dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
resourceloader: Optimize module registry sent in the startup module
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * Internationalisation code.
4 *
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
19 *
20 * @file
21 * @ingroup Language
22 */
23
24 /**
25 * @defgroup Language Language
26 */
27
28 if ( !defined( 'MEDIAWIKI' ) ) {
29 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
30 exit( 1 );
31 }
32
33 if ( function_exists( 'mb_strtoupper' ) ) {
34 mb_internal_encoding( 'UTF-8' );
35 }
36
37 /**
38 * a fake language converter
39 *
40 * @ingroup Language
41 */
42 class FakeConverter {
43 /**
44 * @var Language
45 */
46 public $mLang;
47 function __construct( $langobj ) { $this->mLang = $langobj; }
48 function autoConvert( $text, $variant = false ) { return $text; }
49 function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
50 function convert( $t ) { return $t; }
51 function convertTo( $text, $variant ) { return $text; }
52 function convertTitle( $t ) { return $t->getPrefixedText(); }
53 function convertNamespace( $ns ) { return $this->mLang->getFormattedNsText( $ns ); }
54 function getVariants() { return array( $this->mLang->getCode() ); }
55 function getVariantFallbacks( $variant ) { return $this->mLang->getCode(); }
56 function getPreferredVariant() { return $this->mLang->getCode(); }
57 function getDefaultVariant() { return $this->mLang->getCode(); }
58 function getURLVariant() { return ''; }
59 function getConvRuleTitle() { return false; }
60 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
61 function getExtraHashOptions() { return ''; }
62 function getParsedTitle() { return ''; }
63 function markNoConversion( $text, $noParse = false ) { return $text; }
64 function convertCategoryKey( $key ) { return $key; }
65 /** @deprecated since 1.22 is no longer used */
66 function armourMath( $text ) { return $text; }
67 function validateVariant( $variant = null ) { return $variant === $this->mLang->getCode() ? $variant : null; }
68 function translate( $text, $variant ) { return $text; }
69 }
70
71 /**
72 * Internationalisation code
73 * @ingroup Language
74 */
75 class Language {
76
77 /**
78 * @var LanguageConverter
79 */
80 public $mConverter;
81
82 public $mVariants, $mCode, $mLoaded = false;
83 public $mMagicExtensions = array(), $mMagicHookDone = false;
84 private $mHtmlCode = null, $mParentLanguage = false;
85
86 public $dateFormatStrings = array();
87 public $mExtendedSpecialPageAliases;
88
89 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
90
91 /**
92 * ReplacementArray object caches
93 */
94 public $transformData = array();
95
96 /**
97 * @var LocalisationCache
98 */
99 static public $dataCache;
100
101 static public $mLangObjCache = array();
102
103 static public $mWeekdayMsgs = array(
104 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
105 'friday', 'saturday'
106 );
107
108 static public $mWeekdayAbbrevMsgs = array(
109 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
110 );
111
112 static public $mMonthMsgs = array(
113 'january', 'february', 'march', 'april', 'may_long', 'june',
114 'july', 'august', 'september', 'october', 'november',
115 'december'
116 );
117 static public $mMonthGenMsgs = array(
118 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
119 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
120 'december-gen'
121 );
122 static public $mMonthAbbrevMsgs = array(
123 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
124 'sep', 'oct', 'nov', 'dec'
125 );
126
127 static public $mIranianCalendarMonthMsgs = array(
128 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
129 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
130 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
131 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
132 );
133
134 static public $mHebrewCalendarMonthMsgs = array(
135 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
136 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
137 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
138 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
139 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
140 );
141
142 static public $mHebrewCalendarMonthGenMsgs = array(
143 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
144 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
145 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
146 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
147 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
148 );
149
150 static public $mHijriCalendarMonthMsgs = array(
151 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
152 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
153 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
154 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
155 );
156
157 /**
158 * @since 1.20
159 * @var array
160 */
161 static public $durationIntervals = array(
162 'millennia' => 31556952000,
163 'centuries' => 3155695200,
164 'decades' => 315569520,
165 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
166 'weeks' => 604800,
167 'days' => 86400,
168 'hours' => 3600,
169 'minutes' => 60,
170 'seconds' => 1,
171 );
172
173 /**
174 * Cache for language fallbacks.
175 * @see Language::getFallbacksIncludingSiteLanguage
176 * @since 1.21
177 * @var array
178 */
179 static private $fallbackLanguageCache = array();
180
181 /**
182 * Get a cached or new language object for a given language code
183 * @param $code String
184 * @return Language
185 */
186 static function factory( $code ) {
187 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
188
189 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
190 $code = $wgDummyLanguageCodes[$code];
191 }
192
193 // get the language object to process
194 $langObj = isset( self::$mLangObjCache[$code] )
195 ? self::$mLangObjCache[$code]
196 : self::newFromCode( $code );
197
198 // merge the language object in to get it up front in the cache
199 self::$mLangObjCache = array_merge( array( $code => $langObj ), self::$mLangObjCache );
200 // get rid of the oldest ones in case we have an overflow
201 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
202
203 return $langObj;
204 }
205
206 /**
207 * Create a language object for a given language code
208 * @param $code String
209 * @throws MWException
210 * @return Language
211 */
212 protected static function newFromCode( $code ) {
213 // Protect against path traversal below
214 if ( !Language::isValidCode( $code )
215 || strcspn( $code, ":/\\\000" ) !== strlen( $code )
216 ) {
217 throw new MWException( "Invalid language code \"$code\"" );
218 }
219
220 if ( !Language::isValidBuiltInCode( $code ) ) {
221 // It's not possible to customise this code with class files, so
222 // just return a Language object. This is to support uselang= hacks.
223 $lang = new Language;
224 $lang->setCode( $code );
225 return $lang;
226 }
227
228 // Check if there is a language class for the code
229 $class = self::classFromCode( $code );
230 self::preloadLanguageClass( $class );
231 if ( class_exists( $class ) ) {
232 $lang = new $class;
233 return $lang;
234 }
235
236 // Keep trying the fallback list until we find an existing class
237 $fallbacks = Language::getFallbacksFor( $code );
238 foreach ( $fallbacks as $fallbackCode ) {
239 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
240 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
241 }
242
243 $class = self::classFromCode( $fallbackCode );
244 self::preloadLanguageClass( $class );
245 if ( class_exists( $class ) ) {
246 $lang = Language::newFromCode( $fallbackCode );
247 $lang->setCode( $code );
248 return $lang;
249 }
250 }
251
252 throw new MWException( "Invalid fallback sequence for language '$code'" );
253 }
254
255 /**
256 * Checks whether any localisation is available for that language tag
257 * in MediaWiki (MessagesXx.php exists).
258 *
259 * @param string $code Language tag (in lower case)
260 * @return bool Whether language is supported
261 * @since 1.21
262 */
263 public static function isSupportedLanguage( $code ) {
264 return self::isValidBuiltInCode( $code )
265 && ( is_readable( self::getMessagesFileName( $code ) )
266 || is_readable( self::getJsonMessagesFileName( $code ) )
267 );
268 }
269
270 /**
271 * Returns true if a language code string is a well-formed language tag
272 * according to RFC 5646.
273 * This function only checks well-formedness; it doesn't check that
274 * language, script or variant codes actually exist in the repositories.
275 *
276 * Based on regexes by Mark Davis of the Unicode Consortium:
277 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
278 *
279 * @param $code string
280 * @param $lenient boolean Whether to allow '_' as separator. The default is only '-'.
281 *
282 * @return bool
283 * @since 1.21
284 */
285 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
286 $alpha = '[a-z]';
287 $digit = '[0-9]';
288 $alphanum = '[a-z0-9]';
289 $x = 'x'; # private use singleton
290 $singleton = '[a-wy-z]'; # other singleton
291 $s = $lenient ? '[-_]' : '-';
292
293 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
294 $script = "$alpha{4}"; # ISO 15924
295 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
296 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
297 $extension = "$singleton(?:$s$alphanum{2,8})+";
298 $privateUse = "$x(?:$s$alphanum{1,8})+";
299
300 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
301 # Since these are limited, this is safe even later changes to the registry --
302 # the only oddity is that it might change the type of the tag, and thus
303 # the results from the capturing groups.
304 # http://www.iana.org/assignments/language-subtag-registry
305
306 $grandfathered = "en{$s}GB{$s}oed"
307 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
308 . "|no{$s}(?:bok|nyn)"
309 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
310 . "|zh{$s}min{$s}nan";
311
312 $variantList = "$variant(?:$s$variant)*";
313 $extensionList = "$extension(?:$s$extension)*";
314
315 $langtag = "(?:($language)"
316 . "(?:$s$script)?"
317 . "(?:$s$region)?"
318 . "(?:$s$variantList)?"
319 . "(?:$s$extensionList)?"
320 . "(?:$s$privateUse)?)";
321
322 # The final breakdown, with capturing groups for each of these components
323 # The variants, extensions, grandfathered, and private-use may have interior '-'
324
325 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
326
327 return (bool)preg_match( "/$root/", strtolower( $code ) );
328 }
329
330 /**
331 * Returns true if a language code string is of a valid form, whether or
332 * not it exists. This includes codes which are used solely for
333 * customisation via the MediaWiki namespace.
334 *
335 * @param $code string
336 *
337 * @return bool
338 */
339 public static function isValidCode( $code ) {
340 static $cache = array();
341 if ( isset( $cache[$code] ) ) {
342 return $cache[$code];
343 }
344 // People think language codes are html safe, so enforce it.
345 // Ideally we should only allow a-zA-Z0-9-
346 // but, .+ and other chars are often used for {{int:}} hacks
347 // see bugs 37564, 37587, 36938
348 $cache[$code] =
349 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
350 && !preg_match( Title::getTitleInvalidRegex(), $code );
351
352 return $cache[$code];
353 }
354
355 /**
356 * Returns true if a language code is of a valid form for the purposes of
357 * internal customisation of MediaWiki, via Messages*.php or *.json.
358 *
359 * @param $code string
360 *
361 * @throws MWException
362 * @since 1.18
363 * @return bool
364 */
365 public static function isValidBuiltInCode( $code ) {
366
367 if ( !is_string( $code ) ) {
368 if ( is_object( $code ) ) {
369 $addmsg = " of class " . get_class( $code );
370 } else {
371 $addmsg = '';
372 }
373 $type = gettype( $code );
374 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
375 }
376
377 return (bool)preg_match( '/^[a-z0-9-]{2,}$/i', $code );
378 }
379
380 /**
381 * Returns true if a language code is an IETF tag known to MediaWiki.
382 *
383 * @param $code string
384 *
385 * @since 1.21
386 * @return bool
387 */
388 public static function isKnownLanguageTag( $tag ) {
389 static $coreLanguageNames;
390
391 // Quick escape for invalid input to avoid exceptions down the line
392 // when code tries to process tags which are not valid at all.
393 if ( !self::isValidBuiltInCode( $tag ) ) {
394 return false;
395 }
396
397 if ( $coreLanguageNames === null ) {
398 global $IP;
399 include "$IP/languages/Names.php";
400 }
401
402 if ( isset( $coreLanguageNames[$tag] )
403 || self::fetchLanguageName( $tag, $tag ) !== ''
404 ) {
405 return true;
406 }
407
408 return false;
409 }
410
411 /**
412 * @param $code
413 * @return String Name of the language class
414 */
415 public static function classFromCode( $code ) {
416 if ( $code == 'en' ) {
417 return 'Language';
418 } else {
419 return 'Language' . str_replace( '-', '_', ucfirst( $code ) );
420 }
421 }
422
423 /**
424 * Includes language class files
425 *
426 * @param $class string Name of the language class
427 */
428 public static function preloadLanguageClass( $class ) {
429 global $IP;
430
431 if ( $class === 'Language' ) {
432 return;
433 }
434
435 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
436 include_once "$IP/languages/classes/$class.php";
437 }
438 }
439
440 /**
441 * Get the LocalisationCache instance
442 *
443 * @return LocalisationCache
444 */
445 public static function getLocalisationCache() {
446 if ( is_null( self::$dataCache ) ) {
447 global $wgLocalisationCacheConf;
448 $class = $wgLocalisationCacheConf['class'];
449 self::$dataCache = new $class( $wgLocalisationCacheConf );
450 }
451 return self::$dataCache;
452 }
453
454 function __construct() {
455 $this->mConverter = new FakeConverter( $this );
456 // Set the code to the name of the descendant
457 if ( get_class( $this ) == 'Language' ) {
458 $this->mCode = 'en';
459 } else {
460 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
461 }
462 self::getLocalisationCache();
463 }
464
465 /**
466 * Reduce memory usage
467 */
468 function __destruct() {
469 foreach ( $this as $name => $value ) {
470 unset( $this->$name );
471 }
472 }
473
474 /**
475 * Hook which will be called if this is the content language.
476 * Descendants can use this to register hook functions or modify globals
477 */
478 function initContLang() { }
479
480 /**
481 * Same as getFallbacksFor for current language.
482 * @return array|bool
483 * @deprecated in 1.19
484 */
485 function getFallbackLanguageCode() {
486 wfDeprecated( __METHOD__, '1.19' );
487 return self::getFallbackFor( $this->mCode );
488 }
489
490 /**
491 * @return array
492 * @since 1.19
493 */
494 function getFallbackLanguages() {
495 return self::getFallbacksFor( $this->mCode );
496 }
497
498 /**
499 * Exports $wgBookstoreListEn
500 * @return array
501 */
502 function getBookstoreList() {
503 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
504 }
505
506 /**
507 * Returns an array of localised namespaces indexed by their numbers. If the namespace is not
508 * available in localised form, it will be included in English.
509 *
510 * @return array
511 */
512 public function getNamespaces() {
513 if ( is_null( $this->namespaceNames ) ) {
514 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
515
516 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
517 $validNamespaces = MWNamespace::getCanonicalNamespaces();
518
519 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
520
521 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
522 if ( $wgMetaNamespaceTalk ) {
523 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
524 } else {
525 $talk = $this->namespaceNames[NS_PROJECT_TALK];
526 $this->namespaceNames[NS_PROJECT_TALK] =
527 $this->fixVariableInNamespace( $talk );
528 }
529
530 # Sometimes a language will be localised but not actually exist on this wiki.
531 foreach ( $this->namespaceNames as $key => $text ) {
532 if ( !isset( $validNamespaces[$key] ) ) {
533 unset( $this->namespaceNames[$key] );
534 }
535 }
536
537 # The above mixing may leave namespaces out of canonical order.
538 # Re-order by namespace ID number...
539 ksort( $this->namespaceNames );
540
541 wfRunHooks( 'LanguageGetNamespaces', array( &$this->namespaceNames ) );
542 }
543 return $this->namespaceNames;
544 }
545
546 /**
547 * Arbitrarily set all of the namespace names at once. Mainly used for testing
548 * @param $namespaces Array of namespaces (id => name)
549 */
550 public function setNamespaces( array $namespaces ) {
551 $this->namespaceNames = $namespaces;
552 $this->mNamespaceIds = null;
553 }
554
555 /**
556 * Resets all of the namespace caches. Mainly used for testing
557 */
558 public function resetNamespaces() {
559 $this->namespaceNames = null;
560 $this->mNamespaceIds = null;
561 $this->namespaceAliases = null;
562 }
563
564 /**
565 * A convenience function that returns the same thing as
566 * getNamespaces() except with the array values changed to ' '
567 * where it found '_', useful for producing output to be displayed
568 * e.g. in <select> forms.
569 *
570 * @return array
571 */
572 function getFormattedNamespaces() {
573 $ns = $this->getNamespaces();
574 foreach ( $ns as $k => $v ) {
575 $ns[$k] = strtr( $v, '_', ' ' );
576 }
577 return $ns;
578 }
579
580 /**
581 * Get a namespace value by key
582 * <code>
583 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
584 * echo $mw_ns; // prints 'MediaWiki'
585 * </code>
586 *
587 * @param $index Int: the array key of the namespace to return
588 * @return mixed, string if the namespace value exists, otherwise false
589 */
590 function getNsText( $index ) {
591 $ns = $this->getNamespaces();
592 return isset( $ns[$index] ) ? $ns[$index] : false;
593 }
594
595 /**
596 * A convenience function that returns the same thing as
597 * getNsText() except with '_' changed to ' ', useful for
598 * producing output.
599 *
600 * <code>
601 * $mw_ns = $wgContLang->getFormattedNsText( NS_MEDIAWIKI_TALK );
602 * echo $mw_ns; // prints 'MediaWiki talk'
603 * </code>
604 *
605 * @param int $index The array key of the namespace to return
606 * @return string Namespace name without underscores (empty string if namespace does not exist)
607 */
608 function getFormattedNsText( $index ) {
609 $ns = $this->getNsText( $index );
610 return strtr( $ns, '_', ' ' );
611 }
612
613 /**
614 * Returns gender-dependent namespace alias if available.
615 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
616 * @param $index Int: namespace index
617 * @param $gender String: gender key (male, female... )
618 * @return String
619 * @since 1.18
620 */
621 function getGenderNsText( $index, $gender ) {
622 global $wgExtraGenderNamespaces;
623
624 $ns = $wgExtraGenderNamespaces + self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
625 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
626 }
627
628 /**
629 * Whether this language uses gender-dependent namespace aliases.
630 * See https://www.mediawiki.org/wiki/Manual:$wgExtraGenderNamespaces
631 * @return bool
632 * @since 1.18
633 */
634 function needsGenderDistinction() {
635 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
636 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
637 // $wgExtraGenderNamespaces overrides everything
638 return true;
639 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
640 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
641 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
642 return false;
643 } else {
644 // Check what is in i18n files
645 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
646 return count( $aliases ) > 0;
647 }
648 }
649
650 /**
651 * Get a namespace key by value, case insensitive.
652 * Only matches namespace names for the current language, not the
653 * canonical ones defined in Namespace.php.
654 *
655 * @param $text String
656 * @return mixed An integer if $text is a valid value otherwise false
657 */
658 function getLocalNsIndex( $text ) {
659 $lctext = $this->lc( $text );
660 $ids = $this->getNamespaceIds();
661 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
662 }
663
664 /**
665 * @return array
666 */
667 function getNamespaceAliases() {
668 if ( is_null( $this->namespaceAliases ) ) {
669 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
670 if ( !$aliases ) {
671 $aliases = array();
672 } else {
673 foreach ( $aliases as $name => $index ) {
674 if ( $index === NS_PROJECT_TALK ) {
675 unset( $aliases[$name] );
676 $name = $this->fixVariableInNamespace( $name );
677 $aliases[$name] = $index;
678 }
679 }
680 }
681
682 global $wgExtraGenderNamespaces;
683 $genders = $wgExtraGenderNamespaces + (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
684 foreach ( $genders as $index => $forms ) {
685 foreach ( $forms as $alias ) {
686 $aliases[$alias] = $index;
687 }
688 }
689
690 # Also add converted namespace names as aliases, to avoid confusion.
691 $convertedNames = array();
692 foreach ( $this->getVariants() as $variant ) {
693 if ( $variant === $this->mCode ) {
694 continue;
695 }
696 foreach ( $this->getNamespaces() as $ns => $_ ) {
697 $convertedNames[$this->getConverter()->convertNamespace( $ns, $variant )] = $ns;
698 }
699 }
700
701 $this->namespaceAliases = $aliases + $convertedNames;
702 }
703 return $this->namespaceAliases;
704 }
705
706 /**
707 * @return array
708 */
709 function getNamespaceIds() {
710 if ( is_null( $this->mNamespaceIds ) ) {
711 global $wgNamespaceAliases;
712 # Put namespace names and aliases into a hashtable.
713 # If this is too slow, then we should arrange it so that it is done
714 # before caching. The catch is that at pre-cache time, the above
715 # class-specific fixup hasn't been done.
716 $this->mNamespaceIds = array();
717 foreach ( $this->getNamespaces() as $index => $name ) {
718 $this->mNamespaceIds[$this->lc( $name )] = $index;
719 }
720 foreach ( $this->getNamespaceAliases() as $name => $index ) {
721 $this->mNamespaceIds[$this->lc( $name )] = $index;
722 }
723 if ( $wgNamespaceAliases ) {
724 foreach ( $wgNamespaceAliases as $name => $index ) {
725 $this->mNamespaceIds[$this->lc( $name )] = $index;
726 }
727 }
728 }
729 return $this->mNamespaceIds;
730 }
731
732 /**
733 * Get a namespace key by value, case insensitive. Canonical namespace
734 * names override custom ones defined for the current language.
735 *
736 * @param $text String
737 * @return mixed An integer if $text is a valid value otherwise false
738 */
739 function getNsIndex( $text ) {
740 $lctext = $this->lc( $text );
741 $ns = MWNamespace::getCanonicalIndex( $lctext );
742 if ( $ns !== null ) {
743 return $ns;
744 }
745 $ids = $this->getNamespaceIds();
746 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
747 }
748
749 /**
750 * short names for language variants used for language conversion links.
751 *
752 * @param $code String
753 * @param $usemsg bool Use the "variantname-xyz" message if it exists
754 * @return string
755 */
756 function getVariantname( $code, $usemsg = true ) {
757 $msg = "variantname-$code";
758 if ( $usemsg && wfMessage( $msg )->exists() ) {
759 return $this->getMessageFromDB( $msg );
760 }
761 $name = self::fetchLanguageName( $code );
762 if ( $name ) {
763 return $name; # if it's defined as a language name, show that
764 } else {
765 # otherwise, output the language code
766 return $code;
767 }
768 }
769
770 /**
771 * @param $name string
772 * @return string
773 */
774 function specialPage( $name ) {
775 $aliases = $this->getSpecialPageAliases();
776 if ( isset( $aliases[$name][0] ) ) {
777 $name = $aliases[$name][0];
778 }
779 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
780 }
781
782 /**
783 * @return array
784 */
785 function getDatePreferences() {
786 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
787 }
788
789 /**
790 * @return array
791 */
792 function getDateFormats() {
793 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
794 }
795
796 /**
797 * @return array|string
798 */
799 function getDefaultDateFormat() {
800 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
801 if ( $df === 'dmy or mdy' ) {
802 global $wgAmericanDates;
803 return $wgAmericanDates ? 'mdy' : 'dmy';
804 } else {
805 return $df;
806 }
807 }
808
809 /**
810 * @return array
811 */
812 function getDatePreferenceMigrationMap() {
813 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
814 }
815
816 /**
817 * @param $image
818 * @return array|null
819 */
820 function getImageFile( $image ) {
821 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
822 }
823
824 /**
825 * @return array
826 */
827 function getExtraUserToggles() {
828 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
829 }
830
831 /**
832 * @param $tog
833 * @return string
834 */
835 function getUserToggle( $tog ) {
836 return $this->getMessageFromDB( "tog-$tog" );
837 }
838
839 /**
840 * Get native language names, indexed by code.
841 * Only those defined in MediaWiki, no other data like CLDR.
842 * If $customisedOnly is true, only returns codes with a messages file
843 *
844 * @param $customisedOnly bool
845 *
846 * @return array
847 * @deprecated in 1.20, use fetchLanguageNames()
848 */
849 public static function getLanguageNames( $customisedOnly = false ) {
850 return self::fetchLanguageNames( null, $customisedOnly ? 'mwfile' : 'mw' );
851 }
852
853 /**
854 * Get translated language names. This is done on best effort and
855 * by default this is exactly the same as Language::getLanguageNames.
856 * The CLDR extension provides translated names.
857 * @param $code String Language code.
858 * @return Array language code => language name
859 * @since 1.18.0
860 * @deprecated in 1.20, use fetchLanguageNames()
861 */
862 public static function getTranslatedLanguageNames( $code ) {
863 return self::fetchLanguageNames( $code, 'all' );
864 }
865
866 /**
867 * Get an array of language names, indexed by code.
868 * @param $inLanguage null|string: Code of language in which to return the names
869 * Use null for autonyms (native names)
870 * @param $include string:
871 * 'all' all available languages
872 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
873 * 'mwfile' only if the language is in 'mw' *and* has a message file
874 * @return array: language code => language name
875 * @since 1.20
876 */
877 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
878 global $wgExtraLanguageNames;
879 static $coreLanguageNames;
880
881 if ( $coreLanguageNames === null ) {
882 global $IP;
883 include "$IP/languages/Names.php";
884 }
885
886 $names = array();
887
888 if ( $inLanguage ) {
889 # TODO: also include when $inLanguage is null, when this code is more efficient
890 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
891 }
892
893 $mwNames = $wgExtraLanguageNames + $coreLanguageNames;
894 foreach ( $mwNames as $mwCode => $mwName ) {
895 # - Prefer own MediaWiki native name when not using the hook
896 # - For other names just add if not added through the hook
897 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
898 $names[$mwCode] = $mwName;
899 }
900 }
901
902 if ( $include === 'all' ) {
903 return $names;
904 }
905
906 $returnMw = array();
907 $coreCodes = array_keys( $mwNames );
908 foreach ( $coreCodes as $coreCode ) {
909 $returnMw[$coreCode] = $names[$coreCode];
910 }
911
912 if ( $include === 'mwfile' ) {
913 $namesMwFile = array();
914 # We do this using a foreach over the codes instead of a directory
915 # loop so that messages files in extensions will work correctly.
916 foreach ( $returnMw as $code => $value ) {
917 if ( is_readable( self::getMessagesFileName( $code ) )
918 || is_readable( self::getJsonMessagesFileName( $code ) )
919 ) {
920 $namesMwFile[$code] = $names[$code];
921 }
922 }
923
924 return $namesMwFile;
925 }
926
927 # 'mw' option; default if it's not one of the other two options (all/mwfile)
928 return $returnMw;
929 }
930
931 /**
932 * @param $code string: The code of the language for which to get the name
933 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
934 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
935 * @return string: Language name or empty
936 * @since 1.20
937 */
938 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
939 $code = strtolower( $code );
940 $array = self::fetchLanguageNames( $inLanguage, $include );
941 return !array_key_exists( $code, $array ) ? '' : $array[$code];
942 }
943
944 /**
945 * Get a message from the MediaWiki namespace.
946 *
947 * @param $msg String: message name
948 * @return string
949 */
950 function getMessageFromDB( $msg ) {
951 return wfMessage( $msg )->inLanguage( $this )->text();
952 }
953
954 /**
955 * Get the native language name of $code.
956 * Only if defined in MediaWiki, no other data like CLDR.
957 * @param $code string
958 * @return string
959 * @deprecated in 1.20, use fetchLanguageName()
960 */
961 function getLanguageName( $code ) {
962 return self::fetchLanguageName( $code );
963 }
964
965 /**
966 * @param $key string
967 * @return string
968 */
969 function getMonthName( $key ) {
970 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
971 }
972
973 /**
974 * @return array
975 */
976 function getMonthNamesArray() {
977 $monthNames = array( '' );
978 for ( $i = 1; $i < 13; $i++ ) {
979 $monthNames[] = $this->getMonthName( $i );
980 }
981 return $monthNames;
982 }
983
984 /**
985 * @param $key string
986 * @return string
987 */
988 function getMonthNameGen( $key ) {
989 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
990 }
991
992 /**
993 * @param $key string
994 * @return string
995 */
996 function getMonthAbbreviation( $key ) {
997 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
998 }
999
1000 /**
1001 * @return array
1002 */
1003 function getMonthAbbreviationsArray() {
1004 $monthNames = array( '' );
1005 for ( $i = 1; $i < 13; $i++ ) {
1006 $monthNames[] = $this->getMonthAbbreviation( $i );
1007 }
1008 return $monthNames;
1009 }
1010
1011 /**
1012 * @param $key string
1013 * @return string
1014 */
1015 function getWeekdayName( $key ) {
1016 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
1017 }
1018
1019 /**
1020 * @param $key string
1021 * @return string
1022 */
1023 function getWeekdayAbbreviation( $key ) {
1024 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
1025 }
1026
1027 /**
1028 * @param $key string
1029 * @return string
1030 */
1031 function getIranianCalendarMonthName( $key ) {
1032 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
1033 }
1034
1035 /**
1036 * @param $key string
1037 * @return string
1038 */
1039 function getHebrewCalendarMonthName( $key ) {
1040 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
1041 }
1042
1043 /**
1044 * @param $key string
1045 * @return string
1046 */
1047 function getHebrewCalendarMonthNameGen( $key ) {
1048 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1049 }
1050
1051 /**
1052 * @param $key string
1053 * @return string
1054 */
1055 function getHijriCalendarMonthName( $key ) {
1056 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1057 }
1058
1059 /**
1060 * This is a workalike of PHP's date() function, but with better
1061 * internationalisation, a reduced set of format characters, and a better
1062 * escaping format.
1063 *
1064 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1065 * the PHP manual for definitions. There are a number of extensions, which
1066 * start with "x":
1067 *
1068 * xn Do not translate digits of the next numeric format character
1069 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1070 * xr Use roman numerals for the next numeric format character
1071 * xh Use hebrew numerals for the next numeric format character
1072 * xx Literal x
1073 * xg Genitive month name
1074 *
1075 * xij j (day number) in Iranian calendar
1076 * xiF F (month name) in Iranian calendar
1077 * xin n (month number) in Iranian calendar
1078 * xiy y (two digit year) in Iranian calendar
1079 * xiY Y (full year) in Iranian calendar
1080 *
1081 * xjj j (day number) in Hebrew calendar
1082 * xjF F (month name) in Hebrew calendar
1083 * xjt t (days in month) in Hebrew calendar
1084 * xjx xg (genitive month name) in Hebrew calendar
1085 * xjn n (month number) in Hebrew calendar
1086 * xjY Y (full year) in Hebrew calendar
1087 *
1088 * xmj j (day number) in Hijri calendar
1089 * xmF F (month name) in Hijri calendar
1090 * xmn n (month number) in Hijri calendar
1091 * xmY Y (full year) in Hijri calendar
1092 *
1093 * xkY Y (full year) in Thai solar calendar. Months and days are
1094 * identical to the Gregorian calendar
1095 * xoY Y (full year) in Minguo calendar or Juche year.
1096 * Months and days are identical to the
1097 * Gregorian calendar
1098 * xtY Y (full year) in Japanese nengo. Months and days are
1099 * identical to the Gregorian calendar
1100 *
1101 * Characters enclosed in double quotes will be considered literal (with
1102 * the quotes themselves removed). Unmatched quotes will be considered
1103 * literal quotes. Example:
1104 *
1105 * "The month is" F => The month is January
1106 * i's" => 20'11"
1107 *
1108 * Backslash escaping is also supported.
1109 *
1110 * Input timestamp is assumed to be pre-normalized to the desired local
1111 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1112 * $ts is UTC if $zone is not given.
1113 *
1114 * @param $format String
1115 * @param $ts String: 14-character timestamp
1116 * YYYYMMDDHHMMSS
1117 * 01234567890123
1118 * @param $zone DateTimeZone: Timezone of $ts
1119 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1120 *
1121 * @throws MWException
1122 * @return string
1123 */
1124 function sprintfDate( $format, $ts, DateTimeZone $zone = null ) {
1125 $s = '';
1126 $raw = false;
1127 $roman = false;
1128 $hebrewNum = false;
1129 $dateTimeObj = false;
1130 $rawToggle = false;
1131 $iranian = false;
1132 $hebrew = false;
1133 $hijri = false;
1134 $thai = false;
1135 $minguo = false;
1136 $tenno = false;
1137
1138 if ( strlen( $ts ) !== 14 ) {
1139 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1140 }
1141
1142 if ( !ctype_digit( $ts ) ) {
1143 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1144 }
1145
1146 for ( $p = 0; $p < strlen( $format ); $p++ ) {
1147 $num = false;
1148 $code = $format[$p];
1149 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
1150 $code .= $format[++$p];
1151 }
1152
1153 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
1154 $code .= $format[++$p];
1155 }
1156
1157 switch ( $code ) {
1158 case 'xx':
1159 $s .= 'x';
1160 break;
1161 case 'xn':
1162 $raw = true;
1163 break;
1164 case 'xN':
1165 $rawToggle = !$rawToggle;
1166 break;
1167 case 'xr':
1168 $roman = true;
1169 break;
1170 case 'xh':
1171 $hebrewNum = true;
1172 break;
1173 case 'xg':
1174 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1175 break;
1176 case 'xjx':
1177 if ( !$hebrew ) {
1178 $hebrew = self::tsToHebrew( $ts );
1179 }
1180 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1181 break;
1182 case 'd':
1183 $num = substr( $ts, 6, 2 );
1184 break;
1185 case 'D':
1186 if ( !$dateTimeObj ) {
1187 $dateTimeObj = DateTime::createFromFormat(
1188 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1189 );
1190 }
1191 $s .= $this->getWeekdayAbbreviation( $dateTimeObj->format( 'w' ) + 1 );
1192 break;
1193 case 'j':
1194 $num = intval( substr( $ts, 6, 2 ) );
1195 break;
1196 case 'xij':
1197 if ( !$iranian ) {
1198 $iranian = self::tsToIranian( $ts );
1199 }
1200 $num = $iranian[2];
1201 break;
1202 case 'xmj':
1203 if ( !$hijri ) {
1204 $hijri = self::tsToHijri( $ts );
1205 }
1206 $num = $hijri[2];
1207 break;
1208 case 'xjj':
1209 if ( !$hebrew ) {
1210 $hebrew = self::tsToHebrew( $ts );
1211 }
1212 $num = $hebrew[2];
1213 break;
1214 case 'l':
1215 if ( !$dateTimeObj ) {
1216 $dateTimeObj = DateTime::createFromFormat(
1217 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1218 );
1219 }
1220 $s .= $this->getWeekdayName( $dateTimeObj->format( 'w' ) + 1 );
1221 break;
1222 case 'F':
1223 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1224 break;
1225 case 'xiF':
1226 if ( !$iranian ) {
1227 $iranian = self::tsToIranian( $ts );
1228 }
1229 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1230 break;
1231 case 'xmF':
1232 if ( !$hijri ) {
1233 $hijri = self::tsToHijri( $ts );
1234 }
1235 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1236 break;
1237 case 'xjF':
1238 if ( !$hebrew ) {
1239 $hebrew = self::tsToHebrew( $ts );
1240 }
1241 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1242 break;
1243 case 'm':
1244 $num = substr( $ts, 4, 2 );
1245 break;
1246 case 'M':
1247 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1248 break;
1249 case 'n':
1250 $num = intval( substr( $ts, 4, 2 ) );
1251 break;
1252 case 'xin':
1253 if ( !$iranian ) {
1254 $iranian = self::tsToIranian( $ts );
1255 }
1256 $num = $iranian[1];
1257 break;
1258 case 'xmn':
1259 if ( !$hijri ) {
1260 $hijri = self::tsToHijri ( $ts );
1261 }
1262 $num = $hijri[1];
1263 break;
1264 case 'xjn':
1265 if ( !$hebrew ) {
1266 $hebrew = self::tsToHebrew( $ts );
1267 }
1268 $num = $hebrew[1];
1269 break;
1270 case 'xjt':
1271 if ( !$hebrew ) {
1272 $hebrew = self::tsToHebrew( $ts );
1273 }
1274 $num = $hebrew[3];
1275 break;
1276 case 'Y':
1277 $num = substr( $ts, 0, 4 );
1278 break;
1279 case 'xiY':
1280 if ( !$iranian ) {
1281 $iranian = self::tsToIranian( $ts );
1282 }
1283 $num = $iranian[0];
1284 break;
1285 case 'xmY':
1286 if ( !$hijri ) {
1287 $hijri = self::tsToHijri( $ts );
1288 }
1289 $num = $hijri[0];
1290 break;
1291 case 'xjY':
1292 if ( !$hebrew ) {
1293 $hebrew = self::tsToHebrew( $ts );
1294 }
1295 $num = $hebrew[0];
1296 break;
1297 case 'xkY':
1298 if ( !$thai ) {
1299 $thai = self::tsToYear( $ts, 'thai' );
1300 }
1301 $num = $thai[0];
1302 break;
1303 case 'xoY':
1304 if ( !$minguo ) {
1305 $minguo = self::tsToYear( $ts, 'minguo' );
1306 }
1307 $num = $minguo[0];
1308 break;
1309 case 'xtY':
1310 if ( !$tenno ) {
1311 $tenno = self::tsToYear( $ts, 'tenno' );
1312 }
1313 $num = $tenno[0];
1314 break;
1315 case 'y':
1316 $num = substr( $ts, 2, 2 );
1317 break;
1318 case 'xiy':
1319 if ( !$iranian ) {
1320 $iranian = self::tsToIranian( $ts );
1321 }
1322 $num = substr( $iranian[0], -2 );
1323 break;
1324 case 'a':
1325 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1326 break;
1327 case 'A':
1328 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1329 break;
1330 case 'g':
1331 $h = substr( $ts, 8, 2 );
1332 $num = $h % 12 ? $h % 12 : 12;
1333 break;
1334 case 'G':
1335 $num = intval( substr( $ts, 8, 2 ) );
1336 break;
1337 case 'h':
1338 $h = substr( $ts, 8, 2 );
1339 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1340 break;
1341 case 'H':
1342 $num = substr( $ts, 8, 2 );
1343 break;
1344 case 'i':
1345 $num = substr( $ts, 10, 2 );
1346 break;
1347 case 's':
1348 $num = substr( $ts, 12, 2 );
1349 break;
1350 case 'c':
1351 case 'r':
1352 case 'e':
1353 case 'O':
1354 case 'P':
1355 case 'T':
1356 // Pass through string from $dateTimeObj->format()
1357 if ( !$dateTimeObj ) {
1358 $dateTimeObj = DateTime::createFromFormat(
1359 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1360 );
1361 }
1362 $s .= $dateTimeObj->format( $code );
1363 break;
1364 case 'w':
1365 case 'N':
1366 case 'z':
1367 case 'W':
1368 case 't':
1369 case 'L':
1370 case 'o':
1371 case 'U':
1372 case 'I':
1373 case 'Z':
1374 // Pass through number from $dateTimeObj->format()
1375 if ( !$dateTimeObj ) {
1376 $dateTimeObj = DateTime::createFromFormat(
1377 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1378 );
1379 }
1380 $num = $dateTimeObj->format( $code );
1381 break;
1382 case '\\':
1383 # Backslash escaping
1384 if ( $p < strlen( $format ) - 1 ) {
1385 $s .= $format[++$p];
1386 } else {
1387 $s .= '\\';
1388 }
1389 break;
1390 case '"':
1391 # Quoted literal
1392 if ( $p < strlen( $format ) - 1 ) {
1393 $endQuote = strpos( $format, '"', $p + 1 );
1394 if ( $endQuote === false ) {
1395 # No terminating quote, assume literal "
1396 $s .= '"';
1397 } else {
1398 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1399 $p = $endQuote;
1400 }
1401 } else {
1402 # Quote at end of string, assume literal "
1403 $s .= '"';
1404 }
1405 break;
1406 default:
1407 $s .= $format[$p];
1408 }
1409 if ( $num !== false ) {
1410 if ( $rawToggle || $raw ) {
1411 $s .= $num;
1412 $raw = false;
1413 } elseif ( $roman ) {
1414 $s .= Language::romanNumeral( $num );
1415 $roman = false;
1416 } elseif ( $hebrewNum ) {
1417 $s .= self::hebrewNumeral( $num );
1418 $hebrewNum = false;
1419 } else {
1420 $s .= $this->formatNum( $num, true );
1421 }
1422 }
1423 }
1424 return $s;
1425 }
1426
1427 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1428 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1429
1430 /**
1431 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1432 * Gregorian dates to Iranian dates. Originally written in C, it
1433 * is released under the terms of GNU Lesser General Public
1434 * License. Conversion to PHP was performed by Niklas Laxström.
1435 *
1436 * Link: http://www.farsiweb.info/jalali/jalali.c
1437 *
1438 * @param $ts string
1439 *
1440 * @return string
1441 */
1442 private static function tsToIranian( $ts ) {
1443 $gy = substr( $ts, 0, 4 ) -1600;
1444 $gm = substr( $ts, 4, 2 ) -1;
1445 $gd = substr( $ts, 6, 2 ) -1;
1446
1447 # Days passed from the beginning (including leap years)
1448 $gDayNo = 365 * $gy
1449 + floor( ( $gy + 3 ) / 4 )
1450 - floor( ( $gy + 99 ) / 100 )
1451 + floor( ( $gy + 399 ) / 400 );
1452
1453 // Add days of the past months of this year
1454 for ( $i = 0; $i < $gm; $i++ ) {
1455 $gDayNo += self::$GREG_DAYS[$i];
1456 }
1457
1458 // Leap years
1459 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1460 $gDayNo++;
1461 }
1462
1463 // Days passed in current month
1464 $gDayNo += (int)$gd;
1465
1466 $jDayNo = $gDayNo - 79;
1467
1468 $jNp = floor( $jDayNo / 12053 );
1469 $jDayNo %= 12053;
1470
1471 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1472 $jDayNo %= 1461;
1473
1474 if ( $jDayNo >= 366 ) {
1475 $jy += floor( ( $jDayNo - 1 ) / 365 );
1476 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1477 }
1478
1479 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1480 $jDayNo -= self::$IRANIAN_DAYS[$i];
1481 }
1482
1483 $jm = $i + 1;
1484 $jd = $jDayNo + 1;
1485
1486 return array( $jy, $jm, $jd );
1487 }
1488
1489 /**
1490 * Converting Gregorian dates to Hijri dates.
1491 *
1492 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1493 *
1494 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1495 *
1496 * @param $ts string
1497 *
1498 * @return string
1499 */
1500 private static function tsToHijri( $ts ) {
1501 $year = substr( $ts, 0, 4 );
1502 $month = substr( $ts, 4, 2 );
1503 $day = substr( $ts, 6, 2 );
1504
1505 $zyr = $year;
1506 $zd = $day;
1507 $zm = $month;
1508 $zy = $zyr;
1509
1510 if (
1511 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1512 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1513 ) {
1514 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1515 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1516 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1517 $zd - 32075;
1518 } else {
1519 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1520 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1521 }
1522
1523 $zl = $zjd -1948440 + 10632;
1524 $zn = (int)( ( $zl - 1 ) / 10631 );
1525 $zl = $zl - 10631 * $zn + 354;
1526 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1527 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1528 $zm = (int)( ( 24 * $zl ) / 709 );
1529 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1530 $zy = 30 * $zn + $zj - 30;
1531
1532 return array( $zy, $zm, $zd );
1533 }
1534
1535 /**
1536 * Converting Gregorian dates to Hebrew dates.
1537 *
1538 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1539 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1540 * to translate the relevant functions into PHP and release them under
1541 * GNU GPL.
1542 *
1543 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1544 * and Adar II is 14. In a non-leap year, Adar is 6.
1545 *
1546 * @param $ts string
1547 *
1548 * @return string
1549 */
1550 private static function tsToHebrew( $ts ) {
1551 # Parse date
1552 $year = substr( $ts, 0, 4 );
1553 $month = substr( $ts, 4, 2 );
1554 $day = substr( $ts, 6, 2 );
1555
1556 # Calculate Hebrew year
1557 $hebrewYear = $year + 3760;
1558
1559 # Month number when September = 1, August = 12
1560 $month += 4;
1561 if ( $month > 12 ) {
1562 # Next year
1563 $month -= 12;
1564 $year++;
1565 $hebrewYear++;
1566 }
1567
1568 # Calculate day of year from 1 September
1569 $dayOfYear = $day;
1570 for ( $i = 1; $i < $month; $i++ ) {
1571 if ( $i == 6 ) {
1572 # February
1573 $dayOfYear += 28;
1574 # Check if the year is leap
1575 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1576 $dayOfYear++;
1577 }
1578 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1579 $dayOfYear += 30;
1580 } else {
1581 $dayOfYear += 31;
1582 }
1583 }
1584
1585 # Calculate the start of the Hebrew year
1586 $start = self::hebrewYearStart( $hebrewYear );
1587
1588 # Calculate next year's start
1589 if ( $dayOfYear <= $start ) {
1590 # Day is before the start of the year - it is the previous year
1591 # Next year's start
1592 $nextStart = $start;
1593 # Previous year
1594 $year--;
1595 $hebrewYear--;
1596 # Add days since previous year's 1 September
1597 $dayOfYear += 365;
1598 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1599 # Leap year
1600 $dayOfYear++;
1601 }
1602 # Start of the new (previous) year
1603 $start = self::hebrewYearStart( $hebrewYear );
1604 } else {
1605 # Next year's start
1606 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1607 }
1608
1609 # Calculate Hebrew day of year
1610 $hebrewDayOfYear = $dayOfYear - $start;
1611
1612 # Difference between year's days
1613 $diff = $nextStart - $start;
1614 # Add 12 (or 13 for leap years) days to ignore the difference between
1615 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1616 # difference is only about the year type
1617 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1618 $diff += 13;
1619 } else {
1620 $diff += 12;
1621 }
1622
1623 # Check the year pattern, and is leap year
1624 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1625 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1626 # and non-leap years
1627 $yearPattern = $diff % 30;
1628 # Check if leap year
1629 $isLeap = $diff >= 30;
1630
1631 # Calculate day in the month from number of day in the Hebrew year
1632 # Don't check Adar - if the day is not in Adar, we will stop before;
1633 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1634 $hebrewDay = $hebrewDayOfYear;
1635 $hebrewMonth = 1;
1636 $days = 0;
1637 while ( $hebrewMonth <= 12 ) {
1638 # Calculate days in this month
1639 if ( $isLeap && $hebrewMonth == 6 ) {
1640 # Adar in a leap year
1641 if ( $isLeap ) {
1642 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1643 $days = 30;
1644 if ( $hebrewDay <= $days ) {
1645 # Day in Adar I
1646 $hebrewMonth = 13;
1647 } else {
1648 # Subtract the days of Adar I
1649 $hebrewDay -= $days;
1650 # Try Adar II
1651 $days = 29;
1652 if ( $hebrewDay <= $days ) {
1653 # Day in Adar II
1654 $hebrewMonth = 14;
1655 }
1656 }
1657 }
1658 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1659 # Cheshvan in a complete year (otherwise as the rule below)
1660 $days = 30;
1661 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1662 # Kislev in an incomplete year (otherwise as the rule below)
1663 $days = 29;
1664 } else {
1665 # Odd months have 30 days, even have 29
1666 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1667 }
1668 if ( $hebrewDay <= $days ) {
1669 # In the current month
1670 break;
1671 } else {
1672 # Subtract the days of the current month
1673 $hebrewDay -= $days;
1674 # Try in the next month
1675 $hebrewMonth++;
1676 }
1677 }
1678
1679 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1680 }
1681
1682 /**
1683 * This calculates the Hebrew year start, as days since 1 September.
1684 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1685 * Used for Hebrew date.
1686 *
1687 * @param $year int
1688 *
1689 * @return string
1690 */
1691 private static function hebrewYearStart( $year ) {
1692 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1693 $b = intval( ( $year - 1 ) % 4 );
1694 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1695 if ( $m < 0 ) {
1696 $m--;
1697 }
1698 $Mar = intval( $m );
1699 if ( $m < 0 ) {
1700 $m++;
1701 }
1702 $m -= $Mar;
1703
1704 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1705 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1706 $Mar++;
1707 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1708 $Mar += 2;
1709 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1710 $Mar++;
1711 }
1712
1713 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1714 return $Mar;
1715 }
1716
1717 /**
1718 * Algorithm to convert Gregorian dates to Thai solar dates,
1719 * Minguo dates or Minguo dates.
1720 *
1721 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1722 * http://en.wikipedia.org/wiki/Minguo_calendar
1723 * http://en.wikipedia.org/wiki/Japanese_era_name
1724 *
1725 * @param $ts String: 14-character timestamp
1726 * @param $cName String: calender name
1727 * @return Array: converted year, month, day
1728 */
1729 private static function tsToYear( $ts, $cName ) {
1730 $gy = substr( $ts, 0, 4 );
1731 $gm = substr( $ts, 4, 2 );
1732 $gd = substr( $ts, 6, 2 );
1733
1734 if ( !strcmp( $cName, 'thai' ) ) {
1735 # Thai solar dates
1736 # Add 543 years to the Gregorian calendar
1737 # Months and days are identical
1738 $gy_offset = $gy + 543;
1739 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1740 # Minguo dates
1741 # Deduct 1911 years from the Gregorian calendar
1742 # Months and days are identical
1743 $gy_offset = $gy - 1911;
1744 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1745 # Nengō dates up to Meiji period
1746 # Deduct years from the Gregorian calendar
1747 # depending on the nengo periods
1748 # Months and days are identical
1749 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1750 # Meiji period
1751 $gy_gannen = $gy - 1868 + 1;
1752 $gy_offset = $gy_gannen;
1753 if ( $gy_gannen == 1 ) {
1754 $gy_offset = '元';
1755 }
1756 $gy_offset = '明治' . $gy_offset;
1757 } elseif (
1758 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1759 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1760 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1761 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1762 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1763 ) {
1764 # Taishō period
1765 $gy_gannen = $gy - 1912 + 1;
1766 $gy_offset = $gy_gannen;
1767 if ( $gy_gannen == 1 ) {
1768 $gy_offset = '元';
1769 }
1770 $gy_offset = '大正' . $gy_offset;
1771 } elseif (
1772 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1773 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1774 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1775 ) {
1776 # Shōwa period
1777 $gy_gannen = $gy - 1926 + 1;
1778 $gy_offset = $gy_gannen;
1779 if ( $gy_gannen == 1 ) {
1780 $gy_offset = '元';
1781 }
1782 $gy_offset = '昭和' . $gy_offset;
1783 } else {
1784 # Heisei period
1785 $gy_gannen = $gy - 1989 + 1;
1786 $gy_offset = $gy_gannen;
1787 if ( $gy_gannen == 1 ) {
1788 $gy_offset = '元';
1789 }
1790 $gy_offset = '平成' . $gy_offset;
1791 }
1792 } else {
1793 $gy_offset = $gy;
1794 }
1795
1796 return array( $gy_offset, $gm, $gd );
1797 }
1798
1799 /**
1800 * Roman number formatting up to 10000
1801 *
1802 * @param $num int
1803 *
1804 * @return string
1805 */
1806 static function romanNumeral( $num ) {
1807 static $table = array(
1808 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1809 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1810 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1811 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM', 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1812 );
1813
1814 $num = intval( $num );
1815 if ( $num > 10000 || $num <= 0 ) {
1816 return $num;
1817 }
1818
1819 $s = '';
1820 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1821 if ( $num >= $pow10 ) {
1822 $s .= $table[$i][(int)floor( $num / $pow10 )];
1823 }
1824 $num = $num % $pow10;
1825 }
1826 return $s;
1827 }
1828
1829 /**
1830 * Hebrew Gematria number formatting up to 9999
1831 *
1832 * @param $num int
1833 *
1834 * @return string
1835 */
1836 static function hebrewNumeral( $num ) {
1837 static $table = array(
1838 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1839 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1840 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1841 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1842 );
1843
1844 $num = intval( $num );
1845 if ( $num > 9999 || $num <= 0 ) {
1846 return $num;
1847 }
1848
1849 $s = '';
1850 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1851 if ( $num >= $pow10 ) {
1852 if ( $num == 15 || $num == 16 ) {
1853 $s .= $table[0][9] . $table[0][$num - 9];
1854 $num = 0;
1855 } else {
1856 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1857 if ( $pow10 == 1000 ) {
1858 $s .= "'";
1859 }
1860 }
1861 }
1862 $num = $num % $pow10;
1863 }
1864 if ( strlen( $s ) == 2 ) {
1865 $str = $s . "'";
1866 } else {
1867 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1868 $str .= substr( $s, strlen( $s ) - 2, 2 );
1869 }
1870 $start = substr( $str, 0, strlen( $str ) - 2 );
1871 $end = substr( $str, strlen( $str ) - 2 );
1872 switch ( $end ) {
1873 case 'כ':
1874 $str = $start . 'ך';
1875 break;
1876 case 'מ':
1877 $str = $start . 'ם';
1878 break;
1879 case 'נ':
1880 $str = $start . 'ן';
1881 break;
1882 case 'פ':
1883 $str = $start . 'ף';
1884 break;
1885 case 'צ':
1886 $str = $start . 'ץ';
1887 break;
1888 }
1889 return $str;
1890 }
1891
1892 /**
1893 * Used by date() and time() to adjust the time output.
1894 *
1895 * @param $ts Int the time in date('YmdHis') format
1896 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1897 * get user timecorrection setting)
1898 * @return int
1899 */
1900 function userAdjust( $ts, $tz = false ) {
1901 global $wgUser, $wgLocalTZoffset;
1902
1903 if ( $tz === false ) {
1904 $tz = $wgUser->getOption( 'timecorrection' );
1905 }
1906
1907 $data = explode( '|', $tz, 3 );
1908
1909 if ( $data[0] == 'ZoneInfo' ) {
1910 wfSuppressWarnings();
1911 $userTZ = timezone_open( $data[2] );
1912 wfRestoreWarnings();
1913 if ( $userTZ !== false ) {
1914 $date = date_create( $ts, timezone_open( 'UTC' ) );
1915 date_timezone_set( $date, $userTZ );
1916 $date = date_format( $date, 'YmdHis' );
1917 return $date;
1918 }
1919 # Unrecognized timezone, default to 'Offset' with the stored offset.
1920 $data[0] = 'Offset';
1921 }
1922
1923 $minDiff = 0;
1924 if ( $data[0] == 'System' || $tz == '' ) {
1925 #  Global offset in minutes.
1926 if ( isset( $wgLocalTZoffset ) ) {
1927 $minDiff = $wgLocalTZoffset;
1928 }
1929 } elseif ( $data[0] == 'Offset' ) {
1930 $minDiff = intval( $data[1] );
1931 } else {
1932 $data = explode( ':', $tz );
1933 if ( count( $data ) == 2 ) {
1934 $data[0] = intval( $data[0] );
1935 $data[1] = intval( $data[1] );
1936 $minDiff = abs( $data[0] ) * 60 + $data[1];
1937 if ( $data[0] < 0 ) {
1938 $minDiff = -$minDiff;
1939 }
1940 } else {
1941 $minDiff = intval( $data[0] ) * 60;
1942 }
1943 }
1944
1945 # No difference ? Return time unchanged
1946 if ( 0 == $minDiff ) {
1947 return $ts;
1948 }
1949
1950 wfSuppressWarnings(); // E_STRICT system time bitching
1951 # Generate an adjusted date; take advantage of the fact that mktime
1952 # will normalize out-of-range values so we don't have to split $minDiff
1953 # into hours and minutes.
1954 $t = mktime( (
1955 (int)substr( $ts, 8, 2 ) ), # Hours
1956 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
1957 (int)substr( $ts, 12, 2 ), # Seconds
1958 (int)substr( $ts, 4, 2 ), # Month
1959 (int)substr( $ts, 6, 2 ), # Day
1960 (int)substr( $ts, 0, 4 ) ); # Year
1961
1962 $date = date( 'YmdHis', $t );
1963 wfRestoreWarnings();
1964
1965 return $date;
1966 }
1967
1968 /**
1969 * This is meant to be used by time(), date(), and timeanddate() to get
1970 * the date preference they're supposed to use, it should be used in
1971 * all children.
1972 *
1973 *<code>
1974 * function timeanddate([...], $format = true) {
1975 * $datePreference = $this->dateFormat($format);
1976 * [...]
1977 * }
1978 *</code>
1979 *
1980 * @param $usePrefs Mixed: if true, the user's preference is used
1981 * if false, the site/language default is used
1982 * if int/string, assumed to be a format.
1983 * @return string
1984 */
1985 function dateFormat( $usePrefs = true ) {
1986 global $wgUser;
1987
1988 if ( is_bool( $usePrefs ) ) {
1989 if ( $usePrefs ) {
1990 $datePreference = $wgUser->getDatePreference();
1991 } else {
1992 $datePreference = (string)User::getDefaultOption( 'date' );
1993 }
1994 } else {
1995 $datePreference = (string)$usePrefs;
1996 }
1997
1998 // return int
1999 if ( $datePreference == '' ) {
2000 return 'default';
2001 }
2002
2003 return $datePreference;
2004 }
2005
2006 /**
2007 * Get a format string for a given type and preference
2008 * @param $type string May be date, time or both
2009 * @param $pref string The format name as it appears in Messages*.php
2010 *
2011 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
2012 *
2013 * @return string
2014 */
2015 function getDateFormatString( $type, $pref ) {
2016 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2017 if ( $pref == 'default' ) {
2018 $pref = $this->getDefaultDateFormat();
2019 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2020 } else {
2021 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2022
2023 if ( $type === 'pretty' && $df === null ) {
2024 $df = $this->getDateFormatString( 'date', $pref );
2025 }
2026
2027 if ( $df === null ) {
2028 $pref = $this->getDefaultDateFormat();
2029 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2030 }
2031 }
2032 $this->dateFormatStrings[$type][$pref] = $df;
2033 }
2034 return $this->dateFormatStrings[$type][$pref];
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 date( $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( 'date', $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: true to use user's date format preference
2062 * @param $timecorrection String|bool the time offset as returned by
2063 * validateTimeZone() in Special:Preferences
2064 * @return string
2065 */
2066 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2067 $ts = wfTimestamp( TS_MW, $ts );
2068 if ( $adj ) {
2069 $ts = $this->userAdjust( $ts, $timecorrection );
2070 }
2071 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2072 return $this->sprintfDate( $df, $ts );
2073 }
2074
2075 /**
2076 * @param $ts Mixed: the time format which needs to be turned into a
2077 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2078 * @param $adj Bool: whether to adjust the time output according to the
2079 * user configured offset ($timecorrection)
2080 * @param $format Mixed: what format to return, if it's false output the
2081 * default one (default true)
2082 * @param $timecorrection String|bool the time offset as returned by
2083 * validateTimeZone() in Special:Preferences
2084 * @return string
2085 */
2086 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2087 $ts = wfTimestamp( TS_MW, $ts );
2088 if ( $adj ) {
2089 $ts = $this->userAdjust( $ts, $timecorrection );
2090 }
2091 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2092 return $this->sprintfDate( $df, $ts );
2093 }
2094
2095 /**
2096 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2097 *
2098 * @since 1.20
2099 *
2100 * @param integer $seconds The amount of seconds.
2101 * @param array $chosenIntervals The intervals to enable.
2102 *
2103 * @return string
2104 */
2105 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2106 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2107
2108 $segments = array();
2109
2110 foreach ( $intervals as $intervalName => $intervalValue ) {
2111 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2112 // duration-years, duration-decades, duration-centuries, duration-millennia
2113 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2114 $segments[] = $message->inLanguage( $this )->escaped();
2115 }
2116
2117 return $this->listToText( $segments );
2118 }
2119
2120 /**
2121 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2122 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2123 *
2124 * @since 1.20
2125 *
2126 * @param integer $seconds The amount of seconds.
2127 * @param array $chosenIntervals The intervals to enable.
2128 *
2129 * @return array
2130 */
2131 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2132 if ( empty( $chosenIntervals ) ) {
2133 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
2134 }
2135
2136 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2137 $sortedNames = array_keys( $intervals );
2138 $smallestInterval = array_pop( $sortedNames );
2139
2140 $segments = array();
2141
2142 foreach ( $intervals as $name => $length ) {
2143 $value = floor( $seconds / $length );
2144
2145 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2146 $seconds -= $value * $length;
2147 $segments[$name] = $value;
2148 }
2149 }
2150
2151 return $segments;
2152 }
2153
2154 /**
2155 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2156 *
2157 * @param $type String: can be 'date', 'time' or 'both'
2158 * @param $ts Mixed: the time format which needs to be turned into a
2159 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2160 * @param $user User object used to get preferences for timezone and format
2161 * @param $options Array, can contain the following keys:
2162 * - 'timecorrection': time correction, can have the following values:
2163 * - true: use user's preference
2164 * - false: don't use time correction
2165 * - integer: value of time correction in minutes
2166 * - 'format': format to use, can have the following values:
2167 * - true: use user's preference
2168 * - false: use default preference
2169 * - string: format to use
2170 * @since 1.19
2171 * @return String
2172 */
2173 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2174 $ts = wfTimestamp( TS_MW, $ts );
2175 $options += array( 'timecorrection' => true, 'format' => true );
2176 if ( $options['timecorrection'] !== false ) {
2177 if ( $options['timecorrection'] === true ) {
2178 $offset = $user->getOption( 'timecorrection' );
2179 } else {
2180 $offset = $options['timecorrection'];
2181 }
2182 $ts = $this->userAdjust( $ts, $offset );
2183 }
2184 if ( $options['format'] === true ) {
2185 $format = $user->getDatePreference();
2186 } else {
2187 $format = $options['format'];
2188 }
2189 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2190 return $this->sprintfDate( $df, $ts );
2191 }
2192
2193 /**
2194 * Get the formatted date for the given timestamp and formatted for
2195 * the given user.
2196 *
2197 * @param $ts Mixed: the time format which needs to be turned into a
2198 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2199 * @param $user User object used to get preferences for timezone and format
2200 * @param $options Array, can contain the following keys:
2201 * - 'timecorrection': time correction, can have the following values:
2202 * - true: use user's preference
2203 * - false: don't use time correction
2204 * - integer: value of time correction in minutes
2205 * - 'format': format to use, can have the following values:
2206 * - true: use user's preference
2207 * - false: use default preference
2208 * - string: format to use
2209 * @since 1.19
2210 * @return String
2211 */
2212 public function userDate( $ts, User $user, array $options = array() ) {
2213 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2214 }
2215
2216 /**
2217 * Get the formatted time for the given timestamp and formatted for
2218 * the given user.
2219 *
2220 * @param $ts Mixed: the time format which needs to be turned into a
2221 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2222 * @param $user User object used to get preferences for timezone and format
2223 * @param $options Array, can contain the following keys:
2224 * - 'timecorrection': time correction, can have the following values:
2225 * - true: use user's preference
2226 * - false: don't use time correction
2227 * - integer: value of time correction in minutes
2228 * - 'format': format to use, can have the following values:
2229 * - true: use user's preference
2230 * - false: use default preference
2231 * - string: format to use
2232 * @since 1.19
2233 * @return String
2234 */
2235 public function userTime( $ts, User $user, array $options = array() ) {
2236 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2237 }
2238
2239 /**
2240 * Get the formatted date and time for the given timestamp and formatted for
2241 * the given user.
2242 *
2243 * @param $ts Mixed: the time format which needs to be turned into a
2244 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2245 * @param $user User object used to get preferences for timezone and format
2246 * @param $options Array, can contain the following keys:
2247 * - 'timecorrection': time correction, can have the following values:
2248 * - true: use user's preference
2249 * - false: don't use time correction
2250 * - integer: value of time correction in minutes
2251 * - 'format': format to use, can have the following values:
2252 * - true: use user's preference
2253 * - false: use default preference
2254 * - string: format to use
2255 * @since 1.19
2256 * @return String
2257 */
2258 public function userTimeAndDate( $ts, User $user, array $options = array() ) {
2259 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2260 }
2261
2262 /**
2263 * Convert an MWTimestamp into a pretty human-readable timestamp using
2264 * the given user preferences and relative base time.
2265 *
2266 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2267 * on your timestamp object, which will then call this function. Calling
2268 * this function directly will cause hooks to be skipped over.
2269 *
2270 * @see MWTimestamp::getHumanTimestamp
2271 * @param MWTimestamp $ts Timestamp to prettify
2272 * @param MWTimestamp $relativeTo Base timestamp
2273 * @param User $user User preferences to use
2274 * @return string Human timestamp
2275 * @since 1.22
2276 */
2277 public function getHumanTimestamp( MWTimestamp $ts, MWTimestamp $relativeTo, User $user ) {
2278 $diff = $ts->diff( $relativeTo );
2279 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) - (int)$relativeTo->timestamp->format( 'w' ) );
2280 $days = $diff->days ?: (int)$diffDay;
2281 if ( $diff->invert || $days > 5 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' ) ) {
2282 // Timestamps are in different years: use full timestamp
2283 // Also do full timestamp for future dates
2284 /**
2285 * @FIXME Add better handling of future timestamps.
2286 */
2287 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2288 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2289 } elseif ( $days > 5 ) {
2290 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2291 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2292 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2293 } elseif ( $days > 1 ) {
2294 // Timestamp within the past week: show the day of the week and time
2295 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2296 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2297 // Messages:
2298 // sunday-at, monday-at, tuesday-at, wednesday-at, thursday-at, friday-at, saturday-at
2299 $ts = wfMessage( "$weekday-at" )
2300 ->inLanguage( $this )
2301 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2302 ->text();
2303 } elseif ( $days == 1 ) {
2304 // Timestamp was yesterday: say 'yesterday' and the time.
2305 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2306 $ts = wfMessage( 'yesterday-at' )
2307 ->inLanguage( $this )
2308 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2309 ->text();
2310 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2311 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2312 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2313 $ts = wfMessage( 'today-at' )
2314 ->inLanguage( $this )
2315 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2316 ->text();
2317
2318 // From here on in, the timestamp was soon enough ago so that we can simply say
2319 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2320 } elseif ( $diff->h == 1 ) {
2321 // Less than 90 minutes, but more than an hour ago.
2322 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2323 } elseif ( $diff->i >= 1 ) {
2324 // A few minutes ago.
2325 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2326 } elseif ( $diff->s >= 30 ) {
2327 // Less than a minute, but more than 30 sec ago.
2328 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2329 } else {
2330 // Less than 30 seconds ago.
2331 $ts = wfMessage( 'just-now' )->text();
2332 }
2333
2334 return $ts;
2335 }
2336
2337 /**
2338 * @param $key string
2339 * @return array|null
2340 */
2341 function getMessage( $key ) {
2342 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2343 }
2344
2345 /**
2346 * @return array
2347 */
2348 function getAllMessages() {
2349 return self::$dataCache->getItem( $this->mCode, 'messages' );
2350 }
2351
2352 /**
2353 * @param $in
2354 * @param $out
2355 * @param $string
2356 * @return string
2357 */
2358 function iconv( $in, $out, $string ) {
2359 # This is a wrapper for iconv in all languages except esperanto,
2360 # which does some nasty x-conversions beforehand
2361
2362 # Even with //IGNORE iconv can whine about illegal characters in
2363 # *input* string. We just ignore those too.
2364 # REF: http://bugs.php.net/bug.php?id=37166
2365 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2366 wfSuppressWarnings();
2367 $text = iconv( $in, $out . '//IGNORE', $string );
2368 wfRestoreWarnings();
2369 return $text;
2370 }
2371
2372 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2373
2374 /**
2375 * @param $matches array
2376 * @return mixed|string
2377 */
2378 function ucwordbreaksCallbackAscii( $matches ) {
2379 return $this->ucfirst( $matches[1] );
2380 }
2381
2382 /**
2383 * @param $matches array
2384 * @return string
2385 */
2386 function ucwordbreaksCallbackMB( $matches ) {
2387 return mb_strtoupper( $matches[0] );
2388 }
2389
2390 /**
2391 * @param $matches array
2392 * @return string
2393 */
2394 function ucCallback( $matches ) {
2395 list( $wikiUpperChars ) = self::getCaseMaps();
2396 return strtr( $matches[1], $wikiUpperChars );
2397 }
2398
2399 /**
2400 * @param $matches array
2401 * @return string
2402 */
2403 function lcCallback( $matches ) {
2404 list( , $wikiLowerChars ) = self::getCaseMaps();
2405 return strtr( $matches[1], $wikiLowerChars );
2406 }
2407
2408 /**
2409 * @param $matches array
2410 * @return string
2411 */
2412 function ucwordsCallbackMB( $matches ) {
2413 return mb_strtoupper( $matches[0] );
2414 }
2415
2416 /**
2417 * @param $matches array
2418 * @return string
2419 */
2420 function ucwordsCallbackWiki( $matches ) {
2421 list( $wikiUpperChars ) = self::getCaseMaps();
2422 return strtr( $matches[0], $wikiUpperChars );
2423 }
2424
2425 /**
2426 * Make a string's first character uppercase
2427 *
2428 * @param $str string
2429 *
2430 * @return string
2431 */
2432 function ucfirst( $str ) {
2433 $o = ord( $str );
2434 if ( $o < 96 ) { // if already uppercase...
2435 return $str;
2436 } elseif ( $o < 128 ) {
2437 return ucfirst( $str ); // use PHP's ucfirst()
2438 } else {
2439 // fall back to more complex logic in case of multibyte strings
2440 return $this->uc( $str, true );
2441 }
2442 }
2443
2444 /**
2445 * Convert a string to uppercase
2446 *
2447 * @param $str string
2448 * @param $first bool
2449 *
2450 * @return string
2451 */
2452 function uc( $str, $first = false ) {
2453 if ( function_exists( 'mb_strtoupper' ) ) {
2454 if ( $first ) {
2455 if ( $this->isMultibyte( $str ) ) {
2456 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2457 } else {
2458 return ucfirst( $str );
2459 }
2460 } else {
2461 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2462 }
2463 } else {
2464 if ( $this->isMultibyte( $str ) ) {
2465 $x = $first ? '^' : '';
2466 return preg_replace_callback(
2467 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2468 array( $this, 'ucCallback' ),
2469 $str
2470 );
2471 } else {
2472 return $first ? ucfirst( $str ) : strtoupper( $str );
2473 }
2474 }
2475 }
2476
2477 /**
2478 * @param $str string
2479 * @return mixed|string
2480 */
2481 function lcfirst( $str ) {
2482 $o = ord( $str );
2483 if ( !$o ) {
2484 return strval( $str );
2485 } elseif ( $o >= 128 ) {
2486 return $this->lc( $str, true );
2487 } elseif ( $o > 96 ) {
2488 return $str;
2489 } else {
2490 $str[0] = strtolower( $str[0] );
2491 return $str;
2492 }
2493 }
2494
2495 /**
2496 * @param $str string
2497 * @param $first bool
2498 * @return mixed|string
2499 */
2500 function lc( $str, $first = false ) {
2501 if ( function_exists( 'mb_strtolower' ) ) {
2502 if ( $first ) {
2503 if ( $this->isMultibyte( $str ) ) {
2504 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2505 } else {
2506 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2507 }
2508 } else {
2509 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2510 }
2511 } else {
2512 if ( $this->isMultibyte( $str ) ) {
2513 $x = $first ? '^' : '';
2514 return preg_replace_callback(
2515 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2516 array( $this, 'lcCallback' ),
2517 $str
2518 );
2519 } else {
2520 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2521 }
2522 }
2523 }
2524
2525 /**
2526 * @param $str string
2527 * @return bool
2528 */
2529 function isMultibyte( $str ) {
2530 return (bool)preg_match( '/[\x80-\xff]/', $str );
2531 }
2532
2533 /**
2534 * @param $str string
2535 * @return mixed|string
2536 */
2537 function ucwords( $str ) {
2538 if ( $this->isMultibyte( $str ) ) {
2539 $str = $this->lc( $str );
2540
2541 // regexp to find first letter in each word (i.e. after each space)
2542 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2543
2544 // function to use to capitalize a single char
2545 if ( function_exists( 'mb_strtoupper' ) ) {
2546 return preg_replace_callback(
2547 $replaceRegexp,
2548 array( $this, 'ucwordsCallbackMB' ),
2549 $str
2550 );
2551 } else {
2552 return preg_replace_callback(
2553 $replaceRegexp,
2554 array( $this, 'ucwordsCallbackWiki' ),
2555 $str
2556 );
2557 }
2558 } else {
2559 return ucwords( strtolower( $str ) );
2560 }
2561 }
2562
2563 /**
2564 * capitalize words at word breaks
2565 *
2566 * @param $str string
2567 * @return mixed
2568 */
2569 function ucwordbreaks( $str ) {
2570 if ( $this->isMultibyte( $str ) ) {
2571 $str = $this->lc( $str );
2572
2573 // since \b doesn't work for UTF-8, we explicitely define word break chars
2574 $breaks = "[ \-\(\)\}\{\.,\?!]";
2575
2576 // find first letter after word break
2577 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2578
2579 if ( function_exists( 'mb_strtoupper' ) ) {
2580 return preg_replace_callback(
2581 $replaceRegexp,
2582 array( $this, 'ucwordbreaksCallbackMB' ),
2583 $str
2584 );
2585 } else {
2586 return preg_replace_callback(
2587 $replaceRegexp,
2588 array( $this, 'ucwordsCallbackWiki' ),
2589 $str
2590 );
2591 }
2592 } else {
2593 return preg_replace_callback(
2594 '/\b([\w\x80-\xff]+)\b/',
2595 array( $this, 'ucwordbreaksCallbackAscii' ),
2596 $str
2597 );
2598 }
2599 }
2600
2601 /**
2602 * Return a case-folded representation of $s
2603 *
2604 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2605 * and $s2 are the same except for the case of their characters. It is not
2606 * necessary for the value returned to make sense when displayed.
2607 *
2608 * Do *not* perform any other normalisation in this function. If a caller
2609 * uses this function when it should be using a more general normalisation
2610 * function, then fix the caller.
2611 *
2612 * @param $s string
2613 *
2614 * @return string
2615 */
2616 function caseFold( $s ) {
2617 return $this->uc( $s );
2618 }
2619
2620 /**
2621 * @param $s string
2622 * @return string
2623 */
2624 function checkTitleEncoding( $s ) {
2625 if ( is_array( $s ) ) {
2626 throw new MWException( 'Given array to checkTitleEncoding.' );
2627 }
2628 if ( StringUtils::isUtf8( $s ) ) {
2629 return $s;
2630 }
2631
2632 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2633 }
2634
2635 /**
2636 * @return array
2637 */
2638 function fallback8bitEncoding() {
2639 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2640 }
2641
2642 /**
2643 * Most writing systems use whitespace to break up words.
2644 * Some languages such as Chinese don't conventionally do this,
2645 * which requires special handling when breaking up words for
2646 * searching etc.
2647 *
2648 * @return bool
2649 */
2650 function hasWordBreaks() {
2651 return true;
2652 }
2653
2654 /**
2655 * Some languages such as Chinese require word segmentation,
2656 * Specify such segmentation when overridden in derived class.
2657 *
2658 * @param $string String
2659 * @return String
2660 */
2661 function segmentByWord( $string ) {
2662 return $string;
2663 }
2664
2665 /**
2666 * Some languages have special punctuation need to be normalized.
2667 * Make such changes here.
2668 *
2669 * @param $string String
2670 * @return String
2671 */
2672 function normalizeForSearch( $string ) {
2673 return self::convertDoubleWidth( $string );
2674 }
2675
2676 /**
2677 * convert double-width roman characters to single-width.
2678 * range: ff00-ff5f ~= 0020-007f
2679 *
2680 * @param $string string
2681 *
2682 * @return string
2683 */
2684 protected static function convertDoubleWidth( $string ) {
2685 static $full = null;
2686 static $half = null;
2687
2688 if ( $full === null ) {
2689 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2690 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2691 $full = str_split( $fullWidth, 3 );
2692 $half = str_split( $halfWidth );
2693 }
2694
2695 $string = str_replace( $full, $half, $string );
2696 return $string;
2697 }
2698
2699 /**
2700 * @param $string string
2701 * @param $pattern string
2702 * @return string
2703 */
2704 protected static function insertSpace( $string, $pattern ) {
2705 $string = preg_replace( $pattern, " $1 ", $string );
2706 $string = preg_replace( '/ +/', ' ', $string );
2707 return $string;
2708 }
2709
2710 /**
2711 * @param $termsArray array
2712 * @return array
2713 */
2714 function convertForSearchResult( $termsArray ) {
2715 # some languages, e.g. Chinese, need to do a conversion
2716 # in order for search results to be displayed correctly
2717 return $termsArray;
2718 }
2719
2720 /**
2721 * Get the first character of a string.
2722 *
2723 * @param $s string
2724 * @return string
2725 */
2726 function firstChar( $s ) {
2727 $matches = array();
2728 preg_match(
2729 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2730 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2731 $s,
2732 $matches
2733 );
2734
2735 if ( isset( $matches[1] ) ) {
2736 if ( strlen( $matches[1] ) != 3 ) {
2737 return $matches[1];
2738 }
2739
2740 // Break down Hangul syllables to grab the first jamo
2741 $code = utf8ToCodepoint( $matches[1] );
2742 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2743 return $matches[1];
2744 } elseif ( $code < 0xb098 ) {
2745 return "\xe3\x84\xb1";
2746 } elseif ( $code < 0xb2e4 ) {
2747 return "\xe3\x84\xb4";
2748 } elseif ( $code < 0xb77c ) {
2749 return "\xe3\x84\xb7";
2750 } elseif ( $code < 0xb9c8 ) {
2751 return "\xe3\x84\xb9";
2752 } elseif ( $code < 0xbc14 ) {
2753 return "\xe3\x85\x81";
2754 } elseif ( $code < 0xc0ac ) {
2755 return "\xe3\x85\x82";
2756 } elseif ( $code < 0xc544 ) {
2757 return "\xe3\x85\x85";
2758 } elseif ( $code < 0xc790 ) {
2759 return "\xe3\x85\x87";
2760 } elseif ( $code < 0xcc28 ) {
2761 return "\xe3\x85\x88";
2762 } elseif ( $code < 0xce74 ) {
2763 return "\xe3\x85\x8a";
2764 } elseif ( $code < 0xd0c0 ) {
2765 return "\xe3\x85\x8b";
2766 } elseif ( $code < 0xd30c ) {
2767 return "\xe3\x85\x8c";
2768 } elseif ( $code < 0xd558 ) {
2769 return "\xe3\x85\x8d";
2770 } else {
2771 return "\xe3\x85\x8e";
2772 }
2773 } else {
2774 return '';
2775 }
2776 }
2777
2778 function initEncoding() {
2779 # Some languages may have an alternate char encoding option
2780 # (Esperanto X-coding, Japanese furigana conversion, etc)
2781 # If this language is used as the primary content language,
2782 # an override to the defaults can be set here on startup.
2783 }
2784
2785 /**
2786 * @param $s string
2787 * @return string
2788 */
2789 function recodeForEdit( $s ) {
2790 # For some languages we'll want to explicitly specify
2791 # which characters make it into the edit box raw
2792 # or are converted in some way or another.
2793 global $wgEditEncoding;
2794 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
2795 return $s;
2796 } else {
2797 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2798 }
2799 }
2800
2801 /**
2802 * @param $s string
2803 * @return string
2804 */
2805 function recodeInput( $s ) {
2806 # Take the previous into account.
2807 global $wgEditEncoding;
2808 if ( $wgEditEncoding != '' ) {
2809 $enc = $wgEditEncoding;
2810 } else {
2811 $enc = 'UTF-8';
2812 }
2813 if ( $enc == 'UTF-8' ) {
2814 return $s;
2815 } else {
2816 return $this->iconv( $enc, 'UTF-8', $s );
2817 }
2818 }
2819
2820 /**
2821 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2822 * also cleans up certain backwards-compatible sequences, converting them
2823 * to the modern Unicode equivalent.
2824 *
2825 * This is language-specific for performance reasons only.
2826 *
2827 * @param $s string
2828 *
2829 * @return string
2830 */
2831 function normalize( $s ) {
2832 global $wgAllUnicodeFixes;
2833 $s = UtfNormal::cleanUp( $s );
2834 if ( $wgAllUnicodeFixes ) {
2835 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2836 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2837 }
2838
2839 return $s;
2840 }
2841
2842 /**
2843 * Transform a string using serialized data stored in the given file (which
2844 * must be in the serialized subdirectory of $IP). The file contains pairs
2845 * mapping source characters to destination characters.
2846 *
2847 * The data is cached in process memory. This will go faster if you have the
2848 * FastStringSearch extension.
2849 *
2850 * @param $file string
2851 * @param $string string
2852 *
2853 * @throws MWException
2854 * @return string
2855 */
2856 function transformUsingPairFile( $file, $string ) {
2857 if ( !isset( $this->transformData[$file] ) ) {
2858 $data = wfGetPrecompiledData( $file );
2859 if ( $data === false ) {
2860 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2861 }
2862 $this->transformData[$file] = new ReplacementArray( $data );
2863 }
2864 return $this->transformData[$file]->replace( $string );
2865 }
2866
2867 /**
2868 * For right-to-left language support
2869 *
2870 * @return bool
2871 */
2872 function isRTL() {
2873 return self::$dataCache->getItem( $this->mCode, 'rtl' );
2874 }
2875
2876 /**
2877 * Return the correct HTML 'dir' attribute value for this language.
2878 * @return String
2879 */
2880 function getDir() {
2881 return $this->isRTL() ? 'rtl' : 'ltr';
2882 }
2883
2884 /**
2885 * Return 'left' or 'right' as appropriate alignment for line-start
2886 * for this language's text direction.
2887 *
2888 * Should be equivalent to CSS3 'start' text-align value....
2889 *
2890 * @return String
2891 */
2892 function alignStart() {
2893 return $this->isRTL() ? 'right' : 'left';
2894 }
2895
2896 /**
2897 * Return 'right' or 'left' as appropriate alignment for line-end
2898 * for this language's text direction.
2899 *
2900 * Should be equivalent to CSS3 'end' text-align value....
2901 *
2902 * @return String
2903 */
2904 function alignEnd() {
2905 return $this->isRTL() ? 'left' : 'right';
2906 }
2907
2908 /**
2909 * A hidden direction mark (LRM or RLM), depending on the language direction.
2910 * Unlike getDirMark(), this function returns the character as an HTML entity.
2911 * This function should be used when the output is guaranteed to be HTML,
2912 * because it makes the output HTML source code more readable. When
2913 * the output is plain text or can be escaped, getDirMark() should be used.
2914 *
2915 * @param $opposite Boolean Get the direction mark opposite to your language
2916 * @return string
2917 * @since 1.20
2918 */
2919 function getDirMarkEntity( $opposite = false ) {
2920 if ( $opposite ) {
2921 return $this->isRTL() ? '&lrm;' : '&rlm;';
2922 }
2923 return $this->isRTL() ? '&rlm;' : '&lrm;';
2924 }
2925
2926 /**
2927 * A hidden direction mark (LRM or RLM), depending on the language direction.
2928 * This function produces them as invisible Unicode characters and
2929 * the output may be hard to read and debug, so it should only be used
2930 * when the output is plain text or can be escaped. When the output is
2931 * HTML, use getDirMarkEntity() instead.
2932 *
2933 * @param $opposite Boolean Get the direction mark opposite to your language
2934 * @return string
2935 */
2936 function getDirMark( $opposite = false ) {
2937 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2938 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2939 if ( $opposite ) {
2940 return $this->isRTL() ? $lrm : $rlm;
2941 }
2942 return $this->isRTL() ? $rlm : $lrm;
2943 }
2944
2945 /**
2946 * @return array
2947 */
2948 function capitalizeAllNouns() {
2949 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2950 }
2951
2952 /**
2953 * An arrow, depending on the language direction.
2954 *
2955 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2956 * @return string
2957 */
2958 function getArrow( $direction = 'forwards' ) {
2959 switch ( $direction ) {
2960 case 'forwards':
2961 return $this->isRTL() ? '←' : '→';
2962 case 'backwards':
2963 return $this->isRTL() ? '→' : '←';
2964 case 'left':
2965 return '←';
2966 case 'right':
2967 return '→';
2968 case 'up':
2969 return '↑';
2970 case 'down':
2971 return '↓';
2972 }
2973 }
2974
2975 /**
2976 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2977 *
2978 * @return bool
2979 */
2980 function linkPrefixExtension() {
2981 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2982 }
2983
2984 /**
2985 * Get all magic words from cache.
2986 * @return array
2987 */
2988 function getMagicWords() {
2989 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2990 }
2991
2992 /**
2993 * Run the LanguageGetMagic hook once.
2994 */
2995 protected function doMagicHook() {
2996 if ( $this->mMagicHookDone ) {
2997 return;
2998 }
2999 $this->mMagicHookDone = true;
3000 wfProfileIn( 'LanguageGetMagic' );
3001 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
3002 wfProfileOut( 'LanguageGetMagic' );
3003 }
3004
3005 /**
3006 * Fill a MagicWord object with data from here
3007 *
3008 * @param $mw
3009 */
3010 function getMagic( $mw ) {
3011 // Saves a function call
3012 if ( ! $this->mMagicHookDone ) {
3013 $this->doMagicHook();
3014 }
3015
3016 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3017 $rawEntry = $this->mMagicExtensions[$mw->mId];
3018 } else {
3019 $rawEntry = self::$dataCache->getSubitem(
3020 $this->mCode, 'magicWords', $mw->mId );
3021 }
3022
3023 if ( !is_array( $rawEntry ) ) {
3024 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3025 } else {
3026 $mw->mCaseSensitive = $rawEntry[0];
3027 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3028 }
3029 }
3030
3031 /**
3032 * Add magic words to the extension array
3033 *
3034 * @param $newWords array
3035 */
3036 function addMagicWordsByLang( $newWords ) {
3037 $fallbackChain = $this->getFallbackLanguages();
3038 $fallbackChain = array_reverse( $fallbackChain );
3039 foreach ( $fallbackChain as $code ) {
3040 if ( isset( $newWords[$code] ) ) {
3041 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3042 }
3043 }
3044 }
3045
3046 /**
3047 * Get special page names, as an associative array
3048 * case folded alias => real name
3049 */
3050 function getSpecialPageAliases() {
3051 // Cache aliases because it may be slow to load them
3052 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3053 // Initialise array
3054 $this->mExtendedSpecialPageAliases =
3055 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3056 wfRunHooks( 'LanguageGetSpecialPageAliases',
3057 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
3058 }
3059
3060 return $this->mExtendedSpecialPageAliases;
3061 }
3062
3063 /**
3064 * Italic is unsuitable for some languages
3065 *
3066 * @param $text String: the text to be emphasized.
3067 * @return string
3068 */
3069 function emphasize( $text ) {
3070 return "<em>$text</em>";
3071 }
3072
3073 /**
3074 * Normally we output all numbers in plain en_US style, that is
3075 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3076 * point twohundredthirtyfive. However this is not suitable for all
3077 * languages, some such as Punjabi want ੨੯੩,੨੯੫.੨੩੫ and others such as
3078 * Icelandic just want to use commas instead of dots, and dots instead
3079 * of commas like "293.291,235".
3080 *
3081 * An example of this function being called:
3082 * <code>
3083 * wfMessage( 'message' )->numParams( $num )->text()
3084 * </code>
3085 *
3086 * See $separatorTransformTable on MessageIs.php for
3087 * the , => . and . => , implementation.
3088 *
3089 * @todo check if it's viable to use localeconv() for the decimal
3090 * separator thing.
3091 * @param $number Mixed: the string to be formatted, should be an integer
3092 * or a floating point number.
3093 * @param $nocommafy Bool: set to true for special numbers like dates
3094 * @return string
3095 */
3096 public function formatNum( $number, $nocommafy = false ) {
3097 global $wgTranslateNumerals;
3098 if ( !$nocommafy ) {
3099 $number = $this->commafy( $number );
3100 $s = $this->separatorTransformTable();
3101 if ( $s ) {
3102 $number = strtr( $number, $s );
3103 }
3104 }
3105
3106 if ( $wgTranslateNumerals ) {
3107 $s = $this->digitTransformTable();
3108 if ( $s ) {
3109 $number = strtr( $number, $s );
3110 }
3111 }
3112
3113 return $number;
3114 }
3115
3116 /**
3117 * Front-end for non-commafied formatNum
3118 *
3119 * @param mixed $number the string to be formatted, should be an integer
3120 * or a floating point number.
3121 * @since 1.21
3122 * @return string
3123 */
3124 public function formatNumNoSeparators( $number ) {
3125 return $this->formatNum( $number, true );
3126 }
3127
3128 /**
3129 * @param $number string
3130 * @return string
3131 */
3132 function parseFormattedNumber( $number ) {
3133 $s = $this->digitTransformTable();
3134 if ( $s ) {
3135 $number = strtr( $number, array_flip( $s ) );
3136 }
3137
3138 $s = $this->separatorTransformTable();
3139 if ( $s ) {
3140 $number = strtr( $number, array_flip( $s ) );
3141 }
3142
3143 $number = strtr( $number, array( ',' => '' ) );
3144 return $number;
3145 }
3146
3147 /**
3148 * Adds commas to a given number
3149 * @since 1.19
3150 * @param $number mixed
3151 * @return string
3152 */
3153 function commafy( $number ) {
3154 $digitGroupingPattern = $this->digitGroupingPattern();
3155 if ( $number === null ) {
3156 return '';
3157 }
3158
3159 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3160 // default grouping is at thousands, use the same for ###,###,### pattern too.
3161 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3162 } else {
3163 // Ref: http://cldr.unicode.org/translation/number-patterns
3164 $sign = "";
3165 if ( intval( $number ) < 0 ) {
3166 // For negative numbers apply the algorithm like positive number and add sign.
3167 $sign = "-";
3168 $number = substr( $number, 1 );
3169 }
3170 $integerPart = array();
3171 $decimalPart = array();
3172 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3173 preg_match( "/\d+/", $number, $integerPart );
3174 preg_match( "/\.\d*/", $number, $decimalPart );
3175 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3176 if ( $groupedNumber === $number ) {
3177 // the string does not have any number part. Eg: .12345
3178 return $sign . $groupedNumber;
3179 }
3180 $start = $end = strlen( $integerPart[0] );
3181 while ( $start > 0 ) {
3182 $match = $matches[0][$numMatches - 1];
3183 $matchLen = strlen( $match );
3184 $start = $end - $matchLen;
3185 if ( $start < 0 ) {
3186 $start = 0;
3187 }
3188 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3189 $end = $start;
3190 if ( $numMatches > 1 ) {
3191 // use the last pattern for the rest of the number
3192 $numMatches--;
3193 }
3194 if ( $start > 0 ) {
3195 $groupedNumber = "," . $groupedNumber;
3196 }
3197 }
3198 return $sign . $groupedNumber;
3199 }
3200 }
3201
3202 /**
3203 * @return String
3204 */
3205 function digitGroupingPattern() {
3206 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3207 }
3208
3209 /**
3210 * @return array
3211 */
3212 function digitTransformTable() {
3213 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3214 }
3215
3216 /**
3217 * @return array
3218 */
3219 function separatorTransformTable() {
3220 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3221 }
3222
3223 /**
3224 * Take a list of strings and build a locale-friendly comma-separated
3225 * list, using the local comma-separator message.
3226 * The last two strings are chained with an "and".
3227 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3228 *
3229 * @param $l Array
3230 * @return string
3231 */
3232 function listToText( array $l ) {
3233 $m = count( $l ) - 1;
3234 if ( $m < 0 ) {
3235 return '';
3236 }
3237 if ( $m > 0 ) {
3238 $and = $this->getMessageFromDB( 'and' );
3239 $space = $this->getMessageFromDB( 'word-separator' );
3240 if ( $m > 1 ) {
3241 $comma = $this->getMessageFromDB( 'comma-separator' );
3242 }
3243 }
3244 $s = $l[$m];
3245 for ( $i = $m - 1; $i >= 0; $i-- ) {
3246 if ( $i == $m - 1 ) {
3247 $s = $l[$i] . $and . $space . $s;
3248 } else {
3249 $s = $l[$i] . $comma . $s;
3250 }
3251 }
3252 return $s;
3253 }
3254
3255 /**
3256 * Take a list of strings and build a locale-friendly comma-separated
3257 * list, using the local comma-separator message.
3258 * @param $list array of strings to put in a comma list
3259 * @return string
3260 */
3261 function commaList( array $list ) {
3262 return implode(
3263 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3264 $list
3265 );
3266 }
3267
3268 /**
3269 * Take a list of strings and build a locale-friendly semicolon-separated
3270 * list, using the local semicolon-separator message.
3271 * @param $list array of strings to put in a semicolon list
3272 * @return string
3273 */
3274 function semicolonList( array $list ) {
3275 return implode(
3276 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3277 $list
3278 );
3279 }
3280
3281 /**
3282 * Same as commaList, but separate it with the pipe instead.
3283 * @param $list array of strings to put in a pipe list
3284 * @return string
3285 */
3286 function pipeList( array $list ) {
3287 return implode(
3288 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3289 $list
3290 );
3291 }
3292
3293 /**
3294 * Truncate a string to a specified length in bytes, appending an optional
3295 * string (e.g. for ellipses)
3296 *
3297 * The database offers limited byte lengths for some columns in the database;
3298 * multi-byte character sets mean we need to ensure that only whole characters
3299 * are included, otherwise broken characters can be passed to the user
3300 *
3301 * If $length is negative, the string will be truncated from the beginning
3302 *
3303 * @param $string String to truncate
3304 * @param $length Int: maximum length (including ellipses)
3305 * @param $ellipsis String to append to the truncated text
3306 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3307 * $adjustLength was introduced in 1.18, before that behaved as if false.
3308 * @return string
3309 */
3310 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3311 # Use the localized ellipsis character
3312 if ( $ellipsis == '...' ) {
3313 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3314 }
3315 # Check if there is no need to truncate
3316 if ( $length == 0 ) {
3317 return $ellipsis; // convention
3318 } elseif ( strlen( $string ) <= abs( $length ) ) {
3319 return $string; // no need to truncate
3320 }
3321 $stringOriginal = $string;
3322 # If ellipsis length is >= $length then we can't apply $adjustLength
3323 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3324 $string = $ellipsis; // this can be slightly unexpected
3325 # Otherwise, truncate and add ellipsis...
3326 } else {
3327 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3328 if ( $length > 0 ) {
3329 $length -= $eLength;
3330 $string = substr( $string, 0, $length ); // xyz...
3331 $string = $this->removeBadCharLast( $string );
3332 $string = rtrim( $string );
3333 $string = $string . $ellipsis;
3334 } else {
3335 $length += $eLength;
3336 $string = substr( $string, $length ); // ...xyz
3337 $string = $this->removeBadCharFirst( $string );
3338 $string = ltrim( $string );
3339 $string = $ellipsis . $string;
3340 }
3341 }
3342 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3343 # This check is *not* redundant if $adjustLength, due to the single case where
3344 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3345 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3346 return $string;
3347 } else {
3348 return $stringOriginal;
3349 }
3350 }
3351
3352 /**
3353 * Remove bytes that represent an incomplete Unicode character
3354 * at the end of string (e.g. bytes of the char are missing)
3355 *
3356 * @param $string String
3357 * @return string
3358 */
3359 protected function removeBadCharLast( $string ) {
3360 if ( $string != '' ) {
3361 $char = ord( $string[strlen( $string ) - 1] );
3362 $m = array();
3363 if ( $char >= 0xc0 ) {
3364 # We got the first byte only of a multibyte char; remove it.
3365 $string = substr( $string, 0, -1 );
3366 } elseif ( $char >= 0x80 &&
3367 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3368 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3369 ) {
3370 # We chopped in the middle of a character; remove it
3371 $string = $m[1];
3372 }
3373 }
3374 return $string;
3375 }
3376
3377 /**
3378 * Remove bytes that represent an incomplete Unicode character
3379 * at the start of string (e.g. bytes of the char are missing)
3380 *
3381 * @param $string String
3382 * @return string
3383 */
3384 protected function removeBadCharFirst( $string ) {
3385 if ( $string != '' ) {
3386 $char = ord( $string[0] );
3387 if ( $char >= 0x80 && $char < 0xc0 ) {
3388 # We chopped in the middle of a character; remove the whole thing
3389 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3390 }
3391 }
3392 return $string;
3393 }
3394
3395 /**
3396 * Truncate a string of valid HTML to a specified length in bytes,
3397 * appending an optional string (e.g. for ellipses), and return valid HTML
3398 *
3399 * This is only intended for styled/linked text, such as HTML with
3400 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3401 * Also, this will not detect things like "display:none" CSS.
3402 *
3403 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3404 *
3405 * @param string $text HTML string to truncate
3406 * @param int $length (zero/positive) Maximum length (including ellipses)
3407 * @param string $ellipsis String to append to the truncated text
3408 * @return string
3409 */
3410 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3411 # Use the localized ellipsis character
3412 if ( $ellipsis == '...' ) {
3413 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3414 }
3415 # Check if there is clearly no need to truncate
3416 if ( $length <= 0 ) {
3417 return $ellipsis; // no text shown, nothing to format (convention)
3418 } elseif ( strlen( $text ) <= $length ) {
3419 return $text; // string short enough even *with* HTML (short-circuit)
3420 }
3421
3422 $dispLen = 0; // innerHTML legth so far
3423 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3424 $tagType = 0; // 0-open, 1-close
3425 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3426 $entityState = 0; // 0-not entity, 1-entity
3427 $tag = $ret = ''; // accumulated tag name, accumulated result string
3428 $openTags = array(); // open tag stack
3429 $maybeState = null; // possible truncation state
3430
3431 $textLen = strlen( $text );
3432 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3433 for ( $pos = 0; true; ++$pos ) {
3434 # Consider truncation once the display length has reached the maximim.
3435 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3436 # Check that we're not in the middle of a bracket/entity...
3437 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3438 if ( !$testingEllipsis ) {
3439 $testingEllipsis = true;
3440 # Save where we are; we will truncate here unless there turn out to
3441 # be so few remaining characters that truncation is not necessary.
3442 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3443 $maybeState = array( $ret, $openTags ); // save state
3444 }
3445 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3446 # String in fact does need truncation, the truncation point was OK.
3447 list( $ret, $openTags ) = $maybeState; // reload state
3448 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3449 $ret .= $ellipsis; // add ellipsis
3450 break;
3451 }
3452 }
3453 if ( $pos >= $textLen ) {
3454 break; // extra iteration just for above checks
3455 }
3456
3457 # Read the next char...
3458 $ch = $text[$pos];
3459 $lastCh = $pos ? $text[$pos - 1] : '';
3460 $ret .= $ch; // add to result string
3461 if ( $ch == '<' ) {
3462 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3463 $entityState = 0; // for bad HTML
3464 $bracketState = 1; // tag started (checking for backslash)
3465 } elseif ( $ch == '>' ) {
3466 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3467 $entityState = 0; // for bad HTML
3468 $bracketState = 0; // out of brackets
3469 } elseif ( $bracketState == 1 ) {
3470 if ( $ch == '/' ) {
3471 $tagType = 1; // close tag (e.g. "</span>")
3472 } else {
3473 $tagType = 0; // open tag (e.g. "<span>")
3474 $tag .= $ch;
3475 }
3476 $bracketState = 2; // building tag name
3477 } elseif ( $bracketState == 2 ) {
3478 if ( $ch != ' ' ) {
3479 $tag .= $ch;
3480 } else {
3481 // Name found (e.g. "<a href=..."), add on tag attributes...
3482 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3483 }
3484 } elseif ( $bracketState == 0 ) {
3485 if ( $entityState ) {
3486 if ( $ch == ';' ) {
3487 $entityState = 0;
3488 $dispLen++; // entity is one displayed char
3489 }
3490 } else {
3491 if ( $neLength == 0 && !$maybeState ) {
3492 // Save state without $ch. We want to *hit* the first
3493 // display char (to get tags) but not *use* it if truncating.
3494 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3495 }
3496 if ( $ch == '&' ) {
3497 $entityState = 1; // entity found, (e.g. "&#160;")
3498 } else {
3499 $dispLen++; // this char is displayed
3500 // Add the next $max display text chars after this in one swoop...
3501 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3502 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3503 $dispLen += $skipped;
3504 $pos += $skipped;
3505 }
3506 }
3507 }
3508 }
3509 // Close the last tag if left unclosed by bad HTML
3510 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3511 while ( count( $openTags ) > 0 ) {
3512 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3513 }
3514 return $ret;
3515 }
3516
3517 /**
3518 * truncateHtml() helper function
3519 * like strcspn() but adds the skipped chars to $ret
3520 *
3521 * @param $ret
3522 * @param $text
3523 * @param $search
3524 * @param $start
3525 * @param $len
3526 * @return int
3527 */
3528 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3529 if ( $len === null ) {
3530 $len = -1; // -1 means "no limit" for strcspn
3531 } elseif ( $len < 0 ) {
3532 $len = 0; // sanity
3533 }
3534 $skipCount = 0;
3535 if ( $start < strlen( $text ) ) {
3536 $skipCount = strcspn( $text, $search, $start, $len );
3537 $ret .= substr( $text, $start, $skipCount );
3538 }
3539 return $skipCount;
3540 }
3541
3542 /**
3543 * truncateHtml() helper function
3544 * (a) push or pop $tag from $openTags as needed
3545 * (b) clear $tag value
3546 * @param &$tag string Current HTML tag name we are looking at
3547 * @param $tagType int (0-open tag, 1-close tag)
3548 * @param $lastCh string Character before the '>' that ended this tag
3549 * @param &$openTags array Open tag stack (not accounting for $tag)
3550 */
3551 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3552 $tag = ltrim( $tag );
3553 if ( $tag != '' ) {
3554 if ( $tagType == 0 && $lastCh != '/' ) {
3555 $openTags[] = $tag; // tag opened (didn't close itself)
3556 } elseif ( $tagType == 1 ) {
3557 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3558 array_pop( $openTags ); // tag closed
3559 }
3560 }
3561 $tag = '';
3562 }
3563 }
3564
3565 /**
3566 * Grammatical transformations, needed for inflected languages
3567 * Invoked by putting {{grammar:case|word}} in a message
3568 *
3569 * @param $word string
3570 * @param $case string
3571 * @return string
3572 */
3573 function convertGrammar( $word, $case ) {
3574 global $wgGrammarForms;
3575 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3576 return $wgGrammarForms[$this->getCode()][$case][$word];
3577 }
3578 return $word;
3579 }
3580 /**
3581 * Get the grammar forms for the content language
3582 * @return array of grammar forms
3583 * @since 1.20
3584 */
3585 function getGrammarForms() {
3586 global $wgGrammarForms;
3587 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3588 return $wgGrammarForms[$this->getCode()];
3589 }
3590 return array();
3591 }
3592 /**
3593 * Provides an alternative text depending on specified gender.
3594 * Usage {{gender:username|masculine|feminine|unknown}}.
3595 * username is optional, in which case the gender of current user is used,
3596 * but only in (some) interface messages; otherwise default gender is used.
3597 *
3598 * If no forms are given, an empty string is returned. If only one form is
3599 * given, it will be returned unconditionally. These details are implied by
3600 * the caller and cannot be overridden in subclasses.
3601 *
3602 * If three forms are given, the default is to use the third (unknown) form.
3603 * If fewer than three forms are given, the default is to use the first (masculine) form.
3604 * These details can be overridden in subclasses.
3605 *
3606 * @param $gender string
3607 * @param $forms array
3608 *
3609 * @return string
3610 */
3611 function gender( $gender, $forms ) {
3612 if ( !count( $forms ) ) {
3613 return '';
3614 }
3615 $forms = $this->preConvertPlural( $forms, 2 );
3616 if ( $gender === 'male' ) {
3617 return $forms[0];
3618 }
3619 if ( $gender === 'female' ) {
3620 return $forms[1];
3621 }
3622 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3623 }
3624
3625 /**
3626 * Plural form transformations, needed for some languages.
3627 * For example, there are 3 form of plural in Russian and Polish,
3628 * depending on "count mod 10". See [[w:Plural]]
3629 * For English it is pretty simple.
3630 *
3631 * Invoked by putting {{plural:count|wordform1|wordform2}}
3632 * or {{plural:count|wordform1|wordform2|wordform3}}
3633 *
3634 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3635 *
3636 * @param $count Integer: non-localized number
3637 * @param $forms Array: different plural forms
3638 * @return string Correct form of plural for $count in this language
3639 */
3640 function convertPlural( $count, $forms ) {
3641 // Handle explicit n=pluralform cases
3642 $forms = $this->handleExplicitPluralForms( $count, $forms );
3643 if ( is_string( $forms ) ) {
3644 return $forms;
3645 }
3646 if ( !count( $forms ) ) {
3647 return '';
3648 }
3649
3650 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3651 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3652 return $forms[$pluralForm];
3653 }
3654
3655 /**
3656 * Handles explicit plural forms for Language::convertPlural()
3657 *
3658 * In {{PLURAL:$1|0=nothing|one|many}}, 0=nothing will be returned if $1 equals zero.
3659 * If an explicitly defined plural form matches the $count, then
3660 * string value returned, otherwise array returned for further consideration
3661 * by CLDR rules or overridden convertPlural().
3662 *
3663 * @since 1.23
3664 *
3665 * @param int $count non-localized number
3666 * @param array $forms different plural forms
3667 *
3668 * @return array|string
3669 */
3670 protected function handleExplicitPluralForms( $count, array $forms ) {
3671 foreach ( $forms as $index => $form ) {
3672 if ( preg_match( '/\d+=/i', $form ) ) {
3673 $pos = strpos( $form, '=' );
3674 if ( substr( $form, 0, $pos ) === (string) $count ) {
3675 return substr( $form, $pos + 1 );
3676 }
3677 unset( $forms[$index] );
3678 }
3679 }
3680 return array_values( $forms );
3681 }
3682
3683 /**
3684 * Checks that convertPlural was given an array and pads it to requested
3685 * amount of forms by copying the last one.
3686 *
3687 * @param $count Integer: How many forms should there be at least
3688 * @param $forms Array of forms given to convertPlural
3689 * @return array Padded array of forms or an exception if not an array
3690 */
3691 protected function preConvertPlural( /* Array */ $forms, $count ) {
3692 while ( count( $forms ) < $count ) {
3693 $forms[] = $forms[count( $forms ) - 1];
3694 }
3695 return $forms;
3696 }
3697
3698 /**
3699 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3700 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3701 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3702 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3703 * match up with it.
3704 *
3705 * @param $str String: the validated block duration in English
3706 * @return string Somehow translated block duration
3707 * @see LanguageFi.php for example implementation
3708 */
3709 function translateBlockExpiry( $str ) {
3710 $duration = SpecialBlock::getSuggestedDurations( $this );
3711 foreach ( $duration as $show => $value ) {
3712 if ( strcmp( $str, $value ) == 0 ) {
3713 return htmlspecialchars( trim( $show ) );
3714 }
3715 }
3716
3717 // Since usually only infinite or indefinite is only on list, so try
3718 // equivalents if still here.
3719 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3720 if ( in_array( $str, $indefs ) ) {
3721 foreach ( $indefs as $val ) {
3722 $show = array_search( $val, $duration, true );
3723 if ( $show !== false ) {
3724 return htmlspecialchars( trim( $show ) );
3725 }
3726 }
3727 }
3728
3729 // If all else fails, return a standard duration or timestamp description.
3730 $time = strtotime( $str, 0 );
3731 if ( $time === false ) { // Unknown format. Return it as-is in case.
3732 return $str;
3733 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3734 // $time is relative to 0 so it's a duration length.
3735 return $this->formatDuration( $time );
3736 } else { // It's an absolute timestamp.
3737 if ( $time === 0 ) {
3738 // wfTimestamp() handles 0 as current time instead of epoch.
3739 return $this->timeanddate( '19700101000000' );
3740 } else {
3741 return $this->timeanddate( $time );
3742 }
3743 }
3744 }
3745
3746 /**
3747 * languages like Chinese need to be segmented in order for the diff
3748 * to be of any use
3749 *
3750 * @param $text String
3751 * @return String
3752 */
3753 public function segmentForDiff( $text ) {
3754 return $text;
3755 }
3756
3757 /**
3758 * and unsegment to show the result
3759 *
3760 * @param $text String
3761 * @return String
3762 */
3763 public function unsegmentForDiff( $text ) {
3764 return $text;
3765 }
3766
3767 /**
3768 * Return the LanguageConverter used in the Language
3769 *
3770 * @since 1.19
3771 * @return LanguageConverter
3772 */
3773 public function getConverter() {
3774 return $this->mConverter;
3775 }
3776
3777 /**
3778 * convert text to all supported variants
3779 *
3780 * @param $text string
3781 * @return array
3782 */
3783 public function autoConvertToAllVariants( $text ) {
3784 return $this->mConverter->autoConvertToAllVariants( $text );
3785 }
3786
3787 /**
3788 * convert text to different variants of a language.
3789 *
3790 * @param $text string
3791 * @return string
3792 */
3793 public function convert( $text ) {
3794 return $this->mConverter->convert( $text );
3795 }
3796
3797 /**
3798 * Convert a Title object to a string in the preferred variant
3799 *
3800 * @param $title Title
3801 * @return string
3802 */
3803 public function convertTitle( $title ) {
3804 return $this->mConverter->convertTitle( $title );
3805 }
3806
3807 /**
3808 * Convert a namespace index to a string in the preferred variant
3809 *
3810 * @param $ns int
3811 * @return string
3812 */
3813 public function convertNamespace( $ns ) {
3814 return $this->mConverter->convertNamespace( $ns );
3815 }
3816
3817 /**
3818 * Check if this is a language with variants
3819 *
3820 * @return bool
3821 */
3822 public function hasVariants() {
3823 return count( $this->getVariants() ) > 1;
3824 }
3825
3826 /**
3827 * Check if the language has the specific variant
3828 *
3829 * @since 1.19
3830 * @param $variant string
3831 * @return bool
3832 */
3833 public function hasVariant( $variant ) {
3834 return (bool)$this->mConverter->validateVariant( $variant );
3835 }
3836
3837 /**
3838 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3839 *
3840 * @param $text string
3841 * @return string
3842 * @deprecated since 1.22 is no longer used
3843 */
3844 public function armourMath( $text ) {
3845 return $this->mConverter->armourMath( $text );
3846 }
3847
3848 /**
3849 * Perform output conversion on a string, and encode for safe HTML output.
3850 * @param $text String text to be converted
3851 * @param $isTitle Bool whether this conversion is for the article title
3852 * @return string
3853 * @todo this should get integrated somewhere sane
3854 */
3855 public function convertHtml( $text, $isTitle = false ) {
3856 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3857 }
3858
3859 /**
3860 * @param $key string
3861 * @return string
3862 */
3863 public function convertCategoryKey( $key ) {
3864 return $this->mConverter->convertCategoryKey( $key );
3865 }
3866
3867 /**
3868 * Get the list of variants supported by this language
3869 * see sample implementation in LanguageZh.php
3870 *
3871 * @return array an array of language codes
3872 */
3873 public function getVariants() {
3874 return $this->mConverter->getVariants();
3875 }
3876
3877 /**
3878 * @return string
3879 */
3880 public function getPreferredVariant() {
3881 return $this->mConverter->getPreferredVariant();
3882 }
3883
3884 /**
3885 * @return string
3886 */
3887 public function getDefaultVariant() {
3888 return $this->mConverter->getDefaultVariant();
3889 }
3890
3891 /**
3892 * @return string
3893 */
3894 public function getURLVariant() {
3895 return $this->mConverter->getURLVariant();
3896 }
3897
3898 /**
3899 * If a language supports multiple variants, it is
3900 * possible that non-existing link in one variant
3901 * actually exists in another variant. this function
3902 * tries to find it. See e.g. LanguageZh.php
3903 *
3904 * @param $link String: the name of the link
3905 * @param $nt Mixed: the title object of the link
3906 * @param $ignoreOtherCond Boolean: to disable other conditions when
3907 * we need to transclude a template or update a category's link
3908 * @return null the input parameters may be modified upon return
3909 */
3910 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3911 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
3912 }
3913
3914 /**
3915 * returns language specific options used by User::getPageRenderHash()
3916 * for example, the preferred language variant
3917 *
3918 * @return string
3919 */
3920 function getExtraHashOptions() {
3921 return $this->mConverter->getExtraHashOptions();
3922 }
3923
3924 /**
3925 * For languages that support multiple variants, the title of an
3926 * article may be displayed differently in different variants. this
3927 * function returns the apporiate title defined in the body of the article.
3928 *
3929 * @return string
3930 */
3931 public function getParsedTitle() {
3932 return $this->mConverter->getParsedTitle();
3933 }
3934
3935 /**
3936 * Prepare external link text for conversion. When the text is
3937 * a URL, it shouldn't be converted, and it'll be wrapped in
3938 * the "raw" tag (-{R| }-) to prevent conversion.
3939 *
3940 * This function is called "markNoConversion" for historical
3941 * reasons.
3942 *
3943 * @param $text String: text to be used for external link
3944 * @param $noParse bool: wrap it without confirming it's a real URL first
3945 * @return string the tagged text
3946 */
3947 public function markNoConversion( $text, $noParse = false ) {
3948 // Excluding protocal-relative URLs may avoid many false positives.
3949 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
3950 return $this->mConverter->markNoConversion( $text );
3951 } else {
3952 return $text;
3953 }
3954 }
3955
3956 /**
3957 * A regular expression to match legal word-trailing characters
3958 * which should be merged onto a link of the form [[foo]]bar.
3959 *
3960 * @return string
3961 */
3962 public function linkTrail() {
3963 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
3964 }
3965
3966 /**
3967 * A regular expression character set to match legal word-prefixing
3968 * characters which should be merged onto a link of the form foo[[bar]].
3969 *
3970 * @return string
3971 */
3972 public function linkPrefixCharset() {
3973 return self::$dataCache->getItem( $this->mCode, 'linkPrefixCharset' );
3974 }
3975
3976 /**
3977 * @return Language
3978 */
3979 function getLangObj() {
3980 return $this;
3981 }
3982
3983 /**
3984 * Get the "parent" language which has a converter to convert a "compatible" language
3985 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
3986 *
3987 * @return Language|null
3988 * @since 1.22
3989 */
3990 public function getParentLanguage() {
3991 if ( $this->mParentLanguage !== false ) {
3992 return $this->mParentLanguage;
3993 }
3994
3995 $pieces = explode( '-', $this->getCode() );
3996 $code = $pieces[0];
3997 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
3998 $this->mParentLanguage = null;
3999 return null;
4000 }
4001 $lang = Language::factory( $code );
4002 if ( !$lang->hasVariant( $this->getCode() ) ) {
4003 $this->mParentLanguage = null;
4004 return null;
4005 }
4006
4007 $this->mParentLanguage = $lang;
4008 return $lang;
4009 }
4010
4011 /**
4012 * Get the RFC 3066 code for this language object
4013 *
4014 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4015 * htmlspecialchars() or similar
4016 *
4017 * @return string
4018 */
4019 public function getCode() {
4020 return $this->mCode;
4021 }
4022
4023 /**
4024 * Get the code in Bcp47 format which we can use
4025 * inside of html lang="" tags.
4026 *
4027 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
4028 * htmlspecialchars() or similar.
4029 *
4030 * @since 1.19
4031 * @return string
4032 */
4033 public function getHtmlCode() {
4034 if ( is_null( $this->mHtmlCode ) ) {
4035 $this->mHtmlCode = wfBCP47( $this->getCode() );
4036 }
4037 return $this->mHtmlCode;
4038 }
4039
4040 /**
4041 * @param $code string
4042 */
4043 public function setCode( $code ) {
4044 $this->mCode = $code;
4045 // Ensure we don't leave incorrect cached data lying around
4046 $this->mHtmlCode = null;
4047 $this->mParentLanguage = false;
4048 }
4049
4050 /**
4051 * Get the name of a file for a certain language code
4052 * @param $prefix string Prepend this to the filename
4053 * @param $code string Language code
4054 * @param $suffix string Append this to the filename
4055 * @throws MWException
4056 * @return string $prefix . $mangledCode . $suffix
4057 */
4058 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4059 if ( !self::isValidBuiltInCode( $code ) ) {
4060 throw new MWException( "Invalid language code \"$code\"" );
4061 }
4062
4063 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4064 }
4065
4066 /**
4067 * Get the language code from a file name. Inverse of getFileName()
4068 * @param $filename string $prefix . $languageCode . $suffix
4069 * @param $prefix string Prefix before the language code
4070 * @param $suffix string Suffix after the language code
4071 * @return string Language code, or false if $prefix or $suffix isn't found
4072 */
4073 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4074 $m = null;
4075 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4076 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4077 if ( !count( $m ) ) {
4078 return false;
4079 }
4080 return str_replace( '_', '-', strtolower( $m[1] ) );
4081 }
4082
4083 /**
4084 * @param $code string
4085 * @return string
4086 */
4087 public static function getMessagesFileName( $code ) {
4088 global $IP;
4089 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4090 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
4091 return $file;
4092 }
4093
4094 /**
4095 * @param $code string
4096 * @return string
4097 * @since 1.23
4098 */
4099 public static function getJsonMessagesFileName( $code ) {
4100 global $IP;
4101
4102 if ( !self::isValidBuiltInCode( $code ) ) {
4103 throw new MWException( "Invalid language code \"$code\"" );
4104 }
4105
4106 return "$IP/languages/i18n/$code.json" ;
4107 }
4108
4109 /**
4110 * @param $code string
4111 * @return string
4112 */
4113 public static function getClassFileName( $code ) {
4114 global $IP;
4115 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4116 }
4117
4118 /**
4119 * Get the first fallback for a given language.
4120 *
4121 * @param $code string
4122 *
4123 * @return bool|string
4124 */
4125 public static function getFallbackFor( $code ) {
4126 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4127 return false;
4128 } else {
4129 $fallbacks = self::getFallbacksFor( $code );
4130 $first = array_shift( $fallbacks );
4131 return $first;
4132 }
4133 }
4134
4135 /**
4136 * Get the ordered list of fallback languages.
4137 *
4138 * @since 1.19
4139 * @param $code string Language code
4140 * @return array
4141 */
4142 public static function getFallbacksFor( $code ) {
4143 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4144 return array();
4145 } else {
4146 $v = self::getLocalisationCache()->getItem( $code, 'fallback' );
4147 $v = array_map( 'trim', explode( ',', $v ) );
4148 if ( $v[count( $v ) - 1] !== 'en' ) {
4149 $v[] = 'en';
4150 }
4151 return $v;
4152 }
4153 }
4154
4155 /**
4156 * Get the ordered list of fallback languages, ending with the fallback
4157 * language chain for the site language.
4158 *
4159 * @since 1.22
4160 * @param string $code Language code
4161 * @return array array( fallbacks, site fallbacks )
4162 */
4163 public static function getFallbacksIncludingSiteLanguage( $code ) {
4164 global $wgLanguageCode;
4165
4166 // Usually, we will only store a tiny number of fallback chains, so we
4167 // keep them in static memory.
4168 $cacheKey = "{$code}-{$wgLanguageCode}";
4169
4170 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4171 $fallbacks = self::getFallbacksFor( $code );
4172
4173 // Append the site's fallback chain, including the site language itself
4174 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4175 array_unshift( $siteFallbacks, $wgLanguageCode );
4176
4177 // Eliminate any languages already included in the chain
4178 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4179
4180 self::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4181 }
4182 return self::$fallbackLanguageCache[$cacheKey];
4183 }
4184
4185 /**
4186 * Get all messages for a given language
4187 * WARNING: this may take a long time. If you just need all message *keys*
4188 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4189 *
4190 * @param $code string
4191 *
4192 * @return array
4193 */
4194 public static function getMessagesFor( $code ) {
4195 return self::getLocalisationCache()->getItem( $code, 'messages' );
4196 }
4197
4198 /**
4199 * Get a message for a given language
4200 *
4201 * @param $key string
4202 * @param $code string
4203 *
4204 * @return string
4205 */
4206 public static function getMessageFor( $key, $code ) {
4207 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4208 }
4209
4210 /**
4211 * Get all message keys for a given language. This is a faster alternative to
4212 * array_keys( Language::getMessagesFor( $code ) )
4213 *
4214 * @since 1.19
4215 * @param $code string Language code
4216 * @return array of message keys (strings)
4217 */
4218 public static function getMessageKeysFor( $code ) {
4219 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
4220 }
4221
4222 /**
4223 * @param $talk
4224 * @return mixed
4225 */
4226 function fixVariableInNamespace( $talk ) {
4227 if ( strpos( $talk, '$1' ) === false ) {
4228 return $talk;
4229 }
4230
4231 global $wgMetaNamespace;
4232 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4233
4234 # Allow grammar transformations
4235 # Allowing full message-style parsing would make simple requests
4236 # such as action=raw much more expensive than they need to be.
4237 # This will hopefully cover most cases.
4238 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4239 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4240 return str_replace( ' ', '_', $talk );
4241 }
4242
4243 /**
4244 * @param $m string
4245 * @return string
4246 */
4247 function replaceGrammarInNamespace( $m ) {
4248 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4249 }
4250
4251 /**
4252 * @throws MWException
4253 * @return array
4254 */
4255 static function getCaseMaps() {
4256 static $wikiUpperChars, $wikiLowerChars;
4257 if ( isset( $wikiUpperChars ) ) {
4258 return array( $wikiUpperChars, $wikiLowerChars );
4259 }
4260
4261 wfProfileIn( __METHOD__ );
4262 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4263 if ( $arr === false ) {
4264 throw new MWException(
4265 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4266 }
4267 $wikiUpperChars = $arr['wikiUpperChars'];
4268 $wikiLowerChars = $arr['wikiLowerChars'];
4269 wfProfileOut( __METHOD__ );
4270 return array( $wikiUpperChars, $wikiLowerChars );
4271 }
4272
4273 /**
4274 * Decode an expiry (block, protection, etc) which has come from the DB
4275 *
4276 * @todo FIXME: why are we returnings DBMS-dependent strings???
4277 *
4278 * @param $expiry String: Database expiry String
4279 * @param $format Bool|Int true to process using language functions, or TS_ constant
4280 * to return the expiry in a given timestamp
4281 * @return String
4282 * @since 1.18
4283 */
4284 public function formatExpiry( $expiry, $format = true ) {
4285 static $infinity;
4286 if ( $infinity === null ) {
4287 $infinity = wfGetDB( DB_SLAVE )->getInfinity();
4288 }
4289
4290 if ( $expiry == '' || $expiry == $infinity ) {
4291 return $format === true
4292 ? $this->getMessageFromDB( 'infiniteblock' )
4293 : $infinity;
4294 } else {
4295 return $format === true
4296 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4297 : wfTimestamp( $format, $expiry );
4298 }
4299 }
4300
4301 /**
4302 * @todo Document
4303 * @param $seconds int|float
4304 * @param $format Array Optional
4305 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
4306 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
4307 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
4308 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
4309 * @return string
4310 */
4311 function formatTimePeriod( $seconds, $format = array() ) {
4312 if ( !is_array( $format ) ) {
4313 $format = array( 'avoid' => $format ); // For backwards compatibility
4314 }
4315 if ( !isset( $format['avoid'] ) ) {
4316 $format['avoid'] = false;
4317 }
4318 if ( !isset( $format['noabbrevs' ] ) ) {
4319 $format['noabbrevs'] = false;
4320 }
4321 $secondsMsg = wfMessage(
4322 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4323 $minutesMsg = wfMessage(
4324 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4325 $hoursMsg = wfMessage(
4326 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4327 $daysMsg = wfMessage(
4328 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4329
4330 if ( round( $seconds * 10 ) < 100 ) {
4331 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4332 $s = $secondsMsg->params( $s )->text();
4333 } elseif ( round( $seconds ) < 60 ) {
4334 $s = $this->formatNum( round( $seconds ) );
4335 $s = $secondsMsg->params( $s )->text();
4336 } elseif ( round( $seconds ) < 3600 ) {
4337 $minutes = floor( $seconds / 60 );
4338 $secondsPart = round( fmod( $seconds, 60 ) );
4339 if ( $secondsPart == 60 ) {
4340 $secondsPart = 0;
4341 $minutes++;
4342 }
4343 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4344 $s .= ' ';
4345 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4346 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4347 $hours = floor( $seconds / 3600 );
4348 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4349 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4350 if ( $secondsPart == 60 ) {
4351 $secondsPart = 0;
4352 $minutes++;
4353 }
4354 if ( $minutes == 60 ) {
4355 $minutes = 0;
4356 $hours++;
4357 }
4358 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4359 $s .= ' ';
4360 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4361 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4362 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4363 }
4364 } else {
4365 $days = floor( $seconds / 86400 );
4366 if ( $format['avoid'] === 'avoidminutes' ) {
4367 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4368 if ( $hours == 24 ) {
4369 $hours = 0;
4370 $days++;
4371 }
4372 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4373 $s .= ' ';
4374 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4375 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4376 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4377 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4378 if ( $minutes == 60 ) {
4379 $minutes = 0;
4380 $hours++;
4381 }
4382 if ( $hours == 24 ) {
4383 $hours = 0;
4384 $days++;
4385 }
4386 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4387 $s .= ' ';
4388 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4389 $s .= ' ';
4390 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4391 } else {
4392 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4393 $s .= ' ';
4394 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4395 }
4396 }
4397 return $s;
4398 }
4399
4400 /**
4401 * Format a bitrate for output, using an appropriate
4402 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4403 *
4404 * This use base 1000. For base 1024 use formatSize(), for another base
4405 * see formatComputingNumbers()
4406 *
4407 * @param $bps int
4408 * @return string
4409 */
4410 function formatBitrate( $bps ) {
4411 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4412 }
4413
4414 /**
4415 * @param $size int Size of the unit
4416 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4417 * @param $messageKey string Message key to be uesd
4418 * @return string
4419 */
4420 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4421 if ( $size <= 0 ) {
4422 return str_replace( '$1', $this->formatNum( $size ),
4423 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4424 );
4425 }
4426 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4427 $index = 0;
4428
4429 $maxIndex = count( $sizes ) - 1;
4430 while ( $size >= $boundary && $index < $maxIndex ) {
4431 $index++;
4432 $size /= $boundary;
4433 }
4434
4435 // For small sizes no decimal places necessary
4436 $round = 0;
4437 if ( $index > 1 ) {
4438 // For MB and bigger two decimal places are smarter
4439 $round = 2;
4440 }
4441 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4442
4443 $size = round( $size, $round );
4444 $text = $this->getMessageFromDB( $msg );
4445 return str_replace( '$1', $this->formatNum( $size ), $text );
4446 }
4447
4448 /**
4449 * Format a size in bytes for output, using an appropriate
4450 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4451 *
4452 * This method use base 1024. For base 1000 use formatBitrate(), for
4453 * another base see formatComputingNumbers()
4454 *
4455 * @param $size int Size to format
4456 * @return string Plain text (not HTML)
4457 */
4458 function formatSize( $size ) {
4459 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4460 }
4461
4462 /**
4463 * Make a list item, used by various special pages
4464 *
4465 * @param $page String Page link
4466 * @param $details String Text between brackets
4467 * @param $oppositedm Boolean Add the direction mark opposite to your
4468 * language, to display text properly
4469 * @return String
4470 */
4471 function specialList( $page, $details, $oppositedm = true ) {
4472 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) .
4473 $this->getDirMark();
4474 $details = $details ? $dirmark . $this->getMessageFromDB( 'word-separator' ) .
4475 wfMessage( 'parentheses' )->rawParams( $details )->inLanguage( $this )->escaped() : '';
4476 return $page . $details;
4477 }
4478
4479 /**
4480 * Generate (prev x| next x) (20|50|100...) type links for paging
4481 *
4482 * @param $title Title object to link
4483 * @param $offset Integer offset parameter
4484 * @param $limit Integer limit parameter
4485 * @param $query array|String optional URL query parameter string
4486 * @param $atend Bool optional param for specified if this is the last page
4487 * @return String
4488 */
4489 public function viewPrevNext( Title $title, $offset, $limit, array $query = array(), $atend = false ) {
4490 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4491
4492 # Make 'previous' link
4493 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4494 if ( $offset > 0 ) {
4495 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4496 $query, $prev, 'prevn-title', 'mw-prevlink' );
4497 } else {
4498 $plink = htmlspecialchars( $prev );
4499 }
4500
4501 # Make 'next' link
4502 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4503 if ( $atend ) {
4504 $nlink = htmlspecialchars( $next );
4505 } else {
4506 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4507 $query, $next, 'nextn-title', 'mw-nextlink' );
4508 }
4509
4510 # Make links to set number of items per page
4511 $numLinks = array();
4512 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4513 $numLinks[] = $this->numLink( $title, $offset, $num,
4514 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4515 }
4516
4517 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4518 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4519 }
4520
4521 /**
4522 * Helper function for viewPrevNext() that generates links
4523 *
4524 * @param $title Title object to link
4525 * @param $offset Integer offset parameter
4526 * @param $limit Integer limit parameter
4527 * @param $query Array extra query parameters
4528 * @param $link String text to use for the link; will be escaped
4529 * @param $tooltipMsg String name of the message to use as tooltip
4530 * @param $class String value of the "class" attribute of the link
4531 * @return String HTML fragment
4532 */
4533 private function numLink( Title $title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4534 $query = array( 'limit' => $limit, 'offset' => $offset ) + $query;
4535 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4536 return Html::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4537 'title' => $tooltip, 'class' => $class ), $link );
4538 }
4539
4540 /**
4541 * Get the conversion rule title, if any.
4542 *
4543 * @return string
4544 */
4545 public function getConvRuleTitle() {
4546 return $this->mConverter->getConvRuleTitle();
4547 }
4548
4549 /**
4550 * Get the compiled plural rules for the language
4551 * @since 1.20
4552 * @return array Associative array with plural form, and plural rule as key-value pairs
4553 */
4554 public function getCompiledPluralRules() {
4555 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4556 $fallbacks = Language::getFallbacksFor( $this->mCode );
4557 if ( !$pluralRules ) {
4558 foreach ( $fallbacks as $fallbackCode ) {
4559 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4560 if ( $pluralRules ) {
4561 break;
4562 }
4563 }
4564 }
4565 return $pluralRules;
4566 }
4567
4568 /**
4569 * Get the plural rules for the language
4570 * @since 1.20
4571 * @return array Associative array with plural form number and plural rule as key-value pairs
4572 */
4573 public function getPluralRules() {
4574 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4575 $fallbacks = Language::getFallbacksFor( $this->mCode );
4576 if ( !$pluralRules ) {
4577 foreach ( $fallbacks as $fallbackCode ) {
4578 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4579 if ( $pluralRules ) {
4580 break;
4581 }
4582 }
4583 }
4584 return $pluralRules;
4585 }
4586
4587 /**
4588 * Get the plural rule types for the language
4589 * @since 1.22
4590 * @return array Associative array with plural form number and plural rule type as key-value pairs
4591 */
4592 public function getPluralRuleTypes() {
4593 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4594 $fallbacks = Language::getFallbacksFor( $this->mCode );
4595 if ( !$pluralRuleTypes ) {
4596 foreach ( $fallbacks as $fallbackCode ) {
4597 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4598 if ( $pluralRuleTypes ) {
4599 break;
4600 }
4601 }
4602 }
4603 return $pluralRuleTypes;
4604 }
4605
4606 /**
4607 * Find the index number of the plural rule appropriate for the given number
4608 * @return int The index number of the plural rule
4609 */
4610 public function getPluralRuleIndexNumber( $number ) {
4611 $pluralRules = $this->getCompiledPluralRules();
4612 $form = CLDRPluralRuleEvaluator::evaluateCompiled( $number, $pluralRules );
4613 return $form;
4614 }
4615
4616 /**
4617 * Find the plural rule type appropriate for the given number
4618 * For example, if the language is set to Arabic, getPluralType(5) should
4619 * return 'few'.
4620 * @since 1.22
4621 * @return string The name of the plural rule type, e.g. one, two, few, many
4622 */
4623 public function getPluralRuleType( $number ) {
4624 $index = $this->getPluralRuleIndexNumber( $number );
4625 $pluralRuleTypes = $this->getPluralRuleTypes();
4626 if ( isset( $pluralRuleTypes[$index] ) ) {
4627 return $pluralRuleTypes[$index];
4628 } else {
4629 return 'other';
4630 }
4631 }
4632 }
Wiklou, le Wiki du Wiklou