dépôts / lhc / web / wiklou.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
history | raw | HEAD
* (bug 17160) Gender specific display text for User namespace
[lhc/web/wiklou.git] / languages / Language.php
1 <?php
2 /**
3 * Internationalisation code
4 *
5 * @file
6 * @ingroup Language
7 */
8
9 /**
10 * @defgroup Language Language
11 */
12
13 if ( !defined( 'MEDIAWIKI' ) ) {
14 echo "This file is part of MediaWiki, it is not a valid entry point.\n";
15 exit( 1 );
16 }
17
18 # Read language names
19 global $wgLanguageNames;
20 require_once( dirname( __FILE__ ) . '/Names.php' );
21
22 global $wgInputEncoding, $wgOutputEncoding;
23
24 /**
25 * These are always UTF-8, they exist only for backwards compatibility
26 */
27 $wgInputEncoding = 'UTF-8';
28 $wgOutputEncoding = 'UTF-8';
29
30 if ( function_exists( 'mb_strtoupper' ) ) {
31 mb_internal_encoding( 'UTF-8' );
32 }
33
34 /**
35 * a fake language converter
36 *
37 * @ingroup Language
38 */
39 class FakeConverter {
40 var $mLang;
41 function __construct( $langobj ) { $this->mLang = $langobj; }
42 function autoConvertToAllVariants( $text ) { return array( $this->mLang->getCode() => $text ); }
43 function convert( $t ) { return $t; }
44 function convertTitle( $t ) { return $t->getPrefixedText(); }
45 function getVariants() { return array( $this->mLang->getCode() ); }
46 function getPreferredVariant() { return $this->mLang->getCode(); }
47 function getDefaultVariant() { return $this->mLang->getCode(); }
48 function getURLVariant() { return ''; }
49 function getConvRuleTitle() { return false; }
50 function findVariantLink( &$l, &$n, $ignoreOtherCond = false ) { }
51 function getExtraHashOptions() { return ''; }
52 function getParsedTitle() { return ''; }
53 function markNoConversion( $text, $noParse = false ) { return $text; }
54 function convertCategoryKey( $key ) { return $key; }
55 function convertLinkToAllVariants( $text ) { return $this->autoConvertToAllVariants( $text ); }
56 function armourMath( $text ) { return $text; }
57 }
58
59 /**
60 * Internationalisation code
61 * @ingroup Language
62 */
63 class Language {
64 var $mConverter, $mVariants, $mCode, $mLoaded = false;
65 var $mMagicExtensions = array(), $mMagicHookDone = false;
66
67 var $mNamespaceIds, $namespaceNames, $namespaceAliases;
68 var $dateFormatStrings = array();
69 var $mExtendedSpecialPageAliases;
70
71 /**
72 * ReplacementArray object caches
73 */
74 var $transformData = array();
75
76 static public $dataCache;
77 static public $mLangObjCache = array();
78
79 static public $mWeekdayMsgs = array(
80 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
81 'friday', 'saturday'
82 );
83
84 static public $mWeekdayAbbrevMsgs = array(
85 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
86 );
87
88 static public $mMonthMsgs = array(
89 'january', 'february', 'march', 'april', 'may_long', 'june',
90 'july', 'august', 'september', 'october', 'november',
91 'december'
92 );
93 static public $mMonthGenMsgs = array(
94 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
95 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
96 'december-gen'
97 );
98 static public $mMonthAbbrevMsgs = array(
99 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
100 'sep', 'oct', 'nov', 'dec'
101 );
102
103 static public $mIranianCalendarMonthMsgs = array(
104 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
105 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
106 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
107 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
108 );
109
110 static public $mHebrewCalendarMonthMsgs = array(
111 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
112 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
113 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
114 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
115 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
116 );
117
118 static public $mHebrewCalendarMonthGenMsgs = array(
119 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
120 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
121 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
122 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
123 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
124 );
125
126 static public $mHijriCalendarMonthMsgs = array(
127 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
128 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
129 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
130 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
131 );
132
133 /**
134 * Get a cached language object for a given language code
135 * @param $code String
136 * @return Language
137 */
138 static function factory( $code ) {
139 if ( !isset( self::$mLangObjCache[$code] ) ) {
140 if ( count( self::$mLangObjCache ) > 10 ) {
141 // Don't keep a billion objects around, that's stupid.
142 self::$mLangObjCache = array();
143 }
144 self::$mLangObjCache[$code] = self::newFromCode( $code );
145 }
146 return self::$mLangObjCache[$code];
147 }
148
149 /**
150 * Create a language object for a given language code
151 * @param $code String
152 * @return Language
153 */
154 protected static function newFromCode( $code ) {
155 global $IP;
156 static $recursionLevel = 0;
157
158 // Protect against path traversal below
159 if ( !Language::isValidCode( $code ) ) {
160 throw new MWException( "Invalid language code \"$code\"" );
161 }
162
163 if ( $code == 'en' ) {
164 $class = 'Language';
165 } else {
166 $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
167 // Preload base classes to work around APC/PHP5 bug
168 if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
169 include_once( "$IP/languages/classes/$class.deps.php" );
170 }
171 if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
172 include_once( "$IP/languages/classes/$class.php" );
173 }
174 }
175
176 if ( $recursionLevel > 5 ) {
177 throw new MWException( "Language fallback loop detected when creating class $class\n" );
178 }
179
180 if ( !class_exists( $class ) ) {
181 $fallback = Language::getFallbackFor( $code );
182 ++$recursionLevel;
183 $lang = Language::newFromCode( $fallback );
184 --$recursionLevel;
185 $lang->setCode( $code );
186 } else {
187 $lang = new $class;
188 }
189 return $lang;
190 }
191
192 /**
193 * Returns true if a language code string is of a valid form, whether or
194 * not it exists.
195 */
196 public static function isValidCode( $code ) {
197 return strcspn( $code, "/\\\000" ) === strlen( $code );
198 }
199
200 /**
201 * Get the LocalisationCache instance
202 */
203 public static function getLocalisationCache() {
204 if ( is_null( self::$dataCache ) ) {
205 global $wgLocalisationCacheConf;
206 $class = $wgLocalisationCacheConf['class'];
207 self::$dataCache = new $class( $wgLocalisationCacheConf );
208 }
209 return self::$dataCache;
210 }
211
212 function __construct() {
213 $this->mConverter = new FakeConverter( $this );
214 // Set the code to the name of the descendant
215 if ( get_class( $this ) == 'Language' ) {
216 $this->mCode = 'en';
217 } else {
218 $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
219 }
220 self::getLocalisationCache();
221 }
222
223 /**
224 * Reduce memory usage
225 */
226 function __destruct() {
227 foreach ( $this as $name => $value ) {
228 unset( $this->$name );
229 }
230 }
231
232 /**
233 * Hook which will be called if this is the content language.
234 * Descendants can use this to register hook functions or modify globals
235 */
236 function initContLang() { }
237
238 /**
239 * @deprecated Use User::getDefaultOptions()
240 * @return array
241 */
242 function getDefaultUserOptions() {
243 wfDeprecated( __METHOD__ );
244 return User::getDefaultOptions();
245 }
246
247 function getFallbackLanguageCode() {
248 if ( $this->mCode === 'en' ) {
249 return false;
250 } else {
251 return self::$dataCache->getItem( $this->mCode, 'fallback' );
252 }
253 }
254
255 /**
256 * Exports $wgBookstoreListEn
257 * @return array
258 */
259 function getBookstoreList() {
260 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
261 }
262
263 /**
264 * @return array
265 */
266 function getNamespaces() {
267 if ( is_null( $this->namespaceNames ) ) {
268 global $wgMetaNamespace, $wgMetaNamespaceTalk, $wgExtraNamespaces;
269
270 $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
271 $validNamespaces = MWNamespace::getCanonicalNamespaces();
272
273 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames + $validNamespaces;
274
275 $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
276 if ( $wgMetaNamespaceTalk ) {
277 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
278 } else {
279 $talk = $this->namespaceNames[NS_PROJECT_TALK];
280 $this->namespaceNames[NS_PROJECT_TALK] =
281 $this->fixVariableInNamespace( $talk );
282 }
283
284 # Sometimes a language will be localised but not actually exist on this wiki.
285 foreach( $this->namespaceNames as $key => $text ) {
286 if ( !isset( $validNamespaces[$key] ) ) {
287 unset( $this->namespaceNames[$key] );
288 }
289 }
290
291 # The above mixing may leave namespaces out of canonical order.
292 # Re-order by namespace ID number...
293 ksort( $this->namespaceNames );
294 }
295 return $this->namespaceNames;
296 }
297
298 /**
299 * A convenience function that returns the same thing as
300 * getNamespaces() except with the array values changed to ' '
301 * where it found '_', useful for producing output to be displayed
302 * e.g. in <select> forms.
303 *
304 * @return array
305 */
306 function getFormattedNamespaces() {
307 $ns = $this->getNamespaces();
308 foreach ( $ns as $k => $v ) {
309 $ns[$k] = strtr( $v, '_', ' ' );
310 }
311 return $ns;
312 }
313
314 /**
315 * Get a namespace value by key
316 * <code>
317 * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
318 * echo $mw_ns; // prints 'MediaWiki'
319 * </code>
320 *
321 * @param $index Int: the array key of the namespace to return
322 * @return mixed, string if the namespace value exists, otherwise false
323 */
324 function getNsText( $index ) {
325 $ns = $this->getNamespaces();
326 return isset( $ns[$index] ) ? $ns[$index] : false;
327 }
328
329 /**
330 * A convenience function that returns the same thing as
331 * getNsText() except with '_' changed to ' ', useful for
332 * producing output.
333 *
334 * @return array
335 */
336 function getFormattedNsText( $index ) {
337 $ns = $this->getNsText( $index );
338 return strtr( $ns, '_', ' ' );
339 }
340
341 /**
342 * Returns gender-dependent namespace alias if available.
343 * @param $index Int: namespace index
344 * @param $gender String: gender key (male, female... )
345 * @return String
346 * @since 1.18
347 */
348 function getGenderNsText( $index, $gender ) {
349 $ns = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
350 return isset( $ns[$index][$gender] ) ? $ns[$index][$gender] : $this->getNsText( $index );
351 }
352
353 /**
354 * Whether this language makes distinguishes genders for example in
355 * namespaces.
356 * @return bool
357 * @since 1.18
358 */
359 function needsGenderDistinction() {
360 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
361 return count( $aliases ) > 0;
362 }
363
364 /**
365 * Get a namespace key by value, case insensitive.
366 * Only matches namespace names for the current language, not the
367 * canonical ones defined in Namespace.php.
368 *
369 * @param $text String
370 * @return mixed An integer if $text is a valid value otherwise false
371 */
372 function getLocalNsIndex( $text ) {
373 $lctext = $this->lc( $text );
374 $ids = $this->getNamespaceIds();
375 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
376 }
377
378 function getNamespaceAliases() {
379 if ( is_null( $this->namespaceAliases ) ) {
380 $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
381 if ( !$aliases ) {
382 $aliases = array();
383 } else {
384 foreach ( $aliases as $name => $index ) {
385 if ( $index === NS_PROJECT_TALK ) {
386 unset( $aliases[$name] );
387 $name = $this->fixVariableInNamespace( $name );
388 $aliases[$name] = $index;
389 }
390 }
391 }
392
393 $genders = self::$dataCache->getItem( $this->mCode, 'namespaceGenderAliases' );
394 foreach ( $genders as $index => $forms ) {
395 foreach ( $forms as $alias ) {
396 $aliases[$alias] = $index;
397 }
398 }
399
400 $this->namespaceAliases = $aliases;
401 }
402 return $this->namespaceAliases;
403 }
404
405 function getNamespaceIds() {
406 if ( is_null( $this->mNamespaceIds ) ) {
407 global $wgNamespaceAliases;
408 # Put namespace names and aliases into a hashtable.
409 # If this is too slow, then we should arrange it so that it is done
410 # before caching. The catch is that at pre-cache time, the above
411 # class-specific fixup hasn't been done.
412 $this->mNamespaceIds = array();
413 foreach ( $this->getNamespaces() as $index => $name ) {
414 $this->mNamespaceIds[$this->lc( $name )] = $index;
415 }
416 foreach ( $this->getNamespaceAliases() as $name => $index ) {
417 $this->mNamespaceIds[$this->lc( $name )] = $index;
418 }
419 if ( $wgNamespaceAliases ) {
420 foreach ( $wgNamespaceAliases as $name => $index ) {
421 $this->mNamespaceIds[$this->lc( $name )] = $index;
422 }
423 }
424 }
425 return $this->mNamespaceIds;
426 }
427
428
429 /**
430 * Get a namespace key by value, case insensitive. Canonical namespace
431 * names override custom ones defined for the current language.
432 *
433 * @param $text String
434 * @return mixed An integer if $text is a valid value otherwise false
435 */
436 function getNsIndex( $text ) {
437 $lctext = $this->lc( $text );
438 if ( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) {
439 return $ns;
440 }
441 $ids = $this->getNamespaceIds();
442 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
443 }
444
445 /**
446 * short names for language variants used for language conversion links.
447 *
448 * @param $code String
449 * @return string
450 */
451 function getVariantname( $code ) {
452 return $this->getMessageFromDB( "variantname-$code" );
453 }
454
455 function specialPage( $name ) {
456 $aliases = $this->getSpecialPageAliases();
457 if ( isset( $aliases[$name][0] ) ) {
458 $name = $aliases[$name][0];
459 }
460 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
461 }
462
463 function getQuickbarSettings() {
464 return array(
465 $this->getMessage( 'qbsettings-none' ),
466 $this->getMessage( 'qbsettings-fixedleft' ),
467 $this->getMessage( 'qbsettings-fixedright' ),
468 $this->getMessage( 'qbsettings-floatingleft' ),
469 $this->getMessage( 'qbsettings-floatingright' )
470 );
471 }
472
473 function getMathNames() {
474 return self::$dataCache->getItem( $this->mCode, 'mathNames' );
475 }
476
477 function getDatePreferences() {
478 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
479 }
480
481 function getDateFormats() {
482 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
483 }
484
485 function getDefaultDateFormat() {
486 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
487 if ( $df === 'dmy or mdy' ) {
488 global $wgAmericanDates;
489 return $wgAmericanDates ? 'mdy' : 'dmy';
490 } else {
491 return $df;
492 }
493 }
494
495 function getDatePreferenceMigrationMap() {
496 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
497 }
498
499 function getImageFile( $image ) {
500 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
501 }
502
503 function getDefaultUserOptionOverrides() {
504 return self::$dataCache->getItem( $this->mCode, 'defaultUserOptionOverrides' );
505 }
506
507 function getExtraUserToggles() {
508 return self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
509 }
510
511 function getUserToggle( $tog ) {
512 return $this->getMessageFromDB( "tog-$tog" );
513 }
514
515 /**
516 * Get language names, indexed by code.
517 * If $customisedOnly is true, only returns codes with a messages file
518 */
519 public static function getLanguageNames( $customisedOnly = false ) {
520 global $wgLanguageNames, $wgExtraLanguageNames;
521 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
522 if ( !$customisedOnly ) {
523 return $allNames;
524 }
525
526 global $IP;
527 $names = array();
528 $dir = opendir( "$IP/languages/messages" );
529 while ( false !== ( $file = readdir( $dir ) ) ) {
530 $code = self::getCodeFromFileName( $file, 'Messages' );
531 if ( $code && isset( $allNames[$code] ) ) {
532 $names[$code] = $allNames[$code];
533 }
534 }
535 closedir( $dir );
536 return $names;
537 }
538
539 /**
540 * Get translated language names. This is done on best effort and
541 * by default this is exactly the same as Language::getLanguageNames.
542 * The CLDR extension provides translated names.
543 * @param $code String Language code.
544 * @return Array language code => language name
545 * @since 1.18.0
546 */
547 public static function getTranslatedLanguageNames( $code ) {
548 $names = array();
549 wfRunHooks( 'LanguageGetTranslatedLanguageNames', array( &$names, $code ) );
550
551 foreach ( self::getLanguageNames() as $code => $name ) {
552 if ( !isset( $names[$code] ) ) $names[$code] = $name;
553 }
554
555 return $names;
556 }
557
558 /**
559 * Get a message from the MediaWiki namespace.
560 *
561 * @param $msg String: message name
562 * @return string
563 */
564 function getMessageFromDB( $msg ) {
565 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
566 }
567
568 function getLanguageName( $code ) {
569 $names = self::getLanguageNames();
570 if ( !array_key_exists( $code, $names ) ) {
571 return '';
572 }
573 return $names[$code];
574 }
575
576 function getMonthName( $key ) {
577 return $this->getMessageFromDB( self::$mMonthMsgs[$key - 1] );
578 }
579
580 function getMonthNameGen( $key ) {
581 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key - 1] );
582 }
583
584 function getMonthAbbreviation( $key ) {
585 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key - 1] );
586 }
587
588 function getWeekdayName( $key ) {
589 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key - 1] );
590 }
591
592 function getWeekdayAbbreviation( $key ) {
593 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key - 1] );
594 }
595
596 function getIranianCalendarMonthName( $key ) {
597 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key - 1] );
598 }
599
600 function getHebrewCalendarMonthName( $key ) {
601 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key - 1] );
602 }
603
604 function getHebrewCalendarMonthNameGen( $key ) {
605 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key - 1] );
606 }
607
608 function getHijriCalendarMonthName( $key ) {
609 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key - 1] );
610 }
611
612 /**
613 * Used by date() and time() to adjust the time output.
614 *
615 * @param $ts Int the time in date('YmdHis') format
616 * @param $tz Mixed: adjust the time by this amount (default false, mean we
617 * get user timecorrection setting)
618 * @return int
619 */
620 function userAdjust( $ts, $tz = false ) {
621 global $wgUser, $wgLocalTZoffset;
622
623 if ( $tz === false ) {
624 $tz = $wgUser->getOption( 'timecorrection' );
625 }
626
627 $data = explode( '|', $tz, 3 );
628
629 if ( $data[0] == 'ZoneInfo' ) {
630 if ( function_exists( 'timezone_open' ) && @timezone_open( $data[2] ) !== false ) {
631 $date = date_create( $ts, timezone_open( 'UTC' ) );
632 date_timezone_set( $date, timezone_open( $data[2] ) );
633 $date = date_format( $date, 'YmdHis' );
634 return $date;
635 }
636 # Unrecognized timezone, default to 'Offset' with the stored offset.
637 $data[0] = 'Offset';
638 }
639
640 $minDiff = 0;
641 if ( $data[0] == 'System' || $tz == '' ) {
642 #  Global offset in minutes.
643 if ( isset( $wgLocalTZoffset ) ) {
644 $minDiff = $wgLocalTZoffset;
645 }
646 } else if ( $data[0] == 'Offset' ) {
647 $minDiff = intval( $data[1] );
648 } else {
649 $data = explode( ':', $tz );
650 if ( count( $data ) == 2 ) {
651 $data[0] = intval( $data[0] );
652 $data[1] = intval( $data[1] );
653 $minDiff = abs( $data[0] ) * 60 + $data[1];
654 if ( $data[0] < 0 ) {
655 $minDiff = -$minDiff;
656 }
657 } else {
658 $minDiff = intval( $data[0] ) * 60;
659 }
660 }
661
662 # No difference ? Return time unchanged
663 if ( 0 == $minDiff ) {
664 return $ts;
665 }
666
667 wfSuppressWarnings(); // E_STRICT system time bitching
668 # Generate an adjusted date; take advantage of the fact that mktime
669 # will normalize out-of-range values so we don't have to split $minDiff
670 # into hours and minutes.
671 $t = mktime( (
672 (int)substr( $ts, 8, 2 ) ), # Hours
673 (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
674 (int)substr( $ts, 12, 2 ), # Seconds
675 (int)substr( $ts, 4, 2 ), # Month
676 (int)substr( $ts, 6, 2 ), # Day
677 (int)substr( $ts, 0, 4 ) ); # Year
678
679 $date = date( 'YmdHis', $t );
680 wfRestoreWarnings();
681
682 return $date;
683 }
684
685 /**
686 * This is a workalike of PHP's date() function, but with better
687 * internationalisation, a reduced set of format characters, and a better
688 * escaping format.
689 *
690 * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
691 * PHP manual for definitions. There are a number of extensions, which
692 * start with "x":
693 *
694 * xn Do not translate digits of the next numeric format character
695 * xN Toggle raw digit (xn) flag, stays set until explicitly unset
696 * xr Use roman numerals for the next numeric format character
697 * xh Use hebrew numerals for the next numeric format character
698 * xx Literal x
699 * xg Genitive month name
700 *
701 * xij j (day number) in Iranian calendar
702 * xiF F (month name) in Iranian calendar
703 * xin n (month number) in Iranian calendar
704 * xiY Y (full year) in Iranian calendar
705 *
706 * xjj j (day number) in Hebrew calendar
707 * xjF F (month name) in Hebrew calendar
708 * xjt t (days in month) in Hebrew calendar
709 * xjx xg (genitive month name) in Hebrew calendar
710 * xjn n (month number) in Hebrew calendar
711 * xjY Y (full year) in Hebrew calendar
712 *
713 * xmj j (day number) in Hijri calendar
714 * xmF F (month name) in Hijri calendar
715 * xmn n (month number) in Hijri calendar
716 * xmY Y (full year) in Hijri calendar
717 *
718 * xkY Y (full year) in Thai solar calendar. Months and days are
719 * identical to the Gregorian calendar
720 * xoY Y (full year) in Minguo calendar or Juche year.
721 * Months and days are identical to the
722 * Gregorian calendar
723 * xtY Y (full year) in Japanese nengo. Months and days are
724 * identical to the Gregorian calendar
725 *
726 * Characters enclosed in double quotes will be considered literal (with
727 * the quotes themselves removed). Unmatched quotes will be considered
728 * literal quotes. Example:
729 *
730 * "The month is" F => The month is January
731 * i's" => 20'11"
732 *
733 * Backslash escaping is also supported.
734 *
735 * Input timestamp is assumed to be pre-normalized to the desired local
736 * time zone, if any.
737 *
738 * @param $format String
739 * @param $ts String: 14-character timestamp
740 * YYYYMMDDHHMMSS
741 * 01234567890123
742 * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
743 */
744 function sprintfDate( $format, $ts ) {
745 $s = '';
746 $raw = false;
747 $roman = false;
748 $hebrewNum = false;
749 $unix = false;
750 $rawToggle = false;
751 $iranian = false;
752 $hebrew = false;
753 $hijri = false;
754 $thai = false;
755 $minguo = false;
756 $tenno = false;
757 for ( $p = 0; $p < strlen( $format ); $p++ ) {
758 $num = false;
759 $code = $format[$p];
760 if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
761 $code .= $format[++$p];
762 }
763
764 if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
765 $code .= $format[++$p];
766 }
767
768 switch ( $code ) {
769 case 'xx':
770 $s .= 'x';
771 break;
772 case 'xn':
773 $raw = true;
774 break;
775 case 'xN':
776 $rawToggle = !$rawToggle;
777 break;
778 case 'xr':
779 $roman = true;
780 break;
781 case 'xh':
782 $hebrewNum = true;
783 break;
784 case 'xg':
785 $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
786 break;
787 case 'xjx':
788 if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
789 $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
790 break;
791 case 'd':
792 $num = substr( $ts, 6, 2 );
793 break;
794 case 'D':
795 if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
796 $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
797 break;
798 case 'j':
799 $num = intval( substr( $ts, 6, 2 ) );
800 break;
801 case 'xij':
802 if ( !$iranian ) {
803 $iranian = self::tsToIranian( $ts );
804 }
805 $num = $iranian[2];
806 break;
807 case 'xmj':
808 if ( !$hijri ) {
809 $hijri = self::tsToHijri( $ts );
810 }
811 $num = $hijri[2];
812 break;
813 case 'xjj':
814 if ( !$hebrew ) {
815 $hebrew = self::tsToHebrew( $ts );
816 }
817 $num = $hebrew[2];
818 break;
819 case 'l':
820 if ( !$unix ) {
821 $unix = wfTimestamp( TS_UNIX, $ts );
822 }
823 $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
824 break;
825 case 'N':
826 if ( !$unix ) {
827 $unix = wfTimestamp( TS_UNIX, $ts );
828 }
829 $w = gmdate( 'w', $unix );
830 $num = $w ? $w : 7;
831 break;
832 case 'w':
833 if ( !$unix ) {
834 $unix = wfTimestamp( TS_UNIX, $ts );
835 }
836 $num = gmdate( 'w', $unix );
837 break;
838 case 'z':
839 if ( !$unix ) {
840 $unix = wfTimestamp( TS_UNIX, $ts );
841 }
842 $num = gmdate( 'z', $unix );
843 break;
844 case 'W':
845 if ( !$unix ) {
846 $unix = wfTimestamp( TS_UNIX, $ts );
847 }
848 $num = gmdate( 'W', $unix );
849 break;
850 case 'F':
851 $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
852 break;
853 case 'xiF':
854 if ( !$iranian ) {
855 $iranian = self::tsToIranian( $ts );
856 }
857 $s .= $this->getIranianCalendarMonthName( $iranian[1] );
858 break;
859 case 'xmF':
860 if ( !$hijri ) {
861 $hijri = self::tsToHijri( $ts );
862 }
863 $s .= $this->getHijriCalendarMonthName( $hijri[1] );
864 break;
865 case 'xjF':
866 if ( !$hebrew ) {
867 $hebrew = self::tsToHebrew( $ts );
868 }
869 $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
870 break;
871 case 'm':
872 $num = substr( $ts, 4, 2 );
873 break;
874 case 'M':
875 $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
876 break;
877 case 'n':
878 $num = intval( substr( $ts, 4, 2 ) );
879 break;
880 case 'xin':
881 if ( !$iranian ) {
882 $iranian = self::tsToIranian( $ts );
883 }
884 $num = $iranian[1];
885 break;
886 case 'xmn':
887 if ( !$hijri ) {
888 $hijri = self::tsToHijri ( $ts );
889 }
890 $num = $hijri[1];
891 break;
892 case 'xjn':
893 if ( !$hebrew ) {
894 $hebrew = self::tsToHebrew( $ts );
895 }
896 $num = $hebrew[1];
897 break;
898 case 't':
899 if ( !$unix ) {
900 $unix = wfTimestamp( TS_UNIX, $ts );
901 }
902 $num = gmdate( 't', $unix );
903 break;
904 case 'xjt':
905 if ( !$hebrew ) {
906 $hebrew = self::tsToHebrew( $ts );
907 }
908 $num = $hebrew[3];
909 break;
910 case 'L':
911 if ( !$unix ) {
912 $unix = wfTimestamp( TS_UNIX, $ts );
913 }
914 $num = gmdate( 'L', $unix );
915 break;
916 case 'o':
917 if ( !$unix ) {
918 $unix = wfTimestamp( TS_UNIX, $ts );
919 }
920 $num = date( 'o', $unix );
921 break;
922 case 'Y':
923 $num = substr( $ts, 0, 4 );
924 break;
925 case 'xiY':
926 if ( !$iranian ) {
927 $iranian = self::tsToIranian( $ts );
928 }
929 $num = $iranian[0];
930 break;
931 case 'xmY':
932 if ( !$hijri ) {
933 $hijri = self::tsToHijri( $ts );
934 }
935 $num = $hijri[0];
936 break;
937 case 'xjY':
938 if ( !$hebrew ) {
939 $hebrew = self::tsToHebrew( $ts );
940 }
941 $num = $hebrew[0];
942 break;
943 case 'xkY':
944 if ( !$thai ) {
945 $thai = self::tsToYear( $ts, 'thai' );
946 }
947 $num = $thai[0];
948 break;
949 case 'xoY':
950 if ( !$minguo ) {
951 $minguo = self::tsToYear( $ts, 'minguo' );
952 }
953 $num = $minguo[0];
954 break;
955 case 'xtY':
956 if ( !$tenno ) {
957 $tenno = self::tsToYear( $ts, 'tenno' );
958 }
959 $num = $tenno[0];
960 break;
961 case 'y':
962 $num = substr( $ts, 2, 2 );
963 break;
964 case 'a':
965 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
966 break;
967 case 'A':
968 $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
969 break;
970 case 'g':
971 $h = substr( $ts, 8, 2 );
972 $num = $h % 12 ? $h % 12 : 12;
973 break;
974 case 'G':
975 $num = intval( substr( $ts, 8, 2 ) );
976 break;
977 case 'h':
978 $h = substr( $ts, 8, 2 );
979 $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
980 break;
981 case 'H':
982 $num = substr( $ts, 8, 2 );
983 break;
984 case 'i':
985 $num = substr( $ts, 10, 2 );
986 break;
987 case 's':
988 $num = substr( $ts, 12, 2 );
989 break;
990 case 'c':
991 if ( !$unix ) {
992 $unix = wfTimestamp( TS_UNIX, $ts );
993 }
994 $s .= gmdate( 'c', $unix );
995 break;
996 case 'r':
997 if ( !$unix ) {
998 $unix = wfTimestamp( TS_UNIX, $ts );
999 }
1000 $s .= gmdate( 'r', $unix );
1001 break;
1002 case 'U':
1003 if ( !$unix ) {
1004 $unix = wfTimestamp( TS_UNIX, $ts );
1005 }
1006 $num = $unix;
1007 break;
1008 case '\\':
1009 # Backslash escaping
1010 if ( $p < strlen( $format ) - 1 ) {
1011 $s .= $format[++$p];
1012 } else {
1013 $s .= '\\';
1014 }
1015 break;
1016 case '"':
1017 # Quoted literal
1018 if ( $p < strlen( $format ) - 1 ) {
1019 $endQuote = strpos( $format, '"', $p + 1 );
1020 if ( $endQuote === false ) {
1021 # No terminating quote, assume literal "
1022 $s .= '"';
1023 } else {
1024 $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
1025 $p = $endQuote;
1026 }
1027 } else {
1028 # Quote at end of string, assume literal "
1029 $s .= '"';
1030 }
1031 break;
1032 default:
1033 $s .= $format[$p];
1034 }
1035 if ( $num !== false ) {
1036 if ( $rawToggle || $raw ) {
1037 $s .= $num;
1038 $raw = false;
1039 } elseif ( $roman ) {
1040 $s .= self::romanNumeral( $num );
1041 $roman = false;
1042 } elseif ( $hebrewNum ) {
1043 $s .= self::hebrewNumeral( $num );
1044 $hebrewNum = false;
1045 } else {
1046 $s .= $this->formatNum( $num, true );
1047 }
1048 }
1049 }
1050 return $s;
1051 }
1052
1053 private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
1054 private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
1055 /**
1056 * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
1057 * Gregorian dates to Iranian dates. Originally written in C, it
1058 * is released under the terms of GNU Lesser General Public
1059 * License. Conversion to PHP was performed by Niklas Laxström.
1060 *
1061 * Link: http://www.farsiweb.info/jalali/jalali.c
1062 */
1063 private static function tsToIranian( $ts ) {
1064 $gy = substr( $ts, 0, 4 ) -1600;
1065 $gm = substr( $ts, 4, 2 ) -1;
1066 $gd = substr( $ts, 6, 2 ) -1;
1067
1068 # Days passed from the beginning (including leap years)
1069 $gDayNo = 365 * $gy
1070 + floor( ( $gy + 3 ) / 4 )
1071 - floor( ( $gy + 99 ) / 100 )
1072 + floor( ( $gy + 399 ) / 400 );
1073
1074
1075 // Add days of the past months of this year
1076 for ( $i = 0; $i < $gm; $i++ ) {
1077 $gDayNo += self::$GREG_DAYS[$i];
1078 }
1079
1080 // Leap years
1081 if ( $gm > 1 && ( ( $gy % 4 === 0 && $gy % 100 !== 0 || ( $gy % 400 == 0 ) ) ) ) {
1082 $gDayNo++;
1083 }
1084
1085 // Days passed in current month
1086 $gDayNo += $gd;
1087
1088 $jDayNo = $gDayNo - 79;
1089
1090 $jNp = floor( $jDayNo / 12053 );
1091 $jDayNo %= 12053;
1092
1093 $jy = 979 + 33 * $jNp + 4 * floor( $jDayNo / 1461 );
1094 $jDayNo %= 1461;
1095
1096 if ( $jDayNo >= 366 ) {
1097 $jy += floor( ( $jDayNo - 1 ) / 365 );
1098 $jDayNo = floor( ( $jDayNo - 1 ) % 365 );
1099 }
1100
1101 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
1102 $jDayNo -= self::$IRANIAN_DAYS[$i];
1103 }
1104
1105 $jm = $i + 1;
1106 $jd = $jDayNo + 1;
1107
1108 return array( $jy, $jm, $jd );
1109 }
1110
1111 /**
1112 * Converting Gregorian dates to Hijri dates.
1113 *
1114 * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
1115 *
1116 * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
1117 */
1118 private static function tsToHijri( $ts ) {
1119 $year = substr( $ts, 0, 4 );
1120 $month = substr( $ts, 4, 2 );
1121 $day = substr( $ts, 6, 2 );
1122
1123 $zyr = $year;
1124 $zd = $day;
1125 $zm = $month;
1126 $zy = $zyr;
1127
1128 if (
1129 ( $zy > 1582 ) || ( ( $zy == 1582 ) && ( $zm > 10 ) ) ||
1130 ( ( $zy == 1582 ) && ( $zm == 10 ) && ( $zd > 14 ) )
1131 )
1132 {
1133 $zjd = (int)( ( 1461 * ( $zy + 4800 + (int)( ( $zm - 14 ) / 12 ) ) ) / 4 ) +
1134 (int)( ( 367 * ( $zm - 2 - 12 * ( (int)( ( $zm - 14 ) / 12 ) ) ) ) / 12 ) -
1135 (int)( ( 3 * (int)( ( ( $zy + 4900 + (int)( ( $zm - 14 ) / 12 ) ) / 100 ) ) ) / 4 ) +
1136 $zd - 32075;
1137 } else {
1138 $zjd = 367 * $zy - (int)( ( 7 * ( $zy + 5001 + (int)( ( $zm - 9 ) / 7 ) ) ) / 4 ) +
1139 (int)( ( 275 * $zm ) / 9 ) + $zd + 1729777;
1140 }
1141
1142 $zl = $zjd -1948440 + 10632;
1143 $zn = (int)( ( $zl - 1 ) / 10631 );
1144 $zl = $zl - 10631 * $zn + 354;
1145 $zj = ( (int)( ( 10985 - $zl ) / 5316 ) ) * ( (int)( ( 50 * $zl ) / 17719 ) ) + ( (int)( $zl / 5670 ) ) * ( (int)( ( 43 * $zl ) / 15238 ) );
1146 $zl = $zl - ( (int)( ( 30 - $zj ) / 15 ) ) * ( (int)( ( 17719 * $zj ) / 50 ) ) - ( (int)( $zj / 16 ) ) * ( (int)( ( 15238 * $zj ) / 43 ) ) + 29;
1147 $zm = (int)( ( 24 * $zl ) / 709 );
1148 $zd = $zl - (int)( ( 709 * $zm ) / 24 );
1149 $zy = 30 * $zn + $zj - 30;
1150
1151 return array( $zy, $zm, $zd );
1152 }
1153
1154 /**
1155 * Converting Gregorian dates to Hebrew dates.
1156 *
1157 * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1158 * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1159 * to translate the relevant functions into PHP and release them under
1160 * GNU GPL.
1161 *
1162 * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1163 * and Adar II is 14. In a non-leap year, Adar is 6.
1164 */
1165 private static function tsToHebrew( $ts ) {
1166 # Parse date
1167 $year = substr( $ts, 0, 4 );
1168 $month = substr( $ts, 4, 2 );
1169 $day = substr( $ts, 6, 2 );
1170
1171 # Calculate Hebrew year
1172 $hebrewYear = $year + 3760;
1173
1174 # Month number when September = 1, August = 12
1175 $month += 4;
1176 if ( $month > 12 ) {
1177 # Next year
1178 $month -= 12;
1179 $year++;
1180 $hebrewYear++;
1181 }
1182
1183 # Calculate day of year from 1 September
1184 $dayOfYear = $day;
1185 for ( $i = 1; $i < $month; $i++ ) {
1186 if ( $i == 6 ) {
1187 # February
1188 $dayOfYear += 28;
1189 # Check if the year is leap
1190 if ( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1191 $dayOfYear++;
1192 }
1193 } elseif ( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1194 $dayOfYear += 30;
1195 } else {
1196 $dayOfYear += 31;
1197 }
1198 }
1199
1200 # Calculate the start of the Hebrew year
1201 $start = self::hebrewYearStart( $hebrewYear );
1202
1203 # Calculate next year's start
1204 if ( $dayOfYear <= $start ) {
1205 # Day is before the start of the year - it is the previous year
1206 # Next year's start
1207 $nextStart = $start;
1208 # Previous year
1209 $year--;
1210 $hebrewYear--;
1211 # Add days since previous year's 1 September
1212 $dayOfYear += 365;
1213 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1214 # Leap year
1215 $dayOfYear++;
1216 }
1217 # Start of the new (previous) year
1218 $start = self::hebrewYearStart( $hebrewYear );
1219 } else {
1220 # Next year's start
1221 $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1222 }
1223
1224 # Calculate Hebrew day of year
1225 $hebrewDayOfYear = $dayOfYear - $start;
1226
1227 # Difference between year's days
1228 $diff = $nextStart - $start;
1229 # Add 12 (or 13 for leap years) days to ignore the difference between
1230 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1231 # difference is only about the year type
1232 if ( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1233 $diff += 13;
1234 } else {
1235 $diff += 12;
1236 }
1237
1238 # Check the year pattern, and is leap year
1239 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1240 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1241 # and non-leap years
1242 $yearPattern = $diff % 30;
1243 # Check if leap year
1244 $isLeap = $diff >= 30;
1245
1246 # Calculate day in the month from number of day in the Hebrew year
1247 # Don't check Adar - if the day is not in Adar, we will stop before;
1248 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1249 $hebrewDay = $hebrewDayOfYear;
1250 $hebrewMonth = 1;
1251 $days = 0;
1252 while ( $hebrewMonth <= 12 ) {
1253 # Calculate days in this month
1254 if ( $isLeap && $hebrewMonth == 6 ) {
1255 # Adar in a leap year
1256 if ( $isLeap ) {
1257 # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1258 $days = 30;
1259 if ( $hebrewDay <= $days ) {
1260 # Day in Adar I
1261 $hebrewMonth = 13;
1262 } else {
1263 # Subtract the days of Adar I
1264 $hebrewDay -= $days;
1265 # Try Adar II
1266 $days = 29;
1267 if ( $hebrewDay <= $days ) {
1268 # Day in Adar II
1269 $hebrewMonth = 14;
1270 }
1271 }
1272 }
1273 } elseif ( $hebrewMonth == 2 && $yearPattern == 2 ) {
1274 # Cheshvan in a complete year (otherwise as the rule below)
1275 $days = 30;
1276 } elseif ( $hebrewMonth == 3 && $yearPattern == 0 ) {
1277 # Kislev in an incomplete year (otherwise as the rule below)
1278 $days = 29;
1279 } else {
1280 # Odd months have 30 days, even have 29
1281 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1282 }
1283 if ( $hebrewDay <= $days ) {
1284 # In the current month
1285 break;
1286 } else {
1287 # Subtract the days of the current month
1288 $hebrewDay -= $days;
1289 # Try in the next month
1290 $hebrewMonth++;
1291 }
1292 }
1293
1294 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1295 }
1296
1297 /**
1298 * This calculates the Hebrew year start, as days since 1 September.
1299 * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1300 * Used for Hebrew date.
1301 */
1302 private static function hebrewYearStart( $year ) {
1303 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1304 $b = intval( ( $year - 1 ) % 4 );
1305 $m = 32.044093161144 + 1.5542417966212 * $a + $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1306 if ( $m < 0 ) {
1307 $m--;
1308 }
1309 $Mar = intval( $m );
1310 if ( $m < 0 ) {
1311 $m++;
1312 }
1313 $m -= $Mar;
1314
1315 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7 );
1316 if ( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1317 $Mar++;
1318 } else if ( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1319 $Mar += 2;
1320 } else if ( $c == 2 || $c == 4 || $c == 6 ) {
1321 $Mar++;
1322 }
1323
1324 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1325 return $Mar;
1326 }
1327
1328 /**
1329 * Algorithm to convert Gregorian dates to Thai solar dates,
1330 * Minguo dates or Minguo dates.
1331 *
1332 * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1333 * http://en.wikipedia.org/wiki/Minguo_calendar
1334 * http://en.wikipedia.org/wiki/Japanese_era_name
1335 *
1336 * @param $ts String: 14-character timestamp
1337 * @param $cName String: calender name
1338 * @return Array: converted year, month, day
1339 */
1340 private static function tsToYear( $ts, $cName ) {
1341 $gy = substr( $ts, 0, 4 );
1342 $gm = substr( $ts, 4, 2 );
1343 $gd = substr( $ts, 6, 2 );
1344
1345 if ( !strcmp( $cName, 'thai' ) ) {
1346 # Thai solar dates
1347 # Add 543 years to the Gregorian calendar
1348 # Months and days are identical
1349 $gy_offset = $gy + 543;
1350 } else if ( ( !strcmp( $cName, 'minguo' ) ) || !strcmp( $cName, 'juche' ) ) {
1351 # Minguo dates
1352 # Deduct 1911 years from the Gregorian calendar
1353 # Months and days are identical
1354 $gy_offset = $gy - 1911;
1355 } else if ( !strcmp( $cName, 'tenno' ) ) {
1356 # Nengō dates up to Meiji period
1357 # Deduct years from the Gregorian calendar
1358 # depending on the nengo periods
1359 # Months and days are identical
1360 if ( ( $gy < 1912 ) || ( ( $gy == 1912 ) && ( $gm < 7 ) ) || ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd < 31 ) ) ) {
1361 # Meiji period
1362 $gy_gannen = $gy - 1868 + 1;
1363 $gy_offset = $gy_gannen;
1364 if ( $gy_gannen == 1 ) {
1365 $gy_offset = '元';
1366 }
1367 $gy_offset = '明治' . $gy_offset;
1368 } else if (
1369 ( ( $gy == 1912 ) && ( $gm == 7 ) && ( $gd == 31 ) ) ||
1370 ( ( $gy == 1912 ) && ( $gm >= 8 ) ) ||
1371 ( ( $gy > 1912 ) && ( $gy < 1926 ) ) ||
1372 ( ( $gy == 1926 ) && ( $gm < 12 ) ) ||
1373 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd < 26 ) )
1374 )
1375 {
1376 # Taishō period
1377 $gy_gannen = $gy - 1912 + 1;
1378 $gy_offset = $gy_gannen;
1379 if ( $gy_gannen == 1 ) {
1380 $gy_offset = '元';
1381 }
1382 $gy_offset = '大正' . $gy_offset;
1383 } else if (
1384 ( ( $gy == 1926 ) && ( $gm == 12 ) && ( $gd >= 26 ) ) ||
1385 ( ( $gy > 1926 ) && ( $gy < 1989 ) ) ||
1386 ( ( $gy == 1989 ) && ( $gm == 1 ) && ( $gd < 8 ) )
1387 )
1388 {
1389 # Shōwa period
1390 $gy_gannen = $gy - 1926 + 1;
1391 $gy_offset = $gy_gannen;
1392 if ( $gy_gannen == 1 ) {
1393 $gy_offset = '元';
1394 }
1395 $gy_offset = '昭和' . $gy_offset;
1396 } else {
1397 # Heisei period
1398 $gy_gannen = $gy - 1989 + 1;
1399 $gy_offset = $gy_gannen;
1400 if ( $gy_gannen == 1 ) {
1401 $gy_offset = '元';
1402 }
1403 $gy_offset = '平成' . $gy_offset;
1404 }
1405 } else {
1406 $gy_offset = $gy;
1407 }
1408
1409 return array( $gy_offset, $gm, $gd );
1410 }
1411
1412 /**
1413 * Roman number formatting up to 3000
1414 */
1415 static function romanNumeral( $num ) {
1416 static $table = array(
1417 array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1418 array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1419 array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1420 array( '', 'M', 'MM', 'MMM' )
1421 );
1422
1423 $num = intval( $num );
1424 if ( $num > 3000 || $num <= 0 ) {
1425 return $num;
1426 }
1427
1428 $s = '';
1429 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1430 if ( $num >= $pow10 ) {
1431 $s .= $table[$i][floor( $num / $pow10 )];
1432 }
1433 $num = $num % $pow10;
1434 }
1435 return $s;
1436 }
1437
1438 /**
1439 * Hebrew Gematria number formatting up to 9999
1440 */
1441 static function hebrewNumeral( $num ) {
1442 static $table = array(
1443 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1444 array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1445 array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1446 array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1447 );
1448
1449 $num = intval( $num );
1450 if ( $num > 9999 || $num <= 0 ) {
1451 return $num;
1452 }
1453
1454 $s = '';
1455 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1456 if ( $num >= $pow10 ) {
1457 if ( $num == 15 || $num == 16 ) {
1458 $s .= $table[0][9] . $table[0][$num - 9];
1459 $num = 0;
1460 } else {
1461 $s .= $table[$i][intval( ( $num / $pow10 ) )];
1462 if ( $pow10 == 1000 ) {
1463 $s .= "'";
1464 }
1465 }
1466 }
1467 $num = $num % $pow10;
1468 }
1469 if ( strlen( $s ) == 2 ) {
1470 $str = $s . "'";
1471 } else {
1472 $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1473 $str .= substr( $s, strlen( $s ) - 2, 2 );
1474 }
1475 $start = substr( $str, 0, strlen( $str ) - 2 );
1476 $end = substr( $str, strlen( $str ) - 2 );
1477 switch( $end ) {
1478 case 'כ':
1479 $str = $start . 'ך';
1480 break;
1481 case 'מ':
1482 $str = $start . 'ם';
1483 break;
1484 case 'נ':
1485 $str = $start . 'ן';
1486 break;
1487 case 'פ':
1488 $str = $start . 'ף';
1489 break;
1490 case 'צ':
1491 $str = $start . 'ץ';
1492 break;
1493 }
1494 return $str;
1495 }
1496
1497 /**
1498 * This is meant to be used by time(), date(), and timeanddate() to get
1499 * the date preference they're supposed to use, it should be used in
1500 * all children.
1501 *
1502 *<code>
1503 * function timeanddate([...], $format = true) {
1504 * $datePreference = $this->dateFormat($format);
1505 * [...]
1506 * }
1507 *</code>
1508 *
1509 * @param $usePrefs Mixed: if true, the user's preference is used
1510 * if false, the site/language default is used
1511 * if int/string, assumed to be a format.
1512 * @return string
1513 */
1514 function dateFormat( $usePrefs = true ) {
1515 global $wgUser;
1516
1517 if ( is_bool( $usePrefs ) ) {
1518 if ( $usePrefs ) {
1519 $datePreference = $wgUser->getDatePreference();
1520 } else {
1521 $datePreference = (string)User::getDefaultOption( 'date' );
1522 }
1523 } else {
1524 $datePreference = (string)$usePrefs;
1525 }
1526
1527 // return int
1528 if ( $datePreference == '' ) {
1529 return 'default';
1530 }
1531
1532 return $datePreference;
1533 }
1534
1535 /**
1536 * Get a format string for a given type and preference
1537 * @param $type May be date, time or both
1538 * @param $pref The format name as it appears in Messages*.php
1539 */
1540 function getDateFormatString( $type, $pref ) {
1541 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
1542 if ( $pref == 'default' ) {
1543 $pref = $this->getDefaultDateFormat();
1544 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1545 } else {
1546 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1547 if ( is_null( $df ) ) {
1548 $pref = $this->getDefaultDateFormat();
1549 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1550 }
1551 }
1552 $this->dateFormatStrings[$type][$pref] = $df;
1553 }
1554 return $this->dateFormatStrings[$type][$pref];
1555 }
1556
1557 /**
1558 * @param $ts Mixed: the time format which needs to be turned into a
1559 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1560 * @param $adj Bool: whether to adjust the time output according to the
1561 * user configured offset ($timecorrection)
1562 * @param $format Mixed: true to use user's date format preference
1563 * @param $timecorrection String: the time offset as returned by
1564 * validateTimeZone() in Special:Preferences
1565 * @return string
1566 */
1567 function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1568 $ts = wfTimestamp( TS_MW, $ts );
1569 if ( $adj ) {
1570 $ts = $this->userAdjust( $ts, $timecorrection );
1571 }
1572 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1573 return $this->sprintfDate( $df, $ts );
1574 }
1575
1576 /**
1577 * @param $ts Mixed: the time format which needs to be turned into a
1578 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1579 * @param $adj Bool: whether to adjust the time output according to the
1580 * user configured offset ($timecorrection)
1581 * @param $format Mixed: true to use user's date format preference
1582 * @param $timecorrection String: the time offset as returned by
1583 * validateTimeZone() in Special:Preferences
1584 * @return string
1585 */
1586 function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1587 $ts = wfTimestamp( TS_MW, $ts );
1588 if ( $adj ) {
1589 $ts = $this->userAdjust( $ts, $timecorrection );
1590 }
1591 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1592 return $this->sprintfDate( $df, $ts );
1593 }
1594
1595 /**
1596 * @param $ts Mixed: the time format which needs to be turned into a
1597 * date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1598 * @param $adj Bool: whether to adjust the time output according to the
1599 * user configured offset ($timecorrection)
1600 * @param $format Mixed: what format to return, if it's false output the
1601 * default one (default true)
1602 * @param $timecorrection String: the time offset as returned by
1603 * validateTimeZone() in Special:Preferences
1604 * @return string
1605 */
1606 function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false ) {
1607 $ts = wfTimestamp( TS_MW, $ts );
1608 if ( $adj ) {
1609 $ts = $this->userAdjust( $ts, $timecorrection );
1610 }
1611 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1612 return $this->sprintfDate( $df, $ts );
1613 }
1614
1615 function getMessage( $key ) {
1616 // Don't change getPreferredVariant() to getCode() / mCode, because:
1617
1618 // 1. Some language like Chinese has multiple variant languages. Only
1619 // getPreferredVariant() (in LanguageConverter) could return a
1620 // sub-language which would be more suitable for the user.
1621 // 2. To languages without multiple variants, getPreferredVariant()
1622 // (in FakeConverter) functions exactly same as getCode() / mCode,
1623 // it won't break anything.
1624
1625 // The same below.
1626 return self::$dataCache->getSubitem( $this->getPreferredVariant(), 'messages', $key );
1627 }
1628
1629 function getAllMessages() {
1630 return self::$dataCache->getItem( $this->getPreferredVariant(), 'messages' );
1631 }
1632
1633 function iconv( $in, $out, $string ) {
1634 # This is a wrapper for iconv in all languages except esperanto,
1635 # which does some nasty x-conversions beforehand
1636
1637 # Even with //IGNORE iconv can whine about illegal characters in
1638 # *input* string. We just ignore those too.
1639 # REF: http://bugs.php.net/bug.php?id=37166
1640 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1641 wfSuppressWarnings();
1642 $text = iconv( $in, $out . '//IGNORE', $string );
1643 wfRestoreWarnings();
1644 return $text;
1645 }
1646
1647 // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1648 function ucwordbreaksCallbackAscii( $matches ) {
1649 return $this->ucfirst( $matches[1] );
1650 }
1651
1652 function ucwordbreaksCallbackMB( $matches ) {
1653 return mb_strtoupper( $matches[0] );
1654 }
1655
1656 function ucCallback( $matches ) {
1657 list( $wikiUpperChars ) = self::getCaseMaps();
1658 return strtr( $matches[1], $wikiUpperChars );
1659 }
1660
1661 function lcCallback( $matches ) {
1662 list( , $wikiLowerChars ) = self::getCaseMaps();
1663 return strtr( $matches[1], $wikiLowerChars );
1664 }
1665
1666 function ucwordsCallbackMB( $matches ) {
1667 return mb_strtoupper( $matches[0] );
1668 }
1669
1670 function ucwordsCallbackWiki( $matches ) {
1671 list( $wikiUpperChars ) = self::getCaseMaps();
1672 return strtr( $matches[0], $wikiUpperChars );
1673 }
1674
1675 /**
1676 * Make a string's first character uppercase
1677 */
1678 function ucfirst( $str ) {
1679 $o = ord( $str );
1680 if ( $o < 96 ) { // if already uppercase...
1681 return $str;
1682 } elseif ( $o < 128 ) {
1683 return ucfirst( $str ); // use PHP's ucfirst()
1684 } else {
1685 // fall back to more complex logic in case of multibyte strings
1686 return $this->uc( $str, true );
1687 }
1688 }
1689
1690 /**
1691 * Convert a string to uppercase
1692 */
1693 function uc( $str, $first = false ) {
1694 if ( function_exists( 'mb_strtoupper' ) ) {
1695 if ( $first ) {
1696 if ( $this->isMultibyte( $str ) ) {
1697 return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1698 } else {
1699 return ucfirst( $str );
1700 }
1701 } else {
1702 return $this->isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1703 }
1704 } else {
1705 if ( $this->isMultibyte( $str ) ) {
1706 $x = $first ? '^' : '';
1707 return preg_replace_callback(
1708 "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1709 array( $this, 'ucCallback' ),
1710 $str
1711 );
1712 } else {
1713 return $first ? ucfirst( $str ) : strtoupper( $str );
1714 }
1715 }
1716 }
1717
1718 function lcfirst( $str ) {
1719 $o = ord( $str );
1720 if ( !$o ) {
1721 return strval( $str );
1722 } elseif ( $o >= 128 ) {
1723 return $this->lc( $str, true );
1724 } elseif ( $o > 96 ) {
1725 return $str;
1726 } else {
1727 $str[0] = strtolower( $str[0] );
1728 return $str;
1729 }
1730 }
1731
1732 function lc( $str, $first = false ) {
1733 if ( function_exists( 'mb_strtolower' ) ) {
1734 if ( $first ) {
1735 if ( $this->isMultibyte( $str ) ) {
1736 return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1737 } else {
1738 return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1739 }
1740 } else {
1741 return $this->isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1742 }
1743 } else {
1744 if ( $this->isMultibyte( $str ) ) {
1745 $x = $first ? '^' : '';
1746 return preg_replace_callback(
1747 "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1748 array( $this, 'lcCallback' ),
1749 $str
1750 );
1751 } else {
1752 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1753 }
1754 }
1755 }
1756
1757 function isMultibyte( $str ) {
1758 return (bool)preg_match( '/[\x80-\xff]/', $str );
1759 }
1760
1761 function ucwords( $str ) {
1762 if ( $this->isMultibyte( $str ) ) {
1763 $str = $this->lc( $str );
1764
1765 // regexp to find first letter in each word (i.e. after each space)
1766 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1767
1768 // function to use to capitalize a single char
1769 if ( function_exists( 'mb_strtoupper' ) ) {
1770 return preg_replace_callback(
1771 $replaceRegexp,
1772 array( $this, 'ucwordsCallbackMB' ),
1773 $str
1774 );
1775 } else {
1776 return preg_replace_callback(
1777 $replaceRegexp,
1778 array( $this, 'ucwordsCallbackWiki' ),
1779 $str
1780 );
1781 }
1782 } else {
1783 return ucwords( strtolower( $str ) );
1784 }
1785 }
1786
1787 # capitalize words at word breaks
1788 function ucwordbreaks( $str ) {
1789 if ( $this->isMultibyte( $str ) ) {
1790 $str = $this->lc( $str );
1791
1792 // since \b doesn't work for UTF-8, we explicitely define word break chars
1793 $breaks = "[ \-\(\)\}\{\.,\?!]";
1794
1795 // find first letter after word break
1796 $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1797
1798 if ( function_exists( 'mb_strtoupper' ) ) {
1799 return preg_replace_callback(
1800 $replaceRegexp,
1801 array( $this, 'ucwordbreaksCallbackMB' ),
1802 $str
1803 );
1804 } else {
1805 return preg_replace_callback(
1806 $replaceRegexp,
1807 array( $this, 'ucwordsCallbackWiki' ),
1808 $str
1809 );
1810 }
1811 } else {
1812 return preg_replace_callback(
1813 '/\b([\w\x80-\xff]+)\b/',
1814 array( $this, 'ucwordbreaksCallbackAscii' ),
1815 $str
1816 );
1817 }
1818 }
1819
1820 /**
1821 * Return a case-folded representation of $s
1822 *
1823 * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1824 * and $s2 are the same except for the case of their characters. It is not
1825 * necessary for the value returned to make sense when displayed.
1826 *
1827 * Do *not* perform any other normalisation in this function. If a caller
1828 * uses this function when it should be using a more general normalisation
1829 * function, then fix the caller.
1830 */
1831 function caseFold( $s ) {
1832 return $this->uc( $s );
1833 }
1834
1835 function checkTitleEncoding( $s ) {
1836 if ( is_array( $s ) ) {
1837 wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1838 }
1839 # Check for non-UTF-8 URLs
1840 $ishigh = preg_match( '/[\x80-\xff]/', $s );
1841 if ( !$ishigh ) {
1842 return $s;
1843 }
1844
1845 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1846 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1847 if ( $isutf8 ) {
1848 return $s;
1849 }
1850
1851 return $this->iconv( $this->fallback8bitEncoding(), 'utf-8', $s );
1852 }
1853
1854 function fallback8bitEncoding() {
1855 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
1856 }
1857
1858 /**
1859 * Most writing systems use whitespace to break up words.
1860 * Some languages such as Chinese don't conventionally do this,
1861 * which requires special handling when breaking up words for
1862 * searching etc.
1863 */
1864 function hasWordBreaks() {
1865 return true;
1866 }
1867
1868 /**
1869 * Some languages such as Chinese require word segmentation,
1870 * Specify such segmentation when overridden in derived class.
1871 *
1872 * @param $string String
1873 * @return String
1874 */
1875 function segmentByWord( $string ) {
1876 return $string;
1877 }
1878
1879 /**
1880 * Some languages have special punctuation need to be normalized.
1881 * Make such changes here.
1882 *
1883 * @param $string String
1884 * @return String
1885 */
1886 function normalizeForSearch( $string ) {
1887 return self::convertDoubleWidth( $string );
1888 }
1889
1890 /**
1891 * convert double-width roman characters to single-width.
1892 * range: ff00-ff5f ~= 0020-007f
1893 */
1894 protected static function convertDoubleWidth( $string ) {
1895 static $full = null;
1896 static $half = null;
1897
1898 if ( $full === null ) {
1899 $fullWidth = "０１２３４５６７８９ＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚ";
1900 $halfWidth = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz";
1901 $full = str_split( $fullWidth, 3 );
1902 $half = str_split( $halfWidth );
1903 }
1904
1905 $string = str_replace( $full, $half, $string );
1906 return $string;
1907 }
1908
1909 protected static function insertSpace( $string, $pattern ) {
1910 $string = preg_replace( $pattern, " $1 ", $string );
1911 $string = preg_replace( '/ +/', ' ', $string );
1912 return $string;
1913 }
1914
1915 function convertForSearchResult( $termsArray ) {
1916 # some languages, e.g. Chinese, need to do a conversion
1917 # in order for search results to be displayed correctly
1918 return $termsArray;
1919 }
1920
1921 /**
1922 * Get the first character of a string.
1923 *
1924 * @param $s string
1925 * @return string
1926 */
1927 function firstChar( $s ) {
1928 $matches = array();
1929 preg_match(
1930 '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1931 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/',
1932 $s,
1933 $matches
1934 );
1935
1936 if ( isset( $matches[1] ) ) {
1937 if ( strlen( $matches[1] ) != 3 ) {
1938 return $matches[1];
1939 }
1940
1941 // Break down Hangul syllables to grab the first jamo
1942 $code = utf8ToCodepoint( $matches[1] );
1943 if ( $code < 0xac00 || 0xd7a4 <= $code ) {
1944 return $matches[1];
1945 } elseif ( $code < 0xb098 ) {
1946 return "\xe3\x84\xb1";
1947 } elseif ( $code < 0xb2e4 ) {
1948 return "\xe3\x84\xb4";
1949 } elseif ( $code < 0xb77c ) {
1950 return "\xe3\x84\xb7";
1951 } elseif ( $code < 0xb9c8 ) {
1952 return "\xe3\x84\xb9";
1953 } elseif ( $code < 0xbc14 ) {
1954 return "\xe3\x85\x81";
1955 } elseif ( $code < 0xc0ac ) {
1956 return "\xe3\x85\x82";
1957 } elseif ( $code < 0xc544 ) {
1958 return "\xe3\x85\x85";
1959 } elseif ( $code < 0xc790 ) {
1960 return "\xe3\x85\x87";
1961 } elseif ( $code < 0xcc28 ) {
1962 return "\xe3\x85\x88";
1963 } elseif ( $code < 0xce74 ) {
1964 return "\xe3\x85\x8a";
1965 } elseif ( $code < 0xd0c0 ) {
1966 return "\xe3\x85\x8b";
1967 } elseif ( $code < 0xd30c ) {
1968 return "\xe3\x85\x8c";
1969 } elseif ( $code < 0xd558 ) {
1970 return "\xe3\x85\x8d";
1971 } else {
1972 return "\xe3\x85\x8e";
1973 }
1974 } else {
1975 return '';
1976 }
1977 }
1978
1979 function initEncoding() {
1980 # Some languages may have an alternate char encoding option
1981 # (Esperanto X-coding, Japanese furigana conversion, etc)
1982 # If this language is used as the primary content language,
1983 # an override to the defaults can be set here on startup.
1984 }
1985
1986 function recodeForEdit( $s ) {
1987 # For some languages we'll want to explicitly specify
1988 # which characters make it into the edit box raw
1989 # or are converted in some way or another.
1990 # Note that if wgOutputEncoding is different from
1991 # wgInputEncoding, this text will be further converted
1992 # to wgOutputEncoding.
1993 global $wgEditEncoding;
1994 if ( $wgEditEncoding == '' || $wgEditEncoding == 'UTF-8' ) {
1995 return $s;
1996 } else {
1997 return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1998 }
1999 }
2000
2001 function recodeInput( $s ) {
2002 # Take the previous into account.
2003 global $wgEditEncoding;
2004 if ( $wgEditEncoding != '' ) {
2005 $enc = $wgEditEncoding;
2006 } else {
2007 $enc = 'UTF-8';
2008 }
2009 if ( $enc == 'UTF-8' ) {
2010 return $s;
2011 } else {
2012 return $this->iconv( $enc, 'UTF-8', $s );
2013 }
2014 }
2015
2016 /**
2017 * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
2018 * also cleans up certain backwards-compatible sequences, converting them
2019 * to the modern Unicode equivalent.
2020 *
2021 * This is language-specific for performance reasons only.
2022 */
2023 function normalize( $s ) {
2024 global $wgAllUnicodeFixes;
2025 $s = UtfNormal::cleanUp( $s );
2026 if ( $wgAllUnicodeFixes ) {
2027 $s = $this->transformUsingPairFile( 'normalize-ar.ser', $s );
2028 $s = $this->transformUsingPairFile( 'normalize-ml.ser', $s );
2029 }
2030
2031 return $s;
2032 }
2033
2034 /**
2035 * Transform a string using serialized data stored in the given file (which
2036 * must be in the serialized subdirectory of $IP). The file contains pairs
2037 * mapping source characters to destination characters.
2038 *
2039 * The data is cached in process memory. This will go faster if you have the
2040 * FastStringSearch extension.
2041 */
2042 function transformUsingPairFile( $file, $string ) {
2043 if ( !isset( $this->transformData[$file] ) ) {
2044 $data = wfGetPrecompiledData( $file );
2045 if ( $data === false ) {
2046 throw new MWException( __METHOD__ . ": The transformation file $file is missing" );
2047 }
2048 $this->transformData[$file] = new ReplacementArray( $data );
2049 }
2050 return $this->transformData[$file]->replace( $string );
2051 }
2052
2053 /**
2054 * For right-to-left language support
2055 *
2056 * @return bool
2057 */
2058 function isRTL() {
2059 return self::$dataCache->getItem( $this->mCode, 'rtl' );
2060 }
2061
2062 /**
2063 * Return the correct HTML 'dir' attribute value for this language.
2064 * @return String
2065 */
2066 function getDir() {
2067 return $this->isRTL() ? 'rtl' : 'ltr';
2068 }
2069
2070 /**
2071 * Return 'left' or 'right' as appropriate alignment for line-start
2072 * for this language's text direction.
2073 *
2074 * Should be equivalent to CSS3 'start' text-align value....
2075 *
2076 * @return String
2077 */
2078 function alignStart() {
2079 return $this->isRTL() ? 'right' : 'left';
2080 }
2081
2082 /**
2083 * Return 'right' or 'left' as appropriate alignment for line-end
2084 * for this language's text direction.
2085 *
2086 * Should be equivalent to CSS3 'end' text-align value....
2087 *
2088 * @return String
2089 */
2090 function alignEnd() {
2091 return $this->isRTL() ? 'left' : 'right';
2092 }
2093
2094 /**
2095 * A hidden direction mark (LRM or RLM), depending on the language direction
2096 *
2097 * @return string
2098 */
2099 function getDirMark() {
2100 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
2101 }
2102
2103 function capitalizeAllNouns() {
2104 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
2105 }
2106
2107 /**
2108 * An arrow, depending on the language direction
2109 *
2110 * @return string
2111 */
2112 function getArrow() {
2113 return $this->isRTL() ? '←' : '→';
2114 }
2115
2116 /**
2117 * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
2118 *
2119 * @return bool
2120 */
2121 function linkPrefixExtension() {
2122 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
2123 }
2124
2125 function getMagicWords() {
2126 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
2127 }
2128
2129 protected function doMagicHook() {
2130 if ( $this->mMagicHookDone ) {
2131 return;
2132 }
2133 $this->mMagicHookDone = true;
2134 wfProfileIn( 'LanguageGetMagic' );
2135 wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
2136 wfProfileOut( 'LanguageGetMagic' );
2137 }
2138
2139 # Fill a MagicWord object with data from here
2140 function getMagic( $mw ) {
2141 $this->doMagicHook();
2142
2143 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
2144 $rawEntry = $this->mMagicExtensions[$mw->mId];
2145 } else {
2146 $magicWords = $this->getMagicWords();
2147 if ( isset( $magicWords[$mw->mId] ) ) {
2148 $rawEntry = $magicWords[$mw->mId];
2149 } else {
2150 $rawEntry = false;
2151 }
2152 }
2153
2154 if ( !is_array( $rawEntry ) ) {
2155 error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
2156 } else {
2157 $mw->mCaseSensitive = $rawEntry[0];
2158 $mw->mSynonyms = array_slice( $rawEntry, 1 );
2159 }
2160 }
2161
2162 /**
2163 * Add magic words to the extension array
2164 */
2165 function addMagicWordsByLang( $newWords ) {
2166 $code = $this->getCode();
2167 $fallbackChain = array();
2168 while ( $code && !in_array( $code, $fallbackChain ) ) {
2169 $fallbackChain[] = $code;
2170 $code = self::getFallbackFor( $code );
2171 }
2172 if ( !in_array( 'en', $fallbackChain ) ) {
2173 $fallbackChain[] = 'en';
2174 }
2175 $fallbackChain = array_reverse( $fallbackChain );
2176 foreach ( $fallbackChain as $code ) {
2177 if ( isset( $newWords[$code] ) ) {
2178 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
2179 }
2180 }
2181 }
2182
2183 /**
2184 * Get special page names, as an associative array
2185 * case folded alias => real name
2186 */
2187 function getSpecialPageAliases() {
2188 // Cache aliases because it may be slow to load them
2189 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
2190 // Initialise array
2191 $this->mExtendedSpecialPageAliases =
2192 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
2193 wfRunHooks( 'LanguageGetSpecialPageAliases',
2194 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
2195 }
2196
2197 return $this->mExtendedSpecialPageAliases;
2198 }
2199
2200 /**
2201 * Italic is unsuitable for some languages
2202 *
2203 * @param $text String: the text to be emphasized.
2204 * @return string
2205 */
2206 function emphasize( $text ) {
2207 return "<em>$text</em>";
2208 }
2209
2210 /**
2211 * Normally we output all numbers in plain en_US style, that is
2212 * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2213 * point twohundredthirtyfive. However this is not sutable for all
2214 * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2215 * Icelandic just want to use commas instead of dots, and dots instead
2216 * of commas like "293.291,235".
2217 *
2218 * An example of this function being called:
2219 * <code>
2220 * wfMsg( 'message', $wgLang->formatNum( $num ) )
2221 * </code>
2222 *
2223 * See LanguageGu.php for the Gujarati implementation and
2224 * $separatorTransformTable on MessageIs.php for
2225 * the , => . and . => , implementation.
2226 *
2227 * @todo check if it's viable to use localeconv() for the decimal
2228 * separator thing.
2229 * @param $number Mixed: the string to be formatted, should be an integer
2230 * or a floating point number.
2231 * @param $nocommafy Bool: set to true for special numbers like dates
2232 * @return string
2233 */
2234 function formatNum( $number, $nocommafy = false ) {
2235 global $wgTranslateNumerals;
2236 if ( !$nocommafy ) {
2237 $number = $this->commafy( $number );
2238 $s = $this->separatorTransformTable();
2239 if ( $s ) {
2240 $number = strtr( $number, $s );
2241 }
2242 }
2243
2244 if ( $wgTranslateNumerals ) {
2245 $s = $this->digitTransformTable();
2246 if ( $s ) {
2247 $number = strtr( $number, $s );
2248 }
2249 }
2250
2251 return $number;
2252 }
2253
2254 function parseFormattedNumber( $number ) {
2255 $s = $this->digitTransformTable();
2256 if ( $s ) {
2257 $number = strtr( $number, array_flip( $s ) );
2258 }
2259
2260 $s = $this->separatorTransformTable();
2261 if ( $s ) {
2262 $number = strtr( $number, array_flip( $s ) );
2263 }
2264
2265 $number = strtr( $number, array( ',' => '' ) );
2266 return $number;
2267 }
2268
2269 /**
2270 * Adds commas to a given number
2271 *
2272 * @param $_ mixed
2273 * @return string
2274 */
2275 function commafy( $_ ) {
2276 return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
2277 }
2278
2279 function digitTransformTable() {
2280 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
2281 }
2282
2283 function separatorTransformTable() {
2284 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
2285 }
2286
2287 /**
2288 * Take a list of strings and build a locale-friendly comma-separated
2289 * list, using the local comma-separator message.
2290 * The last two strings are chained with an "and".
2291 *
2292 * @param $l Array
2293 * @return string
2294 */
2295 function listToText( $l ) {
2296 $s = '';
2297 $m = count( $l ) - 1;
2298 if ( $m == 1 ) {
2299 return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2300 } else {
2301 for ( $i = $m; $i >= 0; $i-- ) {
2302 if ( $i == $m ) {
2303 $s = $l[$i];
2304 } else if ( $i == $m - 1 ) {
2305 $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2306 } else {
2307 $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2308 }
2309 }
2310 return $s;
2311 }
2312 }
2313
2314 /**
2315 * Take a list of strings and build a locale-friendly comma-separated
2316 * list, using the local comma-separator message.
2317 * @param $list array of strings to put in a comma list
2318 * @return string
2319 */
2320 function commaList( $list ) {
2321 return implode(
2322 $list,
2323 wfMsgExt(
2324 'comma-separator',
2325 array( 'parsemag', 'escapenoentities', 'language' => $this )
2326 )
2327 );
2328 }
2329
2330 /**
2331 * Take a list of strings and build a locale-friendly semicolon-separated
2332 * list, using the local semicolon-separator message.
2333 * @param $list array of strings to put in a semicolon list
2334 * @return string
2335 */
2336 function semicolonList( $list ) {
2337 return implode(
2338 $list,
2339 wfMsgExt(
2340 'semicolon-separator',
2341 array( 'parsemag', 'escapenoentities', 'language' => $this )
2342 )
2343 );
2344 }
2345
2346 /**
2347 * Same as commaList, but separate it with the pipe instead.
2348 * @param $list array of strings to put in a pipe list
2349 * @return string
2350 */
2351 function pipeList( $list ) {
2352 return implode(
2353 $list,
2354 wfMsgExt(
2355 'pipe-separator',
2356 array( 'escapenoentities', 'language' => $this )
2357 )
2358 );
2359 }
2360
2361 /**
2362 * Truncate a string to a specified length in bytes, appending an optional
2363 * string (e.g. for ellipses)
2364 *
2365 * The database offers limited byte lengths for some columns in the database;
2366 * multi-byte character sets mean we need to ensure that only whole characters
2367 * are included, otherwise broken characters can be passed to the user
2368 *
2369 * If $length is negative, the string will be truncated from the beginning
2370 *
2371 * @param $string String to truncate
2372 * @param $length Int: maximum length (excluding ellipses)
2373 * @param $ellipsis String to append to the truncated text
2374 * @return string
2375 */
2376 function truncate( $string, $length, $ellipsis = '...' ) {
2377 # Use the localized ellipsis character
2378 if ( $ellipsis == '...' ) {
2379 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2380 }
2381 # Check if there is no need to truncate
2382 if ( $length == 0 ) {
2383 return $ellipsis;
2384 } elseif ( strlen( $string ) <= abs( $length ) ) {
2385 return $string;
2386 }
2387 $stringOriginal = $string;
2388 if ( $length > 0 ) {
2389 $string = substr( $string, 0, $length ); // xyz...
2390 $string = $this->removeBadCharLast( $string );
2391 $string = $string . $ellipsis;
2392 } else {
2393 $string = substr( $string, $length ); // ...xyz
2394 $string = $this->removeBadCharFirst( $string );
2395 $string = $ellipsis . $string;
2396 }
2397 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181)
2398 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
2399 return $string;
2400 } else {
2401 return $stringOriginal;
2402 }
2403 }
2404
2405 /**
2406 * Remove bytes that represent an incomplete Unicode character
2407 * at the end of string (e.g. bytes of the char are missing)
2408 *
2409 * @param $string String
2410 * @return string
2411 */
2412 protected function removeBadCharLast( $string ) {
2413 $char = ord( $string[strlen( $string ) - 1] );
2414 $m = array();
2415 if ( $char >= 0xc0 ) {
2416 # We got the first byte only of a multibyte char; remove it.
2417 $string = substr( $string, 0, -1 );
2418 } elseif ( $char >= 0x80 &&
2419 preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2420 '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
2421 {
2422 # We chopped in the middle of a character; remove it
2423 $string = $m[1];
2424 }
2425 return $string;
2426 }
2427
2428 /**
2429 * Remove bytes that represent an incomplete Unicode character
2430 * at the start of string (e.g. bytes of the char are missing)
2431 *
2432 * @param $string String
2433 * @return string
2434 */
2435 protected function removeBadCharFirst( $string ) {
2436 $char = ord( $string[0] );
2437 if ( $char >= 0x80 && $char < 0xc0 ) {
2438 # We chopped in the middle of a character; remove the whole thing
2439 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2440 }
2441 return $string;
2442 }
2443
2444 /*
2445 * Truncate a string of valid HTML to a specified length in bytes,
2446 * appending an optional string (e.g. for ellipses), and return valid HTML
2447 *
2448 * This is only intended for styled/linked text, such as HTML with
2449 * tags like <span> and <a>, were the tags are self-contained (valid HTML)
2450 *
2451 * Note: tries to fix broken HTML with MWTidy
2452 *
2453 * @param string $text HTML string to truncate
2454 * @param int $length (zero/positive) Maximum length (excluding ellipses)
2455 * @param string $ellipsis String to append to the truncated text
2456 * @returns string
2457 */
2458 function truncateHtml( $text, $length, $ellipsis = '...' ) {
2459 # Use the localized ellipsis character
2460 if ( $ellipsis == '...' ) {
2461 $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2462 }
2463 # Check if there is no need to truncate
2464 if ( $length <= 0 ) {
2465 return $ellipsis; // no text shown, nothing to format
2466 } elseif ( strlen( $text ) <= $length ) {
2467 return $text; // string short enough even *with* HTML
2468 }
2469 $text = MWTidy::tidy( $text ); // fix tags
2470 $displayLen = 0; // innerHTML legth so far
2471 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
2472 $tagType = 0; // 0-open, 1-close
2473 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
2474 $entityState = 0; // 0-not entity, 1-entity
2475 $tag = $ret = '';
2476 $openTags = array(); // open tag stack
2477 $textLen = strlen( $text );
2478 for ( $pos = 0; $pos < $textLen; ++$pos ) {
2479 $ch = $text[$pos];
2480 $lastCh = $pos ? $text[$pos - 1] : '';
2481 $ret .= $ch; // add to result string
2482 if ( $ch == '<' ) {
2483 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
2484 $entityState = 0; // for bad HTML
2485 $bracketState = 1; // tag started (checking for backslash)
2486 } elseif ( $ch == '>' ) {
2487 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
2488 $entityState = 0; // for bad HTML
2489 $bracketState = 0; // out of brackets
2490 } elseif ( $bracketState == 1 ) {
2491 if ( $ch == '/' ) {
2492 $tagType = 1; // close tag (e.g. "</span>")
2493 } else {
2494 $tagType = 0; // open tag (e.g. "<span>")
2495 $tag .= $ch;
2496 }
2497 $bracketState = 2; // building tag name
2498 } elseif ( $bracketState == 2 ) {
2499 if ( $ch != ' ' ) {
2500 $tag .= $ch;
2501 } else {
2502 // Name found (e.g. "<a href=..."), add on tag attributes...
2503 $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
2504 }
2505 } elseif ( $bracketState == 0 ) {
2506 if ( $entityState ) {
2507 if ( $ch == ';' ) {
2508 $entityState = 0;
2509 $displayLen++; // entity is one displayed char
2510 }
2511 } else {
2512 if ( $ch == '&' ) {
2513 $entityState = 1; // entity found, (e.g. "&#160;")
2514 } else {
2515 $displayLen++; // this char is displayed
2516 // Add on the other display text after this...
2517 $skipped = $this->truncate_skip(
2518 $ret, $text, "<>&", $pos + 1, $length - $displayLen );
2519 $displayLen += $skipped;
2520 $pos += $skipped;
2521 }
2522 }
2523 }
2524 # Consider truncation once the display length has reached the maximim.
2525 # Double-check that we're not in the middle of a bracket/entity...
2526 if ( $displayLen >= $length && $bracketState == 0 && $entityState == 0 ) {
2527 if ( !$testingEllipsis ) {
2528 $testingEllipsis = true;
2529 # Save where we are; we will truncate here unless
2530 # the ellipsis actually makes the string longer.
2531 $pOpenTags = $openTags; // save state
2532 $pRet = $ret; // save state
2533 } elseif ( $displayLen > ( $length + strlen( $ellipsis ) ) ) {
2534 # Ellipsis won't make string longer/equal, the truncation point was OK.
2535 $openTags = $pOpenTags; // reload state
2536 $ret = $this->removeBadCharLast( $pRet ); // reload state, multi-byte char fix
2537 $ret .= $ellipsis; // add ellipsis
2538 break;
2539 }
2540 }
2541 }
2542 if ( $displayLen == 0 ) {
2543 return ''; // no text shown, nothing to format
2544 }
2545 // Close the last tag if left unclosed by bad HTML
2546 $this->truncate_endBracket( $tag, $text[$textLen - 1], $tagType, $openTags );
2547 while ( count( $openTags ) > 0 ) {
2548 $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
2549 }
2550 return $ret;
2551 }
2552
2553 // truncateHtml() helper function
2554 // like strcspn() but adds the skipped chars to $ret
2555 private function truncate_skip( &$ret, $text, $search, $start, $len = -1 ) {
2556 $skipCount = 0;
2557 if ( $start < strlen( $text ) ) {
2558 $skipCount = strcspn( $text, $search, $start, $len );
2559 $ret .= substr( $text, $start, $skipCount );
2560 }
2561 return $skipCount;
2562 }
2563
2564 /*
2565 * truncateHtml() helper function
2566 * (a) push or pop $tag from $openTags as needed
2567 * (b) clear $tag value
2568 * @param String &$tag Current HTML tag name we are looking at
2569 * @param int $tagType (0-open tag, 1-close tag)
2570 * @param char $lastCh Character before the '>' that ended this tag
2571 * @param array &$openTags Open tag stack (not accounting for $tag)
2572 */
2573 private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
2574 $tag = ltrim( $tag );
2575 if ( $tag != '' ) {
2576 if ( $tagType == 0 && $lastCh != '/' ) {
2577 $openTags[] = $tag; // tag opened (didn't close itself)
2578 } else if ( $tagType == 1 ) {
2579 if ( $openTags && $tag == $openTags[count( $openTags ) - 1] ) {
2580 array_pop( $openTags ); // tag closed
2581 }
2582 }
2583 $tag = '';
2584 }
2585 }
2586
2587 /**
2588 * Grammatical transformations, needed for inflected languages
2589 * Invoked by putting {{grammar:case|word}} in a message
2590 *
2591 * @param $word string
2592 * @param $case string
2593 * @return string
2594 */
2595 function convertGrammar( $word, $case ) {
2596 global $wgGrammarForms;
2597 if ( isset( $wgGrammarForms[$this->getCode()][$case][$word] ) ) {
2598 return $wgGrammarForms[$this->getCode()][$case][$word];
2599 }
2600 return $word;
2601 }
2602
2603 /**
2604 * Provides an alternative text depending on specified gender.
2605 * Usage {{gender:username|masculine|feminine|neutral}}.
2606 * username is optional, in which case the gender of current user is used,
2607 * but only in (some) interface messages; otherwise default gender is used.
2608 * If second or third parameter are not specified, masculine is used.
2609 * These details may be overriden per language.
2610 */
2611 function gender( $gender, $forms ) {
2612 if ( !count( $forms ) ) {
2613 return '';
2614 }
2615 $forms = $this->preConvertPlural( $forms, 2 );
2616 if ( $gender === 'male' ) {
2617 return $forms[0];
2618 }
2619 if ( $gender === 'female' ) {
2620 return $forms[1];
2621 }
2622 return isset( $forms[2] ) ? $forms[2] : $forms[0];
2623 }
2624
2625 /**
2626 * Plural form transformations, needed for some languages.
2627 * For example, there are 3 form of plural in Russian and Polish,
2628 * depending on "count mod 10". See [[w:Plural]]
2629 * For English it is pretty simple.
2630 *
2631 * Invoked by putting {{plural:count|wordform1|wordform2}}
2632 * or {{plural:count|wordform1|wordform2|wordform3}}
2633 *
2634 * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2635 *
2636 * @param $count Integer: non-localized number
2637 * @param $forms Array: different plural forms
2638 * @return string Correct form of plural for $count in this language
2639 */
2640 function convertPlural( $count, $forms ) {
2641 if ( !count( $forms ) ) {
2642 return '';
2643 }
2644 $forms = $this->preConvertPlural( $forms, 2 );
2645
2646 return ( $count == 1 ) ? $forms[0] : $forms[1];
2647 }
2648
2649 /**
2650 * Checks that convertPlural was given an array and pads it to requested
2651 * amound of forms by copying the last one.
2652 *
2653 * @param $count Integer: How many forms should there be at least
2654 * @param $forms Array of forms given to convertPlural
2655 * @return array Padded array of forms or an exception if not an array
2656 */
2657 protected function preConvertPlural( /* Array */ $forms, $count ) {
2658 while ( count( $forms ) < $count ) {
2659 $forms[] = $forms[count( $forms ) - 1];
2660 }
2661 return $forms;
2662 }
2663
2664 /**
2665 * For translating of expiry times
2666 * @param $str String: the validated block time in English
2667 * @return Somehow translated block time
2668 * @see LanguageFi.php for example implementation
2669 */
2670 function translateBlockExpiry( $str ) {
2671 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2672
2673 if ( $scBlockExpiryOptions == '-' ) {
2674 return $str;
2675 }
2676
2677 foreach ( explode( ',', $scBlockExpiryOptions ) as $option ) {
2678 if ( strpos( $option, ':' ) === false ) {
2679 continue;
2680 }
2681 list( $show, $value ) = explode( ':', $option );
2682 if ( strcmp( $str, $value ) == 0 ) {
2683 return htmlspecialchars( trim( $show ) );
2684 }
2685 }
2686
2687 return $str;
2688 }
2689
2690 /**
2691 * languages like Chinese need to be segmented in order for the diff
2692 * to be of any use
2693 *
2694 * @param $text String
2695 * @return String
2696 */
2697 function segmentForDiff( $text ) {
2698 return $text;
2699 }
2700
2701 /**
2702 * and unsegment to show the result
2703 *
2704 * @param $text String
2705 * @return String
2706 */
2707 function unsegmentForDiff( $text ) {
2708 return $text;
2709 }
2710
2711 # convert text to all supported variants
2712 function autoConvertToAllVariants( $text ) {
2713 return $this->mConverter->autoConvertToAllVariants( $text );
2714 }
2715
2716 # convert text to different variants of a language.
2717 function convert( $text ) {
2718 return $this->mConverter->convert( $text );
2719 }
2720
2721 # Convert a Title object to a string in the preferred variant
2722 function convertTitle( $title ) {
2723 return $this->mConverter->convertTitle( $title );
2724 }
2725
2726 # Check if this is a language with variants
2727 function hasVariants() {
2728 return sizeof( $this->getVariants() ) > 1;
2729 }
2730
2731 # Put custom tags (e.g. -{ }-) around math to prevent conversion
2732 function armourMath( $text ) {
2733 return $this->mConverter->armourMath( $text );
2734 }
2735
2736 /**
2737 * Perform output conversion on a string, and encode for safe HTML output.
2738 * @param $text String text to be converted
2739 * @param $isTitle Bool whether this conversion is for the article title
2740 * @return string
2741 * @todo this should get integrated somewhere sane
2742 */
2743 function convertHtml( $text, $isTitle = false ) {
2744 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2745 }
2746
2747 function convertCategoryKey( $key ) {
2748 return $this->mConverter->convertCategoryKey( $key );
2749 }
2750
2751 /**
2752 * Get the list of variants supported by this language
2753 * see sample implementation in LanguageZh.php
2754 *
2755 * @return array an array of language codes
2756 */
2757 function getVariants() {
2758 return $this->mConverter->getVariants();
2759 }
2760
2761 function getPreferredVariant() {
2762 return $this->mConverter->getPreferredVariant();
2763 }
2764
2765 function getDefaultVariant() {
2766 return $this->mConverter->getDefaultVariant();
2767 }
2768
2769 function getURLVariant() {
2770 return $this->mConverter->getURLVariant();
2771 }
2772
2773 /**
2774 * If a language supports multiple variants, it is
2775 * possible that non-existing link in one variant
2776 * actually exists in another variant. this function
2777 * tries to find it. See e.g. LanguageZh.php
2778 *
2779 * @param $link String: the name of the link
2780 * @param $nt Mixed: the title object of the link
2781 * @param $ignoreOtherCond Boolean: to disable other conditions when
2782 * we need to transclude a template or update a category's link
2783 * @return null the input parameters may be modified upon return
2784 */
2785 function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
2786 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
2787 }
2788
2789 /**
2790 * If a language supports multiple variants, converts text
2791 * into an array of all possible variants of the text:
2792 * 'variant' => text in that variant
2793 *
2794 * @deprecated Use autoConvertToAllVariants()
2795 */
2796 function convertLinkToAllVariants( $text ) {
2797 return $this->mConverter->convertLinkToAllVariants( $text );
2798 }
2799
2800 /**
2801 * returns language specific options used by User::getPageRenderHash()
2802 * for example, the preferred language variant
2803 *
2804 * @return string
2805 */
2806 function getExtraHashOptions() {
2807 return $this->mConverter->getExtraHashOptions();
2808 }
2809
2810 /**
2811 * For languages that support multiple variants, the title of an
2812 * article may be displayed differently in different variants. this
2813 * function returns the apporiate title defined in the body of the article.
2814 *
2815 * @return string
2816 */
2817 function getParsedTitle() {
2818 return $this->mConverter->getParsedTitle();
2819 }
2820
2821 /**
2822 * Enclose a string with the "no conversion" tag. This is used by
2823 * various functions in the Parser
2824 *
2825 * @param $text String: text to be tagged for no conversion
2826 * @param $noParse
2827 * @return string the tagged text
2828 */
2829 function markNoConversion( $text, $noParse = false ) {
2830 return $this->mConverter->markNoConversion( $text, $noParse );
2831 }
2832
2833 /**
2834 * A regular expression to match legal word-trailing characters
2835 * which should be merged onto a link of the form [[foo]]bar.
2836 *
2837 * @return string
2838 */
2839 function linkTrail() {
2840 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
2841 }
2842
2843 function getLangObj() {
2844 return $this;
2845 }
2846
2847 /**
2848 * Get the RFC 3066 code for this language object
2849 */
2850 function getCode() {
2851 return $this->mCode;
2852 }
2853
2854 function setCode( $code ) {
2855 $this->mCode = $code;
2856 }
2857
2858 /**
2859 * Get the name of a file for a certain language code
2860 * @param $prefix string Prepend this to the filename
2861 * @param $code string Language code
2862 * @param $suffix string Append this to the filename
2863 * @return string $prefix . $mangledCode . $suffix
2864 */
2865 static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2866 // Protect against path traversal
2867 if ( !Language::isValidCode( $code ) ) {
2868 throw new MWException( "Invalid language code \"$code\"" );
2869 }
2870
2871 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2872 }
2873
2874 /**
2875 * Get the language code from a file name. Inverse of getFileName()
2876 * @param $filename string $prefix . $languageCode . $suffix
2877 * @param $prefix string Prefix before the language code
2878 * @param $suffix string Suffix after the language code
2879 * @return Language code, or false if $prefix or $suffix isn't found
2880 */
2881 static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
2882 $m = null;
2883 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
2884 preg_quote( $suffix, '/' ) . '/', $filename, $m );
2885 if ( !count( $m ) ) {
2886 return false;
2887 }
2888 return str_replace( '_', '-', strtolower( $m[1] ) );
2889 }
2890
2891 static function getMessagesFileName( $code ) {
2892 global $IP;
2893 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2894 }
2895
2896 static function getClassFileName( $code ) {
2897 global $IP;
2898 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2899 }
2900
2901 /**
2902 * Get the fallback for a given language
2903 */
2904 static function getFallbackFor( $code ) {
2905 if ( $code === 'en' ) {
2906 // Shortcut
2907 return false;
2908 } else {
2909 return self::getLocalisationCache()->getItem( $code, 'fallback' );
2910 }
2911 }
2912
2913 /**
2914 * Get all messages for a given language
2915 * WARNING: this may take a long time
2916 */
2917 static function getMessagesFor( $code ) {
2918 return self::getLocalisationCache()->getItem( $code, 'messages' );
2919 }
2920
2921 /**
2922 * Get a message for a given language
2923 */
2924 static function getMessageFor( $key, $code ) {
2925 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
2926 }
2927
2928 function fixVariableInNamespace( $talk ) {
2929 if ( strpos( $talk, '$1' ) === false ) {
2930 return $talk;
2931 }
2932
2933 global $wgMetaNamespace;
2934 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2935
2936 # Allow grammar transformations
2937 # Allowing full message-style parsing would make simple requests
2938 # such as action=raw much more expensive than they need to be.
2939 # This will hopefully cover most cases.
2940 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2941 array( &$this, 'replaceGrammarInNamespace' ), $talk );
2942 return str_replace( ' ', '_', $talk );
2943 }
2944
2945 function replaceGrammarInNamespace( $m ) {
2946 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2947 }
2948
2949 static function getCaseMaps() {
2950 static $wikiUpperChars, $wikiLowerChars;
2951 if ( isset( $wikiUpperChars ) ) {
2952 return array( $wikiUpperChars, $wikiLowerChars );
2953 }
2954
2955 wfProfileIn( __METHOD__ );
2956 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2957 if ( $arr === false ) {
2958 throw new MWException(
2959 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2960 }
2961 $wikiUpperChars = $arr['wikiUpperChars'];
2962 $wikiLowerChars = $arr['wikiLowerChars'];
2963 wfProfileOut( __METHOD__ );
2964 return array( $wikiUpperChars, $wikiLowerChars );
2965 }
2966
2967 function formatTimePeriod( $seconds ) {
2968 if ( round( $seconds * 10 ) < 100 ) {
2969 return $this->formatNum( sprintf( "%.1f", round( $seconds * 10 ) / 10 ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2970 } elseif ( round( $seconds ) < 60 ) {
2971 return $this->formatNum( round( $seconds ) ) . $this->getMessageFromDB( 'seconds-abbrev' );
2972 } elseif ( round( $seconds ) < 3600 ) {
2973 $minutes = floor( $seconds / 60 );
2974 $secondsPart = round( fmod( $seconds, 60 ) );
2975 if ( $secondsPart == 60 ) {
2976 $secondsPart = 0;
2977 $minutes++;
2978 }
2979 return $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2980 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2981 } else {
2982 $hours = floor( $seconds / 3600 );
2983 $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2984 $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2985 if ( $secondsPart == 60 ) {
2986 $secondsPart = 0;
2987 $minutes++;
2988 }
2989 if ( $minutes == 60 ) {
2990 $minutes = 0;
2991 $hours++;
2992 }
2993 return $this->formatNum( $hours ) . $this->getMessageFromDB( 'hours-abbrev' ) . ' ' .
2994 $this->formatNum( $minutes ) . $this->getMessageFromDB( 'minutes-abbrev' ) . ' ' .
2995 $this->formatNum( $secondsPart ) . $this->getMessageFromDB( 'seconds-abbrev' );
2996 }
2997 }
2998
2999 function formatBitrate( $bps ) {
3000 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
3001 if ( $bps <= 0 ) {
3002 return $this->formatNum( $bps ) . $units[0];
3003 }
3004 $unitIndex = floor( log10( $bps ) / 3 );
3005 $mantissa = $bps / pow( 1000, $unitIndex );
3006 if ( $mantissa < 10 ) {
3007 $mantissa = round( $mantissa, 1 );
3008 } else {
3009 $mantissa = round( $mantissa );
3010 }
3011 return $this->formatNum( $mantissa ) . $units[$unitIndex];
3012 }
3013
3014 /**
3015 * Format a size in bytes for output, using an appropriate
3016 * unit (B, KB, MB or GB) according to the magnitude in question
3017 *
3018 * @param $size Size to format
3019 * @return string Plain text (not HTML)
3020 */
3021 function formatSize( $size ) {
3022 // For small sizes no decimal places necessary
3023 $round = 0;
3024 if ( $size > 1024 ) {
3025 $size = $size / 1024;
3026 if ( $size > 1024 ) {
3027 $size = $size / 1024;
3028 // For MB and bigger two decimal places are smarter
3029 $round = 2;
3030 if ( $size > 1024 ) {
3031 $size = $size / 1024;
3032 $msg = 'size-gigabytes';
3033 } else {
3034 $msg = 'size-megabytes';
3035 }
3036 } else {
3037 $msg = 'size-kilobytes';
3038 }
3039 } else {
3040 $msg = 'size-bytes';
3041 }
3042 $size = round( $size, $round );
3043 $text = $this->getMessageFromDB( $msg );
3044 return str_replace( '$1', $this->formatNum( $size ), $text );
3045 }
3046
3047 /**
3048 * Get the conversion rule title, if any.
3049 */
3050 function getConvRuleTitle() {
3051 return $this->mConverter->getConvRuleTitle();
3052 }
3053 }
Wiklou, le Wiki du Wiklou