includes/Parser.php

   1 <?php
   2
   3 // require_once('Tokenizer.php');
   4
   5 # PHP Parser
   6 #
   7 # Processes wiki markup
   8 #
   9 # There are two main entry points into the Parser class: parse() and preSaveTransform().
  10 # The parse() function produces HTML output, preSaveTransform() produces altered wiki markup.
  11 #
  12 # Globals used:
  13 #    objects:   $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
  14 #
  15 # NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
  16 #
  17 #    settings:  $wgUseTex*, $wgUseCategoryMagic*, $wgUseDynamicDates*, $wgInterwikiMagic*,
  18 #               $wgNamespacesWithSubpages, $wgLanguageCode, $wgAllowExternalImages*,
  19 #               $wgLocaltimezone
  20 #
  21 #      * only within ParserOptions
  22 #
  23 #
  24 #----------------------------------------
  25 #    Variable substitution O(N^2) attack
  26 #-----------------------------------------
  27 # Without countermeasures, it would be possible to attack the parser by saving a page
  28 # filled with a large number of inclusions of large pages. The size of the generated
  29 # page would be proportional to the square of the input size. Hence, we limit the number
  30 # of inclusions of any given page, thus bringing any attack back to O(N).
  31 #
  32
  33 define( "MAX_INCLUDE_REPEAT", 5 );
  34
  35 # Allowed values for $mOutputType
  36 define( "OT_HTML", 1 );
  37 define( "OT_WIKI", 2 );
  38 define( "OT_MSG", 3 );
  39
  40 # string parameter for extractTags which will cause it
  41 # to strip HTML comments in addition to regular
  42 # <XML>-style tags. This should not be anything we
  43 # may want to use in wikisyntax
  44 define( "STRIP_COMMENTS", "HTMLCommentStrip" );
  45
  46 # prefix for escaping, used in two functions at least
  47 define( "UNIQ_PREFIX", "NaodW29");
  48
  49 class Parser
  50 {
  51         # Persistent:
  52         var $mTagHooks;
  53
  54         # Cleared with clearState():
  55         var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
  56         var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
  57
  58         # Temporary:
  59         var $mOptions, $mTitle, $mOutputType;
  60
  61         function Parser() {
  62                 $this->mTagHooks = array();
  63                 $this->clearState();
  64         }
  65
  66         function clearState() {
  67                 $this->mOutput = new ParserOutput;
  68                 $this->mAutonumber = 0;
  69                 $this->mLastSection = "";
  70                 $this->mDTopen = false;
  71                 $this->mVariables = false;
  72                 $this->mIncludeCount = array();
  73                 $this->mStripState = array();
  74                 $this->mArgStack = array();
  75                 $this->mInPre = false;
  76         }
  77
  78         # First pass--just handle <nowiki> sections, pass the rest off
  79         # to internalParse() which does all the real work.
  80         #
  81         # Returns a ParserOutput
  82         #
  83         function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
  84                 global $wgUseTidy;
  85                 $fname = "Parser::parse";
  86                 wfProfileIn( $fname );
  87
  88                 if ( $clearState ) {
  89                         $this->clearState();
  90                 }
  91
  92                 $this->mOptions = $options;
  93                 $this->mTitle =& $title;
  94                 $this->mOutputType = OT_HTML;
  95
  96                 $stripState = NULL;
  97                 $text = $this->strip( $text, $this->mStripState );
  98                 $text = $this->internalParse( $text, $linestart );
  99                 $text = $this->unstrip( $text, $this->mStripState );
 100                 # Clean up special characters, only run once, next-to-last before doBlockLevels
 101                 if(!$wgUseTidy) {
 102                         $fixtags = array(
 103                                 # french spaces, last one Guillemet-left
 104                                 # only if there is something before the space
 105                                 '/(.) (\\?|:|!|\\302\\273)/i' => '\\1&nbsp;\\2',
 106                                 # french spaces, Guillemet-right
 107                                 "/(\\302\\253) /i"=>"\\1&nbsp;",
 108                                 '/<hr *>/i' => '<hr />',
 109                                 '/<br *>/i' => '<br />',
 110                                 '/<center *>/i' => '<div class="center">',
 111                                 '/<\\/center *>/i' => '</div>',
 112                                 # Clean up spare ampersands; note that we probably ought to be
 113                                 # more careful about named entities.
 114                                 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&amp;'
 115                         );
 116                         $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
 117                 } else {
 118                         $fixtags = array(
 119                                 # french spaces, last one Guillemet-left
 120                                 '/ (\\?|:|!|\\302\\273)/i' => '&nbsp;\\1',
 121                                 # french spaces, Guillemet-right
 122                                 '/(\\302\\253) /i' => '\\1&nbsp;',
 123                                 '/<center *>/i' => '<div class="center">',
 124                                 '/<\\/center *>/i' => '</div>'
 125                         );
 126                         $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
 127                 }
 128                 # only once and last
 129                 $text = $this->doBlockLevels( $text, $linestart );
 130                 $text = $this->unstripNoWiki( $text, $this->mStripState );
 131                 if($wgUseTidy) {
 132                         $text = $this->tidy($text);
 133                 }
 134                 $this->mOutput->setText( $text );
 135                 wfProfileOut( $fname );
 136                 return $this->mOutput;
 137         }
 138
 139         /* static */ function getRandomString() {
 140                 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
 141         }
 142
 143         # Replaces all occurrences of <$tag>content</$tag> in the text
 144         # with a random marker and returns the new text. the output parameter
 145         # $content will be an associative array filled with data on the form
 146         # $unique_marker => content.
 147
 148         # If $content is already set, the additional entries will be appended
 149
 150         # If $tag is set to STRIP_COMMENTS, the function will extract
 151         # <!-- HTML comments -->
 152
 153         /* static */ function extractTags($tag, $text, &$content, $uniq_prefix = ""){
 154                 $rnd = $uniq_prefix . '-' . $tag . Parser::getRandomString();
 155                 if ( !$content ) {
 156                         $content = array( );
 157                 }
 158                 $n = 1;
 159                 $stripped = '';
 160
 161                 while ( '' != $text ) {
 162                         if($tag==STRIP_COMMENTS) {
 163                                 $p = preg_split( '/<!--/i', $text, 2 );
 164                         } else {
 165                                 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
 166                         }
 167                         $stripped .= $p[0];
 168                         if ( ( count( $p ) < 2 ) || ( '' == $p[1] ) ) {
 169                                 $text = '';
 170                         } else {
 171                                 if($tag==STRIP_COMMENTS) {
 172                                         $q = preg_split( '/-->/i', $p[1], 2 );
 173                                 } else {
 174                                         $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
 175                                 }
 176                                 $marker = $rnd . sprintf('%08X', $n++);
 177                                 $content[$marker] = $q[0];
 178                                 $stripped .= $marker;
 179                                 $text = $q[1];
 180                         }
 181                 }
 182                 return $stripped;
 183         }
 184
 185         # Strips and renders <nowiki>, <pre>, <math>, <hiero>
 186         # If $render is set, performs necessary rendering operations on plugins
 187         # Returns the text, and fills an array with data needed in unstrip()
 188         # If the $state is already a valid strip state, it adds to the state
 189
 190         # When $stripcomments is set, HTML comments <!-- like this -->
 191         # will be stripped in addition to other tags. This is important
 192         # for section editing, where these comments cause confusion when
 193         # counting the sections in the wikisource
 194         function strip( $text, &$state, $stripcomments = false ) {
 195                 $render = ($this->mOutputType == OT_HTML);
 196                 $nowiki_content = array();
 197                 $math_content = array();
 198                 $pre_content = array();
 199                 $comment_content = array();
 200                 $ext_content = array();
 201
 202                 # Replace any instances of the placeholders
 203                 $uniq_prefix = UNIQ_PREFIX;
 204                 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
 205
 206
 207                 # nowiki
 208                 $text = Parser::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
 209                 foreach( $nowiki_content as $marker => $content ){
 210                         if( $render ){
 211                                 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
 212                         } else {
 213                                 $nowiki_content[$marker] = "<nowiki>$content</nowiki>";
 214                         }
 215                 }
 216
 217                 # math
 218                 $text = Parser::extractTags('math', $text, $math_content, $uniq_prefix);
 219                 foreach( $math_content as $marker => $content ){
 220                         if( $render ) {
 221                                 if( $this->mOptions->getUseTeX() ) {
 222                                         $math_content[$marker] = renderMath( $content );
 223                                 } else {
 224                                         $math_content[$marker] = "&lt;math&gt;$content&lt;math&gt;";
 225                                 }
 226                         } else {
 227                                 $math_content[$marker] = "<math>$content</math>";
 228                         }
 229                 }
 230
 231                 # pre
 232                 $text = Parser::extractTags('pre', $text, $pre_content, $uniq_prefix);
 233                 foreach( $pre_content as $marker => $content ){
 234                         if( $render ){
 235                                 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
 236                         } else {
 237                                 $pre_content[$marker] = "<pre>$content</pre>";
 238                         }
 239                 }
 240
 241                 # Comments
 242                 if($stripcomments) {
 243                         $text = Parser::extractTags(STRIP_COMMENTS, $text, $comment_content, $uniq_prefix);
 244                         foreach( $comment_content as $marker => $content ){
 245                                 $comment_content[$marker] = "<!--$content-->";
 246                         }
 247                 }
 248
 249                 # Extensions
 250                 foreach ( $this->mTagHooks as $tag => $callback ) {
 251                         $ext_contents[$tag] = array();
 252                         $text = Parser::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
 253                         foreach( $ext_content[$tag] as $marker => $content ) {
 254                                 if ( $render ) {
 255                                         $ext_content[$tag][$marker] = $callback( $content );
 256                                 } else {
 257                                         $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
 258                                 }
 259                         }
 260                 }
 261
 262                 # Merge state with the pre-existing state, if there is one
 263                 if ( $state ) {
 264                         $state['nowiki'] = $state['nowiki'] + $nowiki_content;
 265                         $state['math'] = $state['math'] + $math_content;
 266                         $state['pre'] = $state['pre'] + $pre_content;
 267                         $state['comment'] = $state['comment'] + $comment_content;
 268
 269                         foreach( $ext_content as $tag => $array ) {
 270                                 if ( array_key_exists( $tag, $state ) ) {
 271                                         $state[$tag] = $state[$tag] + $array;
 272                                 }
 273                         }
 274                 } else {
 275                         $state = array(
 276                           'nowiki' => $nowiki_content,
 277                           'math' => $math_content,
 278                           'pre' => $pre_content,
 279                           'comment' => $comment_content,
 280                         ) + $ext_content;
 281                 }
 282                 return $text;
 283         }
 284
 285         # always call unstripNoWiki() after this one
 286         function unstrip( $text, &$state ) {
 287                 # Must expand in reverse order, otherwise nested tags will be corrupted
 288                 $contentDict = end( $state );
 289                 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
 290                         if( key($state) != 'nowiki') {
 291                                 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
 292                                         $text = str_replace( key( $contentDict ), $content, $text );
 293                                 }
 294                         }
 295                 }
 296
 297                 return $text;
 298         }
 299         # always call this after unstrip() to preserve the order
 300         function unstripNoWiki( $text, &$state ) {
 301                 # Must expand in reverse order, otherwise nested tags will be corrupted
 302                 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
 303                         $text = str_replace( key( $state['nowiki'] ), $content, $text );
 304                 }
 305
 306                 return $text;
 307         }
 308
 309         # Add an item to the strip state
 310         # Returns the unique tag which must be inserted into the stripped text
 311         # The tag will be replaced with the original text in unstrip()
 312
 313         function insertStripItem( $text, &$state ) {
 314                 $rnd = UNIQ_PREFIX . '-item' . Parser::getRandomString();
 315                 if ( !$state ) {
 316                         $state = array(
 317                           'nowiki' => array(),
 318                           'math' => array(),
 319                           'pre' => array()
 320                         );
 321                 }
 322                 $state['item'][$rnd] = $text;
 323                 return $rnd;
 324         }
 325
 326         # categoryMagic
 327         # generate a list of subcategories and pages for a category
 328         # depending on wfMsg("usenewcategorypage") it either calls the new
 329         # or the old code. The new code will not work properly for some
 330         # languages due to sorting issues, so they might want to turn it
 331         # off.
 332         function categoryMagic() {
 333                 $msg = wfMsg('usenewcategorypage');
 334                 if ( '0' == @$msg[0] )
 335                 {
 336                         return $this->oldCategoryMagic();
 337                 } else {
 338                         return $this->newCategoryMagic();
 339                 }
 340         }
 341
 342         # This method generates the list of subcategories and pages for a category
 343         function oldCategoryMagic () {
 344                 global $wgLang , $wgUser ;
 345                 if ( !$this->mOptions->getUseCategoryMagic() ) return ; # Doesn't use categories at all
 346
 347                 $cns = Namespace::getCategory() ;
 348                 if ( $this->mTitle->getNamespace() != $cns ) return "" ; # This ain't a category page
 349
 350                 $r = "<br style=\"clear:both;\"/>\n";
 351
 352
 353                 $sk =& $wgUser->getSkin() ;
 354
 355                 $articles = array() ;
 356                 $children = array() ;
 357                 $data = array () ;
 358                 $id = $this->mTitle->getArticleID() ;
 359
 360                 # FIXME: add limits
 361                 $t = wfStrencode( $this->mTitle->getDBKey() );
 362                 $sql = "SELECT DISTINCT cur_title,cur_namespace FROM cur,categorylinks WHERE cl_to='$t' AND cl_from=cur_id ORDER BY cl_sortkey" ;
 363                 $res = wfQuery ( $sql, DB_READ ) ;
 364                 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
 365
 366                 # For all pages that link to this category
 367                 foreach ( $data AS $x )
 368                 {
 369                         $t = $wgLang->getNsText ( $x->cur_namespace ) ;
 370                         if ( $t != "" ) $t .= ":" ;
 371                         $t .= $x->cur_title ;
 372
 373                         if ( $x->cur_namespace == $cns ) {
 374                                 array_push ( $children , $sk->makeLink ( $t ) ) ; # Subcategory
 375                         } else {
 376                                 array_push ( $articles , $sk->makeLink ( $t ) ) ; # Page in this category
 377                         }
 378                 }
 379                 wfFreeResult ( $res ) ;
 380
 381                 # Showing subcategories
 382                 if ( count ( $children ) > 0 ) {
 383                         $r .= '<h2>'.wfMsg('subcategories')."</h2>\n" ;
 384                         $r .= implode ( ', ' , $children ) ;
 385                 }
 386
 387                 # Showing pages in this category
 388                 if ( count ( $articles ) > 0 ) {
 389                         $ti = $this->mTitle->getText() ;
 390                         $h =  wfMsg( 'category_header', $ti );
 391                         $r .= "<h2>{$h}</h2>\n" ;
 392                         $r .= implode ( ', ' , $articles ) ;
 393                 }
 394
 395
 396                 return $r ;
 397         }
 398
 399
 400
 401         function newCategoryMagic () {
 402                 global $wgLang , $wgUser ;
 403                 if ( !$this->mOptions->getUseCategoryMagic() ) return ; # Doesn't use categories at all
 404
 405                 $cns = Namespace::getCategory() ;
 406                 if ( $this->mTitle->getNamespace() != $cns ) return '' ; # This ain't a category page
 407
 408                 $r = "<br style=\"clear:both;\"/>\n";
 409
 410
 411                 $sk =& $wgUser->getSkin() ;
 412
 413                 $articles = array() ;
 414                 $articles_start_char = array();
 415                 $children = array() ;
 416                 $children_start_char = array();
 417                 $data = array () ;
 418                 $id = $this->mTitle->getArticleID() ;
 419
 420                 # FIXME: add limits
 421                 $t = wfStrencode( $this->mTitle->getDBKey() );
 422                 $sql = "SELECT DISTINCT cur_title,cur_namespace,cl_sortkey FROM
 423 cur,categorylinks WHERE cl_to='$t' AND cl_from=cur_id ORDER BY
 424 cl_sortkey" ;
 425                 $res = wfQuery ( $sql, DB_READ ) ;
 426                 while ( $x = wfFetchObject ( $res ) )
 427                 {
 428                         $t = $ns = $wgLang->getNsText ( $x->cur_namespace ) ;
 429                         if ( $t != '' ) $t .= ':' ;
 430                         $t .= $x->cur_title ;
 431
 432                         if ( $x->cur_namespace == $cns ) {
 433                                 $ctitle = str_replace( '_',' ',$x->cur_title );
 434                                 array_push ( $children, $sk->makeKnownLink ( $t, $ctitle ) ) ; # Subcategory
 435
 436                                 // If there's a link from Category:A to Category:B, the sortkey of the resulting
 437                                 // entry in the categorylinks table is Category:A, not A, which it SHOULD be.
 438                                 // Workaround: If sortkey == "Category:".$title, than use $title for sorting,
 439                                 // else use sortkey...
 440                                 if ( ($ns.":".$ctitle) ==  $x->cl_sortkey ) {
 441                                         array_push ( $children_start_char, $wgLang->firstChar( $x->cur_title ) );
 442                                 } else {
 443                                         array_push ( $children_start_char, $wgLang->firstChar( $x->cl_sortkey ) ) ;
 444                                 }
 445                         } else {
 446                                 array_push ( $articles , $sk->makeLink ( $t ) ) ; # Page in this category
 447                                 array_push ( $articles_start_char, $wgLang->firstChar( $x->cl_sortkey ) ) ;
 448                         }
 449                 }
 450                 wfFreeResult ( $res ) ;
 451
 452                 $ti = $this->mTitle->getText() ;
 453
 454                 # Don't show subcategories section if there are none.
 455                 if ( count ( $children ) > 0 )
 456                 {
 457                         # Showing subcategories
 458                         $r .= '<h2>' . wfMsg( 'subcategories' ) . "</h2>\n"
 459                                 . wfMsg( 'subcategorycount', count( $children ) );
 460                         if ( count ( $children ) > 6 ) {
 461
 462                                 // divide list into three equal chunks
 463                                 $chunk = (int) (count ( $children ) / 3);
 464
 465                                 // get and display header
 466                                 $r .= '<table width="100%"><tr valign="top">';
 467
 468                                 $startChunk = 0;
 469                                 $endChunk = $chunk;
 470
 471                                 // loop through the chunks
 472                                 for($startChunk = 0, $endChunk = $chunk, $chunkIndex = 0;
 473                                         $chunkIndex < 3;
 474                                         $chunkIndex++, $startChunk = $endChunk, $endChunk += $chunk + 1)
 475                                 {
 476
 477                                         $r .= '<td><ul>';
 478                                         // output all subcategories to category
 479                                         for ($index = $startChunk ;
 480                                                 $index < $endChunk && $index < count($children);
 481                                                 $index++ )
 482                                         {
 483                                                 // check for change of starting letter or begging of chunk
 484                                                 if ( ($children_start_char[$index] != $children_start_char[$index - 1])
 485                                                         || ($index == $startChunk) )
 486                                                 {
 487                                                         $r .= "</ul><h3>{$children_start_char[$index]}</h3>\n<ul>";
 488                                                 }
 489
 490                                                 $r .= "<li>{$children[$index]}</li>";
 491                                         }
 492                                         $r .= '</ul></td>';
 493
 494
 495                                 }
 496                                 $r .= '</tr></table>';
 497                         } else {
 498                                 // for short lists of subcategories to category.
 499
 500                                 $r .= "<h3>{$children_start_char[0]}</h3>\n";
 501                                 $r .= '<ul><li>'.$children[0].'</li>';
 502                                 for ($index = 1; $index < count($children); $index++ )
 503                                 {
 504                                         if ($children_start_char[$index] != $children_start_char[$index - 1])
 505                                         {
 506                                                 $r .= "</ul><h3>{$children_start_char[$index]}</h3>\n<ul>";
 507                                         }
 508
 509                                         $r .= "<li>{$children[$index]}</li>";
 510                                 }
 511                                 $r .= '</ul>';
 512                         }
 513                 } # END of if ( count($children) > 0 )
 514
 515                 $r .= '<h2>' . wfMsg( 'category_header', $ti ) . "</h2>\n" .
 516                         wfMsg( 'categoryarticlecount', count( $articles ) );
 517
 518                 # Showing articles in this category
 519                 if ( count ( $articles ) > 6) {
 520                         $ti = $this->mTitle->getText() ;
 521
 522                         // divide list into three equal chunks
 523                         $chunk = (int) (count ( $articles ) / 3);
 524
 525                         // get and display header
 526                         $r .= '<table width="100%"><tr valign="top">';
 527
 528                         // loop through the chunks
 529                         for($startChunk = 0, $endChunk = $chunk, $chunkIndex = 0;
 530                                 $chunkIndex < 3;
 531                                 $chunkIndex++, $startChunk = $endChunk, $endChunk += $chunk + 1)
 532                         {
 533
 534                                 $r .= '<td><ul>';
 535
 536                                 // output all articles in category
 537                                 for ($index = $startChunk ;
 538                                         $index < $endChunk && $index < count($articles);
 539                                         $index++ )
 540                                 {
 541                                         // check for change of starting letter or begging of chunk
 542                                         if ( ($articles_start_char[$index] != $articles_start_char[$index - 1])
 543                                                 || ($index == $startChunk) )
 544                                         {
 545                                                 $r .= "</ul><h3>{$articles_start_char[$index]}</h3>\n<ul>";
 546                                         }
 547
 548                                         $r .= "<li>{$articles[$index]}</li>";
 549                                 }
 550                                 $r .= '</ul></td>';
 551
 552
 553                         }
 554                         $r .= '</tr></table>';
 555                 } elseif ( count ( $articles )  > 0) {
 556                         // for short lists of articles in categories.
 557                         $ti = $this->mTitle->getText() ;
 558
 559                         $r .= '<h3>'.$articles_start_char[0]."</h3>\n";
 560                         $r .= '<ul><li>'.$articles[0].'</li>';
 561                         for ($index = 1; $index < count($articles); $index++ )
 562                         {
 563                                 if ($articles_start_char[$index] != $articles_start_char[$index - 1])
 564                                 {
 565                                         $r .= "</ul><h3>{$articles_start_char[$index]}</h3>\n<ul>";
 566                                 }
 567
 568                                 $r .= "<li>{$articles[$index]}</li>";
 569                         }
 570                         $r .= '</ul>';
 571                 }
 572
 573
 574                 return $r ;
 575         }
 576
 577         # Return allowed HTML attributes
 578         function getHTMLattrs () {
 579                 $htmlattrs = array( # Allowed attributes--no scripting, etc.
 580                                 'title', 'align', 'lang', 'dir', 'width', 'height',
 581                                 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
 582                                 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
 583                                 /* FONT */ 'type', 'start', 'value', 'compact',
 584                                 /* For various lists, mostly deprecated but safe */
 585                                 'summary', 'width', 'border', 'frame', 'rules',
 586                                 'cellspacing', 'cellpadding', 'valign', 'char',
 587                                 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
 588                                 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
 589                                 'id', 'class', 'name', 'style' /* For CSS */
 590                                 );
 591                 return $htmlattrs ;
 592         }
 593
 594         # Remove non approved attributes and javascript in css
 595         function fixTagAttributes ( $t ) {
 596                 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
 597                 $htmlattrs = $this->getHTMLattrs() ;
 598
 599                 # Strip non-approved attributes from the tag
 600                 $t = preg_replace(
 601                         '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
 602                         "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
 603                         $t);
 604                 # Strip javascript "expression" from stylesheets. Brute force approach:
 605                 # If anythin offensive is found, all attributes of the HTML tag are dropped
 606
 607                 if( preg_match(
 608                         '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
 609                         wfMungeToUtf8( $t ) ) )
 610                 {
 611                         $t='';
 612                 }
 613
 614                 return trim ( $t ) ;
 615         }
 616
 617         # interface with html tidy, used if $wgUseTidy = true
 618         function tidy ( $text ) {
 619                 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
 620                 global $wgInputEncoding, $wgOutputEncoding;
 621                 $fname = 'Parser::tidy';
 622                 wfProfileIn( $fname );
 623
 624                 $cleansource = '';
 625                 switch(strtoupper($wgOutputEncoding)) {
 626                         case 'ISO-8859-1':
 627                                 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)? ' -latin1':' -raw';
 628                                 break;
 629                         case 'UTF-8':
 630                                 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)? ' -utf8':' -raw';
 631                                 break;
 632                         default:
 633                                 $wgTidyOpts .= ' -raw';
 634                         }
 635
 636                 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
 637 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
 638 '<head><title>test</title></head><body>'.$text.'</body></html>';
 639                 $descriptorspec = array(
 640                         0 => array('pipe', 'r'),
 641                         1 => array('pipe', 'w'),
 642                         2 => array('file', '/dev/null', 'a')
 643                 );
 644                 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts", $descriptorspec, $pipes);
 645                 if (is_resource($process)) {
 646                         fwrite($pipes[0], $wrappedtext);
 647                         fclose($pipes[0]);
 648                         while (!feof($pipes[1])) {
 649                                 $cleansource .= fgets($pipes[1], 1024);
 650                         }
 651                         fclose($pipes[1]);
 652                         $return_value = proc_close($process);
 653                 }
 654
 655                 wfProfileOut( $fname );
 656
 657                 if( $cleansource == '' && $text != '') {
 658                         wfDebug( "Tidy error detected!\n" );
 659                         return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
 660                 } else {
 661                         return $cleansource;
 662                 }
 663         }
 664
 665         # parse the wiki syntax used to render tables
 666         function doTableStuff ( $t ) {
 667                 $t = explode ( "\n" , $t ) ;
 668                 $td = array () ; # Is currently a td tag open?
 669                         $ltd = array () ; # Was it TD or TH?
 670                         $tr = array () ; # Is currently a tr tag open?
 671                         $ltr = array () ; # tr attributes
 672                         foreach ( $t AS $k => $x )
 673                         {
 674                                 $x = trim ( $x ) ;
 675                                 $fc = substr ( $x , 0 , 1 ) ;
 676                                 if ( '{|' == substr ( $x , 0 , 2 ) )
 677                                 {
 678                                         $t[$k] = "\n<table " . $this->fixTagAttributes ( substr ( $x , 3 ) ) . '>' ;
 679                                         array_push ( $td , false ) ;
 680                                         array_push ( $ltd , '' ) ;
 681                                         array_push ( $tr , false ) ;
 682                                         array_push ( $ltr , '' ) ;
 683                                 }
 684                                 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
 685                                 else if ( '|}' == substr ( $x , 0 , 2 ) )
 686                                 {
 687                                         $z = "</table>\n" ;
 688                                         $l = array_pop ( $ltd ) ;
 689                                         if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 690                                         if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
 691                                         array_pop ( $ltr ) ;
 692                                         $t[$k] = $z ;
 693                                 }
 694                                 /*      else if ( "|_" == substr ( $x , 0 , 2 ) ) # Caption
 695                                                 {
 696                                                 $z = trim ( substr ( $x , 2 ) ) ;
 697                                                 $t[$k] = "<caption>{$z}</caption>\n" ;
 698                                                 }*/
 699                                 else if ( '|-' == substr ( $x , 0 , 2 ) ) # Allows for |---------------
 700                                 {
 701                                         $x = substr ( $x , 1 ) ;
 702                                         while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
 703                                         $z = '' ;
 704                                         $l = array_pop ( $ltd ) ;
 705                                         if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 706                                         if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
 707                                         array_pop ( $ltr ) ;
 708                                         $t[$k] = $z ;
 709                                         array_push ( $tr , false ) ;
 710                                         array_push ( $td , false ) ;
 711                                         array_push ( $ltd , '' ) ;
 712                                         array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
 713                                 }
 714                                 else if ( '|' == $fc || '!' == $fc || '|+' == substr ( $x , 0 , 2 ) ) # Caption
 715                                 {
 716                                         if ( '|+' == substr ( $x , 0 , 2 ) )
 717                                         {
 718                                                 $fc = '+' ;
 719                                                 $x = substr ( $x , 1 ) ;
 720                                         }
 721                                         $after = substr ( $x , 1 ) ;
 722                                         if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
 723                                         $after = explode ( '||' , $after ) ;
 724                                         $t[$k] = '' ;
 725                                         foreach ( $after AS $theline )
 726                                         {
 727                                                 $z = '' ;
 728                                                 if ( $fc != '+' )
 729                                                 {
 730                                                         $tra = array_pop ( $ltr ) ;
 731                                                         if ( !array_pop ( $tr ) ) $z = "<tr {$tra}>\n" ;
 732                                                         array_push ( $tr , true ) ;
 733                                                         array_push ( $ltr , '' ) ;
 734                                                 }
 735
 736                                                 $l = array_pop ( $ltd ) ;
 737                                                 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
 738                                                 if ( $fc == '|' ) $l = 'td' ;
 739                                                 else if ( $fc == '!' ) $l = 'th' ;
 740                                                 else if ( $fc == '+' ) $l = 'caption' ;
 741                                                 else $l = '' ;
 742                                                 array_push ( $ltd , $l ) ;
 743                                                 $y = explode ( '|' , $theline , 2 ) ;
 744                                                 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
 745                                                 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
 746                                                 $t[$k] .= $y ;
 747                                                 array_push ( $td , true ) ;
 748                                         }
 749                                 }
 750                         }
 751
 752                 # Closing open td, tr && table
 753                 while ( count ( $td ) > 0 )
 754                 {
 755                         if ( array_pop ( $td ) ) $t[] = '</td>' ;
 756                         if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
 757                         $t[] = '</table>' ;
 758                 }
 759
 760                 $t = implode ( "\n" , $t ) ;
 761                 #               $t = $this->removeHTMLtags( $t );
 762                 return $t ;
 763         }
 764
 765         # Parses the text and adds the result to the strip state
 766         # Returns the strip tag
 767         function stripParse( $text, $newline, $args )
 768         {
 769                 $text = $this->strip( $text, $this->mStripState );
 770                 $text = $this->internalParse( $text, (bool)$newline, $args, false );
 771                 return $newline.$this->insertStripItem( $text, $this->mStripState );
 772         }
 773
 774         function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
 775                 $fname = 'Parser::internalParse';
 776                 wfProfileIn( $fname );
 777
 778                 $text = $this->removeHTMLtags( $text );
 779                 $text = $this->replaceVariables( $text, $args );
 780
 781                 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
 782
 783                 $text = $this->doHeadings( $text );
 784                 if($this->mOptions->getUseDynamicDates()) {
 785                         global $wgDateFormatter;
 786                         $text = $wgDateFormatter->reformat( $this->mOptions->getDateFormat(), $text );
 787                 }
 788                 $text = $this->doAllQuotes( $text );
 789                 $text = $this->replaceExternalLinks( $text );
 790                 $text = $this->replaceInternalLinks ( $text );
 791                 $text = $this->replaceInternalLinks ( $text );
 792                 //$text = $this->doTokenizedParser ( $text );
 793                 $text = $this->doTableStuff ( $text ) ;
 794                 $text = $this->magicISBN( $text );
 795                 $text = $this->magicRFC( $text );
 796                 $text = $this->formatHeadings( $text, $isMain );
 797                 $sk =& $this->mOptions->getSkin();
 798                 $text = $sk->transformContent( $text );
 799
 800                 if ( !isset ( $this->categoryMagicDone ) ) {
 801                         $text .= $this->categoryMagic () ;
 802                         $this->categoryMagicDone = true ;
 803                 }
 804
 805                 wfProfileOut( $fname );
 806                 return $text;
 807         }
 808
 809     # Parse headers and return html
 810         /* private */ function doHeadings( $text ) {
 811                 $fname = 'Parser::doHeadings';
 812                 wfProfileIn( $fname );
 813                 for ( $i = 6; $i >= 1; --$i ) {
 814                         $h = substr( '======', 0, $i );
 815                         $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
 816                           "<h{$i}>\\1</h{$i}>\\2", $text );
 817                 }
 818                 wfProfileOut( $fname );
 819                 return $text;
 820         }
 821
 822         /* private */ function doAllQuotes( $text ) {
 823                 $fname = 'Parser::doAllQuotes';
 824                 wfProfileIn( $fname );
 825                 $outtext = '';
 826                 $lines = explode( "\n", $text );
 827                 foreach ( $lines as $line ) {
 828                         $outtext .= $this->doQuotes ( '', $line, '' ) . "\n";
 829                 }
 830                 $outtext = substr($outtext, 0,-1);
 831                 wfProfileOut( $fname );
 832                 return $outtext;
 833         }
 834
 835         /* private */ function doQuotes( $pre, $text, $mode ) {
 836                 if ( preg_match( "/^(.*)''(.*)$/sU", $text, $m ) ) {
 837                         $m1_strong = ($m[1] == "") ? "" : "<strong>{$m[1]}</strong>";
 838                         $m1_em = ($m[1] == "") ? "" : "<em>{$m[1]}</em>";
 839                         if ( substr ($m[2], 0, 1) == '\'' ) {
 840                                 $m[2] = substr ($m[2], 1);
 841                                 if ($mode == 'em') {
 842                                         return $this->doQuotes ( $m[1], $m[2], ($m[1] == '') ? 'both' : 'emstrong' );
 843                                 } else if ($mode == 'strong') {
 844                                         return $m1_strong . $this->doQuotes ( '', $m[2], '' );
 845                                 } else if (($mode == 'emstrong') || ($mode == 'both')) {
 846                                         return $this->doQuotes ( '', $pre.$m1_strong.$m[2], 'em' );
 847                                 } else if ($mode == 'strongem') {
 848                                         return "<strong>{$pre}{$m1_em}</strong>" . $this->doQuotes ( '', $m[2], 'em' );
 849                                 } else {
 850                                         return $m[1] . $this->doQuotes ( '', $m[2], 'strong' );
 851                                 }
 852                         } else {
 853                                 if ($mode == 'strong') {
 854                                         return $this->doQuotes ( $m[1], $m[2], ($m[1] == '') ? 'both' : 'strongem' );
 855                                 } else if ($mode == 'em') {
 856                                         return $m1_em . $this->doQuotes ( '', $m[2], '' );
 857                                 } else if ($mode == 'emstrong') {
 858                                         return "<em>{$pre}{$m1_strong}</em>" . $this->doQuotes ( '', $m[2], 'strong' );
 859                                 } else if (($mode == 'strongem') || ($mode == 'both')) {
 860                                         return $this->doQuotes ( '', $pre.$m1_em.$m[2], 'strong' );
 861                                 } else {
 862                                         return $m[1] . $this->doQuotes ( '', $m[2], 'em' );
 863                                 }
 864                         }
 865                 } else {
 866                         $text_strong = ($text == '') ? '' : "<strong>{$text}</strong>";
 867                         $text_em = ($text == '') ? '' : "<em>{$text}</em>";
 868                         if ($mode == '') {
 869                                 return $pre . $text;
 870                         } else if ($mode == 'em') {
 871                                 return $pre . $text_em;
 872                         } else if ($mode == 'strong') {
 873                                 return $pre . $text_strong;
 874                         } else if ($mode == 'strongem') {
 875                                 return (($pre == '') && ($text == '')) ? '' : "<strong>{$pre}{$text_em}</strong>";
 876                         } else {
 877                                 return (($pre == '') && ($text == '')) ? '' : "<em>{$pre}{$text_strong}</em>";
 878                         }
 879                 }
 880         }
 881
 882         # Note: we have to do external links before the internal ones,
 883         # and otherwise take great care in the order of things here, so
 884         # that we don't end up interpreting some URLs twice.
 885
 886         /* private */ function replaceExternalLinks( $text ) {
 887                 $fname = 'Parser::replaceExternalLinks';
 888                 wfProfileIn( $fname );
 889                 $text = $this->subReplaceExternalLinks( $text, 'http', true );
 890                 $text = $this->subReplaceExternalLinks( $text, 'https', true );
 891                 $text = $this->subReplaceExternalLinks( $text, 'ftp', false );
 892                 $text = $this->subReplaceExternalLinks( $text, 'irc', false );
 893                 $text = $this->subReplaceExternalLinks( $text, 'gopher', false );
 894                 $text = $this->subReplaceExternalLinks( $text, 'news', false );
 895                 $text = $this->subReplaceExternalLinks( $text, 'mailto', false );
 896                 wfProfileOut( $fname );
 897                 return $text;
 898         }
 899
 900         /* private */ function subReplaceExternalLinks( $s, $protocol, $autonumber ) {
 901                 $unique = '4jzAfzB8hNvf4sqyO9Edd8pSmk9rE2in0Tgw3';
 902                 $uc = "A-Za-z0-9_\\/~%\\-+&*#?!=()@\\x80-\\xFF";
 903
 904                 # this is  the list of separators that should be ignored if they
 905                 # are the last character of an URL but that should be included
 906                 # if they occur within the URL, e.g. "go to www.foo.com, where .."
 907                 # in this case, the last comma should not become part of the URL,
 908                 # but in "www.foo.com/123,2342,32.htm" it should.
 909                 $sep = ",;\.:";
 910                 $fnc = 'A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF';
 911                 $images = 'gif|png|jpg|jpeg';
 912
 913                 # PLEASE NOTE: The curly braces { } are not part of the regex,
 914                 # they are interpreted as part of the string (used to tell PHP
 915                 # that the content of the string should be inserted there).
 916                 $e1 = "/(^|[^\\[])({$protocol}:)([{$uc}{$sep}]+)\\/([{$fnc}]+)\\." .
 917                   "((?i){$images})([^{$uc}]|$)/";
 918
 919                 $e2 = "/(^|[^\\[])({$protocol}:)(([".$uc."]|[".$sep."][".$uc."])+)([^". $uc . $sep. "]|[".$sep."]|$)/";
 920                 $sk =& $this->mOptions->getSkin();
 921
 922                 if ( $autonumber and $this->mOptions->getAllowExternalImages() ) { # Use img tags only for HTTP urls
 923                         $s = preg_replace( $e1, '\\1' . $sk->makeImage( "{$unique}:\\3" .
 924                           '/\\4.\\5', '\\4.\\5' ) . '\\6', $s );
 925                 }
 926                 $s = preg_replace( $e2, '\\1' . "<a href=\"{$unique}:\\3\"" .
 927                   $sk->getExternalLinkAttributes( "{$unique}:\\3", wfEscapeHTML(
 928                   "{$unique}:\\3" ) ) . ">" . wfEscapeHTML( "{$unique}:\\3" ) .
 929                   '</a>\\5', $s );
 930                 $s = str_replace( $unique, $protocol, $s );
 931
 932                 $a = explode( "[{$protocol}:", " " . $s );
 933                 $s = array_shift( $a );
 934                 $s = substr( $s, 1 );
 935
 936                 $e1 = "/^([{$uc}"."{$sep}]+)](.*)\$/sD";
 937                 $e2 = "/^([{$uc}"."{$sep}]+)\\s+([^\\]]+)](.*)\$/sD";
 938
 939                 foreach ( $a as $line ) {
 940                         if ( preg_match( $e1, $line, $m ) ) {
 941                                 $link = "{$protocol}:{$m[1]}";
 942                                 $trail = $m[2];
 943                                 if ( $autonumber ) { $text = "[" . ++$this->mAutonumber . "]"; }
 944                                 else { $text = wfEscapeHTML( $link ); }
 945                         } else if ( preg_match( $e2, $line, $m ) ) {
 946                                 $link = "{$protocol}:{$m[1]}";
 947                                 $text = $m[2];
 948                                 $trail = $m[3];
 949                         } else {
 950                                 $s .= "[{$protocol}:" . $line;
 951                                 continue;
 952                         }
 953                         if( $link == $text || preg_match( "!$protocol://" . preg_quote( $text, "/" ) . "/?$!", $link ) ) {
 954                                 $paren = '';
 955                         } else {
 956                                 # Expand the URL for printable version
 957                                 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $link ) . "</i>)</span>";
 958                         }
 959                         $la = $sk->getExternalLinkAttributes( $link, $text );
 960                         $s .= "<a href='{$link}'{$la}>{$text}</a>{$paren}{$trail}";
 961
 962                 }
 963                 return $s;
 964         }
 965
 966
 967         /* private */ function replaceInternalLinks( $s ) {
 968                 global $wgLang, $wgLinkCache;
 969                 global $wgNamespacesWithSubpages, $wgLanguageCode;
 970                 static $fname = 'Parser::replaceInternalLinks' ;
 971                 wfProfileIn( $fname );
 972
 973                 wfProfileIn( $fname.'-setup' );
 974                 static $tc = FALSE;
 975                 # the % is needed to support urlencoded titles as well
 976                 if ( !$tc ) { $tc = Title::legalChars() . '#%'; }
 977                 $sk =& $this->mOptions->getSkin();
 978
 979                 $a = explode( '[[', ' ' . $s );
 980                 $s = array_shift( $a );
 981                 $s = substr( $s, 1 );
 982
 983                 # Match a link having the form [[namespace:link|alternate]]trail
 984                 static $e1 = FALSE;
 985                 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
 986                 # Match the end of a line for a word that's not followed by whitespace,
 987                 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
 988                 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
 989
 990                 $useLinkPrefixExtension = $wgLang->linkPrefixExtension();
 991                 # Special and Media are pseudo-namespaces; no pages actually exist in them
 992                 static $image = FALSE;
 993                 static $special = FALSE;
 994                 static $media = FALSE;
 995                 static $category = FALSE;
 996                 if ( !$image ) { $image = Namespace::getImage(); }
 997                 if ( !$special ) { $special = Namespace::getSpecial(); }
 998                 if ( !$media ) { $media = Namespace::getMedia(); }
 999                 if ( !$category ) { $category = Namespace::getCategory(); }
1000
1001                 $nottalk = !Namespace::isTalk( $this->mTitle->getNamespace() );
1002
1003                 if ( $useLinkPrefixExtension ) {
1004                         if ( preg_match( $e2, $s, $m ) ) {
1005                                 $first_prefix = $m[2];
1006                                 $s = $m[1];
1007                         } else {
1008                                 $first_prefix = false;
1009                         }
1010                 } else {
1011                         $prefix = '';
1012                 }
1013
1014                 wfProfileOut( $fname.'-setup' );
1015
1016                 foreach ( $a as $line ) {
1017                         wfProfileIn( $fname.'-prefixhandling' );
1018                         if ( $useLinkPrefixExtension ) {
1019                                 if ( preg_match( $e2, $s, $m ) ) {
1020                                         $prefix = $m[2];
1021                                         $s = $m[1];
1022                                 } else {
1023                                         $prefix='';
1024                                 }
1025                                 # first link
1026                                 if($first_prefix) {
1027                                         $prefix = $first_prefix;
1028                                         $first_prefix = false;
1029                                 }
1030                         }
1031                         wfProfileOut( $fname.'-prefixhandling' );
1032
1033                         if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1034                                 $text = $m[2];
1035                                 # fix up urlencoded title texts
1036                                 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1037                                 $trail = $m[3];
1038                         } else { # Invalid form; output directly
1039                                 $s .= $prefix . '[[' . $line ;
1040                                 continue;
1041                         }
1042
1043                         /* Valid link forms:
1044                         Foobar -- normal
1045                         :Foobar -- override special treatment of prefix (images, language links)
1046                         /Foobar -- convert to CurrentPage/Foobar
1047                         /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1048                         */
1049                         $c = substr($m[1],0,1);
1050                         $noforce = ($c != ':');
1051                         if( $c == '/' ) { # subpage
1052                                 if(substr($m[1],-1,1)=='/') {                 # / at end means we don't want the slash to be shown
1053                                         $m[1]=substr($m[1],1,strlen($m[1])-2);
1054                                         $noslash=$m[1];
1055                                 } else {
1056                                         $noslash=substr($m[1],1);
1057                                 }
1058                                 if(!empty($wgNamespacesWithSubpages[$this->mTitle->getNamespace()])) { # subpages allowed here
1059                                         $link = $this->mTitle->getPrefixedText(). '/' . trim($noslash);
1060                                         if( '' == $text ) {
1061                                                 $text= $m[1];
1062                                         } # this might be changed for ugliness reasons
1063                                 } else {
1064                                         $link = $noslash; # no subpage allowed, use standard link
1065                                 }
1066                         } elseif( $noforce ) { # no subpage
1067                                 $link = $m[1];
1068                         } else {
1069                                 $link = substr( $m[1], 1 );
1070                         }
1071                         $wasblank = ( '' == $text );
1072                         if( $wasblank )
1073                         $text = $link;
1074
1075                         $nt = Title::newFromText( $link );
1076                         if( !$nt ) {
1077                                 $s .= $prefix . '[[' . $line;
1078                                 continue;
1079                         }
1080                         $ns = $nt->getNamespace();
1081                         $iw = $nt->getInterWiki();
1082                         if( $noforce ) {
1083                                 if( $iw && $this->mOptions->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
1084                                         array_push( $this->mOutput->mLanguageLinks, $nt->getPrefixedText() );
1085                                         $tmp = $prefix . $trail ;
1086                                         $s .= (trim($tmp) == '')? '': $tmp;
1087                                         continue;
1088                                 }
1089                                 if ( $ns == $image ) {
1090                                         $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1091                                         $wgLinkCache->addImageLinkObj( $nt );
1092                                         continue;
1093                                 }
1094                                 if ( $ns == $category ) {
1095                                         $t = $nt->getText() ;
1096                                         $nnt = Title::newFromText ( Namespace::getCanonicalName($category).":".$t ) ;
1097
1098                                         $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1099                                         $t = $sk->makeLinkObj( $nnt, $t, '', '' , $prefix );
1100                                         $wgLinkCache->resume();
1101
1102                                         $sortkey = $wasblank ? $this->mTitle->getPrefixedText() : $text;
1103                                         $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1104                                         $this->mOutput->mCategoryLinks[] = $t ;
1105                                         $s .= $prefix . $trail ;
1106                                         continue;
1107                                 }
1108                         }
1109                         if( ( $nt->getPrefixedText() == $this->mTitle->getPrefixedText() ) &&
1110                         ( strpos( $link, '#' ) == FALSE ) ) {
1111                                 # Self-links are handled specially; generally de-link and change to bold.
1112                                 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1113                                 continue;
1114                         }
1115
1116                         if( $ns == $media ) {
1117                                 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1118                                 $wgLinkCache->addImageLinkObj( $nt );
1119                                 continue;
1120                         } elseif( $ns == $special ) {
1121                                 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1122                                 continue;
1123                         }
1124                         $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1125                 }
1126                 wfProfileOut( $fname );
1127                 return $s;
1128         }
1129
1130         # Some functions here used by doBlockLevels()
1131         #
1132         /* private */ function closeParagraph() {
1133                 $result = '';
1134                 if ( '' != $this->mLastSection ) {
1135                         $result = '</' . $this->mLastSection  . ">\n";
1136                 }
1137                 $this->mInPre = false;
1138                 $this->mLastSection = '';
1139                 return $result;
1140         }
1141         # getCommon() returns the length of the longest common substring
1142         # of both arguments, starting at the beginning of both.
1143         #
1144         /* private */ function getCommon( $st1, $st2 ) {
1145                 $fl = strlen( $st1 );
1146                 $shorter = strlen( $st2 );
1147                 if ( $fl < $shorter ) { $shorter = $fl; }
1148
1149                 for ( $i = 0; $i < $shorter; ++$i ) {
1150                         if ( $st1{$i} != $st2{$i} ) { break; }
1151                 }
1152                 return $i;
1153         }
1154         # These next three functions open, continue, and close the list
1155         # element appropriate to the prefix character passed into them.
1156         #
1157         /* private */ function openList( $char )
1158     {
1159                 $result = $this->closeParagraph();
1160
1161                 if ( '*' == $char ) { $result .= '<ul><li>'; }
1162                 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1163                 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1164                 else if ( ';' == $char ) {
1165                         $result .= '<dl><dt>';
1166                         $this->mDTopen = true;
1167                 }
1168                 else { $result = '<!-- ERR 1 -->'; }
1169
1170                 return $result;
1171         }
1172
1173         /* private */ function nextItem( $char ) {
1174                 if ( '*' == $char || '#' == $char ) { return '</li><li>'; }
1175                 else if ( ':' == $char || ';' == $char ) {
1176                         $close = "</dd>";
1177                         if ( $this->mDTopen ) { $close = '</dt>'; }
1178                         if ( ';' == $char ) {
1179                                 $this->mDTopen = true;
1180                                 return $close . '<dt>';
1181                         } else {
1182                                 $this->mDTopen = false;
1183                                 return $close . '<dd>';
1184                         }
1185                 }
1186                 return '<!-- ERR 2 -->';
1187         }
1188
1189         /* private */function closeList( $char ) {
1190                 if ( '*' == $char ) { $text = '</li></ul>'; }
1191                 else if ( '#' == $char ) { $text = '</li></ol>'; }
1192                 else if ( ':' == $char ) {
1193                         if ( $this->mDTopen ) {
1194                                 $this->mDTopen = false;
1195                                 $text = '</dt></dl>';
1196                         } else {
1197                                 $text = '</dd></dl>';
1198                         }
1199                 }
1200                 else {  return '<!-- ERR 3 -->'; }
1201                 return $text."\n";
1202         }
1203
1204         /* private */ function doBlockLevels( $text, $linestart ) {
1205                 $fname = 'Parser::doBlockLevels';
1206                 wfProfileIn( $fname );
1207
1208                 # Parsing through the text line by line.  The main thing
1209                 # happening here is handling of block-level elements p, pre,
1210                 # and making lists from lines starting with * # : etc.
1211                 #
1212                 $textLines = explode( "\n", $text );
1213
1214                 $lastPrefix = $output = $lastLine = '';
1215                 $this->mDTopen = $inBlockElem = false;
1216                 $prefixLength = 0;
1217                 $paragraphStack = false;
1218
1219                 if ( !$linestart ) {
1220                         $output .= array_shift( $textLines );
1221                 }
1222                 foreach ( $textLines as $oLine ) {
1223                         $lastPrefixLength = strlen( $lastPrefix );
1224                         $preCloseMatch = preg_match("/<\\/pre/i", $oLine );
1225                         $preOpenMatch = preg_match("/<pre/i", $oLine );
1226                         if (!$this->mInPre) {
1227                                 $this->mInPre = !empty($preOpenMatch);
1228                         }
1229                         if ( !$this->mInPre ) {
1230                                 # Multiple prefixes may abut each other for nested lists.
1231                                 $prefixLength = strspn( $oLine, '*#:;' );
1232                                 $pref = substr( $oLine, 0, $prefixLength );
1233
1234                                 # eh?
1235                                 $pref2 = str_replace( ';', ':', $pref );
1236                                 $t = substr( $oLine, $prefixLength );
1237                         } else {
1238                                 # Don't interpret any other prefixes in preformatted text
1239                                 $prefixLength = 0;
1240                                 $pref = $pref2 = '';
1241                                 $t = $oLine;
1242                         }
1243
1244                         # List generation
1245                         if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1246                                 # Same as the last item, so no need to deal with nesting or opening stuff
1247                                 $output .= $this->nextItem( substr( $pref, -1 ) );
1248                                 $paragraphStack = false;
1249
1250                                 if ( ";" == substr( $pref, -1 ) ) {
1251                                         # The one nasty exception: definition lists work like this:
1252                                         # ; title : definition text
1253                                         # So we check for : in the remainder text to split up the
1254                                         # title and definition, without b0rking links.
1255                                         # FIXME: This is not foolproof. Something better in Tokenizer might help.
1256                                         if( preg_match( '/^(.*?(?:\s|&nbsp;)):(.*)$/', $t, $match ) ) {
1257                                                 $term = $match[1];
1258                                                 $output .= $term . $this->nextItem( ':' );
1259                                                 $t = $match[2];
1260                                         }
1261                                 }
1262                         } elseif( $prefixLength || $lastPrefixLength ) {
1263                                 # Either open or close a level...
1264                                 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1265                                 $paragraphStack = false;
1266
1267                                 while( $commonPrefixLength < $lastPrefixLength ) {
1268                                         $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1269                                         --$lastPrefixLength;
1270                                 }
1271                                 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1272                                         $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1273                                 }
1274                                 while ( $prefixLength > $commonPrefixLength ) {
1275                                         $char = substr( $pref, $commonPrefixLength, 1 );
1276                                         $output .= $this->openList( $char );
1277
1278                                         if ( ';' == $char ) {
1279                                                 # FIXME: This is dupe of code above
1280                                                 if( preg_match( '/^(.*?(?:\s|&nbsp;)):(.*)$/', $t, $match ) ) {
1281                                                         $term = $match[1];
1282                                                         $output .= $term . $this->nextItem( ":" );
1283                                                         $t = $match[2];
1284                                                 }
1285                                         }
1286                                         ++$commonPrefixLength;
1287                                 }
1288                                 $lastPrefix = $pref2;
1289                         }
1290                         if( 0 == $prefixLength ) {
1291                                 # No prefix (not in list)--go to paragraph mode
1292                                 $uniq_prefix = UNIQ_PREFIX;
1293                                 // XXX: use a stack for nestable elements like span, table and div
1294                                 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1295                                 $closematch = preg_match(
1296                                         '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1297                                         '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1298                                 if ( $openmatch or $closematch ) {
1299                                         $paragraphStack = false;
1300                                         $output .= $this->closeParagraph();
1301                                         if($preOpenMatch and !$preCloseMatch) {
1302                                                 $this->mInPre = true;
1303                                         }
1304                                         if ( $closematch  ) {
1305                                                 $inBlockElem = false;
1306                                         } else {
1307                                                 $inBlockElem = true;
1308                                         }
1309                                 } else if ( !$inBlockElem && !$this->mInPre ) {
1310                                         if ( " " == $t{0} and trim($t) != '' ) {
1311                                                 // pre
1312                                                 if ($this->mLastSection != 'pre') {
1313                                                         $paragraphStack = false;
1314                                                         $output .= $this->closeParagraph().'<pre>';
1315                                                         $this->mLastSection = 'pre';
1316                                                 }
1317                                         } else {
1318                                                 // paragraph
1319                                                 if ( '' == trim($t) ) {
1320                                                         if ( $paragraphStack ) {
1321                                                                 $output .= $paragraphStack.'<br />';
1322                                                                 $paragraphStack = false;
1323                                                                 $this->mLastSection = 'p';
1324                                                         } else {
1325                                                                 if ($this->mLastSection != 'p' ) {
1326                                                                         $output .= $this->closeParagraph();
1327                                                                         $this->mLastSection = '';
1328                                                                         $paragraphStack = '<p>';
1329                                                                 } else {
1330                                                                         $paragraphStack = '</p><p>';
1331                                                                 }
1332                                                         }
1333                                                 } else {
1334                                                         if ( $paragraphStack ) {
1335                                                                 $output .= $paragraphStack;
1336                                                                 $paragraphStack = false;
1337                                                                 $this->mLastSection = 'p';
1338                                                         } else if ($this->mLastSection != 'p') {
1339                                                                 $output .= $this->closeParagraph().'<p>';
1340                                                                 $this->mLastSection = 'p';
1341                                                         }
1342                                                 }
1343                                         }
1344                                 }
1345                         }
1346                         if ($paragraphStack === false) {
1347                                 $output .= $t."\n";
1348                         }
1349                 }
1350                 while ( $prefixLength ) {
1351                         $output .= $this->closeList( $pref2{$prefixLength-1} );
1352                         --$prefixLength;
1353                 }
1354                 if ( '' != $this->mLastSection ) {
1355                         $output .= '</' . $this->mLastSection . '>';
1356                         $this->mLastSection = '';
1357                 }
1358
1359                 wfProfileOut( $fname );
1360                 return $output;
1361         }
1362
1363         # Return value of a magic variable (like PAGENAME)
1364         function getVariableValue( $index ) {
1365                 global $wgLang, $wgSitename, $wgServer;
1366
1367                 switch ( $index ) {
1368                         case MAG_CURRENTMONTH:
1369                                 return date( 'm' );
1370                         case MAG_CURRENTMONTHNAME:
1371                                 return $wgLang->getMonthName( date('n') );
1372                         case MAG_CURRENTMONTHNAMEGEN:
1373                                 return $wgLang->getMonthNameGen( date('n') );
1374                         case MAG_CURRENTDAY:
1375                                 return date('j');
1376                         case MAG_PAGENAME:
1377                                 return $this->mTitle->getText();
1378                         case MAG_NAMESPACE:
1379                                 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1380                                 return $wgLang->getNsText($this->mTitle->getNamespace()); // Patch  by Dori
1381                         case MAG_CURRENTDAYNAME:
1382                                 return $wgLang->getWeekdayName( date('w')+1 );
1383                         case MAG_CURRENTYEAR:
1384                                 return date( 'Y' );
1385                         case MAG_CURRENTTIME:
1386                                 return $wgLang->time( wfTimestampNow(), false );
1387                         case MAG_NUMBEROFARTICLES:
1388                                 return wfNumberOfArticles();
1389                         case MAG_SITENAME:
1390                                 return $wgSitename;
1391                         case MAG_SERVER:
1392                                 return $wgServer;
1393                         default:
1394                                 return NULL;
1395                 }
1396         }
1397
1398         # initialise the magic variables (like CURRENTMONTHNAME)
1399         function initialiseVariables() {
1400                 global $wgVariableIDs;
1401                 $this->mVariables = array();
1402                 foreach ( $wgVariableIDs as $id ) {
1403                         $mw =& MagicWord::get( $id );
1404                         $mw->addToArray( $this->mVariables, $this->getVariableValue( $id ) );
1405                 }
1406         }
1407
1408         /* private */ function replaceVariables( $text, $args = array() ) {
1409                 global $wgLang, $wgScript, $wgArticlePath;
1410
1411                 $fname = 'Parser::replaceVariables';
1412                 wfProfileIn( $fname );
1413
1414                 $bail = false;
1415                 if ( !$this->mVariables ) {
1416                         $this->initialiseVariables();
1417                 }
1418                 $titleChars = Title::legalChars();
1419                 $nonBraceChars = str_replace( array( '{', '}' ), array( '', '' ), $titleChars );
1420
1421                 # This function is called recursively. To keep track of arguments we need a stack:
1422                 array_push( $this->mArgStack, $args );
1423
1424                 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1425                 $GLOBALS['wgCurParser'] =& $this;
1426
1427
1428                 if ( $this->mOutputType == OT_HTML ) {
1429                         # Variable substitution
1430                         $text = preg_replace_callback( "/{{([$nonBraceChars]*?)}}/", 'wfVariableSubstitution', $text );
1431
1432                         # Argument substitution
1433                         $text = preg_replace_callback( "/(\\n?){{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1434                 }
1435                 # Template substitution
1436                 $regex = '/(\\n?){{(['.$nonBraceChars.']*)(\\|.*?|)}}/s';
1437                 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1438
1439                 array_pop( $this->mArgStack );
1440
1441                 wfProfileOut( $fname );
1442                 return $text;
1443         }
1444
1445         function variableSubstitution( $matches ) {
1446                 if ( array_key_exists( $matches[1], $this->mVariables ) ) {
1447                         $text = $this->mVariables[$matches[1]];
1448                         $this->mOutput->mContainsOldMagic = true;
1449                 } else {
1450                         $text = $matches[0];
1451                 }
1452                 return $text;
1453         }
1454
1455         function braceSubstitution( $matches ) {
1456                 global $wgLinkCache, $wgLang;
1457                 $fname = 'Parser::braceSubstitution';
1458                 $found = false;
1459                 $nowiki = false;
1460                 $noparse = false;
1461
1462                 $title = NULL;
1463
1464                 # $newline is an optional newline character before the braces
1465                 # $part1 is the bit before the first |, and must contain only title characters
1466                 # $args is a list of arguments, starting from index 0, not including $part1
1467
1468                 $newline = $matches[1];
1469                 $part1 = $matches[2];
1470                 # If the third subpattern matched anything, it will start with |
1471                 if ( $matches[3] !== '' ) {
1472                         $args = explode( '|', substr( $matches[3], 1 ) );
1473                 } else {
1474                         $args = array();
1475                 }
1476                 $argc = count( $args );
1477
1478                 # {{{}}}
1479                 if ( strpos( $matches[0], '{{{' ) !== false ) {
1480                         $text = $matches[0];
1481                         $found = true;
1482                         $noparse = true;
1483                 }
1484
1485                 # SUBST
1486                 if ( !$found ) {
1487                         $mwSubst =& MagicWord::get( MAG_SUBST );
1488                         if ( $mwSubst->matchStartAndRemove( $part1 ) ) {
1489                                 if ( $this->mOutputType != OT_WIKI ) {
1490                                         # Invalid SUBST not replaced at PST time
1491                                         # Return without further processing
1492                                         $text = $matches[0];
1493                                         $found = true;
1494                                         $noparse= true;
1495                                 }
1496                         } elseif ( $this->mOutputType == OT_WIKI ) {
1497                                 # SUBST not found in PST pass, do nothing
1498                                 $text = $matches[0];
1499                                 $found = true;
1500                         }
1501                 }
1502
1503                 # MSG, MSGNW and INT
1504                 if ( !$found ) {
1505                         # Check for MSGNW:
1506                         $mwMsgnw =& MagicWord::get( MAG_MSGNW );
1507                         if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1508                                 $nowiki = true;
1509                         } else {
1510                                 # Remove obsolete MSG:
1511                                 $mwMsg =& MagicWord::get( MAG_MSG );
1512                                 $mwMsg->matchStartAndRemove( $part1 );
1513                         }
1514
1515                         # Check if it is an internal message
1516                         $mwInt =& MagicWord::get( MAG_INT );
1517                         if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1518                                 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1519                                         $text = wfMsgReal( $part1, $args, true );
1520                                         $found = true;
1521                                 }
1522                         }
1523                 }
1524
1525                 # NS
1526                 if ( !$found ) {
1527                         # Check for NS: (namespace expansion)
1528                         $mwNs = MagicWord::get( MAG_NS );
1529                         if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1530                                 if ( intval( $part1 ) ) {
1531                                         $text = $wgLang->getNsText( intval( $part1 ) );
1532                                         $found = true;
1533                                 } else {
1534                                         $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1535                                         if ( !is_null( $index ) ) {
1536                                                 $text = $wgLang->getNsText( $index );
1537                                                 $found = true;
1538                                         }
1539                                 }
1540                         }
1541                 }
1542
1543                 # LOCALURL and LOCALURLE
1544                 if ( !$found ) {
1545                         $mwLocal = MagicWord::get( MAG_LOCALURL );
1546                         $mwLocalE = MagicWord::get( MAG_LOCALURLE );
1547
1548                         if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1549                                 $func = 'getLocalURL';
1550                         } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1551                                 $func = 'escapeLocalURL';
1552                         } else {
1553                                 $func = '';
1554                         }
1555
1556                         if ( $func !== '' ) {
1557                                 $title = Title::newFromText( $part1 );
1558                                 if ( !is_null( $title ) ) {
1559                                         if ( $argc > 0 ) {
1560                                                 $text = $title->$func( $args[0] );
1561                                         } else {
1562                                                 $text = $title->$func();
1563                                         }
1564                                         $found = true;
1565                                 }
1566                         }
1567                 }
1568
1569                 # Internal variables
1570                 if ( !$found && array_key_exists( $part1, $this->mVariables ) ) {
1571                         $text = $this->mVariables[$part1];
1572                         $found = true;
1573                         $this->mOutput->mContainsOldMagic = true;
1574                 }
1575 /*
1576                 # Arguments input from the caller
1577                 $inputArgs = end( $this->mArgStack );
1578                 if ( !$found && array_key_exists( $part1, $inputArgs ) ) {
1579                         $text = $inputArgs[$part1];
1580                         $found = true;
1581                 }
1582 */
1583                 # Load from database
1584                 if ( !$found ) {
1585                         $title = Title::newFromText( $part1, NS_TEMPLATE );
1586                         if ( !is_null( $title ) && !$title->isExternal() ) {
1587                                 # Check for excessive inclusion
1588                                 $dbk = $title->getPrefixedDBkey();
1589                                 if ( $this->incrementIncludeCount( $dbk ) ) {
1590                                         $article = new Article( $title );
1591                                         $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1592                                         if ( $articleContent !== false ) {
1593                                                 $found = true;
1594                                                 $text = $articleContent;
1595
1596                                         }
1597                                 }
1598
1599                                 # If the title is valid but undisplayable, make a link to it
1600                                 if ( $this->mOutputType == OT_HTML && !$found ) {
1601                                         $text = '[[' . $title->getPrefixedText() . ']]';
1602                                         $found = true;
1603                                 }
1604                         }
1605                 }
1606
1607                 # Recursive parsing, escaping and link table handling
1608                 # Only for HTML output
1609                 if ( $nowiki && $found && $this->mOutputType == OT_HTML ) {
1610                         $text = wfEscapeWikiText( $text );
1611                 } elseif ( $this->mOutputType == OT_HTML && $found && !$noparse) {
1612                         # Clean up argument array
1613                         $assocArgs = array();
1614                         $index = 1;
1615                         foreach( $args as $arg ) {
1616                                 $eqpos = strpos( $arg, '=' );
1617                                 if ( $eqpos === false ) {
1618                                         $assocArgs[$index++] = $arg;
1619                                 } else {
1620                                         $name = trim( substr( $arg, 0, $eqpos ) );
1621                                         $value = trim( substr( $arg, $eqpos+1 ) );
1622                                         if ( $value === false ) {
1623                                                 $value = '';
1624                                         }
1625                                         if ( $name !== false ) {
1626                                                 $assocArgs[$name] = $value;
1627                                         }
1628                                 }
1629                         }
1630
1631                         # Do not enter included links in link table
1632                         if ( !is_null( $title ) ) {
1633                                 $wgLinkCache->suspend();
1634                         }
1635
1636                         # Run full parser on the included text
1637                         $text = $this->stripParse( $text, $newline, $assocArgs );
1638
1639                         # Resume the link cache and register the inclusion as a link
1640                         if ( !is_null( $title ) ) {
1641                                 $wgLinkCache->resume();
1642                                 $wgLinkCache->addLinkObj( $title );
1643                         }
1644                 }
1645
1646                 if ( !$found ) {
1647                         return $matches[0];
1648                 } else {
1649                         return $text;
1650                 }
1651         }
1652
1653         # Triple brace replacement -- used for template arguments
1654         function argSubstitution( $matches ) {
1655                 $newline = $matches[1];
1656                 $arg = trim( $matches[2] );
1657                 $text = $matches[0];
1658                 $inputArgs = end( $this->mArgStack );
1659
1660                 if ( array_key_exists( $arg, $inputArgs ) ) {
1661                         $text = $this->stripParse( $inputArgs[$arg], $newline, array() );
1662                 }
1663
1664                 return $text;
1665         }
1666
1667         # Returns true if the function is allowed to include this entity
1668         function incrementIncludeCount( $dbk ) {
1669                 if ( !array_key_exists( $dbk, $this->mIncludeCount ) ) {
1670                         $this->mIncludeCount[$dbk] = 0;
1671                 }
1672                 if ( ++$this->mIncludeCount[$dbk] <= MAX_INCLUDE_REPEAT ) {
1673                         return true;
1674                 } else {
1675                         return false;
1676                 }
1677         }
1678
1679
1680         # Cleans up HTML, removes dangerous tags and attributes
1681         /* private */ function removeHTMLtags( $text ) {
1682                 global $wgUseTidy, $wgUserHtml;
1683                 $fname = 'Parser::removeHTMLtags';
1684                 wfProfileIn( $fname );
1685
1686                 if( $wgUserHtml ) {
1687                         $htmlpairs = array( # Tags that must be closed
1688                                 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
1689                                 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
1690                                 'strike', 'strong', 'tt', 'var', 'div', 'center',
1691                                 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
1692                                 'ruby', 'rt' , 'rb' , 'rp', 'p'
1693                         );
1694                         $htmlsingle = array(
1695                                 'br', 'hr', 'li', 'dt', 'dd'
1696                         );
1697                         $htmlnest = array( # Tags that can be nested--??
1698                                 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
1699                                 'dl', 'font', 'big', 'small', 'sub', 'sup'
1700                         );
1701                         $tabletags = array( # Can only appear inside table
1702                                 'td', 'th', 'tr'
1703                         );
1704                 } else {
1705                         $htmlpairs = array();
1706                         $htmlsingle = array();
1707                         $htmlnest = array();
1708                         $tabletags = array();
1709                 }
1710
1711                 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1712                 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1713
1714                 $htmlattrs = $this->getHTMLattrs () ;
1715
1716                 # Remove HTML comments
1717                 $text = preg_replace( '/(\\n *<!--.*--> *(?=\\n)|<!--.*-->)/sU', '$2', $text );
1718
1719                 $bits = explode( '<', $text );
1720                 $text = array_shift( $bits );
1721                 if(!$wgUseTidy) {
1722                         $tagstack = array(); $tablestack = array();
1723                         foreach ( $bits as $x ) {
1724                                 $prev = error_reporting( E_ALL & ~( E_NOTICE | E_WARNING ) );
1725                                 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1726                                 $x, $regs );
1727                                 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1728                                 error_reporting( $prev );
1729
1730                                 $badtag = 0 ;
1731                                 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1732                                         # Check our stack
1733                                         if ( $slash ) {
1734                                                 # Closing a tag...
1735                                                 if ( ! in_array( $t, $htmlsingle ) &&
1736                                                 ( $ot = @array_pop( $tagstack ) ) != $t ) {
1737                                                         @array_push( $tagstack, $ot );
1738                                                         $badtag = 1;
1739                                                 } else {
1740                                                         if ( $t == 'table' ) {
1741                                                                 $tagstack = array_pop( $tablestack );
1742                                                         }
1743                                                         $newparams = '';
1744                                                 }
1745                                         } else {
1746                                                 # Keep track for later
1747                                                 if ( in_array( $t, $tabletags ) &&
1748                                                 ! in_array( 'table', $tagstack ) ) {
1749                                                         $badtag = 1;
1750                                                 } else if ( in_array( $t, $tagstack ) &&
1751                                                 ! in_array ( $t , $htmlnest ) ) {
1752                                                         $badtag = 1 ;
1753                                                 } else if ( ! in_array( $t, $htmlsingle ) ) {
1754                                                         if ( $t == 'table' ) {
1755                                                                 array_push( $tablestack, $tagstack );
1756                                                                 $tagstack = array();
1757                                                         }
1758                                                         array_push( $tagstack, $t );
1759                                                 }
1760                                                 # Strip non-approved attributes from the tag
1761                                                 $newparams = $this->fixTagAttributes($params);
1762
1763                                         }
1764                                         if ( ! $badtag ) {
1765                                                 $rest = str_replace( '>', '&gt;', $rest );
1766                                                 $text .= "<$slash$t $newparams$brace$rest";
1767                                                 continue;
1768                                         }
1769                                 }
1770                                 $text .= '&lt;' . str_replace( '>', '&gt;', $x);
1771                         }
1772                         # Close off any remaining tags
1773                         while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
1774                                 $text .= "</$t>\n";
1775                                 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
1776                         }
1777                 } else {
1778                         # this might be possible using tidy itself
1779                         foreach ( $bits as $x ) {
1780                                 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1781                                 $x, $regs );
1782                                 @list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1783                                 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1784                                         $newparams = $this->fixTagAttributes($params);
1785                                         $rest = str_replace( '>', '&gt;', $rest );
1786                                         $text .= "<$slash$t $newparams$brace$rest";
1787                                 } else {
1788                                         $text .= '&lt;' . str_replace( '>', '&gt;', $x);
1789                                 }
1790                         }
1791                 }
1792                 wfProfileOut( $fname );
1793                 return $text;
1794         }
1795
1796
1797 /*
1798  *
1799  * This function accomplishes several tasks:
1800  * 1) Auto-number headings if that option is enabled
1801  * 2) Add an [edit] link to sections for logged in users who have enabled the option
1802  * 3) Add a Table of contents on the top for users who have enabled the option
1803  * 4) Auto-anchor headings
1804  *
1805  * It loops through all headlines, collects the necessary data, then splits up the
1806  * string and re-inserts the newly formatted headlines.
1807  *
1808  */
1809
1810         /* private */ function formatHeadings( $text, $isMain=true ) {
1811                 global $wgInputEncoding;
1812
1813                 $doNumberHeadings = $this->mOptions->getNumberHeadings();
1814                 $doShowToc = $this->mOptions->getShowToc();
1815                 if( !$this->mTitle->userCanEdit() ) {
1816                         $showEditLink = 0;
1817                         $rightClickHack = 0;
1818                 } else {
1819                         $showEditLink = $this->mOptions->getEditSection();
1820                         $rightClickHack = $this->mOptions->getEditSectionOnRightClick();
1821                 }
1822
1823                 # Inhibit editsection links if requested in the page
1824                 $esw =& MagicWord::get( MAG_NOEDITSECTION );
1825                 if( $esw->matchAndRemove( $text ) ) {
1826                         $showEditLink = 0;
1827                 }
1828                 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1829                 # do not add TOC
1830                 $mw =& MagicWord::get( MAG_NOTOC );
1831                 if( $mw->matchAndRemove( $text ) ) {
1832                         $doShowToc = 0;
1833                 }
1834
1835                 # never add the TOC to the Main Page. This is an entry page that should not
1836                 # be more than 1-2 screens large anyway
1837                 if( $this->mTitle->getPrefixedText() == wfMsg('mainpage') ) {
1838                         $doShowToc = 0;
1839                 }
1840
1841                 # Get all headlines for numbering them and adding funky stuff like [edit]
1842                 # links - this is for later, but we need the number of headlines right now
1843                 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
1844
1845                 # if there are fewer than 4 headlines in the article, do not show TOC
1846                 if( $numMatches < 4 ) {
1847                         $doShowToc = 0;
1848                 }
1849
1850                 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
1851                 # override above conditions and always show TOC
1852                 $mw =& MagicWord::get( MAG_FORCETOC );
1853                 if ($mw->matchAndRemove( $text ) ) {
1854                         $doShowToc = 1;
1855                 }
1856
1857
1858                 # We need this to perform operations on the HTML
1859                 $sk =& $this->mOptions->getSkin();
1860
1861                 # headline counter
1862                 $headlineCount = 0;
1863
1864                 # Ugh .. the TOC should have neat indentation levels which can be
1865                 # passed to the skin functions. These are determined here
1866                 $toclevel = 0;
1867                 $toc = '';
1868                 $full = '';
1869                 $head = array();
1870                 $sublevelCount = array();
1871                 $level = 0;
1872                 $prevlevel = 0;
1873                 foreach( $matches[3] as $headline ) {
1874                         $numbering = '';
1875                         if( $level ) {
1876                                 $prevlevel = $level;
1877                         }
1878                         $level = $matches[1][$headlineCount];
1879                         if( ( $doNumberHeadings || $doShowToc ) && $prevlevel && $level > $prevlevel ) {
1880                                 # reset when we enter a new level
1881                                 $sublevelCount[$level] = 0;
1882                                 $toc .= $sk->tocIndent( $level - $prevlevel );
1883                                 $toclevel += $level - $prevlevel;
1884                         }
1885                         if( ( $doNumberHeadings || $doShowToc ) && $level < $prevlevel ) {
1886                                 # reset when we step back a level
1887                                 $sublevelCount[$level+1]=0;
1888                                 $toc .= $sk->tocUnindent( $prevlevel - $level );
1889                                 $toclevel -= $prevlevel - $level;
1890                         }
1891                         # count number of headlines for each level
1892                         @$sublevelCount[$level]++;
1893                         if( $doNumberHeadings || $doShowToc ) {
1894                                 $dot = 0;
1895                                 for( $i = 1; $i <= $level; $i++ ) {
1896                                         if( !empty( $sublevelCount[$i] ) ) {
1897                                                 if( $dot ) {
1898                                                         $numbering .= '.';
1899                                                 }
1900                                                 $numbering .= $sublevelCount[$i];
1901                                                 $dot = 1;
1902                                         }
1903                                 }
1904                         }
1905
1906                         # The canonized header is a version of the header text safe to use for links
1907                         # Avoid insertion of weird stuff like <math> by expanding the relevant sections
1908                         $canonized_headline = $this->unstrip( $headline, $this->mStripState );
1909                         $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState );
1910
1911                         # strip out HTML
1912                         $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
1913                         $tocline = trim( $canonized_headline );
1914                         $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT, $wgInputEncoding ) );
1915                         $replacearray = array(
1916                                 '%3A' => ':',
1917                                 '%' => '.'
1918                         );
1919                         $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
1920                         $refer[$headlineCount] = $canonized_headline;
1921
1922                         # count how many in assoc. array so we can track dupes in anchors
1923                         @$refers[$canonized_headline]++;
1924                         $refcount[$headlineCount]=$refers[$canonized_headline];
1925
1926                         # Prepend the number to the heading text
1927
1928                         if( $doNumberHeadings || $doShowToc ) {
1929                                 $tocline = $numbering . ' ' . $tocline;
1930
1931                                 # Don't number the heading if it is the only one (looks silly)
1932                                 if( $doNumberHeadings && count( $matches[3] ) > 1) {
1933                                         # the two are different if the line contains a link
1934                                         $headline=$numbering . ' ' . $headline;
1935                                 }
1936                         }
1937
1938                         # Create the anchor for linking from the TOC to the section
1939                         $anchor = $canonized_headline;
1940                         if($refcount[$headlineCount] > 1 ) {
1941                                 $anchor .= '_' . $refcount[$headlineCount];
1942                         }
1943                         if( $doShowToc ) {
1944                                 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
1945                         }
1946                         if( $showEditLink ) {
1947                                 if ( empty( $head[$headlineCount] ) ) {
1948                                         $head[$headlineCount] = '';
1949                                 }
1950                                 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+1);
1951                         }
1952
1953                         # Add the edit section span
1954                         if( $rightClickHack ) {
1955                                 $headline = $sk->editSectionScript($headlineCount+1,$headline);
1956                         }
1957
1958                         # give headline the correct <h#> tag
1959                         @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline."</h".$level.">";
1960
1961                         $headlineCount++;
1962                 }
1963
1964                 if( $doShowToc ) {
1965                         $toclines = $headlineCount;
1966                         $toc .= $sk->tocUnindent( $toclevel );
1967                         $toc = $sk->tocTable( $toc );
1968                 }
1969
1970                 # split up and insert constructed headlines
1971
1972                 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
1973                 $i = 0;
1974
1975                 foreach( $blocks as $block ) {
1976                         if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
1977                             # This is the [edit] link that appears for the top block of text when
1978                                 # section editing is enabled
1979
1980                                 # Disabled because it broke block formatting
1981                                 # For example, a bullet point in the top line
1982                                 # $full .= $sk->editSectionLink(0);
1983                         }
1984                         $full .= $block;
1985                         if( $doShowToc && !$i && $isMain) {
1986                         # Top anchor now in skin
1987                                 $full = $full.$toc;
1988                         }
1989
1990                         if( !empty( $head[$i] ) ) {
1991                                 $full .= $head[$i];
1992                         }
1993                         $i++;
1994                 }
1995
1996                 return $full;
1997         }
1998
1999         # Return an HTML link for the "ISBN 123456" text
2000         /* private */ function magicISBN( $text ) {
2001                 global $wgLang;
2002
2003                 $a = split( 'ISBN ', " $text" );
2004                 if ( count ( $a ) < 2 ) return $text;
2005                 $text = substr( array_shift( $a ), 1);
2006                 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2007
2008                 foreach ( $a as $x ) {
2009                         $isbn = $blank = '' ;
2010                         while ( ' ' == $x{0} ) {
2011                                 $blank .= ' ';
2012                                 $x = substr( $x, 1 );
2013                         }
2014                         while ( strstr( $valid, $x{0} ) != false ) {
2015                                 $isbn .= $x{0};
2016                                 $x = substr( $x, 1 );
2017                         }
2018                         $num = str_replace( '-', '', $isbn );
2019                         $num = str_replace( ' ', '', $num );
2020
2021                         if ( '' == $num ) {
2022                                 $text .= "ISBN $blank$x";
2023                         } else {
2024                                 $titleObj = Title::makeTitle( NS_SPECIAL, 'Booksources' );
2025                                 $text .= '<a href="' .
2026                                 $titleObj->escapeLocalUrl( "isbn={$num}" ) .
2027                                         "\" class=\"internal\">ISBN $isbn</a>";
2028                                 $text .= $x;
2029                         }
2030                 }
2031                 return $text;
2032         }
2033
2034         # Return an HTML link for the "RFC 1234" text
2035         /* private */ function magicRFC( $text ) {
2036                 global $wgLang;
2037
2038                 $a = split( 'RFC ', ' '.$text );
2039                 if ( count ( $a ) < 2 ) return $text;
2040                 $text = substr( array_shift( $a ), 1);
2041                 $valid = '0123456789';
2042
2043                 foreach ( $a as $x ) {
2044                         $rfc = $blank = '' ;
2045                         while ( ' ' == $x{0} ) {
2046                                 $blank .= ' ';
2047                                 $x = substr( $x, 1 );
2048                         }
2049                         while ( strstr( $valid, $x{0} ) != false ) {
2050                                 $rfc .= $x{0};
2051                                 $x = substr( $x, 1 );
2052                         }
2053
2054                         if ( '' == $rfc ) {
2055                                 $text .= "RFC $blank$x";
2056                         } else {
2057                                 $url = wfmsg( 'rfcurl' );
2058                                 $url = str_replace( '$1', $rfc, $url);
2059                                 $sk =& $this->mOptions->getSkin();
2060                                 $la = $sk->getExternalLinkAttributes( $url, "RFC {$rfc}" );
2061                                 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2062                         }
2063                 }
2064                 return $text;
2065         }
2066
2067         function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2068                 $this->mOptions = $options;
2069                 $this->mTitle =& $title;
2070                 $this->mOutputType = OT_WIKI;
2071
2072                 if ( $clearState ) {
2073                         $this->clearState();
2074                 }
2075
2076                 $stripState = false;
2077                 $pairs = array(
2078                         "\r\n" => "\n",
2079                         );
2080                 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2081                 // now with regexes
2082                 /*
2083                 $pairs = array(
2084                         "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2085                         "/<br *?>/i" => "<br />",
2086                 );
2087                 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2088                 */
2089                 $text = $this->strip( $text, $stripState, false );
2090                 $text = $this->pstPass2( $text, $user );
2091                 $text = $this->unstrip( $text, $stripState );
2092                 $text = $this->unstripNoWiki( $text, $stripState );
2093                 return $text;
2094         }
2095
2096         /* private */ function pstPass2( $text, &$user ) {
2097                 global $wgLang, $wgLocaltimezone, $wgCurParser;
2098
2099                 # Variable replacement
2100                 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2101                 $text = $this->replaceVariables( $text );
2102
2103                 # Signatures
2104                 #
2105                 $n = $user->getName();
2106                 $k = $user->getOption( 'nickname' );
2107                 if ( '' == $k ) { $k = $n; }
2108                 if(isset($wgLocaltimezone)) {
2109                         $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2110                 }
2111                 /* Note: this is an ugly timezone hack for the European wikis */
2112                 $d = $wgLang->timeanddate( date( 'YmdHis' ), false ) .
2113                   ' (' . date( 'T' ) . ')';
2114                 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2115
2116                 $text = preg_replace( '/~~~~~/', $d, $text );
2117                 $text = preg_replace( '/~~~~/', '[[' . $wgLang->getNsText(
2118                   Namespace::getUser() ) . ":$n|$k]] $d", $text );
2119                 $text = preg_replace( '/~~~/', '[[' . $wgLang->getNsText(
2120                   Namespace::getUser() ) . ":$n|$k]]", $text );
2121
2122                 # Context links: [[|name]] and [[name (context)|]]
2123                 #
2124                 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2125                 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2126                 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2127                 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2128
2129                 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/";             # [[page (context)|]]
2130                 $p2 = "/\[\[\\|({$tc}+)]]/";                                    # [[|page]]
2131                 $p3 = "/\[\[($namespacechar+):({$np}+)\\|]]/";          # [[namespace:page|]]
2132                 $p4 = "/\[\[($namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/";
2133                                                                                                                 # [[ns:page (cont)|]]
2134                 $context = "";
2135                 $t = $this->mTitle->getText();
2136                 if ( preg_match( $conpat, $t, $m ) ) {
2137                         $context = $m[2];
2138                 }
2139                 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2140                 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2141                 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2142
2143                 if ( '' == $context ) {
2144                         $text = preg_replace( $p2, '[[\\1]]', $text );
2145                 } else {
2146                         $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2147                 }
2148
2149                 /*
2150                 $mw =& MagicWord::get( MAG_SUBST );
2151                 $wgCurParser = $this->fork();
2152                 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
2153                 $this->merge( $wgCurParser );
2154                 */
2155
2156                 # Trim trailing whitespace
2157                 # MAG_END (__END__) tag allows for trailing
2158                 # whitespace to be deliberately included
2159                 $text = rtrim( $text );
2160                 $mw =& MagicWord::get( MAG_END );
2161                 $mw->matchAndRemove( $text );
2162
2163                 return $text;
2164         }
2165
2166         # Set up some variables which are usually set up in parse()
2167         # so that an external function can call some class members with confidence
2168         function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2169                 $this->mTitle =& $title;
2170                 $this->mOptions = $options;
2171                 $this->mOutputType = $outputType;
2172                 if ( $clearState ) {
2173                         $this->clearState();
2174                 }
2175         }
2176
2177         function transformMsg( $text, $options ) {
2178                 global $wgTitle;
2179                 static $executing = false;
2180
2181                 # Guard against infinite recursion
2182                 if ( $executing ) {
2183                         return $text;
2184                 }
2185                 $executing = true;
2186
2187                 $this->mTitle = $wgTitle;
2188                 $this->mOptions = $options;
2189                 $this->mOutputType = OT_MSG;
2190                 $this->clearState();
2191                 $text = $this->replaceVariables( $text );
2192
2193                 $executing = false;
2194                 return $text;
2195         }
2196
2197         # Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2198         # Callback will be called with the text within
2199         # Transform and return the text within
2200         function setHook( $tag, $callback ) {
2201                 $oldVal = @$this->mTagHooks[$tag];
2202                 $this->mTagHooks[$tag] = $callback;
2203                 return $oldVal;
2204         }
2205 }
2206
2207 class ParserOutput
2208 {
2209         var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2210         var $mCacheTime; # Used in ParserCache
2211
2212         function ParserOutput( $text = "", $languageLinks = array(), $categoryLinks = array(),
2213                 $containsOldMagic = false )
2214         {
2215                 $this->mText = $text;
2216                 $this->mLanguageLinks = $languageLinks;
2217                 $this->mCategoryLinks = $categoryLinks;
2218                 $this->mContainsOldMagic = $containsOldMagic;
2219                 $this->mCacheTime = "";
2220         }
2221
2222         function getText() { return $this->mText; }
2223         function getLanguageLinks() { return $this->mLanguageLinks; }
2224         function getCategoryLinks() { return $this->mCategoryLinks; }
2225         function getCacheTime() { return $this->mCacheTime; }
2226         function containsOldMagic() { return $this->mContainsOldMagic; }
2227         function setText( $text ) { return wfSetVar( $this->mText, $text ); }
2228         function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks, $ll ); }
2229         function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks, $cl ); }
2230         function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic, $com ); }
2231         function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime, $t ); }
2232
2233         function merge( $other ) {
2234                 $this->mLanguageLinks = array_merge( $this->mLanguageLinks, $other->mLanguageLinks );
2235                 $this->mCategoryLinks = array_merge( $this->mCategoryLinks, $this->mLanguageLinks );
2236                 $this->mContainsOldMagic = $this->mContainsOldMagic || $other->mContainsOldMagic;
2237         }
2238
2239 }
2240
2241 class ParserOptions
2242 {
2243         # All variables are private
2244         var $mUseTeX;                    # Use texvc to expand <math> tags
2245         var $mUseCategoryMagic;          # Treat [[Category:xxxx]] tags specially
2246         var $mUseDynamicDates;           # Use $wgDateFormatter to format dates
2247         var $mInterwikiMagic;            # Interlanguage links are removed and returned in an array
2248         var $mAllowExternalImages;       # Allow external images inline
2249         var $mSkin;                      # Reference to the preferred skin
2250         var $mDateFormat;                # Date format index
2251         var $mEditSection;               # Create "edit section" links
2252         var $mEditSectionOnRightClick;   # Generate JavaScript to edit section on right click
2253         var $mNumberHeadings;            # Automatically number headings
2254         var $mShowToc;                   # Show table of contents
2255
2256         function getUseTeX()                        { return $this->mUseTeX; }
2257         function getUseCategoryMagic()              { return $this->mUseCategoryMagic; }
2258         function getUseDynamicDates()               { return $this->mUseDynamicDates; }
2259         function getInterwikiMagic()                { return $this->mInterwikiMagic; }
2260         function getAllowExternalImages()           { return $this->mAllowExternalImages; }
2261         function getSkin()                          { return $this->mSkin; }
2262         function getDateFormat()                    { return $this->mDateFormat; }
2263         function getEditSection()                   { return $this->mEditSection; }
2264         function getEditSectionOnRightClick()       { return $this->mEditSectionOnRightClick; }
2265         function getNumberHeadings()                { return $this->mNumberHeadings; }
2266         function getShowToc()                       { return $this->mShowToc; }
2267
2268         function setUseTeX( $x )                    { return wfSetVar( $this->mUseTeX, $x ); }
2269         function setUseCategoryMagic( $x )          { return wfSetVar( $this->mUseCategoryMagic, $x ); }
2270         function setUseDynamicDates( $x )           { return wfSetVar( $this->mUseDynamicDates, $x ); }
2271         function setInterwikiMagic( $x )            { return wfSetVar( $this->mInterwikiMagic, $x ); }
2272         function setAllowExternalImages( $x )       { return wfSetVar( $this->mAllowExternalImages, $x ); }
2273         function setDateFormat( $x )                { return wfSetVar( $this->mDateFormat, $x ); }
2274         function setEditSection( $x )               { return wfSetVar( $this->mEditSection, $x ); }
2275         function setEditSectionOnRightClick( $x )   { return wfSetVar( $this->mEditSectionOnRightClick, $x ); }
2276         function setNumberHeadings( $x )            { return wfSetVar( $this->mNumberHeadings, $x ); }
2277         function setShowToc( $x )                   { return wfSetVar( $this->mShowToc, $x ); }
2278
2279     function setSkin( &$x ) { $this->mSkin =& $x; }
2280
2281         /* static */ function newFromUser( &$user ) {
2282                 $popts = new ParserOptions;
2283                 $popts->initialiseFromUser( $user );
2284                 return $popts;
2285         }
2286
2287         function initialiseFromUser( &$userInput ) {
2288                 global $wgUseTeX, $wgUseCategoryMagic, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2289
2290                 if ( !$userInput ) {
2291                         $user = new User;
2292                         $user->setLoaded( true );
2293                 } else {
2294                         $user =& $userInput;
2295                 }
2296
2297                 $this->mUseTeX = $wgUseTeX;
2298                 $this->mUseCategoryMagic = $wgUseCategoryMagic;
2299                 $this->mUseDynamicDates = $wgUseDynamicDates;
2300                 $this->mInterwikiMagic = $wgInterwikiMagic;
2301                 $this->mAllowExternalImages = $wgAllowExternalImages;
2302                 $this->mSkin =& $user->getSkin();
2303                 $this->mDateFormat = $user->getOption( 'date' );
2304                 $this->mEditSection = $user->getOption( 'editsection' );
2305                 $this->mEditSectionOnRightClick = $user->getOption( 'editsectiononrightclick' );
2306                 $this->mNumberHeadings = $user->getOption( 'numberheadings' );
2307                 $this->mShowToc = $user->getOption( 'showtoc' );
2308         }
2309
2310
2311 }
2312
2313 # Regex callbacks, used in Parser::replaceVariables
2314 function wfBraceSubstitution( $matches )
2315 {
2316         global $wgCurParser;
2317         return $wgCurParser->braceSubstitution( $matches );
2318 }
2319
2320 function wfArgSubstitution( $matches )
2321 {
2322         global $wgCurParser;
2323         return $wgCurParser->argSubstitution( $matches );
2324 }
2325
2326 function wfVariableSubstitution( $matches )
2327 {
2328         global $wgCurParser;
2329         return $wgCurParser->variableSubstitution( $matches );
2330 }
2331
2332 ?>