3 * File for Parser and related classes
9 require_once( 'Sanitizer.php' );
12 * Update this version number when the ParserOutput format
13 * changes in an incompatible way, so the parser cache
14 * can automatically discard old data.
16 define( 'MW_PARSER_VERSION', '1.5.0' );
19 * Variable substitution O(N^2) attack
21 * Without countermeasures, it would be possible to attack the parser by saving
22 * a page filled with a large number of inclusions of large pages. The size of
23 * the generated page would be proportional to the square of the input size.
24 * Hence, we limit the number of inclusions of any given page, thus bringing any
25 * attack back to O(N).
28 define( 'MAX_INCLUDE_REPEAT', 100 );
29 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
31 define( 'RLH_FOR_UPDATE', 1 );
33 # Allowed values for $mOutputType
34 define( 'OT_HTML', 1 );
35 define( 'OT_WIKI', 2 );
36 define( 'OT_MSG' , 3 );
38 # string parameter for extractTags which will cause it
39 # to strip HTML comments in addition to regular
40 # <XML>-style tags. This should not be anything we
41 # may want to use in wikisyntax
42 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
44 # prefix for escaping, used in two functions at least
45 define( 'UNIQ_PREFIX', 'NaodW29');
47 # Constants needed for external link processing
48 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
49 define( 'HTTP_PROTOCOLS', 'http|https' );
50 # Everything except bracket, space, or control characters
51 define( 'EXT_LINK_URL_CLASS', '[^]<>"\\x00-\\x20\\x7F]' );
53 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
54 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
55 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
56 define( 'EXT_LINK_BRACKETED', '/\[(\b('.URL_PROTOCOLS
.'):'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
57 define( 'EXT_IMAGE_REGEX',
58 '/^('.HTTP_PROTOCOLS
.':)'. # Protocol
59 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
60 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
66 * Processes wiki markup
69 * There are three main entry points into the Parser class:
71 * produces HTML output
73 * produces altered wiki markup.
75 * performs brace substitution on MediaWiki messages
78 * objects: $wgLang, $wgDateFormatter, $wgLinkCache
80 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
83 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
84 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
87 * * only within ParserOptions
100 # Cleared with clearState():
101 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
102 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
105 var $mOptions, $mTitle, $mOutputType,
106 $mTemplates, // cache of already loaded templates, avoids
107 // multiple SQL queries for the same string
108 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
109 // in this path. Used for loop detection.
119 $this->mTemplates
= array();
120 $this->mTemplatePath
= array();
121 $this->mTagHooks
= array();
130 function clearState() {
131 $this->mOutput
= new ParserOutput
;
132 $this->mAutonumber
= 0;
133 $this->mLastSection
= '';
134 $this->mDTopen
= false;
135 $this->mVariables
= false;
136 $this->mIncludeCount
= array();
137 $this->mStripState
= array();
138 $this->mArgStack
= array();
139 $this->mInPre
= false;
143 * First pass--just handle <nowiki> sections, pass the rest off
144 * to internalParse() which does all the real work.
147 * @param string $text Text we want to parse
148 * @param Title &$title A title object
149 * @param array $options
150 * @param boolean $linestart
151 * @param boolean $clearState
152 * @return ParserOutput a ParserOutput
154 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
155 global $wgUseTidy, $wgContLang;
156 $fname = 'Parser::parse';
157 wfProfileIn( $fname );
163 $this->mOptions
= $options;
164 $this->mTitle
=& $title;
165 $this->mOutputType
= OT_HTML
;
168 global $fnord; $fnord = 1;
169 //$text = $this->strip( $text, $this->mStripState );
170 // VOODOO MAGIC FIX! Sometimes the above segfaults in PHP5.
171 $x =& $this->mStripState
;
172 $text = $this->strip( $text, $x );
174 $text = $this->internalParse( $text, $linestart );
176 $dashReplace = array(
177 '/ - /' => " – ", # N dash
178 '/(?<=[0-9])-(?=[0-9])/' => "–", # N dash between numbers
179 '/ -- /' => " — " # M dash
181 $text = preg_replace( array_keys($dashReplace), array_values($dashReplace), $text );
184 $text = $this->unstrip( $text, $this->mStripState
);
185 # Clean up special characters, only run once, next-to-last before doBlockLevels
189 # french spaces, last one Guillemet-left
190 # only if there is something before the space
191 '/(.) (?=\\?|:|;|!|\\302\\273)/' => '\\1 \\2',
192 # french spaces, Guillemet-right
193 '/(\\302\\253) /' => '\\1 ',
194 '/<hr *>/i' => '<hr />',
195 '/<br *>/i' => '<br />',
196 '/<center *>/i' => '<div class="center">',
197 '/<\\/center *>/i' => '</div>',
199 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
200 $text = Sanitizer
::normalizeCharReferences( $text );
203 # french spaces, last one Guillemet-left
204 '/ (\\?|:|;|!|\\302\\273)/' => ' \\1',
205 # french spaces, Guillemet-right
206 '/(\\302\\253) /' => '\\1 ',
207 '/<center *>/i' => '<div class="center">',
208 '/<\\/center *>/i' => '</div>'
210 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
213 $text = $this->doBlockLevels( $text, $linestart );
215 $this->replaceLinkHolders( $text );
216 $text = $wgContLang->convert($text);
218 $text = $this->unstripNoWiki( $text, $this->mStripState
);
220 $text = Parser
::tidy($text);
223 $this->mOutput
->setText( $text );
224 wfProfileOut( $fname );
225 return $this->mOutput
;
229 * Get a random string
234 function getRandomString() {
235 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
239 * Replaces all occurrences of <$tag>content</$tag> in the text
240 * with a random marker and returns the new text. the output parameter
241 * $content will be an associative array filled with data on the form
242 * $unique_marker => content.
244 * If $content is already set, the additional entries will be appended
245 * If $tag is set to STRIP_COMMENTS, the function will extract
246 * <!-- HTML comments -->
251 function extractTags($tag, $text, &$content, $uniq_prefix = ''){
252 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
259 while ( '' != $text ) {
260 if($tag==STRIP_COMMENTS
) {
261 $p = preg_split( '/<!--/', $text, 2 );
263 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
266 if ( ( count( $p ) < 2 ) ||
( '' == $p[1] ) ) {
269 if($tag==STRIP_COMMENTS
) {
270 $q = preg_split( '/-->/i', $p[1], 2 );
272 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
274 $marker = $rnd . sprintf('%08X', $n++
);
275 $content[$marker] = $q[0];
276 $stripped .= $marker;
284 * Strips and renders nowiki, pre, math, hiero
285 * If $render is set, performs necessary rendering operations on plugins
286 * Returns the text, and fills an array with data needed in unstrip()
287 * If the $state is already a valid strip state, it adds to the state
289 * @param bool $stripcomments when set, HTML comments <!-- like this -->
290 * will be stripped in addition to other tags. This is important
291 * for section editing, where these comments cause confusion when
292 * counting the sections in the wikisource
296 function strip( $text, &$state, $stripcomments = false ) {
297 $render = ($this->mOutputType
== OT_HTML
);
298 $html_content = array();
299 $nowiki_content = array();
300 $math_content = array();
301 $pre_content = array();
302 $comment_content = array();
303 $ext_content = array();
304 $gallery_content = array();
306 # Replace any instances of the placeholders
307 $uniq_prefix = UNIQ_PREFIX
;
308 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
311 global $wgRawHtml, $wgWhitelistEdit;
312 if( $wgRawHtml && $wgWhitelistEdit ) {
313 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
314 foreach( $html_content as $marker => $content ) {
316 # Raw and unchecked for validity.
317 $html_content[$marker] = $content;
319 $html_content[$marker] = '<html>'.$content.'</html>';
325 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
326 foreach( $nowiki_content as $marker => $content ) {
328 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
330 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
335 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
336 foreach( $math_content as $marker => $content ){
338 if( $this->mOptions
->getUseTeX() ) {
339 $math_content[$marker] = renderMath( $content );
341 $math_content[$marker] = '<math>'.$content.'<math>';
344 $math_content[$marker] = '<math>'.$content.'</math>';
349 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
350 foreach( $pre_content as $marker => $content ){
352 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
354 $pre_content[$marker] = '<pre>'.$content.'</pre>';
359 $text = Parser
::extractTags('gallery', $text, $gallery_content, $uniq_prefix);
360 foreach( $gallery_content as $marker => $content ) {
361 require_once( 'ImageGallery.php' );
363 $gallery_content[$marker] = Parser
::renderImageGallery( $content );
365 $gallery_content[$marker] = '<gallery>'.$content.'</gallery>';
371 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
372 foreach( $comment_content as $marker => $content ){
373 $comment_content[$marker] = '<!--'.$content.'-->';
378 foreach ( $this->mTagHooks
as $tag => $callback ) {
379 $ext_contents[$tag] = array();
380 $text = Parser
::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
381 foreach( $ext_content[$tag] as $marker => $content ) {
383 $ext_content[$tag][$marker] = $callback( $content );
385 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
390 # Merge state with the pre-existing state, if there is one
392 $state['html'] = $state['html'] +
$html_content;
393 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
394 $state['math'] = $state['math'] +
$math_content;
395 $state['pre'] = $state['pre'] +
$pre_content;
396 $state['comment'] = $state['comment'] +
$comment_content;
397 $state['gallery'] = $state['gallery'] +
$gallery_content;
399 foreach( $ext_content as $tag => $array ) {
400 if ( array_key_exists( $tag, $state ) ) {
401 $state[$tag] = $state[$tag] +
$array;
406 'html' => $html_content,
407 'nowiki' => $nowiki_content,
408 'math' => $math_content,
409 'pre' => $pre_content,
410 'comment' => $comment_content,
411 'gallery' => $gallery_content,
418 * restores pre, math, and hiero removed by strip()
420 * always call unstripNoWiki() after this one
423 function unstrip( $text, &$state ) {
424 # Must expand in reverse order, otherwise nested tags will be corrupted
425 $contentDict = end( $state );
426 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
427 if( key($state) != 'nowiki' && key($state) != 'html') {
428 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
429 $text = str_replace( key( $contentDict ), $content, $text );
438 * always call this after unstrip() to preserve the order
442 function unstripNoWiki( $text, &$state ) {
443 # Must expand in reverse order, otherwise nested tags will be corrupted
444 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
445 $text = str_replace( key( $state['nowiki'] ), $content, $text );
450 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
451 $text = str_replace( key( $state['html'] ), $content, $text );
459 * Add an item to the strip state
460 * Returns the unique tag which must be inserted into the stripped text
461 * The tag will be replaced with the original text in unstrip()
465 function insertStripItem( $text, &$state ) {
466 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
475 $state['item'][$rnd] = $text;
480 * interface with html tidy, used if $wgUseTidy = true
485 function tidy ( $text ) {
486 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
487 $fname = 'Parser::tidy';
488 wfProfileIn( $fname );
493 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
494 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
495 '<head><title>test</title></head><body>'.$text.'</body></html>';
496 $descriptorspec = array(
497 0 => array('pipe', 'r'),
498 1 => array('pipe', 'w'),
499 2 => array('file', '/dev/null', 'a')
501 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
502 if (is_resource($process)) {
503 fwrite($pipes[0], $wrappedtext);
505 while (!feof($pipes[1])) {
506 $cleansource .= fgets($pipes[1], 1024);
509 $return_value = proc_close($process);
512 wfProfileOut( $fname );
514 if( $cleansource == '' && $text != '') {
515 wfDebug( "Tidy error detected!\n" );
516 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
523 * parse the wiki syntax used to render tables
527 function doTableStuff ( $t ) {
528 $fname = 'Parser::doTableStuff';
529 wfProfileIn( $fname );
531 $t = explode ( "\n" , $t ) ;
532 $td = array () ; # Is currently a td tag open?
533 $ltd = array () ; # Was it TD or TH?
534 $tr = array () ; # Is currently a tr tag open?
535 $ltr = array () ; # tr attributes
536 $indent_level = 0; # indent level of the table
537 foreach ( $t AS $k => $x )
540 $fc = substr ( $x , 0 , 1 ) ;
541 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
542 $indent_level = strlen( $matches[1] );
544 str_repeat( '<dl><dd>', $indent_level ) .
545 '<table' . Sanitizer
::fixTagAttributes ( $matches[2], 'table' ) . '>' ;
546 array_push ( $td , false ) ;
547 array_push ( $ltd , '' ) ;
548 array_push ( $tr , false ) ;
549 array_push ( $ltr , '' ) ;
551 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
552 else if ( '|}' == substr ( $x , 0 , 2 ) ) {
553 $z = "</table>" . substr ( $x , 2) . "\n";
554 $l = array_pop ( $ltd ) ;
555 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
556 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
558 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
560 else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
561 $x = substr ( $x , 1 ) ;
562 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
564 $l = array_pop ( $ltd ) ;
565 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
566 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
569 array_push ( $tr , false ) ;
570 array_push ( $td , false ) ;
571 array_push ( $ltd , '' ) ;
572 array_push ( $ltr , Sanitizer
::fixTagAttributes ( $x, 'tr' ) ) ;
574 else if ( '|' == $fc ||
'!' == $fc ||
'|+' == substr ( $x , 0 , 2 ) ) { # Caption
576 if ( '|+' == substr ( $x , 0 , 2 ) ) {
578 $x = substr ( $x , 1 ) ;
580 $after = substr ( $x , 1 ) ;
581 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
582 $after = explode ( '||' , $after ) ;
585 # Loop through each table cell
586 foreach ( $after AS $theline )
591 $tra = array_pop ( $ltr ) ;
592 if ( !array_pop ( $tr ) ) $z = '<tr'.$tra.">\n" ;
593 array_push ( $tr , true ) ;
594 array_push ( $ltr , '' ) ;
597 $l = array_pop ( $ltd ) ;
598 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
599 if ( $fc == '|' ) $l = 'td' ;
600 else if ( $fc == '!' ) $l = 'th' ;
601 else if ( $fc == '+' ) $l = 'caption' ;
603 array_push ( $ltd , $l ) ;
606 $y = explode ( '|' , $theline , 2 ) ;
607 # Note that a '|' inside an invalid link should not
608 # be mistaken as delimiting cell parameters
609 if ( strpos( $y[0], '[[' ) !== false ) {
610 $y = array ($theline);
612 if ( count ( $y ) == 1 )
613 $y = "{$z}<{$l}>{$y[0]}" ;
614 else $y = $y = "{$z}<{$l}".Sanitizer
::fixTagAttributes($y[0], $l).">{$y[1]}" ;
616 array_push ( $td , true ) ;
621 # Closing open td, tr && table
622 while ( count ( $td ) > 0 )
624 if ( array_pop ( $td ) ) $t[] = '</td>' ;
625 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
629 $t = implode ( "\n" , $t ) ;
630 # $t = Sanitizer::removeHTMLtags( $t );
631 wfProfileOut( $fname );
636 * Helper function for parse() that transforms wiki markup into
637 * HTML. Only called for $mOutputType == OT_HTML.
641 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
644 $fname = 'Parser::internalParse';
645 wfProfileIn( $fname );
647 $text = Sanitizer
::removeHTMLtags( $text );
648 $text = $this->replaceVariables( $text, $args );
650 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
652 $text = $this->doHeadings( $text );
653 if($this->mOptions
->getUseDynamicDates()) {
654 global $wgDateFormatter;
655 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
657 $text = $this->doAllQuotes( $text );
658 $text = $this->replaceInternalLinks( $text );
659 $text = $this->replaceExternalLinks( $text );
661 # replaceInternalLinks may sometimes leave behind
662 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
663 $text = str_replace("http-noparse://","http://",$text);
665 $text = $this->doMagicLinks( $text );
666 $text = $this->doTableStuff( $text );
667 $text = $this->formatHeadings( $text, $isMain );
669 wfProfileOut( $fname );
674 * Replace special strings like "ISBN xxx" and "RFC xxx" with
675 * magic external links.
679 function &doMagicLinks( &$text ) {
680 $text = $this->magicISBN( $text );
681 $text = $this->magicRFC( $text, 'RFC ', 'rfcurl' );
682 $text = $this->magicRFC( $text, 'PMID ', 'pubmedurl' );
687 * Parse ^^ tokens and return html
691 function doExponent( $text ) {
692 $fname = 'Parser::doExponent';
693 wfProfileIn( $fname );
694 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
695 wfProfileOut( $fname );
700 * Parse headers and return html
704 function doHeadings( $text ) {
705 $fname = 'Parser::doHeadings';
706 wfProfileIn( $fname );
707 for ( $i = 6; $i >= 1; --$i ) {
708 $h = substr( '======', 0, $i );
709 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
710 "<h{$i}>\\1</h{$i}>\\2", $text );
712 wfProfileOut( $fname );
717 * Replace single quotes with HTML markup
719 * @return string the altered text
721 function doAllQuotes( $text ) {
722 $fname = 'Parser::doAllQuotes';
723 wfProfileIn( $fname );
725 $lines = explode( "\n", $text );
726 foreach ( $lines as $line ) {
727 $outtext .= $this->doQuotes ( $line ) . "\n";
729 $outtext = substr($outtext, 0,-1);
730 wfProfileOut( $fname );
735 * Helper function for doAllQuotes()
738 function doQuotes( $text ) {
739 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
740 if ( count( $arr ) == 1 )
744 # First, do some preliminary work. This may shift some apostrophes from
745 # being mark-up to being text. It also counts the number of occurrences
746 # of bold and italics mark-ups.
750 foreach ( $arr as $r )
752 if ( ( $i %
2 ) == 1 )
754 # If there are ever four apostrophes, assume the first is supposed to
755 # be text, and the remaining three constitute mark-up for bold text.
756 if ( strlen( $arr[$i] ) == 4 )
761 # If there are more than 5 apostrophes in a row, assume they're all
762 # text except for the last 5.
763 else if ( strlen( $arr[$i] ) > 5 )
765 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
768 # Count the number of occurrences of bold and italics mark-ups.
769 # We are not counting sequences of five apostrophes.
770 if ( strlen( $arr[$i] ) == 2 ) $numitalics++
; else
771 if ( strlen( $arr[$i] ) == 3 ) $numbold++
; else
772 if ( strlen( $arr[$i] ) == 5 ) { $numitalics++
; $numbold++
; }
777 # If there is an odd number of both bold and italics, it is likely
778 # that one of the bold ones was meant to be an apostrophe followed
779 # by italics. Which one we cannot know for certain, but it is more
780 # likely to be one that has a single-letter word before it.
781 if ( ( $numbold %
2 == 1 ) && ( $numitalics %
2 == 1 ) )
784 $firstsingleletterword = -1;
785 $firstmultiletterword = -1;
787 foreach ( $arr as $r )
789 if ( ( $i %
2 == 1 ) and ( strlen( $r ) == 3 ) )
791 $x1 = substr ($arr[$i-1], -1);
792 $x2 = substr ($arr[$i-1], -2, 1);
794 if ($firstspace == -1) $firstspace = $i;
795 } else if ($x2 == ' ') {
796 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
798 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
804 # If there is a single-letter word, use it!
805 if ($firstsingleletterword > -1)
807 $arr [ $firstsingleletterword ] = "''";
808 $arr [ $firstsingleletterword-1 ] .= "'";
810 # If not, but there's a multi-letter word, use that one.
811 else if ($firstmultiletterword > -1)
813 $arr [ $firstmultiletterword ] = "''";
814 $arr [ $firstmultiletterword-1 ] .= "'";
816 # ... otherwise use the first one that has neither.
817 # (notice that it is possible for all three to be -1 if, for example,
818 # there is only one pentuple-apostrophe in the line)
819 else if ($firstspace > -1)
821 $arr [ $firstspace ] = "''";
822 $arr [ $firstspace-1 ] .= "'";
826 # Now let's actually convert our apostrophic mush to HTML!
835 if ($state == 'both')
842 if (strlen ($r) == 2)
845 { $output .= '</i>'; $state = ''; }
846 else if ($state == 'bi')
847 { $output .= '</i>'; $state = 'b'; }
848 else if ($state == 'ib')
849 { $output .= '</b></i><b>'; $state = 'b'; }
850 else if ($state == 'both')
851 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
852 else # $state can be 'b' or ''
853 { $output .= '<i>'; $state .= 'i'; }
855 else if (strlen ($r) == 3)
858 { $output .= '</b>'; $state = ''; }
859 else if ($state == 'bi')
860 { $output .= '</i></b><i>'; $state = 'i'; }
861 else if ($state == 'ib')
862 { $output .= '</b>'; $state = 'i'; }
863 else if ($state == 'both')
864 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
865 else # $state can be 'i' or ''
866 { $output .= '<b>'; $state .= 'b'; }
868 else if (strlen ($r) == 5)
871 { $output .= '</b><i>'; $state = 'i'; }
872 else if ($state == 'i')
873 { $output .= '</i><b>'; $state = 'b'; }
874 else if ($state == 'bi')
875 { $output .= '</i></b>'; $state = ''; }
876 else if ($state == 'ib')
877 { $output .= '</b></i>'; $state = ''; }
878 else if ($state == 'both')
879 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
880 else # ($state == '')
881 { $buffer = ''; $state = 'both'; }
886 # Now close all remaining tags. Notice that the order is important.
887 if ($state == 'b' ||
$state == 'ib')
889 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
893 if ($state == 'both')
894 $output .= '<b><i>'.$buffer.'</i></b>';
900 * Replace external links
902 * Note: this is all very hackish and the order of execution matters a lot.
903 * Make sure to run maintenance/parserTests.php if you change this code.
907 function replaceExternalLinks( $text ) {
908 $fname = 'Parser::replaceExternalLinks';
909 wfProfileIn( $fname );
911 $sk =& $this->mOptions
->getSkin();
913 $linktrail = $wgContLang->linkTrail();
915 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
917 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
920 while ( $i<count( $bits ) ) {
922 $protocol = $bits[$i++
];
924 $trail = $bits[$i++
];
926 # The characters '<' and '>' (which were escaped by
927 # removeHTMLtags()) should not be included in
928 # URLs, per RFC 2396.
929 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
930 $text = substr($url, $m2[0][1]) . ' ' . $text;
931 $url = substr($url, 0, $m2[0][1]);
934 # If the link text is an image URL, replace it with an <img> tag
935 # This happened by accident in the original parser, but some people used it extensively
936 $img = $this->maybeMakeImageLink( $text );
937 if ( $img !== false ) {
943 # Set linktype for CSS - if URL==text, link is essentially free
944 $linktype = ($text == $url) ?
'free' : 'text';
946 # No link text, e.g. [http://domain.tld/some.link]
948 # Autonumber if allowed
949 if ( strpos( HTTP_PROTOCOLS
, $protocol ) !== false ) {
950 $text = '[' . ++
$this->mAutonumber
. ']';
951 $linktype = 'autonumber';
953 # Otherwise just use the URL
954 $text = htmlspecialchars( $url );
958 # Have link text, e.g. [http://domain.tld/some.link text]s
960 if ( preg_match( $linktrail, $trail, $m2 ) ) {
966 # Replace & from obsolete syntax with &.
967 # All HTML entities will be escaped by makeExternalLink()
968 # or maybeMakeImageLink()
969 $url = str_replace( '&', '&', $url );
971 # Process the trail (i.e. everything after this link up until start of the next link),
972 # replacing any non-bracketed links
973 $trail = $this->replaceFreeExternalLinks( $trail );
975 # Use the encoded URL
976 # This means that users can paste URLs directly into the text
977 # Funny characters like ö aren't valid in URLs anyway
978 # This was changed in August 2004
979 $s .= $sk->makeExternalLink( $url, $text, false, $linktype ) . $dtrail . $trail;
982 wfProfileOut( $fname );
987 * Replace anything that looks like a URL with a link
990 function replaceFreeExternalLinks( $text ) {
991 $fname = 'Parser::replaceFreeExternalLinks';
992 wfProfileIn( $fname );
994 $bits = preg_split( '/(\b(?:'.URL_PROTOCOLS
.'):)/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
995 $s = array_shift( $bits );
998 $sk =& $this->mOptions
->getSkin();
1000 while ( $i < count( $bits ) ){
1001 $protocol = $bits[$i++
];
1002 $remainder = $bits[$i++
];
1004 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1005 # Found some characters after the protocol that look promising
1006 $url = $protocol . $m[1];
1009 # The characters '<' and '>' (which were escaped by
1010 # removeHTMLtags()) should not be included in
1011 # URLs, per RFC 2396.
1012 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1013 $trail = substr($url, $m2[0][1]) . $trail;
1014 $url = substr($url, 0, $m2[0][1]);
1017 # Move trailing punctuation to $trail
1019 # If there is no left bracket, then consider right brackets fair game too
1020 if ( strpos( $url, '(' ) === false ) {
1024 $numSepChars = strspn( strrev( $url ), $sep );
1025 if ( $numSepChars ) {
1026 $trail = substr( $url, -$numSepChars ) . $trail;
1027 $url = substr( $url, 0, -$numSepChars );
1030 # Replace & from obsolete syntax with &.
1031 # All HTML entities will be escaped by makeExternalLink()
1032 # or maybeMakeImageLink()
1033 $url = str_replace( '&', '&', $url );
1035 # Is this an external image?
1036 $text = $this->maybeMakeImageLink( $url );
1037 if ( $text === false ) {
1038 # Not an image, make a link
1039 $text = $sk->makeExternalLink( $url, $url, true, 'free' );
1041 $s .= $text . $trail;
1043 $s .= $protocol . $remainder;
1051 * make an image if it's allowed
1054 function maybeMakeImageLink( $url ) {
1055 $sk =& $this->mOptions
->getSkin();
1057 if ( $this->mOptions
->getAllowExternalImages() ) {
1058 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1060 $text = $sk->makeImage( htmlspecialchars( $url ) );
1067 * Process [[ ]] wikilinks
1071 function replaceInternalLinks( $s ) {
1072 global $wgLang, $wgContLang, $wgLinkCache;
1073 global $wgDisableLangConversion;
1074 static $fname = 'Parser::replaceInternalLinks' ;
1076 wfProfileIn( $fname );
1078 wfProfileIn( $fname.'-setup' );
1080 # the % is needed to support urlencoded titles as well
1081 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1083 $sk =& $this->mOptions
->getSkin();
1084 global $wgUseOldExistenceCheck;
1085 # "Post-parse link colour check" works only on wiki text since it's now
1086 # in Parser. Enable it, then disable it when we're done.
1087 $saveParseColour = $sk->postParseLinkColour( !$wgUseOldExistenceCheck );
1089 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1091 #split the entire text string on occurences of [[
1092 $a = explode( '[[', ' ' . $s );
1093 #get the first element (all text up to first [[), and remove the space we added
1094 $s = array_shift( $a );
1095 $s = substr( $s, 1 );
1097 # Match a link having the form [[namespace:link|alternate]]trail
1099 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
1100 # Match cases where there is no "]]", which might still be images
1101 static $e1_img = FALSE;
1102 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1103 # Match the end of a line for a word that's not followed by whitespace,
1104 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1105 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1107 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1109 if( is_null( $this->mTitle
) ) {
1110 wfDebugDieBacktrace( 'nooo' );
1112 $nottalk = !$this->mTitle
->isTalkPage();
1114 if ( $useLinkPrefixExtension ) {
1115 if ( preg_match( $e2, $s, $m ) ) {
1116 $first_prefix = $m[2];
1119 $first_prefix = false;
1125 $selflink = $this->mTitle
->getPrefixedText();
1126 wfProfileOut( $fname.'-setup' );
1128 $checkVariantLink = sizeof($wgContLang->getVariants())>1;
1129 $useSubpages = $this->areSubpagesAllowed();
1131 # Loop for each link
1132 for ($k = 0; isset( $a[$k] ); $k++
) {
1134 if ( $useLinkPrefixExtension ) {
1135 wfProfileIn( $fname.'-prefixhandling' );
1136 if ( preg_match( $e2, $s, $m ) ) {
1144 $prefix = $first_prefix;
1145 $first_prefix = false;
1147 wfProfileOut( $fname.'-prefixhandling' );
1150 $might_be_img = false;
1152 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1154 # fix up urlencoded title texts
1155 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1157 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1158 $might_be_img = true;
1160 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1162 } else { # Invalid form; output directly
1163 $s .= $prefix . '[[' . $line ;
1167 # Don't allow internal links to pages containing
1168 # PROTO: where PROTO is a valid URL protocol; these
1169 # should be external links.
1170 if (preg_match('/^(\b(?:'.URL_PROTOCOLS
.'):)/', $m[1])) {
1171 $s .= $prefix . '[[' . $line ;
1175 # Make subpage if necessary
1176 if( $useSubpages ) {
1177 $link = $this->maybeDoSubpageLink( $m[1], $text );
1182 $noforce = (substr($m[1], 0, 1) != ':');
1184 # Strip off leading ':'
1185 $link = substr($link, 1);
1188 $nt =& Title
::newFromText( $this->unstripNoWiki($link, $this->mStripState
) );
1190 $s .= $prefix . '[[' . $line;
1194 #check other language variants of the link
1195 #if the article does not exist
1196 if( $checkVariantLink
1197 && $nt->getArticleID() == 0 ) {
1198 $wgContLang->findVariantLink($link, $nt);
1201 $ns = $nt->getNamespace();
1202 $iw = $nt->getInterWiki();
1204 if ($might_be_img) { # if this is actually an invalid link
1205 if ($ns == NS_IMAGE
&& $noforce) { #but might be an image
1207 while (isset ($a[$k+
1]) ) {
1208 #look at the next 'line' to see if we can close it there
1209 $next_line = array_shift(array_splice( $a, $k +
1, 1) );
1210 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1211 # the first ]] closes the inner link, the second the image
1213 $text .= '[[' . $m[1];
1216 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1217 #if there's exactly one ]] that's fine, we'll keep looking
1218 $text .= '[[' . $m[0];
1220 #if $next_line is invalid too, we need look no further
1221 $text .= '[[' . $next_line;
1226 # we couldn't find the end of this imageLink, so output it raw
1227 #but don't ignore what might be perfectly normal links in the text we've examined
1228 $text = $this->replaceInternalLinks($text);
1229 $s .= $prefix . '[[' . $link . '|' . $text;
1230 # note: no $trail, because without an end, there *is* no trail
1233 } else { #it's not an image, so output it raw
1234 $s .= $prefix . '[[' . $link . '|' . $text;
1235 # note: no $trail, because without an end, there *is* no trail
1240 $wasblank = ( '' == $text );
1241 if( $wasblank ) $text = $link;
1244 # Link not escaped by : , create the various objects
1248 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1249 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1250 $tmp = $prefix . $trail ;
1251 $s .= (trim($tmp) == '')?
'': $tmp;
1255 if ( $ns == NS_IMAGE
) {
1256 wfProfileIn( "$fname-image" );
1258 # recursively parse links inside the image caption
1259 # actually, this will parse them in any other parameters, too,
1260 # but it might be hard to fix that, and it doesn't matter ATM
1261 $text = $this->replaceExternalLinks($text);
1262 $text = $this->replaceInternalLinks($text);
1264 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1265 $s .= $prefix . str_replace('http://', 'http-noparse://', $sk->makeImageLinkObj( $nt, $text ) ) . $trail;
1266 $wgLinkCache->addImageLinkObj( $nt );
1268 wfProfileOut( "$fname-image" );
1272 if ( $ns == NS_CATEGORY
) {
1273 wfProfileIn( "$fname-category" );
1274 $t = $nt->getText();
1276 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1277 $pPLC=$sk->postParseLinkColour();
1278 $sk->postParseLinkColour( false );
1279 $t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
1280 $sk->postParseLinkColour( $pPLC );
1281 $wgLinkCache->resume();
1284 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1285 $sortkey = $this->mTitle
->getText();
1287 $sortkey = $this->mTitle
->getPrefixedText();
1292 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1293 $this->mOutput
->addCategoryLink( $t );
1294 $s .= $prefix . $trail ;
1296 wfProfileOut( "$fname-category" );
1301 if( ( $nt->getPrefixedText() === $selflink ) &&
1302 ( $nt->getFragment() === '' ) ) {
1303 # Self-links are handled specially; generally de-link and change to bold.
1304 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1308 # Special and Media are pseudo-namespaces; no pages actually exist in them
1309 if( $ns == NS_MEDIA
) {
1310 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text, true ) . $trail;
1311 $wgLinkCache->addImageLinkObj( $nt );
1313 } elseif( $ns == NS_SPECIAL
) {
1314 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1317 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1319 $sk->postParseLinkColour( $saveParseColour );
1320 wfProfileOut( $fname );
1325 * Return true if subpage links should be expanded on this page.
1328 function areSubpagesAllowed() {
1329 # Some namespaces don't allow subpages
1330 global $wgNamespacesWithSubpages;
1331 return !empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]);
1335 * Handle link to subpage if necessary
1336 * @param string $target the source of the link
1337 * @param string &$text the link text, modified as necessary
1338 * @return string the full name of the link
1341 function maybeDoSubpageLink($target, &$text) {
1344 # :Foobar -- override special treatment of prefix (images, language links)
1345 # /Foobar -- convert to CurrentPage/Foobar
1346 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1347 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1348 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1350 $fname = 'Parser::maybeDoSubpageLink';
1351 wfProfileIn( $fname );
1352 $ret = $target; # default return value is no change
1354 # Some namespaces don't allow subpages,
1355 # so only perform processing if subpages are allowed
1356 if( $this->areSubpagesAllowed() ) {
1357 # Look at the first character
1358 if( $target != '' && $target{0} == '/' ) {
1359 # / at end means we don't want the slash to be shown
1360 if( substr( $target, -1, 1 ) == '/' ) {
1361 $target = substr( $target, 1, -1 );
1364 $noslash = substr( $target, 1 );
1367 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1368 if( '' === $text ) {
1370 } # this might be changed for ugliness reasons
1372 # check for .. subpage backlinks
1374 $nodotdot = $target;
1375 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1377 $nodotdot = substr( $nodotdot, 3 );
1379 if($dotdotcount > 0) {
1380 $exploded = explode( '/', $this->mTitle
->GetPrefixedText() );
1381 if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1382 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1383 # / at the end means don't show full path
1384 if( substr( $nodotdot, -1, 1 ) == '/' ) {
1385 $nodotdot = substr( $nodotdot, 0, -1 );
1386 if( '' === $text ) {
1390 $nodotdot = trim( $nodotdot );
1391 if( $nodotdot != '' ) {
1392 $ret .= '/' . $nodotdot;
1399 wfProfileOut( $fname );
1404 * Used by doBlockLevels()
1407 /* private */ function closeParagraph() {
1409 if ( '' != $this->mLastSection
) {
1410 $result = '</' . $this->mLastSection
. ">\n";
1412 $this->mInPre
= false;
1413 $this->mLastSection
= '';
1416 # getCommon() returns the length of the longest common substring
1417 # of both arguments, starting at the beginning of both.
1419 /* private */ function getCommon( $st1, $st2 ) {
1420 $fl = strlen( $st1 );
1421 $shorter = strlen( $st2 );
1422 if ( $fl < $shorter ) { $shorter = $fl; }
1424 for ( $i = 0; $i < $shorter; ++
$i ) {
1425 if ( $st1{$i} != $st2{$i} ) { break; }
1429 # These next three functions open, continue, and close the list
1430 # element appropriate to the prefix character passed into them.
1432 /* private */ function openList( $char ) {
1433 $result = $this->closeParagraph();
1435 if ( '*' == $char ) { $result .= '<ul><li>'; }
1436 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1437 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1438 else if ( ';' == $char ) {
1439 $result .= '<dl><dt>';
1440 $this->mDTopen
= true;
1442 else { $result = '<!-- ERR 1 -->'; }
1447 /* private */ function nextItem( $char ) {
1448 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1449 else if ( ':' == $char ||
';' == $char ) {
1451 if ( $this->mDTopen
) { $close = '</dt>'; }
1452 if ( ';' == $char ) {
1453 $this->mDTopen
= true;
1454 return $close . '<dt>';
1456 $this->mDTopen
= false;
1457 return $close . '<dd>';
1460 return '<!-- ERR 2 -->';
1463 /* private */ function closeList( $char ) {
1464 if ( '*' == $char ) { $text = '</li></ul>'; }
1465 else if ( '#' == $char ) { $text = '</li></ol>'; }
1466 else if ( ':' == $char ) {
1467 if ( $this->mDTopen
) {
1468 $this->mDTopen
= false;
1469 $text = '</dt></dl>';
1471 $text = '</dd></dl>';
1474 else { return '<!-- ERR 3 -->'; }
1480 * Make lists from lines starting with ':', '*', '#', etc.
1483 * @return string the lists rendered as HTML
1485 function doBlockLevels( $text, $linestart ) {
1486 $fname = 'Parser::doBlockLevels';
1487 wfProfileIn( $fname );
1489 # Parsing through the text line by line. The main thing
1490 # happening here is handling of block-level elements p, pre,
1491 # and making lists from lines starting with * # : etc.
1493 $textLines = explode( "\n", $text );
1495 $lastPrefix = $output = $lastLine = '';
1496 $this->mDTopen
= $inBlockElem = false;
1498 $paragraphStack = false;
1500 if ( !$linestart ) {
1501 $output .= array_shift( $textLines );
1503 foreach ( $textLines as $oLine ) {
1504 $lastPrefixLength = strlen( $lastPrefix );
1505 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1506 $preOpenMatch = preg_match('/<pre/i', $oLine );
1507 if ( !$this->mInPre
) {
1508 # Multiple prefixes may abut each other for nested lists.
1509 $prefixLength = strspn( $oLine, '*#:;' );
1510 $pref = substr( $oLine, 0, $prefixLength );
1513 $pref2 = str_replace( ';', ':', $pref );
1514 $t = substr( $oLine, $prefixLength );
1515 $this->mInPre
= !empty($preOpenMatch);
1517 # Don't interpret any other prefixes in preformatted text
1519 $pref = $pref2 = '';
1524 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1525 # Same as the last item, so no need to deal with nesting or opening stuff
1526 $output .= $this->nextItem( substr( $pref, -1 ) );
1527 $paragraphStack = false;
1529 if ( substr( $pref, -1 ) == ';') {
1530 # The one nasty exception: definition lists work like this:
1531 # ; title : definition text
1532 # So we check for : in the remainder text to split up the
1533 # title and definition, without b0rking links.
1534 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1536 $output .= $term . $this->nextItem( ':' );
1539 } elseif( $prefixLength ||
$lastPrefixLength ) {
1540 # Either open or close a level...
1541 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1542 $paragraphStack = false;
1544 while( $commonPrefixLength < $lastPrefixLength ) {
1545 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1546 --$lastPrefixLength;
1548 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1549 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1551 while ( $prefixLength > $commonPrefixLength ) {
1552 $char = substr( $pref, $commonPrefixLength, 1 );
1553 $output .= $this->openList( $char );
1555 if ( ';' == $char ) {
1556 # FIXME: This is dupe of code above
1557 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1559 $output .= $term . $this->nextItem( ':' );
1562 ++
$commonPrefixLength;
1564 $lastPrefix = $pref2;
1566 if( 0 == $prefixLength ) {
1567 wfProfileIn( "$fname-paragraph" );
1568 # No prefix (not in list)--go to paragraph mode
1569 $uniq_prefix = UNIQ_PREFIX
;
1570 // XXX: use a stack for nestable elements like span, table and div
1571 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
1572 $closematch = preg_match(
1573 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1574 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/iS', $t );
1575 if ( $openmatch or $closematch ) {
1576 $paragraphStack = false;
1577 $output .= $this->closeParagraph();
1578 if($preOpenMatch and !$preCloseMatch) {
1579 $this->mInPre
= true;
1581 if ( $closematch ) {
1582 $inBlockElem = false;
1584 $inBlockElem = true;
1586 } else if ( !$inBlockElem && !$this->mInPre
) {
1587 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1589 if ($this->mLastSection
!= 'pre') {
1590 $paragraphStack = false;
1591 $output .= $this->closeParagraph().'<pre>';
1592 $this->mLastSection
= 'pre';
1594 $t = substr( $t, 1 );
1597 if ( '' == trim($t) ) {
1598 if ( $paragraphStack ) {
1599 $output .= $paragraphStack.'<br />';
1600 $paragraphStack = false;
1601 $this->mLastSection
= 'p';
1603 if ($this->mLastSection
!= 'p' ) {
1604 $output .= $this->closeParagraph();
1605 $this->mLastSection
= '';
1606 $paragraphStack = '<p>';
1608 $paragraphStack = '</p><p>';
1612 if ( $paragraphStack ) {
1613 $output .= $paragraphStack;
1614 $paragraphStack = false;
1615 $this->mLastSection
= 'p';
1616 } else if ($this->mLastSection
!= 'p') {
1617 $output .= $this->closeParagraph().'<p>';
1618 $this->mLastSection
= 'p';
1623 wfProfileOut( "$fname-paragraph" );
1625 if ($paragraphStack === false) {
1629 while ( $prefixLength ) {
1630 $output .= $this->closeList( $pref2{$prefixLength-1} );
1633 if ( '' != $this->mLastSection
) {
1634 $output .= '</' . $this->mLastSection
. '>';
1635 $this->mLastSection
= '';
1638 wfProfileOut( $fname );
1643 * Split up a string on ':', ignoring any occurences inside
1644 * <a>..</a> or <span>...</span>
1645 * @param string $str the string to split
1646 * @param string &$before set to everything before the ':'
1647 * @param string &$after set to everything after the ':'
1648 * return string the position of the ':', or false if none found
1650 function findColonNoLinks($str, &$before, &$after) {
1651 # I wonder if we should make this count all tags, not just <a>
1652 # and <span>. That would prevent us from matching a ':' that
1653 # comes in the middle of italics other such formatting....
1655 $fname = 'Parser::findColonNoLinks';
1656 wfProfileIn( $fname );
1659 $colon = strpos($str, ':', $pos);
1661 if ($colon !== false) {
1662 $before = substr($str, 0, $colon);
1663 $after = substr($str, $colon +
1);
1665 # Skip any ':' within <a> or <span> pairs
1666 $a = substr_count($before, '<a');
1667 $s = substr_count($before, '<span');
1668 $ca = substr_count($before, '</a>');
1669 $cs = substr_count($before, '</span>');
1671 if ($a <= $ca and $s <= $cs) {
1672 # Tags are balanced before ':'; ok
1677 } while ($colon !== false);
1678 wfProfileOut( $fname );
1683 * Return value of a magic variable (like PAGENAME)
1687 function getVariableValue( $index ) {
1688 global $wgContLang, $wgSitename, $wgServer, $wgArticle;
1691 * Some of these require message or data lookups and can be
1692 * expensive to check many times.
1694 static $varCache = array();
1695 if( isset( $varCache[$index] ) ) return $varCache[$index];
1698 case MAG_CURRENTMONTH
:
1699 return $varCache[$index] = $wgContLang->formatNum( date( 'm' ) );
1700 case MAG_CURRENTMONTHNAME
:
1701 return $varCache[$index] = $wgContLang->getMonthName( date('n') );
1702 case MAG_CURRENTMONTHNAMEGEN
:
1703 return $varCache[$index] = $wgContLang->getMonthNameGen( date('n') );
1704 case MAG_CURRENTMONTHABBREV
:
1705 return $varCache[$index] = $wgContLang->getMonthAbbreviation( date('n') );
1706 case MAG_CURRENTDAY
:
1707 return $varCache[$index] = $wgContLang->formatNum( date('j') );
1709 return $this->mTitle
->getText();
1711 return $this->mTitle
->getPartialURL();
1712 case MAG_REVISIONID
:
1713 return $wgArticle->getRevIdFetched();
1715 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1716 return $wgContLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1717 case MAG_CURRENTDAYNAME
:
1718 return $varCache[$index] = $wgContLang->getWeekdayName( date('w')+
1 );
1719 case MAG_CURRENTYEAR
:
1720 return $varCache[$index] = $wgContLang->formatNum( date( 'Y' ) );
1721 case MAG_CURRENTTIME
:
1722 return $varCache[$index] = $wgContLang->time( wfTimestampNow(), false );
1723 case MAG_CURRENTWEEK
:
1724 return $varCache[$index] = $wgContLang->formatNum( date('W') );
1725 case MAG_CURRENTDOW
:
1726 return $varCache[$index] = $wgContLang->formatNum( date('w') );
1727 case MAG_NUMBEROFARTICLES
:
1728 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfArticles() );
1739 * initialise the magic variables (like CURRENTMONTHNAME)
1743 function initialiseVariables() {
1744 $fname = 'Parser::initialiseVariables';
1745 wfProfileIn( $fname );
1746 global $wgVariableIDs;
1747 $this->mVariables
= array();
1748 foreach ( $wgVariableIDs as $id ) {
1749 $mw =& MagicWord
::get( $id );
1750 $mw->addToArray( $this->mVariables
, $id );
1752 wfProfileOut( $fname );
1756 * Replace magic variables, templates, and template arguments
1757 * with the appropriate text. Templates are substituted recursively,
1758 * taking care to avoid infinite loops.
1760 * Note that the substitution depends on value of $mOutputType:
1761 * OT_WIKI: only {{subst:}} templates
1762 * OT_MSG: only magic variables
1763 * OT_HTML: all templates and magic variables
1765 * @param string $tex The text to transform
1766 * @param array $args Key-value pairs representing template parameters to substitute
1769 function replaceVariables( $text, $args = array() ) {
1770 global $wgLang, $wgScript, $wgArticlePath;
1772 # Prevent too big inclusions
1773 if( strlen( $text ) > MAX_INCLUDE_SIZE
) {
1777 $fname = 'Parser::replaceVariables';
1778 wfProfileIn( $fname );
1780 $titleChars = Title
::legalChars();
1782 # This function is called recursively. To keep track of arguments we need a stack:
1783 array_push( $this->mArgStack
, $args );
1785 # Variable substitution
1786 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", array( &$this, 'variableSubstitution' ), $text );
1788 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1789 # Argument substitution
1790 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", array( &$this, 'argSubstitution' ), $text );
1792 # Template substitution
1793 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1794 $text = preg_replace_callback( $regex, array( &$this, 'braceSubstitution' ), $text );
1796 array_pop( $this->mArgStack
);
1798 wfProfileOut( $fname );
1803 * Replace magic variables
1806 function variableSubstitution( $matches ) {
1807 $fname = 'parser::variableSubstitution';
1808 $varname = $matches[1];
1809 wfProfileIn( $fname );
1810 if ( !$this->mVariables
) {
1811 $this->initialiseVariables();
1814 if ( $this->mOutputType
== OT_WIKI
) {
1815 # Do only magic variables prefixed by SUBST
1816 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1817 if (!$mwSubst->matchStartAndRemove( $varname ))
1819 # Note that if we don't substitute the variable below,
1820 # we don't remove the {{subst:}} magic word, in case
1821 # it is a template rather than a magic variable.
1823 if ( !$skip && array_key_exists( $varname, $this->mVariables
) ) {
1824 $id = $this->mVariables
[$varname];
1825 $text = $this->getVariableValue( $id );
1826 $this->mOutput
->mContainsOldMagic
= true;
1828 $text = $matches[0];
1830 wfProfileOut( $fname );
1834 # Split template arguments
1835 function getTemplateArgs( $argsString ) {
1836 if ( $argsString === '' ) {
1840 $args = explode( '|', substr( $argsString, 1 ) );
1842 # If any of the arguments contains a '[[' but no ']]', it needs to be
1843 # merged with the next arg because the '|' character between belongs
1844 # to the link syntax and not the template parameter syntax.
1845 $argc = count($args);
1847 for ( $i = 0; $i < $argc-1; $i++
) {
1848 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1849 $args[$i] .= '|'.$args[$i+
1];
1850 array_splice($args, $i+
1, 1);
1860 * Return the text of a template, after recursively
1861 * replacing any variables or templates within the template.
1863 * @param array $matches The parts of the template
1864 * $matches[1]: the title, i.e. the part before the |
1865 * $matches[2]: the parameters (including a leading |), if any
1866 * @return string the text of the template
1869 function braceSubstitution( $matches ) {
1870 global $wgLinkCache, $wgContLang;
1871 $fname = 'Parser::braceSubstitution';
1872 wfProfileIn( $fname );
1880 # Need to know if the template comes at the start of a line,
1881 # to treat the beginning of the template like the beginning
1882 # of a line for tables and block-level elements.
1883 $linestart = $matches[1];
1885 # $part1 is the bit before the first |, and must contain only title characters
1886 # $args is a list of arguments, starting from index 0, not including $part1
1888 $part1 = $matches[2];
1889 # If the third subpattern matched anything, it will start with |
1891 $args = $this->getTemplateArgs($matches[3]);
1892 $argc = count( $args );
1894 # Don't parse {{{}}} because that's only for template arguments
1895 if ( $linestart === '{' ) {
1896 $text = $matches[0];
1903 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1904 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
1905 # One of two possibilities is true:
1906 # 1) Found SUBST but not in the PST phase
1907 # 2) Didn't find SUBST and in the PST phase
1908 # In either case, return without further processing
1909 $text = $matches[0];
1915 # MSG, MSGNW and INT
1918 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1919 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1922 # Remove obsolete MSG:
1923 $mwMsg =& MagicWord
::get( MAG_MSG
);
1924 $mwMsg->matchStartAndRemove( $part1 );
1927 # Check if it is an internal message
1928 $mwInt =& MagicWord
::get( MAG_INT
);
1929 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1930 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1931 $text = $linestart . wfMsgReal( $part1, $args, true );
1939 # Check for NS: (namespace expansion)
1940 $mwNs = MagicWord
::get( MAG_NS
);
1941 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1942 if ( intval( $part1 ) ) {
1943 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1946 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1947 if ( !is_null( $index ) ) {
1948 $text = $linestart . $wgContLang->getNsText( $index );
1955 # LOCALURL and LOCALURLE
1957 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1958 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1960 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1961 $func = 'getLocalURL';
1962 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1963 $func = 'escapeLocalURL';
1968 if ( $func !== '' ) {
1969 $title = Title
::newFromText( $part1 );
1970 if ( !is_null( $title ) ) {
1972 $text = $linestart . $title->$func( $args[0] );
1974 $text = $linestart . $title->$func();
1982 if ( !$found && $argc == 1 ) {
1983 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
1984 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1985 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
1990 # Template table test
1992 # Did we encounter this template already? If yes, it is in the cache
1993 # and we need to check for loops.
1994 if ( !$found && isset( $this->mTemplates
[$part1] ) ) {
1997 # Infinite loop test
1998 if ( isset( $this->mTemplatePath
[$part1] ) ) {
2001 $text = $linestart .
2003 '<!-- WARNING: template loop detected -->';
2004 wfDebug( "$fname: template loop broken at '$part1'\n" );
2006 # set $text to cached message.
2007 $text = $linestart . $this->mTemplates
[$part1];
2011 # Load from database
2012 $itcamefromthedatabase = false;
2013 $lastPathLevel = $this->mTemplatePath
;
2016 $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
2017 if ($subpage !== '') {
2018 $ns = $this->mTitle
->getNamespace();
2020 $title = Title
::newFromText( $part1, $ns );
2021 if ( !is_null( $title ) && !$title->isExternal() ) {
2022 # Check for excessive inclusion
2023 $dbk = $title->getPrefixedDBkey();
2024 if ( $this->incrementIncludeCount( $dbk ) ) {
2025 # This should never be reached.
2026 $article = new Article( $title );
2027 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
2028 if ( $articleContent !== false ) {
2030 $text = $linestart . $articleContent;
2031 $itcamefromthedatabase = true;
2035 # If the title is valid but undisplayable, make a link to it
2036 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
2037 $text = $linestart . '[['.$title->getPrefixedText().']]';
2041 # Template cache array insertion
2043 $this->mTemplates
[$part1] = $text;
2048 # Recursive parsing, escaping and link table handling
2049 # Only for HTML output
2050 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
2051 $text = wfEscapeWikiText( $text );
2052 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
2053 # Clean up argument array
2054 $assocArgs = array();
2056 foreach( $args as $arg ) {
2057 $eqpos = strpos( $arg, '=' );
2058 if ( $eqpos === false ) {
2059 $assocArgs[$index++
] = $arg;
2061 $name = trim( substr( $arg, 0, $eqpos ) );
2062 $value = trim( substr( $arg, $eqpos+
1 ) );
2063 if ( $value === false ) {
2066 if ( $name !== false ) {
2067 $assocArgs[$name] = $value;
2072 # Add a new element to the templace recursion path
2073 $this->mTemplatePath
[$part1] = 1;
2075 $text = $this->strip( $text, $this->mStripState
);
2076 $text = Sanitizer
::removeHTMLtags( $text );
2077 $text = $this->replaceVariables( $text, $assocArgs );
2079 # Resume the link cache and register the inclusion as a link
2080 if ( $this->mOutputType
== OT_HTML
&& !is_null( $title ) ) {
2081 $wgLinkCache->addLinkObj( $title );
2084 # If the template begins with a table or block-level
2085 # element, it should be treated as beginning a new line.
2086 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
2087 $text = "\n" . $text;
2090 # Prune lower levels off the recursion check path
2091 $this->mTemplatePath
= $lastPathLevel;
2094 wfProfileOut( $fname );
2097 # replace ==section headers==
2098 # XXX this needs to go away once we have a better parser.
2099 if ( $this->mOutputType
!= OT_WIKI
&& $itcamefromthedatabase ) {
2100 if( !is_null( $title ) )
2101 $encodedname = base64_encode($title->getPrefixedDBkey());
2103 $encodedname = base64_encode("");
2104 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
2105 PREG_SPLIT_DELIM_CAPTURE
);
2108 for( $i = 0; $i < count($m); $i +
= 2 ) {
2110 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
2112 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
2116 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
2117 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
2118 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
2124 # Prune lower levels off the recursion check path
2125 $this->mTemplatePath
= $lastPathLevel;
2128 wfProfileOut( $fname );
2131 wfProfileOut( $fname );
2137 * Triple brace replacement -- used for template arguments
2140 function argSubstitution( $matches ) {
2141 $arg = trim( $matches[1] );
2142 $text = $matches[0];
2143 $inputArgs = end( $this->mArgStack
);
2145 if ( array_key_exists( $arg, $inputArgs ) ) {
2146 $text = $inputArgs[$arg];
2153 * Returns true if the function is allowed to include this entity
2156 function incrementIncludeCount( $dbk ) {
2157 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
2158 $this->mIncludeCount
[$dbk] = 0;
2160 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
2168 * This function accomplishes several tasks:
2169 * 1) Auto-number headings if that option is enabled
2170 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2171 * 3) Add a Table of contents on the top for users who have enabled the option
2172 * 4) Auto-anchor headings
2174 * It loops through all headlines, collects the necessary data, then splits up the
2175 * string and re-inserts the newly formatted headlines.
2177 * @param string $text
2178 * @param boolean $isMain
2181 function formatHeadings( $text, $isMain=true ) {
2182 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders, $wgInterwikiLinkHolders;
2184 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2186 $forceTocHere = false;
2187 if( !$this->mTitle
->userCanEdit() ) {
2190 $showEditLink = $this->mOptions
->getEditSection();
2193 # Inhibit editsection links if requested in the page
2194 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2195 if( $esw->matchAndRemove( $text ) ) {
2198 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2200 $mw =& MagicWord
::get( MAG_NOTOC
);
2201 if( $mw->matchAndRemove( $text ) ) {
2205 # Get all headlines for numbering them and adding funky stuff like [edit]
2206 # links - this is for later, but we need the number of headlines right now
2207 $numMatches = preg_match_all( '/<H([1-6])(.*?'.'>)(.*?)<\/H[1-6] *>/i', $text, $matches );
2209 # if there are fewer than 4 headlines in the article, do not show TOC
2210 if( $numMatches < 4 ) {
2214 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2215 # override above conditions and always show TOC at that place
2217 $mw =& MagicWord
::get( MAG_TOC
);
2218 if($mw->match( $text ) ) {
2220 $forceTocHere = true;
2222 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2223 # override above conditions and always show TOC above first header
2224 $mw =& MagicWord
::get( MAG_FORCETOC
);
2225 if ($mw->matchAndRemove( $text ) ) {
2230 # Never ever show TOC if no headers
2231 if( $numMatches < 1 ) {
2235 # We need this to perform operations on the HTML
2236 $sk =& $this->mOptions
->getSkin();
2240 $sectionCount = 0; # headlineCount excluding template sections
2242 # Ugh .. the TOC should have neat indentation levels which can be
2243 # passed to the skin functions. These are determined here
2247 $sublevelCount = array();
2248 $levelCount = array();
2255 foreach( $matches[3] as $headline ) {
2257 $templatetitle = '';
2258 $templatesection = 0;
2261 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2263 $templatetitle = base64_decode($mat[1]);
2264 $templatesection = 1 +
(int)base64_decode($mat[2]);
2265 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2269 $prevlevel = $level;
2270 $prevtoclevel = $toclevel;
2272 $level = $matches[1][$headlineCount];
2274 if( $doNumberHeadings ||
$doShowToc ) {
2276 if ( $level > $prevlevel ) {
2277 # Increase TOC level
2279 $sublevelCount[$toclevel] = 0;
2280 $toc .= $sk->tocIndent();
2282 elseif ( $level < $prevlevel && $toclevel > 1 ) {
2283 # Decrease TOC level, find level to jump to
2285 if ( $toclevel == 2 && $level <= $levelCount[1] ) {
2286 # Can only go down to level 1
2289 for ($i = $toclevel; $i > 0; $i--) {
2290 if ( $levelCount[$i] == $level ) {
2291 # Found last matching level
2295 elseif ( $levelCount[$i] < $level ) {
2296 # Found first matching level below current level
2303 $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
2306 # No change in level, end TOC line
2307 $toc .= $sk->tocLineEnd();
2310 $levelCount[$toclevel] = $level;
2312 # count number of headlines for each level
2313 @$sublevelCount[$toclevel]++
;
2315 for( $i = 1; $i <= $toclevel; $i++
) {
2316 if( !empty( $sublevelCount[$i] ) ) {
2320 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2326 # The canonized header is a version of the header text safe to use for links
2327 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2328 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2329 $canonized_headline = $this->unstripNoWiki( $canonized_headline, $this->mStripState
);
2331 # Remove link placeholders by the link text.
2332 # <!--LINK number-->
2334 # link text with suffix
2335 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2336 "\$wgLinkHolders['texts'][\$1]",
2337 $canonized_headline );
2338 $canonized_headline = preg_replace( '/<!--IWLINK ([0-9]*)-->/e',
2339 "\$wgInterwikiLinkHolders[\$1]",
2340 $canonized_headline );
2343 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2344 $tocline = trim( $canonized_headline );
2345 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2346 $replacearray = array(
2350 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2351 $refer[$headlineCount] = $canonized_headline;
2353 # count how many in assoc. array so we can track dupes in anchors
2354 @$refers[$canonized_headline]++
;
2355 $refcount[$headlineCount]=$refers[$canonized_headline];
2357 # Don't number the heading if it is the only one (looks silly)
2358 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2359 # the two are different if the line contains a link
2360 $headline=$numbering . ' ' . $headline;
2363 # Create the anchor for linking from the TOC to the section
2364 $anchor = $canonized_headline;
2365 if($refcount[$headlineCount] > 1 ) {
2366 $anchor .= '_' . $refcount[$headlineCount];
2368 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2369 $toc .= $sk->tocLine($anchor, $tocline, $numbering, $toclevel);
2371 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2372 if ( empty( $head[$headlineCount] ) ) {
2373 $head[$headlineCount] = '';
2376 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2378 $head[$headlineCount] .= $sk->editSectionLink($this->mTitle
, $sectionCount+
1);
2381 # give headline the correct <h#> tag
2382 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2390 $toclines = $headlineCount;
2391 $toc .= $sk->tocUnindent( $toclevel - 1 );
2392 $toc = $sk->tocList( $toc );
2395 # split up and insert constructed headlines
2397 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2400 foreach( $blocks as $block ) {
2401 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2402 # This is the [edit] link that appears for the top block of text when
2403 # section editing is enabled
2405 # Disabled because it broke block formatting
2406 # For example, a bullet point in the top line
2407 # $full .= $sk->editSectionLink(0);
2410 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2411 # Top anchor now in skin
2415 if( !empty( $head[$i] ) ) {
2421 $mw =& MagicWord
::get( MAG_TOC
);
2422 return $mw->replace( $toc, $full );
2429 * Return an HTML link for the "ISBN 123456" text
2432 function magicISBN( $text ) {
2434 $fname = 'Parser::magicISBN';
2435 wfProfileIn( $fname );
2437 $a = split( 'ISBN ', ' '.$text );
2438 if ( count ( $a ) < 2 ) {
2439 wfProfileOut( $fname );
2442 $text = substr( array_shift( $a ), 1);
2443 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2445 foreach ( $a as $x ) {
2446 $isbn = $blank = '' ;
2447 while ( ' ' == $x{0} ) {
2449 $x = substr( $x, 1 );
2451 if ( $x == '' ) { # blank isbn
2452 $text .= "ISBN $blank";
2455 while ( strstr( $valid, $x{0} ) != false ) {
2457 $x = substr( $x, 1 );
2459 $num = str_replace( '-', '', $isbn );
2460 $num = str_replace( ' ', '', $num );
2463 $text .= "ISBN $blank$x";
2465 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2466 $text .= '<a href="' .
2467 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2468 "\" class=\"internal\">ISBN $isbn</a>";
2472 wfProfileOut( $fname );
2477 * Return an HTML link for the "RFC 1234" text
2480 * @param string $text Text to be processed
2481 * @param string $keyword Magic keyword to use (default RFC)
2482 * @param string $urlmsg Interface message to use (default rfcurl)
2485 function magicRFC( $text, $keyword='RFC ', $urlmsg='rfcurl' ) {
2488 $valid = '0123456789';
2491 $a = split( $keyword, ' '.$text );
2492 if ( count ( $a ) < 2 ) {
2495 $text = substr( array_shift( $a ), 1);
2497 /* Check if keyword is preceed by [[.
2498 * This test is made here cause of the array_shift above
2499 * that prevent the test to be done in the foreach.
2501 if ( substr( $text, -2 ) == '[[' ) {
2505 foreach ( $a as $x ) {
2506 /* token might be empty if we have RFC RFC 1234 */
2514 /** remove and save whitespaces in $blank */
2515 while ( $x{0} == ' ' ) {
2517 $x = substr( $x, 1 );
2520 /** remove and save the rfc number in $id */
2521 while ( strstr( $valid, $x{0} ) != false ) {
2523 $x = substr( $x, 1 );
2527 /* call back stripped spaces*/
2528 $text .= $keyword.$blank.$x;
2529 } elseif( $internal ) {
2531 $text .= $keyword.$id.$x;
2533 /* build the external link*/
2534 $url = wfMsg( $urlmsg, $id);
2535 $sk =& $this->mOptions
->getSkin();
2536 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
2537 $text .= "<a href='{$url}'{$la}>{$keyword}{$id}</a>{$x}";
2540 /* Check if the next RFC keyword is preceed by [[ */
2541 $internal = ( substr($x,-2) == '[[' );
2547 * Transform wiki markup when saving a page by doing \r\n -> \n
2548 * conversion, substitting signatures, {{subst:}} templates, etc.
2550 * @param string $text the text to transform
2551 * @param Title &$title the Title object for the current article
2552 * @param User &$user the User object describing the current user
2553 * @param ParserOptions $options parsing options
2554 * @param bool $clearState whether to clear the parser state first
2555 * @return string the altered wiki markup
2558 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2559 $this->mOptions
= $options;
2560 $this->mTitle
=& $title;
2561 $this->mOutputType
= OT_WIKI
;
2563 if ( $clearState ) {
2564 $this->clearState();
2567 $stripState = false;
2571 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
2572 $text = $this->strip( $text, $stripState, false );
2573 $text = $this->pstPass2( $text, $user );
2574 $text = $this->unstrip( $text, $stripState );
2575 $text = $this->unstripNoWiki( $text, $stripState );
2580 * Pre-save transform helper function
2583 function pstPass2( $text, &$user ) {
2584 global $wgLang, $wgContLang, $wgLocaltimezone;
2586 # Variable replacement
2587 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2588 $text = $this->replaceVariables( $text );
2592 $n = $user->getName();
2593 $k = $user->getOption( 'nickname' );
2594 if ( '' == $k ) { $k = $n; }
2595 if ( isset( $wgLocaltimezone ) ) {
2596 $oldtz = getenv( 'TZ' );
2597 putenv( 'TZ='.$wgLocaltimezone );
2600 /* Note: This is the timestamp saved as hardcoded wikitext to
2601 * the database, we use $wgContLang here in order to give
2602 * everyone the same signiture and use the default one rather
2603 * than the one selected in each users preferences.
2605 $d = $wgContLang->timeanddate( wfTimestampNow(), false, false) .
2606 ' (' . date( 'T' ) . ')';
2607 if ( isset( $wgLocaltimezone ) ) {
2608 putenv( 'TZ='.$oldtzs );
2611 if( $user->getOption( 'fancysig' ) ) {
2614 $sigText = '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]]";
2616 $text = preg_replace( '/~~~~~/', $d, $text );
2617 $text = preg_replace( '/~~~~/', "$sigText $d", $text );
2618 $text = preg_replace( '/~~~/', $sigText, $text );
2620 # Context links: [[|name]] and [[name (context)|]]
2622 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2623 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2624 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2625 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2627 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2628 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2629 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2630 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2632 $t = $this->mTitle
->getText();
2633 if ( preg_match( $conpat, $t, $m ) ) {
2636 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2637 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2638 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2640 if ( '' == $context ) {
2641 $text = preg_replace( $p2, '[[\\1]]', $text );
2643 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2646 # Trim trailing whitespace
2647 # MAG_END (__END__) tag allows for trailing
2648 # whitespace to be deliberately included
2649 $text = rtrim( $text );
2650 $mw =& MagicWord
::get( MAG_END
);
2651 $mw->matchAndRemove( $text );
2657 * Set up some variables which are usually set up in parse()
2658 * so that an external function can call some class members with confidence
2661 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2662 $this->mTitle
=& $title;
2663 $this->mOptions
= $options;
2664 $this->mOutputType
= $outputType;
2665 if ( $clearState ) {
2666 $this->clearState();
2671 * Transform a MediaWiki message by replacing magic variables.
2673 * @param string $text the text to transform
2674 * @param ParserOptions $options options
2675 * @return string the text with variables substituted
2678 function transformMsg( $text, $options ) {
2680 static $executing = false;
2682 # Guard against infinite recursion
2688 $this->mTitle
= $wgTitle;
2689 $this->mOptions
= $options;
2690 $this->mOutputType
= OT_MSG
;
2691 $this->clearState();
2692 $text = $this->replaceVariables( $text );
2699 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2700 * Callback will be called with the text within
2701 * Transform and return the text within
2704 function setHook( $tag, $callback ) {
2705 $oldVal = @$this->mTagHooks
[$tag];
2706 $this->mTagHooks
[$tag] = $callback;
2711 * Replace <!--LINK--> link placeholders with actual links, in the buffer
2712 * Placeholders created in Skin::makeLinkObj()
2713 * Returns an array of links found, indexed by PDBK:
2717 * $options is a bit field, RLH_FOR_UPDATE to select for update
2719 function replaceLinkHolders( &$text, $options = 0 ) {
2720 global $wgUser, $wgLinkCache, $wgUseOldExistenceCheck, $wgLinkHolders;
2721 global $wgInterwikiLinkHolders;
2722 global $outputReplace;
2724 if ( $wgUseOldExistenceCheck ) {
2728 $fname = 'Parser::replaceLinkHolders';
2729 wfProfileIn( $fname );
2734 #if ( !empty( $tmpLinks[0] ) ) { #TODO
2735 if ( !empty( $wgLinkHolders['namespaces'] ) ) {
2736 wfProfileIn( $fname.'-check' );
2737 $dbr =& wfGetDB( DB_SLAVE
);
2738 $page = $dbr->tableName( 'page' );
2739 $sk = $wgUser->getSkin();
2740 $threshold = $wgUser->getOption('stubthreshold');
2743 asort( $wgLinkHolders['namespaces'] );
2747 foreach ( $wgLinkHolders['namespaces'] as $key => $val ) {
2749 $title = $wgLinkHolders['titles'][$key];
2751 # Skip invalid entries.
2752 # Result will be ugly, but prevents crash.
2753 if ( is_null( $title ) ) {
2756 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
2758 # Check if it's in the link cache already
2759 if ( $wgLinkCache->getGoodLinkID( $pdbk ) ) {
2760 $colours[$pdbk] = 1;
2761 } elseif ( $wgLinkCache->isBadLink( $pdbk ) ) {
2762 $colours[$pdbk] = 0;
2764 # Not in the link cache, add it to the query
2765 if ( !isset( $current ) ) {
2767 $query = "SELECT page_id, page_namespace, page_title";
2768 if ( $threshold > 0 ) {
2769 $query .= ', page_len, page_is_redirect';
2771 $query .= " FROM $page WHERE (page_namespace=$val AND page_title IN(";
2772 } elseif ( $current != $val ) {
2774 $query .= ")) OR (page_namespace=$val AND page_title IN(";
2779 $query .= $dbr->addQuotes( $wgLinkHolders['dbkeys'][$key] );
2784 if ( $options & RLH_FOR_UPDATE
) {
2785 $query .= ' FOR UPDATE';
2788 $res = $dbr->query( $query, $fname );
2790 # Fetch data and form into an associative array
2791 # non-existent = broken
2794 while ( $s = $dbr->fetchObject($res) ) {
2795 $title = Title
::makeTitle( $s->page_namespace
, $s->page_title
);
2796 $pdbk = $title->getPrefixedDBkey();
2797 $wgLinkCache->addGoodLink( $s->page_id
, $pdbk );
2799 if ( $threshold > 0 ) {
2800 $size = $s->page_len
;
2801 if ( $s->page_is_redirect ||
$s->page_namespace
!= 0 ||
$size >= $threshold ) {
2802 $colours[$pdbk] = 1;
2804 $colours[$pdbk] = 2;
2807 $colours[$pdbk] = 1;
2811 wfProfileOut( $fname.'-check' );
2813 # Construct search and replace arrays
2814 wfProfileIn( $fname.'-construct' );
2815 $outputReplace = array();
2816 foreach ( $wgLinkHolders['namespaces'] as $key => $ns ) {
2817 $pdbk = $pdbks[$key];
2818 $searchkey = '<!--LINK '.$key.'-->';
2819 $title = $wgLinkHolders['titles'][$key];
2820 if ( empty( $colours[$pdbk] ) ) {
2821 $wgLinkCache->addBadLink( $pdbk );
2822 $colours[$pdbk] = 0;
2823 $outputReplace[$searchkey] = $sk->makeBrokenLinkObj( $title,
2824 $wgLinkHolders['texts'][$key],
2825 $wgLinkHolders['queries'][$key] );
2826 } elseif ( $colours[$pdbk] == 1 ) {
2827 $outputReplace[$searchkey] = $sk->makeKnownLinkObj( $title,
2828 $wgLinkHolders['texts'][$key],
2829 $wgLinkHolders['queries'][$key] );
2830 } elseif ( $colours[$pdbk] == 2 ) {
2831 $outputReplace[$searchkey] = $sk->makeStubLinkObj( $title,
2832 $wgLinkHolders['texts'][$key],
2833 $wgLinkHolders['queries'][$key] );
2836 wfProfileOut( $fname.'-construct' );
2839 wfProfileIn( $fname.'-replace' );
2841 $text = preg_replace_callback(
2842 '/(<!--LINK .*?-->)/',
2843 "outputReplaceMatches",
2845 wfProfileOut( $fname.'-replace' );
2848 if ( !empty( $wgInterwikiLinkHolders ) ) {
2849 wfProfileIn( $fname.'-interwiki' );
2850 $outputReplace = $wgInterwikiLinkHolders;
2851 $text = preg_replace_callback(
2852 '/<!--IWLINK (.*?)-->/',
2853 "outputReplaceMatches",
2855 wfProfileOut( $fname.'-interwiki' );
2858 wfProfileOut( $fname );
2863 * Renders an image gallery from a text with one line per image.
2864 * text labels may be given by using |-style alternative text. E.g.
2865 * Image:one.jpg|The number "1"
2866 * Image:tree.jpg|A tree
2867 * given as text will return the HTML of a gallery with two images,
2868 * labeled 'The number "1"' and
2871 function renderImageGallery( $text ) {
2873 global $wgUser, $wgParser, $wgTitle;
2874 $parserOptions = ParserOptions
::newFromUser( $wgUser );
2876 global $wgLinkCache;
2877 $ig = new ImageGallery();
2878 $ig->setShowBytes( false );
2879 $ig->setShowFilename( false );
2880 $lines = explode( "\n", $text );
2882 foreach ( $lines as $line ) {
2883 # match lines like these:
2884 # Image:someimage.jpg|This is some image
2885 preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
2887 if ( count( $matches ) == 0 ) {
2890 $nt = Title
::newFromURL( $matches[1] );
2891 if( is_null( $nt ) ) {
2892 # Bogus title. Ignore these so we don't bomb out later.
2895 if ( isset( $matches[3] ) ) {
2896 $label = $matches[3];
2901 $html = $wgParser->parse( $label , $wgTitle, $parserOptions );
2902 $html = $html->mText
;
2904 $ig->add( Image
::newFromTitle( $nt ), $html );
2905 $wgLinkCache->addImageLinkObj( $nt );
2907 return $ig->toHTML();
2913 * @package MediaWiki
2917 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2918 var $mCacheTime; # Used in ParserCache
2919 var $mVersion; # Compatibility check
2921 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
2922 $containsOldMagic = false )
2924 $this->mText
= $text;
2925 $this->mLanguageLinks
= $languageLinks;
2926 $this->mCategoryLinks
= $categoryLinks;
2927 $this->mContainsOldMagic
= $containsOldMagic;
2928 $this->mCacheTime
= '';
2929 $this->mVersion
= MW_PARSER_VERSION
;
2932 function getText() { return $this->mText
; }
2933 function getLanguageLinks() { return $this->mLanguageLinks
; }
2934 function getCategoryLinks() { return array_keys( $this->mCategoryLinks
); }
2935 function getCacheTime() { return $this->mCacheTime
; }
2936 function containsOldMagic() { return $this->mContainsOldMagic
; }
2937 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2938 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2939 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2940 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2941 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
2942 function addCategoryLink( $c ) { $this->mCategoryLinks
[$c] = 1; }
2944 function merge( $other ) {
2945 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2946 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2947 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2951 * Return true if this cached output object predates the global or
2952 * per-article cache invalidation timestamps, or if it comes from
2953 * an incompatible older version.
2955 * @param string $touched the affected article's last touched timestamp
2959 function expired( $touched ) {
2960 global $wgCacheEpoch;
2961 return $this->getCacheTime() <= $touched ||
2962 $this->getCacheTime() <= $wgCacheEpoch ||
2963 !isset( $this->mVersion
) ||
2964 version_compare( $this->mVersion
, MW_PARSER_VERSION
, "lt" );
2969 * Set options of the Parser
2971 * @package MediaWiki
2975 # All variables are private
2976 var $mUseTeX; # Use texvc to expand <math> tags
2977 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2978 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2979 var $mAllowExternalImages; # Allow external images inline
2980 var $mSkin; # Reference to the preferred skin
2981 var $mDateFormat; # Date format index
2982 var $mEditSection; # Create "edit section" links
2983 var $mNumberHeadings; # Automatically number headings
2985 function getUseTeX() { return $this->mUseTeX
; }
2986 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2987 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2988 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2989 function getSkin() { return $this->mSkin
; }
2990 function getDateFormat() { return $this->mDateFormat
; }
2991 function getEditSection() { return $this->mEditSection
; }
2992 function getNumberHeadings() { return $this->mNumberHeadings
; }
2994 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2995 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2996 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2997 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2998 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2999 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
3000 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
3002 function setSkin( &$x ) { $this->mSkin
=& $x; }
3005 * Get parser options
3008 function newFromUser( &$user ) {
3009 $popts = new ParserOptions
;
3010 $popts->initialiseFromUser( $user );
3014 /** Get user options */
3015 function initialiseFromUser( &$userInput ) {
3016 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
3017 $fname = 'ParserOptions::initialiseFromUser';
3018 wfProfileIn( $fname );
3019 if ( !$userInput ) {
3021 $user->setLoaded( true );
3023 $user =& $userInput;
3026 $this->mUseTeX
= $wgUseTeX;
3027 $this->mUseDynamicDates
= $wgUseDynamicDates;
3028 $this->mInterwikiMagic
= $wgInterwikiMagic;
3029 $this->mAllowExternalImages
= $wgAllowExternalImages;
3030 wfProfileIn( $fname.'-skin' );
3031 $this->mSkin
=& $user->getSkin();
3032 wfProfileOut( $fname.'-skin' );
3033 $this->mDateFormat
= $user->getOption( 'date' );
3034 $this->mEditSection
= $user->getOption( 'editsection' );
3035 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
3036 wfProfileOut( $fname );
3041 * Callback function used by Parser::replaceLinkHolders()
3042 * to substitute link placeholders.
3044 function &outputReplaceMatches( $matches ) {
3045 global $outputReplace;
3046 return $outputReplace[$matches[1]];
3050 * Return the total number of articles
3052 function wfNumberOfArticles() {
3053 global $wgNumberOfArticles;
3056 return $wgNumberOfArticles;
3060 * Get various statistics from the database
3063 function wfLoadSiteStats() {
3064 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
3065 $fname = 'wfLoadSiteStats';
3067 if ( -1 != $wgNumberOfArticles ) return;
3068 $dbr =& wfGetDB( DB_SLAVE
);
3069 $s = $dbr->selectRow( 'site_stats',
3070 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
3071 array( 'ss_row_id' => 1 ), $fname
3074 if ( $s === false ) {
3077 $wgTotalViews = $s->ss_total_views
;
3078 $wgTotalEdits = $s->ss_total_edits
;
3079 $wgNumberOfArticles = $s->ss_good_articles
;
3085 * Basicly replacing " > and < with HTML entities ( ", >, <)
3087 * @param string $in Text that might contain HTML tags
3088 * @return string Escaped string
3090 function wfEscapeHTMLTagsOnly( $in ) {
3092 array( '"', '>', '<' ),
3093 array( '"', '>', '<' ),