= $eLength;
3311 $string = substr( $string, 0, $length ); // xyz...
3312 $string = $this->removeBadCharLast( $string );
3313 $string = $string . $ellipsis;
3314 } else {
3315 $length += $eLength;
3316 $string = substr( $string, $length ); // ...xyz
3317 $string = $this->removeBadCharFirst( $string );
3318 $string = $ellipsis . $string;
3319 }
3320 }
3321 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3322 # This check is *not* redundant if $adjustLength, due to the single case where
3323 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3324 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3325 return $string;
3326 } else {
3327 return $stringOriginal;
3328 }
3329 }
3330
3331 /**
3332 * Remove bytes that represent an incomplete Unicode character
3333 * at the end of string (e.g. bytes of the char are missing)
3334 *
3335 * @param $string String
3336 * @return string
3337 */
3338 protected function removeBadCharLast( $string ) {
3339 if ( $string != '' ) {
3340 $char = ord( $string[strlen( $string ) - 1] );
3341 $m = array();
3342 if ( $char >= 0xc0 ) {
3343 # We got the first byte only of a multibyte char; remove it.
3344 $string = substr( $string, 0, -1 );
3345 } elseif ( $char >= 0x80 &&
3346 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3347 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3348 ) {
3349 # We chopped in the middle of a character; remove it
3350 $string = $m[1];
3351 }
3352 }
3353 return $string;
3354 }
3355
3356 /**
3357 * Remove bytes that represent an incomplete Unicode character
3358 * at the start of string (e.g. bytes of the char are missing)
3359 *
3360 * @param $string String
3361 * @return string
3362 */
3363 protected function removeBadCharFirst( $string ) {
3364 if ( $string != '' ) {
3365 $char = ord( $string[0] );
3366 if ( $char >= 0x80 && $char < 0xc0 ) {
3367 # We chopped in the middle of a character; remove the whole thing
3368 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3369 }
3370 }
3371 return $string;
3372 }
3373
3374 /**
3375 * Truncate a string of valid HTML to a specified length in bytes,
3376 * appending an optional string (e.g. for ellipses), and return valid HTML
3377 *
3378 * This is only intended for styled/linked text, such as HTML with
3379 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3380 * Also, this will not detect things like "display:none" CSS.
3381 *
3382 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3383 *
3384 * @param string $text HTML string to truncate
3385 * @param int $length (zero/positive) Maximum length (including ellipses)
3386 * @param string $ellipsis String to append to the truncated text
3387 * @return string
3388 */
3389 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3390 # Use the localized ellipsis character
3391 if ( $ellipsis == '...' ) {
3392 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3393 }
3394 # Check if there is clearly no need to truncate
3395 if ( $length <= 0 ) {
3396 return $ellipsis; // no text shown, nothing to format (convention)
3397 } elseif ( strlen( $text ) <= $length ) {
3398 return $text; // string short enough even *with* HTML (short-circuit)
3399 }
3400
3401 $dispLen = 0; // innerHTML legth so far
3402 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3403 $tagType = 0; // 0-open, 1-close
3404 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3405 $entityState = 0; // 0-not entity, 1-entity
3406 $tag = $ret = ''; // accumulated tag name, accumulated result string
3407 $openTags = array(); // open tag stack
3408 $maybeState = null; // possible truncation state
3409
3410 $textLen = strlen( $text );
3411 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3412 for ( $pos = 0; true; ++$pos ) {
3413 # Consider truncation once the display length has reached the maximim.
3414 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3415 # Check that we're not in the middle of a bracket/entity...
3416 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3417 if ( !$testingEllipsis ) {
3418 $testingEllipsis = true;
3419 # Save where we are; we will truncate here unless there turn out to
3420 # be so few remaining characters that truncation is not necessary.
3421 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3422 $maybeState = array( $ret, $openTags ); // save state
3423 }
3424 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3425 # String in fact does need truncation, the truncation point was OK.
3426 list( $ret, $openTags ) = $maybeState; // reload state
3427 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3428 $ret .= $ellipsis; // add ellipsis
3429 break;
3430 }
3431 }
3432 if ( $pos >= $textLen ) {
3433 break; // extra iteration just for above checks
3434 }
3435
3436 # Read the next char...
3437 $ch = $text[$pos];
3438 $lastCh = $pos ? $text[$pos - 1] : '';
3439 $ret .= $ch; // add to result string
3440 if ( $ch == '<' ) {
3441 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3442 $entityState = 0; // for bad HTML
3443 $bracketState = 1; // tag started (checking for backslash)
3444 } elseif ( $ch == '>' ) {
3445 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3446 $entityState = 0; // for bad HTML
3447 $bracketState = 0; // out of brackets
3448 } elseif ( $bracketState == 1 ) {
3449 if ( $ch == '/' ) {
3450 $tagType = 1; // close tag (e.g. "</span>")
3451 } else {
3452 $tagType = 0; // open tag (e.g. "<span>")
3453 $tag .= $ch;
3454 }
3455 $bracketState = 2; // building tag name
3456 } elseif ( $bracketState == 2 ) {
3457 if ( $ch != ' ' ) {
3458 $tag .= $ch;
3459 } else {
3460 // Name found (e.g. "<a href=..."), add on tag attributes...
3461 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3462 }
3463 } elseif ( $bracketState == 0 ) {
3464 if ( $entityState ) {
3465 if ( $ch == ';' ) {
3466 $entityState = 0;
3467 $dispLen++; // entity is one displayed char
3468 }
3469 } else {
3470 if ( $neLength == 0 && !$maybeState ) {
3471 // Save state without $ch. We want to *hit* the first
3472 // display char (to get tags) but not *use* it if truncating.
3473 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3474 }
3475 if ( $ch == '&' ) {
3476 $entityState = 1; // entity found, (e.g. "&#160;")
3477 } else {
3478 $dispLen++; // this char is displayed
3479 // Add the next $max display text chars after this in one swoop...
3480 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3481 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3482 $dispLen += $skipped;
3483 $pos += $skipped;
3484 }
3485 }
3486 }
3487 }
3488 // Close the last tag if left unclosed by bad HTML
3489 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3490 while ( count( $openTags ) > 0 ) {
3491 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3492 }
3493 return $ret;
3494 }
3495
3496 /**
3497 * truncateHtml() helper function
3498 * like strcspn() but adds the skipped chars to $ret
3499 *
3500 * @param $ret
3501 * @param $text
3502 * @param $search
3503 * @param $start
3504 * @param $len
3505 * @return int
3506 */
3507 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3508 if ( $len === null ) {
3509 $len = -1; // -1 means "no limit" for strcspn
3510 } elseif ( $len < 0 ) {
3511 $len = 0; // sanity
3512 }
3513 $skipCount = 0;
3514 if ( $start < strlen( $text ) ) {
3515 $skipCount = strcspn( $text, $search, $start, $len );
3516 $ret .= substr( $text, $start, $skipCount );
3517 }
3518 return $skipCount;
3519 }
3520
3521 /**
3522 * truncateHtml() helper function
3523 * (a) push or pop $tag from $openTags as needed
3524 * (b) clear $tag value
3525 * @param &$tag string Current HTML tag name we are looking at
3526 * @param $tagType int (0-open tag, 1-close tag)
3527 * @param $lastCh string Character before the '>' that ended this tag
3528 * @param &$openTags array Open tag stack (not accounting for $tag)
3529 */
3530 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3531 $tag = ltrim( $tag );
3532 if ( $tag != '' ) {
3533 if ( $tagType == 0 && $lastCh != '/' ) {
3534 $openTags[] = $tag; // tag opened (didn't close itself)
3535 } elseif ( $tagType == 1 ) {
3536 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3537 array_pop( $openTags ); // tag closed
3538 }
3539 }
3540 $tag = '';
3541 }
3542 }
3543
3544 /**
3545 * Grammatical transformations, needed for inflected languages
3546 * Invoked by putting {{grammar:case|word}} in a message
3547 *
3548 * @param $word string
3549 * @param $case string
3550 * @return string
3551 */
3552 function convertGrammar( $word, $case ) {
3553 global $wgGrammarForms;
3554 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3555 return $wgGrammarForms[$this->getCode()][$case][$word];
3556 }
3557 return $word;
3558 }
3559 /**
3560 * Get the grammar forms for the content language
3561 * @return array of grammar forms
3562 * @since 1.20
3563 */
3564 function getGrammarForms() {
3565 global $wgGrammarForms;
3566 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3567 return $wgGrammarForms[$this->getCode()];
3568 }
3569 return array();
3570 }
3571 /**
3572 * Provides an alternative text depending on specified gender.
3573 * Usage {{gender:username|masculine|feminine|neutral}}.
3574 * username is optional, in which case the gender of current user is used,
3575 * but only in (some) interface messages; otherwise default gender is used.
3576 *
3577 * If no forms are given, an empty string is returned. If only one form is
3578 * given, it will be returned unconditionally. These details are implied by
3579 * the caller and cannot be overridden in subclasses.
3580 *
3581 * If more than one form is given, the default is to use the neutral one
3582 * if it is specified, and to use the masculine one otherwise. These
3583 * details can be overridden in subclasses.
3584 *
3585 * @param $gender string
3586 * @param $forms array
3587 *
3588 * @return string
3589 */
3590 function gender( $gender, $forms ) {
3591 if ( !count( $forms ) ) {
3592 return '';
3593 }
3594 $forms = $this->preConvertPlural( $forms, 2 );
3595 if ( $gender === 'male' ) {
3596 return $forms[0];
3597 }
3598 if ( $gender === 'female' ) {
3599 return $forms[1];
3600 }
3601 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3602 }
3603
3604 /**
3605 * Plural form transformations, needed for some languages.
3606 * For example, there are 3 form of plural in Russian and Polish,
3607 * depending on "count mod 10". See [[w:Plural]]
3608 * For English it is pretty simple.
3609 *
3610 * Invoked by putting {{plural:count|wordform1|wordform2}}
3611 * or {{plural:count|wordform1|wordform2|wordform3}}
3612 *
3613 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3614 *
3615 * @param $count Integer: non-localized number
3616 * @param $forms Array: different plural forms
3617 * @return string Correct form of plural for $count in this language
3618 */
3619 function convertPlural( $count, $forms ) {
3620 // Handle explicit n=pluralform cases
3621 foreach ( $forms as $index => $form ) {
3622 if ( preg_match( '/^\d+=/i', $form ) ) {
3623 $pos = strpos( $form, '=' );
3624 if ( substr( $form, 0, $pos ) === (string) $count ) {
3625 return substr( $form, $pos + 1 );
3626 }
3627 unset( $forms[$index] );
3628 }
3629 }
3630
3631 $forms = array_values( $forms );
3632 if ( !count( $forms ) ) {
3633 return '';
3634 }
3635
3636 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3637 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3638 return $forms[$pluralForm];
3639 }
3640
3641 /**
3642 * Checks that convertPlural was given an array and pads it to requested
3643 * amount of forms by copying the last one.
3644 *
3645 * @param $count Integer: How many forms should there be at least
3646 * @param $forms Array of forms given to convertPlural
3647 * @return array Padded array of forms or an exception if not an array
3648 */
3649 protected function preConvertPlural( /* Array */ $forms, $count ) {
3650 while ( count( $forms ) < $count ) {
3651 $forms[] = $forms[count( $forms ) - 1];
3652 }
3653 return $forms;
3654 }
3655
3656 /**
3657 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3658 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3659 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3660 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3661 * match up with it.
3662 *
3663 * @param $str String: the validated block duration in English
3664 * @return string Somehow translated block duration
3665 * @see LanguageFi.php for example implementation
3666 */
3667 function translateBlockExpiry( $str ) {
3668 $duration = SpecialBlock::getSuggestedDurations( $this );
3669 foreach ( $duration as $show => $value ) {
3670 if ( strcmp( $str, $value ) == 0 ) {
3671 return htmlspecialchars( trim( $show ) );
3672 }
3673 }
3674
3675 // Since usually only infinite or indefinite is only on list, so try
3676 // equivalents if still here.
3677 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3678 if ( in_array( $str, $indefs ) ) {
3679 foreach ( $indefs as $val ) {
3680 $show = array_search( $val, $duration, true );
3681 if ( $show !== false ) {
3682 return htmlspecialchars( trim( $show ) );
3683 }
3684 }
3685 }
3686
3687 // If all else fails, return a standard duration or timestamp description.
3688 $time = strtotime( $str, 0 );
3689 if ( $time === false ) { // Unknown format. Return it as-is in case.
3690 return $str;
3691 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3692 // $time is relative to 0 so it's a duration length.
3693 return $this->formatDuration( $time );
3694 } else { // It's an absolute timestamp.
3695 if ( $time === 0 ) {
3696 // wfTimestamp() handles 0 as current time instead of epoch.
3697 return $this->timeanddate( '19700101000000' );
3698 } else {
3699 return $this->timeanddate( $time );
3700 }
3701 }
3702 }
3703
3704 /**
3705 * languages like Chinese need to be segmented in order for the diff
3706 * to be of any use
3707 *
3708 * @param $text String
3709 * @return String
3710 */
3711 public function segmentForDiff( $text ) {
3712 return $text;
3713 }
3714
3715 /**
3716 * and unsegment to show the result
3717 *
3718 * @param $text String
3719 * @return String
3720 */
3721 public function unsegmentForDiff( $text ) {
3722 return $text;
3723 }
3724
3725 /**
3726 * Return the LanguageConverter used in the Language
3727 *
3728 * @since 1.19
3729 * @return LanguageConverter
3730 */
3731 public function getConverter() {
3732 return $this->mConverter;
3733 }
3734
3735 /**
3736 * convert text to all supported variants
3737 *
3738 * @param $text string
3739 * @return array
3740 */
3741 public function autoConvertToAllVariants( $text ) {
3742 return $this->mConverter->autoConvertToAllVariants( $text );
3743 }
3744
3745 /**
3746 * convert text to different variants of a language.
3747 *
3748 * @param $text string
3749 * @return string
3750 */
3751 public function convert( $text ) {
3752 return $this->mConverter->convert( $text );
3753 }
3754
3755 /**
3756 * Convert a Title object to a string in the preferred variant
3757 *
3758 * @param $title Title
3759 * @return string
3760 */
3761 public function convertTitle( $title ) {
3762 return $this->mConverter->convertTitle( $title );
3763 }
3764
3765 /**
3766 * Convert a namespace index to a string in the preferred variant
3767 *
3768 * @param $ns int
3769 * @return string
3770 */
3771 public function convertNamespace( $ns ) {
3772 return $this->mConverter->convertNamespace( $ns );
3773 }
3774
3775 /**
3776 * Check if this is a language with variants
3777 *
3778 * @return bool
3779 */
3780 public function hasVariants() {
3781 return count( $this->getVariants() ) > 1;
3782 }
3783
3784 /**
3785 * Check if the language has the specific variant
3786 *
3787 * @since 1.19
3788 * @param $variant string
3789 * @return bool
3790 */
3791 public function hasVariant( $variant ) {
3792 return (bool)$this->mConverter->validateVariant( $variant );
3793 }
3794
3795 /**
3796 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3797 *
3798 * @param $text string
3799 * @return string
3800 */
3801 public function armourMath( $text ) {
3802 return $this->mConverter->armourMath( $text );
3803 }
3804
3805 /**
3806 * Perform output conversion on a string, and encode for safe HTML output.
3807 * @param $text String text to be converted
3808 * @param $isTitle Bool whether this conversion is for the article title
3809 * @return string
3810 * @todo this should get integrated somewhere sane
3811 */
3812 public function convertHtml( $text, $isTitle = false ) {
3813 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3814 }
3815
3816 /**
3817 * @param $key string
3818 * @return string
3819 */
3820 public function convertCategoryKey( $key ) {
3821 return $this->mConverter->convertCategoryKey( $key );
3822 }
3823
3824 /**
3825 * Get the list of variants supported by this language
3826 * see sample implementation in LanguageZh.php
3827 *
3828 * @return array an array of language codes
3829 */
3830 public function getVariants() {
3831 return $this->mConverter->getVariants();
3832 }
3833
3834 /**
3835 * @return string
3836 */
3837 public function getPreferredVariant() {
3838 return $this->mConverter->getPreferredVariant();
3839 }
3840
3841 /**
3842 * @return string
3843 */
3844 public function getDefaultVariant() {
3845 return $this->mConverter->getDefaultVariant();
3846 }
3847
3848 /**
3849 * @return string
3850 */
3851 public function getURLVariant() {
3852 return $this->mConverter->getURLVariant();
3853 }
3854
3855 /**
3856 * If a language supports multiple variants, it is
3857 * possible that non-existing link in one variant
3858 * actually exists in another variant. this function
3859 * tries to find it. See e.g. LanguageZh.php
3860 *
3861 * @param $link String: the name of the link
3862 * @param $nt Mixed: the title object of the link
3863 * @param $ignoreOtherCond Boolean: to disable other conditions when
3864 * we need to transclude a template or update a category's link
3865 * @return null the input parameters may be modified upon return
3866 */
3867 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3868 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
3869 }
3870
3871 /**
3872 * If a language supports multiple variants, converts text
3873 * into an array of all possible variants of the text:
3874 * 'variant' => text in that variant
3875 *
3876 * @deprecated since 1.17 Use autoConvertToAllVariants()
3877 *
3878 * @param $text string
3879 *
3880 * @return string
3881 */
3882 public function convertLinkToAllVariants( $text ) {
3883 return $this->mConverter->convertLinkToAllVariants( $text );
3884 }
3885
3886 /**
3887 * returns language specific options used by User::getPageRenderHash()
3888 * for example, the preferred language variant
3889 *
3890 * @return string
3891 */
3892 function getExtraHashOptions() {
3893 return $this->mConverter->getExtraHashOptions();
3894 }
3895
3896 /**
3897 * For languages that support multiple variants, the title of an
3898 * article may be displayed differently in different variants. this
3899 * function returns the apporiate title defined in the body of the article.
3900 *
3901 * @return string
3902 */
3903 public function getParsedTitle() {
3904 return $this->mConverter->getParsedTitle();
3905 }
3906
3907 /**
3908 * Prepare external link text for conversion. When the text is
3909 * a URL, it shouldn't be converted, and it'll be wrapped in
3910 * the "raw" tag (-{R| }-) to prevent conversion.
3911 *
3912 * This function is called "markNoConversion" for historical
3913 * reasons.
3914 *
3915 * @param $text String: text to be used for external link
3916 * @param $noParse bool: wrap it without confirming it's a real URL first
3917 * @return string the tagged text
3918 */
3919 public function markNoConversion( $text, $noParse = false ) {
3920 // Excluding protocal-relative URLs may avoid many false positives.
3921 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
3922 return $this->mConverter->markNoConversion( $text );
3923 } else {
3924 return $text;
3925 }
3926 }
3927
3928 /**
3929 * A regular expression to match legal word-trailing characters
3930 * which should be merged onto a link of the form [[foo]]bar.
3931 *
3932 * @return string
3933 */
3934 public function linkTrail() {
3935 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
3936 }
3937
3938 /**
3939 * @return Language
3940 */
3941 function getLangObj() {
3942 return $this;
3943 }
3944
3945 /**
3946 * Get the "parent" language which has a converter to convert a "compatible" language
3947 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
3948 *
3949 * @return Language|null
3950 * @since 1.22
3951 */
3952 public function getParentLanguage() {
3953 if ( $this->mParentLanguage !== false ) {
3954 return $this->mParentLanguage;
3955 }
3956
3957 $pieces = explode( '-', $this->getCode() );
3958 $code = $pieces[0];
3959 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
3960 $this->mParentLanguage = null;
3961 return null;
3962 }
3963 $lang = Language::factory( $code );
3964 if ( !$lang->hasVariant( $this->getCode() ) ) {
3965 $this->mParentLanguage = null;
3966 return null;
3967 }
3968
3969 $this->mParentLanguage = $lang;
3970 return $lang;
3971 }
3972
3973 /**
3974 * Get the RFC 3066 code for this language object
3975 *
3976 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3977 * htmlspecialchars() or similar
3978 *
3979 * @return string
3980 */
3981 public function getCode() {
3982 return $this->mCode;
3983 }
3984
3985 /**
3986 * Get the code in Bcp47 format which we can use
3987 * inside of html lang="" tags.
3988 *
3989 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3990 * htmlspecialchars() or similar.
3991 *
3992 * @since 1.19
3993 * @return string
3994 */
3995 public function getHtmlCode() {
3996 if ( is_null( $this->mHtmlCode ) ) {
3997 $this->mHtmlCode = wfBCP47( $this->getCode() );
3998 }
3999 return $this->mHtmlCode;
4000 }
4001
4002 /**
4003 * @param $code string
4004 */
4005 public function setCode( $code ) {
4006 $this->mCode = $code;
4007 // Ensure we don't leave incorrect cached data lying around
4008 $this->mHtmlCode = null;
4009 $this->mParentLanguage = false;
4010 }
4011
4012 /**
4013 * Get the name of a file for a certain language code
4014 * @param $prefix string Prepend this to the filename
4015 * @param $code string Language code
4016 * @param $suffix string Append this to the filename
4017 * @throws MWException
4018 * @return string $prefix . $mangledCode . $suffix
4019 */
4020 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4021 // Protect against path traversal
4022 if ( !Language::isValidCode( $code )
4023 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
4024 {
4025 throw new MWException( "Invalid language code \"$code\"" );
4026 }
4027
4028 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4029 }
4030
4031 /**
4032 * Get the language code from a file name. Inverse of getFileName()
4033 * @param $filename string $prefix . $languageCode . $suffix
4034 * @param $prefix string Prefix before the language code
4035 * @param $suffix string Suffix after the language code
4036 * @return string Language code, or false if $prefix or $suffix isn't found
4037 */
4038 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4039 $m = null;
4040 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4041 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4042 if ( !count( $m ) ) {
4043 return false;
4044 }
4045 return str_replace( '_', '-', strtolower( $m[1] ) );
4046 }
4047
4048 /**
4049 * @param $code string
4050 * @return string
4051 */
4052 public static function getMessagesFileName( $code ) {
4053 global $IP;
4054 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4055 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
4056 return $file;
4057 }
4058
4059 /**
4060 * @param $code string
4061 * @return string
4062 */
4063 public static function getClassFileName( $code ) {
4064 global $IP;
4065 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4066 }
4067
4068 /**
4069 * Get the first fallback for a given language.
4070 *
4071 * @param $code string
4072 *
4073 * @return bool|string
4074 */
4075 public static function getFallbackFor( $code ) {
4076 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4077 return false;
4078 } else {
4079 $fallbacks = self::getFallbacksFor( $code );
4080 $first = array_shift( $fallbacks );
4081 return $first;
4082 }
4083 }
4084
4085 /**
4086 * Get the ordered list of fallback languages.
4087 *
4088 * @since 1.19
4089 * @param $code string Language code
4090 * @return array
4091 */
4092 public static function getFallbacksFor( $code ) {
4093 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4094 return array();
4095 } else {
4096 $v = self::getLocalisationCache()->getItem( $code, 'fallback' );
4097 $v = array_map( 'trim', explode( ',', $v ) );
4098 if ( $v[count( $v ) - 1] !== 'en' ) {
4099 $v[] = 'en';
4100 }
4101 return $v;
4102 }
4103 }
4104
4105 /**
4106 * Get the ordered list of fallback languages, ending with the fallback
4107 * language chain for the site language.
4108 *
4109 * @since 1.22
4110 * @param string $code Language code
4111 * @return array array( fallbacks, site fallbacks )
4112 */
4113 public static function getFallbacksIncludingSiteLanguage( $code ) {
4114 global $wgLanguageCode;
4115
4116 // Usually, we will only store a tiny number of fallback chains, so we
4117 // keep them in static memory.
4118 $cacheKey = "{$code}-{$wgLanguageCode}";
4119
4120 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4121 $fallbacks = self::getFallbacksFor( $code );
4122
4123 // Append the site's fallback chain, including the site language itself
4124 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4125 array_unshift( $siteFallbacks, $wgLanguageCode );
4126
4127 // Eliminate any languages already included in the chain
4128 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4129
4130 self::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4131 }
4132 return self::$fallbackLanguageCache[$cacheKey];
4133 }
4134
4135 /**
4136 * Get all messages for a given language
4137 * WARNING: this may take a long time. If you just need all message *keys*
4138 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4139 *
4140 * @param $code string
4141 *
4142 * @return array
4143 */
4144 public static function getMessagesFor( $code ) {
4145 return self::getLocalisationCache()->getItem( $code, 'messages' );
4146 }
4147
4148 /**
4149 * Get a message for a given language
4150 *
4151 * @param $key string
4152 * @param $code string
4153 *
4154 * @return string
4155 */
4156 public static function getMessageFor( $key, $code ) {
4157 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4158 }
4159
4160 /**
4161 * Get all message keys for a given language. This is a faster alternative to
4162 * array_keys( Language::getMessagesFor( $code ) )
4163 *
4164 * @since 1.19
4165 * @param $code string Language code
4166 * @return array of message keys (strings)
4167 */
4168 public static function getMessageKeysFor( $code ) {
4169 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
4170 }
4171
4172 /**
4173 * @param $talk
4174 * @return mixed
4175 */
4176 function fixVariableInNamespace( $talk ) {
4177 if ( strpos( $talk, '$1' ) === false ) {
4178 return $talk;
4179 }
4180
4181 global $wgMetaNamespace;
4182 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4183
4184 # Allow grammar transformations
4185 # Allowing full message-style parsing would make simple requests
4186 # such as action=raw much more expensive than they need to be.
4187 # This will hopefully cover most cases.
4188 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4189 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4190 return str_replace( ' ', '_', $talk );
4191 }
4192
4193 /**
4194 * @param $m string
4195 * @return string
4196 */
4197 function replaceGrammarInNamespace( $m ) {
4198 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4199 }
4200
4201 /**
4202 * @throws MWException
4203 * @return array
4204 */
4205 static function getCaseMaps() {
4206 static $wikiUpperChars, $wikiLowerChars;
4207 if ( isset( $wikiUpperChars ) ) {
4208 return array( $wikiUpperChars, $wikiLowerChars );
4209 }
4210
4211 wfProfileIn( __METHOD__ );
4212 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4213 if ( $arr === false ) {
4214 throw new MWException(
4215 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4216 }
4217 $wikiUpperChars = $arr['wikiUpperChars'];
4218 $wikiLowerChars = $arr['wikiLowerChars'];
4219 wfProfileOut( __METHOD__ );
4220 return array( $wikiUpperChars, $wikiLowerChars );
4221 }
4222
4223 /**
4224 * Decode an expiry (block, protection, etc) which has come from the DB
4225 *
4226 * @todo FIXME: why are we returnings DBMS-dependent strings???
4227 *
4228 * @param $expiry String: Database expiry String
4229 * @param $format Bool|Int true to process using language functions, or TS_ constant
4230 * to return the expiry in a given timestamp
4231 * @return String
4232 * @since 1.18
4233 */
4234 public function formatExpiry( $expiry, $format = true ) {
4235 static $infinity;
4236 if ( $infinity === null ) {
4237 $infinity = wfGetDB( DB_SLAVE )->getInfinity();
4238 }
4239
4240 if ( $expiry == '' || $expiry == $infinity ) {
4241 return $format === true
4242 ? $this->getMessageFromDB( 'infiniteblock' )
4243 : $infinity;
4244 } else {
4245 return $format === true
4246 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4247 : wfTimestamp( $format, $expiry );
4248 }
4249 }
4250
4251 /**
4252 * @todo Document
4253 * @param $seconds int|float
4254 * @param $format Array Optional
4255 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
4256 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
4257 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
4258 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
4259 * @return string
4260 */
4261 function formatTimePeriod( $seconds, $format = array() ) {
4262 if ( !is_array( $format ) ) {
4263 $format = array( 'avoid' => $format ); // For backwards compatibility
4264 }
4265 if ( !isset( $format['avoid'] ) ) {
4266 $format['avoid'] = false;
4267 }
4268 if ( !isset( $format['noabbrevs' ] ) ) {
4269 $format['noabbrevs'] = false;
4270 }
4271 $secondsMsg = wfMessage(
4272 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4273 $minutesMsg = wfMessage(
4274 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4275 $hoursMsg = wfMessage(
4276 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4277 $daysMsg = wfMessage(
4278 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4279
4280 if ( round( $seconds * 10 ) < 100 ) {
4281 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4282 $s = $secondsMsg->params( $s )->text();
4283 } elseif ( round( $seconds ) < 60 ) {
4284 $s = $this->formatNum( round( $seconds ) );
4285 $s = $secondsMsg->params( $s )->text();
4286 } elseif ( round( $seconds ) < 3600 ) {
4287 $minutes = floor( $seconds / 60 );
4288 $secondsPart = round( fmod( $seconds, 60 ) );
4289 if ( $secondsPart == 60 ) {
4290 $secondsPart = 0;
4291 $minutes++;
4292 }
4293 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4294 $s .= ' ';
4295 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4296 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4297 $hours = floor( $seconds / 3600 );
4298 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4299 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4300 if ( $secondsPart == 60 ) {
4301 $secondsPart = 0;
4302 $minutes++;
4303 }
4304 if ( $minutes == 60 ) {
4305 $minutes = 0;
4306 $hours++;
4307 }
4308 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4309 $s .= ' ';
4310 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4311 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4312 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4313 }
4314 } else {
4315 $days = floor( $seconds / 86400 );
4316 if ( $format['avoid'] === 'avoidminutes' ) {
4317 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4318 if ( $hours == 24 ) {
4319 $hours = 0;
4320 $days++;
4321 }
4322 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4323 $s .= ' ';
4324 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4325 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4326 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4327 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4328 if ( $minutes == 60 ) {
4329 $minutes = 0;
4330 $hours++;
4331 }
4332 if ( $hours == 24 ) {
4333 $hours = 0;
4334 $days++;
4335 }
4336 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4337 $s .= ' ';
4338 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4339 $s .= ' ';
4340 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4341 } else {
4342 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4343 $s .= ' ';
4344 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4345 }
4346 }
4347 return $s;
4348 }
4349
4350 /**
4351 * Format a bitrate for output, using an appropriate
4352 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4353 *
4354 * This use base 1000. For base 1024 use formatSize(), for another base
4355 * see formatComputingNumbers()
4356 *
4357 * @param $bps int
4358 * @return string
4359 */
4360 function formatBitrate( $bps ) {
4361 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4362 }
4363
4364 /**
4365 * @param $size int Size of the unit
4366 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4367 * @param $messageKey string Message key to be uesd
4368 * @return string
4369 */
4370 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4371 if ( $size <= 0 ) {
4372 return str_replace( '$1', $this->formatNum( $size ),
4373 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4374 );
4375 }
4376 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4377 $index = 0;
4378
4379 $maxIndex = count( $sizes ) - 1;
4380 while ( $size >= $boundary && $index < $maxIndex ) {
4381 $index++;
4382 $size /= $boundary;
4383 }
4384
4385 // For small sizes no decimal places necessary
4386 $round = 0;
4387 if ( $index > 1 ) {
4388 // For MB and bigger two decimal places are smarter
4389 $round = 2;
4390 }
4391 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4392
4393 $size = round( $size, $round );
4394 $text = $this->getMessageFromDB( $msg );
4395 return str_replace( '$1', $this->formatNum( $size ), $text );
4396 }
4397
4398 /**
4399 * Format a size in bytes for output, using an appropriate
4400 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4401 *
4402 * This method use base 1024. For base 1000 use formatBitrate(), for
4403 * another base see formatComputingNumbers()
4404 *
4405 * @param $size int Size to format
4406 * @return string Plain text (not HTML)
4407 */
4408 function formatSize( $size ) {
4409 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4410 }
4411
4412 /**
4413 * Make a list item, used by various special pages
4414 *
4415 * @param $page String Page link
4416 * @param $details String Text between brackets
4417 * @param $oppositedm Boolean Add the direction mark opposite to your
4418 * language, to display text properly
4419 * @return String
4420 */
4421 function specialList( $page, $details, $oppositedm = true ) {
4422 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) .
4423 $this->getDirMark();
4424 $details = $details ? $dirmark . $this->getMessageFromDB( 'word-separator' ) .
4425 wfMessage( 'parentheses' )->rawParams( $details )->inLanguage( $this )->escaped() : '';
4426 return $page . $details;
4427 }
4428
4429 /**
4430 * Generate (prev x| next x) (20|50|100...) type links for paging
4431 *
4432 * @param $title Title object to link
4433 * @param $offset Integer offset parameter
4434 * @param $limit Integer limit parameter
4435 * @param $query array|String optional URL query parameter string
4436 * @param $atend Bool optional param for specified if this is the last page
4437 * @return String
4438 */
4439 public function viewPrevNext( Title $title, $offset, $limit, array $query = array(), $atend = false ) {
4440 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4441
4442 # Make 'previous' link
4443 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4444 if ( $offset > 0 ) {
4445 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4446 $query, $prev, 'prevn-title', 'mw-prevlink' );
4447 } else {
4448 $plink = htmlspecialchars( $prev );
4449 }
4450
4451 # Make 'next' link
4452 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4453 if ( $atend ) {
4454 $nlink = htmlspecialchars( $next );
4455 } else {
4456 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4457 $query, $next, 'prevn-title', 'mw-nextlink' );
4458 }
4459
4460 # Make links to set number of items per page
4461 $numLinks = array();
4462 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4463 $numLinks[] = $this->numLink( $title, $offset, $num,
4464 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4465 }
4466
4467 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4468 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4469 }
4470
4471 /**
4472 * Helper function for viewPrevNext() that generates links
4473 *
4474 * @param $title Title object to link
4475 * @param $offset Integer offset parameter
4476 * @param $limit Integer limit parameter
4477 * @param $query Array extra query parameters
4478 * @param $link String text to use for the link; will be escaped
4479 * @param $tooltipMsg String name of the message to use as tooltip
4480 * @param $class String value of the "class" attribute of the link
4481 * @return String HTML fragment
4482 */
4483 private function numLink( Title $title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4484 $query = array( 'limit' => $limit, 'offset' => $offset ) + $query;
4485 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4486 return Html::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4487 'title' => $tooltip, 'class' => $class ), $link );
4488 }
4489
4490 /**
4491 * Get the conversion rule title, if any.
4492 *
4493 * @return string
4494 */
4495 public function getConvRuleTitle() {
4496 return $this->mConverter->getConvRuleTitle();
4497 }
4498
4499 /**
4500 * Get the compiled plural rules for the language
4501 * @since 1.20
4502 * @return array Associative array with plural form, and plural rule as key-value pairs
4503 */
4504 public function getCompiledPluralRules() {
4505 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4506 $fallbacks = Language::getFallbacksFor( $this->mCode );
4507 if ( !$pluralRules ) {
4508 foreach ( $fallbacks as $fallbackCode ) {
4509 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4510 if ( $pluralRules ) {
4511 break;
4512 }
4513 }
4514 }
4515 return $pluralRules;
4516 }
4517
4518 /**
4519 * Get the plural rules for the language
4520 * @since 1.20
4521 * @return array Associative array with plural form number and plural rule as key-value pairs
4522 */
4523 public function getPluralRules() {
4524 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4525 $fallbacks = Language::getFallbacksFor( $this->mCode );
4526 if ( !$pluralRules ) {
4527 foreach ( $fallbacks as $fallbackCode ) {
4528 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4529 if ( $pluralRules ) {
4530 break;
4531 }
4532 }
4533 }
4534 return $pluralRules;
4535 }
4536
4537 /**
4538 * Get the plural rule types for the language
4539 * @since 1.22
4540 * @return array Associative array with plural form number and plural rule type as key-value pairs
4541 */
4542 public function getPluralRuleTypes() {
4543 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4544 $fallbacks = Language::getFallbacksFor( $this->mCode );
4545 if ( !$pluralRuleTypes ) {
4546 foreach ( $fallbacks as $fallbackCode ) {
4547 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4548 if ( $pluralRuleTypes ) {
4549 break;
4550 }
4551 }
4552 }
4553 return $pluralRuleTypes;
4554 }
4555
4556 /**
4557 * Find the index number of the plural rule appropriate for the given number
4558 * @return int The index number of the plural rule
4559 */
4560 public function getPluralRuleIndexNumber( $number ) {
4561 $pluralRules = $this->getCompiledPluralRules();
4562 $form = CLDRPluralRuleEvaluator::evaluateCompiled( $number, $pluralRules );
4563 return $form;
4564 }
4565
4566 /**
4567 * Find the plural rule type appropriate for the given number
4568 * For example, if the language is set to Arabic, getPluralType(5) should
4569 * return 'few'.
4570 * @since 1.22
4571 * @return string The name of the plural rule type, e.g. one, two, few, many
4572 */
4573 public function getPluralRuleType( $number ) {
4574 $index = $this->getPluralRuleIndexNumber( $number );
4575 $pluralRuleTypes = $this->getPluralRuleTypes();
4576 if ( isset( $pluralRuleTypes[$index] ) ) {
4577 return $pluralRuleTypes[$index];
4578 } else {
4579 return 'other';
4580 }
4581 }
4582 }
Wiklou, le Wiki du Wiklou