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', '/\[(\b('.URL_PROTOCOLS
.'):'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/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 # Set linktype for CSS - if URL==text, link is essentially free
996 $linktype = ($text == $url) ?
'free' : 'text';
998 # No link text, e.g. [http://domain.tld/some.link]
1000 # Autonumber if allowed
1001 if ( strpos( HTTP_PROTOCOLS
, $protocol ) !== false ) {
1002 $text = '[' . ++
$this->mAutonumber
. ']';
1003 $linktype = 'autonumber';
1005 # Otherwise just use the URL
1006 $text = htmlspecialchars( $url );
1010 # Have link text, e.g. [http://domain.tld/some.link text]s
1012 if ( preg_match( $linktrail, $trail, $m2 ) ) {
1018 # Process the trail (i.e. everything after this link up until start of the next link),
1019 # replacing any non-bracketed links
1020 $trail = $this->replaceFreeExternalLinks( $trail );
1022 # Use the encoded URL
1023 # This means that users can paste URLs directly into the text
1024 # Funny characters like ö aren't valid in URLs anyway
1025 # This was changed in August 2004
1026 $s .= $sk->makeExternalLink( $url, $text, false, $linktype ) . $dtrail . $trail;
1029 wfProfileOut( $fname );
1034 * Replace anything that looks like a URL with a link
1037 function replaceFreeExternalLinks( $text ) {
1038 $fname = 'Parser::replaceFreeExternalLinks';
1039 wfProfileIn( $fname );
1041 $bits = preg_split( '/(\b(?:'.URL_PROTOCOLS
.'):)/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1042 $s = array_shift( $bits );
1045 $sk =& $this->mOptions
->getSkin();
1047 while ( $i < count( $bits ) ){
1048 $protocol = $bits[$i++
];
1049 $remainder = $bits[$i++
];
1051 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1052 # Found some characters after the protocol that look promising
1053 $url = $protocol . $m[1];
1056 # The characters '<' and '>' (which were escaped by
1057 # removeHTMLtags()) should not be included in
1058 # URLs, per RFC 2396.
1059 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1060 $trail = substr($url, $m2[0][1]) . $trail;
1061 $url = substr($url, 0, $m2[0][1]);
1064 # Move trailing punctuation to $trail
1066 # If there is no left bracket, then consider right brackets fair game too
1067 if ( strpos( $url, '(' ) === false ) {
1071 $numSepChars = strspn( strrev( $url ), $sep );
1072 if ( $numSepChars ) {
1073 $trail = substr( $url, -$numSepChars ) . $trail;
1074 $url = substr( $url, 0, -$numSepChars );
1077 # Replace & from obsolete syntax with &.
1078 # All HTML entities will be escaped by makeExternalLink()
1079 # or maybeMakeImageLink()
1080 $url = str_replace( '&', '&', $url );
1082 # Is this an external image?
1083 $text = $this->maybeMakeImageLink( $url );
1084 if ( $text === false ) {
1085 # Not an image, make a link
1086 $text = $sk->makeExternalLink( $url, $url, true, 'free' );
1088 $s .= $text . $trail;
1090 $s .= $protocol . $remainder;
1098 * make an image if it's allowed
1101 function maybeMakeImageLink( $url ) {
1102 $sk =& $this->mOptions
->getSkin();
1104 if ( $this->mOptions
->getAllowExternalImages() ) {
1105 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1107 $text = $sk->makeImage( htmlspecialchars( $url ) );
1114 * Process [[ ]] wikilinks
1119 function replaceInternalLinks( $s ) {
1120 global $wgLang, $wgContLang, $wgLinkCache;
1121 global $wgDisableLangConversion;
1122 static $fname = 'Parser::replaceInternalLinks' ;
1124 wfProfileIn( $fname );
1126 wfProfileIn( $fname.'-setup' );
1128 # the % is needed to support urlencoded titles as well
1129 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1131 $sk =& $this->mOptions
->getSkin();
1132 global $wgUseOldExistenceCheck;
1133 # "Post-parse link colour check" works only on wiki text since it's now
1134 # in Parser. Enable it, then disable it when we're done.
1135 $saveParseColour = $sk->postParseLinkColour( !$wgUseOldExistenceCheck );
1137 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1139 #split the entire text string on occurences of [[
1140 $a = explode( '[[', ' ' . $s );
1141 #get the first element (all text up to first [[), and remove the space we added
1142 $s = array_shift( $a );
1143 $s = substr( $s, 1 );
1145 # Match a link having the form [[namespace:link|alternate]]trail
1147 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
1148 # Match cases where there is no "]]", which might still be images
1149 static $e1_img = FALSE;
1150 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1151 # Match the end of a line for a word that's not followed by whitespace,
1152 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1153 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1155 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1157 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1159 if ( $useLinkPrefixExtension ) {
1160 if ( preg_match( $e2, $s, $m ) ) {
1161 $first_prefix = $m[2];
1164 $first_prefix = false;
1170 $selflink = $this->mTitle
->getPrefixedText();
1171 wfProfileOut( $fname.'-setup' );
1173 $checkVariantLink = sizeof($wgContLang->getVariants())>1;
1174 $useSubpages = $this->areSubpagesAllowed();
1176 # Loop for each link
1177 for ($k = 0; isset( $a[$k] ); $k++
) {
1179 if ( $useLinkPrefixExtension ) {
1180 wfProfileIn( $fname.'-prefixhandling' );
1181 if ( preg_match( $e2, $s, $m ) ) {
1189 $prefix = $first_prefix;
1190 $first_prefix = false;
1192 wfProfileOut( $fname.'-prefixhandling' );
1195 $might_be_img = false;
1197 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1199 # fix up urlencoded title texts
1200 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1202 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1203 $might_be_img = true;
1205 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1207 } else { # Invalid form; output directly
1208 $s .= $prefix . '[[' . $line ;
1212 # Don't allow internal links to pages containing
1213 # PROTO: where PROTO is a valid URL protocol; these
1214 # should be external links.
1215 if (preg_match('/^(\b(?:'.URL_PROTOCOLS
.'):)/', $m[1])) {
1216 $s .= $prefix . '[[' . $line ;
1220 # Make subpage if necessary
1221 if( $useSubpages ) {
1222 $link = $this->maybeDoSubpageLink( $m[1], $text );
1227 $noforce = (substr($m[1], 0, 1) != ':');
1229 # Strip off leading ':'
1230 $link = substr($link, 1);
1233 $nt =& Title
::newFromText( $this->unstripNoWiki($link, $this->mStripState
) );
1235 $s .= $prefix . '[[' . $line;
1239 #check other language variants of the link
1240 #if the article does not exist
1241 if( $checkVariantLink
1242 && $nt->getArticleID() == 0 ) {
1243 $wgContLang->findVariantLink($link, $nt);
1246 $ns = $nt->getNamespace();
1247 $iw = $nt->getInterWiki();
1249 if ($might_be_img) { # if this is actually an invalid link
1250 if ($ns == NS_IMAGE
&& $noforce) { #but might be an image
1252 while (isset ($a[$k+
1]) ) {
1253 #look at the next 'line' to see if we can close it there
1254 $next_line = array_shift(array_splice( $a, $k +
1, 1) );
1255 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1256 # the first ]] closes the inner link, the second the image
1258 $text .= '[[' . $m[1];
1261 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1262 #if there's exactly one ]] that's fine, we'll keep looking
1263 $text .= '[[' . $m[0];
1265 #if $next_line is invalid too, we need look no further
1266 $text .= '[[' . $next_line;
1271 # we couldn't find the end of this imageLink, so output it raw
1272 #but don't ignore what might be perfectly normal links in the text we've examined
1273 $text = $this->replaceInternalLinks($text);
1274 $s .= $prefix . '[[' . $link . '|' . $text;
1275 # note: no $trail, because without an end, there *is* no trail
1278 } else { #it's not an image, so output it raw
1279 $s .= $prefix . '[[' . $link . '|' . $text;
1280 # note: no $trail, because without an end, there *is* no trail
1285 $wasblank = ( '' == $text );
1286 if( $wasblank ) $text = $link;
1289 # Link not escaped by : , create the various objects
1293 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1294 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1295 $tmp = $prefix . $trail ;
1296 $s .= (trim($tmp) == '')?
'': $tmp;
1300 if ( $ns == NS_IMAGE
) {
1301 wfProfileIn( "$fname-image" );
1303 # recursively parse links inside the image caption
1304 # actually, this will parse them in any other parameters, too,
1305 # but it might be hard to fix that, and it doesn't matter ATM
1306 $text = $this->replaceExternalLinks($text);
1307 $text = $this->replaceInternalLinks($text);
1309 # replace the image with a link-holder so that replaceExternalLinks() can't mess with it
1310 $s .= $prefix . $this->insertStripItem( $sk->makeImageLinkObj( $nt, $text ), $this->mStripState
) . $trail;
1311 $wgLinkCache->addImageLinkObj( $nt );
1313 wfProfileOut( "$fname-image" );
1317 if ( $ns == NS_CATEGORY
) {
1318 wfProfileIn( "$fname-category" );
1319 $t = $nt->getText();
1321 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1322 $pPLC=$sk->postParseLinkColour();
1323 $sk->postParseLinkColour( false );
1324 $t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
1325 $sk->postParseLinkColour( $pPLC );
1326 $wgLinkCache->resume();
1329 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1330 $sortkey = $this->mTitle
->getText();
1332 $sortkey = $this->mTitle
->getPrefixedText();
1337 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1338 $this->mOutput
->addCategoryLink( $t );
1339 $s .= $prefix . $trail ;
1341 wfProfileOut( "$fname-category" );
1346 if( ( $nt->getPrefixedText() === $selflink ) &&
1347 ( $nt->getFragment() === '' ) ) {
1348 # Self-links are handled specially; generally de-link and change to bold.
1349 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1353 # Special and Media are pseudo-namespaces; no pages actually exist in them
1354 if( $ns == NS_MEDIA
) {
1355 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text, true ) . $trail;
1356 $wgLinkCache->addImageLinkObj( $nt );
1358 } elseif( $ns == NS_SPECIAL
) {
1359 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1362 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1364 $sk->postParseLinkColour( $saveParseColour );
1365 wfProfileOut( $fname );
1370 * Return true if subpage links should be expanded on this page.
1373 function areSubpagesAllowed() {
1374 # Some namespaces don't allow subpages
1375 global $wgNamespacesWithSubpages;
1376 return !empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]);
1380 * Handle link to subpage if necessary
1381 * @param string $target the source of the link
1382 * @param string &$text the link text, modified as necessary
1383 * @return string the full name of the link
1386 function maybeDoSubpageLink($target, &$text) {
1389 # :Foobar -- override special treatment of prefix (images, language links)
1390 # /Foobar -- convert to CurrentPage/Foobar
1391 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1392 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1393 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1395 $fname = 'Parser::maybeDoSubpageLink';
1396 wfProfileIn( $fname );
1397 $ret = $target; # default return value is no change
1399 # Some namespaces don't allow subpages,
1400 # so only perform processing if subpages are allowed
1401 if( $this->areSubpagesAllowed() ) {
1402 # Look at the first character
1403 if( $target != '' && $target{0} == '/' ) {
1404 # / at end means we don't want the slash to be shown
1405 if( substr( $target, -1, 1 ) == '/' ) {
1406 $target = substr( $target, 1, -1 );
1409 $noslash = substr( $target, 1 );
1412 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1413 if( '' === $text ) {
1415 } # this might be changed for ugliness reasons
1417 # check for .. subpage backlinks
1419 $nodotdot = $target;
1420 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1422 $nodotdot = substr( $nodotdot, 3 );
1424 if($dotdotcount > 0) {
1425 $exploded = explode( '/', $this->mTitle
->GetPrefixedText() );
1426 if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1427 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1428 # / at the end means don't show full path
1429 if( substr( $nodotdot, -1, 1 ) == '/' ) {
1430 $nodotdot = substr( $nodotdot, 0, -1 );
1431 if( '' === $text ) {
1435 $nodotdot = trim( $nodotdot );
1436 if( $nodotdot != '' ) {
1437 $ret .= '/' . $nodotdot;
1444 wfProfileOut( $fname );
1449 * Used by doBlockLevels()
1452 /* private */ function closeParagraph() {
1454 if ( '' != $this->mLastSection
) {
1455 $result = '</' . $this->mLastSection
. ">\n";
1457 $this->mInPre
= false;
1458 $this->mLastSection
= '';
1461 # getCommon() returns the length of the longest common substring
1462 # of both arguments, starting at the beginning of both.
1464 /* private */ function getCommon( $st1, $st2 ) {
1465 $fl = strlen( $st1 );
1466 $shorter = strlen( $st2 );
1467 if ( $fl < $shorter ) { $shorter = $fl; }
1469 for ( $i = 0; $i < $shorter; ++
$i ) {
1470 if ( $st1{$i} != $st2{$i} ) { break; }
1474 # These next three functions open, continue, and close the list
1475 # element appropriate to the prefix character passed into them.
1477 /* private */ function openList( $char ) {
1478 $result = $this->closeParagraph();
1480 if ( '*' == $char ) { $result .= '<ul><li>'; }
1481 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1482 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1483 else if ( ';' == $char ) {
1484 $result .= '<dl><dt>';
1485 $this->mDTopen
= true;
1487 else { $result = '<!-- ERR 1 -->'; }
1492 /* private */ function nextItem( $char ) {
1493 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1494 else if ( ':' == $char ||
';' == $char ) {
1496 if ( $this->mDTopen
) { $close = '</dt>'; }
1497 if ( ';' == $char ) {
1498 $this->mDTopen
= true;
1499 return $close . '<dt>';
1501 $this->mDTopen
= false;
1502 return $close . '<dd>';
1505 return '<!-- ERR 2 -->';
1508 /* private */ function closeList( $char ) {
1509 if ( '*' == $char ) { $text = '</li></ul>'; }
1510 else if ( '#' == $char ) { $text = '</li></ol>'; }
1511 else if ( ':' == $char ) {
1512 if ( $this->mDTopen
) {
1513 $this->mDTopen
= false;
1514 $text = '</dt></dl>';
1516 $text = '</dd></dl>';
1519 else { return '<!-- ERR 3 -->'; }
1525 * Make lists from lines starting with ':', '*', '#', etc.
1528 * @return string the lists rendered as HTML
1530 function doBlockLevels( $text, $linestart ) {
1531 $fname = 'Parser::doBlockLevels';
1532 wfProfileIn( $fname );
1534 # Parsing through the text line by line. The main thing
1535 # happening here is handling of block-level elements p, pre,
1536 # and making lists from lines starting with * # : etc.
1538 $textLines = explode( "\n", $text );
1540 $lastPrefix = $output = $lastLine = '';
1541 $this->mDTopen
= $inBlockElem = false;
1543 $paragraphStack = false;
1545 if ( !$linestart ) {
1546 $output .= array_shift( $textLines );
1548 foreach ( $textLines as $oLine ) {
1549 $lastPrefixLength = strlen( $lastPrefix );
1550 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1551 $preOpenMatch = preg_match('/<pre/i', $oLine );
1552 if ( !$this->mInPre
) {
1553 # Multiple prefixes may abut each other for nested lists.
1554 $prefixLength = strspn( $oLine, '*#:;' );
1555 $pref = substr( $oLine, 0, $prefixLength );
1558 $pref2 = str_replace( ';', ':', $pref );
1559 $t = substr( $oLine, $prefixLength );
1560 $this->mInPre
= !empty($preOpenMatch);
1562 # Don't interpret any other prefixes in preformatted text
1564 $pref = $pref2 = '';
1569 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1570 # Same as the last item, so no need to deal with nesting or opening stuff
1571 $output .= $this->nextItem( substr( $pref, -1 ) );
1572 $paragraphStack = false;
1574 if ( substr( $pref, -1 ) == ';') {
1575 # The one nasty exception: definition lists work like this:
1576 # ; title : definition text
1577 # So we check for : in the remainder text to split up the
1578 # title and definition, without b0rking links.
1579 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1581 $output .= $term . $this->nextItem( ':' );
1584 } elseif( $prefixLength ||
$lastPrefixLength ) {
1585 # Either open or close a level...
1586 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1587 $paragraphStack = false;
1589 while( $commonPrefixLength < $lastPrefixLength ) {
1590 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1591 --$lastPrefixLength;
1593 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1594 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1596 while ( $prefixLength > $commonPrefixLength ) {
1597 $char = substr( $pref, $commonPrefixLength, 1 );
1598 $output .= $this->openList( $char );
1600 if ( ';' == $char ) {
1601 # FIXME: This is dupe of code above
1602 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1604 $output .= $term . $this->nextItem( ':' );
1607 ++
$commonPrefixLength;
1609 $lastPrefix = $pref2;
1611 if( 0 == $prefixLength ) {
1612 wfProfileIn( "$fname-paragraph" );
1613 # No prefix (not in list)--go to paragraph mode
1614 $uniq_prefix = UNIQ_PREFIX
;
1615 // XXX: use a stack for nestable elements like span, table and div
1616 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
1617 $closematch = preg_match(
1618 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1619 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/iS', $t );
1620 if ( $openmatch or $closematch ) {
1621 $paragraphStack = false;
1622 $output .= $this->closeParagraph();
1623 if($preOpenMatch and !$preCloseMatch) {
1624 $this->mInPre
= true;
1626 if ( $closematch ) {
1627 $inBlockElem = false;
1629 $inBlockElem = true;
1631 } else if ( !$inBlockElem && !$this->mInPre
) {
1632 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1634 if ($this->mLastSection
!= 'pre') {
1635 $paragraphStack = false;
1636 $output .= $this->closeParagraph().'<pre>';
1637 $this->mLastSection
= 'pre';
1639 $t = substr( $t, 1 );
1642 if ( '' == trim($t) ) {
1643 if ( $paragraphStack ) {
1644 $output .= $paragraphStack.'<br />';
1645 $paragraphStack = false;
1646 $this->mLastSection
= 'p';
1648 if ($this->mLastSection
!= 'p' ) {
1649 $output .= $this->closeParagraph();
1650 $this->mLastSection
= '';
1651 $paragraphStack = '<p>';
1653 $paragraphStack = '</p><p>';
1657 if ( $paragraphStack ) {
1658 $output .= $paragraphStack;
1659 $paragraphStack = false;
1660 $this->mLastSection
= 'p';
1661 } else if ($this->mLastSection
!= 'p') {
1662 $output .= $this->closeParagraph().'<p>';
1663 $this->mLastSection
= 'p';
1668 wfProfileOut( "$fname-paragraph" );
1670 if ($paragraphStack === false) {
1674 while ( $prefixLength ) {
1675 $output .= $this->closeList( $pref2{$prefixLength-1} );
1678 if ( '' != $this->mLastSection
) {
1679 $output .= '</' . $this->mLastSection
. '>';
1680 $this->mLastSection
= '';
1683 wfProfileOut( $fname );
1688 * Split up a string on ':', ignoring any occurences inside
1689 * <a>..</a> or <span>...</span>
1690 * @param string $str the string to split
1691 * @param string &$before set to everything before the ':'
1692 * @param string &$after set to everything after the ':'
1693 * return string the position of the ':', or false if none found
1695 function findColonNoLinks($str, &$before, &$after) {
1696 # I wonder if we should make this count all tags, not just <a>
1697 # and <span>. That would prevent us from matching a ':' that
1698 # comes in the middle of italics other such formatting....
1700 $fname = 'Parser::findColonNoLinks';
1701 wfProfileIn( $fname );
1704 $colon = strpos($str, ':', $pos);
1706 if ($colon !== false) {
1707 $before = substr($str, 0, $colon);
1708 $after = substr($str, $colon +
1);
1710 # Skip any ':' within <a> or <span> pairs
1711 $a = substr_count($before, '<a');
1712 $s = substr_count($before, '<span');
1713 $ca = substr_count($before, '</a>');
1714 $cs = substr_count($before, '</span>');
1716 if ($a <= $ca and $s <= $cs) {
1717 # Tags are balanced before ':'; ok
1722 } while ($colon !== false);
1723 wfProfileOut( $fname );
1728 * Return value of a magic variable (like PAGENAME)
1732 function getVariableValue( $index ) {
1733 global $wgContLang, $wgSitename, $wgServer;
1736 * Some of these require message or data lookups and can be
1737 * expensive to check many times.
1739 static $varCache = array();
1740 if( isset( $varCache[$index] ) ) return $varCache[$index];
1743 case MAG_CURRENTMONTH
:
1744 return $varCache[$index] = $wgContLang->formatNum( date( 'm' ) );
1745 case MAG_CURRENTMONTHNAME
:
1746 return $varCache[$index] = $wgContLang->getMonthName( date('n') );
1747 case MAG_CURRENTMONTHNAMEGEN
:
1748 return $varCache[$index] = $wgContLang->getMonthNameGen( date('n') );
1749 case MAG_CURRENTDAY
:
1750 return $varCache[$index] = $wgContLang->formatNum( date('j') );
1752 return $this->mTitle
->getText();
1754 return $this->mTitle
->getPartialURL();
1756 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1757 return $wgContLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1758 case MAG_CURRENTDAYNAME
:
1759 return $varCache[$index] = $wgContLang->getWeekdayName( date('w')+
1 );
1760 case MAG_CURRENTYEAR
:
1761 return $varCache[$index] = $wgContLang->formatNum( date( 'Y' ) );
1762 case MAG_CURRENTTIME
:
1763 return $varCache[$index] = $wgContLang->time( wfTimestampNow(), false );
1764 case MAG_CURRENTWEEK
:
1765 return $varCache[$index] = $wgContLang->formatNum( date('W') );
1766 case MAG_CURRENTDOW
:
1767 return $varCache[$index] = $wgContLang->formatNum( date('w') );
1768 case MAG_NUMBEROFARTICLES
:
1769 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfArticles() );
1780 * initialise the magic variables (like CURRENTMONTHNAME)
1784 function initialiseVariables() {
1785 $fname = 'Parser::initialiseVariables';
1786 wfProfileIn( $fname );
1787 global $wgVariableIDs;
1788 $this->mVariables
= array();
1789 foreach ( $wgVariableIDs as $id ) {
1790 $mw =& MagicWord
::get( $id );
1791 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1793 wfProfileOut( $fname );
1797 * Replace magic variables, templates, and template arguments
1798 * with the appropriate text. Templates are substituted recursively,
1799 * taking care to avoid infinite loops.
1801 * Note that the substitution depends on value of $mOutputType:
1802 * OT_WIKI: only {{subst:}} templates
1803 * OT_MSG: only magic variables
1804 * OT_HTML: all templates and magic variables
1806 * @param string $tex The text to transform
1807 * @param array $args Key-value pairs representing template parameters to substitute
1810 function replaceVariables( $text, $args = array() ) {
1811 global $wgLang, $wgScript, $wgArticlePath;
1813 # Prevent too big inclusions
1814 if( strlen( $text ) > MAX_INCLUDE_SIZE
) {
1818 $fname = 'Parser::replaceVariables';
1819 wfProfileIn( $fname );
1821 $titleChars = Title
::legalChars();
1823 # This function is called recursively. To keep track of arguments we need a stack:
1824 array_push( $this->mArgStack
, $args );
1826 # Variable substitution
1827 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", array( &$this, 'variableSubstitution' ), $text );
1829 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1830 # Argument substitution
1831 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", array( &$this, 'argSubstitution' ), $text );
1833 # Template substitution
1834 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1835 $text = preg_replace_callback( $regex, array( &$this, 'braceSubstitution' ), $text );
1837 array_pop( $this->mArgStack
);
1839 wfProfileOut( $fname );
1844 * Replace magic variables
1847 function variableSubstitution( $matches ) {
1848 $fname = 'parser::variableSubstitution';
1849 wfProfileIn( $fname );
1850 if ( !$this->mVariables
) {
1851 $this->initialiseVariables();
1854 if ( $this->mOutputType
== OT_WIKI
) {
1855 # Do only magic variables prefixed by SUBST
1856 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1857 if (!$mwSubst->matchStartAndRemove( $matches[1] ))
1859 # Note that if we don't substitute the variable below,
1860 # we don't remove the {{subst:}} magic word, in case
1861 # it is a template rather than a magic variable.
1863 if ( !$skip && array_key_exists( $matches[1], $this->mVariables
) ) {
1864 $text = $this->mVariables
[$matches[1]];
1865 $this->mOutput
->mContainsOldMagic
= true;
1867 $text = $matches[0];
1869 wfProfileOut( $fname );
1873 # Split template arguments
1874 function getTemplateArgs( $argsString ) {
1875 if ( $argsString === '' ) {
1879 $args = explode( '|', substr( $argsString, 1 ) );
1881 # If any of the arguments contains a '[[' but no ']]', it needs to be
1882 # merged with the next arg because the '|' character between belongs
1883 # to the link syntax and not the template parameter syntax.
1884 $argc = count($args);
1886 for ( $i = 0; $i < $argc-1; $i++
) {
1887 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1888 $args[$i] .= '|'.$args[$i+
1];
1889 array_splice($args, $i+
1, 1);
1899 * Return the text of a template, after recursively
1900 * replacing any variables or templates within the template.
1902 * @param array $matches The parts of the template
1903 * $matches[1]: the title, i.e. the part before the |
1904 * $matches[2]: the parameters (including a leading |), if any
1905 * @return string the text of the template
1908 function braceSubstitution( $matches ) {
1909 global $wgLinkCache, $wgContLang;
1910 $fname = 'Parser::braceSubstitution';
1911 wfProfileIn( $fname );
1919 # Need to know if the template comes at the start of a line,
1920 # to treat the beginning of the template like the beginning
1921 # of a line for tables and block-level elements.
1922 $linestart = $matches[1];
1924 # $part1 is the bit before the first |, and must contain only title characters
1925 # $args is a list of arguments, starting from index 0, not including $part1
1927 $part1 = $matches[2];
1928 # If the third subpattern matched anything, it will start with |
1930 $args = $this->getTemplateArgs($matches[3]);
1931 $argc = count( $args );
1933 # Don't parse {{{}}} because that's only for template arguments
1934 if ( $linestart === '{' ) {
1935 $text = $matches[0];
1942 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1943 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
1944 # One of two possibilities is true:
1945 # 1) Found SUBST but not in the PST phase
1946 # 2) Didn't find SUBST and in the PST phase
1947 # In either case, return without further processing
1948 $text = $matches[0];
1954 # MSG, MSGNW and INT
1957 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1958 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1961 # Remove obsolete MSG:
1962 $mwMsg =& MagicWord
::get( MAG_MSG
);
1963 $mwMsg->matchStartAndRemove( $part1 );
1966 # Check if it is an internal message
1967 $mwInt =& MagicWord
::get( MAG_INT
);
1968 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1969 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1970 $text = $linestart . wfMsgReal( $part1, $args, true );
1978 # Check for NS: (namespace expansion)
1979 $mwNs = MagicWord
::get( MAG_NS
);
1980 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1981 if ( intval( $part1 ) ) {
1982 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1985 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1986 if ( !is_null( $index ) ) {
1987 $text = $linestart . $wgContLang->getNsText( $index );
1994 # LOCALURL and LOCALURLE
1996 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1997 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1999 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
2000 $func = 'getLocalURL';
2001 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
2002 $func = 'escapeLocalURL';
2007 if ( $func !== '' ) {
2008 $title = Title
::newFromText( $part1 );
2009 if ( !is_null( $title ) ) {
2011 $text = $linestart . $title->$func( $args[0] );
2013 $text = $linestart . $title->$func();
2021 if ( !$found && $argc == 1 ) {
2022 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
2023 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
2024 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
2029 # Template table test
2031 # Did we encounter this template already? If yes, it is in the cache
2032 # and we need to check for loops.
2033 if ( !$found && isset( $this->mTemplates
[$part1] ) ) {
2034 # set $text to cached message.
2035 $text = $linestart . $this->mTemplates
[$part1];
2038 # Infinite loop test
2039 if ( isset( $this->mTemplatePath
[$part1] ) ) {
2042 $text .= '<!-- WARNING: template loop detected -->';
2046 # Load from database
2047 $itcamefromthedatabase = false;
2050 $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
2051 if ($subpage !== '') {
2052 $ns = $this->mTitle
->getNamespace();
2054 $title = Title
::newFromText( $part1, $ns );
2055 if ( !is_null( $title ) && !$title->isExternal() ) {
2056 # Check for excessive inclusion
2057 $dbk = $title->getPrefixedDBkey();
2058 if ( $this->incrementIncludeCount( $dbk ) ) {
2059 # This should never be reached.
2060 $article = new Article( $title );
2061 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
2062 if ( $articleContent !== false ) {
2064 $text = $linestart . $articleContent;
2065 $itcamefromthedatabase = true;
2069 # If the title is valid but undisplayable, make a link to it
2070 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
2071 $text = $linestart . '[['.$title->getPrefixedText().']]';
2075 # Template cache array insertion
2076 $this->mTemplates
[$part1] = $text;
2080 # Recursive parsing, escaping and link table handling
2081 # Only for HTML output
2082 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
2083 $text = wfEscapeWikiText( $text );
2084 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
2085 # Clean up argument array
2086 $assocArgs = array();
2088 foreach( $args as $arg ) {
2089 $eqpos = strpos( $arg, '=' );
2090 if ( $eqpos === false ) {
2091 $assocArgs[$index++
] = $arg;
2093 $name = trim( substr( $arg, 0, $eqpos ) );
2094 $value = trim( substr( $arg, $eqpos+
1 ) );
2095 if ( $value === false ) {
2098 if ( $name !== false ) {
2099 $assocArgs[$name] = $value;
2104 # Add a new element to the templace recursion path
2105 $this->mTemplatePath
[$part1] = 1;
2107 $text = $this->strip( $text, $this->mStripState
);
2108 $text = $this->removeHTMLtags( $text );
2109 $text = $this->replaceVariables( $text, $assocArgs );
2111 # Resume the link cache and register the inclusion as a link
2112 if ( $this->mOutputType
== OT_HTML
&& !is_null( $title ) ) {
2113 $wgLinkCache->addLinkObj( $title );
2116 # If the template begins with a table or block-level
2117 # element, it should be treated as beginning a new line.
2118 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
2119 $text = "\n" . $text;
2123 # Empties the template path
2124 $this->mTemplatePath
= array();
2126 wfProfileOut( $fname );
2129 # replace ==section headers==
2130 # XXX this needs to go away once we have a better parser.
2131 if ( $this->mOutputType
!= OT_WIKI
&& $itcamefromthedatabase ) {
2132 if( !is_null( $title ) )
2133 $encodedname = base64_encode($title->getPrefixedDBkey());
2135 $encodedname = base64_encode("");
2136 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
2137 PREG_SPLIT_DELIM_CAPTURE
);
2140 for( $i = 0; $i < count($m); $i +
= 2 ) {
2142 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
2144 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
2148 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
2149 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
2150 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
2157 # Empties the template path
2158 $this->mTemplatePath
= array();
2161 wfProfileOut( $fname );
2164 wfProfileOut( $fname );
2170 * Triple brace replacement -- used for template arguments
2173 function argSubstitution( $matches ) {
2174 $arg = trim( $matches[1] );
2175 $text = $matches[0];
2176 $inputArgs = end( $this->mArgStack
);
2178 if ( array_key_exists( $arg, $inputArgs ) ) {
2179 $text = $inputArgs[$arg];
2186 * Returns true if the function is allowed to include this entity
2189 function incrementIncludeCount( $dbk ) {
2190 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
2191 $this->mIncludeCount
[$dbk] = 0;
2193 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
2202 * Cleans up HTML, removes dangerous tags and attributes, and
2203 * removes HTML comments
2206 function removeHTMLtags( $text ) {
2207 global $wgUseTidy, $wgUserHtml;
2208 $fname = 'Parser::removeHTMLtags';
2209 wfProfileIn( $fname );
2212 $htmlpairs = array( # Tags that must be closed
2213 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
2214 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
2215 'strike', 'strong', 'tt', 'var', 'div', 'center',
2216 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
2217 'ruby', 'rt' , 'rb' , 'rp', 'p', 'span'
2219 $htmlsingle = array(
2220 'br', 'hr', 'li', 'dt', 'dd'
2222 $htmlnest = array( # Tags that can be nested--??
2223 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
2224 'dl', 'font', 'big', 'small', 'sub', 'sup', 'span'
2226 $tabletags = array( # Can only appear inside table
2230 $htmlpairs = array();
2231 $htmlsingle = array();
2232 $htmlnest = array();
2233 $tabletags = array();
2236 $htmlsingle = array_merge( $tabletags, $htmlsingle );
2237 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
2239 $htmlattrs = $this->getHTMLattrs () ;
2241 # Remove HTML comments
2242 $text = $this->removeHTMLcomments( $text );
2244 $bits = explode( '<', $text );
2245 $text = array_shift( $bits );
2247 $tagstack = array(); $tablestack = array();
2248 foreach ( $bits as $x ) {
2249 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
2250 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2252 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2253 error_reporting( $prev );
2256 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2260 if ( ! in_array( $t, $htmlsingle ) &&
2261 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
2262 @array_push
( $tagstack, $ot );
2265 if ( $t == 'table' ) {
2266 $tagstack = array_pop( $tablestack );
2271 # Keep track for later
2272 if ( in_array( $t, $tabletags ) &&
2273 ! in_array( 'table', $tagstack ) ) {
2275 } else if ( in_array( $t, $tagstack ) &&
2276 ! in_array ( $t , $htmlnest ) ) {
2278 } else if ( ! in_array( $t, $htmlsingle ) ) {
2279 if ( $t == 'table' ) {
2280 array_push( $tablestack, $tagstack );
2281 $tagstack = array();
2283 array_push( $tagstack, $t );
2285 # Strip non-approved attributes from the tag
2286 $newparams = $this->fixTagAttributes($params);
2290 $rest = str_replace( '>', '>', $rest );
2291 $text .= "<$slash$t $newparams$brace$rest";
2295 $text .= '<' . str_replace( '>', '>', $x);
2297 # Close off any remaining tags
2298 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2300 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2303 # this might be possible using tidy itself
2304 foreach ( $bits as $x ) {
2305 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2307 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2308 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2309 $newparams = $this->fixTagAttributes($params);
2310 $rest = str_replace( '>', '>', $rest );
2311 $text .= "<$slash$t $newparams$brace$rest";
2313 $text .= '<' . str_replace( '>', '>', $x);
2317 wfProfileOut( $fname );
2322 * Remove '<!--', '-->', and everything between.
2323 * To avoid leaving blank lines, when a comment is both preceded
2324 * and followed by a newline (ignoring spaces), trim leading and
2325 * trailing spaces and one of the newlines.
2329 function removeHTMLcomments( $text ) {
2330 $fname='Parser::removeHTMLcomments';
2331 wfProfileIn( $fname );
2332 while (($start = strpos($text, '<!--')) !== false) {
2333 $end = strpos($text, '-->', $start +
4);
2334 if ($end === false) {
2335 # Unterminated comment; bail out
2341 # Trim space and newline if the comment is both
2342 # preceded and followed by a newline
2343 $spaceStart = max($start - 1, 0);
2344 $spaceLen = $end - $spaceStart;
2345 while (substr($text, $spaceStart, 1) === ' ' && $spaceStart > 0) {
2349 while (substr($text, $spaceStart +
$spaceLen, 1) === ' ')
2351 if (substr($text, $spaceStart, 1) === "\n" and substr($text, $spaceStart +
$spaceLen, 1) === "\n") {
2352 # Remove the comment, leading and trailing
2353 # spaces, and leave only one newline.
2354 $text = substr_replace($text, "\n", $spaceStart, $spaceLen +
1);
2357 # Remove just the comment.
2358 $text = substr_replace($text, '', $start, $end - $start);
2361 wfProfileOut( $fname );
2366 * This function accomplishes several tasks:
2367 * 1) Auto-number headings if that option is enabled
2368 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2369 * 3) Add a Table of contents on the top for users who have enabled the option
2370 * 4) Auto-anchor headings
2372 * It loops through all headlines, collects the necessary data, then splits up the
2373 * string and re-inserts the newly formatted headlines.
2376 /* private */ function formatHeadings( $text, $isMain=true ) {
2377 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders;
2379 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2380 $doShowToc = $this->mOptions
->getShowToc();
2381 $forceTocHere = false;
2382 if( !$this->mTitle
->userCanEdit() ) {
2384 $rightClickHack = 0;
2386 $showEditLink = $this->mOptions
->getEditSection();
2387 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
2390 # Inhibit editsection links if requested in the page
2391 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2392 if( $esw->matchAndRemove( $text ) ) {
2395 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2397 $mw =& MagicWord
::get( MAG_NOTOC
);
2398 if( $mw->matchAndRemove( $text ) ) {
2402 # never add the TOC to the Main Page. This is an entry page that should not
2403 # be more than 1-2 screens large anyway
2404 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
2408 # Get all headlines for numbering them and adding funky stuff like [edit]
2409 # links - this is for later, but we need the number of headlines right now
2410 $numMatches = preg_match_all( '/<H([1-6])(.*?'.'>)(.*?)<\/H[1-6] *>/i', $text, $matches );
2412 # if there are fewer than 4 headlines in the article, do not show TOC
2413 if( $numMatches < 4 ) {
2417 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2418 # override above conditions and always show TOC at that place
2420 $mw =& MagicWord
::get( MAG_TOC
);
2421 if($mw->match( $text ) ) {
2423 $forceTocHere = true;
2425 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2426 # override above conditions and always show TOC above first header
2427 $mw =& MagicWord
::get( MAG_FORCETOC
);
2428 if ($mw->matchAndRemove( $text ) ) {
2433 # Never ever show TOC if no headers
2434 if( $numMatches < 1 ) {
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
2450 $sublevelCount = array();
2451 $levelCount = array();
2458 foreach( $matches[3] as $headline ) {
2460 $templatetitle = '';
2461 $templatesection = 0;
2464 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2466 $templatetitle = base64_decode($mat[1]);
2467 $templatesection = 1 +
(int)base64_decode($mat[2]);
2468 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2472 $prevlevel = $level;
2473 $prevtoclevel = $toclevel;
2475 $level = $matches[1][$headlineCount];
2477 if( $doNumberHeadings ||
$doShowToc ) {
2479 if ( $level > $prevlevel ) {
2480 # Increase TOC level
2482 $sublevelCount[$toclevel] = 0;
2483 $toc .= $sk->tocIndent();
2485 elseif ( $level < $prevlevel && $toclevel > 1 ) {
2486 # Decrease TOC level, find level to jump to
2488 if ( $toclevel == 2 && $level <= $levelCount[1] ) {
2489 # Can only go down to level 1
2492 for ($i = $toclevel; $i > 0; $i--) {
2493 if ( $levelCount[$i] == $level ) {
2494 # Found last matching level
2498 elseif ( $levelCount[$i] < $level ) {
2499 # Found first matching level below current level
2506 $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
2509 # No change in level, end TOC line
2510 $toc .= $sk->tocLineEnd();
2513 $levelCount[$toclevel] = $level;
2515 # count number of headlines for each level
2516 @$sublevelCount[$toclevel]++
;
2518 for( $i = 1; $i <= $toclevel; $i++
) {
2519 if( !empty( $sublevelCount[$i] ) ) {
2523 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2529 # The canonized header is a version of the header text safe to use for links
2530 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2531 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2532 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2534 # Remove link placeholders by the link text.
2535 # <!--LINK number-->
2537 # link text with suffix
2538 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2539 "\$wgLinkHolders['texts'][\$1]",
2540 $canonized_headline );
2543 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2544 $tocline = trim( $canonized_headline );
2545 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2546 $replacearray = array(
2550 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2551 $refer[$headlineCount] = $canonized_headline;
2553 # count how many in assoc. array so we can track dupes in anchors
2554 @$refers[$canonized_headline]++
;
2555 $refcount[$headlineCount]=$refers[$canonized_headline];
2557 # Don't number the heading if it is the only one (looks silly)
2558 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2559 # the two are different if the line contains a link
2560 $headline=$numbering . ' ' . $headline;
2563 # Create the anchor for linking from the TOC to the section
2564 $anchor = $canonized_headline;
2565 if($refcount[$headlineCount] > 1 ) {
2566 $anchor .= '_' . $refcount[$headlineCount];
2568 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2569 $toc .= $sk->tocLine($anchor, $tocline, $numbering, $toclevel);
2571 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2572 if ( empty( $head[$headlineCount] ) ) {
2573 $head[$headlineCount] = '';
2576 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2578 $head[$headlineCount] .= $sk->editSectionLink($this->mTitle
, $sectionCount+
1);
2581 # Add the edit section span
2582 if( $rightClickHack ) {
2584 $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
2586 $headline = $sk->editSectionScript($this->mTitle
, $sectionCount+
1,$headline);
2589 # give headline the correct <h#> tag
2590 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2598 $toclines = $headlineCount;
2599 $toc .= $sk->tocUnindent( $toclevel - 1 );
2600 $toc = $sk->tocList( $toc );
2603 # split up and insert constructed headlines
2605 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2608 foreach( $blocks as $block ) {
2609 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2610 # This is the [edit] link that appears for the top block of text when
2611 # section editing is enabled
2613 # Disabled because it broke block formatting
2614 # For example, a bullet point in the top line
2615 # $full .= $sk->editSectionLink(0);
2618 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2619 # Top anchor now in skin
2623 if( !empty( $head[$i] ) ) {
2629 $mw =& MagicWord
::get( MAG_TOC
);
2630 return $mw->replace( $toc, $full );
2637 * Return an HTML link for the "ISBN 123456" text
2640 function magicISBN( $text ) {
2642 $fname = 'Parser::magicISBN';
2643 wfProfileIn( $fname );
2645 $a = split( 'ISBN ', ' '.$text );
2646 if ( count ( $a ) < 2 ) {
2647 wfProfileOut( $fname );
2650 $text = substr( array_shift( $a ), 1);
2651 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2653 foreach ( $a as $x ) {
2654 $isbn = $blank = '' ;
2655 while ( ' ' == $x{0} ) {
2657 $x = substr( $x, 1 );
2659 if ( $x == '' ) { # blank isbn
2660 $text .= "ISBN $blank";
2663 while ( strstr( $valid, $x{0} ) != false ) {
2665 $x = substr( $x, 1 );
2667 $num = str_replace( '-', '', $isbn );
2668 $num = str_replace( ' ', '', $num );
2671 $text .= "ISBN $blank$x";
2673 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2674 $text .= '<a href="' .
2675 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2676 "\" class=\"internal\">ISBN $isbn</a>";
2680 wfProfileOut( $fname );
2685 * Return an HTML link for the "GEO ..." text
2688 function magicGEO( $text ) {
2689 global $wgLang, $wgUseGeoMode;
2690 $fname = 'Parser::magicGEO';
2691 wfProfileIn( $fname );
2693 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2694 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2695 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2696 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2697 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2698 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2700 $a = split( 'GEO ', ' '.$text );
2701 if ( count ( $a ) < 2 ) {
2702 wfProfileOut( $fname );
2705 $text = substr( array_shift( $a ), 1);
2706 $valid = '0123456789.+-:';
2708 foreach ( $a as $x ) {
2709 $geo = $blank = '' ;
2710 while ( ' ' == $x{0} ) {
2712 $x = substr( $x, 1 );
2714 while ( strstr( $valid, $x{0} ) != false ) {
2716 $x = substr( $x, 1 );
2718 $num = str_replace( '+', '', $geo );
2719 $num = str_replace( ' ', '', $num );
2721 if ( '' == $num ||
count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2722 $text .= "GEO $blank$x";
2724 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2725 $text .= '<a href="' .
2726 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2727 "\" class=\"internal\">GEO $geo</a>";
2731 wfProfileOut( $fname );
2736 * Return an HTML link for the "RFC 1234" text
2738 * @param string $text text to be processed
2740 function magicRFC( $text, $keyword='RFC ', $urlmsg='rfcurl' ) {
2743 $valid = '0123456789';
2746 $a = split( $keyword, ' '.$text );
2747 if ( count ( $a ) < 2 ) {
2750 $text = substr( array_shift( $a ), 1);
2752 /* Check if keyword is preceed by [[.
2753 * This test is made here cause of the array_shift above
2754 * that prevent the test to be done in the foreach.
2756 if ( substr( $text, -2 ) == '[[' ) {
2760 foreach ( $a as $x ) {
2761 /* token might be empty if we have RFC RFC 1234 */
2769 /** remove and save whitespaces in $blank */
2770 while ( $x{0} == ' ' ) {
2772 $x = substr( $x, 1 );
2775 /** remove and save the rfc number in $id */
2776 while ( strstr( $valid, $x{0} ) != false ) {
2778 $x = substr( $x, 1 );
2782 /* call back stripped spaces*/
2783 $text .= $keyword.$blank.$x;
2784 } elseif( $internal ) {
2786 $text .= $keyword.$id.$x;
2788 /* build the external link*/
2789 $url = wfmsg( $urlmsg );
2790 $url = str_replace( '$1', $id, $url);
2791 $sk =& $this->mOptions
->getSkin();
2792 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
2793 $text .= "<a href='{$url}'{$la}>{$keyword}{$id}</a>{$x}";
2796 /* Check if the next RFC keyword is preceed by [[ */
2797 $internal = ( substr($x,-2) == '[[' );
2803 * Transform wiki markup when saving a page by doing \r\n -> \n
2804 * conversion, substitting signatures, {{subst:}} templates, etc.
2806 * @param string $text the text to transform
2807 * @param Title &$title the Title object for the current article
2808 * @param User &$user the User object describing the current user
2809 * @param ParserOptions $options parsing options
2810 * @param bool $clearState whether to clear the parser state first
2811 * @return string the altered wiki markup
2814 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2815 $this->mOptions
= $options;
2816 $this->mTitle
=& $title;
2817 $this->mOutputType
= OT_WIKI
;
2819 if ( $clearState ) {
2820 $this->clearState();
2823 $stripState = false;
2827 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
2828 $text = $this->strip( $text, $stripState, false );
2829 $text = $this->pstPass2( $text, $user );
2830 $text = $this->unstrip( $text, $stripState );
2831 $text = $this->unstripNoWiki( $text, $stripState );
2836 * Pre-save transform helper function
2839 function pstPass2( $text, &$user ) {
2840 global $wgLang, $wgContLang, $wgLocaltimezone;
2842 # Variable replacement
2843 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2844 $text = $this->replaceVariables( $text );
2848 $n = $user->getName();
2849 $k = $user->getOption( 'nickname' );
2850 if ( '' == $k ) { $k = $n; }
2851 if ( isset( $wgLocaltimezone ) ) {
2852 $oldtz = getenv( 'TZ' );
2853 putenv( 'TZ='.$wgLocaltimezone );
2855 /* Note: this is an ugly timezone hack for the European wikis */
2856 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
2857 ' (' . date( 'T' ) . ')';
2858 if ( isset( $wgLocaltimezone ) ) {
2859 putenv( 'TZ='.$oldtzs );
2862 if( $user->getOption( 'fancysig' ) ) {
2865 $sigText = '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]]";
2867 $text = preg_replace( '/~~~~~/', $d, $text );
2868 $text = preg_replace( '/~~~~/', "$sigText $d", $text );
2869 $text = preg_replace( '/~~~/', $sigText, $text );
2871 # Context links: [[|name]] and [[name (context)|]]
2873 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2874 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2875 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2876 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2878 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2879 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2880 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2881 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2883 $t = $this->mTitle
->getText();
2884 if ( preg_match( $conpat, $t, $m ) ) {
2887 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2888 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2889 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2891 if ( '' == $context ) {
2892 $text = preg_replace( $p2, '[[\\1]]', $text );
2894 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2897 # Trim trailing whitespace
2898 # MAG_END (__END__) tag allows for trailing
2899 # whitespace to be deliberately included
2900 $text = rtrim( $text );
2901 $mw =& MagicWord
::get( MAG_END
);
2902 $mw->matchAndRemove( $text );
2908 * Set up some variables which are usually set up in parse()
2909 * so that an external function can call some class members with confidence
2912 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2913 $this->mTitle
=& $title;
2914 $this->mOptions
= $options;
2915 $this->mOutputType
= $outputType;
2916 if ( $clearState ) {
2917 $this->clearState();
2922 * Transform a MediaWiki message by replacing magic variables.
2924 * @param string $text the text to transform
2925 * @param ParserOptions $options options
2926 * @return string the text with variables substituted
2929 function transformMsg( $text, $options ) {
2931 static $executing = false;
2933 # Guard against infinite recursion
2939 $this->mTitle
= $wgTitle;
2940 $this->mOptions
= $options;
2941 $this->mOutputType
= OT_MSG
;
2942 $this->clearState();
2943 $text = $this->replaceVariables( $text );
2950 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2951 * Callback will be called with the text within
2952 * Transform and return the text within
2955 function setHook( $tag, $callback ) {
2956 $oldVal = @$this->mTagHooks
[$tag];
2957 $this->mTagHooks
[$tag] = $callback;
2962 * Replace <!--LINK--> link placeholders with actual links, in the buffer
2963 * Placeholders created in Skin::makeLinkObj()
2964 * Returns an array of links found, indexed by PDBK:
2968 * $options is a bit field, RLH_FOR_UPDATE to select for update
2970 function replaceLinkHolders( &$text, $options = 0 ) {
2971 global $wgUser, $wgLinkCache, $wgUseOldExistenceCheck, $wgLinkHolders;
2972 global $wgInterwikiLinkHolders;
2973 global $outputReplace;
2975 if ( $wgUseOldExistenceCheck ) {
2979 $fname = 'Parser::replaceLinkHolders';
2980 wfProfileIn( $fname );
2985 #if ( !empty( $tmpLinks[0] ) ) { #TODO
2986 if ( !empty( $wgLinkHolders['namespaces'] ) ) {
2987 wfProfileIn( $fname.'-check' );
2988 $dbr =& wfGetDB( DB_SLAVE
);
2989 $page = $dbr->tableName( 'page' );
2990 $sk = $wgUser->getSkin();
2991 $threshold = $wgUser->getOption('stubthreshold');
2994 asort( $wgLinkHolders['namespaces'] );
2998 foreach ( $wgLinkHolders['namespaces'] as $key => $val ) {
3000 $title = $wgLinkHolders['titles'][$key];
3002 # Skip invalid entries.
3003 # Result will be ugly, but prevents crash.
3004 if ( is_null( $title ) ) {
3007 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
3009 # Check if it's in the link cache already
3010 if ( $wgLinkCache->getGoodLinkID( $pdbk ) ) {
3011 $colours[$pdbk] = 1;
3012 } elseif ( $wgLinkCache->isBadLink( $pdbk ) ) {
3013 $colours[$pdbk] = 0;
3015 # Not in the link cache, add it to the query
3016 if ( !isset( $current ) ) {
3020 $query = "SELECT page_id, page_namespace, page_title";
3021 if ( $threshold > 0 ) {
3022 $textTable = $dbr->tableName( 'text' );
3023 $query .= ', LENGTH(old_text) AS page_len, page_is_redirect';
3024 $tables .= ", $textTable";
3025 $join = 'page_latest=old_id AND';
3027 $query .= " FROM $tables WHERE $join (page_namespace=$val AND page_title IN(";
3028 } elseif ( $current != $val ) {
3030 $query .= ")) OR (page_namespace=$val AND page_title IN(";
3035 $query .= $dbr->addQuotes( $wgLinkHolders['dbkeys'][$key] );
3040 if ( $options & RLH_FOR_UPDATE
) {
3041 $query .= ' FOR UPDATE';
3044 $res = $dbr->query( $query, $fname );
3046 # Fetch data and form into an associative array
3047 # non-existent = broken
3050 while ( $s = $dbr->fetchObject($res) ) {
3051 $title = Title
::makeTitle( $s->page_namespace
, $s->page_title
);
3052 $pdbk = $title->getPrefixedDBkey();
3053 $wgLinkCache->addGoodLink( $s->page_id
, $pdbk );
3055 if ( $threshold > 0 ) {
3056 $size = $s->page_len
;
3057 if ( $s->page_is_redirect ||
$s->page_namespace
!= 0 ||
$length < $threshold ) {
3058 $colours[$pdbk] = 1;
3060 $colours[$pdbk] = 2;
3063 $colours[$pdbk] = 1;
3067 wfProfileOut( $fname.'-check' );
3069 # Construct search and replace arrays
3070 wfProfileIn( $fname.'-construct' );
3071 $outputReplace = array();
3072 foreach ( $wgLinkHolders['namespaces'] as $key => $ns ) {
3073 $pdbk = $pdbks[$key];
3074 $searchkey = '<!--LINK '.$key.'-->';
3075 $title = $wgLinkHolders['titles'][$key];
3076 if ( empty( $colours[$pdbk] ) ) {
3077 $wgLinkCache->addBadLink( $pdbk );
3078 $colours[$pdbk] = 0;
3079 $outputReplace[$searchkey] = $sk->makeBrokenLinkObj( $title,
3080 $wgLinkHolders['texts'][$key],
3081 $wgLinkHolders['queries'][$key] );
3082 } elseif ( $colours[$pdbk] == 1 ) {
3083 $outputReplace[$searchkey] = $sk->makeKnownLinkObj( $title,
3084 $wgLinkHolders['texts'][$key],
3085 $wgLinkHolders['queries'][$key] );
3086 } elseif ( $colours[$pdbk] == 2 ) {
3087 $outputReplace[$searchkey] = $sk->makeStubLinkObj( $title,
3088 $wgLinkHolders['texts'][$key],
3089 $wgLinkHolders['queries'][$key] );
3092 wfProfileOut( $fname.'-construct' );
3095 wfProfileIn( $fname.'-replace' );
3097 $text = preg_replace_callback(
3098 '/(<!--LINK .*?-->)/',
3099 "outputReplaceMatches",
3101 wfProfileOut( $fname.'-replace' );
3104 if ( !empty( $wgInterwikiLinkHolders ) ) {
3105 wfProfileIn( $fname.'-interwiki' );
3106 $outputReplace = $wgInterwikiLinkHolders;
3107 $text = preg_replace_callback(
3108 '/<!--IWLINK (.*?)-->/',
3109 "outputReplaceMatches",
3111 wfProfileOut( $fname.'-interwiki' );
3114 wfProfileOut( $fname );
3119 * Renders an image gallery from a text with one line per image.
3120 * text labels may be given by using |-style alternative text. E.g.
3121 * Image:one.jpg|The number "1"
3122 * Image:tree.jpg|A tree
3123 * given as text will return the HTML of a gallery with two images,
3124 * labeled 'The number "1"' and
3127 function renderImageGallery( $text ) {
3128 global $wgLinkCache;
3129 $ig = new ImageGallery();
3130 $ig->setShowBytes( false );
3131 $ig->setShowFilename( false );
3132 $lines = explode( "\n", $text );
3134 foreach ( $lines as $line ) {
3135 # match lines like these:
3136 # Image:someimage.jpg|This is some image
3137 preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
3139 if ( count( $matches ) == 0 ) {
3142 $nt = Title
::newFromURL( $matches[1] );
3143 if( is_null( $nt ) ) {
3144 # Bogus title. Ignore these so we don't bomb out later.
3147 if ( isset( $matches[3] ) ) {
3148 $label = $matches[3];
3153 # FIXME: Use the full wiki parser and add its links
3154 # to the page's links.
3155 $html = $this->mOptions
->mSkin
->formatComment( $label );
3157 $ig->add( Image
::newFromTitle( $nt ), $html );
3158 $wgLinkCache->addImageLinkObj( $nt );
3160 return $ig->toHTML();
3166 * @package MediaWiki
3170 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
3171 var $mCacheTime; # Used in ParserCache
3172 var $mVersion; # Compatibility check
3174 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
3175 $containsOldMagic = false )
3177 $this->mText
= $text;
3178 $this->mLanguageLinks
= $languageLinks;
3179 $this->mCategoryLinks
= $categoryLinks;
3180 $this->mContainsOldMagic
= $containsOldMagic;
3181 $this->mCacheTime
= '';
3182 $this->mVersion
= MW_PARSER_VERSION
;
3185 function getText() { return $this->mText
; }
3186 function getLanguageLinks() { return $this->mLanguageLinks
; }
3187 function getCategoryLinks() { return array_keys( $this->mCategoryLinks
); }
3188 function getCacheTime() { return $this->mCacheTime
; }
3189 function containsOldMagic() { return $this->mContainsOldMagic
; }
3190 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
3191 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
3192 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
3193 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
3194 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
3195 function addCategoryLink( $c ) { $this->mCategoryLinks
[$c] = 1; }
3197 function merge( $other ) {
3198 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
3199 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
3200 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
3204 * Return true if this cached output object predates the global or
3205 * per-article cache invalidation timestamps, or if it comes from
3206 * an incompatible older version.
3208 * @param string $touched the affected article's last touched timestamp
3212 function expired( $touched ) {
3213 global $wgCacheEpoch;
3214 return $this->getCacheTime() <= $touched ||
3215 $this->getCacheTime() <= $wgCacheEpoch ||
3216 !isset( $this->mVersion
) ||
3217 version_compare( $this->mVersion
, MW_PARSER_VERSION
, "lt" );
3222 * Set options of the Parser
3224 * @package MediaWiki
3228 # All variables are private
3229 var $mUseTeX; # Use texvc to expand <math> tags
3230 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
3231 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
3232 var $mAllowExternalImages; # Allow external images inline
3233 var $mSkin; # Reference to the preferred skin
3234 var $mDateFormat; # Date format index
3235 var $mEditSection; # Create "edit section" links
3236 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
3237 var $mNumberHeadings; # Automatically number headings
3238 var $mShowToc; # Show table of contents
3240 function getUseTeX() { return $this->mUseTeX
; }
3241 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
3242 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
3243 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
3244 function getSkin() { return $this->mSkin
; }
3245 function getDateFormat() { return $this->mDateFormat
; }
3246 function getEditSection() { return $this->mEditSection
; }
3247 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
3248 function getNumberHeadings() { return $this->mNumberHeadings
; }
3249 function getShowToc() { return $this->mShowToc
; }
3251 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
3252 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
3253 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
3254 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
3255 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
3256 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
3257 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
3258 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
3259 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
3261 function setSkin( &$x ) { $this->mSkin
=& $x; }
3263 # Get parser options
3264 /* static */ function newFromUser( &$user ) {
3265 $popts = new ParserOptions
;
3266 $popts->initialiseFromUser( $user );
3271 function initialiseFromUser( &$userInput ) {
3272 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
3273 $fname = 'ParserOptions::initialiseFromUser';
3274 wfProfileIn( $fname );
3275 if ( !$userInput ) {
3277 $user->setLoaded( true );
3279 $user =& $userInput;
3282 $this->mUseTeX
= $wgUseTeX;
3283 $this->mUseDynamicDates
= $wgUseDynamicDates;
3284 $this->mInterwikiMagic
= $wgInterwikiMagic;
3285 $this->mAllowExternalImages
= $wgAllowExternalImages;
3286 wfProfileIn( $fname.'-skin' );
3287 $this->mSkin
=& $user->getSkin();
3288 wfProfileOut( $fname.'-skin' );
3289 $this->mDateFormat
= $user->getOption( 'date' );
3290 $this->mEditSection
= $user->getOption( 'editsection' );
3291 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
3292 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
3293 $this->mShowToc
= $user->getOption( 'showtoc' );
3294 wfProfileOut( $fname );
3301 * Callback function used by Parser::replaceLinkHolders()
3302 * to substitute link placeholders.
3304 function &outputReplaceMatches( $matches ) {
3305 global $outputReplace;
3306 return $outputReplace[$matches[1]];
3310 * Return the total number of articles
3312 function wfNumberOfArticles() {
3313 global $wgNumberOfArticles;
3316 return $wgNumberOfArticles;
3320 * Get various statistics from the database
3323 function wfLoadSiteStats() {
3324 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
3325 $fname = 'wfLoadSiteStats';
3327 if ( -1 != $wgNumberOfArticles ) return;
3328 $dbr =& wfGetDB( DB_SLAVE
);
3329 $s = $dbr->selectRow( 'site_stats',
3330 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
3331 array( 'ss_row_id' => 1 ), $fname
3334 if ( $s === false ) {
3337 $wgTotalViews = $s->ss_total_views
;
3338 $wgTotalEdits = $s->ss_total_edits
;
3339 $wgNumberOfArticles = $s->ss_good_articles
;
3343 function wfEscapeHTMLTagsOnly( $in ) {
3345 array( '"', '>', '<' ),
3346 array( '"', '>', '<' ),