dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Add/update comments for grep
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * Internationalisation code.
4 *
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
19 *
20 * @file
21 * @ingroup Language
22 */
23
24 /**
25 * @defgroup Language Language
26 */
27
28 if ( !defined( 'MEDIAWIKI' ) ) {
29 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
30 exit( 1 );
31 }
32
33 # Read language names
34 global $wgLanguageNames;
35 require_once __DIR__ . '/Names.php';
36
37 if ( function_exists( 'mb_strtoupper' ) ) {
38 mb_internal_encoding( 'UTF-8' );
39 }
40
41 /**
42 * a fake language converter
43 *
44 * @ingroup Language
45 */
46 class FakeConverter {
47 /**
48 * @var Language
49 */
50 public $mLang;
51 function __construct( $langobj ) { $this->mLang = $langobj; }
52 function autoConvert( $text, $variant = false ) { return $text; }
53 function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
54 function convert( $t ) { return $t; }
55 function convertTo( $text, $variant ) { return $text; }
56 function convertTitle( $t ) { return $t->getPrefixedText(); }
57 function convertNamespace( $ns ) { return $this->mLang->getFormattedNsText( $ns ); }
58 function getVariants() { return array( $this->mLang->getCode() ); }
59 function getVariantFallbacks( $variant ) { return $this->mLang->getCode(); }
60 function getPreferredVariant() { return $this->mLang->getCode(); }
61 function getDefaultVariant() { return $this->mLang->getCode(); }
62 function getURLVariant() { return ''; }
63 function getConvRuleTitle() { return false; }
64 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
65 function getExtraHashOptions() { return ''; }
66 function getParsedTitle() { return ''; }
67 function markNoConversion( $text, $noParse = false ) { return $text; }
68 function convertCategoryKey( $key ) { return $key; }
69 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
70 function armourMath( $text ) { return $text; }
71 function validateVariant( $variant = null ) { return $variant === $this->mLang->getCode() ? $variant : null; }
72 function translate( $text, $variant ) { return $text; }
73 }
74
75 /**
76 * Internationalisation code
77 * @ingroup Language
78 */
79 class Language {
80
81 /**
82 * @var LanguageConverter
83 */
84 public $mConverter;
85
86 public $mVariants, $mCode, $mLoaded = false;
87 public $mMagicExtensions = array(), $mMagicHookDone = false;
88 private $mHtmlCode = null, $mParentLanguage = false;
89
90 public $dateFormatStrings = array();
91 public $mExtendedSpecialPageAliases;
92
93 protected $namespaceNames, $mNamespaceIds, $namespaceAliases;
94
95 /**
96 * ReplacementArray object caches
97 */
98 public $transformData = array();
99
100 /**
101 * @var LocalisationCache
102 */
103 static public $dataCache;
104
105 static public $mLangObjCache = array();
106
107 static public $mWeekdayMsgs = array(
108 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
109 'friday', 'saturday'
110 );
111
112 static public $mWeekdayAbbrevMsgs = array(
113 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
114 );
115
116 static public $mMonthMsgs = array(
117 'january', 'february', 'march', 'april', 'may_long', 'june',
118 'july', 'august', 'september', 'october', 'november',
119 'december'
120 );
121 static public $mMonthGenMsgs = array(
122 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
123 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
124 'december-gen'
125 );
126 static public $mMonthAbbrevMsgs = array(
127 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
128 'sep', 'oct', 'nov', 'dec'
129 );
130
131 static public $mIranianCalendarMonthMsgs = array(
132 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
133 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
134 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
135 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
136 );
137
138 static public $mHebrewCalendarMonthMsgs = array(
139 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
140 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
141 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
142 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
143 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
144 );
145
146 static public $mHebrewCalendarMonthGenMsgs = array(
147 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
148 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
149 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
150 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
151 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
152 );
153
154 static public $mHijriCalendarMonthMsgs = array(
155 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
156 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
157 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
158 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
159 );
160
161 /**
162 * @since 1.20
163 * @var array
164 */
165 static public $durationIntervals = array(
166 'millennia' => 31556952000,
167 'centuries' => 3155695200,
168 'decades' => 315569520,
169 'years' => 31556952, // 86400 * ( 365 + ( 24 * 3 + 25 ) / 400 )
170 'weeks' => 604800,
171 'days' => 86400,
172 'hours' => 3600,
173 'minutes' => 60,
174 'seconds' => 1,
175 );
176
177 /**
178 * Cache for language fallbacks.
179 * @see Language::getFallbacksIncludingSiteLanguage
180 * @since 1.21
181 * @var array
182 */
183 static private $fallbackLanguageCache = array();
184
185 /**
186 * Get a cached or new language object for a given language code
187 * @param $code String
188 * @return Language
189 */
190 static function factory( $code ) {
191 global $wgDummyLanguageCodes, $wgLangObjCacheSize;
192
193 if ( isset( $wgDummyLanguageCodes[$code] ) ) {
194 $code = $wgDummyLanguageCodes[$code];
195 }
196
197 // get the language object to process
198 $langObj = isset( self::$mLangObjCache[$code] )
199 ? self::$mLangObjCache[$code]
200 : self::newFromCode( $code );
201
202 // merge the language object in to get it up front in the cache
203 self::$mLangObjCache = array_merge( array( $code => $langObj ), self::$mLangObjCache );
204 // get rid of the oldest ones in case we have an overflow
205 self::$mLangObjCache = array_slice( self::$mLangObjCache, 0, $wgLangObjCacheSize, true );
206
207 return $langObj;
208 }
209
210 /**
211 * Create a language object for a given language code
212 * @param $code String
213 * @throws MWException
214 * @return Language
215 */
216 protected static function newFromCode( $code ) {
217 // Protect against path traversal below
218 if ( !Language::isValidCode( $code )
219 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
220 {
221 throw new MWException( "Invalid language code \"$code\"" );
222 }
223
224 if ( !Language::isValidBuiltInCode( $code ) ) {
225 // It's not possible to customise this code with class files, so
226 // just return a Language object. This is to support uselang= hacks.
227 $lang = new Language;
228 $lang->setCode( $code );
229 return $lang;
230 }
231
232 // Check if there is a language class for the code
233 $class = self::classFromCode( $code );
234 self::preloadLanguageClass( $class );
235 if ( MWInit::classExists( $class ) ) {
236 $lang = new $class;
237 return $lang;
238 }
239
240 // Keep trying the fallback list until we find an existing class
241 $fallbacks = Language::getFallbacksFor( $code );
242 foreach ( $fallbacks as $fallbackCode ) {
243 if ( !Language::isValidBuiltInCode( $fallbackCode ) ) {
244 throw new MWException( "Invalid fallback '$fallbackCode' in fallback sequence for '$code'" );
245 }
246
247 $class = self::classFromCode( $fallbackCode );
248 self::preloadLanguageClass( $class );
249 if ( MWInit::classExists( $class ) ) {
250 $lang = Language::newFromCode( $fallbackCode );
251 $lang->setCode( $code );
252 return $lang;
253 }
254 }
255
256 throw new MWException( "Invalid fallback sequence for language '$code'" );
257 }
258
259 /**
260 * Checks whether any localisation is available for that language tag
261 * in MediaWiki (MessagesXx.php exists).
262 *
263 * @param string $code Language tag (in lower case)
264 * @return bool Whether language is supported
265 * @since 1.21
266 */
267 public static function isSupportedLanguage( $code ) {
268 return $code === strtolower( $code ) && is_readable( self::getMessagesFileName( $code ) );
269 }
270
271 /**
272 * Returns true if a language code string is a well-formed language tag
273 * according to RFC 5646.
274 * This function only checks well-formedness; it doesn't check that
275 * language, script or variant codes actually exist in the repositories.
276 *
277 * Based on regexes by Mark Davis of the Unicode Consortium:
278 * http://unicode.org/repos/cldr/trunk/tools/java/org/unicode/cldr/util/data/langtagRegex.txt
279 *
280 * @param $code string
281 * @param $lenient boolean Whether to allow '_' as separator. The default is only '-'.
282 *
283 * @return bool
284 * @since 1.21
285 */
286 public static function isWellFormedLanguageTag( $code, $lenient = false ) {
287 $alpha = '[a-z]';
288 $digit = '[0-9]';
289 $alphanum = '[a-z0-9]';
290 $x = 'x'; # private use singleton
291 $singleton = '[a-wy-z]'; # other singleton
292 $s = $lenient ? '[-_]' : '-';
293
294 $language = "$alpha{2,8}|$alpha{2,3}$s$alpha{3}";
295 $script = "$alpha{4}"; # ISO 15924
296 $region = "(?:$alpha{2}|$digit{3})"; # ISO 3166-1 alpha-2 or UN M.49
297 $variant = "(?:$alphanum{5,8}|$digit$alphanum{3})";
298 $extension = "$singleton(?:$s$alphanum{2,8})+";
299 $privateUse = "$x(?:$s$alphanum{1,8})+";
300
301 # Define certain grandfathered codes, since otherwise the regex is pretty useless.
302 # Since these are limited, this is safe even later changes to the registry --
303 # the only oddity is that it might change the type of the tag, and thus
304 # the results from the capturing groups.
305 # http://www.iana.org/assignments/language-subtag-registry
306
307 $grandfathered = "en{$s}GB{$s}oed"
308 . "|i{$s}(?:ami|bnn|default|enochian|hak|klingon|lux|mingo|navajo|pwn|tao|tay|tsu)"
309 . "|no{$s}(?:bok|nyn)"
310 . "|sgn{$s}(?:BE{$s}(?:fr|nl)|CH{$s}de)"
311 . "|zh{$s}min{$s}nan";
312
313 $variantList = "$variant(?:$s$variant)*";
314 $extensionList = "$extension(?:$s$extension)*";
315
316 $langtag = "(?:($language)"
317 . "(?:$s$script)?"
318 . "(?:$s$region)?"
319 . "(?:$s$variantList)?"
320 . "(?:$s$extensionList)?"
321 . "(?:$s$privateUse)?)";
322
323 # The final breakdown, with capturing groups for each of these components
324 # The variants, extensions, grandfathered, and private-use may have interior '-'
325
326 $root = "^(?:$langtag|$privateUse|$grandfathered)$";
327
328 return (bool)preg_match( "/$root/", strtolower( $code ) );
329 }
330
331 /**
332 * Returns true if a language code string is of a valid form, whether or
333 * not it exists. This includes codes which are used solely for
334 * customisation via the MediaWiki namespace.
335 *
336 * @param $code string
337 *
338 * @return bool
339 */
340 public static function isValidCode( $code ) {
341 static $cache = array();
342 if ( isset( $cache[$code] ) ) {
343 return $cache[$code];
344 }
345 // People think language codes are html safe, so enforce it.
346 // Ideally we should only allow a-zA-Z0-9-
347 // but, .+ and other chars are often used for {{int:}} hacks
348 // see bugs 37564, 37587, 36938
349 $cache[$code] =
350 strcspn( $code, ":/\\\000&<>'\"" ) === strlen( $code )
351 && !preg_match( Title::getTitleInvalidRegex(), $code );
352
353 return $cache[$code];
354 }
355
356 /**
357 * Returns true if a language code is of a valid form for the purposes of
358 * internal customisation of MediaWiki, via Messages*.php.
359 *
360 * @param $code string
361 *
362 * @throws MWException
363 * @since 1.18
364 * @return bool
365 */
366 public static function isValidBuiltInCode( $code ) {
367
368 if ( !is_string( $code ) ) {
369 if ( is_object( $code ) ) {
370 $addmsg = " of class " . get_class( $code );
371 } else {
372 $addmsg = '';
373 }
374 $type = gettype( $code );
375 throw new MWException( __METHOD__ . " must be passed a string, $type given$addmsg" );
376 }
377
378 return (bool)preg_match( '/^[a-z0-9-]{2,}$/i', $code );
379 }
380
381 /**
382 * Returns true if a language code is an IETF tag known to MediaWiki.
383 *
384 * @param $code string
385 *
386 * @since 1.21
387 * @return bool
388 */
389 public static function isKnownLanguageTag( $tag ) {
390 static $coreLanguageNames;
391
392 // Quick escape for invalid input to avoid exceptions down the line
393 // when code tries to process tags which are not valid at all.
394 if ( !self::isValidBuiltInCode( $tag ) ) {
395 return false;
396 }
397
398 if ( $coreLanguageNames === null ) {
399 include MWInit::compiledPath( '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 * @param $index Int: namespace index
616 * @param $gender String: gender key (male, female... )
617 * @return String
618 * @since 1.18
619 */
620 function getGenderNsText( $index, $gender ) {
621 global $wgExtraGenderNamespaces;
622
623 $ns = $wgExtraGenderNamespaces + self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
624 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
625 }
626
627 /**
628 * Whether this language makes distinguishes genders for example in
629 * namespaces.
630 * @return bool
631 * @since 1.18
632 */
633 function needsGenderDistinction() {
634 global $wgExtraGenderNamespaces, $wgExtraNamespaces;
635 if ( count( $wgExtraGenderNamespaces ) > 0 ) {
636 // $wgExtraGenderNamespaces overrides everything
637 return true;
638 } elseif ( isset( $wgExtraNamespaces[NS_USER] ) && isset( $wgExtraNamespaces[NS_USER_TALK] ) ) {
639 /// @todo There may be other gender namespace than NS_USER & NS_USER_TALK in the future
640 // $wgExtraNamespaces overrides any gender aliases specified in i18n files
641 return false;
642 } else {
643 // Check what is in i18n files
644 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
645 return count( $aliases ) > 0;
646 }
647 }
648
649 /**
650 * Get a namespace key by value, case insensitive.
651 * Only matches namespace names for the current language, not the
652 * canonical ones defined in Namespace.php.
653 *
654 * @param $text String
655 * @return mixed An integer if $text is a valid value otherwise false
656 */
657 function getLocalNsIndex( $text ) {
658 $lctext = $this->lc( $text );
659 $ids = $this->getNamespaceIds();
660 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
661 }
662
663 /**
664 * @return array
665 */
666 function getNamespaceAliases() {
667 if ( is_null( $this->namespaceAliases ) ) {
668 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
669 if ( !$aliases ) {
670 $aliases = array();
671 } else {
672 foreach ( $aliases as $name => $index ) {
673 if ( $index === NS_PROJECT_TALK ) {
674 unset( $aliases[$name] );
675 $name = $this->fixVariableInNamespace( $name );
676 $aliases[$name] = $index;
677 }
678 }
679 }
680
681 global $wgExtraGenderNamespaces;
682 $genders = $wgExtraGenderNamespaces + (array)self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
683 foreach ( $genders as $index => $forms ) {
684 foreach ( $forms as $alias ) {
685 $aliases[$alias] = $index;
686 }
687 }
688
689 $this->namespaceAliases = $aliases;
690 }
691 return $this->namespaceAliases;
692 }
693
694 /**
695 * @return array
696 */
697 function getNamespaceIds() {
698 if ( is_null( $this->mNamespaceIds ) ) {
699 global $wgNamespaceAliases;
700 # Put namespace names and aliases into a hashtable.
701 # If this is too slow, then we should arrange it so that it is done
702 # before caching. The catch is that at pre-cache time, the above
703 # class-specific fixup hasn't been done.
704 $this->mNamespaceIds = array();
705 foreach ( $this->getNamespaces() as $index => $name ) {
706 $this->mNamespaceIds[$this->lc( $name )] = $index;
707 }
708 foreach ( $this->getNamespaceAliases() as $name => $index ) {
709 $this->mNamespaceIds[$this->lc( $name )] = $index;
710 }
711 if ( $wgNamespaceAliases ) {
712 foreach ( $wgNamespaceAliases as $name => $index ) {
713 $this->mNamespaceIds[$this->lc( $name )] = $index;
714 }
715 }
716 }
717 return $this->mNamespaceIds;
718 }
719
720 /**
721 * Get a namespace key by value, case insensitive. Canonical namespace
722 * names override custom ones defined for the current language.
723 *
724 * @param $text String
725 * @return mixed An integer if $text is a valid value otherwise false
726 */
727 function getNsIndex( $text ) {
728 $lctext = $this->lc( $text );
729 $ns = MWNamespace::getCanonicalIndex( $lctext );
730 if ( $ns !== null ) {
731 return $ns;
732 }
733 $ids = $this->getNamespaceIds();
734 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
735 }
736
737 /**
738 * short names for language variants used for language conversion links.
739 *
740 * @param $code String
741 * @param $usemsg bool Use the "variantname-xyz" message if it exists
742 * @return string
743 */
744 function getVariantname( $code, $usemsg = true ) {
745 $msg = "variantname-$code";
746 if ( $usemsg && wfMessage( $msg )->exists() ) {
747 return $this->getMessageFromDB( $msg );
748 }
749 $name = self::fetchLanguageName( $code );
750 if ( $name ) {
751 return $name; # if it's defined as a language name, show that
752 } else {
753 # otherwise, output the language code
754 return $code;
755 }
756 }
757
758 /**
759 * @param $name string
760 * @return string
761 */
762 function specialPage( $name ) {
763 $aliases = $this->getSpecialPageAliases();
764 if ( isset( $aliases[$name][0] ) ) {
765 $name = $aliases[$name][0];
766 }
767 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
768 }
769
770 /**
771 * @return array
772 */
773 function getDatePreferences() {
774 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
775 }
776
777 /**
778 * @return array
779 */
780 function getDateFormats() {
781 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
782 }
783
784 /**
785 * @return array|string
786 */
787 function getDefaultDateFormat() {
788 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
789 if ( $df === 'dmy or mdy' ) {
790 global $wgAmericanDates;
791 return $wgAmericanDates ? 'mdy' : 'dmy';
792 } else {
793 return $df;
794 }
795 }
796
797 /**
798 * @return array
799 */
800 function getDatePreferenceMigrationMap() {
801 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
802 }
803
804 /**
805 * @param $image
806 * @return array|null
807 */
808 function getImageFile( $image ) {
809 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
810 }
811
812 /**
813 * @return array
814 */
815 function getExtraUserToggles() {
816 return (array)self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
817 }
818
819 /**
820 * @param $tog
821 * @return string
822 */
823 function getUserToggle( $tog ) {
824 return $this->getMessageFromDB( "tog-$tog" );
825 }
826
827 /**
828 * Get native language names, indexed by code.
829 * Only those defined in MediaWiki, no other data like CLDR.
830 * If $customisedOnly is true, only returns codes with a messages file
831 *
832 * @param $customisedOnly bool
833 *
834 * @return array
835 * @deprecated in 1.20, use fetchLanguageNames()
836 */
837 public static function getLanguageNames( $customisedOnly = false ) {
838 return self::fetchLanguageNames( null, $customisedOnly ? 'mwfile' : 'mw' );
839 }
840
841 /**
842 * Get translated language names. This is done on best effort and
843 * by default this is exactly the same as Language::getLanguageNames.
844 * The CLDR extension provides translated names.
845 * @param $code String Language code.
846 * @return Array language code => language name
847 * @since 1.18.0
848 * @deprecated in 1.20, use fetchLanguageNames()
849 */
850 public static function getTranslatedLanguageNames( $code ) {
851 return self::fetchLanguageNames( $code, 'all' );
852 }
853
854 /**
855 * Get an array of language names, indexed by code.
856 * @param $inLanguage null|string: Code of language in which to return the names
857 * Use null for autonyms (native names)
858 * @param $include string:
859 * 'all' all available languages
860 * 'mw' only if the language is defined in MediaWiki or wgExtraLanguageNames (default)
861 * 'mwfile' only if the language is in 'mw' *and* has a message file
862 * @return array: language code => language name
863 * @since 1.20
864 */
865 public static function fetchLanguageNames( $inLanguage = null, $include = 'mw' ) {
866 global $wgExtraLanguageNames;
867 static $coreLanguageNames;
868
869 if ( $coreLanguageNames === null ) {
870 include MWInit::compiledPath( 'languages/Names.php' );
871 }
872
873 $names = array();
874
875 if ( $inLanguage ) {
876 # TODO: also include when $inLanguage is null, when this code is more efficient
877 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $inLanguage ) );
878 }
879
880 $mwNames = $wgExtraLanguageNames + $coreLanguageNames;
881 foreach ( $mwNames as $mwCode => $mwName ) {
882 # - Prefer own MediaWiki native name when not using the hook
883 # - For other names just add if not added through the hook
884 if ( $mwCode === $inLanguage || !isset( $names[$mwCode] ) ) {
885 $names[$mwCode] = $mwName;
886 }
887 }
888
889 if ( $include === 'all' ) {
890 return $names;
891 }
892
893 $returnMw = array();
894 $coreCodes = array_keys( $mwNames );
895 foreach ( $coreCodes as $coreCode ) {
896 $returnMw[$coreCode] = $names[$coreCode];
897 }
898
899 if ( $include === 'mwfile' ) {
900 $namesMwFile = array();
901 # We do this using a foreach over the codes instead of a directory
902 # loop so that messages files in extensions will work correctly.
903 foreach ( $returnMw as $code => $value ) {
904 if ( is_readable( self::getMessagesFileName( $code ) ) ) {
905 $namesMwFile[$code] = $names[$code];
906 }
907 }
908 return $namesMwFile;
909 }
910 # 'mw' option; default if it's not one of the other two options (all/mwfile)
911 return $returnMw;
912 }
913
914 /**
915 * @param $code string: The code of the language for which to get the name
916 * @param $inLanguage null|string: Code of language in which to return the name (null for autonyms)
917 * @param $include string: 'all', 'mw' or 'mwfile'; see fetchLanguageNames()
918 * @return string: Language name or empty
919 * @since 1.20
920 */
921 public static function fetchLanguageName( $code, $inLanguage = null, $include = 'all' ) {
922 $array = self::fetchLanguageNames( $inLanguage, $include );
923 return !array_key_exists( $code, $array ) ? '' : $array[$code];
924 }
925
926 /**
927 * Get a message from the MediaWiki namespace.
928 *
929 * @param $msg String: message name
930 * @return string
931 */
932 function getMessageFromDB( $msg ) {
933 return wfMessage( $msg )->inLanguage( $this )->text();
934 }
935
936 /**
937 * Get the native language name of $code.
938 * Only if defined in MediaWiki, no other data like CLDR.
939 * @param $code string
940 * @return string
941 * @deprecated in 1.20, use fetchLanguageName()
942 */
943 function getLanguageName( $code ) {
944 return self::fetchLanguageName( $code );
945 }
946
947 /**
948 * @param $key string
949 * @return string
950 */
951 function getMonthName( $key ) {
952 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
953 }
954
955 /**
956 * @return array
957 */
958 function getMonthNamesArray() {
959 $monthNames = array( '' );
960 for ( $i = 1; $i < 13; $i++ ) {
961 $monthNames[] = $this->getMonthName( $i );
962 }
963 return $monthNames;
964 }
965
966 /**
967 * @param $key string
968 * @return string
969 */
970 function getMonthNameGen( $key ) {
971 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
972 }
973
974 /**
975 * @param $key string
976 * @return string
977 */
978 function getMonthAbbreviation( $key ) {
979 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
980 }
981
982 /**
983 * @return array
984 */
985 function getMonthAbbreviationsArray() {
986 $monthNames = array( '' );
987 for ( $i = 1; $i < 13; $i++ ) {
988 $monthNames[] = $this->getMonthAbbreviation( $i );
989 }
990 return $monthNames;
991 }
992
993 /**
994 * @param $key string
995 * @return string
996 */
997 function getWeekdayName( $key ) {
998 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
999 }
1000
1001 /**
1002 * @param $key string
1003 * @return string
1004 */
1005 function getWeekdayAbbreviation( $key ) {
1006 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
1007 }
1008
1009 /**
1010 * @param $key string
1011 * @return string
1012 */
1013 function getIranianCalendarMonthName( $key ) {
1014 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
1015 }
1016
1017 /**
1018 * @param $key string
1019 * @return string
1020 */
1021 function getHebrewCalendarMonthName( $key ) {
1022 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
1023 }
1024
1025 /**
1026 * @param $key string
1027 * @return string
1028 */
1029 function getHebrewCalendarMonthNameGen( $key ) {
1030 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
1031 }
1032
1033 /**
1034 * @param $key string
1035 * @return string
1036 */
1037 function getHijriCalendarMonthName( $key ) {
1038 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
1039 }
1040
1041 /**
1042 * This is a workalike of PHP's date() function, but with better
1043 * internationalisation, a reduced set of format characters, and a better
1044 * escaping format.
1045 *
1046 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrUeIOPTZ. See
1047 * the PHP manual for definitions. There are a number of extensions, which
1048 * start with "x":
1049 *
1050 * xn Do not translate digits of the next numeric format character
1051 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
1052 * xr Use roman numerals for the next numeric format character
1053 * xh Use hebrew numerals for the next numeric format character
1054 * xx Literal x
1055 * xg Genitive month name
1056 *
1057 * xij j (day number) in Iranian calendar
1058 * xiF F (month name) in Iranian calendar
1059 * xin n (month number) in Iranian calendar
1060 * xiy y (two digit year) in Iranian calendar
1061 * xiY Y (full year) in Iranian calendar
1062 *
1063 * xjj j (day number) in Hebrew calendar
1064 * xjF F (month name) in Hebrew calendar
1065 * xjt t (days in month) in Hebrew calendar
1066 * xjx xg (genitive month name) in Hebrew calendar
1067 * xjn n (month number) in Hebrew calendar
1068 * xjY Y (full year) in Hebrew calendar
1069 *
1070 * xmj j (day number) in Hijri calendar
1071 * xmF F (month name) in Hijri calendar
1072 * xmn n (month number) in Hijri calendar
1073 * xmY Y (full year) in Hijri calendar
1074 *
1075 * xkY Y (full year) in Thai solar calendar. Months and days are
1076 * identical to the Gregorian calendar
1077 * xoY Y (full year) in Minguo calendar or Juche year.
1078 * Months and days are identical to the
1079 * Gregorian calendar
1080 * xtY Y (full year) in Japanese nengo. Months and days are
1081 * identical to the Gregorian calendar
1082 *
1083 * Characters enclosed in double quotes will be considered literal (with
1084 * the quotes themselves removed). Unmatched quotes will be considered
1085 * literal quotes. Example:
1086 *
1087 * "The month is" F => The month is January
1088 * i's" => 20'11"
1089 *
1090 * Backslash escaping is also supported.
1091 *
1092 * Input timestamp is assumed to be pre-normalized to the desired local
1093 * time zone, if any. Note that the format characters crUeIOPTZ will assume
1094 * $ts is UTC if $zone is not given.
1095 *
1096 * @param $format String
1097 * @param $ts String: 14-character timestamp
1098 * YYYYMMDDHHMMSS
1099 * 01234567890123
1100 * @param $zone DateTimeZone: Timezone of $ts
1101 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
1102 *
1103 * @throws MWException
1104 * @return string
1105 */
1106 function sprintfDate( $format, $ts, DateTimeZone $zone = null ) {
1107 $s = '';
1108 $raw = false;
1109 $roman = false;
1110 $hebrewNum = false;
1111 $dateTimeObj = false;
1112 $rawToggle = false;
1113 $iranian = false;
1114 $hebrew = false;
1115 $hijri = false;
1116 $thai = false;
1117 $minguo = false;
1118 $tenno = false;
1119
1120 if ( strlen( $ts ) !== 14 ) {
1121 throw new MWException( __METHOD__ . ": The timestamp $ts should have 14 characters" );
1122 }
1123
1124 if ( !ctype_digit( $ts ) ) {
1125 throw new MWException( __METHOD__ . ": The timestamp $ts should be a number" );
1126 }
1127
1128 for ( $p = 0; $p < strlen( $format ); $p++ ) {
1129 $num = false;
1130 $code = $format[$p];
1131 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
1132 $code .= $format[++$p];
1133 }
1134
1135 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
1136 $code .= $format[++$p];
1137 }
1138
1139 switch ( $code ) {
1140 case 'xx':
1141 $s .= 'x';
1142 break;
1143 case 'xn':
1144 $raw = true;
1145 break;
1146 case 'xN':
1147 $rawToggle = !$rawToggle;
1148 break;
1149 case 'xr':
1150 $roman = true;
1151 break;
1152 case 'xh':
1153 $hebrewNum = true;
1154 break;
1155 case 'xg':
1156 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
1157 break;
1158 case 'xjx':
1159 if ( !$hebrew ) {
1160 $hebrew = self::tsToHebrew( $ts );
1161 }
1162 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
1163 break;
1164 case 'd':
1165 $num = substr( $ts, 6, 2 );
1166 break;
1167 case 'D':
1168 if ( !$dateTimeObj ) {
1169 $dateTimeObj = DateTime::createFromFormat(
1170 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1171 );
1172 }
1173 $s .= $this->getWeekdayAbbreviation( $dateTimeObj->format( 'w' ) + 1 );
1174 break;
1175 case 'j':
1176 $num = intval( substr( $ts, 6, 2 ) );
1177 break;
1178 case 'xij':
1179 if ( !$iranian ) {
1180 $iranian = self::tsToIranian( $ts );
1181 }
1182 $num = $iranian[2];
1183 break;
1184 case 'xmj':
1185 if ( !$hijri ) {
1186 $hijri = self::tsToHijri( $ts );
1187 }
1188 $num = $hijri[2];
1189 break;
1190 case 'xjj':
1191 if ( !$hebrew ) {
1192 $hebrew = self::tsToHebrew( $ts );
1193 }
1194 $num = $hebrew[2];
1195 break;
1196 case 'l':
1197 if ( !$dateTimeObj ) {
1198 $dateTimeObj = DateTime::createFromFormat(
1199 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1200 );
1201 }
1202 $s .= $this->getWeekdayName( $dateTimeObj->format( 'w' ) + 1 );
1203 break;
1204 case 'F':
1205 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
1206 break;
1207 case 'xiF':
1208 if ( !$iranian ) {
1209 $iranian = self::tsToIranian( $ts );
1210 }
1211 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
1212 break;
1213 case 'xmF':
1214 if ( !$hijri ) {
1215 $hijri = self::tsToHijri( $ts );
1216 }
1217 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
1218 break;
1219 case 'xjF':
1220 if ( !$hebrew ) {
1221 $hebrew = self::tsToHebrew( $ts );
1222 }
1223 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
1224 break;
1225 case 'm':
1226 $num = substr( $ts, 4, 2 );
1227 break;
1228 case 'M':
1229 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
1230 break;
1231 case 'n':
1232 $num = intval( substr( $ts, 4, 2 ) );
1233 break;
1234 case 'xin':
1235 if ( !$iranian ) {
1236 $iranian = self::tsToIranian( $ts );
1237 }
1238 $num = $iranian[1];
1239 break;
1240 case 'xmn':
1241 if ( !$hijri ) {
1242 $hijri = self::tsToHijri ( $ts );
1243 }
1244 $num = $hijri[1];
1245 break;
1246 case 'xjn':
1247 if ( !$hebrew ) {
1248 $hebrew = self::tsToHebrew( $ts );
1249 }
1250 $num = $hebrew[1];
1251 break;
1252 case 'xjt':
1253 if ( !$hebrew ) {
1254 $hebrew = self::tsToHebrew( $ts );
1255 }
1256 $num = $hebrew[3];
1257 break;
1258 case 'Y':
1259 $num = substr( $ts, 0, 4 );
1260 break;
1261 case 'xiY':
1262 if ( !$iranian ) {
1263 $iranian = self::tsToIranian( $ts );
1264 }
1265 $num = $iranian[0];
1266 break;
1267 case 'xmY':
1268 if ( !$hijri ) {
1269 $hijri = self::tsToHijri( $ts );
1270 }
1271 $num = $hijri[0];
1272 break;
1273 case 'xjY':
1274 if ( !$hebrew ) {
1275 $hebrew = self::tsToHebrew( $ts );
1276 }
1277 $num = $hebrew[0];
1278 break;
1279 case 'xkY':
1280 if ( !$thai ) {
1281 $thai = self::tsToYear( $ts, 'thai' );
1282 }
1283 $num = $thai[0];
1284 break;
1285 case 'xoY':
1286 if ( !$minguo ) {
1287 $minguo = self::tsToYear( $ts, 'minguo' );
1288 }
1289 $num = $minguo[0];
1290 break;
1291 case 'xtY':
1292 if ( !$tenno ) {
1293 $tenno = self::tsToYear( $ts, 'tenno' );
1294 }
1295 $num = $tenno[0];
1296 break;
1297 case 'y':
1298 $num = substr( $ts, 2, 2 );
1299 break;
1300 case 'xiy':
1301 if ( !$iranian ) {
1302 $iranian = self::tsToIranian( $ts );
1303 }
1304 $num = substr( $iranian[0], -2 );
1305 break;
1306 case 'a':
1307 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
1308 break;
1309 case 'A':
1310 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
1311 break;
1312 case 'g':
1313 $h = substr( $ts, 8, 2 );
1314 $num = $h % 12 ? $h % 12 : 12;
1315 break;
1316 case 'G':
1317 $num = intval( substr( $ts, 8, 2 ) );
1318 break;
1319 case 'h':
1320 $h = substr( $ts, 8, 2 );
1321 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
1322 break;
1323 case 'H':
1324 $num = substr( $ts, 8, 2 );
1325 break;
1326 case 'i':
1327 $num = substr( $ts, 10, 2 );
1328 break;
1329 case 's':
1330 $num = substr( $ts, 12, 2 );
1331 break;
1332 case 'c':
1333 case 'r':
1334 case 'e':
1335 case 'O':
1336 case 'P':
1337 case 'T':
1338 // Pass through string from $dateTimeObj->format()
1339 if ( !$dateTimeObj ) {
1340 $dateTimeObj = DateTime::createFromFormat(
1341 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1342 );
1343 }
1344 $s .= $dateTimeObj->format( $code );
1345 break;
1346 case 'w':
1347 case 'N':
1348 case 'z':
1349 case 'W':
1350 case 't':
1351 case 'L':
1352 case 'o':
1353 case 'U':
1354 case 'I':
1355 case 'Z':
1356 // Pass through number from $dateTimeObj->format()
1357 if ( !$dateTimeObj ) {
1358 $dateTimeObj = DateTime::createFromFormat(
1359 'YmdHis', $ts, $zone ?: new DateTimeZone( 'UTC' )
1360 );
1361 }
1362 $num = $dateTimeObj->format( $code );
1363 break;
1364 case '\\':
1365 # Backslash escaping
1366 if ( $p < strlen( $format ) - 1 ) {
1367 $s .= $format[++$p];
1368 } else {
1369 $s .= '\\';
1370 }
1371 break;
1372 case '"':
1373 # Quoted literal
1374 if ( $p < strlen( $format ) - 1 ) {
1375 $endQuote = strpos( $format, '"', $p + 1 );
1376 if ( $endQuote === false ) {
1377 # No terminating quote, assume literal "
1378 $s .= '"';
1379 } else {
1380 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1381 $p = $endQuote;
1382 }
1383 } else {
1384 # Quote at end of string, assume literal "
1385 $s .= '"';
1386 }
1387 break;
1388 default:
1389 $s .= $format[$p];
1390 }
1391 if ( $num !== false ) {
1392 if ( $rawToggle || $raw ) {
1393 $s .= $num;
1394 $raw = false;
1395 } elseif ( $roman ) {
1396 $s .= Language::romanNumeral( $num );
1397 $roman = false;
1398 } elseif ( $hebrewNum ) {
1399 $s .= self::hebrewNumeral( $num );
1400 $hebrewNum = false;
1401 } else {
1402 $s .= $this->formatNum( $num, true );
1403 }
1404 }
1405 }
1406 return $s;
1407 }
1408
1409 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1410 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1411
1412 /**
1413 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1414 * Gregorian dates to Iranian dates. Originally written in C, it
1415 * is released under the terms of GNU Lesser General Public
1416 * License. Conversion to PHP was performed by Niklas Laxström.
1417 *
1418 * Link: http://www.farsiweb.info/jalali/jalali.c
1419 *
1420 * @param $ts string
1421 *
1422 * @return string
1423 */
1424 private static function tsToIranian( $ts ) {
1425 $gy = substr( $ts, 0, 4 ) -1600;
1426 $gm = substr( $ts, 4, 2 ) -1;
1427 $gd = substr( $ts, 6, 2 ) -1;
1428
1429 # Days passed from the beginning (including leap years)
1430 $gDayNo = 365 * $gy
1431 + floor( ( $gy + 3 ) / 4 )
1432 - floor( ( $gy + 99 ) / 100 )
1433 + floor( ( $gy + 399 ) / 400 );
1434
1435 // Add days of the past months of this year
1436 for ( $i = 0; $i < $gm; $i++ ) {
1437 $gDayNo += self::$GREG_DAYS[$i];
1438 }
1439
1440 // Leap years
1441 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1442 $gDayNo++;
1443 }
1444
1445 // Days passed in current month
1446 $gDayNo += (int)$gd;
1447
1448 $jDayNo = $gDayNo - 79;
1449
1450 $jNp = floor( $jDayNo / 12053 );
1451 $jDayNo %= 12053;
1452
1453 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1454 $jDayNo %= 1461;
1455
1456 if ( $jDayNo >= 366 ) {
1457 $jy += floor( ( $jDayNo - 1 ) / 365 );
1458 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1459 }
1460
1461 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1462 $jDayNo -= self::$IRANIAN_DAYS[$i];
1463 }
1464
1465 $jm = $i + 1;
1466 $jd = $jDayNo + 1;
1467
1468 return array( $jy, $jm, $jd );
1469 }
1470
1471 /**
1472 * Converting Gregorian dates to Hijri dates.
1473 *
1474 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1475 *
1476 * @see http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1477 *
1478 * @param $ts string
1479 *
1480 * @return string
1481 */
1482 private static function tsToHijri( $ts ) {
1483 $year = substr( $ts, 0, 4 );
1484 $month = substr( $ts, 4, 2 );
1485 $day = substr( $ts, 6, 2 );
1486
1487 $zyr = $year;
1488 $zd = $day;
1489 $zm = $month;
1490 $zy = $zyr;
1491
1492 if (
1493 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1494 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1495 )
1496 {
1497 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1498 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1499 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1500 $zd - 32075;
1501 } else {
1502 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1503 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1504 }
1505
1506 $zl = $zjd -1948440 + 10632;
1507 $zn = (int)( ( $zl - 1 ) / 10631 );
1508 $zl = $zl - 10631 * $zn + 354;
1509 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1510 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1511 $zm = (int)( ( 24 * $zl ) / 709 );
1512 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1513 $zy = 30 * $zn + $zj - 30;
1514
1515 return array( $zy, $zm, $zd );
1516 }
1517
1518 /**
1519 * Converting Gregorian dates to Hebrew dates.
1520 *
1521 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1522 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1523 * to translate the relevant functions into PHP and release them under
1524 * GNU GPL.
1525 *
1526 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1527 * and Adar II is 14. In a non-leap year, Adar is 6.
1528 *
1529 * @param $ts string
1530 *
1531 * @return string
1532 */
1533 private static function tsToHebrew( $ts ) {
1534 # Parse date
1535 $year = substr( $ts, 0, 4 );
1536 $month = substr( $ts, 4, 2 );
1537 $day = substr( $ts, 6, 2 );
1538
1539 # Calculate Hebrew year
1540 $hebrewYear = $year + 3760;
1541
1542 # Month number when September = 1, August = 12
1543 $month += 4;
1544 if ( $month > 12 ) {
1545 # Next year
1546 $month -= 12;
1547 $year++;
1548 $hebrewYear++;
1549 }
1550
1551 # Calculate day of year from 1 September
1552 $dayOfYear = $day;
1553 for ( $i = 1; $i < $month; $i++ ) {
1554 if ( $i == 6 ) {
1555 # February
1556 $dayOfYear += 28;
1557 # Check if the year is leap
1558 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1559 $dayOfYear++;
1560 }
1561 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1562 $dayOfYear += 30;
1563 } else {
1564 $dayOfYear += 31;
1565 }
1566 }
1567
1568 # Calculate the start of the Hebrew year
1569 $start = self::hebrewYearStart( $hebrewYear );
1570
1571 # Calculate next year's start
1572 if ( $dayOfYear <= $start ) {
1573 # Day is before the start of the year - it is the previous year
1574 # Next year's start
1575 $nextStart = $start;
1576 # Previous year
1577 $year--;
1578 $hebrewYear--;
1579 # Add days since previous year's 1 September
1580 $dayOfYear += 365;
1581 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1582 # Leap year
1583 $dayOfYear++;
1584 }
1585 # Start of the new (previous) year
1586 $start = self::hebrewYearStart( $hebrewYear );
1587 } else {
1588 # Next year's start
1589 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1590 }
1591
1592 # Calculate Hebrew day of year
1593 $hebrewDayOfYear = $dayOfYear - $start;
1594
1595 # Difference between year's days
1596 $diff = $nextStart - $start;
1597 # Add 12 (or 13 for leap years) days to ignore the difference between
1598 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1599 # difference is only about the year type
1600 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1601 $diff += 13;
1602 } else {
1603 $diff += 12;
1604 }
1605
1606 # Check the year pattern, and is leap year
1607 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1608 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1609 # and non-leap years
1610 $yearPattern = $diff % 30;
1611 # Check if leap year
1612 $isLeap = $diff >= 30;
1613
1614 # Calculate day in the month from number of day in the Hebrew year
1615 # Don't check Adar - if the day is not in Adar, we will stop before;
1616 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1617 $hebrewDay = $hebrewDayOfYear;
1618 $hebrewMonth = 1;
1619 $days = 0;
1620 while ( $hebrewMonth <= 12 ) {
1621 # Calculate days in this month
1622 if ( $isLeap && $hebrewMonth == 6 ) {
1623 # Adar in a leap year
1624 if ( $isLeap ) {
1625 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1626 $days = 30;
1627 if ( $hebrewDay <= $days ) {
1628 # Day in Adar I
1629 $hebrewMonth = 13;
1630 } else {
1631 # Subtract the days of Adar I
1632 $hebrewDay -= $days;
1633 # Try Adar II
1634 $days = 29;
1635 if ( $hebrewDay <= $days ) {
1636 # Day in Adar II
1637 $hebrewMonth = 14;
1638 }
1639 }
1640 }
1641 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1642 # Cheshvan in a complete year (otherwise as the rule below)
1643 $days = 30;
1644 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1645 # Kislev in an incomplete year (otherwise as the rule below)
1646 $days = 29;
1647 } else {
1648 # Odd months have 30 days, even have 29
1649 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1650 }
1651 if ( $hebrewDay <= $days ) {
1652 # In the current month
1653 break;
1654 } else {
1655 # Subtract the days of the current month
1656 $hebrewDay -= $days;
1657 # Try in the next month
1658 $hebrewMonth++;
1659 }
1660 }
1661
1662 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1663 }
1664
1665 /**
1666 * This calculates the Hebrew year start, as days since 1 September.
1667 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1668 * Used for Hebrew date.
1669 *
1670 * @param $year int
1671 *
1672 * @return string
1673 */
1674 private static function hebrewYearStart( $year ) {
1675 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1676 $b = intval( ( $year - 1 ) % 4 );
1677 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1678 if ( $m < 0 ) {
1679 $m--;
1680 }
1681 $Mar = intval( $m );
1682 if ( $m < 0 ) {
1683 $m++;
1684 }
1685 $m -= $Mar;
1686
1687 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1688 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1689 $Mar++;
1690 } elseif ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1691 $Mar += 2;
1692 } elseif ( $c == 2 || $c == 4 || $c == 6 ) {
1693 $Mar++;
1694 }
1695
1696 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1697 return $Mar;
1698 }
1699
1700 /**
1701 * Algorithm to convert Gregorian dates to Thai solar dates,
1702 * Minguo dates or Minguo dates.
1703 *
1704 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1705 * http://en.wikipedia.org/wiki/Minguo_calendar
1706 * http://en.wikipedia.org/wiki/Japanese_era_name
1707 *
1708 * @param $ts String: 14-character timestamp
1709 * @param $cName String: calender name
1710 * @return Array: converted year, month, day
1711 */
1712 private static function tsToYear( $ts, $cName ) {
1713 $gy = substr( $ts, 0, 4 );
1714 $gm = substr( $ts, 4, 2 );
1715 $gd = substr( $ts, 6, 2 );
1716
1717 if ( !strcmp( $cName, 'thai' ) ) {
1718 # Thai solar dates
1719 # Add 543 years to the Gregorian calendar
1720 # Months and days are identical
1721 $gy_offset = $gy + 543;
1722 } elseif ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1723 # Minguo dates
1724 # Deduct 1911 years from the Gregorian calendar
1725 # Months and days are identical
1726 $gy_offset = $gy - 1911;
1727 } elseif ( !strcmp( $cName, 'tenno' ) ) {
1728 # Nengō dates up to Meiji period
1729 # Deduct years from the Gregorian calendar
1730 # depending on the nengo periods
1731 # Months and days are identical
1732 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1733 # Meiji period
1734 $gy_gannen = $gy - 1868 + 1;
1735 $gy_offset = $gy_gannen;
1736 if ( $gy_gannen == 1 ) {
1737 $gy_offset = '元';
1738 }
1739 $gy_offset = '明治' . $gy_offset;
1740 } elseif (
1741 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1742 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1743 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1744 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1745 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1746 )
1747 {
1748 # Taishō period
1749 $gy_gannen = $gy - 1912 + 1;
1750 $gy_offset = $gy_gannen;
1751 if ( $gy_gannen == 1 ) {
1752 $gy_offset = '元';
1753 }
1754 $gy_offset = '大正' . $gy_offset;
1755 } elseif (
1756 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1757 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1758 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1759 )
1760 {
1761 # Shōwa period
1762 $gy_gannen = $gy - 1926 + 1;
1763 $gy_offset = $gy_gannen;
1764 if ( $gy_gannen == 1 ) {
1765 $gy_offset = '元';
1766 }
1767 $gy_offset = '昭和' . $gy_offset;
1768 } else {
1769 # Heisei period
1770 $gy_gannen = $gy - 1989 + 1;
1771 $gy_offset = $gy_gannen;
1772 if ( $gy_gannen == 1 ) {
1773 $gy_offset = '元';
1774 }
1775 $gy_offset = '平成' . $gy_offset;
1776 }
1777 } else {
1778 $gy_offset = $gy;
1779 }
1780
1781 return array( $gy_offset, $gm, $gd );
1782 }
1783
1784 /**
1785 * Roman number formatting up to 10000
1786 *
1787 * @param $num int
1788 *
1789 * @return string
1790 */
1791 static function romanNumeral( $num ) {
1792 static $table = array(
1793 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1794 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1795 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1796 array( '', 'M', 'MM', 'MMM', 'MMMM', 'MMMMM', 'MMMMMM', 'MMMMMMM', 'MMMMMMMM', 'MMMMMMMMM', 'MMMMMMMMMM' )
1797 );
1798
1799 $num = intval( $num );
1800 if ( $num > 10000 || $num <= 0 ) {
1801 return $num;
1802 }
1803
1804 $s = '';
1805 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1806 if ( $num >= $pow10 ) {
1807 $s .= $table[$i][(int)floor( $num / $pow10 )];
1808 }
1809 $num = $num % $pow10;
1810 }
1811 return $s;
1812 }
1813
1814 /**
1815 * Hebrew Gematria number formatting up to 9999
1816 *
1817 * @param $num int
1818 *
1819 * @return string
1820 */
1821 static function hebrewNumeral( $num ) {
1822 static $table = array(
1823 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1824 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1825 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1826 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1827 );
1828
1829 $num = intval( $num );
1830 if ( $num > 9999 || $num <= 0 ) {
1831 return $num;
1832 }
1833
1834 $s = '';
1835 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1836 if ( $num >= $pow10 ) {
1837 if ( $num == 15 || $num == 16 ) {
1838 $s .= $table[0][9] . $table[0][$num - 9];
1839 $num = 0;
1840 } else {
1841 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1842 if ( $pow10 == 1000 ) {
1843 $s .= "'";
1844 }
1845 }
1846 }
1847 $num = $num % $pow10;
1848 }
1849 if ( strlen( $s ) == 2 ) {
1850 $str = $s . "'";
1851 } else {
1852 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1853 $str .= substr( $s, strlen( $s ) - 2, 2 );
1854 }
1855 $start = substr( $str, 0, strlen( $str ) - 2 );
1856 $end = substr( $str, strlen( $str ) - 2 );
1857 switch ( $end ) {
1858 case 'כ':
1859 $str = $start . 'ך';
1860 break;
1861 case 'מ':
1862 $str = $start . 'ם';
1863 break;
1864 case 'נ':
1865 $str = $start . 'ן';
1866 break;
1867 case 'פ':
1868 $str = $start . 'ף';
1869 break;
1870 case 'צ':
1871 $str = $start . 'ץ';
1872 break;
1873 }
1874 return $str;
1875 }
1876
1877 /**
1878 * Used by date() and time() to adjust the time output.
1879 *
1880 * @param $ts Int the time in date('YmdHis') format
1881 * @param $tz Mixed: adjust the time by this amount (default false, mean we
1882 * get user timecorrection setting)
1883 * @return int
1884 */
1885 function userAdjust( $ts, $tz = false ) {
1886 global $wgUser, $wgLocalTZoffset;
1887
1888 if ( $tz === false ) {
1889 $tz = $wgUser->getOption( 'timecorrection' );
1890 }
1891
1892 $data = explode( '|', $tz, 3 );
1893
1894 if ( $data[0] == 'ZoneInfo' ) {
1895 wfSuppressWarnings();
1896 $userTZ = timezone_open( $data[2] );
1897 wfRestoreWarnings();
1898 if ( $userTZ !== false ) {
1899 $date = date_create( $ts, timezone_open( 'UTC' ) );
1900 date_timezone_set( $date, $userTZ );
1901 $date = date_format( $date, 'YmdHis' );
1902 return $date;
1903 }
1904 # Unrecognized timezone, default to 'Offset' with the stored offset.
1905 $data[0] = 'Offset';
1906 }
1907
1908 $minDiff = 0;
1909 if ( $data[0] == 'System' || $tz == '' ) {
1910 #  Global offset in minutes.
1911 if ( isset( $wgLocalTZoffset ) ) {
1912 $minDiff = $wgLocalTZoffset;
1913 }
1914 } elseif ( $data[0] == 'Offset' ) {
1915 $minDiff = intval( $data[1] );
1916 } else {
1917 $data = explode( ':', $tz );
1918 if ( count( $data ) == 2 ) {
1919 $data[0] = intval( $data[0] );
1920 $data[1] = intval( $data[1] );
1921 $minDiff = abs( $data[0] ) * 60 + $data[1];
1922 if ( $data[0] < 0 ) {
1923 $minDiff = -$minDiff;
1924 }
1925 } else {
1926 $minDiff = intval( $data[0] ) * 60;
1927 }
1928 }
1929
1930 # No difference ? Return time unchanged
1931 if ( 0 == $minDiff ) {
1932 return $ts;
1933 }
1934
1935 wfSuppressWarnings(); // E_STRICT system time bitching
1936 # Generate an adjusted date; take advantage of the fact that mktime
1937 # will normalize out-of-range values so we don't have to split $minDiff
1938 # into hours and minutes.
1939 $t = mktime( (
1940 (int)substr( $ts, 8, 2 ) ), # Hours
1941 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
1942 (int)substr( $ts, 12, 2 ), # Seconds
1943 (int)substr( $ts, 4, 2 ), # Month
1944 (int)substr( $ts, 6, 2 ), # Day
1945 (int)substr( $ts, 0, 4 ) ); # Year
1946
1947 $date = date( 'YmdHis', $t );
1948 wfRestoreWarnings();
1949
1950 return $date;
1951 }
1952
1953 /**
1954 * This is meant to be used by time(), date(), and timeanddate() to get
1955 * the date preference they're supposed to use, it should be used in
1956 * all children.
1957 *
1958 *<code>
1959 * function timeanddate([...], $format = true) {
1960 * $datePreference = $this->dateFormat($format);
1961 * [...]
1962 * }
1963 *</code>
1964 *
1965 * @param $usePrefs Mixed: if true, the user's preference is used
1966 * if false, the site/language default is used
1967 * if int/string, assumed to be a format.
1968 * @return string
1969 */
1970 function dateFormat( $usePrefs = true ) {
1971 global $wgUser;
1972
1973 if ( is_bool( $usePrefs ) ) {
1974 if ( $usePrefs ) {
1975 $datePreference = $wgUser->getDatePreference();
1976 } else {
1977 $datePreference = (string)User::getDefaultOption( 'date' );
1978 }
1979 } else {
1980 $datePreference = (string)$usePrefs;
1981 }
1982
1983 // return int
1984 if ( $datePreference == '' ) {
1985 return 'default';
1986 }
1987
1988 return $datePreference;
1989 }
1990
1991 /**
1992 * Get a format string for a given type and preference
1993 * @param $type string May be date, time or both
1994 * @param $pref string The format name as it appears in Messages*.php
1995 *
1996 * @since 1.22 New type 'pretty' that provides a more readable timestamp format
1997 *
1998 * @return string
1999 */
2000 function getDateFormatString( $type, $pref ) {
2001 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
2002 if ( $pref == 'default' ) {
2003 $pref = $this->getDefaultDateFormat();
2004 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2005 } else {
2006 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2007
2008 if ( $type === 'pretty' && $df === null ) {
2009 $df = $this->getDateFormatString( 'date', $pref );
2010 }
2011
2012 if ( $df === null ) {
2013 $pref = $this->getDefaultDateFormat();
2014 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
2015 }
2016 }
2017 $this->dateFormatStrings[$type][$pref] = $df;
2018 }
2019 return $this->dateFormatStrings[$type][$pref];
2020 }
2021
2022 /**
2023 * @param $ts Mixed: the time format which needs to be turned into a
2024 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2025 * @param $adj Bool: whether to adjust the time output according to the
2026 * user configured offset ($timecorrection)
2027 * @param $format Mixed: true to use user's date format preference
2028 * @param $timecorrection String|bool the time offset as returned by
2029 * validateTimeZone() in Special:Preferences
2030 * @return string
2031 */
2032 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
2033 $ts = wfTimestamp( TS_MW, $ts );
2034 if ( $adj ) {
2035 $ts = $this->userAdjust( $ts, $timecorrection );
2036 }
2037 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
2038 return $this->sprintfDate( $df, $ts );
2039 }
2040
2041 /**
2042 * @param $ts Mixed: the time format which needs to be turned into a
2043 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2044 * @param $adj Bool: whether to adjust the time output according to the
2045 * user configured offset ($timecorrection)
2046 * @param $format Mixed: true to use user's date format preference
2047 * @param $timecorrection String|bool the time offset as returned by
2048 * validateTimeZone() in Special:Preferences
2049 * @return string
2050 */
2051 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
2052 $ts = wfTimestamp( TS_MW, $ts );
2053 if ( $adj ) {
2054 $ts = $this->userAdjust( $ts, $timecorrection );
2055 }
2056 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
2057 return $this->sprintfDate( $df, $ts );
2058 }
2059
2060 /**
2061 * @param $ts Mixed: the time format which needs to be turned into a
2062 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2063 * @param $adj Bool: whether to adjust the time output according to the
2064 * user configured offset ($timecorrection)
2065 * @param $format Mixed: what format to return, if it's false output the
2066 * default one (default true)
2067 * @param $timecorrection String|bool the time offset as returned by
2068 * validateTimeZone() in Special:Preferences
2069 * @return string
2070 */
2071 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
2072 $ts = wfTimestamp( TS_MW, $ts );
2073 if ( $adj ) {
2074 $ts = $this->userAdjust( $ts, $timecorrection );
2075 }
2076 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
2077 return $this->sprintfDate( $df, $ts );
2078 }
2079
2080 /**
2081 * Takes a number of seconds and turns it into a text using values such as hours and minutes.
2082 *
2083 * @since 1.20
2084 *
2085 * @param integer $seconds The amount of seconds.
2086 * @param array $chosenIntervals The intervals to enable.
2087 *
2088 * @return string
2089 */
2090 public function formatDuration( $seconds, array $chosenIntervals = array() ) {
2091 $intervals = $this->getDurationIntervals( $seconds, $chosenIntervals );
2092
2093 $segments = array();
2094
2095 foreach ( $intervals as $intervalName => $intervalValue ) {
2096 // Messages: duration-seconds, duration-minutes, duration-hours, duration-days, duration-weeks,
2097 // duration-years, duration-decades, duration-centuries, duration-millennia
2098 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2099 $segments[] = $message->inLanguage( $this )->escaped();
2100 }
2101
2102 return $this->listToText( $segments );
2103 }
2104
2105 /**
2106 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2107 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2108 *
2109 * @since 1.20
2110 *
2111 * @param integer $seconds The amount of seconds.
2112 * @param array $chosenIntervals The intervals to enable.
2113 *
2114 * @return array
2115 */
2116 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2117 if ( empty( $chosenIntervals ) ) {
2118 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
2119 }
2120
2121 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2122 $sortedNames = array_keys( $intervals );
2123 $smallestInterval = array_pop( $sortedNames );
2124
2125 $segments = array();
2126
2127 foreach ( $intervals as $name => $length ) {
2128 $value = floor( $seconds / $length );
2129
2130 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2131 $seconds -= $value * $length;
2132 $segments[$name] = $value;
2133 }
2134 }
2135
2136 return $segments;
2137 }
2138
2139 /**
2140 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2141 *
2142 * @param $type String: can be 'date', 'time' or 'both'
2143 * @param $ts Mixed: the time format which needs to be turned into a
2144 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2145 * @param $user User object used to get preferences for timezone and format
2146 * @param $options Array, can contain the following keys:
2147 * - 'timecorrection': time correction, can have the following values:
2148 * - true: use user's preference
2149 * - false: don't use time correction
2150 * - integer: value of time correction in minutes
2151 * - 'format': format to use, can have the following values:
2152 * - true: use user's preference
2153 * - false: use default preference
2154 * - string: format to use
2155 * @since 1.19
2156 * @return String
2157 */
2158 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2159 $ts = wfTimestamp( TS_MW, $ts );
2160 $options += array( 'timecorrection' => true, 'format' => true );
2161 if ( $options['timecorrection'] !== false ) {
2162 if ( $options['timecorrection'] === true ) {
2163 $offset = $user->getOption( 'timecorrection' );
2164 } else {
2165 $offset = $options['timecorrection'];
2166 }
2167 $ts = $this->userAdjust( $ts, $offset );
2168 }
2169 if ( $options['format'] === true ) {
2170 $format = $user->getDatePreference();
2171 } else {
2172 $format = $options['format'];
2173 }
2174 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2175 return $this->sprintfDate( $df, $ts );
2176 }
2177
2178 /**
2179 * Get the formatted date for the given timestamp and formatted for
2180 * the given user.
2181 *
2182 * @param $ts Mixed: the time format which needs to be turned into a
2183 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2184 * @param $user User object used to get preferences for timezone and format
2185 * @param $options Array, can contain the following keys:
2186 * - 'timecorrection': time correction, can have the following values:
2187 * - true: use user's preference
2188 * - false: don't use time correction
2189 * - integer: value of time correction in minutes
2190 * - 'format': format to use, can have the following values:
2191 * - true: use user's preference
2192 * - false: use default preference
2193 * - string: format to use
2194 * @since 1.19
2195 * @return String
2196 */
2197 public function userDate( $ts, User $user, array $options = array() ) {
2198 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2199 }
2200
2201 /**
2202 * Get the formatted time for the given timestamp and formatted for
2203 * the given user.
2204 *
2205 * @param $ts Mixed: the time format which needs to be turned into a
2206 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2207 * @param $user User object used to get preferences for timezone and format
2208 * @param $options Array, can contain the following keys:
2209 * - 'timecorrection': time correction, can have the following values:
2210 * - true: use user's preference
2211 * - false: don't use time correction
2212 * - integer: value of time correction in minutes
2213 * - 'format': format to use, can have the following values:
2214 * - true: use user's preference
2215 * - false: use default preference
2216 * - string: format to use
2217 * @since 1.19
2218 * @return String
2219 */
2220 public function userTime( $ts, User $user, array $options = array() ) {
2221 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2222 }
2223
2224 /**
2225 * Get the formatted date and time for the given timestamp and formatted for
2226 * the given user.
2227 *
2228 * @param $ts Mixed: the time format which needs to be turned into a
2229 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2230 * @param $user User object used to get preferences for timezone and format
2231 * @param $options Array, can contain the following keys:
2232 * - 'timecorrection': time correction, can have the following values:
2233 * - true: use user's preference
2234 * - false: don't use time correction
2235 * - integer: value of time correction in minutes
2236 * - 'format': format to use, can have the following values:
2237 * - true: use user's preference
2238 * - false: use default preference
2239 * - string: format to use
2240 * @since 1.19
2241 * @return String
2242 */
2243 public function userTimeAndDate( $ts, User $user, array $options = array() ) {
2244 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2245 }
2246
2247 /**
2248 * Convert an MWTimestamp into a pretty human-readable timestamp using
2249 * the given user preferences and relative base time.
2250 *
2251 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2252 * on your timestamp object, which will then call this function. Calling
2253 * this function directly will cause hooks to be skipped over.
2254 *
2255 * @see MWTimestamp::getHumanTimestamp
2256 * @param MWTimestamp $ts Timestamp to prettify
2257 * @param MWTimestamp $relativeTo Base timestamp
2258 * @param User $user User preferences to use
2259 * @return string Human timestamp
2260 * @since 1.22
2261 */
2262 public function getHumanTimestamp( MWTimestamp $ts, MWTimestamp $relativeTo, User $user ) {
2263 $diff = $ts->diff( $relativeTo );
2264 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) - (int)$relativeTo->timestamp->format( 'w' ) );
2265 $days = $diff->days ?: (int)$diffDay;
2266 if ( $diff->invert || $days > 5 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' ) ) {
2267 // Timestamps are in different years: use full timestamp
2268 // Also do full timestamp for future dates
2269 /**
2270 * @FIXME Add better handling of future timestamps.
2271 */
2272 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2273 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2274 } elseif ( $days > 5 ) {
2275 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2276 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2277 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2278 } elseif ( $days > 1 ) {
2279 // Timestamp within the past week: show the day of the week and time
2280 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2281 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2282 $ts = wfMessage( "$weekday-at" )
2283 ->inLanguage( $this )
2284 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2285 ->text();
2286 } elseif ( $days == 1 ) {
2287 // Timestamp was yesterday: say 'yesterday' and the time.
2288 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2289 $ts = wfMessage( 'yesterday-at' )
2290 ->inLanguage( $this )
2291 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2292 ->text();
2293 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2294 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2295 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2296 $ts = wfMessage( 'today-at' )
2297 ->inLanguage( $this )
2298 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2299 ->text();
2300
2301 // From here on in, the timestamp was soon enough ago so that we can simply say
2302 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2303 } elseif ( $diff->h == 1 ) {
2304 // Less than 90 minutes, but more than an hour ago.
2305 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2306 } elseif ( $diff->i >= 1 ) {
2307 // A few minutes ago.
2308 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2309 } elseif ( $diff->s >= 30 ) {
2310 // Less than a minute, but more than 30 sec ago.
2311 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2312 } else {
2313 // Less than 30 seconds ago.
2314 $ts = wfMessage( 'just-now' )->text();
2315 }
2316
2317 return $ts;
2318 }
2319
2320 /**
2321 * @param $key string
2322 * @return array|null
2323 */
2324 function getMessage( $key ) {
2325 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2326 }
2327
2328 /**
2329 * @return array
2330 */
2331 function getAllMessages() {
2332 return self::$dataCache->getItem( $this->mCode, 'messages' );
2333 }
2334
2335 /**
2336 * @param $in
2337 * @param $out
2338 * @param $string
2339 * @return string
2340 */
2341 function iconv( $in, $out, $string ) {
2342 # This is a wrapper for iconv in all languages except esperanto,
2343 # which does some nasty x-conversions beforehand
2344
2345 # Even with //IGNORE iconv can whine about illegal characters in
2346 # *input* string. We just ignore those too.
2347 # REF: http://bugs.php.net/bug.php?id=37166
2348 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2349 wfSuppressWarnings();
2350 $text = iconv( $in, $out . '//IGNORE', $string );
2351 wfRestoreWarnings();
2352 return $text;
2353 }
2354
2355 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2356
2357 /**
2358 * @param $matches array
2359 * @return mixed|string
2360 */
2361 function ucwordbreaksCallbackAscii( $matches ) {
2362 return $this->ucfirst( $matches[1] );
2363 }
2364
2365 /**
2366 * @param $matches array
2367 * @return string
2368 */
2369 function ucwordbreaksCallbackMB( $matches ) {
2370 return mb_strtoupper( $matches[0] );
2371 }
2372
2373 /**
2374 * @param $matches array
2375 * @return string
2376 */
2377 function ucCallback( $matches ) {
2378 list( $wikiUpperChars ) = self::getCaseMaps();
2379 return strtr( $matches[1], $wikiUpperChars );
2380 }
2381
2382 /**
2383 * @param $matches array
2384 * @return string
2385 */
2386 function lcCallback( $matches ) {
2387 list( , $wikiLowerChars ) = self::getCaseMaps();
2388 return strtr( $matches[1], $wikiLowerChars );
2389 }
2390
2391 /**
2392 * @param $matches array
2393 * @return string
2394 */
2395 function ucwordsCallbackMB( $matches ) {
2396 return mb_strtoupper( $matches[0] );
2397 }
2398
2399 /**
2400 * @param $matches array
2401 * @return string
2402 */
2403 function ucwordsCallbackWiki( $matches ) {
2404 list( $wikiUpperChars ) = self::getCaseMaps();
2405 return strtr( $matches[0], $wikiUpperChars );
2406 }
2407
2408 /**
2409 * Make a string's first character uppercase
2410 *
2411 * @param $str string
2412 *
2413 * @return string
2414 */
2415 function ucfirst( $str ) {
2416 $o = ord( $str );
2417 if ( $o < 96 ) { // if already uppercase...
2418 return $str;
2419 } elseif ( $o < 128 ) {
2420 return ucfirst( $str ); // use PHP's ucfirst()
2421 } else {
2422 // fall back to more complex logic in case of multibyte strings
2423 return $this->uc( $str, true );
2424 }
2425 }
2426
2427 /**
2428 * Convert a string to uppercase
2429 *
2430 * @param $str string
2431 * @param $first bool
2432 *
2433 * @return string
2434 */
2435 function uc( $str, $first = false ) {
2436 if ( function_exists( 'mb_strtoupper' ) ) {
2437 if ( $first ) {
2438 if ( $this->isMultibyte( $str ) ) {
2439 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2440 } else {
2441 return ucfirst( $str );
2442 }
2443 } else {
2444 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2445 }
2446 } else {
2447 if ( $this->isMultibyte( $str ) ) {
2448 $x = $first ? '^' : '';
2449 return preg_replace_callback(
2450 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2451 array( $this, 'ucCallback' ),
2452 $str
2453 );
2454 } else {
2455 return $first ? ucfirst( $str ) : strtoupper( $str );
2456 }
2457 }
2458 }
2459
2460 /**
2461 * @param $str string
2462 * @return mixed|string
2463 */
2464 function lcfirst( $str ) {
2465 $o = ord( $str );
2466 if ( !$o ) {
2467 return strval( $str );
2468 } elseif ( $o >= 128 ) {
2469 return $this->lc( $str, true );
2470 } elseif ( $o > 96 ) {
2471 return $str;
2472 } else {
2473 $str[0] = strtolower( $str[0] );
2474 return $str;
2475 }
2476 }
2477
2478 /**
2479 * @param $str string
2480 * @param $first bool
2481 * @return mixed|string
2482 */
2483 function lc( $str, $first = false ) {
2484 if ( function_exists( 'mb_strtolower' ) ) {
2485 if ( $first ) {
2486 if ( $this->isMultibyte( $str ) ) {
2487 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2488 } else {
2489 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2490 }
2491 } else {
2492 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2493 }
2494 } else {
2495 if ( $this->isMultibyte( $str ) ) {
2496 $x = $first ? '^' : '';
2497 return preg_replace_callback(
2498 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2499 array( $this, 'lcCallback' ),
2500 $str
2501 );
2502 } else {
2503 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2504 }
2505 }
2506 }
2507
2508 /**
2509 * @param $str string
2510 * @return bool
2511 */
2512 function isMultibyte( $str ) {
2513 return (bool)preg_match( '/[\x80-\xff]/', $str );
2514 }
2515
2516 /**
2517 * @param $str string
2518 * @return mixed|string
2519 */
2520 function ucwords( $str ) {
2521 if ( $this->isMultibyte( $str ) ) {
2522 $str = $this->lc( $str );
2523
2524 // regexp to find first letter in each word (i.e. after each space)
2525 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2526
2527 // function to use to capitalize a single char
2528 if ( function_exists( 'mb_strtoupper' ) ) {
2529 return preg_replace_callback(
2530 $replaceRegexp,
2531 array( $this, 'ucwordsCallbackMB' ),
2532 $str
2533 );
2534 } else {
2535 return preg_replace_callback(
2536 $replaceRegexp,
2537 array( $this, 'ucwordsCallbackWiki' ),
2538 $str
2539 );
2540 }
2541 } else {
2542 return ucwords( strtolower( $str ) );
2543 }
2544 }
2545
2546 /**
2547 * capitalize words at word breaks
2548 *
2549 * @param $str string
2550 * @return mixed
2551 */
2552 function ucwordbreaks( $str ) {
2553 if ( $this->isMultibyte( $str ) ) {
2554 $str = $this->lc( $str );
2555
2556 // since \b doesn't work for UTF-8, we explicitely define word break chars
2557 $breaks = "[ \-\(\)\}\{\.,\?!]";
2558
2559 // find first letter after word break
2560 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2561
2562 if ( function_exists( 'mb_strtoupper' ) ) {
2563 return preg_replace_callback(
2564 $replaceRegexp,
2565 array( $this, 'ucwordbreaksCallbackMB' ),
2566 $str
2567 );
2568 } else {
2569 return preg_replace_callback(
2570 $replaceRegexp,
2571 array( $this, 'ucwordsCallbackWiki' ),
2572 $str
2573 );
2574 }
2575 } else {
2576 return preg_replace_callback(
2577 '/\b([\w\x80-\xff]+)\b/',
2578 array( $this, 'ucwordbreaksCallbackAscii' ),
2579 $str
2580 );
2581 }
2582 }
2583
2584 /**
2585 * Return a case-folded representation of $s
2586 *
2587 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2588 * and $s2 are the same except for the case of their characters. It is not
2589 * necessary for the value returned to make sense when displayed.
2590 *
2591 * Do *not* perform any other normalisation in this function. If a caller
2592 * uses this function when it should be using a more general normalisation
2593 * function, then fix the caller.
2594 *
2595 * @param $s string
2596 *
2597 * @return string
2598 */
2599 function caseFold( $s ) {
2600 return $this->uc( $s );
2601 }
2602
2603 /**
2604 * @param $s string
2605 * @return string
2606 */
2607 function checkTitleEncoding( $s ) {
2608 if ( is_array( $s ) ) {
2609 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2610 }
2611 if ( StringUtils::isUtf8( $s ) ) {
2612 return $s;
2613 }
2614
2615 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2616 }
2617
2618 /**
2619 * @return array
2620 */
2621 function fallback8bitEncoding() {
2622 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2623 }
2624
2625 /**
2626 * Most writing systems use whitespace to break up words.
2627 * Some languages such as Chinese don't conventionally do this,
2628 * which requires special handling when breaking up words for
2629 * searching etc.
2630 *
2631 * @return bool
2632 */
2633 function hasWordBreaks() {
2634 return true;
2635 }
2636
2637 /**
2638 * Some languages such as Chinese require word segmentation,
2639 * Specify such segmentation when overridden in derived class.
2640 *
2641 * @param $string String
2642 * @return String
2643 */
2644 function segmentByWord( $string ) {
2645 return $string;
2646 }
2647
2648 /**
2649 * Some languages have special punctuation need to be normalized.
2650 * Make such changes here.
2651 *
2652 * @param $string String
2653 * @return String
2654 */
2655 function normalizeForSearch( $string ) {
2656 return self::convertDoubleWidth( $string );
2657 }
2658
2659 /**
2660 * convert double-width roman characters to single-width.
2661 * range: ff00-ff5f ~= 0020-007f
2662 *
2663 * @param $string string
2664 *
2665 * @return string
2666 */
2667 protected static function convertDoubleWidth( $string ) {
2668 static $full = null;
2669 static $half = null;
2670
2671 if ( $full === null ) {
2672 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2673 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2674 $full = str_split( $fullWidth, 3 );
2675 $half = str_split( $halfWidth );
2676 }
2677
2678 $string = str_replace( $full, $half, $string );
2679 return $string;
2680 }
2681
2682 /**
2683 * @param $string string
2684 * @param $pattern string
2685 * @return string
2686 */
2687 protected static function insertSpace( $string, $pattern ) {
2688 $string = preg_replace( $pattern, " $1 ", $string );
2689 $string = preg_replace( '/ +/', ' ', $string );
2690 return $string;
2691 }
2692
2693 /**
2694 * @param $termsArray array
2695 * @return array
2696 */
2697 function convertForSearchResult( $termsArray ) {
2698 # some languages, e.g. Chinese, need to do a conversion
2699 # in order for search results to be displayed correctly
2700 return $termsArray;
2701 }
2702
2703 /**
2704 * Get the first character of a string.
2705 *
2706 * @param $s string
2707 * @return string
2708 */
2709 function firstChar( $s ) {
2710 $matches = array();
2711 preg_match(
2712 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2713 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2714 $s,
2715 $matches
2716 );
2717
2718 if ( isset( $matches[1] ) ) {
2719 if ( strlen( $matches[1] ) != 3 ) {
2720 return $matches[1];
2721 }
2722
2723 // Break down Hangul syllables to grab the first jamo
2724 $code = utf8ToCodepoint( $matches[1] );
2725 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2726 return $matches[1];
2727 } elseif ( $code < 0xb098 ) {
2728 return "\xe3\x84\xb1";
2729 } elseif ( $code < 0xb2e4 ) {
2730 return "\xe3\x84\xb4";
2731 } elseif ( $code < 0xb77c ) {
2732 return "\xe3\x84\xb7";
2733 } elseif ( $code < 0xb9c8 ) {
2734 return "\xe3\x84\xb9";
2735 } elseif ( $code < 0xbc14 ) {
2736 return "\xe3\x85\x81";
2737 } elseif ( $code < 0xc0ac ) {
2738 return "\xe3\x85\x82";
2739 } elseif ( $code < 0xc544 ) {
2740 return "\xe3\x85\x85";
2741 } elseif ( $code < 0xc790 ) {
2742 return "\xe3\x85\x87";
2743 } elseif ( $code < 0xcc28 ) {
2744 return "\xe3\x85\x88";
2745 } elseif ( $code < 0xce74 ) {
2746 return "\xe3\x85\x8a";
2747 } elseif ( $code < 0xd0c0 ) {
2748 return "\xe3\x85\x8b";
2749 } elseif ( $code < 0xd30c ) {
2750 return "\xe3\x85\x8c";
2751 } elseif ( $code < 0xd558 ) {
2752 return "\xe3\x85\x8d";
2753 } else {
2754 return "\xe3\x85\x8e";
2755 }
2756 } else {
2757 return '';
2758 }
2759 }
2760
2761 function initEncoding() {
2762 # Some languages may have an alternate char encoding option
2763 # (Esperanto X-coding, Japanese furigana conversion, etc)
2764 # If this language is used as the primary content language,
2765 # an override to the defaults can be set here on startup.
2766 }
2767
2768 /**
2769 * @param $s string
2770 * @return string
2771 */
2772 function recodeForEdit( $s ) {
2773 # For some languages we'll want to explicitly specify
2774 # which characters make it into the edit box raw
2775 # or are converted in some way or another.
2776 global $wgEditEncoding;
2777 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
2778 return $s;
2779 } else {
2780 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2781 }
2782 }
2783
2784 /**
2785 * @param $s string
2786 * @return string
2787 */
2788 function recodeInput( $s ) {
2789 # Take the previous into account.
2790 global $wgEditEncoding;
2791 if ( $wgEditEncoding != '' ) {
2792 $enc = $wgEditEncoding;
2793 } else {
2794 $enc = 'UTF-8';
2795 }
2796 if ( $enc == 'UTF-8' ) {
2797 return $s;
2798 } else {
2799 return $this->iconv( $enc, 'UTF-8', $s );
2800 }
2801 }
2802
2803 /**
2804 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2805 * also cleans up certain backwards-compatible sequences, converting them
2806 * to the modern Unicode equivalent.
2807 *
2808 * This is language-specific for performance reasons only.
2809 *
2810 * @param $s string
2811 *
2812 * @return string
2813 */
2814 function normalize( $s ) {
2815 global $wgAllUnicodeFixes;
2816 $s = UtfNormal::cleanUp( $s );
2817 if ( $wgAllUnicodeFixes ) {
2818 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2819 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2820 }
2821
2822 return $s;
2823 }
2824
2825 /**
2826 * Transform a string using serialized data stored in the given file (which
2827 * must be in the serialized subdirectory of $IP). The file contains pairs
2828 * mapping source characters to destination characters.
2829 *
2830 * The data is cached in process memory. This will go faster if you have the
2831 * FastStringSearch extension.
2832 *
2833 * @param $file string
2834 * @param $string string
2835 *
2836 * @throws MWException
2837 * @return string
2838 */
2839 function transformUsingPairFile( $file, $string ) {
2840 if ( !isset( $this->transformData[$file] ) ) {
2841 $data = wfGetPrecompiledData( $file );
2842 if ( $data === false ) {
2843 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2844 }
2845 $this->transformData[$file] = new ReplacementArray( $data );
2846 }
2847 return $this->transformData[$file]->replace( $string );
2848 }
2849
2850 /**
2851 * For right-to-left language support
2852 *
2853 * @return bool
2854 */
2855 function isRTL() {
2856 return self::$dataCache->getItem( $this->mCode, 'rtl' );
2857 }
2858
2859 /**
2860 * Return the correct HTML 'dir' attribute value for this language.
2861 * @return String
2862 */
2863 function getDir() {
2864 return $this->isRTL() ? 'rtl' : 'ltr';
2865 }
2866
2867 /**
2868 * Return 'left' or 'right' as appropriate alignment for line-start
2869 * for this language's text direction.
2870 *
2871 * Should be equivalent to CSS3 'start' text-align value....
2872 *
2873 * @return String
2874 */
2875 function alignStart() {
2876 return $this->isRTL() ? 'right' : 'left';
2877 }
2878
2879 /**
2880 * Return 'right' or 'left' as appropriate alignment for line-end
2881 * for this language's text direction.
2882 *
2883 * Should be equivalent to CSS3 'end' text-align value....
2884 *
2885 * @return String
2886 */
2887 function alignEnd() {
2888 return $this->isRTL() ? 'left' : 'right';
2889 }
2890
2891 /**
2892 * A hidden direction mark (LRM or RLM), depending on the language direction.
2893 * Unlike getDirMark(), this function returns the character as an HTML entity.
2894 * This function should be used when the output is guaranteed to be HTML,
2895 * because it makes the output HTML source code more readable. When
2896 * the output is plain text or can be escaped, getDirMark() should be used.
2897 *
2898 * @param $opposite Boolean Get the direction mark opposite to your language
2899 * @return string
2900 * @since 1.20
2901 */
2902 function getDirMarkEntity( $opposite = false ) {
2903 if ( $opposite ) {
2904 return $this->isRTL() ? '&lrm;' : '&rlm;';
2905 }
2906 return $this->isRTL() ? '&rlm;' : '&lrm;';
2907 }
2908
2909 /**
2910 * A hidden direction mark (LRM or RLM), depending on the language direction.
2911 * This function produces them as invisible Unicode characters and
2912 * the output may be hard to read and debug, so it should only be used
2913 * when the output is plain text or can be escaped. When the output is
2914 * HTML, use getDirMarkEntity() instead.
2915 *
2916 * @param $opposite Boolean Get the direction mark opposite to your language
2917 * @return string
2918 */
2919 function getDirMark( $opposite = false ) {
2920 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2921 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2922 if ( $opposite ) {
2923 return $this->isRTL() ? $lrm : $rlm;
2924 }
2925 return $this->isRTL() ? $rlm : $lrm;
2926 }
2927
2928 /**
2929 * @return array
2930 */
2931 function capitalizeAllNouns() {
2932 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2933 }
2934
2935 /**
2936 * An arrow, depending on the language direction.
2937 *
2938 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2939 * @return string
2940 */
2941 function getArrow( $direction = 'forwards' ) {
2942 switch ( $direction ) {
2943 case 'forwards':
2944 return $this->isRTL() ? '←' : '→';
2945 case 'backwards':
2946 return $this->isRTL() ? '→' : '←';
2947 case 'left':
2948 return '←';
2949 case 'right':
2950 return '→';
2951 case 'up':
2952 return '↑';
2953 case 'down':
2954 return '↓';
2955 }
2956 }
2957
2958 /**
2959 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2960 *
2961 * @return bool
2962 */
2963 function linkPrefixExtension() {
2964 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2965 }
2966
2967 /**
2968 * Get all magic words from cache.
2969 * @return array
2970 */
2971 function getMagicWords() {
2972 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2973 }
2974
2975 /**
2976 * Run the LanguageGetMagic hook once.
2977 */
2978 protected function doMagicHook() {
2979 if ( $this->mMagicHookDone ) {
2980 return;
2981 }
2982 $this->mMagicHookDone = true;
2983 wfProfileIn( 'LanguageGetMagic' );
2984 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2985 wfProfileOut( 'LanguageGetMagic' );
2986 }
2987
2988 /**
2989 * Fill a MagicWord object with data from here
2990 *
2991 * @param $mw
2992 */
2993 function getMagic( $mw ) {
2994 // Saves a function call
2995 if ( ! $this->mMagicHookDone ) {
2996 $this->doMagicHook();
2997 }
2998
2999 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3000 $rawEntry = $this->mMagicExtensions[$mw->mId];
3001 } else {
3002 $rawEntry = self::$dataCache->getSubitem(
3003 $this->mCode, 'magicWords', $mw->mId );
3004 }
3005
3006 if ( !is_array( $rawEntry ) ) {
3007 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3008 } else {
3009 $mw->mCaseSensitive = $rawEntry[0];
3010 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3011 }
3012 }
3013
3014 /**
3015 * Add magic words to the extension array
3016 *
3017 * @param $newWords array
3018 */
3019 function addMagicWordsByLang( $newWords ) {
3020 $fallbackChain = $this->getFallbackLanguages();
3021 $fallbackChain = array_reverse( $fallbackChain );
3022 foreach ( $fallbackChain as $code ) {
3023 if ( isset( $newWords[$code] ) ) {
3024 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3025 }
3026 }
3027 }
3028
3029 /**
3030 * Get special page names, as an associative array
3031 * case folded alias => real name
3032 */
3033 function getSpecialPageAliases() {
3034 // Cache aliases because it may be slow to load them
3035 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3036 // Initialise array
3037 $this->mExtendedSpecialPageAliases =
3038 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3039 wfRunHooks( 'LanguageGetSpecialPageAliases',
3040 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
3041 }
3042
3043 return $this->mExtendedSpecialPageAliases;
3044 }
3045
3046 /**
3047 * Italic is unsuitable for some languages
3048 *
3049 * @param $text String: the text to be emphasized.
3050 * @return string
3051 */
3052 function emphasize( $text ) {
3053 return "<em>$text</em>";
3054 }
3055
3056 /**
3057 * Normally we output all numbers in plain en_US style, that is
3058 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3059 * point twohundredthirtyfive. However this is not suitable for all
3060 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
3061 * Icelandic just want to use commas instead of dots, and dots instead
3062 * of commas like "293.291,235".
3063 *
3064 * An example of this function being called:
3065 * <code>
3066 * wfMessage( 'message' )->numParams( $num )->text()
3067 * </code>
3068 *
3069 * See LanguageGu.php for the Gujarati implementation and
3070 * $separatorTransformTable on MessageIs.php for
3071 * the , => . and . => , implementation.
3072 *
3073 * @todo check if it's viable to use localeconv() for the decimal
3074 * separator thing.
3075 * @param $number Mixed: the string to be formatted, should be an integer
3076 * or a floating point number.
3077 * @param $nocommafy Bool: set to true for special numbers like dates
3078 * @return string
3079 */
3080 public function formatNum( $number, $nocommafy = false ) {
3081 global $wgTranslateNumerals;
3082 if ( !$nocommafy ) {
3083 $number = $this->commafy( $number );
3084 $s = $this->separatorTransformTable();
3085 if ( $s ) {
3086 $number = strtr( $number, $s );
3087 }
3088 }
3089
3090 if ( $wgTranslateNumerals ) {
3091 $s = $this->digitTransformTable();
3092 if ( $s ) {
3093 $number = strtr( $number, $s );
3094 }
3095 }
3096
3097 return $number;
3098 }
3099
3100 /**
3101 * Front-end for non-commafied formatNum
3102 *
3103 * @param mixed $number the string to be formatted, should be an integer
3104 * or a floating point number.
3105 * @since 1.21
3106 * @return string
3107 */
3108 public function formatNumNoSeparators( $number ) {
3109 return $this->formatNum( $number, true );
3110 }
3111
3112 /**
3113 * @param $number string
3114 * @return string
3115 */
3116 function parseFormattedNumber( $number ) {
3117 $s = $this->digitTransformTable();
3118 if ( $s ) {
3119 $number = strtr( $number, array_flip( $s ) );
3120 }
3121
3122 $s = $this->separatorTransformTable();
3123 if ( $s ) {
3124 $number = strtr( $number, array_flip( $s ) );
3125 }
3126
3127 $number = strtr( $number, array( ',' => '' ) );
3128 return $number;
3129 }
3130
3131 /**
3132 * Adds commas to a given number
3133 * @since 1.19
3134 * @param $number mixed
3135 * @return string
3136 */
3137 function commafy( $number ) {
3138 $digitGroupingPattern = $this->digitGroupingPattern();
3139 if ( $number === null ) {
3140 return '';
3141 }
3142
3143 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3144 // default grouping is at thousands, use the same for ###,###,### pattern too.
3145 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3146 } else {
3147 // Ref: http://cldr.unicode.org/translation/number-patterns
3148 $sign = "";
3149 if ( intval( $number ) < 0 ) {
3150 // For negative numbers apply the algorithm like positive number and add sign.
3151 $sign = "-";
3152 $number = substr( $number, 1 );
3153 }
3154 $integerPart = array();
3155 $decimalPart = array();
3156 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3157 preg_match( "/\d+/", $number, $integerPart );
3158 preg_match( "/\.\d*/", $number, $decimalPart );
3159 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3160 if ( $groupedNumber === $number ) {
3161 // the string does not have any number part. Eg: .12345
3162 return $sign . $groupedNumber;
3163 }
3164 $start = $end = strlen( $integerPart[0] );
3165 while ( $start > 0 ) {
3166 $match = $matches[0][$numMatches - 1];
3167 $matchLen = strlen( $match );
3168 $start = $end - $matchLen;
3169 if ( $start < 0 ) {
3170 $start = 0;
3171 }
3172 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3173 $end = $start;
3174 if ( $numMatches > 1 ) {
3175 // use the last pattern for the rest of the number
3176 $numMatches--;
3177 }
3178 if ( $start > 0 ) {
3179 $groupedNumber = "," . $groupedNumber;
3180 }
3181 }
3182 return $sign . $groupedNumber;
3183 }
3184 }
3185
3186 /**
3187 * @return String
3188 */
3189 function digitGroupingPattern() {
3190 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3191 }
3192
3193 /**
3194 * @return array
3195 */
3196 function digitTransformTable() {
3197 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3198 }
3199
3200 /**
3201 * @return array
3202 */
3203 function separatorTransformTable() {
3204 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3205 }
3206
3207 /**
3208 * Take a list of strings and build a locale-friendly comma-separated
3209 * list, using the local comma-separator message.
3210 * The last two strings are chained with an "and".
3211 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3212 *
3213 * @param $l Array
3214 * @return string
3215 */
3216 function listToText( array $l ) {
3217 $m = count( $l ) - 1;
3218 if ( $m < 0 ) {
3219 return '';
3220 }
3221 if ( $m > 0 ) {
3222 $and = $this->getMessageFromDB( 'and' );
3223 $space = $this->getMessageFromDB( 'word-separator' );
3224 if ( $m > 1 ) {
3225 $comma = $this->getMessageFromDB( 'comma-separator' );
3226 }
3227 }
3228 $s = $l[$m];
3229 for ( $i = $m - 1; $i >= 0; $i-- ) {
3230 if ( $i == $m - 1 ) {
3231 $s = $l[$i] . $and . $space . $s;
3232 } else {
3233 $s = $l[$i] . $comma . $s;
3234 }
3235 }
3236 return $s;
3237 }
3238
3239 /**
3240 * Take a list of strings and build a locale-friendly comma-separated
3241 * list, using the local comma-separator message.
3242 * @param $list array of strings to put in a comma list
3243 * @return string
3244 */
3245 function commaList( array $list ) {
3246 return implode(
3247 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3248 $list
3249 );
3250 }
3251
3252 /**
3253 * Take a list of strings and build a locale-friendly semicolon-separated
3254 * list, using the local semicolon-separator message.
3255 * @param $list array of strings to put in a semicolon list
3256 * @return string
3257 */
3258 function semicolonList( array $list ) {
3259 return implode(
3260 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3261 $list
3262 );
3263 }
3264
3265 /**
3266 * Same as commaList, but separate it with the pipe instead.
3267 * @param $list array of strings to put in a pipe list
3268 * @return string
3269 */
3270 function pipeList( array $list ) {
3271 return implode(
3272 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3273 $list
3274 );
3275 }
3276
3277 /**
3278 * Truncate a string to a specified length in bytes, appending an optional
3279 * string (e.g. for ellipses)
3280 *
3281 * The database offers limited byte lengths for some columns in the database;
3282 * multi-byte character sets mean we need to ensure that only whole characters
3283 * are included, otherwise broken characters can be passed to the user
3284 *
3285 * If $length is negative, the string will be truncated from the beginning
3286 *
3287 * @param $string String to truncate
3288 * @param $length Int: maximum length (including ellipses)
3289 * @param $ellipsis String to append to the truncated text
3290 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3291 * $adjustLength was introduced in 1.18, before that behaved as if false.
3292 * @return string
3293 */
3294 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3295 # Use the localized ellipsis character
3296 if ( $ellipsis == '...' ) {
3297 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3298 }
3299 # Check if there is no need to truncate
3300 if ( $length == 0 ) {
3301 return $ellipsis; // convention
3302 } elseif ( strlen( $string ) <= abs( $length ) ) {
3303 return $string; // no need to truncate
3304 }
3305 $stringOriginal = $string;
3306 # If ellipsis length is >= $length then we can't apply $adjustLength
3307 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3308 $string = $ellipsis; // this can be slightly unexpected
3309 # Otherwise, truncate and add ellipsis...
3310 } else {
3311 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3312 if ( $length > 0 ) {
3313 $length -= $eLength;
3314 $string = substr( $string, 0, $length ); // xyz...
3315 $string = $this->removeBadCharLast( $string );
3316 $string = $string . $ellipsis;
3317 } else {
3318 $length += $eLength;
3319 $string = substr( $string, $length ); // ...xyz
3320 $string = $this->removeBadCharFirst( $string );
3321 $string = $ellipsis . $string;
3322 }
3323 }
3324 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3325 # This check is *not* redundant if $adjustLength, due to the single case where
3326 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3327 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3328 return $string;
3329 } else {
3330 return $stringOriginal;
3331 }
3332 }
3333
3334 /**
3335 * Remove bytes that represent an incomplete Unicode character
3336 * at the end of string (e.g. bytes of the char are missing)
3337 *
3338 * @param $string String
3339 * @return string
3340 */
3341 protected function removeBadCharLast( $string ) {
3342 if ( $string != '' ) {
3343 $char = ord( $string[strlen( $string ) - 1] );
3344 $m = array();
3345 if ( $char >= 0xc0 ) {
3346 # We got the first byte only of a multibyte char; remove it.
3347 $string = substr( $string, 0, -1 );
3348 } elseif ( $char >= 0x80 &&
3349 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3350 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3351 ) {
3352 # We chopped in the middle of a character; remove it
3353 $string = $m[1];
3354 }
3355 }
3356 return $string;
3357 }
3358
3359 /**
3360 * Remove bytes that represent an incomplete Unicode character
3361 * at the start of string (e.g. bytes of the char are missing)
3362 *
3363 * @param $string String
3364 * @return string
3365 */
3366 protected function removeBadCharFirst( $string ) {
3367 if ( $string != '' ) {
3368 $char = ord( $string[0] );
3369 if ( $char >= 0x80 && $char < 0xc0 ) {
3370 # We chopped in the middle of a character; remove the whole thing
3371 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3372 }
3373 }
3374 return $string;
3375 }
3376
3377 /**
3378 * Truncate a string of valid HTML to a specified length in bytes,
3379 * appending an optional string (e.g. for ellipses), and return valid HTML
3380 *
3381 * This is only intended for styled/linked text, such as HTML with
3382 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3383 * Also, this will not detect things like "display:none" CSS.
3384 *
3385 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3386 *
3387 * @param string $text HTML string to truncate
3388 * @param int $length (zero/positive) Maximum length (including ellipses)
3389 * @param string $ellipsis String to append to the truncated text
3390 * @return string
3391 */
3392 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3393 # Use the localized ellipsis character
3394 if ( $ellipsis == '...' ) {
3395 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3396 }
3397 # Check if there is clearly no need to truncate
3398 if ( $length <= 0 ) {
3399 return $ellipsis; // no text shown, nothing to format (convention)
3400 } elseif ( strlen( $text ) <= $length ) {
3401 return $text; // string short enough even *with* HTML (short-circuit)
3402 }
3403
3404 $dispLen = 0; // innerHTML legth so far
3405 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3406 $tagType = 0; // 0-open, 1-close
3407 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3408 $entityState = 0; // 0-not entity, 1-entity
3409 $tag = $ret = ''; // accumulated tag name, accumulated result string
3410 $openTags = array(); // open tag stack
3411 $maybeState = null; // possible truncation state
3412
3413 $textLen = strlen( $text );
3414 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3415 for ( $pos = 0; true; ++$pos ) {
3416 # Consider truncation once the display length has reached the maximim.
3417 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3418 # Check that we're not in the middle of a bracket/entity...
3419 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3420 if ( !$testingEllipsis ) {
3421 $testingEllipsis = true;
3422 # Save where we are; we will truncate here unless there turn out to
3423 # be so few remaining characters that truncation is not necessary.
3424 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3425 $maybeState = array( $ret, $openTags ); // save state
3426 }
3427 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3428 # String in fact does need truncation, the truncation point was OK.
3429 list( $ret, $openTags ) = $maybeState; // reload state
3430 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3431 $ret .= $ellipsis; // add ellipsis
3432 break;
3433 }
3434 }
3435 if ( $pos >= $textLen ) {
3436 break; // extra iteration just for above checks
3437 }
3438
3439 # Read the next char...
3440 $ch = $text[$pos];
3441 $lastCh = $pos ? $text[$pos - 1] : '';
3442 $ret .= $ch; // add to result string
3443 if ( $ch == '<' ) {
3444 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3445 $entityState = 0; // for bad HTML
3446 $bracketState = 1; // tag started (checking for backslash)
3447 } elseif ( $ch == '>' ) {
3448 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3449 $entityState = 0; // for bad HTML
3450 $bracketState = 0; // out of brackets
3451 } elseif ( $bracketState == 1 ) {
3452 if ( $ch == '/' ) {
3453 $tagType = 1; // close tag (e.g. "</span>")
3454 } else {
3455 $tagType = 0; // open tag (e.g. "<span>")
3456 $tag .= $ch;
3457 }
3458 $bracketState = 2; // building tag name
3459 } elseif ( $bracketState == 2 ) {
3460 if ( $ch != ' ' ) {
3461 $tag .= $ch;
3462 } else {
3463 // Name found (e.g. "<a href=..."), add on tag attributes...
3464 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3465 }
3466 } elseif ( $bracketState == 0 ) {
3467 if ( $entityState ) {
3468 if ( $ch == ';' ) {
3469 $entityState = 0;
3470 $dispLen++; // entity is one displayed char
3471 }
3472 } else {
3473 if ( $neLength == 0 && !$maybeState ) {
3474 // Save state without $ch. We want to *hit* the first
3475 // display char (to get tags) but not *use* it if truncating.
3476 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3477 }
3478 if ( $ch == '&' ) {
3479 $entityState = 1; // entity found, (e.g. "&#160;")
3480 } else {
3481 $dispLen++; // this char is displayed
3482 // Add the next $max display text chars after this in one swoop...
3483 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3484 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3485 $dispLen += $skipped;
3486 $pos += $skipped;
3487 }
3488 }
3489 }
3490 }
3491 // Close the last tag if left unclosed by bad HTML
3492 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3493 while ( count( $openTags ) > 0 ) {
3494 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3495 }
3496 return $ret;
3497 }
3498
3499 /**
3500 * truncateHtml() helper function
3501 * like strcspn() but adds the skipped chars to $ret
3502 *
3503 * @param $ret
3504 * @param $text
3505 * @param $search
3506 * @param $start
3507 * @param $len
3508 * @return int
3509 */
3510 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3511 if ( $len === null ) {
3512 $len = -1; // -1 means "no limit" for strcspn
3513 } elseif ( $len < 0 ) {
3514 $len = 0; // sanity
3515 }
3516 $skipCount = 0;
3517 if ( $start < strlen( $text ) ) {
3518 $skipCount = strcspn( $text, $search, $start, $len );
3519 $ret .= substr( $text, $start, $skipCount );
3520 }
3521 return $skipCount;
3522 }
3523
3524 /**
3525 * truncateHtml() helper function
3526 * (a) push or pop $tag from $openTags as needed
3527 * (b) clear $tag value
3528 * @param &$tag string Current HTML tag name we are looking at
3529 * @param $tagType int (0-open tag, 1-close tag)
3530 * @param $lastCh string Character before the '>' that ended this tag
3531 * @param &$openTags array Open tag stack (not accounting for $tag)
3532 */
3533 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3534 $tag = ltrim( $tag );
3535 if ( $tag != '' ) {
3536 if ( $tagType == 0 && $lastCh != '/' ) {
3537 $openTags[] = $tag; // tag opened (didn't close itself)
3538 } elseif ( $tagType == 1 ) {
3539 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3540 array_pop( $openTags ); // tag closed
3541 }
3542 }
3543 $tag = '';
3544 }
3545 }
3546
3547 /**
3548 * Grammatical transformations, needed for inflected languages
3549 * Invoked by putting {{grammar:case|word}} in a message
3550 *
3551 * @param $word string
3552 * @param $case string
3553 * @return string
3554 */
3555 function convertGrammar( $word, $case ) {
3556 global $wgGrammarForms;
3557 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3558 return $wgGrammarForms[$this->getCode()][$case][$word];
3559 }
3560 return $word;
3561 }
3562 /**
3563 * Get the grammar forms for the content language
3564 * @return array of grammar forms
3565 * @since 1.20
3566 */
3567 function getGrammarForms() {
3568 global $wgGrammarForms;
3569 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3570 return $wgGrammarForms[$this->getCode()];
3571 }
3572 return array();
3573 }
3574 /**
3575 * Provides an alternative text depending on specified gender.
3576 * Usage {{gender:username|masculine|feminine|neutral}}.
3577 * username is optional, in which case the gender of current user is used,
3578 * but only in (some) interface messages; otherwise default gender is used.
3579 *
3580 * If no forms are given, an empty string is returned. If only one form is
3581 * given, it will be returned unconditionally. These details are implied by
3582 * the caller and cannot be overridden in subclasses.
3583 *
3584 * If more than one form is given, the default is to use the neutral one
3585 * if it is specified, and to use the masculine one otherwise. These
3586 * details can be overridden in subclasses.
3587 *
3588 * @param $gender string
3589 * @param $forms array
3590 *
3591 * @return string
3592 */
3593 function gender( $gender, $forms ) {
3594 if ( !count( $forms ) ) {
3595 return '';
3596 }
3597 $forms = $this->preConvertPlural( $forms, 2 );
3598 if ( $gender === 'male' ) {
3599 return $forms[0];
3600 }
3601 if ( $gender === 'female' ) {
3602 return $forms[1];
3603 }
3604 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3605 }
3606
3607 /**
3608 * Plural form transformations, needed for some languages.
3609 * For example, there are 3 form of plural in Russian and Polish,
3610 * depending on "count mod 10". See [[w:Plural]]
3611 * For English it is pretty simple.
3612 *
3613 * Invoked by putting {{plural:count|wordform1|wordform2}}
3614 * or {{plural:count|wordform1|wordform2|wordform3}}
3615 *
3616 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3617 *
3618 * @param $count Integer: non-localized number
3619 * @param $forms Array: different plural forms
3620 * @return string Correct form of plural for $count in this language
3621 */
3622 function convertPlural( $count, $forms ) {
3623 // Handle explicit n=pluralform cases
3624 foreach ( $forms as $index => $form ) {
3625 if ( preg_match( '/^\d+=/i', $form ) ) {
3626 $pos = strpos( $form, '=' );
3627 if ( substr( $form, 0, $pos ) === (string) $count ) {
3628 return substr( $form, $pos + 1 );
3629 }
3630 unset( $forms[$index] );
3631 }
3632 }
3633
3634 $forms = array_values( $forms );
3635 if ( !count( $forms ) ) {
3636 return '';
3637 }
3638
3639 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3640 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3641 return $forms[$pluralForm];
3642 }
3643
3644 /**
3645 * Checks that convertPlural was given an array and pads it to requested
3646 * amount of forms by copying the last one.
3647 *
3648 * @param $count Integer: How many forms should there be at least
3649 * @param $forms Array of forms given to convertPlural
3650 * @return array Padded array of forms or an exception if not an array
3651 */
3652 protected function preConvertPlural( /* Array */ $forms, $count ) {
3653 while ( count( $forms ) < $count ) {
3654 $forms[] = $forms[count( $forms ) - 1];
3655 }
3656 return $forms;
3657 }
3658
3659 /**
3660 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3661 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3662 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3663 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3664 * match up with it.
3665 *
3666 * @param $str String: the validated block duration in English
3667 * @return string Somehow translated block duration
3668 * @see LanguageFi.php for example implementation
3669 */
3670 function translateBlockExpiry( $str ) {
3671 $duration = SpecialBlock::getSuggestedDurations( $this );
3672 foreach ( $duration as $show => $value ) {
3673 if ( strcmp( $str, $value ) == 0 ) {
3674 return htmlspecialchars( trim( $show ) );
3675 }
3676 }
3677
3678 // Since usually only infinite or indefinite is only on list, so try
3679 // equivalents if still here.
3680 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3681 if ( in_array( $str, $indefs ) ) {
3682 foreach ( $indefs as $val ) {
3683 $show = array_search( $val, $duration, true );
3684 if ( $show !== false ) {
3685 return htmlspecialchars( trim( $show ) );
3686 }
3687 }
3688 }
3689
3690 // If all else fails, return a standard duration or timestamp description.
3691 $time = strtotime( $str, 0 );
3692 if ( $time === false ) { // Unknown format. Return it as-is in case.
3693 return $str;
3694 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3695 // $time is relative to 0 so it's a duration length.
3696 return $this->formatDuration( $time );
3697 } else { // It's an absolute timestamp.
3698 if ( $time === 0 ) {
3699 // wfTimestamp() handles 0 as current time instead of epoch.
3700 return $this->timeanddate( '19700101000000' );
3701 } else {
3702 return $this->timeanddate( $time );
3703 }
3704 }
3705 }
3706
3707 /**
3708 * languages like Chinese need to be segmented in order for the diff
3709 * to be of any use
3710 *
3711 * @param $text String
3712 * @return String
3713 */
3714 public function segmentForDiff( $text ) {
3715 return $text;
3716 }
3717
3718 /**
3719 * and unsegment to show the result
3720 *
3721 * @param $text String
3722 * @return String
3723 */
3724 public function unsegmentForDiff( $text ) {
3725 return $text;
3726 }
3727
3728 /**
3729 * Return the LanguageConverter used in the Language
3730 *
3731 * @since 1.19
3732 * @return LanguageConverter
3733 */
3734 public function getConverter() {
3735 return $this->mConverter;
3736 }
3737
3738 /**
3739 * convert text to all supported variants
3740 *
3741 * @param $text string
3742 * @return array
3743 */
3744 public function autoConvertToAllVariants( $text ) {
3745 return $this->mConverter->autoConvertToAllVariants( $text );
3746 }
3747
3748 /**
3749 * convert text to different variants of a language.
3750 *
3751 * @param $text string
3752 * @return string
3753 */
3754 public function convert( $text ) {
3755 return $this->mConverter->convert( $text );
3756 }
3757
3758 /**
3759 * Convert a Title object to a string in the preferred variant
3760 *
3761 * @param $title Title
3762 * @return string
3763 */
3764 public function convertTitle( $title ) {
3765 return $this->mConverter->convertTitle( $title );
3766 }
3767
3768 /**
3769 * Convert a namespace index to a string in the preferred variant
3770 *
3771 * @param $ns int
3772 * @return string
3773 */
3774 public function convertNamespace( $ns ) {
3775 return $this->mConverter->convertNamespace( $ns );
3776 }
3777
3778 /**
3779 * Check if this is a language with variants
3780 *
3781 * @return bool
3782 */
3783 public function hasVariants() {
3784 return count( $this->getVariants() ) > 1;
3785 }
3786
3787 /**
3788 * Check if the language has the specific variant
3789 *
3790 * @since 1.19
3791 * @param $variant string
3792 * @return bool
3793 */
3794 public function hasVariant( $variant ) {
3795 return (bool)$this->mConverter->validateVariant( $variant );
3796 }
3797
3798 /**
3799 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3800 *
3801 * @param $text string
3802 * @return string
3803 */
3804 public function armourMath( $text ) {
3805 return $this->mConverter->armourMath( $text );
3806 }
3807
3808 /**
3809 * Perform output conversion on a string, and encode for safe HTML output.
3810 * @param $text String text to be converted
3811 * @param $isTitle Bool whether this conversion is for the article title
3812 * @return string
3813 * @todo this should get integrated somewhere sane
3814 */
3815 public function convertHtml( $text, $isTitle = false ) {
3816 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3817 }
3818
3819 /**
3820 * @param $key string
3821 * @return string
3822 */
3823 public function convertCategoryKey( $key ) {
3824 return $this->mConverter->convertCategoryKey( $key );
3825 }
3826
3827 /**
3828 * Get the list of variants supported by this language
3829 * see sample implementation in LanguageZh.php
3830 *
3831 * @return array an array of language codes
3832 */
3833 public function getVariants() {
3834 return $this->mConverter->getVariants();
3835 }
3836
3837 /**
3838 * @return string
3839 */
3840 public function getPreferredVariant() {
3841 return $this->mConverter->getPreferredVariant();
3842 }
3843
3844 /**
3845 * @return string
3846 */
3847 public function getDefaultVariant() {
3848 return $this->mConverter->getDefaultVariant();
3849 }
3850
3851 /**
3852 * @return string
3853 */
3854 public function getURLVariant() {
3855 return $this->mConverter->getURLVariant();
3856 }
3857
3858 /**
3859 * If a language supports multiple variants, it is
3860 * possible that non-existing link in one variant
3861 * actually exists in another variant. this function
3862 * tries to find it. See e.g. LanguageZh.php
3863 *
3864 * @param $link String: the name of the link
3865 * @param $nt Mixed: the title object of the link
3866 * @param $ignoreOtherCond Boolean: to disable other conditions when
3867 * we need to transclude a template or update a category's link
3868 * @return null the input parameters may be modified upon return
3869 */
3870 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3871 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
3872 }
3873
3874 /**
3875 * If a language supports multiple variants, converts text
3876 * into an array of all possible variants of the text:
3877 * 'variant' => text in that variant
3878 *
3879 * @deprecated since 1.17 Use autoConvertToAllVariants()
3880 *
3881 * @param $text string
3882 *
3883 * @return string
3884 */
3885 public function convertLinkToAllVariants( $text ) {
3886 return $this->mConverter->convertLinkToAllVariants( $text );
3887 }
3888
3889 /**
3890 * returns language specific options used by User::getPageRenderHash()
3891 * for example, the preferred language variant
3892 *
3893 * @return string
3894 */
3895 function getExtraHashOptions() {
3896 return $this->mConverter->getExtraHashOptions();
3897 }
3898
3899 /**
3900 * For languages that support multiple variants, the title of an
3901 * article may be displayed differently in different variants. this
3902 * function returns the apporiate title defined in the body of the article.
3903 *
3904 * @return string
3905 */
3906 public function getParsedTitle() {
3907 return $this->mConverter->getParsedTitle();
3908 }
3909
3910 /**
3911 * Prepare external link text for conversion. When the text is
3912 * a URL, it shouldn't be converted, and it'll be wrapped in
3913 * the "raw" tag (-{R| }-) to prevent conversion.
3914 *
3915 * This function is called "markNoConversion" for historical
3916 * reasons.
3917 *
3918 * @param $text String: text to be used for external link
3919 * @param $noParse bool: wrap it without confirming it's a real URL first
3920 * @return string the tagged text
3921 */
3922 public function markNoConversion( $text, $noParse = false ) {
3923 // Excluding protocal-relative URLs may avoid many false positives.
3924 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
3925 return $this->mConverter->markNoConversion( $text );
3926 } else {
3927 return $text;
3928 }
3929 }
3930
3931 /**
3932 * A regular expression to match legal word-trailing characters
3933 * which should be merged onto a link of the form [[foo]]bar.
3934 *
3935 * @return string
3936 */
3937 public function linkTrail() {
3938 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
3939 }
3940
3941 /**
3942 * @return Language
3943 */
3944 function getLangObj() {
3945 return $this;
3946 }
3947
3948 /**
3949 * Get the "parent" language which has a converter to convert a "compatible" language
3950 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
3951 *
3952 * @return Language|null
3953 * @since 1.22
3954 */
3955 public function getParentLanguage() {
3956 if ( $this->mParentLanguage !== false ) {
3957 return $this->mParentLanguage;
3958 }
3959
3960 $pieces = explode( '-', $this->getCode() );
3961 $code = $pieces[0];
3962 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
3963 $this->mParentLanguage = null;
3964 return null;
3965 }
3966 $lang = Language::factory( $code );
3967 if ( !$lang->hasVariant( $this->getCode() ) ) {
3968 $this->mParentLanguage = null;
3969 return null;
3970 }
3971
3972 $this->mParentLanguage = $lang;
3973 return $lang;
3974 }
3975
3976 /**
3977 * Get the RFC 3066 code for this language object
3978 *
3979 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3980 * htmlspecialchars() or similar
3981 *
3982 * @return string
3983 */
3984 public function getCode() {
3985 return $this->mCode;
3986 }
3987
3988 /**
3989 * Get the code in Bcp47 format which we can use
3990 * inside of html lang="" tags.
3991 *
3992 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3993 * htmlspecialchars() or similar.
3994 *
3995 * @since 1.19
3996 * @return string
3997 */
3998 public function getHtmlCode() {
3999 if ( is_null( $this->mHtmlCode ) ) {
4000 $this->mHtmlCode = wfBCP47( $this->getCode() );
4001 }
4002 return $this->mHtmlCode;
4003 }
4004
4005 /**
4006 * @param $code string
4007 */
4008 public function setCode( $code ) {
4009 $this->mCode = $code;
4010 // Ensure we don't leave incorrect cached data lying around
4011 $this->mHtmlCode = null;
4012 $this->mParentLanguage = false;
4013 }
4014
4015 /**
4016 * Get the name of a file for a certain language code
4017 * @param $prefix string Prepend this to the filename
4018 * @param $code string Language code
4019 * @param $suffix string Append this to the filename
4020 * @throws MWException
4021 * @return string $prefix . $mangledCode . $suffix
4022 */
4023 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4024 // Protect against path traversal
4025 if ( !Language::isValidCode( $code )
4026 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
4027 {
4028 throw new MWException( "Invalid language code \"$code\"" );
4029 }
4030
4031 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4032 }
4033
4034 /**
4035 * Get the language code from a file name. Inverse of getFileName()
4036 * @param $filename string $prefix . $languageCode . $suffix
4037 * @param $prefix string Prefix before the language code
4038 * @param $suffix string Suffix after the language code
4039 * @return string Language code, or false if $prefix or $suffix isn't found
4040 */
4041 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4042 $m = null;
4043 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4044 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4045 if ( !count( $m ) ) {
4046 return false;
4047 }
4048 return str_replace( '_', '-', strtolower( $m[1] ) );
4049 }
4050
4051 /**
4052 * @param $code string
4053 * @return string
4054 */
4055 public static function getMessagesFileName( $code ) {
4056 global $IP;
4057 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4058 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
4059 return $file;
4060 }
4061
4062 /**
4063 * @param $code string
4064 * @return string
4065 */
4066 public static function getClassFileName( $code ) {
4067 global $IP;
4068 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4069 }
4070
4071 /**
4072 * Get the first fallback for a given language.
4073 *
4074 * @param $code string
4075 *
4076 * @return bool|string
4077 */
4078 public static function getFallbackFor( $code ) {
4079 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4080 return false;
4081 } else {
4082 $fallbacks = self::getFallbacksFor( $code );
4083 $first = array_shift( $fallbacks );
4084 return $first;
4085 }
4086 }
4087
4088 /**
4089 * Get the ordered list of fallback languages.
4090 *
4091 * @since 1.19
4092 * @param $code string Language code
4093 * @return array
4094 */
4095 public static function getFallbacksFor( $code ) {
4096 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4097 return array();
4098 } else {
4099 $v = self::getLocalisationCache()->getItem( $code, 'fallback' );
4100 $v = array_map( 'trim', explode( ',', $v ) );
4101 if ( $v[count( $v ) - 1] !== 'en' ) {
4102 $v[] = 'en';
4103 }
4104 return $v;
4105 }
4106 }
4107
4108 /**
4109 * Get the ordered list of fallback languages, ending with the fallback
4110 * language chain for the site language.
4111 *
4112 * @since 1.22
4113 * @param string $code Language code
4114 * @return array array( fallbacks, site fallbacks )
4115 */
4116 public static function getFallbacksIncludingSiteLanguage( $code ) {
4117 global $wgLanguageCode;
4118
4119 // Usually, we will only store a tiny number of fallback chains, so we
4120 // keep them in static memory.
4121 $cacheKey = "{$code}-{$wgLanguageCode}";
4122
4123 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4124 $fallbacks = self::getFallbacksFor( $code );
4125
4126 // Append the site's fallback chain, including the site language itself
4127 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4128 array_unshift( $siteFallbacks, $wgLanguageCode );
4129
4130 // Eliminate any languages already included in the chain
4131 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4132
4133 self::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4134 }
4135 return self::$fallbackLanguageCache[$cacheKey];
4136 }
4137
4138 /**
4139 * Get all messages for a given language
4140 * WARNING: this may take a long time. If you just need all message *keys*
4141 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4142 *
4143 * @param $code string
4144 *
4145 * @return array
4146 */
4147 public static function getMessagesFor( $code ) {
4148 return self::getLocalisationCache()->getItem( $code, 'messages' );
4149 }
4150
4151 /**
4152 * Get a message for a given language
4153 *
4154 * @param $key string
4155 * @param $code string
4156 *
4157 * @return string
4158 */
4159 public static function getMessageFor( $key, $code ) {
4160 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4161 }
4162
4163 /**
4164 * Get all message keys for a given language. This is a faster alternative to
4165 * array_keys( Language::getMessagesFor( $code ) )
4166 *
4167 * @since 1.19
4168 * @param $code string Language code
4169 * @return array of message keys (strings)
4170 */
4171 public static function getMessageKeysFor( $code ) {
4172 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
4173 }
4174
4175 /**
4176 * @param $talk
4177 * @return mixed
4178 */
4179 function fixVariableInNamespace( $talk ) {
4180 if ( strpos( $talk, '$1' ) === false ) {
4181 return $talk;
4182 }
4183
4184 global $wgMetaNamespace;
4185 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4186
4187 # Allow grammar transformations
4188 # Allowing full message-style parsing would make simple requests
4189 # such as action=raw much more expensive than they need to be.
4190 # This will hopefully cover most cases.
4191 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4192 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4193 return str_replace( ' ', '_', $talk );
4194 }
4195
4196 /**
4197 * @param $m string
4198 * @return string
4199 */
4200 function replaceGrammarInNamespace( $m ) {
4201 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4202 }
4203
4204 /**
4205 * @throws MWException
4206 * @return array
4207 */
4208 static function getCaseMaps() {
4209 static $wikiUpperChars, $wikiLowerChars;
4210 if ( isset( $wikiUpperChars ) ) {
4211 return array( $wikiUpperChars, $wikiLowerChars );
4212 }
4213
4214 wfProfileIn( __METHOD__ );
4215 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4216 if ( $arr === false ) {
4217 throw new MWException(
4218 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4219 }
4220 $wikiUpperChars = $arr['wikiUpperChars'];
4221 $wikiLowerChars = $arr['wikiLowerChars'];
4222 wfProfileOut( __METHOD__ );
4223 return array( $wikiUpperChars, $wikiLowerChars );
4224 }
4225
4226 /**
4227 * Decode an expiry (block, protection, etc) which has come from the DB
4228 *
4229 * @todo FIXME: why are we returnings DBMS-dependent strings???
4230 *
4231 * @param $expiry String: Database expiry String
4232 * @param $format Bool|Int true to process using language functions, or TS_ constant
4233 * to return the expiry in a given timestamp
4234 * @return String
4235 * @since 1.18
4236 */
4237 public function formatExpiry( $expiry, $format = true ) {
4238 static $infinity;
4239 if ( $infinity === null ) {
4240 $infinity = wfGetDB( DB_SLAVE )->getInfinity();
4241 }
4242
4243 if ( $expiry == '' || $expiry == $infinity ) {
4244 return $format === true
4245 ? $this->getMessageFromDB( 'infiniteblock' )
4246 : $infinity;
4247 } else {
4248 return $format === true
4249 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4250 : wfTimestamp( $format, $expiry );
4251 }
4252 }
4253
4254 /**
4255 * @todo Document
4256 * @param $seconds int|float
4257 * @param $format Array Optional
4258 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
4259 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
4260 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
4261 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
4262 * @return string
4263 */
4264 function formatTimePeriod( $seconds, $format = array() ) {
4265 if ( !is_array( $format ) ) {
4266 $format = array( 'avoid' => $format ); // For backwards compatibility
4267 }
4268 if ( !isset( $format['avoid'] ) ) {
4269 $format['avoid'] = false;
4270 }
4271 if ( !isset( $format['noabbrevs' ] ) ) {
4272 $format['noabbrevs'] = false;
4273 }
4274 $secondsMsg = wfMessage(
4275 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4276 $minutesMsg = wfMessage(
4277 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4278 $hoursMsg = wfMessage(
4279 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4280 $daysMsg = wfMessage(
4281 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4282
4283 if ( round( $seconds * 10 ) < 100 ) {
4284 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4285 $s = $secondsMsg->params( $s )->text();
4286 } elseif ( round( $seconds ) < 60 ) {
4287 $s = $this->formatNum( round( $seconds ) );
4288 $s = $secondsMsg->params( $s )->text();
4289 } elseif ( round( $seconds ) < 3600 ) {
4290 $minutes = floor( $seconds / 60 );
4291 $secondsPart = round( fmod( $seconds, 60 ) );
4292 if ( $secondsPart == 60 ) {
4293 $secondsPart = 0;
4294 $minutes++;
4295 }
4296 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4297 $s .= ' ';
4298 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4299 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4300 $hours = floor( $seconds / 3600 );
4301 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4302 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4303 if ( $secondsPart == 60 ) {
4304 $secondsPart = 0;
4305 $minutes++;
4306 }
4307 if ( $minutes == 60 ) {
4308 $minutes = 0;
4309 $hours++;
4310 }
4311 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4312 $s .= ' ';
4313 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4314 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4315 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4316 }
4317 } else {
4318 $days = floor( $seconds / 86400 );
4319 if ( $format['avoid'] === 'avoidminutes' ) {
4320 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4321 if ( $hours == 24 ) {
4322 $hours = 0;
4323 $days++;
4324 }
4325 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4326 $s .= ' ';
4327 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4328 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4329 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4330 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4331 if ( $minutes == 60 ) {
4332 $minutes = 0;
4333 $hours++;
4334 }
4335 if ( $hours == 24 ) {
4336 $hours = 0;
4337 $days++;
4338 }
4339 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4340 $s .= ' ';
4341 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4342 $s .= ' ';
4343 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4344 } else {
4345 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4346 $s .= ' ';
4347 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4348 }
4349 }
4350 return $s;
4351 }
4352
4353 /**
4354 * Format a bitrate for output, using an appropriate
4355 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4356 *
4357 * This use base 1000. For base 1024 use formatSize(), for another base
4358 * see formatComputingNumbers()
4359 *
4360 * @param $bps int
4361 * @return string
4362 */
4363 function formatBitrate( $bps ) {
4364 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4365 }
4366
4367 /**
4368 * @param $size int Size of the unit
4369 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4370 * @param $messageKey string Message key to be uesd
4371 * @return string
4372 */
4373 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4374 if ( $size <= 0 ) {
4375 return str_replace( '$1', $this->formatNum( $size ),
4376 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4377 );
4378 }
4379 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4380 $index = 0;
4381
4382 $maxIndex = count( $sizes ) - 1;
4383 while ( $size >= $boundary && $index < $maxIndex ) {
4384 $index++;
4385 $size /= $boundary;
4386 }
4387
4388 // For small sizes no decimal places necessary
4389 $round = 0;
4390 if ( $index > 1 ) {
4391 // For MB and bigger two decimal places are smarter
4392 $round = 2;
4393 }
4394 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4395
4396 $size = round( $size, $round );
4397 $text = $this->getMessageFromDB( $msg );
4398 return str_replace( '$1', $this->formatNum( $size ), $text );
4399 }
4400
4401 /**
4402 * Format a size in bytes for output, using an appropriate
4403 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4404 *
4405 * This method use base 1024. For base 1000 use formatBitrate(), for
4406 * another base see formatComputingNumbers()
4407 *
4408 * @param $size int Size to format
4409 * @return string Plain text (not HTML)
4410 */
4411 function formatSize( $size ) {
4412 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4413 }
4414
4415 /**
4416 * Make a list item, used by various special pages
4417 *
4418 * @param $page String Page link
4419 * @param $details String Text between brackets
4420 * @param $oppositedm Boolean Add the direction mark opposite to your
4421 * language, to display text properly
4422 * @return String
4423 */
4424 function specialList( $page, $details, $oppositedm = true ) {
4425 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) .
4426 $this->getDirMark();
4427 $details = $details ? $dirmark . $this->getMessageFromDB( 'word-separator' ) .
4428 wfMessage( 'parentheses' )->rawParams( $details )->inLanguage( $this )->escaped() : '';
4429 return $page . $details;
4430 }
4431
4432 /**
4433 * Generate (prev x| next x) (20|50|100...) type links for paging
4434 *
4435 * @param $title Title object to link
4436 * @param $offset Integer offset parameter
4437 * @param $limit Integer limit parameter
4438 * @param $query array|String optional URL query parameter string
4439 * @param $atend Bool optional param for specified if this is the last page
4440 * @return String
4441 */
4442 public function viewPrevNext( Title $title, $offset, $limit, array $query = array(), $atend = false ) {
4443 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4444
4445 # Make 'previous' link
4446 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4447 if ( $offset > 0 ) {
4448 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4449 $query, $prev, 'prevn-title', 'mw-prevlink' );
4450 } else {
4451 $plink = htmlspecialchars( $prev );
4452 }
4453
4454 # Make 'next' link
4455 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4456 if ( $atend ) {
4457 $nlink = htmlspecialchars( $next );
4458 } else {
4459 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4460 $query, $next, 'prevn-title', 'mw-nextlink' );
4461 }
4462
4463 # Make links to set number of items per page
4464 $numLinks = array();
4465 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4466 $numLinks[] = $this->numLink( $title, $offset, $num,
4467 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4468 }
4469
4470 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4471 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4472 }
4473
4474 /**
4475 * Helper function for viewPrevNext() that generates links
4476 *
4477 * @param $title Title object to link
4478 * @param $offset Integer offset parameter
4479 * @param $limit Integer limit parameter
4480 * @param $query Array extra query parameters
4481 * @param $link String text to use for the link; will be escaped
4482 * @param $tooltipMsg String name of the message to use as tooltip
4483 * @param $class String value of the "class" attribute of the link
4484 * @return String HTML fragment
4485 */
4486 private function numLink( Title $title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4487 $query = array( 'limit' => $limit, 'offset' => $offset ) + $query;
4488 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4489 return Html::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4490 'title' => $tooltip, 'class' => $class ), $link );
4491 }
4492
4493 /**
4494 * Get the conversion rule title, if any.
4495 *
4496 * @return string
4497 */
4498 public function getConvRuleTitle() {
4499 return $this->mConverter->getConvRuleTitle();
4500 }
4501
4502 /**
4503 * Get the compiled plural rules for the language
4504 * @since 1.20
4505 * @return array Associative array with plural form, and plural rule as key-value pairs
4506 */
4507 public function getCompiledPluralRules() {
4508 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4509 $fallbacks = Language::getFallbacksFor( $this->mCode );
4510 if ( !$pluralRules ) {
4511 foreach ( $fallbacks as $fallbackCode ) {
4512 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4513 if ( $pluralRules ) {
4514 break;
4515 }
4516 }
4517 }
4518 return $pluralRules;
4519 }
4520
4521 /**
4522 * Get the plural rules for the language
4523 * @since 1.20
4524 * @return array Associative array with plural form number and plural rule as key-value pairs
4525 */
4526 public function getPluralRules() {
4527 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4528 $fallbacks = Language::getFallbacksFor( $this->mCode );
4529 if ( !$pluralRules ) {
4530 foreach ( $fallbacks as $fallbackCode ) {
4531 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4532 if ( $pluralRules ) {
4533 break;
4534 }
4535 }
4536 }
4537 return $pluralRules;
4538 }
4539
4540 /**
4541 * Get the plural rule types for the language
4542 * @since 1.22
4543 * @return array Associative array with plural form number and plural rule type as key-value pairs
4544 */
4545 public function getPluralRuleTypes() {
4546 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4547 $fallbacks = Language::getFallbacksFor( $this->mCode );
4548 if ( !$pluralRuleTypes ) {
4549 foreach ( $fallbacks as $fallbackCode ) {
4550 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4551 if ( $pluralRuleTypes ) {
4552 break;
4553 }
4554 }
4555 }
4556 return $pluralRuleTypes;
4557 }
4558
4559 /**
4560 * Find the index number of the plural rule appropriate for the given number
4561 * @return int The index number of the plural rule
4562 */
4563 public function getPluralRuleIndexNumber( $number ) {
4564 $pluralRules = $this->getCompiledPluralRules();
4565 $form = CLDRPluralRuleEvaluator::evaluateCompiled( $number, $pluralRules );
4566 return $form;
4567 }
4568
4569 /**
4570 * Find the plural rule type appropriate for the given number
4571 * For example, if the language is set to Arabic, getPluralType(5) should
4572 * return 'few'.
4573 * @since 1.22
4574 * @return string The name of the plural rule type, e.g. one, two, few, many
4575 */
4576 public function getPluralRuleType( $number ) {
4577 $index = $this->getPluralRuleIndexNumber( $number );
4578 $pluralRuleTypes = $this->getPluralRuleTypes();
4579 if ( isset( $pluralRuleTypes[$index] ) ) {
4580 return $pluralRuleTypes[$index];
4581 } else {
4582 return 'other';
4583 }
4584 }
4585 }
Wiklou, le Wiki du Wiklou