includes/Parser.php

   1 <?php
   2 /**
   3  * File for Parser and related classes
   4  *
   5  * @package MediaWiki
   6  * @subpackage Parser
   7  */
   8
   9 /** */
  10 require_once( 'Sanitizer.php' );
  11 require_once( 'HttpFunctions.php' );
  12
  13 /**
  14  * Update this version number when the ParserOutput format
  15  * changes in an incompatible way, so the parser cache
  16  * can automatically discard old data.
  17  */
  18 define( 'MW_PARSER_VERSION', '1.5.0' );
  19
  20 /**
  21  * Variable substitution O(N^2) attack
  22  *
  23  * Without countermeasures, it would be possible to attack the parser by saving
  24  * a page filled with a large number of inclusions of large pages. The size of
  25  * the generated page would be proportional to the square of the input size.
  26  * Hence, we limit the number of inclusions of any given page, thus bringing any
  27  * attack back to O(N).
  28  */
  29
  30 define( 'MAX_INCLUDE_REPEAT', 100 );
  31 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
  32
  33 define( 'RLH_FOR_UPDATE', 1 );
  34
  35 # Allowed values for $mOutputType
  36 define( 'OT_HTML', 1 );
  37 define( 'OT_WIKI', 2 );
  38 define( 'OT_MSG' , 3 );
  39
  40 # string parameter for extractTags which will cause it
  41 # to strip HTML comments in addition to regular
  42 # <XML>-style tags. This should not be anything we
  43 # may want to use in wikisyntax
  44 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
  45
  46 # prefix for escaping, used in two functions at least
  47 define( 'UNIQ_PREFIX', 'NaodW29');
  48
  49 # Constants needed for external link processing
  50 define( 'HTTP_PROTOCOLS', 'http:\/\/|https:\/\/' );
  51 # Everything except bracket, space, or control characters
  52 define( 'EXT_LINK_URL_CLASS', '[^]<>"\\x00-\\x20\\x7F]' );
  53 # Including space
  54 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
  55 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
  56 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
  57 define( 'EXT_LINK_BRACKETED',  '/\[(\b(' . wfUrlProtocols() . ')'.EXT_LINK_URL_CLASS.'+) *('.EXT_LINK_TEXT_CLASS.'*?)\]/S' );
  58 define( 'EXT_IMAGE_REGEX',
  59         '/^('.HTTP_PROTOCOLS.')'.  # Protocol
  60         '('.EXT_LINK_URL_CLASS.'+)\\/'.  # Hostname and path
  61         '('.EXT_IMAGE_FNAME_CLASS.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS.')$/S' # Filename
  62 );
  63
  64 /**
  65  * PHP Parser
  66  *
  67  * Processes wiki markup
  68  *
  69  * <pre>
  70  * There are three main entry points into the Parser class:
  71  * parse()
  72  *   produces HTML output
  73  * preSaveTransform().
  74  *   produces altered wiki markup.
  75  * transformMsg()
  76  *   performs brace substitution on MediaWiki messages
  77  *
  78  * Globals used:
  79  *    objects:   $wgLang, $wgLinkCache
  80  *
  81  * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
  82  *
  83  * settings:
  84  *  $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
  85  *  $wgNamespacesWithSubpages, $wgAllowExternalImages*,
  86  *  $wgLocaltimezone, $wgAllowSpecialInclusion*
  87  *
  88  *  * only within ParserOptions
  89  * </pre>
  90  *
  91  * @package MediaWiki
  92  */
  93 class Parser
  94 {
  95         /**#@+
  96          * @access private
  97          */
  98         # Persistent:
  99         var $mTagHooks;
 100
 101         # Cleared with clearState():
 102         var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
 103         var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
 104         var $mInterwikiLinkHolders, $mLinkHolders;
 105
 106         # Temporary:
 107         var $mOptions, $mTitle, $mOutputType,
 108             $mTemplates,        // cache of already loaded templates, avoids
 109                                 // multiple SQL queries for the same string
 110             $mTemplatePath;     // stores an unsorted hash of all the templates already loaded
 111                                 // in this path. Used for loop detection.
 112
 113         var $mIWTransData = array();
 114
 115         /**#@-*/
 116
 117         /**
 118          * Constructor
 119          *
 120          * @access public
 121          */
 122         function Parser() {
 123                 $this->mTemplates = array();
 124                 $this->mTemplatePath = array();
 125                 $this->mTagHooks = array();
 126                 $this->clearState();
 127         }
 128
 129         /**
 130          * Clear Parser state
 131          *
 132          * @access private
 133          */
 134         function clearState() {
 135                 $this->mOutput = new ParserOutput;
 136                 $this->mAutonumber = 0;
 137                 $this->mLastSection = '';
 138                 $this->mDTopen = false;
 139                 $this->mVariables = false;
 140                 $this->mIncludeCount = array();
 141                 $this->mStripState = array();
 142                 $this->mArgStack = array();
 143                 $this->mInPre = false;
 144                 $this->mInterwikiLinkHolders = array(
 145                         'texts' => array(),
 146                         'titles' => array()
 147                 );
 148                 $this->mLinkHolders = array(
 149                         'namespaces' => array(),
 150                         'dbkeys' => array(),
 151                         'queries' => array(),
 152                         'texts' => array(),
 153                         'titles' => array()
 154                 );
 155                 $this->mRevisionId = null;
 156         }
 157
 158         /**
 159          * First pass--just handle <nowiki> sections, pass the rest off
 160          * to internalParse() which does all the real work.
 161          *
 162          * @access private
 163          * @param string $text Text we want to parse
 164          * @param Title &$title A title object
 165          * @param array $options
 166          * @param boolean $linestart
 167          * @param boolean $clearState
 168          * @param int $revid number to pass in {{REVISIONID}}
 169          * @return ParserOutput a ParserOutput
 170          */
 171         function parse( $text, &$title, $options, $linestart = true, $clearState = true, $revid = null ) {
 172                 global $wgUseTidy, $wgContLang;
 173                 $fname = 'Parser::parse';
 174                 wfProfileIn( $fname );
 175
 176                 if ( $clearState ) {
 177                         $this->clearState();
 178                 }
 179
 180                 $this->mOptions = $options;
 181                 $this->mTitle =& $title;
 182                 $this->mRevisionId = $revid;
 183                 $this->mOutputType = OT_HTML;
 184
 185                 $this->mStripState = NULL;
 186
 187                 //$text = $this->strip( $text, $this->mStripState );
 188                 // VOODOO MAGIC FIX! Sometimes the above segfaults in PHP5.
 189                 $x =& $this->mStripState;
 190
 191                 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$x ) );
 192                 $text = $this->strip( $text, $x );
 193                 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$x ) );
 194
 195                 $text = $this->internalParse( $text );
 196
 197                 $text = $this->unstrip( $text, $this->mStripState );
 198
 199                 # Clean up special characters, only run once, next-to-last before doBlockLevels
 200                 $fixtags = array(
 201                         # french spaces, last one Guillemet-left
 202                         # only if there is something before the space
 203                         '/(.) (?=\\?|:|;|!|\\302\\273)/' => '\\1&nbsp;\\2',
 204                         # french spaces, Guillemet-right
 205                         '/(\\302\\253) /' => '\\1&nbsp;',
 206                         '/<center *>(.*)<\\/center *>/i' => '<div class="center">\\1</div>',
 207                 );
 208                 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
 209
 210                 # only once and last
 211                 $text = $this->doBlockLevels( $text, $linestart );
 212
 213                 $this->replaceLinkHolders( $text );
 214
 215                 # the position of the convert() call should not be changed. it
 216                 # assumes that the links are all replaces and the only thing left
 217                 # is the <nowiki> mark.
 218                 $text = $wgContLang->convert($text);
 219                 $this->mOutput->setTitleText($wgContLang->getParsedTitle());
 220
 221                 $text = $this->unstripNoWiki( $text, $this->mStripState );
 222
 223                 wfRunHooks( 'ParserBeforeTidy', array( &$this, &$text ) );
 224
 225                 $text = Sanitizer::normalizeCharReferences( $text );
 226
 227                 if ($wgUseTidy) {
 228                         $text = Parser::tidy($text);
 229                 }
 230
 231                 wfRunHooks( 'ParserAfterTidy', array( &$this, &$text ) );
 232
 233                 $this->mOutput->setText( $text );
 234                 wfProfileOut( $fname );
 235                 return $this->mOutput;
 236         }
 237
 238         /**
 239          * Get a random string
 240          *
 241          * @access private
 242          * @static
 243          */
 244         function getRandomString() {
 245                 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
 246         }
 247
 248         /**
 249          * Replaces all occurrences of <$tag>content</$tag> in the text
 250          * with a random marker and returns the new text. the output parameter
 251          * $content will be an associative array filled with data on the form
 252          * $unique_marker => content.
 253          *
 254          * If $content is already set, the additional entries will be appended
 255          * If $tag is set to STRIP_COMMENTS, the function will extract
 256          * <!-- HTML comments -->
 257          *
 258          * @access private
 259          * @static
 260          */
 261         function extractTagsAndParams($tag, $text, &$content, &$tags, &$params, $uniq_prefix = ''){
 262                 $rnd = $uniq_prefix . '-' . $tag . Parser::getRandomString();
 263                 if ( !$content ) {
 264                         $content = array( );
 265                 }
 266                 $n = 1;
 267                 $stripped = '';
 268
 269                 if ( !$tags ) {
 270                         $tags = array( );
 271                 }
 272
 273                 if ( !$params ) {
 274                         $params = array( );
 275                 }
 276
 277                 if( $tag == STRIP_COMMENTS ) {
 278                         $start = '/<!--()()/';
 279                         $end   = '/-->/';
 280                 } else {
 281                         $start = "/<$tag(\\s+[^\\/>]*|\\s*)(\\/?)>/i";
 282                         $end   = "/<\\/$tag\\s*>/i";
 283                 }
 284
 285                 while ( '' != $text ) {
 286                         $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE );
 287                         $stripped .= $p[0];
 288                         if( count( $p ) < 4 ) {
 289                                 break;
 290                         }
 291                         $attributes = $p[1];
 292                         $empty      = $p[2];
 293                         $inside     = $p[3];
 294
 295                         $marker = $rnd . sprintf('%08X', $n++);
 296                         $stripped .= $marker;
 297
 298                         $tags[$marker] = "<$tag$attributes$empty>";
 299                         $params[$marker] = Sanitizer::decodeTagAttributes( $attributes );
 300
 301                         if ( $empty === '/' ) {
 302                                 // Empty element tag, <tag />
 303                                 $content[$marker] = null;
 304                                 $text = $inside;
 305                         } else {
 306                                 $q = preg_split( $end, $inside, 2 );
 307                                 $content[$marker] = $q[0];
 308                                 if( count( $q ) < 2 ) {
 309                                         # No end tag -- let it run out to the end of the text.
 310                                         break;
 311                                 } else {
 312                                         $text = $q[1];
 313                                 }
 314                         }
 315                 }
 316                 return $stripped;
 317         }
 318
 319         /**
 320          * Wrapper function for extractTagsAndParams
 321          * for cases where $tags and $params isn't needed
 322          * i.e. where tags will never have params, like <nowiki>
 323          *
 324          * @access private
 325          * @static
 326          */
 327         function extractTags( $tag, $text, &$content, $uniq_prefix = '' ) {
 328                 $dummy_tags = array();
 329                 $dummy_params = array();
 330
 331                 return Parser::extractTagsAndParams( $tag, $text, $content,
 332                         $dummy_tags, $dummy_params, $uniq_prefix );
 333         }
 334
 335         /**
 336          * Strips and renders nowiki, pre, math, hiero
 337          * If $render is set, performs necessary rendering operations on plugins
 338          * Returns the text, and fills an array with data needed in unstrip()
 339          * If the $state is already a valid strip state, it adds to the state
 340          *
 341          * @param bool $stripcomments when set, HTML comments <!-- like this -->
 342          *  will be stripped in addition to other tags. This is important
 343          *  for section editing, where these comments cause confusion when
 344          *  counting the sections in the wikisource
 345          *
 346          * @access private
 347          */
 348         function strip( $text, &$state, $stripcomments = false ) {
 349                 $render = ($this->mOutputType == OT_HTML);
 350                 $html_content = array();
 351                 $nowiki_content = array();
 352                 $math_content = array();
 353                 $pre_content = array();
 354                 $comment_content = array();
 355                 $ext_content = array();
 356                 $ext_tags = array();
 357                 $ext_params = array();
 358                 $gallery_content = array();
 359
 360                 # Replace any instances of the placeholders
 361                 $uniq_prefix = UNIQ_PREFIX;
 362                 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
 363
 364                 # html
 365                 global $wgRawHtml;
 366                 if( $wgRawHtml ) {
 367                         $text = Parser::extractTags('html', $text, $html_content, $uniq_prefix);
 368                         foreach( $html_content as $marker => $content ) {
 369                                 if ($render ) {
 370                                         # Raw and unchecked for validity.
 371                                         $html_content[$marker] = $content;
 372                                 } else {
 373                                         $html_content[$marker] = '<html>'.$content.'</html>';
 374                                 }
 375                         }
 376                 }
 377
 378                 # nowiki
 379                 $text = Parser::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
 380                 foreach( $nowiki_content as $marker => $content ) {
 381                         if( $render ){
 382                                 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
 383                         } else {
 384                                 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
 385                         }
 386                 }
 387
 388                 # math
 389                 if( $this->mOptions->getUseTeX() ) {
 390                         $text = Parser::extractTags('math', $text, $math_content, $uniq_prefix);
 391                         foreach( $math_content as $marker => $content ){
 392                                 if( $render ) {
 393                                         $math_content[$marker] = renderMath( $content );
 394                                 } else {
 395                                         $math_content[$marker] = '<math>'.$content.'</math>';
 396                                 }
 397                         }
 398                 }
 399
 400                 # pre
 401                 $text = Parser::extractTags('pre', $text, $pre_content, $uniq_prefix);
 402                 foreach( $pre_content as $marker => $content ){
 403                         if( $render ){
 404                                 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
 405                         } else {
 406                                 $pre_content[$marker] = '<pre>'.$content.'</pre>';
 407                         }
 408                 }
 409
 410                 # gallery
 411                 $text = Parser::extractTags('gallery', $text, $gallery_content, $uniq_prefix);
 412                 foreach( $gallery_content as $marker => $content ) {
 413                         require_once( 'ImageGallery.php' );
 414                         if ( $render ) {
 415                                 $gallery_content[$marker] = Parser::renderImageGallery( $content );
 416                         } else {
 417                                 $gallery_content[$marker] = '<gallery>'.$content.'</gallery>';
 418                         }
 419                 }
 420
 421                 # Comments
 422                 if($stripcomments) {
 423                         $text = Parser::extractTags(STRIP_COMMENTS, $text, $comment_content, $uniq_prefix);
 424                         foreach( $comment_content as $marker => $content ){
 425                                 $comment_content[$marker] = '<!--'.$content.'-->';
 426                         }
 427                 }
 428
 429                 # Extensions
 430                 foreach ( $this->mTagHooks as $tag => $callback ) {
 431                         $ext_content[$tag] = array();
 432                         $text = Parser::extractTagsAndParams( $tag, $text, $ext_content[$tag],
 433                                 $ext_tags[$tag], $ext_params[$tag], $uniq_prefix );
 434                         foreach( $ext_content[$tag] as $marker => $content ) {
 435                                 $full_tag = $ext_tags[$tag][$marker];
 436                                 $params = $ext_params[$tag][$marker];
 437                                 if ( $render )
 438                                         $ext_content[$tag][$marker] = call_user_func_array( $callback, array( $content, $params, &$this ) );
 439                                 else {
 440                                         if ( is_null( $content ) ) {
 441                                                 // Empty element tag
 442                                                 $ext_content[$tag][$marker] = $full_tag;
 443                                         } else {
 444                                                 $ext_content[$tag][$marker] = "$full_tag$content</$tag>";
 445                                         }
 446                                 }
 447                         }
 448                 }
 449
 450                 # Merge state with the pre-existing state, if there is one
 451                 if ( $state ) {
 452                         $state['html'] = $state['html'] + $html_content;
 453                         $state['nowiki'] = $state['nowiki'] + $nowiki_content;
 454                         $state['math'] = $state['math'] + $math_content;
 455                         $state['pre'] = $state['pre'] + $pre_content;
 456                         $state['comment'] = $state['comment'] + $comment_content;
 457                         $state['gallery'] = $state['gallery'] + $gallery_content;
 458
 459                         foreach( $ext_content as $tag => $array ) {
 460                                 if ( array_key_exists( $tag, $state ) ) {
 461                                         $state[$tag] = $state[$tag] + $array;
 462                                 }
 463                         }
 464                 } else {
 465                         $state = array(
 466                           'html' => $html_content,
 467                           'nowiki' => $nowiki_content,
 468                           'math' => $math_content,
 469                           'pre' => $pre_content,
 470                           'comment' => $comment_content,
 471                           'gallery' => $gallery_content,
 472                         ) + $ext_content;
 473                 }
 474                 return $text;
 475         }
 476
 477         /**
 478          * restores pre, math, and hiero removed by strip()
 479          *
 480          * always call unstripNoWiki() after this one
 481          * @access private
 482          */
 483         function unstrip( $text, &$state ) {
 484                 if ( !is_array( $state ) ) {
 485                         return $text;
 486                 }
 487
 488                 # Must expand in reverse order, otherwise nested tags will be corrupted
 489                 foreach( array_reverse( $state, true ) as $tag => $contentDict ) {
 490                         if( $tag != 'nowiki' && $tag != 'html' ) {
 491                                 foreach( array_reverse( $contentDict, true ) as $uniq => $content ) {
 492                                         $text = str_replace( $uniq, $content, $text );
 493                                 }
 494                         }
 495                 }
 496
 497                 return $text;
 498         }
 499
 500         /**
 501          * always call this after unstrip() to preserve the order
 502          *
 503          * @access private
 504          */
 505         function unstripNoWiki( $text, &$state ) {
 506                 if ( !is_array( $state ) ) {
 507                         return $text;
 508                 }
 509
 510                 # Must expand in reverse order, otherwise nested tags will be corrupted
 511                 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
 512                         $text = str_replace( key( $state['nowiki'] ), $content, $text );
 513                 }
 514
 515                 global $wgRawHtml;
 516                 if ($wgRawHtml) {
 517                         for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
 518                                 $text = str_replace( key( $state['html'] ), $content, $text );
 519                         }
 520                 }
 521
 522                 return $text;
 523         }
 524
 525         /**
 526          * Add an item to the strip state
 527          * Returns the unique tag which must be inserted into the stripped text
 528          * The tag will be replaced with the original text in unstrip()
 529          *
 530          * @access private
 531          */
 532         function insertStripItem( $text, &$state ) {
 533                 $rnd = UNIQ_PREFIX . '-item' . Parser::getRandomString();
 534                 if ( !$state ) {
 535                         $state = array(
 536                           'html' => array(),
 537                           'nowiki' => array(),
 538                           'math' => array(),
 539                           'pre' => array(),
 540                           'comment' => array(),
 541                           'gallery' => array(),
 542                         );
 543                 }
 544                 $state['item'][$rnd] = $text;
 545                 return $rnd;
 546         }
 547
 548         /**
 549          * Interface with html tidy, used if $wgUseTidy = true.
 550          * If tidy isn't able to correct the markup, the original will be
 551          * returned in all its glory with a warning comment appended.
 552          *
 553          * Either the external tidy program or the in-process tidy extension
 554          * will be used depending on availability. Override the default
 555          * $wgTidyInternal setting to disable the internal if it's not working.
 556          *
 557          * @param string $text Hideous HTML input
 558          * @return string Corrected HTML output
 559          * @access public
 560          * @static
 561          */
 562         function tidy( $text ) {
 563                 global $wgTidyInternal;
 564                 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
 565 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
 566 '<head><title>test</title></head><body>'.$text.'</body></html>';
 567                 if( $wgTidyInternal ) {
 568                         $correctedtext = Parser::internalTidy( $wrappedtext );
 569                 } else {
 570                         $correctedtext = Parser::externalTidy( $wrappedtext );
 571                 }
 572                 if( is_null( $correctedtext ) ) {
 573                         wfDebug( "Tidy error detected!\n" );
 574                         return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
 575                 }
 576                 return $correctedtext;
 577         }
 578
 579         /**
 580          * Spawn an external HTML tidy process and get corrected markup back from it.
 581          *
 582          * @access private
 583          * @static
 584          */
 585         function externalTidy( $text ) {
 586                 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
 587                 $fname = 'Parser::externalTidy';
 588                 wfProfileIn( $fname );
 589
 590                 $cleansource = '';
 591                 $opts = ' -utf8';
 592
 593                 $descriptorspec = array(
 594                         0 => array('pipe', 'r'),
 595                         1 => array('pipe', 'w'),
 596                         2 => array('file', '/dev/null', 'a')
 597                 );
 598                 $pipes = array();
 599                 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
 600                 if (is_resource($process)) {
 601                         fwrite($pipes[0], $text);
 602                         fclose($pipes[0]);
 603                         while (!feof($pipes[1])) {
 604                                 $cleansource .= fgets($pipes[1], 1024);
 605                         }
 606                         fclose($pipes[1]);
 607                         proc_close($process);
 608                 }
 609
 610                 wfProfileOut( $fname );
 611
 612                 if( $cleansource == '' && $text != '') {
 613                         // Some kind of error happened, so we couldn't get the corrected text.
 614                         // Just give up; we'll use the source text and append a warning.
 615                         return null;
 616                 } else {
 617                         return $cleansource;
 618                 }
 619         }
 620
 621         /**
 622          * Use the HTML tidy PECL extension to use the tidy library in-process,
 623          * saving the overhead of spawning a new process. Currently written to
 624          * the PHP 4.3.x version of the extension, may not work on PHP 5.
 625          *
 626          * 'pear install tidy' should be able to compile the extension module.
 627          *
 628          * @access private
 629          * @static
 630          */
 631         function internalTidy( $text ) {
 632                 global $wgTidyConf;
 633                 $fname = 'Parser::internalTidy';
 634                 wfProfileIn( $fname );
 635
 636                 tidy_load_config( $wgTidyConf );
 637                 tidy_set_encoding( 'utf8' );
 638                 tidy_parse_string( $text );
 639                 tidy_clean_repair();
 640                 if( tidy_get_status() == 2 ) {
 641                         // 2 is magic number for fatal error
 642                         // http://www.php.net/manual/en/function.tidy-get-status.php
 643                         $cleansource = null;
 644                 } else {
 645                         $cleansource = tidy_get_output();
 646                 }
 647                 wfProfileOut( $fname );
 648                 return $cleansource;
 649         }
 650
 651         /**
 652          * parse the wiki syntax used to render tables
 653          *
 654          * @access private
 655          */
 656         function doTableStuff ( $t ) {
 657                 $fname = 'Parser::doTableStuff';
 658                 wfProfileIn( $fname );
 659
 660                 $t = explode ( "\n" , $t ) ;
 661                 $td = array () ; # Is currently a td tag open?
 662                 $ltd = array () ; # Was it TD or TH?
 663                 $tr = array () ; # Is currently a tr tag open?
 664                 $ltr = array () ; # tr attributes
 665                 $indent_level = 0; # indent level of the table
 666                 foreach ( $t AS $k => $x )
 667                 {
 668                         $x = trim ( $x ) ;
 669                         $fc = substr ( $x , 0 , 1 ) ;
 670                         if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
 671                                 $indent_level = strlen( $matches[1] );
 672
 673                                 $attributes = $this->unstripForHTML( $matches[2] );
 674
 675                                 $t[$k] = str_repeat( '<dl><dd>', $indent_level ) .
 676                                         '<table' . Sanitizer::fixTagAttributes ( $attributes, 'table' ) . '>' ;
 677                                 array_push ( $td , false ) ;
 678                                 array_push ( $ltd , '' ) ;
 679                                 array_push ( $tr , false ) ;
 680                                 array_push ( $ltr , '' ) ;
 681                         }
 682                         else if ( count ( $td ) == 0 ) { } # Don't do any of the following
 683                         else if ( '|}' == substr ( $x , 0 , 2 ) ) {
 684                                 $z = "</table>" . substr ( $x , 2);
 685                                 $l = array_pop ( $ltd ) ;
 686                                 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 687                                 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 688                                 array_pop ( $ltr ) ;
 689                                 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
 690                         }
 691                         else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
 692                                 $x = substr ( $x , 1 ) ;
 693                                 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
 694                                 $z = '' ;
 695                                 $l = array_pop ( $ltd ) ;
 696                                 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 697                                 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 698                                 array_pop ( $ltr ) ;
 699                                 $t[$k] = $z ;
 700                                 array_push ( $tr , false ) ;
 701                                 array_push ( $td , false ) ;
 702                                 array_push ( $ltd , '' ) ;
 703                                 $attributes = $this->unstripForHTML( $x );
 704                                 array_push ( $ltr , Sanitizer::fixTagAttributes ( $attributes, 'tr' ) ) ;
 705                         }
 706                         else if ( '|' == $fc || '!' == $fc || '|+' == substr ( $x , 0 , 2 ) ) { # Caption
 707                                 # $x is a table row
 708                                 if ( '|+' == substr ( $x , 0 , 2 ) ) {
 709                                         $fc = '+' ;
 710                                         $x = substr ( $x , 1 ) ;
 711                                 }
 712                                 $after = substr ( $x , 1 ) ;
 713                                 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
 714                                 $after = explode ( '||' , $after ) ;
 715                                 $t[$k] = '' ;
 716
 717                                 # Loop through each table cell
 718                                 foreach ( $after AS $theline )
 719                                 {
 720                                         $z = '' ;
 721                                         if ( $fc != '+' )
 722                                         {
 723                                                 $tra = array_pop ( $ltr ) ;
 724                                                 if ( !array_pop ( $tr ) ) $z = '<tr'.$tra.">\n" ;
 725                                                 array_push ( $tr , true ) ;
 726                                                 array_push ( $ltr , '' ) ;
 727                                         }
 728
 729                                         $l = array_pop ( $ltd ) ;
 730                                         if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 731                                         if ( $fc == '|' ) $l = 'td' ;
 732                                         else if ( $fc == '!' ) $l = 'th' ;
 733                                         else if ( $fc == '+' ) $l = 'caption' ;
 734                                         else $l = '' ;
 735                                         array_push ( $ltd , $l ) ;
 736
 737                                         # Cell parameters
 738                                         $y = explode ( '|' , $theline , 2 ) ;
 739                                         # Note that a '|' inside an invalid link should not
 740                                         # be mistaken as delimiting cell parameters
 741                                         if ( strpos( $y[0], '[[' ) !== false ) {
 742                                                 $y = array ($theline);
 743                                         }
 744                                         if ( count ( $y ) == 1 )
 745                                                 $y = "{$z}<{$l}>{$y[0]}" ;
 746                                         else {
 747                                                 $attributes = $this->unstripForHTML( $y[0] );
 748                                                 $y = "{$z}<{$l}".Sanitizer::fixTagAttributes($attributes, $l).">{$y[1]}" ;
 749                                         }
 750                                         $t[$k] .= $y ;
 751                                         array_push ( $td , true ) ;
 752                                 }
 753                         }
 754                 }
 755
 756                 # Closing open td, tr && table
 757                 while ( count ( $td ) > 0 )
 758                 {
 759                         if ( array_pop ( $td ) ) $t[] = '</td>' ;
 760                         if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
 761                         $t[] = '</table>' ;
 762                 }
 763
 764                 $t = implode ( "\n" , $t ) ;
 765                 wfProfileOut( $fname );
 766                 return $t ;
 767         }
 768
 769         /**
 770          * Helper function for parse() that transforms wiki markup into
 771          * HTML. Only called for $mOutputType == OT_HTML.
 772          *
 773          * @access private
 774          */
 775         function internalParse( $text ) {
 776                 global $wgContLang;
 777                 $args = array();
 778                 $isMain = true;
 779                 $fname = 'Parser::internalParse';
 780                 wfProfileIn( $fname );
 781
 782                 # Remove <noinclude> tags and <includeonly> sections
 783                 $text = strtr( $text, array( '<onlyinclude>' => '' , '</onlyinclude>' => '' ) );
 784                 $text = strtr( $text, array( '<noinclude>' => '', '</noinclude>' => '') );
 785                 $text = preg_replace( '/<includeonly>.*?<\/includeonly>/s', '', $text );
 786
 787                 $text = Sanitizer::removeHTMLtags( $text, array( &$this, 'attributeStripCallback' ) );
 788                 $text = $this->replaceVariables( $text, $args );
 789
 790                 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
 791
 792                 $text = $this->doHeadings( $text );
 793                 if($this->mOptions->getUseDynamicDates()) {
 794                         $df =& DateFormatter::getInstance();
 795                         $text = $df->reformat( $this->mOptions->getDateFormat(), $text );
 796                 }
 797                 $text = $this->doAllQuotes( $text );
 798                 $text = $this->replaceInternalLinks( $text );
 799                 $text = $this->replaceExternalLinks( $text );
 800
 801                 # replaceInternalLinks may sometimes leave behind
 802                 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
 803                 $text = str_replace(UNIQ_PREFIX."NOPARSE", "", $text);
 804
 805                 $text = $this->doMagicLinks( $text );
 806                 $text = $this->doTableStuff( $text );
 807                 $text = $this->formatHeadings( $text, $isMain );
 808
 809                 $regex = '/<!--IW_TRANSCLUDE (\d+)-->/';
 810                 $text = preg_replace_callback($regex, array(&$this, 'scarySubstitution'), $text);
 811
 812                 wfProfileOut( $fname );
 813                 return $text;
 814         }
 815
 816         function scarySubstitution($matches) {
 817 #               return "[[".$matches[0]."]]";
 818                 return $this->mIWTransData[(int)$matches[0]];
 819         }
 820
 821         /**
 822          * Replace special strings like "ISBN xxx" and "RFC xxx" with
 823          * magic external links.
 824          *
 825          * @access private
 826          */
 827         function &doMagicLinks( &$text ) {
 828                 $text = $this->magicISBN( $text );
 829                 $text = $this->magicRFC( $text, 'RFC ', 'rfcurl' );
 830                 $text = $this->magicRFC( $text, 'PMID ', 'pubmedurl' );
 831                 return $text;
 832         }
 833
 834         /**
 835          * Parse headers and return html
 836          *
 837          * @access private
 838          */
 839         function doHeadings( $text ) {
 840                 $fname = 'Parser::doHeadings';
 841                 wfProfileIn( $fname );
 842                 for ( $i = 6; $i >= 1; --$i ) {
 843                         $h = str_repeat( '=', $i );
 844                         $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
 845                           "<h{$i}>\\1</h{$i}>\\2", $text );
 846                 }
 847                 wfProfileOut( $fname );
 848                 return $text;
 849         }
 850
 851         /**
 852          * Replace single quotes with HTML markup
 853          * @access private
 854          * @return string the altered text
 855          */
 856         function doAllQuotes( $text ) {
 857                 $fname = 'Parser::doAllQuotes';
 858                 wfProfileIn( $fname );
 859                 $outtext = '';
 860                 $lines = explode( "\n", $text );
 861                 foreach ( $lines as $line ) {
 862                         $outtext .= $this->doQuotes ( $line ) . "\n";
 863                 }
 864                 $outtext = substr($outtext, 0,-1);
 865                 wfProfileOut( $fname );
 866                 return $outtext;
 867         }
 868
 869         /**
 870          * Helper function for doAllQuotes()
 871          * @access private
 872          */
 873         function doQuotes( $text ) {
 874                 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE );
 875                 if ( count( $arr ) == 1 )
 876                         return $text;
 877                 else
 878                 {
 879                         # First, do some preliminary work. This may shift some apostrophes from
 880                         # being mark-up to being text. It also counts the number of occurrences
 881                         # of bold and italics mark-ups.
 882                         $i = 0;
 883                         $numbold = 0;
 884                         $numitalics = 0;
 885                         foreach ( $arr as $r )
 886                         {
 887                                 if ( ( $i % 2 ) == 1 )
 888                                 {
 889                                         # If there are ever four apostrophes, assume the first is supposed to
 890                                         # be text, and the remaining three constitute mark-up for bold text.
 891                                         if ( strlen( $arr[$i] ) == 4 )
 892                                         {
 893                                                 $arr[$i-1] .= "'";
 894                                                 $arr[$i] = "'''";
 895                                         }
 896                                         # If there are more than 5 apostrophes in a row, assume they're all
 897                                         # text except for the last 5.
 898                                         else if ( strlen( $arr[$i] ) > 5 )
 899                                         {
 900                                                 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
 901                                                 $arr[$i] = "'''''";
 902                                         }
 903                                         # Count the number of occurrences of bold and italics mark-ups.
 904                                         # We are not counting sequences of five apostrophes.
 905                                         if ( strlen( $arr[$i] ) == 2 ) $numitalics++;  else
 906                                         if ( strlen( $arr[$i] ) == 3 ) $numbold++;     else
 907                                         if ( strlen( $arr[$i] ) == 5 ) { $numitalics++; $numbold++; }
 908                                 }
 909                                 $i++;
 910                         }
 911
 912                         # If there is an odd number of both bold and italics, it is likely
 913                         # that one of the bold ones was meant to be an apostrophe followed
 914                         # by italics. Which one we cannot know for certain, but it is more
 915                         # likely to be one that has a single-letter word before it.
 916                         if ( ( $numbold % 2 == 1 ) && ( $numitalics % 2 == 1 ) )
 917                         {
 918                                 $i = 0;
 919                                 $firstsingleletterword = -1;
 920                                 $firstmultiletterword = -1;
 921                                 $firstspace = -1;
 922                                 foreach ( $arr as $r )
 923                                 {
 924                                         if ( ( $i % 2 == 1 ) and ( strlen( $r ) == 3 ) )
 925                                         {
 926                                                 $x1 = substr ($arr[$i-1], -1);
 927                                                 $x2 = substr ($arr[$i-1], -2, 1);
 928                                                 if ($x1 == ' ') {
 929                                                         if ($firstspace == -1) $firstspace = $i;
 930                                                 } else if ($x2 == ' ') {
 931                                                         if ($firstsingleletterword == -1) $firstsingleletterword = $i;
 932                                                 } else {
 933                                                         if ($firstmultiletterword == -1) $firstmultiletterword = $i;
 934                                                 }
 935                                         }
 936                                         $i++;
 937                                 }
 938
 939                                 # If there is a single-letter word, use it!
 940                                 if ($firstsingleletterword > -1)
 941                                 {
 942                                         $arr [ $firstsingleletterword ] = "''";
 943                                         $arr [ $firstsingleletterword-1 ] .= "'";
 944                                 }
 945                                 # If not, but there's a multi-letter word, use that one.
 946                                 else if ($firstmultiletterword > -1)
 947                                 {
 948                                         $arr [ $firstmultiletterword ] = "''";
 949                                         $arr [ $firstmultiletterword-1 ] .= "'";
 950                                 }
 951                                 # ... otherwise use the first one that has neither.
 952                                 # (notice that it is possible for all three to be -1 if, for example,
 953                                 # there is only one pentuple-apostrophe in the line)
 954                                 else if ($firstspace > -1)
 955                                 {
 956                                         $arr [ $firstspace ] = "''";
 957                                         $arr [ $firstspace-1 ] .= "'";
 958                                 }
 959                         }
 960
 961                         # Now let's actually convert our apostrophic mush to HTML!
 962                         $output = '';
 963                         $buffer = '';
 964                         $state = '';
 965                         $i = 0;
 966                         foreach ($arr as $r)
 967                         {
 968                                 if (($i % 2) == 0)
 969                                 {
 970                                         if ($state == 'both')
 971                                                 $buffer .= $r;
 972                                         else
 973                                                 $output .= $r;
 974                                 }
 975                                 else
 976                                 {
 977                                         if (strlen ($r) == 2)
 978                                         {
 979                                                 if ($state == 'i')
 980                                                 { $output .= '</i>'; $state = ''; }
 981                                                 else if ($state == 'bi')
 982                                                 { $output .= '</i>'; $state = 'b'; }
 983                                                 else if ($state == 'ib')
 984                                                 { $output .= '</b></i><b>'; $state = 'b'; }
 985                                                 else if ($state == 'both')
 986                                                 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
 987                                                 else # $state can be 'b' or ''
 988                                                 { $output .= '<i>'; $state .= 'i'; }
 989                                         }
 990                                         else if (strlen ($r) == 3)
 991                                         {
 992                                                 if ($state == 'b')
 993                                                 { $output .= '</b>'; $state = ''; }
 994                                                 else if ($state == 'bi')
 995                                                 { $output .= '</i></b><i>'; $state = 'i'; }
 996                                                 else if ($state == 'ib')
 997                                                 { $output .= '</b>'; $state = 'i'; }
 998                                                 else if ($state == 'both')
 999                                                 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
1000                                                 else # $state can be 'i' or ''
1001                                                 { $output .= '<b>'; $state .= 'b'; }
1002                                         }
1003                                         else if (strlen ($r) == 5)
1004                                         {
1005                                                 if ($state == 'b')
1006                                                 { $output .= '</b><i>'; $state = 'i'; }
1007                                                 else if ($state == 'i')
1008                                                 { $output .= '</i><b>'; $state = 'b'; }
1009                                                 else if ($state == 'bi')
1010                                                 { $output .= '</i></b>'; $state = ''; }
1011                                                 else if ($state == 'ib')
1012                                                 { $output .= '</b></i>'; $state = ''; }
1013                                                 else if ($state == 'both')
1014                                                 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
1015                                                 else # ($state == '')
1016                                                 { $buffer = ''; $state = 'both'; }
1017                                         }
1018                                 }
1019                                 $i++;
1020                         }
1021                         # Now close all remaining tags.  Notice that the order is important.
1022                         if ($state == 'b' || $state == 'ib')
1023                                 $output .= '</b>';
1024                         if ($state == 'i' || $state == 'bi' || $state == 'ib')
1025                                 $output .= '</i>';
1026                         if ($state == 'bi')
1027                                 $output .= '</b>';
1028                         if ($state == 'both')
1029                                 $output .= '<b><i>'.$buffer.'</i></b>';
1030                         return $output;
1031                 }
1032         }
1033
1034         /**
1035          * Replace external links
1036          *
1037          * Note: this is all very hackish and the order of execution matters a lot.
1038          * Make sure to run maintenance/parserTests.php if you change this code.
1039          *
1040          * @access private
1041          */
1042         function replaceExternalLinks( $text ) {
1043                 global $wgContLang;
1044                 $fname = 'Parser::replaceExternalLinks';
1045                 wfProfileIn( $fname );
1046
1047                 $sk =& $this->mOptions->getSkin();
1048
1049                 $bits = preg_split( EXT_LINK_BRACKETED, $text, -1, PREG_SPLIT_DELIM_CAPTURE );
1050
1051                 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
1052
1053                 $i = 0;
1054                 while ( $i<count( $bits ) ) {
1055                         $url = $bits[$i++];
1056                         $protocol = $bits[$i++];
1057                         $text = $bits[$i++];
1058                         $trail = $bits[$i++];
1059
1060                         # The characters '<' and '>' (which were escaped by
1061                         # removeHTMLtags()) should not be included in
1062                         # URLs, per RFC 2396.
1063                         if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE)) {
1064                                 $text = substr($url, $m2[0][1]) . ' ' . $text;
1065                                 $url = substr($url, 0, $m2[0][1]);
1066                         }
1067
1068                         # If the link text is an image URL, replace it with an <img> tag
1069                         # This happened by accident in the original parser, but some people used it extensively
1070                         $img = $this->maybeMakeExternalImage( $text );
1071                         if ( $img !== false ) {
1072                                 $text = $img;
1073                         }
1074
1075                         $dtrail = '';
1076
1077                         # Set linktype for CSS - if URL==text, link is essentially free
1078                         $linktype = ($text == $url) ? 'free' : 'text';
1079
1080                         # No link text, e.g. [http://domain.tld/some.link]
1081                         if ( $text == '' ) {
1082                                 # Autonumber if allowed
1083                                 if ( strpos( HTTP_PROTOCOLS, str_replace('/','\/', $protocol) ) !== false ) {
1084                                         $text = '[' . ++$this->mAutonumber . ']';
1085                                         $linktype = 'autonumber';
1086                                 } else {
1087                                         # Otherwise just use the URL
1088                                         $text = htmlspecialchars( $url );
1089                                         $linktype = 'free';
1090                                 }
1091                         } else {
1092                                 # Have link text, e.g. [http://domain.tld/some.link text]s
1093                                 # Check for trail
1094                                 list( $dtrail, $trail ) = Linker::splitTrail( $trail );
1095                         }
1096
1097                         $text = $wgContLang->markNoConversion($text);
1098
1099                         # Replace &amp; from obsolete syntax with &.
1100                         # All HTML entities will be escaped by makeExternalLink()
1101                         # or maybeMakeExternalImage()
1102                         $url = str_replace( '&amp;', '&', $url );
1103
1104                         # Process the trail (i.e. everything after this link up until start of the next link),
1105                         # replacing any non-bracketed links
1106                         $trail = $this->replaceFreeExternalLinks( $trail );
1107
1108
1109                         # Use the encoded URL
1110                         # This means that users can paste URLs directly into the text
1111                         # Funny characters like &ouml; aren't valid in URLs anyway
1112                         # This was changed in August 2004
1113                         $s .= $sk->makeExternalLink( $url, $text, false, $linktype ) . $dtrail . $trail;
1114                 }
1115
1116                 wfProfileOut( $fname );
1117                 return $s;
1118         }
1119
1120         /**
1121          * Replace anything that looks like a URL with a link
1122          * @access private
1123          */
1124         function replaceFreeExternalLinks( $text ) {
1125                 global $wgContLang;
1126                 $fname = 'Parser::replaceFreeExternalLinks';
1127                 wfProfileIn( $fname );
1128
1129                 $bits = preg_split( '/(\b(?:' . wfUrlProtocols() . '))/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE );
1130                 $s = array_shift( $bits );
1131                 $i = 0;
1132
1133                 $sk =& $this->mOptions->getSkin();
1134
1135                 while ( $i < count( $bits ) ){
1136                         $protocol = $bits[$i++];
1137                         $remainder = $bits[$i++];
1138
1139                         if ( preg_match( '/^('.EXT_LINK_URL_CLASS.'+)(.*)$/s', $remainder, $m ) ) {
1140                                 # Found some characters after the protocol that look promising
1141                                 $url = $protocol . $m[1];
1142                                 $trail = $m[2];
1143
1144                                 # The characters '<' and '>' (which were escaped by
1145                                 # removeHTMLtags()) should not be included in
1146                                 # URLs, per RFC 2396.
1147                                 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE)) {
1148                                         $trail = substr($url, $m2[0][1]) . $trail;
1149                                         $url = substr($url, 0, $m2[0][1]);
1150                                 }
1151
1152                                 # Move trailing punctuation to $trail
1153                                 $sep = ',;\.:!?';
1154                                 # If there is no left bracket, then consider right brackets fair game too
1155                                 if ( strpos( $url, '(' ) === false ) {
1156                                         $sep .= ')';
1157                                 }
1158
1159                                 $numSepChars = strspn( strrev( $url ), $sep );
1160                                 if ( $numSepChars ) {
1161                                         $trail = substr( $url, -$numSepChars ) . $trail;
1162                                         $url = substr( $url, 0, -$numSepChars );
1163                                 }
1164
1165                                 # Replace &amp; from obsolete syntax with &.
1166                                 # All HTML entities will be escaped by makeExternalLink()
1167                                 # or maybeMakeExternalImage()
1168                                 $url = str_replace( '&amp;', '&', $url );
1169
1170                                 # Is this an external image?
1171                                 $text = $this->maybeMakeExternalImage( $url );
1172                                 if ( $text === false ) {
1173                                         # Not an image, make a link
1174                                         $text = $sk->makeExternalLink( $url, $wgContLang->markNoConversion($url), true, 'free' );
1175                                 }
1176                                 $s .= $text . $trail;
1177                         } else {
1178                                 $s .= $protocol . $remainder;
1179                         }
1180                 }
1181                 wfProfileOut( $fname );
1182                 return $s;
1183         }
1184
1185         /**
1186          * make an image if it's allowed, either through the global
1187          * option or through the exception
1188          * @access private
1189          */
1190         function maybeMakeExternalImage( $url ) {
1191                 $sk =& $this->mOptions->getSkin();
1192                 $imagesfrom = $this->mOptions->getAllowExternalImagesFrom();
1193                 $imagesexception = !empty($imagesfrom);
1194                 $text = false;
1195                 if ( $this->mOptions->getAllowExternalImages()
1196                      || ( $imagesexception && strpos( $url, $imagesfrom ) === 0 ) ) {
1197                         if ( preg_match( EXT_IMAGE_REGEX, $url ) ) {
1198                                 # Image found
1199                                 $text = $sk->makeExternalImage( htmlspecialchars( $url ) );
1200                         }
1201                 }
1202                 return $text;
1203         }
1204
1205         /**
1206          * Process [[ ]] wikilinks
1207          *
1208          * @access private
1209          */
1210         function replaceInternalLinks( $s ) {
1211                 global $wgContLang, $wgLinkCache;
1212                 static $fname = 'Parser::replaceInternalLinks' ;
1213
1214                 wfProfileIn( $fname );
1215
1216                 wfProfileIn( $fname.'-setup' );
1217                 static $tc = FALSE;
1218                 # the % is needed to support urlencoded titles as well
1219                 if ( !$tc ) { $tc = Title::legalChars() . '#%'; }
1220
1221                 $sk =& $this->mOptions->getSkin();
1222
1223                 #split the entire text string on occurences of [[
1224                 $a = explode( '[[', ' ' . $s );
1225                 #get the first element (all text up to first [[), and remove the space we added
1226                 $s = array_shift( $a );
1227                 $s = substr( $s, 1 );
1228
1229                 # Match a link having the form [[namespace:link|alternate]]trail
1230                 static $e1 = FALSE;
1231                 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
1232                 # Match cases where there is no "]]", which might still be images
1233                 static $e1_img = FALSE;
1234                 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1235                 # Match the end of a line for a word that's not followed by whitespace,
1236                 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1237                 $e2 = wfMsgForContent( 'linkprefix' );
1238
1239                 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1240
1241                 if( is_null( $this->mTitle ) ) {
1242                         wfDebugDieBacktrace( 'nooo' );
1243                 }
1244                 $nottalk = !$this->mTitle->isTalkPage();
1245
1246                 if ( $useLinkPrefixExtension ) {
1247                         if ( preg_match( $e2, $s, $m ) ) {
1248                                 $first_prefix = $m[2];
1249                         } else {
1250                                 $first_prefix = false;
1251                         }
1252                 } else {
1253                         $prefix = '';
1254                 }
1255
1256                 $selflink = $this->mTitle->getPrefixedText();
1257                 wfProfileOut( $fname.'-setup' );
1258
1259                 $checkVariantLink = sizeof($wgContLang->getVariants())>1;
1260                 $useSubpages = $this->areSubpagesAllowed();
1261
1262                 # Loop for each link
1263                 for ($k = 0; isset( $a[$k] ); $k++) {
1264                         $line = $a[$k];
1265                         if ( $useLinkPrefixExtension ) {
1266                                 wfProfileIn( $fname.'-prefixhandling' );
1267                                 if ( preg_match( $e2, $s, $m ) ) {
1268                                         $prefix = $m[2];
1269                                         $s = $m[1];
1270                                 } else {
1271                                         $prefix='';
1272                                 }
1273                                 # first link
1274                                 if($first_prefix) {
1275                                         $prefix = $first_prefix;
1276                                         $first_prefix = false;
1277                                 }
1278                                 wfProfileOut( $fname.'-prefixhandling' );
1279                         }
1280
1281                         $might_be_img = false;
1282
1283                         if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1284                                 $text = $m[2];
1285                                 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
1286                                 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
1287                                 # the real problem is with the $e1 regex
1288                                 # See bug 1300.
1289                                 #
1290                                 # Still some problems for cases where the ] is meant to be outside punctuation,
1291                                 # and no image is in sight. See bug 2095.
1292                                 #
1293                                 if( $text !== '' && preg_match( "/^\](.*)/s", $m[3], $n ) ) {
1294                                         $text .= ']'; # so that replaceExternalLinks($text) works later
1295                                         $m[3] = $n[1];
1296                                 }
1297                                 # fix up urlencoded title texts
1298                                 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1299                                 $trail = $m[3];
1300                         } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1301                                 $might_be_img = true;
1302                                 $text = $m[2];
1303                                 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1304                                 $trail = "";
1305                         } else { # Invalid form; output directly
1306                                 $s .= $prefix . '[[' . $line ;
1307                                 continue;
1308                         }
1309
1310                         # Don't allow internal links to pages containing
1311                         # PROTO: where PROTO is a valid URL protocol; these
1312                         # should be external links.
1313                         if (preg_match('/^(\b(?:' . wfUrlProtocols() . '))/', $m[1])) {
1314                                 $s .= $prefix . '[[' . $line ;
1315                                 continue;
1316                         }
1317
1318                         # Make subpage if necessary
1319                         if( $useSubpages ) {
1320                                 $link = $this->maybeDoSubpageLink( $m[1], $text );
1321                         } else {
1322                                 $link = $m[1];
1323                         }
1324
1325                         $noforce = (substr($m[1], 0, 1) != ':');
1326                         if (!$noforce) {
1327                                 # Strip off leading ':'
1328                                 $link = substr($link, 1);
1329                         }
1330
1331                         $nt = Title::newFromText( $this->unstripNoWiki($link, $this->mStripState) );
1332                         if( !$nt ) {
1333                                 $s .= $prefix . '[[' . $line;
1334                                 continue;
1335                         }
1336
1337                         #check other language variants of the link
1338                         #if the article does not exist
1339                         if( $checkVariantLink
1340                             && $nt->getArticleID() == 0 ) {
1341                                 $wgContLang->findVariantLink($link, $nt);
1342                         }
1343
1344                         $ns = $nt->getNamespace();
1345                         $iw = $nt->getInterWiki();
1346
1347                         if ($might_be_img) { # if this is actually an invalid link
1348                                 if ($ns == NS_IMAGE && $noforce) { #but might be an image
1349                                         $found = false;
1350                                         while (isset ($a[$k+1]) ) {
1351                                                 #look at the next 'line' to see if we can close it there
1352                                                 $spliced = array_splice( $a, $k + 1, 1 );
1353                                                 $next_line = array_shift( $spliced );
1354                                                 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1355                                                 # the first ]] closes the inner link, the second the image
1356                                                         $found = true;
1357                                                         $text .= '[[' . $m[1];
1358                                                         $trail = $m[2];
1359                                                         break;
1360                                                 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1361                                                         #if there's exactly one ]] that's fine, we'll keep looking
1362                                                         $text .= '[[' . $m[0];
1363                                                 } else {
1364                                                         #if $next_line is invalid too, we need look no further
1365                                                         $text .= '[[' . $next_line;
1366                                                         break;
1367                                                 }
1368                                         }
1369                                         if ( !$found ) {
1370                                                 # we couldn't find the end of this imageLink, so output it raw
1371                                                 #but don't ignore what might be perfectly normal links in the text we've examined
1372                                                 $text = $this->replaceInternalLinks($text);
1373                                                 $s .= $prefix . '[[' . $link . '|' . $text;
1374                                                 # note: no $trail, because without an end, there *is* no trail
1375                                                 continue;
1376                                         }
1377                                 } else { #it's not an image, so output it raw
1378                                         $s .= $prefix . '[[' . $link . '|' . $text;
1379                                         # note: no $trail, because without an end, there *is* no trail
1380                                         continue;
1381                                 }
1382                         }
1383
1384                         $wasblank = ( '' == $text );
1385                         if( $wasblank ) $text = $link;
1386
1387
1388                         # Link not escaped by : , create the various objects
1389                         if( $noforce ) {
1390
1391                                 # Interwikis
1392                                 if( $iw && $this->mOptions->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1393                                         array_push( $this->mOutput->mLanguageLinks, $nt->getFullText() );
1394                                         $s = rtrim($s . "\n");
1395                                         $s .= trim($prefix . $trail, "\n") == '' ? '': $prefix . $trail;
1396                                         continue;
1397                                 }
1398
1399                                 if ( $ns == NS_IMAGE ) {
1400                                         wfProfileIn( "$fname-image" );
1401                                         if ( !wfIsBadImage( $nt->getDBkey() ) ) {
1402                                                 # recursively parse links inside the image caption
1403                                                 # actually, this will parse them in any other parameters, too,
1404                                                 # but it might be hard to fix that, and it doesn't matter ATM
1405                                                 $text = $this->replaceExternalLinks($text);
1406                                                 $text = $this->replaceInternalLinks($text);
1407
1408                                                 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1409                                                 $s .= $prefix . preg_replace( "/\b(" . wfUrlProtocols() . ')/', UNIQ_PREFIX."NOPARSE$1", $this->makeImage( $nt, $text) ) . $trail;
1410                                                 $wgLinkCache->addImageLinkObj( $nt );
1411
1412                                                 wfProfileOut( "$fname-image" );
1413                                                 continue;
1414                                         }
1415                                         wfProfileOut( "$fname-image" );
1416
1417                                 }
1418
1419                                 if ( $ns == NS_CATEGORY ) {
1420                                         wfProfileIn( "$fname-category" );
1421                                         $t = $wgContLang->convertHtml( $nt->getText() );
1422                                         $s = rtrim($s . "\n"); # bug 87
1423
1424                                         $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1425                                         $t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
1426                                         $wgLinkCache->resume();
1427
1428                                         if ( $wasblank ) {
1429                                                 if ( $this->mTitle->getNamespace() == NS_CATEGORY ) {
1430                                                         $sortkey = $this->mTitle->getText();
1431                                                 } else {
1432                                                         $sortkey = $this->mTitle->getPrefixedText();
1433                                                 }
1434                                         } else {
1435                                                 $sortkey = $text;
1436                                         }
1437                                         $sortkey = Sanitizer::decodeCharReferences( $sortkey );
1438                                         $sortkey = $wgContLang->convertCategoryKey( $sortkey );
1439                                         $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1440                                         $this->mOutput->addCategoryLink( $t );
1441
1442                                         /**
1443                                          * Strip the whitespace Category links produce, see bug 87
1444                                          * @todo We might want to use trim($tmp, "\n") here.
1445                                          */
1446                                         $s .= trim($prefix . $trail, "\n") == '' ? '': $prefix . $trail;
1447
1448                                         wfProfileOut( "$fname-category" );
1449                                         continue;
1450                                 }
1451                         }
1452
1453                         if( ( $nt->getPrefixedText() === $selflink ) &&
1454                             ( $nt->getFragment() === '' ) ) {
1455                                 # Self-links are handled specially; generally de-link and change to bold.
1456                                 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1457                                 continue;
1458                         }
1459
1460                         # Special and Media are pseudo-namespaces; no pages actually exist in them
1461                         if( $ns == NS_MEDIA ) {
1462                                 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text, true ) . $trail;
1463                                 $wgLinkCache->addImageLinkObj( $nt );
1464                                 continue;
1465                         } elseif( $ns == NS_SPECIAL ) {
1466                                 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1467                                 continue;
1468                         } elseif( $ns == NS_IMAGE ) {
1469                                 $img = Image::newFromTitle( $nt );
1470                                 if( $img->exists() ) {
1471                                         // Force a blue link if the file exists; may be a remote
1472                                         // upload on the shared repository, and we want to see its
1473                                         // auto-generated page.
1474                                         $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1475                                         continue;
1476                                 }
1477                         }
1478                         $s .= $this->makeLinkHolder( $nt, $text, '', $trail, $prefix );
1479                 }
1480                 wfProfileOut( $fname );
1481                 return $s;
1482         }
1483
1484         /**
1485          * Make a link placeholder. The text returned can be later resolved to a real link with
1486          * replaceLinkHolders(). This is done for two reasons: firstly to avoid further
1487          * parsing of interwiki links, and secondly to allow all extistence checks and
1488          * article length checks (for stub links) to be bundled into a single query.
1489          *
1490          */
1491         function makeLinkHolder( &$nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1492                 if ( ! is_object($nt) ) {
1493                         # Fail gracefully
1494                         $retVal = "<!-- ERROR -->{$prefix}{$text}{$trail}";
1495                 } else {
1496                         # Separate the link trail from the rest of the link
1497                         list( $inside, $trail ) = Linker::splitTrail( $trail );
1498
1499                         if ( $nt->isExternal() ) {
1500                                 $nr = array_push( $this->mInterwikiLinkHolders['texts'], $prefix.$text.$inside );
1501                                 $this->mInterwikiLinkHolders['titles'][] = $nt;
1502                                 $retVal = '<!--IWLINK '. ($nr-1) ."-->{$trail}";
1503                         } else {
1504                                 $nr = array_push( $this->mLinkHolders['namespaces'], $nt->getNamespace() );
1505                                 $this->mLinkHolders['dbkeys'][] = $nt->getDBkey();
1506                                 $this->mLinkHolders['queries'][] = $query;
1507                                 $this->mLinkHolders['texts'][] = $prefix.$text.$inside;
1508                                 $this->mLinkHolders['titles'][] = $nt;
1509
1510                                 $retVal = '<!--LINK '. ($nr-1) ."-->{$trail}";
1511                         }
1512                 }
1513                 return $retVal;
1514         }
1515
1516         /**
1517          * Return true if subpage links should be expanded on this page.
1518          * @return bool
1519          */
1520         function areSubpagesAllowed() {
1521                 # Some namespaces don't allow subpages
1522                 global $wgNamespacesWithSubpages;
1523                 return !empty($wgNamespacesWithSubpages[$this->mTitle->getNamespace()]);
1524         }
1525
1526         /**
1527          * Handle link to subpage if necessary
1528          * @param string $target the source of the link
1529          * @param string &$text the link text, modified as necessary
1530          * @return string the full name of the link
1531          * @access private
1532          */
1533         function maybeDoSubpageLink($target, &$text) {
1534                 # Valid link forms:
1535                 # Foobar -- normal
1536                 # :Foobar -- override special treatment of prefix (images, language links)
1537                 # /Foobar -- convert to CurrentPage/Foobar
1538                 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1539                 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1540                 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1541
1542                 $fname = 'Parser::maybeDoSubpageLink';
1543                 wfProfileIn( $fname );
1544                 $ret = $target; # default return value is no change
1545
1546                 # Some namespaces don't allow subpages,
1547                 # so only perform processing if subpages are allowed
1548                 if( $this->areSubpagesAllowed() ) {
1549                         # Look at the first character
1550                         if( $target != '' && $target{0} == '/' ) {
1551                                 # / at end means we don't want the slash to be shown
1552                                 if( substr( $target, -1, 1 ) == '/' ) {
1553                                         $target = substr( $target, 1, -1 );
1554                                         $noslash = $target;
1555                                 } else {
1556                                         $noslash = substr( $target, 1 );
1557                                 }
1558
1559                                 $ret = $this->mTitle->getPrefixedText(). '/' . trim($noslash);
1560                                 if( '' === $text ) {
1561                                         $text = $target;
1562                                 } # this might be changed for ugliness reasons
1563                         } else {
1564                                 # check for .. subpage backlinks
1565                                 $dotdotcount = 0;
1566                                 $nodotdot = $target;
1567                                 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1568                                         ++$dotdotcount;
1569                                         $nodotdot = substr( $nodotdot, 3 );
1570                                 }
1571                                 if($dotdotcount > 0) {
1572                                         $exploded = explode( '/', $this->mTitle->GetPrefixedText() );
1573                                         if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1574                                                 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1575                                                 # / at the end means don't show full path
1576                                                 if( substr( $nodotdot, -1, 1 ) == '/' ) {
1577                                                         $nodotdot = substr( $nodotdot, 0, -1 );
1578                                                         if( '' === $text ) {
1579                                                                 $text = $nodotdot;
1580                                                         }
1581                                                 }
1582                                                 $nodotdot = trim( $nodotdot );
1583                                                 if( $nodotdot != '' ) {
1584                                                         $ret .= '/' . $nodotdot;
1585                                                 }
1586                                         }
1587                                 }
1588                         }
1589                 }
1590
1591                 wfProfileOut( $fname );
1592                 return $ret;
1593         }
1594
1595         /**#@+
1596          * Used by doBlockLevels()
1597          * @access private
1598          */
1599         /* private */ function closeParagraph() {
1600                 $result = '';
1601                 if ( '' != $this->mLastSection ) {
1602                         $result = '</' . $this->mLastSection  . ">\n";
1603                 }
1604                 $this->mInPre = false;
1605                 $this->mLastSection = '';
1606                 return $result;
1607         }
1608         # getCommon() returns the length of the longest common substring
1609         # of both arguments, starting at the beginning of both.
1610         #
1611         /* private */ function getCommon( $st1, $st2 ) {
1612                 $fl = strlen( $st1 );
1613                 $shorter = strlen( $st2 );
1614                 if ( $fl < $shorter ) { $shorter = $fl; }
1615
1616                 for ( $i = 0; $i < $shorter; ++$i ) {
1617                         if ( $st1{$i} != $st2{$i} ) { break; }
1618                 }
1619                 return $i;
1620         }
1621         # These next three functions open, continue, and close the list
1622         # element appropriate to the prefix character passed into them.
1623         #
1624         /* private */ function openList( $char ) {
1625                 $result = $this->closeParagraph();
1626
1627                 if ( '*' == $char ) { $result .= '<ul><li>'; }
1628                 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1629                 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1630                 else if ( ';' == $char ) {
1631                         $result .= '<dl><dt>';
1632                         $this->mDTopen = true;
1633                 }
1634                 else { $result = '<!-- ERR 1 -->'; }
1635
1636                 return $result;
1637         }
1638
1639         /* private */ function nextItem( $char ) {
1640                 if ( '*' == $char || '#' == $char ) { return '</li><li>'; }
1641                 else if ( ':' == $char || ';' == $char ) {
1642                         $close = '</dd>';
1643                         if ( $this->mDTopen ) { $close = '</dt>'; }
1644                         if ( ';' == $char ) {
1645                                 $this->mDTopen = true;
1646                                 return $close . '<dt>';
1647                         } else {
1648                                 $this->mDTopen = false;
1649                                 return $close . '<dd>';
1650                         }
1651                 }
1652                 return '<!-- ERR 2 -->';
1653         }
1654
1655         /* private */ function closeList( $char ) {
1656                 if ( '*' == $char ) { $text = '</li></ul>'; }
1657                 else if ( '#' == $char ) { $text = '</li></ol>'; }
1658                 else if ( ':' == $char ) {
1659                         if ( $this->mDTopen ) {
1660                                 $this->mDTopen = false;
1661                                 $text = '</dt></dl>';
1662                         } else {
1663                                 $text = '</dd></dl>';
1664                         }
1665                 }
1666                 else {  return '<!-- ERR 3 -->'; }
1667                 return $text."\n";
1668         }
1669         /**#@-*/
1670
1671         /**
1672          * Make lists from lines starting with ':', '*', '#', etc.
1673          *
1674          * @access private
1675          * @return string the lists rendered as HTML
1676          */
1677         function doBlockLevels( $text, $linestart ) {
1678                 $fname = 'Parser::doBlockLevels';
1679                 wfProfileIn( $fname );
1680
1681                 # Parsing through the text line by line.  The main thing
1682                 # happening here is handling of block-level elements p, pre,
1683                 # and making lists from lines starting with * # : etc.
1684                 #
1685                 $textLines = explode( "\n", $text );
1686
1687                 $lastPrefix = $output = '';
1688                 $this->mDTopen = $inBlockElem = false;
1689                 $prefixLength = 0;
1690                 $paragraphStack = false;
1691
1692                 if ( !$linestart ) {
1693                         $output .= array_shift( $textLines );
1694                 }
1695                 foreach ( $textLines as $oLine ) {
1696                         $lastPrefixLength = strlen( $lastPrefix );
1697                         $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1698                         $preOpenMatch = preg_match('/<pre/i', $oLine );
1699                         if ( !$this->mInPre ) {
1700                                 # Multiple prefixes may abut each other for nested lists.
1701                                 $prefixLength = strspn( $oLine, '*#:;' );
1702                                 $pref = substr( $oLine, 0, $prefixLength );
1703
1704                                 # eh?
1705                                 $pref2 = str_replace( ';', ':', $pref );
1706                                 $t = substr( $oLine, $prefixLength );
1707                                 $this->mInPre = !empty($preOpenMatch);
1708                         } else {
1709                                 # Don't interpret any other prefixes in preformatted text
1710                                 $prefixLength = 0;
1711                                 $pref = $pref2 = '';
1712                                 $t = $oLine;
1713                         }
1714
1715                         # List generation
1716                         if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1717                                 # Same as the last item, so no need to deal with nesting or opening stuff
1718                                 $output .= $this->nextItem( substr( $pref, -1 ) );
1719                                 $paragraphStack = false;
1720
1721                                 if ( substr( $pref, -1 ) == ';') {
1722                                         # The one nasty exception: definition lists work like this:
1723                                         # ; title : definition text
1724                                         # So we check for : in the remainder text to split up the
1725                                         # title and definition, without b0rking links.
1726                                         $term = $t2 = '';
1727                                         if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1728                                                 $t = $t2;
1729                                                 $output .= $term . $this->nextItem( ':' );
1730                                         }
1731                                 }
1732                         } elseif( $prefixLength || $lastPrefixLength ) {
1733                                 # Either open or close a level...
1734                                 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1735                                 $paragraphStack = false;
1736
1737                                 while( $commonPrefixLength < $lastPrefixLength ) {
1738                                         $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1739                                         --$lastPrefixLength;
1740                                 }
1741                                 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1742                                         $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1743                                 }
1744                                 while ( $prefixLength > $commonPrefixLength ) {
1745                                         $char = substr( $pref, $commonPrefixLength, 1 );
1746                                         $output .= $this->openList( $char );
1747
1748                                         if ( ';' == $char ) {
1749                                                 # FIXME: This is dupe of code above
1750                                                 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1751                                                         $t = $t2;
1752                                                         $output .= $term . $this->nextItem( ':' );
1753                                                 }
1754                                         }
1755                                         ++$commonPrefixLength;
1756                                 }
1757                                 $lastPrefix = $pref2;
1758                         }
1759                         if( 0 == $prefixLength ) {
1760                                 wfProfileIn( "$fname-paragraph" );
1761                                 # No prefix (not in list)--go to paragraph mode
1762                                 $uniq_prefix = UNIQ_PREFIX;
1763                                 // XXX: use a stack for nestable elements like span, table and div
1764                                 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
1765                                 $closematch = preg_match(
1766                                         '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1767                                         '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/iS', $t );
1768                                 if ( $openmatch or $closematch ) {
1769                                         $paragraphStack = false;
1770                                         $output .= $this->closeParagraph();
1771                                         if ( $preOpenMatch and !$preCloseMatch ) {
1772                                                 $this->mInPre = true;
1773                                         }
1774                                         if ( $closematch ) {
1775                                                 $inBlockElem = false;
1776                                         } else {
1777                                                 $inBlockElem = true;
1778                                         }
1779                                 } else if ( !$inBlockElem && !$this->mInPre ) {
1780                                         if ( ' ' == $t{0} and ( $this->mLastSection == 'pre' or trim($t) != '' ) ) {
1781                                                 // pre
1782                                                 if ($this->mLastSection != 'pre') {
1783                                                         $paragraphStack = false;
1784                                                         $output .= $this->closeParagraph().'<pre>';
1785                                                         $this->mLastSection = 'pre';
1786                                                 }
1787                                                 $t = substr( $t, 1 );
1788                                         } else {
1789                                                 // paragraph
1790                                                 if ( '' == trim($t) ) {
1791                                                         if ( $paragraphStack ) {
1792                                                                 $output .= $paragraphStack.'<br />';
1793                                                                 $paragraphStack = false;
1794                                                                 $this->mLastSection = 'p';
1795                                                         } else {
1796                                                                 if ($this->mLastSection != 'p' ) {
1797                                                                         $output .= $this->closeParagraph();
1798                                                                         $this->mLastSection = '';
1799                                                                         $paragraphStack = '<p>';
1800                                                                 } else {
1801                                                                         $paragraphStack = '</p><p>';
1802                                                                 }
1803                                                         }
1804                                                 } else {
1805                                                         if ( $paragraphStack ) {
1806                                                                 $output .= $paragraphStack;
1807                                                                 $paragraphStack = false;
1808                                                                 $this->mLastSection = 'p';
1809                                                         } else if ($this->mLastSection != 'p') {
1810                                                                 $output .= $this->closeParagraph().'<p>';
1811                                                                 $this->mLastSection = 'p';
1812                                                         }
1813                                                 }
1814                                         }
1815                                 }
1816                                 wfProfileOut( "$fname-paragraph" );
1817                         }
1818                         // somewhere above we forget to get out of pre block (bug 785)
1819                         if($preCloseMatch && $this->mInPre) {
1820                                 $this->mInPre = false;
1821                         }
1822                         if ($paragraphStack === false) {
1823                                 $output .= $t."\n";
1824                         }
1825                 }
1826                 while ( $prefixLength ) {
1827                         $output .= $this->closeList( $pref2{$prefixLength-1} );
1828                         --$prefixLength;
1829                 }
1830                 if ( '' != $this->mLastSection ) {
1831                         $output .= '</' . $this->mLastSection . '>';
1832                         $this->mLastSection = '';
1833                 }
1834
1835                 wfProfileOut( $fname );
1836                 return $output;
1837         }
1838
1839         /**
1840          * Split up a string on ':', ignoring any occurences inside
1841          * <a>..</a> or <span>...</span>
1842          * @param string $str the string to split
1843          * @param string &$before set to everything before the ':'
1844          * @param string &$after set to everything after the ':'
1845          * return string the position of the ':', or false if none found
1846          */
1847         function findColonNoLinks($str, &$before, &$after) {
1848                 # I wonder if we should make this count all tags, not just <a>
1849                 # and <span>. That would prevent us from matching a ':' that
1850                 # comes in the middle of italics other such formatting....
1851                 # -- Wil
1852                 $fname = 'Parser::findColonNoLinks';
1853                 wfProfileIn( $fname );
1854                 $pos = 0;
1855                 do {
1856                         $colon = strpos($str, ':', $pos);
1857
1858                         if ($colon !== false) {
1859                                 $before = substr($str, 0, $colon);
1860                                 $after = substr($str, $colon + 1);
1861
1862                                 # Skip any ':' within <a> or <span> pairs
1863                                 $a = substr_count($before, '<a');
1864                                 $s = substr_count($before, '<span');
1865                                 $ca = substr_count($before, '</a>');
1866                                 $cs = substr_count($before, '</span>');
1867
1868                                 if ($a <= $ca and $s <= $cs) {
1869                                         # Tags are balanced before ':'; ok
1870                                         break;
1871                                 }
1872                                 $pos = $colon + 1;
1873                         }
1874                 } while ($colon !== false);
1875                 wfProfileOut( $fname );
1876                 return $colon;
1877         }
1878
1879         /**
1880          * Return value of a magic variable (like PAGENAME)
1881          *
1882          * @access private
1883          */
1884         function getVariableValue( $index ) {
1885                 global $wgContLang, $wgSitename, $wgServer, $wgServerName, $wgScriptPath;
1886
1887                 /**
1888                  * Some of these require message or data lookups and can be
1889                  * expensive to check many times.
1890                  */
1891                 static $varCache = array();
1892                 if ( wfRunHooks( 'ParserGetVariableValueVarCache', array( &$this, &$varCache ) ) )
1893                         if ( isset( $varCache[$index] ) )
1894                                 return $varCache[$index];
1895
1896                 $ts = time();
1897                 wfRunHooks( 'ParserGetVariableValueTs', array( &$this, &$ts ) );
1898
1899                 switch ( $index ) {
1900                         case MAG_CURRENTMONTH:
1901                                 return $varCache[$index] = $wgContLang->formatNum( date( 'm', $ts ) );
1902                         case MAG_CURRENTMONTHNAME:
1903                                 return $varCache[$index] = $wgContLang->getMonthName( date( 'n', $ts ) );
1904                         case MAG_CURRENTMONTHNAMEGEN:
1905                                 return $varCache[$index] = $wgContLang->getMonthNameGen( date( 'n', $ts ) );
1906                         case MAG_CURRENTMONTHABBREV:
1907                                 return $varCache[$index] = $wgContLang->getMonthAbbreviation( date( 'n', $ts ) );
1908                         case MAG_CURRENTDAY:
1909                                 return $varCache[$index] = $wgContLang->formatNum( date( 'j', $ts ) );
1910                         case MAG_PAGENAME:
1911                                 return $this->mTitle->getText();
1912                         case MAG_PAGENAMEE:
1913                                 return $this->mTitle->getPartialURL();
1914                         case MAG_FULLPAGENAME:
1915                                 return $this->mTitle->getPrefixedText();
1916                         case MAG_FULLPAGENAMEE:
1917                                 return wfUrlencode( $this->mTitle->getPrefixedText() );
1918                         case MAG_REVISIONID:
1919                                 return $this->mRevisionId;
1920                         case MAG_NAMESPACE:
1921                                 return $wgContLang->getNsText( $this->mTitle->getNamespace() );
1922                         case MAG_NAMESPACEE:
1923                                 return wfUrlencode( $wgContLang->getNsText( $this->mTitle->getNamespace() ) );
1924                         case MAG_CURRENTDAYNAME:
1925                                 return $varCache[$index] = $wgContLang->getWeekdayName( date( 'w', $ts ) + 1 );
1926                         case MAG_CURRENTYEAR:
1927                                 return $varCache[$index] = $wgContLang->formatNum( date( 'Y', $ts ), true );
1928                         case MAG_CURRENTTIME:
1929                                 return $varCache[$index] = $wgContLang->time( wfTimestamp( TS_MW, $ts ), false, false );
1930                         case MAG_CURRENTWEEK:
1931                                 return $varCache[$index] = $wgContLang->formatNum( date( 'W', $ts ) );
1932                         case MAG_CURRENTDOW:
1933                                 return $varCache[$index] = $wgContLang->formatNum( date( 'w', $ts ) );
1934                         case MAG_NUMBEROFARTICLES:
1935                                 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfArticles() );
1936                         case MAG_NUMBEROFFILES:
1937                                 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfFiles() );
1938                         case MAG_SITENAME:
1939                                 return $wgSitename;
1940                         case MAG_SERVER:
1941                                 return $wgServer;
1942                         case MAG_SERVERNAME:
1943                                 return $wgServerName;
1944                         case MAG_SCRIPTPATH:
1945                                 return $wgScriptPath;
1946                         default:
1947                                 $ret = null;
1948                                 if ( wfRunHooks( 'ParserGetVariableValueSwitch', array( &$this, &$varCache, &$index, &$ret ) ) )
1949                                         return $ret;
1950                                 else
1951                                         return null;
1952                 }
1953         }
1954
1955         /**
1956          * initialise the magic variables (like CURRENTMONTHNAME)
1957          *
1958          * @access private
1959          */
1960         function initialiseVariables() {
1961                 $fname = 'Parser::initialiseVariables';
1962                 wfProfileIn( $fname );
1963                 global $wgVariableIDs;
1964                 $this->mVariables = array();
1965                 foreach ( $wgVariableIDs as $id ) {
1966                         $mw =& MagicWord::get( $id );
1967                         $mw->addToArray( $this->mVariables, $id );
1968                 }
1969                 wfProfileOut( $fname );
1970         }
1971
1972         /**
1973          * parse any parentheses in format ((title|part|part))
1974          * and call callbacks to get a replacement text for any found piece
1975          *
1976          * @param string $text The text to parse
1977          * @param array $callbacks rules in form:
1978          *     '{' => array(                            # opening parentheses
1979          *                                      'end' => '}',   # closing parentheses
1980          *                                      'cb' => array(2 => callback,    # replacement callback to call if {{..}} is found
1981          *                                                                4 => callback         # replacement callback to call if {{{{..}}}} is found
1982          *                                                                )
1983          *                                      )
1984          * @access private
1985          */
1986         function replace_callback ($text, $callbacks) {
1987                 $openingBraceStack = array();   # this array will hold a stack of parentheses which are not closed yet
1988                 $lastOpeningBrace = -1;                 # last not closed parentheses
1989
1990                 for ($i = 0; $i < strlen($text); $i++) {
1991                         # check for any opening brace
1992                         $rule = null;
1993                         $nextPos = -1;
1994                         foreach ($callbacks as $key => $value) {
1995                                 $pos = strpos ($text, $key, $i);
1996                                 if (false !== $pos && (-1 == $nextPos || $pos < $nextPos)) {
1997                                         $rule = $value;
1998                                         $nextPos = $pos;
1999                                 }
2000                         }
2001
2002                         if ($lastOpeningBrace >= 0) {
2003                                 $pos = strpos ($text, $openingBraceStack[$lastOpeningBrace]['braceEnd'], $i);
2004
2005                                 if (false !== $pos && (-1 == $nextPos || $pos < $nextPos)){
2006                                         $rule = null;
2007                                         $nextPos = $pos;
2008                                 }
2009
2010                                 $pos = strpos ($text, '|', $i);
2011
2012                                 if (false !== $pos && (-1 == $nextPos || $pos < $nextPos)){
2013                                         $rule = null;
2014                                         $nextPos = $pos;
2015                                 }
2016                         }
2017
2018                         if ($nextPos == -1)
2019                                 break;
2020
2021                         $i = $nextPos;
2022
2023                         # found openning brace, lets add it to parentheses stack
2024                         if (null != $rule) {
2025                                 $piece = array('brace' => $text[$i],
2026                                                            'braceEnd' => $rule['end'],
2027                                                            'count' => 1,
2028                                                            'title' => '',
2029                                                            'parts' => null);
2030
2031                                 # count openning brace characters
2032                                 while ($i+1 < strlen($text) && $text[$i+1] == $piece['brace']) {
2033                                         $piece['count']++;
2034                                         $i++;
2035                                 }
2036
2037                                 $piece['startAt'] = $i+1;
2038                                 $piece['partStart'] = $i+1;
2039
2040                                 # we need to add to stack only if openning brace count is enough for any given rule
2041                                 foreach ($rule['cb'] as $cnt => $fn) {
2042                                         if ($piece['count'] >= $cnt) {
2043                                                 $lastOpeningBrace ++;
2044                                                 $openingBraceStack[$lastOpeningBrace] = $piece;
2045                                                 break;
2046                                         }
2047                                 }
2048
2049                                 continue;
2050                         }
2051                         else if ($lastOpeningBrace >= 0) {
2052                                 # first check if it is a closing brace
2053                                 if ($openingBraceStack[$lastOpeningBrace]['braceEnd'] == $text[$i]) {
2054                                         # lets check if it is enough characters for closing brace
2055                                         $count = 1;
2056                                         while ($i+$count < strlen($text) && $text[$i+$count] == $text[$i])
2057                                                 $count++;
2058
2059                                         # if there are more closing parentheses than opening ones, we parse less
2060                                         if ($openingBraceStack[$lastOpeningBrace]['count'] < $count)
2061                                                 $count = $openingBraceStack[$lastOpeningBrace]['count'];
2062
2063                                         # check for maximum matching characters (if there are 5 closing characters, we will probably need only 3 - depending on the rules)
2064                                         $matchingCount = 0;
2065                                         $matchingCallback = null;
2066                                         foreach ($callbacks[$openingBraceStack[$lastOpeningBrace]['brace']]['cb'] as $cnt => $fn) {
2067                                                 if ($count >= $cnt && $matchingCount < $cnt) {
2068                                                         $matchingCount = $cnt;
2069                                                         $matchingCallback = $fn;
2070                                                 }
2071                                         }
2072
2073                                         if ($matchingCount == 0) {
2074                                                 $i += $count - 1;
2075                                                 continue;
2076                                         }
2077
2078                                         # lets set a title or last part (if '|' was found)
2079                                         if (null === $openingBraceStack[$lastOpeningBrace]['parts'])
2080                                                 $openingBraceStack[$lastOpeningBrace]['title'] = substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'], $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2081                                         else
2082                                                 $openingBraceStack[$lastOpeningBrace]['parts'][] = substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'], $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2083
2084                                         $pieceStart = $openingBraceStack[$lastOpeningBrace]['startAt'] - $matchingCount;
2085                                         $pieceEnd = $i + $matchingCount;
2086
2087                                         if( is_callable( $matchingCallback ) ) {
2088                                                 $cbArgs = array (
2089                                                                                  'text' => substr($text, $pieceStart, $pieceEnd - $pieceStart),
2090                                                                                  'title' => trim($openingBraceStack[$lastOpeningBrace]['title']),
2091                                                                                  'parts' => $openingBraceStack[$lastOpeningBrace]['parts'],
2092                                                                                  'lineStart' => (($pieceStart > 0) && ($text[$pieceStart-1] == '\n')),
2093                                                                                  );
2094                                                 # finally we can call a user callback and replace piece of text
2095                                                 $replaceWith = call_user_func( $matchingCallback, $cbArgs );
2096                                                 $text = substr($text, 0, $pieceStart) . $replaceWith . substr($text, $pieceEnd);
2097                                                 $i = $pieceStart + strlen($replaceWith) - 1;
2098                                         }
2099                                         else {
2100                                                 # null value for callback means that parentheses should be parsed, but not replaced
2101                                                 $i += $matchingCount - 1;
2102                                         }
2103
2104                                         # reset last openning parentheses, but keep it in case there are unused characters
2105                                         $piece = array('brace' => $openingBraceStack[$lastOpeningBrace]['brace'],
2106                                                                    'braceEnd' => $openingBraceStack[$lastOpeningBrace]['braceEnd'],
2107                                                                    'count' => $openingBraceStack[$lastOpeningBrace]['count'],
2108                                                                    'title' => '',
2109                                                                    'parts' => null,
2110                                                                    'startAt' => $openingBraceStack[$lastOpeningBrace]['startAt']);
2111                                         $openingBraceStack[$lastOpeningBrace--] = null;
2112
2113                                         if ($matchingCount < $piece['count']) {
2114                                                 $piece['count'] -= $matchingCount;
2115                                                 $piece['startAt'] -= $matchingCount;
2116                                                 $piece['partStart'] = $piece['startAt'];
2117                                                 # do we still qualify for any callback with remaining count?
2118                                                 foreach ($callbacks[$piece['brace']]['cb'] as $cnt => $fn) {
2119                                                         if ($piece['count'] >= $cnt) {
2120                                                                 $lastOpeningBrace ++;
2121                                                                 $openingBraceStack[$lastOpeningBrace] = $piece;
2122                                                                 break;
2123                                                         }
2124                                                 }
2125                                         }
2126                                         continue;
2127                                 }
2128
2129                                 # lets set a title if it is a first separator, or next part otherwise
2130                                 if ($text[$i] == '|') {
2131                                         if (null === $openingBraceStack[$lastOpeningBrace]['parts']) {
2132                                                 $openingBraceStack[$lastOpeningBrace]['title'] = substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'], $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2133                                                 $openingBraceStack[$lastOpeningBrace]['parts'] = array();
2134                                         }
2135                                         else
2136                                                 $openingBraceStack[$lastOpeningBrace]['parts'][] = substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'], $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2137
2138                                         $openingBraceStack[$lastOpeningBrace]['partStart'] = $i + 1;
2139                                 }
2140                         }
2141                 }
2142
2143                 return $text;
2144         }
2145
2146         /**
2147          * Replace magic variables, templates, and template arguments
2148          * with the appropriate text. Templates are substituted recursively,
2149          * taking care to avoid infinite loops.
2150          *
2151          * Note that the substitution depends on value of $mOutputType:
2152          *  OT_WIKI: only {{subst:}} templates
2153          *  OT_MSG: only magic variables
2154          *  OT_HTML: all templates and magic variables
2155          *
2156          * @param string $tex The text to transform
2157          * @param array $args Key-value pairs representing template parameters to substitute
2158          * @access private
2159          */
2160         function replaceVariables( $text, $args = array() ) {
2161                 # Prevent too big inclusions
2162                 if( strlen( $text ) > MAX_INCLUDE_SIZE ) {
2163                         return $text;
2164                 }
2165
2166                 $fname = 'Parser::replaceVariables';
2167                 wfProfileIn( $fname );
2168
2169                 # This function is called recursively. To keep track of arguments we need a stack:
2170                 array_push( $this->mArgStack, $args );
2171
2172                 $braceCallbacks = array();
2173                 $braceCallbacks[2] = array( &$this, 'braceSubstitution' );
2174                 if ( $this->mOutputType == OT_HTML || $this->mOutputType == OT_WIKI ) {
2175                         $braceCallbacks[3] = array( &$this, 'argSubstitution' );
2176                 }
2177                 $callbacks = array();
2178                 $callbacks['{'] = array('end' => '}', 'cb' => $braceCallbacks);
2179                 $callbacks['['] = array('end' => ']', 'cb' => array(2=>null));
2180                 $text = $this->replace_callback ($text, $callbacks);
2181
2182                 array_pop( $this->mArgStack );
2183
2184                 wfProfileOut( $fname );
2185                 return $text;
2186         }
2187
2188         /**
2189          * Replace magic variables
2190          * @access private
2191          */
2192         function variableSubstitution( $matches ) {
2193                 $fname = 'parser::variableSubstitution';
2194                 $varname = $matches[1];
2195                 wfProfileIn( $fname );
2196                 if ( !$this->mVariables ) {
2197                         $this->initialiseVariables();
2198                 }
2199                 $skip = false;
2200                 if ( $this->mOutputType == OT_WIKI ) {
2201                         # Do only magic variables prefixed by SUBST
2202                         $mwSubst =& MagicWord::get( MAG_SUBST );
2203                         if (!$mwSubst->matchStartAndRemove( $varname ))
2204                                 $skip = true;
2205                         # Note that if we don't substitute the variable below,
2206                         # we don't remove the {{subst:}} magic word, in case
2207                         # it is a template rather than a magic variable.
2208                 }
2209                 if ( !$skip && array_key_exists( $varname, $this->mVariables ) ) {
2210                         $id = $this->mVariables[$varname];
2211                         $text = $this->getVariableValue( $id );
2212                         $this->mOutput->mContainsOldMagic = true;
2213                 } else {
2214                         $text = $matches[0];
2215                 }
2216                 wfProfileOut( $fname );
2217                 return $text;
2218         }
2219
2220         # Split template arguments
2221         function getTemplateArgs( $argsString ) {
2222                 if ( $argsString === '' ) {
2223                         return array();
2224                 }
2225
2226                 $args = explode( '|', substr( $argsString, 1 ) );
2227
2228                 # If any of the arguments contains a '[[' but no ']]', it needs to be
2229                 # merged with the next arg because the '|' character between belongs
2230                 # to the link syntax and not the template parameter syntax.
2231                 $argc = count($args);
2232
2233                 for ( $i = 0; $i < $argc-1; $i++ ) {
2234                         if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
2235                                 $args[$i] .= '|'.$args[$i+1];
2236                                 array_splice($args, $i+1, 1);
2237                                 $i--;
2238                                 $argc--;
2239                         }
2240                 }
2241
2242                 return $args;
2243         }
2244
2245         /**
2246          * Return the text of a template, after recursively
2247          * replacing any variables or templates within the template.
2248          *
2249          * @param array $piece The parts of the template
2250          *  $piece['text']: matched text
2251          *  $piece['title']: the title, i.e. the part before the |
2252          *  $piece['parts']: the parameter array
2253          * @return string the text of the template
2254          * @access private
2255          */
2256         function braceSubstitution( $piece ) {
2257                 global $wgLinkCache, $wgContLang;
2258                 $fname = 'Parser::braceSubstitution';
2259                 wfProfileIn( $fname );
2260
2261                 $found = false;
2262                 $nowiki = false;
2263                 $noparse = false;
2264
2265                 $title = NULL;
2266
2267                 $linestart = '';
2268
2269                 # $part1 is the bit before the first |, and must contain only title characters
2270                 # $args is a list of arguments, starting from index 0, not including $part1
2271
2272                 $part1 = $piece['title'];
2273                 # If the third subpattern matched anything, it will start with |
2274
2275                 if (null == $piece['parts']) {
2276                         $replaceWith = $this->variableSubstitution (array ($piece['text'], $piece['title']));
2277                         if ($replaceWith != $piece['text']) {
2278                                 $text = $replaceWith;
2279                                 $found = true;
2280                                 $noparse = true;
2281                         }
2282                 }
2283
2284                 $args = (null == $piece['parts']) ? array() : $piece['parts'];
2285                 $argc = count( $args );
2286
2287                 # SUBST
2288                 if ( !$found ) {
2289                         $mwSubst =& MagicWord::get( MAG_SUBST );
2290                         if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType == OT_WIKI) ) {
2291                                 # One of two possibilities is true:
2292                                 # 1) Found SUBST but not in the PST phase
2293                                 # 2) Didn't find SUBST and in the PST phase
2294                                 # In either case, return without further processing
2295                                 $text = $piece['text'];
2296                                 $found = true;
2297                                 $noparse = true;
2298                         }
2299                 }
2300
2301                 # MSG, MSGNW and INT
2302                 if ( !$found ) {
2303                         # Check for MSGNW:
2304                         $mwMsgnw =& MagicWord::get( MAG_MSGNW );
2305                         if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
2306                                 $nowiki = true;
2307                         } else {
2308                                 # Remove obsolete MSG:
2309                                 $mwMsg =& MagicWord::get( MAG_MSG );
2310                                 $mwMsg->matchStartAndRemove( $part1 );
2311                         }
2312
2313                         # Check if it is an internal message
2314                         $mwInt =& MagicWord::get( MAG_INT );
2315                         if ( $mwInt->matchStartAndRemove( $part1 ) ) {
2316                                 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
2317                                         $text = $linestart . wfMsgReal( $part1, $args, true );
2318                                         $found = true;
2319                                 }
2320                         }
2321                 }
2322
2323                 # NS
2324                 if ( !$found ) {
2325                         # Check for NS: (namespace expansion)
2326                         $mwNs = MagicWord::get( MAG_NS );
2327                         if ( $mwNs->matchStartAndRemove( $part1 ) ) {
2328                                 if ( intval( $part1 ) ) {
2329                                         $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
2330                                         $found = true;
2331                                 } else {
2332                                         $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
2333                                         if ( !is_null( $index ) ) {
2334                                                 $text = $linestart . $wgContLang->getNsText( $index );
2335                                                 $found = true;
2336                                         }
2337                                 }
2338                         }
2339                 }
2340
2341                 # LCFIRST, UCFIRST, LC and UC
2342                 if ( !$found ) {
2343                         $lcfirst =& MagicWord::get( MAG_LCFIRST );
2344                         $ucfirst =& MagicWord::get( MAG_UCFIRST );
2345                         $lc =& MagicWord::get( MAG_LC );
2346                         $uc =& MagicWord::get( MAG_UC );
2347                         if ( $lcfirst->matchStartAndRemove( $part1 ) ) {
2348                                 $text = $linestart . $wgContLang->lcfirst( $part1 );
2349                                 $found = true;
2350                         } else if ( $ucfirst->matchStartAndRemove( $part1 ) ) {
2351                                 $text = $linestart . $wgContLang->ucfirst( $part1 );
2352                                 $found = true;
2353                         } else if ( $lc->matchStartAndRemove( $part1 ) ) {
2354                                 $text = $linestart . $wgContLang->lc( $part1 );
2355                                 $found = true;
2356                         } else if ( $uc->matchStartAndRemove( $part1 ) ) {
2357                                  $text = $linestart . $wgContLang->uc( $part1 );
2358                                  $found = true;
2359                         }
2360                 }
2361
2362                 # LOCALURL and FULLURL
2363                 if ( !$found ) {
2364                         $mwLocal =& MagicWord::get( MAG_LOCALURL );
2365                         $mwLocalE =& MagicWord::get( MAG_LOCALURLE );
2366                         $mwFull =& MagicWord::get( MAG_FULLURL );
2367                         $mwFullE =& MagicWord::get( MAG_FULLURLE );
2368
2369
2370                         if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
2371                                 $func = 'getLocalURL';
2372                         } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
2373                                 $func = 'escapeLocalURL';
2374                         } elseif ( $mwFull->matchStartAndRemove( $part1 ) ) {
2375                                 $func = 'getFullURL';
2376                         } elseif ( $mwFullE->matchStartAndRemove( $part1 ) ) {
2377                                 $func = 'escapeFullURL';
2378                         } else {
2379                                 $func = false;
2380                         }
2381
2382                         if ( $func !== false ) {
2383                                 $title = Title::newFromText( $part1 );
2384                                 if ( !is_null( $title ) ) {
2385                                         if ( $argc > 0 ) {
2386                                                 $text = $linestart . $title->$func( $args[0] );
2387                                         } else {
2388                                                 $text = $linestart . $title->$func();
2389                                         }
2390                                         $found = true;
2391                                 }
2392                         }
2393                 }
2394
2395                 # GRAMMAR
2396                 if ( !$found && $argc == 1 ) {
2397                         $mwGrammar =& MagicWord::get( MAG_GRAMMAR );
2398                         if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
2399                                 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
2400                                 $found = true;
2401                         }
2402                 }
2403
2404                 # PLURAL
2405                 if ( !$found && $argc >= 2 ) {
2406                         $mwPluralForm =& MagicWord::get( MAG_PLURAL );
2407                         if ( $mwPluralForm->matchStartAndRemove( $part1 ) ) {
2408                                 if ($argc==2) {$args[2]=$args[1];}
2409                                 $text = $linestart . $wgContLang->convertPlural( $part1, $args[0], $args[1], $args[2]);
2410                                 $found = true;
2411                         }
2412                 }
2413
2414                 # Template table test
2415
2416                 # Did we encounter this template already? If yes, it is in the cache
2417                 # and we need to check for loops.
2418                 if ( !$found && isset( $this->mTemplates[$part1] ) ) {
2419                         $found = true;
2420
2421                         # Infinite loop test
2422                         if ( isset( $this->mTemplatePath[$part1] ) ) {
2423                                 $noparse = true;
2424                                 $found = true;
2425                                 $text = $linestart .
2426                                         '{{' . $part1 . '}}' .
2427                                         '<!-- WARNING: template loop detected -->';
2428                                 wfDebug( "$fname: template loop broken at '$part1'\n" );
2429                         } else {
2430                                 # set $text to cached message.
2431                                 $text = $linestart . $this->mTemplates[$part1];
2432                         }
2433                 }
2434
2435                 # Load from database
2436                 $replaceHeadings = false;
2437                 $isHTML = false;
2438                 $lastPathLevel = $this->mTemplatePath;
2439                 if ( !$found ) {
2440                         $ns = NS_TEMPLATE;
2441                         $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
2442                         if ($subpage !== '') {
2443                                 $ns = $this->mTitle->getNamespace();
2444                         }
2445                         $title = Title::newFromText( $part1, $ns );
2446
2447                         if ($title) {
2448                                 $interwiki = Title::getInterwikiLink($title->getInterwiki());
2449                                 if ($interwiki != '' && $title->isTrans()) {
2450                                         return $this->scarytransclude($title, $interwiki);
2451                                 }
2452                         }
2453
2454                         if ( !is_null( $title ) && !$title->isExternal() ) {
2455                                 # Check for excessive inclusion
2456                                 $dbk = $title->getPrefixedDBkey();
2457                                 if ( $this->incrementIncludeCount( $dbk ) ) {
2458                                         if ( $title->getNamespace() == NS_SPECIAL && $this->mOptions->getAllowSpecialInclusion() ) {
2459                                                 # Capture special page output
2460                                                 $text = SpecialPage::capturePath( $title );
2461                                                 if ( is_string( $text ) ) {
2462                                                         $found = true;
2463                                                         $noparse = true;
2464                                                         $isHTML = true;
2465                                                         $this->disableCache();
2466                                                 }
2467                                         } else {
2468                                                 $article = new Article( $title );
2469                                                 $articleContent = $article->fetchContent(0, false);
2470                                                 if ( $articleContent !== false ) {
2471                                                         $found = true;
2472                                                         $text = $articleContent;
2473                                                         $replaceHeadings = true;
2474                                                 }
2475                                         }
2476                                 }
2477
2478                                 # If the title is valid but undisplayable, make a link to it
2479                                 if ( $this->mOutputType == OT_HTML && !$found ) {
2480                                         $text = '[['.$title->getPrefixedText().']]';
2481                                         $found = true;
2482                                 }
2483
2484                                 # Template cache array insertion
2485                                 if( $found ) {
2486                                         $this->mTemplates[$part1] = $text;
2487                                         $text = $linestart . $text;
2488                                 }
2489                         }
2490                 }
2491
2492                 # Recursive parsing, escaping and link table handling
2493                 # Only for HTML output
2494                 if ( $nowiki && $found && $this->mOutputType == OT_HTML ) {
2495                         $text = wfEscapeWikiText( $text );
2496                 } elseif ( ($this->mOutputType == OT_HTML || $this->mOutputType == OT_WIKI) && $found && !$noparse) {
2497                         # Clean up argument array
2498                         $assocArgs = array();
2499                         $index = 1;
2500                         foreach( $args as $arg ) {
2501                                 $eqpos = strpos( $arg, '=' );
2502                                 if ( $eqpos === false ) {
2503                                         $assocArgs[$index++] = $arg;
2504                                 } else {
2505                                         $name = trim( substr( $arg, 0, $eqpos ) );
2506                                         $value = trim( substr( $arg, $eqpos+1 ) );
2507                                         if ( $value === false ) {
2508                                                 $value = '';
2509                                         }
2510                                         if ( $name !== false ) {
2511                                                 $assocArgs[$name] = $value;
2512                                         }
2513                                 }
2514                         }
2515
2516                         # Add a new element to the templace recursion path
2517                         $this->mTemplatePath[$part1] = 1;
2518
2519                         # If there are any <onlyinclude> tags, only include them
2520                         if ( in_string( '<onlyinclude>', $text ) && in_string( '</onlyinclude>', $text ) ) {
2521                                 preg_match_all( '/<onlyinclude>(.*?)\n?<\/onlyinclude>/s', $text, $m );
2522                                 $text = '';
2523                                 foreach ($m[1] as $piece)
2524                                         $text .= $piece;
2525                         }
2526                         # Remove <noinclude> sections and <includeonly> tags
2527                         $text = preg_replace( '/<noinclude>.*?<\/noinclude>/s', '', $text );
2528                         $text = strtr( $text, array( '<includeonly>' => '' , '</includeonly>' => '' ) );
2529
2530                         if( $this->mOutputType == OT_HTML ) {
2531                                 # Strip <nowiki>, <pre>, etc.
2532                                 $text = $this->strip( $text, $this->mStripState );
2533                                 $text = Sanitizer::removeHTMLtags( $text, array( &$this, 'replaceVariables' ), $assocArgs );
2534                         }
2535                         $text = $this->replaceVariables( $text, $assocArgs );
2536
2537                         # Resume the link cache and register the inclusion as a link
2538                         if ( $this->mOutputType == OT_HTML && !is_null( $title ) ) {
2539                                 $wgLinkCache->addLinkObj( $title );
2540                         }
2541
2542                         # If the template begins with a table or block-level
2543                         # element, it should be treated as beginning a new line.
2544                         if (!$piece['lineStart'] && preg_match('/^({\\||:|;|#|\*)/', $text)) {
2545                                 $text = "\n" . $text;
2546                         }
2547                 }
2548                 # Prune lower levels off the recursion check path
2549                 $this->mTemplatePath = $lastPathLevel;
2550
2551                 if ( !$found ) {
2552                         wfProfileOut( $fname );
2553                         return $piece['text'];
2554                 } else {
2555                         if ( $isHTML ) {
2556                                 # Replace raw HTML by a placeholder
2557                                 # Add a blank line preceding, to prevent it from mucking up
2558                                 # immediately preceding headings
2559                                 $text = "\n\n" . $this->insertStripItem( $text, $this->mStripState );
2560                         } else {
2561                                 # replace ==section headers==
2562                                 # XXX this needs to go away once we have a better parser.
2563                                 if ( $this->mOutputType != OT_WIKI && $replaceHeadings ) {
2564                                         if( !is_null( $title ) )
2565                                                 $encodedname = base64_encode($title->getPrefixedDBkey());
2566                                         else
2567                                                 $encodedname = base64_encode("");
2568                                         $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
2569                                                 PREG_SPLIT_DELIM_CAPTURE);
2570                                         $text = '';
2571                                         $nsec = 0;
2572                                         for( $i = 0; $i < count($m); $i += 2 ) {
2573                                                 $text .= $m[$i];
2574                                                 if (!isset($m[$i + 1]) || $m[$i + 1] == "") continue;
2575                                                 $hl = $m[$i + 1];
2576                                                 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
2577                                                         $text .= $hl;
2578                                                         continue;
2579                                                 }
2580                                                 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
2581                                                 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
2582                                                         . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
2583
2584                                                 $nsec++;
2585                                         }
2586                                 }
2587                         }
2588                 }
2589
2590                 # Prune lower levels off the recursion check path
2591                 $this->mTemplatePath = $lastPathLevel;
2592
2593                 if ( !$found ) {
2594                         wfProfileOut( $fname );
2595                         return $piece['text'];
2596                 } else {
2597                         wfProfileOut( $fname );
2598                         return $text;
2599                 }
2600         }
2601
2602         /**
2603          * Translude an interwiki link.
2604          */
2605         function scarytransclude($title, $interwiki) {
2606                 global $wgEnableScaryTranscluding;
2607
2608                 if (!$wgEnableScaryTranscluding)
2609                         return wfMsg('scarytranscludedisabled');
2610
2611                 $articlename = "Template:" . $title->getDBkey();
2612                 $url = str_replace('$1', urlencode($articlename), $interwiki);
2613                 if (strlen($url) > 255)
2614                         return wfMsg('scarytranscludetoolong');
2615                 $text = $this->fetchScaryTemplateMaybeFromCache($url);
2616                 $this->mIWTransData[] = $text;
2617                 return "<!--IW_TRANSCLUDE ".(count($this->mIWTransData) - 1)."-->";
2618         }
2619
2620         function fetchScaryTemplateMaybeFromCache($url) {
2621                 $dbr =& wfGetDB(DB_SLAVE);
2622                 $obj = $dbr->selectRow('transcache', array('tc_time', 'tc_contents'),
2623                                 array('tc_url' => $url));
2624                 if ($obj) {
2625                         $time = $obj->tc_time;
2626                         $text = $obj->tc_contents;
2627                         if ($time && $time < (time() + (60*60))) {
2628                                 return $text;
2629                         }
2630                 }
2631
2632                 $text = wfGetHTTP($url . '?action=render');
2633                 if (!$text)
2634                         return wfMsg('scarytranscludefailed', $url);
2635
2636                 $dbw =& wfGetDB(DB_MASTER);
2637                 $dbw->replace('transcache', array(), array(
2638                         'tc_url' => $url,
2639                         'tc_time' => time(),
2640                         'tc_contents' => $text));
2641                 return $text;
2642         }
2643
2644
2645         /**
2646          * Triple brace replacement -- used for template arguments
2647          * @access private
2648          */
2649         function argSubstitution( $matches ) {
2650                 $arg = trim( $matches['title'] );
2651                 $text = $matches['text'];
2652                 $inputArgs = end( $this->mArgStack );
2653
2654                 if ( array_key_exists( $arg, $inputArgs ) ) {
2655                         $text = $inputArgs[$arg];
2656                 } else if ($this->mOutputType == OT_HTML && null != $matches['parts'] && count($matches['parts']) > 0) {
2657                         $text = $matches['parts'][0];
2658                 }
2659
2660                 return $text;
2661         }
2662
2663         /**
2664          * Returns true if the function is allowed to include this entity
2665          * @access private
2666          */
2667         function incrementIncludeCount( $dbk ) {
2668                 if ( !array_key_exists( $dbk, $this->mIncludeCount ) ) {
2669                         $this->mIncludeCount[$dbk] = 0;
2670                 }
2671                 if ( ++$this->mIncludeCount[$dbk] <= MAX_INCLUDE_REPEAT ) {
2672                         return true;
2673                 } else {
2674                         return false;
2675                 }
2676         }
2677
2678         /**
2679          * This function accomplishes several tasks:
2680          * 1) Auto-number headings if that option is enabled
2681          * 2) Add an [edit] link to sections for logged in users who have enabled the option
2682          * 3) Add a Table of contents on the top for users who have enabled the option
2683          * 4) Auto-anchor headings
2684          *
2685          * It loops through all headlines, collects the necessary data, then splits up the
2686          * string and re-inserts the newly formatted headlines.
2687          *
2688          * @param string $text
2689          * @param boolean $isMain
2690          * @access private
2691          */
2692         function formatHeadings( $text, $isMain=true ) {
2693                 global $wgMaxTocLevel, $wgContLang;
2694
2695                 $doNumberHeadings = $this->mOptions->getNumberHeadings();
2696                 $doShowToc = true;
2697                 $forceTocHere = false;
2698                 if( !$this->mTitle->userCanEdit() ) {
2699                         $showEditLink = 0;
2700                 } else {
2701                         $showEditLink = $this->mOptions->getEditSection();
2702                 }
2703
2704                 # Inhibit editsection links if requested in the page
2705                 $esw =& MagicWord::get( MAG_NOEDITSECTION );
2706                 if( $esw->matchAndRemove( $text ) ) {
2707                         $showEditLink = 0;
2708                 }
2709                 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2710                 # do not add TOC
2711                 $mw =& MagicWord::get( MAG_NOTOC );
2712                 if( $mw->matchAndRemove( $text ) ) {
2713                         $doShowToc = false;
2714                 }
2715
2716                 # Get all headlines for numbering them and adding funky stuff like [edit]
2717                 # links - this is for later, but we need the number of headlines right now
2718                 $numMatches = preg_match_all( '/<H([1-6])(.*?'.'>)(.*?)<\/H[1-6] *>/i', $text, $matches );
2719
2720                 # if there are fewer than 4 headlines in the article, do not show TOC
2721                 if( $numMatches < 4 ) {
2722                         $doShowToc = false;
2723                 }
2724
2725                 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2726                 # override above conditions and always show TOC at that place
2727
2728                 $mw =& MagicWord::get( MAG_TOC );
2729                 if($mw->match( $text ) ) {
2730                         $doShowToc = true;
2731                         $forceTocHere = true;
2732                 } else {
2733                         # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2734                         # override above conditions and always show TOC above first header
2735                         $mw =& MagicWord::get( MAG_FORCETOC );
2736                         if ($mw->matchAndRemove( $text ) ) {
2737                                 $doShowToc = true;
2738                         }
2739                 }
2740
2741                 # Never ever show TOC if no headers
2742                 if( $numMatches < 1 ) {
2743                         $doShowToc = false;
2744                 }
2745
2746                 # We need this to perform operations on the HTML
2747                 $sk =& $this->mOptions->getSkin();
2748
2749                 # headline counter
2750                 $headlineCount = 0;
2751                 $sectionCount = 0; # headlineCount excluding template sections
2752
2753                 # Ugh .. the TOC should have neat indentation levels which can be
2754                 # passed to the skin functions. These are determined here
2755                 $toc = '';
2756                 $full = '';
2757                 $head = array();
2758                 $sublevelCount = array();
2759                 $levelCount = array();
2760                 $toclevel = 0;
2761                 $level = 0;
2762                 $prevlevel = 0;
2763                 $toclevel = 0;
2764                 $prevtoclevel = 0;
2765
2766                 foreach( $matches[3] as $headline ) {
2767                         $istemplate = 0;
2768                         $templatetitle = '';
2769                         $templatesection = 0;
2770                         $numbering = '';
2771
2772                         if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2773                                 $istemplate = 1;
2774                                 $templatetitle = base64_decode($mat[1]);
2775                                 $templatesection = 1 + (int)base64_decode($mat[2]);
2776                                 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2777                         }
2778
2779                         if( $toclevel ) {
2780                                 $prevlevel = $level;
2781                                 $prevtoclevel = $toclevel;
2782                         }
2783                         $level = $matches[1][$headlineCount];
2784
2785                         if( $doNumberHeadings || $doShowToc ) {
2786
2787                                 if ( $level > $prevlevel ) {
2788                                         # Increase TOC level
2789                                         $toclevel++;
2790                                         $sublevelCount[$toclevel] = 0;
2791                                         $toc .= $sk->tocIndent();
2792                                 }
2793                                 elseif ( $level < $prevlevel && $toclevel > 1 ) {
2794                                         # Decrease TOC level, find level to jump to
2795
2796                                         if ( $toclevel == 2 && $level <= $levelCount[1] ) {
2797                                                 # Can only go down to level 1
2798                                                 $toclevel = 1;
2799                                         } else {
2800                                                 for ($i = $toclevel; $i > 0; $i--) {
2801                                                         if ( $levelCount[$i] == $level ) {
2802                                                                 # Found last matching level
2803                                                                 $toclevel = $i;
2804                                                                 break;
2805                                                         }
2806                                                         elseif ( $levelCount[$i] < $level ) {
2807                                                                 # Found first matching level below current level
2808                                                                 $toclevel = $i + 1;
2809                                                                 break;
2810                                                         }
2811                                                 }
2812                                         }
2813
2814                                         $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
2815                                 }
2816                                 else {
2817                                         # No change in level, end TOC line
2818                                         $toc .= $sk->tocLineEnd();
2819                                 }
2820
2821                                 $levelCount[$toclevel] = $level;
2822
2823                                 # count number of headlines for each level
2824                                 @$sublevelCount[$toclevel]++;
2825                                 $dot = 0;
2826                                 for( $i = 1; $i <= $toclevel; $i++ ) {
2827                                         if( !empty( $sublevelCount[$i] ) ) {
2828                                                 if( $dot ) {
2829                                                         $numbering .= '.';
2830                                                 }
2831                                                 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2832                                                 $dot = 1;
2833                                         }
2834                                 }
2835                         }
2836
2837                         # The canonized header is a version of the header text safe to use for links
2838                         # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2839                         $canonized_headline = $this->unstrip( $headline, $this->mStripState );
2840                         $canonized_headline = $this->unstripNoWiki( $canonized_headline, $this->mStripState );
2841
2842                         # Remove link placeholders by the link text.
2843                         #     <!--LINK number-->
2844                         # turns into
2845                         #     link text with suffix
2846                         $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2847                                                             "\$this->mLinkHolders['texts'][\$1]",
2848                                                             $canonized_headline );
2849                         $canonized_headline = preg_replace( '/<!--IWLINK ([0-9]*)-->/e',
2850                                                             "\$this->mInterwikiLinkHolders['texts'][\$1]",
2851                                                             $canonized_headline );
2852
2853                         # strip out HTML
2854                         $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2855                         $tocline = trim( $canonized_headline );
2856                         $canonized_headline = urlencode( Sanitizer::decodeCharReferences( str_replace(' ', '_', $tocline) ) );
2857                         $replacearray = array(
2858                                 '%3A' => ':',
2859                                 '%' => '.'
2860                         );
2861                         $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2862                         $refers[$headlineCount] = $canonized_headline;
2863
2864                         # count how many in assoc. array so we can track dupes in anchors
2865                         @$refers[$canonized_headline]++;
2866                         $refcount[$headlineCount]=$refers[$canonized_headline];
2867
2868                         # Don't number the heading if it is the only one (looks silly)
2869                         if( $doNumberHeadings && count( $matches[3] ) > 1) {
2870                                 # the two are different if the line contains a link
2871                                 $headline=$numbering . ' ' . $headline;
2872                         }
2873
2874                         # Create the anchor for linking from the TOC to the section
2875                         $anchor = $canonized_headline;
2876                         if($refcount[$headlineCount] > 1 ) {
2877                                 $anchor .= '_' . $refcount[$headlineCount];
2878                         }
2879                         if( $doShowToc && ( !isset($wgMaxTocLevel) || $toclevel<$wgMaxTocLevel ) ) {
2880                                 $toc .= $sk->tocLine($anchor, $tocline, $numbering, $toclevel);
2881                         }
2882                         if( $showEditLink && ( !$istemplate || $templatetitle !== "" ) ) {
2883                                 if ( empty( $head[$headlineCount] ) ) {
2884                                         $head[$headlineCount] = '';
2885                                 }
2886                                 if( $istemplate )
2887                                         $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2888                                 else
2889                                         $head[$headlineCount] .= $sk->editSectionLink($this->mTitle, $sectionCount+1);
2890                         }
2891
2892                         # give headline the correct <h#> tag
2893                         @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2894
2895                         $headlineCount++;
2896                         if( !$istemplate )
2897                                 $sectionCount++;
2898                 }
2899
2900                 if( $doShowToc ) {
2901                         $toc .= $sk->tocUnindent( $toclevel - 1 );
2902                         $toc = $sk->tocList( $toc );
2903                 }
2904
2905                 # split up and insert constructed headlines
2906
2907                 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2908                 $i = 0;
2909
2910                 foreach( $blocks as $block ) {
2911                         if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2912                                 # This is the [edit] link that appears for the top block of text when
2913                                 # section editing is enabled
2914
2915                                 # Disabled because it broke block formatting
2916                                 # For example, a bullet point in the top line
2917                                 # $full .= $sk->editSectionLink(0);
2918                         }
2919                         $full .= $block;
2920                         if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2921                         # Top anchor now in skin
2922                                 $full = $full.$toc;
2923                         }
2924
2925                         if( !empty( $head[$i] ) ) {
2926                                 $full .= $head[$i];
2927                         }
2928                         $i++;
2929                 }
2930                 if($forceTocHere) {
2931                         $mw =& MagicWord::get( MAG_TOC );
2932                         return $mw->replace( $toc, $full );
2933                 } else {
2934                         return $full;
2935                 }
2936         }
2937
2938         /**
2939          * Return an HTML link for the "ISBN 123456" text
2940          * @access private
2941          */
2942         function magicISBN( $text ) {
2943                 $fname = 'Parser::magicISBN';
2944                 wfProfileIn( $fname );
2945
2946                 $a = split( 'ISBN ', ' '.$text );
2947                 if ( count ( $a ) < 2 ) {
2948                         wfProfileOut( $fname );
2949                         return $text;
2950                 }
2951                 $text = substr( array_shift( $a ), 1);
2952                 $valid = '0123456789-Xx';
2953
2954                 foreach ( $a as $x ) {
2955                         $isbn = $blank = '' ;
2956                         while ( ' ' == $x{0} ) {
2957                                 $blank .= ' ';
2958                                 $x = substr( $x, 1 );
2959                         }
2960                         if ( $x == '' ) { # blank isbn
2961                                 $text .= "ISBN $blank";
2962                                 continue;
2963                         }
2964                         while ( strstr( $valid, $x{0} ) != false ) {
2965                                 $isbn .= $x{0};
2966                                 $x = substr( $x, 1 );
2967                         }
2968                         $num = str_replace( '-', '', $isbn );
2969                         $num = str_replace( ' ', '', $num );
2970                         $num = str_replace( 'x', 'X', $num );
2971
2972                         if ( '' == $num ) {
2973                                 $text .= "ISBN $blank$x";
2974                         } else {
2975                                 $titleObj = Title::makeTitle( NS_SPECIAL, 'Booksources' );
2976                                 $text .= '<a href="' .
2977                                 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2978                                         "\" class=\"internal\">ISBN $isbn</a>";
2979                                 $text .= $x;
2980                         }
2981                 }
2982                 wfProfileOut( $fname );
2983                 return $text;
2984         }
2985
2986         /**
2987          * Return an HTML link for the "RFC 1234" text
2988          *
2989          * @access private
2990          * @param string $text     Text to be processed
2991          * @param string $keyword  Magic keyword to use (default RFC)
2992          * @param string $urlmsg   Interface message to use (default rfcurl)
2993          * @return string
2994          */
2995         function magicRFC( $text, $keyword='RFC ', $urlmsg='rfcurl'  ) {
2996
2997                 $valid = '0123456789';
2998                 $internal = false;
2999
3000                 $a = split( $keyword, ' '.$text );
3001                 if ( count ( $a ) < 2 ) {
3002                         return $text;
3003                 }
3004                 $text = substr( array_shift( $a ), 1);
3005
3006                 /* Check if keyword is preceed by [[.
3007                  * This test is made here cause of the array_shift above
3008                  * that prevent the test to be done in the foreach.
3009                  */
3010                 if ( substr( $text, -2 ) == '[[' ) {
3011                         $internal = true;
3012                 }
3013
3014                 foreach ( $a as $x ) {
3015                         /* token might be empty if we have RFC RFC 1234 */
3016                         if ( $x=='' ) {
3017                                 $text.=$keyword;
3018                                 continue;
3019                                 }
3020
3021                         $id = $blank = '' ;
3022
3023                         /** remove and save whitespaces in $blank */
3024                         while ( $x{0} == ' ' ) {
3025                                 $blank .= ' ';
3026                                 $x = substr( $x, 1 );
3027                         }
3028
3029                         /** remove and save the rfc number in $id */
3030                         while ( strstr( $valid, $x{0} ) != false ) {
3031                                 $id .= $x{0};
3032                                 $x = substr( $x, 1 );
3033                         }
3034
3035                         if ( $id == '' ) {
3036                                 /* call back stripped spaces*/
3037                                 $text .= $keyword.$blank.$x;
3038                         } elseif( $internal ) {
3039                                 /* normal link */
3040                                 $text .= $keyword.$id.$x;
3041                         } else {
3042                                 /* build the external link*/
3043                                 $url = wfMsg( $urlmsg, $id);
3044                                 $sk =& $this->mOptions->getSkin();
3045                                 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
3046                                 $text .= "<a href='{$url}'{$la}>{$keyword}{$id}</a>{$x}";
3047                         }
3048
3049                         /* Check if the next RFC keyword is preceed by [[ */
3050                         $internal = ( substr($x,-2) == '[[' );
3051                 }
3052                 return $text;
3053         }
3054
3055         /**
3056          * Transform wiki markup when saving a page by doing \r\n -> \n
3057          * conversion, substitting signatures, {{subst:}} templates, etc.
3058          *
3059          * @param string $text the text to transform
3060          * @param Title &$title the Title object for the current article
3061          * @param User &$user the User object describing the current user
3062          * @param ParserOptions $options parsing options
3063          * @param bool $clearState whether to clear the parser state first
3064          * @return string the altered wiki markup
3065          * @access public
3066          */
3067         function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
3068                 $this->mOptions = $options;
3069                 $this->mTitle =& $title;
3070                 $this->mOutputType = OT_WIKI;
3071
3072                 if ( $clearState ) {
3073                         $this->clearState();
3074                 }
3075
3076                 $stripState = false;
3077                 $pairs = array(
3078                         "\r\n" => "\n",
3079                 );
3080                 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
3081                 $text = $this->strip( $text, $stripState, true );
3082                 $text = $this->pstPass2( $text, $user );
3083                 $text = $this->unstrip( $text, $stripState );
3084                 $text = $this->unstripNoWiki( $text, $stripState );
3085                 return $text;
3086         }
3087
3088         /**
3089          * Pre-save transform helper function
3090          * @access private
3091          */
3092         function pstPass2( $text, &$user ) {
3093                 global $wgContLang, $wgLocaltimezone;
3094
3095                 # Variable replacement
3096                 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
3097                 $text = $this->replaceVariables( $text );
3098
3099                 # Signatures
3100                 #
3101                 $sigText = $this->getUserSig( $user );
3102
3103                 /* Note: This is the timestamp saved as hardcoded wikitext to
3104                  * the database, we use $wgContLang here in order to give
3105                  * everyone the same signiture and use the default one rather
3106                  * than the one selected in each users preferences.
3107                  */
3108                 if ( isset( $wgLocaltimezone ) ) {
3109                         $oldtz = getenv( 'TZ' );
3110                         putenv( 'TZ='.$wgLocaltimezone );
3111                 }
3112                 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false, false) .
3113                   ' (' . date( 'T' ) . ')';
3114                 if ( isset( $wgLocaltimezone ) ) {
3115                         putenv( 'TZ='.$oldtz );
3116                 }
3117
3118                 $text = preg_replace( '/~~~~~/', $d, $text );
3119                 $text = preg_replace( '/~~~~/', "$sigText $d", $text );
3120                 $text = preg_replace( '/~~~/', $sigText, $text );
3121
3122                 # Context links: [[|name]] and [[name (context)|]]
3123                 #
3124                 global $wgLegalTitleChars;
3125                 $tc = "[$wgLegalTitleChars]";
3126                 $np = str_replace( array( '(', ')' ), array( '', '' ), $tc ); # No parens
3127
3128                 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
3129                 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
3130
3131                 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/";             # [[page (context)|]]
3132                 $p2 = "/\[\[\\|({$tc}+)]]/";                                    # [[|page]]
3133                 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/";                # [[namespace:page|]] and [[:namespace:page|]]
3134                 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
3135                 $context = '';
3136                 $t = $this->mTitle->getText();
3137                 if ( preg_match( $conpat, $t, $m ) ) {
3138                         $context = $m[2];
3139                 }
3140                 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
3141                 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
3142                 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
3143
3144                 if ( '' == $context ) {
3145                         $text = preg_replace( $p2, '[[\\1]]', $text );
3146                 } else {
3147                         $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
3148                 }
3149
3150                 # Trim trailing whitespace
3151                 # MAG_END (__END__) tag allows for trailing
3152                 # whitespace to be deliberately included
3153                 $text = rtrim( $text );
3154                 $mw =& MagicWord::get( MAG_END );
3155                 $mw->matchAndRemove( $text );
3156
3157                 return $text;
3158         }
3159
3160         /**
3161          * Fetch the user's signature text, if any, and normalize to
3162          * validated, ready-to-insert wikitext.
3163          *
3164          * @param User $user
3165          * @return string
3166          * @access private
3167          */
3168         function getUserSig( &$user ) {
3169                 $name = $user->getName();
3170                 $nick = trim( $user->getOption( 'nickname' ) );
3171                 if ( '' == $nick ) {
3172                         $nick = $name;
3173                 }
3174
3175                 if( $user->getOption( 'fancysig' ) ) {
3176                         // A wikitext signature.
3177                         $valid = $this->validateSig( $nick );
3178                         if( $valid === false ) {
3179                                 // Fall back to default sig
3180                                 $nick = $name;
3181                                 wfDebug( "Parser::getUserSig: $name has bad XML tags in signature.\n" );
3182                         } else {
3183                                 return $nick;
3184                         }
3185                 }
3186
3187                 // Plain text linking to the user's homepage
3188                 global $wgContLang;
3189                 $page = $user->getUserPage();
3190                 return '[[' .
3191                         $page->getPrefixedText() .
3192                         "|" .
3193                         wfEscapeWikIText( $nick ) .
3194                         "]]";
3195         }
3196
3197         /**
3198          * We want to enforce two rules on wikitext sigs here:
3199          * 1) Expand any templates at save time (forced subst:)
3200          * 2) Check for unbalanced XML tags, and reject if so.
3201          *
3202          * @param string $text
3203          * @return mixed An expanded string, or false if invalid.
3204          *
3205          * @todo Run brace substitutions
3206          * @todo ?? Check for unbalanced '' and ''' quotes, etc
3207          */
3208         function validateSig( $text ) {
3209                 if( wfIsWellFormedXmlFragment( $text ) ) {
3210                         return $text;
3211                 } else {
3212                         return false;
3213                 }
3214         }
3215
3216         /**
3217          * Set up some variables which are usually set up in parse()
3218          * so that an external function can call some class members with confidence
3219          * @access public
3220          */
3221         function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
3222                 $this->mTitle =& $title;
3223                 $this->mOptions = $options;
3224                 $this->mOutputType = $outputType;
3225                 if ( $clearState ) {
3226                         $this->clearState();
3227                 }
3228         }
3229
3230         /**
3231          * Transform a MediaWiki message by replacing magic variables.
3232          *
3233          * @param string $text the text to transform
3234          * @param ParserOptions $options  options
3235          * @return string the text with variables substituted
3236          * @access public
3237          */
3238         function transformMsg( $text, $options ) {
3239                 global $wgTitle;
3240                 static $executing = false;
3241
3242                 # Guard against infinite recursion
3243                 if ( $executing ) {
3244                         return $text;
3245                 }
3246                 $executing = true;
3247
3248                 $this->mTitle = $wgTitle;
3249                 $this->mOptions = $options;
3250                 $this->mOutputType = OT_MSG;
3251                 $this->clearState();
3252                 $text = $this->replaceVariables( $text );
3253
3254                 $executing = false;
3255                 return $text;
3256         }
3257
3258         /**
3259          * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
3260          * Callback will be called with the text within
3261          * Transform and return the text within
3262          *
3263          * @access public
3264          *
3265          * @param mixed $tag The tag to use, e.g. 'hook' for <hook>
3266          * @param mixed $callback The callback function (and object) to use for the tag
3267          *
3268          * @return The old value of the mTagHooks array associated with the hook
3269          */
3270         function setHook( $tag, $callback ) {
3271                 $oldVal = @$this->mTagHooks[$tag];
3272                 $this->mTagHooks[$tag] = $callback;
3273
3274                 return $oldVal;
3275         }
3276
3277         /**
3278          * Replace <!--LINK--> link placeholders with actual links, in the buffer
3279          * Placeholders created in Skin::makeLinkObj()
3280          * Returns an array of links found, indexed by PDBK:
3281          *  0 - broken
3282          *  1 - normal link
3283          *  2 - stub
3284          * $options is a bit field, RLH_FOR_UPDATE to select for update
3285          */
3286         function replaceLinkHolders( &$text, $options = 0 ) {
3287                 global $wgUser, $wgLinkCache;
3288                 global $wgOutputReplace;
3289
3290                 $fname = 'Parser::replaceLinkHolders';
3291                 wfProfileIn( $fname );
3292
3293                 $pdbks = array();
3294                 $colours = array();
3295                 $sk = $this->mOptions->getSkin();
3296
3297                 if ( !empty( $this->mLinkHolders['namespaces'] ) ) {
3298                         wfProfileIn( $fname.'-check' );
3299                         $dbr =& wfGetDB( DB_SLAVE );
3300                         $page = $dbr->tableName( 'page' );
3301                         $threshold = $wgUser->getOption('stubthreshold');
3302
3303                         # Sort by namespace
3304                         asort( $this->mLinkHolders['namespaces'] );
3305
3306                         # Generate query
3307                         $query = false;
3308                         foreach ( $this->mLinkHolders['namespaces'] as $key => $val ) {
3309                                 # Make title object
3310                                 $title = $this->mLinkHolders['titles'][$key];
3311
3312                                 # Skip invalid entries.
3313                                 # Result will be ugly, but prevents crash.
3314                                 if ( is_null( $title ) ) {
3315                                         continue;
3316                                 }
3317                                 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
3318
3319                                 # Check if it's in the link cache already
3320                                 if ( $title->isAlwaysKnown() || $wgLinkCache->getGoodLinkID( $pdbk ) ) {
3321                                         $colours[$pdbk] = 1;
3322                                 } elseif ( $wgLinkCache->isBadLink( $pdbk ) ) {
3323                                         $colours[$pdbk] = 0;
3324                                 } else {
3325                                         # Not in the link cache, add it to the query
3326                                         if ( !isset( $current ) ) {
3327                                                 $current = $val;
3328                                                 $query =  "SELECT page_id, page_namespace, page_title";
3329                                                 if ( $threshold > 0 ) {
3330                                                         $query .= ', page_len, page_is_redirect';
3331                                                 }
3332                                                 $query .= " FROM $page WHERE (page_namespace=$val AND page_title IN(";
3333                                         } elseif ( $current != $val ) {
3334                                                 $current = $val;
3335                                                 $query .= ")) OR (page_namespace=$val AND page_title IN(";
3336                                         } else {
3337                                                 $query .= ', ';
3338                                         }
3339
3340                                         $query .= $dbr->addQuotes( $this->mLinkHolders['dbkeys'][$key] );
3341                                 }
3342                         }
3343                         if ( $query ) {
3344                                 $query .= '))';
3345                                 if ( $options & RLH_FOR_UPDATE ) {
3346                                         $query .= ' FOR UPDATE';
3347                                 }
3348
3349                                 $res = $dbr->query( $query, $fname );
3350
3351                                 # Fetch data and form into an associative array
3352                                 # non-existent = broken
3353                                 # 1 = known
3354                                 # 2 = stub
3355                                 while ( $s = $dbr->fetchObject($res) ) {
3356                                         $title = Title::makeTitle( $s->page_namespace, $s->page_title );
3357                                         $pdbk = $title->getPrefixedDBkey();
3358                                         $wgLinkCache->addGoodLinkObj( $s->page_id, $title );
3359
3360                                         if ( $threshold >  0 ) {
3361                                                 $size = $s->page_len;
3362                                                 if ( $s->page_is_redirect || $s->page_namespace != 0 || $size >= $threshold ) {
3363                                                         $colours[$pdbk] = 1;
3364                                                 } else {
3365                                                         $colours[$pdbk] = 2;
3366                                                 }
3367                                         } else {
3368                                                 $colours[$pdbk] = 1;
3369                                         }
3370                                 }
3371                         }
3372                         wfProfileOut( $fname.'-check' );
3373
3374                         # Construct search and replace arrays
3375                         wfProfileIn( $fname.'-construct' );
3376                         $wgOutputReplace = array();
3377                         foreach ( $this->mLinkHolders['namespaces'] as $key => $ns ) {
3378                                 $pdbk = $pdbks[$key];
3379                                 $searchkey = "<!--LINK $key-->";
3380                                 $title = $this->mLinkHolders['titles'][$key];
3381                                 if ( empty( $colours[$pdbk] ) ) {
3382                                         $wgLinkCache->addBadLinkObj( $title );
3383                                         $colours[$pdbk] = 0;
3384                                         $wgOutputReplace[$searchkey] = $sk->makeBrokenLinkObj( $title,
3385                                                                         $this->mLinkHolders['texts'][$key],
3386                                                                         $this->mLinkHolders['queries'][$key] );
3387                                 } elseif ( $colours[$pdbk] == 1 ) {
3388                                         $wgOutputReplace[$searchkey] = $sk->makeKnownLinkObj( $title,
3389                                                                         $this->mLinkHolders['texts'][$key],
3390                                                                         $this->mLinkHolders['queries'][$key] );
3391                                 } elseif ( $colours[$pdbk] == 2 ) {
3392                                         $wgOutputReplace[$searchkey] = $sk->makeStubLinkObj( $title,
3393                                                                         $this->mLinkHolders['texts'][$key],
3394                                                                         $this->mLinkHolders['queries'][$key] );
3395                                 }
3396                         }
3397                         wfProfileOut( $fname.'-construct' );
3398
3399                         # Do the thing
3400                         wfProfileIn( $fname.'-replace' );
3401
3402                         $text = preg_replace_callback(
3403                                 '/(<!--LINK .*?-->)/',
3404                                 "wfOutputReplaceMatches",
3405                                 $text);
3406
3407                         wfProfileOut( $fname.'-replace' );
3408                 }
3409
3410                 # Now process interwiki link holders
3411                 # This is quite a bit simpler than internal links
3412                 if ( !empty( $this->mInterwikiLinkHolders['texts'] ) ) {
3413                         wfProfileIn( $fname.'-interwiki' );
3414                         # Make interwiki link HTML
3415                         $wgOutputReplace = array();
3416                         foreach( $this->mInterwikiLinkHolders['texts'] as $key => $link ) {
3417                                 $title = $this->mInterwikiLinkHolders['titles'][$key];
3418                                 $wgOutputReplace[$key] = $sk->makeLinkObj( $title, $link );
3419                         }
3420
3421                         $text = preg_replace_callback(
3422                                 '/<!--IWLINK (.*?)-->/',
3423                                 "wfOutputReplaceMatches",
3424                                 $text );
3425                         wfProfileOut( $fname.'-interwiki' );
3426                 }
3427
3428                 wfProfileOut( $fname );
3429                 return $colours;
3430         }
3431
3432         /**
3433          * Replace <!--LINK--> link placeholders with plain text of links
3434          * (not HTML-formatted).
3435          * @param string $text
3436          * @return string
3437          */
3438         function replaceLinkHoldersText( $text ) {
3439                 global $wgUser, $wgLinkCache;
3440                 global $wgOutputReplace;
3441
3442                 $fname = 'Parser::replaceLinkHoldersText';
3443                 wfProfileIn( $fname );
3444
3445                 $text = preg_replace_callback(
3446                         '/<!--(LINK|IWLINK) (.*?)-->/',
3447                         array( &$this, 'replaceLinkHoldersTextCallback' ),
3448                         $text );
3449
3450                 wfProfileOut( $fname );
3451                 return $text;
3452         }
3453
3454         /**
3455          * @param array $matches
3456          * @return string
3457          * @access private
3458          */
3459         function replaceLinkHoldersTextCallback( $matches ) {
3460                 $type = $matches[1];
3461                 $key  = $matches[2];
3462                 if( $type == 'LINK' ) {
3463                         if( isset( $this->mLinkHolders['texts'][$key] ) ) {
3464                                 return $this->mLinkHolders['texts'][$key];
3465                         }
3466                 } elseif( $type == 'IWLINK' ) {
3467                         if( isset( $this->mInterwikiLinkHolders['texts'][$key] ) ) {
3468                                 return $this->mInterwikiLinkHolders['texts'][$key];
3469                         }
3470                 }
3471                 return $matches[0];
3472         }
3473
3474         /**
3475          * Renders an image gallery from a text with one line per image.
3476          * text labels may be given by using |-style alternative text. E.g.
3477          *   Image:one.jpg|The number "1"
3478          *   Image:tree.jpg|A tree
3479          * given as text will return the HTML of a gallery with two images,
3480          * labeled 'The number "1"' and
3481          * 'A tree'.
3482          *
3483          * @static
3484          */
3485         function renderImageGallery( $text ) {
3486                 # Setup the parser
3487                 global $wgUser, $wgTitle;
3488                 $parserOptions = ParserOptions::newFromUser( $wgUser );
3489                 $localParser = new Parser();
3490
3491                 global $wgLinkCache;
3492                 $ig = new ImageGallery();
3493                 $ig->setShowBytes( false );
3494                 $ig->setShowFilename( false );
3495                 $lines = explode( "\n", $text );
3496
3497                 foreach ( $lines as $line ) {
3498                         # match lines like these:
3499                         # Image:someimage.jpg|This is some image
3500                         preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
3501                         # Skip empty lines
3502                         if ( count( $matches ) == 0 ) {
3503                                 continue;
3504                         }
3505                         $nt = Title::newFromURL( $matches[1] );
3506                         if( is_null( $nt ) ) {
3507                                 # Bogus title. Ignore these so we don't bomb out later.
3508                                 continue;
3509                         }
3510                         if ( isset( $matches[3] ) ) {
3511                                 $label = $matches[3];
3512                         } else {
3513                                 $label = '';
3514                         }
3515
3516                         $html = $localParser->parse( $label , $wgTitle, $parserOptions );
3517                         $html = $html->mText;
3518
3519                         $ig->add( new Image( $nt ), $html );
3520                         $wgLinkCache->addImageLinkObj( $nt );
3521                 }
3522                 return $ig->toHTML();
3523         }
3524
3525         /**
3526          * Parse image options text and use it to make an image
3527          */
3528         function makeImage( &$nt, $options ) {
3529                 global $wgContLang, $wgUseImageResize, $wgUser;
3530
3531                 $align = '';
3532
3533                 # Check if the options text is of the form "options|alt text"
3534                 # Options are:
3535                 #  * thumbnail          make a thumbnail with enlarge-icon and caption, alignment depends on lang
3536                 #  * left               no resizing, just left align. label is used for alt= only
3537                 #  * right              same, but right aligned
3538                 #  * none               same, but not aligned
3539                 #  * ___px              scale to ___ pixels width, no aligning. e.g. use in taxobox
3540                 #  * center             center the image
3541                 #  * framed             Keep original image size, no magnify-button.
3542
3543                 $part = explode( '|', $options);
3544
3545                 $mwThumb  =& MagicWord::get( MAG_IMG_THUMBNAIL );
3546                 $mwManualThumb =& MagicWord::get( MAG_IMG_MANUALTHUMB );
3547                 $mwLeft   =& MagicWord::get( MAG_IMG_LEFT );
3548                 $mwRight  =& MagicWord::get( MAG_IMG_RIGHT );
3549                 $mwNone   =& MagicWord::get( MAG_IMG_NONE );
3550                 $mwWidth  =& MagicWord::get( MAG_IMG_WIDTH );
3551                 $mwCenter =& MagicWord::get( MAG_IMG_CENTER );
3552                 $mwFramed =& MagicWord::get( MAG_IMG_FRAMED );
3553                 $caption = '';
3554
3555                 $width = $height = $framed = $thumb = false;
3556                 $manual_thumb = '' ;
3557
3558                 foreach( $part as $key => $val ) {
3559                         if ( $wgUseImageResize && ! is_null( $mwThumb->matchVariableStartToEnd($val) ) ) {
3560                                 $thumb=true;
3561                         } elseif ( ! is_null( $match = $mwManualThumb->matchVariableStartToEnd($val) ) ) {
3562                                 # use manually specified thumbnail
3563                                 $thumb=true;
3564                                 $manual_thumb = $match;
3565                         } elseif ( ! is_null( $mwRight->matchVariableStartToEnd($val) ) ) {
3566                                 # remember to set an alignment, don't render immediately
3567                                 $align = 'right';
3568                         } elseif ( ! is_null( $mwLeft->matchVariableStartToEnd($val) ) ) {
3569                                 # remember to set an alignment, don't render immediately
3570                                 $align = 'left';
3571                         } elseif ( ! is_null( $mwCenter->matchVariableStartToEnd($val) ) ) {
3572                                 # remember to set an alignment, don't render immediately
3573                                 $align = 'center';
3574                         } elseif ( ! is_null( $mwNone->matchVariableStartToEnd($val) ) ) {
3575                                 # remember to set an alignment, don't render immediately
3576                                 $align = 'none';
3577                         } elseif ( $wgUseImageResize && ! is_null( $match = $mwWidth->matchVariableStartToEnd($val) ) ) {
3578                                 wfDebug( "MAG_IMG_WIDTH match: $match\n" );
3579                                 # $match is the image width in pixels
3580                                 if ( preg_match( '/^([0-9]*)x([0-9]*)$/', $match, $m ) ) {
3581                                         $width = intval( $m[1] );
3582                                         $height = intval( $m[2] );
3583                                 } else {
3584                                         $width = intval($match);
3585                                 }
3586                         } elseif ( ! is_null( $mwFramed->matchVariableStartToEnd($val) ) ) {
3587                                 $framed=true;
3588                         } else {
3589                                 $caption = $val;
3590                         }
3591                 }
3592                 # Strip bad stuff out of the alt text
3593                 $alt = $this->replaceLinkHoldersText( $caption );
3594                 $alt = Sanitizer::stripAllTags( $alt );
3595
3596                 # Linker does the rest
3597                 $sk =& $this->mOptions->getSkin();
3598                 return $sk->makeImageLinkObj( $nt, $caption, $alt, $align, $width, $height, $framed, $thumb, $manual_thumb );
3599         }
3600
3601         /**
3602          * Set a flag in the output object indicating that the content is dynamic and
3603          * shouldn't be cached.
3604          */
3605         function disableCache() {
3606                 $this->mOutput->mCacheTime = -1;
3607         }
3608
3609         /**
3610          * Callback from the Sanitizer for expanding items found in HTML attribute
3611          * values, so they can be safely tested and escaped.
3612          * @param string $text
3613          * @param array $args
3614          * @return string
3615          * @access private
3616          */
3617         function attributeStripCallback( &$text, $args ) {
3618                 $text = $this->replaceVariables( $text, $args );
3619                 $text = $this->unstripForHTML( $text );
3620                 return $text;
3621         }
3622
3623         function unstripForHTML( $text ) {
3624                 $text = $this->unstrip( $text, $this->mStripState );
3625                 $text = $this->unstripNoWiki( $text, $this->mStripState );
3626                 return $text;
3627         }
3628 }
3629
3630 /**
3631  * @todo document
3632  * @package MediaWiki
3633  */
3634 class ParserOutput
3635 {
3636         var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
3637         var $mCacheTime; # Timestamp on this article, or -1 for uncacheable. Used in ParserCache.
3638         var $mVersion;   # Compatibility check
3639         var $mTitleText; # title text of the chosen language variant
3640
3641         function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
3642                 $containsOldMagic = false, $titletext = '' )
3643         {
3644                 $this->mText = $text;
3645                 $this->mLanguageLinks = $languageLinks;
3646                 $this->mCategoryLinks = $categoryLinks;
3647                 $this->mContainsOldMagic = $containsOldMagic;
3648                 $this->mCacheTime = '';
3649                 $this->mVersion = MW_PARSER_VERSION;
3650                 $this->mTitleText = $titletext;
3651         }
3652
3653         function getText()                   { return $this->mText; }
3654         function getLanguageLinks()          { return $this->mLanguageLinks; }
3655         function getCategoryLinks()          { return array_keys( $this->mCategoryLinks ); }
3656         function getCacheTime()              { return $this->mCacheTime; }
3657         function getTitleText()              { return $this->mTitleText; }
3658         function containsOldMagic()          { return $this->mContainsOldMagic; }
3659         function setText( $text )            { return wfSetVar( $this->mText, $text ); }
3660         function setLanguageLinks( $ll )     { return wfSetVar( $this->mLanguageLinks, $ll ); }
3661         function setCategoryLinks( $cl )     { return wfSetVar( $this->mCategoryLinks, $cl ); }
3662         function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic, $com ); }
3663         function setCacheTime( $t )          { return wfSetVar( $this->mCacheTime, $t ); }
3664         function setTitleText( $t )          { return wfSetVar ($this->mTitleText, $t); }
3665
3666         function addCategoryLink( $c )       { $this->mCategoryLinks[$c] = 1; }
3667
3668         function merge( $other ) {
3669                 $this->mLanguageLinks = array_merge( $this->mLanguageLinks, $other->mLanguageLinks );
3670                 $this->mCategoryLinks = array_merge( $this->mCategoryLinks, $this->mLanguageLinks );
3671                 $this->mContainsOldMagic = $this->mContainsOldMagic || $other->mContainsOldMagic;
3672         }
3673
3674         /**
3675          * Return true if this cached output object predates the global or
3676          * per-article cache invalidation timestamps, or if it comes from
3677          * an incompatible older version.
3678          *
3679          * @param string $touched the affected article's last touched timestamp
3680          * @return bool
3681          * @access public
3682          */
3683         function expired( $touched ) {
3684                 global $wgCacheEpoch;
3685                 return $this->getCacheTime() == -1 || // parser says it's uncacheable
3686                        $this->getCacheTime() < $touched ||
3687                        $this->getCacheTime() <= $wgCacheEpoch ||
3688                        !isset( $this->mVersion ) ||
3689                        version_compare( $this->mVersion, MW_PARSER_VERSION, "lt" );
3690         }
3691 }
3692
3693 /**
3694  * Set options of the Parser
3695  * @todo document
3696  * @package MediaWiki
3697  */
3698 class ParserOptions
3699 {
3700         # All variables are private
3701         var $mUseTeX;                    # Use texvc to expand <math> tags
3702         var $mUseDynamicDates;           # Use DateFormatter to format dates
3703         var $mInterwikiMagic;            # Interlanguage links are removed and returned in an array
3704         var $mAllowExternalImages;       # Allow external images inline
3705         var $mAllowExternalImagesFrom;   # If not, any exception?
3706         var $mSkin;                      # Reference to the preferred skin
3707         var $mDateFormat;                # Date format index
3708         var $mEditSection;               # Create "edit section" links
3709         var $mNumberHeadings;            # Automatically number headings
3710         var $mAllowSpecialInclusion;     # Allow inclusion of special pages
3711
3712         function getUseTeX()                        { return $this->mUseTeX; }
3713         function getUseDynamicDates()               { return $this->mUseDynamicDates; }
3714         function getInterwikiMagic()                { return $this->mInterwikiMagic; }
3715         function getAllowExternalImages()           { return $this->mAllowExternalImages; }
3716         function getAllowExternalImagesFrom()       { return $this->mAllowExternalImagesFrom; }
3717         function &getSkin()                         { return $this->mSkin; }
3718         function getDateFormat()                    { return $this->mDateFormat; }
3719         function getEditSection()                   { return $this->mEditSection; }
3720         function getNumberHeadings()                { return $this->mNumberHeadings; }
3721         function getAllowSpecialInclusion()         { return $this->mAllowSpecialInclusion; }
3722
3723
3724         function setUseTeX( $x )                    { return wfSetVar( $this->mUseTeX, $x ); }
3725         function setUseDynamicDates( $x )           { return wfSetVar( $this->mUseDynamicDates, $x ); }
3726         function setInterwikiMagic( $x )            { return wfSetVar( $this->mInterwikiMagic, $x ); }
3727         function setAllowExternalImages( $x )       { return wfSetVar( $this->mAllowExternalImages, $x ); }
3728         function setAllowExternalImagesFrom( $x )   { return wfSetVar( $this->mAllowExternalImagesFrom, $x ); }
3729         function setDateFormat( $x )                { return wfSetVar( $this->mDateFormat, $x ); }
3730         function setEditSection( $x )               { return wfSetVar( $this->mEditSection, $x ); }
3731         function setNumberHeadings( $x )            { return wfSetVar( $this->mNumberHeadings, $x ); }
3732         function setAllowSpecialInclusion( $x )     { return wfSetVar( $this->mAllowSpecialInclusion, $x ); }
3733
3734         function setSkin( &$x ) { $this->mSkin =& $x; }
3735
3736         function ParserOptions() {
3737                 global $wgUser;
3738                 $this->initialiseFromUser( $wgUser );
3739         }
3740
3741         /**
3742          * Get parser options
3743          * @static
3744          */
3745         function newFromUser( &$user ) {
3746                 $popts = new ParserOptions;
3747                 $popts->initialiseFromUser( $user );
3748                 return $popts;
3749         }
3750
3751         /** Get user options */
3752         function initialiseFromUser( &$userInput ) {
3753                 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages,
3754                        $wgAllowExternalImagesFrom, $wgAllowSpecialInclusion;
3755                 $fname = 'ParserOptions::initialiseFromUser';
3756                 wfProfileIn( $fname );
3757                 if ( !$userInput ) {
3758                         $user = new User;
3759                         $user->setLoaded( true );
3760                 } else {
3761                         $user =& $userInput;
3762                 }
3763
3764                 $this->mUseTeX = $wgUseTeX;
3765                 $this->mUseDynamicDates = $wgUseDynamicDates;
3766                 $this->mInterwikiMagic = $wgInterwikiMagic;
3767                 $this->mAllowExternalImages = $wgAllowExternalImages;
3768                 $this->mAllowExternalImagesFrom = $wgAllowExternalImagesFrom;
3769                 wfProfileIn( $fname.'-skin' );
3770                 $this->mSkin =& $user->getSkin();
3771                 wfProfileOut( $fname.'-skin' );
3772                 $this->mDateFormat = $user->getOption( 'date' );
3773                 $this->mEditSection = true;
3774                 $this->mNumberHeadings = $user->getOption( 'numberheadings' );
3775                 $this->mAllowSpecialInclusion = $wgAllowSpecialInclusion;
3776                 wfProfileOut( $fname );
3777         }
3778 }
3779
3780 /**
3781  * Callback function used by Parser::replaceLinkHolders()
3782  * to substitute link placeholders.
3783  */
3784 function &wfOutputReplaceMatches( $matches ) {
3785         global $wgOutputReplace;
3786         return $wgOutputReplace[$matches[1]];
3787 }
3788
3789 /**
3790  * Return the total number of articles
3791  */
3792 function wfNumberOfArticles() {
3793         global $wgNumberOfArticles;
3794
3795         wfLoadSiteStats();
3796         return $wgNumberOfArticles;
3797 }
3798
3799 /**
3800  * Return the number of files
3801  */
3802 function wfNumberOfFiles() {
3803         $fname = 'Parser::wfNumberOfFiles';
3804
3805         wfProfileIn( $fname );
3806         $dbr =& wfGetDB( DB_SLAVE );
3807         $res = $dbr->selectField('image', 'COUNT(*)', array(), $fname );
3808         wfProfileOut( $fname );
3809
3810         return $res;
3811 }
3812
3813 /**
3814  * Get various statistics from the database
3815  * @private
3816  */
3817 function wfLoadSiteStats() {
3818         global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
3819         $fname = 'wfLoadSiteStats';
3820
3821         if ( -1 != $wgNumberOfArticles ) return;
3822         $dbr =& wfGetDB( DB_SLAVE );
3823         $s = $dbr->selectRow( 'site_stats',
3824                 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
3825                 array( 'ss_row_id' => 1 ), $fname
3826         );
3827
3828         if ( $s === false ) {
3829                 return;
3830         } else {
3831                 $wgTotalViews = $s->ss_total_views;
3832                 $wgTotalEdits = $s->ss_total_edits;
3833                 $wgNumberOfArticles = $s->ss_good_articles;
3834         }
3835 }
3836
3837 /**
3838  * Escape html tags
3839  * Basicly replacing " > and < with HTML entities ( &quot;, &gt;, &lt;)
3840  *
3841  * @param string $in Text that might contain HTML tags
3842  * @return string Escaped string
3843  */
3844 function wfEscapeHTMLTagsOnly( $in ) {
3845         return str_replace(
3846                 array( '"', '>', '<' ),
3847                 array( '&quot;', '&gt;', '&lt;' ),
3848                 $in );
3849 }
3850
3851 ?>