dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
Merge "Make FakeConverter more realistic"
[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 // Give grep a chance to find the usages:
2097 // duration-centuries, duration-decades, duration-years, duration-days,
2098 // duration-hours, duration-minutes, duration-seconds
2099 $message = wfMessage( 'duration-' . $intervalName )->numParams( $intervalValue );
2100 $segments[] = $message->inLanguage( $this )->escaped();
2101 }
2102
2103 return $this->listToText( $segments );
2104 }
2105
2106 /**
2107 * Takes a number of seconds and returns an array with a set of corresponding intervals.
2108 * For example 65 will be turned into array( minutes => 1, seconds => 5 ).
2109 *
2110 * @since 1.20
2111 *
2112 * @param integer $seconds The amount of seconds.
2113 * @param array $chosenIntervals The intervals to enable.
2114 *
2115 * @return array
2116 */
2117 public function getDurationIntervals( $seconds, array $chosenIntervals = array() ) {
2118 if ( empty( $chosenIntervals ) ) {
2119 $chosenIntervals = array( 'millennia', 'centuries', 'decades', 'years', 'days', 'hours', 'minutes', 'seconds' );
2120 }
2121
2122 $intervals = array_intersect_key( self::$durationIntervals, array_flip( $chosenIntervals ) );
2123 $sortedNames = array_keys( $intervals );
2124 $smallestInterval = array_pop( $sortedNames );
2125
2126 $segments = array();
2127
2128 foreach ( $intervals as $name => $length ) {
2129 $value = floor( $seconds / $length );
2130
2131 if ( $value > 0 || ( $name == $smallestInterval && empty( $segments ) ) ) {
2132 $seconds -= $value * $length;
2133 $segments[$name] = $value;
2134 }
2135 }
2136
2137 return $segments;
2138 }
2139
2140 /**
2141 * Internal helper function for userDate(), userTime() and userTimeAndDate()
2142 *
2143 * @param $type String: can be 'date', 'time' or 'both'
2144 * @param $ts Mixed: the time format which needs to be turned into a
2145 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2146 * @param $user User object used to get preferences for timezone and format
2147 * @param $options Array, can contain the following keys:
2148 * - 'timecorrection': time correction, can have the following values:
2149 * - true: use user's preference
2150 * - false: don't use time correction
2151 * - integer: value of time correction in minutes
2152 * - 'format': format to use, can have the following values:
2153 * - true: use user's preference
2154 * - false: use default preference
2155 * - string: format to use
2156 * @since 1.19
2157 * @return String
2158 */
2159 private function internalUserTimeAndDate( $type, $ts, User $user, array $options ) {
2160 $ts = wfTimestamp( TS_MW, $ts );
2161 $options += array( 'timecorrection' => true, 'format' => true );
2162 if ( $options['timecorrection'] !== false ) {
2163 if ( $options['timecorrection'] === true ) {
2164 $offset = $user->getOption( 'timecorrection' );
2165 } else {
2166 $offset = $options['timecorrection'];
2167 }
2168 $ts = $this->userAdjust( $ts, $offset );
2169 }
2170 if ( $options['format'] === true ) {
2171 $format = $user->getDatePreference();
2172 } else {
2173 $format = $options['format'];
2174 }
2175 $df = $this->getDateFormatString( $type, $this->dateFormat( $format ) );
2176 return $this->sprintfDate( $df, $ts );
2177 }
2178
2179 /**
2180 * Get the formatted date for the given timestamp and formatted for
2181 * the given user.
2182 *
2183 * @param $ts Mixed: the time format which needs to be turned into a
2184 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2185 * @param $user User object used to get preferences for timezone and format
2186 * @param $options Array, can contain the following keys:
2187 * - 'timecorrection': time correction, can have the following values:
2188 * - true: use user's preference
2189 * - false: don't use time correction
2190 * - integer: value of time correction in minutes
2191 * - 'format': format to use, can have the following values:
2192 * - true: use user's preference
2193 * - false: use default preference
2194 * - string: format to use
2195 * @since 1.19
2196 * @return String
2197 */
2198 public function userDate( $ts, User $user, array $options = array() ) {
2199 return $this->internalUserTimeAndDate( 'date', $ts, $user, $options );
2200 }
2201
2202 /**
2203 * Get the formatted time for the given timestamp and formatted for
2204 * the given user.
2205 *
2206 * @param $ts Mixed: the time format which needs to be turned into a
2207 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2208 * @param $user User object used to get preferences for timezone and format
2209 * @param $options Array, can contain the following keys:
2210 * - 'timecorrection': time correction, can have the following values:
2211 * - true: use user's preference
2212 * - false: don't use time correction
2213 * - integer: value of time correction in minutes
2214 * - 'format': format to use, can have the following values:
2215 * - true: use user's preference
2216 * - false: use default preference
2217 * - string: format to use
2218 * @since 1.19
2219 * @return String
2220 */
2221 public function userTime( $ts, User $user, array $options = array() ) {
2222 return $this->internalUserTimeAndDate( 'time', $ts, $user, $options );
2223 }
2224
2225 /**
2226 * Get the formatted date and time for the given timestamp and formatted for
2227 * the given user.
2228 *
2229 * @param $ts Mixed: the time format which needs to be turned into a
2230 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
2231 * @param $user User object used to get preferences for timezone and format
2232 * @param $options Array, can contain the following keys:
2233 * - 'timecorrection': time correction, can have the following values:
2234 * - true: use user's preference
2235 * - false: don't use time correction
2236 * - integer: value of time correction in minutes
2237 * - 'format': format to use, can have the following values:
2238 * - true: use user's preference
2239 * - false: use default preference
2240 * - string: format to use
2241 * @since 1.19
2242 * @return String
2243 */
2244 public function userTimeAndDate( $ts, User $user, array $options = array() ) {
2245 return $this->internalUserTimeAndDate( 'both', $ts, $user, $options );
2246 }
2247
2248 /**
2249 * Convert an MWTimestamp into a pretty human-readable timestamp using
2250 * the given user preferences and relative base time.
2251 *
2252 * DO NOT USE THIS FUNCTION DIRECTLY. Instead, call MWTimestamp::getHumanTimestamp
2253 * on your timestamp object, which will then call this function. Calling
2254 * this function directly will cause hooks to be skipped over.
2255 *
2256 * @see MWTimestamp::getHumanTimestamp
2257 * @param MWTimestamp $ts Timestamp to prettify
2258 * @param MWTimestamp $relativeTo Base timestamp
2259 * @param User $user User preferences to use
2260 * @return string Human timestamp
2261 * @since 1.22
2262 */
2263 public function getHumanTimestamp( MWTimestamp $ts, MWTimestamp $relativeTo, User $user ) {
2264 $diff = $ts->diff( $relativeTo );
2265 $diffDay = (bool)( (int)$ts->timestamp->format( 'w' ) - (int)$relativeTo->timestamp->format( 'w' ) );
2266 $days = $diff->days ?: (int)$diffDay;
2267 if ( $diff->invert || $days > 5 && $ts->timestamp->format( 'Y' ) !== $relativeTo->timestamp->format( 'Y' ) ) {
2268 // Timestamps are in different years: use full timestamp
2269 // Also do full timestamp for future dates
2270 /**
2271 * @FIXME Add better handling of future timestamps.
2272 */
2273 $format = $this->getDateFormatString( 'both', $user->getDatePreference() ?: 'default' );
2274 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2275 } elseif ( $days > 5 ) {
2276 // Timestamps are in same year, but more than 5 days ago: show day and month only.
2277 $format = $this->getDateFormatString( 'pretty', $user->getDatePreference() ?: 'default' );
2278 $ts = $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) );
2279 } elseif ( $days > 1 ) {
2280 // Timestamp within the past week: show the day of the week and time
2281 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2282 $weekday = self::$mWeekdayMsgs[$ts->timestamp->format( 'w' )];
2283 $ts = wfMessage( "$weekday-at" )
2284 ->inLanguage( $this )
2285 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2286 ->text();
2287 } elseif ( $days == 1 ) {
2288 // Timestamp was yesterday: say 'yesterday' and the time.
2289 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2290 $ts = wfMessage( 'yesterday-at' )
2291 ->inLanguage( $this )
2292 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2293 ->text();
2294 } elseif ( $diff->h > 1 || $diff->h == 1 && $diff->i > 30 ) {
2295 // Timestamp was today, but more than 90 minutes ago: say 'today' and the time.
2296 $format = $this->getDateFormatString( 'time', $user->getDatePreference() ?: 'default' );
2297 $ts = wfMessage( 'today-at' )
2298 ->inLanguage( $this )
2299 ->params( $this->sprintfDate( $format, $ts->getTimestamp( TS_MW ) ) )
2300 ->text();
2301
2302 // From here on in, the timestamp was soon enough ago so that we can simply say
2303 // XX units ago, e.g., "2 hours ago" or "5 minutes ago"
2304 } elseif ( $diff->h == 1 ) {
2305 // Less than 90 minutes, but more than an hour ago.
2306 $ts = wfMessage( 'hours-ago' )->inLanguage( $this )->numParams( 1 )->text();
2307 } elseif ( $diff->i >= 1 ) {
2308 // A few minutes ago.
2309 $ts = wfMessage( 'minutes-ago' )->inLanguage( $this )->numParams( $diff->i )->text();
2310 } elseif ( $diff->s >= 30 ) {
2311 // Less than a minute, but more than 30 sec ago.
2312 $ts = wfMessage( 'seconds-ago' )->inLanguage( $this )->numParams( $diff->s )->text();
2313 } else {
2314 // Less than 30 seconds ago.
2315 $ts = wfMessage( 'just-now' )->text();
2316 }
2317
2318 return $ts;
2319 }
2320
2321 /**
2322 * @param $key string
2323 * @return array|null
2324 */
2325 function getMessage( $key ) {
2326 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
2327 }
2328
2329 /**
2330 * @return array
2331 */
2332 function getAllMessages() {
2333 return self::$dataCache->getItem( $this->mCode, 'messages' );
2334 }
2335
2336 /**
2337 * @param $in
2338 * @param $out
2339 * @param $string
2340 * @return string
2341 */
2342 function iconv( $in, $out, $string ) {
2343 # This is a wrapper for iconv in all languages except esperanto,
2344 # which does some nasty x-conversions beforehand
2345
2346 # Even with //IGNORE iconv can whine about illegal characters in
2347 # *input* string. We just ignore those too.
2348 # REF: http://bugs.php.net/bug.php?id=37166
2349 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
2350 wfSuppressWarnings();
2351 $text = iconv( $in, $out . '//IGNORE', $string );
2352 wfRestoreWarnings();
2353 return $text;
2354 }
2355
2356 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
2357
2358 /**
2359 * @param $matches array
2360 * @return mixed|string
2361 */
2362 function ucwordbreaksCallbackAscii( $matches ) {
2363 return $this->ucfirst( $matches[1] );
2364 }
2365
2366 /**
2367 * @param $matches array
2368 * @return string
2369 */
2370 function ucwordbreaksCallbackMB( $matches ) {
2371 return mb_strtoupper( $matches[0] );
2372 }
2373
2374 /**
2375 * @param $matches array
2376 * @return string
2377 */
2378 function ucCallback( $matches ) {
2379 list( $wikiUpperChars ) = self::getCaseMaps();
2380 return strtr( $matches[1], $wikiUpperChars );
2381 }
2382
2383 /**
2384 * @param $matches array
2385 * @return string
2386 */
2387 function lcCallback( $matches ) {
2388 list( , $wikiLowerChars ) = self::getCaseMaps();
2389 return strtr( $matches[1], $wikiLowerChars );
2390 }
2391
2392 /**
2393 * @param $matches array
2394 * @return string
2395 */
2396 function ucwordsCallbackMB( $matches ) {
2397 return mb_strtoupper( $matches[0] );
2398 }
2399
2400 /**
2401 * @param $matches array
2402 * @return string
2403 */
2404 function ucwordsCallbackWiki( $matches ) {
2405 list( $wikiUpperChars ) = self::getCaseMaps();
2406 return strtr( $matches[0], $wikiUpperChars );
2407 }
2408
2409 /**
2410 * Make a string's first character uppercase
2411 *
2412 * @param $str string
2413 *
2414 * @return string
2415 */
2416 function ucfirst( $str ) {
2417 $o = ord( $str );
2418 if ( $o < 96 ) { // if already uppercase...
2419 return $str;
2420 } elseif ( $o < 128 ) {
2421 return ucfirst( $str ); // use PHP's ucfirst()
2422 } else {
2423 // fall back to more complex logic in case of multibyte strings
2424 return $this->uc( $str, true );
2425 }
2426 }
2427
2428 /**
2429 * Convert a string to uppercase
2430 *
2431 * @param $str string
2432 * @param $first bool
2433 *
2434 * @return string
2435 */
2436 function uc( $str, $first = false ) {
2437 if ( function_exists( 'mb_strtoupper' ) ) {
2438 if ( $first ) {
2439 if ( $this->isMultibyte( $str ) ) {
2440 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2441 } else {
2442 return ucfirst( $str );
2443 }
2444 } else {
2445 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
2446 }
2447 } else {
2448 if ( $this->isMultibyte( $str ) ) {
2449 $x = $first ? '^' : '';
2450 return preg_replace_callback(
2451 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2452 array( $this, 'ucCallback' ),
2453 $str
2454 );
2455 } else {
2456 return $first ? ucfirst( $str ) : strtoupper( $str );
2457 }
2458 }
2459 }
2460
2461 /**
2462 * @param $str string
2463 * @return mixed|string
2464 */
2465 function lcfirst( $str ) {
2466 $o = ord( $str );
2467 if ( !$o ) {
2468 return strval( $str );
2469 } elseif ( $o >= 128 ) {
2470 return $this->lc( $str, true );
2471 } elseif ( $o > 96 ) {
2472 return $str;
2473 } else {
2474 $str[0] = strtolower( $str[0] );
2475 return $str;
2476 }
2477 }
2478
2479 /**
2480 * @param $str string
2481 * @param $first bool
2482 * @return mixed|string
2483 */
2484 function lc( $str, $first = false ) {
2485 if ( function_exists( 'mb_strtolower' ) ) {
2486 if ( $first ) {
2487 if ( $this->isMultibyte( $str ) ) {
2488 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
2489 } else {
2490 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
2491 }
2492 } else {
2493 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
2494 }
2495 } else {
2496 if ( $this->isMultibyte( $str ) ) {
2497 $x = $first ? '^' : '';
2498 return preg_replace_callback(
2499 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
2500 array( $this, 'lcCallback' ),
2501 $str
2502 );
2503 } else {
2504 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
2505 }
2506 }
2507 }
2508
2509 /**
2510 * @param $str string
2511 * @return bool
2512 */
2513 function isMultibyte( $str ) {
2514 return (bool)preg_match( '/[\x80-\xff]/', $str );
2515 }
2516
2517 /**
2518 * @param $str string
2519 * @return mixed|string
2520 */
2521 function ucwords( $str ) {
2522 if ( $this->isMultibyte( $str ) ) {
2523 $str = $this->lc( $str );
2524
2525 // regexp to find first letter in each word (i.e. after each space)
2526 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2527
2528 // function to use to capitalize a single char
2529 if ( function_exists( 'mb_strtoupper' ) ) {
2530 return preg_replace_callback(
2531 $replaceRegexp,
2532 array( $this, 'ucwordsCallbackMB' ),
2533 $str
2534 );
2535 } else {
2536 return preg_replace_callback(
2537 $replaceRegexp,
2538 array( $this, 'ucwordsCallbackWiki' ),
2539 $str
2540 );
2541 }
2542 } else {
2543 return ucwords( strtolower( $str ) );
2544 }
2545 }
2546
2547 /**
2548 * capitalize words at word breaks
2549 *
2550 * @param $str string
2551 * @return mixed
2552 */
2553 function ucwordbreaks( $str ) {
2554 if ( $this->isMultibyte( $str ) ) {
2555 $str = $this->lc( $str );
2556
2557 // since \b doesn't work for UTF-8, we explicitely define word break chars
2558 $breaks = "[ \-\(\)\}\{\.,\?!]";
2559
2560 // find first letter after word break
2561 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
2562
2563 if ( function_exists( 'mb_strtoupper' ) ) {
2564 return preg_replace_callback(
2565 $replaceRegexp,
2566 array( $this, 'ucwordbreaksCallbackMB' ),
2567 $str
2568 );
2569 } else {
2570 return preg_replace_callback(
2571 $replaceRegexp,
2572 array( $this, 'ucwordsCallbackWiki' ),
2573 $str
2574 );
2575 }
2576 } else {
2577 return preg_replace_callback(
2578 '/\b([\w\x80-\xff]+)\b/',
2579 array( $this, 'ucwordbreaksCallbackAscii' ),
2580 $str
2581 );
2582 }
2583 }
2584
2585 /**
2586 * Return a case-folded representation of $s
2587 *
2588 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
2589 * and $s2 are the same except for the case of their characters. It is not
2590 * necessary for the value returned to make sense when displayed.
2591 *
2592 * Do *not* perform any other normalisation in this function. If a caller
2593 * uses this function when it should be using a more general normalisation
2594 * function, then fix the caller.
2595 *
2596 * @param $s string
2597 *
2598 * @return string
2599 */
2600 function caseFold( $s ) {
2601 return $this->uc( $s );
2602 }
2603
2604 /**
2605 * @param $s string
2606 * @return string
2607 */
2608 function checkTitleEncoding( $s ) {
2609 if ( is_array( $s ) ) {
2610 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
2611 }
2612 if ( StringUtils::isUtf8( $s ) ) {
2613 return $s;
2614 }
2615
2616 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
2617 }
2618
2619 /**
2620 * @return array
2621 */
2622 function fallback8bitEncoding() {
2623 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
2624 }
2625
2626 /**
2627 * Most writing systems use whitespace to break up words.
2628 * Some languages such as Chinese don't conventionally do this,
2629 * which requires special handling when breaking up words for
2630 * searching etc.
2631 *
2632 * @return bool
2633 */
2634 function hasWordBreaks() {
2635 return true;
2636 }
2637
2638 /**
2639 * Some languages such as Chinese require word segmentation,
2640 * Specify such segmentation when overridden in derived class.
2641 *
2642 * @param $string String
2643 * @return String
2644 */
2645 function segmentByWord( $string ) {
2646 return $string;
2647 }
2648
2649 /**
2650 * Some languages have special punctuation need to be normalized.
2651 * Make such changes here.
2652 *
2653 * @param $string String
2654 * @return String
2655 */
2656 function normalizeForSearch( $string ) {
2657 return self::convertDoubleWidth( $string );
2658 }
2659
2660 /**
2661 * convert double-width roman characters to single-width.
2662 * range: ff00-ff5f ~= 0020-007f
2663 *
2664 * @param $string string
2665 *
2666 * @return string
2667 */
2668 protected static function convertDoubleWidth( $string ) {
2669 static $full = null;
2670 static $half = null;
2671
2672 if ( $full === null ) {
2673 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
2674 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
2675 $full = str_split( $fullWidth, 3 );
2676 $half = str_split( $halfWidth );
2677 }
2678
2679 $string = str_replace( $full, $half, $string );
2680 return $string;
2681 }
2682
2683 /**
2684 * @param $string string
2685 * @param $pattern string
2686 * @return string
2687 */
2688 protected static function insertSpace( $string, $pattern ) {
2689 $string = preg_replace( $pattern, " $1 ", $string );
2690 $string = preg_replace( '/ +/', ' ', $string );
2691 return $string;
2692 }
2693
2694 /**
2695 * @param $termsArray array
2696 * @return array
2697 */
2698 function convertForSearchResult( $termsArray ) {
2699 # some languages, e.g. Chinese, need to do a conversion
2700 # in order for search results to be displayed correctly
2701 return $termsArray;
2702 }
2703
2704 /**
2705 * Get the first character of a string.
2706 *
2707 * @param $s string
2708 * @return string
2709 */
2710 function firstChar( $s ) {
2711 $matches = array();
2712 preg_match(
2713 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
2714 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
2715 $s,
2716 $matches
2717 );
2718
2719 if ( isset( $matches[1] ) ) {
2720 if ( strlen( $matches[1] ) != 3 ) {
2721 return $matches[1];
2722 }
2723
2724 // Break down Hangul syllables to grab the first jamo
2725 $code = utf8ToCodepoint( $matches[1] );
2726 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
2727 return $matches[1];
2728 } elseif ( $code < 0xb098 ) {
2729 return "\xe3\x84\xb1";
2730 } elseif ( $code < 0xb2e4 ) {
2731 return "\xe3\x84\xb4";
2732 } elseif ( $code < 0xb77c ) {
2733 return "\xe3\x84\xb7";
2734 } elseif ( $code < 0xb9c8 ) {
2735 return "\xe3\x84\xb9";
2736 } elseif ( $code < 0xbc14 ) {
2737 return "\xe3\x85\x81";
2738 } elseif ( $code < 0xc0ac ) {
2739 return "\xe3\x85\x82";
2740 } elseif ( $code < 0xc544 ) {
2741 return "\xe3\x85\x85";
2742 } elseif ( $code < 0xc790 ) {
2743 return "\xe3\x85\x87";
2744 } elseif ( $code < 0xcc28 ) {
2745 return "\xe3\x85\x88";
2746 } elseif ( $code < 0xce74 ) {
2747 return "\xe3\x85\x8a";
2748 } elseif ( $code < 0xd0c0 ) {
2749 return "\xe3\x85\x8b";
2750 } elseif ( $code < 0xd30c ) {
2751 return "\xe3\x85\x8c";
2752 } elseif ( $code < 0xd558 ) {
2753 return "\xe3\x85\x8d";
2754 } else {
2755 return "\xe3\x85\x8e";
2756 }
2757 } else {
2758 return '';
2759 }
2760 }
2761
2762 function initEncoding() {
2763 # Some languages may have an alternate char encoding option
2764 # (Esperanto X-coding, Japanese furigana conversion, etc)
2765 # If this language is used as the primary content language,
2766 # an override to the defaults can be set here on startup.
2767 }
2768
2769 /**
2770 * @param $s string
2771 * @return string
2772 */
2773 function recodeForEdit( $s ) {
2774 # For some languages we'll want to explicitly specify
2775 # which characters make it into the edit box raw
2776 # or are converted in some way or another.
2777 global $wgEditEncoding;
2778 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
2779 return $s;
2780 } else {
2781 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
2782 }
2783 }
2784
2785 /**
2786 * @param $s string
2787 * @return string
2788 */
2789 function recodeInput( $s ) {
2790 # Take the previous into account.
2791 global $wgEditEncoding;
2792 if ( $wgEditEncoding != '' ) {
2793 $enc = $wgEditEncoding;
2794 } else {
2795 $enc = 'UTF-8';
2796 }
2797 if ( $enc == 'UTF-8' ) {
2798 return $s;
2799 } else {
2800 return $this->iconv( $enc, 'UTF-8', $s );
2801 }
2802 }
2803
2804 /**
2805 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2806 * also cleans up certain backwards-compatible sequences, converting them
2807 * to the modern Unicode equivalent.
2808 *
2809 * This is language-specific for performance reasons only.
2810 *
2811 * @param $s string
2812 *
2813 * @return string
2814 */
2815 function normalize( $s ) {
2816 global $wgAllUnicodeFixes;
2817 $s = UtfNormal::cleanUp( $s );
2818 if ( $wgAllUnicodeFixes ) {
2819 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2820 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2821 }
2822
2823 return $s;
2824 }
2825
2826 /**
2827 * Transform a string using serialized data stored in the given file (which
2828 * must be in the serialized subdirectory of $IP). The file contains pairs
2829 * mapping source characters to destination characters.
2830 *
2831 * The data is cached in process memory. This will go faster if you have the
2832 * FastStringSearch extension.
2833 *
2834 * @param $file string
2835 * @param $string string
2836 *
2837 * @throws MWException
2838 * @return string
2839 */
2840 function transformUsingPairFile( $file, $string ) {
2841 if ( !isset( $this->transformData[$file] ) ) {
2842 $data = wfGetPrecompiledData( $file );
2843 if ( $data === false ) {
2844 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2845 }
2846 $this->transformData[$file] = new ReplacementArray( $data );
2847 }
2848 return $this->transformData[$file]->replace( $string );
2849 }
2850
2851 /**
2852 * For right-to-left language support
2853 *
2854 * @return bool
2855 */
2856 function isRTL() {
2857 return self::$dataCache->getItem( $this->mCode, 'rtl' );
2858 }
2859
2860 /**
2861 * Return the correct HTML 'dir' attribute value for this language.
2862 * @return String
2863 */
2864 function getDir() {
2865 return $this->isRTL() ? 'rtl' : 'ltr';
2866 }
2867
2868 /**
2869 * Return 'left' or 'right' as appropriate alignment for line-start
2870 * for this language's text direction.
2871 *
2872 * Should be equivalent to CSS3 'start' text-align value....
2873 *
2874 * @return String
2875 */
2876 function alignStart() {
2877 return $this->isRTL() ? 'right' : 'left';
2878 }
2879
2880 /**
2881 * Return 'right' or 'left' as appropriate alignment for line-end
2882 * for this language's text direction.
2883 *
2884 * Should be equivalent to CSS3 'end' text-align value....
2885 *
2886 * @return String
2887 */
2888 function alignEnd() {
2889 return $this->isRTL() ? 'left' : 'right';
2890 }
2891
2892 /**
2893 * A hidden direction mark (LRM or RLM), depending on the language direction.
2894 * Unlike getDirMark(), this function returns the character as an HTML entity.
2895 * This function should be used when the output is guaranteed to be HTML,
2896 * because it makes the output HTML source code more readable. When
2897 * the output is plain text or can be escaped, getDirMark() should be used.
2898 *
2899 * @param $opposite Boolean Get the direction mark opposite to your language
2900 * @return string
2901 * @since 1.20
2902 */
2903 function getDirMarkEntity( $opposite = false ) {
2904 if ( $opposite ) {
2905 return $this->isRTL() ? '&lrm;' : '&rlm;';
2906 }
2907 return $this->isRTL() ? '&rlm;' : '&lrm;';
2908 }
2909
2910 /**
2911 * A hidden direction mark (LRM or RLM), depending on the language direction.
2912 * This function produces them as invisible Unicode characters and
2913 * the output may be hard to read and debug, so it should only be used
2914 * when the output is plain text or can be escaped. When the output is
2915 * HTML, use getDirMarkEntity() instead.
2916 *
2917 * @param $opposite Boolean Get the direction mark opposite to your language
2918 * @return string
2919 */
2920 function getDirMark( $opposite = false ) {
2921 $lrm = "\xE2\x80\x8E"; # LEFT-TO-RIGHT MARK, commonly abbreviated LRM
2922 $rlm = "\xE2\x80\x8F"; # RIGHT-TO-LEFT MARK, commonly abbreviated RLM
2923 if ( $opposite ) {
2924 return $this->isRTL() ? $lrm : $rlm;
2925 }
2926 return $this->isRTL() ? $rlm : $lrm;
2927 }
2928
2929 /**
2930 * @return array
2931 */
2932 function capitalizeAllNouns() {
2933 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2934 }
2935
2936 /**
2937 * An arrow, depending on the language direction.
2938 *
2939 * @param $direction String: the direction of the arrow: forwards (default), backwards, left, right, up, down.
2940 * @return string
2941 */
2942 function getArrow( $direction = 'forwards' ) {
2943 switch ( $direction ) {
2944 case 'forwards':
2945 return $this->isRTL() ? '←' : '→';
2946 case 'backwards':
2947 return $this->isRTL() ? '→' : '←';
2948 case 'left':
2949 return '←';
2950 case 'right':
2951 return '→';
2952 case 'up':
2953 return '↑';
2954 case 'down':
2955 return '↓';
2956 }
2957 }
2958
2959 /**
2960 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2961 *
2962 * @return bool
2963 */
2964 function linkPrefixExtension() {
2965 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2966 }
2967
2968 /**
2969 * Get all magic words from cache.
2970 * @return array
2971 */
2972 function getMagicWords() {
2973 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2974 }
2975
2976 /**
2977 * Run the LanguageGetMagic hook once.
2978 */
2979 protected function doMagicHook() {
2980 if ( $this->mMagicHookDone ) {
2981 return;
2982 }
2983 $this->mMagicHookDone = true;
2984 wfProfileIn( 'LanguageGetMagic' );
2985 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2986 wfProfileOut( 'LanguageGetMagic' );
2987 }
2988
2989 /**
2990 * Fill a MagicWord object with data from here
2991 *
2992 * @param $mw
2993 */
2994 function getMagic( $mw ) {
2995 // Saves a function call
2996 if ( ! $this->mMagicHookDone ) {
2997 $this->doMagicHook();
2998 }
2999
3000 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
3001 $rawEntry = $this->mMagicExtensions[$mw->mId];
3002 } else {
3003 $rawEntry = self::$dataCache->getSubitem(
3004 $this->mCode, 'magicWords', $mw->mId );
3005 }
3006
3007 if ( !is_array( $rawEntry ) ) {
3008 error_log( "\"$rawEntry\" is not a valid magic word for \"$mw->mId\"" );
3009 } else {
3010 $mw->mCaseSensitive = $rawEntry[0];
3011 $mw->mSynonyms = array_slice( $rawEntry, 1 );
3012 }
3013 }
3014
3015 /**
3016 * Add magic words to the extension array
3017 *
3018 * @param $newWords array
3019 */
3020 function addMagicWordsByLang( $newWords ) {
3021 $fallbackChain = $this->getFallbackLanguages();
3022 $fallbackChain = array_reverse( $fallbackChain );
3023 foreach ( $fallbackChain as $code ) {
3024 if ( isset( $newWords[$code] ) ) {
3025 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
3026 }
3027 }
3028 }
3029
3030 /**
3031 * Get special page names, as an associative array
3032 * case folded alias => real name
3033 */
3034 function getSpecialPageAliases() {
3035 // Cache aliases because it may be slow to load them
3036 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
3037 // Initialise array
3038 $this->mExtendedSpecialPageAliases =
3039 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
3040 wfRunHooks( 'LanguageGetSpecialPageAliases',
3041 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
3042 }
3043
3044 return $this->mExtendedSpecialPageAliases;
3045 }
3046
3047 /**
3048 * Italic is unsuitable for some languages
3049 *
3050 * @param $text String: the text to be emphasized.
3051 * @return string
3052 */
3053 function emphasize( $text ) {
3054 return "<em>$text</em>";
3055 }
3056
3057 /**
3058 * Normally we output all numbers in plain en_US style, that is
3059 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
3060 * point twohundredthirtyfive. However this is not suitable for all
3061 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
3062 * Icelandic just want to use commas instead of dots, and dots instead
3063 * of commas like "293.291,235".
3064 *
3065 * An example of this function being called:
3066 * <code>
3067 * wfMessage( 'message' )->numParams( $num )->text()
3068 * </code>
3069 *
3070 * See LanguageGu.php for the Gujarati implementation and
3071 * $separatorTransformTable on MessageIs.php for
3072 * the , => . and . => , implementation.
3073 *
3074 * @todo check if it's viable to use localeconv() for the decimal
3075 * separator thing.
3076 * @param $number Mixed: the string to be formatted, should be an integer
3077 * or a floating point number.
3078 * @param $nocommafy Bool: set to true for special numbers like dates
3079 * @return string
3080 */
3081 public function formatNum( $number, $nocommafy = false ) {
3082 global $wgTranslateNumerals;
3083 if ( !$nocommafy ) {
3084 $number = $this->commafy( $number );
3085 $s = $this->separatorTransformTable();
3086 if ( $s ) {
3087 $number = strtr( $number, $s );
3088 }
3089 }
3090
3091 if ( $wgTranslateNumerals ) {
3092 $s = $this->digitTransformTable();
3093 if ( $s ) {
3094 $number = strtr( $number, $s );
3095 }
3096 }
3097
3098 return $number;
3099 }
3100
3101 /**
3102 * Front-end for non-commafied formatNum
3103 *
3104 * @param mixed $number the string to be formatted, should be an integer
3105 * or a floating point number.
3106 * @since 1.21
3107 * @return string
3108 */
3109 public function formatNumNoSeparators( $number ) {
3110 return $this->formatNum( $number, true );
3111 }
3112
3113 /**
3114 * @param $number string
3115 * @return string
3116 */
3117 function parseFormattedNumber( $number ) {
3118 $s = $this->digitTransformTable();
3119 if ( $s ) {
3120 $number = strtr( $number, array_flip( $s ) );
3121 }
3122
3123 $s = $this->separatorTransformTable();
3124 if ( $s ) {
3125 $number = strtr( $number, array_flip( $s ) );
3126 }
3127
3128 $number = strtr( $number, array( ',' => '' ) );
3129 return $number;
3130 }
3131
3132 /**
3133 * Adds commas to a given number
3134 * @since 1.19
3135 * @param $number mixed
3136 * @return string
3137 */
3138 function commafy( $number ) {
3139 $digitGroupingPattern = $this->digitGroupingPattern();
3140 if ( $number === null ) {
3141 return '';
3142 }
3143
3144 if ( !$digitGroupingPattern || $digitGroupingPattern === "###,###,###" ) {
3145 // default grouping is at thousands, use the same for ###,###,### pattern too.
3146 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $number ) ) );
3147 } else {
3148 // Ref: http://cldr.unicode.org/translation/number-patterns
3149 $sign = "";
3150 if ( intval( $number ) < 0 ) {
3151 // For negative numbers apply the algorithm like positive number and add sign.
3152 $sign = "-";
3153 $number = substr( $number, 1 );
3154 }
3155 $integerPart = array();
3156 $decimalPart = array();
3157 $numMatches = preg_match_all( "/(#+)/", $digitGroupingPattern, $matches );
3158 preg_match( "/\d+/", $number, $integerPart );
3159 preg_match( "/\.\d*/", $number, $decimalPart );
3160 $groupedNumber = ( count( $decimalPart ) > 0 ) ? $decimalPart[0] : "";
3161 if ( $groupedNumber === $number ) {
3162 // the string does not have any number part. Eg: .12345
3163 return $sign . $groupedNumber;
3164 }
3165 $start = $end = strlen( $integerPart[0] );
3166 while ( $start > 0 ) {
3167 $match = $matches[0][$numMatches - 1];
3168 $matchLen = strlen( $match );
3169 $start = $end - $matchLen;
3170 if ( $start < 0 ) {
3171 $start = 0;
3172 }
3173 $groupedNumber = substr( $number, $start, $end -$start ) . $groupedNumber;
3174 $end = $start;
3175 if ( $numMatches > 1 ) {
3176 // use the last pattern for the rest of the number
3177 $numMatches--;
3178 }
3179 if ( $start > 0 ) {
3180 $groupedNumber = "," . $groupedNumber;
3181 }
3182 }
3183 return $sign . $groupedNumber;
3184 }
3185 }
3186
3187 /**
3188 * @return String
3189 */
3190 function digitGroupingPattern() {
3191 return self::$dataCache->getItem( $this->mCode, 'digitGroupingPattern' );
3192 }
3193
3194 /**
3195 * @return array
3196 */
3197 function digitTransformTable() {
3198 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
3199 }
3200
3201 /**
3202 * @return array
3203 */
3204 function separatorTransformTable() {
3205 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
3206 }
3207
3208 /**
3209 * Take a list of strings and build a locale-friendly comma-separated
3210 * list, using the local comma-separator message.
3211 * The last two strings are chained with an "and".
3212 * NOTE: This function will only work with standard numeric array keys (0, 1, 2…)
3213 *
3214 * @param $l Array
3215 * @return string
3216 */
3217 function listToText( array $l ) {
3218 $m = count( $l ) - 1;
3219 if ( $m < 0 ) {
3220 return '';
3221 }
3222 if ( $m > 0 ) {
3223 $and = $this->getMessageFromDB( 'and' );
3224 $space = $this->getMessageFromDB( 'word-separator' );
3225 if ( $m > 1 ) {
3226 $comma = $this->getMessageFromDB( 'comma-separator' );
3227 }
3228 }
3229 $s = $l[$m];
3230 for ( $i = $m - 1; $i >= 0; $i-- ) {
3231 if ( $i == $m - 1 ) {
3232 $s = $l[$i] . $and . $space . $s;
3233 } else {
3234 $s = $l[$i] . $comma . $s;
3235 }
3236 }
3237 return $s;
3238 }
3239
3240 /**
3241 * Take a list of strings and build a locale-friendly comma-separated
3242 * list, using the local comma-separator message.
3243 * @param $list array of strings to put in a comma list
3244 * @return string
3245 */
3246 function commaList( array $list ) {
3247 return implode(
3248 wfMessage( 'comma-separator' )->inLanguage( $this )->escaped(),
3249 $list
3250 );
3251 }
3252
3253 /**
3254 * Take a list of strings and build a locale-friendly semicolon-separated
3255 * list, using the local semicolon-separator message.
3256 * @param $list array of strings to put in a semicolon list
3257 * @return string
3258 */
3259 function semicolonList( array $list ) {
3260 return implode(
3261 wfMessage( 'semicolon-separator' )->inLanguage( $this )->escaped(),
3262 $list
3263 );
3264 }
3265
3266 /**
3267 * Same as commaList, but separate it with the pipe instead.
3268 * @param $list array of strings to put in a pipe list
3269 * @return string
3270 */
3271 function pipeList( array $list ) {
3272 return implode(
3273 wfMessage( 'pipe-separator' )->inLanguage( $this )->escaped(),
3274 $list
3275 );
3276 }
3277
3278 /**
3279 * Truncate a string to a specified length in bytes, appending an optional
3280 * string (e.g. for ellipses)
3281 *
3282 * The database offers limited byte lengths for some columns in the database;
3283 * multi-byte character sets mean we need to ensure that only whole characters
3284 * are included, otherwise broken characters can be passed to the user
3285 *
3286 * If $length is negative, the string will be truncated from the beginning
3287 *
3288 * @param $string String to truncate
3289 * @param $length Int: maximum length (including ellipses)
3290 * @param $ellipsis String to append to the truncated text
3291 * @param $adjustLength Boolean: Subtract length of ellipsis from $length.
3292 * $adjustLength was introduced in 1.18, before that behaved as if false.
3293 * @return string
3294 */
3295 function truncate( $string, $length, $ellipsis = '...', $adjustLength = true ) {
3296 # Use the localized ellipsis character
3297 if ( $ellipsis == '...' ) {
3298 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3299 }
3300 # Check if there is no need to truncate
3301 if ( $length == 0 ) {
3302 return $ellipsis; // convention
3303 } elseif ( strlen( $string ) <= abs( $length ) ) {
3304 return $string; // no need to truncate
3305 }
3306 $stringOriginal = $string;
3307 # If ellipsis length is >= $length then we can't apply $adjustLength
3308 if ( $adjustLength && strlen( $ellipsis ) >= abs( $length ) ) {
3309 $string = $ellipsis; // this can be slightly unexpected
3310 # Otherwise, truncate and add ellipsis...
3311 } else {
3312 $eLength = $adjustLength ? strlen( $ellipsis ) : 0;
3313 if ( $length > 0 ) {
3314 $length -= $eLength;
3315 $string = substr( $string, 0, $length ); // xyz...
3316 $string = $this->removeBadCharLast( $string );
3317 $string = $string . $ellipsis;
3318 } else {
3319 $length += $eLength;
3320 $string = substr( $string, $length ); // ...xyz
3321 $string = $this->removeBadCharFirst( $string );
3322 $string = $ellipsis . $string;
3323 }
3324 }
3325 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181).
3326 # This check is *not* redundant if $adjustLength, due to the single case where
3327 # LEN($ellipsis) > ABS($limit arg); $stringOriginal could be shorter than $string.
3328 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
3329 return $string;
3330 } else {
3331 return $stringOriginal;
3332 }
3333 }
3334
3335 /**
3336 * Remove bytes that represent an incomplete Unicode character
3337 * at the end of string (e.g. bytes of the char are missing)
3338 *
3339 * @param $string String
3340 * @return string
3341 */
3342 protected function removeBadCharLast( $string ) {
3343 if ( $string != '' ) {
3344 $char = ord( $string[strlen( $string ) - 1] );
3345 $m = array();
3346 if ( $char >= 0xc0 ) {
3347 # We got the first byte only of a multibyte char; remove it.
3348 $string = substr( $string, 0, -1 );
3349 } elseif ( $char >= 0x80 &&
3350 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
3351 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m )
3352 ) {
3353 # We chopped in the middle of a character; remove it
3354 $string = $m[1];
3355 }
3356 }
3357 return $string;
3358 }
3359
3360 /**
3361 * Remove bytes that represent an incomplete Unicode character
3362 * at the start of string (e.g. bytes of the char are missing)
3363 *
3364 * @param $string String
3365 * @return string
3366 */
3367 protected function removeBadCharFirst( $string ) {
3368 if ( $string != '' ) {
3369 $char = ord( $string[0] );
3370 if ( $char >= 0x80 && $char < 0xc0 ) {
3371 # We chopped in the middle of a character; remove the whole thing
3372 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
3373 }
3374 }
3375 return $string;
3376 }
3377
3378 /**
3379 * Truncate a string of valid HTML to a specified length in bytes,
3380 * appending an optional string (e.g. for ellipses), and return valid HTML
3381 *
3382 * This is only intended for styled/linked text, such as HTML with
3383 * tags like <span> and <a>, were the tags are self-contained (valid HTML).
3384 * Also, this will not detect things like "display:none" CSS.
3385 *
3386 * Note: since 1.18 you do not need to leave extra room in $length for ellipses.
3387 *
3388 * @param string $text HTML string to truncate
3389 * @param int $length (zero/positive) Maximum length (including ellipses)
3390 * @param string $ellipsis String to append to the truncated text
3391 * @return string
3392 */
3393 function truncateHtml( $text, $length, $ellipsis = '...' ) {
3394 # Use the localized ellipsis character
3395 if ( $ellipsis == '...' ) {
3396 $ellipsis = wfMessage( 'ellipsis' )->inLanguage( $this )->escaped();
3397 }
3398 # Check if there is clearly no need to truncate
3399 if ( $length <= 0 ) {
3400 return $ellipsis; // no text shown, nothing to format (convention)
3401 } elseif ( strlen( $text ) <= $length ) {
3402 return $text; // string short enough even *with* HTML (short-circuit)
3403 }
3404
3405 $dispLen = 0; // innerHTML legth so far
3406 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
3407 $tagType = 0; // 0-open, 1-close
3408 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
3409 $entityState = 0; // 0-not entity, 1-entity
3410 $tag = $ret = ''; // accumulated tag name, accumulated result string
3411 $openTags = array(); // open tag stack
3412 $maybeState = null; // possible truncation state
3413
3414 $textLen = strlen( $text );
3415 $neLength = max( 0, $length - strlen( $ellipsis ) ); // non-ellipsis len if truncated
3416 for ( $pos = 0; true; ++$pos ) {
3417 # Consider truncation once the display length has reached the maximim.
3418 # We check if $dispLen > 0 to grab tags for the $neLength = 0 case.
3419 # Check that we're not in the middle of a bracket/entity...
3420 if ( $dispLen && $dispLen >= $neLength && $bracketState == 0 && !$entityState ) {
3421 if ( !$testingEllipsis ) {
3422 $testingEllipsis = true;
3423 # Save where we are; we will truncate here unless there turn out to
3424 # be so few remaining characters that truncation is not necessary.
3425 if ( !$maybeState ) { // already saved? ($neLength = 0 case)
3426 $maybeState = array( $ret, $openTags ); // save state
3427 }
3428 } elseif ( $dispLen > $length && $dispLen > strlen( $ellipsis ) ) {
3429 # String in fact does need truncation, the truncation point was OK.
3430 list( $ret, $openTags ) = $maybeState; // reload state
3431 $ret = $this->removeBadCharLast( $ret ); // multi-byte char fix
3432 $ret .= $ellipsis; // add ellipsis
3433 break;
3434 }
3435 }
3436 if ( $pos >= $textLen ) {
3437 break; // extra iteration just for above checks
3438 }
3439
3440 # Read the next char...
3441 $ch = $text[$pos];
3442 $lastCh = $pos ? $text[$pos - 1] : '';
3443 $ret .= $ch; // add to result string
3444 if ( $ch == '<' ) {
3445 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
3446 $entityState = 0; // for bad HTML
3447 $bracketState = 1; // tag started (checking for backslash)
3448 } elseif ( $ch == '>' ) {
3449 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
3450 $entityState = 0; // for bad HTML
3451 $bracketState = 0; // out of brackets
3452 } elseif ( $bracketState == 1 ) {
3453 if ( $ch == '/' ) {
3454 $tagType = 1; // close tag (e.g. "</span>")
3455 } else {
3456 $tagType = 0; // open tag (e.g. "<span>")
3457 $tag .= $ch;
3458 }
3459 $bracketState = 2; // building tag name
3460 } elseif ( $bracketState == 2 ) {
3461 if ( $ch != ' ' ) {
3462 $tag .= $ch;
3463 } else {
3464 // Name found (e.g. "<a href=..."), add on tag attributes...
3465 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
3466 }
3467 } elseif ( $bracketState == 0 ) {
3468 if ( $entityState ) {
3469 if ( $ch == ';' ) {
3470 $entityState = 0;
3471 $dispLen++; // entity is one displayed char
3472 }
3473 } else {
3474 if ( $neLength == 0 && !$maybeState ) {
3475 // Save state without $ch. We want to *hit* the first
3476 // display char (to get tags) but not *use* it if truncating.
3477 $maybeState = array( substr( $ret, 0, -1 ), $openTags );
3478 }
3479 if ( $ch == '&' ) {
3480 $entityState = 1; // entity found, (e.g. "&#160;")
3481 } else {
3482 $dispLen++; // this char is displayed
3483 // Add the next $max display text chars after this in one swoop...
3484 $max = ( $testingEllipsis ? $length : $neLength ) - $dispLen;
3485 $skipped = $this->truncate_skip( $ret, $text, "<>&", $pos + 1, $max );
3486 $dispLen += $skipped;
3487 $pos += $skipped;
3488 }
3489 }
3490 }
3491 }
3492 // Close the last tag if left unclosed by bad HTML
3493 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
3494 while ( count( $openTags ) > 0 ) {
3495 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
3496 }
3497 return $ret;
3498 }
3499
3500 /**
3501 * truncateHtml() helper function
3502 * like strcspn() but adds the skipped chars to $ret
3503 *
3504 * @param $ret
3505 * @param $text
3506 * @param $search
3507 * @param $start
3508 * @param $len
3509 * @return int
3510 */
3511 private function truncate_skip( &$ret, $text, $search, $start, $len = null ) {
3512 if ( $len === null ) {
3513 $len = -1; // -1 means "no limit" for strcspn
3514 } elseif ( $len < 0 ) {
3515 $len = 0; // sanity
3516 }
3517 $skipCount = 0;
3518 if ( $start < strlen( $text ) ) {
3519 $skipCount = strcspn( $text, $search, $start, $len );
3520 $ret .= substr( $text, $start, $skipCount );
3521 }
3522 return $skipCount;
3523 }
3524
3525 /**
3526 * truncateHtml() helper function
3527 * (a) push or pop $tag from $openTags as needed
3528 * (b) clear $tag value
3529 * @param &$tag string Current HTML tag name we are looking at
3530 * @param $tagType int (0-open tag, 1-close tag)
3531 * @param $lastCh string Character before the '>' that ended this tag
3532 * @param &$openTags array Open tag stack (not accounting for $tag)
3533 */
3534 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
3535 $tag = ltrim( $tag );
3536 if ( $tag != '' ) {
3537 if ( $tagType == 0 && $lastCh != '/' ) {
3538 $openTags[] = $tag; // tag opened (didn't close itself)
3539 } elseif ( $tagType == 1 ) {
3540 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
3541 array_pop( $openTags ); // tag closed
3542 }
3543 }
3544 $tag = '';
3545 }
3546 }
3547
3548 /**
3549 * Grammatical transformations, needed for inflected languages
3550 * Invoked by putting {{grammar:case|word}} in a message
3551 *
3552 * @param $word string
3553 * @param $case string
3554 * @return string
3555 */
3556 function convertGrammar( $word, $case ) {
3557 global $wgGrammarForms;
3558 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
3559 return $wgGrammarForms[$this->getCode()][$case][$word];
3560 }
3561 return $word;
3562 }
3563 /**
3564 * Get the grammar forms for the content language
3565 * @return array of grammar forms
3566 * @since 1.20
3567 */
3568 function getGrammarForms() {
3569 global $wgGrammarForms;
3570 if ( isset( $wgGrammarForms[$this->getCode()] ) && is_array( $wgGrammarForms[$this->getCode()] ) ) {
3571 return $wgGrammarForms[$this->getCode()];
3572 }
3573 return array();
3574 }
3575 /**
3576 * Provides an alternative text depending on specified gender.
3577 * Usage {{gender:username|masculine|feminine|neutral}}.
3578 * username is optional, in which case the gender of current user is used,
3579 * but only in (some) interface messages; otherwise default gender is used.
3580 *
3581 * If no forms are given, an empty string is returned. If only one form is
3582 * given, it will be returned unconditionally. These details are implied by
3583 * the caller and cannot be overridden in subclasses.
3584 *
3585 * If more than one form is given, the default is to use the neutral one
3586 * if it is specified, and to use the masculine one otherwise. These
3587 * details can be overridden in subclasses.
3588 *
3589 * @param $gender string
3590 * @param $forms array
3591 *
3592 * @return string
3593 */
3594 function gender( $gender, $forms ) {
3595 if ( !count( $forms ) ) {
3596 return '';
3597 }
3598 $forms = $this->preConvertPlural( $forms, 2 );
3599 if ( $gender === 'male' ) {
3600 return $forms[0];
3601 }
3602 if ( $gender === 'female' ) {
3603 return $forms[1];
3604 }
3605 return isset( $forms[2] ) ? $forms[2] : $forms[0];
3606 }
3607
3608 /**
3609 * Plural form transformations, needed for some languages.
3610 * For example, there are 3 form of plural in Russian and Polish,
3611 * depending on "count mod 10". See [[w:Plural]]
3612 * For English it is pretty simple.
3613 *
3614 * Invoked by putting {{plural:count|wordform1|wordform2}}
3615 * or {{plural:count|wordform1|wordform2|wordform3}}
3616 *
3617 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
3618 *
3619 * @param $count Integer: non-localized number
3620 * @param $forms Array: different plural forms
3621 * @return string Correct form of plural for $count in this language
3622 */
3623 function convertPlural( $count, $forms ) {
3624 // Handle explicit n=pluralform cases
3625 foreach ( $forms as $index => $form ) {
3626 if ( preg_match( '/^\d+=/i', $form ) ) {
3627 $pos = strpos( $form, '=' );
3628 if ( substr( $form, 0, $pos ) === (string) $count ) {
3629 return substr( $form, $pos + 1 );
3630 }
3631 unset( $forms[$index] );
3632 }
3633 }
3634
3635 $forms = array_values( $forms );
3636 if ( !count( $forms ) ) {
3637 return '';
3638 }
3639
3640 $pluralForm = $this->getPluralRuleIndexNumber( $count );
3641 $pluralForm = min( $pluralForm, count( $forms ) - 1 );
3642 return $forms[$pluralForm];
3643 }
3644
3645 /**
3646 * Checks that convertPlural was given an array and pads it to requested
3647 * amount of forms by copying the last one.
3648 *
3649 * @param $count Integer: How many forms should there be at least
3650 * @param $forms Array of forms given to convertPlural
3651 * @return array Padded array of forms or an exception if not an array
3652 */
3653 protected function preConvertPlural( /* Array */ $forms, $count ) {
3654 while ( count( $forms ) < $count ) {
3655 $forms[] = $forms[count( $forms ) - 1];
3656 }
3657 return $forms;
3658 }
3659
3660 /**
3661 * @todo Maybe translate block durations. Note that this function is somewhat misnamed: it
3662 * deals with translating the *duration* ("1 week", "4 days", etc), not the expiry time
3663 * (which is an absolute timestamp). Please note: do NOT add this blindly, as it is used
3664 * on old expiry lengths recorded in log entries. You'd need to provide the start date to
3665 * match up with it.
3666 *
3667 * @param $str String: the validated block duration in English
3668 * @return string Somehow translated block duration
3669 * @see LanguageFi.php for example implementation
3670 */
3671 function translateBlockExpiry( $str ) {
3672 $duration = SpecialBlock::getSuggestedDurations( $this );
3673 foreach ( $duration as $show => $value ) {
3674 if ( strcmp( $str, $value ) == 0 ) {
3675 return htmlspecialchars( trim( $show ) );
3676 }
3677 }
3678
3679 // Since usually only infinite or indefinite is only on list, so try
3680 // equivalents if still here.
3681 $indefs = array( 'infinite', 'infinity', 'indefinite' );
3682 if ( in_array( $str, $indefs ) ) {
3683 foreach ( $indefs as $val ) {
3684 $show = array_search( $val, $duration, true );
3685 if ( $show !== false ) {
3686 return htmlspecialchars( trim( $show ) );
3687 }
3688 }
3689 }
3690
3691 // If all else fails, return a standard duration or timestamp description.
3692 $time = strtotime( $str, 0 );
3693 if ( $time === false ) { // Unknown format. Return it as-is in case.
3694 return $str;
3695 } elseif ( $time !== strtotime( $str, 1 ) ) { // It's a relative timestamp.
3696 // $time is relative to 0 so it's a duration length.
3697 return $this->formatDuration( $time );
3698 } else { // It's an absolute timestamp.
3699 if ( $time === 0 ) {
3700 // wfTimestamp() handles 0 as current time instead of epoch.
3701 return $this->timeanddate( '19700101000000' );
3702 } else {
3703 return $this->timeanddate( $time );
3704 }
3705 }
3706 }
3707
3708 /**
3709 * languages like Chinese need to be segmented in order for the diff
3710 * to be of any use
3711 *
3712 * @param $text String
3713 * @return String
3714 */
3715 public function segmentForDiff( $text ) {
3716 return $text;
3717 }
3718
3719 /**
3720 * and unsegment to show the result
3721 *
3722 * @param $text String
3723 * @return String
3724 */
3725 public function unsegmentForDiff( $text ) {
3726 return $text;
3727 }
3728
3729 /**
3730 * Return the LanguageConverter used in the Language
3731 *
3732 * @since 1.19
3733 * @return LanguageConverter
3734 */
3735 public function getConverter() {
3736 return $this->mConverter;
3737 }
3738
3739 /**
3740 * convert text to all supported variants
3741 *
3742 * @param $text string
3743 * @return array
3744 */
3745 public function autoConvertToAllVariants( $text ) {
3746 return $this->mConverter->autoConvertToAllVariants( $text );
3747 }
3748
3749 /**
3750 * convert text to different variants of a language.
3751 *
3752 * @param $text string
3753 * @return string
3754 */
3755 public function convert( $text ) {
3756 return $this->mConverter->convert( $text );
3757 }
3758
3759 /**
3760 * Convert a Title object to a string in the preferred variant
3761 *
3762 * @param $title Title
3763 * @return string
3764 */
3765 public function convertTitle( $title ) {
3766 return $this->mConverter->convertTitle( $title );
3767 }
3768
3769 /**
3770 * Convert a namespace index to a string in the preferred variant
3771 *
3772 * @param $ns int
3773 * @return string
3774 */
3775 public function convertNamespace( $ns ) {
3776 return $this->mConverter->convertNamespace( $ns );
3777 }
3778
3779 /**
3780 * Check if this is a language with variants
3781 *
3782 * @return bool
3783 */
3784 public function hasVariants() {
3785 return count( $this->getVariants() ) > 1;
3786 }
3787
3788 /**
3789 * Check if the language has the specific variant
3790 *
3791 * @since 1.19
3792 * @param $variant string
3793 * @return bool
3794 */
3795 public function hasVariant( $variant ) {
3796 return (bool)$this->mConverter->validateVariant( $variant );
3797 }
3798
3799 /**
3800 * Put custom tags (e.g. -{ }-) around math to prevent conversion
3801 *
3802 * @param $text string
3803 * @return string
3804 */
3805 public function armourMath( $text ) {
3806 return $this->mConverter->armourMath( $text );
3807 }
3808
3809 /**
3810 * Perform output conversion on a string, and encode for safe HTML output.
3811 * @param $text String text to be converted
3812 * @param $isTitle Bool whether this conversion is for the article title
3813 * @return string
3814 * @todo this should get integrated somewhere sane
3815 */
3816 public function convertHtml( $text, $isTitle = false ) {
3817 return htmlspecialchars( $this->convert( $text, $isTitle ) );
3818 }
3819
3820 /**
3821 * @param $key string
3822 * @return string
3823 */
3824 public function convertCategoryKey( $key ) {
3825 return $this->mConverter->convertCategoryKey( $key );
3826 }
3827
3828 /**
3829 * Get the list of variants supported by this language
3830 * see sample implementation in LanguageZh.php
3831 *
3832 * @return array an array of language codes
3833 */
3834 public function getVariants() {
3835 return $this->mConverter->getVariants();
3836 }
3837
3838 /**
3839 * @return string
3840 */
3841 public function getPreferredVariant() {
3842 return $this->mConverter->getPreferredVariant();
3843 }
3844
3845 /**
3846 * @return string
3847 */
3848 public function getDefaultVariant() {
3849 return $this->mConverter->getDefaultVariant();
3850 }
3851
3852 /**
3853 * @return string
3854 */
3855 public function getURLVariant() {
3856 return $this->mConverter->getURLVariant();
3857 }
3858
3859 /**
3860 * If a language supports multiple variants, it is
3861 * possible that non-existing link in one variant
3862 * actually exists in another variant. this function
3863 * tries to find it. See e.g. LanguageZh.php
3864 *
3865 * @param $link String: the name of the link
3866 * @param $nt Mixed: the title object of the link
3867 * @param $ignoreOtherCond Boolean: to disable other conditions when
3868 * we need to transclude a template or update a category's link
3869 * @return null the input parameters may be modified upon return
3870 */
3871 public function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
3872 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
3873 }
3874
3875 /**
3876 * If a language supports multiple variants, converts text
3877 * into an array of all possible variants of the text:
3878 * 'variant' => text in that variant
3879 *
3880 * @deprecated since 1.17 Use autoConvertToAllVariants()
3881 *
3882 * @param $text string
3883 *
3884 * @return string
3885 */
3886 public function convertLinkToAllVariants( $text ) {
3887 return $this->mConverter->convertLinkToAllVariants( $text );
3888 }
3889
3890 /**
3891 * returns language specific options used by User::getPageRenderHash()
3892 * for example, the preferred language variant
3893 *
3894 * @return string
3895 */
3896 function getExtraHashOptions() {
3897 return $this->mConverter->getExtraHashOptions();
3898 }
3899
3900 /**
3901 * For languages that support multiple variants, the title of an
3902 * article may be displayed differently in different variants. this
3903 * function returns the apporiate title defined in the body of the article.
3904 *
3905 * @return string
3906 */
3907 public function getParsedTitle() {
3908 return $this->mConverter->getParsedTitle();
3909 }
3910
3911 /**
3912 * Prepare external link text for conversion. When the text is
3913 * a URL, it shouldn't be converted, and it'll be wrapped in
3914 * the "raw" tag (-{R| }-) to prevent conversion.
3915 *
3916 * This function is called "markNoConversion" for historical
3917 * reasons.
3918 *
3919 * @param $text String: text to be used for external link
3920 * @param $noParse bool: wrap it without confirming it's a real URL first
3921 * @return string the tagged text
3922 */
3923 public function markNoConversion( $text, $noParse = false ) {
3924 // Excluding protocal-relative URLs may avoid many false positives.
3925 if ( $noParse || preg_match( '/^(?:' . wfUrlProtocolsWithoutProtRel() . ')/', $text ) ) {
3926 return $this->mConverter->markNoConversion( $text );
3927 } else {
3928 return $text;
3929 }
3930 }
3931
3932 /**
3933 * A regular expression to match legal word-trailing characters
3934 * which should be merged onto a link of the form [[foo]]bar.
3935 *
3936 * @return string
3937 */
3938 public function linkTrail() {
3939 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
3940 }
3941
3942 /**
3943 * @return Language
3944 */
3945 function getLangObj() {
3946 return $this;
3947 }
3948
3949 /**
3950 * Get the "parent" language which has a converter to convert a "compatible" language
3951 * (in another variant) to this language (eg. zh for zh-cn, but not en for en-gb).
3952 *
3953 * @return Language|null
3954 * @since 1.22
3955 */
3956 public function getParentLanguage() {
3957 if ( $this->mParentLanguage !== false ) {
3958 return $this->mParentLanguage;
3959 }
3960
3961 $pieces = explode( '-', $this->getCode() );
3962 $code = $pieces[0];
3963 if ( !in_array( $code, LanguageConverter::$languagesWithVariants ) ) {
3964 $this->mParentLanguage = null;
3965 return null;
3966 }
3967 $lang = Language::factory( $code );
3968 if ( !$lang->hasVariant( $this->getCode() ) ) {
3969 $this->mParentLanguage = null;
3970 return null;
3971 }
3972
3973 $this->mParentLanguage = $lang;
3974 return $lang;
3975 }
3976
3977 /**
3978 * Get the RFC 3066 code for this language object
3979 *
3980 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3981 * htmlspecialchars() or similar
3982 *
3983 * @return string
3984 */
3985 public function getCode() {
3986 return $this->mCode;
3987 }
3988
3989 /**
3990 * Get the code in Bcp47 format which we can use
3991 * inside of html lang="" tags.
3992 *
3993 * NOTE: The return value of this function is NOT HTML-safe and must be escaped with
3994 * htmlspecialchars() or similar.
3995 *
3996 * @since 1.19
3997 * @return string
3998 */
3999 public function getHtmlCode() {
4000 if ( is_null( $this->mHtmlCode ) ) {
4001 $this->mHtmlCode = wfBCP47( $this->getCode() );
4002 }
4003 return $this->mHtmlCode;
4004 }
4005
4006 /**
4007 * @param $code string
4008 */
4009 public function setCode( $code ) {
4010 $this->mCode = $code;
4011 // Ensure we don't leave incorrect cached data lying around
4012 $this->mHtmlCode = null;
4013 $this->mParentLanguage = false;
4014 }
4015
4016 /**
4017 * Get the name of a file for a certain language code
4018 * @param $prefix string Prepend this to the filename
4019 * @param $code string Language code
4020 * @param $suffix string Append this to the filename
4021 * @throws MWException
4022 * @return string $prefix . $mangledCode . $suffix
4023 */
4024 public static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
4025 // Protect against path traversal
4026 if ( !Language::isValidCode( $code )
4027 || strcspn( $code, ":/\\\000" ) !== strlen( $code ) )
4028 {
4029 throw new MWException( "Invalid language code \"$code\"" );
4030 }
4031
4032 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
4033 }
4034
4035 /**
4036 * Get the language code from a file name. Inverse of getFileName()
4037 * @param $filename string $prefix . $languageCode . $suffix
4038 * @param $prefix string Prefix before the language code
4039 * @param $suffix string Suffix after the language code
4040 * @return string Language code, or false if $prefix or $suffix isn't found
4041 */
4042 public static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
4043 $m = null;
4044 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
4045 preg_quote( $suffix, '/' ) . '/', $filename, $m );
4046 if ( !count( $m ) ) {
4047 return false;
4048 }
4049 return str_replace( '_', '-', strtolower( $m[1] ) );
4050 }
4051
4052 /**
4053 * @param $code string
4054 * @return string
4055 */
4056 public static function getMessagesFileName( $code ) {
4057 global $IP;
4058 $file = self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
4059 wfRunHooks( 'Language::getMessagesFileName', array( $code, &$file ) );
4060 return $file;
4061 }
4062
4063 /**
4064 * @param $code string
4065 * @return string
4066 */
4067 public static function getClassFileName( $code ) {
4068 global $IP;
4069 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
4070 }
4071
4072 /**
4073 * Get the first fallback for a given language.
4074 *
4075 * @param $code string
4076 *
4077 * @return bool|string
4078 */
4079 public static function getFallbackFor( $code ) {
4080 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4081 return false;
4082 } else {
4083 $fallbacks = self::getFallbacksFor( $code );
4084 $first = array_shift( $fallbacks );
4085 return $first;
4086 }
4087 }
4088
4089 /**
4090 * Get the ordered list of fallback languages.
4091 *
4092 * @since 1.19
4093 * @param $code string Language code
4094 * @return array
4095 */
4096 public static function getFallbacksFor( $code ) {
4097 if ( $code === 'en' || !Language::isValidBuiltInCode( $code ) ) {
4098 return array();
4099 } else {
4100 $v = self::getLocalisationCache()->getItem( $code, 'fallback' );
4101 $v = array_map( 'trim', explode( ',', $v ) );
4102 if ( $v[count( $v ) - 1] !== 'en' ) {
4103 $v[] = 'en';
4104 }
4105 return $v;
4106 }
4107 }
4108
4109 /**
4110 * Get the ordered list of fallback languages, ending with the fallback
4111 * language chain for the site language.
4112 *
4113 * @since 1.22
4114 * @param string $code Language code
4115 * @return array array( fallbacks, site fallbacks )
4116 */
4117 public static function getFallbacksIncludingSiteLanguage( $code ) {
4118 global $wgLanguageCode;
4119
4120 // Usually, we will only store a tiny number of fallback chains, so we
4121 // keep them in static memory.
4122 $cacheKey = "{$code}-{$wgLanguageCode}";
4123
4124 if ( !array_key_exists( $cacheKey, self::$fallbackLanguageCache ) ) {
4125 $fallbacks = self::getFallbacksFor( $code );
4126
4127 // Append the site's fallback chain, including the site language itself
4128 $siteFallbacks = self::getFallbacksFor( $wgLanguageCode );
4129 array_unshift( $siteFallbacks, $wgLanguageCode );
4130
4131 // Eliminate any languages already included in the chain
4132 $siteFallbacks = array_diff( $siteFallbacks, $fallbacks );
4133
4134 self::$fallbackLanguageCache[$cacheKey] = array( $fallbacks, $siteFallbacks );
4135 }
4136 return self::$fallbackLanguageCache[$cacheKey];
4137 }
4138
4139 /**
4140 * Get all messages for a given language
4141 * WARNING: this may take a long time. If you just need all message *keys*
4142 * but need the *contents* of only a few messages, consider using getMessageKeysFor().
4143 *
4144 * @param $code string
4145 *
4146 * @return array
4147 */
4148 public static function getMessagesFor( $code ) {
4149 return self::getLocalisationCache()->getItem( $code, 'messages' );
4150 }
4151
4152 /**
4153 * Get a message for a given language
4154 *
4155 * @param $key string
4156 * @param $code string
4157 *
4158 * @return string
4159 */
4160 public static function getMessageFor( $key, $code ) {
4161 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
4162 }
4163
4164 /**
4165 * Get all message keys for a given language. This is a faster alternative to
4166 * array_keys( Language::getMessagesFor( $code ) )
4167 *
4168 * @since 1.19
4169 * @param $code string Language code
4170 * @return array of message keys (strings)
4171 */
4172 public static function getMessageKeysFor( $code ) {
4173 return self::getLocalisationCache()->getSubItemList( $code, 'messages' );
4174 }
4175
4176 /**
4177 * @param $talk
4178 * @return mixed
4179 */
4180 function fixVariableInNamespace( $talk ) {
4181 if ( strpos( $talk, '$1' ) === false ) {
4182 return $talk;
4183 }
4184
4185 global $wgMetaNamespace;
4186 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
4187
4188 # Allow grammar transformations
4189 # Allowing full message-style parsing would make simple requests
4190 # such as action=raw much more expensive than they need to be.
4191 # This will hopefully cover most cases.
4192 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
4193 array( &$this, 'replaceGrammarInNamespace' ), $talk );
4194 return str_replace( ' ', '_', $talk );
4195 }
4196
4197 /**
4198 * @param $m string
4199 * @return string
4200 */
4201 function replaceGrammarInNamespace( $m ) {
4202 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
4203 }
4204
4205 /**
4206 * @throws MWException
4207 * @return array
4208 */
4209 static function getCaseMaps() {
4210 static $wikiUpperChars, $wikiLowerChars;
4211 if ( isset( $wikiUpperChars ) ) {
4212 return array( $wikiUpperChars, $wikiLowerChars );
4213 }
4214
4215 wfProfileIn( __METHOD__ );
4216 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
4217 if ( $arr === false ) {
4218 throw new MWException(
4219 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
4220 }
4221 $wikiUpperChars = $arr['wikiUpperChars'];
4222 $wikiLowerChars = $arr['wikiLowerChars'];
4223 wfProfileOut( __METHOD__ );
4224 return array( $wikiUpperChars, $wikiLowerChars );
4225 }
4226
4227 /**
4228 * Decode an expiry (block, protection, etc) which has come from the DB
4229 *
4230 * @todo FIXME: why are we returnings DBMS-dependent strings???
4231 *
4232 * @param $expiry String: Database expiry String
4233 * @param $format Bool|Int true to process using language functions, or TS_ constant
4234 * to return the expiry in a given timestamp
4235 * @return String
4236 * @since 1.18
4237 */
4238 public function formatExpiry( $expiry, $format = true ) {
4239 static $infinity;
4240 if ( $infinity === null ) {
4241 $infinity = wfGetDB( DB_SLAVE )->getInfinity();
4242 }
4243
4244 if ( $expiry == '' || $expiry == $infinity ) {
4245 return $format === true
4246 ? $this->getMessageFromDB( 'infiniteblock' )
4247 : $infinity;
4248 } else {
4249 return $format === true
4250 ? $this->timeanddate( $expiry, /* User preference timezone */ true )
4251 : wfTimestamp( $format, $expiry );
4252 }
4253 }
4254
4255 /**
4256 * @todo Document
4257 * @param $seconds int|float
4258 * @param $format Array Optional
4259 * If $format['avoid'] == 'avoidseconds' - don't mention seconds if $seconds >= 1 hour
4260 * If $format['avoid'] == 'avoidminutes' - don't mention seconds/minutes if $seconds > 48 hours
4261 * If $format['noabbrevs'] is true - use 'seconds' and friends instead of 'seconds-abbrev' and friends
4262 * For backwards compatibility, $format may also be one of the strings 'avoidseconds' or 'avoidminutes'
4263 * @return string
4264 */
4265 function formatTimePeriod( $seconds, $format = array() ) {
4266 if ( !is_array( $format ) ) {
4267 $format = array( 'avoid' => $format ); // For backwards compatibility
4268 }
4269 if ( !isset( $format['avoid'] ) ) {
4270 $format['avoid'] = false;
4271 }
4272 if ( !isset( $format['noabbrevs' ] ) ) {
4273 $format['noabbrevs'] = false;
4274 }
4275 $secondsMsg = wfMessage(
4276 $format['noabbrevs'] ? 'seconds' : 'seconds-abbrev' )->inLanguage( $this );
4277 $minutesMsg = wfMessage(
4278 $format['noabbrevs'] ? 'minutes' : 'minutes-abbrev' )->inLanguage( $this );
4279 $hoursMsg = wfMessage(
4280 $format['noabbrevs'] ? 'hours' : 'hours-abbrev' )->inLanguage( $this );
4281 $daysMsg = wfMessage(
4282 $format['noabbrevs'] ? 'days' : 'days-abbrev' )->inLanguage( $this );
4283
4284 if ( round( $seconds * 10 ) < 100 ) {
4285 $s = $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) );
4286 $s = $secondsMsg->params( $s )->text();
4287 } elseif ( round( $seconds ) < 60 ) {
4288 $s = $this->formatNum( round( $seconds ) );
4289 $s = $secondsMsg->params( $s )->text();
4290 } elseif ( round( $seconds ) < 3600 ) {
4291 $minutes = floor( $seconds / 60 );
4292 $secondsPart = round( fmod( $seconds, 60 ) );
4293 if ( $secondsPart == 60 ) {
4294 $secondsPart = 0;
4295 $minutes++;
4296 }
4297 $s = $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4298 $s .= ' ';
4299 $s .= $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4300 } elseif ( round( $seconds ) <= 2 * 86400 ) {
4301 $hours = floor( $seconds / 3600 );
4302 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
4303 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
4304 if ( $secondsPart == 60 ) {
4305 $secondsPart = 0;
4306 $minutes++;
4307 }
4308 if ( $minutes == 60 ) {
4309 $minutes = 0;
4310 $hours++;
4311 }
4312 $s = $hoursMsg->params( $this->formatNum( $hours ) )->text();
4313 $s .= ' ';
4314 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4315 if ( !in_array( $format['avoid'], array( 'avoidseconds', 'avoidminutes' ) ) ) {
4316 $s .= ' ' . $secondsMsg->params( $this->formatNum( $secondsPart ) )->text();
4317 }
4318 } else {
4319 $days = floor( $seconds / 86400 );
4320 if ( $format['avoid'] === 'avoidminutes' ) {
4321 $hours = round( ( $seconds - $days * 86400 ) / 3600 );
4322 if ( $hours == 24 ) {
4323 $hours = 0;
4324 $days++;
4325 }
4326 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4327 $s .= ' ';
4328 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4329 } elseif ( $format['avoid'] === 'avoidseconds' ) {
4330 $hours = floor( ( $seconds - $days * 86400 ) / 3600 );
4331 $minutes = round( ( $seconds - $days * 86400 - $hours * 3600 ) / 60 );
4332 if ( $minutes == 60 ) {
4333 $minutes = 0;
4334 $hours++;
4335 }
4336 if ( $hours == 24 ) {
4337 $hours = 0;
4338 $days++;
4339 }
4340 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4341 $s .= ' ';
4342 $s .= $hoursMsg->params( $this->formatNum( $hours ) )->text();
4343 $s .= ' ';
4344 $s .= $minutesMsg->params( $this->formatNum( $minutes ) )->text();
4345 } else {
4346 $s = $daysMsg->params( $this->formatNum( $days ) )->text();
4347 $s .= ' ';
4348 $s .= $this->formatTimePeriod( $seconds - $days * 86400, $format );
4349 }
4350 }
4351 return $s;
4352 }
4353
4354 /**
4355 * Format a bitrate for output, using an appropriate
4356 * unit (bps, kbps, Mbps, Gbps, Tbps, Pbps, Ebps, Zbps or Ybps) according to the magnitude in question
4357 *
4358 * This use base 1000. For base 1024 use formatSize(), for another base
4359 * see formatComputingNumbers()
4360 *
4361 * @param $bps int
4362 * @return string
4363 */
4364 function formatBitrate( $bps ) {
4365 return $this->formatComputingNumbers( $bps, 1000, "bitrate-$1bits" );
4366 }
4367
4368 /**
4369 * @param $size int Size of the unit
4370 * @param $boundary int Size boundary (1000, or 1024 in most cases)
4371 * @param $messageKey string Message key to be uesd
4372 * @return string
4373 */
4374 function formatComputingNumbers( $size, $boundary, $messageKey ) {
4375 if ( $size <= 0 ) {
4376 return str_replace( '$1', $this->formatNum( $size ),
4377 $this->getMessageFromDB( str_replace( '$1', '', $messageKey ) )
4378 );
4379 }
4380 $sizes = array( '', 'kilo', 'mega', 'giga', 'tera', 'peta', 'exa', 'zeta', 'yotta' );
4381 $index = 0;
4382
4383 $maxIndex = count( $sizes ) - 1;
4384 while ( $size >= $boundary && $index < $maxIndex ) {
4385 $index++;
4386 $size /= $boundary;
4387 }
4388
4389 // For small sizes no decimal places necessary
4390 $round = 0;
4391 if ( $index > 1 ) {
4392 // For MB and bigger two decimal places are smarter
4393 $round = 2;
4394 }
4395 $msg = str_replace( '$1', $sizes[$index], $messageKey );
4396
4397 $size = round( $size, $round );
4398 $text = $this->getMessageFromDB( $msg );
4399 return str_replace( '$1', $this->formatNum( $size ), $text );
4400 }
4401
4402 /**
4403 * Format a size in bytes for output, using an appropriate
4404 * unit (B, KB, MB, GB, TB, PB, EB, ZB or YB) according to the magnitude in question
4405 *
4406 * This method use base 1024. For base 1000 use formatBitrate(), for
4407 * another base see formatComputingNumbers()
4408 *
4409 * @param $size int Size to format
4410 * @return string Plain text (not HTML)
4411 */
4412 function formatSize( $size ) {
4413 return $this->formatComputingNumbers( $size, 1024, "size-$1bytes" );
4414 }
4415
4416 /**
4417 * Make a list item, used by various special pages
4418 *
4419 * @param $page String Page link
4420 * @param $details String Text between brackets
4421 * @param $oppositedm Boolean Add the direction mark opposite to your
4422 * language, to display text properly
4423 * @return String
4424 */
4425 function specialList( $page, $details, $oppositedm = true ) {
4426 $dirmark = ( $oppositedm ? $this->getDirMark( true ) : '' ) .
4427 $this->getDirMark();
4428 $details = $details ? $dirmark . $this->getMessageFromDB( 'word-separator' ) .
4429 wfMessage( 'parentheses' )->rawParams( $details )->inLanguage( $this )->escaped() : '';
4430 return $page . $details;
4431 }
4432
4433 /**
4434 * Generate (prev x| next x) (20|50|100...) type links for paging
4435 *
4436 * @param $title Title object to link
4437 * @param $offset Integer offset parameter
4438 * @param $limit Integer limit parameter
4439 * @param $query array|String optional URL query parameter string
4440 * @param $atend Bool optional param for specified if this is the last page
4441 * @return String
4442 */
4443 public function viewPrevNext( Title $title, $offset, $limit, array $query = array(), $atend = false ) {
4444 // @todo FIXME: Why on earth this needs one message for the text and another one for tooltip?
4445
4446 # Make 'previous' link
4447 $prev = wfMessage( 'prevn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4448 if ( $offset > 0 ) {
4449 $plink = $this->numLink( $title, max( $offset - $limit, 0 ), $limit,
4450 $query, $prev, 'prevn-title', 'mw-prevlink' );
4451 } else {
4452 $plink = htmlspecialchars( $prev );
4453 }
4454
4455 # Make 'next' link
4456 $next = wfMessage( 'nextn' )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4457 if ( $atend ) {
4458 $nlink = htmlspecialchars( $next );
4459 } else {
4460 $nlink = $this->numLink( $title, $offset + $limit, $limit,
4461 $query, $next, 'prevn-title', 'mw-nextlink' );
4462 }
4463
4464 # Make links to set number of items per page
4465 $numLinks = array();
4466 foreach ( array( 20, 50, 100, 250, 500 ) as $num ) {
4467 $numLinks[] = $this->numLink( $title, $offset, $num,
4468 $query, $this->formatNum( $num ), 'shown-title', 'mw-numlink' );
4469 }
4470
4471 return wfMessage( 'viewprevnext' )->inLanguage( $this )->title( $title
4472 )->rawParams( $plink, $nlink, $this->pipeList( $numLinks ) )->escaped();
4473 }
4474
4475 /**
4476 * Helper function for viewPrevNext() that generates links
4477 *
4478 * @param $title Title object to link
4479 * @param $offset Integer offset parameter
4480 * @param $limit Integer limit parameter
4481 * @param $query Array extra query parameters
4482 * @param $link String text to use for the link; will be escaped
4483 * @param $tooltipMsg String name of the message to use as tooltip
4484 * @param $class String value of the "class" attribute of the link
4485 * @return String HTML fragment
4486 */
4487 private function numLink( Title $title, $offset, $limit, array $query, $link, $tooltipMsg, $class ) {
4488 $query = array( 'limit' => $limit, 'offset' => $offset ) + $query;
4489 $tooltip = wfMessage( $tooltipMsg )->inLanguage( $this )->title( $title )->numParams( $limit )->text();
4490 return Html::element( 'a', array( 'href' => $title->getLocalURL( $query ),
4491 'title' => $tooltip, 'class' => $class ), $link );
4492 }
4493
4494 /**
4495 * Get the conversion rule title, if any.
4496 *
4497 * @return string
4498 */
4499 public function getConvRuleTitle() {
4500 return $this->mConverter->getConvRuleTitle();
4501 }
4502
4503 /**
4504 * Get the compiled plural rules for the language
4505 * @since 1.20
4506 * @return array Associative array with plural form, and plural rule as key-value pairs
4507 */
4508 public function getCompiledPluralRules() {
4509 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'compiledPluralRules' );
4510 $fallbacks = Language::getFallbacksFor( $this->mCode );
4511 if ( !$pluralRules ) {
4512 foreach ( $fallbacks as $fallbackCode ) {
4513 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'compiledPluralRules' );
4514 if ( $pluralRules ) {
4515 break;
4516 }
4517 }
4518 }
4519 return $pluralRules;
4520 }
4521
4522 /**
4523 * Get the plural rules for the language
4524 * @since 1.20
4525 * @return array Associative array with plural form number and plural rule as key-value pairs
4526 */
4527 public function getPluralRules() {
4528 $pluralRules = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRules' );
4529 $fallbacks = Language::getFallbacksFor( $this->mCode );
4530 if ( !$pluralRules ) {
4531 foreach ( $fallbacks as $fallbackCode ) {
4532 $pluralRules = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRules' );
4533 if ( $pluralRules ) {
4534 break;
4535 }
4536 }
4537 }
4538 return $pluralRules;
4539 }
4540
4541 /**
4542 * Get the plural rule types for the language
4543 * @since 1.22
4544 * @return array Associative array with plural form number and plural rule type as key-value pairs
4545 */
4546 public function getPluralRuleTypes() {
4547 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $this->mCode ), 'pluralRuleTypes' );
4548 $fallbacks = Language::getFallbacksFor( $this->mCode );
4549 if ( !$pluralRuleTypes ) {
4550 foreach ( $fallbacks as $fallbackCode ) {
4551 $pluralRuleTypes = self::$dataCache->getItem( strtolower( $fallbackCode ), 'pluralRuleTypes' );
4552 if ( $pluralRuleTypes ) {
4553 break;
4554 }
4555 }
4556 }
4557 return $pluralRuleTypes;
4558 }
4559
4560 /**
4561 * Find the index number of the plural rule appropriate for the given number
4562 * @return int The index number of the plural rule
4563 */
4564 public function getPluralRuleIndexNumber( $number ) {
4565 $pluralRules = $this->getCompiledPluralRules();
4566 $form = CLDRPluralRuleEvaluator::evaluateCompiled( $number, $pluralRules );
4567 return $form;
4568 }
4569
4570 /**
4571 * Find the plural rule type appropriate for the given number
4572 * For example, if the language is set to Arabic, getPluralType(5) should
4573 * return 'few'.
4574 * @since 1.22
4575 * @return string The name of the plural rule type, e.g. one, two, few, many
4576 */
4577 public function getPluralRuleType( $number ) {
4578 $index = $this->getPluralRuleIndexNumber( $number );
4579 $pluralRuleTypes = $this->getPluralRuleTypes();
4580 if ( isset( $pluralRuleTypes[$index] ) ) {
4581 return $pluralRuleTypes[$index];
4582 } else {
4583 return 'other';
4584 }
4585 }
4586 }
Wiklou, le Wiki du Wiklou