includes/Parser.php

   1 <?php
   2 /**
   3  * File for Parser and related classes
   4  *
   5  * @package MediaWiki
   6  * @subpackage Parser
   7  */
   8
   9 /**
  10  * Update this version number when the ParserOutput format
  11  * changes in an incompatible way, so the parser cache
  12  * can automatically discard old data.
  13  */
  14 define( 'MW_PARSER_VERSION', '1.6.1' );
  15
  16 /**
  17  * Variable substitution O(N^2) attack
  18  *
  19  * Without countermeasures, it would be possible to attack the parser by saving
  20  * a page filled with a large number of inclusions of large pages. The size of
  21  * the generated page would be proportional to the square of the input size.
  22  * Hence, we limit the number of inclusions of any given page, thus bringing any
  23  * attack back to O(N).
  24  */
  25
  26 define( 'MAX_INCLUDE_REPEAT', 100 );
  27 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
  28
  29 define( 'RLH_FOR_UPDATE', 1 );
  30
  31 # Allowed values for $mOutputType
  32 define( 'OT_HTML', 1 );
  33 define( 'OT_WIKI', 2 );
  34 define( 'OT_MSG' , 3 );
  35
  36 # string parameter for extractTags which will cause it
  37 # to strip HTML comments in addition to regular
  38 # <XML>-style tags. This should not be anything we
  39 # may want to use in wikisyntax
  40 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
  41
  42 # Constants needed for external link processing
  43 define( 'HTTP_PROTOCOLS', 'http:\/\/|https:\/\/' );
  44 # Everything except bracket, space, or control characters
  45 define( 'EXT_LINK_URL_CLASS', '[^][<>"\\x00-\\x20\\x7F]' );
  46 # Including space, but excluding newlines
  47 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x0a\\x0d]' );
  48 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
  49 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
  50 define( 'EXT_LINK_BRACKETED',  '/\[(\b(' . wfUrlProtocols() . ')'.
  51         EXT_LINK_URL_CLASS.'+) *('.EXT_LINK_TEXT_CLASS.'*?)\]/S' );
  52 define( 'EXT_IMAGE_REGEX',
  53         '/^('.HTTP_PROTOCOLS.')'.  # Protocol
  54         '('.EXT_LINK_URL_CLASS.'+)\\/'.  # Hostname and path
  55         '('.EXT_IMAGE_FNAME_CLASS.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS.')$/S' # Filename
  56 );
  57
  58 // State constants for the definition list colon extraction
  59 define( 'MW_COLON_STATE_TEXT', 0 );
  60 define( 'MW_COLON_STATE_TAG', 1 );
  61 define( 'MW_COLON_STATE_TAGSTART', 2 );
  62 define( 'MW_COLON_STATE_CLOSETAG', 3 );
  63 define( 'MW_COLON_STATE_TAGSLASH', 4 );
  64 define( 'MW_COLON_STATE_COMMENT', 5 );
  65 define( 'MW_COLON_STATE_COMMENTDASH', 6 );
  66 define( 'MW_COLON_STATE_COMMENTDASHDASH', 7 );
  67
  68 /**
  69  * PHP Parser
  70  *
  71  * Processes wiki markup
  72  *
  73  * <pre>
  74  * There are three main entry points into the Parser class:
  75  * parse()
  76  *   produces HTML output
  77  * preSaveTransform().
  78  *   produces altered wiki markup.
  79  * transformMsg()
  80  *   performs brace substitution on MediaWiki messages
  81  *
  82  * Globals used:
  83  *    objects:   $wgLang, $wgContLang
  84  *
  85  * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
  86  *
  87  * settings:
  88  *  $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
  89  *  $wgNamespacesWithSubpages, $wgAllowExternalImages*,
  90  *  $wgLocaltimezone, $wgAllowSpecialInclusion*
  91  *
  92  *  * only within ParserOptions
  93  * </pre>
  94  *
  95  * @package MediaWiki
  96  */
  97 class Parser
  98 {
  99         /**#@+
 100          * @private
 101          */
 102         # Persistent:
 103         var $mTagHooks, $mFunctionHooks;
 104
 105         # Cleared with clearState():
 106         var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
 107         var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
 108         var $mInterwikiLinkHolders, $mLinkHolders, $mUniqPrefix;
 109         var $mTemplates,        // cache of already loaded templates, avoids
 110                                 // multiple SQL queries for the same string
 111             $mTemplatePath;     // stores an unsorted hash of all the templates already loaded
 112                                 // in this path. Used for loop detection.
 113
 114         # Temporary
 115         # These are variables reset at least once per parse regardless of $clearState
 116         var $mOptions,      // ParserOptions object
 117                 $mTitle,        // Title context, used for self-link rendering and similar things
 118                 $mOutputType,   // Output type, one of the OT_xxx constants
 119                 $mRevisionId;   // ID to display in {{REVISIONID}} tags
 120
 121         /**#@-*/
 122
 123         /**
 124          * Constructor
 125          *
 126          * @public
 127          */
 128         function Parser() {
 129                 $this->mTagHooks = array();
 130                 $this->mFunctionHooks = array();
 131                 $this->mFunctionSynonyms = array();
 132                 $this->clearState();
 133                 $this->setHook( 'pre', array( $this, 'renderPreTag' ) );
 134         }
 135
 136         /**
 137          * Clear Parser state
 138          *
 139          * @private
 140          */
 141         function clearState() {
 142                 $this->mOutput = new ParserOutput;
 143                 $this->mAutonumber = 0;
 144                 $this->mLastSection = '';
 145                 $this->mDTopen = false;
 146                 $this->mVariables = false;
 147                 $this->mIncludeCount = array();
 148                 $this->mStripState = array();
 149                 $this->mArgStack = array();
 150                 $this->mInPre = false;
 151                 $this->mInterwikiLinkHolders = array(
 152                         'texts' => array(),
 153                         'titles' => array()
 154                 );
 155                 $this->mLinkHolders = array(
 156                         'namespaces' => array(),
 157                         'dbkeys' => array(),
 158                         'queries' => array(),
 159                         'texts' => array(),
 160                         'titles' => array()
 161                 );
 162                 $this->mRevisionId = null;
 163
 164                 /**
 165                  * Prefix for temporary replacement strings for the multipass parser.
 166                  * \x07 should never appear in input as it's disallowed in XML.
 167                  * Using it at the front also gives us a little extra robustness
 168                  * since it shouldn't match when butted up against identifier-like
 169                  * string constructs.
 170                  */
 171                 $this->mUniqPrefix = "\x07UNIQ" . Parser::getRandomString();
 172
 173                 # Clear these on every parse, bug 4549
 174                 $this->mTemplates = array();
 175                 $this->mTemplatePath = array();
 176
 177                 $this->mShowToc = true;
 178                 $this->mForceTocPosition = false;
 179
 180                 wfRunHooks( 'ParserClearState', array( &$this ) );
 181         }
 182
 183         /**
 184          * Accessor for mUniqPrefix.
 185          *
 186          * @public
 187          */
 188         function UniqPrefix() {
 189                 return $this->mUniqPrefix;
 190         }
 191
 192         /**
 193          * Convert wikitext to HTML
 194          * Do not call this function recursively.
 195          *
 196          * @private
 197          * @param string $text Text we want to parse
 198          * @param Title &$title A title object
 199          * @param array $options
 200          * @param boolean $linestart
 201          * @param boolean $clearState
 202          * @param int $revid number to pass in {{REVISIONID}}
 203          * @return ParserOutput a ParserOutput
 204          */
 205         function parse( $text, &$title, $options, $linestart = true, $clearState = true, $revid = null ) {
 206                 /**
 207                  * First pass--just handle <nowiki> sections, pass the rest off
 208                  * to internalParse() which does all the real work.
 209                  */
 210
 211                 global $wgUseTidy, $wgAlwaysUseTidy, $wgContLang;
 212                 $fname = 'Parser::parse';
 213                 wfProfileIn( $fname );
 214
 215                 if ( $clearState ) {
 216                         $this->clearState();
 217                 }
 218
 219                 $this->mOptions = $options;
 220                 $this->mTitle =& $title;
 221                 $this->mRevisionId = $revid;
 222                 $this->mOutputType = OT_HTML;
 223
 224                 //$text = $this->strip( $text, $this->mStripState );
 225                 // VOODOO MAGIC FIX! Sometimes the above segfaults in PHP5.
 226                 $x =& $this->mStripState;
 227
 228                 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$x ) );
 229                 $text = $this->strip( $text, $x );
 230                 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$x ) );
 231
 232                 # Hook to suspend the parser in this state
 233                 if ( !wfRunHooks( 'ParserBeforeInternalParse', array( &$this, &$text, &$x ) ) ) {
 234                         wfProfileOut( $fname );
 235                         return $text ;
 236                 }
 237
 238                 $text = $this->internalParse( $text );
 239
 240                 $text = $this->unstrip( $text, $this->mStripState );
 241
 242                 # Clean up special characters, only run once, next-to-last before doBlockLevels
 243                 $fixtags = array(
 244                         # french spaces, last one Guillemet-left
 245                         # only if there is something before the space
 246                         '/(.) (?=\\?|:|;|!|\\302\\273)/' => '\\1&nbsp;\\2',
 247                         # french spaces, Guillemet-right
 248                         '/(\\302\\253) /' => '\\1&nbsp;',
 249                 );
 250                 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
 251
 252                 # only once and last
 253                 $text = $this->doBlockLevels( $text, $linestart );
 254
 255                 $this->replaceLinkHolders( $text );
 256
 257                 # the position of the parserConvert() call should not be changed. it
 258                 # assumes that the links are all replaced and the only thing left
 259                 # is the <nowiki> mark.
 260                 # Side-effects: this calls $this->mOutput->setTitleText()
 261                 $text = $wgContLang->parserConvert( $text, $this );
 262
 263                 $text = $this->unstripNoWiki( $text, $this->mStripState );
 264
 265                 wfRunHooks( 'ParserBeforeTidy', array( &$this, &$text ) );
 266
 267                 $text = Sanitizer::normalizeCharReferences( $text );
 268
 269                 if (($wgUseTidy and $this->mOptions->mTidy) or $wgAlwaysUseTidy) {
 270                         $text = Parser::tidy($text);
 271                 } else {
 272                         # attempt to sanitize at least some nesting problems
 273                         # (bug #2702 and quite a few others)
 274                         $tidyregs = array(
 275                                 # ''Something [http://www.cool.com cool''] -->
 276                                 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
 277                                 '/(<([bi])>)(<([bi])>)?([^<]*)(<\/?a[^<]*>)([^<]*)(<\/\\4>)?(<\/\\2>)/' =>
 278                                 '\\1\\3\\5\\8\\9\\6\\1\\3\\7\\8\\9',
 279                                 # fix up an anchor inside another anchor, only
 280                                 # at least for a single single nested link (bug 3695)
 281                                 '/(<a[^>]+>)([^<]*)(<a[^>]+>[^<]*)<\/a>(.*)<\/a>/' =>
 282                                 '\\1\\2</a>\\3</a>\\1\\4</a>',
 283                                 # fix div inside inline elements- doBlockLevels won't wrap a line which
 284                                 # contains a div, so fix it up here; replace
 285                                 # div with escaped text
 286                                 '/(<([aib]) [^>]+>)([^<]*)(<div([^>]*)>)(.*)(<\/div>)([^<]*)(<\/\\2>)/' =>
 287                                 '\\1\\3&lt;div\\5&gt;\\6&lt;/div&gt;\\8\\9',
 288                                 # remove empty italic or bold tag pairs, some
 289                                 # introduced by rules above
 290                                 '/<([bi])><\/\\1>/' => ''
 291                         );
 292
 293                         $text = preg_replace(
 294                                 array_keys( $tidyregs ),
 295                                 array_values( $tidyregs ),
 296                                 $text );
 297                 }
 298
 299                 wfRunHooks( 'ParserAfterTidy', array( &$this, &$text ) );
 300
 301                 $this->mOutput->setText( $text );
 302                 wfProfileOut( $fname );
 303
 304                 return $this->mOutput;
 305         }
 306
 307         /**
 308          * Get a random string
 309          *
 310          * @private
 311          * @static
 312          */
 313         function getRandomString() {
 314                 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
 315         }
 316
 317         function &getTitle() { return $this->mTitle; }
 318         function getOptions() { return $this->mOptions; }
 319
 320         /**
 321          * Replaces all occurrences of HTML-style comments and the given tags
 322          * in the text with a random marker and returns teh next text. The output
 323          * parameter $matches will be an associative array filled with data in
 324          * the form:
 325          *   'UNIQ-xxxxx' => array(
 326          *     'element',
 327          *     'tag content',
 328          *     array( 'param' => 'x' ),
 329          *     '<element param="x">tag content</element>' ) )
 330          *
 331          * @param $elements list of element names. Comments are always extracted.
 332          * @param $text Source text string.
 333          * @param $uniq_prefix
 334          *
 335          * @private
 336          * @static
 337          */
 338         function extractTagsAndParams($elements, $text, &$matches, $uniq_prefix = ''){
 339                 $rand = Parser::getRandomString();
 340                 $n = 1;
 341                 $stripped = '';
 342                 $matches = array();
 343
 344                 $taglist = implode( '|', $elements );
 345                 $start = "/<($taglist)(\\s+[^>]*?|\\s*?)(\/?>)|<(!--)/i";
 346
 347                 while ( '' != $text ) {
 348                         $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE );
 349                         $stripped .= $p[0];
 350                         if( count( $p ) < 5 ) {
 351                                 break;
 352                         }
 353                         if( count( $p ) > 5 ) {
 354                                 // comment
 355                                 $element    = $p[4];
 356                                 $attributes = '';
 357                                 $close      = '';
 358                                 $inside     = $p[5];
 359                         } else {
 360                                 // tag
 361                                 $element    = $p[1];
 362                                 $attributes = $p[2];
 363                                 $close      = $p[3];
 364                                 $inside     = $p[4];
 365                         }
 366
 367                         $marker = "$uniq_prefix-$element-$rand" . sprintf('%08X', $n++) . '-QINU';
 368                         $stripped .= $marker;
 369
 370                         if ( $close === '/>' ) {
 371                                 // Empty element tag, <tag />
 372                                 $content = null;
 373                                 $text = $inside;
 374                                 $tail = null;
 375                         } else {
 376                                 if( $element == '!--' ) {
 377                                         $end = '/(-->)/';
 378                                 } else {
 379                                         $end = "/(<\\/$element\\s*>)/i";
 380                                 }
 381                                 $q = preg_split( $end, $inside, 2, PREG_SPLIT_DELIM_CAPTURE );
 382                                 $content = $q[0];
 383                                 if( count( $q ) < 3 ) {
 384                                         # No end tag -- let it run out to the end of the text.
 385                                         $tail = '';
 386                                         $text = '';
 387                                 } else {
 388                                         $tail = $q[1];
 389                                         $text = $q[2];
 390                                 }
 391                         }
 392
 393                         $matches[$marker] = array( $element,
 394                                 $content,
 395                                 Sanitizer::decodeTagAttributes( $attributes ),
 396                                 "<$element$attributes$close$content$tail" );
 397                 }
 398                 return $stripped;
 399         }
 400
 401         /**
 402          * Strips and renders nowiki, pre, math, hiero
 403          * If $render is set, performs necessary rendering operations on plugins
 404          * Returns the text, and fills an array with data needed in unstrip()
 405          * If the $state is already a valid strip state, it adds to the state
 406          *
 407          * @param bool $stripcomments when set, HTML comments <!-- like this -->
 408          *  will be stripped in addition to other tags. This is important
 409          *  for section editing, where these comments cause confusion when
 410          *  counting the sections in the wikisource
 411          *
 412          * @param array dontstrip contains tags which should not be stripped;
 413          *  used to prevent stipping of <gallery> when saving (fixes bug 2700)
 414          *
 415          * @private
 416          */
 417         function strip( $text, &$state, $stripcomments = false , $dontstrip = array () ) {
 418                 $render = ($this->mOutputType == OT_HTML);
 419
 420                 # Replace any instances of the placeholders
 421                 $uniq_prefix = $this->mUniqPrefix;
 422                 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
 423                 $commentState = array();
 424
 425                 $elements = array_merge(
 426                         array( 'nowiki', 'gallery' ),
 427                         array_keys( $this->mTagHooks ) );
 428                 global $wgRawHtml;
 429                 if( $wgRawHtml ) {
 430                         $elements[] = 'html';
 431                 }
 432                 if( $this->mOptions->getUseTeX() ) {
 433                         $elements[] = 'math';
 434                 }
 435
 436                 # Removing $dontstrip tags from $elements list (currently only 'gallery', fixing bug 2700)
 437                 foreach ( $elements AS $k => $v ) {
 438                         if ( !in_array ( $v , $dontstrip ) ) continue;
 439                         unset ( $elements[$k] );
 440                 }
 441
 442                 $matches = array();
 443                 $text = Parser::extractTagsAndParams( $elements, $text, $matches, $uniq_prefix );
 444
 445                 foreach( $matches as $marker => $data ) {
 446                         list( $element, $content, $params, $tag ) = $data;
 447                         if( $render ) {
 448                                 $tagName = strtolower( $element );
 449                                 switch( $tagName ) {
 450                                 case '!--':
 451                                         // Comment
 452                                         if( substr( $tag, -3 ) == '-->' ) {
 453                                                 $output = $tag;
 454                                         } else {
 455                                                 // Unclosed comment in input.
 456                                                 // Close it so later stripping can remove it
 457                                                 $output = "$tag-->";
 458                                         }
 459                                         break;
 460                                 case 'html':
 461                                         if( $wgRawHtml ) {
 462                                                 $output = $content;
 463                                                 break;
 464                                         }
 465                                         // Shouldn't happen otherwise. :)
 466                                 case 'nowiki':
 467                                         $output = wfEscapeHTMLTagsOnly( $content );
 468                                         break;
 469                                 case 'math':
 470                                         $output = MathRenderer::renderMath( $content );
 471                                         break;
 472                                 case 'gallery':
 473                                         $output = $this->renderImageGallery( $content, $params );
 474                                         break;
 475                                 default:
 476                                         if( isset( $this->mTagHooks[$tagName] ) ) {
 477                                                 $output = call_user_func_array( $this->mTagHooks[$tagName],
 478                                                         array( $content, $params, $this ) );
 479                                         } else {
 480                                                 throw new MWException( "Invalid call hook $element" );
 481                                         }
 482                                 }
 483                         } else {
 484                                 // Just stripping tags; keep the source
 485                                 $output = $tag;
 486                         }
 487                         if( !$stripcomments && $element == '!--' ) {
 488                                 $commentState[$marker] = $output;
 489                         } else {
 490                                 $state[$element][$marker] = $output;
 491                         }
 492                 }
 493
 494                 # Unstrip comments unless explicitly told otherwise.
 495                 # (The comments are always stripped prior to this point, so as to
 496                 # not invoke any extension tags / parser hooks contained within
 497                 # a comment.)
 498                 if ( !$stripcomments ) {
 499                         // Put them all back and forget them
 500                         $text = strtr( $text, $commentState );
 501                 }
 502
 503                 return $text;
 504         }
 505
 506         /**
 507          * Restores pre, math, and other extensions removed by strip()
 508          *
 509          * always call unstripNoWiki() after this one
 510          * @private
 511          */
 512         function unstrip( $text, &$state ) {
 513                 if ( !is_array( $state ) ) {
 514                         return $text;
 515                 }
 516
 517                 $replacements = array();
 518                 foreach( $state as $tag => $contentDict ) {
 519                         if( $tag != 'nowiki' && $tag != 'html' ) {
 520                                 foreach( $contentDict as $uniq => $content ) {
 521                                         $replacements[$uniq] = $content;
 522                                 }
 523                         }
 524                 }
 525                 $text = strtr( $text, $replacements );
 526
 527                 return $text;
 528         }
 529
 530         /**
 531          * Always call this after unstrip() to preserve the order
 532          *
 533          * @private
 534          */
 535         function unstripNoWiki( $text, &$state ) {
 536                 if ( !is_array( $state ) ) {
 537                         return $text;
 538                 }
 539
 540                 $replacements = array();
 541                 foreach( $state as $tag => $contentDict ) {
 542                         if( $tag == 'nowiki' || $tag == 'html' ) {
 543                                 foreach( $contentDict as $uniq => $content ) {
 544                                         $replacements[$uniq] = $content;
 545                                 }
 546                         }
 547                 }
 548                 $text = strtr( $text, $replacements );
 549
 550                 return $text;
 551         }
 552
 553         /**
 554          * Add an item to the strip state
 555          * Returns the unique tag which must be inserted into the stripped text
 556          * The tag will be replaced with the original text in unstrip()
 557          *
 558          * @private
 559          */
 560         function insertStripItem( $text, &$state ) {
 561                 $rnd = $this->mUniqPrefix . '-item' . Parser::getRandomString();
 562                 if ( !$state ) {
 563                         $state = array();
 564                 }
 565                 $state['item'][$rnd] = $text;
 566                 return $rnd;
 567         }
 568
 569         /**
 570          * Interface with html tidy, used if $wgUseTidy = true.
 571          * If tidy isn't able to correct the markup, the original will be
 572          * returned in all its glory with a warning comment appended.
 573          *
 574          * Either the external tidy program or the in-process tidy extension
 575          * will be used depending on availability. Override the default
 576          * $wgTidyInternal setting to disable the internal if it's not working.
 577          *
 578          * @param string $text Hideous HTML input
 579          * @return string Corrected HTML output
 580          * @public
 581          * @static
 582          */
 583         function tidy( $text ) {
 584                 global $wgTidyInternal;
 585                 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
 586 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
 587 '<head><title>test</title></head><body>'.$text.'</body></html>';
 588                 if( $wgTidyInternal ) {
 589                         $correctedtext = Parser::internalTidy( $wrappedtext );
 590                 } else {
 591                         $correctedtext = Parser::externalTidy( $wrappedtext );
 592                 }
 593                 if( is_null( $correctedtext ) ) {
 594                         wfDebug( "Tidy error detected!\n" );
 595                         return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
 596                 }
 597                 return $correctedtext;
 598         }
 599
 600         /**
 601          * Spawn an external HTML tidy process and get corrected markup back from it.
 602          *
 603          * @private
 604          * @static
 605          */
 606         function externalTidy( $text ) {
 607                 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
 608                 $fname = 'Parser::externalTidy';
 609                 wfProfileIn( $fname );
 610
 611                 $cleansource = '';
 612                 $opts = ' -utf8';
 613
 614                 $descriptorspec = array(
 615                         0 => array('pipe', 'r'),
 616                         1 => array('pipe', 'w'),
 617                         2 => array('file', '/dev/null', 'a')
 618                 );
 619                 $pipes = array();
 620                 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
 621                 if (is_resource($process)) {
 622                         // Theoretically, this style of communication could cause a deadlock
 623                         // here. If the stdout buffer fills up, then writes to stdin could
 624                         // block. This doesn't appear to happen with tidy, because tidy only
 625                         // writes to stdout after it's finished reading from stdin. Search
 626                         // for tidyParseStdin and tidySaveStdout in console/tidy.c
 627                         fwrite($pipes[0], $text);
 628                         fclose($pipes[0]);
 629                         while (!feof($pipes[1])) {
 630                                 $cleansource .= fgets($pipes[1], 1024);
 631                         }
 632                         fclose($pipes[1]);
 633                         proc_close($process);
 634                 }
 635
 636                 wfProfileOut( $fname );
 637
 638                 if( $cleansource == '' && $text != '') {
 639                         // Some kind of error happened, so we couldn't get the corrected text.
 640                         // Just give up; we'll use the source text and append a warning.
 641                         return null;
 642                 } else {
 643                         return $cleansource;
 644                 }
 645         }
 646
 647         /**
 648          * Use the HTML tidy PECL extension to use the tidy library in-process,
 649          * saving the overhead of spawning a new process. Currently written to
 650          * the PHP 4.3.x version of the extension, may not work on PHP 5.
 651          *
 652          * 'pear install tidy' should be able to compile the extension module.
 653          *
 654          * @private
 655          * @static
 656          */
 657         function internalTidy( $text ) {
 658                 global $wgTidyConf;
 659                 $fname = 'Parser::internalTidy';
 660                 wfProfileIn( $fname );
 661
 662                 tidy_load_config( $wgTidyConf );
 663                 tidy_set_encoding( 'utf8' );
 664                 tidy_parse_string( $text );
 665                 tidy_clean_repair();
 666                 if( tidy_get_status() == 2 ) {
 667                         // 2 is magic number for fatal error
 668                         // http://www.php.net/manual/en/function.tidy-get-status.php
 669                         $cleansource = null;
 670                 } else {
 671                         $cleansource = tidy_get_output();
 672                 }
 673                 wfProfileOut( $fname );
 674                 return $cleansource;
 675         }
 676
 677         /**
 678          * parse the wiki syntax used to render tables
 679          *
 680          * @private
 681          */
 682         function doTableStuff ( $t ) {
 683                 $fname = 'Parser::doTableStuff';
 684                 wfProfileIn( $fname );
 685
 686                 $t = explode ( "\n" , $t ) ;
 687                 $td = array () ; # Is currently a td tag open?
 688                 $ltd = array () ; # Was it TD or TH?
 689                 $tr = array () ; # Is currently a tr tag open?
 690                 $ltr = array () ; # tr attributes
 691                 $has_opened_tr = array(); # Did this table open a <tr> element?
 692                 $indent_level = 0; # indent level of the table
 693                 foreach ( $t AS $k => $x )
 694                 {
 695                         $x = trim ( $x ) ;
 696                         $fc = substr ( $x , 0 , 1 ) ;
 697                         if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
 698                                 $indent_level = strlen( $matches[1] );
 699
 700                                 $attributes = $this->unstripForHTML( $matches[2] );
 701
 702                                 $t[$k] = str_repeat( '<dl><dd>', $indent_level ) .
 703                                         '<table' . Sanitizer::fixTagAttributes ( $attributes, 'table' ) . '>' ;
 704                                 array_push ( $td , false ) ;
 705                                 array_push ( $ltd , '' ) ;
 706                                 array_push ( $tr , false ) ;
 707                                 array_push ( $ltr , '' ) ;
 708                                 array_push ( $has_opened_tr, false );
 709                         }
 710                         else if ( count ( $td ) == 0 ) { } # Don't do any of the following
 711                         else if ( '|}' == substr ( $x , 0 , 2 ) ) {
 712                                 $z = "</table>" . substr ( $x , 2);
 713                                 $l = array_pop ( $ltd ) ;
 714                                 if ( !array_pop ( $has_opened_tr ) ) $z = "<tr><td></td></tr>" . $z ;
 715                                 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 716                                 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 717                                 array_pop ( $ltr ) ;
 718                                 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
 719                         }
 720                         else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
 721                                 $x = substr ( $x , 1 ) ;
 722                                 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
 723                                 $z = '' ;
 724                                 $l = array_pop ( $ltd ) ;
 725                                 array_pop ( $has_opened_tr );
 726                                 array_push ( $has_opened_tr , true ) ;
 727                                 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
 728                                 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 729                                 array_pop ( $ltr ) ;
 730                                 $t[$k] = $z ;
 731                                 array_push ( $tr , false ) ;
 732                                 array_push ( $td , false ) ;
 733                                 array_push ( $ltd , '' ) ;
 734                                 $attributes = $this->unstripForHTML( $x );
 735                                 array_push ( $ltr , Sanitizer::fixTagAttributes ( $attributes, 'tr' ) ) ;
 736                         }
 737                         else if ( '|' == $fc || '!' == $fc || '|+' == substr ( $x , 0 , 2 ) ) { # Caption
 738                                 # $x is a table row
 739                                 if ( '|+' == substr ( $x , 0 , 2 ) ) {
 740                                         $fc = '+' ;
 741                                         $x = substr ( $x , 1 ) ;
 742                                 }
 743                                 $after = substr ( $x , 1 ) ;
 744                                 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
 745
 746                                 // Split up multiple cells on the same line.
 747                                 // FIXME: This can result in improper nesting of tags processed
 748                                 // by earlier parser steps, but should avoid splitting up eg
 749                                 // attribute values containing literal "||".
 750                                 $after = wfExplodeMarkup( '||', $after );
 751
 752                                 $t[$k] = '' ;
 753
 754                                 # Loop through each table cell
 755                                 foreach ( $after AS $theline )
 756                                 {
 757                                         $z = '' ;
 758                                         if ( $fc != '+' )
 759                                         {
 760                                                 $tra = array_pop ( $ltr ) ;
 761                                                 if ( !array_pop ( $tr ) ) $z = '<tr'.$tra.">\n" ;
 762                                                 array_push ( $tr , true ) ;
 763                                                 array_push ( $ltr , '' ) ;
 764                                                 array_pop ( $has_opened_tr );
 765                                                 array_push ( $has_opened_tr , true ) ;
 766                                         }
 767
 768                                         $l = array_pop ( $ltd ) ;
 769                                         if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
 770                                         if ( $fc == '|' ) $l = 'td' ;
 771                                         else if ( $fc == '!' ) $l = 'th' ;
 772                                         else if ( $fc == '+' ) $l = 'caption' ;
 773                                         else $l = '' ;
 774                                         array_push ( $ltd , $l ) ;
 775
 776                                         # Cell parameters
 777                                         $y = explode ( '|' , $theline , 2 ) ;
 778                                         # Note that a '|' inside an invalid link should not
 779                                         # be mistaken as delimiting cell parameters
 780                                         if ( strpos( $y[0], '[[' ) !== false ) {
 781                                                 $y = array ($theline);
 782                                         }
 783                                         if ( count ( $y ) == 1 )
 784                                                 $y = "{$z}<{$l}>{$y[0]}" ;
 785                                         else {
 786                                                 $attributes = $this->unstripForHTML( $y[0] );
 787                                                 $y = "{$z}<{$l}".Sanitizer::fixTagAttributes($attributes, $l).">{$y[1]}" ;
 788                                         }
 789                                         $t[$k] .= $y ;
 790                                         array_push ( $td , true ) ;
 791                                 }
 792                         }
 793                 }
 794
 795                 # Closing open td, tr && table
 796                 while ( count ( $td ) > 0 )
 797                 {
 798                         $l = array_pop ( $ltd ) ;
 799                         if ( array_pop ( $td ) ) $t[] = '</td>' ;
 800                         if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
 801                         if ( !array_pop ( $has_opened_tr ) ) $t[] = "<tr><td></td></tr>" ;
 802                         $t[] = '</table>' ;
 803                 }
 804
 805                 $t = implode ( "\n" , $t ) ;
 806                 # special case: don't return empty table
 807                 if($t == "<table>\n<tr><td></td></tr>\n</table>")
 808                         $t = '';
 809                 wfProfileOut( $fname );
 810                 return $t ;
 811         }
 812
 813         /**
 814          * Helper function for parse() that transforms wiki markup into
 815          * HTML. Only called for $mOutputType == OT_HTML.
 816          *
 817          * @private
 818          */
 819         function internalParse( $text ) {
 820                 $args = array();
 821                 $isMain = true;
 822                 $fname = 'Parser::internalParse';
 823                 wfProfileIn( $fname );
 824
 825                 # Remove <noinclude> tags and <includeonly> sections
 826                 $text = strtr( $text, array( '<onlyinclude>' => '' , '</onlyinclude>' => '' ) );
 827                 $text = strtr( $text, array( '<noinclude>' => '', '</noinclude>' => '') );
 828                 $text = preg_replace( '/<includeonly>.*?<\/includeonly>/s', '', $text );
 829
 830                 $text = Sanitizer::removeHTMLtags( $text, array( &$this, 'attributeStripCallback' ) );
 831
 832                 $text = $this->replaceVariables( $text, $args );
 833
 834                 // Tables need to come after variable replacement for things to work
 835                 // properly; putting them before other transformations should keep
 836                 // exciting things like link expansions from showing up in surprising
 837                 // places.
 838                 $text = $this->doTableStuff( $text );
 839
 840                 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
 841
 842                 $text = $this->stripToc( $text );
 843                 $this->stripNoGallery( $text );
 844                 $text = $this->doHeadings( $text );
 845                 if($this->mOptions->getUseDynamicDates()) {
 846                         $df =& DateFormatter::getInstance();
 847                         $text = $df->reformat( $this->mOptions->getDateFormat(), $text );
 848                 }
 849                 $text = $this->doAllQuotes( $text );
 850                 $text = $this->replaceInternalLinks( $text );
 851                 $text = $this->replaceExternalLinks( $text );
 852
 853                 # replaceInternalLinks may sometimes leave behind
 854                 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
 855                 $text = str_replace($this->mUniqPrefix."NOPARSE", "", $text);
 856
 857                 $text = $this->doMagicLinks( $text );
 858                 $text = $this->formatHeadings( $text, $isMain );
 859
 860                 wfProfileOut( $fname );
 861                 return $text;
 862         }
 863
 864         /**
 865          * Replace special strings like "ISBN xxx" and "RFC xxx" with
 866          * magic external links.
 867          *
 868          * @private
 869          */
 870         function &doMagicLinks( &$text ) {
 871                 $text = $this->magicISBN( $text );
 872                 $text = $this->magicRFC( $text, 'RFC ', 'rfcurl' );
 873                 $text = $this->magicRFC( $text, 'PMID ', 'pubmedurl' );
 874                 return $text;
 875         }
 876
 877         /**
 878          * Parse headers and return html
 879          *
 880          * @private
 881          */
 882         function doHeadings( $text ) {
 883                 $fname = 'Parser::doHeadings';
 884                 wfProfileIn( $fname );
 885                 for ( $i = 6; $i >= 1; --$i ) {
 886                         $h = str_repeat( '=', $i );
 887                         $text = preg_replace( "/^{$h}(.+){$h}\\s*$/m",
 888                           "<h{$i}>\\1</h{$i}>\\2", $text );
 889                 }
 890                 wfProfileOut( $fname );
 891                 return $text;
 892         }
 893
 894         /**
 895          * Replace single quotes with HTML markup
 896          * @private
 897          * @return string the altered text
 898          */
 899         function doAllQuotes( $text ) {
 900                 $fname = 'Parser::doAllQuotes';
 901                 wfProfileIn( $fname );
 902                 $outtext = '';
 903                 $lines = explode( "\n", $text );
 904                 foreach ( $lines as $line ) {
 905                         $outtext .= $this->doQuotes ( $line ) . "\n";
 906                 }
 907                 $outtext = substr($outtext, 0,-1);
 908                 wfProfileOut( $fname );
 909                 return $outtext;
 910         }
 911
 912         /**
 913          * Helper function for doAllQuotes()
 914          * @private
 915          */
 916         function doQuotes( $text ) {
 917                 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE );
 918                 if ( count( $arr ) == 1 )
 919                         return $text;
 920                 else
 921                 {
 922                         # First, do some preliminary work. This may shift some apostrophes from
 923                         # being mark-up to being text. It also counts the number of occurrences
 924                         # of bold and italics mark-ups.
 925                         $i = 0;
 926                         $numbold = 0;
 927                         $numitalics = 0;
 928                         foreach ( $arr as $r )
 929                         {
 930                                 if ( ( $i % 2 ) == 1 )
 931                                 {
 932                                         # If there are ever four apostrophes, assume the first is supposed to
 933                                         # be text, and the remaining three constitute mark-up for bold text.
 934                                         if ( strlen( $arr[$i] ) == 4 )
 935                                         {
 936                                                 $arr[$i-1] .= "'";
 937                                                 $arr[$i] = "'''";
 938                                         }
 939                                         # If there are more than 5 apostrophes in a row, assume they're all
 940                                         # text except for the last 5.
 941                                         else if ( strlen( $arr[$i] ) > 5 )
 942                                         {
 943                                                 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
 944                                                 $arr[$i] = "'''''";
 945                                         }
 946                                         # Count the number of occurrences of bold and italics mark-ups.
 947                                         # We are not counting sequences of five apostrophes.
 948                                         if ( strlen( $arr[$i] ) == 2 ) $numitalics++;  else
 949                                         if ( strlen( $arr[$i] ) == 3 ) $numbold++;     else
 950                                         if ( strlen( $arr[$i] ) == 5 ) { $numitalics++; $numbold++; }
 951                                 }
 952                                 $i++;
 953                         }
 954
 955                         # If there is an odd number of both bold and italics, it is likely
 956                         # that one of the bold ones was meant to be an apostrophe followed
 957                         # by italics. Which one we cannot know for certain, but it is more
 958                         # likely to be one that has a single-letter word before it.
 959                         if ( ( $numbold % 2 == 1 ) && ( $numitalics % 2 == 1 ) )
 960                         {
 961                                 $i = 0;
 962                                 $firstsingleletterword = -1;
 963                                 $firstmultiletterword = -1;
 964                                 $firstspace = -1;
 965                                 foreach ( $arr as $r )
 966                                 {
 967                                         if ( ( $i % 2 == 1 ) and ( strlen( $r ) == 3 ) )
 968                                         {
 969                                                 $x1 = substr ($arr[$i-1], -1);
 970                                                 $x2 = substr ($arr[$i-1], -2, 1);
 971                                                 if ($x1 == ' ') {
 972                                                         if ($firstspace == -1) $firstspace = $i;
 973                                                 } else if ($x2 == ' ') {
 974                                                         if ($firstsingleletterword == -1) $firstsingleletterword = $i;
 975                                                 } else {
 976                                                         if ($firstmultiletterword == -1) $firstmultiletterword = $i;
 977                                                 }
 978                                         }
 979                                         $i++;
 980                                 }
 981
 982                                 # If there is a single-letter word, use it!
 983                                 if ($firstsingleletterword > -1)
 984                                 {
 985                                         $arr [ $firstsingleletterword ] = "''";
 986                                         $arr [ $firstsingleletterword-1 ] .= "'";
 987                                 }
 988                                 # If not, but there's a multi-letter word, use that one.
 989                                 else if ($firstmultiletterword > -1)
 990                                 {
 991                                         $arr [ $firstmultiletterword ] = "''";
 992                                         $arr [ $firstmultiletterword-1 ] .= "'";
 993                                 }
 994                                 # ... otherwise use the first one that has neither.
 995                                 # (notice that it is possible for all three to be -1 if, for example,
 996                                 # there is only one pentuple-apostrophe in the line)
 997                                 else if ($firstspace > -1)
 998                                 {
 999                                         $arr [ $firstspace ] = "''";
1000                                         $arr [ $firstspace-1 ] .= "'";
1001                                 }
1002                         }
1003
1004                         # Now let's actually convert our apostrophic mush to HTML!
1005                         $output = '';
1006                         $buffer = '';
1007                         $state = '';
1008                         $i = 0;
1009                         foreach ($arr as $r)
1010                         {
1011                                 if (($i % 2) == 0)
1012                                 {
1013                                         if ($state == 'both')
1014                                                 $buffer .= $r;
1015                                         else
1016                                                 $output .= $r;
1017                                 }
1018                                 else
1019                                 {
1020                                         if (strlen ($r) == 2)
1021                                         {
1022                                                 if ($state == 'i')
1023                                                 { $output .= '</i>'; $state = ''; }
1024                                                 else if ($state == 'bi')
1025                                                 { $output .= '</i>'; $state = 'b'; }
1026                                                 else if ($state == 'ib')
1027                                                 { $output .= '</b></i><b>'; $state = 'b'; }
1028                                                 else if ($state == 'both')
1029                                                 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
1030                                                 else # $state can be 'b' or ''
1031                                                 { $output .= '<i>'; $state .= 'i'; }
1032                                         }
1033                                         else if (strlen ($r) == 3)
1034                                         {
1035                                                 if ($state == 'b')
1036                                                 { $output .= '</b>'; $state = ''; }
1037                                                 else if ($state == 'bi')
1038                                                 { $output .= '</i></b><i>'; $state = 'i'; }
1039                                                 else if ($state == 'ib')
1040                                                 { $output .= '</b>'; $state = 'i'; }
1041                                                 else if ($state == 'both')
1042                                                 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
1043                                                 else # $state can be 'i' or ''
1044                                                 { $output .= '<b>'; $state .= 'b'; }
1045                                         }
1046                                         else if (strlen ($r) == 5)
1047                                         {
1048                                                 if ($state == 'b')
1049                                                 { $output .= '</b><i>'; $state = 'i'; }
1050                                                 else if ($state == 'i')
1051                                                 { $output .= '</i><b>'; $state = 'b'; }
1052                                                 else if ($state == 'bi')
1053                                                 { $output .= '</i></b>'; $state = ''; }
1054                                                 else if ($state == 'ib')
1055                                                 { $output .= '</b></i>'; $state = ''; }
1056                                                 else if ($state == 'both')
1057                                                 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
1058                                                 else # ($state == '')
1059                                                 { $buffer = ''; $state = 'both'; }
1060                                         }
1061                                 }
1062                                 $i++;
1063                         }
1064                         # Now close all remaining tags.  Notice that the order is important.
1065                         if ($state == 'b' || $state == 'ib')
1066                                 $output .= '</b>';
1067                         if ($state == 'i' || $state == 'bi' || $state == 'ib')
1068                                 $output .= '</i>';
1069                         if ($state == 'bi')
1070                                 $output .= '</b>';
1071                         if ($state == 'both')
1072                                 $output .= '<b><i>'.$buffer.'</i></b>';
1073                         return $output;
1074                 }
1075         }
1076
1077         /**
1078          * Replace external links
1079          *
1080          * Note: this is all very hackish and the order of execution matters a lot.
1081          * Make sure to run maintenance/parserTests.php if you change this code.
1082          *
1083          * @private
1084          */
1085         function replaceExternalLinks( $text ) {
1086                 global $wgContLang;
1087                 $fname = 'Parser::replaceExternalLinks';
1088                 wfProfileIn( $fname );
1089
1090                 $sk =& $this->mOptions->getSkin();
1091
1092                 $bits = preg_split( EXT_LINK_BRACKETED, $text, -1, PREG_SPLIT_DELIM_CAPTURE );
1093
1094                 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
1095
1096                 $i = 0;
1097                 while ( $i<count( $bits ) ) {
1098                         $url = $bits[$i++];
1099                         $protocol = $bits[$i++];
1100                         $text = $bits[$i++];
1101                         $trail = $bits[$i++];
1102
1103                         # The characters '<' and '>' (which were escaped by
1104                         # removeHTMLtags()) should not be included in
1105                         # URLs, per RFC 2396.
1106                         if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE)) {
1107                                 $text = substr($url, $m2[0][1]) . ' ' . $text;
1108                                 $url = substr($url, 0, $m2[0][1]);
1109                         }
1110
1111                         # If the link text is an image URL, replace it with an <img> tag
1112                         # This happened by accident in the original parser, but some people used it extensively
1113                         $img = $this->maybeMakeExternalImage( $text );
1114                         if ( $img !== false ) {
1115                                 $text = $img;
1116                         }
1117
1118                         $dtrail = '';
1119
1120                         # Set linktype for CSS - if URL==text, link is essentially free
1121                         $linktype = ($text == $url) ? 'free' : 'text';
1122
1123                         # No link text, e.g. [http://domain.tld/some.link]
1124                         if ( $text == '' ) {
1125                                 # Autonumber if allowed. See bug #5918
1126                                 if ( strpos( wfUrlProtocols(), substr($protocol, 0, strpos($protocol, ':')) ) !== false ) {
1127                                         $text = '[' . ++$this->mAutonumber . ']';
1128                                         $linktype = 'autonumber';
1129                                 } else {
1130                                         # Otherwise just use the URL
1131                                         $text = htmlspecialchars( $url );
1132                                         $linktype = 'free';
1133                                 }
1134                         } else {
1135                                 # Have link text, e.g. [http://domain.tld/some.link text]s
1136                                 # Check for trail
1137                                 list( $dtrail, $trail ) = Linker::splitTrail( $trail );
1138                         }
1139
1140                         $text = $wgContLang->markNoConversion($text);
1141
1142                         # Normalize any HTML entities in input. They will be
1143                         # re-escaped by makeExternalLink().
1144                         $url = Sanitizer::decodeCharReferences( $url );
1145
1146                         # Escape any control characters introduced by the above step
1147                         $url = preg_replace( '/[\][<>"\\x00-\\x20\\x7F]/e', "urlencode('\\0')", $url );
1148
1149                         # Process the trail (i.e. everything after this link up until start of the next link),
1150                         # replacing any non-bracketed links
1151                         $trail = $this->replaceFreeExternalLinks( $trail );
1152
1153                         # Use the encoded URL
1154                         # This means that users can paste URLs directly into the text
1155                         # Funny characters like &ouml; aren't valid in URLs anyway
1156                         # This was changed in August 2004
1157                         $s .= $sk->makeExternalLink( $url, $text, false, $linktype, $this->mTitle->getNamespace() ) . $dtrail . $trail;
1158
1159                         # Register link in the output object.
1160                         # Replace unnecessary URL escape codes with the referenced character
1161                         # This prevents spammers from hiding links from the filters
1162                         $pasteurized = Parser::replaceUnusualEscapes( $url );
1163                         $this->mOutput->addExternalLink( $pasteurized );
1164                 }
1165
1166                 wfProfileOut( $fname );
1167                 return $s;
1168         }
1169
1170         /**
1171          * Replace anything that looks like a URL with a link
1172          * @private
1173          */
1174         function replaceFreeExternalLinks( $text ) {
1175                 global $wgContLang;
1176                 $fname = 'Parser::replaceFreeExternalLinks';
1177                 wfProfileIn( $fname );
1178
1179                 $bits = preg_split( '/(\b(?:' . wfUrlProtocols() . '))/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE );
1180                 $s = array_shift( $bits );
1181                 $i = 0;
1182
1183                 $sk =& $this->mOptions->getSkin();
1184
1185                 while ( $i < count( $bits ) ){
1186                         $protocol = $bits[$i++];
1187                         $remainder = $bits[$i++];
1188
1189                         if ( preg_match( '/^('.EXT_LINK_URL_CLASS.'+)(.*)$/s', $remainder, $m ) ) {
1190                                 # Found some characters after the protocol that look promising
1191                                 $url = $protocol . $m[1];
1192                                 $trail = $m[2];
1193
1194                                 # special case: handle urls as url args:
1195                                 # http://www.example.com/foo?=http://www.example.com/bar
1196                                 if(strlen($trail) == 0 &&
1197                                         isset($bits[$i]) &&
1198                                         preg_match('/^'. wfUrlProtocols() . '$/S', $bits[$i]) &&
1199                                         preg_match( '/^('.EXT_LINK_URL_CLASS.'+)(.*)$/s', $bits[$i + 1], $m ))
1200                                 {
1201                                         # add protocol, arg
1202                                         $url .= $bits[$i] . $m[1]; # protocol, url as arg to previous link
1203                                         $i += 2;
1204                                         $trail = $m[2];
1205                                 }
1206
1207                                 # The characters '<' and '>' (which were escaped by
1208                                 # removeHTMLtags()) should not be included in
1209                                 # URLs, per RFC 2396.
1210                                 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE)) {
1211                                         $trail = substr($url, $m2[0][1]) . $trail;
1212                                         $url = substr($url, 0, $m2[0][1]);
1213                                 }
1214
1215                                 # Move trailing punctuation to $trail
1216                                 $sep = ',;\.:!?';
1217                                 # If there is no left bracket, then consider right brackets fair game too
1218                                 if ( strpos( $url, '(' ) === false ) {
1219                                         $sep .= ')';
1220                                 }
1221
1222                                 $numSepChars = strspn( strrev( $url ), $sep );
1223                                 if ( $numSepChars ) {
1224                                         $trail = substr( $url, -$numSepChars ) . $trail;
1225                                         $url = substr( $url, 0, -$numSepChars );
1226                                 }
1227
1228                                 # Normalize any HTML entities in input. They will be
1229                                 # re-escaped by makeExternalLink() or maybeMakeExternalImage()
1230                                 $url = Sanitizer::decodeCharReferences( $url );
1231
1232                                 # Escape any control characters introduced by the above step
1233                                 $url = preg_replace( '/[\][<>"\\x00-\\x20\\x7F]/e', "urlencode('\\0')", $url );
1234
1235                                 # Is this an external image?
1236                                 $text = $this->maybeMakeExternalImage( $url );
1237                                 if ( $text === false ) {
1238                                         # Not an image, make a link
1239                                         $text = $sk->makeExternalLink( $url, $wgContLang->markNoConversion($url), true, 'free', $this->mTitle->getNamespace() );
1240                                         # Register it in the output object...
1241                                         # Replace unnecessary URL escape codes with their equivalent characters
1242                                         $pasteurized = Parser::replaceUnusualEscapes( $url );
1243                                         $this->mOutput->addExternalLink( $pasteurized );
1244                                 }
1245                                 $s .= $text . $trail;
1246                         } else {
1247                                 $s .= $protocol . $remainder;
1248                         }
1249                 }
1250                 wfProfileOut( $fname );
1251                 return $s;
1252         }
1253
1254         /**
1255          * Replace unusual URL escape codes with their equivalent characters
1256          * @param string
1257          * @return string
1258          * @static
1259          * @fixme This can merge genuinely required bits in the path or query string,
1260          *        breaking legit URLs. A proper fix would treat the various parts of
1261          *        the URL differently; as a workaround, just use the output for
1262          *        statistical records, not for actual linking/output.
1263          */
1264         function replaceUnusualEscapes( $url ) {
1265                 return preg_replace_callback( '/%[0-9A-Fa-f]{2}/',
1266                         array( 'Parser', 'replaceUnusualEscapesCallback' ), $url );
1267         }
1268
1269         /**
1270          * Callback function used in replaceUnusualEscapes().
1271          * Replaces unusual URL escape codes with their equivalent character
1272          * @static
1273          * @private
1274          */
1275         function replaceUnusualEscapesCallback( $matches ) {
1276                 $char = urldecode( $matches[0] );
1277                 $ord = ord( $char );
1278                 // Is it an unsafe or HTTP reserved character according to RFC 1738?
1279                 if ( $ord > 32 && $ord < 127 && strpos( '<>"#{}|\^~[]`;/?', $char ) === false ) {
1280                         // No, shouldn't be escaped
1281                         return $char;
1282                 } else {
1283                         // Yes, leave it escaped
1284                         return $matches[0];
1285                 }
1286         }
1287
1288         /**
1289          * make an image if it's allowed, either through the global
1290          * option or through the exception
1291          * @private
1292          */
1293         function maybeMakeExternalImage( $url ) {
1294                 $sk =& $this->mOptions->getSkin();
1295                 $imagesfrom = $this->mOptions->getAllowExternalImagesFrom();
1296                 $imagesexception = !empty($imagesfrom);
1297                 $text = false;
1298                 if ( $this->mOptions->getAllowExternalImages()
1299                      || ( $imagesexception && strpos( $url, $imagesfrom ) === 0 ) ) {
1300                         if ( preg_match( EXT_IMAGE_REGEX, $url ) ) {
1301                                 # Image found
1302                                 $text = $sk->makeExternalImage( htmlspecialchars( $url ) );
1303                         }
1304                 }
1305                 return $text;
1306         }
1307
1308         /**
1309          * Process [[ ]] wikilinks
1310          *
1311          * @private
1312          */
1313         function replaceInternalLinks( $s ) {
1314                 global $wgContLang;
1315                 static $fname = 'Parser::replaceInternalLinks' ;
1316
1317                 wfProfileIn( $fname );
1318
1319                 wfProfileIn( $fname.'-setup' );
1320                 static $tc = FALSE;
1321                 # the % is needed to support urlencoded titles as well
1322                 if ( !$tc ) { $tc = Title::legalChars() . '#%'; }
1323
1324                 $sk =& $this->mOptions->getSkin();
1325
1326                 #split the entire text string on occurences of [[
1327                 $a = explode( '[[', ' ' . $s );
1328                 #get the first element (all text up to first [[), and remove the space we added
1329                 $s = array_shift( $a );
1330                 $s = substr( $s, 1 );
1331
1332                 # Match a link having the form [[namespace:link|alternate]]trail
1333                 static $e1 = FALSE;
1334                 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
1335                 # Match cases where there is no "]]", which might still be images
1336                 static $e1_img = FALSE;
1337                 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1338                 # Match the end of a line for a word that's not followed by whitespace,
1339                 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1340                 $e2 = wfMsgForContent( 'linkprefix' );
1341
1342                 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1343
1344                 if( is_null( $this->mTitle ) ) {
1345                         throw new MWException( 'nooo' );
1346                 }
1347                 $nottalk = !$this->mTitle->isTalkPage();
1348
1349                 if ( $useLinkPrefixExtension ) {
1350                         if ( preg_match( $e2, $s, $m ) ) {
1351                                 $first_prefix = $m[2];
1352                         } else {
1353                                 $first_prefix = false;
1354                         }
1355                 } else {
1356                         $prefix = '';
1357                 }
1358
1359                 $selflink = $this->mTitle->getPrefixedText();
1360                 wfProfileOut( $fname.'-setup' );
1361
1362                 $checkVariantLink = sizeof($wgContLang->getVariants())>1;
1363                 $useSubpages = $this->areSubpagesAllowed();
1364
1365                 # Loop for each link
1366                 for ($k = 0; isset( $a[$k] ); $k++) {
1367                         $line = $a[$k];
1368                         if ( $useLinkPrefixExtension ) {
1369                                 wfProfileIn( $fname.'-prefixhandling' );
1370                                 if ( preg_match( $e2, $s, $m ) ) {
1371                                         $prefix = $m[2];
1372                                         $s = $m[1];
1373                                 } else {
1374                                         $prefix='';
1375                                 }
1376                                 # first link
1377                                 if($first_prefix) {
1378                                         $prefix = $first_prefix;
1379                                         $first_prefix = false;
1380                                 }
1381                                 wfProfileOut( $fname.'-prefixhandling' );
1382                         }
1383
1384                         $might_be_img = false;
1385
1386                         if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1387                                 $text = $m[2];
1388                                 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
1389                                 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
1390                                 # the real problem is with the $e1 regex
1391                                 # See bug 1300.
1392                                 #
1393                                 # Still some problems for cases where the ] is meant to be outside punctuation,
1394                                 # and no image is in sight. See bug 2095.
1395                                 #
1396                                 if( $text !== '' &&
1397                                         preg_match( "/^\](.*)/s", $m[3], $n ) &&
1398                                         strpos($text, '[') !== false
1399                                 )
1400                                 {
1401                                         $text .= ']'; # so that replaceExternalLinks($text) works later
1402                                         $m[3] = $n[1];
1403                                 }
1404                                 # fix up urlencoded title texts
1405                                 if(preg_match('/%/', $m[1] ))
1406                                         # Should anchors '#' also be rejected?
1407                                         $m[1] = str_replace( array('<', '>'), array('&lt;', '&gt;'), urldecode($m[1]) );
1408                                 $trail = $m[3];
1409                         } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1410                                 $might_be_img = true;
1411                                 $text = $m[2];
1412                                 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1413                                 $trail = "";
1414                         } else { # Invalid form; output directly
1415                                 $s .= $prefix . '[[' . $line ;
1416                                 continue;
1417                         }
1418
1419                         # Don't allow internal links to pages containing
1420                         # PROTO: where PROTO is a valid URL protocol; these
1421                         # should be external links.
1422                         if (preg_match('/^(\b(?:' . wfUrlProtocols() . '))/', $m[1])) {
1423                                 $s .= $prefix . '[[' . $line ;
1424                                 continue;
1425                         }
1426
1427                         # Make subpage if necessary
1428                         if( $useSubpages ) {
1429                                 $link = $this->maybeDoSubpageLink( $m[1], $text );
1430                         } else {
1431                                 $link = $m[1];
1432                         }
1433
1434                         $noforce = (substr($m[1], 0, 1) != ':');
1435                         if (!$noforce) {
1436                                 # Strip off leading ':'
1437                                 $link = substr($link, 1);
1438                         }
1439
1440                         $nt = Title::newFromText( $this->unstripNoWiki($link, $this->mStripState) );
1441                         if( !$nt ) {
1442                                 $s .= $prefix . '[[' . $line;
1443                                 continue;
1444                         }
1445
1446                         #check other language variants of the link
1447                         #if the article does not exist
1448                         if( $checkVariantLink
1449                             && $nt->getArticleID() == 0 ) {
1450                                 $wgContLang->findVariantLink($link, $nt);
1451                         }
1452
1453                         $ns = $nt->getNamespace();
1454                         $iw = $nt->getInterWiki();
1455
1456                         if ($might_be_img) { # if this is actually an invalid link
1457                                 if ($ns == NS_IMAGE && $noforce) { #but might be an image
1458                                         $found = false;
1459                                         while (isset ($a[$k+1]) ) {
1460                                                 #look at the next 'line' to see if we can close it there
1461                                                 $spliced = array_splice( $a, $k + 1, 1 );
1462                                                 $next_line = array_shift( $spliced );
1463                                                 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1464                                                 # the first ]] closes the inner link, the second the image
1465                                                         $found = true;
1466                                                         $text .= '[[' . $m[1];
1467                                                         $trail = $m[2];
1468                                                         break;
1469                                                 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1470                                                         #if there's exactly one ]] that's fine, we'll keep looking
1471                                                         $text .= '[[' . $m[0];
1472                                                 } else {
1473                                                         #if $next_line is invalid too, we need look no further
1474                                                         $text .= '[[' . $next_line;
1475                                                         break;
1476                                                 }
1477                                         }
1478                                         if ( !$found ) {
1479                                                 # we couldn't find the end of this imageLink, so output it raw
1480                                                 #but don't ignore what might be perfectly normal links in the text we've examined
1481                                                 $text = $this->replaceInternalLinks($text);
1482                                                 $s .= $prefix . '[[' . $link . '|' . $text;
1483                                                 # note: no $trail, because without an end, there *is* no trail
1484                                                 continue;
1485                                         }
1486                                 } else { #it's not an image, so output it raw
1487                                         $s .= $prefix . '[[' . $link . '|' . $text;
1488                                         # note: no $trail, because without an end, there *is* no trail
1489                                         continue;
1490                                 }
1491                         }
1492
1493                         $wasblank = ( '' == $text );
1494                         if( $wasblank ) $text = $link;
1495
1496
1497                         # Link not escaped by : , create the various objects
1498                         if( $noforce ) {
1499
1500                                 # Interwikis
1501                                 if( $iw && $this->mOptions->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1502                                         $this->mOutput->addLanguageLink( $nt->getFullText() );
1503                                         $s = rtrim($s . "\n");
1504                                         $s .= trim($prefix . $trail, "\n") == '' ? '': $prefix . $trail;
1505                                         continue;
1506                                 }
1507
1508                                 if ( $ns == NS_IMAGE ) {
1509                                         wfProfileIn( "$fname-image" );
1510                                         if ( !wfIsBadImage( $nt->getDBkey() ) ) {
1511                                                 # recursively parse links inside the image caption
1512                                                 # actually, this will parse them in any other parameters, too,
1513                                                 # but it might be hard to fix that, and it doesn't matter ATM
1514                                                 $text = $this->replaceExternalLinks($text);
1515                                                 $text = $this->replaceInternalLinks($text);
1516
1517                                                 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1518                                                 $s .= $prefix . $this->armorLinks( $this->makeImage( $nt, $text ) ) . $trail;
1519                                                 $this->mOutput->addImage( $nt->getDBkey() );
1520
1521                                                 wfProfileOut( "$fname-image" );
1522                                                 continue;
1523                                         }
1524                                         wfProfileOut( "$fname-image" );
1525
1526                                 }
1527
1528                                 if ( $ns == NS_CATEGORY ) {
1529                                         wfProfileIn( "$fname-category" );
1530                                         $s = rtrim($s . "\n"); # bug 87
1531
1532                                         if ( $wasblank ) {
1533                                                 if ( $this->mTitle->getNamespace() == NS_CATEGORY ) {
1534                                                         $sortkey = $this->mTitle->getText();
1535                                                 } else {
1536                                                         $sortkey = $this->mTitle->getPrefixedText();
1537                                                 }
1538                                         } else {
1539                                                 $sortkey = $text;
1540                                         }
1541                                         $sortkey = Sanitizer::decodeCharReferences( $sortkey );
1542                                         $sortkey = str_replace( "\n", '', $sortkey );
1543                                         $sortkey = $wgContLang->convertCategoryKey( $sortkey );
1544                                         $this->mOutput->addCategory( $nt->getDBkey(), $sortkey );
1545
1546                                         /**
1547                                          * Strip the whitespace Category links produce, see bug 87
1548                                          * @todo We might want to use trim($tmp, "\n") here.
1549                                          */
1550                                         $s .= trim($prefix . $trail, "\n") == '' ? '': $prefix . $trail;
1551
1552                                         wfProfileOut( "$fname-category" );
1553                                         continue;
1554                                 }
1555                         }
1556
1557                         if( ( $nt->getPrefixedText() === $selflink ) &&
1558                             ( $nt->getFragment() === '' ) ) {
1559                                 # Self-links are handled specially; generally de-link and change to bold.
1560                                 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1561                                 continue;
1562                         }
1563
1564                         # Special and Media are pseudo-namespaces; no pages actually exist in them
1565                         if( $ns == NS_MEDIA ) {
1566                                 $link = $sk->makeMediaLinkObj( $nt, $text );
1567                                 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
1568                                 $s .= $prefix . $this->armorLinks( $link ) . $trail;
1569                                 $this->mOutput->addImage( $nt->getDBkey() );
1570                                 continue;
1571                         } elseif( $ns == NS_SPECIAL ) {
1572                                 $s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
1573                                 continue;
1574                         } elseif( $ns == NS_IMAGE ) {
1575                                 $img = Image::newFromTitle( $nt );
1576                                 if( $img->exists() ) {
1577                                         // Force a blue link if the file exists; may be a remote
1578                                         // upload on the shared repository, and we want to see its
1579                                         // auto-generated page.
1580                                         $s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
1581                                         continue;
1582                                 }
1583                         }
1584                         $s .= $this->makeLinkHolder( $nt, $text, '', $trail, $prefix );
1585                 }
1586                 wfProfileOut( $fname );
1587                 return $s;
1588         }
1589
1590         /**
1591          * Make a link placeholder. The text returned can be later resolved to a real link with
1592          * replaceLinkHolders(). This is done for two reasons: firstly to avoid further
1593          * parsing of interwiki links, and secondly to allow all extistence checks and
1594          * article length checks (for stub links) to be bundled into a single query.
1595          *
1596          */
1597         function makeLinkHolder( &$nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1598                 if ( ! is_object($nt) ) {
1599                         # Fail gracefully
1600                         $retVal = "<!-- ERROR -->{$prefix}{$text}{$trail}";
1601                 } else {
1602                         # Separate the link trail from the rest of the link
1603                         list( $inside, $trail ) = Linker::splitTrail( $trail );
1604
1605                         if ( $nt->isExternal() ) {
1606                                 $nr = array_push( $this->mInterwikiLinkHolders['texts'], $prefix.$text.$inside );
1607                                 $this->mInterwikiLinkHolders['titles'][] = $nt;
1608                                 $retVal = '<!--IWLINK '. ($nr-1) ."-->{$trail}";
1609                         } else {
1610                                 $nr = array_push( $this->mLinkHolders['namespaces'], $nt->getNamespace() );
1611                                 $this->mLinkHolders['dbkeys'][] = $nt->getDBkey();
1612                                 $this->mLinkHolders['queries'][] = $query;
1613                                 $this->mLinkHolders['texts'][] = $prefix.$text.$inside;
1614                                 $this->mLinkHolders['titles'][] = $nt;
1615
1616                                 $retVal = '<!--LINK '. ($nr-1) ."-->{$trail}";
1617                         }
1618                 }
1619                 return $retVal;
1620         }
1621
1622         /**
1623          * Render a forced-blue link inline; protect against double expansion of
1624          * URLs if we're in a mode that prepends full URL prefixes to internal links.
1625          * Since this little disaster has to split off the trail text to avoid
1626          * breaking URLs in the following text without breaking trails on the
1627          * wiki links, it's been made into a horrible function.
1628          *
1629          * @param Title $nt
1630          * @param string $text
1631          * @param string $query
1632          * @param string $trail
1633          * @param string $prefix
1634          * @return string HTML-wikitext mix oh yuck
1635          */
1636         function makeKnownLinkHolder( $nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1637                 list( $inside, $trail ) = Linker::splitTrail( $trail );
1638                 $sk =& $this->mOptions->getSkin();
1639                 $link = $sk->makeKnownLinkObj( $nt, $text, $query, $inside, $prefix );
1640                 return $this->armorLinks( $link ) . $trail;
1641         }
1642
1643         /**
1644          * Insert a NOPARSE hacky thing into any inline links in a chunk that's
1645          * going to go through further parsing steps before inline URL expansion.
1646          *
1647          * In particular this is important when using action=render, which causes
1648          * full URLs to be included.
1649          *
1650          * Oh man I hate our multi-layer parser!
1651          *
1652          * @param string more-or-less HTML
1653          * @return string less-or-more HTML with NOPARSE bits
1654          */
1655         function armorLinks( $text ) {
1656                 return preg_replace( "/\b(" . wfUrlProtocols() . ')/',
1657                         "{$this->mUniqPrefix}NOPARSE$1", $text );
1658         }
1659
1660         /**
1661          * Return true if subpage links should be expanded on this page.
1662          * @return bool
1663          */
1664         function areSubpagesAllowed() {
1665                 # Some namespaces don't allow subpages
1666                 global $wgNamespacesWithSubpages;
1667                 return !empty($wgNamespacesWithSubpages[$this->mTitle->getNamespace()]);
1668         }
1669
1670         /**
1671          * Handle link to subpage if necessary
1672          * @param string $target the source of the link
1673          * @param string &$text the link text, modified as necessary
1674          * @return string the full name of the link
1675          * @private
1676          */
1677         function maybeDoSubpageLink($target, &$text) {
1678                 # Valid link forms:
1679                 # Foobar -- normal
1680                 # :Foobar -- override special treatment of prefix (images, language links)
1681                 # /Foobar -- convert to CurrentPage/Foobar
1682                 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1683                 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1684                 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1685
1686                 $fname = 'Parser::maybeDoSubpageLink';
1687                 wfProfileIn( $fname );
1688                 $ret = $target; # default return value is no change
1689
1690                 # Some namespaces don't allow subpages,
1691                 # so only perform processing if subpages are allowed
1692                 if( $this->areSubpagesAllowed() ) {
1693                         # Look at the first character
1694                         if( $target != '' && $target{0} == '/' ) {
1695                                 # / at end means we don't want the slash to be shown
1696                                 if( substr( $target, -1, 1 ) == '/' ) {
1697                                         $target = substr( $target, 1, -1 );
1698                                         $noslash = $target;
1699                                 } else {
1700                                         $noslash = substr( $target, 1 );
1701                                 }
1702
1703                                 $ret = $this->mTitle->getPrefixedText(). '/' . trim($noslash);
1704                                 if( '' === $text ) {
1705                                         $text = $target;
1706                                 } # this might be changed for ugliness reasons
1707                         } else {
1708                                 # check for .. subpage backlinks
1709                                 $dotdotcount = 0;
1710                                 $nodotdot = $target;
1711                                 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1712                                         ++$dotdotcount;
1713                                         $nodotdot = substr( $nodotdot, 3 );
1714                                 }
1715                                 if($dotdotcount > 0) {
1716                                         $exploded = explode( '/', $this->mTitle->GetPrefixedText() );
1717                                         if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1718                                                 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1719                                                 # / at the end means don't show full path
1720                                                 if( substr( $nodotdot, -1, 1 ) == '/' ) {
1721                                                         $nodotdot = substr( $nodotdot, 0, -1 );
1722                                                         if( '' === $text ) {
1723                                                                 $text = $nodotdot;
1724                                                         }
1725                                                 }
1726                                                 $nodotdot = trim( $nodotdot );
1727                                                 if( $nodotdot != '' ) {
1728                                                         $ret .= '/' . $nodotdot;
1729                                                 }
1730                                         }
1731                                 }
1732                         }
1733                 }
1734
1735                 wfProfileOut( $fname );
1736                 return $ret;
1737         }
1738
1739         /**#@+
1740          * Used by doBlockLevels()
1741          * @private
1742          */
1743         /* private */ function closeParagraph() {
1744                 $result = '';
1745                 if ( '' != $this->mLastSection ) {
1746                         $result = '</' . $this->mLastSection  . ">\n";
1747                 }
1748                 $this->mInPre = false;
1749                 $this->mLastSection = '';
1750                 return $result;
1751         }
1752         # getCommon() returns the length of the longest common substring
1753         # of both arguments, starting at the beginning of both.
1754         #
1755         /* private */ function getCommon( $st1, $st2 ) {
1756                 $fl = strlen( $st1 );
1757                 $shorter = strlen( $st2 );
1758                 if ( $fl < $shorter ) { $shorter = $fl; }
1759
1760                 for ( $i = 0; $i < $shorter; ++$i ) {
1761                         if ( $st1{$i} != $st2{$i} ) { break; }
1762                 }
1763                 return $i;
1764         }
1765         # These next three functions open, continue, and close the list
1766         # element appropriate to the prefix character passed into them.
1767         #
1768         /* private */ function openList( $char ) {
1769                 $result = $this->closeParagraph();
1770
1771                 if ( '*' == $char ) { $result .= '<ul><li>'; }
1772                 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1773                 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1774                 else if ( ';' == $char ) {
1775                         $result .= '<dl><dt>';
1776                         $this->mDTopen = true;
1777                 }
1778                 else { $result = '<!-- ERR 1 -->'; }
1779
1780                 return $result;
1781         }
1782
1783         /* private */ function nextItem( $char ) {
1784                 if ( '*' == $char || '#' == $char ) { return '</li><li>'; }
1785                 else if ( ':' == $char || ';' == $char ) {
1786                         $close = '</dd>';
1787                         if ( $this->mDTopen ) { $close = '</dt>'; }
1788                         if ( ';' == $char ) {
1789                                 $this->mDTopen = true;
1790                                 return $close . '<dt>';
1791                         } else {
1792                                 $this->mDTopen = false;
1793                                 return $close . '<dd>';
1794                         }
1795                 }
1796                 return '<!-- ERR 2 -->';
1797         }
1798
1799         /* private */ function closeList( $char ) {
1800                 if ( '*' == $char ) { $text = '</li></ul>'; }
1801                 else if ( '#' == $char ) { $text = '</li></ol>'; }
1802                 else if ( ':' == $char ) {
1803                         if ( $this->mDTopen ) {
1804                                 $this->mDTopen = false;
1805                                 $text = '</dt></dl>';
1806                         } else {
1807                                 $text = '</dd></dl>';
1808                         }
1809                 }
1810                 else {  return '<!-- ERR 3 -->'; }
1811                 return $text."\n";
1812         }
1813         /**#@-*/
1814
1815         /**
1816          * Make lists from lines starting with ':', '*', '#', etc.
1817          *
1818          * @private
1819          * @return string the lists rendered as HTML
1820          */
1821         function doBlockLevels( $text, $linestart ) {
1822                 $fname = 'Parser::doBlockLevels';
1823                 wfProfileIn( $fname );
1824
1825                 # Parsing through the text line by line.  The main thing
1826                 # happening here is handling of block-level elements p, pre,
1827                 # and making lists from lines starting with * # : etc.
1828                 #
1829                 $textLines = explode( "\n", $text );
1830
1831                 $lastPrefix = $output = '';
1832                 $this->mDTopen = $inBlockElem = false;
1833                 $prefixLength = 0;
1834                 $paragraphStack = false;
1835
1836                 if ( !$linestart ) {
1837                         $output .= array_shift( $textLines );
1838                 }
1839                 foreach ( $textLines as $oLine ) {
1840                         $lastPrefixLength = strlen( $lastPrefix );
1841                         $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1842                         $preOpenMatch = preg_match('/<pre/i', $oLine );
1843                         if ( !$this->mInPre ) {
1844                                 # Multiple prefixes may abut each other for nested lists.
1845                                 $prefixLength = strspn( $oLine, '*#:;' );
1846                                 $pref = substr( $oLine, 0, $prefixLength );
1847
1848                                 # eh?
1849                                 $pref2 = str_replace( ';', ':', $pref );
1850                                 $t = substr( $oLine, $prefixLength );
1851                                 $this->mInPre = !empty($preOpenMatch);
1852                         } else {
1853                                 # Don't interpret any other prefixes in preformatted text
1854                                 $prefixLength = 0;
1855                                 $pref = $pref2 = '';
1856                                 $t = $oLine;
1857                         }
1858
1859                         # List generation
1860                         if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1861                                 # Same as the last item, so no need to deal with nesting or opening stuff
1862                                 $output .= $this->nextItem( substr( $pref, -1 ) );
1863                                 $paragraphStack = false;
1864
1865                                 if ( substr( $pref, -1 ) == ';') {
1866                                         # The one nasty exception: definition lists work like this:
1867                                         # ; title : definition text
1868                                         # So we check for : in the remainder text to split up the
1869                                         # title and definition, without b0rking links.
1870                                         $term = $t2 = '';
1871                                         if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1872                                                 $t = $t2;
1873                                                 $output .= $term . $this->nextItem( ':' );
1874                                         }
1875                                 }
1876                         } elseif( $prefixLength || $lastPrefixLength ) {
1877                                 # Either open or close a level...
1878                                 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1879                                 $paragraphStack = false;
1880
1881                                 while( $commonPrefixLength < $lastPrefixLength ) {
1882                                         $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1883                                         --$lastPrefixLength;
1884                                 }
1885                                 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1886                                         $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1887                                 }
1888                                 while ( $prefixLength > $commonPrefixLength ) {
1889                                         $char = substr( $pref, $commonPrefixLength, 1 );
1890                                         $output .= $this->openList( $char );
1891
1892                                         if ( ';' == $char ) {
1893                                                 # FIXME: This is dupe of code above
1894                                                 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1895                                                         $t = $t2;
1896                                                         $output .= $term . $this->nextItem( ':' );
1897                                                 }
1898                                         }
1899                                         ++$commonPrefixLength;
1900                                 }
1901                                 $lastPrefix = $pref2;
1902                         }
1903                         if( 0 == $prefixLength ) {
1904                                 wfProfileIn( "$fname-paragraph" );
1905                                 # No prefix (not in list)--go to paragraph mode
1906                                 // XXX: use a stack for nestable elements like span, table and div
1907                                 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<ol|<li|<\\/center|<\\/tr|<\\/td|<\\/th)/iS', $t );
1908                                 $closematch = preg_match(
1909                                         '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1910                                         '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$this->mUniqPrefix.'-pre|<\\/li|<\\/ul|<\\/ol|<center)/iS', $t );
1911                                 if ( $openmatch or $closematch ) {
1912                                         $paragraphStack = false;
1913                                         # TODO bug 5718: paragraph closed
1914                                         $output .= $this->closeParagraph();
1915                                         if ( $preOpenMatch and !$preCloseMatch ) {
1916                                                 $this->mInPre = true;
1917                                         }
1918                                         if ( $closematch ) {
1919                                                 $inBlockElem = false;
1920                                         } else {
1921                                                 $inBlockElem = true;
1922                                         }
1923                                 } else if ( !$inBlockElem && !$this->mInPre ) {
1924                                         if ( ' ' == $t{0} and ( $this->mLastSection == 'pre' or trim($t) != '' ) ) {
1925                                                 // pre
1926                                                 if ($this->mLastSection != 'pre') {
1927                                                         $paragraphStack = false;
1928                                                         $output .= $this->closeParagraph().'<pre>';
1929                                                         $this->mLastSection = 'pre';
1930                                                 }
1931                                                 $t = substr( $t, 1 );
1932                                         } else {
1933                                                 // paragraph
1934                                                 if ( '' == trim($t) ) {
1935                                                         if ( $paragraphStack ) {
1936                                                                 $output .= $paragraphStack.'<br />';
1937                                                                 $paragraphStack = false;
1938                                                                 $this->mLastSection = 'p';
1939                                                         } else {
1940                                                                 if ($this->mLastSection != 'p' ) {
1941                                                                         $output .= $this->closeParagraph();
1942                                                                         $this->mLastSection = '';
1943                                                                         $paragraphStack = '<p>';
1944                                                                 } else {
1945                                                                         $paragraphStack = '</p><p>';
1946                                                                 }
1947                                                         }
1948                                                 } else {
1949                                                         if ( $paragraphStack ) {
1950                                                                 $output .= $paragraphStack;
1951                                                                 $paragraphStack = false;
1952                                                                 $this->mLastSection = 'p';
1953                                                         } else if ($this->mLastSection != 'p') {
1954                                                                 $output .= $this->closeParagraph().'<p>';
1955                                                                 $this->mLastSection = 'p';
1956                                                         }
1957                                                 }
1958                                         }
1959                                 }
1960                                 wfProfileOut( "$fname-paragraph" );
1961                         }
1962                         // somewhere above we forget to get out of pre block (bug 785)
1963                         if($preCloseMatch && $this->mInPre) {
1964                                 $this->mInPre = false;
1965                         }
1966                         if ($paragraphStack === false) {
1967                                 $output .= $t."\n";
1968                         }
1969                 }
1970                 while ( $prefixLength ) {
1971                         $output .= $this->closeList( $pref2{$prefixLength-1} );
1972                         --$prefixLength;
1973                 }
1974                 if ( '' != $this->mLastSection ) {
1975                         $output .= '</' . $this->mLastSection . '>';
1976                         $this->mLastSection = '';
1977                 }
1978
1979                 wfProfileOut( $fname );
1980                 return $output;
1981         }
1982
1983         /**
1984          * Split up a string on ':', ignoring any occurences inside tags
1985          * to prevent illegal overlapping.
1986          * @param string $str the string to split
1987          * @param string &$before set to everything before the ':'
1988          * @param string &$after set to everything after the ':'
1989          * return string the position of the ':', or false if none found
1990          */
1991         function findColonNoLinks($str, &$before, &$after) {
1992                 $fname = 'Parser::findColonNoLinks';
1993                 wfProfileIn( $fname );
1994
1995                 $pos = strpos( $str, ':' );
1996                 if( $pos === false ) {
1997                         // Nothing to find!
1998                         wfProfileOut( $fname );
1999                         return false;
2000                 }
2001
2002                 $lt = strpos( $str, '<' );
2003                 if( $lt === false || $lt > $pos ) {
2004                         // Easy; no tag nesting to worry about
2005                         $before = substr( $str, 0, $pos );
2006                         $after = substr( $str, $pos+1 );
2007                         wfProfileOut( $fname );
2008                         return $pos;
2009                 }
2010
2011                 // Ugly state machine to walk through avoiding tags.
2012                 $state = MW_COLON_STATE_TEXT;
2013                 $stack = 0;
2014                 $len = strlen( $str );
2015                 for( $i = 0; $i < $len; $i++ ) {
2016                         $c = $str{$i};
2017
2018                         switch( $state ) {
2019                         // (Using the number is a performance hack for common cases)
2020                         case 0: // MW_COLON_STATE_TEXT:
2021                                 switch( $c ) {
2022                                 case "<":
2023                                         // Could be either a <start> tag or an </end> tag
2024                                         $state = MW_COLON_STATE_TAGSTART;
2025                                         break;
2026                                 case ":":
2027                                         if( $stack == 0 ) {
2028                                                 // We found it!
2029                                                 $before = substr( $str, 0, $i );
2030                                                 $after = substr( $str, $i + 1 );
2031                                                 wfProfileOut( $fname );
2032                                                 return $i;
2033                                         }
2034                                         // Embedded in a tag; don't break it.
2035                                         break;
2036                                 default:
2037                                         // Skip ahead looking for something interesting
2038                                         $colon = strpos( $str, ':', $i );
2039                                         if( $colon === false ) {
2040                                                 // Nothing else interesting
2041                                                 wfProfileOut( $fname );
2042                                                 return false;
2043                                         }
2044                                         $lt = strpos( $str, '<', $i );
2045                                         if( $stack === 0 ) {
2046                                                 if( $lt === false || $colon < $lt ) {
2047                                                         // We found it!
2048                                                         $before = substr( $str, 0, $colon );
2049                                                         $after = substr( $str, $colon + 1 );
2050                                                         wfProfileOut( $fname );
2051                                                         return $i;
2052                                                 }
2053                                         }
2054                                         if( $lt === false ) {
2055                                                 // Nothing else interesting to find; abort!
2056                                                 // We're nested, but there's no close tags left. Abort!
2057                                                 break 2;
2058                                         }
2059                                         // Skip ahead to next tag start
2060                                         $i = $lt;
2061                                         $state = MW_COLON_STATE_TAGSTART;
2062                                 }
2063                                 break;
2064                         case 1: // MW_COLON_STATE_TAG:
2065                                 // In a <tag>
2066                                 switch( $c ) {
2067                                 case ">":
2068                                         $stack++;
2069                                         $state = MW_COLON_STATE_TEXT;
2070                                         break;
2071                                 case "/":
2072                                         // Slash may be followed by >?
2073                                         $state = MW_COLON_STATE_TAGSLASH;
2074                                         break;
2075                                 default:
2076                                         // ignore
2077                                 }
2078                                 break;
2079                         case 2: // MW_COLON_STATE_TAGSTART:
2080                                 switch( $c ) {
2081                                 case "/":
2082                                         $state = MW_COLON_STATE_CLOSETAG;
2083                                         break;
2084                                 case "!":
2085                                         $state = MW_COLON_STATE_COMMENT;
2086                                         break;
2087                                 case ">":
2088                                         // Illegal early close? This shouldn't happen D:
2089                                         $state = MW_COLON_STATE_TEXT;
2090                                         break;
2091                                 default:
2092                                         $state = MW_COLON_STATE_TAG;
2093                                 }
2094                                 break;
2095                         case 3: // MW_COLON_STATE_CLOSETAG:
2096                                 // In a </tag>
2097                                 if( $c == ">" ) {
2098                                         $stack--;
2099                                         if( $stack < 0 ) {
2100                                                 wfDebug( "Invalid input in $fname; too many close tags\n" );
2101                                                 wfProfileOut( $fname );
2102                                                 return false;
2103                                         }
2104                                         $state = MW_COLON_STATE_TEXT;
2105                                 }
2106                                 break;
2107                         case MW_COLON_STATE_TAGSLASH:
2108                                 if( $c == ">" ) {
2109                                         // Yes, a self-closed tag <blah/>
2110                                         $state = MW_COLON_STATE_TEXT;
2111                                 } else {
2112                                         // Probably we're jumping the gun, and this is an attribute
2113                                         $state = MW_COLON_STATE_TAG;
2114                                 }
2115                                 break;
2116                         case 5: // MW_COLON_STATE_COMMENT:
2117                                 if( $c == "-" ) {
2118                                         $state = MW_COLON_STATE_COMMENTDASH;
2119                                 }
2120                                 break;
2121                         case MW_COLON_STATE_COMMENTDASH:
2122                                 if( $c == "-" ) {
2123                                         $state = MW_COLON_STATE_COMMENTDASHDASH;
2124                                 } else {
2125                                         $state = MW_COLON_STATE_COMMENT;
2126                                 }
2127                                 break;
2128                         case MW_COLON_STATE_COMMENTDASHDASH:
2129                                 if( $c == ">" ) {
2130                                         $state = MW_COLON_STATE_TEXT;
2131                                 } else {
2132                                         $state = MW_COLON_STATE_COMMENT;
2133                                 }
2134                                 break;
2135                         default:
2136                                 throw new MWException( "State machine error in $fname" );
2137                         }
2138                 }
2139                 if( $stack > 0 ) {
2140                         wfDebug( "Invalid input in $fname; not enough close tags (stack $stack, state $state)\n" );
2141                         return false;
2142                 }
2143                 wfProfileOut( $fname );
2144                 return false;
2145         }
2146
2147         /**
2148          * Return value of a magic variable (like PAGENAME)
2149          *
2150          * @private
2151          */
2152         function getVariableValue( $index ) {
2153                 global $wgContLang, $wgSitename, $wgServer, $wgServerName, $wgScriptPath;
2154
2155                 /**
2156                  * Some of these require message or data lookups and can be
2157                  * expensive to check many times.
2158                  */
2159                 static $varCache = array();
2160                 if ( wfRunHooks( 'ParserGetVariableValueVarCache', array( &$this, &$varCache ) ) )
2161                         if ( isset( $varCache[$index] ) )
2162                                 return $varCache[$index];
2163
2164                 $ts = time();
2165                 wfRunHooks( 'ParserGetVariableValueTs', array( &$this, &$ts ) );
2166
2167                 switch ( $index ) {
2168                         case MAG_CURRENTMONTH:
2169                                 return $varCache[$index] = $wgContLang->formatNum( date( 'm', $ts ) );
2170                         case MAG_CURRENTMONTHNAME:
2171                                 return $varCache[$index] = $wgContLang->getMonthName( date( 'n', $ts ) );
2172                         case MAG_CURRENTMONTHNAMEGEN:
2173                                 return $varCache[$index] = $wgContLang->getMonthNameGen( date( 'n', $ts ) );
2174                         case MAG_CURRENTMONTHABBREV:
2175                                 return $varCache[$index] = $wgContLang->getMonthAbbreviation( date( 'n', $ts ) );
2176                         case MAG_CURRENTDAY:
2177                                 return $varCache[$index] = $wgContLang->formatNum( date( 'j', $ts ) );
2178                         case MAG_CURRENTDAY2:
2179                                 return $varCache[$index] = $wgContLang->formatNum( date( 'd', $ts ) );
2180                         case MAG_PAGENAME:
2181                                 return $this->mTitle->getText();
2182                         case MAG_PAGENAMEE:
2183                                 return $this->mTitle->getPartialURL();
2184                         case MAG_FULLPAGENAME:
2185                                 return $this->mTitle->getPrefixedText();
2186                         case MAG_FULLPAGENAMEE:
2187                                 return $this->mTitle->getPrefixedURL();
2188                         case MAG_SUBPAGENAME:
2189                                 return $this->mTitle->getSubpageText();
2190                         case MAG_SUBPAGENAMEE:
2191                                 return $this->mTitle->getSubpageUrlForm();
2192                         case MAG_BASEPAGENAME:
2193                                 return $this->mTitle->getBaseText();
2194                         case MAG_BASEPAGENAMEE:
2195                                 return wfUrlEncode( str_replace( ' ', '_', $this->mTitle->getBaseText() ) );
2196                         case MAG_TALKPAGENAME:
2197                                 if( $this->mTitle->canTalk() ) {
2198                                         $talkPage = $this->mTitle->getTalkPage();
2199                                         return $talkPage->getPrefixedText();
2200                                 } else {
2201                                         return '';
2202                                 }
2203                         case MAG_TALKPAGENAMEE:
2204                                 if( $this->mTitle->canTalk() ) {
2205                                         $talkPage = $this->mTitle->getTalkPage();
2206                                         return $talkPage->getPrefixedUrl();
2207                                 } else {
2208                                         return '';
2209                                 }
2210                         case MAG_SUBJECTPAGENAME:
2211                                 $subjPage = $this->mTitle->getSubjectPage();
2212                                 return $subjPage->getPrefixedText();
2213                         case MAG_SUBJECTPAGENAMEE:
2214                                 $subjPage = $this->mTitle->getSubjectPage();
2215                                 return $subjPage->getPrefixedUrl();
2216                         case MAG_REVISIONID:
2217                                 return $this->mRevisionId;
2218                         case MAG_NAMESPACE:
2219                                 return str_replace('_',' ',$wgContLang->getNsText( $this->mTitle->getNamespace() ) );
2220                         case MAG_NAMESPACEE:
2221                                 return wfUrlencode( $wgContLang->getNsText( $this->mTitle->getNamespace() ) );
2222                         case MAG_TALKSPACE:
2223                                 return $this->mTitle->canTalk() ? str_replace('_',' ',$this->mTitle->getTalkNsText()) : '';
2224                         case MAG_TALKSPACEE:
2225                                 return $this->mTitle->canTalk() ? wfUrlencode( $this->mTitle->getTalkNsText() ) : '';
2226                         case MAG_SUBJECTSPACE:
2227                                 return $this->mTitle->getSubjectNsText();
2228                         case MAG_SUBJECTSPACEE:
2229                                 return( wfUrlencode( $this->mTitle->getSubjectNsText() ) );
2230                         case MAG_CURRENTDAYNAME:
2231                                 return $varCache[$index] = $wgContLang->getWeekdayName( date( 'w', $ts ) + 1 );
2232                         case MAG_CURRENTYEAR:
2233                                 return $varCache[$index] = $wgContLang->formatNum( date( 'Y', $ts ), true );
2234                         case MAG_CURRENTTIME:
2235                                 return $varCache[$index] = $wgContLang->time( wfTimestamp( TS_MW, $ts ), false, false );
2236                         case MAG_CURRENTWEEK:
2237                                 // @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2238                                 // int to remove the padding
2239                                 return $varCache[$index] = $wgContLang->formatNum( (int)date( 'W', $ts ) );
2240                         case MAG_CURRENTDOW:
2241                                 return $varCache[$index] = $wgContLang->formatNum( date( 'w', $ts ) );
2242                         case MAG_NUMBEROFARTICLES:
2243                                 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfArticles() );
2244                         case MAG_NUMBEROFFILES:
2245                                 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfFiles() );
2246                         case MAG_NUMBEROFUSERS:
2247                                 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfUsers() );
2248                         case MAG_NUMBEROFPAGES:
2249                                 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfPages() );
2250                         case MAG_NUMBEROFADMINS:
2251                                 return $varCache[$index]  = $wgContLang->formatNum( wfNumberOfAdmins() );
2252                         case MAG_CURRENTTIMESTAMP:
2253                                 return $varCache[$index] = wfTimestampNow();
2254                         case MAG_CURRENTVERSION:
2255                                 global $wgVersion;
2256                                 return $wgVersion;
2257                         case MAG_SITENAME:
2258                                 return $wgSitename;
2259                         case MAG_SERVER:
2260                                 return $wgServer;
2261                         case MAG_SERVERNAME:
2262                                 return $wgServerName;
2263                         case MAG_SCRIPTPATH:
2264                                 return $wgScriptPath;
2265                         case MAG_DIRECTIONMARK:
2266                                 return $wgContLang->getDirMark();
2267                         case MAG_CONTENTLANGUAGE:
2268                                 global $wgContLanguageCode;
2269                                 return $wgContLanguageCode;
2270                         default:
2271                                 $ret = null;
2272                                 if ( wfRunHooks( 'ParserGetVariableValueSwitch', array( &$this, &$varCache, &$index, &$ret ) ) )
2273                                         return $ret;
2274                                 else
2275                                         return null;
2276                 }
2277         }
2278
2279         /**
2280          * initialise the magic variables (like CURRENTMONTHNAME)
2281          *
2282          * @private
2283          */
2284         function initialiseVariables() {
2285                 $fname = 'Parser::initialiseVariables';
2286                 wfProfileIn( $fname );
2287                 global $wgVariableIDs;
2288                 $this->mVariables = array();
2289                 foreach ( $wgVariableIDs as $id ) {
2290                         $mw =& MagicWord::get( $id );
2291                         $mw->addToArray( $this->mVariables, $id );
2292                 }
2293                 wfProfileOut( $fname );
2294         }
2295
2296         /**
2297          * parse any parentheses in format ((title|part|part))
2298          * and call callbacks to get a replacement text for any found piece
2299          *
2300          * @param string $text The text to parse
2301          * @param array $callbacks rules in form:
2302          *     '{' => array(                            # opening parentheses
2303          *                                      'end' => '}',   # closing parentheses
2304          *                                      'cb' => array(2 => callback,    # replacement callback to call if {{..}} is found
2305          *                                                                4 => callback         # replacement callback to call if {{{{..}}}} is found
2306          *                                                                )
2307          *                                      )
2308          * @private
2309          */
2310         function replace_callback ($text, $callbacks) {
2311                 $openingBraceStack = array();   # this array will hold a stack of parentheses which are not closed yet
2312                 $lastOpeningBrace = -1;                 # last not closed parentheses
2313
2314                 for ($i = 0; $i < strlen($text); $i++) {
2315                         # check for any opening brace
2316                         $rule = null;
2317                         $nextPos = -1;
2318                         foreach ($callbacks as $key => $value) {
2319                                 $pos = strpos ($text, $key, $i);
2320                                 if (false !== $pos && (-1 == $nextPos || $pos < $nextPos)) {
2321                                         $rule = $value;
2322                                         $nextPos = $pos;
2323                                 }
2324                         }
2325
2326                         if ($lastOpeningBrace >= 0) {
2327                                 $pos = strpos ($text, $openingBraceStack[$lastOpeningBrace]['braceEnd'], $i);
2328
2329                                 if (false !== $pos && (-1 == $nextPos || $pos < $nextPos)){
2330                                         $rule = null;
2331                                         $nextPos = $pos;
2332                                 }
2333
2334                                 $pos = strpos ($text, '|', $i);
2335
2336                                 if (false !== $pos && (-1 == $nextPos || $pos < $nextPos)){
2337                                         $rule = null;
2338                                         $nextPos = $pos;
2339                                 }
2340                         }
2341
2342                         if ($nextPos == -1)
2343                                 break;
2344
2345                         $i = $nextPos;
2346
2347                         # found openning brace, lets add it to parentheses stack
2348                         if (null != $rule) {
2349                                 $piece = array('brace' => $text[$i],
2350                                                            'braceEnd' => $rule['end'],
2351                                                            'count' => 1,
2352                                                            'title' => '',
2353                                                            'parts' => null);
2354
2355                                 # count openning brace characters
2356                                 while ($i+1 < strlen($text) && $text[$i+1] == $piece['brace']) {
2357                                         $piece['count']++;
2358                                         $i++;
2359                                 }
2360
2361                                 $piece['startAt'] = $i+1;
2362                                 $piece['partStart'] = $i+1;
2363
2364                                 # we need to add to stack only if openning brace count is enough for any given rule
2365                                 foreach ($rule['cb'] as $cnt => $fn) {
2366                                         if ($piece['count'] >= $cnt) {
2367                                                 $lastOpeningBrace ++;
2368                                                 $openingBraceStack[$lastOpeningBrace] = $piece;
2369                                                 break;
2370                                         }
2371                                 }
2372
2373                                 continue;
2374                         }
2375                         else if ($lastOpeningBrace >= 0) {
2376                                 # first check if it is a closing brace
2377                                 if ($openingBraceStack[$lastOpeningBrace]['braceEnd'] == $text[$i]) {
2378                                         # lets check if it is enough characters for closing brace
2379                                         $count = 1;
2380                                         while ($i+$count < strlen($text) && $text[$i+$count] == $text[$i])
2381                                                 $count++;
2382
2383                                         # if there are more closing parentheses than opening ones, we parse less
2384                                         if ($openingBraceStack[$lastOpeningBrace]['count'] < $count)
2385                                                 $count = $openingBraceStack[$lastOpeningBrace]['count'];
2386
2387                                         # check for maximum matching characters (if there are 5 closing characters, we will probably need only 3 - depending on the rules)
2388                                         $matchingCount = 0;
2389                                         $matchingCallback = null;
2390                                         foreach ($callbacks[$openingBraceStack[$lastOpeningBrace]['brace']]['cb'] as $cnt => $fn) {
2391                                                 if ($count >= $cnt && $matchingCount < $cnt) {
2392                                                         $matchingCount = $cnt;
2393                                                         $matchingCallback = $fn;
2394                                                 }
2395                                         }
2396
2397                                         if ($matchingCount == 0) {
2398                                                 $i += $count - 1;
2399                                                 continue;
2400                                         }
2401
2402                                         # lets set a title or last part (if '|' was found)
2403                                         if (null === $openingBraceStack[$lastOpeningBrace]['parts'])
2404                                                 $openingBraceStack[$lastOpeningBrace]['title'] = substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'], $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2405                                         else
2406                                                 $openingBraceStack[$lastOpeningBrace]['parts'][] = substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'], $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2407
2408                                         $pieceStart = $openingBraceStack[$lastOpeningBrace]['startAt'] - $matchingCount;
2409                                         $pieceEnd = $i + $matchingCount;
2410
2411                                         if( is_callable( $matchingCallback ) ) {
2412                                                 $cbArgs = array (
2413                                                                                  'text' => substr($text, $pieceStart, $pieceEnd - $pieceStart),
2414                                                                                  'title' => trim($openingBraceStack[$lastOpeningBrace]['title']),
2415                                                                                  'parts' => $openingBraceStack[$lastOpeningBrace]['parts'],
2416                                                                                  'lineStart' => (($pieceStart > 0) && ($text[$pieceStart-1] == "\n")),
2417                                                                                  );
2418                                                 # finally we can call a user callback and replace piece of text
2419                                                 $replaceWith = call_user_func( $matchingCallback, $cbArgs );
2420                                                 $text = substr($text, 0, $pieceStart) . $replaceWith . substr($text, $pieceEnd);
2421                                                 $i = $pieceStart + strlen($replaceWith) - 1;
2422                                         }
2423                                         else {
2424                                                 # null value for callback means that parentheses should be parsed, but not replaced
2425                                                 $i += $matchingCount - 1;
2426                                         }
2427
2428                                         # reset last openning parentheses, but keep it in case there are unused characters
2429                                         $piece = array('brace' => $openingBraceStack[$lastOpeningBrace]['brace'],
2430                                                                    'braceEnd' => $openingBraceStack[$lastOpeningBrace]['braceEnd'],
2431                                                                    'count' => $openingBraceStack[$lastOpeningBrace]['count'],
2432                                                                    'title' => '',
2433                                                                    'parts' => null,
2434                                                                    'startAt' => $openingBraceStack[$lastOpeningBrace]['startAt']);
2435                                         $openingBraceStack[$lastOpeningBrace--] = null;
2436
2437                                         if ($matchingCount < $piece['count']) {
2438                                                 $piece['count'] -= $matchingCount;
2439                                                 $piece['startAt'] -= $matchingCount;
2440                                                 $piece['partStart'] = $piece['startAt'];
2441                                                 # do we still qualify for any callback with remaining count?
2442                                                 foreach ($callbacks[$piece['brace']]['cb'] as $cnt => $fn) {
2443                                                         if ($piece['count'] >= $cnt) {
2444                                                                 $lastOpeningBrace ++;
2445                                                                 $openingBraceStack[$lastOpeningBrace] = $piece;
2446                                                                 break;
2447                                                         }
2448                                                 }
2449                                         }
2450                                         continue;
2451                                 }
2452
2453                                 # lets set a title if it is a first separator, or next part otherwise
2454                                 if ($text[$i] == '|') {
2455                                         if (null === $openingBraceStack[$lastOpeningBrace]['parts']) {
2456                                                 $openingBraceStack[$lastOpeningBrace]['title'] = substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'], $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2457                                                 $openingBraceStack[$lastOpeningBrace]['parts'] = array();
2458                                         }
2459                                         else
2460                                                 $openingBraceStack[$lastOpeningBrace]['parts'][] = substr($text, $openingBraceStack[$lastOpeningBrace]['partStart'], $i - $openingBraceStack[$lastOpeningBrace]['partStart']);
2461
2462                                         $openingBraceStack[$lastOpeningBrace]['partStart'] = $i + 1;
2463                                 }
2464                         }
2465                 }
2466
2467                 return $text;
2468         }
2469
2470         /**
2471          * Replace magic variables, templates, and template arguments
2472          * with the appropriate text. Templates are substituted recursively,
2473          * taking care to avoid infinite loops.
2474          *
2475          * Note that the substitution depends on value of $mOutputType:
2476          *  OT_WIKI: only {{subst:}} templates
2477          *  OT_MSG: only magic variables
2478          *  OT_HTML: all templates and magic variables
2479          *
2480          * @param string $tex The text to transform
2481          * @param array $args Key-value pairs representing template parameters to substitute
2482          * @param bool $argsOnly Only do argument (triple-brace) expansion, not double-brace expansion
2483          * @private
2484          */
2485         function replaceVariables( $text, $args = array(), $argsOnly = false ) {
2486                 # Prevent too big inclusions
2487                 if( strlen( $text ) > MAX_INCLUDE_SIZE ) {
2488                         return $text;
2489                 }
2490
2491                 $fname = 'Parser::replaceVariables';
2492                 wfProfileIn( $fname );
2493
2494                 # This function is called recursively. To keep track of arguments we need a stack:
2495                 array_push( $this->mArgStack, $args );
2496
2497                 $braceCallbacks = array();
2498                 if ( !$argsOnly ) {
2499                         $braceCallbacks[2] = array( &$this, 'braceSubstitution' );
2500                 }
2501                 if ( $this->mOutputType == OT_HTML || $this->mOutputType == OT_WIKI ) {
2502                         $braceCallbacks[3] = array( &$this, 'argSubstitution' );
2503                 }
2504                 $callbacks = array();
2505                 $callbacks['{'] = array('end' => '}', 'cb' => $braceCallbacks);
2506                 $callbacks['['] = array('end' => ']', 'cb' => array(2=>null));
2507                 $text = $this->replace_callback ($text, $callbacks);
2508
2509                 array_pop( $this->mArgStack );
2510
2511                 wfProfileOut( $fname );
2512                 return $text;
2513         }
2514
2515         /**
2516          * Replace magic variables
2517          * @private
2518          */
2519         function variableSubstitution( $matches ) {
2520                 $fname = 'Parser::variableSubstitution';
2521                 $varname = $matches[1];
2522                 wfProfileIn( $fname );
2523                 if ( !$this->mVariables ) {
2524                         $this->initialiseVariables();
2525                 }
2526                 $skip = false;
2527                 if ( $this->mOutputType == OT_WIKI ) {
2528                         # Do only magic variables prefixed by SUBST
2529                         $mwSubst =& MagicWord::get( MAG_SUBST );
2530                         if (!$mwSubst->matchStartAndRemove( $varname ))
2531                                 $skip = true;
2532                         # Note that if we don't substitute the variable below,
2533                         # we don't remove the {{subst:}} magic word, in case
2534                         # it is a template rather than a magic variable.
2535                 }
2536                 if ( !$skip && array_key_exists( $varname, $this->mVariables ) ) {
2537                         $id = $this->mVariables[$varname];
2538                         $text = $this->getVariableValue( $id );
2539                         $this->mOutput->mContainsOldMagic = true;
2540                 } else {
2541                         $text = $matches[0];
2542                 }
2543                 wfProfileOut( $fname );
2544                 return $text;
2545         }
2546
2547         # Split template arguments
2548         function getTemplateArgs( $argsString ) {
2549                 if ( $argsString === '' ) {
2550                         return array();
2551                 }
2552
2553                 $args = explode( '|', substr( $argsString, 1 ) );
2554
2555                 # If any of the arguments contains a '[[' but no ']]', it needs to be
2556                 # merged with the next arg because the '|' character between belongs
2557                 # to the link syntax and not the template parameter syntax.
2558                 $argc = count($args);
2559
2560                 for ( $i = 0; $i < $argc-1; $i++ ) {
2561                         if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
2562                                 $args[$i] .= '|'.$args[$i+1];
2563                                 array_splice($args, $i+1, 1);
2564                                 $i--;
2565                                 $argc--;
2566                         }
2567                 }
2568
2569                 return $args;
2570         }
2571
2572         /**
2573          * Return the text of a template, after recursively
2574          * replacing any variables or templates within the template.
2575          *
2576          * @param array $piece The parts of the template
2577          *  $piece['text']: matched text
2578          *  $piece['title']: the title, i.e. the part before the |
2579          *  $piece['parts']: the parameter array
2580          * @return string the text of the template
2581          * @private
2582          */
2583         function braceSubstitution( $piece ) {
2584                 global $wgContLang, $wgLang, $wgAllowDisplayTitle, $action;
2585                 $fname = 'Parser::braceSubstitution';
2586                 wfProfileIn( $fname );
2587
2588                 # Flags
2589                 $found = false;             # $text has been filled
2590                 $nowiki = false;            # wiki markup in $text should be escaped
2591                 $noparse = false;           # Unsafe HTML tags should not be stripped, etc.
2592                 $noargs = false;            # Don't replace triple-brace arguments in $text
2593                 $replaceHeadings = false;   # Make the edit section links go to the template not the article
2594                 $isHTML = false;            # $text is HTML, armour it against wikitext transformation
2595                 $forceRawInterwiki = false; # Force interwiki transclusion to be done in raw mode not rendered
2596
2597                 # Title object, where $text came from
2598                 $title = NULL;
2599
2600                 $linestart = '';
2601
2602                 # $part1 is the bit before the first |, and must contain only title characters
2603                 # $args is a list of arguments, starting from index 0, not including $part1
2604
2605                 $part1 = $piece['title'];
2606                 # If the third subpattern matched anything, it will start with |
2607
2608                 if (null == $piece['parts']) {
2609                         $replaceWith = $this->variableSubstitution (array ($piece['text'], $piece['title']));
2610                         if ($replaceWith != $piece['text']) {
2611                                 $text = $replaceWith;
2612                                 $found = true;
2613                                 $noparse = true;
2614                                 $noargs = true;
2615                         }
2616                 }
2617
2618                 $args = (null == $piece['parts']) ? array() : $piece['parts'];
2619                 $argc = count( $args );
2620
2621                 # SUBST
2622                 if ( !$found ) {
2623                         $mwSubst =& MagicWord::get( MAG_SUBST );
2624                         if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType == OT_WIKI) ) {
2625                                 # One of two possibilities is true:
2626                                 # 1) Found SUBST but not in the PST phase
2627                                 # 2) Didn't find SUBST and in the PST phase
2628                                 # In either case, return without further processing
2629                                 $text = $piece['text'];
2630                                 $found = true;
2631                                 $noparse = true;
2632                                 $noargs = true;
2633                         }
2634                 }
2635
2636                 # MSG, MSGNW, INT and RAW
2637                 if ( !$found ) {
2638                         # Check for MSGNW:
2639                         $mwMsgnw =& MagicWord::get( MAG_MSGNW );
2640                         if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
2641                                 $nowiki = true;
2642                         } else {
2643                                 # Remove obsolete MSG:
2644                                 $mwMsg =& MagicWord::get( MAG_MSG );
2645                                 $mwMsg->matchStartAndRemove( $part1 );
2646                         }
2647
2648                         # Check for RAW:
2649                         $mwRaw =& MagicWord::get( MAG_RAW );
2650                         if ( $mwRaw->matchStartAndRemove( $part1 ) ) {
2651                                 $forceRawInterwiki = true;
2652                         }
2653
2654                         # Check if it is an internal message
2655                         $mwInt =& MagicWord::get( MAG_INT );
2656                         if ( $mwInt->matchStartAndRemove( $part1 ) ) {
2657                                 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
2658                                         $text = $linestart . wfMsgReal( $part1, $args, true );
2659                                         $found = true;
2660                                 }
2661                         }
2662                 }
2663
2664                 # NS
2665                 if ( !$found ) {
2666                         # Check for NS: (namespace expansion)
2667                         $mwNs = MagicWord::get( MAG_NS );
2668                         if ( $mwNs->matchStartAndRemove( $part1 ) ) {
2669                                 if ( intval( $part1 ) || $part1 == "0" ) {
2670                                         $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
2671                                         $found = true;
2672                                 } else {
2673                                         $param = str_replace( ' ', '_', strtolower( $part1 ) );
2674                                         $index = Namespace::getCanonicalIndex( strtolower( $param ) );
2675                                         if ( !is_null( $index ) ) {
2676                                                 $text = $linestart . $wgContLang->getNsText( $index );
2677                                                 $found = true;
2678                                         }
2679                                 }
2680                         }
2681                 }
2682
2683                 # URLENCODE
2684                 if( !$found ) {
2685                         $urlencode =& MagicWord::get( MAG_URLENCODE );
2686                         if( $urlencode->matchStartAndRemove( $part1 ) ) {
2687                                 $text = $linestart . urlencode( $part1 );
2688                                 $found = true;
2689                         }
2690                 }
2691
2692                 # LCFIRST, UCFIRST, LC and UC
2693                 if ( !$found ) {
2694                         $lcfirst =& MagicWord::get( MAG_LCFIRST );
2695                         $ucfirst =& MagicWord::get( MAG_UCFIRST );
2696                         $lc =& MagicWord::get( MAG_LC );
2697                         $uc =& MagicWord::get( MAG_UC );
2698                         if ( $lcfirst->matchStartAndRemove( $part1 ) ) {
2699                                 $text = $linestart . $wgContLang->lcfirst( $part1 );
2700                                 $found = true;
2701                         } else if ( $ucfirst->matchStartAndRemove( $part1 ) ) {
2702                                 $text = $linestart . $wgContLang->ucfirst( $part1 );
2703                                 $found = true;
2704                         } else if ( $lc->matchStartAndRemove( $part1 ) ) {
2705                                 $text = $linestart . $wgContLang->lc( $part1 );
2706                                 $found = true;
2707                         } else if ( $uc->matchStartAndRemove( $part1 ) ) {
2708                                  $text = $linestart . $wgContLang->uc( $part1 );
2709                                  $found = true;
2710                         }
2711                 }
2712
2713                 # LOCALURL and FULLURL
2714                 if ( !$found ) {
2715                         $mwLocal =& MagicWord::get( MAG_LOCALURL );
2716                         $mwLocalE =& MagicWord::get( MAG_LOCALURLE );
2717                         $mwFull =& MagicWord::get( MAG_FULLURL );
2718                         $mwFullE =& MagicWord::get( MAG_FULLURLE );
2719
2720
2721                         if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
2722                                 $func = 'getLocalURL';
2723                         } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
2724                                 $func = 'escapeLocalURL';
2725                         } elseif ( $mwFull->matchStartAndRemove( $part1 ) ) {
2726                                 $func = 'getFullURL';
2727                         } elseif ( $mwFullE->matchStartAndRemove( $part1 ) ) {
2728                                 $func = 'escapeFullURL';
2729                         } else {
2730                                 $func = false;
2731                         }
2732
2733                         if ( $func !== false ) {
2734                                 $title = Title::newFromText( $part1 );
2735                                 # Due to order of execution of a lot of bits, the values might be encoded
2736                                 # before arriving here; if that's true, then the title can't be created
2737                                 # and the variable will fail. If we can't get a decent title from the first
2738                                 # attempt, url-decode and try for a second.
2739                                 if( is_null( $title ) )
2740                                         $title = Title::newFromUrl( urldecode( $part1 ) );
2741                                 if ( !is_null( $title ) ) {
2742                                         if ( $argc > 0 ) {
2743                                                 $text = $linestart . $title->$func( $args[0] );
2744                                         } else {
2745                                                 $text = $linestart . $title->$func();
2746                                         }
2747                                         $found = true;
2748                                 }
2749                         }
2750                 }
2751
2752                 $lang = $this->mOptions->getInterfaceMessage() ? $wgLang : $wgContLang;
2753                 if ( !$found && $argc == 1 ) {
2754                         $mwGrammar =& MagicWord::get( MAG_FORMATNUM );
2755                         if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
2756                                 $text = $linestart . $lang->formatNum( $args[0] );
2757                                 $found = true;
2758                         }
2759                 }
2760
2761
2762                 # GRAMMAR
2763                 if ( !$found && $argc == 1 ) {
2764                         $mwGrammar =& MagicWord::get( MAG_GRAMMAR );
2765                         if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
2766                                 $text = $linestart . $lang->convertGrammar( $args[0], $part1 );
2767                                 $found = true;
2768                         }
2769                 }
2770
2771                 # PLURAL
2772                 if ( !$found && $argc >= 2 ) {
2773                         $mwPluralForm =& MagicWord::get( MAG_PLURAL );
2774                         if ( $mwPluralForm->matchStartAndRemove( $part1 ) ) {
2775                                 while ( count($args) < 5 ) { $args[] = $args[count($args)-1]; }
2776                                 $text = $linestart . $lang->convertPlural( $part1, $args[0], $args[1],
2777                                         $args[2], $args[3], $args[4]);
2778                                 $found = true;
2779                         }
2780                 }
2781
2782                 # DISPLAYTITLE
2783                 if ( !$found && $argc == 1 && $wgAllowDisplayTitle ) {
2784                         $mwDT =& MagicWord::get( MAG_DISPLAYTITLE );
2785                         if ( $mwDT->matchStartAndRemove( $part1 ) ) {
2786
2787                                 # Set title in parser output object
2788                                 $param = $args[0];
2789                                 $parserOptions = new ParserOptions;
2790                                 $local_parser = new Parser ();
2791                                 $t2 = $local_parser->parse ( $param, $this->mTitle, $parserOptions, false );
2792                                 $this->mOutput->mHTMLtitle = $t2->GetText();
2793
2794                                 # Add subtitle
2795                                 $t = $this->mTitle->getPrefixedText();
2796                                 $this->mOutput->mSubtitle .= wfMsg('displaytitle', $t);
2797                                 $text = "" ;
2798                                 $found = true ;
2799                         }
2800                 }
2801
2802                 # NUMBEROFPAGES, NUMBEROFUSERS, NUMBEROFARTICLES, and NUMBEROFFILES
2803                 if( !$found ) {
2804                         $mwWordsToCheck = array( MAG_NUMBEROFPAGES => 'wfNumberOfPages',
2805                                                                          MAG_NUMBEROFUSERS => 'wfNumberOfUsers',
2806                                                                          MAG_NUMBEROFARTICLES => 'wfNumberOfArticles',
2807                                                                          MAG_NUMBEROFFILES => 'wfNumberOfFiles',
2808                                                                          MAG_NUMBEROFADMINS => 'wfNumberOfAdmins' );
2809                         foreach( $mwWordsToCheck as $word => $func ) {
2810                                 $mwCurrentWord =& MagicWord::get( $word );
2811                                 if( $mwCurrentWord->matchStartAndRemove( $part1 ) ) {
2812                                         $mwRawSuffix =& MagicWord::get( MAG_RAWSUFFIX );
2813                                         if( isset( $args[0] ) && $mwRawSuffix->match( $args[0] ) ) {
2814                                                 # Raw and unformatted
2815                                                 $text = $linestart . call_user_func( $func );
2816                                         } else {
2817                                                 # Formatted according to the content default
2818                                                 $text = $linestart . $wgContLang->formatNum( call_user_func( $func ) );
2819                                         }
2820                                         $found = true;
2821                                 }
2822                         }
2823                 }
2824
2825                 # PAGESINNAMESPACE
2826                 if( !$found ) {
2827                         $mwPagesInNs =& MagicWord::get( MAG_PAGESINNAMESPACE );
2828                         if( $mwPagesInNs->matchStartAndRemove( $part1 ) ) {
2829                                 $found = true;
2830                                 $count = wfPagesInNs( intval( $part1 ) );
2831                                 $mwRawSuffix =& MagicWord::get( MAG_RAWSUFFIX );
2832                                 if( isset( $args[0] ) && $mwRawSuffix->match( $args[0] ) ) {
2833                                         $text = $linestart . $count;
2834                                 } else {
2835                                         $text = $linestart . $wgContLang->formatNum( $count );
2836                                 }
2837                         }
2838                 }
2839
2840                 # #LANGUAGE:
2841                 if( !$found ) {
2842                         $mwLanguage =& MagicWord::get( MAG_LANGUAGE );
2843                         if( $mwLanguage->matchStartAndRemove( $part1 ) ) {
2844                                 $lang = $wgContLang->getLanguageName( strtolower( $part1 ) );
2845                                 $text = $linestart . ( $lang != '' ? $lang : $part1 );
2846                                 $found = true;
2847                         }
2848                 }
2849
2850                 # Extensions
2851                 if ( !$found && substr( $part1, 0, 1 ) == '#' ) {
2852                         $colonPos = strpos( $part1, ':' );
2853                         if ( $colonPos !== false ) {
2854                                 $function = strtolower( substr( $part1, 1, $colonPos - 1 ) );
2855                                 if ( isset( $this->mFunctionSynonyms[$function] ) ) {
2856                                         $function = $this->mFunctionSynonyms[$function];
2857                                         $funcArgs = array_map( 'trim', $args );
2858                                         $funcArgs = array_merge( array( &$this, trim( substr( $part1, $colonPos + 1 ) ) ), $funcArgs );
2859                                         $result = call_user_func_array( $this->mFunctionHooks[$function], $funcArgs );
2860                                         $found = true;
2861
2862                                         // The text is usually already parsed, doesn't need triple-brace tags expanded, etc.
2863                                         //$noargs = true;
2864                                         //$noparse = true;
2865
2866                                         if ( is_array( $result ) ) {
2867                                                 $text = $linestart . $result[0];
2868                                                 unset( $result[0] );
2869
2870                                                 // Extract flags into the local scope
2871                                                 // This allows callers to set flags such as nowiki, noparse, found, etc.
2872                                                 extract( $result );
2873                                         } else {
2874                                                 $text = $linestart . $result;
2875                                         }
2876                                 }
2877                         }
2878                 }
2879
2880                 # Template table test
2881
2882                 # Did we encounter this template already? If yes, it is in the cache
2883                 # and we need to check for loops.
2884                 if ( !$found && isset( $this->mTemplates[$piece['title']] ) ) {
2885                         $found = true;
2886
2887                         # Infinite loop test
2888                         if ( isset( $this->mTemplatePath[$part1] ) ) {
2889                                 $noparse = true;
2890                                 $noargs = true;
2891                                 $found = true;
2892                                 $text = $linestart .
2893                                         '{{' . $part1 . '}}' .
2894                                         '<!-- WARNING: template loop detected -->';
2895                                 wfDebug( "$fname: template loop broken at '$part1'\n" );
2896                         } else {
2897                                 # set $text to cached message.
2898                                 $text = $linestart . $this->mTemplates[$piece['title']];
2899                         }
2900                 }
2901
2902                 # Load from database
2903                 $lastPathLevel = $this->mTemplatePath;
2904                 if ( !$found ) {
2905                         $ns = NS_TEMPLATE;
2906                         # declaring $subpage directly in the function call
2907                         # does not work correctly with references and breaks
2908                         # {{/subpage}}-style inclusions
2909                         $subpage = '';
2910                         $part1 = $this->maybeDoSubpageLink( $part1, $subpage );
2911                         if ($subpage !== '') {
2912                                 $ns = $this->mTitle->getNamespace();
2913                         }
2914                         $title = Title::newFromText( $part1, $ns );
2915
2916
2917                         if ( !is_null( $title ) ) {
2918                                 $checkVariantLink = sizeof($wgContLang->getVariants())>1;
2919                                 # Check for language variants if the template is not found
2920                                 if($checkVariantLink && $title->getArticleID() == 0){
2921                                         $wgContLang->findVariantLink($part1, $title);
2922                                 }
2923
2924                                 if ( !$title->isExternal() ) {
2925                                         # Check for excessive inclusion
2926                                         $dbk = $title->getPrefixedDBkey();
2927                                         if ( $this->incrementIncludeCount( $dbk ) ) {
2928                                                 if ( $title->getNamespace() == NS_SPECIAL && $this->mOptions->getAllowSpecialInclusion() && $this->mOutputType != OT_WIKI ) {
2929                                                         $text = SpecialPage::capturePath( $title );
2930                                                         if ( is_string( $text ) ) {
2931                                                                 $found = true;
2932                                                                 $noparse = true;
2933                                                                 $noargs = true;
2934                                                                 $isHTML = true;
2935                                                                 $this->disableCache();
2936                                                         }
2937                                                 } else {
2938                                                         $articleContent = $this->fetchTemplate( $title );
2939                                                         if ( $articleContent !== false ) {
2940                                                                 $found = true;
2941                                                                 $text = $articleContent;
2942                                                                 $replaceHeadings = true;
2943                                                         }
2944                                                 }
2945                                         }
2946
2947                                         # If the title is valid but undisplayable, make a link to it
2948                                         if ( $this->mOutputType == OT_HTML && !$found ) {
2949                                                 $text = '[['.$title->getPrefixedText().']]';
2950                                                 $found = true;
2951                                         }
2952                                 } elseif ( $title->isTrans() ) {
2953                                         // Interwiki transclusion
2954                                         if ( $this->mOutputType == OT_HTML && !$forceRawInterwiki ) {
2955                                                 $text = $this->interwikiTransclude( $title, 'render' );
2956                                                 $isHTML = true;
2957                                                 $noparse = true;
2958                                         } else {
2959                                                 $text = $this->interwikiTransclude( $title, 'raw' );
2960                                                 $replaceHeadings = true;
2961                                         }
2962                                         $found = true;
2963                                 }
2964
2965                                 # Template cache array insertion
2966                                 # Use the original $piece['title'] not the mangled $part1, so that
2967                                 # modifiers such as RAW: produce separate cache entries
2968                                 if( $found ) {
2969                                         if( $isHTML ) {
2970                                                 // A special page; don't store it in the template cache.
2971                                         } else {
2972                                                 $this->mTemplates[$piece['title']] = $text;
2973                                         }
2974                                         $text = $linestart . $text;
2975                                 }
2976                         }
2977                 }
2978
2979                 # Recursive parsing, escaping and link table handling
2980                 # Only for HTML output
2981                 if ( $nowiki && $found && $this->mOutputType == OT_HTML ) {
2982                         $text = wfEscapeWikiText( $text );
2983                 } elseif ( ($this->mOutputType == OT_HTML || $this->mOutputType == OT_WIKI) && $found ) {
2984                         if ( $noargs ) {
2985                                 $assocArgs = array();
2986                         } else {
2987                                 # Clean up argument array
2988                                 $assocArgs = array();
2989                                 $index = 1;
2990                                 foreach( $args as $arg ) {
2991                                         $eqpos = strpos( $arg, '=' );
2992                                         if ( $eqpos === false ) {
2993                                                 $assocArgs[$index++] = $arg;
2994                                         } else {
2995                                                 $name = trim( substr( $arg, 0, $eqpos ) );
2996                                                 $value = trim( substr( $arg, $eqpos+1 ) );
2997                                                 if ( $value === false ) {
2998                                                         $value = '';
2999                                                 }
3000                                                 if ( $name !== false ) {
3001                                                         $assocArgs[$name] = $value;
3002                                                 }
3003                                         }
3004                                 }
3005
3006                                 # Add a new element to the templace recursion path
3007                                 $this->mTemplatePath[$part1] = 1;
3008                         }
3009
3010                         if ( !$noparse ) {
3011                                 # If there are any <onlyinclude> tags, only include them
3012                                 if ( in_string( '<onlyinclude>', $text ) && in_string( '</onlyinclude>', $text ) ) {
3013                                         preg_match_all( '/<onlyinclude>(.*?)\n?<\/onlyinclude>/s', $text, $m );
3014                                         $text = '';
3015                                         foreach ($m[1] as $piece)
3016                                                 $text .= $piece;
3017                                 }
3018                                 # Remove <noinclude> sections and <includeonly> tags
3019                                 $text = preg_replace( '/<noinclude>.*?<\/noinclude>/s', '', $text );
3020                                 $text = strtr( $text, array( '<includeonly>' => '' , '</includeonly>' => '' ) );
3021
3022                                 if( $this->mOutputType == OT_HTML ) {
3023                                         # Strip <nowiki>, <pre>, etc.
3024                                         $text = $this->strip( $text, $this->mStripState );
3025                                         $text = Sanitizer::removeHTMLtags( $text, array( &$this, 'replaceVariables' ), $assocArgs );
3026                                 }
3027                                 $text = $this->replaceVariables( $text, $assocArgs );
3028
3029                                 # If the template begins with a table or block-level
3030                                 # element, it should be treated as beginning a new line.
3031                                 if (!$piece['lineStart'] && preg_match('/^({\\||:|;|#|\*)/', $text)) {
3032                                         $text = "\n" . $text;
3033                                 }
3034                         } elseif ( !$noargs ) {
3035                                 # $noparse and !$noargs
3036                                 # Just replace the arguments, not any double-brace items
3037                                 # This is used for rendered interwiki transclusion
3038                                 $text = $this->replaceVariables( $text, $assocArgs, true );
3039                         }
3040                 }
3041                 # Prune lower levels off the recursion check path
3042                 $this->mTemplatePath = $lastPathLevel;
3043
3044                 if ( !$found ) {
3045                         wfProfileOut( $fname );
3046                         return $piece['text'];
3047                 } else {
3048                         if ( $isHTML ) {
3049                                 # Replace raw HTML by a placeholder
3050                                 # Add a blank line preceding, to prevent it from mucking up
3051                                 # immediately preceding headings
3052                                 $text = "\n\n" . $this->insertStripItem( $text, $this->mStripState );
3053                         } else {
3054                                 # replace ==section headers==
3055                                 # XXX this needs to go away once we have a better parser.
3056                                 if ( $this->mOutputType != OT_WIKI && $replaceHeadings ) {
3057                                         if( !is_null( $title ) )
3058                                                 $encodedname = base64_encode($title->getPrefixedDBkey());
3059                                         else
3060                                                 $encodedname = base64_encode("");
3061                                         $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
3062                                                 PREG_SPLIT_DELIM_CAPTURE);
3063                                         $text = '';
3064                                         $nsec = 0;
3065                                         for( $i = 0; $i < count($m); $i += 2 ) {
3066                                                 $text .= $m[$i];
3067                                                 if (!isset($m[$i + 1]) || $m[$i + 1] == "") continue;
3068                                                 $hl = $m[$i + 1];
3069                                                 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
3070                                                         $text .= $hl;
3071                                                         continue;
3072                                                 }
3073                                                 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
3074                                                 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
3075                                                         . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
3076
3077                                                 $nsec++;
3078                                         }
3079                                 }
3080                         }
3081                 }
3082
3083                 # Prune lower levels off the recursion check path
3084                 $this->mTemplatePath = $lastPathLevel;
3085
3086                 if ( !$found ) {
3087                         wfProfileOut( $fname );
3088                         return $piece['text'];
3089                 } else {
3090                         wfProfileOut( $fname );
3091                         return $text;
3092                 }
3093         }
3094
3095         /**
3096          * Fetch the unparsed text of a template and register a reference to it.
3097          */
3098         function fetchTemplate( $title ) {
3099                 $text = false;
3100                 // Loop to fetch the article, with up to 1 redirect
3101                 for ( $i = 0; $i < 2 && is_object( $title ); $i++ ) {
3102                         $rev = Revision::newFromTitle( $title );
3103                         $this->mOutput->addTemplate( $title, $title->getArticleID() );
3104                         if ( !$rev ) {
3105                                 break;
3106                         }
3107                         $text = $rev->getText();
3108                         if ( $text === false ) {
3109                                 break;
3110                         }
3111                         // Redirect?
3112                         $title = Title::newFromRedirect( $text );
3113                 }
3114                 return $text;
3115         }
3116
3117         /**
3118          * Transclude an interwiki link.
3119          */
3120         function interwikiTransclude( $title, $action ) {
3121                 global $wgEnableScaryTranscluding, $wgCanonicalNamespaceNames;
3122
3123                 if (!$wgEnableScaryTranscluding)
3124                         return wfMsg('scarytranscludedisabled');
3125
3126                 // The namespace will actually only be 0 or 10, depending on whether there was a leading :
3127                 // But we'll handle it generally anyway
3128                 if ( $title->getNamespace() ) {
3129                         // Use the canonical namespace, which should work anywhere
3130                         $articleName = $wgCanonicalNamespaceNames[$title->getNamespace()] . ':' . $title->getDBkey();
3131                 } else {
3132                         $articleName = $title->getDBkey();
3133                 }
3134
3135                 $url = str_replace('$1', urlencode($articleName), Title::getInterwikiLink($title->getInterwiki()));
3136                 $url .= "?action=$action";
3137                 if (strlen($url) > 255)
3138                         return wfMsg('scarytranscludetoolong');
3139                 return $this->fetchScaryTemplateMaybeFromCache($url);
3140         }
3141
3142         function fetchScaryTemplateMaybeFromCache($url) {
3143                 global $wgTranscludeCacheExpiry;
3144                 $dbr =& wfGetDB(DB_SLAVE);
3145                 $obj = $dbr->selectRow('transcache', array('tc_time', 'tc_contents'),
3146                                 array('tc_url' => $url));
3147                 if ($obj) {
3148                         $time = $obj->tc_time;
3149                         $text = $obj->tc_contents;
3150                         if ($time && time() < $time + $wgTranscludeCacheExpiry ) {
3151                                 return $text;
3152                         }
3153                 }
3154
3155                 $text = Http::get($url);
3156                 if (!$text)
3157                         return wfMsg('scarytranscludefailed', $url);
3158
3159                 $dbw =& wfGetDB(DB_MASTER);
3160                 $dbw->replace('transcache', array('tc_url'), array(
3161                         'tc_url' => $url,
3162                         'tc_time' => time(),
3163                         'tc_contents' => $text));
3164                 return $text;
3165         }
3166
3167
3168         /**
3169          * Triple brace replacement -- used for template arguments
3170          * @private
3171          */
3172         function argSubstitution( $matches ) {
3173                 $arg = trim( $matches['title'] );
3174                 $text = $matches['text'];
3175                 $inputArgs = end( $this->mArgStack );
3176
3177                 if ( array_key_exists( $arg, $inputArgs ) ) {
3178                         $text = $inputArgs[$arg];
3179                 } else if ($this->mOutputType == OT_HTML && null != $matches['parts'] && count($matches['parts']) > 0) {
3180                         $text = $matches['parts'][0];
3181                 }
3182
3183                 return $text;
3184         }
3185
3186         /**
3187          * Returns true if the function is allowed to include this entity
3188          * @private
3189          */
3190         function incrementIncludeCount( $dbk ) {
3191                 if ( !array_key_exists( $dbk, $this->mIncludeCount ) ) {
3192                         $this->mIncludeCount[$dbk] = 0;
3193                 }
3194                 if ( ++$this->mIncludeCount[$dbk] <= MAX_INCLUDE_REPEAT ) {
3195                         return true;
3196                 } else {
3197                         return false;
3198                 }
3199         }
3200
3201         /**
3202          * Detect __NOGALLERY__ magic word and set a placeholder
3203          */
3204         function stripNoGallery( &$text ) {
3205                 # if the string __NOGALLERY__ (not case-sensitive) occurs in the HTML,
3206                 # do not add TOC
3207                 $mw = MagicWord::get( MAG_NOGALLERY );
3208                 $this->mOutput->mNoGallery = $mw->matchAndRemove( $text ) ;
3209         }
3210
3211         /**
3212          * Detect __TOC__ magic word and set a placeholder
3213          */
3214         function stripToc( $text ) {
3215                 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
3216                 # do not add TOC
3217                 $mw = MagicWord::get( MAG_NOTOC );
3218                 if( $mw->matchAndRemove( $text ) ) {
3219                         $this->mShowToc = false;
3220                 }
3221
3222                 $mw = MagicWord::get( MAG_TOC );
3223                 if( $mw->match( $text ) ) {
3224                         $this->mShowToc = true;
3225                         $this->mForceTocPosition = true;
3226
3227                         // Set a placeholder. At the end we'll fill it in with the TOC.
3228                         $text = $mw->replace( '<!--MWTOC-->', $text, 1 );
3229
3230                         // Only keep the first one.
3231                         $text = $mw->replace( '', $text );
3232                 }
3233                 return $text;
3234         }
3235
3236         /**
3237          * This function accomplishes several tasks:
3238          * 1) Auto-number headings if that option is enabled
3239          * 2) Add an [edit] link to sections for logged in users who have enabled the option
3240          * 3) Add a Table of contents on the top for users who have enabled the option
3241          * 4) Auto-anchor headings
3242          *
3243          * It loops through all headlines, collects the necessary data, then splits up the
3244          * string and re-inserts the newly formatted headlines.
3245          *
3246          * @param string $text
3247          * @param boolean $isMain
3248          * @private
3249          */
3250         function formatHeadings( $text, $isMain=true ) {
3251                 global $wgMaxTocLevel, $wgContLang;
3252
3253                 $doNumberHeadings = $this->mOptions->getNumberHeadings();
3254                 if( !$this->mTitle->userCanEdit() ) {
3255                         $showEditLink = 0;
3256                 } else {
3257                         $showEditLink = $this->mOptions->getEditSection();
3258                 }
3259
3260                 # Inhibit editsection links if requested in the page
3261                 $esw =& MagicWord::get( MAG_NOEDITSECTION );
3262                 if( $esw->matchAndRemove( $text ) ) {
3263                         $showEditLink = 0;
3264                 }
3265
3266                 # Get all headlines for numbering them and adding funky stuff like [edit]
3267                 # links - this is for later, but we need the number of headlines right now
3268                 $numMatches = preg_match_all( '/<H([1-6])(.*?'.'>)(.*?)<\/H[1-6] *>/i', $text, $matches );
3269
3270                 # if there are fewer than 4 headlines in the article, do not show TOC
3271                 # unless it's been explicitly enabled.
3272                 $enoughToc = $this->mShowToc &&
3273                         (($numMatches >= 4) || $this->mForceTocPosition);
3274
3275                 # Allow user to stipulate that a page should have a "new section"
3276                 # link added via __NEWSECTIONLINK__
3277                 $mw =& MagicWord::get( MAG_NEWSECTIONLINK );
3278                 if( $mw->matchAndRemove( $text ) )
3279                         $this->mOutput->setNewSection( true );
3280
3281                 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
3282                 # override above conditions and always show TOC above first header
3283                 $mw =& MagicWord::get( MAG_FORCETOC );
3284                 if ($mw->matchAndRemove( $text ) ) {
3285                         $this->mShowToc = true;
3286                         $enoughToc = true;
3287                 }
3288
3289                 # Never ever show TOC if no headers
3290                 if( $numMatches < 1 ) {
3291                         $enoughToc = false;
3292                 }
3293
3294                 # We need this to perform operations on the HTML
3295                 $sk =& $this->mOptions->getSkin();
3296
3297                 # headline counter
3298                 $headlineCount = 0;
3299                 $sectionCount = 0; # headlineCount excluding template sections
3300
3301                 # Ugh .. the TOC should have neat indentation levels which can be
3302                 # passed to the skin functions. These are determined here
3303                 $toc = '';
3304                 $full = '';
3305                 $head = array();
3306                 $sublevelCount = array();
3307                 $levelCount = array();
3308                 $toclevel = 0;
3309                 $level = 0;
3310                 $prevlevel = 0;
3311                 $toclevel = 0;
3312                 $prevtoclevel = 0;
3313
3314                 foreach( $matches[3] as $headline ) {
3315                         $istemplate = 0;
3316                         $templatetitle = '';
3317                         $templatesection = 0;
3318                         $numbering = '';
3319
3320                         if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
3321                                 $istemplate = 1;
3322                                 $templatetitle = base64_decode($mat[1]);
3323                                 $templatesection = 1 + (int)base64_decode($mat[2]);
3324                                 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
3325                         }
3326
3327                         if( $toclevel ) {
3328                                 $prevlevel = $level;
3329                                 $prevtoclevel = $toclevel;
3330                         }
3331                         $level = $matches[1][$headlineCount];
3332
3333                         if( $doNumberHeadings || $enoughToc ) {
3334
3335                                 if ( $level > $prevlevel ) {
3336                                         # Increase TOC level
3337                                         $toclevel++;
3338                                         $sublevelCount[$toclevel] = 0;
3339                                         if( $toclevel<$wgMaxTocLevel ) {
3340                                                 $toc .= $sk->tocIndent();
3341                                         }
3342                                 }
3343                                 elseif ( $level < $prevlevel && $toclevel > 1 ) {
3344                                         # Decrease TOC level, find level to jump to
3345
3346                                         if ( $toclevel == 2 && $level <= $levelCount[1] ) {
3347                                                 # Can only go down to level 1
3348                                                 $toclevel = 1;
3349                                         } else {
3350                                                 for ($i = $toclevel; $i > 0; $i--) {
3351                                                         if ( $levelCount[$i] == $level ) {
3352                                                                 # Found last matching level
3353                                                                 $toclevel = $i;
3354                                                                 break;
3355                                                         }
3356                                                         elseif ( $levelCount[$i] < $level ) {
3357                                                                 # Found first matching level below current level
3358                                                                 $toclevel = $i + 1;
3359                                                                 break;
3360                                                         }
3361                                                 }
3362                                         }
3363                                         if( $toclevel<$wgMaxTocLevel ) {
3364                                                 $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
3365                                         }
3366                                 }
3367                                 else {
3368                                         # No change in level, end TOC line
3369                                         if( $toclevel<$wgMaxTocLevel ) {
3370                                                 $toc .= $sk->tocLineEnd();
3371                                         }
3372                                 }
3373
3374                                 $levelCount[$toclevel] = $level;
3375
3376                                 # count number of headlines for each level
3377                                 @$sublevelCount[$toclevel]++;
3378                                 $dot = 0;
3379                                 for( $i = 1; $i <= $toclevel; $i++ ) {
3380                                         if( !empty( $sublevelCount[$i] ) ) {
3381                                                 if( $dot ) {
3382                                                         $numbering .= '.';
3383                                                 }
3384                                                 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
3385                                                 $dot = 1;
3386                                         }
3387                                 }
3388                         }
3389
3390                         # The canonized header is a version of the header text safe to use for links
3391                         # Avoid insertion of weird stuff like <math> by expanding the relevant sections
3392                         $canonized_headline = $this->unstrip( $headline, $this->mStripState );
3393                         $canonized_headline = $this->unstripNoWiki( $canonized_headline, $this->mStripState );
3394
3395                         # Remove link placeholders by the link text.
3396                         #     <!--LINK number-->
3397                         # turns into
3398                         #     link text with suffix
3399                         $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
3400                                                             "\$this->mLinkHolders['texts'][\$1]",
3401                                                             $canonized_headline );
3402                         $canonized_headline = preg_replace( '/<!--IWLINK ([0-9]*)-->/e',
3403                                                             "\$this->mInterwikiLinkHolders['texts'][\$1]",
3404                                                             $canonized_headline );
3405
3406                         # strip out HTML
3407                         $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
3408                         $tocline = trim( $canonized_headline );
3409                         # Save headline for section edit hint before it's escaped
3410                         $headline_hint = trim( $canonized_headline );
3411                         $canonized_headline = Sanitizer::escapeId( $tocline );
3412                         $refers[$headlineCount] = $canonized_headline;
3413
3414                         # count how many in assoc. array so we can track dupes in anchors
3415                         @$refers[$canonized_headline]++;
3416                         $refcount[$headlineCount]=$refers[$canonized_headline];
3417
3418                         # Don't number the heading if it is the only one (looks silly)
3419                         if( $doNumberHeadings && count( $matches[3] ) > 1) {
3420                                 # the two are different if the line contains a link
3421                                 $headline=$numbering . ' ' . $headline;
3422                         }
3423
3424                         # Create the anchor for linking from the TOC to the section
3425                         $anchor = $canonized_headline;
3426                         if($refcount[$headlineCount] > 1 ) {
3427                                 $anchor .= '_' . $refcount[$headlineCount];
3428                         }
3429                         if( $enoughToc && ( !isset($wgMaxTocLevel) || $toclevel<$wgMaxTocLevel ) ) {
3430                                 $toc .= $sk->tocLine($anchor, $tocline, $numbering, $toclevel);
3431                         }
3432                         if( $showEditLink && ( !$istemplate || $templatetitle !== "" ) ) {
3433                                 if ( empty( $head[$headlineCount] ) ) {
3434                                         $head[$headlineCount] = '';
3435                                 }
3436                                 if( $istemplate )
3437                                         $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
3438                                 else
3439                                         $head[$headlineCount] .= $sk->editSectionLink($this->mTitle, $sectionCount+1, $headline_hint);
3440                         }
3441
3442                         # give headline the correct <h#> tag
3443                         @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
3444
3445                         $headlineCount++;
3446                         if( !$istemplate )
3447                                 $sectionCount++;
3448                 }
3449
3450                 if( $enoughToc ) {
3451                         if( $toclevel<$wgMaxTocLevel ) {
3452                                 $toc .= $sk->tocUnindent( $toclevel - 1 );
3453                         }
3454                         $toc = $sk->tocList( $toc );
3455                 }
3456
3457                 # split up and insert constructed headlines
3458
3459                 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
3460                 $i = 0;
3461
3462                 foreach( $blocks as $block ) {
3463                         if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
3464                                 # This is the [edit] link that appears for the top block of text when
3465                                 # section editing is enabled
3466
3467                                 # Disabled because it broke block formatting
3468                                 # For example, a bullet point in the top line
3469                                 # $full .= $sk->editSectionLink(0);
3470                         }
3471                         $full .= $block;
3472                         if( $enoughToc && !$i && $isMain && !$this->mForceTocPosition ) {
3473                                 # Top anchor now in skin
3474                                 $full = $full.$toc;
3475                         }
3476
3477                         if( !empty( $head[$i] ) ) {
3478                                 $full .= $head[$i];
3479                         }
3480                         $i++;
3481                 }
3482                 if( $this->mForceTocPosition ) {
3483                         return str_replace( '<!--MWTOC-->', $toc, $full );
3484                 } else {
3485                         return $full;
3486                 }
3487         }
3488
3489         /**
3490          * Return an HTML link for the "ISBN 123456" text
3491          * @private
3492          */
3493         function magicISBN( $text ) {
3494                 $fname = 'Parser::magicISBN';
3495                 wfProfileIn( $fname );
3496
3497                 $a = split( 'ISBN ', ' '.$text );
3498                 if ( count ( $a ) < 2 ) {
3499                         wfProfileOut( $fname );
3500                         return $text;
3501                 }
3502                 $text = substr( array_shift( $a ), 1);
3503                 $valid = '0123456789-Xx';
3504
3505                 foreach ( $a as $x ) {
3506                         # hack: don't replace inside thumbnail title/alt
3507                         # attributes
3508                         if(preg_match('/<[^>]+(alt|title)="[^">]*$/', $text)) {
3509                                 $text .= "ISBN $x";
3510                                 continue;
3511                         }
3512
3513                         $isbn = $blank = '' ;
3514                         while ( ' ' == $x{0} ) {
3515                                 $blank .= ' ';
3516                                 $x = substr( $x, 1 );
3517                         }
3518                         if ( $x == '' ) { # blank isbn
3519                                 $text .= "ISBN $blank";
3520                                 continue;
3521                         }
3522                         while ( strstr( $valid, $x{0} ) != false ) {
3523                                 $isbn .= $x{0};
3524                                 $x = substr( $x, 1 );
3525                         }
3526                         $num = str_replace( '-', '', $isbn );
3527                         $num = str_replace( ' ', '', $num );
3528                         $num = str_replace( 'x', 'X', $num );
3529
3530                         if ( '' == $num ) {
3531                                 $text .= "ISBN $blank$x";
3532                         } else {
3533                                 $titleObj = Title::makeTitle( NS_SPECIAL, 'Booksources' );
3534                                 $text .= '<a href="' .
3535                                         $titleObj->escapeLocalUrl( 'isbn='.$num ) .
3536                                         "\" class=\"internal\">ISBN $isbn</a>";
3537                                 $text .= $x;
3538                         }
3539                 }
3540                 wfProfileOut( $fname );
3541                 return $text;
3542         }
3543
3544         /**
3545          * Return an HTML link for the "RFC 1234" text
3546          *
3547          * @private
3548          * @param string $text     Text to be processed
3549          * @param string $keyword  Magic keyword to use (default RFC)
3550          * @param string $urlmsg   Interface message to use (default rfcurl)
3551          * @return string
3552          */
3553         function magicRFC( $text, $keyword='RFC ', $urlmsg='rfcurl'  ) {
3554
3555                 $valid = '0123456789';
3556                 $internal = false;
3557
3558                 $a = split( $keyword, ' '.$text );
3559                 if ( count ( $a ) < 2 ) {
3560                         return $text;
3561                 }
3562                 $text = substr( array_shift( $a ), 1);
3563
3564                 /* Check if keyword is preceed by [[.
3565                  * This test is made here cause of the array_shift above
3566                  * that prevent the test to be done in the foreach.
3567                  */
3568                 if ( substr( $text, -2 ) == '[[' ) {
3569                         $internal = true;
3570                 }
3571
3572                 foreach ( $a as $x ) {
3573                         /* token might be empty if we have RFC RFC 1234 */
3574                         if ( $x=='' ) {
3575                                 $text.=$keyword;
3576                                 continue;
3577                                 }
3578
3579                         # hack: don't replace inside thumbnail title/alt
3580                         # attributes
3581                         if(preg_match('/<[^>]+(alt|title)="[^">]*$/', $text)) {
3582                                 $text .= $keyword . $x;
3583                                 continue;
3584                         }
3585
3586                         $id = $blank = '' ;
3587
3588                         /** remove and save whitespaces in $blank */
3589                         while ( $x{0} == ' ' ) {
3590                                 $blank .= ' ';
3591                                 $x = substr( $x, 1 );
3592                         }
3593
3594                         /** remove and save the rfc number in $id */
3595                         while ( strstr( $valid, $x{0} ) != false ) {
3596                                 $id .= $x{0};
3597                                 $x = substr( $x, 1 );
3598                         }
3599
3600                         if ( $id == '' ) {
3601                                 /* call back stripped spaces*/
3602                                 $text .= $keyword.$blank.$x;
3603                         } elseif( $internal ) {
3604                                 /* normal link */
3605                                 $text .= $keyword.$id.$x;
3606                         } else {
3607                                 /* build the external link*/
3608                                 $url = wfMsg( $urlmsg, $id);
3609                                 $sk =& $this->mOptions->getSkin();
3610                                 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
3611                                 $text .= "<a href=\"{$url}\"{$la}>{$keyword}{$id}</a>{$x}";
3612                         }
3613
3614                         /* Check if the next RFC keyword is preceed by [[ */
3615                         $internal = ( substr($x,-2) == '[[' );
3616                 }
3617                 return $text;
3618         }
3619
3620         /**
3621          * Transform wiki markup when saving a page by doing \r\n -> \n
3622          * conversion, substitting signatures, {{subst:}} templates, etc.
3623          *
3624          * @param string $text the text to transform
3625          * @param Title &$title the Title object for the current article
3626          * @param User &$user the User object describing the current user
3627          * @param ParserOptions $options parsing options
3628          * @param bool $clearState whether to clear the parser state first
3629          * @return string the altered wiki markup
3630          * @public
3631          */
3632         function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
3633                 $this->mOptions = $options;
3634                 $this->mTitle =& $title;
3635                 $this->mOutputType = OT_WIKI;
3636
3637                 if ( $clearState ) {
3638                         $this->clearState();
3639                 }
3640
3641                 $stripState = false;
3642                 $pairs = array(
3643                         "\r\n" => "\n",
3644                 );
3645                 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
3646                 $text = $this->strip( $text, $stripState, true, array( 'gallery' ) );
3647                 $text = $this->pstPass2( $text, $stripState, $user );
3648                 $text = $this->unstrip( $text, $stripState );
3649                 $text = $this->unstripNoWiki( $text, $stripState );
3650                 return $text;
3651         }
3652
3653         /**
3654          * Pre-save transform helper function
3655          * @private
3656          */
3657         function pstPass2( $text, &$stripState, &$user ) {
3658                 global $wgContLang, $wgLocaltimezone;
3659
3660                 /* Note: This is the timestamp saved as hardcoded wikitext to
3661                  * the database, we use $wgContLang here in order to give
3662                  * everyone the same signature and use the default one rather
3663                  * than the one selected in each user's preferences.
3664                  */
3665                 if ( isset( $wgLocaltimezone ) ) {
3666                         $oldtz = getenv( 'TZ' );
3667                         putenv( 'TZ='.$wgLocaltimezone );
3668                 }
3669                 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false, false) .
3670                   ' (' . date( 'T' ) . ')';
3671                 if ( isset( $wgLocaltimezone ) ) {
3672                         putenv( 'TZ='.$oldtz );
3673                 }
3674
3675                 # Variable replacement
3676                 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
3677                 $text = $this->replaceVariables( $text );
3678
3679                 # Strip out <nowiki> etc. added via replaceVariables
3680                 $text = $this->strip( $text, $stripState, false, array( 'gallery' ) );
3681
3682                 # Signatures
3683                 $sigText = $this->getUserSig( $user );
3684                 $text = strtr( $text, array(
3685                         '~~~~~' => $d,
3686                         '~~~~' => "$sigText $d",
3687                         '~~~' => $sigText
3688                 ) );
3689
3690                 # Context links: [[|name]] and [[name (context)|]]
3691                 #
3692                 global $wgLegalTitleChars;
3693                 $tc = "[$wgLegalTitleChars]";
3694                 $np = str_replace( array( '(', ')' ), array( '', '' ), $tc ); # No parens
3695
3696                 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
3697                 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
3698
3699                 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/";             # [[page (context)|]]
3700                 $p2 = "/\[\[\\|({$tc}+)]]/";                                    # [[|page]]
3701                 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/";                # [[namespace:page|]] and [[:namespace:page|]]
3702                 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
3703                 $context = '';
3704                 $t = $this->mTitle->getText();
3705                 if ( preg_match( $conpat, $t, $m ) ) {
3706                         $context = $m[2];
3707                 }
3708                 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
3709                 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
3710                 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
3711
3712                 if ( '' == $context ) {
3713                         $text = preg_replace( $p2, '[[\\1]]', $text );
3714                 } else {
3715                         $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
3716                 }
3717
3718                 # Trim trailing whitespace
3719                 # MAG_END (__END__) tag allows for trailing
3720                 # whitespace to be deliberately included
3721                 $text = rtrim( $text );
3722                 $mw =& MagicWord::get( MAG_END );
3723                 $mw->matchAndRemove( $text );
3724
3725                 return $text;
3726         }
3727
3728         /**
3729          * Fetch the user's signature text, if any, and normalize to
3730          * validated, ready-to-insert wikitext.
3731          *
3732          * @param User $user
3733          * @return string
3734          * @private
3735          */
3736         function getUserSig( &$user ) {
3737                 $username = $user->getName();
3738                 $nickname = $user->getOption( 'nickname' );
3739                 $nickname = $nickname === '' ? $username : $nickname;
3740
3741                 if( $user->getBoolOption( 'fancysig' ) !== false ) {
3742                         # Sig. might contain markup; validate this
3743                         if( $this->validateSig( $nickname ) !== false ) {
3744                                 # Validated; clean up (if needed) and return it
3745                                 return $this->cleanSig( $nickname, true );
3746                         } else {
3747                                 # Failed to validate; fall back to the default
3748                                 $nickname = $username;
3749                                 wfDebug( "Parser::getUserSig: $username has bad XML tags in signature.\n" );
3750                         }
3751                 }
3752
3753                 // Make sure nickname doesnt get a sig in a sig
3754                 $nickname = $this->cleanSigInSig( $nickname );
3755
3756                 # If we're still here, make it a link to the user page
3757                 $userpage = $user->getUserPage();
3758                 return( '[[' . $userpage->getPrefixedText() . '|' . wfEscapeWikiText( $nickname ) . ']]' );
3759         }
3760
3761         /**
3762          * Check that the user's signature contains no bad XML
3763          *
3764          * @param string $text
3765          * @return mixed An expanded string, or false if invalid.
3766          */
3767         function validateSig( $text ) {
3768                 return( wfIsWellFormedXmlFragment( $text ) ? $text : false );
3769         }
3770
3771         /**
3772          * Clean up signature text
3773          *
3774          * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
3775          * 2) Substitute all transclusions
3776          *
3777          * @param string $text
3778          * @param $parsing Whether we're cleaning (preferences save) or parsing
3779          * @return string Signature text
3780          */
3781         function cleanSig( $text, $parsing = false ) {
3782                 global $wgTitle;
3783                 $this->startExternalParse( $wgTitle, new ParserOptions(), $parsing ? OT_WIKI : OT_MSG );
3784
3785                 $substWord = MagicWord::get( MAG_SUBST );
3786                 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
3787                 $substText = '{{' . $substWord->getSynonym( 0 );
3788
3789                 $text = preg_replace( $substRegex, $substText, $text );
3790                 $text = $this->cleanSigInSig( $text );
3791                 $text = $this->replaceVariables( $text );
3792
3793                 $this->clearState();
3794                 return $text;
3795         }
3796
3797         /**
3798          * Strip ~~~, ~~~~ and ~~~~~ out of signatures
3799          * @param string $text
3800          * @return string Signature text with /~{3,5}/ removed
3801          */
3802         function cleanSigInSig( $text ) {
3803                 $text = preg_replace( '/~{3,5}/', '', $text );
3804                 return $text;
3805         }
3806
3807         /**
3808          * Set up some variables which are usually set up in parse()
3809          * so that an external function can call some class members with confidence
3810          * @public
3811          */
3812         function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
3813                 $this->mTitle =& $title;
3814                 $this->mOptions = $options;
3815                 $this->mOutputType = $outputType;
3816                 if ( $clearState ) {
3817                         $this->clearState();
3818                 }
3819         }
3820
3821         /**
3822          * Transform a MediaWiki message by replacing magic variables.
3823          *
3824          * @param string $text the text to transform
3825          * @param ParserOptions $options  options
3826          * @return string the text with variables substituted
3827          * @public
3828          */
3829         function transformMsg( $text, $options ) {
3830                 global $wgTitle;
3831                 static $executing = false;
3832
3833                 $fname = "Parser::transformMsg";
3834
3835                 # Guard against infinite recursion
3836                 if ( $executing ) {
3837                         return $text;
3838                 }
3839                 $executing = true;
3840
3841                 wfProfileIn($fname);
3842
3843                 $this->mTitle = $wgTitle;
3844                 $this->mOptions = $options;
3845                 $this->mOutputType = OT_MSG;
3846                 $this->clearState();
3847                 $text = $this->replaceVariables( $text );
3848
3849                 $executing = false;
3850                 wfProfileOut($fname);
3851                 return $text;
3852         }
3853
3854         /**
3855          * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
3856          * The callback should have the following form:
3857          *    function myParserHook( $text, $params, &$parser ) { ... }
3858          *
3859          * Transform and return $text. Use $parser for any required context, e.g. use
3860          * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
3861          *
3862          * @public
3863          *
3864          * @param mixed $tag The tag to use, e.g. 'hook' for <hook>
3865          * @param mixed $callback The callback function (and object) to use for the tag
3866          *
3867          * @return The old value of the mTagHooks array associated with the hook
3868          */
3869         function setHook( $tag, $callback ) {
3870                 $tag = strtolower( $tag );
3871                 $oldVal = @$this->mTagHooks[$tag];
3872                 $this->mTagHooks[$tag] = $callback;
3873
3874                 return $oldVal;
3875         }
3876
3877         /**
3878          * Create a function, e.g. {{sum:1|2|3}}
3879          * The callback function should have the form:
3880          *    function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
3881          *
3882          * The callback may either return the text result of the function, or an array with the text
3883          * in element 0, and a number of flags in the other elements. The names of the flags are
3884          * specified in the keys. Valid flags are:
3885          *   found                     The text returned is valid, stop processing the template. This
3886          *                             is on by default.
3887          *   nowiki                    Wiki markup in the return value should be escaped
3888          *   noparse                   Unsafe HTML tags should not be stripped, etc.
3889          *   noargs                    Don't replace triple-brace arguments in the return value
3890          *   isHTML                    The returned text is HTML, armour it against wikitext transformation
3891          *
3892          * @public
3893          *
3894          * @param mixed $id The magic word ID, or (deprecated) the function name. Function names are case-insensitive.
3895          * @param mixed $callback The callback function (and object) to use
3896          *
3897          * @return The old callback function for this name, if any
3898          */
3899         function setFunctionHook( $id, $callback ) {
3900                 if( is_string( $id ) ) {
3901                         $id = strtolower( $id );
3902                 }
3903                 $oldVal = @$this->mFunctionHooks[$id];
3904                 $this->mFunctionHooks[$id] = $callback;
3905                 # Add to function cache
3906                 if ( is_int( $id ) ) {
3907                         $mw = MagicWord::get( $id );
3908                         $mw->addToArray( $this->mFunctionSynonyms, $id );
3909                 } else {
3910                         $this->mFunctionSynonyms[$id] = $id;
3911                 }
3912                 return $oldVal;
3913         }
3914
3915         /**
3916          * Replace <!--LINK--> link placeholders with actual links, in the buffer
3917          * Placeholders created in Skin::makeLinkObj()
3918          * Returns an array of links found, indexed by PDBK:
3919          *  0 - broken
3920          *  1 - normal link
3921          *  2 - stub
3922          * $options is a bit field, RLH_FOR_UPDATE to select for update
3923          */
3924         function replaceLinkHolders( &$text, $options = 0 ) {
3925                 global $wgUser;
3926                 global $wgOutputReplace;
3927
3928                 $fname = 'Parser::replaceLinkHolders';
3929                 wfProfileIn( $fname );
3930
3931                 $pdbks = array();
3932                 $colours = array();
3933                 $sk =& $this->mOptions->getSkin();
3934                 $linkCache =& LinkCache::singleton();
3935
3936                 if ( !empty( $this->mLinkHolders['namespaces'] ) ) {
3937                         wfProfileIn( $fname.'-check' );
3938                         $dbr =& wfGetDB( DB_SLAVE );
3939                         $page = $dbr->tableName( 'page' );
3940                         $threshold = $wgUser->getOption('stubthreshold');
3941
3942                         # Sort by namespace
3943                         asort( $this->mLinkHolders['namespaces'] );
3944
3945                         # Generate query
3946                         $query = false;
3947                         foreach ( $this->mLinkHolders['namespaces'] as $key => $ns ) {
3948                                 # Make title object
3949                                 $title = $this->mLinkHolders['titles'][$key];
3950
3951                                 # Skip invalid entries.
3952                                 # Result will be ugly, but prevents crash.
3953                                 if ( is_null( $title ) ) {
3954                                         continue;
3955                                 }
3956                                 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
3957
3958                                 # Check if it's a static known link, e.g. interwiki
3959                                 if ( $title->isAlwaysKnown() ) {
3960                                         $colours[$pdbk] = 1;
3961                                 } elseif ( ( $id = $linkCache->getGoodLinkID( $pdbk ) ) != 0 ) {
3962                                         $colours[$pdbk] = 1;
3963                                         $this->mOutput->addLink( $title, $id );
3964                                 } elseif ( $linkCache->isBadLink( $pdbk ) ) {
3965                                         $colours[$pdbk] = 0;
3966                                 } else {
3967                                         # Not in the link cache, add it to the query
3968                                         if ( !isset( $current ) ) {
3969                                                 $current = $ns;
3970                                                 $query =  "SELECT page_id, page_namespace, page_title";
3971                                                 if ( $threshold > 0 ) {
3972                                                         $query .= ', page_len, page_is_redirect';
3973                                                 }
3974                                                 $query .= " FROM $page WHERE (page_namespace=$ns AND page_title IN(";
3975                                         } elseif ( $current != $ns ) {
3976                                                 $current = $ns;
3977                                                 $query .= ")) OR (page_namespace=$ns AND page_title IN(";
3978                                         } else {
3979                                                 $query .= ', ';
3980                                         }
3981
3982                                         $query .= $dbr->addQuotes( $this->mLinkHolders['dbkeys'][$key] );
3983                                 }
3984                         }
3985                         if ( $query ) {
3986                                 $query .= '))';
3987                                 if ( $options & RLH_FOR_UPDATE ) {
3988                                         $query .= ' FOR UPDATE';
3989                                 }
3990
3991                                 $res = $dbr->query( $query, $fname );
3992
3993                                 # Fetch data and form into an associative array
3994                                 # non-existent = broken
3995                                 # 1 = known
3996                                 # 2 = stub
3997                                 while ( $s = $dbr->fetchObject($res) ) {
3998                                         $title = Title::makeTitle( $s->page_namespace, $s->page_title );
3999                                         $pdbk = $title->getPrefixedDBkey();
4000                                         $linkCache->addGoodLinkObj( $s->page_id, $title );
4001                                         $this->mOutput->addLink( $title, $s->page_id );
4002
4003                                         if ( $threshold >  0 ) {
4004                                                 $size = $s->page_len;
4005                                                 if ( $s->page_is_redirect || $s->page_namespace != 0 || $size >= $threshold ) {
4006                                                         $colours[$pdbk] = 1;
4007                                                 } else {
4008                                                         $colours[$pdbk] = 2;
4009                                                 }
4010                                         } else {
4011                                                 $colours[$pdbk] = 1;
4012                                         }
4013                                 }
4014                         }
4015                         wfProfileOut( $fname.'-check' );
4016
4017                         # Construct search and replace arrays
4018                         wfProfileIn( $fname.'-construct' );
4019                         $wgOutputReplace = array();
4020                         foreach ( $this->mLinkHolders['namespaces'] as $key => $ns ) {
4021                                 $pdbk = $pdbks[$key];
4022                                 $searchkey = "<!--LINK $key-->";
4023                                 $title = $this->mLinkHolders['titles'][$key];
4024                                 if ( empty( $colours[$pdbk] ) ) {
4025                                         $linkCache->addBadLinkObj( $title );
4026                                         $colours[$pdbk] = 0;
4027                                         $this->mOutput->addLink( $title, 0 );
4028                                         $wgOutputReplace[$searchkey] = $sk->makeBrokenLinkObj( $title,
4029                                                                         $this->mLinkHolders['texts'][$key],
4030                                                                         $this->mLinkHolders['queries'][$key] );
4031                                 } elseif ( $colours[$pdbk] == 1 ) {
4032                                         $wgOutputReplace[$searchkey] = $sk->makeKnownLinkObj( $title,
4033                                                                         $this->mLinkHolders['texts'][$key],
4034                                                                         $this->mLinkHolders['queries'][$key] );
4035                                 } elseif ( $colours[$pdbk] == 2 ) {
4036                                         $wgOutputReplace[$searchkey] = $sk->makeStubLinkObj( $title,
4037                                                                         $this->mLinkHolders['texts'][$key],
4038                                                                         $this->mLinkHolders['queries'][$key] );
4039                                 }
4040                         }
4041                         wfProfileOut( $fname.'-construct' );
4042
4043                         # Do the thing
4044                         wfProfileIn( $fname.'-replace' );
4045
4046                         $text = preg_replace_callback(
4047                                 '/(<!--LINK .*?-->)/',
4048                                 "wfOutputReplaceMatches",
4049                                 $text);
4050
4051                         wfProfileOut( $fname.'-replace' );
4052                 }
4053
4054                 # Now process interwiki link holders
4055                 # This is quite a bit simpler than internal links
4056                 if ( !empty( $this->mInterwikiLinkHolders['texts'] ) ) {
4057                         wfProfileIn( $fname.'-interwiki' );
4058                         # Make interwiki link HTML
4059                         $wgOutputReplace = array();
4060                         foreach( $this->mInterwikiLinkHolders['texts'] as $key => $link ) {
4061                                 $title = $this->mInterwikiLinkHolders['titles'][$key];
4062                                 $wgOutputReplace[$key] = $sk->makeLinkObj( $title, $link );
4063                         }
4064
4065                         $text = preg_replace_callback(
4066                                 '/<!--IWLINK (.*?)-->/',
4067                                 "wfOutputReplaceMatches",
4068                                 $text );
4069                         wfProfileOut( $fname.'-interwiki' );
4070                 }
4071
4072                 wfProfileOut( $fname );
4073                 return $colours;
4074         }
4075
4076         /**
4077          * Replace <!--LINK--> link placeholders with plain text of links
4078          * (not HTML-formatted).
4079          * @param string $text
4080          * @return string
4081          */
4082         function replaceLinkHoldersText( $text ) {
4083                 $fname = 'Parser::replaceLinkHoldersText';
4084                 wfProfileIn( $fname );
4085
4086                 $text = preg_replace_callback(
4087                         '/<!--(LINK|IWLINK) (.*?)-->/',
4088                         array( &$this, 'replaceLinkHoldersTextCallback' ),
4089                         $text );
4090
4091                 wfProfileOut( $fname );
4092                 return $text;
4093         }
4094
4095         /**
4096          * @param array $matches
4097          * @return string
4098          * @private
4099          */
4100         function replaceLinkHoldersTextCallback( $matches ) {
4101                 $type = $matches[1];
4102                 $key  = $matches[2];
4103                 if( $type == 'LINK' ) {
4104                         if( isset( $this->mLinkHolders['texts'][$key] ) ) {
4105                                 return $this->mLinkHolders['texts'][$key];
4106                         }
4107                 } elseif( $type == 'IWLINK' ) {
4108                         if( isset( $this->mInterwikiLinkHolders['texts'][$key] ) ) {
4109                                 return $this->mInterwikiLinkHolders['texts'][$key];
4110                         }
4111                 }
4112                 return $matches[0];
4113         }
4114
4115         /**
4116          * Tag hook handler for 'pre'.
4117          */
4118         function renderPreTag( $text, $attribs, $parser ) {
4119                 // Backwards-compatibility hack
4120                 $content = preg_replace( '!<nowiki>(.*?)</nowiki>!is', '\\1', $text );
4121
4122                 $attribs = Sanitizer::validateTagAttributes( $attribs, 'pre' );
4123                 return wfOpenElement( 'pre', $attribs ) .
4124                         wfEscapeHTMLTagsOnly( $content ) .
4125                         '</pre>';
4126         }
4127
4128         /**
4129          * Renders an image gallery from a text with one line per image.
4130          * text labels may be given by using |-style alternative text. E.g.
4131          *   Image:one.jpg|The number "1"
4132          *   Image:tree.jpg|A tree
4133          * given as text will return the HTML of a gallery with two images,
4134          * labeled 'The number "1"' and
4135          * 'A tree'.
4136          */
4137         function renderImageGallery( $text, $params ) {
4138                 $ig = new ImageGallery();
4139                 $ig->setShowBytes( false );
4140                 $ig->setShowFilename( false );
4141                 $ig->setParsing();
4142                 $ig->useSkin( $this->mOptions->getSkin() );
4143
4144                 if( isset( $params['caption'] ) )
4145                         $ig->setCaption( $params['caption'] );
4146
4147                 $lines = explode( "\n", $text );
4148                 foreach ( $lines as $line ) {
4149                         # match lines like these:
4150                         # Image:someimage.jpg|This is some image
4151                         preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
4152                         # Skip empty lines
4153                         if ( count( $matches ) == 0 ) {
4154                                 continue;
4155                         }
4156                         $nt =& Title::newFromText( $matches[1] );
4157                         if( is_null( $nt ) ) {
4158                                 # Bogus title. Ignore these so we don't bomb out later.
4159                                 continue;
4160                         }
4161                         if ( isset( $matches[3] ) ) {
4162                                 $label = $matches[3];
4163                         } else {
4164                                 $label = '';
4165                         }
4166
4167                         $pout = $this->parse( $label,
4168                                 $this->mTitle,
4169                                 $this->mOptions,
4170                                 false, // Strip whitespace...?
4171                                 false  // Don't clear state!
4172                         );
4173                         $html = $pout->getText();
4174
4175                         $ig->add( new Image( $nt ), $html );
4176
4177                         # Only add real images (bug #5586)
4178                         if ( $nt->getNamespace() == NS_IMAGE ) {
4179                                 $this->mOutput->addImage( $nt->getDBkey() );
4180                         }
4181                 }
4182                 return $ig->toHTML();
4183         }
4184
4185         /**
4186          * Parse image options text and use it to make an image
4187          */
4188         function makeImage( &$nt, $options ) {
4189                 global $wgUseImageResize;
4190
4191                 $align = '';
4192
4193                 # Check if the options text is of the form "options|alt text"
4194                 # Options are:
4195                 #  * thumbnail          make a thumbnail with enlarge-icon and caption, alignment depends on lang
4196                 #  * left               no resizing, just left align. label is used for alt= only
4197                 #  * right              same, but right aligned
4198                 #  * none               same, but not aligned
4199                 #  * ___px              scale to ___ pixels width, no aligning. e.g. use in taxobox
4200                 #  * center             center the image
4201                 #  * framed             Keep original image size, no magnify-button.
4202
4203                 $part = explode( '|', $options);
4204
4205                 $mwThumb  =& MagicWord::get( MAG_IMG_THUMBNAIL );
4206                 $mwManualThumb =& MagicWord::get( MAG_IMG_MANUALTHUMB );
4207                 $mwLeft   =& MagicWord::get( MAG_IMG_LEFT );
4208                 $mwRight  =& MagicWord::get( MAG_IMG_RIGHT );
4209                 $mwNone   =& MagicWord::get( MAG_IMG_NONE );
4210                 $mwWidth  =& MagicWord::get( MAG_IMG_WIDTH );
4211                 $mwCenter =& MagicWord::get( MAG_IMG_CENTER );
4212                 $mwFramed =& MagicWord::get( MAG_IMG_FRAMED );
4213                 $caption = '';
4214
4215                 $width = $height = $framed = $thumb = false;
4216                 $manual_thumb = '' ;
4217
4218                 foreach( $part as $key => $val ) {
4219                         if ( $wgUseImageResize && ! is_null( $mwThumb->matchVariableStartToEnd($val) ) ) {
4220                                 $thumb=true;
4221                         } elseif ( ! is_null( $match = $mwManualThumb->matchVariableStartToEnd($val) ) ) {
4222                                 # use manually specified thumbnail
4223                                 $thumb=true;
4224                                 $manual_thumb = $match;
4225                         } elseif ( ! is_null( $mwRight->matchVariableStartToEnd($val) ) ) {
4226                                 # remember to set an alignment, don't render immediately
4227                                 $align = 'right';
4228                         } elseif ( ! is_null( $mwLeft->matchVariableStartToEnd($val) ) ) {
4229                                 # remember to set an alignment, don't render immediately
4230                                 $align = 'left';
4231                         } elseif ( ! is_null( $mwCenter->matchVariableStartToEnd($val) ) ) {
4232                                 # remember to set an alignment, don't render immediately
4233                                 $align = 'center';
4234                         } elseif ( ! is_null( $mwNone->matchVariableStartToEnd($val) ) ) {
4235                                 # remember to set an alignment, don't render immediately
4236                                 $align = 'none';
4237                         } elseif ( $wgUseImageResize && ! is_null( $match = $mwWidth->matchVariableStartToEnd($val) ) ) {
4238                                 wfDebug( "MAG_IMG_WIDTH match: $match\n" );
4239                                 # $match is the image width in pixels
4240                                 if ( preg_match( '/^([0-9]*)x([0-9]*)$/', $match, $m ) ) {
4241                                         $width = intval( $m[1] );
4242                                         $height = intval( $m[2] );
4243                                 } else {
4244                                         $width = intval($match);
4245                                 }
4246                         } elseif ( ! is_null( $mwFramed->matchVariableStartToEnd($val) ) ) {
4247                                 $framed=true;
4248                         } else {
4249                                 $caption = $val;
4250                         }
4251                 }
4252                 # Strip bad stuff out of the alt text
4253                 $alt = $this->replaceLinkHoldersText( $caption );
4254
4255                 # make sure there are no placeholders in thumbnail attributes
4256                 # that are later expanded to html- so expand them now and
4257                 # remove the tags
4258                 $alt = $this->unstrip($alt, $this->mStripState);
4259                 $alt = Sanitizer::stripAllTags( $alt );
4260
4261                 # Linker does the rest
4262                 $sk =& $this->mOptions->getSkin();
4263                 return $sk->makeImageLinkObj( $nt, $caption, $alt, $align, $width, $height, $framed, $thumb, $manual_thumb );
4264         }
4265
4266         /**
4267          * Set a flag in the output object indicating that the content is dynamic and
4268          * shouldn't be cached.
4269          */
4270         function disableCache() {
4271                 wfDebug( "Parser output marked as uncacheable.\n" );
4272                 $this->mOutput->mCacheTime = -1;
4273         }
4274
4275         /**#@+
4276          * Callback from the Sanitizer for expanding items found in HTML attribute
4277          * values, so they can be safely tested and escaped.
4278          * @param string $text
4279          * @param array $args
4280          * @return string
4281          * @private
4282          */
4283         function attributeStripCallback( &$text, $args ) {
4284                 $text = $this->replaceVariables( $text, $args );
4285                 $text = $this->unstripForHTML( $text );
4286                 return $text;
4287         }
4288
4289         function unstripForHTML( $text ) {
4290                 $text = $this->unstrip( $text, $this->mStripState );
4291                 $text = $this->unstripNoWiki( $text, $this->mStripState );
4292                 return $text;
4293         }
4294         /**#@-*/
4295
4296         /**#@+
4297          * Accessor/mutator
4298          */
4299         function Title( $x = NULL ) { return wfSetVar( $this->mTitle, $x ); }
4300         function Options( $x = NULL ) { return wfSetVar( $this->mOptions, $x ); }
4301         function OutputType( $x = NULL ) { return wfSetVar( $this->mOutputType, $x ); }
4302         /**#@-*/
4303
4304         /**#@+
4305          * Accessor
4306          */
4307         function getTags() { return array_keys( $this->mTagHooks ); }
4308         /**#@-*/
4309
4310
4311         /**
4312          * Break wikitext input into sections, and either pull or replace
4313          * some particular section's text.
4314          *
4315          * External callers should use the getSection and replaceSection methods.
4316          *
4317          * @param $text Page wikitext
4318          * @param $section Numbered section. 0 pulls the text before the first
4319          *                 heading; other numbers will pull the given section
4320          *                 along with its lower-level subsections.
4321          * @param $mode One of "get" or "replace"
4322          * @param $newtext Replacement text for section data.
4323          * @return string for "get", the extracted section text.
4324          *                for "replace", the whole page with the section replaced.
4325          */
4326         private function extractSections( $text, $section, $mode, $newtext='' ) {
4327                 # strip NOWIKI etc. to avoid confusion (true-parameter causes HTML
4328                 # comments to be stripped as well)
4329                 $striparray = array();
4330
4331                 $oldOutputType = $this->mOutputType;
4332                 $oldOptions = $this->mOptions;
4333                 $this->mOptions = new ParserOptions();
4334                 $this->mOutputType = OT_WIKI;
4335
4336                 $striptext = $this->strip( $text, $striparray, true );
4337
4338                 $this->mOutputType = $oldOutputType;
4339                 $this->mOptions = $oldOptions;
4340
4341                 # now that we can be sure that no pseudo-sections are in the source,
4342                 # split it up by section
4343                 $uniq = preg_quote( $this->uniqPrefix(), '/' );
4344                 $comment = "(?:$uniq-!--.*?QINU)";
4345                 $secs = preg_split(
4346                 /*
4347                         "/
4348                         ^(
4349                         (?:$comment|<\/?noinclude>)* # Initial comments will be stripped
4350                         (?:
4351                                 (=+) # Should this be limited to 6?
4352                                 .+?  # Section title...
4353                                 \\2  # Ending = count must match start
4354                         |
4355                                 ^
4356                                 <h([1-6])\b.*?>
4357                                 .*?
4358                                 <\/h\\3\s*>
4359                         )
4360                         (?:$comment|<\/?noinclude>|\s+)* # Trailing whitespace ok
4361                         )$
4362                         /mix",
4363                 */
4364                         "/
4365                         (
4366                                 ^
4367                                 (?:$comment|<\/?noinclude>)* # Initial comments will be stripped
4368                                 (=+) # Should this be limited to 6?
4369                                 .+?  # Section title...
4370                                 \\2  # Ending = count must match start
4371                                 (?:$comment|<\/?noinclude>|[ \\t]+)* # Trailing whitespace ok
4372                                 $
4373                         |
4374                                 <h([1-6])\b.*?>
4375                                 .*?
4376                                 <\/h\\3\s*>
4377                         )
4378                         /mix",
4379                         $striptext, -1,
4380                         PREG_SPLIT_DELIM_CAPTURE);
4381
4382                 if( $mode == "get" ) {
4383                         if( $section == 0 ) {
4384                                 // "Section 0" returns the content before any other section.
4385                                 $rv = $secs[0];
4386                         } else {
4387                                 $rv = "";
4388                         }
4389                 } elseif( $mode == "replace" ) {
4390                         if( $section == 0 ) {
4391                                 $rv = $newtext . "\n\n";
4392                                 $remainder = true;
4393                         } else {
4394                                 $rv = $secs[0];
4395                                 $remainder = false;
4396                         }
4397                 }
4398                 $count = 0;
4399                 $sectionLevel = 0;
4400                 for( $index = 1; $index < count( $secs ); ) {
4401                         $headerLine = $secs[$index++];
4402                         if( $secs[$index] ) {
4403                                 // A wiki header
4404                                 $headerLevel = strlen( $secs[$index++] );
4405                         } else {
4406                                 // An HTML header
4407                                 $index++;
4408                                 $headerLevel = intval( $secs[$index++] );
4409                         }
4410                         $content = $secs[$index++];
4411
4412                         $count++;
4413                         if( $mode == "get" ) {
4414                                 if( $count == $section ) {
4415                                         $rv = $headerLine . $content;
4416                                         $sectionLevel = $headerLevel;
4417                                 } elseif( $count > $section ) {
4418                                         if( $sectionLevel && $headerLevel > $sectionLevel ) {
4419                                                 $rv .= $headerLine . $content;
4420                                         } else {
4421                                                 // Broke out to a higher-level section
4422                                                 break;
4423                                         }
4424                                 }
4425                         } elseif( $mode == "replace" ) {
4426                                 if( $count < $section ) {
4427                                         $rv .= $headerLine . $content;
4428                                 } elseif( $count == $section ) {
4429                                         $rv .= $newtext . "\n\n";
4430                                         $sectionLevel = $headerLevel;
4431                                 } elseif( $count > $section ) {
4432                                         if( $headerLevel <= $sectionLevel ) {
4433                                                 // Passed the section's sub-parts.
4434                                                 $remainder = true;
4435                                         }
4436                                         if( $remainder ) {
4437                                                 $rv .= $headerLine . $content;
4438                                         }
4439                                 }
4440                         }
4441                 }
4442                 # reinsert stripped tags
4443                 $rv = $this->unstrip( $rv, $striparray );
4444                 $rv = $this->unstripNoWiki( $rv, $striparray );
4445                 $rv = trim( $rv );
4446                 return $rv;
4447         }
4448
4449         /**
4450          * This function returns the text of a section, specified by a number ($section).
4451          * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
4452          * the first section before any such heading (section 0).
4453          *
4454          * If a section contains subsections, these are also returned.
4455          *
4456          * @param $text String: text to look in
4457          * @param $section Integer: section number
4458          * @return string text of the requested section
4459          */
4460         function getSection( $text, $section ) {
4461                 return $this->extractSections( $text, $section, "get" );
4462         }
4463
4464         function replaceSection( $oldtext, $section, $text ) {
4465                 return $this->extractSections( $oldtext, $section, "replace", $text );
4466         }
4467
4468 }
4469
4470 /**
4471  * @todo document
4472  * @package MediaWiki
4473  */
4474 class ParserOutput
4475 {
4476         var $mText,             # The output text
4477                 $mLanguageLinks,    # List of the full text of language links, in the order they appear
4478                 $mCategories,       # Map of category names to sort keys
4479                 $mContainsOldMagic, # Boolean variable indicating if the input contained variables like {{CURRENTDAY}}
4480                 $mCacheTime,        # Time when this object was generated, or -1 for uncacheable. Used in ParserCache.
4481                 $mVersion,          # Compatibility check
4482                 $mTitleText,        # title text of the chosen language variant
4483                 $mLinks,            # 2-D map of NS/DBK to ID for the links in the document. ID=zero for broken.
4484                 $mTemplates,        # 2-D map of NS/DBK to ID for the template references. ID=zero for broken.
4485                 $mImages,           # DB keys of the images used, in the array key only
4486                 $mExternalLinks,    # External link URLs, in the key only
4487                 $mHTMLtitle,            # Display HTML title
4488                 $mSubtitle,                     # Additional subtitle
4489                 $mNewSection,           # Show a new section link?
4490                 $mNoGallery;            # No gallery on category page? (__NOGALLERY__)
4491
4492         function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
4493                 $containsOldMagic = false, $titletext = '' )
4494         {
4495                 $this->mText = $text;
4496                 $this->mLanguageLinks = $languageLinks;
4497                 $this->mCategories = $categoryLinks;
4498                 $this->mContainsOldMagic = $containsOldMagic;
4499                 $this->mCacheTime = '';
4500                 $this->mVersion = MW_PARSER_VERSION;
4501                 $this->mTitleText = $titletext;
4502                 $this->mLinks = array();
4503                 $this->mTemplates = array();
4504                 $this->mImages = array();
4505                 $this->mExternalLinks = array();
4506                 $this->mHTMLtitle = "" ;
4507                 $this->mSubtitle = "" ;
4508                 $this->mNewSection = false;
4509                 $this->mNoGallery = false;
4510         }
4511
4512         function getText()                   { return $this->mText; }
4513         function &getLanguageLinks()          { return $this->mLanguageLinks; }
4514         function getCategoryLinks()          { return array_keys( $this->mCategories ); }
4515         function &getCategories()            { return $this->mCategories; }
4516         function getCacheTime()              { return $this->mCacheTime; }
4517         function getTitleText()              { return $this->mTitleText; }
4518         function &getLinks()                 { return $this->mLinks; }
4519         function &getTemplates()             { return $this->mTemplates; }
4520         function &getImages()                { return $this->mImages; }
4521         function &getExternalLinks()         { return $this->mExternalLinks; }
4522         function getNoGallery()              { return $this->mNoGallery; }
4523
4524         function containsOldMagic()          { return $this->mContainsOldMagic; }
4525         function setText( $text )            { return wfSetVar( $this->mText, $text ); }
4526         function setLanguageLinks( $ll )     { return wfSetVar( $this->mLanguageLinks, $ll ); }
4527         function setCategoryLinks( $cl )     { return wfSetVar( $this->mCategories, $cl ); }
4528         function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic, $com ); }
4529         function setCacheTime( $t )          { return wfSetVar( $this->mCacheTime, $t ); }
4530         function setTitleText( $t )          { return wfSetVar ($this->mTitleText, $t); }
4531
4532         function addCategory( $c, $sort )    { $this->mCategories[$c] = $sort; }
4533         function addImage( $name )           { $this->mImages[$name] = 1; }
4534         function addLanguageLink( $t )       { $this->mLanguageLinks[] = $t; }
4535         function addExternalLink( $url )     { $this->mExternalLinks[$url] = 1; }
4536
4537         function setNewSection( $value ) {
4538                 $this->mNewSection = (bool)$value;
4539         }
4540         function getNewSection() {
4541                 return (bool)$this->mNewSection;
4542         }
4543
4544         function addLink( $title, $id ) {
4545                 $ns = $title->getNamespace();
4546                 $dbk = $title->getDBkey();
4547                 if ( !isset( $this->mLinks[$ns] ) ) {
4548                         $this->mLinks[$ns] = array();
4549                 }
4550                 $this->mLinks[$ns][$dbk] = $id;
4551         }
4552
4553         function addTemplate( $title, $id ) {
4554                 $ns = $title->getNamespace();
4555                 $dbk = $title->getDBkey();
4556                 if ( !isset( $this->mTemplates[$ns] ) ) {
4557                         $this->mTemplates[$ns] = array();
4558                 }
4559                 $this->mTemplates[$ns][$dbk] = $id;
4560         }
4561
4562         /**
4563          * Return true if this cached output object predates the global or
4564          * per-article cache invalidation timestamps, or if it comes from
4565          * an incompatible older version.
4566          *
4567          * @param string $touched the affected article's last touched timestamp
4568          * @return bool
4569          * @public
4570          */
4571         function expired( $touched ) {
4572                 global $wgCacheEpoch;
4573                 return $this->getCacheTime() == -1 || // parser says it's uncacheable
4574                        $this->getCacheTime() < $touched ||
4575                        $this->getCacheTime() <= $wgCacheEpoch ||
4576                        !isset( $this->mVersion ) ||
4577                        version_compare( $this->mVersion, MW_PARSER_VERSION, "lt" );
4578         }
4579 }
4580
4581 /**
4582  * Set options of the Parser
4583  * @todo document
4584  * @package MediaWiki
4585  */
4586 class ParserOptions
4587 {
4588         # All variables are private
4589         var $mUseTeX;                    # Use texvc to expand <math> tags
4590         var $mUseDynamicDates;           # Use DateFormatter to format dates
4591         var $mInterwikiMagic;            # Interlanguage links are removed and returned in an array
4592         var $mAllowExternalImages;       # Allow external images inline
4593         var $mAllowExternalImagesFrom;   # If not, any exception?
4594         var $mSkin;                      # Reference to the preferred skin
4595         var $mDateFormat;                # Date format index
4596         var $mEditSection;               # Create "edit section" links
4597         var $mNumberHeadings;            # Automatically number headings
4598         var $mAllowSpecialInclusion;     # Allow inclusion of special pages
4599         var $mTidy;                      # Ask for tidy cleanup
4600         var $mInterfaceMessage;          # Which lang to call for PLURAL and GRAMMAR
4601
4602         var $mUser;                      # Stored user object, just used to initialise the skin
4603
4604         function getUseTeX()                        { return $this->mUseTeX; }
4605         function getUseDynamicDates()               { return $this->mUseDynamicDates; }
4606         function getInterwikiMagic()                { return $this->mInterwikiMagic; }
4607         function getAllowExternalImages()           { return $this->mAllowExternalImages; }
4608         function getAllowExternalImagesFrom()       { return $this->mAllowExternalImagesFrom; }
4609         function getDateFormat()                    { return $this->mDateFormat; }
4610         function getEditSection()                   { return $this->mEditSection; }
4611         function getNumberHeadings()                { return $this->mNumberHeadings; }
4612         function getAllowSpecialInclusion()         { return $this->mAllowSpecialInclusion; }
4613         function getTidy()                          { return $this->mTidy; }
4614         function getInterfaceMessage()              { return $this->mInterfaceMessage; }
4615
4616         function &getSkin() {
4617                 if ( !isset( $this->mSkin ) ) {
4618                         $this->mSkin = $this->mUser->getSkin();
4619                 }
4620                 return $this->mSkin;
4621         }
4622
4623         function setUseTeX( $x )                    { return wfSetVar( $this->mUseTeX, $x ); }
4624         function setUseDynamicDates( $x )           { return wfSetVar( $this->mUseDynamicDates, $x ); }
4625         function setInterwikiMagic( $x )            { return wfSetVar( $this->mInterwikiMagic, $x ); }
4626         function setAllowExternalImages( $x )       { return wfSetVar( $this->mAllowExternalImages, $x ); }
4627         function setAllowExternalImagesFrom( $x )   { return wfSetVar( $this->mAllowExternalImagesFrom, $x ); }
4628         function setDateFormat( $x )                { return wfSetVar( $this->mDateFormat, $x ); }
4629         function setEditSection( $x )               { return wfSetVar( $this->mEditSection, $x ); }
4630         function setNumberHeadings( $x )            { return wfSetVar( $this->mNumberHeadings, $x ); }
4631         function setAllowSpecialInclusion( $x )     { return wfSetVar( $this->mAllowSpecialInclusion, $x ); }
4632         function setTidy( $x )                      { return wfSetVar( $this->mTidy, $x); }
4633         function setSkin( &$x ) { $this->mSkin =& $x; }
4634         function setInterfaceMessage( $x )          { return wfSetVar( $this->mInterfaceMessage, $x); }
4635
4636         function ParserOptions( $user = null ) {
4637                 $this->initialiseFromUser( $user );
4638         }
4639
4640         /**
4641          * Get parser options
4642          * @static
4643          */
4644         function newFromUser( &$user ) {
4645                 return new ParserOptions( $user );
4646         }
4647
4648         /** Get user options */
4649         function initialiseFromUser( &$userInput ) {
4650                 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
4651                 global $wgAllowExternalImagesFrom, $wgAllowSpecialInclusion;
4652                 $fname = 'ParserOptions::initialiseFromUser';
4653                 wfProfileIn( $fname );
4654                 if ( !$userInput ) {
4655                         global $wgUser;
4656                         if ( isset( $wgUser ) ) {
4657                                 $user = $wgUser;
4658                         } else {
4659                                 $user = new User;
4660                                 $user->setLoaded( true );
4661                         }
4662                 } else {
4663                         $user =& $userInput;
4664                 }
4665
4666                 $this->mUser = $user;
4667
4668                 $this->mUseTeX = $wgUseTeX;
4669                 $this->mUseDynamicDates = $wgUseDynamicDates;
4670                 $this->mInterwikiMagic = $wgInterwikiMagic;
4671                 $this->mAllowExternalImages = $wgAllowExternalImages;
4672                 $this->mAllowExternalImagesFrom = $wgAllowExternalImagesFrom;
4673                 $this->mSkin = null; # Deferred
4674                 $this->mDateFormat = $user->getOption( 'date' );
4675                 $this->mEditSection = true;
4676                 $this->mNumberHeadings = $user->getOption( 'numberheadings' );
4677                 $this->mAllowSpecialInclusion = $wgAllowSpecialInclusion;
4678                 $this->mTidy = false;
4679                 $this->mInterfaceMessage = false;
4680                 wfProfileOut( $fname );
4681         }
4682 }
4683
4684 /**
4685  * Callback function used by Parser::replaceLinkHolders()
4686  * to substitute link placeholders.
4687  */
4688 function &wfOutputReplaceMatches( $matches ) {
4689         global $wgOutputReplace;
4690         return $wgOutputReplace[$matches[1]];
4691 }
4692
4693 /**
4694  * Return the total number of articles
4695  */
4696 function wfNumberOfArticles() {
4697         global $wgNumberOfArticles;
4698
4699         wfLoadSiteStats();
4700         return $wgNumberOfArticles;
4701 }
4702
4703 /**
4704  * Return the number of files
4705  */
4706 function wfNumberOfFiles() {
4707         $fname = 'wfNumberOfFiles';
4708
4709         wfProfileIn( $fname );
4710         $dbr =& wfGetDB( DB_SLAVE );
4711         $numImages = $dbr->selectField('site_stats', 'ss_images', array(), $fname );
4712         wfProfileOut( $fname );
4713
4714         return $numImages;
4715 }
4716
4717 /**
4718  * Return the number of user accounts
4719  * @return integer
4720  */
4721 function wfNumberOfUsers() {
4722         wfProfileIn( 'wfNumberOfUsers' );
4723         $dbr =& wfGetDB( DB_SLAVE );
4724         $count = $dbr->selectField( 'site_stats', 'ss_users', array(), 'wfNumberOfUsers' );
4725         wfProfileOut( 'wfNumberOfUsers' );
4726         return (int)$count;
4727 }
4728
4729 /**
4730  * Return the total number of pages
4731  * @return integer
4732  */
4733 function wfNumberOfPages() {
4734         wfProfileIn( 'wfNumberOfPages' );
4735         $dbr =& wfGetDB( DB_SLAVE );
4736         $count = $dbr->selectField( 'site_stats', 'ss_total_pages', array(), 'wfNumberOfPages' );
4737         wfProfileOut( 'wfNumberOfPages' );
4738         return (int)$count;
4739 }
4740
4741 /**
4742  * Return the total number of admins
4743  *
4744  * @return integer
4745  */
4746 function wfNumberOfAdmins() {
4747         static $admins = -1;
4748         wfProfileIn( 'wfNumberOfAdmins' );
4749         if( $admins == -1 ) {
4750                 $dbr =& wfGetDB( DB_SLAVE );
4751                 $admins = $dbr->selectField( 'user_groups', 'COUNT(*)', array( 'ug_group' => 'sysop' ), 'wfNumberOfAdmins' );
4752         }
4753         wfProfileOut( 'wfNumberOfAdmins' );
4754         return (int)$admins;
4755 }
4756
4757 /**
4758  * Count the number of pages in a particular namespace
4759  *
4760  * @param $ns Namespace
4761  * @return integer
4762  */
4763 function wfPagesInNs( $ns ) {
4764         static $pageCount = array();
4765         wfProfileIn( 'wfPagesInNs' );
4766         if( !isset( $pageCount[$ns] ) ) {
4767                 $dbr =& wfGetDB( DB_SLAVE );
4768                 $pageCount[$ns] = $dbr->selectField( 'page', 'COUNT(*)', array( 'page_namespace' => $ns ), 'wfPagesInNs' );
4769         }
4770         wfProfileOut( 'wfPagesInNs' );
4771         return (int)$pageCount[$ns];
4772 }
4773
4774 /**
4775  * Get various statistics from the database
4776  * @private
4777  */
4778 function wfLoadSiteStats() {
4779         global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
4780         $fname = 'wfLoadSiteStats';
4781
4782         if ( -1 != $wgNumberOfArticles ) return;
4783         $dbr =& wfGetDB( DB_SLAVE );
4784         $s = $dbr->selectRow( 'site_stats',
4785                 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
4786                 array( 'ss_row_id' => 1 ), $fname
4787         );
4788
4789         if ( $s === false ) {
4790                 return;
4791         } else {
4792                 $wgTotalViews = $s->ss_total_views;
4793                 $wgTotalEdits = $s->ss_total_edits;
4794                 $wgNumberOfArticles = $s->ss_good_articles;
4795         }
4796 }
4797
4798 /**
4799  * Escape html tags
4800  * Basically replacing " > and < with HTML entities ( &quot;, &gt;, &lt;)
4801  *
4802  * @param $in String: text that might contain HTML tags.
4803  * @return string Escaped string
4804  */
4805 function wfEscapeHTMLTagsOnly( $in ) {
4806         return str_replace(
4807                 array( '"', '>', '<' ),
4808                 array( '&quot;', '&gt;', '&lt;' ),
4809                 $in );
4810 }
4811
4812 ?>