4 * File for Parser and related classes
10 * Update this version number when the ParserOutput format
11 * changes in an incompatible way, so the parser cache
12 * can automatically discard old data.
14 define( 'MW_PARSER_VERSION', '1.4.0' );
17 * Variable substitution O(N^2) attack
19 * Without countermeasures, it would be possible to attack the parser by saving
20 * a page filled with a large number of inclusions of large pages. The size of
21 * the generated page would be proportional to the square of the input size.
22 * Hence, we limit the number of inclusions of any given page, thus bringing any
23 * attack back to O(N).
26 define( 'MAX_INCLUDE_REPEAT', 100 );
27 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
29 define( 'RLH_FOR_UPDATE', 1 );
31 # Allowed values for $mOutputType
32 define( 'OT_HTML', 1 );
33 define( 'OT_WIKI', 2 );
34 define( 'OT_MSG' , 3 );
36 # string parameter for extractTags which will cause it
37 # to strip HTML comments in addition to regular
38 # <XML>-style tags. This should not be anything we
39 # may want to use in wikisyntax
40 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
42 # prefix for escaping, used in two functions at least
43 define( 'UNIQ_PREFIX', 'NaodW29');
45 # Constants needed for external link processing
46 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
47 define( 'HTTP_PROTOCOLS', 'http|https' );
48 # Everything except bracket, space, or control characters
49 define( 'EXT_LINK_URL_CLASS', '[^]<>"\\x00-\\x20\\x7F]' );
51 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
52 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
53 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
54 define( 'EXT_LINK_BRACKETED', '/\[(('.URL_PROTOCOLS
.'):'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
55 define( 'EXT_IMAGE_REGEX',
56 '/^('.HTTP_PROTOCOLS
.':)'. # Protocol
57 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
58 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
64 * Processes wiki markup
67 * There are three main entry points into the Parser class:
69 * produces HTML output
71 * produces altered wiki markup.
73 * performs brace substitution on MediaWiki messages
76 * objects: $wgLang, $wgDateFormatter, $wgLinkCache
78 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
81 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
82 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
85 * * only within ParserOptions
98 # Cleared with clearState():
99 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
100 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
103 var $mOptions, $mTitle, $mOutputType,
104 $mTemplates, // cache of already loaded templates, avoids
105 // multiple SQL queries for the same string
106 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
107 // in this path. Used for loop detection.
117 $this->mTemplates
= array();
118 $this->mTemplatePath
= array();
119 $this->mTagHooks
= array();
128 function clearState() {
129 $this->mOutput
= new ParserOutput
;
130 $this->mAutonumber
= 0;
131 $this->mLastSection
= "";
132 $this->mDTopen
= false;
133 $this->mVariables
= false;
134 $this->mIncludeCount
= array();
135 $this->mStripState
= array();
136 $this->mArgStack
= array();
137 $this->mInPre
= false;
141 * First pass--just handle <nowiki> sections, pass the rest off
142 * to internalParse() which does all the real work.
145 * @return ParserOutput a ParserOutput
147 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
148 global $wgUseTidy, $wgContLang;
149 $fname = 'Parser::parse';
150 wfProfileIn( $fname );
156 $this->mOptions
= $options;
157 $this->mTitle
=& $title;
158 $this->mOutputType
= OT_HTML
;
161 global $fnord; $fnord = 1;
162 //$text = $this->strip( $text, $this->mStripState );
163 // VOODOO MAGIC FIX! Sometimes the above segfaults in PHP5.
164 $x =& $this->mStripState
;
165 $text = $this->strip( $text, $x );
167 $text = $this->internalParse( $text, $linestart );
168 $text = $this->unstrip( $text, $this->mStripState
);
169 # Clean up special characters, only run once, next-to-last before doBlockLevels
172 # french spaces, last one Guillemet-left
173 # only if there is something before the space
174 '/(.) (?=\\?|:|;|!|\\302\\273)/i' => '\\1 \\2',
175 # french spaces, Guillemet-right
176 "/(\\302\\253) /i"=>"\\1 ",
177 '/<hr *>/i' => '<hr />',
178 '/<br *>/i' => '<br />',
179 '/<center *>/i' => '<div class="center">',
180 '/<\\/center *>/i' => '</div>',
181 # Clean up spare ampersands; note that we probably ought to be
182 # more careful about named entities.
183 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
185 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
188 # french spaces, last one Guillemet-left
189 '/ (\\?|:|;|!|\\302\\273)/i' => ' \\1',
190 # french spaces, Guillemet-right
191 '/(\\302\\253) /i' => '\\1 ',
192 '/<center *>/i' => '<div class="center">',
193 '/<\\/center *>/i' => '</div>'
195 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
198 $text = $this->doBlockLevels( $text, $linestart );
200 $this->replaceLinkHolders( $text );
201 $text = $wgContLang->convert($text);
203 $text = $this->unstripNoWiki( $text, $this->mStripState
);
206 $text = Parser
::tidy($text);
209 $this->mOutput
->setText( $text );
210 wfProfileOut( $fname );
211 return $this->mOutput
;
215 * Get a random string
220 function getRandomString() {
221 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
225 * Replaces all occurrences of <$tag>content</$tag> in the text
226 * with a random marker and returns the new text. the output parameter
227 * $content will be an associative array filled with data on the form
228 * $unique_marker => content.
230 * If $content is already set, the additional entries will be appended
231 * If $tag is set to STRIP_COMMENTS, the function will extract
232 * <!-- HTML comments -->
237 function extractTags($tag, $text, &$content, $uniq_prefix = ''){
238 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
245 while ( '' != $text ) {
246 if($tag==STRIP_COMMENTS
) {
247 $p = preg_split( '/<!--/i', $text, 2 );
249 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
252 if ( ( count( $p ) < 2 ) ||
( '' == $p[1] ) ) {
255 if($tag==STRIP_COMMENTS
) {
256 $q = preg_split( '/-->/i', $p[1], 2 );
258 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
260 $marker = $rnd . sprintf('%08X', $n++
);
261 $content[$marker] = $q[0];
262 $stripped .= $marker;
270 * Strips and renders nowiki, pre, math, hiero
271 * If $render is set, performs necessary rendering operations on plugins
272 * Returns the text, and fills an array with data needed in unstrip()
273 * If the $state is already a valid strip state, it adds to the state
275 * @param bool $stripcomments when set, HTML comments <!-- like this -->
276 * will be stripped in addition to other tags. This is important
277 * for section editing, where these comments cause confusion when
278 * counting the sections in the wikisource
282 function strip( $text, &$state, $stripcomments = false ) {
283 $render = ($this->mOutputType
== OT_HTML
);
284 $html_content = array();
285 $nowiki_content = array();
286 $math_content = array();
287 $pre_content = array();
288 $comment_content = array();
289 $ext_content = array();
290 $gallery_content = array();
292 # Replace any instances of the placeholders
293 $uniq_prefix = UNIQ_PREFIX
;
294 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
297 global $wgRawHtml, $wgWhitelistEdit;
298 if( $wgRawHtml && $wgWhitelistEdit ) {
299 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
300 foreach( $html_content as $marker => $content ) {
302 # Raw and unchecked for validity.
303 $html_content[$marker] = $content;
305 $html_content[$marker] = '<html>'.$content.'</html>';
311 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
312 foreach( $nowiki_content as $marker => $content ) {
314 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
316 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
321 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
322 foreach( $math_content as $marker => $content ){
324 if( $this->mOptions
->getUseTeX() ) {
325 $math_content[$marker] = renderMath( $content );
327 $math_content[$marker] = '<math>'.$content.'<math>';
330 $math_content[$marker] = '<math>'.$content.'</math>';
335 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
336 foreach( $pre_content as $marker => $content ){
338 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
340 $pre_content[$marker] = '<pre>'.$content.'</pre>';
345 $text = Parser
::extractTags('gallery', $text, $gallery_content, $uniq_prefix);
346 foreach( $gallery_content as $marker => $content ) {
347 require_once( 'ImageGallery.php' );
349 $gallery_content[$marker] = Parser
::renderImageGallery( $content );
351 $gallery_content[$marker] = '<gallery>'.$content.'</gallery>';
357 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
358 foreach( $comment_content as $marker => $content ){
359 $comment_content[$marker] = '<!--'.$content.'-->';
364 foreach ( $this->mTagHooks
as $tag => $callback ) {
365 $ext_contents[$tag] = array();
366 $text = Parser
::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
367 foreach( $ext_content[$tag] as $marker => $content ) {
369 $ext_content[$tag][$marker] = $callback( $content );
371 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
376 # Merge state with the pre-existing state, if there is one
378 $state['html'] = $state['html'] +
$html_content;
379 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
380 $state['math'] = $state['math'] +
$math_content;
381 $state['pre'] = $state['pre'] +
$pre_content;
382 $state['comment'] = $state['comment'] +
$comment_content;
383 $state['gallery'] = $state['gallery'] +
$gallery_content;
385 foreach( $ext_content as $tag => $array ) {
386 if ( array_key_exists( $tag, $state ) ) {
387 $state[$tag] = $state[$tag] +
$array;
392 'html' => $html_content,
393 'nowiki' => $nowiki_content,
394 'math' => $math_content,
395 'pre' => $pre_content,
396 'comment' => $comment_content,
397 'gallery' => $gallery_content,
404 * restores pre, math, and hiero removed by strip()
406 * always call unstripNoWiki() after this one
409 function unstrip( $text, &$state ) {
410 # Must expand in reverse order, otherwise nested tags will be corrupted
411 $contentDict = end( $state );
412 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
413 if( key($state) != 'nowiki' && key($state) != 'html') {
414 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
415 $text = str_replace( key( $contentDict ), $content, $text );
424 * always call this after unstrip() to preserve the order
428 function unstripNoWiki( $text, &$state ) {
429 # Must expand in reverse order, otherwise nested tags will be corrupted
430 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
431 $text = str_replace( key( $state['nowiki'] ), $content, $text );
436 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
437 $text = str_replace( key( $state['html'] ), $content, $text );
445 * Add an item to the strip state
446 * Returns the unique tag which must be inserted into the stripped text
447 * The tag will be replaced with the original text in unstrip()
451 function insertStripItem( $text, &$state ) {
452 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
461 $state['item'][$rnd] = $text;
466 * Return allowed HTML attributes
470 function getHTMLattrs () {
471 $htmlattrs = array( # Allowed attributes--no scripting, etc.
472 'title', 'align', 'lang', 'dir', 'width', 'height',
473 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
474 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
475 /* FONT */ 'type', 'start', 'value', 'compact',
476 /* For various lists, mostly deprecated but safe */
477 'summary', 'width', 'border', 'frame', 'rules',
478 'cellspacing', 'cellpadding', 'valign', 'char',
479 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
480 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
481 'id', 'class', 'name', 'style' /* For CSS */
487 * Remove non approved attributes and javascript in css
491 function fixTagAttributes ( $t ) {
492 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
493 $htmlattrs = $this->getHTMLattrs() ;
495 # Strip non-approved attributes from the tag
497 '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
498 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
501 $t = str_replace ( '<></>' , '' , $t ) ; # This should fix bug 980557
503 # Strip javascript "expression" from stylesheets. Brute force approach:
504 # If anythin offensive is found, all attributes of the HTML tag are dropped
507 '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
508 wfMungeToUtf8( $t ) ) )
517 * interface with html tidy, used if $wgUseTidy = true
522 function tidy ( $text ) {
523 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
524 global $wgInputEncoding, $wgOutputEncoding;
525 $fname = 'Parser::tidy';
526 wfProfileIn( $fname );
530 switch(strtoupper($wgOutputEncoding)) {
532 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -latin1':' -raw';
535 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -utf8':' -raw';
541 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
542 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
543 '<head><title>test</title></head><body>'.$text.'</body></html>';
544 $descriptorspec = array(
545 0 => array('pipe', 'r'),
546 1 => array('pipe', 'w'),
547 2 => array('file', '/dev/null', 'a')
549 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
550 if (is_resource($process)) {
551 fwrite($pipes[0], $wrappedtext);
553 while (!feof($pipes[1])) {
554 $cleansource .= fgets($pipes[1], 1024);
557 $return_value = proc_close($process);
560 wfProfileOut( $fname );
562 if( $cleansource == '' && $text != '') {
563 wfDebug( "Tidy error detected!\n" );
564 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
571 * parse the wiki syntax used to render tables
575 function doTableStuff ( $t ) {
576 $fname = 'Parser::doTableStuff';
577 wfProfileIn( $fname );
579 $t = explode ( "\n" , $t ) ;
580 $td = array () ; # Is currently a td tag open?
581 $ltd = array () ; # Was it TD or TH?
582 $tr = array () ; # Is currently a tr tag open?
583 $ltr = array () ; # tr attributes
584 $indent_level = 0; # indent level of the table
585 foreach ( $t AS $k => $x )
588 $fc = substr ( $x , 0 , 1 ) ;
589 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
590 $indent_level = strlen( $matches[1] );
592 str_repeat( '<dl><dd>', $indent_level ) .
593 '<table ' . $this->fixTagAttributes ( $matches[2] ) . '>' ;
594 array_push ( $td , false ) ;
595 array_push ( $ltd , '' ) ;
596 array_push ( $tr , false ) ;
597 array_push ( $ltr , '' ) ;
599 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
600 else if ( '|}' == substr ( $x , 0 , 2 ) ) {
602 $l = array_pop ( $ltd ) ;
603 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
604 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
606 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
608 else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
609 $x = substr ( $x , 1 ) ;
610 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
612 $l = array_pop ( $ltd ) ;
613 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
614 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
617 array_push ( $tr , false ) ;
618 array_push ( $td , false ) ;
619 array_push ( $ltd , '' ) ;
620 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
622 else if ( '|' == $fc ||
'!' == $fc ||
'|+' == substr ( $x , 0 , 2 ) ) { # Caption
624 if ( '|+' == substr ( $x , 0 , 2 ) ) {
626 $x = substr ( $x , 1 ) ;
628 $after = substr ( $x , 1 ) ;
629 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
630 $after = explode ( '||' , $after ) ;
633 # Loop through each table cell
634 foreach ( $after AS $theline )
639 $tra = array_pop ( $ltr ) ;
640 if ( !array_pop ( $tr ) ) $z = '<tr '.$tra.">\n" ;
641 array_push ( $tr , true ) ;
642 array_push ( $ltr , '' ) ;
645 $l = array_pop ( $ltd ) ;
646 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
647 if ( $fc == '|' ) $l = 'td' ;
648 else if ( $fc == '!' ) $l = 'th' ;
649 else if ( $fc == '+' ) $l = 'caption' ;
651 array_push ( $ltd , $l ) ;
654 $y = explode ( '|' , $theline , 2 ) ;
655 # Note that a '|' inside an invalid link should not
656 # be mistaken as delimiting cell parameters
657 if ( strpos( $y[0], '[[' ) !== false ) {
658 $y = array ($theline);
660 if ( count ( $y ) == 1 )
661 $y = "{$z}<{$l}>{$y[0]}" ;
662 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
664 array_push ( $td , true ) ;
669 # Closing open td, tr && table
670 while ( count ( $td ) > 0 )
672 if ( array_pop ( $td ) ) $t[] = '</td>' ;
673 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
677 $t = implode ( "\n" , $t ) ;
678 # $t = $this->removeHTMLtags( $t );
679 wfProfileOut( $fname );
684 * Helper function for parse() that transforms wiki markup into
685 * HTML. Only called for $mOutputType == OT_HTML.
689 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
692 $fname = 'Parser::internalParse';
693 wfProfileIn( $fname );
695 $text = $this->removeHTMLtags( $text );
696 $text = $this->replaceVariables( $text, $args );
698 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
700 $text = $this->doHeadings( $text );
701 if($this->mOptions
->getUseDynamicDates()) {
702 global $wgDateFormatter;
703 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
705 $text = $this->doAllQuotes( $text );
706 $text = $this->replaceInternalLinks( $text );
707 $text = $this->replaceExternalLinks( $text );
709 # replaceInternalLinks may sometimes leave behind
710 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
711 $text = str_replace("http-noparse://","http://",$text);
713 $text = $this->doMagicLinks( $text );
714 $text = $this->doTableStuff( $text );
715 $text = $this->formatHeadings( $text, $isMain );
717 wfProfileOut( $fname );
722 * Replace special strings like "ISBN xxx" and "RFC xxx" with
723 * magic external links.
727 function &doMagicLinks( &$text ) {
728 global $wgUseGeoMode;
729 $text = $this->magicISBN( $text );
730 if ( isset( $wgUseGeoMode ) && $wgUseGeoMode ) {
731 $text = $this->magicGEO( $text );
733 $text = $this->magicRFC( $text, 'RFC ', 'rfcurl' );
734 $text = $this->magicRFC( $text, 'PMID ', 'pubmedurl' );
739 * Parse ^^ tokens and return html
743 function doExponent( $text ) {
744 $fname = 'Parser::doExponent';
745 wfProfileIn( $fname );
746 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
747 wfProfileOut( $fname );
752 * Parse headers and return html
756 function doHeadings( $text ) {
757 $fname = 'Parser::doHeadings';
758 wfProfileIn( $fname );
759 for ( $i = 6; $i >= 1; --$i ) {
760 $h = substr( '======', 0, $i );
761 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
762 "<h{$i}>\\1</h{$i}>\\2", $text );
764 wfProfileOut( $fname );
769 * Replace single quotes with HTML markup
771 * @return string the altered text
773 function doAllQuotes( $text ) {
774 $fname = 'Parser::doAllQuotes';
775 wfProfileIn( $fname );
777 $lines = explode( "\n", $text );
778 foreach ( $lines as $line ) {
779 $outtext .= $this->doQuotes ( $line ) . "\n";
781 $outtext = substr($outtext, 0,-1);
782 wfProfileOut( $fname );
787 * Helper function for doAllQuotes()
790 function doQuotes( $text ) {
791 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
792 if ( count( $arr ) == 1 )
796 # First, do some preliminary work. This may shift some apostrophes from
797 # being mark-up to being text. It also counts the number of occurrences
798 # of bold and italics mark-ups.
802 foreach ( $arr as $r )
804 if ( ( $i %
2 ) == 1 )
806 # If there are ever four apostrophes, assume the first is supposed to
807 # be text, and the remaining three constitute mark-up for bold text.
808 if ( strlen( $arr[$i] ) == 4 )
813 # If there are more than 5 apostrophes in a row, assume they're all
814 # text except for the last 5.
815 else if ( strlen( $arr[$i] ) > 5 )
817 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
820 # Count the number of occurrences of bold and italics mark-ups.
821 # We are not counting sequences of five apostrophes.
822 if ( strlen( $arr[$i] ) == 2 ) $numitalics++
; else
823 if ( strlen( $arr[$i] ) == 3 ) $numbold++
; else
824 if ( strlen( $arr[$i] ) == 5 ) { $numitalics++
; $numbold++
; }
829 # If there is an odd number of both bold and italics, it is likely
830 # that one of the bold ones was meant to be an apostrophe followed
831 # by italics. Which one we cannot know for certain, but it is more
832 # likely to be one that has a single-letter word before it.
833 if ( ( $numbold %
2 == 1 ) && ( $numitalics %
2 == 1 ) )
836 $firstsingleletterword = -1;
837 $firstmultiletterword = -1;
839 foreach ( $arr as $r )
841 if ( ( $i %
2 == 1 ) and ( strlen( $r ) == 3 ) )
843 $x1 = substr ($arr[$i-1], -1);
844 $x2 = substr ($arr[$i-1], -2, 1);
846 if ($firstspace == -1) $firstspace = $i;
847 } else if ($x2 == ' ') {
848 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
850 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
856 # If there is a single-letter word, use it!
857 if ($firstsingleletterword > -1)
859 $arr [ $firstsingleletterword ] = "''";
860 $arr [ $firstsingleletterword-1 ] .= "'";
862 # If not, but there's a multi-letter word, use that one.
863 else if ($firstmultiletterword > -1)
865 $arr [ $firstmultiletterword ] = "''";
866 $arr [ $firstmultiletterword-1 ] .= "'";
868 # ... otherwise use the first one that has neither.
869 # (notice that it is possible for all three to be -1 if, for example,
870 # there is only one pentuple-apostrophe in the line)
871 else if ($firstspace > -1)
873 $arr [ $firstspace ] = "''";
874 $arr [ $firstspace-1 ] .= "'";
878 # Now let's actually convert our apostrophic mush to HTML!
887 if ($state == 'both')
894 if (strlen ($r) == 2)
897 { $output .= '</i>'; $state = ''; }
898 else if ($state == 'bi')
899 { $output .= '</i>'; $state = 'b'; }
900 else if ($state == 'ib')
901 { $output .= '</b></i><b>'; $state = 'b'; }
902 else if ($state == 'both')
903 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
904 else # $state can be 'b' or ''
905 { $output .= '<i>'; $state .= 'i'; }
907 else if (strlen ($r) == 3)
910 { $output .= '</b>'; $state = ''; }
911 else if ($state == 'bi')
912 { $output .= '</i></b><i>'; $state = 'i'; }
913 else if ($state == 'ib')
914 { $output .= '</b>'; $state = 'i'; }
915 else if ($state == 'both')
916 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
917 else # $state can be 'i' or ''
918 { $output .= '<b>'; $state .= 'b'; }
920 else if (strlen ($r) == 5)
923 { $output .= '</b><i>'; $state = 'i'; }
924 else if ($state == 'i')
925 { $output .= '</i><b>'; $state = 'b'; }
926 else if ($state == 'bi')
927 { $output .= '</i></b>'; $state = ''; }
928 else if ($state == 'ib')
929 { $output .= '</b></i>'; $state = ''; }
930 else if ($state == 'both')
931 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
932 else # ($state == '')
933 { $buffer = ''; $state = 'both'; }
938 # Now close all remaining tags. Notice that the order is important.
939 if ($state == 'b' ||
$state == 'ib')
941 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
945 if ($state == 'both')
946 $output .= '<b><i>'.$buffer.'</i></b>';
952 * Replace external links
954 * Note: this is all very hackish and the order of execution matters a lot.
955 * Make sure to run maintenance/parserTests.php if you change this code.
959 function replaceExternalLinks( $text ) {
960 $fname = 'Parser::replaceExternalLinks';
961 wfProfileIn( $fname );
963 $sk =& $this->mOptions
->getSkin();
965 $linktrail = $wgContLang->linkTrail();
967 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
969 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
972 while ( $i<count( $bits ) ) {
974 $protocol = $bits[$i++
];
976 $trail = $bits[$i++
];
978 # The characters '<' and '>' (which were escaped by
979 # removeHTMLtags()) should not be included in
980 # URLs, per RFC 2396.
981 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
982 $text = substr($url, $m2[0][1]) . ' ' . $text;
983 $url = substr($url, 0, $m2[0][1]);
986 # If the link text is an image URL, replace it with an <img> tag
987 # This happened by accident in the original parser, but some people used it extensively
988 $img = $this->maybeMakeImageLink( $text );
989 if ( $img !== false ) {
995 # No link text, e.g. [http://domain.tld/some.link]
997 # Autonumber if allowed
998 if ( strpos( HTTP_PROTOCOLS
, $protocol ) !== false ) {
999 $text = '[' . ++
$this->mAutonumber
. ']';
1001 # Otherwise just use the URL
1002 $text = htmlspecialchars( $url );
1005 # Have link text, e.g. [http://domain.tld/some.link text]s
1007 if ( preg_match( $linktrail, $trail, $m2 ) ) {
1013 $encUrl = htmlspecialchars( $url );
1014 # Bit in parentheses showing the URL for the printable version
1015 if( $url == $text ||
preg_match( "!$protocol://" . preg_quote( $text, '/' ) . "/?$!", $url ) ) {
1018 # Expand the URL for printable version
1019 if ( ! $sk->suppressUrlExpansion() ) {
1020 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $encUrl ) . "</i>)</span>";
1026 # Process the trail (i.e. everything after this link up until start of the next link),
1027 # replacing any non-bracketed links
1028 $trail = $this->replaceFreeExternalLinks( $trail );
1030 # Use the encoded URL
1031 # This means that users can paste URLs directly into the text
1032 # Funny characters like ö aren't valid in URLs anyway
1033 # This was changed in August 2004
1034 $s .= $sk->makeExternalLink( $url, $text, false ) . $dtrail. $paren . $trail;
1037 wfProfileOut( $fname );
1042 * Replace anything that looks like a URL with a link
1045 function replaceFreeExternalLinks( $text ) {
1046 $fname = 'Parser::replaceFreeExternalLinks';
1047 wfProfileIn( $fname );
1049 $bits = preg_split( '/((?:'.URL_PROTOCOLS
.'):)/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1050 $s = array_shift( $bits );
1053 $sk =& $this->mOptions
->getSkin();
1055 while ( $i < count( $bits ) ){
1056 $protocol = $bits[$i++
];
1057 $remainder = $bits[$i++
];
1059 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1060 # Found some characters after the protocol that look promising
1061 $url = $protocol . $m[1];
1064 # The characters '<' and '>' (which were escaped by
1065 # removeHTMLtags()) should not be included in
1066 # URLs, per RFC 2396.
1067 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1068 $trail = substr($url, $m2[0][1]) . $trail;
1069 $url = substr($url, 0, $m2[0][1]);
1072 # Move trailing punctuation to $trail
1074 # If there is no left bracket, then consider right brackets fair game too
1075 if ( strpos( $url, '(' ) === false ) {
1079 $numSepChars = strspn( strrev( $url ), $sep );
1080 if ( $numSepChars ) {
1081 $trail = substr( $url, -$numSepChars ) . $trail;
1082 $url = substr( $url, 0, -$numSepChars );
1085 # Replace & from obsolete syntax with &.
1086 # All HTML entities will be escaped by makeExternalLink()
1087 # or maybeMakeImageLink()
1088 $url = str_replace( '&', '&', $url );
1090 # Is this an external image?
1091 $text = $this->maybeMakeImageLink( $url );
1092 if ( $text === false ) {
1093 # Not an image, make a link
1094 $text = $sk->makeExternalLink( $url, $url );
1096 $s .= $text . $trail;
1098 $s .= $protocol . $remainder;
1106 * make an image if it's allowed
1109 function maybeMakeImageLink( $url ) {
1110 $sk =& $this->mOptions
->getSkin();
1112 if ( $this->mOptions
->getAllowExternalImages() ) {
1113 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1115 $text = $sk->makeImage( htmlspecialchars( $url ) );
1122 * Process [[ ]] wikilinks
1127 function replaceInternalLinks( $s ) {
1128 global $wgLang, $wgContLang, $wgLinkCache;
1129 global $wgDisableLangConversion;
1130 static $fname = 'Parser::replaceInternalLinks' ;
1132 wfProfileIn( $fname );
1134 wfProfileIn( $fname.'-setup' );
1136 # the % is needed to support urlencoded titles as well
1137 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1139 $sk =& $this->mOptions
->getSkin();
1140 global $wgUseOldExistenceCheck;
1141 # "Post-parse link colour check" works only on wiki text since it's now
1142 # in Parser. Enable it, then disable it when we're done.
1143 $saveParseColour = $sk->postParseLinkColour( !$wgUseOldExistenceCheck );
1145 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1147 #split the entire text string on occurences of [[
1148 $a = explode( '[[', ' ' . $s );
1149 #get the first element (all text up to first [[), and remove the space we added
1150 $s = array_shift( $a );
1151 $s = substr( $s, 1 );
1153 # Match a link having the form [[namespace:link|alternate]]trail
1155 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1156 # Match cases where there is no "]]", which might still be images
1157 static $e1_img = FALSE;
1158 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1159 # Match the end of a line for a word that's not followed by whitespace,
1160 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1161 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1163 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1165 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1167 if ( $useLinkPrefixExtension ) {
1168 if ( preg_match( $e2, $s, $m ) ) {
1169 $first_prefix = $m[2];
1172 $first_prefix = false;
1178 $selflink = $this->mTitle
->getPrefixedText();
1179 wfProfileOut( $fname.'-setup' );
1181 $checkVariantLink = sizeof($wgContLang->getVariants())>1;
1182 $useSubpages = $this->areSubpagesAllowed();
1184 # Loop for each link
1185 for ($k = 0; isset( $a[$k] ); $k++
) {
1187 if ( $useLinkPrefixExtension ) {
1188 wfProfileIn( $fname.'-prefixhandling' );
1189 if ( preg_match( $e2, $s, $m ) ) {
1197 $prefix = $first_prefix;
1198 $first_prefix = false;
1200 wfProfileOut( $fname.'-prefixhandling' );
1203 $might_be_img = false;
1205 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1207 # fix up urlencoded title texts
1208 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1210 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1211 $might_be_img = true;
1213 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1215 } else { # Invalid form; output directly
1216 $s .= $prefix . '[[' . $line ;
1220 # Don't allow internal links to pages containing
1221 # PROTO: where PROTO is a valid URL protocol; these
1222 # should be external links.
1223 if (preg_match('/^((?:'.URL_PROTOCOLS
.'):)/', $m[1])) {
1224 $s .= $prefix . '[[' . $line ;
1228 # Make subpage if necessary
1229 if( $useSubpages ) {
1230 $link = $this->maybeDoSubpageLink( $m[1], $text );
1235 $noforce = (substr($m[1], 0, 1) != ':');
1237 # Strip off leading ':'
1238 $link = substr($link, 1);
1241 $nt =& Title
::newFromText( $this->unstripNoWiki($link, $this->mStripState
) );
1243 $s .= $prefix . '[[' . $line;
1247 #check other language variants of the link
1248 #if the article does not exist
1249 if( $checkVariantLink
1250 && $nt->getArticleID() == 0 ) {
1251 $wgContLang->findVariantLink($link, $nt);
1254 $ns = $nt->getNamespace();
1255 $iw = $nt->getInterWiki();
1257 if ($might_be_img) { # if this is actually an invalid link
1258 if ($ns == NS_IMAGE
&& $noforce) { #but might be an image
1260 while (isset ($a[$k+
1]) ) {
1261 #look at the next 'line' to see if we can close it there
1262 $next_line = array_shift(array_splice( $a, $k +
1, 1) );
1263 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1264 # the first ]] closes the inner link, the second the image
1266 $text .= '[[' . $m[1];
1269 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1270 #if there's exactly one ]] that's fine, we'll keep looking
1271 $text .= '[[' . $m[0];
1273 #if $next_line is invalid too, we need look no further
1274 $text .= '[[' . $next_line;
1279 # we couldn't find the end of this imageLink, so output it raw
1280 #but don't ignore what might be perfectly normal links in the text we've examined
1281 $text = $this->replaceInternalLinks($text);
1282 $s .= $prefix . '[[' . $link . '|' . $text;
1283 # note: no $trail, because without an end, there *is* no trail
1286 } else { #it's not an image, so output it raw
1287 $s .= $prefix . '[[' . $link . '|' . $text;
1288 # note: no $trail, because without an end, there *is* no trail
1293 $wasblank = ( '' == $text );
1294 if( $wasblank ) $text = $link;
1297 # Link not escaped by : , create the various objects
1301 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1302 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1303 $tmp = $prefix . $trail ;
1304 $s .= (trim($tmp) == '')?
'': $tmp;
1308 if ( $ns == NS_IMAGE
) {
1309 wfProfileIn( "$fname-image" );
1311 # recursively parse links inside the image caption
1312 # actually, this will parse them in any other parameters, too,
1313 # but it might be hard to fix that, and it doesn't matter ATM
1314 $text = $this->replaceExternalLinks($text);
1315 $text = $this->replaceInternalLinks($text);
1317 # replace the image with a link-holder so that replaceExternalLinks() can't mess with it
1318 $s .= $prefix . $this->insertStripItem( $sk->makeImageLinkObj( $nt, $text ), $this->mStripState
) . $trail;
1319 $wgLinkCache->addImageLinkObj( $nt );
1321 wfProfileOut( "$fname-image" );
1325 if ( $ns == NS_CATEGORY
) {
1326 wfProfileIn( "$fname-category" );
1327 $t = $nt->getText();
1329 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1330 $pPLC=$sk->postParseLinkColour();
1331 $sk->postParseLinkColour( false );
1332 $t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
1333 $sk->postParseLinkColour( $pPLC );
1334 $wgLinkCache->resume();
1337 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1338 $sortkey = $this->mTitle
->getText();
1340 $sortkey = $this->mTitle
->getPrefixedText();
1345 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1346 $this->mOutput
->addCategoryLink( $t );
1347 $s .= $prefix . $trail ;
1349 wfProfileOut( "$fname-category" );
1354 if( ( $nt->getPrefixedText() === $selflink ) &&
1355 ( $nt->getFragment() === '' ) ) {
1356 # Self-links are handled specially; generally de-link and change to bold.
1357 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1361 # Special and Media are pseudo-namespaces; no pages actually exist in them
1362 if( $ns == NS_MEDIA
) {
1363 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text, true ) . $trail;
1364 $wgLinkCache->addImageLinkObj( $nt );
1366 } elseif( $ns == NS_SPECIAL
) {
1367 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1370 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1372 $sk->postParseLinkColour( $saveParseColour );
1373 wfProfileOut( $fname );
1378 * Return true if subpage links should be expanded on this page.
1381 function areSubpagesAllowed() {
1382 # Some namespaces don't allow subpages
1383 global $wgNamespacesWithSubpages;
1384 return !empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]);
1388 * Handle link to subpage if necessary
1389 * @param string $target the source of the link
1390 * @param string &$text the link text, modified as necessary
1391 * @return string the full name of the link
1394 function maybeDoSubpageLink($target, &$text) {
1397 # :Foobar -- override special treatment of prefix (images, language links)
1398 # /Foobar -- convert to CurrentPage/Foobar
1399 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1400 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1401 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1403 $fname = 'Parser::maybeDoSubpageLink';
1404 wfProfileIn( $fname );
1405 $ret = $target; # default return value is no change
1407 # Some namespaces don't allow subpages,
1408 # so only perform processing if subpages are allowed
1409 if( $this->areSubpagesAllowed() ) {
1410 # Look at the first character
1411 if( $target != '' && $target{0} == '/' ) {
1412 # / at end means we don't want the slash to be shown
1413 if( substr( $target, -1, 1 ) == '/' ) {
1414 $target = substr( $target, 1, -1 );
1417 $noslash = substr( $target, 1 );
1420 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1421 if( '' === $text ) {
1423 } # this might be changed for ugliness reasons
1425 # check for .. subpage backlinks
1427 $nodotdot = $target;
1428 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1430 $nodotdot = substr( $nodotdot, 3 );
1432 if($dotdotcount > 0) {
1433 $exploded = explode( '/', $this->mTitle
->GetPrefixedText() );
1434 if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1435 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1436 # / at the end means don't show full path
1437 if( substr( $nodotdot, -1, 1 ) == '/' ) {
1438 $nodotdot = substr( $nodotdot, 0, -1 );
1439 if( '' === $text ) {
1443 $nodotdot = trim( $nodotdot );
1444 if( $nodotdot != '' ) {
1445 $ret .= '/' . $nodotdot;
1452 wfProfileOut( $fname );
1457 * Used by doBlockLevels()
1460 /* private */ function closeParagraph() {
1462 if ( '' != $this->mLastSection
) {
1463 $result = '</' . $this->mLastSection
. ">\n";
1465 $this->mInPre
= false;
1466 $this->mLastSection
= '';
1469 # getCommon() returns the length of the longest common substring
1470 # of both arguments, starting at the beginning of both.
1472 /* private */ function getCommon( $st1, $st2 ) {
1473 $fl = strlen( $st1 );
1474 $shorter = strlen( $st2 );
1475 if ( $fl < $shorter ) { $shorter = $fl; }
1477 for ( $i = 0; $i < $shorter; ++
$i ) {
1478 if ( $st1{$i} != $st2{$i} ) { break; }
1482 # These next three functions open, continue, and close the list
1483 # element appropriate to the prefix character passed into them.
1485 /* private */ function openList( $char ) {
1486 $result = $this->closeParagraph();
1488 if ( '*' == $char ) { $result .= '<ul><li>'; }
1489 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1490 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1491 else if ( ';' == $char ) {
1492 $result .= '<dl><dt>';
1493 $this->mDTopen
= true;
1495 else { $result = '<!-- ERR 1 -->'; }
1500 /* private */ function nextItem( $char ) {
1501 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1502 else if ( ':' == $char ||
';' == $char ) {
1504 if ( $this->mDTopen
) { $close = '</dt>'; }
1505 if ( ';' == $char ) {
1506 $this->mDTopen
= true;
1507 return $close . '<dt>';
1509 $this->mDTopen
= false;
1510 return $close . '<dd>';
1513 return '<!-- ERR 2 -->';
1516 /* private */ function closeList( $char ) {
1517 if ( '*' == $char ) { $text = '</li></ul>'; }
1518 else if ( '#' == $char ) { $text = '</li></ol>'; }
1519 else if ( ':' == $char ) {
1520 if ( $this->mDTopen
) {
1521 $this->mDTopen
= false;
1522 $text = '</dt></dl>';
1524 $text = '</dd></dl>';
1527 else { return '<!-- ERR 3 -->'; }
1533 * Make lists from lines starting with ':', '*', '#', etc.
1536 * @return string the lists rendered as HTML
1538 function doBlockLevels( $text, $linestart ) {
1539 $fname = 'Parser::doBlockLevels';
1540 wfProfileIn( $fname );
1542 # Parsing through the text line by line. The main thing
1543 # happening here is handling of block-level elements p, pre,
1544 # and making lists from lines starting with * # : etc.
1546 $textLines = explode( "\n", $text );
1548 $lastPrefix = $output = $lastLine = '';
1549 $this->mDTopen
= $inBlockElem = false;
1551 $paragraphStack = false;
1553 if ( !$linestart ) {
1554 $output .= array_shift( $textLines );
1556 foreach ( $textLines as $oLine ) {
1557 $lastPrefixLength = strlen( $lastPrefix );
1558 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1559 $preOpenMatch = preg_match('/<pre/i', $oLine );
1560 if ( !$this->mInPre
) {
1561 # Multiple prefixes may abut each other for nested lists.
1562 $prefixLength = strspn( $oLine, '*#:;' );
1563 $pref = substr( $oLine, 0, $prefixLength );
1566 $pref2 = str_replace( ';', ':', $pref );
1567 $t = substr( $oLine, $prefixLength );
1568 $this->mInPre
= !empty($preOpenMatch);
1570 # Don't interpret any other prefixes in preformatted text
1572 $pref = $pref2 = '';
1577 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1578 # Same as the last item, so no need to deal with nesting or opening stuff
1579 $output .= $this->nextItem( substr( $pref, -1 ) );
1580 $paragraphStack = false;
1582 if ( substr( $pref, -1 ) == ';') {
1583 # The one nasty exception: definition lists work like this:
1584 # ; title : definition text
1585 # So we check for : in the remainder text to split up the
1586 # title and definition, without b0rking links.
1587 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1589 $output .= $term . $this->nextItem( ':' );
1592 } elseif( $prefixLength ||
$lastPrefixLength ) {
1593 # Either open or close a level...
1594 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1595 $paragraphStack = false;
1597 while( $commonPrefixLength < $lastPrefixLength ) {
1598 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1599 --$lastPrefixLength;
1601 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1602 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1604 while ( $prefixLength > $commonPrefixLength ) {
1605 $char = substr( $pref, $commonPrefixLength, 1 );
1606 $output .= $this->openList( $char );
1608 if ( ';' == $char ) {
1609 # FIXME: This is dupe of code above
1610 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1612 $output .= $term . $this->nextItem( ':' );
1615 ++
$commonPrefixLength;
1617 $lastPrefix = $pref2;
1619 if( 0 == $prefixLength ) {
1620 wfProfileIn( "$fname-paragraph" );
1621 # No prefix (not in list)--go to paragraph mode
1622 $uniq_prefix = UNIQ_PREFIX
;
1623 // XXX: use a stack for nestable elements like span, table and div
1624 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
1625 $closematch = preg_match(
1626 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1627 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/iS', $t );
1628 if ( $openmatch or $closematch ) {
1629 $paragraphStack = false;
1630 $output .= $this->closeParagraph();
1631 if($preOpenMatch and !$preCloseMatch) {
1632 $this->mInPre
= true;
1634 if ( $closematch ) {
1635 $inBlockElem = false;
1637 $inBlockElem = true;
1639 } else if ( !$inBlockElem && !$this->mInPre
) {
1640 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1642 if ($this->mLastSection
!= 'pre') {
1643 $paragraphStack = false;
1644 $output .= $this->closeParagraph().'<pre>';
1645 $this->mLastSection
= 'pre';
1647 $t = substr( $t, 1 );
1650 if ( '' == trim($t) ) {
1651 if ( $paragraphStack ) {
1652 $output .= $paragraphStack.'<br />';
1653 $paragraphStack = false;
1654 $this->mLastSection
= 'p';
1656 if ($this->mLastSection
!= 'p' ) {
1657 $output .= $this->closeParagraph();
1658 $this->mLastSection
= '';
1659 $paragraphStack = '<p>';
1661 $paragraphStack = '</p><p>';
1665 if ( $paragraphStack ) {
1666 $output .= $paragraphStack;
1667 $paragraphStack = false;
1668 $this->mLastSection
= 'p';
1669 } else if ($this->mLastSection
!= 'p') {
1670 $output .= $this->closeParagraph().'<p>';
1671 $this->mLastSection
= 'p';
1676 wfProfileOut( "$fname-paragraph" );
1678 if ($paragraphStack === false) {
1682 while ( $prefixLength ) {
1683 $output .= $this->closeList( $pref2{$prefixLength-1} );
1686 if ( '' != $this->mLastSection
) {
1687 $output .= '</' . $this->mLastSection
. '>';
1688 $this->mLastSection
= '';
1691 wfProfileOut( $fname );
1696 * Split up a string on ':', ignoring any occurences inside
1697 * <a>..</a> or <span>...</span>
1698 * @param string $str the string to split
1699 * @param string &$before set to everything before the ':'
1700 * @param string &$after set to everything after the ':'
1701 * return string the position of the ':', or false if none found
1703 function findColonNoLinks($str, &$before, &$after) {
1704 # I wonder if we should make this count all tags, not just <a>
1705 # and <span>. That would prevent us from matching a ':' that
1706 # comes in the middle of italics other such formatting....
1708 $fname = 'Parser::findColonNoLinks';
1709 wfProfileIn( $fname );
1712 $colon = strpos($str, ':', $pos);
1714 if ($colon !== false) {
1715 $before = substr($str, 0, $colon);
1716 $after = substr($str, $colon +
1);
1718 # Skip any ':' within <a> or <span> pairs
1719 $a = substr_count($before, '<a');
1720 $s = substr_count($before, '<span');
1721 $ca = substr_count($before, '</a>');
1722 $cs = substr_count($before, '</span>');
1724 if ($a <= $ca and $s <= $cs) {
1725 # Tags are balanced before ':'; ok
1730 } while ($colon !== false);
1731 wfProfileOut( $fname );
1736 * Return value of a magic variable (like PAGENAME)
1740 function getVariableValue( $index ) {
1741 global $wgContLang, $wgSitename, $wgServer;
1744 * Some of these require message or data lookups and can be
1745 * expensive to check many times.
1747 static $varCache = array();
1748 if( isset( $varCache[$index] ) ) return $varCache[$index];
1751 case MAG_CURRENTMONTH
:
1752 return $varCache[$index] = $wgContLang->formatNum( date( 'm' ) );
1753 case MAG_CURRENTMONTHNAME
:
1754 return $varCache[$index] = $wgContLang->getMonthName( date('n') );
1755 case MAG_CURRENTMONTHNAMEGEN
:
1756 return $varCache[$index] = $wgContLang->getMonthNameGen( date('n') );
1757 case MAG_CURRENTDAY
:
1758 return $varCache[$index] = $wgContLang->formatNum( date('j') );
1760 return $this->mTitle
->getText();
1762 return $this->mTitle
->getPartialURL();
1764 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1765 return $wgContLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1766 case MAG_CURRENTDAYNAME
:
1767 return $varCache[$index] = $wgContLang->getWeekdayName( date('w')+
1 );
1768 case MAG_CURRENTYEAR
:
1769 return $varCache[$index] = $wgContLang->formatNum( date( 'Y' ) );
1770 case MAG_CURRENTTIME
:
1771 return $varCache[$index] = $wgContLang->time( wfTimestampNow(), false );
1772 case MAG_NUMBEROFARTICLES
:
1773 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfArticles() );
1784 * initialise the magic variables (like CURRENTMONTHNAME)
1788 function initialiseVariables() {
1789 $fname = 'Parser::initialiseVariables';
1790 wfProfileIn( $fname );
1791 global $wgVariableIDs;
1792 $this->mVariables
= array();
1793 foreach ( $wgVariableIDs as $id ) {
1794 $mw =& MagicWord
::get( $id );
1795 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1797 wfProfileOut( $fname );
1801 * Replace magic variables, templates, and template arguments
1802 * with the appropriate text. Templates are substituted recursively,
1803 * taking care to avoid infinite loops.
1805 * Note that the substitution depends on value of $mOutputType:
1806 * OT_WIKI: only {{subst:}} templates
1807 * OT_MSG: only magic variables
1808 * OT_HTML: all templates and magic variables
1810 * @param string $tex The text to transform
1811 * @param array $args Key-value pairs representing template parameters to substitute
1814 function replaceVariables( $text, $args = array() ) {
1815 global $wgLang, $wgScript, $wgArticlePath;
1817 # Prevent too big inclusions
1818 if( strlen( $text ) > MAX_INCLUDE_SIZE
) {
1822 $fname = 'Parser::replaceVariables';
1823 wfProfileIn( $fname );
1825 $titleChars = Title
::legalChars();
1827 # This function is called recursively. To keep track of arguments we need a stack:
1828 array_push( $this->mArgStack
, $args );
1830 # Variable substitution
1831 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", array( &$this, 'variableSubstitution' ), $text );
1833 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1834 # Argument substitution
1835 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", array( &$this, 'argSubstitution' ), $text );
1837 # Template substitution
1838 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1839 $text = preg_replace_callback( $regex, array( &$this, 'braceSubstitution' ), $text );
1841 array_pop( $this->mArgStack
);
1843 wfProfileOut( $fname );
1848 * Replace magic variables
1851 function variableSubstitution( $matches ) {
1852 $fname = 'parser::variableSubstitution';
1853 wfProfileIn( $fname );
1854 if ( !$this->mVariables
) {
1855 $this->initialiseVariables();
1858 if ( $this->mOutputType
== OT_WIKI
) {
1859 # Do only magic variables prefixed by SUBST
1860 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1861 if (!$mwSubst->matchStartAndRemove( $matches[1] ))
1863 # Note that if we don't substitute the variable below,
1864 # we don't remove the {{subst:}} magic word, in case
1865 # it is a template rather than a magic variable.
1867 if ( !$skip && array_key_exists( $matches[1], $this->mVariables
) ) {
1868 $text = $this->mVariables
[$matches[1]];
1869 $this->mOutput
->mContainsOldMagic
= true;
1871 $text = $matches[0];
1873 wfProfileOut( $fname );
1877 # Split template arguments
1878 function getTemplateArgs( $argsString ) {
1879 if ( $argsString === '' ) {
1883 $args = explode( '|', substr( $argsString, 1 ) );
1885 # If any of the arguments contains a '[[' but no ']]', it needs to be
1886 # merged with the next arg because the '|' character between belongs
1887 # to the link syntax and not the template parameter syntax.
1888 $argc = count($args);
1890 for ( $i = 0; $i < $argc-1; $i++
) {
1891 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1892 $args[$i] .= '|'.$args[$i+
1];
1893 array_splice($args, $i+
1, 1);
1903 * Return the text of a template, after recursively
1904 * replacing any variables or templates within the template.
1906 * @param array $matches The parts of the template
1907 * $matches[1]: the title, i.e. the part before the |
1908 * $matches[2]: the parameters (including a leading |), if any
1909 * @return string the text of the template
1912 function braceSubstitution( $matches ) {
1913 global $wgLinkCache, $wgContLang;
1914 $fname = 'Parser::braceSubstitution';
1915 wfProfileIn( $fname );
1923 # Need to know if the template comes at the start of a line,
1924 # to treat the beginning of the template like the beginning
1925 # of a line for tables and block-level elements.
1926 $linestart = $matches[1];
1928 # $part1 is the bit before the first |, and must contain only title characters
1929 # $args is a list of arguments, starting from index 0, not including $part1
1931 $part1 = $matches[2];
1932 # If the third subpattern matched anything, it will start with |
1934 $args = $this->getTemplateArgs($matches[3]);
1935 $argc = count( $args );
1937 # Don't parse {{{}}} because that's only for template arguments
1938 if ( $linestart === '{' ) {
1939 $text = $matches[0];
1946 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1947 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
1948 # One of two possibilities is true:
1949 # 1) Found SUBST but not in the PST phase
1950 # 2) Didn't find SUBST and in the PST phase
1951 # In either case, return without further processing
1952 $text = $matches[0];
1958 # MSG, MSGNW and INT
1961 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1962 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1965 # Remove obsolete MSG:
1966 $mwMsg =& MagicWord
::get( MAG_MSG
);
1967 $mwMsg->matchStartAndRemove( $part1 );
1970 # Check if it is an internal message
1971 $mwInt =& MagicWord
::get( MAG_INT
);
1972 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1973 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1974 $text = $linestart . wfMsgReal( $part1, $args, true );
1982 # Check for NS: (namespace expansion)
1983 $mwNs = MagicWord
::get( MAG_NS
);
1984 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1985 if ( intval( $part1 ) ) {
1986 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1989 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1990 if ( !is_null( $index ) ) {
1991 $text = $linestart . $wgContLang->getNsText( $index );
1998 # LOCALURL and LOCALURLE
2000 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
2001 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
2003 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
2004 $func = 'getLocalURL';
2005 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
2006 $func = 'escapeLocalURL';
2011 if ( $func !== '' ) {
2012 $title = Title
::newFromText( $part1 );
2013 if ( !is_null( $title ) ) {
2015 $text = $linestart . $title->$func( $args[0] );
2017 $text = $linestart . $title->$func();
2025 if ( !$found && $argc == 1 ) {
2026 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
2027 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
2028 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
2033 # Template table test
2035 # Did we encounter this template already? If yes, it is in the cache
2036 # and we need to check for loops.
2037 if ( !$found && isset( $this->mTemplates
[$part1] ) ) {
2038 # set $text to cached message.
2039 $text = $linestart . $this->mTemplates
[$part1];
2042 # Infinite loop test
2043 if ( isset( $this->mTemplatePath
[$part1] ) ) {
2046 $text .= '<!-- WARNING: template loop detected -->';
2050 # Load from database
2051 $itcamefromthedatabase = false;
2054 $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
2055 if ($subpage !== '') {
2056 $ns = $this->mTitle
->getNamespace();
2058 $title = Title
::newFromText( $part1, $ns );
2059 if ( !is_null( $title ) && !$title->isExternal() ) {
2060 # Check for excessive inclusion
2061 $dbk = $title->getPrefixedDBkey();
2062 if ( $this->incrementIncludeCount( $dbk ) ) {
2063 # This should never be reached.
2064 $article = new Article( $title );
2065 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
2066 if ( $articleContent !== false ) {
2068 $text = $linestart . $articleContent;
2069 $itcamefromthedatabase = true;
2073 # If the title is valid but undisplayable, make a link to it
2074 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
2075 $text = $linestart . '[['.$title->getPrefixedText().']]';
2079 # Template cache array insertion
2080 $this->mTemplates
[$part1] = $text;
2084 # Recursive parsing, escaping and link table handling
2085 # Only for HTML output
2086 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
2087 $text = wfEscapeWikiText( $text );
2088 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
2089 # Clean up argument array
2090 $assocArgs = array();
2092 foreach( $args as $arg ) {
2093 $eqpos = strpos( $arg, '=' );
2094 if ( $eqpos === false ) {
2095 $assocArgs[$index++
] = $arg;
2097 $name = trim( substr( $arg, 0, $eqpos ) );
2098 $value = trim( substr( $arg, $eqpos+
1 ) );
2099 if ( $value === false ) {
2102 if ( $name !== false ) {
2103 $assocArgs[$name] = $value;
2108 # Add a new element to the templace recursion path
2109 $this->mTemplatePath
[$part1] = 1;
2111 $text = $this->strip( $text, $this->mStripState
);
2112 $text = $this->removeHTMLtags( $text );
2113 $text = $this->replaceVariables( $text, $assocArgs );
2115 # Resume the link cache and register the inclusion as a link
2116 if ( $this->mOutputType
== OT_HTML
&& !is_null( $title ) ) {
2117 $wgLinkCache->addLinkObj( $title );
2120 # If the template begins with a table or block-level
2121 # element, it should be treated as beginning a new line.
2122 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
2123 $text = "\n" . $text;
2127 # Empties the template path
2128 $this->mTemplatePath
= array();
2130 wfProfileOut( $fname );
2133 # replace ==section headers==
2134 # XXX this needs to go away once we have a better parser.
2135 if ( $this->mOutputType
!= OT_WIKI
&& $itcamefromthedatabase ) {
2136 if( !is_null( $title ) )
2137 $encodedname = base64_encode($title->getPrefixedDBkey());
2139 $encodedname = base64_encode("");
2140 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
2141 PREG_SPLIT_DELIM_CAPTURE
);
2144 for( $i = 0; $i < count($m); $i +
= 2 ) {
2146 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
2148 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
2152 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
2153 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
2154 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
2161 # Empties the template path
2162 $this->mTemplatePath
= array();
2165 wfProfileOut( $fname );
2168 wfProfileOut( $fname );
2174 * Triple brace replacement -- used for template arguments
2177 function argSubstitution( $matches ) {
2178 $arg = trim( $matches[1] );
2179 $text = $matches[0];
2180 $inputArgs = end( $this->mArgStack
);
2182 if ( array_key_exists( $arg, $inputArgs ) ) {
2183 $text = $inputArgs[$arg];
2190 * Returns true if the function is allowed to include this entity
2193 function incrementIncludeCount( $dbk ) {
2194 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
2195 $this->mIncludeCount
[$dbk] = 0;
2197 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
2206 * Cleans up HTML, removes dangerous tags and attributes, and
2207 * removes HTML comments
2210 function removeHTMLtags( $text ) {
2211 global $wgUseTidy, $wgUserHtml;
2212 $fname = 'Parser::removeHTMLtags';
2213 wfProfileIn( $fname );
2216 $htmlpairs = array( # Tags that must be closed
2217 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
2218 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
2219 'strike', 'strong', 'tt', 'var', 'div', 'center',
2220 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
2221 'ruby', 'rt' , 'rb' , 'rp', 'p'
2223 $htmlsingle = array(
2224 'br', 'hr', 'li', 'dt', 'dd'
2226 $htmlnest = array( # Tags that can be nested--??
2227 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
2228 'dl', 'font', 'big', 'small', 'sub', 'sup'
2230 $tabletags = array( # Can only appear inside table
2234 $htmlpairs = array();
2235 $htmlsingle = array();
2236 $htmlnest = array();
2237 $tabletags = array();
2240 $htmlsingle = array_merge( $tabletags, $htmlsingle );
2241 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
2243 $htmlattrs = $this->getHTMLattrs () ;
2245 # Remove HTML comments
2246 $text = $this->removeHTMLcomments( $text );
2248 $bits = explode( '<', $text );
2249 $text = array_shift( $bits );
2251 $tagstack = array(); $tablestack = array();
2252 foreach ( $bits as $x ) {
2253 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
2254 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2256 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2257 error_reporting( $prev );
2260 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2264 if ( ! in_array( $t, $htmlsingle ) &&
2265 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
2266 @array_push
( $tagstack, $ot );
2269 if ( $t == 'table' ) {
2270 $tagstack = array_pop( $tablestack );
2275 # Keep track for later
2276 if ( in_array( $t, $tabletags ) &&
2277 ! in_array( 'table', $tagstack ) ) {
2279 } else if ( in_array( $t, $tagstack ) &&
2280 ! in_array ( $t , $htmlnest ) ) {
2282 } else if ( ! in_array( $t, $htmlsingle ) ) {
2283 if ( $t == 'table' ) {
2284 array_push( $tablestack, $tagstack );
2285 $tagstack = array();
2287 array_push( $tagstack, $t );
2289 # Strip non-approved attributes from the tag
2290 $newparams = $this->fixTagAttributes($params);
2294 $rest = str_replace( '>', '>', $rest );
2295 $text .= "<$slash$t $newparams$brace$rest";
2299 $text .= '<' . str_replace( '>', '>', $x);
2301 # Close off any remaining tags
2302 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2304 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2307 # this might be possible using tidy itself
2308 foreach ( $bits as $x ) {
2309 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2311 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2312 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2313 $newparams = $this->fixTagAttributes($params);
2314 $rest = str_replace( '>', '>', $rest );
2315 $text .= "<$slash$t $newparams$brace$rest";
2317 $text .= '<' . str_replace( '>', '>', $x);
2321 wfProfileOut( $fname );
2326 * Remove '<!--', '-->', and everything between.
2327 * To avoid leaving blank lines, when a comment is both preceded
2328 * and followed by a newline (ignoring spaces), trim leading and
2329 * trailing spaces and one of the newlines.
2333 function removeHTMLcomments( $text ) {
2334 $fname='Parser::removeHTMLcomments';
2335 wfProfileIn( $fname );
2336 while (($start = strpos($text, '<!--')) !== false) {
2337 $end = strpos($text, '-->', $start +
4);
2338 if ($end === false) {
2339 # Unterminated comment; bail out
2345 # Trim space and newline if the comment is both
2346 # preceded and followed by a newline
2347 $spaceStart = max($start - 1, 0);
2348 $spaceLen = $end - $spaceStart;
2349 while (substr($text, $spaceStart, 1) === ' ' && $spaceStart > 0) {
2353 while (substr($text, $spaceStart +
$spaceLen, 1) === ' ')
2355 if (substr($text, $spaceStart, 1) === "\n" and substr($text, $spaceStart +
$spaceLen, 1) === "\n") {
2356 # Remove the comment, leading and trailing
2357 # spaces, and leave only one newline.
2358 $text = substr_replace($text, "\n", $spaceStart, $spaceLen +
1);
2361 # Remove just the comment.
2362 $text = substr_replace($text, '', $start, $end - $start);
2365 wfProfileOut( $fname );
2370 * This function accomplishes several tasks:
2371 * 1) Auto-number headings if that option is enabled
2372 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2373 * 3) Add a Table of contents on the top for users who have enabled the option
2374 * 4) Auto-anchor headings
2376 * It loops through all headlines, collects the necessary data, then splits up the
2377 * string and re-inserts the newly formatted headlines.
2380 /* private */ function formatHeadings( $text, $isMain=true ) {
2381 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders;
2383 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2384 $doShowToc = $this->mOptions
->getShowToc();
2385 $forceTocHere = false;
2386 if( !$this->mTitle
->userCanEdit() ) {
2388 $rightClickHack = 0;
2390 $showEditLink = $this->mOptions
->getEditSection();
2391 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
2394 # Inhibit editsection links if requested in the page
2395 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2396 if( $esw->matchAndRemove( $text ) ) {
2399 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2401 $mw =& MagicWord
::get( MAG_NOTOC
);
2402 if( $mw->matchAndRemove( $text ) ) {
2406 # never add the TOC to the Main Page. This is an entry page that should not
2407 # be more than 1-2 screens large anyway
2408 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
2412 # Get all headlines for numbering them and adding funky stuff like [edit]
2413 # links - this is for later, but we need the number of headlines right now
2414 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
2416 # if there are fewer than 4 headlines in the article, do not show TOC
2417 if( $numMatches < 4 ) {
2421 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2422 # override above conditions and always show TOC at that place
2423 $mw =& MagicWord
::get( MAG_TOC
);
2424 if ($mw->match( $text ) ) {
2426 $forceTocHere = true;
2428 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2429 # override above conditions and always show TOC above first header
2430 $mw =& MagicWord
::get( MAG_FORCETOC
);
2431 if ($mw->matchAndRemove( $text ) ) {
2438 # We need this to perform operations on the HTML
2439 $sk =& $this->mOptions
->getSkin();
2443 $sectionCount = 0; # headlineCount excluding template sections
2445 # Ugh .. the TOC should have neat indentation levels which can be
2446 # passed to the skin functions. These are determined here
2451 $sublevelCount = array();
2454 foreach( $matches[3] as $headline ) {
2456 $templatetitle = "";
2457 $templatesection = 0;
2459 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2461 $templatetitle = base64_decode($mat[1]);
2462 $templatesection = 1 +
(int)base64_decode($mat[2]);
2463 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2468 $prevlevel = $level;
2470 $level = $matches[1][$headlineCount];
2471 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2472 # reset when we enter a new level
2473 $sublevelCount[$level] = 0;
2474 $toc .= $sk->tocIndent( $level - $prevlevel );
2475 $toclevel +
= $level - $prevlevel;
2477 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2478 # reset when we step back a level
2479 $sublevelCount[$level+
1]=0;
2480 $toc .= $sk->tocUnindent( $prevlevel - $level );
2481 $toclevel -= $prevlevel - $level;
2483 # count number of headlines for each level
2484 @$sublevelCount[$level]++
;
2485 if( $doNumberHeadings ||
$doShowToc ) {
2487 for( $i = 1; $i <= $level; $i++
) {
2488 if( !empty( $sublevelCount[$i] ) ) {
2492 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2498 # The canonized header is a version of the header text safe to use for links
2499 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2500 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2501 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2503 # Remove link placeholders by the link text.
2504 # <!--LINK number-->
2506 # link text with suffix
2507 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2508 "\$wgLinkHolders['texts'][\$1]",
2509 $canonized_headline );
2512 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2513 $tocline = trim( $canonized_headline );
2514 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2515 $replacearray = array(
2519 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2520 $refer[$headlineCount] = $canonized_headline;
2522 # count how many in assoc. array so we can track dupes in anchors
2523 @$refers[$canonized_headline]++
;
2524 $refcount[$headlineCount]=$refers[$canonized_headline];
2526 # Prepend the number to the heading text
2528 if( $doNumberHeadings ||
$doShowToc ) {
2529 $tocline = $numbering . ' ' . $tocline;
2531 # Don't number the heading if it is the only one (looks silly)
2532 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2533 # the two are different if the line contains a link
2534 $headline=$numbering . ' ' . $headline;
2538 # Create the anchor for linking from the TOC to the section
2539 $anchor = $canonized_headline;
2540 if($refcount[$headlineCount] > 1 ) {
2541 $anchor .= '_' . $refcount[$headlineCount];
2543 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2544 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2546 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2547 if ( empty( $head[$headlineCount] ) ) {
2548 $head[$headlineCount] = '';
2551 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2553 $head[$headlineCount] .= $sk->editSectionLink($this->mTitle
, $sectionCount+
1);
2556 # Add the edit section span
2557 if( $rightClickHack ) {
2559 $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
2561 $headline = $sk->editSectionScript($this->mTitle
, $sectionCount+
1,$headline);
2564 # give headline the correct <h#> tag
2565 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2573 $toclines = $headlineCount;
2574 $toc .= $sk->tocUnindent( $toclevel );
2575 $toc = $sk->tocTable( $toc );
2578 # split up and insert constructed headlines
2580 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2583 foreach( $blocks as $block ) {
2584 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2585 # This is the [edit] link that appears for the top block of text when
2586 # section editing is enabled
2588 # Disabled because it broke block formatting
2589 # For example, a bullet point in the top line
2590 # $full .= $sk->editSectionLink(0);
2593 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2594 # Top anchor now in skin
2598 if( !empty( $head[$i] ) ) {
2604 $mw =& MagicWord
::get( MAG_TOC
);
2605 return $mw->replace( $toc, $full );
2612 * Return an HTML link for the "ISBN 123456" text
2615 function magicISBN( $text ) {
2617 $fname = 'Parser::magicISBN';
2618 wfProfileIn( $fname );
2620 $a = split( 'ISBN ', ' '.$text );
2621 if ( count ( $a ) < 2 ) {
2622 wfProfileOut( $fname );
2625 $text = substr( array_shift( $a ), 1);
2626 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2628 foreach ( $a as $x ) {
2629 $isbn = $blank = '' ;
2630 while ( ' ' == $x{0} ) {
2632 $x = substr( $x, 1 );
2634 if ( $x == '' ) { # blank isbn
2635 $text .= "ISBN $blank";
2638 while ( strstr( $valid, $x{0} ) != false ) {
2640 $x = substr( $x, 1 );
2642 $num = str_replace( '-', '', $isbn );
2643 $num = str_replace( ' ', '', $num );
2646 $text .= "ISBN $blank$x";
2648 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2649 $text .= '<a href="' .
2650 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2651 "\" class=\"internal\">ISBN $isbn</a>";
2655 wfProfileOut( $fname );
2660 * Return an HTML link for the "GEO ..." text
2663 function magicGEO( $text ) {
2664 global $wgLang, $wgUseGeoMode;
2665 $fname = 'Parser::magicGEO';
2666 wfProfileIn( $fname );
2668 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2669 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2670 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2671 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2672 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2673 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2675 $a = split( 'GEO ', ' '.$text );
2676 if ( count ( $a ) < 2 ) {
2677 wfProfileOut( $fname );
2680 $text = substr( array_shift( $a ), 1);
2681 $valid = '0123456789.+-:';
2683 foreach ( $a as $x ) {
2684 $geo = $blank = '' ;
2685 while ( ' ' == $x{0} ) {
2687 $x = substr( $x, 1 );
2689 while ( strstr( $valid, $x{0} ) != false ) {
2691 $x = substr( $x, 1 );
2693 $num = str_replace( '+', '', $geo );
2694 $num = str_replace( ' ', '', $num );
2696 if ( '' == $num ||
count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2697 $text .= "GEO $blank$x";
2699 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2700 $text .= '<a href="' .
2701 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2702 "\" class=\"internal\">GEO $geo</a>";
2706 wfProfileOut( $fname );
2711 * Return an HTML link for the "RFC 1234" text
2713 * @param string $text text to be processed
2715 function magicRFC( $text, $keyword='RFC ', $urlmsg='rfcurl' ) {
2718 $valid = '0123456789';
2721 $a = split( $keyword, ' '.$text );
2722 if ( count ( $a ) < 2 ) {
2725 $text = substr( array_shift( $a ), 1);
2727 /* Check if keyword is preceed by [[.
2728 * This test is made here cause of the array_shift above
2729 * that prevent the test to be done in the foreach.
2731 if ( substr( $text, -2 ) == '[[' ) {
2735 foreach ( $a as $x ) {
2736 /* token might be empty if we have RFC RFC 1234 */
2744 /** remove and save whitespaces in $blank */
2745 while ( $x{0} == ' ' ) {
2747 $x = substr( $x, 1 );
2750 /** remove and save the rfc number in $id */
2751 while ( strstr( $valid, $x{0} ) != false ) {
2753 $x = substr( $x, 1 );
2757 /* call back stripped spaces*/
2758 $text .= $keyword.$blank.$x;
2759 } elseif( $internal ) {
2761 $text .= $keyword.$id.$x;
2763 /* build the external link*/
2764 $url = wfmsg( $urlmsg );
2765 $url = str_replace( '$1', $id, $url);
2766 $sk =& $this->mOptions
->getSkin();
2767 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
2768 $text .= "<a href='{$url}'{$la}>{$keyword}{$id}</a>{$x}";
2771 /* Check if the next RFC keyword is preceed by [[ */
2772 $internal = ( substr($x,-2) == '[[' );
2778 * Transform wiki markup when saving a page by doing \r\n -> \n
2779 * conversion, substitting signatures, {{subst:}} templates, etc.
2781 * @param string $text the text to transform
2782 * @param Title &$title the Title object for the current article
2783 * @param User &$user the User object describing the current user
2784 * @param ParserOptions $options parsing options
2785 * @param bool $clearState whether to clear the parser state first
2786 * @return string the altered wiki markup
2789 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2790 $this->mOptions
= $options;
2791 $this->mTitle
=& $title;
2792 $this->mOutputType
= OT_WIKI
;
2794 if ( $clearState ) {
2795 $this->clearState();
2798 $stripState = false;
2802 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
2803 $text = $this->strip( $text, $stripState, false );
2804 $text = $this->pstPass2( $text, $user );
2805 $text = $this->unstrip( $text, $stripState );
2806 $text = $this->unstripNoWiki( $text, $stripState );
2811 * Pre-save transform helper function
2814 function pstPass2( $text, &$user ) {
2815 global $wgLang, $wgContLang, $wgLocaltimezone;
2817 # Variable replacement
2818 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2819 $text = $this->replaceVariables( $text );
2823 $n = $user->getName();
2824 $k = $user->getOption( 'nickname' );
2825 if ( '' == $k ) { $k = $n; }
2826 if ( isset( $wgLocaltimezone ) ) {
2827 $oldtz = getenv( 'TZ' );
2828 putenv( 'TZ='.$wgLocaltimezone );
2830 /* Note: this is an ugly timezone hack for the European wikis */
2831 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
2832 ' (' . date( 'T' ) . ')';
2833 if ( isset( $wgLocaltimezone ) ) {
2834 putenv( 'TZ='.$oldtzs );
2837 $text = preg_replace( '/~~~~~~/', $d, $text );
2838 $text = preg_replace( '/~~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]] $d", $text );
2839 $text = preg_replace( '/~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]]", $text );
2841 # Context links: [[|name]] and [[name (context)|]]
2843 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2844 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2845 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2846 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2848 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2849 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2850 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2851 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2853 $t = $this->mTitle
->getText();
2854 if ( preg_match( $conpat, $t, $m ) ) {
2857 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2858 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2859 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2861 if ( '' == $context ) {
2862 $text = preg_replace( $p2, '[[\\1]]', $text );
2864 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2867 # Trim trailing whitespace
2868 # MAG_END (__END__) tag allows for trailing
2869 # whitespace to be deliberately included
2870 $text = rtrim( $text );
2871 $mw =& MagicWord
::get( MAG_END
);
2872 $mw->matchAndRemove( $text );
2878 * Set up some variables which are usually set up in parse()
2879 * so that an external function can call some class members with confidence
2882 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2883 $this->mTitle
=& $title;
2884 $this->mOptions
= $options;
2885 $this->mOutputType
= $outputType;
2886 if ( $clearState ) {
2887 $this->clearState();
2892 * Transform a MediaWiki message by replacing magic variables.
2894 * @param string $text the text to transform
2895 * @param ParserOptions $options options
2896 * @return string the text with variables substituted
2899 function transformMsg( $text, $options ) {
2901 static $executing = false;
2903 # Guard against infinite recursion
2909 $this->mTitle
= $wgTitle;
2910 $this->mOptions
= $options;
2911 $this->mOutputType
= OT_MSG
;
2912 $this->clearState();
2913 $text = $this->replaceVariables( $text );
2920 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2921 * Callback will be called with the text within
2922 * Transform and return the text within
2925 function setHook( $tag, $callback ) {
2926 $oldVal = @$this->mTagHooks
[$tag];
2927 $this->mTagHooks
[$tag] = $callback;
2932 * Replace <!--LINK--> link placeholders with actual links, in the buffer
2933 * Placeholders created in Skin::makeLinkObj()
2934 * Returns an array of links found, indexed by PDBK:
2938 * $options is a bit field, RLH_FOR_UPDATE to select for update
2940 function replaceLinkHolders( &$text, $options = 0 ) {
2941 global $wgUser, $wgLinkCache, $wgUseOldExistenceCheck, $wgLinkHolders;
2942 global $wgInterwikiLinkHolders;
2943 global $outputReplace;
2945 if ( $wgUseOldExistenceCheck ) {
2949 $fname = 'Parser::replaceLinkHolders';
2950 wfProfileIn( $fname );
2955 #if ( !empty( $tmpLinks[0] ) ) { #TODO
2956 if ( !empty( $wgLinkHolders['namespaces'] ) ) {
2957 wfProfileIn( $fname.'-check' );
2958 $dbr =& wfGetDB( DB_SLAVE
);
2959 $page = $dbr->tableName( 'page' );
2960 $sk = $wgUser->getSkin();
2961 $threshold = $wgUser->getOption('stubthreshold');
2964 asort( $wgLinkHolders['namespaces'] );
2968 foreach ( $wgLinkHolders['namespaces'] as $key => $val ) {
2970 $title = $wgLinkHolders['titles'][$key];
2972 # Skip invalid entries.
2973 # Result will be ugly, but prevents crash.
2974 if ( is_null( $title ) ) {
2977 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
2979 # Check if it's in the link cache already
2980 if ( $wgLinkCache->getGoodLinkID( $pdbk ) ) {
2981 $colours[$pdbk] = 1;
2982 } elseif ( $wgLinkCache->isBadLink( $pdbk ) ) {
2983 $colours[$pdbk] = 0;
2985 # Not in the link cache, add it to the query
2986 if ( !isset( $current ) ) {
2990 $query = "SELECT page_id, page_namespace, page_title";
2991 if ( $threshold > 0 ) {
2992 $textTable = $dbr->tableName( 'text' );
2993 $query .= ', LENGTH(old_text) AS page_len, page_is_redirect';
2994 $tables .= ", $textTable";
2995 $join = 'page_latest=old_id AND';
2997 $query .= " FROM $tables WHERE $join (page_namespace=$val AND page_title IN(";
2998 } elseif ( $current != $val ) {
3000 $query .= ")) OR (page_namespace=$val AND page_title IN(";
3005 $query .= $dbr->addQuotes( $wgLinkHolders['dbkeys'][$key] );
3010 if ( $options & RLH_FOR_UPDATE
) {
3011 $query .= ' FOR UPDATE';
3014 $res = $dbr->query( $query, $fname );
3016 # Fetch data and form into an associative array
3017 # non-existent = broken
3020 while ( $s = $dbr->fetchObject($res) ) {
3021 $title = Title
::makeTitle( $s->page_namespace
, $s->page_title
);
3022 $pdbk = $title->getPrefixedDBkey();
3023 $wgLinkCache->addGoodLink( $s->page_id
, $pdbk );
3025 if ( $threshold > 0 ) {
3026 $size = $s->page_len
;
3027 if ( $s->page_is_redirect ||
$s->page_namespace
!= 0 ||
$length < $threshold ) {
3028 $colours[$pdbk] = 1;
3030 $colours[$pdbk] = 2;
3033 $colours[$pdbk] = 1;
3037 wfProfileOut( $fname.'-check' );
3039 # Construct search and replace arrays
3040 wfProfileIn( $fname.'-construct' );
3041 $outputReplace = array();
3042 foreach ( $wgLinkHolders['namespaces'] as $key => $ns ) {
3043 $pdbk = $pdbks[$key];
3044 $searchkey = '<!--LINK '.$key.'-->';
3045 $title = $wgLinkHolders['titles'][$key];
3046 if ( empty( $colours[$pdbk] ) ) {
3047 $wgLinkCache->addBadLink( $pdbk );
3048 $colours[$pdbk] = 0;
3049 $outputReplace[$searchkey] = $sk->makeBrokenLinkObj( $title,
3050 $wgLinkHolders['texts'][$key],
3051 $wgLinkHolders['queries'][$key] );
3052 } elseif ( $colours[$pdbk] == 1 ) {
3053 $outputReplace[$searchkey] = $sk->makeKnownLinkObj( $title,
3054 $wgLinkHolders['texts'][$key],
3055 $wgLinkHolders['queries'][$key] );
3056 } elseif ( $colours[$pdbk] == 2 ) {
3057 $outputReplace[$searchkey] = $sk->makeStubLinkObj( $title,
3058 $wgLinkHolders['texts'][$key],
3059 $wgLinkHolders['queries'][$key] );
3062 wfProfileOut( $fname.'-construct' );
3065 wfProfileIn( $fname.'-replace' );
3067 $text = preg_replace_callback(
3068 '/(<!--LINK .*?-->)/',
3069 "outputReplaceMatches",
3071 wfProfileOut( $fname.'-replace' );
3074 if ( !empty( $wgInterwikiLinkHolders ) ) {
3075 wfProfileIn( $fname.'-interwiki' );
3076 $outputReplace = $wgInterwikiLinkHolders;
3077 $text = preg_replace_callback(
3078 '/<!--IWLINK (.*?)-->/',
3079 "outputReplaceMatches",
3081 wfProfileOut( $fname.'-interwiki' );
3084 wfProfileOut( $fname );
3089 * Renders an image gallery from a text with one line per image.
3090 * text labels may be given by using |-style alternative text. E.g.
3091 * Image:one.jpg|The number "1"
3092 * Image:tree.jpg|A tree
3093 * given as text will return the HTML of a gallery with two images,
3094 * labeled 'The number "1"' and
3097 function renderImageGallery( $text ) {
3098 global $wgLinkCache;
3099 $ig = new ImageGallery();
3100 $ig->setShowBytes( false );
3101 $ig->setShowFilename( false );
3102 $lines = explode( "\n", $text );
3104 foreach ( $lines as $line ) {
3105 # match lines like these:
3106 # Image:someimage.jpg|This is some image
3107 preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
3109 if ( count( $matches ) == 0 ) {
3112 $nt = Title
::newFromURL( $matches[1] );
3113 if( is_null( $nt ) ) {
3114 # Bogus title. Ignore these so we don't bomb out later.
3117 if ( isset( $matches[3] ) ) {
3118 $label = $matches[3];
3123 # FIXME: Use the full wiki parser and add its links
3124 # to the page's links.
3125 $html = $this->mOptions
->mSkin
->formatComment( $label );
3127 $ig->add( Image
::newFromTitle( $nt ), $html );
3128 $wgLinkCache->addImageLinkObj( $nt );
3130 return $ig->toHTML();
3136 * @package MediaWiki
3140 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
3141 var $mCacheTime; # Used in ParserCache
3142 var $mVersion; # Compatibility check
3144 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
3145 $containsOldMagic = false )
3147 $this->mText
= $text;
3148 $this->mLanguageLinks
= $languageLinks;
3149 $this->mCategoryLinks
= $categoryLinks;
3150 $this->mContainsOldMagic
= $containsOldMagic;
3151 $this->mCacheTime
= '';
3152 $this->mVersion
= MW_PARSER_VERSION
;
3155 function getText() { return $this->mText
; }
3156 function getLanguageLinks() { return $this->mLanguageLinks
; }
3157 function getCategoryLinks() { return array_keys( $this->mCategoryLinks
); }
3158 function getCacheTime() { return $this->mCacheTime
; }
3159 function containsOldMagic() { return $this->mContainsOldMagic
; }
3160 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
3161 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
3162 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
3163 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
3164 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
3165 function addCategoryLink( $c ) { $this->mCategoryLinks
[$c] = 1; }
3167 function merge( $other ) {
3168 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
3169 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
3170 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
3174 * Return true if this cached output object predates the global or
3175 * per-article cache invalidation timestamps, or if it comes from
3176 * an incompatible older version.
3178 * @param string $touched the affected article's last touched timestamp
3182 function expired( $touched ) {
3183 global $wgCacheEpoch;
3184 return $this->getCacheTime() <= $touched ||
3185 $this->getCacheTime() <= $wgCacheEpoch ||
3186 !isset( $this->mVersion
) ||
3187 version_compare( $this->mVersion
, MW_PARSER_VERSION
, "lt" );
3192 * Set options of the Parser
3194 * @package MediaWiki
3198 # All variables are private
3199 var $mUseTeX; # Use texvc to expand <math> tags
3200 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
3201 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
3202 var $mAllowExternalImages; # Allow external images inline
3203 var $mSkin; # Reference to the preferred skin
3204 var $mDateFormat; # Date format index
3205 var $mEditSection; # Create "edit section" links
3206 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
3207 var $mNumberHeadings; # Automatically number headings
3208 var $mShowToc; # Show table of contents
3210 function getUseTeX() { return $this->mUseTeX
; }
3211 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
3212 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
3213 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
3214 function getSkin() { return $this->mSkin
; }
3215 function getDateFormat() { return $this->mDateFormat
; }
3216 function getEditSection() { return $this->mEditSection
; }
3217 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
3218 function getNumberHeadings() { return $this->mNumberHeadings
; }
3219 function getShowToc() { return $this->mShowToc
; }
3221 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
3222 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
3223 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
3224 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
3225 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
3226 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
3227 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
3228 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
3229 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
3231 function setSkin( &$x ) { $this->mSkin
=& $x; }
3233 # Get parser options
3234 /* static */ function newFromUser( &$user ) {
3235 $popts = new ParserOptions
;
3236 $popts->initialiseFromUser( $user );
3241 function initialiseFromUser( &$userInput ) {
3242 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
3243 $fname = 'ParserOptions::initialiseFromUser';
3244 wfProfileIn( $fname );
3245 if ( !$userInput ) {
3247 $user->setLoaded( true );
3249 $user =& $userInput;
3252 $this->mUseTeX
= $wgUseTeX;
3253 $this->mUseDynamicDates
= $wgUseDynamicDates;
3254 $this->mInterwikiMagic
= $wgInterwikiMagic;
3255 $this->mAllowExternalImages
= $wgAllowExternalImages;
3256 wfProfileIn( $fname.'-skin' );
3257 $this->mSkin
=& $user->getSkin();
3258 wfProfileOut( $fname.'-skin' );
3259 $this->mDateFormat
= $user->getOption( 'date' );
3260 $this->mEditSection
= $user->getOption( 'editsection' );
3261 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
3262 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
3263 $this->mShowToc
= $user->getOption( 'showtoc' );
3264 wfProfileOut( $fname );
3271 * Callback function used by Parser::replaceLinkHolders()
3272 * to substitute link placeholders.
3274 function &outputReplaceMatches( $matches ) {
3275 global $outputReplace;
3276 return $outputReplace[$matches[1]];
3280 * Return the total number of articles
3282 function wfNumberOfArticles() {
3283 global $wgNumberOfArticles;
3286 return $wgNumberOfArticles;
3290 * Get various statistics from the database
3293 function wfLoadSiteStats() {
3294 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
3295 $fname = 'wfLoadSiteStats';
3297 if ( -1 != $wgNumberOfArticles ) return;
3298 $dbr =& wfGetDB( DB_SLAVE
);
3299 $s = $dbr->selectRow( 'site_stats',
3300 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
3301 array( 'ss_row_id' => 1 ), $fname
3304 if ( $s === false ) {
3307 $wgTotalViews = $s->ss_total_views
;
3308 $wgTotalEdits = $s->ss_total_edits
;
3309 $wgNumberOfArticles = $s->ss_good_articles
;
3313 function wfEscapeHTMLTagsOnly( $in ) {
3315 array( '"', '>', '<' ),
3316 array( '"', '>', '<' ),