3 // require_once('Tokenizer.php');
7 # Processes wiki markup
9 # There are two main entry points into the Parser class: parse() and preSaveTransform().
10 # The parse() function produces HTML output, preSaveTransform() produces altered wiki markup.
13 # objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
15 # NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
17 # settings: $wgUseTex*, $wgUseCategoryMagic*, $wgUseDynamicDates*, $wgInterwikiMagic*,
18 # $wgNamespacesWithSubpages, $wgLanguageCode, $wgAllowExternalImages*,
21 # * only within ParserOptions
24 #----------------------------------------
25 # Variable substitution O(N^2) attack
26 #-----------------------------------------
27 # Without countermeasures, it would be possible to attack the parser by saving a page
28 # filled with a large number of inclusions of large pages. The size of the generated
29 # page would be proportional to the square of the input size. Hence, we limit the number
30 # of inclusions of any given page, thus bringing any attack back to O(N).
33 define( "MAX_INCLUDE_REPEAT", 5 );
34 define( "MAX_INCLUDE_SIZE", 1000000 ); // 1 Million
36 # Allowed values for $mOutputType
37 define( "OT_HTML", 1 );
38 define( "OT_WIKI", 2 );
39 define( "OT_MSG", 3 );
41 # string parameter for extractTags which will cause it
42 # to strip HTML comments in addition to regular
43 # <XML>-style tags. This should not be anything we
44 # may want to use in wikisyntax
45 define( "STRIP_COMMENTS", "HTMLCommentStrip" );
47 # prefix for escaping, used in two functions at least
48 define( "UNIQ_PREFIX", "NaodW29");
55 # Cleared with clearState():
56 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
57 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
60 var $mOptions, $mTitle, $mOutputType,
61 $mTemplates, // cache of already loaded templates, avoids
62 // multiple SQL queries for the same string
63 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
64 // in this path. Used for loop detection.
67 $this->mTemplates
= array();
68 $this->mTemplatePath
= array();
69 $this->mTagHooks
= array();
73 function clearState() {
74 $this->mOutput
= new ParserOutput
;
75 $this->mAutonumber
= 0;
76 $this->mLastSection
= "";
77 $this->mDTopen
= false;
78 $this->mVariables
= false;
79 $this->mIncludeCount
= array();
80 $this->mStripState
= array();
81 $this->mArgStack
= array();
82 $this->mInPre
= false;
85 # First pass--just handle <nowiki> sections, pass the rest off
86 # to internalParse() which does all the real work.
88 # Returns a ParserOutput
90 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
92 $fname = "Parser::parse";
93 wfProfileIn( $fname );
99 $this->mOptions
= $options;
100 $this->mTitle
=& $title;
101 $this->mOutputType
= OT_HTML
;
104 $text = $this->strip( $text, $this->mStripState
);
105 $text = $this->internalParse( $text, $linestart );
106 $text = $this->unstrip( $text, $this->mStripState
);
107 # Clean up special characters, only run once, next-to-last before doBlockLevels
110 # french spaces, last one Guillemet-left
111 # only if there is something before the space
112 '/(.) (\\?|:|;|!|\\302\\273)/i' => '\\1 \\2',
113 # french spaces, Guillemet-right
114 "/(\\302\\253) /i"=>"\\1 ",
115 '/<hr *>/i' => '<hr />',
116 '/<br *>/i' => '<br />',
117 '/<center *>/i' => '<div class="center">',
118 '/<\\/center *>/i' => '</div>',
119 # Clean up spare ampersands; note that we probably ought to be
120 # more careful about named entities.
121 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
123 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
126 # french spaces, last one Guillemet-left
127 '/ (\\?|:|!|\\302\\273)/i' => ' \\1',
128 # french spaces, Guillemet-right
129 '/(\\302\\253) /i' => '\\1 ',
130 '/([^> ]+(0(1|3|9);)[^< ]*)/i' => '<span class="diacrit">\\1</span>',
131 '/<center *>/i' => '<div class="center">',
132 '/<\\/center *>/i' => '</div>'
134 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
137 $text = $this->doBlockLevels( $text, $linestart );
138 $text = $this->unstripNoWiki( $text, $this->mStripState
);
140 $text = $this->tidy($text);
142 $this->mOutput
->setText( $text );
143 wfProfileOut( $fname );
144 return $this->mOutput
;
147 /* static */ function getRandomString() {
148 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
151 # Replaces all occurrences of <$tag>content</$tag> in the text
152 # with a random marker and returns the new text. the output parameter
153 # $content will be an associative array filled with data on the form
154 # $unique_marker => content.
156 # If $content is already set, the additional entries will be appended
158 # If $tag is set to STRIP_COMMENTS, the function will extract
159 # <!-- HTML comments -->
161 /* static */ function extractTags($tag, $text, &$content, $uniq_prefix = ""){
162 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
169 while ( '' != $text ) {
170 if($tag==STRIP_COMMENTS
) {
171 $p = preg_split( '/<!--/i', $text, 2 );
173 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
176 if ( ( count( $p ) < 2 ) ||
( '' == $p[1] ) ) {
179 if($tag==STRIP_COMMENTS
) {
180 $q = preg_split( '/-->/i', $p[1], 2 );
182 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
184 $marker = $rnd . sprintf('%08X', $n++
);
185 $content[$marker] = $q[0];
186 $stripped .= $marker;
193 # Strips and renders <nowiki>, <pre>, <math>, <hiero>
194 # If $render is set, performs necessary rendering operations on plugins
195 # Returns the text, and fills an array with data needed in unstrip()
196 # If the $state is already a valid strip state, it adds to the state
198 # When $stripcomments is set, HTML comments <!-- like this -->
199 # will be stripped in addition to other tags. This is important
200 # for section editing, where these comments cause confusion when
201 # counting the sections in the wikisource
202 function strip( $text, &$state, $stripcomments = false ) {
203 $render = ($this->mOutputType
== OT_HTML
);
204 $html_content = array();
205 $nowiki_content = array();
206 $math_content = array();
207 $pre_content = array();
208 $comment_content = array();
209 $ext_content = array();
211 # Replace any instances of the placeholders
212 $uniq_prefix = UNIQ_PREFIX
;
213 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
218 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
219 foreach( $html_content as $marker => $content ) {
221 # Raw and unchecked for validity.
222 $html_content[$marker] = $content;
224 $html_content[$marker] = "<html>$content</html>";
230 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
231 foreach( $nowiki_content as $marker => $content ) {
233 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
235 $nowiki_content[$marker] = "<nowiki>$content</nowiki>";
240 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
241 foreach( $math_content as $marker => $content ){
243 if( $this->mOptions
->getUseTeX() ) {
244 $math_content[$marker] = renderMath( $content );
246 $math_content[$marker] = "<math>$content<math>";
249 $math_content[$marker] = "<math>$content</math>";
254 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
255 foreach( $pre_content as $marker => $content ){
257 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
259 $pre_content[$marker] = "<pre>$content</pre>";
265 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
266 foreach( $comment_content as $marker => $content ){
267 $comment_content[$marker] = "<!--$content-->";
272 foreach ( $this->mTagHooks
as $tag => $callback ) {
273 $ext_contents[$tag] = array();
274 $text = Parser
::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
275 foreach( $ext_content[$tag] as $marker => $content ) {
277 $ext_content[$tag][$marker] = $callback( $content );
279 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
284 # Merge state with the pre-existing state, if there is one
286 $state['html'] = $state['html'] +
$html_content;
287 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
288 $state['math'] = $state['math'] +
$math_content;
289 $state['pre'] = $state['pre'] +
$pre_content;
290 $state['comment'] = $state['comment'] +
$comment_content;
292 foreach( $ext_content as $tag => $array ) {
293 if ( array_key_exists( $tag, $state ) ) {
294 $state[$tag] = $state[$tag] +
$array;
299 'html' => $html_content,
300 'nowiki' => $nowiki_content,
301 'math' => $math_content,
302 'pre' => $pre_content,
303 'comment' => $comment_content,
309 # always call unstripNoWiki() after this one
310 function unstrip( $text, &$state ) {
311 # Must expand in reverse order, otherwise nested tags will be corrupted
312 $contentDict = end( $state );
313 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
314 if( key($state) != 'nowiki' && key($state) != 'html') {
315 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
316 $text = str_replace( key( $contentDict ), $content, $text );
323 # always call this after unstrip() to preserve the order
324 function unstripNoWiki( $text, &$state ) {
325 # Must expand in reverse order, otherwise nested tags will be corrupted
326 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
327 $text = str_replace( key( $state['nowiki'] ), $content, $text );
332 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
333 $text = str_replace( key( $state['html'] ), $content, $text );
340 # Add an item to the strip state
341 # Returns the unique tag which must be inserted into the stripped text
342 # The tag will be replaced with the original text in unstrip()
344 function insertStripItem( $text, &$state ) {
345 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
354 $state['item'][$rnd] = $text;
359 # generate a list of subcategories and pages for a category
360 # depending on wfMsg("usenewcategorypage") it either calls the new
361 # or the old code. The new code will not work properly for some
362 # languages due to sorting issues, so they might want to turn it
364 function categoryMagic() {
365 $msg = wfMsg('usenewcategorypage');
366 if ( '0' == @$msg[0] )
368 return $this->oldCategoryMagic();
370 return $this->newCategoryMagic();
374 # This method generates the list of subcategories and pages for a category
375 function oldCategoryMagic () {
376 global $wgLang , $wgUser ;
377 $fname = 'Parser::oldCategoryMagic';
379 if ( !$this->mOptions
->getUseCategoryMagic() ) return ; # Doesn't use categories at all
381 $cns = Namespace::getCategory() ;
382 if ( $this->mTitle
->getNamespace() != $cns ) return "" ; # This ain't a category page
384 $r = "<br style=\"clear:both;\"/>\n";
387 $sk =& $wgUser->getSkin() ;
389 $articles = array() ;
390 $children = array() ;
392 $id = $this->mTitle
->getArticleID() ;
395 $dbr =& wfGetDB( DB_SLAVE
);
396 $cur = $dbr->tableName( 'cur' );
397 $categorylinks = $dbr->tableName( 'categorylinks' );
399 $t = $dbr->strencode( $this->mTitle
->getDBKey() );
400 $sql = "SELECT DISTINCT cur_title,cur_namespace FROM $cur,$categorylinks " .
401 "WHERE cl_to='$t' AND cl_from=cur_id ORDER BY cl_sortkey" ;
402 $res = $dbr->query( $sql, $fname ) ;
403 while ( $x = $dbr->fetchObject ( $res ) ) $data[] = $x ;
405 # For all pages that link to this category
406 foreach ( $data AS $x )
408 $t = $wgLang->getNsText ( $x->cur_namespace
) ;
409 if ( $t != "" ) $t .= ":" ;
410 $t .= $x->cur_title
;
412 if ( $x->cur_namespace
== $cns ) {
413 array_push ( $children , $sk->makeLink ( $t ) ) ; # Subcategory
415 array_push ( $articles , $sk->makeLink ( $t ) ) ; # Page in this category
418 $dbr->freeResult ( $res ) ;
420 # Showing subcategories
421 if ( count ( $children ) > 0 ) {
422 $r .= '<h2>'.wfMsg('subcategories')."</h2>\n" ;
423 $r .= implode ( ', ' , $children ) ;
426 # Showing pages in this category
427 if ( count ( $articles ) > 0 ) {
428 $ti = $this->mTitle
->getText() ;
429 $h = wfMsg( 'category_header', $ti );
430 $r .= "<h2>{$h}</h2>\n" ;
431 $r .= implode ( ', ' , $articles ) ;
439 function newCategoryMagic () {
440 global $wgLang , $wgUser ;
441 if ( !$this->mOptions
->getUseCategoryMagic() ) return ; # Doesn't use categories at all
443 $cns = Namespace::getCategory() ;
444 if ( $this->mTitle
->getNamespace() != $cns ) return '' ; # This ain't a category page
446 $r = "<br style=\"clear:both;\"/>\n";
449 $sk =& $wgUser->getSkin() ;
451 $articles = array() ;
452 $articles_start_char = array();
453 $children = array() ;
454 $children_start_char = array();
456 $id = $this->mTitle
->getArticleID() ;
459 $dbr =& wfGetDB( DB_SLAVE
);
460 $cur = $dbr->tableName( 'cur' );
461 $categorylinks = $dbr->tableName( 'categorylinks' );
463 $t = $dbr->strencode( $this->mTitle
->getDBKey() );
464 $sql = "SELECT DISTINCT cur_title,cur_namespace,cl_sortkey FROM " .
465 "$cur,$categorylinks WHERE cl_to='$t' AND cl_from=cur_id ORDER BY cl_sortkey" ;
466 $res = $dbr->query ( $sql ) ;
467 while ( $x = $dbr->fetchObject ( $res ) )
469 $t = $ns = $wgLang->getNsText ( $x->cur_namespace
) ;
470 if ( $t != '' ) $t .= ':' ;
471 $t .= $x->cur_title
;
473 if ( $x->cur_namespace
== $cns ) {
474 $ctitle = str_replace( '_',' ',$x->cur_title
);
475 array_push ( $children, $sk->makeKnownLink ( $t, $ctitle ) ) ; # Subcategory
477 // If there's a link from Category:A to Category:B, the sortkey of the resulting
478 // entry in the categorylinks table is Category:A, not A, which it SHOULD be.
479 // Workaround: If sortkey == "Category:".$title, than use $title for sorting,
480 // else use sortkey...
481 if ( ($ns.":".$ctitle) == $x->cl_sortkey
) {
482 array_push ( $children_start_char, $wgLang->firstChar( $x->cur_title
) );
484 array_push ( $children_start_char, $wgLang->firstChar( $x->cl_sortkey
) ) ;
487 array_push ( $articles , $sk->makeKnownLink ( $t ) ) ; # Page in this category
488 array_push ( $articles_start_char, $wgLang->firstChar( $x->cl_sortkey
) ) ;
491 $dbr->freeResult ( $res ) ;
493 $ti = $this->mTitle
->getText() ;
495 # Don't show subcategories section if there are none.
496 if ( count ( $children ) > 0 )
498 # Showing subcategories
499 $r .= '<h2>' . wfMsg( 'subcategories' ) . "</h2>\n"
500 . wfMsg( 'subcategorycount', count( $children ) );
501 if ( count ( $children ) > 6 ) {
503 // divide list into three equal chunks
504 $chunk = (int) (count ( $children ) / 3);
506 // get and display header
507 $r .= '<table width="100%"><tr valign="top">';
512 // loop through the chunks
513 for($startChunk = 0, $endChunk = $chunk, $chunkIndex = 0;
515 $chunkIndex++
, $startChunk = $endChunk, $endChunk +
= $chunk +
1)
519 // output all subcategories to category
520 for ($index = $startChunk ;
521 $index < $endChunk && $index < count($children);
524 // check for change of starting letter or begging of chunk
525 if ( ($children_start_char[$index] != $children_start_char[$index - 1])
526 ||
($index == $startChunk) )
528 $r .= "</ul><h3>{$children_start_char[$index]}</h3>\n<ul>";
531 $r .= "<li>{$children[$index]}</li>";
537 $r .= '</tr></table>';
539 // for short lists of subcategories to category.
541 $r .= "<h3>{$children_start_char[0]}</h3>\n";
542 $r .= '<ul><li>'.$children[0].'</li>';
543 for ($index = 1; $index < count($children); $index++
)
545 if ($children_start_char[$index] != $children_start_char[$index - 1])
547 $r .= "</ul><h3>{$children_start_char[$index]}</h3>\n<ul>";
550 $r .= "<li>{$children[$index]}</li>";
554 } # END of if ( count($children) > 0 )
556 $r .= '<h2>' . wfMsg( 'category_header', $ti ) . "</h2>\n" .
557 wfMsg( 'categoryarticlecount', count( $articles ) );
559 # Showing articles in this category
560 if ( count ( $articles ) > 6) {
561 $ti = $this->mTitle
->getText() ;
563 // divide list into three equal chunks
564 $chunk = (int) (count ( $articles ) / 3);
566 // get and display header
567 $r .= '<table width="100%"><tr valign="top">';
569 // loop through the chunks
570 for($startChunk = 0, $endChunk = $chunk, $chunkIndex = 0;
572 $chunkIndex++
, $startChunk = $endChunk, $endChunk +
= $chunk +
1)
577 // output all articles in category
578 for ($index = $startChunk ;
579 $index < $endChunk && $index < count($articles);
582 // check for change of starting letter or begging of chunk
583 if ( ($articles_start_char[$index] != $articles_start_char[$index - 1])
584 ||
($index == $startChunk) )
586 $r .= "</ul><h3>{$articles_start_char[$index]}</h3>\n<ul>";
589 $r .= "<li>{$articles[$index]}</li>";
595 $r .= '</tr></table>';
596 } elseif ( count ( $articles ) > 0) {
597 // for short lists of articles in categories.
598 $ti = $this->mTitle
->getText() ;
600 $r .= '<h3>'.$articles_start_char[0]."</h3>\n";
601 $r .= '<ul><li>'.$articles[0].'</li>';
602 for ($index = 1; $index < count($articles); $index++
)
604 if ($articles_start_char[$index] != $articles_start_char[$index - 1])
606 $r .= "</ul><h3>{$articles_start_char[$index]}</h3>\n<ul>";
609 $r .= "<li>{$articles[$index]}</li>";
618 # Return allowed HTML attributes
619 function getHTMLattrs () {
620 $htmlattrs = array( # Allowed attributes--no scripting, etc.
621 'title', 'align', 'lang', 'dir', 'width', 'height',
622 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
623 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
624 /* FONT */ 'type', 'start', 'value', 'compact',
625 /* For various lists, mostly deprecated but safe */
626 'summary', 'width', 'border', 'frame', 'rules',
627 'cellspacing', 'cellpadding', 'valign', 'char',
628 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
629 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
630 'id', 'class', 'name', 'style' /* For CSS */
635 # Remove non approved attributes and javascript in css
636 function fixTagAttributes ( $t ) {
637 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
638 $htmlattrs = $this->getHTMLattrs() ;
640 # Strip non-approved attributes from the tag
642 '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
643 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
646 $t = str_replace ( "<></>" , "" , $t ) ; # This should fix bug 980557
648 # Strip javascript "expression" from stylesheets. Brute force approach:
649 # If anythin offensive is found, all attributes of the HTML tag are dropped
652 '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
653 wfMungeToUtf8( $t ) ) )
661 # interface with html tidy, used if $wgUseTidy = true
662 function tidy ( $text ) {
663 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
664 global $wgInputEncoding, $wgOutputEncoding;
665 $fname = 'Parser::tidy';
666 wfProfileIn( $fname );
669 switch(strtoupper($wgOutputEncoding)) {
671 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)?
' -latin1':' -raw';
674 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)?
' -utf8':' -raw';
677 $wgTidyOpts .= ' -raw';
680 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
681 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
682 '<head><title>test</title></head><body>'.$text.'</body></html>';
683 $descriptorspec = array(
684 0 => array('pipe', 'r'),
685 1 => array('pipe', 'w'),
686 2 => array('file', '/dev/null', 'a')
688 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts", $descriptorspec, $pipes);
689 if (is_resource($process)) {
690 fwrite($pipes[0], $wrappedtext);
692 while (!feof($pipes[1])) {
693 $cleansource .= fgets($pipes[1], 1024);
696 $return_value = proc_close($process);
699 wfProfileOut( $fname );
701 if( $cleansource == '' && $text != '') {
702 wfDebug( "Tidy error detected!\n" );
703 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
709 # parse the wiki syntax used to render tables
710 function doTableStuff ( $t ) {
711 $fname = 'Parser::doTableStuff';
712 wfProfileIn( $fname );
714 $t = explode ( "\n" , $t ) ;
715 $td = array () ; # Is currently a td tag open?
716 $ltd = array () ; # Was it TD or TH?
717 $tr = array () ; # Is currently a tr tag open?
718 $ltr = array () ; # tr attributes
719 $indent_level = 0; # indent level of the table
720 foreach ( $t AS $k => $x )
723 $fc = substr ( $x , 0 , 1 ) ;
724 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) )
726 $indent_level = strlen( $matches[1] );
728 str_repeat( "<dl><dd>", $indent_level ) .
729 "<table " . $this->fixTagAttributes ( $matches[2] ) . '>' ;
730 array_push ( $td , false ) ;
731 array_push ( $ltd , '' ) ;
732 array_push ( $tr , false ) ;
733 array_push ( $ltr , '' ) ;
735 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
736 else if ( '|}' == substr ( $x , 0 , 2 ) )
739 $l = array_pop ( $ltd ) ;
740 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
741 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
743 $t[$k] = $z . str_repeat( "</dd></dl>", $indent_level );
745 else if ( '|-' == substr ( $x , 0 , 2 ) ) # Allows for |---------------
747 $x = substr ( $x , 1 ) ;
748 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
750 $l = array_pop ( $ltd ) ;
751 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
752 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
755 array_push ( $tr , false ) ;
756 array_push ( $td , false ) ;
757 array_push ( $ltd , '' ) ;
758 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
760 else if ( '|' == $fc ||
'!' == $fc ||
'|+' == substr ( $x , 0 , 2 ) ) # Caption
762 if ( '|+' == substr ( $x , 0 , 2 ) )
765 $x = substr ( $x , 1 ) ;
767 $after = substr ( $x , 1 ) ;
768 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
769 $after = explode ( '||' , $after ) ;
771 foreach ( $after AS $theline )
776 $tra = array_pop ( $ltr ) ;
777 if ( !array_pop ( $tr ) ) $z = "<tr {$tra}>\n" ;
778 array_push ( $tr , true ) ;
779 array_push ( $ltr , '' ) ;
782 $l = array_pop ( $ltd ) ;
783 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
784 if ( $fc == '|' ) $l = 'td' ;
785 else if ( $fc == '!' ) $l = 'th' ;
786 else if ( $fc == '+' ) $l = 'caption' ;
788 array_push ( $ltd , $l ) ;
789 $y = explode ( '|' , $theline , 2 ) ;
790 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
791 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
793 array_push ( $td , true ) ;
798 # Closing open td, tr && table
799 while ( count ( $td ) > 0 )
801 if ( array_pop ( $td ) ) $t[] = '</td>' ;
802 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
806 $t = implode ( "\n" , $t ) ;
807 # $t = $this->removeHTMLtags( $t );
808 wfProfileOut( $fname );
812 # Parses the text and adds the result to the strip state
813 # Returns the strip tag
814 function stripParse( $text, $newline, $args )
816 $text = $this->strip( $text, $this->mStripState
);
817 $text = $this->internalParse( $text, (bool)$newline, $args, false );
818 return $newline.$this->insertStripItem( $text, $this->mStripState
);
821 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
822 $fname = 'Parser::internalParse';
823 wfProfileIn( $fname );
826 if ( $fixLbug ) $text = preg_replace ( '/(l|L)\'/' , '\\1'' , $text ) ;
828 $text = $this->removeHTMLtags( $text );
829 $text = $this->replaceVariables( $text, $args );
831 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
833 $text = $this->doHeadings( $text );
834 if($this->mOptions
->getUseDynamicDates()) {
835 global $wgDateFormatter;
836 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
838 $text = $this->doAllQuotes( $text );
839 // $text = $this->doExponent( $text );
840 $text = $this->replaceExternalLinks( $text );
841 $text = $this->doMagicLinks( $text );
842 $text = $this->replaceInternalLinks ( $text );
843 $text = $this->replaceInternalLinks ( $text );
844 //$text = $this->doTokenizedParser ( $text );
845 $text = $this->doTableStuff( $text );
846 $text = $this->formatHeadings( $text, $isMain );
847 $sk =& $this->mOptions
->getSkin();
848 $text = $sk->transformContent( $text );
850 if ( $isMain && !isset ( $this->categoryMagicDone
) ) {
851 $text .= $this->categoryMagic () ;
852 $this->categoryMagicDone
= true ;
855 wfProfileOut( $fname );
859 /* private */ function &doMagicLinks( &$text ) {
860 $text = $this->magicISBN( $text );
861 $text = $this->magicGEO( $text );
862 $text = $this->magicRFC( $text );
866 # Parse ^^ tokens and return html
867 /* private */ function doExponent ( $text )
869 $fname = 'Parser::doExponent';
870 wfProfileIn( $fname);
871 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
872 wfProfileOut( $fname);
876 # Parse headers and return html
877 /* private */ function doHeadings( $text ) {
878 $fname = 'Parser::doHeadings';
879 wfProfileIn( $fname );
880 for ( $i = 6; $i >= 1; --$i ) {
881 $h = substr( '======', 0, $i );
882 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
883 "<h{$i}>\\1</h{$i}>\\2", $text );
885 wfProfileOut( $fname );
889 /* private */ function doAllQuotes( $text ) {
890 $fname = 'Parser::doAllQuotes';
891 wfProfileIn( $fname );
893 $lines = explode( "\n", $text );
894 foreach ( $lines as $line ) {
895 $outtext .= $this->doQuotes ( '', $line, '' ) . "\n";
897 $outtext = substr($outtext, 0,-1);
898 wfProfileOut( $fname );
902 /* private */ function doQuotes( $pre, $text, $mode ) {
903 if ( preg_match( "/^(.*)''(.*)$/sU", $text, $m ) ) {
904 $m1_strong = ($m[1] == "") ?
"" : "<strong>{$m[1]}</strong>";
905 $m1_em = ($m[1] == "") ?
"" : "<em>{$m[1]}</em>";
906 if ( substr ($m[2], 0, 1) == '\'' ) {
907 $m[2] = substr ($m[2], 1);
909 return $this->doQuotes ( $m[1], $m[2], ($m[1] == '') ?
'both' : 'emstrong' );
910 } else if ($mode == 'strong') {
911 return $m1_strong . $this->doQuotes ( '', $m[2], '' );
912 } else if (($mode == 'emstrong') ||
($mode == 'both')) {
913 return $this->doQuotes ( '', $pre.$m1_strong.$m[2], 'em' );
914 } else if ($mode == 'strongem') {
915 return "<strong>{$pre}{$m1_em}</strong>" . $this->doQuotes ( '', $m[2], 'em' );
917 return $m[1] . $this->doQuotes ( '', $m[2], 'strong' );
920 if ($mode == 'strong') {
921 return $this->doQuotes ( $m[1], $m[2], ($m[1] == '') ?
'both' : 'strongem' );
922 } else if ($mode == 'em') {
923 return $m1_em . $this->doQuotes ( '', $m[2], '' );
924 } else if ($mode == 'emstrong') {
925 return "<em>{$pre}{$m1_strong}</em>" . $this->doQuotes ( '', $m[2], 'strong' );
926 } else if (($mode == 'strongem') ||
($mode == 'both')) {
927 return $this->doQuotes ( '', $pre.$m1_em.$m[2], 'strong' );
929 return $m[1] . $this->doQuotes ( '', $m[2], 'em' );
933 $text_strong = ($text == '') ?
'' : "<strong>{$text}</strong>";
934 $text_em = ($text == '') ?
'' : "<em>{$text}</em>";
937 } else if ($mode == 'em') {
938 return $pre . $text_em;
939 } else if ($mode == 'strong') {
940 return $pre . $text_strong;
941 } else if ($mode == 'strongem') {
942 return (($pre == '') && ($text == '')) ?
'' : "<strong>{$pre}{$text_em}</strong>";
944 return (($pre == '') && ($text == '')) ?
'' : "<em>{$pre}{$text_strong}</em>";
949 # Note: we have to do external links before the internal ones,
950 # and otherwise take great care in the order of things here, so
951 # that we don't end up interpreting some URLs twice.
953 /* private */ function replaceExternalLinks( $text ) {
954 $fname = 'Parser::replaceExternalLinks';
955 wfProfileIn( $fname );
956 $text = $this->subReplaceExternalLinks( $text, 'http', true );
957 $text = $this->subReplaceExternalLinks( $text, 'https', true );
958 $text = $this->subReplaceExternalLinks( $text, 'ftp', false );
959 $text = $this->subReplaceExternalLinks( $text, 'irc', false );
960 $text = $this->subReplaceExternalLinks( $text, 'gopher', false );
961 $text = $this->subReplaceExternalLinks( $text, 'news', false );
962 $text = $this->subReplaceExternalLinks( $text, 'mailto', false );
963 wfProfileOut( $fname );
967 /* private */ function subReplaceExternalLinks( $s, $protocol, $autonumber ) {
968 $unique = '4jzAfzB8hNvf4sqyO9Edd8pSmk9rE2in0Tgw3';
969 $uc = "A-Za-z0-9_\\/~%\\-+&*#?!=()@\\x80-\\xFF";
971 # this is the list of separators that should be ignored if they
972 # are the last character of an URL but that should be included
973 # if they occur within the URL, e.g. "go to www.foo.com, where .."
974 # in this case, the last comma should not become part of the URL,
975 # but in "www.foo.com/123,2342,32.htm" it should.
977 $fnc = 'A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF';
978 $images = 'gif|png|jpg|jpeg';
980 # PLEASE NOTE: The curly braces { } are not part of the regex,
981 # they are interpreted as part of the string (used to tell PHP
982 # that the content of the string should be inserted there).
983 $e1 = "/(^|[^\\[])({$protocol}:)([{$uc}{$sep}]+)\\/([{$fnc}]+)\\." .
984 "((?i){$images})([^{$uc}]|$)/";
986 $e2 = "/(^|[^\\[])({$protocol}:)(([".$uc."]|[".$sep."][".$uc."])+)([^". $uc . $sep. "]|[".$sep."]|$)/";
987 $sk =& $this->mOptions
->getSkin();
989 if ( $autonumber and $this->mOptions
->getAllowExternalImages() ) { # Use img tags only for HTTP urls
990 $s = preg_replace( $e1, '\\1' . $sk->makeImage( "{$unique}:\\3" .
991 '/\\4.\\5', '\\4.\\5' ) . '\\6', $s );
993 $s = preg_replace( $e2, '\\1' . "<a href=\"{$unique}:\\3\"" .
994 $sk->getExternalLinkAttributes( "{$unique}:\\3", wfEscapeHTML(
995 "{$unique}:\\3" ) ) . ">" . wfEscapeHTML( "{$unique}:\\3" ) .
997 $s = str_replace( $unique, $protocol, $s );
999 $a = explode( "[{$protocol}:", " " . $s );
1000 $s = array_shift( $a );
1001 $s = substr( $s, 1 );
1003 # Regexp for URL in square brackets
1004 $e1 = "/^([{$uc}{$sep}]+)\\](.*)\$/sD";
1005 # Regexp for URL with link text in square brackets
1006 $e2 = "/^([{$uc}{$sep}]+)\\s+([^\\]]+)\\](.*)\$/sD";
1008 foreach ( $a as $line ) {
1010 # CASE 1: Link in square brackets, e.g.
1011 # some text [http://domain.tld/some.link] more text
1012 if ( preg_match( $e1, $line, $m ) ) {
1013 $link = "{$protocol}:{$m[1]}";
1015 if ( $autonumber ) { $text = "[" . ++
$this->mAutonumber
. "]"; }
1016 else { $text = wfEscapeHTML( $link ); }
1019 # CASE 2: Link with link text and text directly following it, e.g.
1020 # This is a collection of [http://domain.tld/some.link link]s
1021 else if ( preg_match( $e2, $line, $m ) ) {
1022 $link = "{$protocol}:{$m[1]}";
1026 if ( preg_match( wfMsg ('linktrail'), $trail, $m2 ) ) {
1032 # CASE 3: Nothing matches, just output the source text
1034 $s .= "[{$protocol}:" . $line;
1038 if( $link == $text ||
preg_match( "!$protocol://" . preg_quote( $text, "/" ) . "/?$!", $link ) ) {
1041 # Expand the URL for printable version
1042 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $link ) . "</i>)</span>";
1044 $la = $sk->getExternalLinkAttributes( $link, $text );
1045 $s .= "<a href='{$link}'{$la}>{$text}</a>{$dtrail}{$paren}{$trail}";
1052 /* private */ function replaceInternalLinks( $s ) {
1053 global $wgLang, $wgLinkCache;
1054 global $wgNamespacesWithSubpages, $wgLanguageCode;
1055 static $fname = 'Parser::replaceInternalLinks' ;
1056 wfProfileIn( $fname );
1058 wfProfileIn( $fname.'-setup' );
1060 # the % is needed to support urlencoded titles as well
1061 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1062 $sk =& $this->mOptions
->getSkin();
1064 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1065 $isRedirect = $redirect->matchStart ( strtoupper ( substr ( $s , 0 , 10 ) ) ) ;
1067 $a = explode( '[[', ' ' . $s );
1068 $s = array_shift( $a );
1069 $s = substr( $s, 1 );
1071 # Match a link having the form [[namespace:link|alternate]]trail
1073 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1074 # Match the end of a line for a word that's not followed by whitespace,
1075 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1076 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1078 $useLinkPrefixExtension = $wgLang->linkPrefixExtension();
1079 # Special and Media are pseudo-namespaces; no pages actually exist in them
1080 static $image = FALSE;
1081 static $special = FALSE;
1082 static $media = FALSE;
1083 static $category = FALSE;
1084 if ( !$image ) { $image = Namespace::getImage(); }
1085 if ( !$special ) { $special = Namespace::getSpecial(); }
1086 if ( !$media ) { $media = Namespace::getMedia(); }
1087 if ( !$category ) { $category = Namespace::getCategory(); }
1089 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1091 if ( $useLinkPrefixExtension ) {
1092 if ( preg_match( $e2, $s, $m ) ) {
1093 $first_prefix = $m[2];
1096 $first_prefix = false;
1102 wfProfileOut( $fname.'-setup' );
1104 foreach ( $a as $line ) {
1105 wfProfileIn( $fname.'-prefixhandling' );
1106 if ( $useLinkPrefixExtension ) {
1107 if ( preg_match( $e2, $s, $m ) ) {
1115 $prefix = $first_prefix;
1116 $first_prefix = false;
1119 wfProfileOut( $fname.'-prefixhandling' );
1121 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1123 # fix up urlencoded title texts
1124 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1126 } else { # Invalid form; output directly
1127 $s .= $prefix . '[[' . $line ;
1131 /* Valid link forms:
1133 :Foobar -- override special treatment of prefix (images, language links)
1134 /Foobar -- convert to CurrentPage/Foobar
1135 /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1137 $c = substr($m[1],0,1);
1138 $noforce = ($c != ':');
1139 if( $c == '/' ) { # subpage
1140 if(substr($m[1],-1,1)=='/') { # / at end means we don't want the slash to be shown
1141 $m[1]=substr($m[1],1,strlen($m[1])-2);
1144 $noslash=substr($m[1],1);
1146 if(!empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()])) { # subpages allowed here
1147 $link = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1150 } # this might be changed for ugliness reasons
1152 $link = $noslash; # no subpage allowed, use standard link
1154 } elseif( $noforce ) { # no subpage
1157 $link = substr( $m[1], 1 );
1159 $wasblank = ( '' == $text );
1163 $nt = Title
::newFromText( $link );
1165 $s .= $prefix . '[[' . $line;
1168 $ns = $nt->getNamespace();
1169 $iw = $nt->getInterWiki();
1171 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
1172 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1173 $tmp = $prefix . $trail ;
1174 $s .= (trim($tmp) == '')?
'': $tmp;
1177 if ( $ns == $image ) {
1178 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1179 $wgLinkCache->addImageLinkObj( $nt );
1182 if ( $ns == $category && !$isRedirect ) {
1183 $t = $nt->getText() ;
1184 $nnt = Title
::newFromText ( Namespace::getCanonicalName($category).":".$t ) ;
1186 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1187 $pPLC=$sk->postParseLinkColour();
1188 $sk->postParseLinkColour( false );
1189 $t = $sk->makeLinkObj( $nnt, $t, '', '' , $prefix );
1190 $sk->postParseLinkColour( $pPLC );
1191 $wgLinkCache->resume();
1193 $sortkey = $wasblank ?
$this->mTitle
->getPrefixedText() : $text;
1194 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1195 $this->mOutput
->mCategoryLinks
[] = $t ;
1196 $s .= $prefix . $trail ;
1200 if( ( $nt->getPrefixedText() == $this->mTitle
->getPrefixedText() ) &&
1201 ( strpos( $link, '#' ) == FALSE ) ) {
1202 # Self-links are handled specially; generally de-link and change to bold.
1203 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1207 if( $ns == $media ) {
1208 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1209 $wgLinkCache->addImageLinkObj( $nt );
1211 } elseif( $ns == $special ) {
1212 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1215 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1217 wfProfileOut( $fname );
1221 # Some functions here used by doBlockLevels()
1223 /* private */ function closeParagraph() {
1225 if ( '' != $this->mLastSection
) {
1226 $result = '</' . $this->mLastSection
. ">\n";
1228 $this->mInPre
= false;
1229 $this->mLastSection
= '';
1232 # getCommon() returns the length of the longest common substring
1233 # of both arguments, starting at the beginning of both.
1235 /* private */ function getCommon( $st1, $st2 ) {
1236 $fl = strlen( $st1 );
1237 $shorter = strlen( $st2 );
1238 if ( $fl < $shorter ) { $shorter = $fl; }
1240 for ( $i = 0; $i < $shorter; ++
$i ) {
1241 if ( $st1{$i} != $st2{$i} ) { break; }
1245 # These next three functions open, continue, and close the list
1246 # element appropriate to the prefix character passed into them.
1248 /* private */ function openList( $char )
1250 $result = $this->closeParagraph();
1252 if ( '*' == $char ) { $result .= '<ul><li>'; }
1253 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1254 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1255 else if ( ';' == $char ) {
1256 $result .= '<dl><dt>';
1257 $this->mDTopen
= true;
1259 else { $result = '<!-- ERR 1 -->'; }
1264 /* private */ function nextItem( $char ) {
1265 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1266 else if ( ':' == $char ||
';' == $char ) {
1268 if ( $this->mDTopen
) { $close = '</dt>'; }
1269 if ( ';' == $char ) {
1270 $this->mDTopen
= true;
1271 return $close . '<dt>';
1273 $this->mDTopen
= false;
1274 return $close . '<dd>';
1277 return '<!-- ERR 2 -->';
1280 /* private */function closeList( $char ) {
1281 if ( '*' == $char ) { $text = '</li></ul>'; }
1282 else if ( '#' == $char ) { $text = '</li></ol>'; }
1283 else if ( ':' == $char ) {
1284 if ( $this->mDTopen
) {
1285 $this->mDTopen
= false;
1286 $text = '</dt></dl>';
1288 $text = '</dd></dl>';
1291 else { return '<!-- ERR 3 -->'; }
1295 /* private */ function doBlockLevels( $text, $linestart ) {
1296 $fname = 'Parser::doBlockLevels';
1297 wfProfileIn( $fname );
1299 # Parsing through the text line by line. The main thing
1300 # happening here is handling of block-level elements p, pre,
1301 # and making lists from lines starting with * # : etc.
1303 $textLines = explode( "\n", $text );
1305 $lastPrefix = $output = $lastLine = '';
1306 $this->mDTopen
= $inBlockElem = false;
1308 $paragraphStack = false;
1310 if ( !$linestart ) {
1311 $output .= array_shift( $textLines );
1313 foreach ( $textLines as $oLine ) {
1314 $lastPrefixLength = strlen( $lastPrefix );
1315 $preCloseMatch = preg_match("/<\\/pre/i", $oLine );
1316 $preOpenMatch = preg_match("/<pre/i", $oLine );
1317 if ( !$this->mInPre
) {
1318 # Multiple prefixes may abut each other for nested lists.
1319 $prefixLength = strspn( $oLine, '*#:;' );
1320 $pref = substr( $oLine, 0, $prefixLength );
1323 $pref2 = str_replace( ';', ':', $pref );
1324 $t = substr( $oLine, $prefixLength );
1325 $this->mInPre
= !empty($preOpenMatch);
1327 # Don't interpret any other prefixes in preformatted text
1329 $pref = $pref2 = '';
1334 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1335 # Same as the last item, so no need to deal with nesting or opening stuff
1336 $output .= $this->nextItem( substr( $pref, -1 ) );
1337 $paragraphStack = false;
1339 if ( ";" == substr( $pref, -1 ) ) {
1340 # The one nasty exception: definition lists work like this:
1341 # ; title : definition text
1342 # So we check for : in the remainder text to split up the
1343 # title and definition, without b0rking links.
1344 # FIXME: This is not foolproof. Something better in Tokenizer might help.
1345 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1347 $output .= $term . $this->nextItem( ':' );
1351 } elseif( $prefixLength ||
$lastPrefixLength ) {
1352 # Either open or close a level...
1353 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1354 $paragraphStack = false;
1356 while( $commonPrefixLength < $lastPrefixLength ) {
1357 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1358 --$lastPrefixLength;
1360 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1361 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1363 while ( $prefixLength > $commonPrefixLength ) {
1364 $char = substr( $pref, $commonPrefixLength, 1 );
1365 $output .= $this->openList( $char );
1367 if ( ';' == $char ) {
1368 # FIXME: This is dupe of code above
1369 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1371 $output .= $term . $this->nextItem( ":" );
1375 ++
$commonPrefixLength;
1377 $lastPrefix = $pref2;
1379 if( 0 == $prefixLength ) {
1380 # No prefix (not in list)--go to paragraph mode
1381 $uniq_prefix = UNIQ_PREFIX
;
1382 // XXX: use a stack for nestable elements like span, table and div
1383 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1384 $closematch = preg_match(
1385 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1386 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1387 if ( $openmatch or $closematch ) {
1388 $paragraphStack = false;
1389 $output .= $this->closeParagraph();
1390 if($preOpenMatch and !$preCloseMatch) {
1391 $this->mInPre
= true;
1393 if ( $closematch ) {
1394 $inBlockElem = false;
1396 $inBlockElem = true;
1398 } else if ( !$inBlockElem && !$this->mInPre
) {
1399 if ( " " == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1401 if ($this->mLastSection
!= 'pre') {
1402 $paragraphStack = false;
1403 $output .= $this->closeParagraph().'<pre>';
1404 $this->mLastSection
= 'pre';
1408 if ( '' == trim($t) ) {
1409 if ( $paragraphStack ) {
1410 $output .= $paragraphStack.'<br />';
1411 $paragraphStack = false;
1412 $this->mLastSection
= 'p';
1414 if ($this->mLastSection
!= 'p' ) {
1415 $output .= $this->closeParagraph();
1416 $this->mLastSection
= '';
1417 $paragraphStack = '<p>';
1419 $paragraphStack = '</p><p>';
1423 if ( $paragraphStack ) {
1424 $output .= $paragraphStack;
1425 $paragraphStack = false;
1426 $this->mLastSection
= 'p';
1427 } else if ($this->mLastSection
!= 'p') {
1428 $output .= $this->closeParagraph().'<p>';
1429 $this->mLastSection
= 'p';
1435 if ($paragraphStack === false) {
1439 while ( $prefixLength ) {
1440 $output .= $this->closeList( $pref2{$prefixLength-1} );
1443 if ( '' != $this->mLastSection
) {
1444 $output .= '</' . $this->mLastSection
. '>';
1445 $this->mLastSection
= '';
1448 wfProfileOut( $fname );
1452 # Return value of a magic variable (like PAGENAME)
1453 function getVariableValue( $index ) {
1454 global $wgLang, $wgSitename, $wgServer;
1457 case MAG_CURRENTMONTH
:
1458 return $wgLang->formatNum( date( 'm' ) );
1459 case MAG_CURRENTMONTHNAME
:
1460 return $wgLang->getMonthName( date('n') );
1461 case MAG_CURRENTMONTHNAMEGEN
:
1462 return $wgLang->getMonthNameGen( date('n') );
1463 case MAG_CURRENTDAY
:
1464 return $wgLang->formatNum( date('j') );
1466 return $this->mTitle
->getText();
1468 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1469 return $wgLang->getNsText($this->mTitle
->getNamespace()); // Patch by Dori
1470 case MAG_CURRENTDAYNAME
:
1471 return $wgLang->getWeekdayName( date('w')+
1 );
1472 case MAG_CURRENTYEAR
:
1473 return $wgLang->formatNum( date( 'Y' ) );
1474 case MAG_CURRENTTIME
:
1475 return $wgLang->time( wfTimestampNow(), false );
1476 case MAG_NUMBEROFARTICLES
:
1477 return $wgLang->formatNum( wfNumberOfArticles() );
1487 # initialise the magic variables (like CURRENTMONTHNAME)
1488 function initialiseVariables() {
1489 global $wgVariableIDs;
1490 $this->mVariables
= array();
1491 foreach ( $wgVariableIDs as $id ) {
1492 $mw =& MagicWord
::get( $id );
1493 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1497 /* private */ function replaceVariables( $text, $args = array() ) {
1498 global $wgLang, $wgScript, $wgArticlePath;
1500 # Prevent too big inclusions
1501 if(strlen($text)> MAX_INCLUDE_SIZE
)
1504 $fname = 'Parser::replaceVariables';
1505 wfProfileIn( $fname );
1508 $titleChars = Title
::legalChars();
1509 $nonBraceChars = str_replace( array( '{', '}' ), array( '', '' ), $titleChars );
1511 # This function is called recursively. To keep track of arguments we need a stack:
1512 array_push( $this->mArgStack
, $args );
1514 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1515 $GLOBALS['wgCurParser'] =& $this;
1518 if ( $this->mOutputType
== OT_HTML
) {
1519 # Variable substitution
1520 $text = preg_replace_callback( "/{{([$nonBraceChars]*?)}}/", 'wfVariableSubstitution', $text );
1522 # Argument substitution
1523 $text = preg_replace_callback( "/(\\n?){{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1525 # Template substitution
1526 $regex = '/(\\n?){{(['.$nonBraceChars.']*)(\\|.*?|)}}/s';
1527 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1529 array_pop( $this->mArgStack
);
1531 wfProfileOut( $fname );
1535 function variableSubstitution( $matches ) {
1536 if ( !$this->mVariables
) {
1537 $this->initialiseVariables();
1539 if ( array_key_exists( $matches[1], $this->mVariables
) ) {
1540 $text = $this->mVariables
[$matches[1]];
1541 $this->mOutput
->mContainsOldMagic
= true;
1543 $text = $matches[0];
1548 # Split template arguments
1549 function getTemplateArgs( $argsString ) {
1550 if ( $argsString === '' ) {
1554 $args = explode( '|', substr( $argsString, 1 ) );
1556 # If any of the arguments contains a '[[' but no ']]', it needs to be
1557 # merged with the next arg because the '|' character between belongs
1558 # to the link syntax and not the template parameter syntax.
1559 $argc = count($args);
1561 for ( $i = 0; $i < $argc-1; $i++
) {
1562 if ( substr_count ( $args[$i], "[[" ) != substr_count ( $args[$i], "]]" ) ) {
1563 $args[$i] .= "|".$args[$i+
1];
1564 array_splice($args, $i+
1, 1);
1573 function braceSubstitution( $matches ) {
1574 global $wgLinkCache, $wgLang;
1575 $fname = 'Parser::braceSubstitution';
1582 # $newline is an optional newline character before the braces
1583 # $part1 is the bit before the first |, and must contain only title characters
1584 # $args is a list of arguments, starting from index 0, not including $part1
1586 $newline = $matches[1];
1587 $part1 = $matches[2];
1588 # If the third subpattern matched anything, it will start with |
1590 $args = $this->getTemplateArgs($matches[3]);
1591 $argc = count( $args );
1594 if ( strpos( $matches[0], '{{{' ) !== false ) {
1595 $text = $matches[0];
1602 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1603 if ( $mwSubst->matchStartAndRemove( $part1 ) ) {
1604 if ( $this->mOutputType
!= OT_WIKI
) {
1605 # Invalid SUBST not replaced at PST time
1606 # Return without further processing
1607 $text = $matches[0];
1611 } elseif ( $this->mOutputType
== OT_WIKI
) {
1612 # SUBST not found in PST pass, do nothing
1613 $text = $matches[0];
1618 # MSG, MSGNW and INT
1621 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1622 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1625 # Remove obsolete MSG:
1626 $mwMsg =& MagicWord
::get( MAG_MSG
);
1627 $mwMsg->matchStartAndRemove( $part1 );
1630 # Check if it is an internal message
1631 $mwInt =& MagicWord
::get( MAG_INT
);
1632 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1633 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1634 $text = wfMsgReal( $part1, $args, true );
1642 # Check for NS: (namespace expansion)
1643 $mwNs = MagicWord
::get( MAG_NS
);
1644 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1645 if ( intval( $part1 ) ) {
1646 $text = $wgLang->getNsText( intval( $part1 ) );
1649 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1650 if ( !is_null( $index ) ) {
1651 $text = $wgLang->getNsText( $index );
1658 # LOCALURL and LOCALURLE
1660 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1661 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1663 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1664 $func = 'getLocalURL';
1665 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1666 $func = 'escapeLocalURL';
1671 if ( $func !== '' ) {
1672 $title = Title
::newFromText( $part1 );
1673 if ( !is_null( $title ) ) {
1675 $text = $title->$func( $args[0] );
1677 $text = $title->$func();
1684 # Internal variables
1685 if ( !$this->mVariables
) {
1686 $this->initialiseVariables();
1688 if ( !$found && array_key_exists( $part1, $this->mVariables
) ) {
1689 $text = $this->mVariables
[$part1];
1691 $this->mOutput
->mContainsOldMagic
= true;
1694 # Template table test
1696 # Did we encounter this template already? If yes, it is in the cache
1697 # and we need to check for loops.
1698 if ( isset( $this->mTemplates
[$part1] ) ) {
1699 # Infinite loop test
1700 if ( isset( $this->mTemplatePath
[$part1] ) ) {
1704 # set $text to cached message.
1705 $text = $this->mTemplates
[$part1];
1709 # Load from database
1711 $title = Title
::newFromText( $part1, NS_TEMPLATE
);
1712 if ( !is_null( $title ) && !$title->isExternal() ) {
1713 # Check for excessive inclusion
1714 $dbk = $title->getPrefixedDBkey();
1715 if ( $this->incrementIncludeCount( $dbk ) ) {
1716 # This should never be reached.
1717 $article = new Article( $title );
1718 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1719 if ( $articleContent !== false ) {
1721 $text = $articleContent;
1726 # If the title is valid but undisplayable, make a link to it
1727 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1728 $text = '[[' . $title->getPrefixedText() . ']]';
1732 # Template cache array insertion
1733 $this->mTemplates
[$part1] = $text;
1737 # Recursive parsing, escaping and link table handling
1738 # Only for HTML output
1739 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1740 $text = wfEscapeWikiText( $text );
1741 } elseif ( $this->mOutputType
== OT_HTML
&& $found && !$noparse) {
1742 # Clean up argument array
1743 $assocArgs = array();
1745 foreach( $args as $arg ) {
1746 $eqpos = strpos( $arg, '=' );
1747 if ( $eqpos === false ) {
1748 $assocArgs[$index++
] = $arg;
1750 $name = trim( substr( $arg, 0, $eqpos ) );
1751 $value = trim( substr( $arg, $eqpos+
1 ) );
1752 if ( $value === false ) {
1755 if ( $name !== false ) {
1756 $assocArgs[$name] = $value;
1761 # Do not enter included links in link table
1762 if ( !is_null( $title ) ) {
1763 $wgLinkCache->suspend();
1766 # Add a new element to the templace recursion path
1767 $this->mTemplatePath
[$part1] = 1;
1769 # Run full parser on the included text
1770 $text = $this->internalParse( $text, $newline, $assocArgs );
1771 # I replaced the line below with the line above, as it former seems to cause several bugs
1772 #$text = $this->stripParse( $text, $newline, $assocArgs );
1774 # Resume the link cache and register the inclusion as a link
1775 if ( !is_null( $title ) ) {
1776 $wgLinkCache->resume();
1777 $wgLinkCache->addLinkObj( $title );
1780 # Empties the template path
1781 $this->mTemplatePath
= array();
1790 # Triple brace replacement -- used for template arguments
1791 function argSubstitution( $matches ) {
1792 $newline = $matches[1];
1793 $arg = trim( $matches[2] );
1794 $text = $matches[0];
1795 $inputArgs = end( $this->mArgStack
);
1797 if ( array_key_exists( $arg, $inputArgs ) ) {
1798 $text = $this->stripParse( $inputArgs[$arg], $newline, array() );
1804 # Returns true if the function is allowed to include this entity
1805 function incrementIncludeCount( $dbk ) {
1806 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1807 $this->mIncludeCount
[$dbk] = 0;
1809 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1817 # Cleans up HTML, removes dangerous tags and attributes
1818 /* private */ function removeHTMLtags( $text ) {
1819 global $wgUseTidy, $wgUserHtml;
1820 $fname = 'Parser::removeHTMLtags';
1821 wfProfileIn( $fname );
1824 $htmlpairs = array( # Tags that must be closed
1825 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
1826 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
1827 'strike', 'strong', 'tt', 'var', 'div', 'center',
1828 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
1829 'ruby', 'rt' , 'rb' , 'rp', 'p'
1831 $htmlsingle = array(
1832 'br', 'hr', 'li', 'dt', 'dd'
1834 $htmlnest = array( # Tags that can be nested--??
1835 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
1836 'dl', 'font', 'big', 'small', 'sub', 'sup'
1838 $tabletags = array( # Can only appear inside table
1842 $htmlpairs = array();
1843 $htmlsingle = array();
1844 $htmlnest = array();
1845 $tabletags = array();
1848 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1849 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1851 $htmlattrs = $this->getHTMLattrs () ;
1853 # Remove HTML comments
1854 $text = preg_replace( '/(\\n *<!--.*--> *(?=\\n)|<!--.*-->)/sU', '$2', $text );
1856 $bits = explode( '<', $text );
1857 $text = array_shift( $bits );
1859 $tagstack = array(); $tablestack = array();
1860 foreach ( $bits as $x ) {
1861 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
1862 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1864 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1865 error_reporting( $prev );
1868 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1872 if ( ! in_array( $t, $htmlsingle ) &&
1873 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
1874 @array_push
( $tagstack, $ot );
1877 if ( $t == 'table' ) {
1878 $tagstack = array_pop( $tablestack );
1883 # Keep track for later
1884 if ( in_array( $t, $tabletags ) &&
1885 ! in_array( 'table', $tagstack ) ) {
1887 } else if ( in_array( $t, $tagstack ) &&
1888 ! in_array ( $t , $htmlnest ) ) {
1890 } else if ( ! in_array( $t, $htmlsingle ) ) {
1891 if ( $t == 'table' ) {
1892 array_push( $tablestack, $tagstack );
1893 $tagstack = array();
1895 array_push( $tagstack, $t );
1897 # Strip non-approved attributes from the tag
1898 $newparams = $this->fixTagAttributes($params);
1902 $rest = str_replace( '>', '>', $rest );
1903 $text .= "<$slash$t $newparams$brace$rest";
1907 $text .= '<' . str_replace( '>', '>', $x);
1909 # Close off any remaining tags
1910 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
1912 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
1915 # this might be possible using tidy itself
1916 foreach ( $bits as $x ) {
1917 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1919 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1920 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1921 $newparams = $this->fixTagAttributes($params);
1922 $rest = str_replace( '>', '>', $rest );
1923 $text .= "<$slash$t $newparams$brace$rest";
1925 $text .= '<' . str_replace( '>', '>', $x);
1929 wfProfileOut( $fname );
1936 * This function accomplishes several tasks:
1937 * 1) Auto-number headings if that option is enabled
1938 * 2) Add an [edit] link to sections for logged in users who have enabled the option
1939 * 3) Add a Table of contents on the top for users who have enabled the option
1940 * 4) Auto-anchor headings
1942 * It loops through all headlines, collects the necessary data, then splits up the
1943 * string and re-inserts the newly formatted headlines.
1947 /* private */ function formatHeadings( $text, $isMain=true ) {
1948 global $wgInputEncoding, $wgMaxTocLevel;
1950 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
1951 $doShowToc = $this->mOptions
->getShowToc();
1952 $forceTocHere = false;
1953 if( !$this->mTitle
->userCanEdit() ) {
1955 $rightClickHack = 0;
1957 $showEditLink = $this->mOptions
->getEditSection();
1958 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
1961 # Inhibit editsection links if requested in the page
1962 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
1963 if( $esw->matchAndRemove( $text ) ) {
1966 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1968 $mw =& MagicWord
::get( MAG_NOTOC
);
1969 if( $mw->matchAndRemove( $text ) ) {
1973 # never add the TOC to the Main Page. This is an entry page that should not
1974 # be more than 1-2 screens large anyway
1975 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
1979 # Get all headlines for numbering them and adding funky stuff like [edit]
1980 # links - this is for later, but we need the number of headlines right now
1981 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
1983 # if there are fewer than 4 headlines in the article, do not show TOC
1984 if( $numMatches < 4 ) {
1988 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
1989 # override above conditions and always show TOC at that place
1990 $mw =& MagicWord
::get( MAG_TOC
);
1991 if ($mw->match( $text ) ) {
1993 $forceTocHere = true;
1995 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
1996 # override above conditions and always show TOC above first header
1997 $mw =& MagicWord
::get( MAG_FORCETOC
);
1998 if ($mw->matchAndRemove( $text ) ) {
2005 # We need this to perform operations on the HTML
2006 $sk =& $this->mOptions
->getSkin();
2011 # Ugh .. the TOC should have neat indentation levels which can be
2012 # passed to the skin functions. These are determined here
2017 $sublevelCount = array();
2020 foreach( $matches[3] as $headline ) {
2023 $prevlevel = $level;
2025 $level = $matches[1][$headlineCount];
2026 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2027 # reset when we enter a new level
2028 $sublevelCount[$level] = 0;
2029 $toc .= $sk->tocIndent( $level - $prevlevel );
2030 $toclevel +
= $level - $prevlevel;
2032 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2033 # reset when we step back a level
2034 $sublevelCount[$level+
1]=0;
2035 $toc .= $sk->tocUnindent( $prevlevel - $level );
2036 $toclevel -= $prevlevel - $level;
2038 # count number of headlines for each level
2039 @$sublevelCount[$level]++
;
2040 if( $doNumberHeadings ||
$doShowToc ) {
2042 for( $i = 1; $i <= $level; $i++
) {
2043 if( !empty( $sublevelCount[$i] ) ) {
2047 $numbering .= $sublevelCount[$i];
2053 # The canonized header is a version of the header text safe to use for links
2054 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2055 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2056 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2059 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2060 $tocline = trim( $canonized_headline );
2061 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2062 $replacearray = array(
2066 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2067 $refer[$headlineCount] = $canonized_headline;
2069 # count how many in assoc. array so we can track dupes in anchors
2070 @$refers[$canonized_headline]++
;
2071 $refcount[$headlineCount]=$refers[$canonized_headline];
2073 # Prepend the number to the heading text
2075 if( $doNumberHeadings ||
$doShowToc ) {
2076 $tocline = $numbering . ' ' . $tocline;
2078 # Don't number the heading if it is the only one (looks silly)
2079 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2080 # the two are different if the line contains a link
2081 $headline=$numbering . ' ' . $headline;
2085 # Create the anchor for linking from the TOC to the section
2086 $anchor = $canonized_headline;
2087 if($refcount[$headlineCount] > 1 ) {
2088 $anchor .= '_' . $refcount[$headlineCount];
2090 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2091 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2093 if( $showEditLink ) {
2094 if ( empty( $head[$headlineCount] ) ) {
2095 $head[$headlineCount] = '';
2097 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+
1);
2100 # Add the edit section span
2101 if( $rightClickHack ) {
2102 $headline = $sk->editSectionScript($headlineCount+
1,$headline);
2105 # give headline the correct <h#> tag
2106 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline."</h".$level.">";
2112 $toclines = $headlineCount;
2113 $toc .= $sk->tocUnindent( $toclevel );
2114 $toc = $sk->tocTable( $toc );
2117 # split up and insert constructed headlines
2119 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2122 foreach( $blocks as $block ) {
2123 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2124 # This is the [edit] link that appears for the top block of text when
2125 # section editing is enabled
2127 # Disabled because it broke block formatting
2128 # For example, a bullet point in the top line
2129 # $full .= $sk->editSectionLink(0);
2132 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2133 # Top anchor now in skin
2137 if( !empty( $head[$i] ) ) {
2143 $mw =& MagicWord
::get( MAG_TOC
);
2144 return $mw->replace( $toc, $full );
2150 # Return an HTML link for the "ISBN 123456" text
2151 /* private */ function magicISBN( $text ) {
2153 $fname = 'Parser::magicISBN';
2154 wfProfileIn( $fname );
2156 $a = split( 'ISBN ', " $text" );
2157 if ( count ( $a ) < 2 ) {
2158 wfProfileOut( $fname );
2161 $text = substr( array_shift( $a ), 1);
2162 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2164 foreach ( $a as $x ) {
2165 $isbn = $blank = '' ;
2166 while ( ' ' == $x{0} ) {
2168 $x = substr( $x, 1 );
2170 while ( strstr( $valid, $x{0} ) != false ) {
2172 $x = substr( $x, 1 );
2174 $num = str_replace( '-', '', $isbn );
2175 $num = str_replace( ' ', '', $num );
2178 $text .= "ISBN $blank$x";
2180 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2181 $text .= '<a href="' .
2182 $titleObj->escapeLocalUrl( "isbn={$num}" ) .
2183 "\" class=\"internal\">ISBN $isbn</a>";
2187 wfProfileOut( $fname );
2191 # Return an HTML link for the "GEO ..." text
2192 /* private */ function magicGEO( $text ) {
2193 global $wgLang, $wgUseGeoMode;
2194 if ( !isset ( $wgUseGeoMode ) ||
!$wgUseGeoMode ) return $text ;
2195 $fname = 'Parser::magicGEO';
2196 wfProfileIn( $fname );
2198 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2199 $directions = array ( "N" => "North" , "S" => "South" , "E" => "East" , "W" => "West" ) ;
2200 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2201 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2202 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2203 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2205 $a = split( 'GEO ', " $text" );
2206 if ( count ( $a ) < 2 ) {
2207 wfProfileOut( $fname );
2210 $text = substr( array_shift( $a ), 1);
2211 $valid = '0123456789.+-:';
2213 foreach ( $a as $x ) {
2214 $geo = $blank = '' ;
2215 while ( ' ' == $x{0} ) {
2217 $x = substr( $x, 1 );
2219 while ( strstr( $valid, $x{0} ) != false ) {
2221 $x = substr( $x, 1 );
2223 $num = str_replace( '+', '', $geo );
2224 $num = str_replace( ' ', '', $num );
2226 if ( '' == $num ||
count ( explode ( ":" , $num , 3 ) ) < 2 ) {
2227 $text .= "GEO $blank$x";
2229 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2230 $text .= '<a href="' .
2231 $titleObj->escapeLocalUrl( "coordinates={$num}" ) .
2232 "\" class=\"internal\">GEO $geo</a>";
2236 wfProfileOut( $fname );
2240 # Return an HTML link for the "RFC 1234" text
2241 /* private */ function magicRFC( $text ) {
2244 $a = split( 'RFC ', ' '.$text );
2245 if ( count ( $a ) < 2 ) return $text;
2246 $text = substr( array_shift( $a ), 1);
2247 $valid = '0123456789';
2249 foreach ( $a as $x ) {
2250 $rfc = $blank = '' ;
2251 while ( ' ' == $x{0} ) {
2253 $x = substr( $x, 1 );
2255 while ( strstr( $valid, $x{0} ) != false ) {
2257 $x = substr( $x, 1 );
2261 $text .= "RFC $blank$x";
2263 $url = wfmsg( 'rfcurl' );
2264 $url = str_replace( '$1', $rfc, $url);
2265 $sk =& $this->mOptions
->getSkin();
2266 $la = $sk->getExternalLinkAttributes( $url, "RFC {$rfc}" );
2267 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2273 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2274 $this->mOptions
= $options;
2275 $this->mTitle
=& $title;
2276 $this->mOutputType
= OT_WIKI
;
2278 if ( $clearState ) {
2279 $this->clearState();
2282 $stripState = false;
2286 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2290 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2291 "/<br *?>/i" => "<br />",
2293 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2295 $text = $this->strip( $text, $stripState, false );
2296 $text = $this->pstPass2( $text, $user );
2297 $text = $this->unstrip( $text, $stripState );
2298 $text = $this->unstripNoWiki( $text, $stripState );
2302 /* private */ function pstPass2( $text, &$user ) {
2303 global $wgLang, $wgLocaltimezone, $wgCurParser;
2305 # Variable replacement
2306 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2307 $text = $this->replaceVariables( $text );
2311 $n = $user->getName();
2312 $k = $user->getOption( 'nickname' );
2313 if ( '' == $k ) { $k = $n; }
2314 if(isset($wgLocaltimezone)) {
2315 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2317 /* Note: this is an ugly timezone hack for the European wikis */
2318 $d = $wgLang->timeanddate( date( 'YmdHis' ), false ) .
2319 ' (' . date( 'T' ) . ')';
2320 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2322 $text = preg_replace( '/~~~~~/', $d, $text );
2323 $text = preg_replace( '/~~~~/', '[[' . $wgLang->getNsText(
2324 Namespace::getUser() ) . ":$n|$k]] $d", $text );
2325 $text = preg_replace( '/~~~/', '[[' . $wgLang->getNsText(
2326 Namespace::getUser() ) . ":$n|$k]]", $text );
2328 # Context links: [[|name]] and [[name (context)|]]
2330 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2331 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2332 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2333 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2335 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2336 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2337 $p3 = "/\[\[($namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]]
2338 $p4 = "/\[\[($namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/";
2339 # [[ns:page (cont)|]]
2341 $t = $this->mTitle
->getText();
2342 if ( preg_match( $conpat, $t, $m ) ) {
2345 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2346 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2347 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2349 if ( '' == $context ) {
2350 $text = preg_replace( $p2, '[[\\1]]', $text );
2352 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2356 $mw =& MagicWord::get( MAG_SUBST );
2357 $wgCurParser = $this->fork();
2358 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
2359 $this->merge( $wgCurParser );
2362 # Trim trailing whitespace
2363 # MAG_END (__END__) tag allows for trailing
2364 # whitespace to be deliberately included
2365 $text = rtrim( $text );
2366 $mw =& MagicWord
::get( MAG_END
);
2367 $mw->matchAndRemove( $text );
2372 # Set up some variables which are usually set up in parse()
2373 # so that an external function can call some class members with confidence
2374 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2375 $this->mTitle
=& $title;
2376 $this->mOptions
= $options;
2377 $this->mOutputType
= $outputType;
2378 if ( $clearState ) {
2379 $this->clearState();
2383 function transformMsg( $text, $options ) {
2385 static $executing = false;
2387 # Guard against infinite recursion
2393 $this->mTitle
= $wgTitle;
2394 $this->mOptions
= $options;
2395 $this->mOutputType
= OT_MSG
;
2396 $this->clearState();
2397 $text = $this->replaceVariables( $text );
2403 # Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2404 # Callback will be called with the text within
2405 # Transform and return the text within
2406 function setHook( $tag, $callback ) {
2407 $oldVal = @$this->mTagHooks
[$tag];
2408 $this->mTagHooks
[$tag] = $callback;
2415 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2416 var $mCacheTime; # Used in ParserCache
2418 function ParserOutput( $text = "", $languageLinks = array(), $categoryLinks = array(),
2419 $containsOldMagic = false )
2421 $this->mText
= $text;
2422 $this->mLanguageLinks
= $languageLinks;
2423 $this->mCategoryLinks
= $categoryLinks;
2424 $this->mContainsOldMagic
= $containsOldMagic;
2425 $this->mCacheTime
= "";
2428 function getText() { return $this->mText
; }
2429 function getLanguageLinks() { return $this->mLanguageLinks
; }
2430 function getCategoryLinks() { return $this->mCategoryLinks
; }
2431 function getCacheTime() { return $this->mCacheTime
; }
2432 function containsOldMagic() { return $this->mContainsOldMagic
; }
2433 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2434 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2435 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2436 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2437 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
2439 function merge( $other ) {
2440 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2441 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2442 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2449 # All variables are private
2450 var $mUseTeX; # Use texvc to expand <math> tags
2451 var $mUseCategoryMagic; # Treat [[Category:xxxx]] tags specially
2452 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2453 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2454 var $mAllowExternalImages; # Allow external images inline
2455 var $mSkin; # Reference to the preferred skin
2456 var $mDateFormat; # Date format index
2457 var $mEditSection; # Create "edit section" links
2458 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2459 var $mNumberHeadings; # Automatically number headings
2460 var $mShowToc; # Show table of contents
2462 function getUseTeX() { return $this->mUseTeX
; }
2463 function getUseCategoryMagic() { return $this->mUseCategoryMagic
; }
2464 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2465 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2466 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2467 function getSkin() { return $this->mSkin
; }
2468 function getDateFormat() { return $this->mDateFormat
; }
2469 function getEditSection() { return $this->mEditSection
; }
2470 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
2471 function getNumberHeadings() { return $this->mNumberHeadings
; }
2472 function getShowToc() { return $this->mShowToc
; }
2474 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2475 function setUseCategoryMagic( $x ) { return wfSetVar( $this->mUseCategoryMagic
, $x ); }
2476 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2477 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2478 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2479 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2480 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
2481 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
2482 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
2483 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
2485 function setSkin( &$x ) { $this->mSkin
=& $x; }
2487 /* static */ function newFromUser( &$user ) {
2488 $popts = new ParserOptions
;
2489 $popts->initialiseFromUser( $user );
2493 function initialiseFromUser( &$userInput ) {
2494 global $wgUseTeX, $wgUseCategoryMagic, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2496 if ( !$userInput ) {
2498 $user->setLoaded( true );
2500 $user =& $userInput;
2503 $this->mUseTeX
= $wgUseTeX;
2504 $this->mUseCategoryMagic
= $wgUseCategoryMagic;
2505 $this->mUseDynamicDates
= $wgUseDynamicDates;
2506 $this->mInterwikiMagic
= $wgInterwikiMagic;
2507 $this->mAllowExternalImages
= $wgAllowExternalImages;
2508 $this->mSkin
=& $user->getSkin();
2509 $this->mDateFormat
= $user->getOption( 'date' );
2510 $this->mEditSection
= $user->getOption( 'editsection' );
2511 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
2512 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
2513 $this->mShowToc
= $user->getOption( 'showtoc' );
2519 # Regex callbacks, used in Parser::replaceVariables
2520 function wfBraceSubstitution( $matches )
2522 global $wgCurParser;
2523 return $wgCurParser->braceSubstitution( $matches );
2526 function wfArgSubstitution( $matches )
2528 global $wgCurParser;
2529 return $wgCurParser->argSubstitution( $matches );
2532 function wfVariableSubstitution( $matches )
2534 global $wgCurParser;
2535 return $wgCurParser->variableSubstitution( $matches );