includes/Parser.php

   1 <?php
   2 /**
   3  * File for Parser and related classes
   4  *
   5  * @package MediaWiki
   6  */
   7
   8 /** */
   9 require_once( 'Sanitizer.php' );
  10
  11 /**
  12  * Update this version number when the ParserOutput format
  13  * changes in an incompatible way, so the parser cache
  14  * can automatically discard old data.
  15  */
  16 define( 'MW_PARSER_VERSION', '1.5.0' );
  17
  18 /**
  19  * Variable substitution O(N^2) attack
  20  *
  21  * Without countermeasures, it would be possible to attack the parser by saving
  22  * a page filled with a large number of inclusions of large pages. The size of
  23  * the generated page would be proportional to the square of the input size.
  24  * Hence, we limit the number of inclusions of any given page, thus bringing any
  25  * attack back to O(N).
  26  */
  27
  28 define( 'MAX_INCLUDE_REPEAT', 100 );
  29 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
  30
  31 define( 'RLH_FOR_UPDATE', 1 );
  32
  33 # Allowed values for $mOutputType
  34 define( 'OT_HTML', 1 );
  35 define( 'OT_WIKI', 2 );
  36 define( 'OT_MSG' , 3 );
  37
  38 # string parameter for extractTags which will cause it
  39 # to strip HTML comments in addition to regular
  40 # <XML>-style tags. This should not be anything we
  41 # may want to use in wikisyntax
  42 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
  43
  44 # prefix for escaping, used in two functions at least
  45 define( 'UNIQ_PREFIX', 'NaodW29');
  46
  47 # Constants needed for external link processing
  48 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
  49 define( 'HTTP_PROTOCOLS', 'http|https' );
  50 # Everything except bracket, space, or control characters
  51 define( 'EXT_LINK_URL_CLASS', '[^]<>"\\x00-\\x20\\x7F]' );
  52 # Including space
  53 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
  54 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
  55 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
  56 define( 'EXT_LINK_BRACKETED',  '/\[(\b('.URL_PROTOCOLS.'):'.EXT_LINK_URL_CLASS.'+) *('.EXT_LINK_TEXT_CLASS.'*?)\]/S' );
  57 define( 'EXT_IMAGE_REGEX',
  58         '/^('.HTTP_PROTOCOLS.':)'.  # Protocol
  59         '('.EXT_LINK_URL_CLASS.'+)\\/'.  # Hostname and path
  60         '('.EXT_IMAGE_FNAME_CLASS.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS.')$/S' # Filename
  61 );
  62
  63 /**
  64  * PHP Parser
  65  *
  66  * Processes wiki markup
  67  *
  68  * <pre>
  69  * There are three main entry points into the Parser class:
  70  * parse()
  71  *   produces HTML output
  72  * preSaveTransform().
  73  *   produces altered wiki markup.
  74  * transformMsg()
  75  *   performs brace substitution on MediaWiki messages
  76  *
  77  * Globals used:
  78  *    objects:   $wgLang, $wgDateFormatter, $wgLinkCache
  79  *
  80  * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
  81  *
  82  * settings:
  83  *  $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
  84  *  $wgNamespacesWithSubpages, $wgAllowExternalImages*,
  85  *  $wgLocaltimezone
  86  *
  87  *  * only within ParserOptions
  88  * </pre>
  89  *
  90  * @package MediaWiki
  91  */
  92 class Parser
  93 {
  94         /**#@+
  95          * @access private
  96          */
  97         # Persistent:
  98         var $mTagHooks;
  99
 100         # Cleared with clearState():
 101         var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
 102         var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
 103
 104         # Temporary:
 105         var $mOptions, $mTitle, $mOutputType,
 106             $mTemplates,        // cache of already loaded templates, avoids
 107                                 // multiple SQL queries for the same string
 108             $mTemplatePath;     // stores an unsorted hash of all the templates already loaded
 109                                 // in this path. Used for loop detection.
 110
 111         /**#@-*/
 112
 113         /**
 114          * Constructor
 115          *
 116          * @access public
 117          */
 118         function Parser() {
 119                 $this->mTemplates = array();
 120                 $this->mTemplatePath = array();
 121                 $this->mTagHooks = array();
 122                 $this->clearState();
 123         }
 124
 125         /**
 126          * Clear Parser state
 127          *
 128          * @access private
 129          */
 130         function clearState() {
 131                 $this->mOutput = new ParserOutput;
 132                 $this->mAutonumber = 0;
 133                 $this->mLastSection = '';
 134                 $this->mDTopen = false;
 135                 $this->mVariables = false;
 136                 $this->mIncludeCount = array();
 137                 $this->mStripState = array();
 138                 $this->mArgStack = array();
 139                 $this->mInPre = false;
 140         }
 141
 142         /**
 143          * First pass--just handle <nowiki> sections, pass the rest off
 144          * to internalParse() which does all the real work.
 145          *
 146          * @access private
 147          * @param string $text Text we want to parse
 148          * @param Title &$title A title object
 149          * @param array $options
 150          * @param boolean $linestart
 151          * @param boolean $clearState
 152          * @return ParserOutput a ParserOutput
 153          */
 154         function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
 155                 global $wgUseTidy, $wgContLang;
 156                 $fname = 'Parser::parse';
 157                 wfProfileIn( $fname );
 158
 159                 if ( $clearState ) {
 160                         $this->clearState();
 161                 }
 162
 163                 $this->mOptions = $options;
 164                 $this->mTitle =& $title;
 165                 $this->mOutputType = OT_HTML;
 166
 167                 $stripState = NULL;
 168                 global $fnord; $fnord = 1;
 169                 //$text = $this->strip( $text, $this->mStripState );
 170                 // VOODOO MAGIC FIX! Sometimes the above segfaults in PHP5.
 171                 $x =& $this->mStripState;
 172                 $text = $this->strip( $text, $x );
 173
 174                 $text = $this->internalParse( $text, $linestart );
 175
 176                 $dashReplace = array(
 177                         '/ - /' => "&nbsp;&ndash; ", # N dash
 178                         '/(?<=[0-9])-(?=[0-9])/' => "&ndash;", # N dash between numbers
 179                         '/ -- /' => "&nbsp;&mdash; " # M dash
 180                 );
 181                 $text = preg_replace( array_keys($dashReplace), array_values($dashReplace), $text );
 182
 183
 184                 $text = $this->unstrip( $text, $this->mStripState );
 185                 # Clean up special characters, only run once, next-to-last before doBlockLevels
 186                 global $wgUseTidy;
 187                 if(!$wgUseTidy) {
 188                         $fixtags = array(
 189                                 # french spaces, last one Guillemet-left
 190                                 # only if there is something before the space
 191                                 '/(.) (?=\\?|:|;|!|\\302\\273)/' => '\\1&nbsp;\\2',
 192                                 # french spaces, Guillemet-right
 193                                 '/(\\302\\253) /' => '\\1&nbsp;',
 194                                 '/<hr *>/i' => '<hr />',
 195                                 '/<br *>/i' => '<br />',
 196                                 '/<center *>/i' => '<div class="center">',
 197                                 '/<\\/center *>/i' => '</div>',
 198                         );
 199                         $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
 200                         $text = Sanitizer::normalizeCharReferences( $text );
 201                 } else {
 202                         $fixtags = array(
 203                                 # french spaces, last one Guillemet-left
 204                                 '/ (\\?|:|;|!|\\302\\273)/' => '&nbsp;\\1',
 205                                 # french spaces, Guillemet-right
 206                                 '/(\\302\\253) /' => '\\1&nbsp;',
 207                                 '/<center *>/i' => '<div class="center">',
 208                                 '/<\\/center *>/i' => '</div>'
 209                         );
 210                         $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
 211                 }
 212                 # only once and last
 213                 $text = $this->doBlockLevels( $text, $linestart );
 214
 215                 $this->replaceLinkHolders( $text );
 216                 $text = $wgContLang->convert($text);
 217
 218                 $text = $this->unstripNoWiki( $text, $this->mStripState );
 219                 if ($wgUseTidy) {
 220                         $text = Parser::tidy($text);
 221                 }
 222
 223                 $this->mOutput->setText( $text );
 224                 wfProfileOut( $fname );
 225                 return $this->mOutput;
 226         }
 227
 228         /**
 229          * Get a random string
 230          *
 231          * @access private
 232          * @static
 233          */
 234         function getRandomString() {
 235                 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
 236         }
 237
 238         /**
 239          * Replaces all occurrences of <$tag>content</$tag> in the text
 240          * with a random marker and returns the new text. the output parameter
 241          * $content will be an associative array filled with data on the form
 242          * $unique_marker => content.
 243          *
 244          * If $content is already set, the additional entries will be appended
 245          * If $tag is set to STRIP_COMMENTS, the function will extract
 246          * <!-- HTML comments -->
 247          *
 248          * @access private
 249          * @static
 250          */
 251         function extractTags($tag, $text, &$content, $uniq_prefix = ''){
 252                 $rnd = $uniq_prefix . '-' . $tag . Parser::getRandomString();
 253                 if ( !$content ) {
 254                         $content = array( );
 255                 }
 256                 $n = 1;
 257                 $stripped = '';
 258
 259                 while ( '' != $text ) {
 260                         if($tag==STRIP_COMMENTS) {
 261                                 $p = preg_split( '/<!--/', $text, 2 );
 262                         } else {
 263                                 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
 264                         }
 265                         $stripped .= $p[0];
 266                         if ( ( count( $p ) < 2 ) || ( '' == $p[1] ) ) {
 267                                 $text = '';
 268                         } else {
 269                                 if($tag==STRIP_COMMENTS) {
 270                                         $q = preg_split( '/-->/i', $p[1], 2 );
 271                                 } else {
 272                                         $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
 273                                 }
 274                                 $marker = $rnd . sprintf('%08X', $n++);
 275                                 $content[$marker] = $q[0];
 276                                 $stripped .= $marker;
 277                                 $text = $q[1];
 278                         }
 279                 }
 280                 return $stripped;
 281         }
 282
 283         /**
 284          * Strips and renders nowiki, pre, math, hiero
 285          * If $render is set, performs necessary rendering operations on plugins
 286          * Returns the text, and fills an array with data needed in unstrip()
 287          * If the $state is already a valid strip state, it adds to the state
 288          *
 289          * @param bool $stripcomments when set, HTML comments <!-- like this -->
 290          *  will be stripped in addition to other tags. This is important
 291          *  for section editing, where these comments cause confusion when
 292          *  counting the sections in the wikisource
 293          *
 294          * @access private
 295          */
 296         function strip( $text, &$state, $stripcomments = false ) {
 297                 $render = ($this->mOutputType == OT_HTML);
 298                 $html_content = array();
 299                 $nowiki_content = array();
 300                 $math_content = array();
 301                 $pre_content = array();
 302                 $comment_content = array();
 303                 $ext_content = array();
 304                 $gallery_content = array();
 305
 306                 # Replace any instances of the placeholders
 307                 $uniq_prefix = UNIQ_PREFIX;
 308                 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
 309
 310                 # html
 311                 global $wgRawHtml, $wgWhitelistEdit;
 312                 if( $wgRawHtml && $wgWhitelistEdit ) {
 313                         $text = Parser::extractTags('html', $text, $html_content, $uniq_prefix);
 314                         foreach( $html_content as $marker => $content ) {
 315                                 if ($render ) {
 316                                         # Raw and unchecked for validity.
 317                                         $html_content[$marker] = $content;
 318                                 } else {
 319                                         $html_content[$marker] = '<html>'.$content.'</html>';
 320                                 }
 321                         }
 322                 }
 323
 324                 # nowiki
 325                 $text = Parser::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
 326                 foreach( $nowiki_content as $marker => $content ) {
 327                         if( $render ){
 328                                 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
 329                         } else {
 330                                 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
 331                         }
 332                 }
 333
 334                 # math
 335                 $text = Parser::extractTags('math', $text, $math_content, $uniq_prefix);
 336                 foreach( $math_content as $marker => $content ){
 337                         if( $render ) {
 338                                 if( $this->mOptions->getUseTeX() ) {
 339                                         $math_content[$marker] = renderMath( $content );
 340                                 } else {
 341                                         $math_content[$marker] = '&lt;math&gt;'.$content.'&lt;math&gt;';
 342                                 }
 343                         } else {
 344                                 $math_content[$marker] = '<math>'.$content.'</math>';
 345                         }
 346                 }
 347
 348                 # pre
 349                 $text = Parser::extractTags('pre', $text, $pre_content, $uniq_prefix);
 350                 foreach( $pre_content as $marker => $content ){
 351                         if( $render ){
 352                                 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
 353                         } else {
 354                                 $pre_content[$marker] = '<pre>'.$content.'</pre>';
 355                         }
 356                 }
 357
 358                 # gallery
 359                 $text = Parser::extractTags('gallery', $text, $gallery_content, $uniq_prefix);
 360                 foreach( $gallery_content as $marker => $content ) {
 361                         require_once( 'ImageGallery.php' );
 362                         if ( $render ) {
 363                                 $gallery_content[$marker] = Parser::renderImageGallery( $content );
 364                         } else {
 365                                 $gallery_content[$marker] = '<gallery>'.$content.'</gallery>';
 366                         }
 367                 }
 368
 369                 # Comments
 370                 if($stripcomments) {
 371                         $text = Parser::extractTags(STRIP_COMMENTS, $text, $comment_content, $uniq_prefix);
 372                         foreach( $comment_content as $marker => $content ){
 373                                 $comment_content[$marker] = '<!--'.$content.'-->';
 374                         }
 375                 }
 376
 377                 # Extensions
 378                 foreach ( $this->mTagHooks as $tag => $callback ) {
 379                         $ext_contents[$tag] = array();
 380                         $text = Parser::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
 381                         foreach( $ext_content[$tag] as $marker => $content ) {
 382                                 if ( $render ) {
 383                                         $ext_content[$tag][$marker] = $callback( $content );
 384                                 } else {
 385                                         $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
 386                                 }
 387                         }
 388                 }
 389
 390                 # Merge state with the pre-existing state, if there is one
 391                 if ( $state ) {
 392                         $state['html'] = $state['html'] + $html_content;
 393                         $state['nowiki'] = $state['nowiki'] + $nowiki_content;
 394                         $state['math'] = $state['math'] + $math_content;
 395                         $state['pre'] = $state['pre'] + $pre_content;
 396                         $state['comment'] = $state['comment'] + $comment_content;
 397                         $state['gallery'] = $state['gallery'] + $gallery_content;
 398
 399                         foreach( $ext_content as $tag => $array ) {
 400                                 if ( array_key_exists( $tag, $state ) ) {
 401                                         $state[$tag] = $state[$tag] + $array;
 402                                 }
 403                         }
 404                 } else {
 405                         $state = array(
 406                           'html' => $html_content,
 407                           'nowiki' => $nowiki_content,
 408                           'math' => $math_content,
 409                           'pre' => $pre_content,
 410                           'comment' => $comment_content,
 411                           'gallery' => $gallery_content,
 412                         ) + $ext_content;
 413                 }
 414                 return $text;
 415         }
 416
 417         /**
 418          * restores pre, math, and hiero removed by strip()
 419          *
 420          * always call unstripNoWiki() after this one
 421          * @access private
 422          */
 423         function unstrip( $text, &$state ) {
 424                 # Must expand in reverse order, otherwise nested tags will be corrupted
 425                 $contentDict = end( $state );
 426                 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
 427                         if( key($state) != 'nowiki' && key($state) != 'html') {
 428                                 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
 429                                         $text = str_replace( key( $contentDict ), $content, $text );
 430                                 }
 431                         }
 432                 }
 433
 434                 return $text;
 435         }
 436
 437         /**
 438          * always call this after unstrip() to preserve the order
 439          *
 440          * @access private
 441          */
 442         function unstripNoWiki( $text, &$state ) {
 443                 # Must expand in reverse order, otherwise nested tags will be corrupted
 444                 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
 445                         $text = str_replace( key( $state['nowiki'] ), $content, $text );
 446                 }
 447
 448                 global $wgRawHtml;
 449                 if ($wgRawHtml) {
 450                         for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
 451                                 $text = str_replace( key( $state['html'] ), $content, $text );
 452                         }
 453                 }
 454
 455                 return $text;
 456         }
 457
 458         /**
 459          * Add an item to the strip state
 460          * Returns the unique tag which must be inserted into the stripped text
 461          * The tag will be replaced with the original text in unstrip()
 462          *
 463          * @access private
 464          */
 465         function insertStripItem( $text, &$state ) {
 466                 $rnd = UNIQ_PREFIX . '-item' . Parser::getRandomString();
 467                 if ( !$state ) {
 468                         $state = array(
 469                           'html' => array(),
 470                           'nowiki' => array(),
 471                           'math' => array(),
 472                           'pre' => array()
 473                         );
 474                 }
 475                 $state['item'][$rnd] = $text;
 476                 return $rnd;
 477         }
 478
 479         /**
 480          * interface with html tidy, used if $wgUseTidy = true
 481          *
 482          * @access public
 483          * @static
 484          */
 485         function tidy ( $text ) {
 486                 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
 487                 $fname = 'Parser::tidy';
 488                 wfProfileIn( $fname );
 489
 490                 $cleansource = '';
 491                 $opts = ' -utf8';
 492
 493                 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
 494 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
 495 '<head><title>test</title></head><body>'.$text.'</body></html>';
 496                 $descriptorspec = array(
 497                         0 => array('pipe', 'r'),
 498                         1 => array('pipe', 'w'),
 499                         2 => array('file', '/dev/null', 'a')
 500                 );
 501                 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
 502                 if (is_resource($process)) {
 503                         fwrite($pipes[0], $wrappedtext);
 504                         fclose($pipes[0]);
 505                         while (!feof($pipes[1])) {
 506                                 $cleansource .= fgets($pipes[1], 1024);
 507                         }
 508                         fclose($pipes[1]);
 509                         $return_value = proc_close($process);
 510                 }
 511
 512                 wfProfileOut( $fname );
 513
 514                 if( $cleansource == '' && $text != '') {
 515                         wfDebug( "Tidy error detected!\n" );
 516                         return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
 517                 } else {
 518                         return $cleansource;
 519                 }
 520         }
 521
 522         /**
 523          * parse the wiki syntax used to render tables
 524          *
 525          * @access private
 526          */
 527         function doTableStuff ( $t ) {
 528                 $fname = 'Parser::doTableStuff';
 529                 wfProfileIn( $fname );
 530
 531                 $t = explode ( "\n" , $t ) ;
 532                 $td = array () ; # Is currently a td tag open?
 533                 $ltd = array () ; # Was it TD or TH?
 534                 $tr = array () ; # Is currently a tr tag open?
 535                 $ltr = array () ; # tr attributes
 536                 $indent_level = 0; # indent level of the table
 537                 foreach ( $t AS $k => $x )
 538                 {
 539                         $x = trim ( $x ) ;
 540                         $fc = substr ( $x , 0 , 1 ) ;
 541                         if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
 542                                 $indent_level = strlen( $matches[1] );
 543                                 $t[$k] = "\n" .
 544                                         str_repeat( '<dl><dd>', $indent_level ) .
 545                                         '<table' . Sanitizer::fixTagAttributes ( $matches[2], 'table' ) . '>' ;
 546                                 array_push ( $td , false ) ;
 547                                 array_push ( $ltd , '' ) ;
 548                                 array_push ( $tr , false ) ;
 549                                 array_push ( $ltr , '' ) ;
 550                         }
 551                         else if ( count ( $td ) == 0 ) { } # Don't do any of the following
 552                         else if ( '|}' == substr ( $x , 0 , 2 ) ) {
 553                                 $z = "</table>" . substr ( $x , 2) . "\n";
 554                                 $l = array_pop ( $ltd ) ;
 555                                 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 556                                 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 557                                 array_pop ( $ltr ) ;
 558                                 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
 559                         }
 560                         else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
 561                                 $x = substr ( $x , 1 ) ;
 562                                 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
 563                                 $z = '' ;
 564                                 $l = array_pop ( $ltd ) ;
 565                                 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 566                                 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 567                                 array_pop ( $ltr ) ;
 568                                 $t[$k] = $z ;
 569                                 array_push ( $tr , false ) ;
 570                                 array_push ( $td , false ) ;
 571                                 array_push ( $ltd , '' ) ;
 572                                 array_push ( $ltr , Sanitizer::fixTagAttributes ( $x, 'tr' ) ) ;
 573                         }
 574                         else if ( '|' == $fc || '!' == $fc || '|+' == substr ( $x , 0 , 2 ) ) { # Caption
 575                                 # $x is a table row
 576                                 if ( '|+' == substr ( $x , 0 , 2 ) ) {
 577                                         $fc = '+' ;
 578                                         $x = substr ( $x , 1 ) ;
 579                                 }
 580                                 $after = substr ( $x , 1 ) ;
 581                                 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
 582                                 $after = explode ( '||' , $after ) ;
 583                                 $t[$k] = '' ;
 584
 585                                 # Loop through each table cell
 586                                 foreach ( $after AS $theline )
 587                                 {
 588                                         $z = '' ;
 589                                         if ( $fc != '+' )
 590                                         {
 591                                                 $tra = array_pop ( $ltr ) ;
 592                                                 if ( !array_pop ( $tr ) ) $z = '<tr'.$tra.">\n" ;
 593                                                 array_push ( $tr , true ) ;
 594                                                 array_push ( $ltr , '' ) ;
 595                                         }
 596
 597                                         $l = array_pop ( $ltd ) ;
 598                                         if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 599                                         if ( $fc == '|' ) $l = 'td' ;
 600                                         else if ( $fc == '!' ) $l = 'th' ;
 601                                         else if ( $fc == '+' ) $l = 'caption' ;
 602                                         else $l = '' ;
 603                                         array_push ( $ltd , $l ) ;
 604
 605                                         # Cell parameters
 606                                         $y = explode ( '|' , $theline , 2 ) ;
 607                                         # Note that a '|' inside an invalid link should not
 608                                         # be mistaken as delimiting cell parameters
 609                                         if ( strpos( $y[0], '[[' ) !== false ) {
 610                                                 $y = array ($theline);
 611                                         }
 612                                         if ( count ( $y ) == 1 )
 613                                                 $y = "{$z}<{$l}>{$y[0]}" ;
 614                                         else $y = $y = "{$z}<{$l}".Sanitizer::fixTagAttributes($y[0], $l).">{$y[1]}" ;
 615                                         $t[$k] .= $y ;
 616                                         array_push ( $td , true ) ;
 617                                 }
 618                         }
 619                 }
 620
 621                 # Closing open td, tr && table
 622                 while ( count ( $td ) > 0 )
 623                 {
 624                         if ( array_pop ( $td ) ) $t[] = '</td>' ;
 625                         if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
 626                         $t[] = '</table>' ;
 627                 }
 628
 629                 $t = implode ( "\n" , $t ) ;
 630                 #               $t = Sanitizer::removeHTMLtags( $t );
 631                 wfProfileOut( $fname );
 632                 return $t ;
 633         }
 634
 635         /**
 636          * Helper function for parse() that transforms wiki markup into
 637          * HTML. Only called for $mOutputType == OT_HTML.
 638          *
 639          * @access private
 640          */
 641         function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
 642                 global $wgContLang;
 643
 644                 $fname = 'Parser::internalParse';
 645                 wfProfileIn( $fname );
 646
 647                 $text = Sanitizer::removeHTMLtags( $text );
 648                 $text = $this->replaceVariables( $text, $args );
 649
 650                 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
 651
 652                 $text = $this->doHeadings( $text );
 653                 if($this->mOptions->getUseDynamicDates()) {
 654                         global $wgDateFormatter;
 655                         $text = $wgDateFormatter->reformat( $this->mOptions->getDateFormat(), $text );
 656                 }
 657                 $text = $this->doAllQuotes( $text );
 658                 $text = $this->replaceInternalLinks( $text );
 659                 $text = $this->replaceExternalLinks( $text );
 660
 661                 # replaceInternalLinks may sometimes leave behind
 662                 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
 663                 $text = str_replace("http-noparse://","http://",$text);
 664
 665                 $text = $this->doMagicLinks( $text );
 666                 $text = $this->doTableStuff( $text );
 667                 $text = $this->formatHeadings( $text, $isMain );
 668
 669                 wfProfileOut( $fname );
 670                 return $text;
 671         }
 672
 673         /**
 674          * Replace special strings like "ISBN xxx" and "RFC xxx" with
 675          * magic external links.
 676          *
 677          * @access private
 678          */
 679         function &doMagicLinks( &$text ) {
 680                 $text = $this->magicISBN( $text );
 681                 $text = $this->magicRFC( $text, 'RFC ', 'rfcurl' );
 682                 $text = $this->magicRFC( $text, 'PMID ', 'pubmedurl' );
 683                 return $text;
 684         }
 685
 686         /**
 687          * Parse ^^ tokens and return html
 688          *
 689          * @access private
 690          */
 691         function doExponent( $text ) {
 692                 $fname = 'Parser::doExponent';
 693                 wfProfileIn( $fname );
 694                 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
 695                 wfProfileOut( $fname );
 696                 return $text;
 697         }
 698
 699         /**
 700          * Parse headers and return html
 701          *
 702          * @access private
 703          */
 704         function doHeadings( $text ) {
 705                 $fname = 'Parser::doHeadings';
 706                 wfProfileIn( $fname );
 707                 for ( $i = 6; $i >= 1; --$i ) {
 708                         $h = substr( '======', 0, $i );
 709                         $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
 710                           "<h{$i}>\\1</h{$i}>\\2", $text );
 711                 }
 712                 wfProfileOut( $fname );
 713                 return $text;
 714         }
 715
 716         /**
 717          * Replace single quotes with HTML markup
 718          * @access private
 719          * @return string the altered text
 720          */
 721         function doAllQuotes( $text ) {
 722                 $fname = 'Parser::doAllQuotes';
 723                 wfProfileIn( $fname );
 724                 $outtext = '';
 725                 $lines = explode( "\n", $text );
 726                 foreach ( $lines as $line ) {
 727                         $outtext .= $this->doQuotes ( $line ) . "\n";
 728                 }
 729                 $outtext = substr($outtext, 0,-1);
 730                 wfProfileOut( $fname );
 731                 return $outtext;
 732         }
 733
 734         /**
 735          * Helper function for doAllQuotes()
 736          * @access private
 737          */
 738         function doQuotes( $text ) {
 739                 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE );
 740                 if ( count( $arr ) == 1 )
 741                         return $text;
 742                 else
 743                 {
 744                         # First, do some preliminary work. This may shift some apostrophes from
 745                         # being mark-up to being text. It also counts the number of occurrences
 746                         # of bold and italics mark-ups.
 747                         $i = 0;
 748                         $numbold = 0;
 749                         $numitalics = 0;
 750                         foreach ( $arr as $r )
 751                         {
 752                                 if ( ( $i % 2 ) == 1 )
 753                                 {
 754                                         # If there are ever four apostrophes, assume the first is supposed to
 755                                         # be text, and the remaining three constitute mark-up for bold text.
 756                                         if ( strlen( $arr[$i] ) == 4 )
 757                                         {
 758                                                 $arr[$i-1] .= "'";
 759                                                 $arr[$i] = "'''";
 760                                         }
 761                                         # If there are more than 5 apostrophes in a row, assume they're all
 762                                         # text except for the last 5.
 763                                         else if ( strlen( $arr[$i] ) > 5 )
 764                                         {
 765                                                 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
 766                                                 $arr[$i] = "'''''";
 767                                         }
 768                                         # Count the number of occurrences of bold and italics mark-ups.
 769                                         # We are not counting sequences of five apostrophes.
 770                                         if ( strlen( $arr[$i] ) == 2 ) $numitalics++;  else
 771                                         if ( strlen( $arr[$i] ) == 3 ) $numbold++;     else
 772                                         if ( strlen( $arr[$i] ) == 5 ) { $numitalics++; $numbold++; }
 773                                 }
 774                                 $i++;
 775                         }
 776
 777                         # If there is an odd number of both bold and italics, it is likely
 778                         # that one of the bold ones was meant to be an apostrophe followed
 779                         # by italics. Which one we cannot know for certain, but it is more
 780                         # likely to be one that has a single-letter word before it.
 781                         if ( ( $numbold % 2 == 1 ) && ( $numitalics % 2 == 1 ) )
 782                         {
 783                                 $i = 0;
 784                                 $firstsingleletterword = -1;
 785                                 $firstmultiletterword = -1;
 786                                 $firstspace = -1;
 787                                 foreach ( $arr as $r )
 788                                 {
 789                                         if ( ( $i % 2 == 1 ) and ( strlen( $r ) == 3 ) )
 790                                         {
 791                                                 $x1 = substr ($arr[$i-1], -1);
 792                                                 $x2 = substr ($arr[$i-1], -2, 1);
 793                                                 if ($x1 == ' ') {
 794                                                         if ($firstspace == -1) $firstspace = $i;
 795                                                 } else if ($x2 == ' ') {
 796                                                         if ($firstsingleletterword == -1) $firstsingleletterword = $i;
 797                                                 } else {
 798                                                         if ($firstmultiletterword == -1) $firstmultiletterword = $i;
 799                                                 }
 800                                         }
 801                                         $i++;
 802                                 }
 803
 804                                 # If there is a single-letter word, use it!
 805                                 if ($firstsingleletterword > -1)
 806                                 {
 807                                         $arr [ $firstsingleletterword ] = "''";
 808                                         $arr [ $firstsingleletterword-1 ] .= "'";
 809                                 }
 810                                 # If not, but there's a multi-letter word, use that one.
 811                                 else if ($firstmultiletterword > -1)
 812                                 {
 813                                         $arr [ $firstmultiletterword ] = "''";
 814                                         $arr [ $firstmultiletterword-1 ] .= "'";
 815                                 }
 816                                 # ... otherwise use the first one that has neither.
 817                                 # (notice that it is possible for all three to be -1 if, for example,
 818                                 # there is only one pentuple-apostrophe in the line)
 819                                 else if ($firstspace > -1)
 820                                 {
 821                                         $arr [ $firstspace ] = "''";
 822                                         $arr [ $firstspace-1 ] .= "'";
 823                                 }
 824                         }
 825
 826                         # Now let's actually convert our apostrophic mush to HTML!
 827                         $output = '';
 828                         $buffer = '';
 829                         $state = '';
 830                         $i = 0;
 831                         foreach ($arr as $r)
 832                         {
 833                                 if (($i % 2) == 0)
 834                                 {
 835                                         if ($state == 'both')
 836                                                 $buffer .= $r;
 837                                         else
 838                                                 $output .= $r;
 839                                 }
 840                                 else
 841                                 {
 842                                         if (strlen ($r) == 2)
 843                                         {
 844                                                 if ($state == 'i')
 845                                                 { $output .= '</i>'; $state = ''; }
 846                                                 else if ($state == 'bi')
 847                                                 { $output .= '</i>'; $state = 'b'; }
 848                                                 else if ($state == 'ib')
 849                                                 { $output .= '</b></i><b>'; $state = 'b'; }
 850                                                 else if ($state == 'both')
 851                                                 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
 852                                                 else # $state can be 'b' or ''
 853                                                 { $output .= '<i>'; $state .= 'i'; }
 854                                         }
 855                                         else if (strlen ($r) == 3)
 856                                         {
 857                                                 if ($state == 'b')
 858                                                 { $output .= '</b>'; $state = ''; }
 859                                                 else if ($state == 'bi')
 860                                                 { $output .= '</i></b><i>'; $state = 'i'; }
 861                                                 else if ($state == 'ib')
 862                                                 { $output .= '</b>'; $state = 'i'; }
 863                                                 else if ($state == 'both')
 864                                                 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
 865                                                 else # $state can be 'i' or ''
 866                                                 { $output .= '<b>'; $state .= 'b'; }
 867                                         }
 868                                         else if (strlen ($r) == 5)
 869                                         {
 870                                                 if ($state == 'b')
 871                                                 { $output .= '</b><i>'; $state = 'i'; }
 872                                                 else if ($state == 'i')
 873                                                 { $output .= '</i><b>'; $state = 'b'; }
 874                                                 else if ($state == 'bi')
 875                                                 { $output .= '</i></b>'; $state = ''; }
 876                                                 else if ($state == 'ib')
 877                                                 { $output .= '</b></i>'; $state = ''; }
 878                                                 else if ($state == 'both')
 879                                                 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
 880                                                 else # ($state == '')
 881                                                 { $buffer = ''; $state = 'both'; }
 882                                         }
 883                                 }
 884                                 $i++;
 885                         }
 886                         # Now close all remaining tags.  Notice that the order is important.
 887                         if ($state == 'b' || $state == 'ib')
 888                                 $output .= '</b>';
 889                         if ($state == 'i' || $state == 'bi' || $state == 'ib')
 890                                 $output .= '</i>';
 891                         if ($state == 'bi')
 892                                 $output .= '</b>';
 893                         if ($state == 'both')
 894                                 $output .= '<b><i>'.$buffer.'</i></b>';
 895                         return $output;
 896                 }
 897         }
 898
 899         /**
 900          * Replace external links
 901          *
 902          * Note: this is all very hackish and the order of execution matters a lot.
 903          * Make sure to run maintenance/parserTests.php if you change this code.
 904          *
 905          * @access private
 906          */
 907         function replaceExternalLinks( $text ) {
 908                 $fname = 'Parser::replaceExternalLinks';
 909                 wfProfileIn( $fname );
 910
 911                 $sk =& $this->mOptions->getSkin();
 912                 global $wgContLang;
 913                 $linktrail = $wgContLang->linkTrail();
 914
 915                 $bits = preg_split( EXT_LINK_BRACKETED, $text, -1, PREG_SPLIT_DELIM_CAPTURE );
 916
 917                 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
 918
 919                 $i = 0;
 920                 while ( $i<count( $bits ) ) {
 921                         $url = $bits[$i++];
 922                         $protocol = $bits[$i++];
 923                         $text = $bits[$i++];
 924                         $trail = $bits[$i++];
 925
 926                         # The characters '<' and '>' (which were escaped by
 927                         # removeHTMLtags()) should not be included in
 928                         # URLs, per RFC 2396.
 929                         if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE)) {
 930                                 $text = substr($url, $m2[0][1]) . ' ' . $text;
 931                                 $url = substr($url, 0, $m2[0][1]);
 932                         }
 933
 934                         # If the link text is an image URL, replace it with an <img> tag
 935                         # This happened by accident in the original parser, but some people used it extensively
 936                         $img = $this->maybeMakeImageLink( $text );
 937                         if ( $img !== false ) {
 938                                 $text = $img;
 939                         }
 940
 941                         $dtrail = '';
 942
 943                         # Set linktype for CSS - if URL==text, link is essentially free
 944                         $linktype = ($text == $url) ? 'free' : 'text';
 945
 946                         # No link text, e.g. [http://domain.tld/some.link]
 947                         if ( $text == '' ) {
 948                                 # Autonumber if allowed
 949                                 if ( strpos( HTTP_PROTOCOLS, $protocol ) !== false ) {
 950                                         $text = '[' . ++$this->mAutonumber . ']';
 951                                         $linktype = 'autonumber';
 952                                 } else {
 953                                         # Otherwise just use the URL
 954                                         $text = htmlspecialchars( $url );
 955                                         $linktype = 'free';
 956                                 }
 957                         } else {
 958                                 # Have link text, e.g. [http://domain.tld/some.link text]s
 959                                 # Check for trail
 960                                 if ( preg_match( $linktrail, $trail, $m2 ) ) {
 961                                         $dtrail = $m2[1];
 962                                         $trail = $m2[2];
 963                                 }
 964                         }
 965
 966                         # Replace &amp; from obsolete syntax with &.
 967                         # All HTML entities will be escaped by makeExternalLink()
 968                         # or maybeMakeImageLink()
 969                         $url = str_replace( '&amp;', '&', $url );
 970
 971                         # Process the trail (i.e. everything after this link up until start of the next link),
 972                         # replacing any non-bracketed links
 973                         $trail = $this->replaceFreeExternalLinks( $trail );
 974
 975                         # Use the encoded URL
 976                         # This means that users can paste URLs directly into the text
 977                         # Funny characters like &ouml; aren't valid in URLs anyway
 978                         # This was changed in August 2004
 979                         $s .= $sk->makeExternalLink( $url, $text, false, $linktype ) . $dtrail . $trail;
 980                 }
 981
 982                 wfProfileOut( $fname );
 983                 return $s;
 984         }
 985
 986         /**
 987          * Replace anything that looks like a URL with a link
 988          * @access private
 989          */
 990         function replaceFreeExternalLinks( $text ) {
 991                 $fname = 'Parser::replaceFreeExternalLinks';
 992                 wfProfileIn( $fname );
 993
 994                 $bits = preg_split( '/(\b(?:'.URL_PROTOCOLS.'):)/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE );
 995                 $s = array_shift( $bits );
 996                 $i = 0;
 997
 998                 $sk =& $this->mOptions->getSkin();
 999
1000                 while ( $i < count( $bits ) ){
1001                         $protocol = $bits[$i++];
1002                         $remainder = $bits[$i++];
1003
1004                         if ( preg_match( '/^('.EXT_LINK_URL_CLASS.'+)(.*)$/s', $remainder, $m ) ) {
1005                                 # Found some characters after the protocol that look promising
1006                                 $url = $protocol . $m[1];
1007                                 $trail = $m[2];
1008
1009                                 # The characters '<' and '>' (which were escaped by
1010                                 # removeHTMLtags()) should not be included in
1011                                 # URLs, per RFC 2396.
1012                                 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE)) {
1013                                         $trail = substr($url, $m2[0][1]) . $trail;
1014                                         $url = substr($url, 0, $m2[0][1]);
1015                                 }
1016
1017                                 # Move trailing punctuation to $trail
1018                                 $sep = ',;\.:!?';
1019                                 # If there is no left bracket, then consider right brackets fair game too
1020                                 if ( strpos( $url, '(' ) === false ) {
1021                                         $sep .= ')';
1022                                 }
1023
1024                                 $numSepChars = strspn( strrev( $url ), $sep );
1025                                 if ( $numSepChars ) {
1026                                         $trail = substr( $url, -$numSepChars ) . $trail;
1027                                         $url = substr( $url, 0, -$numSepChars );
1028                                 }
1029
1030                                 # Replace &amp; from obsolete syntax with &.
1031                                 # All HTML entities will be escaped by makeExternalLink()
1032                                 # or maybeMakeImageLink()
1033                                 $url = str_replace( '&amp;', '&', $url );
1034
1035                                 # Is this an external image?
1036                                 $text = $this->maybeMakeImageLink( $url );
1037                                 if ( $text === false ) {
1038                                         # Not an image, make a link
1039                                         $text = $sk->makeExternalLink( $url, $url, true, 'free' );
1040                                 }
1041                                 $s .= $text . $trail;
1042                         } else {
1043                                 $s .= $protocol . $remainder;
1044                         }
1045                 }
1046                 wfProfileOut();
1047                 return $s;
1048         }
1049
1050         /**
1051          * make an image if it's allowed
1052          * @access private
1053          */
1054         function maybeMakeImageLink( $url ) {
1055                 $sk =& $this->mOptions->getSkin();
1056                 $text = false;
1057                 if ( $this->mOptions->getAllowExternalImages() ) {
1058                         if ( preg_match( EXT_IMAGE_REGEX, $url ) ) {
1059                                 # Image found
1060                                 $text = $sk->makeImage( htmlspecialchars( $url ) );
1061                         }
1062                 }
1063                 return $text;
1064         }
1065
1066         /**
1067          * Process [[ ]] wikilinks
1068          *
1069          * @access private
1070          */
1071         function replaceInternalLinks( $s ) {
1072                 global $wgLang, $wgContLang, $wgLinkCache;
1073                 global $wgDisableLangConversion;
1074                 static $fname = 'Parser::replaceInternalLinks' ;
1075
1076                 wfProfileIn( $fname );
1077
1078                 wfProfileIn( $fname.'-setup' );
1079                 static $tc = FALSE;
1080                 # the % is needed to support urlencoded titles as well
1081                 if ( !$tc ) { $tc = Title::legalChars() . '#%'; }
1082
1083                 $sk =& $this->mOptions->getSkin();
1084                 global $wgUseOldExistenceCheck;
1085                 # "Post-parse link colour check" works only on wiki text since it's now
1086                 # in Parser. Enable it, then disable it when we're done.
1087                 $saveParseColour = $sk->postParseLinkColour( !$wgUseOldExistenceCheck );
1088
1089                 $redirect = MagicWord::get ( MAG_REDIRECT ) ;
1090
1091                 #split the entire text string on occurences of [[
1092                 $a = explode( '[[', ' ' . $s );
1093                 #get the first element (all text up to first [[), and remove the space we added
1094                 $s = array_shift( $a );
1095                 $s = substr( $s, 1 );
1096
1097                 # Match a link having the form [[namespace:link|alternate]]trail
1098                 static $e1 = FALSE;
1099                 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
1100                 # Match cases where there is no "]]", which might still be images
1101                 static $e1_img = FALSE;
1102                 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1103                 # Match the end of a line for a word that's not followed by whitespace,
1104                 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1105                 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1106
1107                 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1108
1109                 if( is_null( $this->mTitle ) ) {
1110                         wfDebugDieBacktrace( 'nooo' );
1111                 }
1112                 $nottalk = !$this->mTitle->isTalkPage();
1113
1114                 if ( $useLinkPrefixExtension ) {
1115                         if ( preg_match( $e2, $s, $m ) ) {
1116                                 $first_prefix = $m[2];
1117                                 $s = $m[1];
1118                         } else {
1119                                 $first_prefix = false;
1120                         }
1121                 } else {
1122                         $prefix = '';
1123                 }
1124
1125                 $selflink = $this->mTitle->getPrefixedText();
1126                 wfProfileOut( $fname.'-setup' );
1127
1128                 $checkVariantLink = sizeof($wgContLang->getVariants())>1;
1129                 $useSubpages = $this->areSubpagesAllowed();
1130
1131                 # Loop for each link
1132                 for ($k = 0; isset( $a[$k] ); $k++) {
1133                         $line = $a[$k];
1134                         if ( $useLinkPrefixExtension ) {
1135                                 wfProfileIn( $fname.'-prefixhandling' );
1136                                 if ( preg_match( $e2, $s, $m ) ) {
1137                                         $prefix = $m[2];
1138                                         $s = $m[1];
1139                                 } else {
1140                                         $prefix='';
1141                                 }
1142                                 # first link
1143                                 if($first_prefix) {
1144                                         $prefix = $first_prefix;
1145                                         $first_prefix = false;
1146                                 }
1147                                 wfProfileOut( $fname.'-prefixhandling' );
1148                         }
1149
1150                         $might_be_img = false;
1151
1152                         if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1153                                 $text = $m[2];
1154                                 # fix up urlencoded title texts
1155                                 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1156                                 $trail = $m[3];
1157                         } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1158                                 $might_be_img = true;
1159                                 $text = $m[2];
1160                                 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1161                                 $trail = "";
1162                         } else { # Invalid form; output directly
1163                                 $s .= $prefix . '[[' . $line ;
1164                                 continue;
1165                         }
1166
1167                         # Don't allow internal links to pages containing
1168                         # PROTO: where PROTO is a valid URL protocol; these
1169                         # should be external links.
1170                         if (preg_match('/^(\b(?:'.URL_PROTOCOLS.'):)/', $m[1])) {
1171                                 $s .= $prefix . '[[' . $line ;
1172                                 continue;
1173                         }
1174
1175                         # Make subpage if necessary
1176                         if( $useSubpages ) {
1177                                 $link = $this->maybeDoSubpageLink( $m[1], $text );
1178                         } else {
1179                                 $link = $m[1];
1180                         }
1181
1182                         $noforce = (substr($m[1], 0, 1) != ':');
1183                         if (!$noforce) {
1184                                 # Strip off leading ':'
1185                                 $link = substr($link, 1);
1186                         }
1187
1188                         $nt =& Title::newFromText( $this->unstripNoWiki($link, $this->mStripState) );
1189                         if( !$nt ) {
1190                                 $s .= $prefix . '[[' . $line;
1191                                 continue;
1192                         }
1193
1194                         #check other language variants of the link
1195                         #if the article does not exist
1196                         if( $checkVariantLink
1197                             && $nt->getArticleID() == 0 ) {
1198                                 $wgContLang->findVariantLink($link, $nt);
1199                         }
1200
1201                         $ns = $nt->getNamespace();
1202                         $iw = $nt->getInterWiki();
1203
1204                         if ($might_be_img) { # if this is actually an invalid link
1205                                 if ($ns == NS_IMAGE && $noforce) { #but might be an image
1206                                         $found = false;
1207                                         while (isset ($a[$k+1]) ) {
1208                                                 #look at the next 'line' to see if we can close it there
1209                                                 $next_line =  array_shift(array_splice( $a, $k + 1, 1) );
1210                                                 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1211                                                 # the first ]] closes the inner link, the second the image
1212                                                         $found = true;
1213                                                         $text .= '[[' . $m[1];
1214                                                         $trail = $m[2];
1215                                                         break;
1216                                                 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1217                                                         #if there's exactly one ]] that's fine, we'll keep looking
1218                                                         $text .= '[[' . $m[0];
1219                                                 } else {
1220                                                         #if $next_line is invalid too, we need look no further
1221                                                         $text .= '[[' . $next_line;
1222                                                         break;
1223                                                 }
1224                                         }
1225                                         if ( !$found ) {
1226                                                 # we couldn't find the end of this imageLink, so output it raw
1227                                                 #but don't ignore what might be perfectly normal links in the text we've examined
1228                                                 $text = $this->replaceInternalLinks($text);
1229                                                 $s .= $prefix . '[[' . $link . '|' . $text;
1230                                                 # note: no $trail, because without an end, there *is* no trail
1231                                                 continue;
1232                                         }
1233                                 } else { #it's not an image, so output it raw
1234                                         $s .= $prefix . '[[' . $link . '|' . $text;
1235                                         # note: no $trail, because without an end, there *is* no trail
1236                                         continue;
1237                                 }
1238                         }
1239
1240                         $wasblank = ( '' == $text );
1241                         if( $wasblank ) $text = $link;
1242
1243
1244                         # Link not escaped by : , create the various objects
1245                         if( $noforce ) {
1246
1247                                 # Interwikis
1248                                 if( $iw && $this->mOptions->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1249                                         array_push( $this->mOutput->mLanguageLinks, $nt->getFullText() );
1250                                         $tmp = $prefix . $trail ;
1251                                         $s .= (trim($tmp) == '')? '': $tmp;
1252                                         continue;
1253                                 }
1254
1255                                 if ( $ns == NS_IMAGE ) {
1256                                         wfProfileIn( "$fname-image" );
1257
1258                                         # recursively parse links inside the image caption
1259                                         # actually, this will parse them in any other parameters, too,
1260                                         # but it might be hard to fix that, and it doesn't matter ATM
1261                                         $text = $this->replaceExternalLinks($text);
1262                                         $text = $this->replaceInternalLinks($text);
1263
1264                                         # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1265                                         $s .= $prefix . str_replace('http://', 'http-noparse://', $sk->makeImageLinkObj( $nt, $text ) ) . $trail;
1266                                         $wgLinkCache->addImageLinkObj( $nt );
1267
1268                                         wfProfileOut( "$fname-image" );
1269                                         continue;
1270                                 }
1271
1272                                 if ( $ns == NS_CATEGORY ) {
1273                                         wfProfileIn( "$fname-category" );
1274                                         $t = $nt->getText();
1275
1276                                         $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1277                                         $pPLC=$sk->postParseLinkColour();
1278                                         $sk->postParseLinkColour( false );
1279                                         $t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
1280                                         $sk->postParseLinkColour( $pPLC );
1281                                         $wgLinkCache->resume();
1282
1283                                         if ( $wasblank ) {
1284                                                 if ( $this->mTitle->getNamespace() == NS_CATEGORY ) {
1285                                                         $sortkey = $this->mTitle->getText();
1286                                                 } else {
1287                                                         $sortkey = $this->mTitle->getPrefixedText();
1288                                                 }
1289                                         } else {
1290                                                 $sortkey = $text;
1291                                         }
1292                                         $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1293                                         $this->mOutput->addCategoryLink( $t );
1294                                         $s .= $prefix . $trail ;
1295
1296                                         wfProfileOut( "$fname-category" );
1297                                         continue;
1298                                 }
1299                         }
1300
1301                         if( ( $nt->getPrefixedText() === $selflink ) &&
1302                             ( $nt->getFragment() === '' ) ) {
1303                                 # Self-links are handled specially; generally de-link and change to bold.
1304                                 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1305                                 continue;
1306                         }
1307
1308                         # Special and Media are pseudo-namespaces; no pages actually exist in them
1309                         if( $ns == NS_MEDIA ) {
1310                                 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text, true ) . $trail;
1311                                 $wgLinkCache->addImageLinkObj( $nt );
1312                                 continue;
1313                         } elseif( $ns == NS_SPECIAL ) {
1314                                 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1315                                 continue;
1316                         }
1317                         $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1318                 }
1319                 $sk->postParseLinkColour( $saveParseColour );
1320                 wfProfileOut( $fname );
1321                 return $s;
1322         }
1323
1324         /**
1325          * Return true if subpage links should be expanded on this page.
1326          * @return bool
1327          */
1328         function areSubpagesAllowed() {
1329                 # Some namespaces don't allow subpages
1330                 global $wgNamespacesWithSubpages;
1331                 return !empty($wgNamespacesWithSubpages[$this->mTitle->getNamespace()]);
1332         }
1333
1334         /**
1335          * Handle link to subpage if necessary
1336          * @param string $target the source of the link
1337          * @param string &$text the link text, modified as necessary
1338          * @return string the full name of the link
1339          * @access private
1340          */
1341         function maybeDoSubpageLink($target, &$text) {
1342                 # Valid link forms:
1343                 # Foobar -- normal
1344                 # :Foobar -- override special treatment of prefix (images, language links)
1345                 # /Foobar -- convert to CurrentPage/Foobar
1346                 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1347                 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1348                 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1349
1350                 $fname = 'Parser::maybeDoSubpageLink';
1351                 wfProfileIn( $fname );
1352                 $ret = $target; # default return value is no change
1353
1354                 # Some namespaces don't allow subpages,
1355                 # so only perform processing if subpages are allowed
1356                 if( $this->areSubpagesAllowed() ) {
1357                         # Look at the first character
1358                         if( $target != '' && $target{0} == '/' ) {
1359                                 # / at end means we don't want the slash to be shown
1360                                 if( substr( $target, -1, 1 ) == '/' ) {
1361                                         $target = substr( $target, 1, -1 );
1362                                         $noslash = $target;
1363                                 } else {
1364                                         $noslash = substr( $target, 1 );
1365                                 }
1366
1367                                 $ret = $this->mTitle->getPrefixedText(). '/' . trim($noslash);
1368                                 if( '' === $text ) {
1369                                         $text = $target;
1370                                 } # this might be changed for ugliness reasons
1371                         } else {
1372                                 # check for .. subpage backlinks
1373                                 $dotdotcount = 0;
1374                                 $nodotdot = $target;
1375                                 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1376                                         ++$dotdotcount;
1377                                         $nodotdot = substr( $nodotdot, 3 );
1378                                 }
1379                                 if($dotdotcount > 0) {
1380                                         $exploded = explode( '/', $this->mTitle->GetPrefixedText() );
1381                                         if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1382                                                 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1383                                                 # / at the end means don't show full path
1384                                                 if( substr( $nodotdot, -1, 1 ) == '/' ) {
1385                                                         $nodotdot = substr( $nodotdot, 0, -1 );
1386                                                         if( '' === $text ) {
1387                                                                 $text = $nodotdot;
1388                                                         }
1389                                                 }
1390                                                 $nodotdot = trim( $nodotdot );
1391                                                 if( $nodotdot != '' ) {
1392                                                         $ret .= '/' . $nodotdot;
1393                                                 }
1394                                         }
1395                                 }
1396                         }
1397                 }
1398
1399                 wfProfileOut( $fname );
1400                 return $ret;
1401         }
1402
1403         /**#@+
1404          * Used by doBlockLevels()
1405          * @access private
1406          */
1407         /* private */ function closeParagraph() {
1408                 $result = '';
1409                 if ( '' != $this->mLastSection ) {
1410                         $result = '</' . $this->mLastSection  . ">\n";
1411                 }
1412                 $this->mInPre = false;
1413                 $this->mLastSection = '';
1414                 return $result;
1415         }
1416         # getCommon() returns the length of the longest common substring
1417         # of both arguments, starting at the beginning of both.
1418         #
1419         /* private */ function getCommon( $st1, $st2 ) {
1420                 $fl = strlen( $st1 );
1421                 $shorter = strlen( $st2 );
1422                 if ( $fl < $shorter ) { $shorter = $fl; }
1423
1424                 for ( $i = 0; $i < $shorter; ++$i ) {
1425                         if ( $st1{$i} != $st2{$i} ) { break; }
1426                 }
1427                 return $i;
1428         }
1429         # These next three functions open, continue, and close the list
1430         # element appropriate to the prefix character passed into them.
1431         #
1432         /* private */ function openList( $char ) {
1433                 $result = $this->closeParagraph();
1434
1435                 if ( '*' == $char ) { $result .= '<ul><li>'; }
1436                 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1437                 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1438                 else if ( ';' == $char ) {
1439                         $result .= '<dl><dt>';
1440                         $this->mDTopen = true;
1441                 }
1442                 else { $result = '<!-- ERR 1 -->'; }
1443
1444                 return $result;
1445         }
1446
1447         /* private */ function nextItem( $char ) {
1448                 if ( '*' == $char || '#' == $char ) { return '</li><li>'; }
1449                 else if ( ':' == $char || ';' == $char ) {
1450                         $close = '</dd>';
1451                         if ( $this->mDTopen ) { $close = '</dt>'; }
1452                         if ( ';' == $char ) {
1453                                 $this->mDTopen = true;
1454                                 return $close . '<dt>';
1455                         } else {
1456                                 $this->mDTopen = false;
1457                                 return $close . '<dd>';
1458                         }
1459                 }
1460                 return '<!-- ERR 2 -->';
1461         }
1462
1463         /* private */ function closeList( $char ) {
1464                 if ( '*' == $char ) { $text = '</li></ul>'; }
1465                 else if ( '#' == $char ) { $text = '</li></ol>'; }
1466                 else if ( ':' == $char ) {
1467                         if ( $this->mDTopen ) {
1468                                 $this->mDTopen = false;
1469                                 $text = '</dt></dl>';
1470                         } else {
1471                                 $text = '</dd></dl>';
1472                         }
1473                 }
1474                 else {  return '<!-- ERR 3 -->'; }
1475                 return $text."\n";
1476         }
1477         /**#@-*/
1478
1479         /**
1480          * Make lists from lines starting with ':', '*', '#', etc.
1481          *
1482          * @access private
1483          * @return string the lists rendered as HTML
1484          */
1485         function doBlockLevels( $text, $linestart ) {
1486                 $fname = 'Parser::doBlockLevels';
1487                 wfProfileIn( $fname );
1488
1489                 # Parsing through the text line by line.  The main thing
1490                 # happening here is handling of block-level elements p, pre,
1491                 # and making lists from lines starting with * # : etc.
1492                 #
1493                 $textLines = explode( "\n", $text );
1494
1495                 $lastPrefix = $output = $lastLine = '';
1496                 $this->mDTopen = $inBlockElem = false;
1497                 $prefixLength = 0;
1498                 $paragraphStack = false;
1499
1500                 if ( !$linestart ) {
1501                         $output .= array_shift( $textLines );
1502                 }
1503                 foreach ( $textLines as $oLine ) {
1504                         $lastPrefixLength = strlen( $lastPrefix );
1505                         $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1506                         $preOpenMatch = preg_match('/<pre/i', $oLine );
1507                         if ( !$this->mInPre ) {
1508                                 # Multiple prefixes may abut each other for nested lists.
1509                                 $prefixLength = strspn( $oLine, '*#:;' );
1510                                 $pref = substr( $oLine, 0, $prefixLength );
1511
1512                                 # eh?
1513                                 $pref2 = str_replace( ';', ':', $pref );
1514                                 $t = substr( $oLine, $prefixLength );
1515                                 $this->mInPre = !empty($preOpenMatch);
1516                         } else {
1517                                 # Don't interpret any other prefixes in preformatted text
1518                                 $prefixLength = 0;
1519                                 $pref = $pref2 = '';
1520                                 $t = $oLine;
1521                         }
1522
1523                         # List generation
1524                         if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1525                                 # Same as the last item, so no need to deal with nesting or opening stuff
1526                                 $output .= $this->nextItem( substr( $pref, -1 ) );
1527                                 $paragraphStack = false;
1528
1529                                 if ( substr( $pref, -1 ) == ';') {
1530                                         # The one nasty exception: definition lists work like this:
1531                                         # ; title : definition text
1532                                         # So we check for : in the remainder text to split up the
1533                                         # title and definition, without b0rking links.
1534                                         if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1535                                                 $t = $t2;
1536                                                 $output .= $term . $this->nextItem( ':' );
1537                                         }
1538                                 }
1539                         } elseif( $prefixLength || $lastPrefixLength ) {
1540                                 # Either open or close a level...
1541                                 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1542                                 $paragraphStack = false;
1543
1544                                 while( $commonPrefixLength < $lastPrefixLength ) {
1545                                         $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1546                                         --$lastPrefixLength;
1547                                 }
1548                                 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1549                                         $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1550                                 }
1551                                 while ( $prefixLength > $commonPrefixLength ) {
1552                                         $char = substr( $pref, $commonPrefixLength, 1 );
1553                                         $output .= $this->openList( $char );
1554
1555                                         if ( ';' == $char ) {
1556                                                 # FIXME: This is dupe of code above
1557                                                 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1558                                                         $t = $t2;
1559                                                         $output .= $term . $this->nextItem( ':' );
1560                                                 }
1561                                         }
1562                                         ++$commonPrefixLength;
1563                                 }
1564                                 $lastPrefix = $pref2;
1565                         }
1566                         if( 0 == $prefixLength ) {
1567                                 wfProfileIn( "$fname-paragraph" );
1568                                 # No prefix (not in list)--go to paragraph mode
1569                                 $uniq_prefix = UNIQ_PREFIX;
1570                                 // XXX: use a stack for nestable elements like span, table and div
1571                                 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
1572                                 $closematch = preg_match(
1573                                         '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1574                                         '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/iS', $t );
1575                                 if ( $openmatch or $closematch ) {
1576                                         $paragraphStack = false;
1577                                         $output .= $this->closeParagraph();
1578                                         if($preOpenMatch and !$preCloseMatch) {
1579                                                 $this->mInPre = true;
1580                                         }
1581                                         if ( $closematch ) {
1582                                                 $inBlockElem = false;
1583                                         } else {
1584                                                 $inBlockElem = true;
1585                                         }
1586                                 } else if ( !$inBlockElem && !$this->mInPre ) {
1587                                         if ( ' ' == $t{0} and ( $this->mLastSection == 'pre' or trim($t) != '' ) ) {
1588                                                 // pre
1589                                                 if ($this->mLastSection != 'pre') {
1590                                                         $paragraphStack = false;
1591                                                         $output .= $this->closeParagraph().'<pre>';
1592                                                         $this->mLastSection = 'pre';
1593                                                 }
1594                                                 $t = substr( $t, 1 );
1595                                         } else {
1596                                                 // paragraph
1597                                                 if ( '' == trim($t) ) {
1598                                                         if ( $paragraphStack ) {
1599                                                                 $output .= $paragraphStack.'<br />';
1600                                                                 $paragraphStack = false;
1601                                                                 $this->mLastSection = 'p';
1602                                                         } else {
1603                                                                 if ($this->mLastSection != 'p' ) {
1604                                                                         $output .= $this->closeParagraph();
1605                                                                         $this->mLastSection = '';
1606                                                                         $paragraphStack = '<p>';
1607                                                                 } else {
1608                                                                         $paragraphStack = '</p><p>';
1609                                                                 }
1610                                                         }
1611                                                 } else {
1612                                                         if ( $paragraphStack ) {
1613                                                                 $output .= $paragraphStack;
1614                                                                 $paragraphStack = false;
1615                                                                 $this->mLastSection = 'p';
1616                                                         } else if ($this->mLastSection != 'p') {
1617                                                                 $output .= $this->closeParagraph().'<p>';
1618                                                                 $this->mLastSection = 'p';
1619                                                         }
1620                                                 }
1621                                         }
1622                                 }
1623                                 wfProfileOut( "$fname-paragraph" );
1624                         }
1625                         if ($paragraphStack === false) {
1626                                 $output .= $t."\n";
1627                         }
1628                 }
1629                 while ( $prefixLength ) {
1630                         $output .= $this->closeList( $pref2{$prefixLength-1} );
1631                         --$prefixLength;
1632                 }
1633                 if ( '' != $this->mLastSection ) {
1634                         $output .= '</' . $this->mLastSection . '>';
1635                         $this->mLastSection = '';
1636                 }
1637
1638                 wfProfileOut( $fname );
1639                 return $output;
1640         }
1641
1642         /**
1643          * Split up a string on ':', ignoring any occurences inside
1644          * <a>..</a> or <span>...</span>
1645          * @param string $str the string to split
1646          * @param string &$before set to everything before the ':'
1647          * @param string &$after set to everything after the ':'
1648          * return string the position of the ':', or false if none found
1649          */
1650         function findColonNoLinks($str, &$before, &$after) {
1651                 # I wonder if we should make this count all tags, not just <a>
1652                 # and <span>. That would prevent us from matching a ':' that
1653                 # comes in the middle of italics other such formatting....
1654                 # -- Wil
1655                 $fname = 'Parser::findColonNoLinks';
1656                 wfProfileIn( $fname );
1657                 $pos = 0;
1658                 do {
1659                         $colon = strpos($str, ':', $pos);
1660
1661                         if ($colon !== false) {
1662                                 $before = substr($str, 0, $colon);
1663                                 $after = substr($str, $colon + 1);
1664
1665                                 # Skip any ':' within <a> or <span> pairs
1666                                 $a = substr_count($before, '<a');
1667                                 $s = substr_count($before, '<span');
1668                                 $ca = substr_count($before, '</a>');
1669                                 $cs = substr_count($before, '</span>');
1670
1671                                 if ($a <= $ca and $s <= $cs) {
1672                                         # Tags are balanced before ':'; ok
1673                                         break;
1674                                 }
1675                                 $pos = $colon + 1;
1676                         }
1677                 } while ($colon !== false);
1678                 wfProfileOut( $fname );
1679                 return $colon;
1680         }
1681
1682         /**
1683          * Return value of a magic variable (like PAGENAME)
1684          *
1685          * @access private
1686          */
1687         function getVariableValue( $index ) {
1688                 global $wgContLang, $wgSitename, $wgServer, $wgArticle;
1689
1690                 /**
1691                  * Some of these require message or data lookups and can be
1692                  * expensive to check many times.
1693                  */
1694                 static $varCache = array();
1695                 if( isset( $varCache[$index] ) ) return $varCache[$index];
1696
1697                 switch ( $index ) {
1698                         case MAG_CURRENTMONTH:
1699                                 return $varCache[$index] = $wgContLang->formatNum( date( 'm' ) );
1700                         case MAG_CURRENTMONTHNAME:
1701                                 return $varCache[$index] = $wgContLang->getMonthName( date('n') );
1702                         case MAG_CURRENTMONTHNAMEGEN:
1703                                 return $varCache[$index] = $wgContLang->getMonthNameGen( date('n') );
1704                         case MAG_CURRENTMONTHABBREV:
1705                                 return $varCache[$index] = $wgContLang->getMonthAbbreviation( date('n') );
1706                         case MAG_CURRENTDAY:
1707                                 return $varCache[$index] = $wgContLang->formatNum( date('j') );
1708                         case MAG_PAGENAME:
1709                                 return $this->mTitle->getText();
1710                         case MAG_PAGENAMEE:
1711                                 return $this->mTitle->getPartialURL();
1712                         case MAG_REVISIONID:
1713                                 return $wgArticle->getRevIdFetched();
1714                         case MAG_NAMESPACE:
1715                                 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1716                                 return $wgContLang->getNsText($this->mTitle->getNamespace()); # Patch by Dori
1717                         case MAG_CURRENTDAYNAME:
1718                                 return $varCache[$index] = $wgContLang->getWeekdayName( date('w')+1 );
1719                         case MAG_CURRENTYEAR:
1720                                 return $varCache[$index] = $wgContLang->formatNum( date( 'Y' ) );
1721                         case MAG_CURRENTTIME:
1722                                 return $varCache[$index] = $wgContLang->time( wfTimestampNow(), false );
1723                         case MAG_CURRENTWEEK:
1724                                 return $varCache[$index] = $wgContLang->formatNum( date('W') );
1725                         case MAG_CURRENTDOW:
1726                                 return $varCache[$index] = $wgContLang->formatNum( date('w') );
1727                         case MAG_NUMBEROFARTICLES:
1728                                 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfArticles() );
1729                         case MAG_SITENAME:
1730                                 return $wgSitename;
1731                         case MAG_SERVER:
1732                                 return $wgServer;
1733                         default:
1734                                 return NULL;
1735                 }
1736         }
1737
1738         /**
1739          * initialise the magic variables (like CURRENTMONTHNAME)
1740          *
1741          * @access private
1742          */
1743         function initialiseVariables() {
1744                 $fname = 'Parser::initialiseVariables';
1745                 wfProfileIn( $fname );
1746                 global $wgVariableIDs;
1747                 $this->mVariables = array();
1748                 foreach ( $wgVariableIDs as $id ) {
1749                         $mw =& MagicWord::get( $id );
1750                         $mw->addToArray( $this->mVariables, $id );
1751                 }
1752                 wfProfileOut( $fname );
1753         }
1754
1755         /**
1756          * Replace magic variables, templates, and template arguments
1757          * with the appropriate text. Templates are substituted recursively,
1758          * taking care to avoid infinite loops.
1759          *
1760          * Note that the substitution depends on value of $mOutputType:
1761          *  OT_WIKI: only {{subst:}} templates
1762          *  OT_MSG: only magic variables
1763          *  OT_HTML: all templates and magic variables
1764          *
1765          * @param string $tex The text to transform
1766          * @param array $args Key-value pairs representing template parameters to substitute
1767          * @access private
1768          */
1769         function replaceVariables( $text, $args = array() ) {
1770                 global $wgLang, $wgScript, $wgArticlePath;
1771
1772                 # Prevent too big inclusions
1773                 if( strlen( $text ) > MAX_INCLUDE_SIZE ) {
1774                         return $text;
1775                 }
1776
1777                 $fname = 'Parser::replaceVariables';
1778                 wfProfileIn( $fname );
1779
1780                 $titleChars = Title::legalChars();
1781
1782                 # This function is called recursively. To keep track of arguments we need a stack:
1783                 array_push( $this->mArgStack, $args );
1784
1785                 # Variable substitution
1786                 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", array( &$this, 'variableSubstitution' ), $text );
1787
1788                 if ( $this->mOutputType == OT_HTML || $this->mOutputType == OT_WIKI ) {
1789                         # Argument substitution
1790                         $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", array( &$this, 'argSubstitution' ), $text );
1791                 }
1792                 # Template substitution
1793                 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1794                 $text = preg_replace_callback( $regex, array( &$this, 'braceSubstitution' ), $text );
1795
1796                 array_pop( $this->mArgStack );
1797
1798                 wfProfileOut( $fname );
1799                 return $text;
1800         }
1801
1802         /**
1803          * Replace magic variables
1804          * @access private
1805          */
1806         function variableSubstitution( $matches ) {
1807                 $fname = 'parser::variableSubstitution';
1808                 $varname = $matches[1];
1809                 wfProfileIn( $fname );
1810                 if ( !$this->mVariables ) {
1811                         $this->initialiseVariables();
1812                 }
1813                 $skip = false;
1814                 if ( $this->mOutputType == OT_WIKI ) {
1815                         # Do only magic variables prefixed by SUBST
1816                         $mwSubst =& MagicWord::get( MAG_SUBST );
1817                         if (!$mwSubst->matchStartAndRemove( $varname ))
1818                                 $skip = true;
1819                         # Note that if we don't substitute the variable below,
1820                         # we don't remove the {{subst:}} magic word, in case
1821                         # it is a template rather than a magic variable.
1822                 }
1823                 if ( !$skip && array_key_exists( $varname, $this->mVariables ) ) {
1824                         $id = $this->mVariables[$varname];
1825                         $text = $this->getVariableValue( $id );
1826                         $this->mOutput->mContainsOldMagic = true;
1827                 } else {
1828                         $text = $matches[0];
1829                 }
1830                 wfProfileOut( $fname );
1831                 return $text;
1832         }
1833
1834         # Split template arguments
1835         function getTemplateArgs( $argsString ) {
1836                 if ( $argsString === '' ) {
1837                         return array();
1838                 }
1839
1840                 $args = explode( '|', substr( $argsString, 1 ) );
1841
1842                 # If any of the arguments contains a '[[' but no ']]', it needs to be
1843                 # merged with the next arg because the '|' character between belongs
1844                 # to the link syntax and not the template parameter syntax.
1845                 $argc = count($args);
1846                 $i = 0;
1847                 for ( $i = 0; $i < $argc-1; $i++ ) {
1848                         if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1849                                 $args[$i] .= '|'.$args[$i+1];
1850                                 array_splice($args, $i+1, 1);
1851                                 $i--;
1852                                 $argc--;
1853                         }
1854                 }
1855
1856                 return $args;
1857         }
1858
1859         /**
1860          * Return the text of a template, after recursively
1861          * replacing any variables or templates within the template.
1862          *
1863          * @param array $matches The parts of the template
1864          *  $matches[1]: the title, i.e. the part before the |
1865          *  $matches[2]: the parameters (including a leading |), if  any
1866          * @return string the text of the template
1867          * @access private
1868          */
1869         function braceSubstitution( $matches ) {
1870                 global $wgLinkCache, $wgContLang;
1871                 $fname = 'Parser::braceSubstitution';
1872                 wfProfileIn( $fname );
1873
1874                 $found = false;
1875                 $nowiki = false;
1876                 $noparse = false;
1877
1878                 $title = NULL;
1879
1880                 # Need to know if the template comes at the start of a line,
1881                 # to treat the beginning of the template like the beginning
1882                 # of a line for tables and block-level elements.
1883                 $linestart = $matches[1];
1884
1885                 # $part1 is the bit before the first |, and must contain only title characters
1886                 # $args is a list of arguments, starting from index 0, not including $part1
1887
1888                 $part1 = $matches[2];
1889                 # If the third subpattern matched anything, it will start with |
1890
1891                 $args = $this->getTemplateArgs($matches[3]);
1892                 $argc = count( $args );
1893
1894                 # Don't parse {{{}}} because that's only for template arguments
1895                 if ( $linestart === '{' ) {
1896                         $text = $matches[0];
1897                         $found = true;
1898                         $noparse = true;
1899                 }
1900
1901                 # SUBST
1902                 if ( !$found ) {
1903                         $mwSubst =& MagicWord::get( MAG_SUBST );
1904                         if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType == OT_WIKI) ) {
1905                                 # One of two possibilities is true:
1906                                 # 1) Found SUBST but not in the PST phase
1907                                 # 2) Didn't find SUBST and in the PST phase
1908                                 # In either case, return without further processing
1909                                 $text = $matches[0];
1910                                 $found = true;
1911                                 $noparse = true;
1912                         }
1913                 }
1914
1915                 # MSG, MSGNW and INT
1916                 if ( !$found ) {
1917                         # Check for MSGNW:
1918                         $mwMsgnw =& MagicWord::get( MAG_MSGNW );
1919                         if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1920                                 $nowiki = true;
1921                         } else {
1922                                 # Remove obsolete MSG:
1923                                 $mwMsg =& MagicWord::get( MAG_MSG );
1924                                 $mwMsg->matchStartAndRemove( $part1 );
1925                         }
1926
1927                         # Check if it is an internal message
1928                         $mwInt =& MagicWord::get( MAG_INT );
1929                         if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1930                                 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1931                                         $text = $linestart . wfMsgReal( $part1, $args, true );
1932                                         $found = true;
1933                                 }
1934                         }
1935                 }
1936
1937                 # NS
1938                 if ( !$found ) {
1939                         # Check for NS: (namespace expansion)
1940                         $mwNs = MagicWord::get( MAG_NS );
1941                         if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1942                                 if ( intval( $part1 ) ) {
1943                                         $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1944                                         $found = true;
1945                                 } else {
1946                                         $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1947                                         if ( !is_null( $index ) ) {
1948                                                 $text = $linestart . $wgContLang->getNsText( $index );
1949                                                 $found = true;
1950                                         }
1951                                 }
1952                         }
1953                 }
1954
1955                 # LOCALURL and LOCALURLE
1956                 if ( !$found ) {
1957                         $mwLocal = MagicWord::get( MAG_LOCALURL );
1958                         $mwLocalE = MagicWord::get( MAG_LOCALURLE );
1959
1960                         if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1961                                 $func = 'getLocalURL';
1962                         } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1963                                 $func = 'escapeLocalURL';
1964                         } else {
1965                                 $func = '';
1966                         }
1967
1968                         if ( $func !== '' ) {
1969                                 $title = Title::newFromText( $part1 );
1970                                 if ( !is_null( $title ) ) {
1971                                         if ( $argc > 0 ) {
1972                                                 $text = $linestart . $title->$func( $args[0] );
1973                                         } else {
1974                                                 $text = $linestart . $title->$func();
1975                                         }
1976                                         $found = true;
1977                                 }
1978                         }
1979                 }
1980
1981                 # GRAMMAR
1982                 if ( !$found && $argc == 1 ) {
1983                         $mwGrammar =& MagicWord::get( MAG_GRAMMAR );
1984                         if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1985                                 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
1986                                 $found = true;
1987                         }
1988                 }
1989
1990                 # Template table test
1991
1992                 # Did we encounter this template already? If yes, it is in the cache
1993                 # and we need to check for loops.
1994                 if ( !$found && isset( $this->mTemplates[$part1] ) ) {
1995                         $found = true;
1996
1997                         # Infinite loop test
1998                         if ( isset( $this->mTemplatePath[$part1] ) ) {
1999                                 $noparse = true;
2000                                 $found = true;
2001                                 $text = $linestart .
2002                                         "\{\{$part1}}" .
2003                                         '<!-- WARNING: template loop detected -->';
2004                                 wfDebug( "$fname: template loop broken at '$part1'\n" );
2005                         } else {
2006                                 # set $text to cached message.
2007                                 $text = $linestart . $this->mTemplates[$part1];
2008                         }
2009                 }
2010
2011                 # Load from database
2012                 $itcamefromthedatabase = false;
2013                 $lastPathLevel = $this->mTemplatePath;
2014                 if ( !$found ) {
2015                         $ns = NS_TEMPLATE;
2016                         $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
2017                         if ($subpage !== '') {
2018                                 $ns = $this->mTitle->getNamespace();
2019                         }
2020                         $title = Title::newFromText( $part1, $ns );
2021                         if ( !is_null( $title ) && !$title->isExternal() ) {
2022                                 # Check for excessive inclusion
2023                                 $dbk = $title->getPrefixedDBkey();
2024                                 if ( $this->incrementIncludeCount( $dbk ) ) {
2025                                         # This should never be reached.
2026                                         $article = new Article( $title );
2027                                         $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
2028                                         if ( $articleContent !== false ) {
2029                                                 $found = true;
2030                                                 $text = $linestart . $articleContent;
2031                                                 $itcamefromthedatabase = true;
2032                                         }
2033                                 }
2034
2035                                 # If the title is valid but undisplayable, make a link to it
2036                                 if ( $this->mOutputType == OT_HTML && !$found ) {
2037                                         $text = $linestart . '[['.$title->getPrefixedText().']]';
2038                                         $found = true;
2039                                 }
2040
2041                                 # Template cache array insertion
2042                                 if( $found ) {
2043                                         $this->mTemplates[$part1] = $text;
2044                                 }
2045                         }
2046                 }
2047
2048                 # Recursive parsing, escaping and link table handling
2049                 # Only for HTML output
2050                 if ( $nowiki && $found && $this->mOutputType == OT_HTML ) {
2051                         $text = wfEscapeWikiText( $text );
2052                 } elseif ( ($this->mOutputType == OT_HTML || $this->mOutputType == OT_WIKI) && $found && !$noparse) {
2053                         # Clean up argument array
2054                         $assocArgs = array();
2055                         $index = 1;
2056                         foreach( $args as $arg ) {
2057                                 $eqpos = strpos( $arg, '=' );
2058                                 if ( $eqpos === false ) {
2059                                         $assocArgs[$index++] = $arg;
2060                                 } else {
2061                                         $name = trim( substr( $arg, 0, $eqpos ) );
2062                                         $value = trim( substr( $arg, $eqpos+1 ) );
2063                                         if ( $value === false ) {
2064                                                 $value = '';
2065                                         }
2066                                         if ( $name !== false ) {
2067                                                 $assocArgs[$name] = $value;
2068                                         }
2069                                 }
2070                         }
2071
2072                         # Add a new element to the templace recursion path
2073                         $this->mTemplatePath[$part1] = 1;
2074
2075                         $text = $this->strip( $text, $this->mStripState );
2076                         $text = Sanitizer::removeHTMLtags( $text );
2077                         $text = $this->replaceVariables( $text, $assocArgs );
2078
2079                         # Resume the link cache and register the inclusion as a link
2080                         if ( $this->mOutputType == OT_HTML && !is_null( $title ) ) {
2081                                 $wgLinkCache->addLinkObj( $title );
2082                         }
2083
2084                         # If the template begins with a table or block-level
2085                         # element, it should be treated as beginning a new line.
2086                         if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
2087                                 $text = "\n" . $text;
2088                         }
2089                 }
2090                 # Prune lower levels off the recursion check path
2091                 $this->mTemplatePath = $lastPathLevel;
2092
2093                 if ( !$found ) {
2094                         wfProfileOut( $fname );
2095                         return $matches[0];
2096                 } else {
2097                         # replace ==section headers==
2098                         # XXX this needs to go away once we have a better parser.
2099                         if ( $this->mOutputType != OT_WIKI && $itcamefromthedatabase ) {
2100                                 if( !is_null( $title ) )
2101                                         $encodedname = base64_encode($title->getPrefixedDBkey());
2102                                 else
2103                                         $encodedname = base64_encode("");
2104                                 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
2105                                         PREG_SPLIT_DELIM_CAPTURE);
2106                                 $text = '';
2107                                 $nsec = 0;
2108                                 for( $i = 0; $i < count($m); $i += 2 ) {
2109                                         $text .= $m[$i];
2110                                         if (!isset($m[$i + 1]) || $m[$i + 1] == "") continue;
2111                                         $hl = $m[$i + 1];
2112                                         if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
2113                                                 $text .= $hl;
2114                                                 continue;
2115                                         }
2116                                         preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
2117                                         $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
2118                                                 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
2119
2120                                         $nsec++;
2121                                 }
2122                         }
2123                 }
2124                 # Prune lower levels off the recursion check path
2125                 $this->mTemplatePath = $lastPathLevel;
2126
2127                 if ( !$found ) {
2128                         wfProfileOut( $fname );
2129                         return $matches[0];
2130                 } else {
2131                         wfProfileOut( $fname );
2132                         return $text;
2133                 }
2134         }
2135
2136         /**
2137          * Triple brace replacement -- used for template arguments
2138          * @access private
2139          */
2140         function argSubstitution( $matches ) {
2141                 $arg = trim( $matches[1] );
2142                 $text = $matches[0];
2143                 $inputArgs = end( $this->mArgStack );
2144
2145                 if ( array_key_exists( $arg, $inputArgs ) ) {
2146                         $text = $inputArgs[$arg];
2147                 }
2148
2149                 return $text;
2150         }
2151
2152         /**
2153          * Returns true if the function is allowed to include this entity
2154          * @access private
2155          */
2156         function incrementIncludeCount( $dbk ) {
2157                 if ( !array_key_exists( $dbk, $this->mIncludeCount ) ) {
2158                         $this->mIncludeCount[$dbk] = 0;
2159                 }
2160                 if ( ++$this->mIncludeCount[$dbk] <= MAX_INCLUDE_REPEAT ) {
2161                         return true;
2162                 } else {
2163                         return false;
2164                 }
2165         }
2166
2167         /**
2168          * This function accomplishes several tasks:
2169          * 1) Auto-number headings if that option is enabled
2170          * 2) Add an [edit] link to sections for logged in users who have enabled the option
2171          * 3) Add a Table of contents on the top for users who have enabled the option
2172          * 4) Auto-anchor headings
2173          *
2174          * It loops through all headlines, collects the necessary data, then splits up the
2175          * string and re-inserts the newly formatted headlines.
2176          *
2177          * @param string $text
2178          * @param boolean $isMain
2179          * @access private
2180          */
2181         function formatHeadings( $text, $isMain=true ) {
2182                 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders, $wgInterwikiLinkHolders;
2183
2184                 $doNumberHeadings = $this->mOptions->getNumberHeadings();
2185                 $doShowToc = true;
2186                 $forceTocHere = false;
2187                 if( !$this->mTitle->userCanEdit() ) {
2188                         $showEditLink = 0;
2189                 } else {
2190                         $showEditLink = $this->mOptions->getEditSection();
2191                 }
2192
2193                 # Inhibit editsection links if requested in the page
2194                 $esw =& MagicWord::get( MAG_NOEDITSECTION );
2195                 if( $esw->matchAndRemove( $text ) ) {
2196                         $showEditLink = 0;
2197                 }
2198                 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2199                 # do not add TOC
2200                 $mw =& MagicWord::get( MAG_NOTOC );
2201                 if( $mw->matchAndRemove( $text ) ) {
2202                         $doShowToc = false;
2203                 }
2204
2205                 # Get all headlines for numbering them and adding funky stuff like [edit]
2206                 # links - this is for later, but we need the number of headlines right now
2207                 $numMatches = preg_match_all( '/<H([1-6])(.*?'.'>)(.*?)<\/H[1-6] *>/i', $text, $matches );
2208
2209                 # if there are fewer than 4 headlines in the article, do not show TOC
2210                 if( $numMatches < 4 ) {
2211                         $doShowToc = false;
2212                 }
2213
2214                 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2215                 # override above conditions and always show TOC at that place
2216
2217                 $mw =& MagicWord::get( MAG_TOC );
2218                 if($mw->match( $text ) ) {
2219                         $doShowToc = true;
2220                         $forceTocHere = true;
2221                 } else {
2222                         # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2223                         # override above conditions and always show TOC above first header
2224                         $mw =& MagicWord::get( MAG_FORCETOC );
2225                         if ($mw->matchAndRemove( $text ) ) {
2226                                 $doShowToc = true;
2227                         }
2228                 }
2229
2230                 # Never ever show TOC if no headers
2231                 if( $numMatches < 1 ) {
2232                         $doShowToc = false;
2233                 }
2234
2235                 # We need this to perform operations on the HTML
2236                 $sk =& $this->mOptions->getSkin();
2237
2238                 # headline counter
2239                 $headlineCount = 0;
2240                 $sectionCount = 0; # headlineCount excluding template sections
2241
2242                 # Ugh .. the TOC should have neat indentation levels which can be
2243                 # passed to the skin functions. These are determined here
2244                 $toc = '';
2245                 $full = '';
2246                 $head = array();
2247                 $sublevelCount = array();
2248                 $levelCount = array();
2249                 $toclevel = 0;
2250                 $level = 0;
2251                 $prevlevel = 0;
2252                 $toclevel = 0;
2253                 $prevtoclevel = 0;
2254
2255                 foreach( $matches[3] as $headline ) {
2256                         $istemplate = 0;
2257                         $templatetitle = '';
2258                         $templatesection = 0;
2259                         $numbering = '';
2260
2261                         if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2262                                 $istemplate = 1;
2263                                 $templatetitle = base64_decode($mat[1]);
2264                                 $templatesection = 1 + (int)base64_decode($mat[2]);
2265                                 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2266                         }
2267
2268                         if( $toclevel ) {
2269                                 $prevlevel = $level;
2270                                 $prevtoclevel = $toclevel;
2271                         }
2272                         $level = $matches[1][$headlineCount];
2273
2274                         if( $doNumberHeadings || $doShowToc ) {
2275
2276                                 if ( $level > $prevlevel ) {
2277                                         # Increase TOC level
2278                                         $toclevel++;
2279                                         $sublevelCount[$toclevel] = 0;
2280                                         $toc .= $sk->tocIndent();
2281                                 }
2282                                 elseif ( $level < $prevlevel && $toclevel > 1 ) {
2283                                         # Decrease TOC level, find level to jump to
2284
2285                                         if ( $toclevel == 2 && $level <= $levelCount[1] ) {
2286                                                 # Can only go down to level 1
2287                                                 $toclevel = 1;
2288                                         } else {
2289                                                 for ($i = $toclevel; $i > 0; $i--) {
2290                                                         if ( $levelCount[$i] == $level ) {
2291                                                                 # Found last matching level
2292                                                                 $toclevel = $i;
2293                                                                 break;
2294                                                         }
2295                                                         elseif ( $levelCount[$i] < $level ) {
2296                                                                 # Found first matching level below current level
2297                                                                 $toclevel = $i + 1;
2298                                                                 break;
2299                                                         }
2300                                                 }
2301                                         }
2302
2303                                         $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
2304                                 }
2305                                 else {
2306                                         # No change in level, end TOC line
2307                                         $toc .= $sk->tocLineEnd();
2308                                 }
2309
2310                                 $levelCount[$toclevel] = $level;
2311
2312                                 # count number of headlines for each level
2313                                 @$sublevelCount[$toclevel]++;
2314                                 $dot = 0;
2315                                 for( $i = 1; $i <= $toclevel; $i++ ) {
2316                                         if( !empty( $sublevelCount[$i] ) ) {
2317                                                 if( $dot ) {
2318                                                         $numbering .= '.';
2319                                                 }
2320                                                 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2321                                                 $dot = 1;
2322                                         }
2323                                 }
2324                         }
2325
2326                         # The canonized header is a version of the header text safe to use for links
2327                         # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2328                         $canonized_headline = $this->unstrip( $headline, $this->mStripState );
2329                         $canonized_headline = $this->unstripNoWiki( $canonized_headline, $this->mStripState );
2330
2331                         # Remove link placeholders by the link text.
2332                         #     <!--LINK number-->
2333                         # turns into
2334                         #     link text with suffix
2335                         $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2336                                                             "\$wgLinkHolders['texts'][\$1]",
2337                                                             $canonized_headline );
2338                         $canonized_headline = preg_replace( '/<!--IWLINK ([0-9]*)-->/e',
2339                                                             "\$wgInterwikiLinkHolders[\$1]",
2340                                                             $canonized_headline );
2341
2342                         # strip out HTML
2343                         $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2344                         $tocline = trim( $canonized_headline );
2345                         $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT, $wgInputEncoding ) );
2346                         $replacearray = array(
2347                                 '%3A' => ':',
2348                                 '%' => '.'
2349                         );
2350                         $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2351                         $refer[$headlineCount] = $canonized_headline;
2352
2353                         # count how many in assoc. array so we can track dupes in anchors
2354                         @$refers[$canonized_headline]++;
2355                         $refcount[$headlineCount]=$refers[$canonized_headline];
2356
2357                         # Don't number the heading if it is the only one (looks silly)
2358                         if( $doNumberHeadings && count( $matches[3] ) > 1) {
2359                                 # the two are different if the line contains a link
2360                                 $headline=$numbering . ' ' . $headline;
2361                         }
2362
2363                         # Create the anchor for linking from the TOC to the section
2364                         $anchor = $canonized_headline;
2365                         if($refcount[$headlineCount] > 1 ) {
2366                                 $anchor .= '_' . $refcount[$headlineCount];
2367                         }
2368                         if( $doShowToc && ( !isset($wgMaxTocLevel) || $toclevel<$wgMaxTocLevel ) ) {
2369                                 $toc .= $sk->tocLine($anchor, $tocline, $numbering, $toclevel);
2370                         }
2371                         if( $showEditLink && ( !$istemplate || $templatetitle !== "" ) ) {
2372                                 if ( empty( $head[$headlineCount] ) ) {
2373                                         $head[$headlineCount] = '';
2374                                 }
2375                                 if( $istemplate )
2376                                         $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2377                                 else
2378                                         $head[$headlineCount] .= $sk->editSectionLink($this->mTitle, $sectionCount+1);
2379                         }
2380
2381                         # give headline the correct <h#> tag
2382                         @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2383
2384                         $headlineCount++;
2385                         if( !$istemplate )
2386                                 $sectionCount++;
2387                 }
2388
2389                 if( $doShowToc ) {
2390                         $toclines = $headlineCount;
2391                         $toc .= $sk->tocUnindent( $toclevel - 1 );
2392                         $toc = $sk->tocList( $toc );
2393                 }
2394
2395                 # split up and insert constructed headlines
2396
2397                 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2398                 $i = 0;
2399
2400                 foreach( $blocks as $block ) {
2401                         if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2402                                 # This is the [edit] link that appears for the top block of text when
2403                                 # section editing is enabled
2404
2405                                 # Disabled because it broke block formatting
2406                                 # For example, a bullet point in the top line
2407                                 # $full .= $sk->editSectionLink(0);
2408                         }
2409                         $full .= $block;
2410                         if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2411                         # Top anchor now in skin
2412                                 $full = $full.$toc;
2413                         }
2414
2415                         if( !empty( $head[$i] ) ) {
2416                                 $full .= $head[$i];
2417                         }
2418                         $i++;
2419                 }
2420                 if($forceTocHere) {
2421                         $mw =& MagicWord::get( MAG_TOC );
2422                         return $mw->replace( $toc, $full );
2423                 } else {
2424                         return $full;
2425                 }
2426         }
2427
2428         /**
2429          * Return an HTML link for the "ISBN 123456" text
2430          * @access private
2431          */
2432         function magicISBN( $text ) {
2433                 global $wgLang;
2434                 $fname = 'Parser::magicISBN';
2435                 wfProfileIn( $fname );
2436
2437                 $a = split( 'ISBN ', ' '.$text );
2438                 if ( count ( $a ) < 2 ) {
2439                         wfProfileOut( $fname );
2440                         return $text;
2441                 }
2442                 $text = substr( array_shift( $a ), 1);
2443                 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2444
2445                 foreach ( $a as $x ) {
2446                         $isbn = $blank = '' ;
2447                         while ( ' ' == $x{0} ) {
2448                                 $blank .= ' ';
2449                                 $x = substr( $x, 1 );
2450                         }
2451                         if ( $x == '' ) { # blank isbn
2452                                 $text .= "ISBN $blank";
2453                                 continue;
2454                         }
2455                         while ( strstr( $valid, $x{0} ) != false ) {
2456                                 $isbn .= $x{0};
2457                                 $x = substr( $x, 1 );
2458                         }
2459                         $num = str_replace( '-', '', $isbn );
2460                         $num = str_replace( ' ', '', $num );
2461
2462                         if ( '' == $num ) {
2463                                 $text .= "ISBN $blank$x";
2464                         } else {
2465                                 $titleObj = Title::makeTitle( NS_SPECIAL, 'Booksources' );
2466                                 $text .= '<a href="' .
2467                                 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2468                                         "\" class=\"internal\">ISBN $isbn</a>";
2469                                 $text .= $x;
2470                         }
2471                 }
2472                 wfProfileOut( $fname );
2473                 return $text;
2474         }
2475
2476         /**
2477          * Return an HTML link for the "RFC 1234" text
2478          *
2479          * @access private
2480          * @param string $text     Text to be processed
2481          * @param string $keyword  Magic keyword to use (default RFC)
2482          * @param string $urlmsg   Interface message to use (default rfcurl)
2483          * @return string
2484          */
2485         function magicRFC( $text, $keyword='RFC ', $urlmsg='rfcurl'  ) {
2486                 global $wgLang;
2487
2488                 $valid = '0123456789';
2489                 $internal = false;
2490
2491                 $a = split( $keyword, ' '.$text );
2492                 if ( count ( $a ) < 2 ) {
2493                         return $text;
2494                 }
2495                 $text = substr( array_shift( $a ), 1);
2496
2497                 /* Check if keyword is preceed by [[.
2498                  * This test is made here cause of the array_shift above
2499                  * that prevent the test to be done in the foreach.
2500                  */
2501                 if ( substr( $text, -2 ) == '[[' ) {
2502                         $internal = true;
2503                 }
2504
2505                 foreach ( $a as $x ) {
2506                         /* token might be empty if we have RFC RFC 1234 */
2507                         if ( $x=='' ) {
2508                                 $text.=$keyword;
2509                                 continue;
2510                                 }
2511
2512                         $id = $blank = '' ;
2513
2514                         /** remove and save whitespaces in $blank */
2515                         while ( $x{0} == ' ' ) {
2516                                 $blank .= ' ';
2517                                 $x = substr( $x, 1 );
2518                         }
2519
2520                         /** remove and save the rfc number in $id */
2521                         while ( strstr( $valid, $x{0} ) != false ) {
2522                                 $id .= $x{0};
2523                                 $x = substr( $x, 1 );
2524                         }
2525
2526                         if ( $id == '' ) {
2527                                 /* call back stripped spaces*/
2528                                 $text .= $keyword.$blank.$x;
2529                         } elseif( $internal ) {
2530                                 /* normal link */
2531                                 $text .= $keyword.$id.$x;
2532                         } else {
2533                                 /* build the external link*/
2534                                 $url = wfMsg( $urlmsg, $id);
2535                                 $sk =& $this->mOptions->getSkin();
2536                                 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
2537                                 $text .= "<a href='{$url}'{$la}>{$keyword}{$id}</a>{$x}";
2538                         }
2539
2540                         /* Check if the next RFC keyword is preceed by [[ */
2541                         $internal = ( substr($x,-2) == '[[' );
2542                 }
2543                 return $text;
2544         }
2545
2546         /**
2547          * Transform wiki markup when saving a page by doing \r\n -> \n
2548          * conversion, substitting signatures, {{subst:}} templates, etc.
2549          *
2550          * @param string $text the text to transform
2551          * @param Title &$title the Title object for the current article
2552          * @param User &$user the User object describing the current user
2553          * @param ParserOptions $options parsing options
2554          * @param bool $clearState whether to clear the parser state first
2555          * @return string the altered wiki markup
2556          * @access public
2557          */
2558         function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2559                 $this->mOptions = $options;
2560                 $this->mTitle =& $title;
2561                 $this->mOutputType = OT_WIKI;
2562
2563                 if ( $clearState ) {
2564                         $this->clearState();
2565                 }
2566
2567                 $stripState = false;
2568                 $pairs = array(
2569                         "\r\n" => "\n",
2570                         );
2571                 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
2572                 $text = $this->strip( $text, $stripState, false );
2573                 $text = $this->pstPass2( $text, $user );
2574                 $text = $this->unstrip( $text, $stripState );
2575                 $text = $this->unstripNoWiki( $text, $stripState );
2576                 return $text;
2577         }
2578
2579         /**
2580          * Pre-save transform helper function
2581          * @access private
2582          */
2583         function pstPass2( $text, &$user ) {
2584                 global $wgLang, $wgContLang, $wgLocaltimezone;
2585
2586                 # Variable replacement
2587                 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2588                 $text = $this->replaceVariables( $text );
2589
2590                 # Signatures
2591                 #
2592                 $n = $user->getName();
2593                 $k = $user->getOption( 'nickname' );
2594                 if ( '' == $k ) { $k = $n; }
2595                 if ( isset( $wgLocaltimezone ) ) {
2596                         $oldtz = getenv( 'TZ' );
2597                         putenv( 'TZ='.$wgLocaltimezone );
2598                 }
2599                 /* Note: this is an ugly timezone hack for the European wikis */
2600                 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
2601                   ' (' . date( 'T' ) . ')';
2602                 if ( isset( $wgLocaltimezone ) ) {
2603                         putenv( 'TZ='.$oldtzs );
2604                 }
2605
2606                 if( $user->getOption( 'fancysig' ) ) {
2607                         $sigText = $k;
2608                 } else {
2609                         $sigText = '[[' . $wgContLang->getNsText( NS_USER ) . ":$n|$k]]";
2610                 }
2611                 $text = preg_replace( '/~~~~~/', $d, $text );
2612                 $text = preg_replace( '/~~~~/', "$sigText $d", $text );
2613                 $text = preg_replace( '/~~~/', $sigText, $text );
2614
2615                 # Context links: [[|name]] and [[name (context)|]]
2616                 #
2617                 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2618                 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2619                 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2620                 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2621
2622                 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/";             # [[page (context)|]]
2623                 $p2 = "/\[\[\\|({$tc}+)]]/";                                    # [[|page]]
2624                 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/";                # [[namespace:page|]] and [[:namespace:page|]]
2625                 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2626                 $context = '';
2627                 $t = $this->mTitle->getText();
2628                 if ( preg_match( $conpat, $t, $m ) ) {
2629                         $context = $m[2];
2630                 }
2631                 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2632                 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2633                 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2634
2635                 if ( '' == $context ) {
2636                         $text = preg_replace( $p2, '[[\\1]]', $text );
2637                 } else {
2638                         $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2639                 }
2640
2641                 # Trim trailing whitespace
2642                 # MAG_END (__END__) tag allows for trailing
2643                 # whitespace to be deliberately included
2644                 $text = rtrim( $text );
2645                 $mw =& MagicWord::get( MAG_END );
2646                 $mw->matchAndRemove( $text );
2647
2648                 return $text;
2649         }
2650
2651         /**
2652          * Set up some variables which are usually set up in parse()
2653          * so that an external function can call some class members with confidence
2654          * @access public
2655          */
2656         function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2657                 $this->mTitle =& $title;
2658                 $this->mOptions = $options;
2659                 $this->mOutputType = $outputType;
2660                 if ( $clearState ) {
2661                         $this->clearState();
2662                 }
2663         }
2664
2665         /**
2666          * Transform a MediaWiki message by replacing magic variables.
2667          *
2668          * @param string $text the text to transform
2669          * @param ParserOptions $options  options
2670          * @return string the text with variables substituted
2671          * @access public
2672          */
2673         function transformMsg( $text, $options ) {
2674                 global $wgTitle;
2675                 static $executing = false;
2676
2677                 # Guard against infinite recursion
2678                 if ( $executing ) {
2679                         return $text;
2680                 }
2681                 $executing = true;
2682
2683                 $this->mTitle = $wgTitle;
2684                 $this->mOptions = $options;
2685                 $this->mOutputType = OT_MSG;
2686                 $this->clearState();
2687                 $text = $this->replaceVariables( $text );
2688
2689                 $executing = false;
2690                 return $text;
2691         }
2692
2693         /**
2694          * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2695          * Callback will be called with the text within
2696          * Transform and return the text within
2697          * @access public
2698          */
2699         function setHook( $tag, $callback ) {
2700                 $oldVal = @$this->mTagHooks[$tag];
2701                 $this->mTagHooks[$tag] = $callback;
2702                 return $oldVal;
2703         }
2704
2705         /**
2706          * Replace <!--LINK--> link placeholders with actual links, in the buffer
2707          * Placeholders created in Skin::makeLinkObj()
2708          * Returns an array of links found, indexed by PDBK:
2709          *  0 - broken
2710          *  1 - normal link
2711          *  2 - stub
2712          * $options is a bit field, RLH_FOR_UPDATE to select for update
2713          */
2714         function replaceLinkHolders( &$text, $options = 0 ) {
2715                 global $wgUser, $wgLinkCache, $wgUseOldExistenceCheck, $wgLinkHolders;
2716                 global $wgInterwikiLinkHolders;
2717                 global $outputReplace;
2718
2719                 if ( $wgUseOldExistenceCheck ) {
2720                         return array();
2721                 }
2722
2723                 $fname = 'Parser::replaceLinkHolders';
2724                 wfProfileIn( $fname );
2725
2726                 $pdbks = array();
2727                 $colours = array();
2728
2729                 #if ( !empty( $tmpLinks[0] ) ) { #TODO
2730                 if ( !empty( $wgLinkHolders['namespaces'] ) ) {
2731                         wfProfileIn( $fname.'-check' );
2732                         $dbr =& wfGetDB( DB_SLAVE );
2733                         $page = $dbr->tableName( 'page' );
2734                         $sk = $wgUser->getSkin();
2735                         $threshold = $wgUser->getOption('stubthreshold');
2736
2737                         # Sort by namespace
2738                         asort( $wgLinkHolders['namespaces'] );
2739
2740                         # Generate query
2741                         $query = false;
2742                         foreach ( $wgLinkHolders['namespaces'] as $key => $val ) {
2743                                 # Make title object
2744                                 $title = $wgLinkHolders['titles'][$key];
2745
2746                                 # Skip invalid entries.
2747                                 # Result will be ugly, but prevents crash.
2748                                 if ( is_null( $title ) ) {
2749                                         continue;
2750                                 }
2751                                 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
2752
2753                                 # Check if it's in the link cache already
2754                                 if ( $wgLinkCache->getGoodLinkID( $pdbk ) ) {
2755                                         $colours[$pdbk] = 1;
2756                                 } elseif ( $wgLinkCache->isBadLink( $pdbk ) ) {
2757                                         $colours[$pdbk] = 0;
2758                                 } else {
2759                                         # Not in the link cache, add it to the query
2760                                         if ( !isset( $current ) ) {
2761                                                 $current = $val;
2762                                                 $query =  "SELECT page_id, page_namespace, page_title";
2763                                                 if ( $threshold > 0 ) {
2764                                                         $query .= ', page_len, page_is_redirect';
2765                                                 }
2766                                                 $query .= " FROM $page WHERE (page_namespace=$val AND page_title IN(";
2767                                         } elseif ( $current != $val ) {
2768                                                 $current = $val;
2769                                                 $query .= ")) OR (page_namespace=$val AND page_title IN(";
2770                                         } else {
2771                                                 $query .= ', ';
2772                                         }
2773
2774                                         $query .= $dbr->addQuotes( $wgLinkHolders['dbkeys'][$key] );
2775                                 }
2776                         }
2777                         if ( $query ) {
2778                                 $query .= '))';
2779                                 if ( $options & RLH_FOR_UPDATE ) {
2780                                         $query .= ' FOR UPDATE';
2781                                 }
2782
2783                                 $res = $dbr->query( $query, $fname );
2784
2785                                 # Fetch data and form into an associative array
2786                                 # non-existent = broken
2787                                 # 1 = known
2788                                 # 2 = stub
2789                                 while ( $s = $dbr->fetchObject($res) ) {
2790                                         $title = Title::makeTitle( $s->page_namespace, $s->page_title );
2791                                         $pdbk = $title->getPrefixedDBkey();
2792                                         $wgLinkCache->addGoodLink( $s->page_id, $pdbk );
2793
2794                                         if ( $threshold >  0 ) {
2795                                                 $size = $s->page_len;
2796                                                 if ( $s->page_is_redirect || $s->page_namespace != 0 || $size >= $threshold ) {
2797                                                         $colours[$pdbk] = 1;
2798                                                 } else {
2799                                                         $colours[$pdbk] = 2;
2800                                                 }
2801                                         } else {
2802                                                 $colours[$pdbk] = 1;
2803                                         }
2804                                 }
2805                         }
2806                         wfProfileOut( $fname.'-check' );
2807
2808                         # Construct search and replace arrays
2809                         wfProfileIn( $fname.'-construct' );
2810                         $outputReplace = array();
2811                         foreach ( $wgLinkHolders['namespaces'] as $key => $ns ) {
2812                                 $pdbk = $pdbks[$key];
2813                                 $searchkey = '<!--LINK '.$key.'-->';
2814                                 $title = $wgLinkHolders['titles'][$key];
2815                                 if ( empty( $colours[$pdbk] ) ) {
2816                                         $wgLinkCache->addBadLink( $pdbk );
2817                                         $colours[$pdbk] = 0;
2818                                         $outputReplace[$searchkey] = $sk->makeBrokenLinkObj( $title,
2819                                                                         $wgLinkHolders['texts'][$key],
2820                                                                         $wgLinkHolders['queries'][$key] );
2821                                 } elseif ( $colours[$pdbk] == 1 ) {
2822                                         $outputReplace[$searchkey] = $sk->makeKnownLinkObj( $title,
2823                                                                         $wgLinkHolders['texts'][$key],
2824                                                                         $wgLinkHolders['queries'][$key] );
2825                                 } elseif ( $colours[$pdbk] == 2 ) {
2826                                         $outputReplace[$searchkey] = $sk->makeStubLinkObj( $title,
2827                                                                         $wgLinkHolders['texts'][$key],
2828                                                                         $wgLinkHolders['queries'][$key] );
2829                                 }
2830                         }
2831                         wfProfileOut( $fname.'-construct' );
2832
2833                         # Do the thing
2834                         wfProfileIn( $fname.'-replace' );
2835
2836                         $text = preg_replace_callback(
2837                                 '/(<!--LINK .*?-->)/',
2838                                 "outputReplaceMatches",
2839                                 $text);
2840                         wfProfileOut( $fname.'-replace' );
2841                 }
2842
2843                 if ( !empty( $wgInterwikiLinkHolders ) ) {
2844                         wfProfileIn( $fname.'-interwiki' );
2845                         $outputReplace = $wgInterwikiLinkHolders;
2846                         $text = preg_replace_callback(
2847                                 '/<!--IWLINK (.*?)-->/',
2848                                 "outputReplaceMatches",
2849                                 $text );
2850                         wfProfileOut( $fname.'-interwiki' );
2851                 }
2852
2853                 wfProfileOut( $fname );
2854                 return $colours;
2855         }
2856
2857         /**
2858          * Renders an image gallery from a text with one line per image.
2859          * text labels may be given by using |-style alternative text. E.g.
2860          *   Image:one.jpg|The number "1"
2861          *   Image:tree.jpg|A tree
2862          * given as text will return the HTML of a gallery with two images,
2863          * labeled 'The number "1"' and
2864          * 'A tree'.
2865          */
2866         function renderImageGallery( $text ) {
2867                 # Setup the parser
2868                 global $wgUser, $wgParser, $wgTitle;
2869                 $parserOptions = ParserOptions::newFromUser( $wgUser );
2870
2871                 global $wgLinkCache;
2872                 $ig = new ImageGallery();
2873                 $ig->setShowBytes( false );
2874                 $ig->setShowFilename( false );
2875                 $lines = explode( "\n", $text );
2876
2877                 foreach ( $lines as $line ) {
2878                         # match lines like these:
2879                         # Image:someimage.jpg|This is some image
2880                         preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
2881                         # Skip empty lines
2882                         if ( count( $matches ) == 0 ) {
2883                                 continue;
2884                         }
2885                         $nt = Title::newFromURL( $matches[1] );
2886                         if( is_null( $nt ) ) {
2887                                 # Bogus title. Ignore these so we don't bomb out later.
2888                                 continue;
2889                         }
2890                         if ( isset( $matches[3] ) ) {
2891                                 $label = $matches[3];
2892                         } else {
2893                                 $label = '';
2894                         }
2895
2896                         $html = $wgParser->parse( $label , $wgTitle, $parserOptions );
2897                         $html = $html->mText;
2898
2899                         $ig->add( Image::newFromTitle( $nt ), $html );
2900                         $wgLinkCache->addImageLinkObj( $nt );
2901                 }
2902                 return $ig->toHTML();
2903         }
2904 }
2905
2906 /**
2907  * @todo document
2908  * @package MediaWiki
2909  */
2910 class ParserOutput
2911 {
2912         var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2913         var $mCacheTime; # Used in ParserCache
2914         var $mVersion;   # Compatibility check
2915
2916         function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
2917                 $containsOldMagic = false )
2918         {
2919                 $this->mText = $text;
2920                 $this->mLanguageLinks = $languageLinks;
2921                 $this->mCategoryLinks = $categoryLinks;
2922                 $this->mContainsOldMagic = $containsOldMagic;
2923                 $this->mCacheTime = '';
2924                 $this->mVersion = MW_PARSER_VERSION;
2925         }
2926
2927         function getText()                   { return $this->mText; }
2928         function getLanguageLinks()          { return $this->mLanguageLinks; }
2929         function getCategoryLinks()          { return array_keys( $this->mCategoryLinks ); }
2930         function getCacheTime()              { return $this->mCacheTime; }
2931         function containsOldMagic()          { return $this->mContainsOldMagic; }
2932         function setText( $text )            { return wfSetVar( $this->mText, $text ); }
2933         function setLanguageLinks( $ll )     { return wfSetVar( $this->mLanguageLinks, $ll ); }
2934         function setCategoryLinks( $cl )     { return wfSetVar( $this->mCategoryLinks, $cl ); }
2935         function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic, $com ); }
2936         function setCacheTime( $t )          { return wfSetVar( $this->mCacheTime, $t ); }
2937         function addCategoryLink( $c )       { $this->mCategoryLinks[$c] = 1; }
2938
2939         function merge( $other ) {
2940                 $this->mLanguageLinks = array_merge( $this->mLanguageLinks, $other->mLanguageLinks );
2941                 $this->mCategoryLinks = array_merge( $this->mCategoryLinks, $this->mLanguageLinks );
2942                 $this->mContainsOldMagic = $this->mContainsOldMagic || $other->mContainsOldMagic;
2943         }
2944
2945         /**
2946          * Return true if this cached output object predates the global or
2947          * per-article cache invalidation timestamps, or if it comes from
2948          * an incompatible older version.
2949          *
2950          * @param string $touched the affected article's last touched timestamp
2951          * @return bool
2952          * @access public
2953          */
2954         function expired( $touched ) {
2955                 global $wgCacheEpoch;
2956                 return $this->getCacheTime() <= $touched ||
2957                        $this->getCacheTime() <= $wgCacheEpoch ||
2958                        !isset( $this->mVersion ) ||
2959                        version_compare( $this->mVersion, MW_PARSER_VERSION, "lt" );
2960         }
2961 }
2962
2963 /**
2964  * Set options of the Parser
2965  * @todo document
2966  * @package MediaWiki
2967  */
2968 class ParserOptions
2969 {
2970         # All variables are private
2971         var $mUseTeX;                    # Use texvc to expand <math> tags
2972         var $mUseDynamicDates;           # Use $wgDateFormatter to format dates
2973         var $mInterwikiMagic;            # Interlanguage links are removed and returned in an array
2974         var $mAllowExternalImages;       # Allow external images inline
2975         var $mSkin;                      # Reference to the preferred skin
2976         var $mDateFormat;                # Date format index
2977         var $mEditSection;               # Create "edit section" links
2978         var $mNumberHeadings;            # Automatically number headings
2979
2980         function getUseTeX()                        { return $this->mUseTeX; }
2981         function getUseDynamicDates()               { return $this->mUseDynamicDates; }
2982         function getInterwikiMagic()                { return $this->mInterwikiMagic; }
2983         function getAllowExternalImages()           { return $this->mAllowExternalImages; }
2984         function getSkin()                          { return $this->mSkin; }
2985         function getDateFormat()                    { return $this->mDateFormat; }
2986         function getEditSection()                   { return $this->mEditSection; }
2987         function getNumberHeadings()                { return $this->mNumberHeadings; }
2988
2989         function setUseTeX( $x )                    { return wfSetVar( $this->mUseTeX, $x ); }
2990         function setUseDynamicDates( $x )           { return wfSetVar( $this->mUseDynamicDates, $x ); }
2991         function setInterwikiMagic( $x )            { return wfSetVar( $this->mInterwikiMagic, $x ); }
2992         function setAllowExternalImages( $x )       { return wfSetVar( $this->mAllowExternalImages, $x ); }
2993         function setDateFormat( $x )                { return wfSetVar( $this->mDateFormat, $x ); }
2994         function setEditSection( $x )               { return wfSetVar( $this->mEditSection, $x ); }
2995         function setNumberHeadings( $x )            { return wfSetVar( $this->mNumberHeadings, $x ); }
2996
2997         function setSkin( &$x ) { $this->mSkin =& $x; }
2998
2999         /**
3000          * Get parser options
3001          * @static
3002          */
3003         function newFromUser( &$user ) {
3004                 $popts = new ParserOptions;
3005                 $popts->initialiseFromUser( $user );
3006                 return $popts;
3007         }
3008
3009         /** Get user options */
3010         function initialiseFromUser( &$userInput ) {
3011                 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
3012                 $fname = 'ParserOptions::initialiseFromUser';
3013                 wfProfileIn( $fname );
3014                 if ( !$userInput ) {
3015                         $user = new User;
3016                         $user->setLoaded( true );
3017                 } else {
3018                         $user =& $userInput;
3019                 }
3020
3021                 $this->mUseTeX = $wgUseTeX;
3022                 $this->mUseDynamicDates = $wgUseDynamicDates;
3023                 $this->mInterwikiMagic = $wgInterwikiMagic;
3024                 $this->mAllowExternalImages = $wgAllowExternalImages;
3025                 wfProfileIn( $fname.'-skin' );
3026                 $this->mSkin =& $user->getSkin();
3027                 wfProfileOut( $fname.'-skin' );
3028                 $this->mDateFormat = $user->getOption( 'date' );
3029                 $this->mEditSection = $user->getOption( 'editsection' );
3030                 $this->mNumberHeadings = $user->getOption( 'numberheadings' );
3031                 wfProfileOut( $fname );
3032         }
3033 }
3034
3035 /**
3036  * Callback function used by Parser::replaceLinkHolders()
3037  * to substitute link placeholders.
3038  */
3039 function &outputReplaceMatches( $matches ) {
3040         global $outputReplace;
3041         return $outputReplace[$matches[1]];
3042 }
3043
3044 /**
3045  * Return the total number of articles
3046  */
3047 function wfNumberOfArticles() {
3048         global $wgNumberOfArticles;
3049
3050         wfLoadSiteStats();
3051         return $wgNumberOfArticles;
3052 }
3053
3054 /**
3055  * Get various statistics from the database
3056  * @private
3057  */
3058 function wfLoadSiteStats() {
3059         global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
3060         $fname = 'wfLoadSiteStats';
3061
3062         if ( -1 != $wgNumberOfArticles ) return;
3063         $dbr =& wfGetDB( DB_SLAVE );
3064         $s = $dbr->selectRow( 'site_stats',
3065                 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
3066                 array( 'ss_row_id' => 1 ), $fname
3067         );
3068
3069         if ( $s === false ) {
3070                 return;
3071         } else {
3072                 $wgTotalViews = $s->ss_total_views;
3073                 $wgTotalEdits = $s->ss_total_edits;
3074                 $wgNumberOfArticles = $s->ss_good_articles;
3075         }
3076 }
3077
3078 /**
3079  * Escape html tags
3080  * Basicly replacing " > and < with HTML entities ( &quot;, &gt;, &lt;)
3081  *
3082  * @param string $in Text that might contain HTML tags
3083  * @return string Escaped string
3084  */
3085 function wfEscapeHTMLTagsOnly( $in ) {
3086         return str_replace(
3087                 array( '"', '>', '<' ),
3088                 array( '&quot;', '&gt;', '&lt;' ),
3089                 $in );
3090 }
3091
3092 ?>