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
1202          * @param $cName String: calender name
1203          * @return Array: converted year, month, day
1204          */
1205         private static function tsToYear( $ts, $cName ) {
1206                 $gy = substr( $ts, 0, 4 );
1207                 $gm = substr( $ts, 4, 2 );
1208                 $gd = substr( $ts, 6, 2 );
1209
1210                 if (!strcmp($cName,'thai')) {
1211                         # Thai solar dates
1212                         # Add 543 years to the Gregorian calendar
1213                         # Months and days are identical
1214                         $gy_offset = $gy + 543;
1215                 } else if ((!strcmp($cName,'minguo')) || !strcmp($cName,'juche')) {
1216                         # Minguo dates
1217                         # Deduct 1911 years from the Gregorian calendar
1218                         # Months and days are identical
1219                         $gy_offset = $gy - 1911;
1220                 } else if (!strcmp($cName,'tenno')) {
1221                         # Nengō dates up to Meiji period
1222                         # Deduct years from the Gregorian calendar
1223                         # depending on the nengo periods
1224                         # Months and days are identical
1225                         if (($gy < 1912) || (($gy == 1912) && ($gm < 7)) || (($gy == 1912) && ($gm == 7) && ($gd < 31))) {
1226                                 # Meiji period
1227                                 $gy_gannen = $gy - 1868 + 1;
1228                                 $gy_offset = $gy_gannen;
1229                                 if ($gy_gannen == 1)
1230                                         $gy_offset = '元';
1231                                 $gy_offset = '明治'.$gy_offset;
1232                         } 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))) {
1233                                 # Taishō period
1234                                 $gy_gannen = $gy - 1912 + 1;
1235                                 $gy_offset = $gy_gannen;
1236                                 if ($gy_gannen == 1)
1237                                         $gy_offset = '元';
1238                                 $gy_offset = '大正'.$gy_offset;
1239                         } else if ((($gy == 1926) && ($gm == 12) && ($gd >= 26)) || (($gy > 1926) && ($gy < 1989)) || (($gy == 1989) && ($gm == 1) && ($gd < 8))) {
1240                                 # Shōwa period
1241                                 $gy_gannen = $gy - 1926 + 1;
1242                                 $gy_offset = $gy_gannen;
1243                                 if ($gy_gannen == 1)
1244                                         $gy_offset = '元';
1245                                 $gy_offset = '昭和'.$gy_offset;
1246                         } else {
1247                                 # Heisei period
1248                                 $gy_gannen = $gy - 1989 + 1;
1249                                 $gy_offset = $gy_gannen;
1250                                 if ($gy_gannen == 1)
1251                                         $gy_offset = '元';
1252                                 $gy_offset = '平成'.$gy_offset;
1253                         }
1254                 } else {
1255                         $gy_offset = $gy;
1256                 }
1257
1258                 return array( $gy_offset, $gm, $gd );
1259         }
1260
1261         /**
1262          * Roman number formatting up to 3000
1263          */
1264         static function romanNumeral( $num ) {
1265                 static $table = array(
1266                         array( '', 'I', 'II', 'III', 'IV', 'V', 'VI', 'VII', 'VIII', 'IX', 'X' ),
1267                         array( '', 'X', 'XX', 'XXX', 'XL', 'L', 'LX', 'LXX', 'LXXX', 'XC', 'C' ),
1268                         array( '', 'C', 'CC', 'CCC', 'CD', 'D', 'DC', 'DCC', 'DCCC', 'CM', 'M' ),
1269                         array( '', 'M', 'MM', 'MMM' )
1270                 );
1271
1272                 $num = intval( $num );
1273                 if ( $num > 3000 || $num <= 0 ) {
1274                         return $num;
1275                 }
1276
1277                 $s = '';
1278                 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1279                         if ( $num >= $pow10 ) {
1280                                 $s .= $table[$i][floor($num / $pow10)];
1281                         }
1282                         $num = $num % $pow10;
1283                 }
1284                 return $s;
1285         }
1286
1287         /**
1288          * Hebrew Gematria number formatting up to 9999
1289          */
1290         static function hebrewNumeral( $num ) {
1291                 static $table = array(
1292                         array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' ),
1293                         array( '', 'י', 'כ', 'ל', 'מ', 'נ', 'ס', 'ע', 'פ', 'צ', 'ק' ),
1294                         array( '', 'ק', 'ר', 'ש', 'ת', 'תק', 'תר', 'תש', 'תת', 'תתק', 'תתר' ),
1295                         array( '', 'א', 'ב', 'ג', 'ד', 'ה', 'ו', 'ז', 'ח', 'ט', 'י' )
1296                 );
1297
1298                 $num = intval( $num );
1299                 if ( $num > 9999 || $num <= 0 ) {
1300                         return $num;
1301                 }
1302
1303                 $s = '';
1304                 for ( $pow10 = 1000, $i = 3; $i >= 0; $pow10 /= 10, $i-- ) {
1305                         if ( $num >= $pow10 ) {
1306                                 if ( $num == 15 || $num == 16 ) {
1307                                         $s .= $table[0][9] . $table[0][$num - 9];
1308                                         $num = 0;
1309                                 } else {
1310                                         $s .= $table[$i][intval( ( $num / $pow10 ) )];
1311                                         if( $pow10 == 1000 ) {
1312                                                 $s .= "'";
1313                                         }
1314                                 }
1315                         }
1316                         $num = $num % $pow10;
1317                 }
1318                 if( strlen( $s ) == 2 ) {
1319                         $str = $s . "'";
1320                 } else  {
1321                         $str = substr( $s, 0, strlen( $s ) - 2 ) . '"';
1322                         $str .= substr( $s, strlen( $s ) - 2, 2 );
1323                 }
1324                 $start = substr( $str, 0, strlen( $str ) - 2 );
1325                 $end = substr( $str, strlen( $str ) - 2 );
1326                 switch( $end ) {
1327                         case 'כ':
1328                                 $str = $start . 'ך';
1329                                 break;
1330                         case 'מ':
1331                                 $str = $start . 'ם';
1332                                 break;
1333                         case 'נ':
1334                                 $str = $start . 'ן';
1335                                 break;
1336                         case 'פ':
1337                                 $str = $start . 'ף';
1338                                 break;
1339                         case 'צ':
1340                                 $str = $start . 'ץ';
1341                                 break;
1342                 }
1343                 return $str;
1344         }
1345
1346         /**
1347          * This is meant to be used by time(), date(), and timeanddate() to get
1348          * the date preference they're supposed to use, it should be used in
1349          * all children.
1350          *
1351          *<code>
1352          * function timeanddate([...], $format = true) {
1353          *      $datePreference = $this->dateFormat($format);
1354          * [...]
1355          * }
1356          *</code>
1357          *
1358          * @param $usePrefs Mixed: if true, the user's preference is used
1359          *                         if false, the site/language default is used
1360          *                         if int/string, assumed to be a format.
1361          * @return string
1362          */
1363         function dateFormat( $usePrefs = true ) {
1364                 global $wgUser;
1365
1366                 if( is_bool( $usePrefs ) ) {
1367                         if( $usePrefs ) {
1368                                 $datePreference = $wgUser->getDatePreference();
1369                         } else {
1370                                 $options = User::getDefaultOptions();
1371                                 $datePreference = (string)$options['date'];
1372                         }
1373                 } else {
1374                         $datePreference = (string)$usePrefs;
1375                 }
1376
1377                 // return int
1378                 if( $datePreference == '' ) {
1379                         return 'default';
1380                 }
1381
1382                 return $datePreference;
1383         }
1384
1385         /**
1386          * Get a format string for a given type and preference
1387          * @param $type May be date, time or both
1388          * @param $pref The format name as it appears in Messages*.php
1389          */
1390         function getDateFormatString( $type, $pref ) {
1391                 if ( !isset( $this->dateFormatStrings[$type][$pref] ) ) {
1392                         if ( $pref == 'default' ) {
1393                                 $pref = $this->getDefaultDateFormat();
1394                                 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1395                         } else {
1396                                 $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1397                                 if ( is_null( $df ) ) {
1398                                         $pref = $this->getDefaultDateFormat();
1399                                         $df = self::$dataCache->getSubitem( $this->mCode, 'dateFormats', "$pref $type" );
1400                                 }
1401                         }
1402                         $this->dateFormatStrings[$type][$pref] = $df;
1403                 }
1404                 return $this->dateFormatStrings[$type][$pref];
1405         }
1406
1407         /**
1408          * @param $ts Mixed: the time format which needs to be turned into a
1409          *            date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1410          * @param $adj Bool: whether to adjust the time output according to the
1411          *             user configured offset ($timecorrection)
1412          * @param $format Mixed: true to use user's date format preference
1413          * @param $timecorrection String: the time offset as returned by
1414          *                        validateTimeZone() in Special:Preferences
1415          * @return string
1416          */
1417         function date( $ts, $adj = false, $format = true, $timecorrection = false ) {
1418                 if ( $adj ) {
1419                         $ts = $this->userAdjust( $ts, $timecorrection );
1420                 }
1421                 $df = $this->getDateFormatString( 'date', $this->dateFormat( $format ) );
1422                 return $this->sprintfDate( $df, $ts );
1423         }
1424
1425         /**
1426          * @param $ts Mixed: the time format which needs to be turned into a
1427          *            date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1428          * @param $adj Bool: whether to adjust the time output according to the
1429          *             user configured offset ($timecorrection)
1430          * @param $format Mixed: true to use user's date format preference
1431          * @param $timecorrection String: the time offset as returned by
1432          *                        validateTimeZone() in Special:Preferences
1433          * @return string
1434          */
1435         function time( $ts, $adj = false, $format = true, $timecorrection = false ) {
1436                 if ( $adj ) {
1437                         $ts = $this->userAdjust( $ts, $timecorrection );
1438                 }
1439                 $df = $this->getDateFormatString( 'time', $this->dateFormat( $format ) );
1440                 return $this->sprintfDate( $df, $ts );
1441         }
1442
1443         /**
1444          * @param $ts Mixed: the time format which needs to be turned into a
1445          *            date('YmdHis') format with wfTimestamp(TS_MW,$ts)
1446          * @param $adj Bool: whether to adjust the time output according to the
1447          *             user configured offset ($timecorrection)
1448          * @param $format Mixed: what format to return, if it's false output the
1449          *                default one (default true)
1450          * @param $timecorrection String: the time offset as returned by
1451          *                        validateTimeZone() in Special:Preferences
1452          * @return string
1453          */
1454         function timeanddate( $ts, $adj = false, $format = true, $timecorrection = false) {
1455                 $ts = wfTimestamp( TS_MW, $ts );
1456                 if ( $adj ) {
1457                         $ts = $this->userAdjust( $ts, $timecorrection );
1458                 }
1459                 $df = $this->getDateFormatString( 'both', $this->dateFormat( $format ) );
1460                 return $this->sprintfDate( $df, $ts );
1461         }
1462
1463         function getMessage( $key ) {
1464                 return self::$dataCache->getSubitem( $this->mCode, 'messages', $key );
1465         }
1466
1467         function getAllMessages() {
1468                 return self::$dataCache->getItem( $this->mCode, 'messages' );
1469         }
1470
1471         function iconv( $in, $out, $string ) {
1472                 # This is a wrapper for iconv in all languages except esperanto,
1473                 # which does some nasty x-conversions beforehand
1474
1475                 # Even with //IGNORE iconv can whine about illegal characters in
1476                 # *input* string. We just ignore those too.
1477                 # REF: http://bugs.php.net/bug.php?id=37166
1478                 # REF: https://bugzilla.wikimedia.org/show_bug.cgi?id=16885
1479                 wfSuppressWarnings();
1480                 $text = iconv( $in, $out . '//IGNORE', $string );
1481                 wfRestoreWarnings();
1482                 return $text;
1483         }
1484
1485         // callback functions for uc(), lc(), ucwords(), ucwordbreaks()
1486         function ucwordbreaksCallbackAscii($matches){
1487                 return $this->ucfirst($matches[1]);
1488         }
1489
1490         function ucwordbreaksCallbackMB($matches){
1491                 return mb_strtoupper($matches[0]);
1492         }
1493
1494         function ucCallback($matches){
1495                 list( $wikiUpperChars ) = self::getCaseMaps();
1496                 return strtr( $matches[1], $wikiUpperChars );
1497         }
1498
1499         function lcCallback($matches){
1500                 list( , $wikiLowerChars ) = self::getCaseMaps();
1501                 return strtr( $matches[1], $wikiLowerChars );
1502         }
1503
1504         function ucwordsCallbackMB($matches){
1505                 return mb_strtoupper($matches[0]);
1506         }
1507
1508         function ucwordsCallbackWiki($matches){
1509                 list( $wikiUpperChars ) = self::getCaseMaps();
1510                 return strtr( $matches[0], $wikiUpperChars );
1511         }
1512
1513         function ucfirst( $str ) {
1514                 $o = ord( $str );
1515                 if ( $o < 96 ) {
1516                         return $str;
1517                 } elseif ( $o < 128 ) {
1518                         return ucfirst($str);
1519                 } else {
1520                         // fall back to more complex logic in case of multibyte strings
1521                         return self::uc($str,true);
1522                 }
1523         }
1524
1525         function uc( $str, $first = false ) {
1526                 if ( function_exists( 'mb_strtoupper' ) ) {
1527                         if ( $first ) {
1528                                 if ( self::isMultibyte( $str ) ) {
1529                                         return mb_strtoupper( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1530                                 } else {
1531                                         return ucfirst( $str );
1532                                 }
1533                         } else {
1534                                 return self::isMultibyte( $str ) ? mb_strtoupper( $str ) : strtoupper( $str );
1535                         }
1536                 } else {
1537                         if ( self::isMultibyte( $str ) ) {
1538                                 list( $wikiUpperChars ) = $this->getCaseMaps();
1539                                 $x = $first ? '^' : '';
1540                                 return preg_replace_callback(
1541                                         "/$x([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1542                                         array($this,"ucCallback"),
1543                                         $str
1544                                 );
1545                         } else {
1546                                 return $first ? ucfirst( $str ) : strtoupper( $str );
1547                         }
1548                 }
1549         }
1550
1551         function lcfirst( $str ) {
1552                 $o = ord( $str );
1553                 if ( !$o ) {
1554                         return strval( $str );
1555                 } elseif ( $o >= 128 ) {
1556                         return self::lc( $str, true );
1557                 } elseif ( $o > 96 ) {
1558                         return $str;
1559                 } else {
1560                         $str[0] = strtolower( $str[0] );
1561                         return $str;
1562                 }
1563         }
1564
1565         function lc( $str, $first = false ) {
1566                 if ( function_exists( 'mb_strtolower' ) )
1567                         if ( $first )
1568                                 if ( self::isMultibyte( $str ) )
1569                                         return mb_strtolower( mb_substr( $str, 0, 1 ) ) . mb_substr( $str, 1 );
1570                                 else
1571                                         return strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 );
1572                         else
1573                                 return self::isMultibyte( $str ) ? mb_strtolower( $str ) : strtolower( $str );
1574                 else
1575                         if ( self::isMultibyte( $str ) ) {
1576                                 list( , $wikiLowerChars ) = self::getCaseMaps();
1577                                 $x = $first ? '^' : '';
1578                                 return preg_replace_callback(
1579                                         "/$x([A-Z]|[\\xc0-\\xff][\\x80-\\xbf]*)/",
1580                                         array($this,"lcCallback"),
1581                                         $str
1582                                 );
1583                         } else
1584                                 return $first ? strtolower( substr( $str, 0, 1 ) ) . substr( $str, 1 ) : strtolower( $str );
1585         }
1586
1587         function isMultibyte( $str ) {
1588                 return (bool)preg_match( '/[\x80-\xff]/', $str );
1589         }
1590
1591         function ucwords($str) {
1592                 if ( self::isMultibyte( $str ) ) {
1593                         $str = self::lc($str);
1594
1595                         // regexp to find first letter in each word (i.e. after each space)
1596                         $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)| ([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1597
1598                         // function to use to capitalize a single char
1599                         if ( function_exists( 'mb_strtoupper' ) )
1600                                 return preg_replace_callback(
1601                                         $replaceRegexp,
1602                                         array($this,"ucwordsCallbackMB"),
1603                                         $str
1604                                 );
1605                         else
1606                                 return preg_replace_callback(
1607                                         $replaceRegexp,
1608                                         array($this,"ucwordsCallbackWiki"),
1609                                         $str
1610                                 );
1611                 }
1612                 else
1613                         return ucwords( strtolower( $str ) );
1614         }
1615
1616   # capitalize words at word breaks
1617         function ucwordbreaks($str){
1618                 if (self::isMultibyte( $str ) ) {
1619                         $str = self::lc($str);
1620
1621                         // since \b doesn't work for UTF-8, we explicitely define word break chars
1622                         $breaks= "[ \-\(\)\}\{\.,\?!]";
1623
1624                         // find first letter after word break
1625                         $replaceRegexp = "/^([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)|$breaks([a-z]|[\\xc0-\\xff][\\x80-\\xbf]*)/";
1626
1627                         if ( function_exists( 'mb_strtoupper' ) )
1628                                 return preg_replace_callback(
1629                                         $replaceRegexp,
1630                                         array($this,"ucwordbreaksCallbackMB"),
1631                                         $str
1632                                 );
1633                         else
1634                                 return preg_replace_callback(
1635                                         $replaceRegexp,
1636                                         array($this,"ucwordsCallbackWiki"),
1637                                         $str
1638                                 );
1639                 }
1640                 else
1641                         return preg_replace_callback(
1642                         '/\b([\w\x80-\xff]+)\b/',
1643                         array($this,"ucwordbreaksCallbackAscii"),
1644                         $str );
1645         }
1646
1647         /**
1648          * Return a case-folded representation of $s
1649          *
1650          * This is a representation such that caseFold($s1)==caseFold($s2) if $s1
1651          * and $s2 are the same except for the case of their characters. It is not
1652          * necessary for the value returned to make sense when displayed.
1653          *
1654          * Do *not* perform any other normalisation in this function. If a caller
1655          * uses this function when it should be using a more general normalisation
1656          * function, then fix the caller.
1657          */
1658         function caseFold( $s ) {
1659                 return $this->uc( $s );
1660         }
1661
1662         function checkTitleEncoding( $s ) {
1663                 if( is_array( $s ) ) {
1664                         wfDebugDieBacktrace( 'Given array to checkTitleEncoding.' );
1665                 }
1666                 # Check for non-UTF-8 URLs
1667                 $ishigh = preg_match( '/[\x80-\xff]/', $s);
1668                 if(!$ishigh) return $s;
1669
1670                 $isutf8 = preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1671                 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})+$/', $s );
1672                 if( $isutf8 ) return $s;
1673
1674                 return $this->iconv( $this->fallback8bitEncoding(), "utf-8", $s );
1675         }
1676
1677         function fallback8bitEncoding() {
1678                 return self::$dataCache->getItem( $this->mCode, 'fallback8bitEncoding' );
1679         }
1680
1681         /**
1682          * Most writing systems use whitespace to break up words.
1683          * Some languages such as Chinese don't conventionally do this,
1684          * which requires special handling when breaking up words for
1685          * searching etc.
1686          */
1687         function hasWordBreaks() {
1688                 return true;
1689         }
1690
1691         /**
1692          * Some languages such as Chinese require word segmentation,
1693          * Specify such segmentation when overridden in derived class.
1694          *
1695          * @param $string String
1696          * @return String
1697          */
1698         function wordSegmentation( $string ) {
1699                 return $string;
1700         }
1701
1702         /**
1703          * Some languages have special punctuation need to be normalized.
1704          * Make such changes here.
1705          *
1706          * @param $string String
1707          * @return String
1708          */
1709         function normalizeForSearch( $string ) {
1710                 return $string;
1711         }
1712
1713         /**
1714          * convert double-width roman characters to single-width.
1715          * range: ff00-ff5f ~= 0020-007f
1716          */
1717         protected static function convertDoubleWidth( $string ) {
1718                 $string = preg_replace( '/\xef\xbc([\x80-\xbf])/e', 'chr((ord("$1") & 0x3f) + 0x20)', $string );
1719                 $string = preg_replace( '/\xef\xbd([\x80-\x99])/e', 'chr((ord("$1") & 0x3f) + 0x60)', $string );
1720                 return $string;
1721         }
1722
1723         protected static function insertSpace( $string, $pattern ) {
1724                 $string = preg_replace( $pattern, " $1 ", $string );
1725                 $string = preg_replace( '/ +/', ' ', $string );
1726                 return $string;
1727         }
1728
1729         function convertForSearchResult( $termsArray ) {
1730                 # some languages, e.g. Chinese, need to do a conversion
1731                 # in order for search results to be displayed correctly
1732                 return $termsArray;
1733         }
1734
1735         /**
1736          * Get the first character of a string.
1737          *
1738          * @param $s string
1739          * @return string
1740          */
1741         function firstChar( $s ) {
1742                 $matches = array();
1743                 preg_match( '/^([\x00-\x7f]|[\xc0-\xdf][\x80-\xbf]|' .
1744                 '[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf7][\x80-\xbf]{3})/', $s, $matches);
1745
1746                 if ( isset( $matches[1] ) ) {
1747                         if ( strlen( $matches[1] ) != 3 ) {
1748                                 return $matches[1];
1749                         }
1750
1751                         // Break down Hangul syllables to grab the first jamo
1752                         $code = utf8ToCodepoint( $matches[1] );
1753                         if ( $code < 0xac00 || 0xd7a4 <= $code) {
1754                                 return $matches[1];
1755                         } elseif ( $code < 0xb098 ) {
1756                                 return "\xe3\x84\xb1";
1757                         } elseif ( $code < 0xb2e4 ) {
1758                                 return "\xe3\x84\xb4";
1759                         } elseif ( $code < 0xb77c ) {
1760                                 return "\xe3\x84\xb7";
1761                         } elseif ( $code < 0xb9c8 ) {
1762                                 return "\xe3\x84\xb9";
1763                         } elseif ( $code < 0xbc14 ) {
1764                                 return "\xe3\x85\x81";
1765                         } elseif ( $code < 0xc0ac ) {
1766                                 return "\xe3\x85\x82";
1767                         } elseif ( $code < 0xc544 ) {
1768                                 return "\xe3\x85\x85";
1769                         } elseif ( $code < 0xc790 ) {
1770                                 return "\xe3\x85\x87";
1771                         } elseif ( $code < 0xcc28 ) {
1772                                 return "\xe3\x85\x88";
1773                         } elseif ( $code < 0xce74 ) {
1774                                 return "\xe3\x85\x8a";
1775                         } elseif ( $code < 0xd0c0 ) {
1776                                 return "\xe3\x85\x8b";
1777                         } elseif ( $code < 0xd30c ) {
1778                                 return "\xe3\x85\x8c";
1779                         } elseif ( $code < 0xd558 ) {
1780                                 return "\xe3\x85\x8d";
1781                         } else {
1782                                 return "\xe3\x85\x8e";
1783                         }
1784                 } else {
1785                         return "";
1786                 }
1787         }
1788
1789         function initEncoding() {
1790                 # Some languages may have an alternate char encoding option
1791                 # (Esperanto X-coding, Japanese furigana conversion, etc)
1792                 # If this language is used as the primary content language,
1793                 # an override to the defaults can be set here on startup.
1794         }
1795
1796         function recodeForEdit( $s ) {
1797                 # For some languages we'll want to explicitly specify
1798                 # which characters make it into the edit box raw
1799                 # or are converted in some way or another.
1800                 # Note that if wgOutputEncoding is different from
1801                 # wgInputEncoding, this text will be further converted
1802                 # to wgOutputEncoding.
1803                 global $wgEditEncoding;
1804                 if( $wgEditEncoding == '' or
1805                   $wgEditEncoding == 'UTF-8' ) {
1806                         return $s;
1807                 } else {
1808                         return $this->iconv( 'UTF-8', $wgEditEncoding, $s );
1809                 }
1810         }
1811
1812         function recodeInput( $s ) {
1813                 # Take the previous into account.
1814                 global $wgEditEncoding;
1815                 if($wgEditEncoding != "") {
1816                         $enc = $wgEditEncoding;
1817                 } else {
1818                         $enc = 'UTF-8';
1819                 }
1820                 if( $enc == 'UTF-8' ) {
1821                         return $s;
1822                 } else {
1823                         return $this->iconv( $enc, 'UTF-8', $s );
1824                 }
1825         }
1826
1827         /**
1828          * Convert a UTF-8 string to normal form C. In Malayalam and Arabic, this
1829          * also cleans up certain backwards-compatible sequences, converting them
1830          * to the modern Unicode equivalent.
1831          *
1832          * This is language-specific for performance reasons only.
1833          */
1834         function normalize( $s ) {
1835                 return UtfNormal::cleanUp( $s );
1836         }
1837
1838         /**
1839          * Transform a string using serialized data stored in the given file (which
1840          * must be in the serialized subdirectory of $IP). The file contains pairs
1841          * mapping source characters to destination characters.
1842          *
1843          * The data is cached in process memory. This will go faster if you have the
1844          * FastStringSearch extension.
1845          */
1846         function transformUsingPairFile( $file, $string ) {
1847                 if ( !isset( $this->transformData[$file] ) ) {
1848                         $data = wfGetPrecompiledData( $file );
1849                         if ( $data === false ) {
1850                                 throw new MWException( __METHOD__.": The transformation file $file is missing" );
1851                         }
1852                         $this->transformData[$file] = new ReplacementArray( $data );
1853                 }
1854                 return $this->transformData[$file]->replace( $string );
1855         }
1856
1857         /**
1858          * For right-to-left language support
1859          *
1860          * @return bool
1861          */
1862         function isRTL() {
1863                 return self::$dataCache->getItem( $this->mCode, 'rtl' );
1864         }
1865
1866         /**
1867          * Return the correct HTML 'dir' attribute value for this language.
1868          * @return String
1869          */
1870         function getDir() {
1871                 return $this->isRTL() ? 'rtl' : 'ltr';
1872         }
1873
1874         /**
1875          * Return 'left' or 'right' as appropriate alignment for line-start
1876          * for this language's text direction.
1877          *
1878          * Should be equivalent to CSS3 'start' text-align value....
1879          *
1880          * @return String
1881          */
1882         function alignStart() {
1883                 return $this->isRTL() ? 'right' : 'left';
1884         }
1885
1886         /**
1887          * Return 'right' or 'left' as appropriate alignment for line-end
1888          * for this language's text direction.
1889          *
1890          * Should be equivalent to CSS3 'end' text-align value....
1891          *
1892          * @return String
1893          */
1894         function alignEnd() {
1895                 return $this->isRTL() ? 'left' : 'right';
1896         }
1897
1898         /**
1899          * A hidden direction mark (LRM or RLM), depending on the language direction
1900          *
1901          * @return string
1902          */
1903         function getDirMark() {
1904                 return $this->isRTL() ? "\xE2\x80\x8F" : "\xE2\x80\x8E";
1905         }
1906
1907         function capitalizeAllNouns() {
1908                 return self::$dataCache->getItem( $this->mCode, 'capitalizeAllNouns' );
1909         }
1910
1911         /**
1912          * An arrow, depending on the language direction
1913          *
1914          * @return string
1915          */
1916         function getArrow() {
1917                 return $this->isRTL() ? '←' : '→';
1918         }
1919
1920         /**
1921          * To allow "foo[[bar]]" to extend the link over the whole word "foobar"
1922          *
1923          * @return bool
1924          */
1925         function linkPrefixExtension() {
1926                 return self::$dataCache->getItem( $this->mCode, 'linkPrefixExtension' );
1927         }
1928
1929         function getMagicWords() {
1930                 return self::$dataCache->getItem( $this->mCode, 'magicWords' );
1931         }
1932
1933         # Fill a MagicWord object with data from here
1934         function getMagic( $mw ) {
1935                 if ( !$this->mMagicHookDone ) {
1936                         $this->mMagicHookDone = true;
1937                         wfProfileIn( 'LanguageGetMagic' );
1938                         wfRunHooks( 'LanguageGetMagic', array( &$this->mMagicExtensions, $this->getCode() ) );
1939                         wfProfileOut( 'LanguageGetMagic' );
1940                 }
1941                 if ( isset( $this->mMagicExtensions[$mw->mId] ) ) {
1942                         $rawEntry = $this->mMagicExtensions[$mw->mId];
1943                 } else {
1944                         $magicWords = $this->getMagicWords();
1945                         if ( isset( $magicWords[$mw->mId] ) ) {
1946                                 $rawEntry = $magicWords[$mw->mId];
1947                         } else {
1948                                 $rawEntry = false;
1949                         }
1950                 }
1951
1952                 if( !is_array( $rawEntry ) ) {
1953                         error_log( "\"$rawEntry\" is not a valid magic thingie for \"$mw->mId\"" );
1954                 } else {
1955                         $mw->mCaseSensitive = $rawEntry[0];
1956                         $mw->mSynonyms = array_slice( $rawEntry, 1 );
1957                 }
1958         }
1959
1960         /**
1961          * Add magic words to the extension array
1962          */
1963         function addMagicWordsByLang( $newWords ) {
1964                 $code = $this->getCode();
1965                 $fallbackChain = array();
1966                 while ( $code && !in_array( $code, $fallbackChain ) ) {
1967                         $fallbackChain[] = $code;
1968                         $code = self::getFallbackFor( $code );
1969                 }
1970                 if ( !in_array( 'en', $fallbackChain ) ) {
1971                         $fallbackChain[] = 'en';
1972                 }
1973                 $fallbackChain = array_reverse( $fallbackChain );
1974                 foreach ( $fallbackChain as $code ) {
1975                         if ( isset( $newWords[$code] ) ) {
1976                                 $this->mMagicExtensions = $newWords[$code] + $this->mMagicExtensions;
1977                         }
1978                 }
1979         }
1980
1981         /**
1982          * Get special page names, as an associative array
1983          *   case folded alias => real name
1984          */
1985         function getSpecialPageAliases() {
1986                 // Cache aliases because it may be slow to load them
1987                 if ( is_null( $this->mExtendedSpecialPageAliases ) ) {
1988                         // Initialise array
1989                         $this->mExtendedSpecialPageAliases =
1990                                 self::$dataCache->getItem( $this->mCode, 'specialPageAliases' );
1991                         wfRunHooks( 'LanguageGetSpecialPageAliases',
1992                                 array( &$this->mExtendedSpecialPageAliases, $this->getCode() ) );
1993                 }
1994
1995                 return $this->mExtendedSpecialPageAliases;
1996         }
1997
1998         /**
1999          * Italic is unsuitable for some languages
2000          *
2001          * @param $text String: the text to be emphasized.
2002          * @return string
2003          */
2004         function emphasize( $text ) {
2005                 return "<em>$text</em>";
2006         }
2007
2008          /**
2009           * Normally we output all numbers in plain en_US style, that is
2010           * 293,291.235 for twohundredninetythreethousand-twohundredninetyone
2011           * point twohundredthirtyfive. However this is not sutable for all
2012           * languages, some such as Pakaran want ੨੯੩,੨੯੫.੨੩੫ and others such as
2013           * Icelandic just want to use commas instead of dots, and dots instead
2014           * of commas like "293.291,235".
2015           *
2016           * An example of this function being called:
2017           * <code>
2018           * wfMsg( 'message', $wgLang->formatNum( $num ) )
2019           * </code>
2020           *
2021           * See LanguageGu.php for the Gujarati implementation and
2022           * $separatorTransformTable on MessageIs.php for
2023           * the , => . and . => , implementation.
2024           *
2025           * @todo check if it's viable to use localeconv() for the decimal
2026           *       separator thing.
2027           * @param $number Mixed: the string to be formatted, should be an integer
2028           *        or a floating point number.
2029           * @param $nocommafy Bool: set to true for special numbers like dates
2030           * @return string
2031           */
2032         function formatNum( $number, $nocommafy = false ) {
2033                 global $wgTranslateNumerals;
2034                 if (!$nocommafy) {
2035                         $number = $this->commafy($number);
2036                         $s = $this->separatorTransformTable();
2037                         if ($s) { $number = strtr($number, $s); }
2038                 }
2039
2040                 if ($wgTranslateNumerals) {
2041                         $s = $this->digitTransformTable();
2042                         if ($s) { $number = strtr($number, $s); }
2043                 }
2044
2045                 return $number;
2046         }
2047
2048         function parseFormattedNumber( $number ) {
2049                 $s = $this->digitTransformTable();
2050                 if ($s) { $number = strtr($number, array_flip($s)); }
2051
2052                 $s = $this->separatorTransformTable();
2053                 if ($s) { $number = strtr($number, array_flip($s)); }
2054
2055                 $number = strtr( $number, array (',' => '') );
2056                 return $number;
2057         }
2058
2059         /**
2060          * Adds commas to a given number
2061          *
2062          * @param $_ mixed
2063          * @return string
2064          */
2065         function commafy($_) {
2066                 return strrev((string)preg_replace('/(\d{3})(?=\d)(?!\d*\.)/','$1,',strrev($_)));
2067         }
2068
2069         function digitTransformTable() {
2070                 return self::$dataCache->getItem( $this->mCode, 'digitTransformTable' );
2071         }
2072
2073         function separatorTransformTable() {
2074                 return self::$dataCache->getItem( $this->mCode, 'separatorTransformTable' );
2075         }
2076
2077
2078         /**
2079          * Take a list of strings and build a locale-friendly comma-separated
2080          * list, using the local comma-separator message.
2081          * The last two strings are chained with an "and".
2082          *
2083          * @param $l Array
2084          * @return string
2085          */
2086         function listToText( $l ) {
2087                 $s = '';
2088                 $m = count( $l ) - 1;
2089                 if( $m == 1 ) {
2090                         return $l[0] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $l[1];
2091                 }
2092                 else {
2093                         for ( $i = $m; $i >= 0; $i-- ) {
2094                                 if ( $i == $m ) {
2095                                         $s = $l[$i];
2096                                 } else if( $i == $m - 1 ) {
2097                                         $s = $l[$i] . $this->getMessageFromDB( 'and' ) . $this->getMessageFromDB( 'word-separator' ) . $s;
2098                                 } else {
2099                                         $s = $l[$i] . $this->getMessageFromDB( 'comma-separator' ) . $s;
2100                                 }
2101                         }
2102                         return $s;
2103                 }
2104         }
2105
2106         /**
2107          * Take a list of strings and build a locale-friendly comma-separated
2108          * list, using the local comma-separator message.
2109          * @param $list array of strings to put in a comma list
2110          * @return string
2111          */
2112         function commaList( $list ) {
2113                 return implode(
2114                         $list,
2115                         wfMsgExt( 'comma-separator', array( 'parsemag', 'escapenoentities', 'language' => $this ) ) );
2116         }
2117
2118         /**
2119          * Take a list of strings and build a locale-friendly semicolon-separated
2120          * list, using the local semicolon-separator message.
2121          * @param $list array of strings to put in a semicolon list
2122          * @return string
2123          */
2124         function semicolonList( $list ) {
2125                 return implode(
2126                         $list,
2127                         wfMsgExt( 'semicolon-separator', array( 'parsemag', 'escapenoentities', 'language' => $this ) ) );
2128         }
2129
2130         /**
2131          * Same as commaList, but separate it with the pipe instead.
2132          * @param $list array of strings to put in a pipe list
2133          * @return string
2134          */
2135         function pipeList( $list ) {
2136                 return implode(
2137                         $list,
2138                         wfMsgExt( 'pipe-separator', array( 'escapenoentities', 'language' => $this ) ) );
2139         }
2140
2141         /**
2142          * Truncate a string to a specified length in bytes, appending an optional
2143          * string (e.g. for ellipses)
2144          *
2145          * The database offers limited byte lengths for some columns in the database;
2146          * multi-byte character sets mean we need to ensure that only whole characters
2147          * are included, otherwise broken characters can be passed to the user
2148          *
2149          * If $length is negative, the string will be truncated from the beginning
2150          *
2151          * @param $string String to truncate
2152          * @param $length Int: maximum length (excluding ellipses)
2153          * @param $ellipsis String to append to the truncated text
2154          * @return string
2155          */
2156         function truncate( $string, $length, $ellipsis = '...' ) {
2157                 # Use the localized ellipsis character
2158                 if ( $ellipsis == '...' ) {
2159                         $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2160                 }
2161                 # Check if there is no need to truncate
2162                 if ( $length == 0 ) {
2163                         return $ellipsis;
2164                 } elseif ( strlen( $string ) <= abs( $length ) ) {
2165                         return $string;
2166                 }
2167                 $stringOriginal = $string;
2168                 if ( $length > 0 ) {
2169                         $string = substr( $string, 0, $length ); // xyz...
2170                         $string = $this->removeBadCharLast( $string );
2171                         $string = $string . $ellipsis;
2172                 } else {
2173                         $string = substr( $string, $length ); // ...xyz
2174                         $string = $this->removeBadCharFirst( $string );
2175                         $string = $ellipsis . $string;
2176                 }
2177                 # Do not truncate if the ellipsis makes the string longer/equal (bug 22181)
2178                 if ( strlen( $string ) < strlen( $stringOriginal ) ) {
2179                         return $string;
2180                 } else {
2181                         return $stringOriginal;
2182                 }
2183         }
2184
2185         /**
2186          * Remove bytes that represent an incomplete Unicode character
2187          * at the end of string (e.g. bytes of the char are missing)
2188          *
2189          * @param $string String
2190          * @return string
2191          */
2192         protected function removeBadCharLast( $string ) {
2193                 $char = ord( $string[strlen( $string ) - 1] );
2194                 $m = array();
2195                 if ( $char >= 0xc0 ) {
2196                         # We got the first byte only of a multibyte char; remove it.
2197                         $string = substr( $string, 0, -1 );
2198                 } elseif ( $char >= 0x80 &&
2199                       preg_match( '/^(.*)(?:[\xe0-\xef][\x80-\xbf]|' .
2200                                   '[\xf0-\xf7][\x80-\xbf]{1,2})$/', $string, $m ) )
2201                 {
2202                         # We chopped in the middle of a character; remove it
2203                         $string = $m[1];
2204                 }
2205                 return $string;
2206         }
2207
2208         /**
2209          * Remove bytes that represent an incomplete Unicode character
2210          * at the start of string (e.g. bytes of the char are missing)
2211          *
2212          * @param $string String
2213          * @return string
2214          */
2215         protected function removeBadCharFirst( $string ) {
2216                 $char = ord( $string[0] );
2217                 if ( $char >= 0x80 && $char < 0xc0 ) {
2218                         # We chopped in the middle of a character; remove the whole thing
2219                         $string = preg_replace( '/^[\x80-\xbf]+/', '', $string );
2220                 }
2221                 return $string;
2222         }
2223
2224         /*
2225          * Truncate a string of valid HTML to a specified length in bytes,
2226          * appending an optional string (e.g. for ellipses), and return valid HTML
2227          *
2228          * This is only intended for styled/linked text, such as HTML with
2229          * tags like <span> and <a>, were the tags are self-contained (valid HTML)
2230          *
2231          * Note: tries to fix broken HTML with MWTidy
2232          *
2233          * @param string $text String to truncate
2234          * @param int $length (zero/positive) Maximum length (excluding ellipses)
2235          * @param string $ellipsis String to append to the truncated text
2236          * @returns string
2237          */
2238         function truncateHtml( $text, $length, $ellipsis = '...' ) {
2239                 # Use the localized ellipsis character
2240                 if ( $ellipsis == '...' ) {
2241                         $ellipsis = wfMsgExt( 'ellipsis', array( 'escapenoentities', 'language' => $this ) );
2242                 }
2243                 # Check if there is no need to truncate
2244                 if ( $length <= 0 ) {
2245                         return $ellipsis; // no text shown, nothing to format
2246                 } elseif ( strlen($text) <= $length ) {
2247                         return $text; // string short enough even *with* HTML
2248                 }
2249                 $text = MWTidy::tidy( $text ); // fix tags
2250                 $displayLen = 0; // innerHTML legth so far
2251                 $testingEllipsis = false; // checking if ellipses will make string longer/equal?
2252                 $tagType = 0; // 0-open, 1-close
2253                 $bracketState = 0; // 1-tag start, 2-tag name, 0-neither
2254                 $entityState = 0; // 0-not entity, 1-entity
2255                 $tag = $ret = $ch = '';
2256                 $openTags = array();
2257                 $textLen = strlen($text);
2258                 for( $pos = 0; $pos < $textLen; ++$pos ) {
2259                         $ch = $text[$pos];
2260                         $lastCh = $pos ? $text[$pos-1] : '';
2261                         $ret .= $ch; // add to result string
2262                         if ( $ch == '<' ) {
2263                                 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags ); // for bad HTML
2264                                 $entityState = 0; // for bad HTML
2265                                 $bracketState = 1; // tag started (checking for backslash)
2266                         } elseif ( $ch == '>' ) {
2267                                 $this->truncate_endBracket( $tag, $tagType, $lastCh, $openTags );
2268                                 $entityState = 0; // for bad HTML
2269                                 $bracketState = 0; // out of brackets
2270                         } elseif ( $bracketState == 1 ) {
2271                                 if ( $ch == '/' ) {
2272                                         $tagType = 1; // close tag (e.g. "</span>")
2273                                 } else {
2274                                         $tagType = 0; // open tag (e.g. "<span>")
2275                                         $tag .= $ch;
2276                                 }
2277                                 $bracketState = 2; // building tag name
2278                         } elseif ( $bracketState == 2 ) {
2279                                 if ( $ch != ' ' ) {
2280                                         $tag .= $ch;
2281                                 } else {
2282                                         // Name found (e.g. "<a href=..."), add on tag attributes...
2283                                         $pos += $this->truncate_skip( $ret, $text, "<>", $pos + 1 );
2284                                 }
2285                         } elseif ( $bracketState == 0 ) {
2286                                 if ( $entityState ) {
2287                                         if ( $ch == ';' ) {
2288                                                 $entityState = 0;
2289                                                 $displayLen++; // entity is one displayed char
2290                                         }
2291                                 } else {
2292                                         if ( $ch == '&' ) {
2293                                                 $entityState = 1; // entity found, (e.g. "&nbsp;")
2294                                         } else {
2295                                                 $displayLen++; // this char is displayed
2296                                                 // Add on the other display text after this...
2297                                                 $skipped = $this->truncate_skip(
2298                                                         $ret, $text, "<>&", $pos + 1, $length - $displayLen );
2299                                                 $displayLen += $skipped;
2300                                                 $pos += $skipped;
2301                                         }
2302                                 }
2303                         }
2304                         # Consider truncation once the display length has reached the maximim.
2305                         # Double-check that we're not in the middle of a bracket/entity...
2306                         if ( $displayLen >= $length && $bracketState == 0 && $entityState == 0 ) {
2307                                 if ( !$testingEllipsis ) {
2308                                         $testingEllipsis = true;
2309                                         # Save were we are; we will truncate here unless
2310                                         # the ellipsis actually makes the string longer.
2311                                         $pOpenTags = $openTags; // save state
2312                                         $pRet = $ret; // save state
2313                                 } elseif ( $displayLen > ($length + strlen($ellipsis)) ) {
2314                                         # Ellipsis won't make string longer/equal, the truncation point was OK.
2315                                         $openTags = $pOpenTags; // reload state
2316                                         $ret = $this->removeBadCharLast( $pRet ); // reload state, multi-byte char fix
2317                                         $ret .= $ellipsis; // add ellipsis
2318                                         break;
2319                                 }
2320                         }
2321                 }
2322                 if ( $displayLen == 0 ) {
2323                         return ''; // no text shown, nothing to format
2324                 }
2325                 $this->truncate_endBracket( $tag, $text[$textLen-1], $tagType, $openTags ); // for bad HTML
2326                 while ( count( $openTags ) > 0 ) {
2327                         $ret .= '</' . array_pop( $openTags ) . '>'; // close open tags
2328                 }
2329                 return $ret;
2330         }
2331
2332         // truncateHtml() helper function
2333         // like strcspn() but adds the skipped chars to $ret
2334         private function truncate_skip( &$ret, $text, $search, $start, $len = -1 ) {
2335                 $skipCount = 0;
2336                 if( $start < strlen($text) ) {
2337                         $skipCount = strcspn( $text, $search, $start, $len );
2338                         $ret .= substr( $text, $start, $skipCount );
2339                 }
2340                 return $skipCount;
2341         }
2342
2343         // truncateHtml() helper function
2344         // (a) push or pop $tag from $openTags as needed
2345         // (b) clear $tag value
2346         private function truncate_endBracket( &$tag, $tagType, $lastCh, &$openTags ) {
2347                 $tag = ltrim( $tag );
2348                 if( $tag != '' ) {
2349                         if( $tagType == 0 && $lastCh != '/' ) {
2350                                 $openTags[] = $tag; // tag opened (didn't close itself)
2351                         } else if( $tagType == 1 ) {
2352                                 if( $openTags && $tag == $openTags[count($openTags)-1] ) {
2353                                         array_pop( $openTags ); // tag closed
2354                                 }
2355                         }
2356                         $tag = '';
2357                 }
2358         }
2359
2360         /**
2361          * Grammatical transformations, needed for inflected languages
2362          * Invoked by putting {{grammar:case|word}} in a message
2363          *
2364          * @param $word string
2365          * @param $case string
2366          * @return string
2367          */
2368         function convertGrammar( $word, $case ) {
2369                 global $wgGrammarForms;
2370                 if ( isset($wgGrammarForms[$this->getCode()][$case][$word]) ) {
2371                         return $wgGrammarForms[$this->getCode()][$case][$word];
2372                 }
2373                 return $word;
2374         }
2375
2376         /**
2377          * Provides an alternative text depending on specified gender.
2378          * Usage {{gender:username|masculine|feminine|neutral}}.
2379          * username is optional, in which case the gender of current user is used,
2380          * but only in (some) interface messages; otherwise default gender is used.
2381          * If second or third parameter are not specified, masculine is used.
2382          * These details may be overriden per language.
2383          */
2384         function gender( $gender, $forms ) {
2385                 if ( !count($forms) ) { return ''; }
2386                 $forms = $this->preConvertPlural( $forms, 2 );
2387                 if ( $gender === 'male' ) return $forms[0];
2388                 if ( $gender === 'female' ) return $forms[1];
2389                 return isset($forms[2]) ? $forms[2] : $forms[0];
2390         }
2391
2392         /**
2393          * Plural form transformations, needed for some languages.
2394          * For example, there are 3 form of plural in Russian and Polish,
2395          * depending on "count mod 10". See [[w:Plural]]
2396          * For English it is pretty simple.
2397          *
2398          * Invoked by putting {{plural:count|wordform1|wordform2}}
2399          * or {{plural:count|wordform1|wordform2|wordform3}}
2400          *
2401          * Example: {{plural:{{NUMBEROFARTICLES}}|article|articles}}
2402          *
2403          * @param $count Integer: non-localized number
2404          * @param $forms Array: different plural forms
2405          * @return string Correct form of plural for $count in this language
2406          */
2407         function convertPlural( $count, $forms ) {
2408                 if ( !count($forms) ) { return ''; }
2409                 $forms = $this->preConvertPlural( $forms, 2 );
2410
2411                 return ( $count == 1 ) ? $forms[0] : $forms[1];
2412         }
2413
2414         /**
2415          * Checks that convertPlural was given an array and pads it to requested
2416          * amound of forms by copying the last one.
2417          *
2418          * @param $count Integer: How many forms should there be at least
2419          * @param $forms Array of forms given to convertPlural
2420          * @return array Padded array of forms or an exception if not an array
2421          */
2422         protected function preConvertPlural( /* Array */ $forms, $count ) {
2423                 while ( count($forms) < $count ) {
2424                         $forms[] = $forms[count($forms)-1];
2425                 }
2426                 return $forms;
2427         }
2428
2429         /**
2430          * For translaing of expiry times
2431          * @param $str String: the validated block time in English
2432          * @return Somehow translated block time
2433          * @see LanguageFi.php for example implementation
2434          */
2435         function translateBlockExpiry( $str ) {
2436
2437                 $scBlockExpiryOptions = $this->getMessageFromDB( 'ipboptions' );
2438
2439                 if ( $scBlockExpiryOptions == '-') {
2440                         return $str;
2441                 }
2442
2443                 foreach (explode(',', $scBlockExpiryOptions) as $option) {
2444                         if ( strpos($option, ":") === false )
2445                                 continue;
2446                         list($show, $value) = explode(":", $option);
2447                         if ( strcmp ( $str, $value) == 0 ) {
2448                                 return htmlspecialchars( trim( $show ) );
2449                         }
2450                 }
2451
2452                 return $str;
2453         }
2454
2455         /**
2456          * languages like Chinese need to be segmented in order for the diff
2457          * to be of any use
2458          *
2459          * @param $text String
2460          * @return String
2461          */
2462         function segmentForDiff( $text ) {
2463                 return $text;
2464         }
2465
2466         /**
2467          * and unsegment to show the result
2468          *
2469          * @param $text String
2470          * @return String
2471          */
2472         function unsegmentForDiff( $text ) {
2473                 return $text;
2474         }
2475
2476         # convert text to all supported variants
2477         function autoConvertToAllVariants($text) {
2478                 return $this->mConverter->autoConvertToAllVariants($text);
2479         }
2480
2481         # convert text to different variants of a language.
2482         function convert( $text, $isTitle = false) {
2483                 return $this->mConverter->convert($text, $isTitle);
2484         }
2485
2486         # Check if this is a language with variants
2487         function hasVariants(){
2488                 return sizeof($this->getVariants())>1;
2489         }
2490
2491         # Put custom tags (e.g. -{ }-) around math to prevent conversion
2492         function armourMath($text){
2493                 return $this->mConverter->armourMath($text);
2494         }
2495
2496
2497         /**
2498          * Perform output conversion on a string, and encode for safe HTML output.
2499          * @param $text String text to be converted
2500          * @param $isTitle Bool whether this conversion is for the article title
2501          * @return string
2502          * @todo this should get integrated somewhere sane
2503          */
2504         function convertHtml( $text, $isTitle = false ) {
2505                 return htmlspecialchars( $this->convert( $text, $isTitle ) );
2506         }
2507
2508         function convertCategoryKey( $key ) {
2509                 return $this->mConverter->convertCategoryKey( $key );
2510         }
2511
2512         /**
2513          * get the list of variants supported by this langauge
2514          * see sample implementation in LanguageZh.php
2515          *
2516          * @return array an array of language codes
2517          */
2518         function getVariants() {
2519                 return $this->mConverter->getVariants();
2520         }
2521
2522
2523         function getPreferredVariant( $fromUser = true, $fromHeader = false ) {
2524                 return $this->mConverter->getPreferredVariant( $fromUser, $fromHeader );
2525         }
2526
2527         /**
2528          * if a language supports multiple variants, it is
2529          * possible that non-existing link in one variant
2530          * actually exists in another variant. this function
2531          * tries to find it. See e.g. LanguageZh.php
2532          *
2533          * @param $link String: the name of the link
2534          * @param $nt Mixed: the title object of the link
2535          * @param $ignoreOtherCond Boolean: to disable other conditions when
2536          *      we need to transclude a template or update a category's link
2537          * @return null the input parameters may be modified upon return
2538          */
2539         function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
2540                 $this->mConverter->findVariantLink( $link, $nt, $ignoreOtherCond );
2541         }
2542
2543         /**
2544          * If a language supports multiple variants, converts text
2545          * into an array of all possible variants of the text:
2546          *  'variant' => text in that variant
2547          */
2548         function convertLinkToAllVariants($text){
2549                 return $this->mConverter->convertLinkToAllVariants($text);
2550         }
2551
2552
2553         /**
2554          * returns language specific options used by User::getPageRenderHash()
2555          * for example, the preferred language variant
2556          *
2557          * @return string
2558          */
2559         function getExtraHashOptions() {
2560                 return $this->mConverter->getExtraHashOptions();
2561         }
2562
2563         /**
2564          * for languages that support multiple variants, the title of an
2565          * article may be displayed differently in different variants. this
2566          * function returns the apporiate title defined in the body of the article.
2567          *
2568          * @return string
2569          */
2570         function getParsedTitle() {
2571                 return $this->mConverter->getParsedTitle();
2572         }
2573
2574         /**
2575          * Enclose a string with the "no conversion" tag. This is used by
2576          * various functions in the Parser
2577          *
2578          * @param $text String: text to be tagged for no conversion
2579          * @param $noParse
2580          * @return string the tagged text
2581          */
2582         function markNoConversion( $text, $noParse=false ) {
2583                 return $this->mConverter->markNoConversion( $text, $noParse );
2584         }
2585
2586         /**
2587          * A regular expression to match legal word-trailing characters
2588          * which should be merged onto a link of the form [[foo]]bar.
2589          *
2590          * @return string
2591          */
2592         function linkTrail() {
2593                 return self::$dataCache->getItem( $this->mCode, 'linkTrail' );
2594         }
2595
2596         function getLangObj() {
2597                 return $this;
2598         }
2599
2600         /**
2601          * Get the RFC 3066 code for this language object
2602          */
2603         function getCode() {
2604                 return $this->mCode;
2605         }
2606
2607         function setCode( $code ) {
2608                 $this->mCode = $code;
2609         }
2610
2611         /**
2612          * Get the name of a file for a certain language code
2613          * @param $prefix string Prepend this to the filename
2614          * @param $code string Language code
2615          * @param $suffix string Append this to the filename
2616          * @return string $prefix . $mangledCode . $suffix
2617          */
2618         static function getFileName( $prefix = 'Language', $code, $suffix = '.php' ) {
2619                 return $prefix . str_replace( '-', '_', ucfirst( $code ) ) . $suffix;
2620         }
2621
2622         /**
2623          * Get the language code from a file name. Inverse of getFileName()
2624          * @param $filename string $prefix . $languageCode . $suffix
2625          * @param $prefix string Prefix before the language code
2626          * @param $suffix string Suffix after the language code
2627          * @return Language code, or false if $prefix or $suffix isn't found
2628          */
2629         static function getCodeFromFileName( $filename, $prefix = 'Language', $suffix = '.php' ) {
2630                 $m = null;
2631                 preg_match( '/' . preg_quote( $prefix, '/' ) . '([A-Z][a-z_]+)' .
2632                         preg_quote( $suffix, '/' ) . '/', $filename, $m );
2633                 if ( !count( $m ) ) {
2634                         return false;
2635                 }
2636                 return str_replace( '_', '-', strtolower( $m[1] ) );
2637         }
2638
2639         static function getMessagesFileName( $code ) {
2640                 global $IP;
2641                 return self::getFileName( "$IP/languages/messages/Messages", $code, '.php' );
2642         }
2643
2644         static function getClassFileName( $code ) {
2645                 global $IP;
2646                 return self::getFileName( "$IP/languages/classes/Language", $code, '.php' );
2647         }
2648
2649         /**
2650          * Get the fallback for a given language
2651          */
2652         static function getFallbackFor( $code ) {
2653                 if ( $code === 'en' ) {
2654                         // Shortcut
2655                         return false;
2656                 } else {
2657                         return self::getLocalisationCache()->getItem( $code, 'fallback' );
2658                 }
2659         }
2660
2661         /**
2662          * Get all messages for a given language
2663          * WARNING: this may take a long time
2664          */
2665         static function getMessagesFor( $code ) {
2666                 return self::getLocalisationCache()->getItem( $code, 'messages' );
2667         }
2668
2669         /**
2670          * Get a message for a given language
2671          */
2672         static function getMessageFor( $key, $code ) {
2673                 return self::getLocalisationCache()->getSubitem( $code, 'messages', $key );
2674         }
2675
2676         function fixVariableInNamespace( $talk ) {
2677                 if ( strpos( $talk, '$1' ) === false ) return $talk;
2678
2679                 global $wgMetaNamespace;
2680                 $talk = str_replace( '$1', $wgMetaNamespace, $talk );
2681
2682                 # Allow grammar transformations
2683                 # Allowing full message-style parsing would make simple requests
2684                 # such as action=raw much more expensive than they need to be.
2685                 # This will hopefully cover most cases.
2686                 $talk = preg_replace_callback( '/{{grammar:(.*?)\|(.*?)}}/i',
2687                         array( &$this, 'replaceGrammarInNamespace' ), $talk );
2688                 return str_replace( ' ', '_', $talk );
2689         }
2690
2691         function replaceGrammarInNamespace( $m ) {
2692                 return $this->convertGrammar( trim( $m[2] ), trim( $m[1] ) );
2693         }
2694
2695         static function getCaseMaps() {
2696                 static $wikiUpperChars, $wikiLowerChars;
2697                 if ( isset( $wikiUpperChars ) ) {
2698                         return array( $wikiUpperChars, $wikiLowerChars );
2699                 }
2700
2701                 wfProfileIn( __METHOD__ );
2702                 $arr = wfGetPrecompiledData( 'Utf8Case.ser' );
2703                 if ( $arr === false ) {
2704                         throw new MWException(
2705                                 "Utf8Case.ser is missing, please run \"make\" in the serialized directory\n" );
2706                 }
2707                 extract( $arr );
2708                 wfProfileOut( __METHOD__ );
2709                 return array( $wikiUpperChars, $wikiLowerChars );
2710         }
2711
2712         function formatTimePeriod( $seconds ) {
2713                 if ( $seconds < 10 ) {
2714                         return $this->formatNum( sprintf( "%.1f", $seconds ) ) . wfMsg( 'seconds-abbrev' );
2715                 } elseif ( $seconds < 60 ) {
2716                         return $this->formatNum( round( $seconds ) ) . wfMsg( 'seconds-abbrev' );
2717                 } elseif ( $seconds < 3600 ) {
2718                         return $this->formatNum( floor( $seconds / 60 ) ) . wfMsg( 'minutes-abbrev' ) .
2719                                 $this->formatNum( round( fmod( $seconds, 60 ) ) ) . wfMsg( 'seconds-abbrev' );
2720                 } else {
2721                         $hours = floor( $seconds / 3600 );
2722                         $minutes = floor( ( $seconds - $hours * 3600 ) / 60 );
2723                         $secondsPart = round( $seconds - $hours * 3600 - $minutes * 60 );
2724                         return $this->formatNum( $hours ) . wfMsg( 'hours-abbrev' ) .
2725                                 $this->formatNum( $minutes ) . wfMsg( 'minutes-abbrev' ) .
2726                                 $this->formatNum( $secondsPart ) . wfMsg( 'seconds-abbrev' );
2727                 }
2728         }
2729
2730         function formatBitrate( $bps ) {
2731                 $units = array( 'bps', 'kbps', 'Mbps', 'Gbps' );
2732                 if ( $bps <= 0 ) {
2733                         return $this->formatNum( $bps ) . $units[0];
2734                 }
2735                 $unitIndex = floor( log10( $bps ) / 3 );
2736                 $mantissa = $bps / pow( 1000, $unitIndex );
2737                 if ( $mantissa < 10 ) {
2738                         $mantissa = round( $mantissa, 1 );
2739                 } else {
2740                         $mantissa = round( $mantissa );
2741                 }
2742                 return $this->formatNum( $mantissa ) . $units[$unitIndex];
2743         }
2744
2745         /**
2746          * Format a size in bytes for output, using an appropriate
2747          * unit (B, KB, MB or GB) according to the magnitude in question
2748          *
2749          * @param $size Size to format
2750          * @return string Plain text (not HTML)
2751          */
2752         function formatSize( $size ) {
2753                 // For small sizes no decimal places necessary
2754                 $round = 0;
2755                 if( $size > 1024 ) {
2756                         $size = $size / 1024;
2757                         if( $size > 1024 ) {
2758                                 $size = $size / 1024;
2759                                 // For MB and bigger two decimal places are smarter
2760                                 $round = 2;
2761                                 if( $size > 1024 ) {
2762                                         $size = $size / 1024;
2763                                         $msg = 'size-gigabytes';
2764                                 } else {
2765                                         $msg = 'size-megabytes';
2766                                 }
2767                         } else {
2768                                 $msg = 'size-kilobytes';
2769                         }
2770                 } else {
2771                         $msg = 'size-bytes';
2772                 }
2773                 $size = round( $size, $round );
2774                 $text = $this->getMessageFromDB( $msg );
2775                 return str_replace( '$1', $this->formatNum( $size ), $text );
2776         }
2777
2778         /**
2779          * Get the conversion rule title, if any.
2780          */
2781         function getConvRuleTitle() {
2782                 return $this->mConverter->getConvRuleTitle();
2783         }
2784 }