3 require_once('Tokenizer.php');
5 if( $GLOBALS['wgUseWikiHiero'] ){
6 require_once('extensions/wikihiero/wikihiero.php');
8 if( $GLOBALS['wgUseTimeline'] ){
9 require_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 )
90 $fname = "Parser::parse";
91 wfProfileIn( $fname );
97 $this->mOptions
= $options;
98 $this->mTitle
=& $title;
99 $this->mOutputType
= OT_HTML
;
102 $text = $this->strip( $text, $this->mStripState
);
103 $text = $this->internalParse( $text, $linestart );
104 $text = $this->unstrip( $text, $this->mStripState
);
105 # Clean up special characters, only run once, next-to-last before doBlockLevels
108 "/<hr *>/i" => '<hr/>',
109 "/<br *>/i" => '<br/>',
110 "/<center *>/i"=>'<div class="center">',
111 "/<\\/center *>/i" => '</div>',
112 # Clean up spare ampersands; note that we probably ought to be
113 # more careful about named entities.
114 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
116 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
119 "/<center *>/i"=>'<div class="center">',
120 "/<\\/center *>/i" => '</div>'
122 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
125 $text = $this->doBlockLevels( $text, $linestart );
127 $text = $this->tidy($text);
129 $this->mOutput
->setText( $text );
130 wfProfileOut( $fname );
131 return $this->mOutput
;
134 /* static */ function getRandomString()
136 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
139 # Replaces all occurrences of <$tag>content</$tag> in the text
140 # with a random marker and returns the new text. the output parameter
141 # $content will be an associative array filled with data on the form
142 # $unique_marker => content.
144 # If $content is already set, the additional entries will be appended
146 # If $tag is set to STRIP_COMMENTS, the function will extract
147 # <!-- HTML comments -->
149 /* static */ function extractTags($tag, $text, &$content, $uniq_prefix = ""){
150 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
157 while ( "" != $text ) {
158 if($tag==STRIP_COMMENTS
) {
159 $p = preg_split( "/<!--/i", $text, 2 );
161 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
164 if ( ( count( $p ) < 2 ) ||
( "" == $p[1] ) ) {
167 if($tag==STRIP_COMMENTS
) {
168 $q = preg_split( "/-->/i", $p[1], 2 );
170 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
172 $marker = $rnd . sprintf("%08X", $n++
);
173 $content[$marker] = $q[0];
174 $stripped .= $marker;
181 # Strips and renders <nowiki>, <pre>, <math>, <hiero>
182 # If $render is set, performs necessary rendering operations on plugins
183 # Returns the text, and fills an array with data needed in unstrip()
184 # If the $state is already a valid strip state, it adds to the state
186 # When $stripcomments is set, HTML comments <!-- like this -->
187 # will be stripped in addition to other tags. This is important
188 # for section editing, where these comments cause confusion when
189 # counting the sections in the wikisource
190 function strip( $text, &$state, $stripcomments = false )
192 $render = ($this->mOutputType
== OT_HTML
);
193 $nowiki_content = array();
194 $hiero_content = array();
195 $math_content = array();
196 $pre_content = array();
197 $comment_content = array();
199 # Replace any instances of the placeholders
200 $uniq_prefix = UNIQ_PREFIX
;
201 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
203 $text = Parser
::extractTags("nowiki", $text, $nowiki_content, $uniq_prefix);
204 foreach( $nowiki_content as $marker => $content ){
206 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
208 $nowiki_content[$marker] = "<nowiki>$content</nowiki>";
212 $text = Parser
::extractTags("hiero", $text, $hiero_content, $uniq_prefix);
213 foreach( $hiero_content as $marker => $content ){
214 if( $render && $GLOBALS['wgUseWikiHiero']){
215 $hiero_content[$marker] = WikiHiero( $content, WH_MODE_HTML
);
217 $hiero_content[$marker] = "<hiero>$content</hiero>";
221 $text = Parser
::extractTags("math", $text, $math_content, $uniq_prefix);
222 foreach( $math_content as $marker => $content ){
224 if( $this->mOptions
->getUseTeX() ) {
225 $math_content[$marker] = renderMath( $content );
227 $math_content[$marker] = "<math>$content<math>";
230 $math_content[$marker] = "<math>$content</math>";
234 $text = Parser
::extractTags("pre", $text, $pre_content, $uniq_prefix);
235 foreach( $pre_content as $marker => $content ){
237 $pre_content[$marker] = "<pre>" . wfEscapeHTMLTagsOnly( $content ) . "</pre>";
239 $pre_content[$marker] = "<pre>$content</pre>";
243 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
244 foreach( $comment_content as $marker => $content ){
245 $comment_content[$marker] = "<!--$content-->";
249 # Merge state with the pre-existing state, if there is one
251 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
252 $state['hiero'] = $state['hiero'] +
$hiero_content;
253 $state['math'] = $state['math'] +
$math_content;
254 $state['pre'] = $state['pre'] +
$pre_content;
255 $state['comment'] = $state['comment'] +
$comment_content;
258 'nowiki' => $nowiki_content,
259 'hiero' => $hiero_content,
260 'math' => $math_content,
261 'pre' => $pre_content,
262 'comment' => $comment_content
268 function unstrip( $text, &$state )
270 # Must expand in reverse order, otherwise nested tags will be corrupted
271 $contentDict = end( $state );
272 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
273 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
274 $text = str_replace( key( $contentDict ), $content, $text );
281 # Add an item to the strip state
282 # Returns the unique tag which must be inserted into the stripped text
283 # The tag will be replaced with the original text in unstrip()
285 function insertStripItem( $text, &$state )
287 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
296 $state['item'][$rnd] = $text;
300 # This method generates the list of subcategories and pages for a category
301 function categoryMagic ()
303 global $wgLang , $wgUser ;
304 if ( !$this->mOptions
->getUseCategoryMagic() ) return ; # Doesn't use categories at all
306 $cns = Namespace::getCategory() ;
307 if ( $this->mTitle
->getNamespace() != $cns ) return "" ; # This ain't a category page
309 $r = "<br style=\"clear:both;\"/>\n";
312 $sk =& $wgUser->getSkin() ;
314 $articles = array() ;
315 $children = array() ;
317 $id = $this->mTitle
->getArticleID() ;
319 # For existing categories
321 $sql = "SELECT DISTINCT cur_title,cur_namespace FROM cur,links WHERE l_to={$id} AND l_from=cur_id";
322 $res = wfQuery ( $sql, DB_READ
) ;
323 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
325 # For non-existing categories
326 $t = wfStrencode( $this->mTitle
->getPrefixedDBKey() );
327 $sql = "SELECT DISTINCT cur_title,cur_namespace FROM cur,brokenlinks WHERE bl_to='$t' AND bl_from=cur_id" ;
328 $res = wfQuery ( $sql, DB_READ
) ;
329 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
332 # For all pages that link to this category
333 foreach ( $data AS $x )
335 $t = $wgLang->getNsText ( $x->cur_namespace
) ;
336 if ( $t != "" ) $t .= ":" ;
337 $t .= $x->cur_title
;
339 if ( $x->cur_namespace
== $cns ) {
340 array_push ( $children , $sk->makeLink ( $t ) ) ; # Subcategory
342 array_push ( $articles , $sk->makeLink ( $t ) ) ; # Page in this category
345 wfFreeResult ( $res ) ;
347 # Showing subcategories
348 if ( count ( $children ) > 0 )
350 asort ( $children ) ;
351 $r .= "<h2>".wfMsg("subcategories")."</h2>\n" ;
352 $r .= implode ( ", " , $children ) ;
355 # Showing pages in this category
356 if ( count ( $articles ) > 0 )
358 $ti = $this->mTitle
->getText() ;
359 asort ( $articles ) ;
360 $h = wfMsg( "category_header", $ti );
361 $r .= "<h2>{$h}</h2>\n" ;
362 $r .= implode ( ", " , $articles ) ;
369 function getHTMLattrs ()
371 $htmlattrs = array( # Allowed attributes--no scripting, etc.
372 "title", "align", "lang", "dir", "width", "height",
373 "bgcolor", "clear", /* BR */ "noshade", /* HR */
374 "cite", /* BLOCKQUOTE, Q */ "size", "face", "color",
375 /* FONT */ "type", "start", "value", "compact",
376 /* For various lists, mostly deprecated but safe */
377 "summary", "width", "border", "frame", "rules",
378 "cellspacing", "cellpadding", "valign", "char",
379 "charoff", "colgroup", "col", "span", "abbr", "axis",
380 "headers", "scope", "rowspan", "colspan", /* Tables */
381 "id", "class", "name", "style" /* For CSS */
386 function fixTagAttributes ( $t )
388 if ( trim ( $t ) == "" ) return "" ; # Saves runtime ;-)
389 $htmlattrs = $this->getHTMLattrs() ;
391 # Strip non-approved attributes from the tag
393 "/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e",
394 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
396 # Strip javascript "expression" from stylesheets. Brute force approach:
397 # If anythin offensive is found, all attributes of the HTML tag are dropped
400 "/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is",
401 wfMungeToUtf8( $t ) ) )
409 /* interface with html tidy, used if $wgUseTidy = true */
410 function tidy ( $text ) {
411 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
412 global $wgInputEncoding, $wgOutputEncoding;
414 switch(strtoupper($wgOutputEncoding)) {
416 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)?
' -latin1':' -raw';
419 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)?
' -utf8':' -raw';
422 $wgTidyOpts .= ' -raw';
425 $text = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
426 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
427 '<head><title>test</title></head><body>'.$text.'</body></html>';
428 $descriptorspec = array(
429 0 => array("pipe", "r"),
430 1 => array("pipe", "w"),
431 2 => array("file", "/dev/null", "a")
433 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts", $descriptorspec, $pipes);
434 if (is_resource($process)) {
435 fwrite($pipes[0], $text);
437 while (!feof($pipes[1])) {
438 $cleansource .= fgets($pipes[1], 1024);
441 $return_value = proc_close($process);
443 if( $cleansource == '' && $text != '') {
444 return '<h2>'.wfMsg('seriousxhtmlerrors').'</h2><pre>'.htmlspecialchars($text).'</pre>';
450 function doTableStuff ( $t )
452 $t = explode ( "\n" , $t ) ;
453 $td = array () ; # Is currently a td tag open?
454 $ltd = array () ; # Was it TD or TH?
455 $tr = array () ; # Is currently a tr tag open?
456 $ltr = array () ; # tr attributes
457 foreach ( $t AS $k => $x )
460 $fc = substr ( $x , 0 , 1 ) ;
461 if ( "{|" == substr ( $x , 0 , 2 ) )
463 $t[$k] = "\n<table " . $this->fixTagAttributes ( substr ( $x , 3 ) ) . ">" ;
464 array_push ( $td , false ) ;
465 array_push ( $ltd , "" ) ;
466 array_push ( $tr , false ) ;
467 array_push ( $ltr , "" ) ;
469 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
470 else if ( "|}" == substr ( $x , 0 , 2 ) )
473 $l = array_pop ( $ltd ) ;
474 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
475 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
479 /* else if ( "|_" == substr ( $x , 0 , 2 ) ) # Caption
481 $z = trim ( substr ( $x , 2 ) ) ;
482 $t[$k] = "<caption>{$z}</caption>\n" ;
484 else if ( "|-" == substr ( $x , 0 , 2 ) ) # Allows for |---------------
486 $x = substr ( $x , 1 ) ;
487 while ( $x != "" && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
489 $l = array_pop ( $ltd ) ;
490 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
491 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
494 array_push ( $tr , false ) ;
495 array_push ( $td , false ) ;
496 array_push ( $ltd , "" ) ;
497 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
499 else if ( "|" == $fc ||
"!" == $fc ||
"|+" == substr ( $x , 0 , 2 ) ) # Caption
501 if ( "|+" == substr ( $x , 0 , 2 ) )
504 $x = substr ( $x , 1 ) ;
506 $after = substr ( $x , 1 ) ;
507 if ( $fc == "!" ) $after = str_replace ( "!!" , "||" , $after ) ;
508 $after = explode ( "||" , $after ) ;
510 foreach ( $after AS $theline )
515 $tra = array_pop ( $ltr ) ;
516 if ( !array_pop ( $tr ) ) $z = "<tr {$tra}>\n" ;
517 array_push ( $tr , true ) ;
518 array_push ( $ltr , "" ) ;
521 $l = array_pop ( $ltd ) ;
522 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
523 if ( $fc == "|" ) $l = "td" ;
524 else if ( $fc == "!" ) $l = "th" ;
525 else if ( $fc == "+" ) $l = "caption" ;
527 array_push ( $ltd , $l ) ;
528 $y = explode ( "|" , $theline , 2 ) ;
529 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
530 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
532 array_push ( $td , true ) ;
537 # Closing open td, tr && table
538 while ( count ( $td ) > 0 )
540 if ( array_pop ( $td ) ) $t[] = "</td>" ;
541 if ( array_pop ( $tr ) ) $t[] = "</tr>" ;
545 $t = implode ( "\n" , $t ) ;
546 # $t = $this->removeHTMLtags( $t );
550 function internalParse( $text, $linestart, $args = array(), $isMain=true )
552 $fname = "Parser::internalParse";
553 wfProfileIn( $fname );
555 $text = $this->removeHTMLtags( $text );
556 $text = $this->replaceVariables( $text, $args );
558 # $text = preg_replace( "/(^|\n)-----*/", "\\1<hr>", $text );
560 $text = $this->doHeadings( $text );
561 if($this->mOptions
->getUseDynamicDates()) {
562 global $wgDateFormatter;
563 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
565 $text = $this->replaceExternalLinks( $text );
566 $text = $this->doTokenizedParser ( $text );
567 $text = $this->doTableStuff ( $text ) ;
568 $text = $this->formatHeadings( $text, $isMain );
569 $sk =& $this->mOptions
->getSkin();
570 $text = $sk->transformContent( $text );
572 if ( !isset ( $this->categoryMagicDone
) ) {
573 $text .= $this->categoryMagic () ;
574 $this->categoryMagicDone
= true ;
577 wfProfileOut( $fname );
582 /* private */ function doHeadings( $text )
584 for ( $i = 6; $i >= 1; --$i ) {
585 $h = substr( "======", 0, $i );
586 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
587 "<h{$i}>\\1</h{$i}>\\2", $text );
592 # Note: we have to do external links before the internal ones,
593 # and otherwise take great care in the order of things here, so
594 # that we don't end up interpreting some URLs twice.
596 /* private */ function replaceExternalLinks( $text )
598 $fname = "Parser::replaceExternalLinks";
599 wfProfileIn( $fname );
600 $text = $this->subReplaceExternalLinks( $text, "http", true );
601 $text = $this->subReplaceExternalLinks( $text, "https", true );
602 $text = $this->subReplaceExternalLinks( $text, "ftp", false );
603 $text = $this->subReplaceExternalLinks( $text, "irc", false );
604 $text = $this->subReplaceExternalLinks( $text, "gopher", false );
605 $text = $this->subReplaceExternalLinks( $text, "news", false );
606 $text = $this->subReplaceExternalLinks( $text, "mailto", false );
607 wfProfileOut( $fname );
611 /* private */ function subReplaceExternalLinks( $s, $protocol, $autonumber )
613 $unique = "4jzAfzB8hNvf4sqyO9Edd8pSmk9rE2in0Tgw3";
614 $uc = "A-Za-z0-9_\\/~%\\-+&*#?!=()@\\x80-\\xFF";
616 # this is the list of separators that should be ignored if they
617 # are the last character of an URL but that should be included
618 # if they occur within the URL, e.g. "go to www.foo.com, where .."
619 # in this case, the last comma should not become part of the URL,
620 # but in "www.foo.com/123,2342,32.htm" it should.
622 $fnc = "A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF";
623 $images = "gif|png|jpg|jpeg";
625 # PLEASE NOTE: The curly braces { } are not part of the regex,
626 # they are interpreted as part of the string (used to tell PHP
627 # that the content of the string should be inserted there).
628 $e1 = "/(^|[^\\[])({$protocol}:)([{$uc}{$sep}]+)\\/([{$fnc}]+)\\." .
629 "((?i){$images})([^{$uc}]|$)/";
631 $e2 = "/(^|[^\\[])({$protocol}:)(([".$uc."]|[".$sep."][".$uc."])+)([^". $uc . $sep. "]|[".$sep."]|$)/";
632 $sk =& $this->mOptions
->getSkin();
634 if ( $autonumber and $this->mOptions
->getAllowExternalImages() ) { # Use img tags only for HTTP urls
635 $s = preg_replace( $e1, "\\1" . $sk->makeImage( "{$unique}:\\3" .
636 "/\\4.\\5", "\\4.\\5" ) . "\\6", $s );
638 $s = preg_replace( $e2, "\\1" . "<a href=\"{$unique}:\\3\"" .
639 $sk->getExternalLinkAttributes( "{$unique}:\\3", wfEscapeHTML(
640 "{$unique}:\\3" ) ) . ">" . wfEscapeHTML( "{$unique}:\\3" ) .
642 $s = str_replace( $unique, $protocol, $s );
644 $a = explode( "[{$protocol}:", " " . $s );
645 $s = array_shift( $a );
646 $s = substr( $s, 1 );
648 $e1 = "/^([{$uc}"."{$sep}]+)](.*)\$/sD";
649 $e2 = "/^([{$uc}"."{$sep}]+)\\s+([^\\]]+)](.*)\$/sD";
651 foreach ( $a as $line ) {
652 if ( preg_match( $e1, $line, $m ) ) {
653 $link = "{$protocol}:{$m[1]}";
655 if ( $autonumber ) { $text = "[" . ++
$this->mAutonumber
. "]"; }
656 else { $text = wfEscapeHTML( $link ); }
657 } else if ( preg_match( $e2, $line, $m ) ) {
658 $link = "{$protocol}:{$m[1]}";
662 $s .= "[{$protocol}:" . $line;
665 if( $link == $text ||
preg_match( "!$protocol://" . preg_quote( $text, "/" ) . "/?$!", $link ) ) {
668 # Expand the URL for printable version
669 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $link ) . "</i>)</span>";
671 $la = $sk->getExternalLinkAttributes( $link, $text );
672 $s .= "<a href='{$link}'{$la}>{$text}</a>{$paren}{$trail}";
678 /* private */ function handle4Quotes( &$state, $token )
680 /* This one makes some assumptions.
681 * '''Caesar''''s army => <strong>Caesar</strong>'s army
682 * ''''Caesar'''' was a roman emperor => '<strong>Caesar</strong>' was a roman emperor
683 * These assumptions might be wrong, but any other assumption might be wrong, too.
685 if ( $state["strong"] !== false ) {
686 return $this->handle3Quotes( $state, $token ) . "'";
688 return "'" . $this->handle3Quotes( $state, $token );
693 /* private */ function handle3Quotes( &$state, $token )
695 if ( $state["strong"] !== false ) {
696 if ( $state["em"] !== false && $state["em"] > $state["strong"] )
698 # ''' lala ''lala '''
699 $s = "</em></strong><em>";
703 $state["strong"] = FALSE;
706 $state["strong"] = $token["pos"];
711 /* private */ function handle2Quotes( &$state, $token )
713 if ( $state["em"] !== false ) {
714 if ( $state["strong"] !== false && $state["strong"] > $state["em"] )
716 # ''lala'''lala'' ....'''
717 $s = "</strong></em><strong>";
721 $state["em"] = FALSE;
724 $state["em"] = $token["pos"];
730 /* private */ function handle5Quotes( &$state, $token )
733 if ( $state["em"] !== false && $state["strong"] !== false ) {
734 if ( $state["em"] < $state["strong"] ) {
735 $s .= "</strong></em>";
737 $s .= "</em></strong>";
739 $state["strong"] = $state["em"] = FALSE;
740 } elseif ( $state["em"] !== false ) {
741 $s .= "</em><strong>";
742 $state["em"] = FALSE;
743 $state["strong"] = $token["pos"];
744 } elseif ( $state["strong"] !== false ) {
745 $s .= "</strong><em>";
746 $state["strong"] = FALSE;
747 $state["em"] = $token["pos"];
748 } else { # not $em and not $strong
749 $s .= "<strong><em>";
750 $state["strong"] = $state["em"] = $token["pos"];
755 /* private */ function doTokenizedParser( $str )
757 global $wgLang; # for language specific parser hook
758 global $wgUploadDirectory, $wgUseTimeline;
760 $tokenizer=Tokenizer
::newFromString( $str );
761 $tokenStack = array();
764 $state["em"] = FALSE;
765 $state["strong"] = FALSE;
769 # The tokenizer splits the text into tokens and returns them one by one.
770 # Every call to the tokenizer returns a new token.
771 while ( $token = $tokenizer->nextToken() )
773 switch ( $token["type"] )
776 # simple text with no further markup
777 $txt = $token["text"];
780 # Text that contains blanks that have to be converted to
781 # non-breakable spaces for French.
782 # U+202F NARROW NO-BREAK SPACE might be a better choice, but
783 # browser support for Unicode spacing is poor.
784 $txt = str_replace( " ", " ", $token["text"] );
787 # remember the tag opened with 3 [
791 # FIXME : Treat orphaned open tags (stack not empty when text is over)
793 array_push( $tokenStack, $token );
800 # get text from stack, glue it together, and call the code to handle a
803 if ( count( $tokenStack ) == 0 )
805 # stack empty. Found a ]] without an opening [[
809 $lastToken = array_pop( $tokenStack );
810 while ( !(($lastToken["type"] == "[[[") or ($lastToken["type"] == "[[")) )
812 if( !empty( $lastToken["text"] ) ) {
813 $linkText = $lastToken["text"] . $linkText;
815 $lastToken = array_pop( $tokenStack );
818 $txt = $linkText ."]]";
820 if( isset( $lastToken["text"] ) ) {
821 $prefix = $lastToken["text"];
825 $nextToken = $tokenizer->previewToken();
826 if ( $nextToken["type"] == "text" )
828 # Preview just looks at it. Now we have to fetch it.
829 $nextToken = $tokenizer->nextToken();
830 $txt .= $nextToken["text"];
832 $txt = $this->handleInternalLink( $this->unstrip($txt,$this->mStripState
), $prefix );
834 # did the tag start with 3 [ ?
836 # show the first as text
842 $tagIsOpen = (count( $tokenStack ) != 0);
848 # This and the four next ones handle quotes
849 $txt = $this->handle3Quotes( $state, $token );
852 $txt = $this->handle2Quotes( $state, $token );
855 $txt = $this->handle5Quotes( $state, $token );
858 $txt = $this->handle4Quotes( $state, $token );
865 #heading- used to close all unbalanced bold or em tags in this section
867 if( $state['em'] !== false and
868 ( $state['strong'] === false or $state['em'] > $state['strong'] ) )
871 $state['em'] = false;
873 if ( $state['strong'] !== false ) $txt .= '</strong>';
874 if ( $state['em'] !== false ) $txt .= '</em>';
875 $state['strong'] = $state['em'] = false;
881 $txt = $this->doMagicRFC( $tokenizer );
888 $txt = $this->doMagicISBN( $tokenizer );
892 if ( $wgUseTimeline &&
893 "" != ( $timelinesrc = $tokenizer->readAllUntil("</timeline>") ) )
895 $txt = renderTimeline( $timelinesrc );
901 # Call language specific Hook.
902 $txt = $wgLang->processToken( $token, $tokenStack );
903 if ( NULL == $txt ) {
904 # An unkown token. Highlight.
905 $txt = "<font color=\"#FF0000\"><b>".$token["type"]."</b></font>";
906 $txt .= "<font color=\"#FFFF00\"><b>".$token["text"]."</b></font>";
910 # If we're parsing the interior of a link, don't append the interior to $s,
911 # but push it to the stack so it can be processed when a ]] token is found.
912 if ( $tagIsOpen && $txt != "" ) {
913 $token["type"] = "text";
914 $token["text"] = $txt;
915 array_push( $tokenStack, $token );
921 # make 100% sure all strong and em tags are closed
922 # doBlockLevels often messes the last bit up though, but invalid nesting is better than unclosed tags
923 # tidy solves this though
924 if( $state['em'] !== false and
925 ( $state['strong'] === false or $state['em'] > $state['strong'] ) )
928 $state['em'] = false;
930 if ( $state['strong'] !== false ) $s .= '</strong>';
931 if ( $state['em'] !== false ) $s .= '</em>';
933 if ( count( $tokenStack ) != 0 )
935 # still objects on stack. opened [[ tag without closing ]] tag.
937 while ( $lastToken = array_pop( $tokenStack ) )
939 if ( $lastToken["type"] == "text" )
941 $txt = $lastToken["text"] . $txt;
943 $txt = $lastToken["type"] . $txt;
951 /* private */ function handleInternalLink( $line, $prefix )
953 global $wgLang, $wgLinkCache;
954 global $wgNamespacesWithSubpages, $wgLanguageCode;
955 static $fname = "Parser::handleInternalLink" ;
956 wfProfileIn( $fname );
958 wfProfileIn( "$fname-setup" );
960 if ( !$tc ) { $tc = Title
::legalChars() . "#"; }
961 $sk =& $this->mOptions
->getSkin();
963 # Match a link having the form [[namespace:link|alternate]]trail
965 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
966 # Match the end of a line for a word that's not followed by whitespace,
967 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
968 #$e2 = "/^(.*)\\b(\\w+)\$/suD";
969 #$e2 = "/^(.*\\s)(\\S+)\$/suD";
970 static $e2 = '/^(.*\s)([a-zA-Z\x80-\xff]+)$/sD';
973 # Special and Media are pseudo-namespaces; no pages actually exist in them
974 static $image = FALSE;
975 static $special = FALSE;
976 static $media = FALSE;
977 static $category = FALSE;
978 if ( !$image ) { $image = Namespace::getImage(); }
979 if ( !$special ) { $special = Namespace::getSpecial(); }
980 if ( !$media ) { $media = Namespace::getMedia(); }
981 if ( !$category ) { $category = Namespace::getCategory(); }
983 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
985 wfProfileOut( "$fname-setup" );
988 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
991 } else { # Invalid form; output directly
992 $s .= $prefix . "[[" . $line ;
998 :Foobar -- override special treatment of prefix (images, language links)
999 /Foobar -- convert to CurrentPage/Foobar
1000 /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1002 $c = substr($m[1],0,1);
1003 $noforce = ($c != ":");
1004 if( $c == "/" ) { # subpage
1005 if(substr($m[1],-1,1)=="/") { # / at end means we don't want the slash to be shown
1006 $m[1]=substr($m[1],1,strlen($m[1])-2);
1009 $noslash=substr($m[1],1);
1011 if($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]) { # subpages allowed here
1012 $link = $this->mTitle
->getPrefixedText(). "/" . trim($noslash);
1015 } # this might be changed for ugliness reasons
1017 $link = $noslash; # no subpage allowed, use standard link
1019 } elseif( $noforce ) { # no subpage
1022 $link = substr( $m[1], 1 );
1027 $nt = Title
::newFromText( $link );
1029 $s .= $prefix . "[[" . $line;
1032 $ns = $nt->getNamespace();
1033 $iw = $nt->getInterWiki();
1035 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
1036 array_push( $this->mOutput
->mLanguageLinks
, $nt->getPrefixedText() );
1037 $s .= $prefix . $trail ;
1038 return (trim($s) == '')?
'': $s;
1040 if( $ns == $image ) {
1041 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1042 $wgLinkCache->addImageLinkObj( $nt );
1045 if ( $ns == $category ) {
1046 $t = $nt->getText() ;
1047 $nnt = Title
::newFromText ( Namespace::getCanonicalName($category).":".$t ) ;
1048 $t = $sk->makeLinkObj( $nnt, $t, "", "" , $prefix );
1049 $this->mOutput
->mCategoryLinks
[] = $t ;
1050 $s .= $prefix . $trail ;
1054 if( ( $nt->getPrefixedText() == $this->mTitle
->getPrefixedText() ) &&
1055 ( strpos( $link, "#" ) == FALSE ) ) {
1056 # Self-links are handled specially; generally de-link and change to bold.
1057 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, "", $trail );
1061 if( $ns == $media ) {
1062 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1063 $wgLinkCache->addImageLinkObj( $nt );
1065 } elseif( $ns == $special ) {
1066 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, "", $trail );
1069 $s .= $sk->makeLinkObj( $nt, $text, "", $trail , $prefix );
1071 wfProfileOut( $fname );
1075 # Some functions here used by doBlockLevels()
1077 /* private */ function closeParagraph()
1080 if ( '' != $this->mLastSection
) {
1081 $result = "</" . $this->mLastSection
. ">\n";
1083 $this->mInPre
= false;
1084 $this->mLastSection
= "";
1087 # getCommon() returns the length of the longest common substring
1088 # of both arguments, starting at the beginning of both.
1090 /* private */ function getCommon( $st1, $st2 )
1092 $fl = strlen( $st1 );
1093 $shorter = strlen( $st2 );
1094 if ( $fl < $shorter ) { $shorter = $fl; }
1096 for ( $i = 0; $i < $shorter; ++
$i ) {
1097 if ( $st1{$i} != $st2{$i} ) { break; }
1101 # These next three functions open, continue, and close the list
1102 # element appropriate to the prefix character passed into them.
1104 /* private */ function openList( $char )
1106 $result = $this->closeParagraph();
1108 if ( "*" == $char ) { $result .= "<ul><li>"; }
1109 else if ( "#" == $char ) { $result .= "<ol><li>"; }
1110 else if ( ":" == $char ) { $result .= "<dl><dd>"; }
1111 else if ( ";" == $char ) {
1112 $result .= "<dl><dt>";
1113 $this->mDTopen
= true;
1115 else { $result = "<!-- ERR 1 -->"; }
1120 /* private */ function nextItem( $char )
1122 if ( "*" == $char ||
"#" == $char ) { return "</li><li>"; }
1123 else if ( ":" == $char ||
";" == $char ) {
1125 if ( $this->mDTopen
) { $close = "</dt>"; }
1126 if ( ";" == $char ) {
1127 $this->mDTopen
= true;
1128 return $close . "<dt>";
1130 $this->mDTopen
= false;
1131 return $close . "<dd>";
1134 return "<!-- ERR 2 -->";
1137 /* private */function closeList( $char )
1139 if ( "*" == $char ) { $text = "</li></ul>"; }
1140 else if ( "#" == $char ) { $text = "</li></ol>"; }
1141 else if ( ":" == $char ) {
1142 if ( $this->mDTopen
) {
1143 $this->mDTopen
= false;
1144 $text = "</dt></dl>";
1146 $text = "</dd></dl>";
1149 else { return "<!-- ERR 3 -->"; }
1153 /* private */ function doBlockLevels( $text, $linestart ) {
1154 $fname = "Parser::doBlockLevels";
1155 wfProfileIn( $fname );
1157 # Parsing through the text line by line. The main thing
1158 # happening here is handling of block-level elements p, pre,
1159 # and making lists from lines starting with * # : etc.
1161 $textLines = explode( "\n", $text );
1163 $lastPrefix = $output = $lastLine = '';
1164 $this->mDTopen
= $inBlockElem = false;
1166 $paragraphStack = false;
1168 if ( !$linestart ) {
1169 $output .= array_shift( $textLines );
1171 foreach ( $textLines as $oLine ) {
1172 $lastPrefixLength = strlen( $lastPrefix );
1173 $preCloseMatch = preg_match("/<\\/pre/i", $oLine );
1174 $preOpenMatch = preg_match("/<pre/i", $oLine );
1175 if (!$this->mInPre
) {
1176 $this->mInPre
= !empty($preOpenMatch);
1178 if ( !$this->mInPre
) {
1179 # Multiple prefixes may abut each other for nested lists.
1180 $prefixLength = strspn( $oLine, "*#:;" );
1181 $pref = substr( $oLine, 0, $prefixLength );
1184 $pref2 = str_replace( ";", ":", $pref );
1185 $t = substr( $oLine, $prefixLength );
1187 # Don't interpret any other prefixes in preformatted text
1189 $pref = $pref2 = '';
1194 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1195 # Same as the last item, so no need to deal with nesting or opening stuff
1196 $output .= $this->nextItem( substr( $pref, -1 ) );
1197 $paragraphStack = false;
1199 if ( ";" == substr( $pref, -1 ) ) {
1200 # The one nasty exception: definition lists work like this:
1201 # ; title : definition text
1202 # So we check for : in the remainder text to split up the
1203 # title and definition, without b0rking links.
1204 # FIXME: This is not foolproof. Something better in Tokenizer might help.
1205 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1207 $output .= $term . $this->nextItem( ":" );
1211 } elseif( $prefixLength ||
$lastPrefixLength ) {
1212 # Either open or close a level...
1213 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1214 $paragraphStack = false;
1216 while( $commonPrefixLength < $lastPrefixLength ) {
1217 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1218 --$lastPrefixLength;
1220 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1221 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1223 while ( $prefixLength > $commonPrefixLength ) {
1224 $char = substr( $pref, $commonPrefixLength, 1 );
1225 $output .= $this->openList( $char );
1227 if ( ";" == $char ) {
1228 # FIXME: This is dupe of code above
1229 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1231 $output .= $term . $this->nextItem( ":" );
1235 ++
$commonPrefixLength;
1237 $lastPrefix = $pref2;
1239 if( 0 == $prefixLength ) {
1240 # No prefix (not in list)--go to paragraph mode
1241 $uniq_prefix = UNIQ_PREFIX
;
1242 // XXX: use a stack for nestable elements like span, table and div
1243 $openmatch = preg_match("/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<div|<pre|<tr|<td|<p|<ul|<li)/i", $t );
1244 $closematch = preg_match(
1245 "/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|".
1246 "<\\/div|<hr|<\\/td|<\\/pre|<\\/p|".$uniq_prefix."-pre|<\\/li|<\\/ul)/i", $t );
1247 if ( $openmatch or $closematch ) {
1248 $paragraphStack = false;
1249 $output .= $this->closeParagraph();
1250 if($preOpenMatch and !$preCloseMatch) {
1251 $this->mInPre
= true;
1253 if ( $closematch ) {
1254 $inBlockElem = false;
1256 $inBlockElem = true;
1258 } else if ( !$inBlockElem ) {
1259 if ( " " == $t{0} ) {
1261 if ($this->mLastSection
!= 'pre') {
1262 $paragraphStack = false;
1263 $output .= $this->closeParagraph().'<pre>';
1264 $this->mLastSection
= 'pre';
1268 if ( '' == trim($t) ) {
1269 if ( $paragraphStack ) {
1270 $output .= $paragraphStack.'<br/>';
1271 $paragraphStack = false;
1272 $this->mLastSection
= 'p';
1274 if ($this->mLastSection
!= 'p' ) {
1275 $output .= $this->closeParagraph();
1276 $this->mLastSection
= '';
1277 $paragraphStack = "<p>";
1279 $paragraphStack = '</p><p>';
1283 if ( $paragraphStack ) {
1284 $output .= $paragraphStack;
1285 $paragraphStack = false;
1286 $this->mLastSection
= 'p';
1287 } else if ($this->mLastSection
!= 'p') {
1288 $output .= $this->closeParagraph().'<p>';
1289 $this->mLastSection
= 'p';
1295 if ($paragraphStack === false) {
1299 while ( $prefixLength ) {
1300 $output .= $this->closeList( $pref2{$prefixLength-1} );
1303 if ( "" != $this->mLastSection
) {
1304 $output .= "</" . $this->mLastSection
. ">";
1305 $this->mLastSection
= "";
1308 wfProfileOut( $fname );
1312 function getVariableValue( $index ) {
1313 global $wgLang, $wgSitename, $wgServer;
1316 case MAG_CURRENTMONTH
:
1318 case MAG_CURRENTMONTHNAME
:
1319 return $wgLang->getMonthName( date("n") );
1320 case MAG_CURRENTMONTHNAMEGEN
:
1321 return $wgLang->getMonthNameGen( date("n") );
1322 case MAG_CURRENTDAY
:
1325 return $this->mTitle
->getText();
1327 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1328 return $wgLang->getNsText($this->mTitle
->getNamespace()); // Patch by Dori
1329 case MAG_CURRENTDAYNAME
:
1330 return $wgLang->getWeekdayName( date("w")+
1 );
1331 case MAG_CURRENTYEAR
:
1333 case MAG_CURRENTTIME
:
1334 return $wgLang->time( wfTimestampNow(), false );
1335 case MAG_NUMBEROFARTICLES
:
1336 return wfNumberOfArticles();
1346 function initialiseVariables()
1348 global $wgVariableIDs;
1349 $this->mVariables
= array();
1350 foreach ( $wgVariableIDs as $id ) {
1351 $mw =& MagicWord
::get( $id );
1352 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1356 /* private */ function replaceVariables( $text, $args = array() )
1358 global $wgLang, $wgScript, $wgArticlePath;
1360 $fname = "Parser::replaceVariables";
1361 wfProfileIn( $fname );
1364 if ( !$this->mVariables
) {
1365 $this->initialiseVariables();
1367 $titleChars = Title
::legalChars();
1368 $regex = "/(\\n?){{([$titleChars]*?)(\\|.*?|)}}/s";
1370 # This function is called recursively. To keep track of arguments we need a stack:
1371 array_push( $this->mArgStack
, $args );
1373 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1374 $GLOBALS['wgCurParser'] =& $this;
1375 $text = preg_replace_callback( $regex, "wfBraceSubstitution", $text );
1377 array_pop( $this->mArgStack
);
1382 function braceSubstitution( $matches )
1384 global $wgLinkCache, $wgLang;
1385 $fname = "Parser::braceSubstitution";
1390 # $newline is an optional newline character before the braces
1391 # $part1 is the bit before the first |, and must contain only title characters
1392 # $args is a list of arguments, starting from index 0, not including $part1
1394 $newline = $matches[1];
1395 $part1 = $matches[2];
1396 # If the third subpattern matched anything, it will start with |
1397 if ( $matches[3] !== "" ) {
1398 $args = explode( "|", substr( $matches[3], 1 ) );
1402 $argc = count( $args );
1405 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1406 if ( $mwSubst->matchStartAndRemove( $part1 ) ) {
1407 if ( $this->mOutputType
!= OT_WIKI
) {
1408 # Invalid SUBST not replaced at PST time
1409 # Return without further processing
1410 $text = $matches[0];
1413 } elseif ( $this->mOutputType
== OT_WIKI
) {
1414 # SUBST not found in PST pass, do nothing
1415 $text = $matches[0];
1419 # MSG, MSGNW and INT
1422 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1423 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1426 # Remove obsolete MSG:
1427 $mwMsg =& MagicWord
::get( MAG_MSG
);
1428 $mwMsg->matchStartAndRemove( $part1 );
1431 # Check if it is an internal message
1432 $mwInt =& MagicWord
::get( MAG_INT
);
1433 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1434 if ( $this->incrementIncludeCount( "int:$part1" ) ) {
1435 $text = wfMsgReal( $part1, $args, true );
1443 # Check for NS: (namespace expansion)
1444 $mwNs = MagicWord
::get( MAG_NS
);
1445 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1446 if ( intval( $part1 ) ) {
1447 $text = $wgLang->getNsText( intval( $part1 ) );
1450 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1451 if ( !is_null( $index ) ) {
1452 $text = $wgLang->getNsText( $index );
1459 # LOCALURL and LOCALURLE
1461 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1462 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1464 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1465 $func = 'getLocalURL';
1466 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1467 $func = 'escapeLocalURL';
1472 if ( $func !== '' ) {
1473 $title = Title
::newFromText( $part1 );
1474 if ( !is_null( $title ) ) {
1476 $text = $title->$func( $args[0] );
1478 $text = $title->$func();
1485 # Internal variables
1486 if ( !$found && array_key_exists( $part1, $this->mVariables
) ) {
1487 $text = $this->mVariables
[$part1];
1489 $this->mOutput
->mContainsOldMagic
= true;
1492 # Arguments input from the caller
1493 $inputArgs = end( $this->mArgStack
);
1494 if ( !$found && array_key_exists( $part1, $inputArgs ) ) {
1495 $text = $inputArgs[$part1];
1499 # Load from database
1501 $title = Title
::newFromText( $part1, NS_TEMPLATE
);
1502 if ( !is_null( $title ) && !$title->isExternal() ) {
1503 # Check for excessive inclusion
1504 $dbk = $title->getPrefixedDBkey();
1505 if ( $this->incrementIncludeCount( $dbk ) ) {
1506 $article = new Article( $title );
1507 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1508 if ( $articleContent !== false ) {
1510 $text = $articleContent;
1515 # If the title is valid but undisplayable, make a link to it
1516 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1517 $text = "[[" . $title->getPrefixedText() . "]]";
1523 # Recursive parsing, escaping and link table handling
1524 # Only for HTML output
1525 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1526 $text = wfEscapeWikiText( $text );
1527 } elseif ( $this->mOutputType
== OT_HTML
&& $found ) {
1528 # Clean up argument array
1529 $assocArgs = array();
1531 foreach( $args as $arg ) {
1532 $eqpos = strpos( $arg, "=" );
1533 if ( $eqpos === false ) {
1534 $assocArgs[$index++
] = $arg;
1536 $name = trim( substr( $arg, 0, $eqpos ) );
1537 $value = trim( substr( $arg, $eqpos+
1 ) );
1538 if ( $value === false ) {
1541 if ( $name !== false ) {
1542 $assocArgs[$name] = $value;
1547 # Do not enter included links in link table
1548 if ( !is_null( $title ) ) {
1549 $wgLinkCache->suspend();
1552 # Run full parser on the included text
1553 $text = $this->strip( $text, $this->mStripState
);
1554 $text = $this->internalParse( $text, (bool)$newline, $assocArgs, false );
1556 # Add the result to the strip state for re-inclusion after
1557 # the rest of the processing
1558 $text = $this->insertStripItem( $text, $this->mStripState
);
1560 # Resume the link cache and register the inclusion as a link
1561 if ( !is_null( $title ) ) {
1562 $wgLinkCache->resume();
1563 $wgLinkCache->addLinkObj( $title );
1574 # Returns true if the function is allowed to include this entity
1575 function incrementIncludeCount( $dbk )
1577 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1578 $this->mIncludeCount
[$dbk] = 0;
1580 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1588 # Cleans up HTML, removes dangerous tags and attributes
1589 /* private */ function removeHTMLtags( $text )
1591 global $wgUseTidy, $wgUserHtml;
1592 $fname = "Parser::removeHTMLtags";
1593 wfProfileIn( $fname );
1596 $htmlpairs = array( # Tags that must be closed
1597 "b", "del", "i", "ins", "u", "font", "big", "small", "sub", "sup", "h1",
1598 "h2", "h3", "h4", "h5", "h6", "cite", "code", "em", "s",
1599 "strike", "strong", "tt", "var", "div", "center",
1600 "blockquote", "ol", "ul", "dl", "table", "caption", "pre",
1601 "ruby", "rt" , "rb" , "rp", "p"
1603 $htmlsingle = array(
1604 "br", "hr", "li", "dt", "dd"
1606 $htmlnest = array( # Tags that can be nested--??
1607 "table", "tr", "td", "th", "div", "blockquote", "ol", "ul",
1608 "dl", "font", "big", "small", "sub", "sup"
1610 $tabletags = array( # Can only appear inside table
1614 $htmlpairs = array();
1615 $htmlsingle = array();
1616 $htmlnest = array();
1617 $tabletags = array();
1620 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1621 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1623 $htmlattrs = $this->getHTMLattrs () ;
1625 # Remove HTML comments
1626 $text = preg_replace( "/(\\n *<!--.*--> *(?=\\n)|<!--.*-->)/sU", "$2", $text );
1628 $bits = explode( "<", $text );
1629 $text = array_shift( $bits );
1631 $tagstack = array(); $tablestack = array();
1632 foreach ( $bits as $x ) {
1633 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
1634 preg_match( "/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/",
1636 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1637 error_reporting( $prev );
1640 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1644 if ( ! in_array( $t, $htmlsingle ) &&
1645 ( count($tagstack) && $ot = array_pop( $tagstack ) ) != $t ) {
1646 if(!empty($ot)) array_push( $tagstack, $ot );
1649 if ( $t == "table" ) {
1650 $tagstack = array_pop( $tablestack );
1655 # Keep track for later
1656 if ( in_array( $t, $tabletags ) &&
1657 ! in_array( "table", $tagstack ) ) {
1659 } else if ( in_array( $t, $tagstack ) &&
1660 ! in_array ( $t , $htmlnest ) ) {
1662 } else if ( ! in_array( $t, $htmlsingle ) ) {
1663 if ( $t == "table" ) {
1664 array_push( $tablestack, $tagstack );
1665 $tagstack = array();
1667 array_push( $tagstack, $t );
1669 # Strip non-approved attributes from the tag
1670 $newparams = $this->fixTagAttributes($params);
1674 $rest = str_replace( ">", ">", $rest );
1675 $text .= "<$slash$t $newparams$brace$rest";
1679 $text .= "<" . str_replace( ">", ">", $x);
1681 # Close off any remaining tags
1682 while ( $t = array_pop( $tagstack ) ) {
1684 if ( $t == "table" ) { $tagstack = array_pop( $tablestack ); }
1687 # this might be possible using tidy itself
1688 foreach ( $bits as $x ) {
1689 preg_match( "/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/",
1691 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1692 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1693 $newparams = $this->fixTagAttributes($params);
1694 $rest = str_replace( ">", ">", $rest );
1695 $text .= "<$slash$t $newparams$brace$rest";
1697 $text .= "<" . str_replace( ">", ">", $x);
1701 wfProfileOut( $fname );
1708 * This function accomplishes several tasks:
1709 * 1) Auto-number headings if that option is enabled
1710 * 2) Add an [edit] link to sections for logged in users who have enabled the option
1711 * 3) Add a Table of contents on the top for users who have enabled the option
1712 * 4) Auto-anchor headings
1714 * It loops through all headlines, collects the necessary data, then splits up the
1715 * string and re-inserts the newly formatted headlines.
1719 /* private */ function formatHeadings( $text, $isMain=true )
1721 global $wgInputEncoding;
1723 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
1724 $doShowToc = $this->mOptions
->getShowToc();
1725 if( !$this->mTitle
->userCanEdit() ) {
1727 $rightClickHack = 0;
1729 $showEditLink = $this->mOptions
->getEditSection();
1730 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
1733 # Inhibit editsection links if requested in the page
1734 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
1735 if( $esw->matchAndRemove( $text ) ) {
1738 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1740 $mw =& MagicWord
::get( MAG_NOTOC
);
1741 if( $mw->matchAndRemove( $text ) ) {
1745 # never add the TOC to the Main Page. This is an entry page that should not
1746 # be more than 1-2 screens large anyway
1747 if( $this->mTitle
->getPrefixedText() == wfMsg("mainpage") ) {
1751 # Get all headlines for numbering them and adding funky stuff like [edit]
1752 # links - this is for later, but we need the number of headlines right now
1753 $numMatches = preg_match_all( "/<H([1-6])(.*?" . ">)(.*?)<\/H[1-6]>/i", $text, $matches );
1755 # if there are fewer than 4 headlines in the article, do not show TOC
1756 if( $numMatches < 4 ) {
1760 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
1761 # override above conditions and always show TOC
1762 $mw =& MagicWord
::get( MAG_FORCETOC
);
1763 if ($mw->matchAndRemove( $text ) ) {
1768 # We need this to perform operations on the HTML
1769 $sk =& $this->mOptions
->getSkin();
1774 # Ugh .. the TOC should have neat indentation levels which can be
1775 # passed to the skin functions. These are determined here
1780 $sublevelCount = array();
1783 foreach( $matches[3] as $headline ) {
1786 $prevlevel = $level;
1788 $level = $matches[1][$headlineCount];
1789 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
1790 # reset when we enter a new level
1791 $sublevelCount[$level] = 0;
1792 $toc .= $sk->tocIndent( $level - $prevlevel );
1793 $toclevel +
= $level - $prevlevel;
1795 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
1796 # reset when we step back a level
1797 $sublevelCount[$level+
1]=0;
1798 $toc .= $sk->tocUnindent( $prevlevel - $level );
1799 $toclevel -= $prevlevel - $level;
1801 # count number of headlines for each level
1802 @$sublevelCount[$level]++
;
1803 if( $doNumberHeadings ||
$doShowToc ) {
1805 for( $i = 1; $i <= $level; $i++
) {
1806 if( !empty( $sublevelCount[$i] ) ) {
1810 $numbering .= $sublevelCount[$i];
1816 # The canonized header is a version of the header text safe to use for links
1817 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
1818 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
1821 $canonized_headline = preg_replace( "/<.*?" . ">/","",$canonized_headline );
1822 $tocline = trim( $canonized_headline );
1823 $canonized_headline = preg_replace("/[ \\?&\\/<>\\(\\)\\[\\]=,+']+/", '_', urlencode( do_html_entity_decode( $tocline, ENT_COMPAT
, $wgInputEncoding ) ) );
1824 $refer[$headlineCount] = $canonized_headline;
1826 # count how many in assoc. array so we can track dupes in anchors
1827 @$refers[$canonized_headline]++
;
1828 $refcount[$headlineCount]=$refers[$canonized_headline];
1830 # Prepend the number to the heading text
1832 if( $doNumberHeadings ||
$doShowToc ) {
1833 $tocline = $numbering . " " . $tocline;
1835 # Don't number the heading if it is the only one (looks silly)
1836 if( $doNumberHeadings && count( $matches[3] ) > 1) {
1837 # the two are different if the line contains a link
1838 $headline=$numbering . " " . $headline;
1842 # Create the anchor for linking from the TOC to the section
1843 $anchor = $canonized_headline;
1844 if($refcount[$headlineCount] > 1 ) {
1845 $anchor .= "_" . $refcount[$headlineCount];
1848 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
1850 if( $showEditLink ) {
1851 if ( empty( $head[$headlineCount] ) ) {
1852 $head[$headlineCount] = "";
1854 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+
1);
1857 # Add the edit section span
1858 if( $rightClickHack ) {
1859 $headline = $sk->editSectionScript($headlineCount+
1,$headline);
1862 # give headline the correct <h#> tag
1863 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline."</h".$level.">";
1869 $toclines = $headlineCount;
1870 $toc .= $sk->tocUnindent( $toclevel );
1871 $toc = $sk->tocTable( $toc );
1874 # split up and insert constructed headlines
1876 $blocks = preg_split( "/<H[1-6].*?" . ">.*?<\/H[1-6]>/i", $text );
1879 foreach( $blocks as $block ) {
1880 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
1881 # This is the [edit] link that appears for the top block of text when
1882 # section editing is enabled
1884 # Disabled because it broke block formatting
1885 # For example, a bullet point in the top line
1886 # $full .= $sk->editSectionLink(0);
1889 if( $doShowToc && !$i && $isMain) {
1890 # Top anchor now in skin
1894 if( !empty( $head[$i] ) ) {
1903 /* private */ function doMagicISBN( &$tokenizer )
1907 # Check whether next token is a text token
1908 # If yes, fetch it and convert the text into a
1909 # Special::BookSources link
1910 $token = $tokenizer->previewToken();
1911 while ( $token["type"] == "" )
1913 $tokenizer->nextToken();
1914 $token = $tokenizer->previewToken();
1916 if ( $token["type"] == "text" )
1918 $token = $tokenizer->nextToken();
1919 $x = $token["text"];
1920 $valid = "0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ";
1922 $isbn = $blank = "" ;
1923 while ( " " == $x{0} ) {
1925 $x = substr( $x, 1 );
1927 while ( strstr( $valid, $x{0} ) != false ) {
1929 $x = substr( $x, 1 );
1931 $num = str_replace( "-", "", $isbn );
1932 $num = str_replace( " ", "", $num );
1935 $text = "ISBN $blank$x";
1937 $titleObj = Title
::makeTitle( NS_SPECIAL
, "Booksources" );
1938 $text = "<a href=\"" .
1939 $titleObj->escapeLocalUrl( "isbn={$num}" ) .
1940 "\" class=\"internal\">ISBN $isbn</a>";
1948 /* private */ function doMagicRFC( &$tokenizer )
1952 # Check whether next token is a text token
1953 # If yes, fetch it and convert the text into a
1954 # link to an RFC source
1955 $token = $tokenizer->previewToken();
1956 while ( $token["type"] == "" )
1958 $tokenizer->nextToken();
1959 $token = $tokenizer->previewToken();
1961 if ( $token["type"] == "text" )
1963 $token = $tokenizer->nextToken();
1964 $x = $token["text"];
1965 $valid = "0123456789";
1967 $rfc = $blank = "" ;
1968 while ( " " == $x{0} ) {
1970 $x = substr( $x, 1 );
1972 while ( strstr( $valid, $x{0} ) != false ) {
1974 $x = substr( $x, 1 );
1978 $text .= "RFC $blank$x";
1980 $url = wfmsg( "rfcurl" );
1981 $url = str_replace( "$1", $rfc, $url);
1982 $sk =& $this->mOptions
->getSkin();
1983 $la = $sk->getExternalLinkAttributes( $url, "RFC {$rfc}" );
1984 $text = "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
1992 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true )
1994 $this->mOptions
= $options;
1995 $this->mTitle
=& $title;
1996 $this->mOutputType
= OT_WIKI
;
1998 if ( $clearState ) {
1999 $this->clearState();
2002 $stripState = false;
2006 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2009 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2010 "/<br *?>/i" => "<br/>",
2012 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2013 $text = $this->strip( $text, $stripState, false );
2014 $text = $this->pstPass2( $text, $user );
2015 $text = $this->unstrip( $text, $stripState );
2019 /* private */ function pstPass2( $text, &$user )
2021 global $wgLang, $wgLocaltimezone, $wgCurParser;
2023 # Variable replacement
2024 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2025 $text = $this->replaceVariables( $text );
2029 $n = $user->getName();
2030 $k = $user->getOption( "nickname" );
2031 if ( "" == $k ) { $k = $n; }
2032 if(isset($wgLocaltimezone)) {
2033 $oldtz = getenv("TZ"); putenv("TZ=$wgLocaltimezone");
2035 /* Note: this is an ugly timezone hack for the European wikis */
2036 $d = $wgLang->timeanddate( date( "YmdHis" ), false ) .
2037 " (" . date( "T" ) . ")";
2038 if(isset($wgLocaltimezone)) putenv("TZ=$oldtz");
2040 $text = preg_replace( "/~~~~~/", $d, $text );
2041 $text = preg_replace( "/~~~~/", "[[" . $wgLang->getNsText(
2042 Namespace::getUser() ) . ":$n|$k]] $d", $text );
2043 $text = preg_replace( "/~~~/", "[[" . $wgLang->getNsText(
2044 Namespace::getUser() ) . ":$n|$k]]", $text );
2046 # Context links: [[|name]] and [[name (context)|]]
2048 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2049 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2050 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2051 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2053 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2054 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2055 $p3 = "/\[\[($namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]]
2056 $p4 = "/\[\[($namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/";
2057 # [[ns:page (cont)|]]
2059 $t = $this->mTitle
->getText();
2060 if ( preg_match( $conpat, $t, $m ) ) {
2063 $text = preg_replace( $p4, "[[\\1:\\2 (\\3)|\\2]]", $text );
2064 $text = preg_replace( $p1, "[[\\1 (\\2)|\\1]]", $text );
2065 $text = preg_replace( $p3, "[[\\1:\\2|\\2]]", $text );
2067 if ( "" == $context ) {
2068 $text = preg_replace( $p2, "[[\\1]]", $text );
2070 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2074 $mw =& MagicWord::get( MAG_SUBST );
2075 $wgCurParser = $this->fork();
2076 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
2077 $this->merge( $wgCurParser );
2080 # Trim trailing whitespace
2081 # MAG_END (__END__) tag allows for trailing
2082 # whitespace to be deliberately included
2083 $text = rtrim( $text );
2084 $mw =& MagicWord
::get( MAG_END
);
2085 $mw->matchAndRemove( $text );
2090 # Set up some variables which are usually set up in parse()
2091 # so that an external function can call some class members with confidence
2092 function startExternalParse( &$title, $options, $outputType, $clearState = true )
2094 $this->mTitle
=& $title;
2095 $this->mOptions
= $options;
2096 $this->mOutputType
= $outputType;
2097 if ( $clearState ) {
2098 $this->clearState();
2102 function transformMsg( $text, $options ) {
2104 static $executing = false;
2106 # Guard against infinite recursion
2112 $this->mTitle
= $wgTitle;
2113 $this->mOptions
= $options;
2114 $this->mOutputType
= OT_MSG
;
2115 $this->clearState();
2116 $text = $this->replaceVariables( $text );
2125 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2127 function ParserOutput( $text = "", $languageLinks = array(), $categoryLinks = array(),
2128 $containsOldMagic = false )
2130 $this->mText
= $text;
2131 $this->mLanguageLinks
= $languageLinks;
2132 $this->mCategoryLinks
= $categoryLinks;
2133 $this->mContainsOldMagic
= $containsOldMagic;
2136 function getText() { return $this->mText
; }
2137 function getLanguageLinks() { return $this->mLanguageLinks
; }
2138 function getCategoryLinks() { return $this->mCategoryLinks
; }
2139 function containsOldMagic() { return $this->mContainsOldMagic
; }
2140 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2141 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2142 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2143 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2145 function merge( $other ) {
2146 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2147 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2148 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2155 # All variables are private
2156 var $mUseTeX; # Use texvc to expand <math> tags
2157 var $mUseCategoryMagic; # Treat [[Category:xxxx]] tags specially
2158 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2159 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2160 var $mAllowExternalImages; # Allow external images inline
2161 var $mSkin; # Reference to the preferred skin
2162 var $mDateFormat; # Date format index
2163 var $mEditSection; # Create "edit section" links
2164 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2165 var $mNumberHeadings; # Automatically number headings
2166 var $mShowToc; # Show table of contents
2168 function getUseTeX() { return $this->mUseTeX
; }
2169 function getUseCategoryMagic() { return $this->mUseCategoryMagic
; }
2170 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2171 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2172 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2173 function getSkin() { return $this->mSkin
; }
2174 function getDateFormat() { return $this->mDateFormat
; }
2175 function getEditSection() { return $this->mEditSection
; }
2176 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
2177 function getNumberHeadings() { return $this->mNumberHeadings
; }
2178 function getShowToc() { return $this->mShowToc
; }
2180 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2181 function setUseCategoryMagic( $x ) { return wfSetVar( $this->mUseCategoryMagic
, $x ); }
2182 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2183 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2184 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2185 function setSkin( $x ) { return wfSetRef( $this->mSkin
, $x ); }
2186 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2187 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
2188 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
2189 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
2190 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
2192 /* static */ function newFromUser( &$user )
2194 $popts = new ParserOptions
;
2195 $popts->initialiseFromUser( $user );
2199 function initialiseFromUser( &$userInput )
2201 global $wgUseTeX, $wgUseCategoryMagic, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2203 if ( !$userInput ) {
2205 $user->setLoaded( true );
2207 $user =& $userInput;
2210 $this->mUseTeX
= $wgUseTeX;
2211 $this->mUseCategoryMagic
= $wgUseCategoryMagic;
2212 $this->mUseDynamicDates
= $wgUseDynamicDates;
2213 $this->mInterwikiMagic
= $wgInterwikiMagic;
2214 $this->mAllowExternalImages
= $wgAllowExternalImages;
2215 $this->mSkin
=& $user->getSkin();
2216 $this->mDateFormat
= $user->getOption( "date" );
2217 $this->mEditSection
= $user->getOption( "editsection" );
2218 $this->mEditSectionOnRightClick
= $user->getOption( "editsectiononrightclick" );
2219 $this->mNumberHeadings
= $user->getOption( "numberheadings" );
2220 $this->mShowToc
= $user->getOption( "showtoc" );
2226 # Regex callbacks, used in Parser::replaceVariables
2227 function wfBraceSubstitution( $matches )
2229 global $wgCurParser;
2230 return $wgCurParser->braceSubstitution( $matches );