3 include_once('Tokenizer.php');
5 if( $GLOBALS['wgUseWikiHiero'] ){
6 include_once('wikihiero.php');
8 if( $GLOBALS['wgUseTimeline'] ){
9 include_once('extensions/timeline/Timeline.php');
14 # Processes wiki markup
16 # There are two main entry points into the Parser class: parse() and preSaveTransform().
17 # The parse() function produces HTML output, preSaveTransform() produces altered wiki markup.
20 # objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
22 # NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
24 # settings: $wgUseTex*, $wgUseCategoryMagic*, $wgUseDynamicDates*, $wgInterwikiMagic*,
25 # $wgNamespacesWithSubpages, $wgLanguageCode, $wgAllowExternalImages*,
28 # * only within ParserOptions
31 #----------------------------------------
32 # Variable substitution O(N^2) attack
33 #-----------------------------------------
34 # Without countermeasures, it would be possible to attack the parser by saving a page
35 # filled with a large number of inclusions of large pages. The size of the generated
36 # page would be proportional to the square of the input size. Hence, we limit the number
37 # of inclusions of any given page, thus bringing any attack back to O(N).
40 define( "MAX_INCLUDE_REPEAT", 5 );
42 # Allowed values for $mOutputType
43 define( "OT_HTML", 1 );
44 define( "OT_WIKI", 2 );
45 define( "OT_MSG", 3 );
47 # string parameter for extractTags which will cause it
48 # to strip HTML comments in addition to regular
49 # <XML>-style tags. This should not be anything we
50 # may want to use in wikisyntax
51 define( "STRIP_COMMENTS", "HTMLCommentStrip" );
53 # prefix for escaping, used in two functions at least
54 define( "UNIQ_PREFIX", "NaodW29");
58 # Cleared with clearState():
59 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
60 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
63 var $mOptions, $mTitle, $mOutputType;
72 $this->mOutput
= new ParserOutput
;
73 $this->mAutonumber
= 0;
74 $this->mLastSection
= "";
75 $this->mDTopen
= false;
76 $this->mVariables
= false;
77 $this->mIncludeCount
= array();
78 $this->mStripState
= array();
79 $this->mArgStack
= array();
82 # First pass--just handle <nowiki> sections, pass the rest off
83 # to internalParse() which does all the real work.
85 # Returns a ParserOutput
87 function parse( $text, &$title, $options, $linestart = true, $clearState = true )
89 $fname = "Parser::parse";
90 wfProfileIn( $fname );
96 $this->mOptions
= $options;
97 $this->mTitle
=& $title;
98 $this->mOutputType
= OT_HTML
;
101 $text = $this->strip( $text, $this->mStripState
);
102 $text = $this->internalParse( $text, $linestart );
103 $text = $this->unstrip( $text, $this->mStripState
);
104 # Clean up special characters, only run once, next-to-last before doBlockLevels
106 "/<hr *>/i" => '<hr/>',
107 "/<br *>/i" => '<br/>',
108 "/<center *>/i"=>'<div class="center">',
109 "/<\\/center *>/i" => '</div>',
110 # Clean up spare ampersands; note that we probably ought to be
111 # more careful about named entities.
112 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
114 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
117 $text = $this->doBlockLevels( $text, $linestart );
119 $this->mOutput
->setText( $text );
120 wfProfileOut( $fname );
121 return $this->mOutput
;
124 /* static */ function getRandomString()
126 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
129 # Replaces all occurrences of <$tag>content</$tag> in the text
130 # with a random marker and returns the new text. the output parameter
131 # $content will be an associative array filled with data on the form
132 # $unique_marker => content.
134 # If $content is already set, the additional entries will be appended
136 # If $tag is set to STRIP_COMMENTS, the function will extract
137 # <!-- HTML comments -->
139 /* static */ function extractTags($tag, $text, &$content, $uniq_prefix = ""){
140 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
147 while ( "" != $text ) {
148 if($tag==STRIP_COMMENTS
) {
149 $p = preg_split( "/<!--/i", $text, 2 );
151 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
154 if ( ( count( $p ) < 2 ) ||
( "" == $p[1] ) ) {
157 if($tag==STRIP_COMMENTS
) {
158 $q = preg_split( "/-->/i", $p[1], 2 );
160 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
162 $marker = $rnd . sprintf("%08X", $n++
);
163 $content[$marker] = $q[0];
164 $stripped .= $marker;
171 # Strips and renders <nowiki>, <pre>, <math>, <hiero>
172 # If $render is set, performs necessary rendering operations on plugins
173 # Returns the text, and fills an array with data needed in unstrip()
174 # If the $state is already a valid strip state, it adds to the state
176 # When $stripcomments is set, HTML comments <!-- like this -->
177 # will be stripped in addition to other tags. This is important
178 # for section editing, where these comments cause confusion when
179 # counting the sections in the wikisource
180 function strip( $text, &$state, $stripcomments = false )
182 $render = ($this->mOutputType
== OT_HTML
);
183 $nowiki_content = array();
184 $hiero_content = array();
185 $math_content = array();
186 $pre_content = array();
187 $comment_content = array();
189 # Replace any instances of the placeholders
190 $uniq_prefix = UNIQ_PREFIX
;
191 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
193 $text = Parser
::extractTags("nowiki", $text, $nowiki_content, $uniq_prefix);
194 foreach( $nowiki_content as $marker => $content ){
196 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
198 $nowiki_content[$marker] = "<nowiki>$content</nowiki>";
202 $text = Parser
::extractTags("hiero", $text, $hiero_content, $uniq_prefix);
203 foreach( $hiero_content as $marker => $content ){
204 if( $render && $GLOBALS['wgUseWikiHiero']){
205 $hiero_content[$marker] = WikiHiero( $content, WH_MODE_HTML
);
207 $hiero_content[$marker] = "<hiero>$content</hiero>";
211 $text = Parser
::extractTags("math", $text, $math_content, $uniq_prefix);
212 foreach( $math_content as $marker => $content ){
213 if( $render && $this->mOptions
->getUseTeX() ){
214 $math_content[$marker] = renderMath( $content );
216 $math_content[$marker] = "<math>$content</math>";
220 $text = Parser
::extractTags("pre", $text, $pre_content, $uniq_prefix);
221 foreach( $pre_content as $marker => $content ){
223 $pre_content[$marker] = "<pre>" . wfEscapeHTMLTagsOnly( $content ) . "</pre>";
225 $pre_content[$marker] = "<pre>$content</pre>";
229 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
230 foreach( $comment_content as $marker => $content ){
231 $comment_content[$marker] = "<!--$content-->";
235 # Merge state with the pre-existing state, if there is one
237 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
238 $state['hiero'] = $state['hiero'] +
$hiero_content;
239 $state['math'] = $state['math'] +
$math_content;
240 $state['pre'] = $state['pre'] +
$pre_content;
241 $state['comment'] = $state['comment'] +
$comment_content;
244 'nowiki' => $nowiki_content,
245 'hiero' => $hiero_content,
246 'math' => $math_content,
247 'pre' => $pre_content,
248 'comment' => $comment_content
254 function unstrip( $text, &$state )
256 # Must expand in reverse order, otherwise nested tags will be corrupted
257 $contentDict = end( $state );
258 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
259 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
260 $text = str_replace( key( $contentDict ), $content, $text );
267 # Add an item to the strip state
268 # Returns the unique tag which must be inserted into the stripped text
269 # The tag will be replaced with the original text in unstrip()
271 function insertStripItem( $text, &$state )
273 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
282 $state['item'][$rnd] = $text;
286 # This method generates the list of subcategories and pages for a category
287 function categoryMagic ()
289 global $wgLang , $wgUser ;
290 if ( !$this->mOptions
->getUseCategoryMagic() ) return ; # Doesn't use categories at all
292 $cns = Namespace::getCategory() ;
293 if ( $this->mTitle
->getNamespace() != $cns ) return "" ; # This ain't a category page
295 $r = "<br style=\"clear:both;\"/>\n";
298 $sk =& $wgUser->getSkin() ;
300 $articles = array() ;
301 $children = array() ;
303 $id = $this->mTitle
->getArticleID() ;
305 # For existing categories
307 $sql = "SELECT DISTINCT cur_title,cur_namespace FROM cur,links WHERE l_to={$id} AND l_from=cur_id";
308 $res = wfQuery ( $sql, DB_READ
) ;
309 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
311 # For non-existing categories
312 $t = wfStrencode( $this->mTitle
->getPrefixedDBKey() );
313 $sql = "SELECT DISTINCT cur_title,cur_namespace FROM cur,brokenlinks WHERE bl_to='$t' AND bl_from=cur_id" ;
314 $res = wfQuery ( $sql, DB_READ
) ;
315 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
318 # For all pages that link to this category
319 foreach ( $data AS $x )
321 $t = $wgLang->getNsText ( $x->cur_namespace
) ;
322 if ( $t != "" ) $t .= ":" ;
323 $t .= $x->cur_title
;
325 if ( $x->cur_namespace
== $cns ) {
326 array_push ( $children , $sk->makeLink ( $t ) ) ; # Subcategory
328 array_push ( $articles , $sk->makeLink ( $t ) ) ; # Page in this category
331 wfFreeResult ( $res ) ;
333 # Showing subcategories
334 if ( count ( $children ) > 0 )
336 asort ( $children ) ;
337 $r .= "<h2>".wfMsg("subcategories")."</h2>\n" ;
338 $r .= implode ( ", " , $children ) ;
341 # Showing pages in this category
342 if ( count ( $articles ) > 0 )
344 $ti = $this->mTitle
->getText() ;
345 asort ( $articles ) ;
346 $h = wfMsg( "category_header", $ti );
347 $r .= "<h2>{$h}</h2>\n" ;
348 $r .= implode ( ", " , $articles ) ;
355 function getHTMLattrs ()
357 $htmlattrs = array( # Allowed attributes--no scripting, etc.
358 "title", "align", "lang", "dir", "width", "height",
359 "bgcolor", "clear", /* BR */ "noshade", /* HR */
360 "cite", /* BLOCKQUOTE, Q */ "size", "face", "color",
361 /* FONT */ "type", "start", "value", "compact",
362 /* For various lists, mostly deprecated but safe */
363 "summary", "width", "border", "frame", "rules",
364 "cellspacing", "cellpadding", "valign", "char",
365 "charoff", "colgroup", "col", "span", "abbr", "axis",
366 "headers", "scope", "rowspan", "colspan", /* Tables */
367 "id", "class", "name", "style" /* For CSS */
372 function fixTagAttributes ( $t )
374 if ( trim ( $t ) == "" ) return "" ; # Saves runtime ;-)
375 $htmlattrs = $this->getHTMLattrs() ;
377 # Strip non-approved attributes from the tag
379 "/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e",
380 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
382 # Strip javascript "expression" from stylesheets. Brute force approach:
383 # If anythin offensive is found, all attributes of the HTML tag are dropped
386 "/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is",
387 wfMungeToUtf8( $t ) ) )
395 function doTableStuff ( $t )
397 $t = explode ( "\n" , $t ) ;
398 $td = array () ; # Is currently a td tag open?
399 $ltd = array () ; # Was it TD or TH?
400 $tr = array () ; # Is currently a tr tag open?
401 $ltr = array () ; # tr attributes
402 foreach ( $t AS $k => $x )
405 $fc = substr ( $x , 0 , 1 ) ;
406 if ( "{|" == substr ( $x , 0 , 2 ) )
408 $t[$k] = "\n<table " . $this->fixTagAttributes ( substr ( $x , 3 ) ) . ">" ;
409 array_push ( $td , false ) ;
410 array_push ( $ltd , "" ) ;
411 array_push ( $tr , false ) ;
412 array_push ( $ltr , "" ) ;
414 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
415 else if ( "|}" == substr ( $x , 0 , 2 ) )
418 $l = array_pop ( $ltd ) ;
419 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
420 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
424 /* else if ( "|_" == substr ( $x , 0 , 2 ) ) # Caption
426 $z = trim ( substr ( $x , 2 ) ) ;
427 $t[$k] = "<caption>{$z}</caption>\n" ;
429 else if ( "|-" == substr ( $x , 0 , 2 ) ) # Allows for |---------------
431 $x = substr ( $x , 1 ) ;
432 while ( $x != "" && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
434 $l = array_pop ( $ltd ) ;
435 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
436 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
439 array_push ( $tr , false ) ;
440 array_push ( $td , false ) ;
441 array_push ( $ltd , "" ) ;
442 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
444 else if ( "|" == $fc ||
"!" == $fc ||
"|+" == substr ( $x , 0 , 2 ) ) # Caption
446 if ( "|+" == substr ( $x , 0 , 2 ) )
449 $x = substr ( $x , 1 ) ;
451 $after = substr ( $x , 1 ) ;
452 if ( $fc == "!" ) $after = str_replace ( "!!" , "||" , $after ) ;
453 $after = explode ( "||" , $after ) ;
455 foreach ( $after AS $theline )
460 $tra = array_pop ( $ltr ) ;
461 if ( !array_pop ( $tr ) ) $z = "<tr {$tra}>\n" ;
462 array_push ( $tr , true ) ;
463 array_push ( $ltr , "" ) ;
466 $l = array_pop ( $ltd ) ;
467 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
468 if ( $fc == "|" ) $l = "td" ;
469 else if ( $fc == "!" ) $l = "th" ;
470 else if ( $fc == "+" ) $l = "caption" ;
472 array_push ( $ltd , $l ) ;
473 $y = explode ( "|" , $theline , 2 ) ;
474 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
475 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
477 array_push ( $td , true ) ;
482 # Closing open td, tr && table
483 while ( count ( $td ) > 0 )
485 if ( array_pop ( $td ) ) $t[] = "</td>" ;
486 if ( array_pop ( $tr ) ) $t[] = "</tr>" ;
490 $t = implode ( "\n" , $t ) ;
491 # $t = $this->removeHTMLtags( $t );
495 function internalParse( $text, $linestart, $args = array() )
497 $fname = "Parser::internalParse";
498 wfProfileIn( $fname );
500 $text = $this->removeHTMLtags( $text );
501 $text = $this->replaceVariables( $text, $args );
503 # $text = preg_replace( "/(^|\n)-----*/", "\\1<hr>", $text );
505 $text = $this->doHeadings( $text );
506 if($this->mOptions
->getUseDynamicDates()) {
507 global $wgDateFormatter;
508 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
510 $text = $this->replaceExternalLinks( $text );
511 $text = $this->doTokenizedParser ( $text );
512 $text = $this->doTableStuff ( $text ) ;
513 $text = $this->formatHeadings( $text );
514 $sk =& $this->mOptions
->getSkin();
515 $text = $sk->transformContent( $text );
517 if ( !isset ( $this->categoryMagicDone
) ) {
518 $text .= $this->categoryMagic () ;
519 $this->categoryMagicDone
= true ;
522 wfProfileOut( $fname );
527 /* private */ function doHeadings( $text )
529 for ( $i = 6; $i >= 1; --$i ) {
530 $h = substr( "======", 0, $i );
531 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
532 "<h{$i}>\\1</h{$i}>\\2", $text );
537 # Note: we have to do external links before the internal ones,
538 # and otherwise take great care in the order of things here, so
539 # that we don't end up interpreting some URLs twice.
541 /* private */ function replaceExternalLinks( $text )
543 $fname = "Parser::replaceExternalLinks";
544 wfProfileIn( $fname );
545 $text = $this->subReplaceExternalLinks( $text, "http", true );
546 $text = $this->subReplaceExternalLinks( $text, "https", true );
547 $text = $this->subReplaceExternalLinks( $text, "ftp", false );
548 $text = $this->subReplaceExternalLinks( $text, "irc", false );
549 $text = $this->subReplaceExternalLinks( $text, "gopher", false );
550 $text = $this->subReplaceExternalLinks( $text, "news", false );
551 $text = $this->subReplaceExternalLinks( $text, "mailto", false );
552 wfProfileOut( $fname );
556 /* private */ function subReplaceExternalLinks( $s, $protocol, $autonumber )
558 $unique = "4jzAfzB8hNvf4sqyO9Edd8pSmk9rE2in0Tgw3";
559 $uc = "A-Za-z0-9_\\/~%\\-+&*#?!=()@\\x80-\\xFF";
561 # this is the list of separators that should be ignored if they
562 # are the last character of an URL but that should be included
563 # if they occur within the URL, e.g. "go to www.foo.com, where .."
564 # in this case, the last comma should not become part of the URL,
565 # but in "www.foo.com/123,2342,32.htm" it should.
567 $fnc = "A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF";
568 $images = "gif|png|jpg|jpeg";
570 # PLEASE NOTE: The curly braces { } are not part of the regex,
571 # they are interpreted as part of the string (used to tell PHP
572 # that the content of the string should be inserted there).
573 $e1 = "/(^|[^\\[])({$protocol}:)([{$uc}{$sep}]+)\\/([{$fnc}]+)\\." .
574 "((?i){$images})([^{$uc}]|$)/";
576 $e2 = "/(^|[^\\[])({$protocol}:)(([".$uc."]|[".$sep."][".$uc."])+)([^". $uc . $sep. "]|[".$sep."]|$)/";
577 $sk =& $this->mOptions
->getSkin();
579 if ( $autonumber and $this->mOptions
->getAllowExternalImages() ) { # Use img tags only for HTTP urls
580 $s = preg_replace( $e1, "\\1" . $sk->makeImage( "{$unique}:\\3" .
581 "/\\4.\\5", "\\4.\\5" ) . "\\6", $s );
583 $s = preg_replace( $e2, "\\1" . "<a href=\"{$unique}:\\3\"" .
584 $sk->getExternalLinkAttributes( "{$unique}:\\3", wfEscapeHTML(
585 "{$unique}:\\3" ) ) . ">" . wfEscapeHTML( "{$unique}:\\3" ) .
587 $s = str_replace( $unique, $protocol, $s );
589 $a = explode( "[{$protocol}:", " " . $s );
590 $s = array_shift( $a );
591 $s = substr( $s, 1 );
593 $e1 = "/^([{$uc}"."{$sep}]+)](.*)\$/sD";
594 $e2 = "/^([{$uc}"."{$sep}]+)\\s+([^\\]]+)](.*)\$/sD";
596 foreach ( $a as $line ) {
597 if ( preg_match( $e1, $line, $m ) ) {
598 $link = "{$protocol}:{$m[1]}";
600 if ( $autonumber ) { $text = "[" . ++
$this->mAutonumber
. "]"; }
601 else { $text = wfEscapeHTML( $link ); }
602 } else if ( preg_match( $e2, $line, $m ) ) {
603 $link = "{$protocol}:{$m[1]}";
607 $s .= "[{$protocol}:" . $line;
610 if( $link == $text ||
preg_match( "!$protocol://" . preg_quote( $text, "/" ) . "/?$!", $link ) ) {
613 # Expand the URL for printable version
614 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $link ) . "</i>)</span>";
616 $la = $sk->getExternalLinkAttributes( $link, $text );
617 $s .= "<a href='{$link}'{$la}>{$text}</a>{$paren}{$trail}";
623 /* private */ function handle3Quotes( &$state, $token )
625 if ( $state["strong"] !== false ) {
626 if ( $state["em"] !== false && $state["em"] > $state["strong"] )
628 # ''' lala ''lala '''
629 $s = "</em></strong><em>";
633 $state["strong"] = FALSE;
636 $state["strong"] = isset($token["pos"]) ?
$token["pos"] : true;
641 /* private */ function handle2Quotes( &$state, $token )
643 if ( $state["em"] !== false ) {
644 if ( $state["strong"] !== false && $state["strong"] > $state["em"] )
646 # ''lala'''lala'' ....'''
647 $s = "</strong></em><strong>";
651 $state["em"] = FALSE;
654 $state["em"] = isset($token["pos"]) ?
$token["pos"] : true;
660 /* private */ function handle5Quotes( &$state, $token )
663 if ( $state["em"] !== false && $state["strong"] !== false ) {
664 if ( $state["em"] < $state["strong"] ) {
665 $s .= "</strong></em>";
667 $s .= "</em></strong>";
669 $state["strong"] = $state["em"] = FALSE;
670 } elseif ( $state["em"] !== false ) {
671 $s .= "</em><strong>";
672 $state["em"] = FALSE;
673 $state["strong"] = $token["pos"];
674 } elseif ( $state["strong"] !== false ) {
675 $s .= "</strong><em>";
676 $state["strong"] = FALSE;
677 $state["em"] = $token["pos"];
678 } else { # not $em and not $strong
679 $s .= "<strong><em>";
680 $state["strong"] = $state["em"] = isset($token["pos"]) ?
$token["pos"] : true;
685 /* private */ function doTokenizedParser( $str )
687 global $wgLang; # for language specific parser hook
688 global $wgUploadDirectory, $wgUseTimeline;
690 $tokenizer=Tokenizer
::newFromString( $str );
691 $tokenStack = array();
694 $state["em"] = FALSE;
695 $state["strong"] = FALSE;
699 # The tokenizer splits the text into tokens and returns them one by one.
700 # Every call to the tokenizer returns a new token.
701 while ( $token = $tokenizer->nextToken() )
703 switch ( $token["type"] )
706 # simple text with no further markup
707 $txt = $token["text"];
710 # Text that contains blanks that have to be converted to
711 # non-breakable spaces for French.
712 # U+202F NARROW NO-BREAK SPACE might be a better choice, but
713 # browser support for Unicode spacing is poor.
714 $txt = str_replace( " ", " ", $token["text"] );
717 # remember the tag opened with 3 [
721 # FIXME : Treat orphaned open tags (stack not empty when text is over)
723 array_push( $tokenStack, $token );
730 # get text from stack, glue it together, and call the code to handle a
733 if ( count( $tokenStack ) == 0 )
735 # stack empty. Found a ]] without an opening [[
739 $lastToken = array_pop( $tokenStack );
740 while ( !(($lastToken["type"] == "[[[") or ($lastToken["type"] == "[[")) )
742 if( !empty( $lastToken["text"] ) ) {
743 $linkText = $lastToken["text"] . $linkText;
745 $lastToken = array_pop( $tokenStack );
748 $txt = $linkText ."]]";
750 if( isset( $lastToken["text"] ) ) {
751 $prefix = $lastToken["text"];
755 $nextToken = $tokenizer->previewToken();
756 if ( $nextToken["type"] == "text" )
758 # Preview just looks at it. Now we have to fetch it.
759 $nextToken = $tokenizer->nextToken();
760 $txt .= $nextToken["text"];
762 $txt = $this->handleInternalLink( $this->unstrip($txt,$this->mStripState
), $prefix );
764 # did the tag start with 3 [ ?
766 # show the first as text
772 $tagIsOpen = (count( $tokenStack ) != 0);
778 # This and the three next ones handle quotes
779 $txt = $this->handle3Quotes( $state, $token );
782 $txt = $this->handle2Quotes( $state, $token );
785 $txt = $this->handle5Quotes( $state, $token );
795 $txt = $this->doMagicRFC( $tokenizer );
802 $txt = $this->doMagicISBN( $tokenizer );
806 if ( $wgUseTimeline &&
807 "" != ( $timelinesrc = $tokenizer->readAllUntil("</timeline>") ) )
809 $txt = renderTimeline( $timelinesrc );
815 # Call language specific Hook.
816 $txt = $wgLang->processToken( $token, $tokenStack );
817 if ( NULL == $txt ) {
818 # An unkown token. Highlight.
819 $txt = "<font color=\"#FF0000\"><b>".$token["type"]."</b></font>";
820 $txt .= "<font color=\"#FFFF00\"><b>".$token["text"]."</b></font>";
824 # If we're parsing the interior of a link, don't append the interior to $s,
825 # but push it to the stack so it can be processed when a ]] token is found.
826 if ( $tagIsOpen && $txt != "" ) {
827 $token["type"] = "text";
828 $token["text"] = $txt;
829 array_push( $tokenStack, $token );
834 if ( count( $tokenStack ) != 0 )
836 # still objects on stack. opened [[ tag without closing ]] tag.
838 while ( $lastToken = array_pop( $tokenStack ) )
840 if ( $lastToken["type"] == "text" )
842 $txt = $lastToken["text"] . $txt;
844 $txt = $lastToken["type"] . $txt;
852 /* private */ function handleInternalLink( $line, $prefix )
854 global $wgLang, $wgLinkCache;
855 global $wgNamespacesWithSubpages, $wgLanguageCode;
856 static $fname = "Parser::handleInternalLink" ;
857 wfProfileIn( $fname );
859 wfProfileIn( "$fname-setup" );
861 if ( !$tc ) { $tc = Title
::legalChars() . "#"; }
862 $sk =& $this->mOptions
->getSkin();
864 # Match a link having the form [[namespace:link|alternate]]trail
866 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
867 # Match the end of a line for a word that's not followed by whitespace,
868 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
869 #$e2 = "/^(.*)\\b(\\w+)\$/suD";
870 #$e2 = "/^(.*\\s)(\\S+)\$/suD";
871 static $e2 = '/^(.*\s)([a-zA-Z\x80-\xff]+)$/sD';
874 # Special and Media are pseudo-namespaces; no pages actually exist in them
875 static $image = FALSE;
876 static $special = FALSE;
877 static $media = FALSE;
878 static $category = FALSE;
879 if ( !$image ) { $image = Namespace::getImage(); }
880 if ( !$special ) { $special = Namespace::getSpecial(); }
881 if ( !$media ) { $media = Namespace::getMedia(); }
882 if ( !$category ) { $category = Namespace::getCategory(); }
884 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
886 wfProfileOut( "$fname-setup" );
889 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
892 } else { # Invalid form; output directly
893 $s .= $prefix . "[[" . $line ;
899 :Foobar -- override special treatment of prefix (images, language links)
900 /Foobar -- convert to CurrentPage/Foobar
901 /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
903 $c = substr($m[1],0,1);
904 $noforce = ($c != ":");
905 if( $c == "/" ) { # subpage
906 if(substr($m[1],-1,1)=="/") { # / at end means we don't want the slash to be shown
907 $m[1]=substr($m[1],1,strlen($m[1])-2);
910 $noslash=substr($m[1],1);
912 if($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]) { # subpages allowed here
913 $link = $this->mTitle
->getPrefixedText(). "/" . trim($noslash);
916 } # this might be changed for ugliness reasons
918 $link = $noslash; # no subpage allowed, use standard link
920 } elseif( $noforce ) { # no subpage
923 $link = substr( $m[1], 1 );
928 $nt = Title
::newFromText( $link );
930 $s .= $prefix . "[[" . $line;
933 $ns = $nt->getNamespace();
934 $iw = $nt->getInterWiki();
936 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
937 array_push( $this->mOutput
->mLanguageLinks
, $nt->getPrefixedText() );
938 $s .= $prefix . $trail ;
939 return (trim($s) == '')?
'': $s;
941 if( $ns == $image ) {
942 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
943 $wgLinkCache->addImageLinkObj( $nt );
946 if ( $ns == $category ) {
947 $t = $nt->getText() ;
948 $nnt = Title
::newFromText ( Namespace::getCanonicalName($category).":".$t ) ;
949 $t = $sk->makeLinkObj( $nnt, $t, "", "" , $prefix );
950 $this->mOutput
->mCategoryLinks
[] = $t ;
951 $s .= $prefix . $trail ;
955 if( ( $nt->getPrefixedText() == $this->mTitle
->getPrefixedText() ) &&
956 ( strpos( $link, "#" ) == FALSE ) ) {
957 # Self-links are handled specially; generally de-link and change to bold.
958 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, "", $trail );
962 if( $ns == $media ) {
963 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
964 $wgLinkCache->addImageLinkObj( $nt );
966 } elseif( $ns == $special ) {
967 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, "", $trail );
970 $s .= $sk->makeLinkObj( $nt, $text, "", $trail , $prefix );
972 wfProfileOut( $fname );
976 # Some functions here used by doBlockLevels()
978 /* private */ function closeParagraph()
981 if ( '' != $this->mLastSection
) {
982 $result = "</" . $this->mLastSection
. ">\n";
984 $this->mInPre
= false;
985 $this->mLastSection
= "";
988 # getCommon() returns the length of the longest common substring
989 # of both arguments, starting at the beginning of both.
991 /* private */ function getCommon( $st1, $st2 )
993 $fl = strlen( $st1 );
994 $shorter = strlen( $st2 );
995 if ( $fl < $shorter ) { $shorter = $fl; }
997 for ( $i = 0; $i < $shorter; ++
$i ) {
998 if ( $st1{$i} != $st2{$i} ) { break; }
1002 # These next three functions open, continue, and close the list
1003 # element appropriate to the prefix character passed into them.
1005 /* private */ function openList( $char )
1007 $result = $this->closeParagraph();
1009 if ( "*" == $char ) { $result .= "<ul><li>"; }
1010 else if ( "#" == $char ) { $result .= "<ol><li>"; }
1011 else if ( ":" == $char ) { $result .= "<dl><dd>"; }
1012 else if ( ";" == $char ) {
1013 $result .= "<dl><dt>";
1014 $this->mDTopen
= true;
1016 else { $result = "<!-- ERR 1 -->"; }
1021 /* private */ function nextItem( $char )
1023 if ( "*" == $char ||
"#" == $char ) { return "</li><li>"; }
1024 else if ( ":" == $char ||
";" == $char ) {
1026 if ( $this->mDTopen
) { $close = "</dt>"; }
1027 if ( ";" == $char ) {
1028 $this->mDTopen
= true;
1029 return $close . "<dt>";
1031 $this->mDTopen
= false;
1032 return $close . "<dd>";
1035 return "<!-- ERR 2 -->";
1038 /* private */function closeList( $char )
1040 if ( "*" == $char ) { $text = "</li></ul>"; }
1041 else if ( "#" == $char ) { $text = "</li></ol>"; }
1042 else if ( ":" == $char ) {
1043 if ( $this->mDTopen
) {
1044 $this->mDTopen
= false;
1045 $text = "</dt></dl>";
1047 $text = "</dd></dl>";
1050 else { return "<!-- ERR 3 -->"; }
1054 /* private */ function doBlockLevels( $text, $linestart ) {
1055 $fname = "Parser::doBlockLevels";
1056 wfProfileIn( $fname );
1058 # Parsing through the text line by line. The main thing
1059 # happening here is handling of block-level elements p, pre,
1060 # and making lists from lines starting with * # : etc.
1062 $textLines = explode( "\n", $text );
1064 $lastPrefix = $output = $lastLine = '';
1065 $this->mDTopen
= $inBlockElem = false;
1067 $paragraphStack = false;
1069 if ( !$linestart ) {
1070 $output .= array_shift( $textLines );
1072 foreach ( $textLines as $oLine ) {
1073 $lastPrefixLength = strlen( $lastPrefix );
1074 $preCloseMatch = preg_match("/<\\/pre/i", $oLine );
1075 $preOpenMatch = preg_match("/<pre/i", $oLine );
1076 if (!$this->mInPre
) {
1077 $this->mInPre
= !empty($preOpenMatch);
1079 if ( !$this->mInPre
) {
1080 # Multiple prefixes may abut each other for nested lists.
1081 $prefixLength = strspn( $oLine, "*#:;" );
1082 $pref = substr( $oLine, 0, $prefixLength );
1085 $pref2 = str_replace( ";", ":", $pref );
1086 $t = substr( $oLine, $prefixLength );
1088 # Don't interpret any other prefixes in preformatted text
1090 $pref = $pref2 = '';
1095 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1096 # Same as the last item, so no need to deal with nesting or opening stuff
1097 $output .= $this->nextItem( substr( $pref, -1 ) );
1098 $paragraphStack = false;
1100 if ( ";" == substr( $pref, -1 ) ) {
1101 # The one nasty exception: definition lists work like this:
1102 # ; title : definition text
1103 # So we check for : in the remainder text to split up the
1104 # title and definition, without b0rking links.
1105 # FIXME: This is not foolproof. Something better in Tokenizer might help.
1106 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1108 $output .= $term . $this->nextItem( ":" );
1112 } elseif( $prefixLength ||
$lastPrefixLength ) {
1113 # Either open or close a level...
1114 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1115 $paragraphStack = false;
1117 while( $commonPrefixLength < $lastPrefixLength ) {
1118 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1119 --$lastPrefixLength;
1121 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1122 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1124 while ( $prefixLength > $commonPrefixLength ) {
1125 $char = substr( $pref, $commonPrefixLength, 1 );
1126 $output .= $this->openList( $char );
1128 if ( ";" == $char ) {
1129 # FIXME: This is dupe of code above
1130 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1132 $output .= $term . $this->nextItem( ":" );
1136 ++
$commonPrefixLength;
1138 $lastPrefix = $pref2;
1140 if( 0 == $prefixLength ) {
1141 # No prefix (not in list)--go to paragraph mode
1142 $uniq_prefix = UNIQ_PREFIX
;
1143 // XXX: use a stack for nestable elements like span, table and div
1144 $openmatch = preg_match("/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<div|<pre|<tr|<td|<p|<ul|<li)/i", $t );
1145 $closematch = preg_match(
1146 "/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|".
1147 "<\\/div|<hr|<\\/td|<\\/pre|<\\/p|".$uniq_prefix."-pre|<\\/li|<\\/ul)/i", $t );
1148 if ( $openmatch or $closematch ) {
1149 $paragraphStack = false;
1150 $output .= $this->closeParagraph();
1151 if($preOpenMatch and !$preCloseMatch) {
1152 $this->mInPre
= true;
1154 if ( $closematch ) {
1155 $inBlockElem = false;
1157 $inBlockElem = true;
1159 } else if ( !$inBlockElem ) {
1160 if ( " " == $t{0} ) {
1162 if ($this->mLastSection
!= 'pre') {
1163 $paragraphStack = false;
1164 $output .= $this->closeParagraph().'<pre>';
1165 $this->mLastSection
= 'pre';
1169 if ( '' == trim($t) ) {
1170 if ( $paragraphStack ) {
1171 $output .= $paragraphStack.'<br/>';
1172 $paragraphStack = false;
1173 $this->mLastSection
= 'p';
1175 if ($this->mLastSection
!= 'p' ) {
1176 $output .= $this->closeParagraph();
1177 $this->mLastSection
= '';
1178 $paragraphStack = "<p>";
1180 $paragraphStack = '</p><p>';
1184 if ( $paragraphStack ) {
1185 $output .= $paragraphStack;
1186 $paragraphStack = false;
1187 $this->mLastSection
= 'p';
1188 } else if ($this->mLastSection
!= 'p') {
1189 $output .= $this->closeParagraph().'<p>';
1190 $this->mLastSection
= 'p';
1196 if ($paragraphStack === false) {
1200 while ( $prefixLength ) {
1201 $output .= $this->closeList( $pref2{$prefixLength-1} );
1204 if ( "" != $this->mLastSection
) {
1205 $output .= "</" . $this->mLastSection
. ">";
1206 $this->mLastSection
= "";
1209 wfProfileOut( $fname );
1213 function getVariableValue( $index ) {
1214 global $wgLang, $wgSitename, $wgServer;
1217 case MAG_CURRENTMONTH
:
1219 case MAG_CURRENTMONTHNAME
:
1220 return $wgLang->getMonthName( date("n") );
1221 case MAG_CURRENTMONTHNAMEGEN
:
1222 return $wgLang->getMonthNameGen( date("n") );
1223 case MAG_CURRENTDAY
:
1226 return $this->mTitle
->getText();
1228 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1229 return $wgLang->getNsText($this->mTitle
->getNamespace()); // Patch by Dori
1230 case MAG_CURRENTDAYNAME
:
1231 return $wgLang->getWeekdayName( date("w")+
1 );
1232 case MAG_CURRENTYEAR
:
1234 case MAG_CURRENTTIME
:
1235 return $wgLang->time( wfTimestampNow(), false );
1236 case MAG_NUMBEROFARTICLES
:
1237 return wfNumberOfArticles();
1247 function initialiseVariables()
1249 global $wgVariableIDs;
1250 $this->mVariables
= array();
1251 foreach ( $wgVariableIDs as $id ) {
1252 $mw =& MagicWord
::get( $id );
1253 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1257 /* private */ function replaceVariables( $text, $args = array() )
1259 global $wgLang, $wgScript, $wgArticlePath;
1261 $fname = "Parser::replaceVariables";
1262 wfProfileIn( $fname );
1265 if ( !$this->mVariables
) {
1266 $this->initialiseVariables();
1268 $titleChars = Title
::legalChars();
1269 $regex = "/(\\n?){{([$titleChars]*?)(\\|.*?|)}}/s";
1271 # This function is called recursively. To keep track of arguments we need a stack:
1272 array_push( $this->mArgStack
, $args );
1274 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1275 $GLOBALS['wgCurParser'] =& $this;
1276 $text = preg_replace_callback( $regex, "wfBraceSubstitution", $text );
1278 array_pop( $this->mArgStack
);
1283 function braceSubstitution( $matches )
1285 global $wgLinkCache, $wgLang;
1286 $fname = "Parser::braceSubstitution";
1291 # $newline is an optional newline character before the braces
1292 # $part1 is the bit before the first |, and must contain only title characters
1293 # $args is a list of arguments, starting from index 0, not including $part1
1295 $newline = $matches[1];
1296 $part1 = $matches[2];
1297 # If the third subpattern matched anything, it will start with |
1298 if ( $matches[3] !== "" ) {
1299 $args = explode( "|", substr( $matches[3], 1 ) );
1303 $argc = count( $args );
1306 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1307 if ( $mwSubst->matchStartAndRemove( $part1 ) ) {
1308 if ( $this->mOutputType
!= OT_WIKI
) {
1309 # Invalid SUBST not replaced at PST time
1310 # Return without further processing
1311 $text = $matches[0];
1314 } elseif ( $this->mOutputType
== OT_WIKI
) {
1315 # SUBST not found in PST pass, do nothing
1316 $text = $matches[0];
1320 # MSG, MSGNW and INT
1323 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1324 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1327 # Remove obsolete MSG:
1328 $mwMsg =& MagicWord
::get( MAG_MSG
);
1329 $mwMsg->matchStartAndRemove( $part1 );
1332 # Check if it is an internal message
1333 $mwInt =& MagicWord
::get( MAG_INT
);
1334 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1335 if ( $this->incrementIncludeCount( "int:$part1" ) ) {
1336 $text = wfMsgReal( $part1, $args, true );
1344 # Check for NS: (namespace expansion)
1345 $mwNs = MagicWord
::get( MAG_NS
);
1346 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1347 if ( intval( $part1 ) ) {
1348 $text = $wgLang->getNsText( intval( $part1 ) );
1351 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1352 if ( !is_null( $index ) ) {
1353 $text = $wgLang->getNsText( $index );
1360 # LOCALURL and LOCALURLE
1362 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1363 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1365 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1366 $func = 'getLocalURL';
1367 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1368 $func = 'escapeLocalURL';
1373 if ( $func !== '' ) {
1374 $title = Title
::newFromText( $part1 );
1375 if ( !is_null( $title ) ) {
1377 $text = $title->$func( $args[0] );
1379 $text = $title->$func();
1386 # Internal variables
1387 if ( !$found && array_key_exists( $part1, $this->mVariables
) ) {
1388 $text = $this->mVariables
[$part1];
1390 $this->mOutput
->mContainsOldMagic
= true;
1393 # Arguments input from the caller
1394 $inputArgs = end( $this->mArgStack
);
1395 if ( !$found && array_key_exists( $part1, $inputArgs ) ) {
1396 $text = $inputArgs[$part1];
1400 # Load from database
1402 $title = Title
::newFromText( $part1, NS_TEMPLATE
);
1403 if ( !is_null( $title ) && !$title->isExternal() ) {
1404 # Check for excessive inclusion
1405 $dbk = $title->getPrefixedDBkey();
1406 if ( $this->incrementIncludeCount( $dbk ) ) {
1407 $article = new Article( $title );
1408 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1409 if ( $articleContent !== false ) {
1411 $text = $articleContent;
1416 # If the title is valid but undisplayable, make a link to it
1417 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1418 $text = "[[" . $title->getPrefixedText() . "]]";
1424 # Recursive parsing, escaping and link table handling
1425 # Only for HTML output
1426 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1427 $text = wfEscapeWikiText( $text );
1428 } elseif ( $this->mOutputType
== OT_HTML
&& $found ) {
1429 # Clean up argument array
1430 $assocArgs = array();
1432 foreach( $args as $arg ) {
1433 $eqpos = strpos( $arg, "=" );
1434 if ( $eqpos === false ) {
1435 $assocArgs[$index++
] = $arg;
1437 $name = trim( substr( $arg, 0, $eqpos ) );
1438 $value = trim( substr( $arg, $eqpos+
1 ) );
1439 if ( $value === false ) {
1442 if ( $name !== false ) {
1443 $assocArgs[$name] = $value;
1448 # Do not enter included links in link table
1449 if ( !is_null( $title ) ) {
1450 $wgLinkCache->suspend();
1453 # Run full parser on the included text
1454 $text = $this->strip( $text, $this->mStripState
);
1455 $text = $this->internalParse( $text, (bool)$newline, $assocArgs );
1457 # Add the result to the strip state for re-inclusion after
1458 # the rest of the processing
1459 $text = $this->insertStripItem( $text, $this->mStripState
);
1461 # Resume the link cache and register the inclusion as a link
1462 if ( !is_null( $title ) ) {
1463 $wgLinkCache->resume();
1464 $wgLinkCache->addLinkObj( $title );
1471 return $newline . $text;
1475 # Returns true if the function is allowed to include this entity
1476 function incrementIncludeCount( $dbk )
1478 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1479 $this->mIncludeCount
[$dbk] = 0;
1481 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1489 # Cleans up HTML, removes dangerous tags and attributes
1490 /* private */ function removeHTMLtags( $text )
1492 $fname = "Parser::removeHTMLtags";
1493 wfProfileIn( $fname );
1494 $htmlpairs = array( # Tags that must be closed
1495 "b", "del", "i", "ins", "u", "font", "big", "small", "sub", "sup", "h1",
1496 "h2", "h3", "h4", "h5", "h6", "cite", "code", "em", "s",
1497 "strike", "strong", "tt", "var", "div", "center",
1498 "blockquote", "ol", "ul", "dl", "table", "caption", "pre",
1499 "ruby", "rt" , "rb" , "rp", "p"
1501 $htmlsingle = array(
1502 "br", "hr", "li", "dt", "dd"
1504 $htmlnest = array( # Tags that can be nested--??
1505 "table", "tr", "td", "th", "div", "blockquote", "ol", "ul",
1506 "dl", "font", "big", "small", "sub", "sup"
1508 $tabletags = array( # Can only appear inside table
1512 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1513 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1515 $htmlattrs = $this->getHTMLattrs () ;
1517 # Remove HTML comments
1518 $text = preg_replace( "/\n?<!--.*-->\n?/sU", "", $text );
1520 $bits = explode( "<", $text );
1521 $text = array_shift( $bits );
1522 $tagstack = array(); $tablestack = array();
1524 foreach ( $bits as $x ) {
1525 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
1526 preg_match( "/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/",
1528 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1529 error_reporting( $prev );
1532 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1536 if ( ! in_array( $t, $htmlsingle ) &&
1537 ( $ot = array_pop( $tagstack ) ) != $t ) {
1538 array_push( $tagstack, $ot );
1541 if ( $t == "table" ) {
1542 $tagstack = array_pop( $tablestack );
1547 # Keep track for later
1548 if ( in_array( $t, $tabletags ) &&
1549 ! in_array( "table", $tagstack ) ) {
1551 } else if ( in_array( $t, $tagstack ) &&
1552 ! in_array ( $t , $htmlnest ) ) {
1554 } else if ( ! in_array( $t, $htmlsingle ) ) {
1555 if ( $t == "table" ) {
1556 array_push( $tablestack, $tagstack );
1557 $tagstack = array();
1559 array_push( $tagstack, $t );
1561 # Strip non-approved attributes from the tag
1562 $newparams = $this->fixTagAttributes($params);
1566 $rest = str_replace( ">", ">", $rest );
1567 $text .= "<$slash$t $newparams$brace$rest";
1571 $text .= "<" . str_replace( ">", ">", $x);
1573 # Close off any remaining tags
1574 while ( $t = array_pop( $tagstack ) ) {
1576 if ( $t == "table" ) { $tagstack = array_pop( $tablestack ); }
1578 wfProfileOut( $fname );
1584 * This function accomplishes several tasks:
1585 * 1) Auto-number headings if that option is enabled
1586 * 2) Add an [edit] link to sections for logged in users who have enabled the option
1587 * 3) Add a Table of contents on the top for users who have enabled the option
1588 * 4) Auto-anchor headings
1590 * It loops through all headlines, collects the necessary data, then splits up the
1591 * string and re-inserts the newly formatted headlines.
1595 /* private */ function formatHeadings( $text )
1597 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
1598 $doShowToc = $this->mOptions
->getShowToc();
1599 if( !$this->mTitle
->userCanEdit() ) {
1601 $rightClickHack = 0;
1603 $showEditLink = $this->mOptions
->getEditSection();
1604 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
1607 # Inhibit editsection links if requested in the page
1608 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
1609 if( $esw->matchAndRemove( $text ) ) {
1612 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1614 $mw =& MagicWord
::get( MAG_NOTOC
);
1615 if( $mw->matchAndRemove( $text ) ) {
1619 # never add the TOC to the Main Page. This is an entry page that should not
1620 # be more than 1-2 screens large anyway
1621 if( $this->mTitle
->getPrefixedText() == wfMsg("mainpage") ) {
1625 # Get all headlines for numbering them and adding funky stuff like [edit]
1626 # links - this is for later, but we need the number of headlines right now
1627 $numMatches = preg_match_all( "/<H([1-6])(.*?" . ">)(.*?)<\/H[1-6]>/i", $text, $matches );
1629 # if there are fewer than 4 headlines in the article, do not show TOC
1630 if( $numMatches < 4 ) {
1634 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
1635 # override above conditions and always show TOC
1636 $mw =& MagicWord
::get( MAG_FORCETOC
);
1637 if ($mw->matchAndRemove( $text ) ) {
1642 # We need this to perform operations on the HTML
1643 $sk =& $this->mOptions
->getSkin();
1648 # Ugh .. the TOC should have neat indentation levels which can be
1649 # passed to the skin functions. These are determined here
1654 $sublevelCount = array();
1657 foreach( $matches[3] as $headline ) {
1660 $prevlevel = $level;
1662 $level = $matches[1][$headlineCount];
1663 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
1664 # reset when we enter a new level
1665 $sublevelCount[$level] = 0;
1666 $toc .= $sk->tocIndent( $level - $prevlevel );
1667 $toclevel +
= $level - $prevlevel;
1669 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
1670 # reset when we step back a level
1671 $sublevelCount[$level+
1]=0;
1672 $toc .= $sk->tocUnindent( $prevlevel - $level );
1673 $toclevel -= $prevlevel - $level;
1675 # count number of headlines for each level
1676 @$sublevelCount[$level]++
;
1677 if( $doNumberHeadings ||
$doShowToc ) {
1679 for( $i = 1; $i <= $level; $i++
) {
1680 if( !empty( $sublevelCount[$i] ) ) {
1684 $numbering .= $sublevelCount[$i];
1690 # The canonized header is a version of the header text safe to use for links
1691 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
1692 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
1695 $canonized_headline = preg_replace( "/<.*?" . ">/","",$canonized_headline );
1696 $tocline = trim( $canonized_headline );
1697 $canonized_headline = preg_replace("/[ \\?&\\/<>\\(\\)\\[\\]=,+']+/", '_', html_entity_decode( $tocline));
1698 $refer[$headlineCount] = $canonized_headline;
1700 # count how many in assoc. array so we can track dupes in anchors
1701 @$refers[$canonized_headline]++
;
1702 $refcount[$headlineCount]=$refers[$canonized_headline];
1704 # Prepend the number to the heading text
1706 if( $doNumberHeadings ||
$doShowToc ) {
1707 $tocline = $numbering . " " . $tocline;
1709 # Don't number the heading if it is the only one (looks silly)
1710 if( $doNumberHeadings && count( $matches[3] ) > 1) {
1711 # the two are different if the line contains a link
1712 $headline=$numbering . " " . $headline;
1716 # Create the anchor for linking from the TOC to the section
1717 $anchor = $canonized_headline;
1718 if($refcount[$headlineCount] > 1 ) {
1719 $anchor .= "_" . $refcount[$headlineCount];
1722 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
1724 if( $showEditLink ) {
1725 if ( empty( $head[$headlineCount] ) ) {
1726 $head[$headlineCount] = "";
1728 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+
1);
1731 # Add the edit section span
1732 if( $rightClickHack ) {
1733 $headline = $sk->editSectionScript($headlineCount+
1,$headline);
1736 # give headline the correct <h#> tag
1737 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline."</h".$level.">";
1743 $toclines = $headlineCount;
1744 $toc .= $sk->tocUnindent( $toclevel );
1745 $toc = $sk->tocTable( $toc );
1748 # split up and insert constructed headlines
1750 $blocks = preg_split( "/<H[1-6].*?" . ">.*?<\/H[1-6]>/i", $text );
1753 foreach( $blocks as $block ) {
1754 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
1755 # This is the [edit] link that appears for the top block of text when
1756 # section editing is enabled
1758 # Disabled because it broke block formatting
1759 # For example, a bullet point in the top line
1760 # $full .= $sk->editSectionLink(0);
1763 if( $doShowToc && !$i) {
1764 # Top anchor now in skin
1768 if( !empty( $head[$i] ) ) {
1777 /* private */ function doMagicISBN( &$tokenizer )
1781 # Check whether next token is a text token
1782 # If yes, fetch it and convert the text into a
1783 # Special::BookSources link
1784 $token = $tokenizer->previewToken();
1785 while ( $token["type"] == "" )
1787 $tokenizer->nextToken();
1788 $token = $tokenizer->previewToken();
1790 if ( $token["type"] == "text" )
1792 $token = $tokenizer->nextToken();
1793 $x = $token["text"];
1794 $valid = "0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ";
1796 $isbn = $blank = "" ;
1797 while ( " " == $x{0} ) {
1799 $x = substr( $x, 1 );
1801 while ( strstr( $valid, $x{0} ) != false ) {
1803 $x = substr( $x, 1 );
1805 $num = str_replace( "-", "", $isbn );
1806 $num = str_replace( " ", "", $num );
1809 $text = "ISBN $blank$x";
1811 $titleObj = Title
::makeTitle( NS_SPECIAL
, "Booksources" );
1812 $text = "<a href=\"" .
1813 $titleObj->escapeLocalUrl( "isbn={$num}" ) .
1814 "\" class=\"internal\">ISBN $isbn</a>";
1822 /* private */ function doMagicRFC( &$tokenizer )
1826 # Check whether next token is a text token
1827 # If yes, fetch it and convert the text into a
1828 # link to an RFC source
1829 $token = $tokenizer->previewToken();
1830 while ( $token["type"] == "" )
1832 $tokenizer->nextToken();
1833 $token = $tokenizer->previewToken();
1835 if ( $token["type"] == "text" )
1837 $token = $tokenizer->nextToken();
1838 $x = $token["text"];
1839 $valid = "0123456789";
1841 $rfc = $blank = "" ;
1842 while ( " " == $x{0} ) {
1844 $x = substr( $x, 1 );
1846 while ( strstr( $valid, $x{0} ) != false ) {
1848 $x = substr( $x, 1 );
1852 $text .= "RFC $blank$x";
1854 $url = wfmsg( "rfcurl" );
1855 $url = str_replace( "$1", $rfc, $url);
1856 $sk =& $this->mOptions
->getSkin();
1857 $la = $sk->getExternalLinkAttributes( $url, "RFC {$rfc}" );
1858 $text = "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
1866 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true )
1868 $this->mOptions
= $options;
1869 $this->mTitle
=& $title;
1870 $this->mOutputType
= OT_WIKI
;
1872 if ( $clearState ) {
1873 $this->clearState();
1876 $stripState = false;
1880 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
1883 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
1884 "/<br *?>/i" => "<br/>",
1886 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
1887 $text = $this->strip( $text, $stripState, false );
1888 $text = $this->pstPass2( $text, $user );
1889 $text = $this->unstrip( $text, $stripState );
1893 /* private */ function pstPass2( $text, &$user )
1895 global $wgLang, $wgLocaltimezone, $wgCurParser;
1897 # Variable replacement
1898 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
1899 $text = $this->replaceVariables( $text );
1903 $n = $user->getName();
1904 $k = $user->getOption( "nickname" );
1905 if ( "" == $k ) { $k = $n; }
1906 if(isset($wgLocaltimezone)) {
1907 $oldtz = getenv("TZ"); putenv("TZ=$wgLocaltimezone");
1909 /* Note: this is an ugly timezone hack for the European wikis */
1910 $d = $wgLang->timeanddate( date( "YmdHis" ), false ) .
1911 " (" . date( "T" ) . ")";
1912 if(isset($wgLocaltimezone)) putenv("TZ=$oldtz");
1914 $text = preg_replace( "/~~~~~/", $d, $text );
1915 $text = preg_replace( "/~~~~/", "[[" . $wgLang->getNsText(
1916 Namespace::getUser() ) . ":$n|$k]] $d", $text );
1917 $text = preg_replace( "/~~~/", "[[" . $wgLang->getNsText(
1918 Namespace::getUser() ) . ":$n|$k]]", $text );
1920 # Context links: [[|name]] and [[name (context)|]]
1922 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
1923 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
1924 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
1925 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
1927 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
1928 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
1929 $p3 = "/\[\[($namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]]
1930 $p4 = "/\[\[($namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/";
1931 # [[ns:page (cont)|]]
1933 $t = $this->mTitle
->getText();
1934 if ( preg_match( $conpat, $t, $m ) ) {
1937 $text = preg_replace( $p4, "[[\\1:\\2 (\\3)|\\2]]", $text );
1938 $text = preg_replace( $p1, "[[\\1 (\\2)|\\1]]", $text );
1939 $text = preg_replace( $p3, "[[\\1:\\2|\\2]]", $text );
1941 if ( "" == $context ) {
1942 $text = preg_replace( $p2, "[[\\1]]", $text );
1944 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
1948 $mw =& MagicWord::get( MAG_SUBST );
1949 $wgCurParser = $this->fork();
1950 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
1951 $this->merge( $wgCurParser );
1954 # Trim trailing whitespace
1955 # MAG_END (__END__) tag allows for trailing
1956 # whitespace to be deliberately included
1957 $text = rtrim( $text );
1958 $mw =& MagicWord
::get( MAG_END
);
1959 $mw->matchAndRemove( $text );
1964 # Set up some variables which are usually set up in parse()
1965 # so that an external function can call some class members with confidence
1966 function startExternalParse( &$title, $options, $outputType, $clearState = true )
1968 $this->mTitle
=& $title;
1969 $this->mOptions
= $options;
1970 $this->mOutputType
= $outputType;
1971 if ( $clearState ) {
1972 $this->clearState();
1976 function transformMsg( $text, $options ) {
1978 static $executing = false;
1980 # Guard against infinite recursion
1986 $this->mTitle
= $wgTitle;
1987 $this->mOptions
= $options;
1988 $this->mOutputType
= OT_MSG
;
1989 $this->clearState();
1990 $text = $this->replaceVariables( $text );
1999 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2001 function ParserOutput( $text = "", $languageLinks = array(), $categoryLinks = array(),
2002 $containsOldMagic = false )
2004 $this->mText
= $text;
2005 $this->mLanguageLinks
= $languageLinks;
2006 $this->mCategoryLinks
= $categoryLinks;
2007 $this->mContainsOldMagic
= $containsOldMagic;
2010 function getText() { return $this->mText
; }
2011 function getLanguageLinks() { return $this->mLanguageLinks
; }
2012 function getCategoryLinks() { return $this->mCategoryLinks
; }
2013 function containsOldMagic() { return $this->mContainsOldMagic
; }
2014 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2015 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2016 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2017 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2019 function merge( $other ) {
2020 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2021 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2022 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2029 # All variables are private
2030 var $mUseTeX; # Use texvc to expand <math> tags
2031 var $mUseCategoryMagic; # Treat [[Category:xxxx]] tags specially
2032 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2033 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2034 var $mAllowExternalImages; # Allow external images inline
2035 var $mSkin; # Reference to the preferred skin
2036 var $mDateFormat; # Date format index
2037 var $mEditSection; # Create "edit section" links
2038 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2039 var $mNumberHeadings; # Automatically number headings
2040 var $mShowToc; # Show table of contents
2042 function getUseTeX() { return $this->mUseTeX
; }
2043 function getUseCategoryMagic() { return $this->mUseCategoryMagic
; }
2044 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2045 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2046 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2047 function getSkin() { return $this->mSkin
; }
2048 function getDateFormat() { return $this->mDateFormat
; }
2049 function getEditSection() { return $this->mEditSection
; }
2050 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
2051 function getNumberHeadings() { return $this->mNumberHeadings
; }
2052 function getShowToc() { return $this->mShowToc
; }
2054 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2055 function setUseCategoryMagic( $x ) { return wfSetVar( $this->mUseCategoryMagic
, $x ); }
2056 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2057 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2058 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2059 function setSkin( $x ) { return wfSetRef( $this->mSkin
, $x ); }
2060 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2061 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
2062 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
2063 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
2064 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
2066 /* static */ function newFromUser( &$user )
2068 $popts = new ParserOptions
;
2069 $popts->initialiseFromUser( $user );
2073 function initialiseFromUser( &$userInput )
2075 global $wgUseTeX, $wgUseCategoryMagic, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2077 if ( !$userInput ) {
2079 $user->setLoaded( true );
2081 $user =& $userInput;
2084 $this->mUseTeX
= $wgUseTeX;
2085 $this->mUseCategoryMagic
= $wgUseCategoryMagic;
2086 $this->mUseDynamicDates
= $wgUseDynamicDates;
2087 $this->mInterwikiMagic
= $wgInterwikiMagic;
2088 $this->mAllowExternalImages
= $wgAllowExternalImages;
2089 $this->mSkin
=& $user->getSkin();
2090 $this->mDateFormat
= $user->getOption( "date" );
2091 $this->mEditSection
= $user->getOption( "editsection" );
2092 $this->mEditSectionOnRightClick
= $user->getOption( "editsectiononrightclick" );
2093 $this->mNumberHeadings
= $user->getOption( "numberheadings" );
2094 $this->mShowToc
= $user->getOption( "showtoc" );
2100 # Regex callbacks, used in Parser::replaceVariables
2101 function wfBraceSubstitution( $matches )
2103 global $wgCurParser;
2104 return $wgCurParser->braceSubstitution( $matches );