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).
36 define( "MAX_INCLUDE_REPEAT", 5 );
38 # Recursion depth of variable/inclusion evaluation
39 define( "MAX_INCLUDE_PASSES", 3 );
41 # Allowed values for $mOutputType
42 define( "OT_HTML", 1 );
43 define( "OT_WIKI", 2 );
44 define( "OT_MSG", 3 );
46 # prefix for escaping, used in two functions at least
47 define( "UNIQ_PREFIX", "NaodW29");
51 # Cleared with clearState():
52 var $mOutput, $mAutonumber, $mLastSection, $mDTopen, $mStripState = array();
53 var $mVariables, $mIncludeCount;
56 var $mOptions, $mTitle, $mOutputType;
65 $this->mOutput
= new ParserOutput
;
66 $this->mAutonumber
= 0;
67 $this->mLastSection
= "";
68 $this->mDTopen
= false;
69 $this->mVariables
= false;
70 $this->mIncludeCount
= array();
71 $this->mStripState
= array();
74 # First pass--just handle <nowiki> sections, pass the rest off
75 # to doWikiPass2() which does all the real work.
77 # Returns a ParserOutput
79 function parse( $text, &$title, $options, $linestart = true, $clearState = true )
81 $fname = "Parser::parse";
82 wfProfileIn( $fname );
88 $this->mOptions
= $options;
89 $this->mTitle
=& $title;
90 $this->mOutputType
= OT_HTML
;
93 $text = $this->strip( $text, $this->mStripState
);
94 $text = $this->doWikiPass2( $text, $linestart );
95 $text = $this->unstrip( $text, $this->mStripState
);
97 $this->mOutput
->setText( $text );
98 wfProfileOut( $fname );
99 return $this->mOutput
;
102 /* static */ function getRandomString()
104 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
107 # Replaces all occurences of <$tag>content</$tag> in the text
108 # with a random marker and returns the new text. the output parameter
109 # $content will be an associative array filled with data on the form
110 # $unique_marker => content.
112 /* static */ function extractTags($tag, $text, &$content, $uniq_prefix = ""){
114 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
119 while ( "" != $text ) {
120 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
122 if ( ( count( $p ) < 2 ) ||
( "" == $p[1] ) ) {
125 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
126 $marker = $rnd . sprintf("%08X", $n++
);
127 $content[$marker] = $q[0];
128 $stripped .= $marker;
135 # Strips <nowiki>, <pre> and <math>
136 # Returns the text, and fills an array with data needed in unstrip()
138 function strip( $text, &$state )
140 $render = ($this->mOutputType
== OT_HTML
);
141 $nowiki_content = array();
142 $hiero_content = array();
143 $math_content = array();
144 $pre_content = array();
146 # Replace any instances of the placeholders
147 $uniq_prefix = UNIQ_PREFIX
;
148 $text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
150 $text = Parser
::extractTags("nowiki", $text, $nowiki_content, $uniq_prefix);
151 foreach( $nowiki_content as $marker => $content ){
153 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
155 $nowiki_content[$marker] = "<nowiki>$content</nowiki>";
159 if( $GLOBALS['wgUseWikiHiero'] ){
160 $text = Parser
::extractTags("hiero", $text, $hiero_content, $uniq_prefix);
161 foreach( $hiero_content as $marker => $content ){
163 $hiero_content[$marker] = WikiHiero( $content, WH_MODE_HTML
);
165 $hiero_content[$marker] = "<hiero>$content</hiero>";
170 if( $this->mOptions
->getUseTeX() ){
171 $text = Parser
::extractTags("math", $text, $math_content, $uniq_prefix);
172 foreach( $math_content as $marker => $content ){
174 $math_content[$marker] = renderMath( $content );
176 $math_content[$marker] = "<math>$content</math>";
181 $text = Parser
::extractTags("pre", $text, $pre_content, $uniq_prefix);
182 foreach( $pre_content as $marker => $content ){
184 $pre_content[$marker] = "<pre>" . wfEscapeHTMLTagsOnly( $content ) . "</pre>";
186 $pre_content[$marker] = "<pre>$content</pre>";
190 # Must expand in reverse order, otherwise nested tags will be corrupted
191 $state = array( $pre_content, $math_content, $hiero_content, $nowiki_content );
195 function unstrip( $text, &$state )
197 foreach( $state as $content_dict ){
198 foreach( $content_dict as $marker => $content ){
199 $text = str_replace( $marker, $content, $text );
205 function categoryMagic ()
207 global $wgLang , $wgUser ;
208 if ( !$this->mOptions
->getUseCategoryMagic() ) return ;
209 $id = $this->mTitle
->getArticleID() ;
210 $cat = $wgLang->ucfirst ( wfMsg ( "category" ) ) ;
211 $ti = $this->mTitle
->getText() ;
212 $ti = explode ( ":" , $ti , 2 ) ;
213 if ( $cat != $ti[0] ) return "" ;
214 $r = "<br break='all' />\n" ;
216 $articles = array() ;
217 $parents = array () ;
218 $children = array() ;
221 # $sk =& $this->mGetSkin();
222 $sk =& $wgUser->getSkin() ;
225 $sql1 = "SELECT DISTINCT cur_title,cur_namespace FROM cur,links WHERE l_to={$id} AND l_from=cur_id";
226 $sql2 = "SELECT DISTINCT cur_title,cur_namespace FROM cur,brokenlinks WHERE bl_to={$id} AND bl_from=cur_id" ;
228 $res = wfQuery ( $sql1, DB_READ
) ;
229 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
231 $res = wfQuery ( $sql2, DB_READ
) ;
232 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
235 foreach ( $data AS $x )
237 $t = $wgLang->getNsText ( $x->cur_namespace
) ;
238 if ( $t != "" ) $t .= ":" ;
239 $t .= $x->cur_title
;
241 $y = explode ( ":" , $t , 2 ) ;
242 if ( count ( $y ) == 2 && $y[0] == $cat ) {
243 array_push ( $children , $sk->makeLink ( $t , $y[1] ) ) ;
245 array_push ( $articles , $sk->makeLink ( $t ) ) ;
248 wfFreeResult ( $res ) ;
251 if ( count ( $children ) > 0 )
253 asort ( $children ) ;
254 $r .= "<h2>".wfMsg("subcategories")."</h2>\n" ;
255 $r .= implode ( ", " , $children ) ;
259 if ( count ( $articles ) > 0 )
261 asort ( $articles ) ;
262 $h = wfMsg( "category_header", $ti[1] );
263 $r .= "<h2>{$h}</h2>\n" ;
264 $r .= implode ( ", " , $articles ) ;
271 function getHTMLattrs ()
273 $htmlattrs = array( # Allowed attributes--no scripting, etc.
274 "title", "align", "lang", "dir", "width", "height",
275 "bgcolor", "clear", /* BR */ "noshade", /* HR */
276 "cite", /* BLOCKQUOTE, Q */ "size", "face", "color",
277 /* FONT */ "type", "start", "value", "compact",
278 /* For various lists, mostly deprecated but safe */
279 "summary", "width", "border", "frame", "rules",
280 "cellspacing", "cellpadding", "valign", "char",
281 "charoff", "colgroup", "col", "span", "abbr", "axis",
282 "headers", "scope", "rowspan", "colspan", /* Tables */
283 "id", "class", "name", "style" /* For CSS */
288 function fixTagAttributes ( $t )
290 if ( trim ( $t ) == "" ) return "" ; # Saves runtime ;-)
291 $htmlattrs = $this->getHTMLattrs() ;
293 # Strip non-approved attributes from the tag
295 "/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e",
296 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
298 # Strip javascript "expression" from stylesheets. Brute force approach:
299 # If anythin offensive is found, all attributes of the HTML tag are dropped
302 "/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is",
303 wfMungeToUtf8( $t ) ) )
311 function doTableStuff ( $t )
313 $t = explode ( "\n" , $t ) ;
314 $td = array () ; # Is currently a td tag open?
315 $ltd = array () ; # Was it TD or TH?
316 $tr = array () ; # Is currently a tr tag open?
317 $ltr = array () ; # tr attributes
318 foreach ( $t AS $k => $x )
321 $fc = substr ( $x , 0 , 1 ) ;
322 if ( "{|" == substr ( $x , 0 , 2 ) )
324 $t[$k] = "\n<table " . $this->fixTagAttributes ( substr ( $x , 3 ) ) . ">" ;
325 array_push ( $td , false ) ;
326 array_push ( $ltd , "" ) ;
327 array_push ( $tr , false ) ;
328 array_push ( $ltr , "" ) ;
330 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
331 else if ( "|}" == substr ( $x , 0 , 2 ) )
334 $l = array_pop ( $ltd ) ;
335 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
336 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
340 /* else if ( "|_" == substr ( $x , 0 , 2 ) ) # Caption
342 $z = trim ( substr ( $x , 2 ) ) ;
343 $t[$k] = "<caption>{$z}</caption>\n" ;
345 else if ( "|-" == substr ( $x , 0 , 2 ) ) # Allows for |---------------
347 $x = substr ( $x , 1 ) ;
348 while ( $x != "" && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
350 $l = array_pop ( $ltd ) ;
351 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
352 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
355 array_push ( $tr , false ) ;
356 array_push ( $td , false ) ;
357 array_push ( $ltd , "" ) ;
358 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
360 else if ( "|" == $fc ||
"!" == $fc ||
"|+" == substr ( $x , 0 , 2 ) ) # Caption
362 if ( "|+" == substr ( $x , 0 , 2 ) )
365 $x = substr ( $x , 1 ) ;
367 $after = substr ( $x , 1 ) ;
368 if ( $fc == "!" ) $after = str_replace ( "!!" , "||" , $after ) ;
369 $after = explode ( "||" , $after ) ;
371 foreach ( $after AS $theline )
376 $tra = array_pop ( $ltr ) ;
377 if ( !array_pop ( $tr ) ) $z = "<tr {$tra}>\n" ;
378 array_push ( $tr , true ) ;
379 array_push ( $ltr , "" ) ;
382 $l = array_pop ( $ltd ) ;
383 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
384 if ( $fc == "|" ) $l = "td" ;
385 else if ( $fc == "!" ) $l = "th" ;
386 else if ( $fc == "+" ) $l = "caption" ;
388 array_push ( $ltd , $l ) ;
389 $y = explode ( "|" , $theline , 2 ) ;
390 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
391 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
393 array_push ( $td , true ) ;
398 # Closing open td, tr && table
399 while ( count ( $td ) > 0 )
401 if ( array_pop ( $td ) ) $t[] = "</td>" ;
402 if ( array_pop ( $tr ) ) $t[] = "</tr>" ;
406 $t = implode ( "\n" , $t ) ;
407 # $t = $this->removeHTMLtags( $t );
411 # Well, OK, it's actually about 14 passes. But since all the
412 # hard lifting is done inside PHP's regex code, it probably
413 # wouldn't speed things up much to add a real parser.
415 function doWikiPass2( $text, $linestart )
417 $fname = "Parser::doWikiPass2";
418 wfProfileIn( $fname );
420 $text = $this->removeHTMLtags( $text );
421 $text = $this->replaceVariables( $text );
423 # $text = preg_replace( "/(^|\n)-----*/", "\\1<hr>", $text );
425 $text = $this->doHeadings( $text );
427 if($this->mOptions
->getUseDynamicDates()) {
428 global $wgDateFormatter;
429 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
432 $text = $this->replaceExternalLinks( $text );
433 $text = $this->doBlockLevels( $text, $linestart );
434 $text = $this->doTokenizedParser ( $text );
435 $text = $this->doTableStuff ( $text ) ;
437 $text = $this->formatHeadings( $text );
439 $sk =& $this->mOptions
->getSkin();
440 $text = $sk->transformContent( $text );
442 "/<hr *>/i" => '<hr/>',
443 "/<br *>/i" => '<br/>',
444 "/<center *>/i"=>'<span style="text-align:center;">',
445 "/<\\/center *>/i" => '</span>'
447 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
449 $text .= $this->categoryMagic () ;
451 wfProfileOut( $fname );
456 /* private */ function doHeadings( $text )
458 for ( $i = 6; $i >= 1; --$i ) {
459 $h = substr( "======", 0, $i );
460 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
461 "<h{$i}>\\1</h{$i}>\\2", $text );
466 # Note: we have to do external links before the internal ones,
467 # and otherwise take great care in the order of things here, so
468 # that we don't end up interpreting some URLs twice.
470 /* private */ function replaceExternalLinks( $text )
472 $fname = "Parser::replaceExternalLinks";
473 wfProfileIn( $fname );
474 $text = $this->subReplaceExternalLinks( $text, "http", true );
475 $text = $this->subReplaceExternalLinks( $text, "https", true );
476 $text = $this->subReplaceExternalLinks( $text, "ftp", false );
477 $text = $this->subReplaceExternalLinks( $text, "irc", false );
478 $text = $this->subReplaceExternalLinks( $text, "gopher", false );
479 $text = $this->subReplaceExternalLinks( $text, "news", false );
480 $text = $this->subReplaceExternalLinks( $text, "mailto", false );
481 wfProfileOut( $fname );
485 /* private */ function subReplaceExternalLinks( $s, $protocol, $autonumber )
487 $unique = "4jzAfzB8hNvf4sqyO9Edd8pSmk9rE2in0Tgw3";
488 $uc = "A-Za-z0-9_\\/~%\\-+&*#?!=()@\\x80-\\xFF";
490 # this is the list of separators that should be ignored if they
491 # are the last character of an URL but that should be included
492 # if they occur within the URL, e.g. "go to www.foo.com, where .."
493 # in this case, the last comma should not become part of the URL,
494 # but in "www.foo.com/123,2342,32.htm" it should.
496 $fnc = "A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF";
497 $images = "gif|png|jpg|jpeg";
499 # PLEASE NOTE: The curly braces { } are not part of the regex,
500 # they are interpreted as part of the string (used to tell PHP
501 # that the content of the string should be inserted there).
502 $e1 = "/(^|[^\\[])({$protocol}:)([{$uc}{$sep}]+)\\/([{$fnc}]+)\\." .
503 "((?i){$images})([^{$uc}]|$)/";
505 $e2 = "/(^|[^\\[])({$protocol}:)(([".$uc."]|[".$sep."][".$uc."])+)([^". $uc . $sep. "]|[".$sep."]|$)/";
506 $sk =& $this->mOptions
->getSkin();
508 if ( $autonumber and $this->mOptions
->getAllowExternalImages() ) { # Use img tags only for HTTP urls
509 $s = preg_replace( $e1, "\\1" . $sk->makeImage( "{$unique}:\\3" .
510 "/\\4.\\5", "\\4.\\5" ) . "\\6", $s );
512 $s = preg_replace( $e2, "\\1" . "<a href=\"{$unique}:\\3\"" .
513 $sk->getExternalLinkAttributes( "{$unique}:\\3", wfEscapeHTML(
514 "{$unique}:\\3" ) ) . ">" . wfEscapeHTML( "{$unique}:\\3" ) .
516 $s = str_replace( $unique, $protocol, $s );
518 $a = explode( "[{$protocol}:", " " . $s );
519 $s = array_shift( $a );
520 $s = substr( $s, 1 );
522 $e1 = "/^([{$uc}"."{$sep}]+)](.*)\$/sD";
523 $e2 = "/^([{$uc}"."{$sep}]+)\\s+([^\\]]+)](.*)\$/sD";
525 foreach ( $a as $line ) {
526 if ( preg_match( $e1, $line, $m ) ) {
527 $link = "{$protocol}:{$m[1]}";
529 if ( $autonumber ) { $text = "[" . ++
$this->mAutonumber
. "]"; }
530 else { $text = wfEscapeHTML( $link ); }
531 } else if ( preg_match( $e2, $line, $m ) ) {
532 $link = "{$protocol}:{$m[1]}";
536 $s .= "[{$protocol}:" . $line;
539 if( $link == $text ||
preg_match( "!$protocol://" . preg_quote( $text, "/" ) . "/?$!", $link ) ) {
542 # Expand the URL for printable version
543 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $link ) . "</i>)</span>";
545 $la = $sk->getExternalLinkAttributes( $link, $text );
546 $s .= "<a href='{$link}'{$la}>{$text}</a>{$paren}{$trail}";
552 /* private */ function handle3Quotes( &$state, $token )
554 if ( $state["strong"] !== false ) {
555 if ( $state["em"] !== false && $state["em"] > $state["strong"] )
557 # ''' lala ''lala '''
558 $s = "</em></strong><em>";
562 $state["strong"] = FALSE;
565 $state["strong"] = $token["pos"];
570 /* private */ function handle2Quotes( &$state, $token )
572 if ( $state["em"] !== false ) {
573 if ( $state["strong"] !== false && $state["strong"] > $state["em"] )
575 # ''lala'''lala'' ....'''
576 $s = "</strong></em><strong>";
580 $state["em"] = FALSE;
583 $state["em"] = $token["pos"];
588 /* private */ function handle5Quotes( &$state, $token )
591 if ( $state["em"] !== false && $state["strong"] ) {
592 if ( $state["em"] < $state["strong"] ) {
593 $s .= "</strong></em>";
595 $s .= "</em></strong>";
597 $state["strong"] = $state["em"] = FALSE;
598 } elseif ( $state["em"] !== false ) {
599 $s .= "</em><strong>";
600 $state["em"] = FALSE;
601 $state["strong"] = $token["pos"];
602 } elseif ( $state["strong"] !== false ) {
603 $s .= "</strong><em>";
604 $state["strong"] = FALSE;
605 $state["em"] = $token["pos"];
606 } else { # not $em and not $strong
607 $s .= "<strong><em>";
608 $state["strong"] = $state["em"] = $token["pos"];
613 /* private */ function doTokenizedParser( $str )
615 global $wgLang; # for language specific parser hook
617 $tokenizer=Tokenizer
::newFromString( $str );
618 $tokenStack = array();
621 $state["em"] = FALSE;
622 $state["strong"] = FALSE;
626 # The tokenizer splits the text into tokens and returns them one by one.
627 # Every call to the tokenizer returns a new token.
628 while ( $token = $tokenizer->nextToken() )
630 switch ( $token["type"] )
633 # simple text with no further markup
634 $txt = $token["text"];
637 # remember the tag opened with 3 [
641 # FIXME : Treat orphaned open tags (stack not empty when text is over)
643 array_push( $tokenStack, $token );
650 # get text from stack, glue it together, and call the code to handle a
653 if ( count( $tokenStack ) == 0 )
655 # stack empty. Found a ]] without an opening [[
659 $lastToken = array_pop( $tokenStack );
660 while ( !(($lastToken["type"] == "[[[") or ($lastToken["type"] == "[[")) )
662 if( !empty( $lastToken["text"] ) ) {
663 $linkText = $lastToken["text"] . $linkText;
665 $lastToken = array_pop( $tokenStack );
668 $txt = $linkText ."]]";
670 if( isset( $lastToken["text"] ) ) {
671 $prefix = $lastToken["text"];
675 $nextToken = $tokenizer->previewToken();
676 if ( $nextToken["type"] == "text" )
678 # Preview just looks at it. Now we have to fetch it.
679 $nextToken = $tokenizer->nextToken();
680 $txt .= $nextToken["text"];
682 $fakestate = $this->mStripState
;
683 $txt = $this->handleInternalLink( $this->unstrip($txt,$fakestate), $prefix );
685 # did the tag start with 3 [ ?
687 # show the first as text
693 $tagIsOpen = (count( $tokenStack ) != 0);
699 # This and the three next ones handle quotes
700 $txt = $this->handle3Quotes( $state, $token );
703 $txt = $this->handle2Quotes( $state, $token );
706 $txt = $this->handle5Quotes( $state, $token );
716 $txt = $this->doMagicRFC( $tokenizer );
723 $txt = $this->doMagicISBN( $tokenizer );
727 # Call language specific Hook.
728 $txt = $wgLang->processToken( $token, $tokenStack );
729 if ( NULL == $txt ) {
730 # An unkown token. Highlight.
731 $txt = "<font color=\"#FF0000\"><b>".$token["type"]."</b></font>";
732 $txt .= "<font color=\"#FFFF00\"><b>".$token["text"]."</b></font>";
736 # If we're parsing the interior of a link, don't append the interior to $s,
737 # but push it to the stack so it can be processed when a ]] token is found.
738 if ( $tagIsOpen && $txt != "" ) {
739 $token["type"] = "text";
740 $token["text"] = $txt;
741 array_push( $tokenStack, $token );
746 if ( count( $tokenStack ) != 0 )
748 # still objects on stack. opened [[ tag without closing ]] tag.
750 while ( $lastToken = array_pop( $tokenStack ) )
752 if ( $lastToken["type"] == "text" )
754 $txt = $lastToken["text"] . $txt;
756 $txt = $lastToken["type"] . $txt;
764 /* private */ function handleInternalLink( $line, $prefix )
766 global $wgLang, $wgLinkCache;
767 global $wgNamespacesWithSubpages, $wgLanguageCode;
768 static $fname = "Parser::handleInternalLink" ;
769 wfProfileIn( $fname );
771 wfProfileIn( "$fname-setup" );
773 if ( !$tc ) { $tc = Title
::legalChars() . "#"; }
774 $sk =& $this->mOptions
->getSkin();
776 # Match a link having the form [[namespace:link|alternate]]trail
778 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
779 # Match the end of a line for a word that's not followed by whitespace,
780 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
781 #$e2 = "/^(.*)\\b(\\w+)\$/suD";
782 #$e2 = "/^(.*\\s)(\\S+)\$/suD";
783 static $e2 = '/^(.*\s)([a-zA-Z\x80-\xff]+)$/sD';
786 # Special and Media are pseudo-namespaces; no pages actually exist in them
787 static $image = FALSE;
788 static $special = FALSE;
789 static $media = FALSE;
790 static $category = FALSE;
791 if ( !$image ) { $image = Namespace::getImage(); }
792 if ( !$special ) { $special = Namespace::getSpecial(); }
793 if ( !$media ) { $media = Namespace::getMedia(); }
794 if ( !$category ) { $category = wfMsg ( "category" ) ; }
796 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
798 wfProfileOut( "$fname-setup" );
801 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
804 } else { # Invalid form; output directly
805 $s .= $prefix . "[[" . $line ;
811 :Foobar -- override special treatment of prefix (images, language links)
812 /Foobar -- convert to CurrentPage/Foobar
813 /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
815 $c = substr($m[1],0,1);
816 $noforce = ($c != ":");
817 if( $c == "/" ) { # subpage
818 if(substr($m[1],-1,1)=="/") { # / at end means we don't want the slash to be shown
819 $m[1]=substr($m[1],1,strlen($m[1])-2);
822 $noslash=substr($m[1],1);
824 if($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]) { # subpages allowed here
825 $link = $this->mTitle
->getPrefixedText(). "/" . trim($noslash);
828 } # this might be changed for ugliness reasons
830 $link = $noslash; # no subpage allowed, use standard link
832 } elseif( $noforce ) { # no subpage
835 $link = substr( $m[1], 1 );
840 $nt = Title
::newFromText( $link );
842 $s .= $prefix . "[[" . $line;
845 $ns = $nt->getNamespace();
846 $iw = $nt->getInterWiki();
848 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
849 array_push( $this->mOutput
->mLanguageLinks
, $nt->getPrefixedText() );
850 return (trim($s) == '')?
'': $s;
852 if( $ns == $image ) {
853 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
854 $wgLinkCache->addImageLinkObj( $nt );
858 if( ( $nt->getPrefixedText() == $this->mTitle
->getPrefixedText() ) &&
859 ( strpos( $link, "#" ) == FALSE ) ) {
860 $s .= $prefix . "<strong>" . $text . "</strong>" . $trail;
865 $catns = strtoupper ( $nt->getDBkey () ) ;
866 $catns = explode ( ":" , $catns ) ;
867 if ( count ( $catns ) > 1 ) $catns = array_shift ( $catns ) ;
869 if ( $catns == strtoupper($category) && $this->mOptions
->getUseCategoryMagic() ) {
870 $t = explode ( ":" , $nt->getText() ) ;
872 $t = implode ( ":" , $t ) ;
873 $t = $wgLang->ucFirst ( $t ) ;
874 $nnt = Title
::newFromText ( $category.":".$t ) ;
875 $t = $sk->makeLinkObj( $nnt, $t, "", $trail , $prefix );
876 $this->mOutput
->mCategoryLinks
[] = $t ;
877 $s .= $prefix . $trail ;
881 if( $ns == $media ) {
882 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
883 $wgLinkCache->addImageLinkObj( $nt );
885 } elseif( $ns == $special ) {
886 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, "", $trail );
889 $s .= $sk->makeLinkObj( $nt, $text, "", $trail , $prefix );
891 wfProfileOut( $fname );
895 # Some functions here used by doBlockLevels()
897 /* private */ function closeParagraph()
900 if ( '' != $this->mLastSection
) {
901 $result = "</" . $this->mLastSection
. ">";
903 $this->mLastSection
= "";
906 # getCommon() returns the length of the longest common substring
907 # of both arguments, starting at the beginning of both.
909 /* private */ function getCommon( $st1, $st2 )
911 $fl = strlen( $st1 );
912 $shorter = strlen( $st2 );
913 if ( $fl < $shorter ) { $shorter = $fl; }
915 for ( $i = 0; $i < $shorter; ++
$i ) {
916 if ( $st1{$i} != $st2{$i} ) { break; }
920 # These next three functions open, continue, and close the list
921 # element appropriate to the prefix character passed into them.
923 /* private */ function openList( $char )
925 $result = $this->closeParagraph();
927 if ( "*" == $char ) { $result .= "<ul><li>"; }
928 else if ( "#" == $char ) { $result .= "<ol><li>"; }
929 else if ( ":" == $char ) { $result .= "<dl><dd>"; }
930 else if ( ";" == $char ) {
931 $result .= "<dl><dt>";
932 $this->mDTopen
= true;
934 else { $result = "<!-- ERR 1 -->"; }
939 /* private */ function nextItem( $char )
941 if ( "*" == $char ||
"#" == $char ) { return "</li><li>"; }
942 else if ( ":" == $char ||
";" == $char ) {
944 if ( $this->mDTopen
) { $close = "</dt>"; }
945 if ( ";" == $char ) {
946 $this->mDTopen
= true;
947 return $close . "<dt>";
949 $this->mDTopen
= false;
950 return $close . "<dd>";
953 return "<!-- ERR 2 -->";
956 /* private */function closeList( $char )
958 if ( "*" == $char ) { $text = "</li></ul>"; }
959 else if ( "#" == $char ) { $text = "</li></ol>"; }
960 else if ( ":" == $char ) {
961 if ( $this->mDTopen
) {
962 $this->mDTopen
= false;
963 $text = "</dt></dl>";
965 $text = "</dd></dl>";
968 else { return "<!-- ERR 3 -->"; }
972 /* private */ function doBlockLevels( $text, $linestart )
974 $fname = "Parser::doBlockLevels";
975 wfProfileIn( $fname );
976 # Parsing through the text line by line. The main thing
977 # happening here is handling of block-level elements p, pre,
978 # and making lists from lines starting with * # : etc.
980 $a = explode( "\n", $text );
981 $lastPref = $text = '';
982 $this->mDTopen
= $inBlockElem = false;
984 if ( ! $linestart ) { $text .= array_shift( $a ); }
985 foreach ( $a as $t ) {
986 if ( "" != $text ) { $text .= "\n"; }
989 $opl = strlen( $lastPref );
990 $npl = strspn( $t, "*#:;" );
991 $pref = substr( $t, 0, $npl );
992 $pref2 = str_replace( ";", ":", $pref );
993 $t = substr( $t, $npl );
995 if ( 0 != $npl && 0 == strcmp( $lastPref, $pref2 ) ) {
996 $text .= $this->nextItem( substr( $pref, -1 ) );
998 if ( ";" == substr( $pref, -1 ) ) {
999 $cpos = strpos( $t, ":" );
1000 if ( ! ( false === $cpos ) ) {
1001 $term = substr( $t, 0, $cpos );
1002 $text .= $term . $this->nextItem( ":" );
1003 $t = substr( $t, $cpos +
1 );
1006 } else if (0 != $npl ||
0 != $opl) {
1007 $cpl = $this->getCommon( $pref, $lastPref );
1009 while ( $cpl < $opl ) {
1010 $text .= $this->closeList( $lastPref{$opl-1} );
1013 if ( $npl <= $cpl && $cpl > 0 ) {
1014 $text .= $this->nextItem( $pref{$cpl-1} );
1016 while ( $npl > $cpl ) {
1017 $char = substr( $pref, $cpl, 1 );
1018 $text .= $this->openList( $char );
1020 if ( ";" == $char ) {
1021 $cpos = strpos( $t, ":" );
1022 if ( ! ( false === $cpos ) ) {
1023 $term = substr( $t, 0, $cpos );
1024 $text .= $term . $this->nextItem( ":" );
1025 $t = substr( $t, $cpos +
1 );
1032 if ( 0 == $npl ) { # No prefix--go to paragraph mode
1033 $uniq_prefix = UNIQ_PREFIX
;
1035 "/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<div)/i", $t ) ) {
1036 $text .= $this->closeParagraph();
1037 $inBlockElem = true;
1038 } else if ( preg_match("/(<hr|<\\/td|".$uniq_prefix."-pre)/i", $t ) ) {
1039 $text .= $this->closeParagraph();
1040 $inBlockElem = false;
1042 if ( ! $inBlockElem ) {
1043 if ( " " == $t{0} ) {
1044 $newSection = "pre";
1045 $text .= $this->closeParagraph();
1046 # $t = wfEscapeHTML( $t );
1048 else { $newSection = "p"; }
1050 if ( ( '' == trim( $oLine ) ) ||
( $this->mLastSection
== $newSection and $newSection != 'p' )) {
1051 $text .= $this->closeParagraph();
1052 $text .= "<" . $newSection . ">";
1053 $this->mLastSection
= $newSection;
1056 if ( $inBlockElem &&
1057 preg_match( "/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|<\\/p<\\/div)/i", $t ) ) {
1058 $inBlockElem = false;
1064 $text .= $this->closeList( $pref2{$npl-1} );
1067 if ( "" != $this->mLastSection
) {
1068 $text .= "</" . $this->mLastSection
. ">";
1069 $this->mLastSection
= "";
1071 wfProfileOut( $fname );
1075 function getVariableValue( $index ) {
1076 global $wgLang, $wgSitename, $wgServer;
1079 case MAG_CURRENTMONTH
:
1081 case MAG_CURRENTMONTHNAME
:
1082 return $wgLang->getMonthName( date("n") );
1083 case MAG_CURRENTMONTHNAMEGEN
:
1084 return $wgLang->getMonthNameGen( date("n") );
1085 case MAG_CURRENTDAY
:
1087 case MAG_CURRENTDAYNAME
:
1088 return $wgLang->getWeekdayName( date("w")+
1 );
1089 case MAG_CURRENTYEAR
:
1091 case MAG_CURRENTTIME
:
1092 return $wgLang->time( wfTimestampNow(), false );
1093 case MAG_NUMBEROFARTICLES
:
1094 return wfNumberOfArticles();
1104 function initialiseVariables()
1106 global $wgVariableIDs;
1107 $this->mVariables
= array();
1108 foreach ( $wgVariableIDs as $id ) {
1109 $mw =& MagicWord
::get( $id );
1110 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1114 /* private */ function replaceVariables( $text )
1116 global $wgLang, $wgCurParser;
1117 global $wgScript, $wgArticlePath;
1119 $fname = "Parser::replaceVariables";
1120 wfProfileIn( $fname );
1123 if ( !$this->mVariables
) {
1124 $this->initialiseVariables();
1126 $titleChars = Title
::legalChars();
1127 $regex = "/{{([$titleChars\\|]*?)}}/s";
1129 # "Recursive" variable expansion: run it through a couple of passes
1130 for ( $i=0; $i<MAX_INCLUDE_REPEAT
&& !$bail; $i++
) {
1133 # It's impossible to rebind a global in PHP
1134 # Instead, we run the substitution on a copy, then merge the changed fields back in
1135 $wgCurParser = $this->fork();
1137 $text = preg_replace_callback( $regex, "wfBraceSubstitution", $text );
1138 if ( $oldText == $text ) {
1141 $this->merge( $wgCurParser );
1147 # Returns a copy of this object except with various variables cleared
1148 # This copy can be re-merged with the parent after operations on the copy
1152 $copy->mOutput
= new ParserOutput
;
1156 # Merges a copy split off with fork()
1157 function merge( &$copy )
1159 $this->mOutput
->merge( $copy->mOutput
);
1161 # Merge include throttling arrays
1162 foreach( $copy->mIncludeCount
as $dbk => $count ) {
1163 if ( array_key_exists( $dbk, $this->mIncludeCount
) ) {
1164 $this->mIncludeCount
[$dbk] +
= $count;
1166 $this->mIncludeCount
[$dbk] = $count;
1171 function braceSubstitution( $matches )
1173 global $wgLinkCache, $wgLang;
1174 $fname = "Parser::braceSubstitution";
1178 $text = $matches[1];
1181 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1182 if ( $mwSubst->matchStartAndRemove( $text ) ) {
1183 if ( $this->mOutputType
!= OT_WIKI
) {
1184 # Invalid SUBST not replaced at PST time
1185 # Return without further processing
1186 $text = $matches[0];
1189 } elseif ( $this->mOutputType
== OT_WIKI
) {
1190 # SUBST not found in PST pass, do nothing
1191 $text = $matches[0];
1195 # MSG, MSGNW and INT
1198 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1199 if ( $mwMsgnw->matchStartAndRemove( $text ) ) {
1202 # Remove obsolete MSG:
1203 $mwMsg =& MagicWord
::get( MAG_MSG
);
1204 $mwMsg->matchStartAndRemove( $text );
1207 # Check if it is an internal message
1208 $mwInt =& MagicWord
::get( MAG_INT
);
1209 if ( $mwInt->matchStartAndRemove( $text ) ) {
1210 $text = wfMsg( $text );
1217 # Check for NS: (namespace expansion)
1218 $mwNs = MagicWord
::get( MAG_NS
);
1219 if ( $mwNs->matchStartAndRemove( $text ) ) {
1220 if ( intval( $text ) ) {
1221 $text = $wgLang->getNsText( intval( $text ) );
1224 $index = Namespace::getCanonicalIndex( strtolower( $text ) );
1225 if ( !is_null( $index ) ) {
1226 $text = $wgLang->getNsText( $index );
1233 # LOCALURL and LOCALURLE
1235 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1236 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1238 if ( $mwLocal->matchStartAndRemove( $text ) ) {
1239 $func = 'getLocalURL';
1240 } elseif ( $mwLocalE->matchStartAndRemove( $text ) ) {
1241 $func = 'escapeLocalURL';
1246 if ( $func !== '' ) {
1247 $args = explode( "|", $text );
1248 $n = count( $args );
1250 $title = Title
::newFromText( $args[0] );
1251 if ( !is_null( $title ) ) {
1253 $text = $title->$func( $args[1] );
1255 $text = $title->$func();
1263 # Check for a match against internal variables
1264 if ( !$found && array_key_exists( $text, $this->mVariables
) ) {
1265 $text = $this->mVariables
[$text];
1267 $this->mOutput
->mContainsOldMagic
= true;
1270 # Load from database
1272 $title = Title
::newFromText( $text, NS_TEMPLATE
);
1273 if ( is_object( $title ) && !$title->isExternal() ) {
1274 # Check for excessive inclusion
1275 $dbk = $title->getPrefixedDBkey();
1276 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1277 $this->mIncludeCount
[$dbk] = 0;
1279 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1280 $article = new Article( $title );
1281 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1282 if ( $articleContent !== false ) {
1284 $text = $articleContent;
1286 # Escaping and link table handling
1287 # Not required for preSaveTransform()
1288 if ( $this->mOutputType
== OT_HTML
) {
1290 $text = wfEscapeWikiText( $text );
1292 $text = $this->removeHTMLtags( $text );
1294 $wgLinkCache->suspend();
1295 $text = $this->doTokenizedParser( $text );
1296 $wgLinkCache->resume();
1297 $wgLinkCache->addLinkObj( $title );
1303 # If the title is valid but undisplayable, make a link to it
1304 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1305 $text = "[[" . $title->getPrefixedText() . "]]";
1318 # Cleans up HTML, removes dangerous tags and attributes
1319 /* private */ function removeHTMLtags( $text )
1321 $fname = "Parser::removeHTMLtags";
1322 wfProfileIn( $fname );
1323 $htmlpairs = array( # Tags that must be closed
1324 "b", "i", "u", "font", "big", "small", "sub", "sup", "h1",
1325 "h2", "h3", "h4", "h5", "h6", "cite", "code", "em", "s",
1326 "strike", "strong", "tt", "var", "div", "center",
1327 "blockquote", "ol", "ul", "dl", "table", "caption", "pre",
1328 "ruby", "rt" , "rb" , "rp", "p"
1330 $htmlsingle = array(
1331 "br", "hr", "li", "dt", "dd"
1333 $htmlnest = array( # Tags that can be nested--??
1334 "table", "tr", "td", "th", "div", "blockquote", "ol", "ul",
1335 "dl", "font", "big", "small", "sub", "sup"
1337 $tabletags = array( # Can only appear inside table
1341 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1342 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1344 $htmlattrs = $this->getHTMLattrs () ;
1346 # Remove HTML comments
1347 $text = preg_replace( "/<!--.*-->/sU", "", $text );
1349 $bits = explode( "<", $text );
1350 $text = array_shift( $bits );
1351 $tagstack = array(); $tablestack = array();
1353 foreach ( $bits as $x ) {
1354 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
1355 preg_match( "/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/",
1357 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1358 error_reporting( $prev );
1361 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1365 if ( ! in_array( $t, $htmlsingle ) &&
1366 ( $ot = array_pop( $tagstack ) ) != $t ) {
1367 array_push( $tagstack, $ot );
1370 if ( $t == "table" ) {
1371 $tagstack = array_pop( $tablestack );
1376 # Keep track for later
1377 if ( in_array( $t, $tabletags ) &&
1378 ! in_array( "table", $tagstack ) ) {
1380 } else if ( in_array( $t, $tagstack ) &&
1381 ! in_array ( $t , $htmlnest ) ) {
1383 } else if ( ! in_array( $t, $htmlsingle ) ) {
1384 if ( $t == "table" ) {
1385 array_push( $tablestack, $tagstack );
1386 $tagstack = array();
1388 array_push( $tagstack, $t );
1390 # Strip non-approved attributes from the tag
1391 $newparams = $this->fixTagAttributes($params);
1395 $rest = str_replace( ">", ">", $rest );
1396 $text .= "<$slash$t $newparams$brace$rest";
1400 $text .= "<" . str_replace( ">", ">", $x);
1402 # Close off any remaining tags
1403 while ( $t = array_pop( $tagstack ) ) {
1405 if ( $t == "table" ) { $tagstack = array_pop( $tablestack ); }
1407 wfProfileOut( $fname );
1413 * This function accomplishes several tasks:
1414 * 1) Auto-number headings if that option is enabled
1415 * 2) Add an [edit] link to sections for logged in users who have enabled the option
1416 * 3) Add a Table of contents on the top for users who have enabled the option
1417 * 4) Auto-anchor headings
1419 * It loops through all headlines, collects the necessary data, then splits up the
1420 * string and re-inserts the newly formatted headlines.
1424 /* private */ function formatHeadings( $text )
1426 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
1427 $doShowToc = $this->mOptions
->getShowToc();
1428 if( !$this->mTitle
->userCanEdit() ) {
1430 $rightClickHack = 0;
1432 $showEditLink = $this->mOptions
->getEditSection();
1433 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
1436 # Inhibit editsection links if requested in the page
1437 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
1438 if( $esw->matchAndRemove( $text ) ) {
1441 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1443 $mw =& MagicWord
::get( MAG_NOTOC
);
1444 if( $mw->matchAndRemove( $text ) ) {
1448 # never add the TOC to the Main Page. This is an entry page that should not
1449 # be more than 1-2 screens large anyway
1450 if( $this->mTitle
->getPrefixedText() == wfMsg("mainpage") ) {
1454 # Get all headlines for numbering them and adding funky stuff like [edit]
1455 # links - this is for later, but we need the number of headlines right now
1456 $numMatches = preg_match_all( "/<H([1-6])(.*?" . ">)(.*?)<\/H[1-6]>/i", $text, $matches );
1458 # if there are fewer than 4 headlines in the article, do not show TOC
1459 if( $numMatches < 4 ) {
1463 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
1464 # override above conditions and always show TOC
1465 $mw =& MagicWord
::get( MAG_FORCETOC
);
1466 if ($mw->matchAndRemove( $text ) ) {
1471 # We need this to perform operations on the HTML
1472 $sk =& $this->mOptions
->getSkin();
1477 # Ugh .. the TOC should have neat indentation levels which can be
1478 # passed to the skin functions. These are determined here
1483 $sublevelCount = array();
1486 foreach( $matches[3] as $headline ) {
1489 $prevlevel = $level;
1491 $level = $matches[1][$headlineCount];
1492 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
1493 # reset when we enter a new level
1494 $sublevelCount[$level] = 0;
1495 $toc .= $sk->tocIndent( $level - $prevlevel );
1496 $toclevel +
= $level - $prevlevel;
1498 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
1499 # reset when we step back a level
1500 $sublevelCount[$level+
1]=0;
1501 $toc .= $sk->tocUnindent( $prevlevel - $level );
1502 $toclevel -= $prevlevel - $level;
1504 # count number of headlines for each level
1505 @$sublevelCount[$level]++
;
1506 if( $doNumberHeadings ||
$doShowToc ) {
1508 for( $i = 1; $i <= $level; $i++
) {
1509 if( !empty( $sublevelCount[$i] ) ) {
1513 $numbering .= $sublevelCount[$i];
1519 # The canonized header is a version of the header text safe to use for links
1520 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
1521 $canonized_headline = Parser
::unstrip( $headline, $this->mStripState
);
1524 $canonized_headline = preg_replace( "/<.*?" . ">/","",$canonized_headline );
1525 $tocline = trim( $canonized_headline );
1526 $canonized_headline = preg_replace("/[ &\\/<>\\(\\)\\[\\]=,+']+/", '_', html_entity_decode( $tocline));
1527 $refer[$headlineCount] = $canonized_headline;
1529 # count how many in assoc. array so we can track dupes in anchors
1530 @$refers[$canonized_headline]++
;
1531 $refcount[$headlineCount]=$refers[$canonized_headline];
1533 # Prepend the number to the heading text
1535 if( $doNumberHeadings ||
$doShowToc ) {
1536 $tocline = $numbering . " " . $tocline;
1538 # Don't number the heading if it is the only one (looks silly)
1539 if( $doNumberHeadings && count( $matches[3] ) > 1) {
1540 # the two are different if the line contains a link
1541 $headline=$numbering . " " . $headline;
1545 # Create the anchor for linking from the TOC to the section
1546 $anchor = $canonized_headline;
1547 if($refcount[$headlineCount] > 1 ) {
1548 $anchor .= "_" . $refcount[$headlineCount];
1551 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
1553 if( $showEditLink ) {
1554 if ( empty( $head[$headlineCount] ) ) {
1555 $head[$headlineCount] = "";
1557 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+
1);
1560 # Add the edit section span
1561 if( $rightClickHack ) {
1562 $headline = $sk->editSectionScript($headlineCount+
1,$headline);
1565 # give headline the correct <h#> tag
1566 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline."</h".$level.">";
1572 $toclines = $headlineCount;
1573 $toc .= $sk->tocUnindent( $toclevel );
1574 $toc = $sk->tocTable( $toc );
1577 # split up and insert constructed headlines
1579 $blocks = preg_split( "/<H[1-6].*?" . ">.*?<\/H[1-6]>/i", $text );
1582 foreach( $blocks as $block ) {
1583 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
1584 # This is the [edit] link that appears for the top block of text when
1585 # section editing is enabled
1586 $full .= $sk->editSectionLink(0);
1589 if( $doShowToc && !$i) {
1590 # Top anchor now in skin
1594 if( !empty( $head[$i] ) ) {
1603 /* private */ function doMagicISBN( &$tokenizer )
1607 # Check whether next token is a text token
1608 # If yes, fetch it and convert the text into a
1609 # Special::BookSources link
1610 $token = $tokenizer->previewToken();
1611 while ( $token["type"] == "" )
1613 $tokenizer->nextToken();
1614 $token = $tokenizer->previewToken();
1616 if ( $token["type"] == "text" )
1618 $token = $tokenizer->nextToken();
1619 $x = $token["text"];
1620 $valid = "0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ";
1622 $isbn = $blank = "" ;
1623 while ( " " == $x{0} ) {
1625 $x = substr( $x, 1 );
1627 while ( strstr( $valid, $x{0} ) != false ) {
1629 $x = substr( $x, 1 );
1631 $num = str_replace( "-", "", $isbn );
1632 $num = str_replace( " ", "", $num );
1635 $text = "ISBN $blank$x";
1637 $titleObj = Title
::makeTitle( NS_SPECIAL
, "Booksources" );
1638 $text = "<a href=\"" .
1639 $titleObj->escapeLocalUrl( "isbn={$num}" ) .
1640 "\" class=\"internal\">ISBN $isbn</a>";
1648 /* private */ function doMagicRFC( &$tokenizer )
1652 # Check whether next token is a text token
1653 # If yes, fetch it and convert the text into a
1654 # link to an RFC source
1655 $token = $tokenizer->previewToken();
1656 while ( $token["type"] == "" )
1658 $tokenizer->nextToken();
1659 $token = $tokenizer->previewToken();
1661 if ( $token["type"] == "text" )
1663 $token = $tokenizer->nextToken();
1664 $x = $token["text"];
1665 $valid = "0123456789";
1667 $rfc = $blank = "" ;
1668 while ( " " == $x{0} ) {
1670 $x = substr( $x, 1 );
1672 while ( strstr( $valid, $x{0} ) != false ) {
1674 $x = substr( $x, 1 );
1678 $text .= "RFC $blank$x";
1680 $url = wfmsg( "rfcurl" );
1681 $url = str_replace( "$1", $rfc, $url);
1682 $sk =& $this->mOptions
->getSkin();
1683 $la = $sk->getExternalLinkAttributes( $url, "RFC {$rfc}" );
1684 $text = "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
1692 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true )
1694 $this->mOptions
= $options;
1695 $this->mTitle
=& $title;
1696 $this->mOutputType
= OT_WIKI
;
1698 if ( $clearState ) {
1699 $this->clearState();
1702 $stripState = false;
1703 $text = str_replace("\r\n", "\n", $text);
1704 $text = $this->strip( $text, $stripState, false );
1705 $text = $this->pstPass2( $text, $user );
1706 $text = $this->unstrip( $text, $stripState );
1710 /* private */ function pstPass2( $text, &$user )
1712 global $wgLang, $wgLocaltimezone, $wgCurParser;
1714 # Variable replacement
1715 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
1716 $text = $this->replaceVariables( $text );
1720 $n = $user->getName();
1721 $k = $user->getOption( "nickname" );
1722 if ( "" == $k ) { $k = $n; }
1723 if(isset($wgLocaltimezone)) {
1724 $oldtz = getenv("TZ"); putenv("TZ=$wgLocaltimezone");
1726 /* Note: this is an ugly timezone hack for the European wikis */
1727 $d = $wgLang->timeanddate( date( "YmdHis" ), false ) .
1728 " (" . date( "T" ) . ")";
1729 if(isset($wgLocaltimezone)) putenv("TZ=$oldtz");
1731 $text = preg_replace( "/~~~~~/", $d, $text );
1732 $text = preg_replace( "/~~~~/", "[[" . $wgLang->getNsText(
1733 Namespace::getUser() ) . ":$n|$k]] $d", $text );
1734 $text = preg_replace( "/~~~/", "[[" . $wgLang->getNsText(
1735 Namespace::getUser() ) . ":$n|$k]]", $text );
1737 # Context links: [[|name]] and [[name (context)|]]
1739 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
1740 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
1741 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
1742 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
1744 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
1745 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
1746 $p3 = "/\[\[($namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]]
1747 $p4 = "/\[\[($namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/";
1748 # [[ns:page (cont)|]]
1750 $t = $this->mTitle
->getText();
1751 if ( preg_match( $conpat, $t, $m ) ) {
1754 $text = preg_replace( $p4, "[[\\1:\\2 (\\3)|\\2]]", $text );
1755 $text = preg_replace( $p1, "[[\\1 (\\2)|\\1]]", $text );
1756 $text = preg_replace( $p3, "[[\\1:\\2|\\2]]", $text );
1758 if ( "" == $context ) {
1759 $text = preg_replace( $p2, "[[\\1]]", $text );
1761 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
1765 $mw =& MagicWord::get( MAG_SUBST );
1766 $wgCurParser = $this->fork();
1767 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
1768 $this->merge( $wgCurParser );
1771 # Trim trailing whitespace
1772 # MAG_END (__END__) tag allows for trailing
1773 # whitespace to be deliberately included
1774 $text = rtrim( $text );
1775 $mw =& MagicWord
::get( MAG_END
);
1776 $mw->matchAndRemove( $text );
1781 # Set up some variables which are usually set up in parse()
1782 # so that an external function can call some class members with confidence
1783 function startExternalParse( &$title, $options, $outputType, $clearState = true )
1785 $this->mTitle
=& $title;
1786 $this->mOptions
= $options;
1787 $this->mOutputType
= $outputType;
1788 if ( $clearState ) {
1789 $this->clearState();
1793 function transformMsg( $text, $options ) {
1795 static $executing = false;
1797 # Guard against infinite recursion
1803 $this->mTitle
= $wgTitle;
1804 $this->mOptions
= $options;
1805 $this->mOutputType
= OT_MSG
;
1806 $this->clearState();
1807 $text = $this->replaceVariables( $text );
1816 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
1818 function ParserOutput( $text = "", $languageLinks = array(), $categoryLinks = array(),
1819 $containsOldMagic = false )
1821 $this->mText
= $text;
1822 $this->mLanguageLinks
= $languageLinks;
1823 $this->mCategoryLinks
= $categoryLinks;
1824 $this->mContainsOldMagic
= $containsOldMagic;
1827 function getText() { return $this->mText
; }
1828 function getLanguageLinks() { return $this->mLanguageLinks
; }
1829 function getCategoryLinks() { return $this->mCategoryLinks
; }
1830 function containsOldMagic() { return $this->mContainsOldMagic
; }
1831 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
1832 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
1833 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
1834 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
1836 function merge( $other ) {
1837 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
1838 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
1839 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
1846 # All variables are private
1847 var $mUseTeX; # Use texvc to expand <math> tags
1848 var $mUseCategoryMagic; # Treat [[Category:xxxx]] tags specially
1849 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
1850 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
1851 var $mAllowExternalImages; # Allow external images inline
1852 var $mSkin; # Reference to the preferred skin
1853 var $mDateFormat; # Date format index
1854 var $mEditSection; # Create "edit section" links
1855 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
1856 var $mNumberHeadings; # Automatically number headings
1857 var $mShowToc; # Show table of contents
1859 function getUseTeX() { return $this->mUseTeX
; }
1860 function getUseCategoryMagic() { return $this->mUseCategoryMagic
; }
1861 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
1862 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
1863 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
1864 function getSkin() { return $this->mSkin
; }
1865 function getDateFormat() { return $this->mDateFormat
; }
1866 function getEditSection() { return $this->mEditSection
; }
1867 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
1868 function getNumberHeadings() { return $this->mNumberHeadings
; }
1869 function getShowToc() { return $this->mShowToc
; }
1871 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
1872 function setUseCategoryMagic( $x ) { return wfSetVar( $this->mUseCategoryMagic
, $x ); }
1873 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
1874 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
1875 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
1876 function setSkin( $x ) { return wfSetRef( $this->mSkin
, $x ); }
1877 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
1878 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
1879 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
1880 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
1881 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
1883 /* static */ function newFromUser( &$user )
1885 $popts = new ParserOptions
;
1886 $popts->initialiseFromUser( &$user );
1890 function initialiseFromUser( &$userInput )
1892 global $wgUseTeX, $wgUseCategoryMagic, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
1894 if ( !$userInput ) {
1896 $user->setLoaded( true );
1898 $user =& $userInput;
1901 $this->mUseTeX
= $wgUseTeX;
1902 $this->mUseCategoryMagic
= $wgUseCategoryMagic;
1903 $this->mUseDynamicDates
= $wgUseDynamicDates;
1904 $this->mInterwikiMagic
= $wgInterwikiMagic;
1905 $this->mAllowExternalImages
= $wgAllowExternalImages;
1906 $this->mSkin
=& $user->getSkin();
1907 $this->mDateFormat
= $user->getOption( "date" );
1908 $this->mEditSection
= $user->getOption( "editsection" );
1909 $this->mEditSectionOnRightClick
= $user->getOption( "editsectiononrightclick" );
1910 $this->mNumberHeadings
= $user->getOption( "numberheadings" );
1911 $this->mShowToc
= $user->getOption( "showtoc" );
1917 # Regex callbacks, used in Parser::replaceVariables
1918 function wfBraceSubstitution( $matches )
1920 global $wgCurParser;
1921 return $wgCurParser->braceSubstitution( $matches );