includes/Parser.php

   1 <?php
   2
   3 // require_once('Tokenizer.php');
   4
   5 /**
   6  * PHP Parser
   7  *
   8  * Processes wiki markup
   9  *
  10  * There are two main entry points into the Parser class:
  11  * parse()
  12  *   produces HTML output
  13  * preSaveTransform().
  14  *   produces altered wiki markup.
  15  *
  16  * Globals used:
  17  *    objects:   $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
  18  *
  19  * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
  20  *
  21  * settings:
  22  *  $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
  23  *  $wgNamespacesWithSubpages, $wgLanguageCode, $wgAllowExternalImages*,
  24  *  $wgLocaltimezone
  25  *
  26  *  * only within ParserOptions
  27  *
  28  * @package MediaWiki
  29  */
  30
  31 /**
  32  * Variable substitution O(N^2) attack
  33  *
  34 * Without countermeasures, it would be possible to attack the parser by saving
  35 * a page filled with a large number of inclusions of large pages. The size of
  36 * the generated page would be proportional to the square of the input size.
  37 * Hence, we limit the number of inclusions of any given page, thus bringing any
  38 * attack back to O(N).
  39 */
  40 define( 'MAX_INCLUDE_REPEAT', 100 );
  41 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
  42
  43 # Allowed values for $mOutputType
  44 define( 'OT_HTML', 1 );
  45 define( 'OT_WIKI', 2 );
  46 define( 'OT_MSG' , 3 );
  47
  48 # string parameter for extractTags which will cause it
  49 # to strip HTML comments in addition to regular
  50 # <XML>-style tags. This should not be anything we
  51 # may want to use in wikisyntax
  52 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
  53
  54 # prefix for escaping, used in two functions at least
  55 define( 'UNIQ_PREFIX', 'NaodW29');
  56
  57 # Constants needed for external link processing
  58 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
  59 define( 'HTTP_PROTOCOLS', 'http|https' );
  60 # Everything except bracket, space, or control characters
  61 define( 'EXT_LINK_URL_CLASS', '[^]\\x00-\\x20\\x7F]' );
  62 define( 'INVERSE_EXT_LINK_URL_CLASS', '[\]\\x00-\\x20\\x7F]' );
  63 # Including space
  64 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
  65 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
  66 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
  67 define( 'EXT_LINK_BRACKETED',  '/\[(('.URL_PROTOCOLS.'):'.EXT_LINK_URL_CLASS.'+) *('.EXT_LINK_TEXT_CLASS.'*?)\]/S' );
  68 define( 'EXT_IMAGE_REGEX',
  69         '/^('.HTTP_PROTOCOLS.':)'.  # Protocol
  70         '('.EXT_LINK_URL_CLASS.'+)\\/'.  # Hostname and path
  71         '('.EXT_IMAGE_FNAME_CLASS.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS.')$/S' # Filename
  72 );
  73
  74 /**
  75  * @todo document
  76  * @package MediaWiki
  77  */
  78 class Parser
  79 {
  80         # Persistent:
  81         var $mTagHooks;
  82
  83         # Cleared with clearState():
  84         var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
  85         var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
  86
  87         # Temporary:
  88         var $mOptions, $mTitle, $mOutputType,
  89             $mTemplates,        // cache of already loaded templates, avoids
  90                                 // multiple SQL queries for the same string
  91             $mTemplatePath;     // stores an unsorted hash of all the templates already loaded
  92                                 // in this path. Used for loop detection.
  93
  94         function Parser() {
  95                 $this->mTemplates = array();
  96                 $this->mTemplatePath = array();
  97                 $this->mTagHooks = array();
  98                 $this->clearState();
  99         }
 100
 101         function clearState() {
 102                 $this->mOutput = new ParserOutput;
 103                 $this->mAutonumber = 0;
 104                 $this->mLastSection = "";
 105                 $this->mDTopen = false;
 106                 $this->mVariables = false;
 107                 $this->mIncludeCount = array();
 108                 $this->mStripState = array();
 109                 $this->mArgStack = array();
 110                 $this->mInPre = false;
 111         }
 112
 113         # First pass--just handle <nowiki> sections, pass the rest off
 114         # to internalParse() which does all the real work.
 115         #
 116         # Returns a ParserOutput
 117         #
 118         function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
 119                 global $wgUseTidy;
 120                 $fname = 'Parser::parse';
 121                 wfProfileIn( $fname );
 122
 123                 if ( $clearState ) {
 124                         $this->clearState();
 125                 }
 126
 127                 $this->mOptions = $options;
 128                 $this->mTitle =& $title;
 129                 $this->mOutputType = OT_HTML;
 130
 131                 $stripState = NULL;
 132                 $text = $this->strip( $text, $this->mStripState );
 133                 $text = $this->internalParse( $text, $linestart );
 134                 $text = $this->unstrip( $text, $this->mStripState );
 135                 # Clean up special characters, only run once, next-to-last before doBlockLevels
 136                 if(!$wgUseTidy) {
 137                         $fixtags = array(
 138                                 # french spaces, last one Guillemet-left
 139                                 # only if there is something before the space
 140                                 '/(.) (?=\\?|:|;|!|\\302\\273)/i' => '\\1&nbsp;\\2',
 141                                 # french spaces, Guillemet-right
 142                                 "/(\\302\\253) /i"=>"\\1&nbsp;",
 143                                 '/<hr *>/i' => '<hr />',
 144                                 '/<br *>/i' => '<br />',
 145                                 '/<center *>/i' => '<div class="center">',
 146                                 '/<\\/center *>/i' => '</div>',
 147                                 # Clean up spare ampersands; note that we probably ought to be
 148                                 # more careful about named entities.
 149                                 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&amp;'
 150                         );
 151                         $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
 152                 } else {
 153                         $fixtags = array(
 154                                 # french spaces, last one Guillemet-left
 155                                 '/ (\\?|:|;|!|\\302\\273)/i' => '&nbsp;\\1',
 156                                 # french spaces, Guillemet-right
 157                                 '/(\\302\\253) /i' => '\\1&nbsp;',
 158                                 '/<center *>/i' => '<div class="center">',
 159                                 '/<\\/center *>/i' => '</div>'
 160                         );
 161                         $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
 162                 }
 163                 # only once and last
 164                 $text = $this->doBlockLevels( $text, $linestart );
 165                 $text = $this->unstripNoWiki( $text, $this->mStripState );
 166                 if($wgUseTidy) {
 167                         $text = $this->tidy($text);
 168                 }
 169                 $this->mOutput->setText( $text );
 170                 wfProfileOut( $fname );
 171                 return $this->mOutput;
 172         }
 173
 174         /* static */ function getRandomString() {
 175                 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
 176         }
 177
 178         # Replaces all occurrences of <$tag>content</$tag> in the text
 179         # with a random marker and returns the new text. the output parameter
 180         # $content will be an associative array filled with data on the form
 181         # $unique_marker => content.
 182
 183         # If $content is already set, the additional entries will be appended
 184
 185         # If $tag is set to STRIP_COMMENTS, the function will extract
 186         # <!-- HTML comments -->
 187
 188         /* static */ function extractTags($tag, $text, &$content, $uniq_prefix = ''){
 189                 $rnd = $uniq_prefix . '-' . $tag . Parser::getRandomString();
 190                 if ( !$content ) {
 191                         $content = array( );
 192                 }
 193                 $n = 1;
 194                 $stripped = '';
 195
 196                 while ( '' != $text ) {
 197                         if($tag==STRIP_COMMENTS) {
 198                                 $p = preg_split( '/<!--/i', $text, 2 );
 199                         } else {
 200                                 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
 201                         }
 202                         $stripped .= $p[0];
 203                         if ( ( count( $p ) < 2 ) || ( '' == $p[1] ) ) {
 204                                 $text = '';
 205                         } else {
 206                                 if($tag==STRIP_COMMENTS) {
 207                                         $q = preg_split( '/-->/i', $p[1], 2 );
 208                                 } else {
 209                                         $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
 210                                 }
 211                                 $marker = $rnd . sprintf('%08X', $n++);
 212                                 $content[$marker] = $q[0];
 213                                 $stripped .= $marker;
 214                                 $text = $q[1];
 215                         }
 216                 }
 217                 return $stripped;
 218         }
 219
 220         # Strips and renders <nowiki>, <pre>, <math>, <hiero>
 221         # If $render is set, performs necessary rendering operations on plugins
 222         # Returns the text, and fills an array with data needed in unstrip()
 223         # If the $state is already a valid strip state, it adds to the state
 224
 225         # When $stripcomments is set, HTML comments <!-- like this -->
 226         # will be stripped in addition to other tags. This is important
 227         # for section editing, where these comments cause confusion when
 228         # counting the sections in the wikisource
 229         function strip( $text, &$state, $stripcomments = false ) {
 230                 $render = ($this->mOutputType == OT_HTML);
 231                 $html_content = array();
 232                 $nowiki_content = array();
 233                 $math_content = array();
 234                 $pre_content = array();
 235                 $comment_content = array();
 236                 $ext_content = array();
 237
 238                 # Replace any instances of the placeholders
 239                 $uniq_prefix = UNIQ_PREFIX;
 240                 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
 241
 242                 # html
 243                 global $wgRawHtml;
 244                 if( $wgRawHtml ) {
 245                         $text = Parser::extractTags('html', $text, $html_content, $uniq_prefix);
 246                         foreach( $html_content as $marker => $content ) {
 247                                 if ($render ) {
 248                                         # Raw and unchecked for validity.
 249                                         $html_content[$marker] = $content;
 250                                 } else {
 251                                         $html_content[$marker] = '<html>'.$content.'</html>';
 252                                 }
 253                         }
 254                 }
 255
 256                 # nowiki
 257                 $text = Parser::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
 258                 foreach( $nowiki_content as $marker => $content ) {
 259                         if( $render ){
 260                                 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
 261                         } else {
 262                                 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
 263                         }
 264                 }
 265
 266                 # math
 267                 $text = Parser::extractTags('math', $text, $math_content, $uniq_prefix);
 268                 foreach( $math_content as $marker => $content ){
 269                         if( $render ) {
 270                                 if( $this->mOptions->getUseTeX() ) {
 271                                         $math_content[$marker] = renderMath( $content );
 272                                 } else {
 273                                         $math_content[$marker] = '&lt;math&gt;'.$content.'&lt;math&gt;';
 274                                 }
 275                         } else {
 276                                 $math_content[$marker] = '<math>'.$content.'</math>';
 277                         }
 278                 }
 279
 280                 # pre
 281                 $text = Parser::extractTags('pre', $text, $pre_content, $uniq_prefix);
 282                 foreach( $pre_content as $marker => $content ){
 283                         if( $render ){
 284                                 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
 285                         } else {
 286                                 $pre_content[$marker] = '<pre>'.$content.'</pre>';
 287                         }
 288                 }
 289
 290                 # Comments
 291                 if($stripcomments) {
 292                         $text = Parser::extractTags(STRIP_COMMENTS, $text, $comment_content, $uniq_prefix);
 293                         foreach( $comment_content as $marker => $content ){
 294                                 $comment_content[$marker] = '<!--'.$content.'-->';
 295                         }
 296                 }
 297
 298                 # Extensions
 299                 foreach ( $this->mTagHooks as $tag => $callback ) {
 300                         $ext_contents[$tag] = array();
 301                         $text = Parser::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
 302                         foreach( $ext_content[$tag] as $marker => $content ) {
 303                                 if ( $render ) {
 304                                         $ext_content[$tag][$marker] = $callback( $content );
 305                                 } else {
 306                                         $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
 307                                 }
 308                         }
 309                 }
 310
 311                 # Merge state with the pre-existing state, if there is one
 312                 if ( $state ) {
 313                         $state['html'] = $state['html'] + $html_content;
 314                         $state['nowiki'] = $state['nowiki'] + $nowiki_content;
 315                         $state['math'] = $state['math'] + $math_content;
 316                         $state['pre'] = $state['pre'] + $pre_content;
 317                         $state['comment'] = $state['comment'] + $comment_content;
 318
 319                         foreach( $ext_content as $tag => $array ) {
 320                                 if ( array_key_exists( $tag, $state ) ) {
 321                                         $state[$tag] = $state[$tag] + $array;
 322                                 }
 323                         }
 324                 } else {
 325                         $state = array(
 326                           'html' => $html_content,
 327                           'nowiki' => $nowiki_content,
 328                           'math' => $math_content,
 329                           'pre' => $pre_content,
 330                           'comment' => $comment_content,
 331                         ) + $ext_content;
 332                 }
 333                 return $text;
 334         }
 335
 336         # always call unstripNoWiki() after this one
 337         function unstrip( $text, &$state ) {
 338                 # Must expand in reverse order, otherwise nested tags will be corrupted
 339                 $contentDict = end( $state );
 340                 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
 341                         if( key($state) != 'nowiki' && key($state) != 'html') {
 342                                 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
 343                                         $text = str_replace( key( $contentDict ), $content, $text );
 344                                 }
 345                         }
 346                 }
 347
 348                 return $text;
 349         }
 350         # always call this after unstrip() to preserve the order
 351         function unstripNoWiki( $text, &$state ) {
 352                 # Must expand in reverse order, otherwise nested tags will be corrupted
 353                 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
 354                         $text = str_replace( key( $state['nowiki'] ), $content, $text );
 355                 }
 356
 357                 global $wgRawHtml;
 358                 if ($wgRawHtml) {
 359                         for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
 360                                 $text = str_replace( key( $state['html'] ), $content, $text );
 361                         }
 362                 }
 363
 364                 return $text;
 365         }
 366
 367         # Add an item to the strip state
 368         # Returns the unique tag which must be inserted into the stripped text
 369         # The tag will be replaced with the original text in unstrip()
 370         function insertStripItem( $text, &$state ) {
 371                 $rnd = UNIQ_PREFIX . '-item' . Parser::getRandomString();
 372                 if ( !$state ) {
 373                         $state = array(
 374                           'html' => array(),
 375                           'nowiki' => array(),
 376                           'math' => array(),
 377                           'pre' => array()
 378                         );
 379                 }
 380                 $state['item'][$rnd] = $text;
 381                 return $rnd;
 382         }
 383
 384         # Return allowed HTML attributes
 385         function getHTMLattrs () {
 386                 $htmlattrs = array( # Allowed attributes--no scripting, etc.
 387                                 'title', 'align', 'lang', 'dir', 'width', 'height',
 388                                 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
 389                                 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
 390                                 /* FONT */ 'type', 'start', 'value', 'compact',
 391                                 /* For various lists, mostly deprecated but safe */
 392                                 'summary', 'width', 'border', 'frame', 'rules',
 393                                 'cellspacing', 'cellpadding', 'valign', 'char',
 394                                 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
 395                                 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
 396                                 'id', 'class', 'name', 'style' /* For CSS */
 397                                 );
 398                 return $htmlattrs ;
 399         }
 400
 401         # Remove non approved attributes and javascript in css
 402         function fixTagAttributes ( $t ) {
 403                 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
 404                 $htmlattrs = $this->getHTMLattrs() ;
 405
 406                 # Strip non-approved attributes from the tag
 407                 $t = preg_replace(
 408                         '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
 409                         "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
 410                         $t);
 411
 412                 $t = str_replace ( '<></>' , '' , $t ) ; # This should fix bug 980557
 413
 414                 # Strip javascript "expression" from stylesheets. Brute force approach:
 415                 # If anythin offensive is found, all attributes of the HTML tag are dropped
 416
 417                 if( preg_match(
 418                         '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
 419                         wfMungeToUtf8( $t ) ) )
 420                 {
 421                         $t='';
 422                 }
 423
 424                 return trim ( $t ) ;
 425         }
 426
 427         # interface with html tidy, used if $wgUseTidy = true
 428         function tidy ( $text ) {
 429                 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
 430                 global $wgInputEncoding, $wgOutputEncoding;
 431                 $fname = 'Parser::tidy';
 432                 wfProfileIn( $fname );
 433
 434                 $cleansource = '';
 435                 switch(strtoupper($wgOutputEncoding)) {
 436                         case 'ISO-8859-1':
 437                                 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)? ' -latin1':' -raw';
 438                                 break;
 439                         case 'UTF-8':
 440                                 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)? ' -utf8':' -raw';
 441                                 break;
 442                         default:
 443                                 $wgTidyOpts .= ' -raw';
 444                         }
 445
 446                 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
 447 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
 448 '<head><title>test</title></head><body>'.$text.'</body></html>';
 449                 $descriptorspec = array(
 450                         0 => array('pipe', 'r'),
 451                         1 => array('pipe', 'w'),
 452                         2 => array('file', '/dev/null', 'a')
 453                 );
 454                 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts", $descriptorspec, $pipes);
 455                 if (is_resource($process)) {
 456                         fwrite($pipes[0], $wrappedtext);
 457                         fclose($pipes[0]);
 458                         while (!feof($pipes[1])) {
 459                                 $cleansource .= fgets($pipes[1], 1024);
 460                         }
 461                         fclose($pipes[1]);
 462                         $return_value = proc_close($process);
 463                 }
 464
 465                 wfProfileOut( $fname );
 466
 467                 if( $cleansource == '' && $text != '') {
 468                         wfDebug( "Tidy error detected!\n" );
 469                         return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
 470                 } else {
 471                         return $cleansource;
 472                 }
 473         }
 474
 475         # parse the wiki syntax used to render tables
 476         function doTableStuff ( $t ) {
 477                 $fname = 'Parser::doTableStuff';
 478                 wfProfileIn( $fname );
 479
 480                 $t = explode ( "\n" , $t ) ;
 481                 $td = array () ; # Is currently a td tag open?
 482                 $ltd = array () ; # Was it TD or TH?
 483                 $tr = array () ; # Is currently a tr tag open?
 484                 $ltr = array () ; # tr attributes
 485                 $indent_level = 0; # indent level of the table
 486                 foreach ( $t AS $k => $x )
 487                 {
 488                         $x = trim ( $x ) ;
 489                         $fc = substr ( $x , 0 , 1 ) ;
 490                         if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
 491                                 $indent_level = strlen( $matches[1] );
 492                                 $t[$k] = "\n" .
 493                                         str_repeat( '<dl><dd>', $indent_level ) .
 494                                         '<table ' . $this->fixTagAttributes ( $matches[2] ) . '>' ;
 495                                 array_push ( $td , false ) ;
 496                                 array_push ( $ltd , '' ) ;
 497                                 array_push ( $tr , false ) ;
 498                                 array_push ( $ltr , '' ) ;
 499                         }
 500                         else if ( count ( $td ) == 0 ) { } # Don't do any of the following
 501                         else if ( '|}' == substr ( $x , 0 , 2 ) ) {
 502                                 $z = "</table>\n" ;
 503                                 $l = array_pop ( $ltd ) ;
 504                                 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 505                                 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 506                                 array_pop ( $ltr ) ;
 507                                 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
 508                         }
 509                         else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
 510                                 $x = substr ( $x , 1 ) ;
 511                                 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
 512                                 $z = '' ;
 513                                 $l = array_pop ( $ltd ) ;
 514                                 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 515                                 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 516                                 array_pop ( $ltr ) ;
 517                                 $t[$k] = $z ;
 518                                 array_push ( $tr , false ) ;
 519                                 array_push ( $td , false ) ;
 520                                 array_push ( $ltd , '' ) ;
 521                                 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
 522                         }
 523                         else if ( '|' == $fc || '!' == $fc || '|+' == substr ( $x , 0 , 2 ) ) { # Caption
 524                                 if ( '|+' == substr ( $x , 0 , 2 ) ) {
 525                                         $fc = '+' ;
 526                                         $x = substr ( $x , 1 ) ;
 527                                 }
 528                                 $after = substr ( $x , 1 ) ;
 529                                 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
 530                                 $after = explode ( '||' , $after ) ;
 531                                 $t[$k] = '' ;
 532                                 foreach ( $after AS $theline )
 533                                 {
 534                                         $z = '' ;
 535                                         if ( $fc != '+' )
 536                                         {
 537                                                 $tra = array_pop ( $ltr ) ;
 538                                                 if ( !array_pop ( $tr ) ) $z = '<tr '.$tra.">\n" ;
 539                                                 array_push ( $tr , true ) ;
 540                                                 array_push ( $ltr , '' ) ;
 541                                         }
 542
 543                                         $l = array_pop ( $ltd ) ;
 544                                         if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 545                                         if ( $fc == '|' ) $l = 'td' ;
 546                                         else if ( $fc == '!' ) $l = 'th' ;
 547                                         else if ( $fc == '+' ) $l = 'caption' ;
 548                                         else $l = '' ;
 549                                         array_push ( $ltd , $l ) ;
 550                                         $y = explode ( '|' , $theline , 2 ) ;
 551                                         if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
 552                                         else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
 553                                         $t[$k] .= $y ;
 554                                         array_push ( $td , true ) ;
 555                                 }
 556                         }
 557                 }
 558
 559                 # Closing open td, tr && table
 560                 while ( count ( $td ) > 0 )
 561                 {
 562                         if ( array_pop ( $td ) ) $t[] = '</td>' ;
 563                         if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
 564                         $t[] = '</table>' ;
 565                 }
 566
 567                 $t = implode ( "\n" , $t ) ;
 568                 #               $t = $this->removeHTMLtags( $t );
 569                 wfProfileOut( $fname );
 570                 return $t ;
 571         }
 572
 573         # Parses the text and adds the result to the strip state
 574         # Returns the strip tag
 575         function stripParse( $text, $newline, $args ) {
 576                 $text = $this->strip( $text, $this->mStripState );
 577                 $text = $this->internalParse( $text, (bool)$newline, $args, false );
 578                 return $newline.$this->insertStripItem( $text, $this->mStripState );
 579         }
 580
 581         function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
 582                 $fname = 'Parser::internalParse';
 583                 wfProfileIn( $fname );
 584
 585                 $text = $this->removeHTMLtags( $text );
 586                 $text = $this->replaceVariables( $text, $args );
 587
 588                 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
 589
 590                 $text = $this->doHeadings( $text );
 591                 if($this->mOptions->getUseDynamicDates()) {
 592                         global $wgDateFormatter;
 593                         $text = $wgDateFormatter->reformat( $this->mOptions->getDateFormat(), $text );
 594                 }
 595                 $text = $this->doAllQuotes( $text );
 596                 $text = $this->replaceExternalLinks( $text );
 597                 $text = $this->doMagicLinks( $text );
 598                 $text = $this->replaceInternalLinks ( $text );
 599                 $text = $this->replaceInternalLinks ( $text );
 600
 601                 $text = $this->unstrip( $text, $this->mStripState );
 602                 $text = $this->unstripNoWiki( $text, $this->mStripState );
 603
 604                 $text = $this->doTableStuff( $text );
 605                 $text = $this->formatHeadings( $text, $isMain );
 606                 $sk =& $this->mOptions->getSkin();
 607                 $text = $sk->transformContent( $text );
 608
 609                 wfProfileOut( $fname );
 610                 return $text;
 611         }
 612
 613         /* private */ function &doMagicLinks( &$text ) {
 614                 global $wgUseGeoMode;
 615                 $text = $this->magicISBN( $text );
 616                 if ( isset( $wgUseGeoMode ) && $wgUseGeoMode ) {
 617                         $text = $this->magicGEO( $text );
 618                 }
 619                 $text = $this->magicRFC( $text );
 620                 return $text;
 621         }
 622
 623         # Parse ^^ tokens and return html
 624         /* private */ function doExponent ( $text ) {
 625                 $fname = 'Parser::doExponent';
 626                 wfProfileIn( $fname);
 627                 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
 628                 wfProfileOut( $fname);
 629                 return $text;
 630         }
 631
 632         # Parse headers and return html
 633         /* private */ function doHeadings( $text ) {
 634                 $fname = 'Parser::doHeadings';
 635                 wfProfileIn( $fname );
 636                 for ( $i = 6; $i >= 1; --$i ) {
 637                         $h = substr( '======', 0, $i );
 638                         $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
 639                           "<h{$i}>\\1</h{$i}>\\2", $text );
 640                 }
 641                 wfProfileOut( $fname );
 642                 return $text;
 643         }
 644
 645         /* private */ function doAllQuotes( $text ) {
 646                 $fname = 'Parser::doAllQuotes';
 647                 wfProfileIn( $fname );
 648                 $outtext = '';
 649                 $lines = explode( "\n", $text );
 650                 foreach ( $lines as $line ) {
 651                         $outtext .= $this->doQuotes ( $line ) . "\n";
 652                 }
 653                 $outtext = substr($outtext, 0,-1);
 654                 wfProfileOut( $fname );
 655                 return $outtext;
 656         }
 657
 658         /* private */ function doQuotes( $text ) {
 659                 $arr = preg_split ("/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE);
 660                 if (count ($arr) == 1)
 661                         return $text;
 662                 else
 663                 {
 664                         # First, do some preliminary work. This may shift some apostrophes from
 665                         # being mark-up to being text. It also counts the number of occurrences
 666                         # of bold and italics mark-ups.
 667                         $i = 0;
 668                         $numbold = 0;
 669                         $numitalics = 0;
 670                         foreach ($arr as $r)
 671                         {
 672                                 if (($i % 2) == 1)
 673                                 {
 674                                         # If there are ever four apostrophes, assume the first is supposed to
 675                                         # be text, and the remaining three constitute mark-up for bold text.
 676                                         if (strlen ($arr[$i]) == 4)
 677                                         {
 678                                                 $arr[$i-1] .= "'";
 679                                                 $arr[$i] = "'''";
 680                                         }
 681                                         # If there are more than 5 apostrophes in a row, assume they're all
 682                                         # text except for the last 5.
 683                                         else if (strlen ($arr[$i]) > 5)
 684                                         {
 685                                                 $arr[$i-1] .= str_repeat ("'", strlen ($arr[$i]) - 5);
 686                                                 $arr[$i] = "'''''";
 687                                         }
 688                                         # Count the number of occurrences of bold and italics mark-ups.
 689                                         # We are not counting sequences of five apostrophes.
 690                                         if (strlen ($arr[$i]) == 2) $numitalics++;  else
 691                                         if (strlen ($arr[$i]) == 3) $numbold++;     else
 692                                         if (strlen ($arr[$i]) == 5) { $numitalics++; $numbold++; }
 693                                 }
 694                                 $i++;
 695                         }
 696
 697                         # If there is an odd number of both bold and italics, it is likely
 698                         # that one of the bold ones was meant to be an apostrophe followed
 699                         # by italics. Which one we cannot know for certain, but it is more
 700                         # likely to be one that has a single-letter word before it.
 701                         if (($numbold % 2 == 1) && ($numitalics % 2 == 1))
 702                         {
 703                                 $i = 0;
 704                                 $firstsingleletterword = -1;
 705                                 $firstmultiletterword = -1;
 706                                 $firstspace = -1;
 707                                 foreach ($arr as $r)
 708                                 {
 709                                         if (($i % 2 == 1) and (strlen ($r) == 3))
 710                                         {
 711                                                 $x1 = substr ($arr[$i-1], -1);
 712                                                 $x2 = substr ($arr[$i-1], -2, 1);
 713                                                 if ($x1 == ' ') {
 714                                                         if ($firstspace == -1) $firstspace = $i;
 715                                                 } else if ($x2 == ' ') {
 716                                                         if ($firstsingleletterword == -1) $firstsingleletterword = $i;
 717                                                 } else {
 718                                                         if ($firstmultiletterword == -1) $firstmultiletterword = $i;
 719                                                 }
 720                                         }
 721                                         $i++;
 722                                 }
 723
 724                                 # If there is a single-letter word, use it!
 725                                 if ($firstsingleletterword > -1)
 726                                 {
 727                                         $arr [ $firstsingleletterword ] = "''";
 728                                         $arr [ $firstsingleletterword-1 ] .= "'";
 729                                 }
 730                                 # If not, but there's a multi-letter word, use that one.
 731                                 else if ($firstmultiletterword > -1)
 732                                 {
 733                                         $arr [ $firstmultiletterword ] = "''";
 734                                         $arr [ $firstmultiletterword-1 ] .= "'";
 735                                 }
 736                                 # ... otherwise use the first one that has neither.
 737                                 # (notice that it is possible for all three to be -1 if, for example,
 738                                 # there is only one pentuple-apostrophe in the line)
 739                                 else if ($firstspace > -1)
 740                                 {
 741                                         $arr [ $firstspace ] = "''";
 742                                         $arr [ $firstspace-1 ] .= "'";
 743                                 }
 744                         }
 745
 746                         # Now let's actually convert our apostrophic mush to HTML!
 747                         $output = '';
 748                         $buffer = '';
 749                         $state = '';
 750                         $i = 0;
 751                         foreach ($arr as $r)
 752                         {
 753                                 if (($i % 2) == 0)
 754                                 {
 755                                         if ($state == 'both')
 756                                                 $buffer .= $r;
 757                                         else
 758                                                 $output .= $r;
 759                                 }
 760                                 else
 761                                 {
 762                                         if (strlen ($r) == 2)
 763                                         {
 764                                                 if ($state == 'em')
 765                                                 { $output .= '</em>'; $state = ''; }
 766                                                 else if ($state == 'strongem')
 767                                                 { $output .= '</em>'; $state = 'strong'; }
 768                                                 else if ($state == 'emstrong')
 769                                                 { $output .= '</strong></em><strong>'; $state = 'strong'; }
 770                                                 else if ($state == 'both')
 771                                                 { $output .= '<strong><em>'.$buffer.'</em>'; $state = 'strong'; }
 772                                                 else # $state can be 'strong' or ''
 773                                                 { $output .= '<em>'; $state .= 'em'; }
 774                                         }
 775                                         else if (strlen ($r) == 3)
 776                                         {
 777                                                 if ($state == 'strong')
 778                                                 { $output .= '</strong>'; $state = ''; }
 779                                                 else if ($state == 'strongem')
 780                                                 { $output .= '</em></strong><em>'; $state = 'em'; }
 781                                                 else if ($state == 'emstrong')
 782                                                 { $output .= '</strong>'; $state = 'em'; }
 783                                                 else if ($state == 'both')
 784                                                 { $output .= '<em><strong>'.$buffer.'</strong>'; $state = 'em'; }
 785                                                 else # $state can be 'em' or ''
 786                                                 { $output .= '<strong>'; $state .= 'strong'; }
 787                                         }
 788                                         else if (strlen ($r) == 5)
 789                                         {
 790                                                 if ($state == 'strong')
 791                                                 { $output .= '</strong><em>'; $state = 'em'; }
 792                                                 else if ($state == 'em')
 793                                                 { $output .= '</em><strong>'; $state = 'strong'; }
 794                                                 else if ($state == 'strongem')
 795                                                 { $output .= '</em></strong>'; $state = ''; }
 796                                                 else if ($state == 'emstrong')
 797                                                 { $output .= '</strong></em>'; $state = ''; }
 798                                                 else if ($state == 'both')
 799                                                 { $output .= '<em><strong>'.$buffer.'</strong></em>'; $state = ''; }
 800                                                 else # ($state == '')
 801                                                 { $buffer = ''; $state = 'both'; }
 802                                         }
 803                                 }
 804                                 $i++;
 805                         }
 806                         # Now close all remaining tags.  Notice that the order is important.
 807                         if ($state == 'strong' || $state == 'emstrong')
 808                                 $output .= '</strong>';
 809                         if ($state == 'em' || $state == 'strongem' || $state == 'emstrong')
 810                                 $output .= '</em>';
 811                         if ($state == 'strongem')
 812                                 $output .= '</strong>';
 813                         if ($state == 'both')
 814                                 $output .= '<strong><em>'.$buffer.'</em></strong>';
 815                         return $output;
 816                 }
 817         }
 818
 819         # Note: we have to do external links before the internal ones,
 820         # and otherwise take great care in the order of things here, so
 821         # that we don't end up interpreting some URLs twice.
 822
 823         /* private */ function replaceExternalLinks( $text ) {
 824                 $fname = 'Parser::replaceExternalLinks';
 825                 wfProfileIn( $fname );
 826
 827                 $sk =& $this->mOptions->getSkin();
 828                 $linktrail = wfMsg('linktrail');
 829                 $bits = preg_split( EXT_LINK_BRACKETED, $text, -1, PREG_SPLIT_DELIM_CAPTURE );
 830
 831                 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
 832
 833                 $i = 0;
 834                 while ( $i<count( $bits ) ) {
 835                         $url = $bits[$i++];
 836                         $protocol = $bits[$i++];
 837                         $text = $bits[$i++];
 838                         $trail = $bits[$i++];
 839
 840                         # If the link text is an image URL, replace it with an <img> tag
 841                         # This happened by accident in the original parser, but some people used it extensively
 842                         $img = $this->maybeMakeImageLink( $text );
 843                         if ( $img !== false ) {
 844                                 $text = $img;
 845                         }
 846
 847                         $dtrail = '';
 848
 849                         # No link text, e.g. [http://domain.tld/some.link]
 850                         if ( $text == '' ) {
 851                                 # Autonumber if allowed
 852                                 if ( strpos( HTTP_PROTOCOLS, $protocol ) !== false ) {
 853                                         $text = '[' . ++$this->mAutonumber . ']';
 854                                 } else {
 855                                         # Otherwise just use the URL
 856                                         $text = htmlspecialchars( $url );
 857                                 }
 858                         } else {
 859                                 # Have link text, e.g. [http://domain.tld/some.link text]s
 860                                 # Check for trail
 861                                 if ( preg_match( $linktrail, $trail, $m2 ) ) {
 862                                         $dtrail = $m2[1];
 863                                         $trail = $m2[2];
 864                                 }
 865                         }
 866
 867                         $encUrl = htmlspecialchars( $url );
 868                         # Bit in parentheses showing the URL for the printable version
 869                         if( $url == $text || preg_match( "!$protocol://" . preg_quote( $text, '/' ) . "/?$!", $url ) ) {
 870                                 $paren = '';
 871                         } else {
 872                                 # Expand the URL for printable version
 873                                 if ( ! $sk->suppressUrlExpansion() ) {
 874                                         $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $encUrl ) . "</i>)</span>";
 875                                 } else {
 876                                         $paren = '';
 877                                 }
 878                         }
 879
 880                         # Process the trail (i.e. everything after this link up until start of the next link),
 881                         # replacing any non-bracketed links
 882                         $trail = $this->replaceFreeExternalLinks( $trail );
 883
 884                         $la = $sk->getExternalLinkAttributes( $url, $text );
 885
 886                         # Use the encoded URL
 887                         # This means that users can paste URLs directly into the text
 888                         # Funny characters like &ouml; aren't valid in URLs anyway
 889                         # This was changed in August 2004
 890                         $s .= "<a href=\"{$url}\" {$la}>{$text}</a>{$dtrail}{$paren}{$trail}";
 891                 }
 892
 893                 wfProfileOut( $fname );
 894                 return $s;
 895         }
 896
 897         # Replace anything that looks like a URL with a link
 898         function replaceFreeExternalLinks( $text ) {
 899                 $bits = preg_split( '/((?:'.URL_PROTOCOLS.'):)/', $text, -1, PREG_SPLIT_DELIM_CAPTURE );
 900                 $s = array_shift( $bits );
 901                 $i = 0;
 902
 903                 $sk =& $this->mOptions->getSkin();
 904
 905                 while ( $i < count( $bits ) ){
 906                         $protocol = $bits[$i++];
 907                         $remainder = $bits[$i++];
 908
 909                         if ( preg_match( '/^('.EXT_LINK_URL_CLASS.'+)(.*)$/s', $remainder, $m ) ) {
 910                                 # Found some characters after the protocol that look promising
 911                                 $url = $protocol . $m[1];
 912                                 $trail = $m[2];
 913
 914                                 # Move trailing punctuation to $trail
 915                                 $sep = ',;\.:!?';
 916                                 # If there is no left bracket, then consider right brackets fair game too
 917                                 if ( strpos( $url, '(' ) === false ) {
 918                                         $sep .= ')';
 919                                 }
 920
 921                                 $numSepChars = strspn( strrev( $url ), $sep );
 922                                 if ( $numSepChars ) {
 923                                         $trail = substr( $url, -$numSepChars ) . $trail;
 924                                         $url = substr( $url, 0, -$numSepChars );
 925                                 }
 926
 927                                 # Replace &amp; from obsolete syntax with &
 928                                 $url = str_replace( '&amp;', '&', $url );
 929
 930                                 # Is this an external image?
 931                                 $text = $this->maybeMakeImageLink( $url );
 932                                 if ( $text === false ) {
 933                                         # Not an image, make a link
 934                                         $text = $sk->makeExternalLink( $url, $url );
 935                                 }
 936                                 $s .= $text . $trail;
 937                         } else {
 938                                 $s .= $protocol . $remainder;
 939                         }
 940                 }
 941                 return $s;
 942         }
 943
 944         # make an image if it's allowed
 945         function maybeMakeImageLink( $url ) {
 946                 $sk =& $this->mOptions->getSkin();
 947                 $text = false;
 948                 if ( $this->mOptions->getAllowExternalImages() ) {
 949                         if ( preg_match( EXT_IMAGE_REGEX, $url ) ) {
 950                                 # Image found
 951                                 $text = $sk->makeImage( htmlspecialchars( $url ) );
 952                         }
 953                 }
 954                 return $text;
 955         }
 956
 957         # The wikilinks [[ ]] are procedeed here.
 958         /* private */ function replaceInternalLinks( $s ) {
 959                 global $wgLang, $wgLinkCache;
 960                 global $wgNamespacesWithSubpages, $wgLanguageCode;
 961                 static $fname = 'Parser::replaceInternalLinks' ;
 962                 wfProfileIn( $fname );
 963
 964                 wfProfileIn( $fname.'-setup' );
 965                 static $tc = FALSE;
 966                 # the % is needed to support urlencoded titles as well
 967                 if ( !$tc ) { $tc = Title::legalChars() . '#%'; }
 968                 $sk =& $this->mOptions->getSkin();
 969
 970                 $redirect = MagicWord::get ( MAG_REDIRECT ) ;
 971
 972                 $a = explode( '[[', ' ' . $s );
 973                 $s = array_shift( $a );
 974                 $s = substr( $s, 1 );
 975
 976                 # Match a link having the form [[namespace:link|alternate]]trail
 977                 static $e1 = FALSE;
 978                 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
 979                 # Match the end of a line for a word that's not followed by whitespace,
 980                 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
 981                 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
 982
 983                 $useLinkPrefixExtension = $wgLang->linkPrefixExtension();
 984                 # Special and Media are pseudo-namespaces; no pages actually exist in them
 985
 986                 $nottalk = !Namespace::isTalk( $this->mTitle->getNamespace() );
 987
 988                 if ( $useLinkPrefixExtension ) {
 989                         if ( preg_match( $e2, $s, $m ) ) {
 990                                 $first_prefix = $m[2];
 991                                 $s = $m[1];
 992                         } else {
 993                                 $first_prefix = false;
 994                         }
 995                 } else {
 996                         $prefix = '';
 997                 }
 998
 999                 wfProfileOut( $fname.'-setup' );
1000
1001                 # start procedeeding each line
1002                 foreach ( $a as $line ) {
1003                         wfProfileIn( $fname.'-prefixhandling' );
1004                         if ( $useLinkPrefixExtension ) {
1005                                 if ( preg_match( $e2, $s, $m ) ) {
1006                                         $prefix = $m[2];
1007                                         $s = $m[1];
1008                                 } else {
1009                                         $prefix='';
1010                                 }
1011                                 # first link
1012                                 if($first_prefix) {
1013                                         $prefix = $first_prefix;
1014                                         $first_prefix = false;
1015                                 }
1016                         }
1017                         wfProfileOut( $fname.'-prefixhandling' );
1018
1019                         if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1020                                 $text = $m[2];
1021                                 # fix up urlencoded title texts
1022                                 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1023                                 $trail = $m[3];
1024                         } else { # Invalid form; output directly
1025                                 $s .= $prefix . '[[' . $line ;
1026                                 continue;
1027                         }
1028
1029                         # Valid link forms:
1030                         # Foobar -- normal
1031                         # :Foobar -- override special treatment of prefix (images, language links)
1032                         # /Foobar -- convert to CurrentPage/Foobar
1033                         # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1034
1035                         # Look at the first character
1036                         $c = substr($m[1],0,1);
1037                         $noforce = ($c != ':');
1038
1039                         # subpage
1040                         if( $c == '/' ) {
1041                                 # / at end means we don't want the slash to be shown
1042                                 if(substr($m[1],-1,1)=='/') {
1043                                         $m[1]=substr($m[1],1,strlen($m[1])-2);
1044                                         $noslash=$m[1];
1045                                 } else {
1046                                         $noslash=substr($m[1],1);
1047                                 }
1048
1049                                 # Some namespaces don't allow subpages
1050                                 if(!empty($wgNamespacesWithSubpages[$this->mTitle->getNamespace()])) {
1051                                         # subpages allowed here
1052                                         $link = $this->mTitle->getPrefixedText(). '/' . trim($noslash);
1053                                         if( '' == $text ) {
1054                                                 $text= $m[1];
1055                                         } # this might be changed for ugliness reasons
1056                                 } else {
1057                                         # no subpage allowed, use standard link
1058                                         $link = $noslash;
1059                                 }
1060
1061                         } elseif( $noforce ) { # no subpage
1062                                 $link = $m[1];
1063                         } else {
1064                                 # We don't want to keep the first character
1065                                 $link = substr( $m[1], 1 );
1066                         }
1067
1068                         $wasblank = ( '' == $text );
1069                         if( $wasblank ) $text = $link;
1070
1071                         $nt = Title::newFromText( $link );
1072                         if( !$nt ) {
1073                                 $s .= $prefix . '[[' . $line;
1074                                 continue;
1075                         }
1076
1077                         $ns = $nt->getNamespace();
1078                         $iw = $nt->getInterWiki();
1079
1080                         # Link not escaped by : , create the various objects
1081                         if( $noforce ) {
1082
1083                                 # Interwikis
1084                                 if( $iw && $this->mOptions->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
1085                                         array_push( $this->mOutput->mLanguageLinks, $nt->getFullText() );
1086                                         $tmp = $prefix . $trail ;
1087                                         $s .= (trim($tmp) == '')? '': $tmp;
1088                                         continue;
1089                                 }
1090
1091                                 if ( $ns == NS_IMAGE ) {
1092                                         $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1093                                         $wgLinkCache->addImageLinkObj( $nt );
1094                                         continue;
1095                                 }
1096
1097                                 if ( $ns == NS_CATEGORY ) {
1098                                         $t = $nt->getText() ;
1099                                         $nnt = Title::newFromText ( Namespace::getCanonicalName(NS_CATEGORY).':'.$t ) ;
1100
1101                                         $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1102                                         $pPLC=$sk->postParseLinkColour();
1103                                         $sk->postParseLinkColour( false );
1104                                         $t = $sk->makeLinkObj( $nnt, $t, '', '' , $prefix );
1105                                         $sk->postParseLinkColour( $pPLC );
1106                                         $wgLinkCache->resume();
1107
1108                                         $sortkey = $wasblank ? $this->mTitle->getPrefixedText() : $text;
1109                                         $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1110                                         $this->mOutput->mCategoryLinks[] = $t ;
1111                                         $s .= $prefix . $trail ;
1112                                         continue;
1113                                 }
1114                         }
1115
1116                         if( ( $nt->getPrefixedText() === $this->mTitle->getPrefixedText() ) &&
1117                             ( strpos( $link, '#' ) === FALSE ) ) {
1118                                 # Self-links are handled specially; generally de-link and change to bold.
1119                                 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1120                                 continue;
1121                         }
1122
1123                         if( $ns == NS_MEDIA ) {
1124                                 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1125                                 $wgLinkCache->addImageLinkObj( $nt );
1126                                 continue;
1127                         } elseif( $ns == NS_SPECIAL ) {
1128                                 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1129                                 continue;
1130                         }
1131                         $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1132                 }
1133                 wfProfileOut( $fname );
1134                 return $s;
1135         }
1136
1137         # Some functions here used by doBlockLevels()
1138         #
1139         /* private */ function closeParagraph() {
1140                 $result = '';
1141                 if ( '' != $this->mLastSection ) {
1142                         $result = '</' . $this->mLastSection  . ">\n";
1143                 }
1144                 $this->mInPre = false;
1145                 $this->mLastSection = '';
1146                 return $result;
1147         }
1148         # getCommon() returns the length of the longest common substring
1149         # of both arguments, starting at the beginning of both.
1150         #
1151         /* private */ function getCommon( $st1, $st2 ) {
1152                 $fl = strlen( $st1 );
1153                 $shorter = strlen( $st2 );
1154                 if ( $fl < $shorter ) { $shorter = $fl; }
1155
1156                 for ( $i = 0; $i < $shorter; ++$i ) {
1157                         if ( $st1{$i} != $st2{$i} ) { break; }
1158                 }
1159                 return $i;
1160         }
1161         # These next three functions open, continue, and close the list
1162         # element appropriate to the prefix character passed into them.
1163         #
1164         /* private */ function openList( $char ) {
1165                 $result = $this->closeParagraph();
1166
1167                 if ( '*' == $char ) { $result .= '<ul><li>'; }
1168                 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1169                 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1170                 else if ( ';' == $char ) {
1171                         $result .= '<dl><dt>';
1172                         $this->mDTopen = true;
1173                 }
1174                 else { $result = '<!-- ERR 1 -->'; }
1175
1176                 return $result;
1177         }
1178
1179         /* private */ function nextItem( $char ) {
1180                 if ( '*' == $char || '#' == $char ) { return '</li><li>'; }
1181                 else if ( ':' == $char || ';' == $char ) {
1182                         $close = '</dd>';
1183                         if ( $this->mDTopen ) { $close = '</dt>'; }
1184                         if ( ';' == $char ) {
1185                                 $this->mDTopen = true;
1186                                 return $close . '<dt>';
1187                         } else {
1188                                 $this->mDTopen = false;
1189                                 return $close . '<dd>';
1190                         }
1191                 }
1192                 return '<!-- ERR 2 -->';
1193         }
1194
1195         /* private */ function closeList( $char ) {
1196                 if ( '*' == $char ) { $text = '</li></ul>'; }
1197                 else if ( '#' == $char ) { $text = '</li></ol>'; }
1198                 else if ( ':' == $char ) {
1199                         if ( $this->mDTopen ) {
1200                                 $this->mDTopen = false;
1201                                 $text = '</dt></dl>';
1202                         } else {
1203                                 $text = '</dd></dl>';
1204                         }
1205                 }
1206                 else {  return '<!-- ERR 3 -->'; }
1207                 return $text."\n";
1208         }
1209
1210         /* private */ function doBlockLevels( $text, $linestart ) {
1211                 $fname = 'Parser::doBlockLevels';
1212                 wfProfileIn( $fname );
1213
1214                 # Parsing through the text line by line.  The main thing
1215                 # happening here is handling of block-level elements p, pre,
1216                 # and making lists from lines starting with * # : etc.
1217                 #
1218                 $textLines = explode( "\n", $text );
1219
1220                 $lastPrefix = $output = $lastLine = '';
1221                 $this->mDTopen = $inBlockElem = false;
1222                 $prefixLength = 0;
1223                 $paragraphStack = false;
1224
1225                 if ( !$linestart ) {
1226                         $output .= array_shift( $textLines );
1227                 }
1228                 foreach ( $textLines as $oLine ) {
1229                         $lastPrefixLength = strlen( $lastPrefix );
1230                         $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1231                         $preOpenMatch = preg_match('/<pre/i', $oLine );
1232                         if ( !$this->mInPre ) {
1233                                 # Multiple prefixes may abut each other for nested lists.
1234                                 $prefixLength = strspn( $oLine, '*#:;' );
1235                                 $pref = substr( $oLine, 0, $prefixLength );
1236
1237                                 # eh?
1238                                 $pref2 = str_replace( ';', ':', $pref );
1239                                 $t = substr( $oLine, $prefixLength );
1240                                 $this->mInPre = !empty($preOpenMatch);
1241                         } else {
1242                                 # Don't interpret any other prefixes in preformatted text
1243                                 $prefixLength = 0;
1244                                 $pref = $pref2 = '';
1245                                 $t = $oLine;
1246                         }
1247
1248                         # List generation
1249                         if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1250                                 # Same as the last item, so no need to deal with nesting or opening stuff
1251                                 $output .= $this->nextItem( substr( $pref, -1 ) );
1252                                 $paragraphStack = false;
1253
1254                                 if ( substr( $pref, -1 ) == ';') {
1255                                         # The one nasty exception: definition lists work like this:
1256                                         # ; title : definition text
1257                                         # So we check for : in the remainder text to split up the
1258                                         # title and definition, without b0rking links.
1259                                         # FIXME: This is not foolproof. Something better in Tokenizer might help.
1260                                         if( preg_match( '/^(.*?(?:\s|&nbsp;)):(.*)$/', $t, $match ) ) {
1261                                                 $term = $match[1];
1262                                                 $output .= $term . $this->nextItem( ':' );
1263                                                 $t = $match[2];
1264                                         }
1265                                 }
1266                         } elseif( $prefixLength || $lastPrefixLength ) {
1267                                 # Either open or close a level...
1268                                 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1269                                 $paragraphStack = false;
1270
1271                                 while( $commonPrefixLength < $lastPrefixLength ) {
1272                                         $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1273                                         --$lastPrefixLength;
1274                                 }
1275                                 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1276                                         $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1277                                 }
1278                                 while ( $prefixLength > $commonPrefixLength ) {
1279                                         $char = substr( $pref, $commonPrefixLength, 1 );
1280                                         $output .= $this->openList( $char );
1281
1282                                         if ( ';' == $char ) {
1283                                                 # FIXME: This is dupe of code above
1284                                                 if( preg_match( '/^(.*?(?:\s|&nbsp;)):(.*)$/', $t, $match ) ) {
1285                                                         $term = $match[1];
1286                                                         $output .= $term . $this->nextItem( ':' );
1287                                                         $t = $match[2];
1288                                                 }
1289                                         }
1290                                         ++$commonPrefixLength;
1291                                 }
1292                                 $lastPrefix = $pref2;
1293                         }
1294                         if( 0 == $prefixLength ) {
1295                                 # No prefix (not in list)--go to paragraph mode
1296                                 $uniq_prefix = UNIQ_PREFIX;
1297                                 // XXX: use a stack for nestable elements like span, table and div
1298                                 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1299                                 $closematch = preg_match(
1300                                         '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1301                                         '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1302                                 if ( $openmatch or $closematch ) {
1303                                         $paragraphStack = false;
1304                                         $output .= $this->closeParagraph();
1305                                         if($preOpenMatch and !$preCloseMatch) {
1306                                                 $this->mInPre = true;
1307                                         }
1308                                         if ( $closematch ) {
1309                                                 $inBlockElem = false;
1310                                         } else {
1311                                                 $inBlockElem = true;
1312                                         }
1313                                 } else if ( !$inBlockElem && !$this->mInPre ) {
1314                                         if ( ' ' == $t{0} and ( $this->mLastSection == 'pre' or trim($t) != '' ) ) {
1315                                                 // pre
1316                                                 if ($this->mLastSection != 'pre') {
1317                                                         $paragraphStack = false;
1318                                                         $output .= $this->closeParagraph().'<pre>';
1319                                                         $this->mLastSection = 'pre';
1320                                                 }
1321                                         } else {
1322                                                 // paragraph
1323                                                 if ( '' == trim($t) ) {
1324                                                         if ( $paragraphStack ) {
1325                                                                 $output .= $paragraphStack.'<br />';
1326                                                                 $paragraphStack = false;
1327                                                                 $this->mLastSection = 'p';
1328                                                         } else {
1329                                                                 if ($this->mLastSection != 'p' ) {
1330                                                                         $output .= $this->closeParagraph();
1331                                                                         $this->mLastSection = '';
1332                                                                         $paragraphStack = '<p>';
1333                                                                 } else {
1334                                                                         $paragraphStack = '</p><p>';
1335                                                                 }
1336                                                         }
1337                                                 } else {
1338                                                         if ( $paragraphStack ) {
1339                                                                 $output .= $paragraphStack;
1340                                                                 $paragraphStack = false;
1341                                                                 $this->mLastSection = 'p';
1342                                                         } else if ($this->mLastSection != 'p') {
1343                                                                 $output .= $this->closeParagraph().'<p>';
1344                                                                 $this->mLastSection = 'p';
1345                                                         }
1346                                                 }
1347                                         }
1348                                 }
1349                         }
1350                         if ($paragraphStack === false) {
1351                                 $output .= $t."\n";
1352                         }
1353                 }
1354                 while ( $prefixLength ) {
1355                         $output .= $this->closeList( $pref2{$prefixLength-1} );
1356                         --$prefixLength;
1357                 }
1358                 if ( '' != $this->mLastSection ) {
1359                         $output .= '</' . $this->mLastSection . '>';
1360                         $this->mLastSection = '';
1361                 }
1362
1363                 wfProfileOut( $fname );
1364                 return $output;
1365         }
1366
1367         # Return value of a magic variable (like PAGENAME)
1368         function getVariableValue( $index ) {
1369                 global $wgLang, $wgSitename, $wgServer;
1370
1371                 switch ( $index ) {
1372                         case MAG_CURRENTMONTH:
1373                                 return $wgLang->formatNum( date( 'm' ) );
1374                         case MAG_CURRENTMONTHNAME:
1375                                 return $wgLang->getMonthName( date('n') );
1376                         case MAG_CURRENTMONTHNAMEGEN:
1377                                 return $wgLang->getMonthNameGen( date('n') );
1378                         case MAG_CURRENTDAY:
1379                                 return $wgLang->formatNum( date('j') );
1380                         case MAG_PAGENAME:
1381                                 return $this->mTitle->getText();
1382                         case MAG_PAGENAMEE:
1383                                 return $this->mTitle->getPartialURL();
1384                         case MAG_NAMESPACE:
1385                                 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1386                                 return $wgLang->getNsText($this->mTitle->getNamespace()); # Patch by Dori
1387                         case MAG_CURRENTDAYNAME:
1388                                 return $wgLang->getWeekdayName( date('w')+1 );
1389                         case MAG_CURRENTYEAR:
1390                                 return $wgLang->formatNum( date( 'Y' ) );
1391                         case MAG_CURRENTTIME:
1392                                 return $wgLang->time( wfTimestampNow(), false );
1393                         case MAG_NUMBEROFARTICLES:
1394                                 return $wgLang->formatNum( wfNumberOfArticles() );
1395                         case MAG_SITENAME:
1396                                 return $wgSitename;
1397                         case MAG_SERVER:
1398                                 return $wgServer;
1399                         default:
1400                                 return NULL;
1401                 }
1402         }
1403
1404         # initialise the magic variables (like CURRENTMONTHNAME)
1405         function initialiseVariables() {
1406                 global $wgVariableIDs;
1407                 $this->mVariables = array();
1408                 foreach ( $wgVariableIDs as $id ) {
1409                         $mw =& MagicWord::get( $id );
1410                         $mw->addToArray( $this->mVariables, $this->getVariableValue( $id ) );
1411                 }
1412         }
1413
1414         /* private */ function replaceVariables( $text, $args = array() ) {
1415                 global $wgLang, $wgScript, $wgArticlePath;
1416
1417                 # Prevent too big inclusions
1418                 if(strlen($text)> MAX_INCLUDE_SIZE)
1419                 return $text;
1420
1421                 $fname = 'Parser::replaceVariables';
1422                 wfProfileIn( $fname );
1423
1424                 $bail = false;
1425                 $titleChars = Title::legalChars();
1426                 $nonBraceChars = str_replace( array( '{', '}' ), array( '', '' ), $titleChars );
1427
1428                 # This function is called recursively. To keep track of arguments we need a stack:
1429                 array_push( $this->mArgStack, $args );
1430
1431                 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1432                 $GLOBALS['wgCurParser'] =& $this;
1433
1434                 if ( $this->mOutputType == OT_HTML || $this->mOutputType == OT_MSG ) {
1435                         # Variable substitution
1436                         $text = preg_replace_callback( "/{{([$nonBraceChars]*?)}}/", 'wfVariableSubstitution', $text );
1437                 }
1438
1439                 if ( $this->mOutputType == OT_HTML ) {
1440                         # Argument substitution
1441                         $text = preg_replace_callback( "/(\\n?){{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1442                 }
1443                 # Template substitution
1444                 $regex = '/(\\n?){{(['.$nonBraceChars.']*)(\\|.*?|)}}/s';
1445                 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1446
1447                 array_pop( $this->mArgStack );
1448
1449                 wfProfileOut( $fname );
1450                 return $text;
1451         }
1452
1453         function variableSubstitution( $matches ) {
1454                 if ( !$this->mVariables ) {
1455                         $this->initialiseVariables();
1456                 }
1457                 if ( array_key_exists( $matches[1], $this->mVariables ) ) {
1458                         $text = $this->mVariables[$matches[1]];
1459                         $this->mOutput->mContainsOldMagic = true;
1460                 } else {
1461                         $text = $matches[0];
1462                 }
1463                 return $text;
1464         }
1465
1466         # Split template arguments
1467         function getTemplateArgs( $argsString ) {
1468                 if ( $argsString === '' ) {
1469                         return array();
1470                 }
1471
1472                 $args = explode( '|', substr( $argsString, 1 ) );
1473
1474                 # If any of the arguments contains a '[[' but no ']]', it needs to be
1475                 # merged with the next arg because the '|' character between belongs
1476                 # to the link syntax and not the template parameter syntax.
1477                 $argc = count($args);
1478                 $i = 0;
1479                 for ( $i = 0; $i < $argc-1; $i++ ) {
1480                         if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1481                                 $args[$i] .= '|'.$args[$i+1];
1482                                 array_splice($args, $i+1, 1);
1483                                 $i--;
1484                                 $argc--;
1485                         }
1486                 }
1487
1488                 return $args;
1489         }
1490
1491         function braceSubstitution( $matches ) {
1492                 global $wgLinkCache, $wgLang;
1493                 $fname = 'Parser::braceSubstitution';
1494                 $found = false;
1495                 $nowiki = false;
1496                 $noparse = false;
1497
1498                 $title = NULL;
1499
1500                 # $newline is an optional newline character before the braces
1501                 # $part1 is the bit before the first |, and must contain only title characters
1502                 # $args is a list of arguments, starting from index 0, not including $part1
1503
1504                 $newline = $matches[1];
1505                 $part1 = $matches[2];
1506                 # If the third subpattern matched anything, it will start with |
1507
1508                 $args = $this->getTemplateArgs($matches[3]);
1509                 $argc = count( $args );
1510
1511                 # {{{}}}
1512                 if ( strpos( $matches[0], '{{{' ) !== false ) {
1513                         $text = $matches[0];
1514                         $found = true;
1515                         $noparse = true;
1516                 }
1517
1518                 # SUBST
1519                 if ( !$found ) {
1520                         $mwSubst =& MagicWord::get( MAG_SUBST );
1521                         if ( $mwSubst->matchStartAndRemove( $part1 ) ) {
1522                                 if ( $this->mOutputType != OT_WIKI ) {
1523                                         # Invalid SUBST not replaced at PST time
1524                                         # Return without further processing
1525                                         $text = $matches[0];
1526                                         $found = true;
1527                                         $noparse= true;
1528                                 }
1529                         } elseif ( $this->mOutputType == OT_WIKI ) {
1530                                 # SUBST not found in PST pass, do nothing
1531                                 $text = $matches[0];
1532                                 $found = true;
1533                         }
1534                 }
1535
1536                 # MSG, MSGNW and INT
1537                 if ( !$found ) {
1538                         # Check for MSGNW:
1539                         $mwMsgnw =& MagicWord::get( MAG_MSGNW );
1540                         if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1541                                 $nowiki = true;
1542                         } else {
1543                                 # Remove obsolete MSG:
1544                                 $mwMsg =& MagicWord::get( MAG_MSG );
1545                                 $mwMsg->matchStartAndRemove( $part1 );
1546                         }
1547
1548                         # Check if it is an internal message
1549                         $mwInt =& MagicWord::get( MAG_INT );
1550                         if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1551                                 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1552                                         $text = wfMsgReal( $part1, $args, true );
1553                                         $found = true;
1554                                 }
1555                         }
1556                 }
1557
1558                 # NS
1559                 if ( !$found ) {
1560                         # Check for NS: (namespace expansion)
1561                         $mwNs = MagicWord::get( MAG_NS );
1562                         if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1563                                 if ( intval( $part1 ) ) {
1564                                         $text = $wgLang->getNsText( intval( $part1 ) );
1565                                         $found = true;
1566                                 } else {
1567                                         $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1568                                         if ( !is_null( $index ) ) {
1569                                                 $text = $wgLang->getNsText( $index );
1570                                                 $found = true;
1571                                         }
1572                                 }
1573                         }
1574                 }
1575
1576                 # LOCALURL and LOCALURLE
1577                 if ( !$found ) {
1578                         $mwLocal = MagicWord::get( MAG_LOCALURL );
1579                         $mwLocalE = MagicWord::get( MAG_LOCALURLE );
1580
1581                         if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1582                                 $func = 'getLocalURL';
1583                         } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1584                                 $func = 'escapeLocalURL';
1585                         } else {
1586                                 $func = '';
1587                         }
1588
1589                         if ( $func !== '' ) {
1590                                 $title = Title::newFromText( $part1 );
1591                                 if ( !is_null( $title ) ) {
1592                                         if ( $argc > 0 ) {
1593                                                 $text = $title->$func( $args[0] );
1594                                         } else {
1595                                                 $text = $title->$func();
1596                                         }
1597                                         $found = true;
1598                                 }
1599                         }
1600                 }
1601
1602                 # Internal variables
1603                 if ( !$this->mVariables ) {
1604                         $this->initialiseVariables();
1605                 }
1606                 if ( !$found && array_key_exists( $part1, $this->mVariables ) ) {
1607                         $text = $this->mVariables[$part1];
1608                         $found = true;
1609                         $this->mOutput->mContainsOldMagic = true;
1610                 }
1611
1612                 # GRAMMAR
1613                 if ( !$found && $argc == 1 ) {
1614                         $mwGrammar =& MagicWord::get( MAG_GRAMMAR );
1615                         if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1616                                 $text = $wgLang->convertGrammar( $args[0], $part1 );
1617                                 $found = true;
1618                         }
1619                 }
1620
1621                 # Template table test
1622
1623                 # Did we encounter this template already? If yes, it is in the cache
1624                 # and we need to check for loops.
1625                 if ( isset( $this->mTemplates[$part1] ) ) {
1626                         # Infinite loop test
1627                         if ( isset( $this->mTemplatePath[$part1] ) ) {
1628                                 $noparse = true;
1629                                 $found = true;
1630                         }
1631                         # set $text to cached message.
1632                         $text = $this->mTemplates[$part1];
1633                         $found = true;
1634                 }
1635
1636                 # Load from database
1637                 if ( !$found ) {
1638                         $title = Title::newFromText( $part1, NS_TEMPLATE );
1639                         if ( !is_null( $title ) && !$title->isExternal() ) {
1640                                 # Check for excessive inclusion
1641                                 $dbk = $title->getPrefixedDBkey();
1642                                 if ( $this->incrementIncludeCount( $dbk ) ) {
1643                                         # This should never be reached.
1644                                         $article = new Article( $title );
1645                                         $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1646                                         if ( $articleContent !== false ) {
1647                                                 $found = true;
1648                                                 $text = $articleContent;
1649                                         }
1650                                 }
1651
1652                                 # If the title is valid but undisplayable, make a link to it
1653                                 if ( $this->mOutputType == OT_HTML && !$found ) {
1654                                         $text = '[['.$title->getPrefixedText().']]';
1655                                         $found = true;
1656                                 }
1657
1658                                 # Template cache array insertion
1659                                 $this->mTemplates[$part1] = $text;
1660                         }
1661                 }
1662
1663                 # Recursive parsing, escaping and link table handling
1664                 # Only for HTML output
1665                 if ( $nowiki && $found && $this->mOutputType == OT_HTML ) {
1666                         $text = wfEscapeWikiText( $text );
1667                 } elseif ( $this->mOutputType == OT_HTML && $found && !$noparse) {
1668                         # Clean up argument array
1669                         $assocArgs = array();
1670                         $index = 1;
1671                         foreach( $args as $arg ) {
1672                                 $eqpos = strpos( $arg, '=' );
1673                                 if ( $eqpos === false ) {
1674                                         $assocArgs[$index++] = $arg;
1675                                 } else {
1676                                         $name = trim( substr( $arg, 0, $eqpos ) );
1677                                         $value = trim( substr( $arg, $eqpos+1 ) );
1678                                         if ( $value === false ) {
1679                                                 $value = '';
1680                                         }
1681                                         if ( $name !== false ) {
1682                                                 $assocArgs[$name] = $value;
1683                                         }
1684                                 }
1685                         }
1686
1687                         # Do not enter included links in link table
1688                         if ( !is_null( $title ) ) {
1689                                 $wgLinkCache->suspend();
1690                         }
1691
1692                         # Add a new element to the templace recursion path
1693                         $this->mTemplatePath[$part1] = 1;
1694
1695                         $text = $this->stripParse( $text, $newline, $assocArgs );
1696
1697                         # Resume the link cache and register the inclusion as a link
1698                         if ( !is_null( $title ) ) {
1699                                 $wgLinkCache->resume();
1700                                 $wgLinkCache->addLinkObj( $title );
1701                         }
1702                 }
1703
1704                 # Empties the template path
1705                 $this->mTemplatePath = array();
1706
1707                 if ( !$found ) {
1708                         return $matches[0];
1709                 } else {
1710                         return $text;
1711                 }
1712         }
1713
1714         # Triple brace replacement -- used for template arguments
1715         function argSubstitution( $matches ) {
1716                 $newline = $matches[1];
1717                 $arg = trim( $matches[2] );
1718                 $text = $matches[0];
1719                 $inputArgs = end( $this->mArgStack );
1720
1721                 if ( array_key_exists( $arg, $inputArgs ) ) {
1722                         $text = $this->stripParse( $inputArgs[$arg], $newline, array() );
1723                 }
1724
1725                 return $text;
1726         }
1727
1728         # Returns true if the function is allowed to include this entity
1729         function incrementIncludeCount( $dbk ) {
1730                 if ( !array_key_exists( $dbk, $this->mIncludeCount ) ) {
1731                         $this->mIncludeCount[$dbk] = 0;
1732                 }
1733                 if ( ++$this->mIncludeCount[$dbk] <= MAX_INCLUDE_REPEAT ) {
1734                         return true;
1735                 } else {
1736                         return false;
1737                 }
1738         }
1739
1740
1741         # Cleans up HTML, removes dangerous tags and attributes
1742         /* private */ function removeHTMLtags( $text ) {
1743                 global $wgUseTidy, $wgUserHtml;
1744                 $fname = 'Parser::removeHTMLtags';
1745                 wfProfileIn( $fname );
1746
1747                 if( $wgUserHtml ) {
1748                         $htmlpairs = array( # Tags that must be closed
1749                                 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
1750                                 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
1751                                 'strike', 'strong', 'tt', 'var', 'div', 'center',
1752                                 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
1753                                 'ruby', 'rt' , 'rb' , 'rp', 'p'
1754                         );
1755                         $htmlsingle = array(
1756                                 'br', 'hr', 'li', 'dt', 'dd'
1757                         );
1758                         $htmlnest = array( # Tags that can be nested--??
1759                                 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
1760                                 'dl', 'font', 'big', 'small', 'sub', 'sup'
1761                         );
1762                         $tabletags = array( # Can only appear inside table
1763                                 'td', 'th', 'tr'
1764                         );
1765                 } else {
1766                         $htmlpairs = array();
1767                         $htmlsingle = array();
1768                         $htmlnest = array();
1769                         $tabletags = array();
1770                 }
1771
1772                 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1773                 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1774
1775                 $htmlattrs = $this->getHTMLattrs () ;
1776
1777                 # Remove HTML comments
1778                 $text = preg_replace( '/(\\n *<!--.*--> *(?=\\n)|<!--.*-->)/sU', '$2', $text );
1779
1780                 $bits = explode( '<', $text );
1781                 $text = array_shift( $bits );
1782                 if(!$wgUseTidy) {
1783                         $tagstack = array(); $tablestack = array();
1784                         foreach ( $bits as $x ) {
1785                                 $prev = error_reporting( E_ALL & ~( E_NOTICE | E_WARNING ) );
1786                                 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1787                                 $x, $regs );
1788                                 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1789                                 error_reporting( $prev );
1790
1791                                 $badtag = 0 ;
1792                                 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1793                                         # Check our stack
1794                                         if ( $slash ) {
1795                                                 # Closing a tag...
1796                                                 if ( ! in_array( $t, $htmlsingle ) &&
1797                                                 ( $ot = @array_pop( $tagstack ) ) != $t ) {
1798                                                         @array_push( $tagstack, $ot );
1799                                                         $badtag = 1;
1800                                                 } else {
1801                                                         if ( $t == 'table' ) {
1802                                                                 $tagstack = array_pop( $tablestack );
1803                                                         }
1804                                                         $newparams = '';
1805                                                 }
1806                                         } else {
1807                                                 # Keep track for later
1808                                                 if ( in_array( $t, $tabletags ) &&
1809                                                 ! in_array( 'table', $tagstack ) ) {
1810                                                         $badtag = 1;
1811                                                 } else if ( in_array( $t, $tagstack ) &&
1812                                                 ! in_array ( $t , $htmlnest ) ) {
1813                                                         $badtag = 1 ;
1814                                                 } else if ( ! in_array( $t, $htmlsingle ) ) {
1815                                                         if ( $t == 'table' ) {
1816                                                                 array_push( $tablestack, $tagstack );
1817                                                                 $tagstack = array();
1818                                                         }
1819                                                         array_push( $tagstack, $t );
1820                                                 }
1821                                                 # Strip non-approved attributes from the tag
1822                                                 $newparams = $this->fixTagAttributes($params);
1823
1824                                         }
1825                                         if ( ! $badtag ) {
1826                                                 $rest = str_replace( '>', '&gt;', $rest );
1827                                                 $text .= "<$slash$t $newparams$brace$rest";
1828                                                 continue;
1829                                         }
1830                                 }
1831                                 $text .= '&lt;' . str_replace( '>', '&gt;', $x);
1832                         }
1833                         # Close off any remaining tags
1834                         while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
1835                                 $text .= "</$t>\n";
1836                                 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
1837                         }
1838                 } else {
1839                         # this might be possible using tidy itself
1840                         foreach ( $bits as $x ) {
1841                                 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1842                                 $x, $regs );
1843                                 @list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1844                                 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1845                                         $newparams = $this->fixTagAttributes($params);
1846                                         $rest = str_replace( '>', '&gt;', $rest );
1847                                         $text .= "<$slash$t $newparams$brace$rest";
1848                                 } else {
1849                                         $text .= '&lt;' . str_replace( '>', '&gt;', $x);
1850                                 }
1851                         }
1852                 }
1853                 wfProfileOut( $fname );
1854                 return $text;
1855         }
1856
1857
1858         # This function accomplishes several tasks:
1859         # 1) Auto-number headings if that option is enabled
1860         # 2) Add an [edit] link to sections for logged in users who have enabled the option
1861         # 3) Add a Table of contents on the top for users who have enabled the option
1862         # 4) Auto-anchor headings
1863         #
1864         # It loops through all headlines, collects the necessary data, then splits up the
1865         # string and re-inserts the newly formatted headlines.
1866         /* private */ function formatHeadings( $text, $isMain=true ) {
1867                 global $wgInputEncoding, $wgMaxTocLevel, $wgLang;
1868
1869                 $doNumberHeadings = $this->mOptions->getNumberHeadings();
1870                 $doShowToc = $this->mOptions->getShowToc();
1871                 $forceTocHere = false;
1872                 if( !$this->mTitle->userCanEdit() ) {
1873                         $showEditLink = 0;
1874                         $rightClickHack = 0;
1875                 } else {
1876                         $showEditLink = $this->mOptions->getEditSection();
1877                         $rightClickHack = $this->mOptions->getEditSectionOnRightClick();
1878                 }
1879
1880                 # Inhibit editsection links if requested in the page
1881                 $esw =& MagicWord::get( MAG_NOEDITSECTION );
1882                 if( $esw->matchAndRemove( $text ) ) {
1883                         $showEditLink = 0;
1884                 }
1885                 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1886                 # do not add TOC
1887                 $mw =& MagicWord::get( MAG_NOTOC );
1888                 if( $mw->matchAndRemove( $text ) ) {
1889                         $doShowToc = 0;
1890                 }
1891
1892                 # never add the TOC to the Main Page. This is an entry page that should not
1893                 # be more than 1-2 screens large anyway
1894                 if( $this->mTitle->getPrefixedText() == wfMsg('mainpage') ) {
1895                         $doShowToc = 0;
1896                 }
1897
1898                 # Get all headlines for numbering them and adding funky stuff like [edit]
1899                 # links - this is for later, but we need the number of headlines right now
1900                 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
1901
1902                 # if there are fewer than 4 headlines in the article, do not show TOC
1903                 if( $numMatches < 4 ) {
1904                         $doShowToc = 0;
1905                 }
1906
1907                 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
1908                 # override above conditions and always show TOC at that place
1909                 $mw =& MagicWord::get( MAG_TOC );
1910                 if ($mw->match( $text ) ) {
1911                         $doShowToc = 1;
1912                         $forceTocHere = true;
1913                 } else {
1914                         # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
1915                         # override above conditions and always show TOC above first header
1916                         $mw =& MagicWord::get( MAG_FORCETOC );
1917                         if ($mw->matchAndRemove( $text ) ) {
1918                                 $doShowToc = 1;
1919                         }
1920                 }
1921
1922
1923
1924                 # We need this to perform operations on the HTML
1925                 $sk =& $this->mOptions->getSkin();
1926
1927                 # headline counter
1928                 $headlineCount = 0;
1929
1930                 # Ugh .. the TOC should have neat indentation levels which can be
1931                 # passed to the skin functions. These are determined here
1932                 $toclevel = 0;
1933                 $toc = '';
1934                 $full = '';
1935                 $head = array();
1936                 $sublevelCount = array();
1937                 $level = 0;
1938                 $prevlevel = 0;
1939                 foreach( $matches[3] as $headline ) {
1940                         $numbering = '';
1941                         if( $level ) {
1942                                 $prevlevel = $level;
1943                         }
1944                         $level = $matches[1][$headlineCount];
1945                         if( ( $doNumberHeadings || $doShowToc ) && $prevlevel && $level > $prevlevel ) {
1946                                 # reset when we enter a new level
1947                                 $sublevelCount[$level] = 0;
1948                                 $toc .= $sk->tocIndent( $level - $prevlevel );
1949                                 $toclevel += $level - $prevlevel;
1950                         }
1951                         if( ( $doNumberHeadings || $doShowToc ) && $level < $prevlevel ) {
1952                                 # reset when we step back a level
1953                                 $sublevelCount[$level+1]=0;
1954                                 $toc .= $sk->tocUnindent( $prevlevel - $level );
1955                                 $toclevel -= $prevlevel - $level;
1956                         }
1957                         # count number of headlines for each level
1958                         @$sublevelCount[$level]++;
1959                         if( $doNumberHeadings || $doShowToc ) {
1960                                 $dot = 0;
1961                                 for( $i = 1; $i <= $level; $i++ ) {
1962                                         if( !empty( $sublevelCount[$i] ) ) {
1963                                                 if( $dot ) {
1964                                                         $numbering .= '.';
1965                                                 }
1966                                                 $numbering .= $wgLang->formatNum( $sublevelCount[$i] );
1967                                                 $dot = 1;
1968                                         }
1969                                 }
1970                         }
1971
1972                         # The canonized header is a version of the header text safe to use for links
1973                         # Avoid insertion of weird stuff like <math> by expanding the relevant sections
1974                         $canonized_headline = $this->unstrip( $headline, $this->mStripState );
1975                         $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState );
1976
1977                         # Remove link placeholders by the link text.
1978                         #     <!--LINK namespace page_title link text with suffix-->
1979                         # turns into
1980                         #     link text with suffix
1981                         $canonized_headline = preg_replace( '/<!--LINK [0-9]* [^ ]* *(.*?)-->/','$1', $canonized_headline );
1982                         # strip out HTML
1983                         $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
1984                         $tocline = trim( $canonized_headline );
1985                         $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT, $wgInputEncoding ) );
1986                         $replacearray = array(
1987                                 '%3A' => ':',
1988                                 '%' => '.'
1989                         );
1990                         $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
1991                         $refer[$headlineCount] = $canonized_headline;
1992
1993                         # count how many in assoc. array so we can track dupes in anchors
1994                         @$refers[$canonized_headline]++;
1995                         $refcount[$headlineCount]=$refers[$canonized_headline];
1996
1997                         # Prepend the number to the heading text
1998
1999                         if( $doNumberHeadings || $doShowToc ) {
2000                                 $tocline = $numbering . ' ' . $tocline;
2001
2002                                 # Don't number the heading if it is the only one (looks silly)
2003                                 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2004                                         # the two are different if the line contains a link
2005                                         $headline=$numbering . ' ' . $headline;
2006                                 }
2007                         }
2008
2009                         # Create the anchor for linking from the TOC to the section
2010                         $anchor = $canonized_headline;
2011                         if($refcount[$headlineCount] > 1 ) {
2012                                 $anchor .= '_' . $refcount[$headlineCount];
2013                         }
2014                         if( $doShowToc && ( !isset($wgMaxTocLevel) || $toclevel<$wgMaxTocLevel ) ) {
2015                                 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2016                         }
2017                         if( $showEditLink ) {
2018                                 if ( empty( $head[$headlineCount] ) ) {
2019                                         $head[$headlineCount] = '';
2020                                 }
2021                                 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+1);
2022                         }
2023
2024                         # Add the edit section span
2025                         if( $rightClickHack ) {
2026                                 $headline = $sk->editSectionScript($headlineCount+1,$headline);
2027                         }
2028
2029                         # give headline the correct <h#> tag
2030                         @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2031
2032                         $headlineCount++;
2033                 }
2034
2035                 if( $doShowToc ) {
2036                         $toclines = $headlineCount;
2037                         $toc .= $sk->tocUnindent( $toclevel );
2038                         $toc = $sk->tocTable( $toc );
2039                 }
2040
2041                 # split up and insert constructed headlines
2042
2043                 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2044                 $i = 0;
2045
2046                 foreach( $blocks as $block ) {
2047                         if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2048                                 # This is the [edit] link that appears for the top block of text when
2049                                 # section editing is enabled
2050
2051                                 # Disabled because it broke block formatting
2052                                 # For example, a bullet point in the top line
2053                                 # $full .= $sk->editSectionLink(0);
2054                         }
2055                         $full .= $block;
2056                         if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2057                         # Top anchor now in skin
2058                                 $full = $full.$toc;
2059                         }
2060
2061                         if( !empty( $head[$i] ) ) {
2062                                 $full .= $head[$i];
2063                         }
2064                         $i++;
2065                 }
2066                 if($forceTocHere) {
2067                         $mw =& MagicWord::get( MAG_TOC );
2068                         return $mw->replace( $toc, $full );
2069                 } else {
2070                         return $full;
2071                 }
2072         }
2073
2074         # Return an HTML link for the "ISBN 123456" text
2075         /* private */ function magicISBN( $text ) {
2076                 global $wgLang;
2077                 $fname = 'Parser::magicISBN';
2078                 wfProfileIn( $fname );
2079
2080                 $a = split( 'ISBN ', ' '.$text );
2081                 if ( count ( $a ) < 2 ) {
2082                         wfProfileOut( $fname );
2083                         return $text;
2084                 }
2085                 $text = substr( array_shift( $a ), 1);
2086                 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2087
2088                 foreach ( $a as $x ) {
2089                         $isbn = $blank = '' ;
2090                         while ( ' ' == $x{0} ) {
2091                                 $blank .= ' ';
2092                                 $x = substr( $x, 1 );
2093                         }
2094                         while ( strstr( $valid, $x{0} ) != false ) {
2095                                 $isbn .= $x{0};
2096                                 $x = substr( $x, 1 );
2097                         }
2098                         $num = str_replace( '-', '', $isbn );
2099                         $num = str_replace( ' ', '', $num );
2100
2101                         if ( '' == $num ) {
2102                                 $text .= "ISBN $blank$x";
2103                         } else {
2104                                 $titleObj = Title::makeTitle( NS_SPECIAL, 'Booksources' );
2105                                 $text .= '<a href="' .
2106                                 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2107                                         "\" class=\"internal\">ISBN $isbn</a>";
2108                                 $text .= $x;
2109                         }
2110                 }
2111                 wfProfileOut( $fname );
2112                 return $text;
2113         }
2114
2115         # Return an HTML link for the "GEO ..." text
2116         /* private */ function magicGEO( $text ) {
2117                 global $wgLang, $wgUseGeoMode;
2118                 $fname = 'Parser::magicGEO';
2119                 wfProfileIn( $fname );
2120
2121                 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2122                 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2123                 $text = preg_replace ( "/(\d+)&deg;(\d+)'(\d+)\" {$directions['N']}, (\d+)&deg;(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2124                 $text = preg_replace ( "/(\d+)&deg;(\d+)'(\d+)\" {$directions['N']}, (\d+)&deg;(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2125                 $text = preg_replace ( "/(\d+)&deg;(\d+)'(\d+)\" {$directions['S']}, (\d+)&deg;(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2126                 $text = preg_replace ( "/(\d+)&deg;(\d+)'(\d+)\" {$directions['S']}, (\d+)&deg;(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2127
2128                 $a = split( 'GEO ', ' '.$text );
2129                 if ( count ( $a ) < 2 ) {
2130                         wfProfileOut( $fname );
2131                         return $text;
2132                 }
2133                 $text = substr( array_shift( $a ), 1);
2134                 $valid = '0123456789.+-:';
2135
2136                 foreach ( $a as $x ) {
2137                         $geo = $blank = '' ;
2138                         while ( ' ' == $x{0} ) {
2139                                 $blank .= ' ';
2140                                 $x = substr( $x, 1 );
2141                         }
2142                         while ( strstr( $valid, $x{0} ) != false ) {
2143                                 $geo .= $x{0};
2144                                 $x = substr( $x, 1 );
2145                         }
2146                         $num = str_replace( '+', '', $geo );
2147                         $num = str_replace( ' ', '', $num );
2148
2149                         if ( '' == $num || count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2150                                 $text .= "GEO $blank$x";
2151                         } else {
2152                                 $titleObj = Title::makeTitle( NS_SPECIAL, 'Geo' );
2153                                 $text .= '<a href="' .
2154                                 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2155                                         "\" class=\"internal\">GEO $geo</a>";
2156                                 $text .= $x;
2157                         }
2158                 }
2159                 wfProfileOut( $fname );
2160                 return $text;
2161         }
2162
2163         # Return an HTML link for the "RFC 1234" text
2164         /* private */ function magicRFC( $text ) {
2165                 global $wgLang;
2166
2167                 $a = split( 'RFC ', ' '.$text );
2168                 if ( count ( $a ) < 2 ) return $text;
2169                 $text = substr( array_shift( $a ), 1);
2170                 $valid = '0123456789';
2171
2172                 foreach ( $a as $x ) {
2173                         $rfc = $blank = '' ;
2174                         while ( ' ' == $x{0} ) {
2175                                 $blank .= ' ';
2176                                 $x = substr( $x, 1 );
2177                         }
2178                         while ( strstr( $valid, $x{0} ) != false ) {
2179                                 $rfc .= $x{0};
2180                                 $x = substr( $x, 1 );
2181                         }
2182
2183                         if ( '' == $rfc ) {
2184                                 $text .= "RFC $blank$x";
2185                         } else {
2186                                 $url = wfmsg( 'rfcurl' );
2187                                 $url = str_replace( '$1', $rfc, $url);
2188                                 $sk =& $this->mOptions->getSkin();
2189                                 $la = $sk->getExternalLinkAttributes( $url, 'RFC '.$rfc );
2190                                 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2191                         }
2192                 }
2193                 return $text;
2194         }
2195
2196         function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2197                 $this->mOptions = $options;
2198                 $this->mTitle =& $title;
2199                 $this->mOutputType = OT_WIKI;
2200
2201                 if ( $clearState ) {
2202                         $this->clearState();
2203                 }
2204
2205                 $stripState = false;
2206                 $pairs = array(
2207                         "\r\n" => "\n",
2208                         );
2209                 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2210                 // now with regexes
2211                 /*
2212                 $pairs = array(
2213                         "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2214                         "/<br *?>/i" => "<br />",
2215                 );
2216                 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2217                 */
2218                 $text = $this->strip( $text, $stripState, false );
2219                 $text = $this->pstPass2( $text, $user );
2220                 $text = $this->unstrip( $text, $stripState );
2221                 $text = $this->unstripNoWiki( $text, $stripState );
2222                 return $text;
2223         }
2224
2225         /* private */ function pstPass2( $text, &$user ) {
2226                 global $wgLang, $wgLocaltimezone, $wgCurParser;
2227
2228                 # Variable replacement
2229                 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2230                 $text = $this->replaceVariables( $text );
2231
2232                 # Signatures
2233                 #
2234                 $n = $user->getName();
2235                 $k = $user->getOption( 'nickname' );
2236                 if ( '' == $k ) { $k = $n; }
2237                 if(isset($wgLocaltimezone)) {
2238                         $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2239                 }
2240                 /* Note: this is an ugly timezone hack for the European wikis */
2241                 $d = $wgLang->timeanddate( date( 'YmdHis' ), false ) .
2242                   ' (' . date( 'T' ) . ')';
2243                 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2244
2245                 $text = preg_replace( '/~~~~~/', $d, $text );
2246                 $text = preg_replace( '/~~~~/', '[[' . $wgLang->getNsText( NS_USER ) . ":$n|$k]] $d", $text );
2247                 $text = preg_replace( '/~~~/', '[[' . $wgLang->getNsText( NS_USER ) . ":$n|$k]]", $text );
2248
2249                 # Context links: [[|name]] and [[name (context)|]]
2250                 #
2251                 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2252                 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2253                 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2254                 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2255
2256                 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/";             # [[page (context)|]]
2257                 $p2 = "/\[\[\\|({$tc}+)]]/";                                    # [[|page]]
2258                 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/";                # [[namespace:page|]] and [[:namespace:page|]]
2259                 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2260                 $context = '';
2261                 $t = $this->mTitle->getText();
2262                 if ( preg_match( $conpat, $t, $m ) ) {
2263                         $context = $m[2];
2264                 }
2265                 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2266                 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2267                 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2268
2269                 if ( '' == $context ) {
2270                         $text = preg_replace( $p2, '[[\\1]]', $text );
2271                 } else {
2272                         $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2273                 }
2274
2275                 /*
2276                 $mw =& MagicWord::get( MAG_SUBST );
2277                 $wgCurParser = $this->fork();
2278                 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
2279                 $this->merge( $wgCurParser );
2280                 */
2281
2282                 # Trim trailing whitespace
2283                 # MAG_END (__END__) tag allows for trailing
2284                 # whitespace to be deliberately included
2285                 $text = rtrim( $text );
2286                 $mw =& MagicWord::get( MAG_END );
2287                 $mw->matchAndRemove( $text );
2288
2289                 return $text;
2290         }
2291
2292         # Set up some variables which are usually set up in parse()
2293         # so that an external function can call some class members with confidence
2294         function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2295                 $this->mTitle =& $title;
2296                 $this->mOptions = $options;
2297                 $this->mOutputType = $outputType;
2298                 if ( $clearState ) {
2299                         $this->clearState();
2300                 }
2301         }
2302
2303         function transformMsg( $text, $options ) {
2304                 global $wgTitle;
2305                 static $executing = false;
2306
2307                 # Guard against infinite recursion
2308                 if ( $executing ) {
2309                         return $text;
2310                 }
2311                 $executing = true;
2312
2313                 $this->mTitle = $wgTitle;
2314                 $this->mOptions = $options;
2315                 $this->mOutputType = OT_MSG;
2316                 $this->clearState();
2317                 $text = $this->replaceVariables( $text );
2318
2319                 $executing = false;
2320                 return $text;
2321         }
2322
2323         # Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2324         # Callback will be called with the text within
2325         # Transform and return the text within
2326         function setHook( $tag, $callback ) {
2327                 $oldVal = @$this->mTagHooks[$tag];
2328                 $this->mTagHooks[$tag] = $callback;
2329                 return $oldVal;
2330         }
2331 }
2332
2333 /**
2334  * @todo document
2335  * @package MediaWiki
2336  */
2337 class ParserOutput
2338 {
2339         var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2340         var $mCacheTime; # Used in ParserCache
2341
2342         function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
2343                 $containsOldMagic = false )
2344         {
2345                 $this->mText = $text;
2346                 $this->mLanguageLinks = $languageLinks;
2347                 $this->mCategoryLinks = $categoryLinks;
2348                 $this->mContainsOldMagic = $containsOldMagic;
2349                 $this->mCacheTime = '';
2350         }
2351
2352         function getText() { return $this->mText; }
2353         function getLanguageLinks() { return $this->mLanguageLinks; }
2354         function getCategoryLinks() { return $this->mCategoryLinks; }
2355         function getCacheTime() { return $this->mCacheTime; }
2356         function containsOldMagic() { return $this->mContainsOldMagic; }
2357         function setText( $text ) { return wfSetVar( $this->mText, $text ); }
2358         function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks, $ll ); }
2359         function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks, $cl ); }
2360         function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic, $com ); }
2361         function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime, $t ); }
2362
2363         function merge( $other ) {
2364                 $this->mLanguageLinks = array_merge( $this->mLanguageLinks, $other->mLanguageLinks );
2365                 $this->mCategoryLinks = array_merge( $this->mCategoryLinks, $this->mLanguageLinks );
2366                 $this->mContainsOldMagic = $this->mContainsOldMagic || $other->mContainsOldMagic;
2367         }
2368
2369 }
2370
2371 /**
2372  * Set options of the Parser
2373  * @todo document
2374  * @package MediaWiki
2375  */
2376 class ParserOptions
2377 {
2378         # All variables are private
2379         var $mUseTeX;                    # Use texvc to expand <math> tags
2380         var $mUseDynamicDates;           # Use $wgDateFormatter to format dates
2381         var $mInterwikiMagic;            # Interlanguage links are removed and returned in an array
2382         var $mAllowExternalImages;       # Allow external images inline
2383         var $mSkin;                      # Reference to the preferred skin
2384         var $mDateFormat;                # Date format index
2385         var $mEditSection;               # Create "edit section" links
2386         var $mEditSectionOnRightClick;   # Generate JavaScript to edit section on right click
2387         var $mNumberHeadings;            # Automatically number headings
2388         var $mShowToc;                   # Show table of contents
2389
2390         function getUseTeX()                        { return $this->mUseTeX; }
2391         function getUseDynamicDates()               { return $this->mUseDynamicDates; }
2392         function getInterwikiMagic()                { return $this->mInterwikiMagic; }
2393         function getAllowExternalImages()           { return $this->mAllowExternalImages; }
2394         function getSkin()                          { return $this->mSkin; }
2395         function getDateFormat()                    { return $this->mDateFormat; }
2396         function getEditSection()                   { return $this->mEditSection; }
2397         function getEditSectionOnRightClick()       { return $this->mEditSectionOnRightClick; }
2398         function getNumberHeadings()                { return $this->mNumberHeadings; }
2399         function getShowToc()                       { return $this->mShowToc; }
2400
2401         function setUseTeX( $x )                    { return wfSetVar( $this->mUseTeX, $x ); }
2402         function setUseDynamicDates( $x )           { return wfSetVar( $this->mUseDynamicDates, $x ); }
2403         function setInterwikiMagic( $x )            { return wfSetVar( $this->mInterwikiMagic, $x ); }
2404         function setAllowExternalImages( $x )       { return wfSetVar( $this->mAllowExternalImages, $x ); }
2405         function setDateFormat( $x )                { return wfSetVar( $this->mDateFormat, $x ); }
2406         function setEditSection( $x )               { return wfSetVar( $this->mEditSection, $x ); }
2407         function setEditSectionOnRightClick( $x )   { return wfSetVar( $this->mEditSectionOnRightClick, $x ); }
2408         function setNumberHeadings( $x )            { return wfSetVar( $this->mNumberHeadings, $x ); }
2409         function setShowToc( $x )                   { return wfSetVar( $this->mShowToc, $x ); }
2410
2411         function setSkin( &$x ) { $this->mSkin =& $x; }
2412
2413         # Get parser options
2414         /* static */ function newFromUser( &$user ) {
2415                 $popts = new ParserOptions;
2416                 $popts->initialiseFromUser( $user );
2417                 return $popts;
2418         }
2419
2420         # Get user options
2421         function initialiseFromUser( &$userInput ) {
2422                 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2423
2424                 $fname = 'ParserOptions::initialiseFromUser';
2425                 wfProfileIn( $fname );
2426                 if ( !$userInput ) {
2427                         $user = new User;
2428                         $user->setLoaded( true );
2429                 } else {
2430                         $user =& $userInput;
2431                 }
2432
2433                 $this->mUseTeX = $wgUseTeX;
2434                 $this->mUseDynamicDates = $wgUseDynamicDates;
2435                 $this->mInterwikiMagic = $wgInterwikiMagic;
2436                 $this->mAllowExternalImages = $wgAllowExternalImages;
2437                 wfProfileIn( $fname.'-skin' );
2438                 $this->mSkin =& $user->getSkin();
2439                 wfProfileOut( $fname.'-skin' );
2440                 $this->mDateFormat = $user->getOption( 'date' );
2441                 $this->mEditSection = $user->getOption( 'editsection' );
2442                 $this->mEditSectionOnRightClick = $user->getOption( 'editsectiononrightclick' );
2443                 $this->mNumberHeadings = $user->getOption( 'numberheadings' );
2444                 $this->mShowToc = $user->getOption( 'showtoc' );
2445                 wfProfileOut( $fname );
2446         }
2447
2448
2449 }
2450
2451 # Regex callbacks, used in Parser::replaceVariables
2452 function wfBraceSubstitution( $matches ) {
2453         global $wgCurParser;
2454         return $wgCurParser->braceSubstitution( $matches );
2455 }
2456
2457 function wfArgSubstitution( $matches ) {
2458         global $wgCurParser;
2459         return $wgCurParser->argSubstitution( $matches );
2460 }
2461
2462 function wfVariableSubstitution( $matches ) {
2463         global $wgCurParser;
2464         return $wgCurParser->variableSubstitution( $matches );
2465 }
2466
2467 /**
2468  * Return the total number of articles
2469  */
2470 function wfNumberOfArticles() {
2471         global $wgNumberOfArticles;
2472
2473         wfLoadSiteStats();
2474         return $wgNumberOfArticles;
2475 }
2476
2477 /**
2478  * Get various statistics from the database
2479  * @private
2480  */
2481 function wfLoadSiteStats() {
2482         global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
2483         $fname = 'wfLoadSiteStats';
2484
2485         if ( -1 != $wgNumberOfArticles ) return;
2486         $dbr =& wfGetDB( DB_SLAVE );
2487         $s = $dbr->getArray( 'site_stats',
2488                 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
2489                 array( 'ss_row_id' => 1 ), $fname
2490         );
2491
2492         if ( $s === false ) {
2493                 return;
2494         } else {
2495                 $wgTotalViews = $s->ss_total_views;
2496                 $wgTotalEdits = $s->ss_total_edits;
2497                 $wgNumberOfArticles = $s->ss_good_articles;
2498         }
2499 }
2500
2501 function wfEscapeHTMLTagsOnly( $in ) {
2502         return str_replace(
2503                 array( '"', '>', '<' ),
2504                 array( '&quot;', '&gt;', '&lt;' ),
2505                 $in );
2506 }
2507
2508
2509 ?>