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 );
825 $text = $this->removeHTMLtags( $text );
826 $text = $this->replaceVariables( $text, $args );
828 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
830 $text = $this->doHeadings( $text );
831 if($this->mOptions
->getUseDynamicDates()) {
832 global $wgDateFormatter;
833 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
835 $text = $this->doAllQuotes( $text );
836 // $text = $this->doExponent( $text );
837 $text = $this->replaceExternalLinks( $text );
838 $text = $this->doMagicLinks( $text );
839 $text = $this->replaceInternalLinks ( $text );
840 $text = $this->replaceInternalLinks ( $text );
841 //$text = $this->doTokenizedParser ( $text );
842 $text = $this->doTableStuff( $text );
843 $text = $this->formatHeadings( $text, $isMain );
844 $sk =& $this->mOptions
->getSkin();
845 $text = $sk->transformContent( $text );
847 if ( $isMain && !isset ( $this->categoryMagicDone
) ) {
848 $text .= $this->categoryMagic () ;
849 $this->categoryMagicDone
= true ;
852 wfProfileOut( $fname );
856 /* private */ function &doMagicLinks( &$text ) {
857 $text = $this->magicISBN( $text );
858 $text = $this->magicGEO( $text );
859 $text = $this->magicRFC( $text );
863 # Parse ^^ tokens and return html
864 /* private */ function doExponent ( $text )
866 $fname = 'Parser::doExponent';
867 wfProfileIn( $fname);
868 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
869 wfProfileOut( $fname);
873 # Parse headers and return html
874 /* private */ function doHeadings( $text ) {
875 $fname = 'Parser::doHeadings';
876 wfProfileIn( $fname );
877 for ( $i = 6; $i >= 1; --$i ) {
878 $h = substr( '======', 0, $i );
879 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
880 "<h{$i}>\\1</h{$i}>\\2", $text );
882 wfProfileOut( $fname );
886 /* private */ function doAllQuotes( $text ) {
887 $fname = 'Parser::doAllQuotes';
888 wfProfileIn( $fname );
890 $lines = explode( "\n", $text );
891 foreach ( $lines as $line ) {
892 $outtext .= $this->doQuotes ( '', $line, '' ) . "\n";
894 $outtext = substr($outtext, 0,-1);
895 wfProfileOut( $fname );
899 /* private */ function doQuotes( $pre, $text, $mode ) {
900 if ( preg_match( "/^(.*)''(.*)$/sU", $text, $m ) ) {
901 $m1_strong = ($m[1] == "") ?
"" : "<strong>{$m[1]}</strong>";
902 $m1_em = ($m[1] == "") ?
"" : "<em>{$m[1]}</em>";
903 if ( substr ($m[2], 0, 1) == '\'' ) {
904 $m[2] = substr ($m[2], 1);
906 return $this->doQuotes ( $m[1], $m[2], ($m[1] == '') ?
'both' : 'emstrong' );
907 } else if ($mode == 'strong') {
908 return $m1_strong . $this->doQuotes ( '', $m[2], '' );
909 } else if (($mode == 'emstrong') ||
($mode == 'both')) {
910 return $this->doQuotes ( '', $pre.$m1_strong.$m[2], 'em' );
911 } else if ($mode == 'strongem') {
912 return "<strong>{$pre}{$m1_em}</strong>" . $this->doQuotes ( '', $m[2], 'em' );
914 return $m[1] . $this->doQuotes ( '', $m[2], 'strong' );
917 if ($mode == 'strong') {
918 return $this->doQuotes ( $m[1], $m[2], ($m[1] == '') ?
'both' : 'strongem' );
919 } else if ($mode == 'em') {
920 return $m1_em . $this->doQuotes ( '', $m[2], '' );
921 } else if ($mode == 'emstrong') {
922 return "<em>{$pre}{$m1_strong}</em>" . $this->doQuotes ( '', $m[2], 'strong' );
923 } else if (($mode == 'strongem') ||
($mode == 'both')) {
924 return $this->doQuotes ( '', $pre.$m1_em.$m[2], 'strong' );
926 return $m[1] . $this->doQuotes ( '', $m[2], 'em' );
930 $text_strong = ($text == '') ?
'' : "<strong>{$text}</strong>";
931 $text_em = ($text == '') ?
'' : "<em>{$text}</em>";
934 } else if ($mode == 'em') {
935 return $pre . $text_em;
936 } else if ($mode == 'strong') {
937 return $pre . $text_strong;
938 } else if ($mode == 'strongem') {
939 return (($pre == '') && ($text == '')) ?
'' : "<strong>{$pre}{$text_em}</strong>";
941 return (($pre == '') && ($text == '')) ?
'' : "<em>{$pre}{$text_strong}</em>";
946 # Note: we have to do external links before the internal ones,
947 # and otherwise take great care in the order of things here, so
948 # that we don't end up interpreting some URLs twice.
950 /* private */ function replaceExternalLinks( $text ) {
951 $fname = 'Parser::replaceExternalLinks';
952 wfProfileIn( $fname );
953 $text = $this->subReplaceExternalLinks( $text, 'http', true );
954 $text = $this->subReplaceExternalLinks( $text, 'https', true );
955 $text = $this->subReplaceExternalLinks( $text, 'ftp', false );
956 $text = $this->subReplaceExternalLinks( $text, 'irc', false );
957 $text = $this->subReplaceExternalLinks( $text, 'gopher', false );
958 $text = $this->subReplaceExternalLinks( $text, 'news', false );
959 $text = $this->subReplaceExternalLinks( $text, 'mailto', false );
960 wfProfileOut( $fname );
964 /* private */ function subReplaceExternalLinks( $s, $protocol, $autonumber ) {
965 $unique = '4jzAfzB8hNvf4sqyO9Edd8pSmk9rE2in0Tgw3';
966 $uc = "A-Za-z0-9_\\/~%\\-+&*#?!=()@\\x80-\\xFF";
968 # this is the list of separators that should be ignored if they
969 # are the last character of an URL but that should be included
970 # if they occur within the URL, e.g. "go to www.foo.com, where .."
971 # in this case, the last comma should not become part of the URL,
972 # but in "www.foo.com/123,2342,32.htm" it should.
974 $fnc = 'A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF';
975 $images = 'gif|png|jpg|jpeg';
977 # PLEASE NOTE: The curly braces { } are not part of the regex,
978 # they are interpreted as part of the string (used to tell PHP
979 # that the content of the string should be inserted there).
980 $e1 = "/(^|[^\\[])({$protocol}:)([{$uc}{$sep}]+)\\/([{$fnc}]+)\\." .
981 "((?i){$images})([^{$uc}]|$)/";
983 $e2 = "/(^|[^\\[])({$protocol}:)(([".$uc."]|[".$sep."][".$uc."])+)([^". $uc . $sep. "]|[".$sep."]|$)/";
984 $sk =& $this->mOptions
->getSkin();
986 if ( $autonumber and $this->mOptions
->getAllowExternalImages() ) { # Use img tags only for HTTP urls
987 $s = preg_replace( $e1, '\\1' . $sk->makeImage( "{$unique}:\\3" .
988 '/\\4.\\5', '\\4.\\5' ) . '\\6', $s );
990 $s = preg_replace( $e2, '\\1' . "<a href=\"{$unique}:\\3\"" .
991 $sk->getExternalLinkAttributes( "{$unique}:\\3", wfEscapeHTML(
992 "{$unique}:\\3" ) ) . ">" . wfEscapeHTML( "{$unique}:\\3" ) .
994 $s = str_replace( $unique, $protocol, $s );
996 $a = explode( "[{$protocol}:", " " . $s );
997 $s = array_shift( $a );
998 $s = substr( $s, 1 );
1000 # Regexp for URL in square brackets
1001 $e1 = "/^([{$uc}{$sep}]+)\\](.*)\$/sD";
1002 # Regexp for URL with link text in square brackets
1003 $e2 = "/^([{$uc}{$sep}]+)\\s+([^\\]]+)\\](.*)\$/sD";
1005 foreach ( $a as $line ) {
1007 # CASE 1: Link in square brackets, e.g.
1008 # some text [http://domain.tld/some.link] more text
1009 if ( preg_match( $e1, $line, $m ) ) {
1010 $link = "{$protocol}:{$m[1]}";
1012 if ( $autonumber ) { $text = "[" . ++
$this->mAutonumber
. "]"; }
1013 else { $text = wfEscapeHTML( $link ); }
1016 # CASE 2: Link with link text and text directly following it, e.g.
1017 # This is a collection of [http://domain.tld/some.link link]s
1018 else if ( preg_match( $e2, $line, $m ) ) {
1019 $link = "{$protocol}:{$m[1]}";
1023 if ( preg_match( wfMsg ('linktrail'), $trail, $m2 ) ) {
1029 # CASE 3: Nothing matches, just output the source text
1031 $s .= "[{$protocol}:" . $line;
1035 if( $link == $text ||
preg_match( "!$protocol://" . preg_quote( $text, "/" ) . "/?$!", $link ) ) {
1038 # Expand the URL for printable version
1039 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $link ) . "</i>)</span>";
1041 $la = $sk->getExternalLinkAttributes( $link, $text );
1042 $s .= "<a href='{$link}'{$la}>{$text}</a>{$dtrail}{$paren}{$trail}";
1049 /* private */ function replaceInternalLinks( $s ) {
1050 global $wgLang, $wgLinkCache;
1051 global $wgNamespacesWithSubpages, $wgLanguageCode;
1052 static $fname = 'Parser::replaceInternalLinks' ;
1053 wfProfileIn( $fname );
1055 wfProfileIn( $fname.'-setup' );
1057 # the % is needed to support urlencoded titles as well
1058 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1059 $sk =& $this->mOptions
->getSkin();
1061 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1062 $isRedirect = $redirect->matchStart ( strtoupper ( substr ( $s , 0 , 10 ) ) ) ;
1064 $a = explode( '[[', ' ' . $s );
1065 $s = array_shift( $a );
1066 $s = substr( $s, 1 );
1068 # Match a link having the form [[namespace:link|alternate]]trail
1070 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1071 # Match the end of a line for a word that's not followed by whitespace,
1072 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1073 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1075 $useLinkPrefixExtension = $wgLang->linkPrefixExtension();
1076 # Special and Media are pseudo-namespaces; no pages actually exist in them
1077 static $image = FALSE;
1078 static $special = FALSE;
1079 static $media = FALSE;
1080 static $category = FALSE;
1081 if ( !$image ) { $image = Namespace::getImage(); }
1082 if ( !$special ) { $special = Namespace::getSpecial(); }
1083 if ( !$media ) { $media = Namespace::getMedia(); }
1084 if ( !$category ) { $category = Namespace::getCategory(); }
1086 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1088 if ( $useLinkPrefixExtension ) {
1089 if ( preg_match( $e2, $s, $m ) ) {
1090 $first_prefix = $m[2];
1093 $first_prefix = false;
1099 wfProfileOut( $fname.'-setup' );
1101 foreach ( $a as $line ) {
1102 wfProfileIn( $fname.'-prefixhandling' );
1103 if ( $useLinkPrefixExtension ) {
1104 if ( preg_match( $e2, $s, $m ) ) {
1112 $prefix = $first_prefix;
1113 $first_prefix = false;
1116 wfProfileOut( $fname.'-prefixhandling' );
1118 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1120 # fix up urlencoded title texts
1121 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1123 } else { # Invalid form; output directly
1124 $s .= $prefix . '[[' . $line ;
1128 /* Valid link forms:
1130 :Foobar -- override special treatment of prefix (images, language links)
1131 /Foobar -- convert to CurrentPage/Foobar
1132 /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1134 $c = substr($m[1],0,1);
1135 $noforce = ($c != ':');
1136 if( $c == '/' ) { # subpage
1137 if(substr($m[1],-1,1)=='/') { # / at end means we don't want the slash to be shown
1138 $m[1]=substr($m[1],1,strlen($m[1])-2);
1141 $noslash=substr($m[1],1);
1143 if(!empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()])) { # subpages allowed here
1144 $link = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1147 } # this might be changed for ugliness reasons
1149 $link = $noslash; # no subpage allowed, use standard link
1151 } elseif( $noforce ) { # no subpage
1154 $link = substr( $m[1], 1 );
1156 $wasblank = ( '' == $text );
1160 $nt = Title
::newFromText( $link );
1162 $s .= $prefix . '[[' . $line;
1165 $ns = $nt->getNamespace();
1166 $iw = $nt->getInterWiki();
1168 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
1169 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1170 $tmp = $prefix . $trail ;
1171 $s .= (trim($tmp) == '')?
'': $tmp;
1174 if ( $ns == $image ) {
1175 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1176 $wgLinkCache->addImageLinkObj( $nt );
1179 if ( $ns == $category && !$isRedirect ) {
1180 $t = $nt->getText() ;
1181 $nnt = Title
::newFromText ( Namespace::getCanonicalName($category).":".$t ) ;
1183 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1184 $t = $sk->makeLinkObj( $nnt, $t, '', '' , $prefix );
1185 $wgLinkCache->resume();
1187 $sortkey = $wasblank ?
$this->mTitle
->getPrefixedText() : $text;
1188 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1189 $this->mOutput
->mCategoryLinks
[] = $t ;
1190 $s .= $prefix . $trail ;
1194 if( ( $nt->getPrefixedText() == $this->mTitle
->getPrefixedText() ) &&
1195 ( strpos( $link, '#' ) == FALSE ) ) {
1196 # Self-links are handled specially; generally de-link and change to bold.
1197 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1201 if( $ns == $media ) {
1202 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1203 $wgLinkCache->addImageLinkObj( $nt );
1205 } elseif( $ns == $special ) {
1206 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1209 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1211 wfProfileOut( $fname );
1215 # Some functions here used by doBlockLevels()
1217 /* private */ function closeParagraph() {
1219 if ( '' != $this->mLastSection
) {
1220 $result = '</' . $this->mLastSection
. ">\n";
1222 $this->mInPre
= false;
1223 $this->mLastSection
= '';
1226 # getCommon() returns the length of the longest common substring
1227 # of both arguments, starting at the beginning of both.
1229 /* private */ function getCommon( $st1, $st2 ) {
1230 $fl = strlen( $st1 );
1231 $shorter = strlen( $st2 );
1232 if ( $fl < $shorter ) { $shorter = $fl; }
1234 for ( $i = 0; $i < $shorter; ++
$i ) {
1235 if ( $st1{$i} != $st2{$i} ) { break; }
1239 # These next three functions open, continue, and close the list
1240 # element appropriate to the prefix character passed into them.
1242 /* private */ function openList( $char )
1244 $result = $this->closeParagraph();
1246 if ( '*' == $char ) { $result .= '<ul><li>'; }
1247 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1248 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1249 else if ( ';' == $char ) {
1250 $result .= '<dl><dt>';
1251 $this->mDTopen
= true;
1253 else { $result = '<!-- ERR 1 -->'; }
1258 /* private */ function nextItem( $char ) {
1259 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1260 else if ( ':' == $char ||
';' == $char ) {
1262 if ( $this->mDTopen
) { $close = '</dt>'; }
1263 if ( ';' == $char ) {
1264 $this->mDTopen
= true;
1265 return $close . '<dt>';
1267 $this->mDTopen
= false;
1268 return $close . '<dd>';
1271 return '<!-- ERR 2 -->';
1274 /* private */function closeList( $char ) {
1275 if ( '*' == $char ) { $text = '</li></ul>'; }
1276 else if ( '#' == $char ) { $text = '</li></ol>'; }
1277 else if ( ':' == $char ) {
1278 if ( $this->mDTopen
) {
1279 $this->mDTopen
= false;
1280 $text = '</dt></dl>';
1282 $text = '</dd></dl>';
1285 else { return '<!-- ERR 3 -->'; }
1289 /* private */ function doBlockLevels( $text, $linestart ) {
1290 $fname = 'Parser::doBlockLevels';
1291 wfProfileIn( $fname );
1293 # Parsing through the text line by line. The main thing
1294 # happening here is handling of block-level elements p, pre,
1295 # and making lists from lines starting with * # : etc.
1297 $textLines = explode( "\n", $text );
1299 $lastPrefix = $output = $lastLine = '';
1300 $this->mDTopen
= $inBlockElem = false;
1302 $paragraphStack = false;
1304 if ( !$linestart ) {
1305 $output .= array_shift( $textLines );
1307 foreach ( $textLines as $oLine ) {
1308 $lastPrefixLength = strlen( $lastPrefix );
1309 $preCloseMatch = preg_match("/<\\/pre/i", $oLine );
1310 $preOpenMatch = preg_match("/<pre/i", $oLine );
1311 if ( !$this->mInPre
) {
1312 # Multiple prefixes may abut each other for nested lists.
1313 $prefixLength = strspn( $oLine, '*#:;' );
1314 $pref = substr( $oLine, 0, $prefixLength );
1317 $pref2 = str_replace( ';', ':', $pref );
1318 $t = substr( $oLine, $prefixLength );
1319 $this->mInPre
= !empty($preOpenMatch);
1321 # Don't interpret any other prefixes in preformatted text
1323 $pref = $pref2 = '';
1328 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1329 # Same as the last item, so no need to deal with nesting or opening stuff
1330 $output .= $this->nextItem( substr( $pref, -1 ) );
1331 $paragraphStack = false;
1333 if ( ";" == substr( $pref, -1 ) ) {
1334 # The one nasty exception: definition lists work like this:
1335 # ; title : definition text
1336 # So we check for : in the remainder text to split up the
1337 # title and definition, without b0rking links.
1338 # FIXME: This is not foolproof. Something better in Tokenizer might help.
1339 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1341 $output .= $term . $this->nextItem( ':' );
1345 } elseif( $prefixLength ||
$lastPrefixLength ) {
1346 # Either open or close a level...
1347 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1348 $paragraphStack = false;
1350 while( $commonPrefixLength < $lastPrefixLength ) {
1351 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1352 --$lastPrefixLength;
1354 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1355 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1357 while ( $prefixLength > $commonPrefixLength ) {
1358 $char = substr( $pref, $commonPrefixLength, 1 );
1359 $output .= $this->openList( $char );
1361 if ( ';' == $char ) {
1362 # FIXME: This is dupe of code above
1363 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1365 $output .= $term . $this->nextItem( ":" );
1369 ++
$commonPrefixLength;
1371 $lastPrefix = $pref2;
1373 if( 0 == $prefixLength ) {
1374 # No prefix (not in list)--go to paragraph mode
1375 $uniq_prefix = UNIQ_PREFIX
;
1376 // XXX: use a stack for nestable elements like span, table and div
1377 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1378 $closematch = preg_match(
1379 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1380 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1381 if ( $openmatch or $closematch ) {
1382 $paragraphStack = false;
1383 $output .= $this->closeParagraph();
1384 if($preOpenMatch and !$preCloseMatch) {
1385 $this->mInPre
= true;
1387 if ( $closematch ) {
1388 $inBlockElem = false;
1390 $inBlockElem = true;
1392 } else if ( !$inBlockElem && !$this->mInPre
) {
1393 if ( " " == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1395 if ($this->mLastSection
!= 'pre') {
1396 $paragraphStack = false;
1397 $output .= $this->closeParagraph().'<pre>';
1398 $this->mLastSection
= 'pre';
1402 if ( '' == trim($t) ) {
1403 if ( $paragraphStack ) {
1404 $output .= $paragraphStack.'<br />';
1405 $paragraphStack = false;
1406 $this->mLastSection
= 'p';
1408 if ($this->mLastSection
!= 'p' ) {
1409 $output .= $this->closeParagraph();
1410 $this->mLastSection
= '';
1411 $paragraphStack = '<p>';
1413 $paragraphStack = '</p><p>';
1417 if ( $paragraphStack ) {
1418 $output .= $paragraphStack;
1419 $paragraphStack = false;
1420 $this->mLastSection
= 'p';
1421 } else if ($this->mLastSection
!= 'p') {
1422 $output .= $this->closeParagraph().'<p>';
1423 $this->mLastSection
= 'p';
1429 if ($paragraphStack === false) {
1433 while ( $prefixLength ) {
1434 $output .= $this->closeList( $pref2{$prefixLength-1} );
1437 if ( '' != $this->mLastSection
) {
1438 $output .= '</' . $this->mLastSection
. '>';
1439 $this->mLastSection
= '';
1442 wfProfileOut( $fname );
1446 # Return value of a magic variable (like PAGENAME)
1447 function getVariableValue( $index ) {
1448 global $wgLang, $wgSitename, $wgServer;
1451 case MAG_CURRENTMONTH
:
1452 return $wgLang->formatNum( date( 'm' ) );
1453 case MAG_CURRENTMONTHNAME
:
1454 return $wgLang->getMonthName( date('n') );
1455 case MAG_CURRENTMONTHNAMEGEN
:
1456 return $wgLang->getMonthNameGen( date('n') );
1457 case MAG_CURRENTDAY
:
1458 return $wgLang->formatNum( date('j') );
1460 return $this->mTitle
->getText();
1462 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1463 return $wgLang->getNsText($this->mTitle
->getNamespace()); // Patch by Dori
1464 case MAG_CURRENTDAYNAME
:
1465 return $wgLang->getWeekdayName( date('w')+
1 );
1466 case MAG_CURRENTYEAR
:
1467 return $wgLang->formatNum( date( 'Y' ) );
1468 case MAG_CURRENTTIME
:
1469 return $wgLang->time( wfTimestampNow(), false );
1470 case MAG_NUMBEROFARTICLES
:
1471 return $wgLang->formatNum( wfNumberOfArticles() );
1481 # initialise the magic variables (like CURRENTMONTHNAME)
1482 function initialiseVariables() {
1483 global $wgVariableIDs;
1484 $this->mVariables
= array();
1485 foreach ( $wgVariableIDs as $id ) {
1486 $mw =& MagicWord
::get( $id );
1487 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1491 /* private */ function replaceVariables( $text, $args = array() ) {
1492 global $wgLang, $wgScript, $wgArticlePath;
1494 # Prevent too big inclusions
1495 if(strlen($text)> MAX_INCLUDE_SIZE
)
1498 $fname = 'Parser::replaceVariables';
1499 wfProfileIn( $fname );
1502 $titleChars = Title
::legalChars();
1503 $nonBraceChars = str_replace( array( '{', '}' ), array( '', '' ), $titleChars );
1505 # This function is called recursively. To keep track of arguments we need a stack:
1506 array_push( $this->mArgStack
, $args );
1508 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1509 $GLOBALS['wgCurParser'] =& $this;
1512 if ( $this->mOutputType
== OT_HTML
) {
1513 # Variable substitution
1514 $text = preg_replace_callback( "/{{([$nonBraceChars]*?)}}/", 'wfVariableSubstitution', $text );
1516 # Argument substitution
1517 $text = preg_replace_callback( "/(\\n?){{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1519 # Template substitution
1520 $regex = '/(\\n?){{(['.$nonBraceChars.']*)(\\|.*?|)}}/s';
1521 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1523 array_pop( $this->mArgStack
);
1525 wfProfileOut( $fname );
1529 function variableSubstitution( $matches ) {
1530 if ( !$this->mVariables
) {
1531 $this->initialiseVariables();
1533 if ( array_key_exists( $matches[1], $this->mVariables
) ) {
1534 $text = $this->mVariables
[$matches[1]];
1535 $this->mOutput
->mContainsOldMagic
= true;
1537 $text = $matches[0];
1542 # Split template arguments
1543 function getTemplateArgs( $argsString ) {
1544 if ( $argsString === '' ) {
1548 $args = explode( '|', substr( $argsString, 1 ) );
1550 # If any of the arguments contains a '[[' but no ']]', it needs to be
1551 # merged with the next arg because the '|' character between belongs
1552 # to the link syntax and not the template parameter syntax.
1553 $argc = count($args);
1555 for ( $i = 0; $i < $argc-1; $i++
) {
1556 if ( substr_count ( $args[$i], "[[" ) != substr_count ( $args[$i], "]]" ) ) {
1557 $args[$i] .= "|".$args[$i+
1];
1558 array_splice($args, $i+
1, 1);
1567 function braceSubstitution( $matches ) {
1568 global $wgLinkCache, $wgLang;
1569 $fname = 'Parser::braceSubstitution';
1576 # $newline is an optional newline character before the braces
1577 # $part1 is the bit before the first |, and must contain only title characters
1578 # $args is a list of arguments, starting from index 0, not including $part1
1580 $newline = $matches[1];
1581 $part1 = $matches[2];
1582 # If the third subpattern matched anything, it will start with |
1584 $args = $this->getTemplateArgs($matches[3]);
1585 $argc = count( $args );
1588 if ( strpos( $matches[0], '{{{' ) !== false ) {
1589 $text = $matches[0];
1596 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1597 if ( $mwSubst->matchStartAndRemove( $part1 ) ) {
1598 if ( $this->mOutputType
!= OT_WIKI
) {
1599 # Invalid SUBST not replaced at PST time
1600 # Return without further processing
1601 $text = $matches[0];
1605 } elseif ( $this->mOutputType
== OT_WIKI
) {
1606 # SUBST not found in PST pass, do nothing
1607 $text = $matches[0];
1612 # MSG, MSGNW and INT
1615 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1616 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1619 # Remove obsolete MSG:
1620 $mwMsg =& MagicWord
::get( MAG_MSG
);
1621 $mwMsg->matchStartAndRemove( $part1 );
1624 # Check if it is an internal message
1625 $mwInt =& MagicWord
::get( MAG_INT
);
1626 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1627 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1628 $text = wfMsgReal( $part1, $args, true );
1636 # Check for NS: (namespace expansion)
1637 $mwNs = MagicWord
::get( MAG_NS
);
1638 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1639 if ( intval( $part1 ) ) {
1640 $text = $wgLang->getNsText( intval( $part1 ) );
1643 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1644 if ( !is_null( $index ) ) {
1645 $text = $wgLang->getNsText( $index );
1652 # LOCALURL and LOCALURLE
1654 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1655 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1657 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1658 $func = 'getLocalURL';
1659 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1660 $func = 'escapeLocalURL';
1665 if ( $func !== '' ) {
1666 $title = Title
::newFromText( $part1 );
1667 if ( !is_null( $title ) ) {
1669 $text = $title->$func( $args[0] );
1671 $text = $title->$func();
1678 # Internal variables
1679 if ( !$this->mVariables
) {
1680 $this->initialiseVariables();
1682 if ( !$found && array_key_exists( $part1, $this->mVariables
) ) {
1683 $text = $this->mVariables
[$part1];
1685 $this->mOutput
->mContainsOldMagic
= true;
1688 # Template table test
1690 # Did we encounter this template already? If yes, it is in the cache
1691 # and we need to check for loops.
1692 if ( isset( $this->mTemplates
[$part1] ) ) {
1693 # Infinite loop test
1694 if ( isset( $this->mTemplatePath
[$part1] ) ) {
1698 # set $text to cached message.
1699 $text = $this->mTemplates
[$part1];
1703 # Load from database
1705 $title = Title
::newFromText( $part1, NS_TEMPLATE
);
1706 if ( !is_null( $title ) && !$title->isExternal() ) {
1707 # Check for excessive inclusion
1708 $dbk = $title->getPrefixedDBkey();
1709 if ( $this->incrementIncludeCount( $dbk ) ) {
1710 # This should never be reached.
1711 $article = new Article( $title );
1712 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1713 if ( $articleContent !== false ) {
1715 $text = $articleContent;
1720 # If the title is valid but undisplayable, make a link to it
1721 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1722 $text = '[[' . $title->getPrefixedText() . ']]';
1726 # Template cache array insertion
1727 $this->mTemplates
[$part1] = $text;
1731 # Recursive parsing, escaping and link table handling
1732 # Only for HTML output
1733 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1734 $text = wfEscapeWikiText( $text );
1735 } elseif ( $this->mOutputType
== OT_HTML
&& $found && !$noparse) {
1736 # Clean up argument array
1737 $assocArgs = array();
1739 foreach( $args as $arg ) {
1740 $eqpos = strpos( $arg, '=' );
1741 if ( $eqpos === false ) {
1742 $assocArgs[$index++
] = $arg;
1744 $name = trim( substr( $arg, 0, $eqpos ) );
1745 $value = trim( substr( $arg, $eqpos+
1 ) );
1746 if ( $value === false ) {
1749 if ( $name !== false ) {
1750 $assocArgs[$name] = $value;
1755 # Do not enter included links in link table
1756 if ( !is_null( $title ) ) {
1757 $wgLinkCache->suspend();
1760 # Add a new element to the templace recursion path
1761 $this->mTemplatePath
[$part1] = 1;
1763 # Run full parser on the included text
1764 $text = $this->internalParse( $text, $newline, $assocArgs );
1765 # I replaced the line below with the line above, as it former seems to cause several bugs
1766 #$text = $this->stripParse( $text, $newline, $assocArgs );
1768 # Resume the link cache and register the inclusion as a link
1769 if ( !is_null( $title ) ) {
1770 $wgLinkCache->resume();
1771 $wgLinkCache->addLinkObj( $title );
1774 # Empties the template path
1775 $this->mTemplatePath
= array();
1784 # Triple brace replacement -- used for template arguments
1785 function argSubstitution( $matches ) {
1786 $newline = $matches[1];
1787 $arg = trim( $matches[2] );
1788 $text = $matches[0];
1789 $inputArgs = end( $this->mArgStack
);
1791 if ( array_key_exists( $arg, $inputArgs ) ) {
1792 $text = $this->stripParse( $inputArgs[$arg], $newline, array() );
1798 # Returns true if the function is allowed to include this entity
1799 function incrementIncludeCount( $dbk ) {
1800 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1801 $this->mIncludeCount
[$dbk] = 0;
1803 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1811 # Cleans up HTML, removes dangerous tags and attributes
1812 /* private */ function removeHTMLtags( $text ) {
1813 global $wgUseTidy, $wgUserHtml;
1814 $fname = 'Parser::removeHTMLtags';
1815 wfProfileIn( $fname );
1818 $htmlpairs = array( # Tags that must be closed
1819 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
1820 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
1821 'strike', 'strong', 'tt', 'var', 'div', 'center',
1822 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
1823 'ruby', 'rt' , 'rb' , 'rp', 'p'
1825 $htmlsingle = array(
1826 'br', 'hr', 'li', 'dt', 'dd'
1828 $htmlnest = array( # Tags that can be nested--??
1829 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
1830 'dl', 'font', 'big', 'small', 'sub', 'sup'
1832 $tabletags = array( # Can only appear inside table
1836 $htmlpairs = array();
1837 $htmlsingle = array();
1838 $htmlnest = array();
1839 $tabletags = array();
1842 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1843 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1845 $htmlattrs = $this->getHTMLattrs () ;
1847 # Remove HTML comments
1848 $text = preg_replace( '/(\\n *<!--.*--> *(?=\\n)|<!--.*-->)/sU', '$2', $text );
1850 $bits = explode( '<', $text );
1851 $text = array_shift( $bits );
1853 $tagstack = array(); $tablestack = array();
1854 foreach ( $bits as $x ) {
1855 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
1856 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1858 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1859 error_reporting( $prev );
1862 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1866 if ( ! in_array( $t, $htmlsingle ) &&
1867 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
1868 @array_push
( $tagstack, $ot );
1871 if ( $t == 'table' ) {
1872 $tagstack = array_pop( $tablestack );
1877 # Keep track for later
1878 if ( in_array( $t, $tabletags ) &&
1879 ! in_array( 'table', $tagstack ) ) {
1881 } else if ( in_array( $t, $tagstack ) &&
1882 ! in_array ( $t , $htmlnest ) ) {
1884 } else if ( ! in_array( $t, $htmlsingle ) ) {
1885 if ( $t == 'table' ) {
1886 array_push( $tablestack, $tagstack );
1887 $tagstack = array();
1889 array_push( $tagstack, $t );
1891 # Strip non-approved attributes from the tag
1892 $newparams = $this->fixTagAttributes($params);
1896 $rest = str_replace( '>', '>', $rest );
1897 $text .= "<$slash$t $newparams$brace$rest";
1901 $text .= '<' . str_replace( '>', '>', $x);
1903 # Close off any remaining tags
1904 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
1906 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
1909 # this might be possible using tidy itself
1910 foreach ( $bits as $x ) {
1911 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1913 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1914 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1915 $newparams = $this->fixTagAttributes($params);
1916 $rest = str_replace( '>', '>', $rest );
1917 $text .= "<$slash$t $newparams$brace$rest";
1919 $text .= '<' . str_replace( '>', '>', $x);
1923 wfProfileOut( $fname );
1930 * This function accomplishes several tasks:
1931 * 1) Auto-number headings if that option is enabled
1932 * 2) Add an [edit] link to sections for logged in users who have enabled the option
1933 * 3) Add a Table of contents on the top for users who have enabled the option
1934 * 4) Auto-anchor headings
1936 * It loops through all headlines, collects the necessary data, then splits up the
1937 * string and re-inserts the newly formatted headlines.
1941 /* private */ function formatHeadings( $text, $isMain=true ) {
1942 global $wgInputEncoding, $wgMaxTocLevel;
1944 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
1945 $doShowToc = $this->mOptions
->getShowToc();
1946 $forceTocHere = false;
1947 if( !$this->mTitle
->userCanEdit() ) {
1949 $rightClickHack = 0;
1951 $showEditLink = $this->mOptions
->getEditSection();
1952 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
1955 # Inhibit editsection links if requested in the page
1956 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
1957 if( $esw->matchAndRemove( $text ) ) {
1960 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1962 $mw =& MagicWord
::get( MAG_NOTOC
);
1963 if( $mw->matchAndRemove( $text ) ) {
1967 # never add the TOC to the Main Page. This is an entry page that should not
1968 # be more than 1-2 screens large anyway
1969 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
1973 # Get all headlines for numbering them and adding funky stuff like [edit]
1974 # links - this is for later, but we need the number of headlines right now
1975 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
1977 # if there are fewer than 4 headlines in the article, do not show TOC
1978 if( $numMatches < 4 ) {
1982 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
1983 # override above conditions and always show TOC at that place
1984 $mw =& MagicWord
::get( MAG_TOC
);
1985 if ($mw->match( $text ) ) {
1987 $forceTocHere = true;
1989 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
1990 # override above conditions and always show TOC above first header
1991 $mw =& MagicWord
::get( MAG_FORCETOC
);
1992 if ($mw->matchAndRemove( $text ) ) {
1999 # We need this to perform operations on the HTML
2000 $sk =& $this->mOptions
->getSkin();
2005 # Ugh .. the TOC should have neat indentation levels which can be
2006 # passed to the skin functions. These are determined here
2011 $sublevelCount = array();
2014 foreach( $matches[3] as $headline ) {
2017 $prevlevel = $level;
2019 $level = $matches[1][$headlineCount];
2020 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2021 # reset when we enter a new level
2022 $sublevelCount[$level] = 0;
2023 $toc .= $sk->tocIndent( $level - $prevlevel );
2024 $toclevel +
= $level - $prevlevel;
2026 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2027 # reset when we step back a level
2028 $sublevelCount[$level+
1]=0;
2029 $toc .= $sk->tocUnindent( $prevlevel - $level );
2030 $toclevel -= $prevlevel - $level;
2032 # count number of headlines for each level
2033 @$sublevelCount[$level]++
;
2034 if( $doNumberHeadings ||
$doShowToc ) {
2036 for( $i = 1; $i <= $level; $i++
) {
2037 if( !empty( $sublevelCount[$i] ) ) {
2041 $numbering .= $sublevelCount[$i];
2047 # The canonized header is a version of the header text safe to use for links
2048 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2049 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2050 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2053 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2054 $tocline = trim( $canonized_headline );
2055 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2056 $replacearray = array(
2060 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2061 $refer[$headlineCount] = $canonized_headline;
2063 # count how many in assoc. array so we can track dupes in anchors
2064 @$refers[$canonized_headline]++
;
2065 $refcount[$headlineCount]=$refers[$canonized_headline];
2067 # Prepend the number to the heading text
2069 if( $doNumberHeadings ||
$doShowToc ) {
2070 $tocline = $numbering . ' ' . $tocline;
2072 # Don't number the heading if it is the only one (looks silly)
2073 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2074 # the two are different if the line contains a link
2075 $headline=$numbering . ' ' . $headline;
2079 # Create the anchor for linking from the TOC to the section
2080 $anchor = $canonized_headline;
2081 if($refcount[$headlineCount] > 1 ) {
2082 $anchor .= '_' . $refcount[$headlineCount];
2084 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2085 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2087 if( $showEditLink ) {
2088 if ( empty( $head[$headlineCount] ) ) {
2089 $head[$headlineCount] = '';
2091 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+
1);
2094 # Add the edit section span
2095 if( $rightClickHack ) {
2096 $headline = $sk->editSectionScript($headlineCount+
1,$headline);
2099 # give headline the correct <h#> tag
2100 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline."</h".$level.">";
2106 $toclines = $headlineCount;
2107 $toc .= $sk->tocUnindent( $toclevel );
2108 $toc = $sk->tocTable( $toc );
2111 # split up and insert constructed headlines
2113 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2116 foreach( $blocks as $block ) {
2117 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2118 # This is the [edit] link that appears for the top block of text when
2119 # section editing is enabled
2121 # Disabled because it broke block formatting
2122 # For example, a bullet point in the top line
2123 # $full .= $sk->editSectionLink(0);
2126 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2127 # Top anchor now in skin
2131 if( !empty( $head[$i] ) ) {
2137 $mw =& MagicWord
::get( MAG_TOC
);
2138 return $mw->replace( $toc, $full );
2144 # Return an HTML link for the "ISBN 123456" text
2145 /* private */ function magicISBN( $text ) {
2147 $fname = 'Parser::magicISBN';
2148 wfProfileIn( $fname );
2150 $a = split( 'ISBN ', " $text" );
2151 if ( count ( $a ) < 2 ) {
2152 wfProfileOut( $fname );
2155 $text = substr( array_shift( $a ), 1);
2156 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2158 foreach ( $a as $x ) {
2159 $isbn = $blank = '' ;
2160 while ( ' ' == $x{0} ) {
2162 $x = substr( $x, 1 );
2164 while ( strstr( $valid, $x{0} ) != false ) {
2166 $x = substr( $x, 1 );
2168 $num = str_replace( '-', '', $isbn );
2169 $num = str_replace( ' ', '', $num );
2172 $text .= "ISBN $blank$x";
2174 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2175 $text .= '<a href="' .
2176 $titleObj->escapeLocalUrl( "isbn={$num}" ) .
2177 "\" class=\"internal\">ISBN $isbn</a>";
2181 wfProfileOut( $fname );
2185 # Return an HTML link for the "GEO ..." text
2186 /* private */ function magicGEO( $text ) {
2187 global $wgLang, $wgUseGeoMode;
2188 if ( !isset ( $wgUseGeoMode ) ||
!$wgUseGeoMode ) return $text ;
2189 $fname = 'Parser::magicGEO';
2190 wfProfileIn( $fname );
2192 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2193 $directions = array ( "N" => "North" , "S" => "South" , "E" => "East" , "W" => "West" ) ;
2194 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2195 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2196 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2197 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2199 $a = split( 'GEO ', " $text" );
2200 if ( count ( $a ) < 2 ) {
2201 wfProfileOut( $fname );
2204 $text = substr( array_shift( $a ), 1);
2205 $valid = '0123456789.+-:';
2207 foreach ( $a as $x ) {
2208 $geo = $blank = '' ;
2209 while ( ' ' == $x{0} ) {
2211 $x = substr( $x, 1 );
2213 while ( strstr( $valid, $x{0} ) != false ) {
2215 $x = substr( $x, 1 );
2217 $num = str_replace( '+', '', $geo );
2218 $num = str_replace( ' ', '', $num );
2220 if ( '' == $num ||
count ( explode ( ":" , $num , 3 ) ) < 2 ) {
2221 $text .= "GEO $blank$x";
2223 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2224 $text .= '<a href="' .
2225 $titleObj->escapeLocalUrl( "coordinates={$num}" ) .
2226 "\" class=\"internal\">GEO $geo</a>";
2230 wfProfileOut( $fname );
2234 # Return an HTML link for the "RFC 1234" text
2235 /* private */ function magicRFC( $text ) {
2238 $a = split( 'RFC ', ' '.$text );
2239 if ( count ( $a ) < 2 ) return $text;
2240 $text = substr( array_shift( $a ), 1);
2241 $valid = '0123456789';
2243 foreach ( $a as $x ) {
2244 $rfc = $blank = '' ;
2245 while ( ' ' == $x{0} ) {
2247 $x = substr( $x, 1 );
2249 while ( strstr( $valid, $x{0} ) != false ) {
2251 $x = substr( $x, 1 );
2255 $text .= "RFC $blank$x";
2257 $url = wfmsg( 'rfcurl' );
2258 $url = str_replace( '$1', $rfc, $url);
2259 $sk =& $this->mOptions
->getSkin();
2260 $la = $sk->getExternalLinkAttributes( $url, "RFC {$rfc}" );
2261 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2267 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2268 $this->mOptions
= $options;
2269 $this->mTitle
=& $title;
2270 $this->mOutputType
= OT_WIKI
;
2272 if ( $clearState ) {
2273 $this->clearState();
2276 $stripState = false;
2280 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2284 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2285 "/<br *?>/i" => "<br />",
2287 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2289 $text = $this->strip( $text, $stripState, false );
2290 $text = $this->pstPass2( $text, $user );
2291 $text = $this->unstrip( $text, $stripState );
2292 $text = $this->unstripNoWiki( $text, $stripState );
2296 /* private */ function pstPass2( $text, &$user ) {
2297 global $wgLang, $wgLocaltimezone, $wgCurParser;
2299 # Variable replacement
2300 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2301 $text = $this->replaceVariables( $text );
2305 $n = $user->getName();
2306 $k = $user->getOption( 'nickname' );
2307 if ( '' == $k ) { $k = $n; }
2308 if(isset($wgLocaltimezone)) {
2309 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2311 /* Note: this is an ugly timezone hack for the European wikis */
2312 $d = $wgLang->timeanddate( date( 'YmdHis' ), false ) .
2313 ' (' . date( 'T' ) . ')';
2314 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2316 $text = preg_replace( '/~~~~~/', $d, $text );
2317 $text = preg_replace( '/~~~~/', '[[' . $wgLang->getNsText(
2318 Namespace::getUser() ) . ":$n|$k]] $d", $text );
2319 $text = preg_replace( '/~~~/', '[[' . $wgLang->getNsText(
2320 Namespace::getUser() ) . ":$n|$k]]", $text );
2322 # Context links: [[|name]] and [[name (context)|]]
2324 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2325 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2326 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2327 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2329 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2330 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2331 $p3 = "/\[\[($namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]]
2332 $p4 = "/\[\[($namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/";
2333 # [[ns:page (cont)|]]
2335 $t = $this->mTitle
->getText();
2336 if ( preg_match( $conpat, $t, $m ) ) {
2339 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2340 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2341 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2343 if ( '' == $context ) {
2344 $text = preg_replace( $p2, '[[\\1]]', $text );
2346 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2350 $mw =& MagicWord::get( MAG_SUBST );
2351 $wgCurParser = $this->fork();
2352 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
2353 $this->merge( $wgCurParser );
2356 # Trim trailing whitespace
2357 # MAG_END (__END__) tag allows for trailing
2358 # whitespace to be deliberately included
2359 $text = rtrim( $text );
2360 $mw =& MagicWord
::get( MAG_END
);
2361 $mw->matchAndRemove( $text );
2366 # Set up some variables which are usually set up in parse()
2367 # so that an external function can call some class members with confidence
2368 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2369 $this->mTitle
=& $title;
2370 $this->mOptions
= $options;
2371 $this->mOutputType
= $outputType;
2372 if ( $clearState ) {
2373 $this->clearState();
2377 function transformMsg( $text, $options ) {
2379 static $executing = false;
2381 # Guard against infinite recursion
2387 $this->mTitle
= $wgTitle;
2388 $this->mOptions
= $options;
2389 $this->mOutputType
= OT_MSG
;
2390 $this->clearState();
2391 $text = $this->replaceVariables( $text );
2397 # Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2398 # Callback will be called with the text within
2399 # Transform and return the text within
2400 function setHook( $tag, $callback ) {
2401 $oldVal = @$this->mTagHooks
[$tag];
2402 $this->mTagHooks
[$tag] = $callback;
2409 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2410 var $mCacheTime; # Used in ParserCache
2412 function ParserOutput( $text = "", $languageLinks = array(), $categoryLinks = array(),
2413 $containsOldMagic = false )
2415 $this->mText
= $text;
2416 $this->mLanguageLinks
= $languageLinks;
2417 $this->mCategoryLinks
= $categoryLinks;
2418 $this->mContainsOldMagic
= $containsOldMagic;
2419 $this->mCacheTime
= "";
2422 function getText() { return $this->mText
; }
2423 function getLanguageLinks() { return $this->mLanguageLinks
; }
2424 function getCategoryLinks() { return $this->mCategoryLinks
; }
2425 function getCacheTime() { return $this->mCacheTime
; }
2426 function containsOldMagic() { return $this->mContainsOldMagic
; }
2427 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2428 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2429 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2430 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2431 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
2433 function merge( $other ) {
2434 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2435 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2436 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2443 # All variables are private
2444 var $mUseTeX; # Use texvc to expand <math> tags
2445 var $mUseCategoryMagic; # Treat [[Category:xxxx]] tags specially
2446 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2447 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2448 var $mAllowExternalImages; # Allow external images inline
2449 var $mSkin; # Reference to the preferred skin
2450 var $mDateFormat; # Date format index
2451 var $mEditSection; # Create "edit section" links
2452 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2453 var $mNumberHeadings; # Automatically number headings
2454 var $mShowToc; # Show table of contents
2456 function getUseTeX() { return $this->mUseTeX
; }
2457 function getUseCategoryMagic() { return $this->mUseCategoryMagic
; }
2458 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2459 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2460 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2461 function getSkin() { return $this->mSkin
; }
2462 function getDateFormat() { return $this->mDateFormat
; }
2463 function getEditSection() { return $this->mEditSection
; }
2464 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
2465 function getNumberHeadings() { return $this->mNumberHeadings
; }
2466 function getShowToc() { return $this->mShowToc
; }
2468 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2469 function setUseCategoryMagic( $x ) { return wfSetVar( $this->mUseCategoryMagic
, $x ); }
2470 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2471 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2472 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2473 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2474 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
2475 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
2476 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
2477 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
2479 function setSkin( &$x ) { $this->mSkin
=& $x; }
2481 /* static */ function newFromUser( &$user ) {
2482 $popts = new ParserOptions
;
2483 $popts->initialiseFromUser( $user );
2487 function initialiseFromUser( &$userInput ) {
2488 global $wgUseTeX, $wgUseCategoryMagic, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2490 if ( !$userInput ) {
2492 $user->setLoaded( true );
2494 $user =& $userInput;
2497 $this->mUseTeX
= $wgUseTeX;
2498 $this->mUseCategoryMagic
= $wgUseCategoryMagic;
2499 $this->mUseDynamicDates
= $wgUseDynamicDates;
2500 $this->mInterwikiMagic
= $wgInterwikiMagic;
2501 $this->mAllowExternalImages
= $wgAllowExternalImages;
2502 $this->mSkin
=& $user->getSkin();
2503 $this->mDateFormat
= $user->getOption( 'date' );
2504 $this->mEditSection
= $user->getOption( 'editsection' );
2505 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
2506 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
2507 $this->mShowToc
= $user->getOption( 'showtoc' );
2513 # Regex callbacks, used in Parser::replaceVariables
2514 function wfBraceSubstitution( $matches )
2516 global $wgCurParser;
2517 return $wgCurParser->braceSubstitution( $matches );
2520 function wfArgSubstitution( $matches )
2522 global $wgCurParser;
2523 return $wgCurParser->argSubstitution( $matches );
2526 function wfVariableSubstitution( $matches )
2528 global $wgCurParser;
2529 return $wgCurParser->variableSubstitution( $matches );