3 include_once('Tokenizer.php');
5 if( $GLOBALS['wgUseWikiHiero'] ){
6 include_once('wikihiero.php');
11 # Processes wiki markup
13 # There are two main entry points into the Parser class: parse() and preSaveTransform().
14 # The parse() function produces HTML output, preSaveTransform() produces altered wiki markup.
17 # objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
19 # NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
21 # settings: $wgUseTex*, $wgUseCategoryMagic*, $wgUseDynamicDates*, $wgInterwikiMagic*,
22 # $wgNamespacesWithSubpages, $wgLanguageCode, $wgAllowExternalImages*,
25 # * only within ParserOptions
28 #----------------------------------------
29 # Variable substitution O(N^2) attack
30 #-----------------------------------------
31 # Without countermeasures, it would be possible to attack the parser by saving a page
32 # filled with a large number of inclusions of large pages. The size of the generated
33 # page would be proportional to the square of the input size. Hence, we limit the number
34 # of inclusions of any given page, thus bringing any attack back to O(N).
37 define( "MAX_INCLUDE_REPEAT", 5 );
39 # Allowed values for $mOutputType
40 define( "OT_HTML", 1 );
41 define( "OT_WIKI", 2 );
42 define( "OT_MSG", 3 );
44 # prefix for escaping, used in two functions at least
45 define( "UNIQ_PREFIX", "NaodW29");
49 # Cleared with clearState():
50 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
51 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
54 var $mOptions, $mTitle, $mOutputType;
63 $this->mOutput
= new ParserOutput
;
64 $this->mAutonumber
= 0;
65 $this->mLastSection
= "";
66 $this->mDTopen
= false;
67 $this->mVariables
= false;
68 $this->mIncludeCount
= array();
69 $this->mStripState
= array();
70 $this->mArgStack
= array();
73 # First pass--just handle <nowiki> sections, pass the rest off
74 # to internalParse() which does all the real work.
76 # Returns a ParserOutput
78 function parse( $text, &$title, $options, $linestart = true, $clearState = true )
80 $fname = "Parser::parse";
81 wfProfileIn( $fname );
87 $this->mOptions
= $options;
88 $this->mTitle
=& $title;
89 $this->mOutputType
= OT_HTML
;
92 $text = $this->strip( $text, $this->mStripState
);
93 $text = $this->internalParse( $text, $linestart );
94 $text = $this->unstrip( $text, $this->mStripState
);
95 # Clean up special characters, only run once, next-to-last before doBlockLevels
97 "/<hr *>/i" => '<hr/>',
98 "/<br *>/i" => '<br/>',
99 "/<center *>/i"=>'<div class="center">',
100 "/<\\/center *>/i" => '</div>',
101 # Clean up spare ampersands; note that we probably ought to be
102 # more careful about named entities.
103 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
105 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
108 $text = $this->doBlockLevels( $text, $linestart );
110 $this->mOutput
->setText( $text );
111 wfProfileOut( $fname );
112 return $this->mOutput
;
115 /* static */ function getRandomString()
117 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
120 # Replaces all occurrences of <$tag>content</$tag> in the text
121 # with a random marker and returns the new text. the output parameter
122 # $content will be an associative array filled with data on the form
123 # $unique_marker => content.
125 # If $content is already set, the additional entries will be appended
127 /* static */ function extractTags($tag, $text, &$content, $uniq_prefix = ""){
128 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
135 while ( "" != $text ) {
136 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
138 if ( ( count( $p ) < 2 ) ||
( "" == $p[1] ) ) {
141 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
142 $marker = $rnd . sprintf("%08X", $n++
);
143 $content[$marker] = $q[0];
144 $stripped .= $marker;
151 # Strips <nowiki>, <pre> and <math>
152 # Returns the text, and fills an array with data needed in unstrip()
153 # If the $state is already a valid strip state, it adds to the state
155 function strip( $text, &$state )
157 $render = ($this->mOutputType
== OT_HTML
);
158 $nowiki_content = array();
159 $hiero_content = array();
160 $math_content = array();
161 $pre_content = array();
162 $item_content = array();
164 # Replace any instances of the placeholders
165 $uniq_prefix = UNIQ_PREFIX
;
166 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
168 $text = Parser
::extractTags("nowiki", $text, $nowiki_content, $uniq_prefix);
169 foreach( $nowiki_content as $marker => $content ){
171 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
173 $nowiki_content[$marker] = "<nowiki>$content</nowiki>";
177 if( $GLOBALS['wgUseWikiHiero'] ){
178 $text = Parser
::extractTags("hiero", $text, $hiero_content, $uniq_prefix);
179 foreach( $hiero_content as $marker => $content ){
181 $hiero_content[$marker] = WikiHiero( $content, WH_MODE_HTML
);
183 $hiero_content[$marker] = "<hiero>$content</hiero>";
188 if( $this->mOptions
->getUseTeX() ){
189 $text = Parser
::extractTags("math", $text, $math_content, $uniq_prefix);
190 foreach( $math_content as $marker => $content ){
192 $math_content[$marker] = renderMath( $content );
194 $math_content[$marker] = "<math>$content</math>";
199 $text = Parser
::extractTags("pre", $text, $pre_content, $uniq_prefix);
200 foreach( $pre_content as $marker => $content ){
202 $pre_content[$marker] = "<pre>" . wfEscapeHTMLTagsOnly( $content ) . "</pre>";
204 $pre_content[$marker] = "<pre>$content</pre>";
208 # Merge state with the pre-existing state, if there is one
210 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
211 $state['hiero'] = $state['hiero'] +
$hiero_content;
212 $state['math'] = $state['math'] +
$math_content;
213 $state['pre'] = $state['pre'] +
$pre_content;
216 'nowiki' => $nowiki_content,
217 'hiero' => $hiero_content,
218 'math' => $math_content,
219 'pre' => $pre_content,
220 'item' => $item_content
226 function unstrip( $text, &$state )
228 # Must expand in reverse order, otherwise nested tags will be corrupted
229 $contentDict = end( $state );
230 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
231 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
232 $text = str_replace( key( $contentDict ), $content, $text );
239 # Add an item to the strip state
240 # Returns the unique tag which must be inserted into the stripped text
241 # The tag will be replaced with the original text in unstrip()
243 function insertStripItem( $text, &$state )
245 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
255 $state['item'][$rnd] = $text;
259 # This method generates the list of subcategories and pages for a category
260 function categoryMagic ()
262 global $wgLang , $wgUser ;
263 if ( !$this->mOptions
->getUseCategoryMagic() ) return ; # Doesn't use categories at all
265 $cns = Namespace::getCategory() ;
266 if ( $this->mTitle
->getNamespace() != $cns ) return "" ; # This ain't a category page
268 $r = "<br style=\"clear:both;\"/>\n";
271 $sk =& $wgUser->getSkin() ;
273 $articles = array() ;
274 $children = array() ;
276 $id = $this->mTitle
->getArticleID() ;
278 # For existing categories
279 $sql = "SELECT DISTINCT cur_title,cur_namespace FROM cur,links WHERE l_to={$id} AND l_from=cur_id";
280 $res = wfQuery ( $sql, DB_READ
) ;
281 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
283 # For non-existing categories
284 $sql = "SELECT DISTINCT cur_title,cur_namespace FROM cur,brokenlinks WHERE bl_to={$id} AND bl_from=cur_id" ;
285 $res = wfQuery ( $sql, DB_READ
) ;
286 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
288 # For all pages that link to this category
289 foreach ( $data AS $x )
291 $t = $wgLang->getNsText ( $x->cur_namespace
) ;
292 if ( $t != "" ) $t .= ":" ;
293 $t .= $x->cur_title
;
295 if ( $x->cur_namespace
== $cns ) {
296 array_push ( $children , $sk->makeLink ( $t ) ) ; # Subcategory
298 array_push ( $articles , $sk->makeLink ( $t ) ) ; # Page in this category
301 wfFreeResult ( $res ) ;
303 # Showing subcategories
304 if ( count ( $children ) > 0 )
306 asort ( $children ) ;
307 $r .= "<h2>".wfMsg("subcategories")."</h2>\n" ;
308 $r .= implode ( ", " , $children ) ;
311 # Showing pages in this category
312 if ( count ( $articles ) > 0 )
314 $ti = $this->mTitle
->getText() ;
315 asort ( $articles ) ;
316 $h = wfMsg( "category_header", $ti );
317 $r .= "<h2>{$h}</h2>\n" ;
318 $r .= implode ( ", " , $articles ) ;
325 function getHTMLattrs ()
327 $htmlattrs = array( # Allowed attributes--no scripting, etc.
328 "title", "align", "lang", "dir", "width", "height",
329 "bgcolor", "clear", /* BR */ "noshade", /* HR */
330 "cite", /* BLOCKQUOTE, Q */ "size", "face", "color",
331 /* FONT */ "type", "start", "value", "compact",
332 /* For various lists, mostly deprecated but safe */
333 "summary", "width", "border", "frame", "rules",
334 "cellspacing", "cellpadding", "valign", "char",
335 "charoff", "colgroup", "col", "span", "abbr", "axis",
336 "headers", "scope", "rowspan", "colspan", /* Tables */
337 "id", "class", "name", "style" /* For CSS */
342 function fixTagAttributes ( $t )
344 if ( trim ( $t ) == "" ) return "" ; # Saves runtime ;-)
345 $htmlattrs = $this->getHTMLattrs() ;
347 # Strip non-approved attributes from the tag
349 "/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e",
350 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
352 # Strip javascript "expression" from stylesheets. Brute force approach:
353 # If anythin offensive is found, all attributes of the HTML tag are dropped
356 "/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is",
357 wfMungeToUtf8( $t ) ) )
365 function doTableStuff ( $t )
367 $t = explode ( "\n" , $t ) ;
368 $td = array () ; # Is currently a td tag open?
369 $ltd = array () ; # Was it TD or TH?
370 $tr = array () ; # Is currently a tr tag open?
371 $ltr = array () ; # tr attributes
372 foreach ( $t AS $k => $x )
375 $fc = substr ( $x , 0 , 1 ) ;
376 if ( "{|" == substr ( $x , 0 , 2 ) )
378 $t[$k] = "\n<table " . $this->fixTagAttributes ( substr ( $x , 3 ) ) . ">" ;
379 array_push ( $td , false ) ;
380 array_push ( $ltd , "" ) ;
381 array_push ( $tr , false ) ;
382 array_push ( $ltr , "" ) ;
384 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
385 else if ( "|}" == substr ( $x , 0 , 2 ) )
388 $l = array_pop ( $ltd ) ;
389 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
390 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
394 /* else if ( "|_" == substr ( $x , 0 , 2 ) ) # Caption
396 $z = trim ( substr ( $x , 2 ) ) ;
397 $t[$k] = "<caption>{$z}</caption>\n" ;
399 else if ( "|-" == substr ( $x , 0 , 2 ) ) # Allows for |---------------
401 $x = substr ( $x , 1 ) ;
402 while ( $x != "" && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
404 $l = array_pop ( $ltd ) ;
405 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
406 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
409 array_push ( $tr , false ) ;
410 array_push ( $td , false ) ;
411 array_push ( $ltd , "" ) ;
412 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
414 else if ( "|" == $fc ||
"!" == $fc ||
"|+" == substr ( $x , 0 , 2 ) ) # Caption
416 if ( "|+" == substr ( $x , 0 , 2 ) )
419 $x = substr ( $x , 1 ) ;
421 $after = substr ( $x , 1 ) ;
422 if ( $fc == "!" ) $after = str_replace ( "!!" , "||" , $after ) ;
423 $after = explode ( "||" , $after ) ;
425 foreach ( $after AS $theline )
430 $tra = array_pop ( $ltr ) ;
431 if ( !array_pop ( $tr ) ) $z = "<tr {$tra}>\n" ;
432 array_push ( $tr , true ) ;
433 array_push ( $ltr , "" ) ;
436 $l = array_pop ( $ltd ) ;
437 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
438 if ( $fc == "|" ) $l = "td" ;
439 else if ( $fc == "!" ) $l = "th" ;
440 else if ( $fc == "+" ) $l = "caption" ;
442 array_push ( $ltd , $l ) ;
443 $y = explode ( "|" , $theline , 2 ) ;
444 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
445 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
447 array_push ( $td , true ) ;
452 # Closing open td, tr && table
453 while ( count ( $td ) > 0 )
455 if ( array_pop ( $td ) ) $t[] = "</td>" ;
456 if ( array_pop ( $tr ) ) $t[] = "</tr>" ;
460 $t = implode ( "\n" , $t ) ;
461 # $t = $this->removeHTMLtags( $t );
465 function internalParse( $text, $linestart, $args = array() )
467 $fname = "Parser::internalParse";
468 wfProfileIn( $fname );
470 $text = $this->removeHTMLtags( $text );
471 $text = $this->replaceVariables( $text, $args );
473 # $text = preg_replace( "/(^|\n)-----*/", "\\1<hr>", $text );
475 $text = $this->doHeadings( $text );
476 if($this->mOptions
->getUseDynamicDates()) {
477 global $wgDateFormatter;
478 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
480 $text = $this->replaceExternalLinks( $text );
481 $text = $this->doTokenizedParser ( $text );
482 $text = $this->doTableStuff ( $text ) ;
483 $text = $this->formatHeadings( $text );
484 $sk =& $this->mOptions
->getSkin();
485 $text = $sk->transformContent( $text );
487 if ( !isset ( $this->categoryMagicDone
) ) {
488 $text .= $this->categoryMagic () ;
489 $this->categoryMagicDone
= true ;
492 wfProfileOut( $fname );
497 /* private */ function doHeadings( $text )
499 for ( $i = 6; $i >= 1; --$i ) {
500 $h = substr( "======", 0, $i );
501 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
502 "<h{$i}>\\1</h{$i}>\\2", $text );
507 # Note: we have to do external links before the internal ones,
508 # and otherwise take great care in the order of things here, so
509 # that we don't end up interpreting some URLs twice.
511 /* private */ function replaceExternalLinks( $text )
513 $fname = "Parser::replaceExternalLinks";
514 wfProfileIn( $fname );
515 $text = $this->subReplaceExternalLinks( $text, "http", true );
516 $text = $this->subReplaceExternalLinks( $text, "https", true );
517 $text = $this->subReplaceExternalLinks( $text, "ftp", false );
518 $text = $this->subReplaceExternalLinks( $text, "irc", false );
519 $text = $this->subReplaceExternalLinks( $text, "gopher", false );
520 $text = $this->subReplaceExternalLinks( $text, "news", false );
521 $text = $this->subReplaceExternalLinks( $text, "mailto", false );
522 wfProfileOut( $fname );
526 /* private */ function subReplaceExternalLinks( $s, $protocol, $autonumber )
528 $unique = "4jzAfzB8hNvf4sqyO9Edd8pSmk9rE2in0Tgw3";
529 $uc = "A-Za-z0-9_\\/~%\\-+&*#?!=()@\\x80-\\xFF";
531 # this is the list of separators that should be ignored if they
532 # are the last character of an URL but that should be included
533 # if they occur within the URL, e.g. "go to www.foo.com, where .."
534 # in this case, the last comma should not become part of the URL,
535 # but in "www.foo.com/123,2342,32.htm" it should.
537 $fnc = "A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF";
538 $images = "gif|png|jpg|jpeg";
540 # PLEASE NOTE: The curly braces { } are not part of the regex,
541 # they are interpreted as part of the string (used to tell PHP
542 # that the content of the string should be inserted there).
543 $e1 = "/(^|[^\\[])({$protocol}:)([{$uc}{$sep}]+)\\/([{$fnc}]+)\\." .
544 "((?i){$images})([^{$uc}]|$)/";
546 $e2 = "/(^|[^\\[])({$protocol}:)(([".$uc."]|[".$sep."][".$uc."])+)([^". $uc . $sep. "]|[".$sep."]|$)/";
547 $sk =& $this->mOptions
->getSkin();
549 if ( $autonumber and $this->mOptions
->getAllowExternalImages() ) { # Use img tags only for HTTP urls
550 $s = preg_replace( $e1, "\\1" . $sk->makeImage( "{$unique}:\\3" .
551 "/\\4.\\5", "\\4.\\5" ) . "\\6", $s );
553 $s = preg_replace( $e2, "\\1" . "<a href=\"{$unique}:\\3\"" .
554 $sk->getExternalLinkAttributes( "{$unique}:\\3", wfEscapeHTML(
555 "{$unique}:\\3" ) ) . ">" . wfEscapeHTML( "{$unique}:\\3" ) .
557 $s = str_replace( $unique, $protocol, $s );
559 $a = explode( "[{$protocol}:", " " . $s );
560 $s = array_shift( $a );
561 $s = substr( $s, 1 );
563 $e1 = "/^([{$uc}"."{$sep}]+)](.*)\$/sD";
564 $e2 = "/^([{$uc}"."{$sep}]+)\\s+([^\\]]+)](.*)\$/sD";
566 foreach ( $a as $line ) {
567 if ( preg_match( $e1, $line, $m ) ) {
568 $link = "{$protocol}:{$m[1]}";
570 if ( $autonumber ) { $text = "[" . ++
$this->mAutonumber
. "]"; }
571 else { $text = wfEscapeHTML( $link ); }
572 } else if ( preg_match( $e2, $line, $m ) ) {
573 $link = "{$protocol}:{$m[1]}";
577 $s .= "[{$protocol}:" . $line;
580 if( $link == $text ||
preg_match( "!$protocol://" . preg_quote( $text, "/" ) . "/?$!", $link ) ) {
583 # Expand the URL for printable version
584 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $link ) . "</i>)</span>";
586 $la = $sk->getExternalLinkAttributes( $link, $text );
587 $s .= "<a href='{$link}'{$la}>{$text}</a>{$paren}{$trail}";
593 /* private */ function handle3Quotes( &$state, $token )
595 if ( $state["strong"] !== false ) {
596 if ( $state["em"] !== false && $state["em"] > $state["strong"] )
598 # ''' lala ''lala '''
599 $s = "</em></strong><em>";
603 $state["strong"] = FALSE;
606 $state["strong"] = isset($token["pos"]) ?
$token["pos"] : true;
611 /* private */ function handle2Quotes( &$state, $token )
613 if ( $state["em"] !== false ) {
614 if ( $state["strong"] !== false && $state["strong"] > $state["em"] )
616 # ''lala'''lala'' ....'''
617 $s = "</strong></em><strong>";
621 $state["em"] = FALSE;
624 $state["em"] = isset($token["pos"]) ?
$token["pos"] : true;
630 /* private */ function handle5Quotes( &$state, $token )
633 if ( $state["em"] !== false && $state["strong"] !== false ) {
634 if ( $state["em"] < $state["strong"] ) {
635 $s .= "</strong></em>";
637 $s .= "</em></strong>";
639 $state["strong"] = $state["em"] = FALSE;
640 } elseif ( $state["em"] !== false ) {
641 $s .= "</em><strong>";
642 $state["em"] = FALSE;
643 $state["strong"] = $token["pos"];
644 } elseif ( $state["strong"] !== false ) {
645 $s .= "</strong><em>";
646 $state["strong"] = FALSE;
647 $state["em"] = $token["pos"];
648 } else { # not $em and not $strong
649 $s .= "<strong><em>";
650 $state["strong"] = $state["em"] = isset($token["pos"]) ?
$token["pos"] : true;
655 /* private */ function doTokenizedParser( $str )
657 global $wgLang; # for language specific parser hook
659 $tokenizer=Tokenizer
::newFromString( $str );
660 $tokenStack = array();
663 $state["em"] = FALSE;
664 $state["strong"] = FALSE;
668 # The tokenizer splits the text into tokens and returns them one by one.
669 # Every call to the tokenizer returns a new token.
670 while ( $token = $tokenizer->nextToken() )
672 switch ( $token["type"] )
675 # simple text with no further markup
676 $txt = $token["text"];
679 # remember the tag opened with 3 [
683 # FIXME : Treat orphaned open tags (stack not empty when text is over)
685 array_push( $tokenStack, $token );
692 # get text from stack, glue it together, and call the code to handle a
695 if ( count( $tokenStack ) == 0 )
697 # stack empty. Found a ]] without an opening [[
701 $lastToken = array_pop( $tokenStack );
702 while ( !(($lastToken["type"] == "[[[") or ($lastToken["type"] == "[[")) )
704 if( !empty( $lastToken["text"] ) ) {
705 $linkText = $lastToken["text"] . $linkText;
707 $lastToken = array_pop( $tokenStack );
710 $txt = $linkText ."]]";
712 if( isset( $lastToken["text"] ) ) {
713 $prefix = $lastToken["text"];
717 $nextToken = $tokenizer->previewToken();
718 if ( $nextToken["type"] == "text" )
720 # Preview just looks at it. Now we have to fetch it.
721 $nextToken = $tokenizer->nextToken();
722 $txt .= $nextToken["text"];
724 $txt = $this->handleInternalLink( $this->unstrip($txt,$this->mStripState
), $prefix );
726 # did the tag start with 3 [ ?
728 # show the first as text
734 $tagIsOpen = (count( $tokenStack ) != 0);
740 # This and the three next ones handle quotes
741 $txt = $this->handle3Quotes( $state, $token );
744 $txt = $this->handle2Quotes( $state, $token );
747 $txt = $this->handle5Quotes( $state, $token );
757 $txt = $this->doMagicRFC( $tokenizer );
764 $txt = $this->doMagicISBN( $tokenizer );
768 # Call language specific Hook.
769 $txt = $wgLang->processToken( $token, $tokenStack );
770 if ( NULL == $txt ) {
771 # An unkown token. Highlight.
772 $txt = "<font color=\"#FF0000\"><b>".$token["type"]."</b></font>";
773 $txt .= "<font color=\"#FFFF00\"><b>".$token["text"]."</b></font>";
777 # If we're parsing the interior of a link, don't append the interior to $s,
778 # but push it to the stack so it can be processed when a ]] token is found.
779 if ( $tagIsOpen && $txt != "" ) {
780 $token["type"] = "text";
781 $token["text"] = $txt;
782 array_push( $tokenStack, $token );
787 if ( count( $tokenStack ) != 0 )
789 # still objects on stack. opened [[ tag without closing ]] tag.
791 while ( $lastToken = array_pop( $tokenStack ) )
793 if ( $lastToken["type"] == "text" )
795 $txt = $lastToken["text"] . $txt;
797 $txt = $lastToken["type"] . $txt;
805 /* private */ function handleInternalLink( $line, $prefix )
807 global $wgLang, $wgLinkCache;
808 global $wgNamespacesWithSubpages, $wgLanguageCode;
809 static $fname = "Parser::handleInternalLink" ;
810 wfProfileIn( $fname );
812 wfProfileIn( "$fname-setup" );
814 if ( !$tc ) { $tc = Title
::legalChars() . "#"; }
815 $sk =& $this->mOptions
->getSkin();
817 # Match a link having the form [[namespace:link|alternate]]trail
819 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
820 # Match the end of a line for a word that's not followed by whitespace,
821 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
822 #$e2 = "/^(.*)\\b(\\w+)\$/suD";
823 #$e2 = "/^(.*\\s)(\\S+)\$/suD";
824 static $e2 = '/^(.*\s)([a-zA-Z\x80-\xff]+)$/sD';
827 # Special and Media are pseudo-namespaces; no pages actually exist in them
828 static $image = FALSE;
829 static $special = FALSE;
830 static $media = FALSE;
831 static $category = FALSE;
832 if ( !$image ) { $image = Namespace::getImage(); }
833 if ( !$special ) { $special = Namespace::getSpecial(); }
834 if ( !$media ) { $media = Namespace::getMedia(); }
835 if ( !$category ) { $category = Namespace::getCategory(); ; }
837 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
839 wfProfileOut( "$fname-setup" );
842 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
845 } else { # Invalid form; output directly
846 $s .= $prefix . "[[" . $line ;
852 :Foobar -- override special treatment of prefix (images, language links)
853 /Foobar -- convert to CurrentPage/Foobar
854 /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
856 $c = substr($m[1],0,1);
857 $noforce = ($c != ":");
858 if( $c == "/" ) { # subpage
859 if(substr($m[1],-1,1)=="/") { # / at end means we don't want the slash to be shown
860 $m[1]=substr($m[1],1,strlen($m[1])-2);
863 $noslash=substr($m[1],1);
865 if($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]) { # subpages allowed here
866 $link = $this->mTitle
->getPrefixedText(). "/" . trim($noslash);
869 } # this might be changed for ugliness reasons
871 $link = $noslash; # no subpage allowed, use standard link
873 } elseif( $noforce ) { # no subpage
876 $link = substr( $m[1], 1 );
881 $nt = Title
::newFromText( $link );
883 $s .= $prefix . "[[" . $line;
886 $ns = $nt->getNamespace();
887 $iw = $nt->getInterWiki();
889 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
890 array_push( $this->mOutput
->mLanguageLinks
, $nt->getPrefixedText() );
891 return (trim($s) == '')?
'': $s;
893 if( $ns == $image ) {
894 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
895 $wgLinkCache->addImageLinkObj( $nt );
898 if ( $ns == $category ) {
899 $t = $nt->getText() ;
900 $nnt = Title
::newFromText ( Namespace::getCanonicalName($category).":".$t ) ;
901 $t = $sk->makeLinkObj( $nnt, $t, "", "" , $prefix );
902 $this->mOutput
->mCategoryLinks
[] = $t ;
903 $s .= $prefix . $trail ;
907 if( ( $nt->getPrefixedText() == $this->mTitle
->getPrefixedText() ) &&
908 ( strpos( $link, "#" ) == FALSE ) ) {
909 $s .= $prefix . "<strong>" . $text . "</strong>" . $trail;
913 if( $ns == $media ) {
914 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
915 $wgLinkCache->addImageLinkObj( $nt );
917 } elseif( $ns == $special ) {
918 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, "", $trail );
921 $s .= $sk->makeLinkObj( $nt, $text, "", $trail , $prefix );
923 wfProfileOut( $fname );
927 # Some functions here used by doBlockLevels()
929 /* private */ function closeParagraph()
932 if ( '' != $this->mLastSection
) {
933 $result = "</" . $this->mLastSection
. ">\n";
935 $this->mInPre
= false;
936 $this->mLastSection
= "";
939 # getCommon() returns the length of the longest common substring
940 # of both arguments, starting at the beginning of both.
942 /* private */ function getCommon( $st1, $st2 )
944 $fl = strlen( $st1 );
945 $shorter = strlen( $st2 );
946 if ( $fl < $shorter ) { $shorter = $fl; }
948 for ( $i = 0; $i < $shorter; ++
$i ) {
949 if ( $st1{$i} != $st2{$i} ) { break; }
953 # These next three functions open, continue, and close the list
954 # element appropriate to the prefix character passed into them.
956 /* private */ function openList( $char )
958 $result = $this->closeParagraph();
960 if ( "*" == $char ) { $result .= "<ul><li>"; }
961 else if ( "#" == $char ) { $result .= "<ol><li>"; }
962 else if ( ":" == $char ) { $result .= "<dl><dd>"; }
963 else if ( ";" == $char ) {
964 $result .= "<dl><dt>";
965 $this->mDTopen
= true;
967 else { $result = "<!-- ERR 1 -->"; }
972 /* private */ function nextItem( $char )
974 if ( "*" == $char ||
"#" == $char ) { return "</li><li>"; }
975 else if ( ":" == $char ||
";" == $char ) {
977 if ( $this->mDTopen
) { $close = "</dt>"; }
978 if ( ";" == $char ) {
979 $this->mDTopen
= true;
980 return $close . "<dt>";
982 $this->mDTopen
= false;
983 return $close . "<dd>";
986 return "<!-- ERR 2 -->";
989 /* private */function closeList( $char )
991 if ( "*" == $char ) { $text = "</li></ul>"; }
992 else if ( "#" == $char ) { $text = "</li></ol>"; }
993 else if ( ":" == $char ) {
994 if ( $this->mDTopen
) {
995 $this->mDTopen
= false;
996 $text = "</dt></dl>";
998 $text = "</dd></dl>";
1001 else { return "<!-- ERR 3 -->"; }
1005 /* private */ function doBlockLevels( $text, $linestart )
1007 $fname = "Parser::doBlockLevels";
1008 wfProfileIn( $fname );
1009 # Parsing through the text line by line. The main thing
1010 # happening here is handling of block-level elements p, pre,
1011 # and making lists from lines starting with * # : etc.
1013 $a = explode( "\n", $text );
1015 $lastPref = $text = $lastLine = '';
1016 $this->mDTopen
= $inBlockElem = false;
1020 if ( ! $linestart ) { $text .= array_shift( $a ); }
1021 foreach ( $a as $t ) {
1023 $opl = strlen( $lastPref );
1024 $preCloseMatch = preg_match("/<\\/pre/i", $t );
1025 $preOpenMatch = preg_match("/<pre/i", $t );
1026 if (!$this->mInPre
) {
1027 $this->mInPre
= !empty($preOpenMatch);
1029 if ( !$this->mInPre
) {
1030 $npl = strspn( $t, "*#:;" );
1031 $pref = substr( $t, 0, $npl );
1032 $pref2 = str_replace( ";", ":", $pref );
1033 $t = substr( $t, $npl );
1036 $pref = $pref2 = '';
1040 if ( 0 != $npl && 0 == strcmp( $lastPref, $pref2 ) ) {
1041 $text .= $this->nextItem( substr( $pref, -1 ) );
1042 if ( $pstack ) { $pstack = false; }
1044 if ( ";" == substr( $pref, -1 ) ) {
1045 $cpos = strpos( $t, ":" );
1046 if ( false !== $cpos ) {
1047 $term = substr( $t, 0, $cpos );
1048 $text .= $term . $this->nextItem( ":" );
1049 $t = substr( $t, $cpos +
1 );
1052 } else if (0 != $npl ||
0 != $opl) {
1053 $cpl = $this->getCommon( $pref, $lastPref );
1054 if ( $pstack ) { $pstack = false; }
1056 while ( $cpl < $opl ) {
1057 $text .= $this->closeList( $lastPref{$opl-1} );
1060 if ( $npl <= $cpl && $cpl > 0 ) {
1061 $text .= $this->nextItem( $pref{$cpl-1} );
1063 while ( $npl > $cpl ) {
1064 $char = substr( $pref, $cpl, 1 );
1065 $text .= $this->openList( $char );
1067 if ( ";" == $char ) {
1068 $cpos = strpos( $t, ":" );
1069 if ( ! ( false === $cpos ) ) {
1070 $term = substr( $t, 0, $cpos );
1071 $text .= $term . $this->nextItem( ":" );
1072 $t = substr( $t, $cpos +
1 );
1079 if ( 0 == $npl ) { # No prefix (not in list)--go to paragraph mode
1080 $uniq_prefix = UNIQ_PREFIX
;
1081 // XXX: use a stack for nestable elements like span, table and div
1082 $openmatch = preg_match("/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<div|<pre|<tr|<td|<p|<ul|<li)/i", $t );
1083 $closematch = preg_match(
1084 "/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|".
1085 "<\\/div|<hr|<\\/td|<\\/pre|<\\/p|".$uniq_prefix."-pre|<\\/li|<\\/ul)/i", $t );
1086 if ( $openmatch or $closematch ) {
1087 if ( $pstack ) { $pstack = false; }
1088 $text .= $this->closeParagraph();
1089 if($preOpenMatch and !$preCloseMatch) {
1090 $this->mInPre
= true;
1092 if ( $closematch ) {
1093 $inBlockElem = false;
1095 $inBlockElem = true;
1097 } else if ( !$inBlockElem ) {
1098 if ( " " == $t{0} ) {
1100 if ($this->mLastSection
!= 'pre') {
1102 $text .= $this->closeParagraph().'<pre>';
1103 $this->mLastSection
= 'pre';
1107 if ( '' == trim($t) ) {
1109 $text .= $pstack.'<br/>';
1111 $this->mLastSection
= 'p';
1113 if ($this->mLastSection
!= 'p' ) {
1114 $text .= $this->closeParagraph();
1115 $this->mLastSection
= '';
1118 $pstack = '</p><p>';
1125 $this->mLastSection
= 'p';
1126 } else if ($this->mLastSection
!= 'p') {
1127 $text .= $this->closeParagraph().'<p>';
1128 $this->mLastSection
= 'p';
1134 if ($pstack === false) {
1139 $text .= $this->closeList( $pref2{$npl-1} );
1142 if ( "" != $this->mLastSection
) {
1143 $text .= "</" . $this->mLastSection
. ">";
1144 $this->mLastSection
= "";
1147 wfProfileOut( $fname );
1151 function getVariableValue( $index ) {
1152 global $wgLang, $wgSitename, $wgServer;
1155 case MAG_CURRENTMONTH
:
1157 case MAG_CURRENTMONTHNAME
:
1158 return $wgLang->getMonthName( date("n") );
1159 case MAG_CURRENTMONTHNAMEGEN
:
1160 return $wgLang->getMonthNameGen( date("n") );
1161 case MAG_CURRENTDAY
:
1163 case MAG_CURRENTDAYNAME
:
1164 return $wgLang->getWeekdayName( date("w")+
1 );
1165 case MAG_CURRENTYEAR
:
1167 case MAG_CURRENTTIME
:
1168 return $wgLang->time( wfTimestampNow(), false );
1169 case MAG_NUMBEROFARTICLES
:
1170 return wfNumberOfArticles();
1180 function initialiseVariables()
1182 global $wgVariableIDs;
1183 $this->mVariables
= array();
1184 foreach ( $wgVariableIDs as $id ) {
1185 $mw =& MagicWord
::get( $id );
1186 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1190 /* private */ function replaceVariables( $text, $args = array() )
1192 global $wgLang, $wgScript, $wgArticlePath;
1194 $fname = "Parser::replaceVariables";
1195 wfProfileIn( $fname );
1198 if ( !$this->mVariables
) {
1199 $this->initialiseVariables();
1201 $titleChars = Title
::legalChars();
1202 $regex = "/(\\n?){{([$titleChars]*?)(\\|.*?|)}}/s";
1204 # This function is called recursively. To keep track of arguments we need a stack:
1205 array_push( $this->mArgStack
, $args );
1207 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1208 $GLOBALS['wgCurParser'] =& $this;
1209 $text = preg_replace_callback( $regex, "wfBraceSubstitution", $text );
1211 array_pop( $this->mArgStack
);
1216 function braceSubstitution( $matches )
1218 global $wgLinkCache, $wgLang;
1219 $fname = "Parser::braceSubstitution";
1224 # $newline is an optional newline character before the braces
1225 # $part1 is the bit before the first |, and must contain only title characters
1226 # $args is a list of arguments, starting from index 0, not including $part1
1228 $newline = $matches[1];
1229 $part1 = $matches[2];
1230 # If the third subpattern matched anything, it will start with |
1231 if ( $matches[3] !== "" ) {
1232 $args = explode( "|", substr( $matches[3], 1 ) );
1236 $argc = count( $args );
1239 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1240 if ( $mwSubst->matchStartAndRemove( $part1 ) ) {
1241 if ( $this->mOutputType
!= OT_WIKI
) {
1242 # Invalid SUBST not replaced at PST time
1243 # Return without further processing
1244 $text = $matches[0];
1247 } elseif ( $this->mOutputType
== OT_WIKI
) {
1248 # SUBST not found in PST pass, do nothing
1249 $text = $matches[0];
1253 # MSG, MSGNW and INT
1256 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1257 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1260 # Remove obsolete MSG:
1261 $mwMsg =& MagicWord
::get( MAG_MSG
);
1262 $mwMsg->matchStartAndRemove( $part1 );
1265 # Check if it is an internal message
1266 $mwInt =& MagicWord
::get( MAG_INT
);
1267 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1268 if ( $this->incrementIncludeCount( "int:$part1" ) ) {
1269 $text = wfMsgReal( $part1, $args, true );
1277 # Check for NS: (namespace expansion)
1278 $mwNs = MagicWord
::get( MAG_NS
);
1279 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1280 if ( intval( $part1 ) ) {
1281 $text = $wgLang->getNsText( intval( $part1 ) );
1284 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1285 if ( !is_null( $index ) ) {
1286 $text = $wgLang->getNsText( $index );
1293 # LOCALURL and LOCALURLE
1295 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1296 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1298 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1299 $func = 'getLocalURL';
1300 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1301 $func = 'escapeLocalURL';
1306 if ( $func !== '' ) {
1307 $title = Title
::newFromText( $part1 );
1308 if ( !is_null( $title ) ) {
1310 $text = $title->$func( $args[0] );
1312 $text = $title->$func();
1319 # Internal variables
1320 if ( !$found && array_key_exists( $part1, $this->mVariables
) ) {
1321 $text = $this->mVariables
[$part1];
1323 $this->mOutput
->mContainsOldMagic
= true;
1326 # Arguments input from the caller
1327 $inputArgs = end( $this->mArgStack
);
1328 if ( !$found && array_key_exists( $part1, $inputArgs ) ) {
1329 $text = $inputArgs[$part1];
1333 # Load from database
1335 $title = Title
::newFromText( $part1, NS_TEMPLATE
);
1336 if ( !is_null( $title ) && !$title->isExternal() ) {
1337 # Check for excessive inclusion
1338 $dbk = $title->getPrefixedDBkey();
1339 if ( $this->incrementIncludeCount( $dbk ) ) {
1340 $article = new Article( $title );
1341 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1342 if ( $articleContent !== false ) {
1344 $text = $articleContent;
1349 # If the title is valid but undisplayable, make a link to it
1350 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1351 $text = "[[" . $title->getPrefixedText() . "]]";
1357 # Recursive parsing, escaping and link table handling
1358 # Only for HTML output
1359 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1360 $text = wfEscapeWikiText( $text );
1361 } elseif ( $this->mOutputType
== OT_HTML
&& $found ) {
1362 # Clean up argument array
1363 $assocArgs = array();
1365 foreach( $args as $arg ) {
1366 $eqpos = strpos( $arg, "=" );
1367 if ( $eqpos === false ) {
1368 $assocArgs[$index++
] = $arg;
1370 $name = trim( substr( $arg, 0, $eqpos ) );
1371 $value = trim( substr( $arg, $eqpos+
1 ) );
1372 if ( $value === false ) {
1375 if ( $name !== false ) {
1376 $assocArgs[$name] = $value;
1381 # Do not enter included links in link table
1382 if ( !is_null( $title ) ) {
1383 $wgLinkCache->suspend();
1386 # Run full parser on the included text
1387 $text = $this->strip( $text, $this->mStripState
);
1388 $text = $this->internalParse( $text, (bool)$newline, $assocArgs );
1390 # Add the result to the strip state for re-inclusion after
1391 # the rest of the processing
1392 $text = $this->insertStripItem( $text, $this->mStripState
);
1394 # Resume the link cache and register the inclusion as a link
1395 if ( !is_null( $title ) ) {
1396 $wgLinkCache->resume();
1397 $wgLinkCache->addLinkObj( $title );
1404 return $newline . $text;
1408 # Returns true if the function is allowed to include this entity
1409 function incrementIncludeCount( $dbk )
1411 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1412 $this->mIncludeCount
[$dbk] = 0;
1414 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1422 # Cleans up HTML, removes dangerous tags and attributes
1423 /* private */ function removeHTMLtags( $text )
1425 $fname = "Parser::removeHTMLtags";
1426 wfProfileIn( $fname );
1427 $htmlpairs = array( # Tags that must be closed
1428 "b", "del", "i", "ins", "u", "font", "big", "small", "sub", "sup", "h1",
1429 "h2", "h3", "h4", "h5", "h6", "cite", "code", "em", "s",
1430 "strike", "strong", "tt", "var", "div", "center",
1431 "blockquote", "ol", "ul", "dl", "table", "caption", "pre",
1432 "ruby", "rt" , "rb" , "rp", "p"
1434 $htmlsingle = array(
1435 "br", "hr", "li", "dt", "dd"
1437 $htmlnest = array( # Tags that can be nested--??
1438 "table", "tr", "td", "th", "div", "blockquote", "ol", "ul",
1439 "dl", "font", "big", "small", "sub", "sup"
1441 $tabletags = array( # Can only appear inside table
1445 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1446 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1448 $htmlattrs = $this->getHTMLattrs () ;
1450 # Remove HTML comments
1451 $text = preg_replace( "/<!--.*-->/sU", "", $text );
1453 $bits = explode( "<", $text );
1454 $text = array_shift( $bits );
1455 $tagstack = array(); $tablestack = array();
1457 foreach ( $bits as $x ) {
1458 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
1459 preg_match( "/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/",
1461 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1462 error_reporting( $prev );
1465 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1469 if ( ! in_array( $t, $htmlsingle ) &&
1470 ( $ot = array_pop( $tagstack ) ) != $t ) {
1471 array_push( $tagstack, $ot );
1474 if ( $t == "table" ) {
1475 $tagstack = array_pop( $tablestack );
1480 # Keep track for later
1481 if ( in_array( $t, $tabletags ) &&
1482 ! in_array( "table", $tagstack ) ) {
1484 } else if ( in_array( $t, $tagstack ) &&
1485 ! in_array ( $t , $htmlnest ) ) {
1487 } else if ( ! in_array( $t, $htmlsingle ) ) {
1488 if ( $t == "table" ) {
1489 array_push( $tablestack, $tagstack );
1490 $tagstack = array();
1492 array_push( $tagstack, $t );
1494 # Strip non-approved attributes from the tag
1495 $newparams = $this->fixTagAttributes($params);
1499 $rest = str_replace( ">", ">", $rest );
1500 $text .= "<$slash$t $newparams$brace$rest";
1504 $text .= "<" . str_replace( ">", ">", $x);
1506 # Close off any remaining tags
1507 while ( $t = array_pop( $tagstack ) ) {
1509 if ( $t == "table" ) { $tagstack = array_pop( $tablestack ); }
1511 wfProfileOut( $fname );
1517 * This function accomplishes several tasks:
1518 * 1) Auto-number headings if that option is enabled
1519 * 2) Add an [edit] link to sections for logged in users who have enabled the option
1520 * 3) Add a Table of contents on the top for users who have enabled the option
1521 * 4) Auto-anchor headings
1523 * It loops through all headlines, collects the necessary data, then splits up the
1524 * string and re-inserts the newly formatted headlines.
1528 /* private */ function formatHeadings( $text )
1530 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
1531 $doShowToc = $this->mOptions
->getShowToc();
1532 if( !$this->mTitle
->userCanEdit() ) {
1534 $rightClickHack = 0;
1536 $showEditLink = $this->mOptions
->getEditSection();
1537 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
1540 # Inhibit editsection links if requested in the page
1541 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
1542 if( $esw->matchAndRemove( $text ) ) {
1545 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1547 $mw =& MagicWord
::get( MAG_NOTOC
);
1548 if( $mw->matchAndRemove( $text ) ) {
1552 # never add the TOC to the Main Page. This is an entry page that should not
1553 # be more than 1-2 screens large anyway
1554 if( $this->mTitle
->getPrefixedText() == wfMsg("mainpage") ) {
1558 # Get all headlines for numbering them and adding funky stuff like [edit]
1559 # links - this is for later, but we need the number of headlines right now
1560 $numMatches = preg_match_all( "/<H([1-6])(.*?" . ">)(.*?)<\/H[1-6]>/i", $text, $matches );
1562 # if there are fewer than 4 headlines in the article, do not show TOC
1563 if( $numMatches < 4 ) {
1567 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
1568 # override above conditions and always show TOC
1569 $mw =& MagicWord
::get( MAG_FORCETOC
);
1570 if ($mw->matchAndRemove( $text ) ) {
1575 # We need this to perform operations on the HTML
1576 $sk =& $this->mOptions
->getSkin();
1581 # Ugh .. the TOC should have neat indentation levels which can be
1582 # passed to the skin functions. These are determined here
1587 $sublevelCount = array();
1590 foreach( $matches[3] as $headline ) {
1593 $prevlevel = $level;
1595 $level = $matches[1][$headlineCount];
1596 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
1597 # reset when we enter a new level
1598 $sublevelCount[$level] = 0;
1599 $toc .= $sk->tocIndent( $level - $prevlevel );
1600 $toclevel +
= $level - $prevlevel;
1602 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
1603 # reset when we step back a level
1604 $sublevelCount[$level+
1]=0;
1605 $toc .= $sk->tocUnindent( $prevlevel - $level );
1606 $toclevel -= $prevlevel - $level;
1608 # count number of headlines for each level
1609 @$sublevelCount[$level]++
;
1610 if( $doNumberHeadings ||
$doShowToc ) {
1612 for( $i = 1; $i <= $level; $i++
) {
1613 if( !empty( $sublevelCount[$i] ) ) {
1617 $numbering .= $sublevelCount[$i];
1623 # The canonized header is a version of the header text safe to use for links
1624 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
1625 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
1628 $canonized_headline = preg_replace( "/<.*?" . ">/","",$canonized_headline );
1629 $tocline = trim( $canonized_headline );
1630 $canonized_headline = preg_replace("/[ \\?&\\/<>\\(\\)\\[\\]=,+']+/", '_', html_entity_decode( $tocline));
1631 $refer[$headlineCount] = $canonized_headline;
1633 # count how many in assoc. array so we can track dupes in anchors
1634 @$refers[$canonized_headline]++
;
1635 $refcount[$headlineCount]=$refers[$canonized_headline];
1637 # Prepend the number to the heading text
1639 if( $doNumberHeadings ||
$doShowToc ) {
1640 $tocline = $numbering . " " . $tocline;
1642 # Don't number the heading if it is the only one (looks silly)
1643 if( $doNumberHeadings && count( $matches[3] ) > 1) {
1644 # the two are different if the line contains a link
1645 $headline=$numbering . " " . $headline;
1649 # Create the anchor for linking from the TOC to the section
1650 $anchor = $canonized_headline;
1651 if($refcount[$headlineCount] > 1 ) {
1652 $anchor .= "_" . $refcount[$headlineCount];
1655 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
1657 if( $showEditLink ) {
1658 if ( empty( $head[$headlineCount] ) ) {
1659 $head[$headlineCount] = "";
1661 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+
1);
1664 # Add the edit section span
1665 if( $rightClickHack ) {
1666 $headline = $sk->editSectionScript($headlineCount+
1,$headline);
1669 # give headline the correct <h#> tag
1670 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline."</h".$level.">";
1676 $toclines = $headlineCount;
1677 $toc .= $sk->tocUnindent( $toclevel );
1678 $toc = $sk->tocTable( $toc );
1681 # split up and insert constructed headlines
1683 $blocks = preg_split( "/<H[1-6].*?" . ">.*?<\/H[1-6]>/i", $text );
1686 foreach( $blocks as $block ) {
1687 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
1688 # This is the [edit] link that appears for the top block of text when
1689 # section editing is enabled
1691 # Disabled because it broke block formatting
1692 # For example, a bullet point in the top line
1693 # $full .= $sk->editSectionLink(0);
1696 if( $doShowToc && !$i) {
1697 # Top anchor now in skin
1701 if( !empty( $head[$i] ) ) {
1710 /* private */ function doMagicISBN( &$tokenizer )
1714 # Check whether next token is a text token
1715 # If yes, fetch it and convert the text into a
1716 # Special::BookSources link
1717 $token = $tokenizer->previewToken();
1718 while ( $token["type"] == "" )
1720 $tokenizer->nextToken();
1721 $token = $tokenizer->previewToken();
1723 if ( $token["type"] == "text" )
1725 $token = $tokenizer->nextToken();
1726 $x = $token["text"];
1727 $valid = "0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ";
1729 $isbn = $blank = "" ;
1730 while ( " " == $x{0} ) {
1732 $x = substr( $x, 1 );
1734 while ( strstr( $valid, $x{0} ) != false ) {
1736 $x = substr( $x, 1 );
1738 $num = str_replace( "-", "", $isbn );
1739 $num = str_replace( " ", "", $num );
1742 $text = "ISBN $blank$x";
1744 $titleObj = Title
::makeTitle( NS_SPECIAL
, "Booksources" );
1745 $text = "<a href=\"" .
1746 $titleObj->escapeLocalUrl( "isbn={$num}" ) .
1747 "\" class=\"internal\">ISBN $isbn</a>";
1755 /* private */ function doMagicRFC( &$tokenizer )
1759 # Check whether next token is a text token
1760 # If yes, fetch it and convert the text into a
1761 # link to an RFC source
1762 $token = $tokenizer->previewToken();
1763 while ( $token["type"] == "" )
1765 $tokenizer->nextToken();
1766 $token = $tokenizer->previewToken();
1768 if ( $token["type"] == "text" )
1770 $token = $tokenizer->nextToken();
1771 $x = $token["text"];
1772 $valid = "0123456789";
1774 $rfc = $blank = "" ;
1775 while ( " " == $x{0} ) {
1777 $x = substr( $x, 1 );
1779 while ( strstr( $valid, $x{0} ) != false ) {
1781 $x = substr( $x, 1 );
1785 $text .= "RFC $blank$x";
1787 $url = wfmsg( "rfcurl" );
1788 $url = str_replace( "$1", $rfc, $url);
1789 $sk =& $this->mOptions
->getSkin();
1790 $la = $sk->getExternalLinkAttributes( $url, "RFC {$rfc}" );
1791 $text = "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
1799 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true )
1801 $this->mOptions
= $options;
1802 $this->mTitle
=& $title;
1803 $this->mOutputType
= OT_WIKI
;
1805 if ( $clearState ) {
1806 $this->clearState();
1809 $stripState = false;
1813 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
1816 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
1817 "/<br *?>/i" => "<br/>",
1819 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
1820 $text = $this->strip( $text, $stripState, false );
1821 $text = $this->pstPass2( $text, $user );
1822 $text = $this->unstrip( $text, $stripState );
1826 /* private */ function pstPass2( $text, &$user )
1828 global $wgLang, $wgLocaltimezone, $wgCurParser;
1830 # Variable replacement
1831 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
1832 $text = $this->replaceVariables( $text );
1836 $n = $user->getName();
1837 $k = $user->getOption( "nickname" );
1838 if ( "" == $k ) { $k = $n; }
1839 if(isset($wgLocaltimezone)) {
1840 $oldtz = getenv("TZ"); putenv("TZ=$wgLocaltimezone");
1842 /* Note: this is an ugly timezone hack for the European wikis */
1843 $d = $wgLang->timeanddate( date( "YmdHis" ), false ) .
1844 " (" . date( "T" ) . ")";
1845 if(isset($wgLocaltimezone)) putenv("TZ=$oldtz");
1847 $text = preg_replace( "/~~~~~/", $d, $text );
1848 $text = preg_replace( "/~~~~/", "[[" . $wgLang->getNsText(
1849 Namespace::getUser() ) . ":$n|$k]] $d", $text );
1850 $text = preg_replace( "/~~~/", "[[" . $wgLang->getNsText(
1851 Namespace::getUser() ) . ":$n|$k]]", $text );
1853 # Context links: [[|name]] and [[name (context)|]]
1855 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
1856 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
1857 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
1858 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
1860 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
1861 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
1862 $p3 = "/\[\[($namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]]
1863 $p4 = "/\[\[($namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/";
1864 # [[ns:page (cont)|]]
1866 $t = $this->mTitle
->getText();
1867 if ( preg_match( $conpat, $t, $m ) ) {
1870 $text = preg_replace( $p4, "[[\\1:\\2 (\\3)|\\2]]", $text );
1871 $text = preg_replace( $p1, "[[\\1 (\\2)|\\1]]", $text );
1872 $text = preg_replace( $p3, "[[\\1:\\2|\\2]]", $text );
1874 if ( "" == $context ) {
1875 $text = preg_replace( $p2, "[[\\1]]", $text );
1877 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
1881 $mw =& MagicWord::get( MAG_SUBST );
1882 $wgCurParser = $this->fork();
1883 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
1884 $this->merge( $wgCurParser );
1887 # Trim trailing whitespace
1888 # MAG_END (__END__) tag allows for trailing
1889 # whitespace to be deliberately included
1890 $text = rtrim( $text );
1891 $mw =& MagicWord
::get( MAG_END
);
1892 $mw->matchAndRemove( $text );
1897 # Set up some variables which are usually set up in parse()
1898 # so that an external function can call some class members with confidence
1899 function startExternalParse( &$title, $options, $outputType, $clearState = true )
1901 $this->mTitle
=& $title;
1902 $this->mOptions
= $options;
1903 $this->mOutputType
= $outputType;
1904 if ( $clearState ) {
1905 $this->clearState();
1909 function transformMsg( $text, $options ) {
1911 static $executing = false;
1913 # Guard against infinite recursion
1919 $this->mTitle
= $wgTitle;
1920 $this->mOptions
= $options;
1921 $this->mOutputType
= OT_MSG
;
1922 $this->clearState();
1923 $text = $this->replaceVariables( $text );
1932 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
1934 function ParserOutput( $text = "", $languageLinks = array(), $categoryLinks = array(),
1935 $containsOldMagic = false )
1937 $this->mText
= $text;
1938 $this->mLanguageLinks
= $languageLinks;
1939 $this->mCategoryLinks
= $categoryLinks;
1940 $this->mContainsOldMagic
= $containsOldMagic;
1943 function getText() { return $this->mText
; }
1944 function getLanguageLinks() { return $this->mLanguageLinks
; }
1945 function getCategoryLinks() { return $this->mCategoryLinks
; }
1946 function containsOldMagic() { return $this->mContainsOldMagic
; }
1947 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
1948 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
1949 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
1950 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
1952 function merge( $other ) {
1953 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
1954 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
1955 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
1962 # All variables are private
1963 var $mUseTeX; # Use texvc to expand <math> tags
1964 var $mUseCategoryMagic; # Treat [[Category:xxxx]] tags specially
1965 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
1966 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
1967 var $mAllowExternalImages; # Allow external images inline
1968 var $mSkin; # Reference to the preferred skin
1969 var $mDateFormat; # Date format index
1970 var $mEditSection; # Create "edit section" links
1971 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
1972 var $mNumberHeadings; # Automatically number headings
1973 var $mShowToc; # Show table of contents
1975 function getUseTeX() { return $this->mUseTeX
; }
1976 function getUseCategoryMagic() { return $this->mUseCategoryMagic
; }
1977 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
1978 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
1979 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
1980 function getSkin() { return $this->mSkin
; }
1981 function getDateFormat() { return $this->mDateFormat
; }
1982 function getEditSection() { return $this->mEditSection
; }
1983 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
1984 function getNumberHeadings() { return $this->mNumberHeadings
; }
1985 function getShowToc() { return $this->mShowToc
; }
1987 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
1988 function setUseCategoryMagic( $x ) { return wfSetVar( $this->mUseCategoryMagic
, $x ); }
1989 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
1990 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
1991 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
1992 function setSkin( $x ) { return wfSetRef( $this->mSkin
, $x ); }
1993 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
1994 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
1995 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
1996 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
1997 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
1999 /* static */ function newFromUser( &$user )
2001 $popts = new ParserOptions
;
2002 $popts->initialiseFromUser( $user );
2006 function initialiseFromUser( &$userInput )
2008 global $wgUseTeX, $wgUseCategoryMagic, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2010 if ( !$userInput ) {
2012 $user->setLoaded( true );
2014 $user =& $userInput;
2017 $this->mUseTeX
= $wgUseTeX;
2018 $this->mUseCategoryMagic
= $wgUseCategoryMagic;
2019 $this->mUseDynamicDates
= $wgUseDynamicDates;
2020 $this->mInterwikiMagic
= $wgInterwikiMagic;
2021 $this->mAllowExternalImages
= $wgAllowExternalImages;
2022 $this->mSkin
=& $user->getSkin();
2023 $this->mDateFormat
= $user->getOption( "date" );
2024 $this->mEditSection
= $user->getOption( "editsection" );
2025 $this->mEditSectionOnRightClick
= $user->getOption( "editsectiononrightclick" );
2026 $this->mNumberHeadings
= $user->getOption( "numberheadings" );
2027 $this->mShowToc
= $user->getOption( "showtoc" );
2033 # Regex callbacks, used in Parser::replaceVariables
2034 function wfBraceSubstitution( $matches )
2036 global $wgCurParser;
2037 return $wgCurParser->braceSubstitution( $matches );