languages/Language.php

   1 <?php
   2 /**
   3  * @defgroup Language Language
   4  *
   5  * @file
   6  * @ingroup Language
   7  */
   8
   9 if( !defined( 'MEDIAWIKI' ) ) {
  10         echo "This file is part of MediaWiki, it is not a valid entry point.\n";
  11         exit( 1 );
  12 }
  13
  14 # Read language names
  15 global $wgLanguageNames;
  16 require_once( dirname(__FILE__) . '/Names.php' ) ;
  17
  18 global $wgInputEncoding, $wgOutputEncoding;
  19
  20 /**
  21  * These are always UTF-8, they exist only for backwards compatibility
  22  */
  23 $wgInputEncoding    = "UTF-8";
  24 $wgOutputEncoding       = "UTF-8";
  25
  26 if( function_exists( 'mb_strtoupper' ) ) {
  27         mb_internal_encoding('UTF-8');
  28 }
  29
  30 /**
  31  * a fake language converter
  32  *
  33  * @ingroup Language
  34  */
  35 class FakeConverter {
  36         var $mLang;
  37         function FakeConverter($langobj) {$this->mLang = $langobj;}
  38         function autoConvertToAllVariants($text) {return $text;}
  39         function convert($t, $i) {return $t;}
  40         function getVariants() { return array( $this->mLang->getCode() ); }
  41         function getPreferredVariant() { return $this->mLang->getCode(); }
  42         function getConvRuleTitle() { return false; }
  43         function findVariantLink(&$l, &$n, $ignoreOtherCond = false) {}
  44         function getExtraHashOptions() {return '';}
  45         function getParsedTitle() {return '';}
  46         function markNoConversion($text, $noParse=false) {return $text;}
  47         function convertCategoryKey( $key ) {return $key; }
  48         function convertLinkToAllVariants($text){ return array( $this->mLang->getCode() => $text); }
  49         function armourMath($text){ return $text; }
  50 }
  51
  52 /**
  53  * Internationalisation code
  54  * @ingroup Language
  55  */
  56 class Language {
  57         var $mConverter, $mVariants, $mCode, $mLoaded = false;
  58         var $mMagicExtensions = array(), $mMagicHookDone = false;
  59
  60         var $mNamespaceIds, $namespaceNames, $namespaceAliases;
  61         var $dateFormatStrings = array();
  62         var $mExtendedSpecialPageAliases;
  63
  64         /**
  65          * ReplacementArray object caches
  66          */
  67         var $transformData = array();
  68
  69         static public $dataCache;
  70         static public $mLangObjCache = array();
  71
  72         static public $mWeekdayMsgs = array(
  73                 'sunday', 'monday', 'tuesday', 'wednesday', 'thursday',
  74                 'friday', 'saturday'
  75         );
  76
  77         static public $mWeekdayAbbrevMsgs = array(
  78                 'sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat'
  79         );
  80
  81         static public $mMonthMsgs = array(
  82                 'january', 'february', 'march', 'april', 'may_long', 'june',
  83                 'july', 'august', 'september', 'october', 'november',
  84                 'december'
  85         );
  86         static public $mMonthGenMsgs = array(
  87                 'january-gen', 'february-gen', 'march-gen', 'april-gen', 'may-gen', 'june-gen',
  88                 'july-gen', 'august-gen', 'september-gen', 'october-gen', 'november-gen',
  89                 'december-gen'
  90         );
  91         static public $mMonthAbbrevMsgs = array(
  92                 'jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', 'aug',
  93                 'sep', 'oct', 'nov', 'dec'
  94         );
  95
  96         static public $mIranianCalendarMonthMsgs = array(
  97                 'iranian-calendar-m1', 'iranian-calendar-m2', 'iranian-calendar-m3',
  98                 'iranian-calendar-m4', 'iranian-calendar-m5', 'iranian-calendar-m6',
  99                 'iranian-calendar-m7', 'iranian-calendar-m8', 'iranian-calendar-m9',
 100                 'iranian-calendar-m10', 'iranian-calendar-m11', 'iranian-calendar-m12'
 101         );
 102
 103         static public $mHebrewCalendarMonthMsgs = array(
 104                 'hebrew-calendar-m1', 'hebrew-calendar-m2', 'hebrew-calendar-m3',
 105                 'hebrew-calendar-m4', 'hebrew-calendar-m5', 'hebrew-calendar-m6',
 106                 'hebrew-calendar-m7', 'hebrew-calendar-m8', 'hebrew-calendar-m9',
 107                 'hebrew-calendar-m10', 'hebrew-calendar-m11', 'hebrew-calendar-m12',
 108                 'hebrew-calendar-m6a', 'hebrew-calendar-m6b'
 109         );
 110
 111         static public $mHebrewCalendarMonthGenMsgs = array(
 112                 'hebrew-calendar-m1-gen', 'hebrew-calendar-m2-gen', 'hebrew-calendar-m3-gen',
 113                 'hebrew-calendar-m4-gen', 'hebrew-calendar-m5-gen', 'hebrew-calendar-m6-gen',
 114                 'hebrew-calendar-m7-gen', 'hebrew-calendar-m8-gen', 'hebrew-calendar-m9-gen',
 115                 'hebrew-calendar-m10-gen', 'hebrew-calendar-m11-gen', 'hebrew-calendar-m12-gen',
 116                 'hebrew-calendar-m6a-gen', 'hebrew-calendar-m6b-gen'
 117         );
 118
 119         static public $mHijriCalendarMonthMsgs = array(
 120                 'hijri-calendar-m1', 'hijri-calendar-m2', 'hijri-calendar-m3',
 121                 'hijri-calendar-m4', 'hijri-calendar-m5', 'hijri-calendar-m6',
 122                 'hijri-calendar-m7', 'hijri-calendar-m8', 'hijri-calendar-m9',
 123                 'hijri-calendar-m10', 'hijri-calendar-m11', 'hijri-calendar-m12'
 124         );
 125
 126         /**
 127          * Get a cached language object for a given language code
 128          */
 129         static function factory( $code ) {
 130                 if ( !isset( self::$mLangObjCache[$code] ) ) {
 131                         if( count( self::$mLangObjCache ) > 10 ) {
 132                                 // Don't keep a billion objects around, that's stupid.
 133                                 self::$mLangObjCache = array();
 134                         }
 135                         self::$mLangObjCache[$code] = self::newFromCode( $code );
 136                 }
 137                 return self::$mLangObjCache[$code];
 138         }
 139
 140         /**
 141          * Create a language object for a given language code
 142          */
 143         protected static function newFromCode( $code ) {
 144                 global $IP;
 145                 static $recursionLevel = 0;
 146                 if ( $code == 'en' ) {
 147                         $class = 'Language';
 148                 } else {
 149                         $class = 'Language' . str_replace( '-', '_', ucfirst( $code ) );
 150                         // Preload base classes to work around APC/PHP5 bug
 151                         if ( file_exists( "$IP/languages/classes/$class.deps.php" ) ) {
 152                                 include_once("$IP/languages/classes/$class.deps.php");
 153                         }
 154                         if ( file_exists( "$IP/languages/classes/$class.php" ) ) {
 155                                 include_once("$IP/languages/classes/$class.php");
 156                         }
 157                 }
 158
 159                 if ( $recursionLevel > 5 ) {
 160                         throw new MWException( "Language fallback loop detected when creating class $class\n" );
 161                 }
 162
 163                 if( ! class_exists( $class ) ) {
 164                         $fallback = Language::getFallbackFor( $code );
 165                         ++$recursionLevel;
 166                         $lang = Language::newFromCode( $fallback );
 167                         --$recursionLevel;
 168                         $lang->setCode( $code );
 169                 } else {
 170                         $lang = new $class;
 171                 }
 172                 return $lang;
 173         }
 174
 175         /**
 176          * Get the LocalisationCache instance
 177          */
 178         public static function getLocalisationCache() {
 179                 if ( is_null( self::$dataCache ) ) {
 180                         global $wgLocalisationCacheConf;
 181                         $class = $wgLocalisationCacheConf['class'];
 182                         self::$dataCache = new $class( $wgLocalisationCacheConf );
 183                 }
 184                 return self::$dataCache;
 185         }
 186
 187         function __construct() {
 188                 $this->mConverter = new FakeConverter($this);
 189                 // Set the code to the name of the descendant
 190                 if ( get_class( $this ) == 'Language' ) {
 191                         $this->mCode = 'en';
 192                 } else {
 193                         $this->mCode = str_replace( '_', '-', strtolower( substr( get_class( $this ), 8 ) ) );
 194                 }
 195                 self::getLocalisationCache();
 196         }
 197
 198         /**
 199          * Reduce memory usage
 200          */
 201         function __destruct() {
 202                 foreach ( $this as $name => $value ) {
 203                         unset( $this->$name );
 204                 }
 205         }
 206
 207         /**
 208          * Hook which will be called if this is the content language.
 209          * Descendants can use this to register hook functions or modify globals
 210          */
 211         function initContLang() {}
 212
 213         /**
 214          * @deprecated Use User::getDefaultOptions()
 215          * @return array
 216          */
 217         function getDefaultUserOptions() {
 218                 wfDeprecated( __METHOD__ );
 219                 return User::getDefaultOptions();
 220         }
 221
 222         function getFallbackLanguageCode() {
 223                 if ( $this->mCode === 'en' ) {
 224                         return false;
 225                 } else {
 226                         return self::$dataCache->getItem( $this->mCode, 'fallback' );
 227                 }
 228         }
 229
 230         /**
 231          * Exports $wgBookstoreListEn
 232          * @return array
 233          */
 234         function getBookstoreList() {
 235                 return self::$dataCache->getItem( $this->mCode, 'bookstoreList' );
 236         }
 237
 238         /**
 239          * @return array
 240          */
 241         function getNamespaces() {
 242                 if ( is_null( $this->namespaceNames ) ) {
 243                         global $wgExtraNamespaces, $wgMetaNamespace, $wgMetaNamespaceTalk;
 244
 245                         $this->namespaceNames = self::$dataCache->getItem( $this->mCode, 'namespaceNames' );
 246                         if ( $wgExtraNamespaces ) {
 247                                 $this->namespaceNames = $wgExtraNamespaces + $this->namespaceNames;
 248                         }
 249
 250                         $this->namespaceNames[NS_PROJECT] = $wgMetaNamespace;
 251                         if ( $wgMetaNamespaceTalk ) {
 252                                 $this->namespaceNames[NS_PROJECT_TALK] = $wgMetaNamespaceTalk;
 253                         } else {
 254                                 $talk = $this->namespaceNames[NS_PROJECT_TALK];
 255                                 $this->namespaceNames[NS_PROJECT_TALK] =
 256                                         $this->fixVariableInNamespace( $talk );
 257                         }
 258
 259                         # The above mixing may leave namespaces out of canonical order.
 260                         # Re-order by namespace ID number...
 261                         ksort( $this->namespaceNames );
 262                 }
 263                 return $this->namespaceNames;
 264         }
 265
 266         /**
 267          * A convenience function that returns the same thing as
 268          * getNamespaces() except with the array values changed to ' '
 269          * where it found '_', useful for producing output to be displayed
 270          * e.g. in <select> forms.
 271          *
 272          * @return array
 273          */
 274         function getFormattedNamespaces() {
 275                 $ns = $this->getNamespaces();
 276                 foreach($ns as $k => $v) {
 277                         $ns[$k] = strtr($v, '_', ' ');
 278                 }
 279                 return $ns;
 280         }
 281
 282         /**
 283          * Get a namespace value by key
 284          * <code>
 285          * $mw_ns = $wgContLang->getNsText( NS_MEDIAWIKI );
 286          * echo $mw_ns; // prints 'MediaWiki'
 287          * </code>
 288          *
 289          * @param $index Int: the array key of the namespace to return
 290          * @return mixed, string if the namespace value exists, otherwise false
 291          */
 292         function getNsText( $index ) {
 293                 $ns = $this->getNamespaces();
 294                 return isset( $ns[$index] ) ? $ns[$index] : false;
 295         }
 296
 297         /**
 298          * A convenience function that returns the same thing as
 299          * getNsText() except with '_' changed to ' ', useful for
 300          * producing output.
 301          *
 302          * @return array
 303          */
 304         function getFormattedNsText( $index ) {
 305                 $ns = $this->getNsText( $index );
 306                 return strtr($ns, '_', ' ');
 307         }
 308
 309         /**
 310          * Get a namespace key by value, case insensitive.
 311          * Only matches namespace names for the current language, not the
 312          * canonical ones defined in Namespace.php.
 313          *
 314          * @param $text String
 315          * @return mixed An integer if $text is a valid value otherwise false
 316          */
 317         function getLocalNsIndex( $text ) {
 318                 $lctext = $this->lc($text);
 319                 $ids = $this->getNamespaceIds();
 320                 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
 321         }
 322
 323         function getNamespaceAliases() {
 324                 if ( is_null( $this->namespaceAliases ) ) {
 325                         $aliases = self::$dataCache->getItem( $this->mCode, 'namespaceAliases' );
 326                         if ( !$aliases ) {
 327                                 $aliases = array();
 328                         } else {
 329                                 foreach ( $aliases as $name => $index ) {
 330                                         if ( $index === NS_PROJECT_TALK ) {
 331                                                 unset( $aliases[$name] );
 332                                                 $name = $this->fixVariableInNamespace( $name );
 333                                                 $aliases[$name] = $index;
 334                                         }
 335                                 }
 336                         }
 337                         $this->namespaceAliases = $aliases;
 338                 }
 339                 return $this->namespaceAliases;
 340         }
 341
 342         function getNamespaceIds() {
 343                 if ( is_null( $this->mNamespaceIds ) ) {
 344                         global $wgNamespaceAliases;
 345                         # Put namespace names and aliases into a hashtable.
 346                         # If this is too slow, then we should arrange it so that it is done
 347                         # before caching. The catch is that at pre-cache time, the above
 348                         # class-specific fixup hasn't been done.
 349                         $this->mNamespaceIds = array();
 350                         foreach ( $this->getNamespaces() as $index => $name ) {
 351                                 $this->mNamespaceIds[$this->lc($name)] = $index;
 352                         }
 353                         foreach ( $this->getNamespaceAliases() as $name => $index ) {
 354                                 $this->mNamespaceIds[$this->lc($name)] = $index;
 355                         }
 356                         if ( $wgNamespaceAliases ) {
 357                                 foreach ( $wgNamespaceAliases as $name => $index ) {
 358                                         $this->mNamespaceIds[$this->lc($name)] = $index;
 359                                 }
 360                         }
 361                 }
 362                 return $this->mNamespaceIds;
 363         }
 364
 365
 366         /**
 367          * Get a namespace key by value, case insensitive.  Canonical namespace
 368          * names override custom ones defined for the current language.
 369          *
 370          * @param $text String
 371          * @return mixed An integer if $text is a valid value otherwise false
 372          */
 373         function getNsIndex( $text ) {
 374                 $lctext = $this->lc($text);
 375                 if ( ( $ns = MWNamespace::getCanonicalIndex( $lctext ) ) !== null ) {
 376                         return $ns;
 377                 }
 378                 $ids = $this->getNamespaceIds();
 379                 return isset( $ids[$lctext] ) ? $ids[$lctext] : false;
 380         }
 381
 382         /**
 383          * short names for language variants used for language conversion links.
 384          *
 385          * @param $code String
 386          * @return string
 387          */
 388         function getVariantname( $code ) {
 389                 return $this->getMessageFromDB( "variantname-$code" );
 390         }
 391
 392         function specialPage( $name ) {
 393                 $aliases = $this->getSpecialPageAliases();
 394                 if ( isset( $aliases[$name][0] ) ) {
 395                         $name = $aliases[$name][0];
 396                 }
 397                 return $this->getNsText( NS_SPECIAL ) . ':' . $name;
 398         }
 399
 400         function getQuickbarSettings() {
 401                 return array(
 402                         $this->getMessage( 'qbsettings-none' ),
 403                         $this->getMessage( 'qbsettings-fixedleft' ),
 404                         $this->getMessage( 'qbsettings-fixedright' ),
 405                         $this->getMessage( 'qbsettings-floatingleft' ),
 406                         $this->getMessage( 'qbsettings-floatingright' )
 407                 );
 408         }
 409
 410         function getMathNames() {
 411                 return self::$dataCache->getItem( $this->mCode, 'mathNames' );
 412         }
 413
 414         function getDatePreferences() {
 415                 return self::$dataCache->getItem( $this->mCode, 'datePreferences' );
 416         }
 417
 418         function getDateFormats() {
 419                 return self::$dataCache->getItem( $this->mCode, 'dateFormats' );
 420         }
 421
 422         function getDefaultDateFormat() {
 423                 $df = self::$dataCache->getItem( $this->mCode, 'defaultDateFormat' );
 424                 if ( $df === 'dmy or mdy' ) {
 425                         global $wgAmericanDates;
 426                         return $wgAmericanDates ? 'mdy' : 'dmy';
 427                 } else {
 428                         return $df;
 429                 }
 430         }
 431
 432         function getDatePreferenceMigrationMap() {
 433                 return self::$dataCache->getItem( $this->mCode, 'datePreferenceMigrationMap' );
 434         }
 435
 436         function getImageFile( $image ) {
 437                 return self::$dataCache->getSubitem( $this->mCode, 'imageFiles', $image );
 438         }
 439
 440         function getDefaultUserOptionOverrides() {
 441                 return self::$dataCache->getItem( $this->mCode, 'defaultUserOptionOverrides' );
 442         }
 443
 444         function getExtraUserToggles() {
 445                 return self::$dataCache->getItem( $this->mCode, 'extraUserToggles' );
 446         }
 447
 448         function getUserToggle( $tog ) {
 449                 return $this->getMessageFromDB( "tog-$tog" );
 450         }
 451
 452         /**
 453          * Get language names, indexed by code.
 454          * If $customisedOnly is true, only returns codes with a messages file
 455          */
 456         public static function getLanguageNames( $customisedOnly = false ) {
 457                 global $wgLanguageNames, $wgExtraLanguageNames;
 458                 $allNames = $wgExtraLanguageNames + $wgLanguageNames;
 459                 if ( !$customisedOnly ) {
 460                         return $allNames;
 461                 }
 462
 463                 global $IP;
 464                 $names = array();
 465                 $dir = opendir( "$IP/languages/messages" );
 466                 while( false !== ( $file = readdir( $dir ) ) ) {
 467                         $code = self::getCodeFromFileName( $file, 'Messages' );
 468                         if ( $code && isset( $allNames[$code] ) ) {
 469                                 $names[$code] = $allNames[$code];
 470                         }
 471                 }
 472                 closedir( $dir );
 473                 return $names;
 474         }
 475
 476         /**
 477          * Get a message from the MediaWiki namespace.
 478          *
 479          * @param $msg String: message name
 480          * @return string
 481          */
 482         function getMessageFromDB( $msg ) {
 483                 return wfMsgExt( $msg, array( 'parsemag', 'language' => $this ) );
 484         }
 485
 486         function getLanguageName( $code ) {
 487                 $names = self::getLanguageNames();
 488                 if ( !array_key_exists( $code, $names ) ) {
 489                         return '';
 490                 }
 491                 return $names[$code];
 492         }
 493
 494         function getMonthName( $key ) {
 495                 return $this->getMessageFromDB( self::$mMonthMsgs[$key-1] );
 496         }
 497
 498         function getMonthNameGen( $key ) {
 499                 return $this->getMessageFromDB( self::$mMonthGenMsgs[$key-1] );
 500         }
 501
 502         function getMonthAbbreviation( $key ) {
 503                 return $this->getMessageFromDB( self::$mMonthAbbrevMsgs[$key-1] );
 504         }
 505
 506         function getWeekdayName( $key ) {
 507                 return $this->getMessageFromDB( self::$mWeekdayMsgs[$key-1] );
 508         }
 509
 510         function getWeekdayAbbreviation( $key ) {
 511                 return $this->getMessageFromDB( self::$mWeekdayAbbrevMsgs[$key-1] );
 512         }
 513
 514         function getIranianCalendarMonthName( $key ) {
 515                 return $this->getMessageFromDB( self::$mIranianCalendarMonthMsgs[$key-1] );
 516         }
 517
 518         function getHebrewCalendarMonthName( $key ) {
 519                 return $this->getMessageFromDB( self::$mHebrewCalendarMonthMsgs[$key-1] );
 520         }
 521
 522         function getHebrewCalendarMonthNameGen( $key ) {
 523                 return $this->getMessageFromDB( self::$mHebrewCalendarMonthGenMsgs[$key-1] );
 524         }
 525
 526         function getHijriCalendarMonthName( $key ) {
 527                 return $this->getMessageFromDB( self::$mHijriCalendarMonthMsgs[$key-1] );
 528         }
 529
 530         /**
 531          * Used by date() and time() to adjust the time output.
 532          *
 533          * @param $ts Int the time in date('YmdHis') format
 534          * @param $tz Mixed: adjust the time by this amount (default false, mean we
 535          *            get user timecorrection setting)
 536          * @return int
 537          */
 538         function userAdjust( $ts, $tz = false ) {
 539                 global $wgUser, $wgLocalTZoffset;
 540
 541                 if ( $tz === false ) {
 542                         $tz = $wgUser->getOption( 'timecorrection' );
 543                 }
 544
 545                 $data = explode( '|', $tz, 3 );
 546
 547                 if ( $data[0] == 'ZoneInfo' ) {
 548                         if ( function_exists( 'timezone_open' ) && @timezone_open( $data[2] ) !== false ) {
 549                                 $date = date_create( $ts, timezone_open( 'UTC' ) );
 550                                 date_timezone_set( $date, timezone_open( $data[2] ) );
 551                                 $date = date_format( $date, 'YmdHis' );
 552                                 return $date;
 553                         }
 554                         # Unrecognized timezone, default to 'Offset' with the stored offset.
 555                         $data[0] = 'Offset';
 556                 }
 557
 558                 $minDiff = 0;
 559                 if ( $data[0] == 'System' || $tz == '' ) {
 560                         # Global offset in minutes.
 561                         if( isset($wgLocalTZoffset) ) $minDiff = $wgLocalTZoffset;
 562                 } else if ( $data[0] == 'Offset' ) {
 563                         $minDiff = intval( $data[1] );
 564                 } else {
 565                         $data = explode( ':', $tz );
 566                         if( count( $data ) == 2 ) {
 567                                 $data[0] = intval( $data[0] );
 568                                 $data[1] = intval( $data[1] );
 569                                 $minDiff = abs( $data[0] ) * 60 + $data[1];
 570                                 if ( $data[0] < 0 ) $minDiff = -$minDiff;
 571                         } else {
 572                                 $minDiff = intval( $data[0] ) * 60;
 573                         }
 574                 }
 575
 576                 # No difference ? Return time unchanged
 577                 if ( 0 == $minDiff ) return $ts;
 578
 579                 wfSuppressWarnings(); // E_STRICT system time bitching
 580                 # Generate an adjusted date; take advantage of the fact that mktime
 581                 # will normalize out-of-range values so we don't have to split $minDiff
 582                 # into hours and minutes.
 583                 $t = mktime( (
 584                   (int)substr( $ts, 8, 2) ), # Hours
 585                   (int)substr( $ts, 10, 2 ) + $minDiff, # Minutes
 586                   (int)substr( $ts, 12, 2 ), # Seconds
 587                   (int)substr( $ts, 4, 2 ), # Month
 588                   (int)substr( $ts, 6, 2 ), # Day
 589                   (int)substr( $ts, 0, 4 ) ); #Year
 590
 591                 $date = date( 'YmdHis', $t );
 592                 wfRestoreWarnings();
 593
 594                 return $date;
 595         }
 596
 597         /**
 598          * This is a workalike of PHP's date() function, but with better
 599          * internationalisation, a reduced set of format characters, and a better
 600          * escaping format.
 601          *
 602          * Supported format characters are dDjlNwzWFmMntLoYyaAgGhHiscrU. See the
 603          * PHP manual for definitions. "o" format character is supported since
 604          * PHP 5.1.0, previous versions return literal o.
 605          * There are a number of extensions, which start with "x":
 606          *
 607          *    xn   Do not translate digits of the next numeric format character
 608          *    xN   Toggle raw digit (xn) flag, stays set until explicitly unset
 609          *    xr   Use roman numerals for the next numeric format character
 610          *    xh   Use hebrew numerals for the next numeric format character
 611          *    xx   Literal x
 612          *    xg   Genitive month name
 613          *
 614          *    xij  j (day number) in Iranian calendar
 615          *    xiF  F (month name) in Iranian calendar
 616          *    xin  n (month number) in Iranian calendar
 617          *    xiY  Y (full year) in Iranian calendar
 618          *
 619          *    xjj  j (day number) in Hebrew calendar
 620          *    xjF  F (month name) in Hebrew calendar
 621          *    xjt  t (days in month) in Hebrew calendar
 622          *    xjx  xg (genitive month name) in Hebrew calendar
 623          *    xjn  n (month number) in Hebrew calendar
 624          *    xjY  Y (full year) in Hebrew calendar
 625          *
 626          *    xmj  j (day number) in Hijri calendar
 627          *    xmF  F (month name) in Hijri calendar
 628          *    xmn  n (month number) in Hijri calendar
 629          *    xmY  Y (full year) in Hijri calendar
 630          *
 631          *    xkY  Y (full year) in Thai solar calendar. Months and days are
 632          *                       identical to the Gregorian calendar
 633          *    xoY  Y (full year) in Minguo calendar or Juche year.
 634          *                       Months and days are identical to the
 635          *                       Gregorian calendar
 636          *    xtY  Y (full year) in Japanese nengo. Months and days are
 637          *                       identical to the Gregorian calendar
 638          *
 639          * Characters enclosed in double quotes will be considered literal (with
 640          * the quotes themselves removed). Unmatched quotes will be considered
 641          * literal quotes. Example:
 642          *
 643          * "The month is" F       => The month is January
 644          * i's"                   => 20'11"
 645          *
 646          * Backslash escaping is also supported.
 647          *
 648          * Input timestamp is assumed to be pre-normalized to the desired local
 649          * time zone, if any.
 650          *
 651          * @param $format String
 652          * @param $ts String: 14-character timestamp
 653          *      YYYYMMDDHHMMSS
 654          *      01234567890123
 655          * @todo emulation of "o" format character for PHP pre 5.1.0
 656          * @todo handling of "o" format character for Iranian, Hebrew, Hijri & Thai?
 657          */
 658         function sprintfDate( $format, $ts ) {
 659                 $s = '';
 660                 $raw = false;
 661                 $roman = false;
 662                 $hebrewNum = false;
 663                 $unix = false;
 664                 $rawToggle = false;
 665                 $iranian = false;
 666                 $hebrew = false;
 667                 $hijri = false;
 668                 $thai = false;
 669                 $minguo = false;
 670                 $tenno = false;
 671                 for ( $p = 0; $p < strlen( $format ); $p++ ) {
 672                         $num = false;
 673                         $code = $format[$p];
 674                         if ( $code == 'x' && $p < strlen( $format ) - 1 ) {
 675                                 $code .= $format[++$p];
 676                         }
 677
 678                         if ( ( $code === 'xi' || $code == 'xj' || $code == 'xk' || $code == 'xm' || $code == 'xo' || $code == 'xt' ) && $p < strlen( $format ) - 1 ) {
 679                                 $code .= $format[++$p];
 680                         }
 681
 682                         switch ( $code ) {
 683                                 case 'xx':
 684                                         $s .= 'x';
 685                                         break;
 686                                 case 'xn':
 687                                         $raw = true;
 688                                         break;
 689                                 case 'xN':
 690                                         $rawToggle = !$rawToggle;
 691                                         break;
 692                                 case 'xr':
 693                                         $roman = true;
 694                                         break;
 695                                 case 'xh':
 696                                         $hebrewNum = true;
 697                                         break;
 698                                 case 'xg':
 699                                         $s .= $this->getMonthNameGen( substr( $ts, 4, 2 ) );
 700                                         break;
 701                                 case 'xjx':
 702                                         if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
 703                                         $s .= $this->getHebrewCalendarMonthNameGen( $hebrew[1] );
 704                                         break;
 705                                 case 'd':
 706                                         $num = substr( $ts, 6, 2 );
 707                                         break;
 708                                 case 'D':
 709                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 710                                         $s .= $this->getWeekdayAbbreviation( gmdate( 'w', $unix ) + 1 );
 711                                         break;
 712                                 case 'j':
 713                                         $num = intval( substr( $ts, 6, 2 ) );
 714                                         break;
 715                                 case 'xij':
 716                                         if ( !$iranian ) $iranian = self::tsToIranian( $ts );
 717                                         $num = $iranian[2];
 718                                         break;
 719                                 case 'xmj':
 720                                         if ( !$hijri ) $hijri = self::tsToHijri( $ts );
 721                                         $num = $hijri[2];
 722                                         break;
 723                                 case 'xjj':
 724                                         if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
 725                                         $num = $hebrew[2];
 726                                         break;
 727                                 case 'l':
 728                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 729                                         $s .= $this->getWeekdayName( gmdate( 'w', $unix ) + 1 );
 730                                         break;
 731                                 case 'N':
 732                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 733                                         $w = gmdate( 'w', $unix );
 734                                         $num = $w ? $w : 7;
 735                                         break;
 736                                 case 'w':
 737                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 738                                         $num = gmdate( 'w', $unix );
 739                                         break;
 740                                 case 'z':
 741                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 742                                         $num = gmdate( 'z', $unix );
 743                                         break;
 744                                 case 'W':
 745                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 746                                         $num = gmdate( 'W', $unix );
 747                                         break;
 748                                 case 'F':
 749                                         $s .= $this->getMonthName( substr( $ts, 4, 2 ) );
 750                                         break;
 751                                 case 'xiF':
 752                                         if ( !$iranian ) $iranian = self::tsToIranian( $ts );
 753                                         $s .= $this->getIranianCalendarMonthName( $iranian[1] );
 754                                         break;
 755                                 case 'xmF':
 756                                         if ( !$hijri ) $hijri = self::tsToHijri( $ts );
 757                                         $s .= $this->getHijriCalendarMonthName( $hijri[1] );
 758                                         break;
 759                                 case 'xjF':
 760                                         if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
 761                                         $s .= $this->getHebrewCalendarMonthName( $hebrew[1] );
 762                                         break;
 763                                 case 'm':
 764                                         $num = substr( $ts, 4, 2 );
 765                                         break;
 766                                 case 'M':
 767                                         $s .= $this->getMonthAbbreviation( substr( $ts, 4, 2 ) );
 768                                         break;
 769                                 case 'n':
 770                                         $num = intval( substr( $ts, 4, 2 ) );
 771                                         break;
 772                                 case 'xin':
 773                                         if ( !$iranian ) $iranian = self::tsToIranian( $ts );
 774                                         $num = $iranian[1];
 775                                         break;
 776                                 case 'xmn':
 777                                         if ( !$hijri ) $hijri = self::tsToHijri ( $ts );
 778                                         $num = $hijri[1];
 779                                         break;
 780                                 case 'xjn':
 781                                         if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
 782                                         $num = $hebrew[1];
 783                                         break;
 784                                 case 't':
 785                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 786                                         $num = gmdate( 't', $unix );
 787                                         break;
 788                                 case 'xjt':
 789                                         if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
 790                                         $num = $hebrew[3];
 791                                         break;
 792                                 case 'L':
 793                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 794                                         $num = gmdate( 'L', $unix );
 795                                         break;
 796                                 # 'o' is supported since PHP 5.1.0
 797                                 # return literal if not supported
 798                                 # TODO: emulation for pre 5.1.0 versions
 799                                 case 'o':
 800                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 801                                         if ( version_compare(PHP_VERSION, '5.1.0') === 1 )
 802                                                 $num = date( 'o', $unix );
 803                                         else
 804                                                 $s .= 'o';
 805                                         break;
 806                                 case 'Y':
 807                                         $num = substr( $ts, 0, 4 );
 808                                         break;
 809                                 case 'xiY':
 810                                         if ( !$iranian ) $iranian = self::tsToIranian( $ts );
 811                                         $num = $iranian[0];
 812                                         break;
 813                                 case 'xmY':
 814                                         if ( !$hijri ) $hijri = self::tsToHijri( $ts );
 815                                         $num = $hijri[0];
 816                                         break;
 817                                 case 'xjY':
 818                                         if ( !$hebrew ) $hebrew = self::tsToHebrew( $ts );
 819                                         $num = $hebrew[0];
 820                                         break;
 821                                 case 'xkY':
 822                                         if ( !$thai ) $thai = self::tsToYear( $ts, 'thai' );
 823                                         $num = $thai[0];
 824                                         break;
 825                                 case 'xoY':
 826                                         if ( !$minguo ) $minguo = self::tsToYear( $ts, 'minguo' );
 827                                         $num = $minguo[0];
 828                                         break;
 829                                 case 'xtY':
 830                                         if ( !$tenno ) $tenno = self::tsToYear( $ts, 'tenno' );
 831                                         $num = $tenno[0];
 832                                         break;
 833                                 case 'y':
 834                                         $num = substr( $ts, 2, 2 );
 835                                         break;
 836                                 case 'a':
 837                                         $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'am' : 'pm';
 838                                         break;
 839                                 case 'A':
 840                                         $s .= intval( substr( $ts, 8, 2 ) ) < 12 ? 'AM' : 'PM';
 841                                         break;
 842                                 case 'g':
 843                                         $h = substr( $ts, 8, 2 );
 844                                         $num = $h % 12 ? $h % 12 : 12;
 845                                         break;
 846                                 case 'G':
 847                                         $num = intval( substr( $ts, 8, 2 ) );
 848                                         break;
 849                                 case 'h':
 850                                         $h = substr( $ts, 8, 2 );
 851                                         $num = sprintf( '%02d', $h % 12 ? $h % 12 : 12 );
 852                                         break;
 853                                 case 'H':
 854                                         $num = substr( $ts, 8, 2 );
 855                                         break;
 856                                 case 'i':
 857                                         $num = substr( $ts, 10, 2 );
 858                                         break;
 859                                 case 's':
 860                                         $num = substr( $ts, 12, 2 );
 861                                         break;
 862                                 case 'c':
 863                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 864                                         $s .= gmdate( 'c', $unix );
 865                                         break;
 866                                 case 'r':
 867                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 868                                         $s .= gmdate( 'r', $unix );
 869                                         break;
 870                                 case 'U':
 871                                         if ( !$unix ) $unix = wfTimestamp( TS_UNIX, $ts );
 872                                         $num = $unix;
 873                                         break;
 874                                 case '\\':
 875                                         # Backslash escaping
 876                                         if ( $p < strlen( $format ) - 1 ) {
 877                                                 $s .= $format[++$p];
 878                                         } else {
 879                                                 $s .= '\\';
 880                                         }
 881                                         break;
 882                                 case '"':
 883                                         # Quoted literal
 884                                         if ( $p < strlen( $format ) - 1 ) {
 885                                                 $endQuote = strpos( $format, '"', $p + 1 );
 886                                                 if ( $endQuote === false ) {
 887                                                         # No terminating quote, assume literal "
 888                                                         $s .= '"';
 889                                                 } else {
 890                                                         $s .= substr( $format, $p + 1, $endQuote - $p - 1 );
 891                                                         $p = $endQuote;
 892                                                 }
 893                                         } else {
 894                                                 # Quote at end of string, assume literal "
 895                                                 $s .= '"';
 896                                         }
 897                                         break;
 898                                 default:
 899                                         $s .= $format[$p];
 900                         }
 901                         if ( $num !== false ) {
 902                                 if ( $rawToggle || $raw ) {
 903                                         $s .= $num;
 904                                         $raw = false;
 905                                 } elseif ( $roman ) {
 906                                         $s .= self::romanNumeral( $num );
 907                                         $roman = false;
 908                                 } elseif( $hebrewNum ) {
 909                                         $s .= self::hebrewNumeral( $num );
 910                                         $hebrewNum = false;
 911                                 } else {
 912                                         $s .= $this->formatNum( $num, true );
 913                                 }
 914                                 $num = false;
 915                         }
 916                 }
 917                 return $s;
 918         }
 919
 920         private static $GREG_DAYS = array( 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 );
 921         private static $IRANIAN_DAYS = array( 31, 31, 31, 31, 31, 31, 30, 30, 30, 30, 30, 29 );
 922         /**
 923          * Algorithm by Roozbeh Pournader and Mohammad Toossi to convert
 924          * Gregorian dates to Iranian dates. Originally written in C, it
 925          * is released under the terms of GNU Lesser General Public
 926          * License. Conversion to PHP was performed by Niklas Laxström.
 927          *
 928          * Link: http://www.farsiweb.info/jalali/jalali.c
 929          */
 930         private static function tsToIranian( $ts ) {
 931                 $gy = substr( $ts, 0, 4 ) -1600;
 932                 $gm = substr( $ts, 4, 2 ) -1;
 933                 $gd = substr( $ts, 6, 2 ) -1;
 934
 935                 # Days passed from the beginning (including leap years)
 936                 $gDayNo = 365*$gy
 937                         + floor(($gy+3) / 4)
 938                         - floor(($gy+99) / 100)
 939                         + floor(($gy+399) / 400);
 940
 941
 942                 // Add days of the past months of this year
 943                 for( $i = 0; $i < $gm; $i++ ) {
 944                         $gDayNo += self::$GREG_DAYS[$i];
 945                 }
 946
 947                 // Leap years
 948                 if ( $gm > 1 && (($gy%4===0 && $gy%100!==0 || ($gy%400==0)))) {
 949                         $gDayNo++;
 950                 }
 951
 952                 // Days passed in current month
 953                 $gDayNo += $gd;
 954
 955                 $jDayNo = $gDayNo - 79;
 956
 957                 $jNp = floor($jDayNo / 12053);
 958                 $jDayNo %= 12053;
 959
 960                 $jy = 979 + 33*$jNp + 4*floor($jDayNo/1461);
 961                 $jDayNo %= 1461;
 962
 963                 if ( $jDayNo >= 366 ) {
 964                         $jy += floor(($jDayNo-1)/365);
 965                         $jDayNo = floor(($jDayNo-1)%365);
 966                 }
 967
 968                 for ( $i = 0; $i < 11 && $jDayNo >= self::$IRANIAN_DAYS[$i]; $i++ ) {
 969                         $jDayNo -= self::$IRANIAN_DAYS[$i];
 970                 }
 971
 972                 $jm= $i+1;
 973                 $jd= $jDayNo+1;
 974
 975                 return array($jy, $jm, $jd);
 976         }
 977         /**
 978          * Converting Gregorian dates to Hijri dates.
 979          *
 980          * Based on a PHP-Nuke block by Sharjeel which is released under GNU/GPL license
 981          *
 982          * @link http://phpnuke.org/modules.php?name=News&file=article&sid=8234&mode=thread&order=0&thold=0
 983          */
 984         private static function tsToHijri ( $ts ) {
 985                 $year = substr( $ts, 0, 4 );
 986                 $month = substr( $ts, 4, 2 );
 987                 $day = substr( $ts, 6, 2 );
 988
 989                 $zyr = $year;
 990                 $zd=$day;
 991                 $zm=$month;
 992                 $zy=$zyr;
 993
 994
 995
 996                 if (($zy>1582)||(($zy==1582)&&($zm>10))||(($zy==1582)&&($zm==10)&&($zd>14)))
 997                         {
 998
 999
1000                                     $zjd=(int)((1461*($zy + 4800 + (int)( ($zm-14) /12) ))/4) + (int)((367*($zm-2-12*((int)(($zm-14)/12))))/12)-(int)((3*(int)(( ($zy+4900+(int)(($zm-14)/12))/100)))/4)+$zd-32075;
1001                                     }
1002                  else
1003                         {
1004                                     $zjd = 367*$zy-(int)((7*($zy+5001+(int)(($zm-9)/7)))/4)+(int)((275*$zm)/9)+$zd+1729777;
1005                         }
1006
1007                 $zl=$zjd-1948440+10632;
1008                 $zn=(int)(($zl-1)/10631);
1009                 $zl=$zl-10631*$zn+354;
1010                 $zj=((int)((10985-$zl)/5316))*((int)((50*$zl)/17719))+((int)($zl/5670))*((int)((43*$zl)/15238));
1011                 $zl=$zl-((int)((30-$zj)/15))*((int)((17719*$zj)/50))-((int)($zj/16))*((int)((15238*$zj)/43))+29;
1012                 $zm=(int)((24*$zl)/709);
1013                 $zd=$zl-(int)((709*$zm)/24);
1014                 $zy=30*$zn+$zj-30;
1015
1016                 return array ($zy, $zm, $zd);
1017         }
1018
1019         /**
1020          * Converting Gregorian dates to Hebrew dates.
1021          *
1022          * Based on a JavaScript code by Abu Mami and Yisrael Hersch
1023          * (abu-mami@kaluach.net, http://www.kaluach.net), who permitted
1024          * to translate the relevant functions into PHP and release them under
1025          * GNU GPL.
1026          *
1027          * The months are counted from Tishrei = 1. In a leap year, Adar I is 13
1028          * and Adar II is 14. In a non-leap year, Adar is 6.
1029          */
1030         private static function tsToHebrew( $ts ) {
1031                 # Parse date
1032                 $year = substr( $ts, 0, 4 );
1033                 $month = substr( $ts, 4, 2 );
1034                 $day = substr( $ts, 6, 2 );
1035
1036                 # Calculate Hebrew year
1037                 $hebrewYear = $year + 3760;
1038
1039                 # Month number when September = 1, August = 12
1040                 $month += 4;
1041                 if( $month > 12 ) {
1042                         # Next year
1043                         $month -= 12;
1044                         $year++;
1045                         $hebrewYear++;
1046                 }
1047
1048                 # Calculate day of year from 1 September
1049                 $dayOfYear = $day;
1050                 for( $i = 1; $i < $month; $i++ ) {
1051                         if( $i == 6 ) {
1052                                 # February
1053                                 $dayOfYear += 28;
1054                                 # Check if the year is leap
1055                                 if( $year % 400 == 0 || ( $year % 4 == 0 && $year % 100 > 0 ) ) {
1056                                         $dayOfYear++;
1057                                 }
1058                         } elseif( $i == 8 || $i == 10 || $i == 1 || $i == 3 ) {
1059                                 $dayOfYear += 30;
1060                         } else {
1061                                 $dayOfYear += 31;
1062                         }
1063                 }
1064
1065                 # Calculate the start of the Hebrew year
1066                 $start = self::hebrewYearStart( $hebrewYear );
1067
1068                 # Calculate next year's start
1069                 if( $dayOfYear <= $start ) {
1070                         # Day is before the start of the year - it is the previous year
1071                         # Next year's start
1072                         $nextStart = $start;
1073                         # Previous year
1074                         $year--;
1075                         $hebrewYear--;
1076                         # Add days since previous year's 1 September
1077                         $dayOfYear += 365;
1078                         if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1079                                 # Leap year
1080                                 $dayOfYear++;
1081                         }
1082                         # Start of the new (previous) year
1083                         $start = self::hebrewYearStart( $hebrewYear );
1084                 } else {
1085                         # Next year's start
1086                         $nextStart = self::hebrewYearStart( $hebrewYear + 1 );
1087                 }
1088
1089                 # Calculate Hebrew day of year
1090                 $hebrewDayOfYear = $dayOfYear - $start;
1091
1092                 # Difference between year's days
1093                 $diff = $nextStart - $start;
1094                 # Add 12 (or 13 for leap years) days to ignore the difference between
1095                 # Hebrew and Gregorian year (353 at least vs. 365/6) - now the
1096                 # difference is only about the year type
1097                 if( ( $year % 400 == 0 ) || ( $year % 100 != 0 && $year % 4 == 0 ) ) {
1098                         $diff += 13;
1099                 } else {
1100                         $diff += 12;
1101                 }
1102
1103                 # Check the year pattern, and is leap year
1104                 # 0 means an incomplete year, 1 means a regular year, 2 means a complete year
1105                 # This is mod 30, to work on both leap years (which add 30 days of Adar I)
1106                 # and non-leap years
1107                 $yearPattern = $diff % 30;
1108                 # Check if leap year
1109                 $isLeap = $diff >= 30;
1110
1111                 # Calculate day in the month from number of day in the Hebrew year
1112                 # Don't check Adar - if the day is not in Adar, we will stop before;
1113                 # if it is in Adar, we will use it to check if it is Adar I or Adar II
1114                 $hebrewDay = $hebrewDayOfYear;
1115                 $hebrewMonth = 1;
1116                 $days = 0;
1117                 while( $hebrewMonth <= 12 ) {
1118                         # Calculate days in this month
1119                         if( $isLeap && $hebrewMonth == 6 ) {
1120                                 # Adar in a leap year
1121                                 if( $isLeap ) {
1122                                         # Leap year - has Adar I, with 30 days, and Adar II, with 29 days
1123                                         $days = 30;
1124                                         if( $hebrewDay <= $days ) {
1125                                                 # Day in Adar I
1126                                                 $hebrewMonth = 13;
1127                                         } else {
1128                                                 # Subtract the days of Adar I
1129                                                 $hebrewDay -= $days;
1130                                                 # Try Adar II
1131                                                 $days = 29;
1132                                                 if( $hebrewDay <= $days ) {
1133                                                         # Day in Adar II
1134                                                         $hebrewMonth = 14;
1135                                                 }
1136                                         }
1137                                 }
1138                         } elseif( $hebrewMonth == 2 && $yearPattern == 2 ) {
1139                                 # Cheshvan in a complete year (otherwise as the rule below)
1140                                 $days = 30;
1141                         } elseif( $hebrewMonth == 3 && $yearPattern == 0 ) {
1142                                 # Kislev in an incomplete year (otherwise as the rule below)
1143                                 $days = 29;
1144                         } else {
1145                                 # Odd months have 30 days, even have 29
1146                                 $days = 30 - ( $hebrewMonth - 1 ) % 2;
1147                         }
1148                         if( $hebrewDay <= $days ) {
1149                                 # In the current month
1150                                 break;
1151                         } else {
1152                                 # Subtract the days of the current month
1153                                 $hebrewDay -= $days;
1154                                 # Try in the next month
1155                                 $hebrewMonth++;
1156                         }
1157                 }
1158
1159                 return array( $hebrewYear, $hebrewMonth, $hebrewDay, $days );
1160         }
1161
1162         /**
1163          * This calculates the Hebrew year start, as days since 1 September.
1164          * Based on Carl Friedrich Gauss algorithm for finding Easter date.
1165          * Used for Hebrew date.
1166          */
1167         private static function hebrewYearStart( $year ) {
1168                 $a = intval( ( 12 * ( $year - 1 ) + 17 ) % 19 );
1169                 $b = intval( ( $year - 1 ) % 4 );
1170                 $m = 32.044093161144 + 1.5542417966212 * $a +  $b / 4.0 - 0.0031777940220923 * ( $year - 1 );
1171                 if( $m < 0 ) {
1172                         $m--;
1173                 }
1174                 $Mar = intval( $m );
1175                 if( $m < 0 ) {
1176                         $m++;
1177                 }
1178                 $m -= $Mar;
1179
1180                 $c = intval( ( $Mar + 3 * ( $year - 1 ) + 5 * $b + 5 ) % 7);
1181                 if( $c == 0 && $a > 11 && $m >= 0.89772376543210 ) {
1182                         $Mar++;
1183                 } else if( $c == 1 && $a > 6 && $m >= 0.63287037037037 ) {
1184                         $Mar += 2;
1185                 } else if( $c == 2 || $c == 4 || $c == 6 ) {
1186                         $Mar++;
1187                 }
1188
1189                 $Mar += intval( ( $year - 3761 ) / 100 ) - intval( ( $year - 3761 ) / 400 ) - 24;
1190                 return $Mar;
1191         }
1192
1193         /**
1194          * Algorithm to convert Gregorian dates to Thai solar dates,
1195          * Minguo dates or Minguo dates.
1196          *
1197          * Link: http://en.wikipedia.org/wiki/Thai_solar_calendar
1198          *       http://en.wikipedia.org/wiki/Minguo_calendar
1199          *       http://en.wikipedia.org/wiki/Japanese_era_name
1200          *
1201          * @param $ts String: 14-character timestamp, calender name
1202          * @return array converted year, month, day
1203          */
1204         private static function tsToYear( $ts, $cName ) {
1205                 $gy = substr( $ts, 0, 4 );
1206                 $gm = substr( $ts, 4, 2 );
1207                 $gd = substr( $ts, 6, 2 );
1208
1209                 if (!strcmp($cName,'thai')) {
1210                         # Thai solar dates
1211                         # Add 543 years to the Gregorian calendar
1212                         # Months and days are identical
1213                         $gy_offset = $gy + 543;
1214                 } else if ((!strcmp($cName,'minguo')) || !strcmp($cName,'juche')) {
1215                         # Minguo dates
1216                         # Deduct 1911 years from the Gregorian calendar
1217                         # Months and days are identical
1218                         $gy_offset = $gy - 1911;
1219                 } else if (!strcmp($cName,'tenno')) {
1220                         # Nengō dates up to Meiji period
1221                         # Deduct years from the Gregorian calendar
1222                         # depending on the nengo periods
1223                         # Months and days are identical
1224                         if (($gy < 1912) || (($gy == 1912) && ($gm < 7)) || (($gy == 1912) && ($gm == 7) && ($gd < 31))) {
1225                                 # Meiji period
1226                                 $gy_gannen = $gy - 1868 + 1;
1227                                 $gy_offset = $gy_gannen;
1228                                 if ($gy_gannen == 1)
1229                                         $gy_offset = '元';
1230                                 $gy_offset = '明治'.$gy_offset;
1231                         } else if ((($gy == 1912) && ($gm == 7) && ($gd == 31)) || (($gy == 1912) && ($gm >= 8)) || (($gy > 1912) && ($gy < 1926)) || (($gy == 1926) && ($gm < 12)) || (($gy == 1926) && ($gm == 12) && ($gd < 26))) {
1232                                 # Taishō period
1233                                 $gy_gannen = $gy - 1912 + 1;
1234                                 $gy_offset = $gy_gannen;
1235                                 if ($gy_gannen == 1)
1236                                         $gy_offset = '元';
1237                                 $gy_offset = '大正'.$gy_offset;
1238                         } else if ((($gy == 1926) && ($gm == 12) && ($gd >= 26)) || (($gy > 1926) && ($gy < 1989)) || (($gy == 1989) && ($gm == 1) && ($gd < 8))) {
1239                                 # Shōwa period
1240                                 $gy_gannen = $gy - 1926 + 1;
1241                                 $gy_offset = $gy_gannen;
1242                                 if ($gy_gannen == 1)
1243                                         $gy_offset = '元';
1244                                 $gy_offset = '昭和'.$gy_offset;
1245                         } else {
1246                                 # Heisei period
1247                                 $gy_gannen = $gy - 1989 + 1;
1248                                 $gy_offset = $gy_gannen;
1249                                 if ($gy_gannen == 1)
1250                                         $gy_offset = '元';
1251                                 $gy_offset = '平成'.$gy_offset;
1252                         }
1253                 } else {
1254                         $gy_offset = $gy;
1255                 }
1256
1257                 return array( $gy_offset, $gm, $gd );
1258         }
1259
1260         /**
1261          * Roman number formatting up to 3000
1262          */
1263         static function romanNumeral( $num ) {
1264                 static $table = array(
1265                         array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1266                         array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1267                         array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1268                         array( '', 'M', 'MM', 'MMM' )
1269                 );
1270
1271                 $num = intval( $num );
1272                 if ( $num > 3000 || $num <= 0 ) {
1273                         return $num;
1274                 }
1275
1276                 $s = '';
1277                 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1278                         if ( $num >= $pow10 ) {
1279                                 $s .= $table[$i][floor($num / $pow10)];
1280                         }
1281                         $num = $num % $pow10;
1282                 }
1283                 return $s;
1284         }
1285
1286         /**
1287          * Hebrew Gematria number formatting up to 9999
1288          */
1289         static function hebrewNumeral( $num ) {
1290                 static $table = array(
1291                         array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1292                         array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1293                         array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1294                         array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1295                 );
1296
1297                 $num = intval( $num );
1298                 if ( $num > 9999 || $num <= 0 ) {
1299                         return $num;
1300                 }
1301
1302                 $s = '';
1303                 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1304                         if ( $num >= $pow10 ) {
1305                                 if ( $num == 15 || $num == 16 ) {
1306                                         $s .= $table[0][9] . $table[0][$num - 9];
1307                                         $num = 0;
1308                                 } else {
1309                                         $s .= $table[$i][intval( ( $num / $pow10 ) )];
1310                                         if( $pow10 == 1000 ) {
1311                                                 $s .= "'";
1312                                         }
1313                                 }
1314                         }
1315                         $num = $num % $pow10;
1316                 }
1317                 if( strlen( $s ) == 2 ) {
1318                         $str = $s . "'";
1319                 } else  {
1320                         $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1321                         $str .= substr( $s, strlen( $s ) - 2, 2 );
1322                 }
1323                 $start = substr( $str, 0, strlen( $str ) - 2 );
1324                 $end = substr( $str, strlen( $str ) - 2 );
1325                 switch( $end ) {
1326                         case 'כ':
1327                                 $str = $start . 'ך';
1328                                 break;
1329                         case 'מ':
1330                                 $str = $start . 'ם';
1331                                 break;
1332                         case 'נ':
1333                                 $str = $start . 'ן';
1334                                 break;
1335                         case 'פ':
1336                                 $str = $start . 'ף';
1337                                 break;
1338                         case 'צ':
1339                                 $str = $start . 'ץ';
1340                                 break;
1341                 }
1342                 return $str;
1343         }
1344
1345         /**
1346          * This is meant to be used by time(), date(), and timeanddate() to get
1347          * the date preference they're supposed to use, it should be used in
1348          * all children.
1349          *
1350          *<code>
1351          * function timeanddate([...], $format = true) {
1352          *      $datePreference = $this->dateFormat($format);
1353          * [...]
1354          * }
1355          *</code>
1356          *
1357          * @param $usePrefs Mixed: if true, the user's preference is used
1358          *                         if false, the site/language default is used
1359          *                         if int/string, assumed to be a format.
1360          * @return string
1361          */
1362         function dateFormat( $usePrefs = true ) {
1363                 global $wgUser;
1364
1365                 if( is_bool( $usePrefs ) ) {
1366                         if( $usePrefs ) {
1367                                 $datePreference = $wgUser->getDatePreference();
1368                         } else {
1369                                 $options = User::getDefaultOptions();
1370                                 $datePreference = (string)$options['date'];
1371                         }
1372                 } else {
1373                         $datePreference = (string)$usePrefs;
1374                 }
1375
1376                 // return int
1377                 if( $datePreference == '' ) {
1378                         return 'default';
1379                 }
1380
1381                 return $datePreference;
1382         }
1383
1384         /**
1385          * Get a format string for a given type and preference
1386          * @param $type May be date, time or both
1387          * @param $pref The format name as it appears in Messages*.php
1388          */
1389         function getDateFormatString( $type, $pref ) {
1390                 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
1391                         if ( $pref == 'default' ) {
1392                                 $pref = $this->getDefaultDateFormat();
1393                                 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1394                         } else {
1395                                 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1396                                 if ( is_null( $df ) ) {
1397                                         $pref = $this->getDefaultDateFormat();
1398                                         $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1399                                 }
1400                         }
1401                         $this->dateFormatStrings[$type][$pref] = $df;
1402                 }
1403                 return $this->dateFormatStrings[$type][$pref];
1404         }
1405
1406         /**
1407          * @param $ts Mixed: the time format which needs to be turned into a
1408          *            date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1409          * @param $adj Bool: whether to adjust the time output according to the
1410          *             user configured offset ($timecorrection)
1411          * @param $format Mixed: true to use user's date format preference
1412          * @param $timecorrection String: the time offset as returned by
1413          *                        validateTimeZone() in Special:Preferences
1414          * @return string
1415          */
1416         function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1417                 if ( $adj ) {
1418                         $ts = $this->userAdjust( $ts, $timecorrection );
1419                 }
1420                 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1421                 return $this->sprintfDate( $df, $ts );
1422         }
1423
1424         /**
1425          * @param $ts Mixed: the time format which needs to be turned into a
1426          *            date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1427          * @param $adj Bool: whether to adjust the time output according to the
1428          *             user configured offset ($timecorrection)
1429          * @param $format Mixed: true to use user's date format preference
1430          * @param $timecorrection String: the time offset as returned by
1431          *                        validateTimeZone() in Special:Preferences
1432          * @return string
1433          */
1434         function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1435                 if ( $adj ) {
1436                         $ts = $this->userAdjust( $ts, $timecorrection );
1437                 }
1438                 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1439                 return $this->sprintfDate( $df, $ts );
1440         }
1441
1442         /**
1443          * @param $ts Mixed: the time format which needs to be turned into a
1444          *            date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1445          * @param $adj Bool: whether to adjust the time output according to the
1446          *             user configured offset ($timecorrection)
1447          * @param $format Mixed: what format to return, if it's false output the
1448          *                default one (default true)
1449          * @param $timecorrection String: the time offset as returned by
1450          *                        validateTimeZone() in Special:Preferences
1451          * @return string
1452          */
1453         function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1454                 $ts = wfTimestamp( TS_MW, $ts );
1455                 if ( $adj ) {
1456                         $ts = $this->userAdjust( $ts, $timecorrection );
1457                 }
1458                 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1459                 return $this->sprintfDate( $df, $ts );
1460         }
1461
1462         function getMessage( $key ) {
1463                 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
1464         }
1465
1466         function getAllMessages() {
1467                 return self::$dataCache->getItem( $this->mCode, 'messages' );
1468         }
1469
1470         function iconv( $in, $out, $string ) {
1471                 # This is a wrapper for iconv in all languages except esperanto,
1472                 # which does some nasty x-conversions beforehand
1473
1474                 # Even with //IGNORE iconv can whine about illegal characters in
1475                 # *input* string. We just ignore those too.
1476                 # REF: http://bugs.php.net/bug.php?id=37166
1477                 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1478                 wfSuppressWarnings();
1479                 $text = iconv( $in, $out . '//IGNORE', $string );
1480                 wfRestoreWarnings();
1481                 return $text;
1482         }
1483
1484         // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1485         function ucwordbreaksCallbackAscii($matches){
1486                 return $this->ucfirst($matches[1]);
1487         }
1488
1489         function ucwordbreaksCallbackMB($matches){
1490                 return mb_strtoupper($matches[0]);
1491         }
1492
1493         function ucCallback($matches){
1494                 list( $wikiUpperChars ) = self::getCaseMaps();
1495                 return strtr( $matches[1], $wikiUpperChars );
1496         }
1497
1498         function lcCallback($matches){
1499                 list( , $wikiLowerChars ) = self::getCaseMaps();
1500                 return strtr( $matches[1], $wikiLowerChars );
1501         }
1502
1503         function ucwordsCallbackMB($matches){
1504                 return mb_strtoupper($matches[0]);
1505         }
1506
1507         function ucwordsCallbackWiki($matches){
1508                 list( $wikiUpperChars ) = self::getCaseMaps();
1509                 return strtr( $matches[0], $wikiUpperChars );
1510         }
1511
1512         function ucfirst( $str ) {
1513                 $o = ord( $str );
1514                 if ( $o < 96 ) {
1515                         return $str;
1516                 } elseif ( $o < 128 ) {
1517                         return ucfirst($str);
1518                 } else {
1519                         // fall back to more complex logic in case of multibyte strings
1520                         return self::uc($str,true);
1521                 }
1522         }
1523
1524         function uc( $str, $first = false ) {
1525                 if ( function_exists( 'mb_strtoupper' ) ) {
1526                         if ( $first ) {
1527                                 if ( self::isMultibyte( $str ) ) {
1528                                         return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1529                                 } else {
1530                                         return ucfirst( $str );
1531                                 }
1532                         } else {
1533                                 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1534                         }
1535                 } else {
1536                         if ( self::isMultibyte( $str ) ) {
1537                                 list( $wikiUpperChars ) = $this->getCaseMaps();
1538                                 $x = $first ? '^' : '';
1539                                 return preg_replace_callback(
1540                                         "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1541                                         array($this,"ucCallback"),
1542                                         $str
1543                                 );
1544                         } else {
1545                                 return $first ? ucfirst( $str ) : strtoupper( $str );
1546                         }
1547                 }
1548         }
1549
1550         function lcfirst( $str ) {
1551                 $o = ord( $str );
1552                 if ( !$o ) {
1553                         return strval( $str );
1554                 } elseif ( $o >= 128 ) {
1555                         return self::lc( $str, true );
1556                 } elseif ( $o > 96 ) {
1557                         return $str;
1558                 } else {
1559                         $str[0] = strtolower( $str[0] );
1560                         return $str;
1561                 }
1562         }
1563
1564         function lc( $str, $first = false ) {
1565                 if ( function_exists( 'mb_strtolower' ) )
1566                         if ( $first )
1567                                 if ( self::isMultibyte( $str ) )
1568                                         return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1569                                 else
1570                                         return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1571                         else
1572                                 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1573                 else
1574                         if ( self::isMultibyte( $str ) ) {
1575                                 list( , $wikiLowerChars ) = self::getCaseMaps();
1576                                 $x = $first ? '^' : '';
1577                                 return preg_replace_callback(
1578                                         "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1579                                         array($this,"lcCallback"),
1580                                         $str
1581                                 );
1582                         } else
1583                                 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1584         }
1585
1586         function isMultibyte( $str ) {
1587                 return (bool)preg_match( '/[\x80-\xff]/', $str );
1588         }
1589
1590         function ucwords($str) {
1591                 if ( self::isMultibyte( $str ) ) {
1592                         $str = self::lc($str);
1593
1594                         // regexp to find first letter in each word (i.e. after each space)
1595                         $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1596
1597                         // function to use to capitalize a single char
1598                         if ( function_exists( 'mb_strtoupper' ) )
1599                                 return preg_replace_callback(
1600                                         $replaceRegexp,
1601                                         array($this,"ucwordsCallbackMB"),
1602                                         $str
1603                                 );
1604                         else
1605                                 return preg_replace_callback(
1606                                         $replaceRegexp,
1607                                         array($this,"ucwordsCallbackWiki"),
1608                                         $str
1609                                 );
1610                 }
1611                 else
1612                         return ucwords( strtolower( $str ) );
1613         }
1614
1615   # capitalize words at word breaks
1616         function ucwordbreaks($str){
1617                 if (self::isMultibyte( $str ) ) {
1618                         $str = self::lc($str);
1619
1620                         // since \b doesn't work for UTF-8, we explicitely define word break chars
1621                         $breaks= "[ \-\(\)\}\{\.,\?!]";
1622
1623                         // find first letter after word break
1624                         $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1625
1626                         if ( function_exists( 'mb_strtoupper' ) )
1627                                 return preg_replace_callback(
1628                                         $replaceRegexp,
1629                                         array($this,"ucwordbreaksCallbackMB"),
1630                                         $str
1631                                 );
1632                         else
1633                                 return preg_replace_callback(
1634                                         $replaceRegexp,
1635                                         array($this,"ucwordsCallbackWiki"),
1636                                         $str
1637                                 );
1638                 }
1639                 else
1640                         return preg_replace_callback(
1641                         '/\b([\w\x80-\xff]+)\b/',
1642                         array($this,"ucwordbreaksCallbackAscii"),
1643                         $str );
1644         }
1645
1646         /**
1647          * Return a case-folded representation of $s
1648          *
1649          * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1650          * and $s2 are the same except for the case of their characters. It is not
1651          * necessary for the value returned to make sense when displayed.
1652          *
1653          * Do *not* perform any other normalisation in this function. If a caller
1654          * uses this function when it should be using a more general normalisation
1655          * function, then fix the caller.
1656          */
1657         function caseFold( $s ) {
1658                 return $this->uc( $s );
1659         }
1660
1661         function checkTitleEncoding( $s ) {
1662                 if( is_array( $s ) ) {
1663                         wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1664                 }
1665                 # Check for non-UTF-8 URLs
1666                 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1667                 if(!$ishigh) return $s;
1668
1669                 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1670                 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1671                 if( $isutf8 ) return $s;
1672
1673                 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1674         }
1675
1676         function fallback8bitEncoding() {
1677                 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
1678         }
1679
1680         /**
1681          * Most writing systems use whitespace to break up words.
1682          * Some languages such as Chinese don't conventionally do this,
1683          * which requires special handling when breaking up words for
1684          * searching etc.
1685          */
1686         function hasWordBreaks() {
1687                 return true;
1688         }
1689
1690         /**
1691          * Some languages have special punctuation to strip out.
1692          * Make such changes here.
1693          *
1694          * @param $string String
1695          * @return String
1696          */
1697         function stripForSearch( $string, $doStrip = true ) {
1698                 return $string;
1699         }
1700
1701         /**
1702          * convert double-width roman characters to single-width.
1703          * range: ff00-ff5f ~= 0020-007f
1704          */
1705         protected static function convertDoubleWidth( $string ) {
1706                 $string = preg_replace( '/\xef\xbc([\x80-\xbf])/e', 'chr((ord("$1") & 0x3f) + 0x20)', $string );
1707                 $string = preg_replace( '/\xef\xbd([\x80-\x99])/e', 'chr((ord("$1") & 0x3f) + 0x60)', $string );
1708                 return $string;
1709         }
1710
1711         protected static function wordSegmentation( $string, $pattern ) {
1712                 $string = preg_replace( $pattern, " $1 ", $string );
1713                 $string = preg_replace( '/ +/', ' ', $string );
1714                 return $string;
1715         }
1716
1717         function convertForSearchResult( $termsArray ) {
1718                 # some languages, e.g. Chinese, need to do a conversion
1719                 # in order for search results to be displayed correctly
1720                 return $termsArray;
1721         }
1722
1723         /**
1724          * Get the first character of a string.
1725          *
1726          * @param $s string
1727          * @return string
1728          */
1729         function firstChar( $s ) {
1730                 $matches = array();
1731                 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1732                 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1733
1734                 if ( isset( $matches[1] ) ) {
1735                         if ( strlen( $matches[1] ) != 3 ) {
1736                                 return $matches[1];
1737                         }
1738
1739                         // Break down Hangul syllables to grab the first jamo
1740                         $code = utf8ToCodepoint( $matches[1] );
1741                         if ( $code < 0xac00 || 0xd7a4 <= $code) {
1742                                 return $matches[1];
1743                         } elseif ( $code < 0xb098 ) {
1744                                 return "\xe3\x84\xb1";
1745                         } elseif ( $code < 0xb2e4 ) {
1746                                 return "\xe3\x84\xb4";
1747                         } elseif ( $code < 0xb77c ) {
1748                                 return "\xe3\x84\xb7";
1749                         } elseif ( $code < 0xb9c8 ) {
1750                                 return "\xe3\x84\xb9";
1751                         } elseif ( $code < 0xbc14 ) {
1752                                 return "\xe3\x85\x81";
1753                         } elseif ( $code < 0xc0ac ) {
1754                                 return "\xe3\x85\x82";
1755                         } elseif ( $code < 0xc544 ) {
1756                                 return "\xe3\x85\x85";
1757                         } elseif ( $code < 0xc790 ) {
1758                                 return "\xe3\x85\x87";
1759                         } elseif ( $code < 0xcc28 ) {
1760                                 return "\xe3\x85\x88";
1761                         } elseif ( $code < 0xce74 ) {
1762                                 return "\xe3\x85\x8a";
1763                         } elseif ( $code < 0xd0c0 ) {
1764                                 return "\xe3\x85\x8b";
1765                         } elseif ( $code < 0xd30c ) {
1766                                 return "\xe3\x85\x8c";
1767                         } elseif ( $code < 0xd558 ) {
1768                                 return "\xe3\x85\x8d";
1769                         } else {
1770                                 return "\xe3\x85\x8e";
1771                         }
1772                 } else {
1773                         return "";
1774                 }
1775         }
1776
1777         function initEncoding() {
1778                 # Some languages may have an alternate char encoding option
1779                 # (Esperanto X-coding, Japanese furigana conversion, etc)
1780                 # If this language is used as the primary content language,
1781                 # an override to the defaults can be set here on startup.
1782         }
1783
1784         function recodeForEdit( $s ) {
1785                 # For some languages we'll want to explicitly specify
1786                 # which characters make it into the edit box raw
1787                 # or are converted in some way or another.
1788                 # Note that if wgOutputEncoding is different from
1789                 # wgInputEncoding, this text will be further converted
1790                 # to wgOutputEncoding.
1791                 global $wgEditEncoding;
1792                 if( $wgEditEncoding == '' or
1793                   $wgEditEncoding == 'UTF-8' ) {
1794                         return $s;
1795                 } else {
1796                         return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1797                 }
1798         }
1799
1800         function recodeInput( $s ) {
1801                 # Take the previous into account.
1802                 global $wgEditEncoding;
1803                 if($wgEditEncoding != "") {
1804                         $enc = $wgEditEncoding;
1805                 } else {
1806                         $enc = 'UTF-8';
1807                 }
1808                 if( $enc == 'UTF-8' ) {
1809                         return $s;
1810                 } else {
1811                         return $this->iconv( $enc, 'UTF-8', $s );
1812                 }
1813         }
1814
1815         /**
1816          * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
1817          * also cleans up certain backwards-compatible sequences, converting them
1818          * to the modern Unicode equivalent.
1819          *
1820          * This is language-specific for performance reasons only.
1821          */
1822         function normalize( $s ) {
1823                 return UtfNormal::cleanUp( $s );
1824         }
1825
1826         /**
1827          * Transform a string using serialized data stored in the given file (which
1828          * must be in the serialized subdirectory of $IP). The file contains pairs
1829          * mapping source characters to destination characters.
1830          *
1831          * The data is cached in process memory. This will go faster if you have the
1832          * FastStringSearch extension.
1833          */
1834         function transformUsingPairFile( $file, $string ) {
1835                 if ( !isset( $this->transformData[$file] ) ) {
1836                         $data = wfGetPrecompiledData( $file );
1837                         if ( $data === false ) {
1838                                 throw new MWException( __METHOD__.": The transformation file $file is missing" );
1839                         }
1840                         $this->transformData[$file] = new ReplacementArray( $data );
1841                 }
1842                 return $this->transformData[$file]->replace( $string );
1843         }
1844
1845         /**
1846          * For right-to-left language support
1847          *
1848          * @return bool
1849          */
1850         function isRTL() {
1851                 return self::$dataCache->getItem( $this->mCode, 'rtl' );
1852         }
1853
1854         /**
1855          * Return the correct HTML 'dir' attribute value for this language.
1856          * @return String
1857          */
1858         function getDir() {
1859                 return $this->isRTL() ? 'rtl' : 'ltr';
1860         }
1861
1862         /**
1863          * Return 'left' or 'right' as appropriate alignment for line-start
1864          * for this language's text direction.
1865          *
1866          * Should be equivalent to CSS3 'start' text-align value....
1867          *
1868          * @return String
1869          */
1870         function alignStart() {
1871                 return $this->isRTL() ? 'right' : 'left';
1872         }
1873
1874         /**
1875          * Return 'right' or 'left' as appropriate alignment for line-end
1876          * for this language's text direction.
1877          *
1878          * Should be equivalent to CSS3 'end' text-align value....
1879          *
1880          * @return String
1881          */
1882         function alignEnd() {
1883                 return $this->isRTL() ? 'left' : 'right';
1884         }
1885
1886         /**
1887          * A hidden direction mark (LRM or RLM), depending on the language direction
1888          *
1889          * @return string
1890          */
1891         function getDirMark() {
1892                 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1893         }
1894
1895         function capitalizeAllNouns() {
1896                 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
1897         }
1898
1899         /**
1900          * An arrow, depending on the language direction
1901          *
1902          * @return string
1903          */
1904         function getArrow() {
1905                 return $this->isRTL() ? '←' : '→';
1906         }
1907
1908         /**
1909          * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1910          *
1911          * @return bool
1912          */
1913         function linkPrefixExtension() {
1914                 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
1915         }
1916
1917         function getMagicWords() {
1918                 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
1919         }
1920
1921         # Fill a MagicWord object with data from here
1922         function getMagic( $mw ) {
1923                 if ( !$this->mMagicHookDone ) {
1924                         $this->mMagicHookDone = true;
1925                         wfProfileIn( 'LanguageGetMagic' );
1926                         wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1927                         wfProfileOut( 'LanguageGetMagic' );
1928                 }
1929                 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1930                         $rawEntry = $this->mMagicExtensions[$mw->mId];
1931                 } else {
1932                         $magicWords = $this->getMagicWords();
1933                         if ( isset( $magicWords[$mw->mId] ) ) {
1934                                 $rawEntry = $magicWords[$mw->mId];
1935                         } else {
1936                                 $rawEntry = false;
1937                         }
1938                 }
1939
1940                 if( !is_array( $rawEntry ) ) {
1941                         error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1942                 } else {
1943                         $mw->mCaseSensitive = $rawEntry[0];
1944                         $mw->mSynonyms = array_slice( $rawEntry, 1 );
1945                 }
1946         }
1947
1948         /**
1949          * Add magic words to the extension array
1950          */
1951         function addMagicWordsByLang( $newWords ) {
1952                 $code = $this->getCode();
1953                 $fallbackChain = array();
1954                 while ( $code && !in_array( $code, $fallbackChain ) ) {
1955                         $fallbackChain[] = $code;
1956                         $code = self::getFallbackFor( $code );
1957                 }
1958                 if ( !in_array( 'en', $fallbackChain ) ) {
1959                         $fallbackChain[] = 'en';
1960                 }
1961                 $fallbackChain = array_reverse( $fallbackChain );
1962                 foreach ( $fallbackChain as $code ) {
1963                         if ( isset( $newWords[$code] ) ) {
1964                                 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1965                         }
1966                 }
1967         }
1968
1969         /**
1970          * Get special page names, as an associative array
1971          *   case folded alias => real name
1972          */
1973         function getSpecialPageAliases() {
1974                 // Cache aliases because it may be slow to load them
1975                 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
1976                         // Initialise array
1977                         $this->mExtendedSpecialPageAliases =
1978                                 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
1979                         wfRunHooks( 'LanguageGetSpecialPageAliases',
1980                                 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1981                 }
1982
1983                 return $this->mExtendedSpecialPageAliases;
1984         }
1985
1986         /**
1987          * Italic is unsuitable for some languages
1988          *
1989          * @param $text String: the text to be emphasized.
1990          * @return string
1991          */
1992         function emphasize( $text ) {
1993                 return "<em>$text</em>";
1994         }
1995
1996          /**
1997           * Normally we output all numbers in plain en_US style, that is
1998           * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
1999           * point twohundredthirtyfive. However this is not sutable for all
2000           * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2001           * Icelandic just want to use commas instead of dots, and dots instead
2002           * of commas like "293.291,235".
2003           *
2004           * An example of this function being called:
2005           * <code>
2006           * wfMsg( 'message', $wgLang->formatNum( $num ) )
2007           * </code>
2008           *
2009           * See LanguageGu.php for the Gujarati implementation and
2010           * $separatorTransformTable on MessageIs.php for
2011           * the , => . and . => , implementation.
2012           *
2013           * @todo check if it's viable to use localeconv() for the decimal
2014           *       separator thing.
2015           * @param $number Mixed: the string to be formatted, should be an integer
2016           *        or a floating point number.
2017           * @param $nocommafy Bool: set to true for special numbers like dates
2018           * @return string
2019           */
2020         function formatNum( $number, $nocommafy = false ) {
2021                 global $wgTranslateNumerals;
2022                 if (!$nocommafy) {
2023                         $number = $this->commafy($number);
2024                         $s = $this->separatorTransformTable();
2025                         if ($s) { $number = strtr($number, $s); }
2026                 }
2027
2028                 if ($wgTranslateNumerals) {
2029                         $s = $this->digitTransformTable();
2030                         if ($s) { $number = strtr($number, $s); }
2031                 }
2032
2033                 return $number;
2034         }
2035
2036         function parseFormattedNumber( $number ) {
2037                 $s = $this->digitTransformTable();
2038                 if ($s) { $number = strtr($number, array_flip($s)); }
2039
2040                 $s = $this->separatorTransformTable();
2041                 if ($s) { $number = strtr($number, array_flip($s)); }
2042
2043                 $number = strtr( $number, array (',' => '') );
2044                 return $number;
2045         }
2046
2047         /**
2048          * Adds commas to a given number
2049          *
2050          * @param $_ mixed
2051          * @return string
2052          */
2053         function commafy($_) {
2054                 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
2055         }
2056
2057         function digitTransformTable() {
2058                 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
2059         }
2060
2061         function separatorTransformTable() {
2062                 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
2063         }
2064
2065
2066         /**
2067          * Take a list of strings and build a locale-friendly comma-separated
2068          * list, using the local comma-separator message.
2069          * The last two strings are chained with an "and".
2070          *
2071          * @param $l Array
2072          * @return string
2073          */
2074         function listToText( $l ) {
2075                 $s = '';
2076                 $m = count( $l ) - 1;
2077                 if( $m == 1 ) {
2078                         return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2079                 }
2080                 else {
2081                         for ( $i = $m; $i >= 0; $i-- ) {
2082                                 if ( $i == $m ) {
2083                                         $s = $l[$i];
2084                                 } else if( $i == $m - 1 ) {
2085                                         $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2086                                 } else {
2087                                         $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2088                                 }
2089                         }
2090                         return $s;
2091                 }
2092         }
2093
2094         /**
2095          * Take a list of strings and build a locale-friendly comma-separated
2096          * list, using the local comma-separator message.
2097          * @param $list array of strings to put in a comma list
2098          * @return string
2099          */
2100         function commaList( $list ) {
2101                 return implode(
2102                         $list,
2103                         wfMsgExt( 'comma-separator', array( 'parsemag', 'escapenoentities', 'language' => $this ) ) );
2104         }
2105
2106         /**
2107          * Take a list of strings and build a locale-friendly semicolon-separated
2108          * list, using the local semicolon-separator message.
2109          * @param $list array of strings to put in a semicolon list
2110          * @return string
2111          */
2112         function semicolonList( $list ) {
2113                 return implode(
2114                         $list,
2115                         wfMsgExt( 'semicolon-separator', array( 'parsemag', 'escapenoentities', 'language' => $this ) ) );
2116         }
2117
2118         /**
2119          * Same as commaList, but separate it with the pipe instead.
2120          * @param $list array of strings to put in a pipe list
2121          * @return string
2122          */
2123         function pipeList( $list ) {
2124                 return implode(
2125                         $list,
2126                         wfMsgExt( 'pipe-separator', array( 'escapenoentities', 'language' => $this ) ) );
2127         }
2128
2129         /**
2130          * Truncate a string to a specified length in bytes, appending an optional
2131          * string (e.g. for ellipses)
2132          *
2133          * The database offers limited byte lengths for some columns in the database;
2134          * multi-byte character sets mean we need to ensure that only whole characters
2135          * are included, otherwise broken characters can be passed to the user
2136          *
2137          * If $length is negative, the string will be truncated from the beginning
2138          *
2139          * @param $string String to truncate
2140          * @param $length Int: maximum length (excluding ellipses)
2141          * @param $ellipsis String to append to the truncated text
2142          * @return string
2143          */
2144         function truncate( $string, $length, $ellipsis = '...' ) {
2145                 # Use the localized ellipsis character
2146                 if( $ellipsis == '...' ) {
2147                         $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2148                 }
2149
2150                 if( $length == 0 ) {
2151                         return $ellipsis;
2152                 }
2153                 if ( strlen( $string ) <= abs( $length ) ) {
2154                         return $string;
2155                 }
2156                 $stringOriginal = $string;
2157                 if( $length > 0 ) {
2158                         $string = substr( $string, 0, $length );
2159                         $char = ord( $string[strlen( $string ) - 1] );
2160                         $m = array();
2161                         if ($char >= 0xc0) {
2162                                 # We got the first byte only of a multibyte char; remove it.
2163                                 $string = substr( $string, 0, -1 );
2164                         } elseif( $char >= 0x80 &&
2165                                   preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2166                                               '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) ) {
2167                                 # We chopped in the middle of a character; remove it
2168                                 $string = $m[1];
2169                         }
2170                         # Do not truncate if the ellipsis actually make the string longer. Bug 22181
2171                         if ( strlen( $string ) + strlen( $ellipsis ) < strlen( $stringOriginal ) ) {
2172                                 return $string . $ellipsis;
2173                         } else {
2174                                 return $stringOriginal;
2175                         }
2176
2177                 } else {
2178                         $string = substr( $string, $length );
2179                         $char = ord( $string[0] );
2180                         if( $char >= 0x80 && $char < 0xc0 ) {
2181                                 # We chopped in the middle of a character; remove the whole thing
2182                                 $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2183                         }
2184                         # Do not truncate if the ellipsis actually make the string longer. Bug 22181
2185                         if ( strlen( $string ) + strlen( $ellipsis ) < strlen( $stringOriginal ) ) {
2186                                 return $ellipsis . $string;
2187                         } else {
2188                                 return $stringOriginal;
2189                         }
2190                 }
2191         }
2192
2193         /**
2194          * Grammatical transformations, needed for inflected languages
2195          * Invoked by putting {{grammar:case|word}} in a message
2196          *
2197          * @param $word string
2198          * @param $case string
2199          * @return string
2200          */
2201         function convertGrammar( $word, $case ) {
2202                 global $wgGrammarForms;
2203                 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
2204                         return $wgGrammarForms[$this->getCode()][$case][$word];
2205                 }
2206                 return $word;
2207         }
2208
2209         /**
2210          * Provides an alternative text depending on specified gender.
2211          * Usage {{gender:username|masculine|feminine|neutral}}.
2212          * username is optional, in which case the gender of current user is used,
2213          * but only in (some) interface messages; otherwise default gender is used.
2214          * If second or third parameter are not specified, masculine is used.
2215          * These details may be overriden per language.
2216          */
2217         function gender( $gender, $forms ) {
2218                 if ( !count($forms) ) { return ''; }
2219                 $forms = $this->preConvertPlural( $forms, 2 );
2220                 if ( $gender === 'male' ) return $forms[0];
2221                 if ( $gender === 'female' ) return $forms[1];
2222                 return isset($forms[2]) ? $forms[2] : $forms[0];
2223         }
2224
2225         /**
2226          * Plural form transformations, needed for some languages.
2227          * For example, there are 3 form of plural in Russian and Polish,
2228          * depending on "count mod 10". See [[w:Plural]]
2229          * For English it is pretty simple.
2230          *
2231          * Invoked by putting {{plural:count|wordform1|wordform2}}
2232          * or {{plural:count|wordform1|wordform2|wordform3}}
2233          *
2234          * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2235          *
2236          * @param $count Integer: non-localized number
2237          * @param $forms Array: different plural forms
2238          * @return string Correct form of plural for $count in this language
2239          */
2240         function convertPlural( $count, $forms ) {
2241                 if ( !count($forms) ) { return ''; }
2242                 $forms = $this->preConvertPlural( $forms, 2 );
2243
2244                 return ( $count == 1 ) ? $forms[0] : $forms[1];
2245         }
2246
2247         /**
2248          * Checks that convertPlural was given an array and pads it to requested
2249          * amound of forms by copying the last one.
2250          *
2251          * @param $count Integer: How many forms should there be at least
2252          * @param $forms Array of forms given to convertPlural
2253          * @return array Padded array of forms or an exception if not an array
2254          */
2255         protected function preConvertPlural( /* Array */ $forms, $count ) {
2256                 while ( count($forms) < $count ) {
2257                         $forms[] = $forms[count($forms)-1];
2258                 }
2259                 return $forms;
2260         }
2261
2262         /**
2263          * For translaing of expiry times
2264          * @param $str String: the validated block time in English
2265          * @return Somehow translated block time
2266          * @see LanguageFi.php for example implementation
2267          */
2268         function translateBlockExpiry( $str ) {
2269
2270                 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2271
2272                 if ( $scBlockExpiryOptions == '-') {
2273                         return $str;
2274                 }
2275
2276                 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2277                         if ( strpos($option, ":") === false )
2278                                 continue;
2279                         list($show, $value) = explode(":", $option);
2280                         if ( strcmp ( $str, $value) == 0 ) {
2281                                 return htmlspecialchars( trim( $show ) );
2282                         }
2283                 }
2284
2285                 return $str;
2286         }
2287
2288         /**
2289          * languages like Chinese need to be segmented in order for the diff
2290          * to be of any use
2291          *
2292          * @param $text String
2293          * @return String
2294          */
2295         function segmentForDiff( $text ) {
2296                 return $text;
2297         }
2298
2299         /**
2300          * and unsegment to show the result
2301          *
2302          * @param $text String
2303          * @return String
2304          */
2305         function unsegmentForDiff( $text ) {
2306                 return $text;
2307         }
2308
2309         # convert text to all supported variants
2310         function autoConvertToAllVariants($text) {
2311                 return $this->mConverter->autoConvertToAllVariants($text);
2312         }
2313
2314         # convert text to different variants of a language.
2315         function convert( $text, $isTitle = false) {
2316                 return $this->mConverter->convert($text, $isTitle);
2317         }
2318
2319         # Check if this is a language with variants
2320         function hasVariants(){
2321                 return sizeof($this->getVariants())>1;
2322         }
2323
2324         # Put custom tags (e.g. -{ }-) around math to prevent conversion
2325         function armourMath($text){
2326                 return $this->mConverter->armourMath($text);
2327         }
2328
2329
2330         /**
2331          * Perform output conversion on a string, and encode for safe HTML output.
2332          * @param $text String text to be converted
2333          * @param $isTitle Bool whether this conversion is for the article title
2334          * @return string
2335          * @todo this should get integrated somewhere sane
2336          */
2337         function convertHtml( $text, $isTitle = false ) {
2338                 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2339         }
2340
2341         function convertCategoryKey( $key ) {
2342                 return $this->mConverter->convertCategoryKey( $key );
2343         }
2344
2345         /**
2346          * get the list of variants supported by this langauge
2347          * see sample implementation in LanguageZh.php
2348          *
2349          * @return array an array of language codes
2350          */
2351         function getVariants() {
2352                 return $this->mConverter->getVariants();
2353         }
2354
2355
2356         function getPreferredVariant( $fromUser = true, $fromHeader = false ) {
2357                 return $this->mConverter->getPreferredVariant( $fromUser, $fromHeader );
2358         }
2359
2360         /**
2361          * if a language supports multiple variants, it is
2362          * possible that non-existing link in one variant
2363          * actually exists in another variant. this function
2364          * tries to find it. See e.g. LanguageZh.php
2365          *
2366          * @param $link String: the name of the link
2367          * @param $nt Mixed: the title object of the link
2368          * @param boolean $ignoreOtherCond: to disable other conditions when
2369          *      we need to transclude a template or update a category's link
2370          * @return null the input parameters may be modified upon return
2371          */
2372         function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
2373                 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
2374         }
2375
2376         /**
2377          * If a language supports multiple variants, converts text
2378          * into an array of all possible variants of the text:
2379          *  'variant' => text in that variant
2380          */
2381         function convertLinkToAllVariants($text){
2382                 return $this->mConverter->convertLinkToAllVariants($text);
2383         }
2384
2385
2386         /**
2387          * returns language specific options used by User::getPageRenderHash()
2388          * for example, the preferred language variant
2389          *
2390          * @return string
2391          */
2392         function getExtraHashOptions() {
2393                 return $this->mConverter->getExtraHashOptions();
2394         }
2395
2396         /**
2397          * for languages that support multiple variants, the title of an
2398          * article may be displayed differently in different variants. this
2399          * function returns the apporiate title defined in the body of the article.
2400          *
2401          * @return string
2402          */
2403         function getParsedTitle() {
2404                 return $this->mConverter->getParsedTitle();
2405         }
2406
2407         /**
2408          * Enclose a string with the "no conversion" tag. This is used by
2409          * various functions in the Parser
2410          *
2411          * @param $text String: text to be tagged for no conversion
2412          * @param $noParse
2413          * @return string the tagged text
2414          */
2415         function markNoConversion( $text, $noParse=false ) {
2416                 return $this->mConverter->markNoConversion( $text, $noParse );
2417         }
2418
2419         /**
2420          * A regular expression to match legal word-trailing characters
2421          * which should be merged onto a link of the form [[foo]]bar.
2422          *
2423          * @return string
2424          */
2425         function linkTrail() {
2426                 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
2427         }
2428
2429         function getLangObj() {
2430                 return $this;
2431         }
2432
2433         /**
2434          * Get the RFC 3066 code for this language object
2435          */
2436         function getCode() {
2437                 return $this->mCode;
2438         }
2439
2440         function setCode( $code ) {
2441                 $this->mCode = $code;
2442         }
2443
2444         /**
2445          * Get the name of a file for a certain language code
2446          * @param $prefix string Prepend this to the filename
2447          * @param $code string Language code
2448          * @param $suffix string Append this to the filename
2449          * @return string $prefix . $mangledCode . $suffix
2450          */
2451         static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2452                 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2453         }
2454
2455         /**
2456          * Get the language code from a file name. Inverse of getFileName()
2457          * @param $filename string $prefix . $languageCode . $suffix
2458          * @param $prefix string Prefix before the language code
2459          * @param $suffix string Suffix after the language code
2460          * @return Language code, or false if $prefix or $suffix isn't found
2461          */
2462         static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
2463                 $m = null;
2464                 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
2465                         preg_quote( $suffix, '/' ) . '/', $filename, $m );
2466                 if ( !count( $m ) ) {
2467                         return false;
2468                 }
2469                 return str_replace( '_', '-', strtolower( $m[1] ) );
2470         }
2471
2472         static function getMessagesFileName( $code ) {
2473                 global $IP;
2474                 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2475         }
2476
2477         static function getClassFileName( $code ) {
2478                 global $IP;
2479                 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2480         }
2481
2482         /**
2483          * Get the fallback for a given language
2484          */
2485         static function getFallbackFor( $code ) {
2486                 if ( $code === 'en' ) {
2487                         // Shortcut
2488                         return false;
2489                 } else {
2490                         return self::getLocalisationCache()->getItem( $code, 'fallback' );
2491                 }
2492         }
2493
2494         /**
2495          * Get all messages for a given language
2496          * WARNING: this may take a long time
2497          */
2498         static function getMessagesFor( $code ) {
2499                 return self::getLocalisationCache()->getItem( $code, 'messages' );
2500         }
2501
2502         /**
2503          * Get a message for a given language
2504          */
2505         static function getMessageFor( $key, $code ) {
2506                 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
2507         }
2508
2509         function fixVariableInNamespace( $talk ) {
2510                 if ( strpos( $talk, '$1' ) === false ) return $talk;
2511
2512                 global $wgMetaNamespace;
2513                 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2514
2515                 # Allow grammar transformations
2516                 # Allowing full message-style parsing would make simple requests
2517                 # such as action=raw much more expensive than they need to be.
2518                 # This will hopefully cover most cases.
2519                 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2520                         array( &$this, 'replaceGrammarInNamespace' ), $talk );
2521                 return str_replace( ' ', '_', $talk );
2522         }
2523
2524         function replaceGrammarInNamespace( $m ) {
2525                 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2526         }
2527
2528         static function getCaseMaps() {
2529                 static $wikiUpperChars, $wikiLowerChars;
2530                 if ( isset( $wikiUpperChars ) ) {
2531                         return array( $wikiUpperChars, $wikiLowerChars );
2532                 }
2533
2534                 wfProfileIn( __METHOD__ );
2535                 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2536                 if ( $arr === false ) {
2537                         throw new MWException(
2538                                 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2539                 }
2540                 extract( $arr );
2541                 wfProfileOut( __METHOD__ );
2542                 return array( $wikiUpperChars, $wikiLowerChars );
2543         }
2544
2545         function formatTimePeriod( $seconds ) {
2546                 if ( $seconds < 10 ) {
2547                         return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2548                 } elseif ( $seconds < 60 ) {
2549                         return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2550                 } elseif ( $seconds < 3600 ) {
2551                         return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2552                                 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2553                 } else {
2554                         $hours = floor( $seconds / 3600 );
2555                         $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2556                         $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2557                         return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2558                                 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2559                                 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2560                 }
2561         }
2562
2563         function formatBitrate( $bps ) {
2564                 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2565                 if ( $bps <= 0 ) {
2566                         return $this->formatNum( $bps ) . $units[0];
2567                 }
2568                 $unitIndex = floor( log10( $bps ) / 3 );
2569                 $mantissa = $bps / pow( 1000, $unitIndex );
2570                 if ( $mantissa < 10 ) {
2571                         $mantissa = round( $mantissa, 1 );
2572                 } else {
2573                         $mantissa = round( $mantissa );
2574                 }
2575                 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2576         }
2577
2578         /**
2579          * Format a size in bytes for output, using an appropriate
2580          * unit (B, KB, MB or GB) according to the magnitude in question
2581          *
2582          * @param $size Size to format
2583          * @return string Plain text (not HTML)
2584          */
2585         function formatSize( $size ) {
2586                 // For small sizes no decimal places necessary
2587                 $round = 0;
2588                 if( $size > 1024 ) {
2589                         $size = $size / 1024;
2590                         if( $size > 1024 ) {
2591                                 $size = $size / 1024;
2592                                 // For MB and bigger two decimal places are smarter
2593                                 $round = 2;
2594                                 if( $size > 1024 ) {
2595                                         $size = $size / 1024;
2596                                         $msg = 'size-gigabytes';
2597                                 } else {
2598                                         $msg = 'size-megabytes';
2599                                 }
2600                         } else {
2601                                 $msg = 'size-kilobytes';
2602                         }
2603                 } else {
2604                         $msg = 'size-bytes';
2605                 }
2606                 $size = round( $size, $round );
2607                 $text = $this->getMessageFromDB( $msg );
2608                 return str_replace( '$1', $this->formatNum( $size ), $text );
2609         }
2610
2611         /**
2612          * Get the conversion rule title, if any.
2613          */
2614         function getConvRuleTitle() {
2615                 return $this->mConverter->getConvRuleTitle();
2616         }
2617 }