4 * File for Parser and related classes
11 * Variable substitution O(N^2) attack
13 * Without countermeasures, it would be possible to attack the parser by saving
14 * a page filled with a large number of inclusions of large pages. The size of
15 * the generated page would be proportional to the square of the input size.
16 * Hence, we limit the number of inclusions of any given page, thus bringing any
17 * attack back to O(N).
20 define( 'MAX_INCLUDE_REPEAT', 100 );
21 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
23 define( 'RLH_FOR_UPDATE', 1 );
25 # Allowed values for $mOutputType
26 define( 'OT_HTML', 1 );
27 define( 'OT_WIKI', 2 );
28 define( 'OT_MSG' , 3 );
30 # string parameter for extractTags which will cause it
31 # to strip HTML comments in addition to regular
32 # <XML>-style tags. This should not be anything we
33 # may want to use in wikisyntax
34 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
36 # prefix for escaping, used in two functions at least
37 define( 'UNIQ_PREFIX', 'NaodW29');
39 # Constants needed for external link processing
40 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
41 define( 'HTTP_PROTOCOLS', 'http|https' );
42 # Everything except bracket, space, or control characters
43 define( 'EXT_LINK_URL_CLASS', '[^]<>"\\x00-\\x20\\x7F]' );
45 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
46 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
47 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
48 define( 'EXT_LINK_BRACKETED', '/\[(('.URL_PROTOCOLS
.'):'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
49 define( 'EXT_IMAGE_REGEX',
50 '/^('.HTTP_PROTOCOLS
.':)'. # Protocol
51 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
52 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
58 * Processes wiki markup
61 * There are three main entry points into the Parser class:
63 * produces HTML output
65 * produces altered wiki markup.
67 * performs brace substitution on MediaWiki messages
70 * objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
72 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
75 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
76 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
79 * * only within ParserOptions
92 # Cleared with clearState():
93 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
94 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
97 var $mOptions, $mTitle, $mOutputType,
98 $mTemplates, // cache of already loaded templates, avoids
99 // multiple SQL queries for the same string
100 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
101 // in this path. Used for loop detection.
111 $this->mTemplates
= array();
112 $this->mTemplatePath
= array();
113 $this->mTagHooks
= array();
122 function clearState() {
123 $this->mOutput
= new ParserOutput
;
124 $this->mAutonumber
= 0;
125 $this->mLastSection
= "";
126 $this->mDTopen
= false;
127 $this->mVariables
= false;
128 $this->mIncludeCount
= array();
129 $this->mStripState
= array();
130 $this->mArgStack
= array();
131 $this->mInPre
= false;
135 * First pass--just handle <nowiki> sections, pass the rest off
136 * to internalParse() which does all the real work.
139 * @return ParserOutput a ParserOutput
141 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
142 global $wgUseTidy, $wgContLang;
143 $fname = 'Parser::parse';
144 wfProfileIn( $fname );
150 $this->mOptions
= $options;
151 $this->mTitle
=& $title;
152 $this->mOutputType
= OT_HTML
;
155 $text = $this->strip( $text, $this->mStripState
);
157 $text = $this->internalParse( $text, $linestart );
158 $text = $this->unstrip( $text, $this->mStripState
);
159 # Clean up special characters, only run once, next-to-last before doBlockLevels
162 # french spaces, last one Guillemet-left
163 # only if there is something before the space
164 '/(.) (?=\\?|:|;|!|\\302\\273)/i' => '\\1 \\2',
165 # french spaces, Guillemet-right
166 "/(\\302\\253) /i"=>"\\1 ",
167 '/<hr *>/i' => '<hr />',
168 '/<br *>/i' => '<br />',
169 '/<center *>/i' => '<div class="center">',
170 '/<\\/center *>/i' => '</div>',
171 # Clean up spare ampersands; note that we probably ought to be
172 # more careful about named entities.
173 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
175 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
178 # french spaces, last one Guillemet-left
179 '/ (\\?|:|;|!|\\302\\273)/i' => ' \\1',
180 # french spaces, Guillemet-right
181 '/(\\302\\253) /i' => '\\1 ',
182 '/<center *>/i' => '<div class="center">',
183 '/<\\/center *>/i' => '</div>'
185 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
188 $text = $this->doBlockLevels( $text, $linestart );
190 $this->replaceLinkHolders( $text );
191 $text = $wgContLang->convert($text);
193 $text = $this->unstripNoWiki( $text, $this->mStripState
);
196 $text = Parser
::tidy($text);
199 $this->mOutput
->setText( $text );
200 wfProfileOut( $fname );
201 return $this->mOutput
;
205 * Get a random string
210 function getRandomString() {
211 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
215 * Replaces all occurrences of <$tag>content</$tag> in the text
216 * with a random marker and returns the new text. the output parameter
217 * $content will be an associative array filled with data on the form
218 * $unique_marker => content.
220 * If $content is already set, the additional entries will be appended
221 * If $tag is set to STRIP_COMMENTS, the function will extract
222 * <!-- HTML comments -->
227 function extractTags($tag, $text, &$content, $uniq_prefix = ''){
228 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
235 while ( '' != $text ) {
236 if($tag==STRIP_COMMENTS
) {
237 $p = preg_split( '/<!--/i', $text, 2 );
239 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
242 if ( ( count( $p ) < 2 ) ||
( '' == $p[1] ) ) {
245 if($tag==STRIP_COMMENTS
) {
246 $q = preg_split( '/-->/i', $p[1], 2 );
248 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
250 $marker = $rnd . sprintf('%08X', $n++
);
251 $content[$marker] = $q[0];
252 $stripped .= $marker;
260 * Strips and renders nowiki, pre, math, hiero
261 * If $render is set, performs necessary rendering operations on plugins
262 * Returns the text, and fills an array with data needed in unstrip()
263 * If the $state is already a valid strip state, it adds to the state
265 * @param bool $stripcomments when set, HTML comments <!-- like this -->
266 * will be stripped in addition to other tags. This is important
267 * for section editing, where these comments cause confusion when
268 * counting the sections in the wikisource
272 function strip( $text, &$state, $stripcomments = false ) {
273 $render = ($this->mOutputType
== OT_HTML
);
274 $html_content = array();
275 $nowiki_content = array();
276 $math_content = array();
277 $pre_content = array();
278 $comment_content = array();
279 $ext_content = array();
281 # Replace any instances of the placeholders
282 $uniq_prefix = UNIQ_PREFIX
;
283 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
286 global $wgRawHtml, $wgWhitelistEdit;
287 if( $wgRawHtml && $wgWhitelistEdit ) {
288 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
289 foreach( $html_content as $marker => $content ) {
291 # Raw and unchecked for validity.
292 $html_content[$marker] = $content;
294 $html_content[$marker] = '<html>'.$content.'</html>';
300 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
301 foreach( $nowiki_content as $marker => $content ) {
303 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
305 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
310 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
311 foreach( $math_content as $marker => $content ){
313 if( $this->mOptions
->getUseTeX() ) {
314 $math_content[$marker] = renderMath( $content );
316 $math_content[$marker] = '<math>'.$content.'<math>';
319 $math_content[$marker] = '<math>'.$content.'</math>';
324 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
325 foreach( $pre_content as $marker => $content ){
327 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
329 $pre_content[$marker] = '<pre>'.$content.'</pre>';
335 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
336 foreach( $comment_content as $marker => $content ){
337 $comment_content[$marker] = '<!--'.$content.'-->';
342 foreach ( $this->mTagHooks
as $tag => $callback ) {
343 $ext_contents[$tag] = array();
344 $text = Parser
::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
345 foreach( $ext_content[$tag] as $marker => $content ) {
347 $ext_content[$tag][$marker] = $callback( $content );
349 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
354 # Merge state with the pre-existing state, if there is one
356 $state['html'] = $state['html'] +
$html_content;
357 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
358 $state['math'] = $state['math'] +
$math_content;
359 $state['pre'] = $state['pre'] +
$pre_content;
360 $state['comment'] = $state['comment'] +
$comment_content;
362 foreach( $ext_content as $tag => $array ) {
363 if ( array_key_exists( $tag, $state ) ) {
364 $state[$tag] = $state[$tag] +
$array;
369 'html' => $html_content,
370 'nowiki' => $nowiki_content,
371 'math' => $math_content,
372 'pre' => $pre_content,
373 'comment' => $comment_content,
380 * restores pre, math, and heiro removed by strip()
382 * always call unstripNoWiki() after this one
385 function unstrip( $text, &$state ) {
386 # Must expand in reverse order, otherwise nested tags will be corrupted
387 $contentDict = end( $state );
388 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
389 if( key($state) != 'nowiki' && key($state) != 'html') {
390 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
391 $text = str_replace( key( $contentDict ), $content, $text );
400 * always call this after unstrip() to preserve the order
404 function unstripNoWiki( $text, &$state ) {
405 # Must expand in reverse order, otherwise nested tags will be corrupted
406 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
407 $text = str_replace( key( $state['nowiki'] ), $content, $text );
412 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
413 $text = str_replace( key( $state['html'] ), $content, $text );
421 * Add an item to the strip state
422 * Returns the unique tag which must be inserted into the stripped text
423 * The tag will be replaced with the original text in unstrip()
427 function insertStripItem( $text, &$state ) {
428 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
437 $state['item'][$rnd] = $text;
442 * Return allowed HTML attributes
446 function getHTMLattrs () {
447 $htmlattrs = array( # Allowed attributes--no scripting, etc.
448 'title', 'align', 'lang', 'dir', 'width', 'height',
449 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
450 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
451 /* FONT */ 'type', 'start', 'value', 'compact',
452 /* For various lists, mostly deprecated but safe */
453 'summary', 'width', 'border', 'frame', 'rules',
454 'cellspacing', 'cellpadding', 'valign', 'char',
455 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
456 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
457 'id', 'class', 'name', 'style' /* For CSS */
463 * Remove non approved attributes and javascript in css
467 function fixTagAttributes ( $t ) {
468 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
469 $htmlattrs = $this->getHTMLattrs() ;
471 # Strip non-approved attributes from the tag
473 '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
474 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
477 $t = str_replace ( '<></>' , '' , $t ) ; # This should fix bug 980557
479 # Strip javascript "expression" from stylesheets. Brute force approach:
480 # If anythin offensive is found, all attributes of the HTML tag are dropped
483 '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
484 wfMungeToUtf8( $t ) ) )
493 * interface with html tidy, used if $wgUseTidy = true
498 function tidy ( $text ) {
499 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
500 global $wgInputEncoding, $wgOutputEncoding;
501 $fname = 'Parser::tidy';
502 wfProfileIn( $fname );
506 switch(strtoupper($wgOutputEncoding)) {
508 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -latin1':' -raw';
511 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -utf8':' -raw';
517 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
518 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
519 '<head><title>test</title></head><body>'.$text.'</body></html>';
520 $descriptorspec = array(
521 0 => array('pipe', 'r'),
522 1 => array('pipe', 'w'),
523 2 => array('file', '/dev/null', 'a')
525 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
526 if (is_resource($process)) {
527 fwrite($pipes[0], $wrappedtext);
529 while (!feof($pipes[1])) {
530 $cleansource .= fgets($pipes[1], 1024);
533 $return_value = proc_close($process);
536 wfProfileOut( $fname );
538 if( $cleansource == '' && $text != '') {
539 wfDebug( "Tidy error detected!\n" );
540 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
547 * parse the wiki syntax used to render tables
551 function doTableStuff ( $t ) {
552 $fname = 'Parser::doTableStuff';
553 wfProfileIn( $fname );
555 $t = explode ( "\n" , $t ) ;
556 $td = array () ; # Is currently a td tag open?
557 $ltd = array () ; # Was it TD or TH?
558 $tr = array () ; # Is currently a tr tag open?
559 $ltr = array () ; # tr attributes
560 $indent_level = 0; # indent level of the table
561 foreach ( $t AS $k => $x )
564 $fc = substr ( $x , 0 , 1 ) ;
565 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
566 $indent_level = strlen( $matches[1] );
568 str_repeat( '<dl><dd>', $indent_level ) .
569 '<table ' . $this->fixTagAttributes ( $matches[2] ) . '>' ;
570 array_push ( $td , false ) ;
571 array_push ( $ltd , '' ) ;
572 array_push ( $tr , false ) ;
573 array_push ( $ltr , '' ) ;
575 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
576 else if ( '|}' == substr ( $x , 0 , 2 ) ) {
578 $l = array_pop ( $ltd ) ;
579 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
580 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
582 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
584 else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
585 $x = substr ( $x , 1 ) ;
586 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
588 $l = array_pop ( $ltd ) ;
589 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
590 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
593 array_push ( $tr , false ) ;
594 array_push ( $td , false ) ;
595 array_push ( $ltd , '' ) ;
596 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
598 else if ( '|' == $fc ||
'!' == $fc ||
'|+' == substr ( $x , 0 , 2 ) ) { # Caption
600 if ( '|+' == substr ( $x , 0 , 2 ) ) {
602 $x = substr ( $x , 1 ) ;
604 $after = substr ( $x , 1 ) ;
605 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
606 $after = explode ( '||' , $after ) ;
609 # Loop through each table cell
610 foreach ( $after AS $theline )
615 $tra = array_pop ( $ltr ) ;
616 if ( !array_pop ( $tr ) ) $z = '<tr '.$tra.">\n" ;
617 array_push ( $tr , true ) ;
618 array_push ( $ltr , '' ) ;
621 $l = array_pop ( $ltd ) ;
622 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
623 if ( $fc == '|' ) $l = 'td' ;
624 else if ( $fc == '!' ) $l = 'th' ;
625 else if ( $fc == '+' ) $l = 'caption' ;
627 array_push ( $ltd , $l ) ;
630 $y = explode ( '|' , $theline , 2 ) ;
631 # Note that a '|' inside an invalid link should not
632 # be mistaken as delimiting cell parameters
633 if ( strpos( $y[0], '[[' ) !== false ) {
634 $y = array ($theline);
636 if ( count ( $y ) == 1 )
637 $y = "{$z}<{$l}>{$y[0]}" ;
638 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
640 array_push ( $td , true ) ;
645 # Closing open td, tr && table
646 while ( count ( $td ) > 0 )
648 if ( array_pop ( $td ) ) $t[] = '</td>' ;
649 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
653 $t = implode ( "\n" , $t ) ;
654 # $t = $this->removeHTMLtags( $t );
655 wfProfileOut( $fname );
660 * Helper function for parse() that transforms wiki markup into
661 * HTML. Only called for $mOutputType == OT_HTML.
665 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
668 $fname = 'Parser::internalParse';
669 wfProfileIn( $fname );
671 $text = $this->removeHTMLtags( $text );
672 $text = $this->replaceVariables( $text, $args );
674 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
676 $text = $this->doHeadings( $text );
677 if($this->mOptions
->getUseDynamicDates()) {
678 global $wgDateFormatter;
679 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
681 $text = $this->doAllQuotes( $text );
682 $text = $this->replaceInternalLinks ( $text );
683 $text = $this->replaceExternalLinks( $text );
684 $text = $this->doMagicLinks( $text );
685 $text = $this->doTableStuff( $text );
686 $text = $this->formatHeadings( $text, $isMain );
687 $sk =& $this->mOptions
->getSkin();
688 $text = $sk->transformContent( $text );
690 wfProfileOut( $fname );
695 * Replace special strings like "ISBN xxx" and "RFC xxx" with
696 * magic external links.
700 function &doMagicLinks( &$text ) {
701 global $wgUseGeoMode;
702 $text = $this->magicISBN( $text );
703 if ( isset( $wgUseGeoMode ) && $wgUseGeoMode ) {
704 $text = $this->magicGEO( $text );
706 $text = $this->magicRFC( $text );
711 * Parse ^^ tokens and return html
715 function doExponent ( $text ) {
716 $fname = 'Parser::doExponent';
717 wfProfileIn( $fname);
718 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
719 wfProfileOut( $fname);
724 * Parse headers and return html
728 function doHeadings( $text ) {
729 $fname = 'Parser::doHeadings';
730 wfProfileIn( $fname );
731 for ( $i = 6; $i >= 1; --$i ) {
732 $h = substr( '======', 0, $i );
733 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
734 "<h{$i}>\\1</h{$i}>\\2", $text );
736 wfProfileOut( $fname );
741 * Replace single quotes with HTML markup
743 * @return string the altered text
745 function doAllQuotes( $text ) {
746 $fname = 'Parser::doAllQuotes';
747 wfProfileIn( $fname );
749 $lines = explode( "\n", $text );
750 foreach ( $lines as $line ) {
751 $outtext .= $this->doQuotes ( $line ) . "\n";
753 $outtext = substr($outtext, 0,-1);
754 wfProfileOut( $fname );
759 * Helper function for doAllQuotes()
762 function doQuotes( $text ) {
763 $arr = preg_split ("/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
764 if (count ($arr) == 1)
768 # First, do some preliminary work. This may shift some apostrophes from
769 # being mark-up to being text. It also counts the number of occurrences
770 # of bold and italics mark-ups.
778 # If there are ever four apostrophes, assume the first is supposed to
779 # be text, and the remaining three constitute mark-up for bold text.
780 if (strlen ($arr[$i]) == 4)
785 # If there are more than 5 apostrophes in a row, assume they're all
786 # text except for the last 5.
787 else if (strlen ($arr[$i]) > 5)
789 $arr[$i-1] .= str_repeat ("'", strlen ($arr[$i]) - 5);
792 # Count the number of occurrences of bold and italics mark-ups.
793 # We are not counting sequences of five apostrophes.
794 if (strlen ($arr[$i]) == 2) $numitalics++
; else
795 if (strlen ($arr[$i]) == 3) $numbold++
; else
796 if (strlen ($arr[$i]) == 5) { $numitalics++
; $numbold++
; }
801 # If there is an odd number of both bold and italics, it is likely
802 # that one of the bold ones was meant to be an apostrophe followed
803 # by italics. Which one we cannot know for certain, but it is more
804 # likely to be one that has a single-letter word before it.
805 if (($numbold %
2 == 1) && ($numitalics %
2 == 1))
808 $firstsingleletterword = -1;
809 $firstmultiletterword = -1;
813 if (($i %
2 == 1) and (strlen ($r) == 3))
815 $x1 = substr ($arr[$i-1], -1);
816 $x2 = substr ($arr[$i-1], -2, 1);
818 if ($firstspace == -1) $firstspace = $i;
819 } else if ($x2 == ' ') {
820 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
822 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
828 # If there is a single-letter word, use it!
829 if ($firstsingleletterword > -1)
831 $arr [ $firstsingleletterword ] = "''";
832 $arr [ $firstsingleletterword-1 ] .= "'";
834 # If not, but there's a multi-letter word, use that one.
835 else if ($firstmultiletterword > -1)
837 $arr [ $firstmultiletterword ] = "''";
838 $arr [ $firstmultiletterword-1 ] .= "'";
840 # ... otherwise use the first one that has neither.
841 # (notice that it is possible for all three to be -1 if, for example,
842 # there is only one pentuple-apostrophe in the line)
843 else if ($firstspace > -1)
845 $arr [ $firstspace ] = "''";
846 $arr [ $firstspace-1 ] .= "'";
850 # Now let's actually convert our apostrophic mush to HTML!
859 if ($state == 'both')
866 if (strlen ($r) == 2)
869 { $output .= '</i>'; $state = ''; }
870 else if ($state == 'bi')
871 { $output .= '</i>'; $state = 'b'; }
872 else if ($state == 'ib')
873 { $output .= '</b></i><b>'; $state = 'b'; }
874 else if ($state == 'both')
875 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
876 else # $state can be 'b' or ''
877 { $output .= '<i>'; $state .= 'i'; }
879 else if (strlen ($r) == 3)
882 { $output .= '</b>'; $state = ''; }
883 else if ($state == 'bi')
884 { $output .= '</i></b><i>'; $state = 'i'; }
885 else if ($state == 'ib')
886 { $output .= '</b>'; $state = 'i'; }
887 else if ($state == 'both')
888 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
889 else # $state can be 'i' or ''
890 { $output .= '<b>'; $state .= 'b'; }
892 else if (strlen ($r) == 5)
895 { $output .= '</b><i>'; $state = 'i'; }
896 else if ($state == 'i')
897 { $output .= '</i><b>'; $state = 'b'; }
898 else if ($state == 'bi')
899 { $output .= '</i></b>'; $state = ''; }
900 else if ($state == 'ib')
901 { $output .= '</b></i>'; $state = ''; }
902 else if ($state == 'both')
903 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
904 else # ($state == '')
905 { $buffer = ''; $state = 'both'; }
910 # Now close all remaining tags. Notice that the order is important.
911 if ($state == 'b' ||
$state == 'ib')
913 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
917 if ($state == 'both')
918 $output .= '<b><i>'.$buffer.'</i></b>';
924 * Replace external links
926 * Note: we have to do external links before the internal ones,
927 * and otherwise take great care in the order of things here, so
928 * that we don't end up interpreting some URLs twice.
932 function replaceExternalLinks( $text ) {
933 $fname = 'Parser::replaceExternalLinks';
934 wfProfileIn( $fname );
936 $sk =& $this->mOptions
->getSkin();
937 $linktrail = wfMsgForContent('linktrail');
938 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
940 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
943 while ( $i<count( $bits ) ) {
945 $protocol = $bits[$i++
];
947 $trail = $bits[$i++
];
949 # The characters '<' and '>' (which were escaped by
950 # removeHTMLtags()) should not be included in
951 # URLs, per RFC 2396.
952 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
953 $text = substr($url, $m2[0][1]) . ' ' . $text;
954 $url = substr($url, 0, $m2[0][1]);
957 # If the link text is an image URL, replace it with an <img> tag
958 # This happened by accident in the original parser, but some people used it extensively
959 $img = $this->maybeMakeImageLink( $text );
960 if ( $img !== false ) {
966 # No link text, e.g. [http://domain.tld/some.link]
968 # Autonumber if allowed
969 if ( strpos( HTTP_PROTOCOLS
, $protocol ) !== false ) {
970 $text = '[' . ++
$this->mAutonumber
. ']';
972 # Otherwise just use the URL
973 $text = htmlspecialchars( $url );
976 # Have link text, e.g. [http://domain.tld/some.link text]s
978 if ( preg_match( $linktrail, $trail, $m2 ) ) {
984 $encUrl = htmlspecialchars( $url );
985 # Bit in parentheses showing the URL for the printable version
986 if( $url == $text ||
preg_match( "!$protocol://" . preg_quote( $text, '/' ) . "/?$!", $url ) ) {
989 # Expand the URL for printable version
990 if ( ! $sk->suppressUrlExpansion() ) {
991 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $encUrl ) . "</i>)</span>";
997 # Process the trail (i.e. everything after this link up until start of the next link),
998 # replacing any non-bracketed links
999 $trail = $this->replaceFreeExternalLinks( $trail );
1001 # Use the encoded URL
1002 # This means that users can paste URLs directly into the text
1003 # Funny characters like ö aren't valid in URLs anyway
1004 # This was changed in August 2004
1005 $s .= $sk->makeExternalLink( $url, $text, false ) . $dtrail. $paren . $trail;
1008 wfProfileOut( $fname );
1013 * Replace anything that looks like a URL with a link
1016 function replaceFreeExternalLinks( $text ) {
1017 $bits = preg_split( '/((?:'.URL_PROTOCOLS
.'):)/', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1018 $s = array_shift( $bits );
1021 $sk =& $this->mOptions
->getSkin();
1023 while ( $i < count( $bits ) ){
1024 $protocol = $bits[$i++
];
1025 $remainder = $bits[$i++
];
1027 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1028 # Found some characters after the protocol that look promising
1029 $url = $protocol . $m[1];
1032 # The characters '<' and '>' (which were escaped by
1033 # removeHTMLtags()) should not be included in
1034 # URLs, per RFC 2396.
1035 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1036 $trail = substr($url, $m2[0][1]) . $trail;
1037 $url = substr($url, 0, $m2[0][1]);
1040 # Move trailing punctuation to $trail
1042 # If there is no left bracket, then consider right brackets fair game too
1043 if ( strpos( $url, '(' ) === false ) {
1047 $numSepChars = strspn( strrev( $url ), $sep );
1048 if ( $numSepChars ) {
1049 $trail = substr( $url, -$numSepChars ) . $trail;
1050 $url = substr( $url, 0, -$numSepChars );
1053 # Replace & from obsolete syntax with &.
1054 # All HTML entities will be escaped by makeExternalLink()
1055 # or maybeMakeImageLink()
1056 $url = str_replace( '&', '&', $url );
1058 # Is this an external image?
1059 $text = $this->maybeMakeImageLink( $url );
1060 if ( $text === false ) {
1061 # Not an image, make a link
1062 $text = $sk->makeExternalLink( $url, $url );
1064 $s .= $text . $trail;
1066 $s .= $protocol . $remainder;
1073 * make an image if it's allowed
1076 function maybeMakeImageLink( $url ) {
1077 $sk =& $this->mOptions
->getSkin();
1079 if ( $this->mOptions
->getAllowExternalImages() ) {
1080 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1082 $text = $sk->makeImage( htmlspecialchars( $url ) );
1089 * Process [[ ]] wikilinks
1094 function replaceInternalLinks( $s ) {
1095 global $wgLang, $wgContLang, $wgLinkCache;
1096 global $wgDisableLangConversion;
1097 static $fname = 'Parser::replaceInternalLinks' ;
1098 # use a counter to prevent too much unknown links from
1099 # being checked for different language variants.
1100 wfProfileIn( $fname );
1102 wfProfileIn( $fname.'-setup' );
1104 # the % is needed to support urlencoded titles as well
1105 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1107 $sk =& $this->mOptions
->getSkin();
1108 global $wgUseOldExistenceCheck;
1109 # "Post-parse link colour check" works only on wiki text since it's now
1110 # in Parser. Enable it, then disable it when we're done.
1111 $saveParseColour = $sk->postParseLinkColour( !$wgUseOldExistenceCheck );
1113 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1115 #split the entire text string on occurences of [[
1116 $a = explode( '[[', ' ' . $s );
1117 #get the first element (all text up to first [[), and remove the space we added
1118 $s = array_shift( $a );
1119 $s = substr( $s, 1 );
1121 # Match a link having the form [[namespace:link|alternate]]trail
1123 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1124 # Match cases where there is no "]]", which might still be images
1125 static $e1_img = FALSE;
1126 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1127 # Match the end of a line for a word that's not followed by whitespace,
1128 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1129 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1131 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1133 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1135 if ( $useLinkPrefixExtension ) {
1136 if ( preg_match( $e2, $s, $m ) ) {
1137 $first_prefix = $m[2];
1140 $first_prefix = false;
1146 wfProfileOut( $fname.'-setup' );
1148 $checkVariantLink = sizeof($wgContLang->getVariants())>1;
1149 # Loop for each link
1150 for ($k = 0; isset( $a[$k] ); $k++
) {
1152 wfProfileIn( $fname.'-prefixhandling' );
1153 if ( $useLinkPrefixExtension ) {
1154 if ( preg_match( $e2, $s, $m ) ) {
1162 $prefix = $first_prefix;
1163 $first_prefix = false;
1166 wfProfileOut( $fname.'-prefixhandling' );
1168 $might_be_img = false;
1170 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1172 # fix up urlencoded title texts
1173 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1175 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1176 $might_be_img = true;
1178 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1180 } else { # Invalid form; output directly
1181 $s .= $prefix . '[[' . $line ;
1185 # Don't allow internal links to pages containing
1186 # PROTO: where PROTO is a valid URL protocol; these
1187 # should be external links.
1188 if (preg_match('/((?:'.URL_PROTOCOLS
.'):)/', $m[1])) {
1189 $s .= $prefix . '[[' . $line ;
1193 # Make subpage if necessary
1194 $link = $this->maybeDoSubpageLink( $m[1], $text );
1196 $noforce = (substr($m[1], 0, 1) != ':');
1198 # Strip off leading ':'
1199 $link = substr($link, 1);
1202 $nt = Title
::newFromText( $this->unstripNoWiki($link, $this->mStripState
) );
1204 $s .= $prefix . '[[' . $line;
1208 #check other language variants of the link
1209 #if the article does not exist
1210 if( $nt->getArticleID() == 0
1211 && $checkVariantLink ) {
1212 $wgContLang->findVariantLink($link, $nt);
1215 $ns = $nt->getNamespace();
1216 $iw = $nt->getInterWiki();
1218 if ($might_be_img) { # if this is actually an invalid link
1219 if ($ns == NS_IMAGE
&& $noforce) { #but might be an image
1221 while (isset ($a[$k+
1]) ) {
1222 #look at the next 'line' to see if we can close it there
1223 $next_line = array_shift(array_splice( $a, $k +
1, 1) );
1224 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1225 # the first ]] closes the inner link, the second the image
1227 $text .= '[[' . $m[1];
1230 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1231 #if there's exactly one ]] that's fine, we'll keep looking
1232 $text .= '[[' . $m[0];
1234 #if $next_line is invalid too, we need look no further
1235 $text .= '[[' . $next_line;
1240 # we couldn't find the end of this imageLink, so output it raw
1241 #but don't ignore what might be perfectly normal links in the text we've examined
1242 $text = $this->replaceInternalLinks($text);
1243 $s .= $prefix . '[[' . $link . '|' . $text;
1244 # note: no $trail, because without an end, there *is* no trail
1247 } else { #it's not an image, so output it raw
1248 $s .= $prefix . '[[' . $link . '|' . $text;
1249 # note: no $trail, because without an end, there *is* no trail
1254 $wasblank = ( '' == $text );
1255 if( $wasblank ) $text = $link;
1258 # Link not escaped by : , create the various objects
1262 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1263 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1264 $tmp = $prefix . $trail ;
1265 $s .= (trim($tmp) == '')?
'': $tmp;
1269 if ( $ns == NS_IMAGE
) {
1270 # recursively parse links inside the image caption
1271 # actually, this will parse them in any other parameters, too,
1272 # but it might be hard to fix that, and it doesn't matter ATM
1273 $text = $this->replaceExternalLinks($text);
1274 $text = $this->replaceInternalLinks($text);
1276 # replace the image with a link-holder so that replaceExternalLinks() can't mess with it
1277 $s .= $prefix . $this->insertStripItem( $sk->makeImageLinkObj( $nt, $text ), $this->mStripState
) . $trail;
1278 $wgLinkCache->addImageLinkObj( $nt );
1282 if ( $ns == NS_CATEGORY
) {
1283 $t = $nt->getText() ;
1285 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1286 $pPLC=$sk->postParseLinkColour();
1287 $sk->postParseLinkColour( false );
1288 $t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
1289 $sk->postParseLinkColour( $pPLC );
1290 $wgLinkCache->resume();
1293 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1294 $sortkey = $this->mTitle
->getText();
1296 $sortkey = $this->mTitle
->getPrefixedText();
1301 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1302 $this->mOutput
->mCategoryLinks
[] = $t ;
1303 $s .= $prefix . $trail ;
1308 if( ( $nt->getPrefixedText() === $this->mTitle
->getPrefixedText() ) &&
1309 ( $nt->getFragment() === '' ) ) {
1310 # Self-links are handled specially; generally de-link and change to bold.
1311 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1315 # Special and Media are pseudo-namespaces; no pages actually exist in them
1316 if( $ns == NS_MEDIA
) {
1317 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1318 $wgLinkCache->addImageLinkObj( $nt );
1320 } elseif( $ns == NS_SPECIAL
) {
1321 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1324 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1326 $sk->postParseLinkColour( $saveParseColour );
1327 wfProfileOut( $fname );
1332 * Handle link to subpage if necessary
1333 * @param string $target the source of the link
1334 * @param string &$text the link text, modified as necessary
1335 * @return string the full name of the link
1338 function maybeDoSubpageLink($target, &$text) {
1341 # :Foobar -- override special treatment of prefix (images, language links)
1342 # /Foobar -- convert to CurrentPage/Foobar
1343 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1344 global $wgNamespacesWithSubpages;
1346 $fname = 'Parser::maybeDoSubpageLink';
1347 wfProfileIn( $fname );
1348 # Look at the first character
1349 if( $target{0} == '/' ) {
1350 # / at end means we don't want the slash to be shown
1351 if(substr($target,-1,1)=='/') {
1352 $target=substr($target,1,-1);
1355 $noslash=substr($target,1);
1358 # Some namespaces don't allow subpages
1359 if(!empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()])) {
1360 # subpages allowed here
1361 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1362 if( '' === $text ) {
1364 } # this might be changed for ugliness reasons
1366 # no subpage allowed, use standard link
1374 wfProfileOut( $fname );
1379 * Used by doBlockLevels()
1382 /* private */ function closeParagraph() {
1384 if ( '' != $this->mLastSection
) {
1385 $result = '</' . $this->mLastSection
. ">\n";
1387 $this->mInPre
= false;
1388 $this->mLastSection
= '';
1391 # getCommon() returns the length of the longest common substring
1392 # of both arguments, starting at the beginning of both.
1394 /* private */ function getCommon( $st1, $st2 ) {
1395 $fl = strlen( $st1 );
1396 $shorter = strlen( $st2 );
1397 if ( $fl < $shorter ) { $shorter = $fl; }
1399 for ( $i = 0; $i < $shorter; ++
$i ) {
1400 if ( $st1{$i} != $st2{$i} ) { break; }
1404 # These next three functions open, continue, and close the list
1405 # element appropriate to the prefix character passed into them.
1407 /* private */ function openList( $char ) {
1408 $result = $this->closeParagraph();
1410 if ( '*' == $char ) { $result .= '<ul><li>'; }
1411 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1412 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1413 else if ( ';' == $char ) {
1414 $result .= '<dl><dt>';
1415 $this->mDTopen
= true;
1417 else { $result = '<!-- ERR 1 -->'; }
1422 /* private */ function nextItem( $char ) {
1423 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1424 else if ( ':' == $char ||
';' == $char ) {
1426 if ( $this->mDTopen
) { $close = '</dt>'; }
1427 if ( ';' == $char ) {
1428 $this->mDTopen
= true;
1429 return $close . '<dt>';
1431 $this->mDTopen
= false;
1432 return $close . '<dd>';
1435 return '<!-- ERR 2 -->';
1438 /* private */ function closeList( $char ) {
1439 if ( '*' == $char ) { $text = '</li></ul>'; }
1440 else if ( '#' == $char ) { $text = '</li></ol>'; }
1441 else if ( ':' == $char ) {
1442 if ( $this->mDTopen
) {
1443 $this->mDTopen
= false;
1444 $text = '</dt></dl>';
1446 $text = '</dd></dl>';
1449 else { return '<!-- ERR 3 -->'; }
1455 * Make lists from lines starting with ':', '*', '#', etc.
1458 * @return string the lists rendered as HTML
1460 function doBlockLevels( $text, $linestart ) {
1461 $fname = 'Parser::doBlockLevels';
1462 wfProfileIn( $fname );
1464 # Parsing through the text line by line. The main thing
1465 # happening here is handling of block-level elements p, pre,
1466 # and making lists from lines starting with * # : etc.
1468 $textLines = explode( "\n", $text );
1470 $lastPrefix = $output = $lastLine = '';
1471 $this->mDTopen
= $inBlockElem = false;
1473 $paragraphStack = false;
1475 if ( !$linestart ) {
1476 $output .= array_shift( $textLines );
1478 foreach ( $textLines as $oLine ) {
1479 $lastPrefixLength = strlen( $lastPrefix );
1480 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1481 $preOpenMatch = preg_match('/<pre/i', $oLine );
1482 if ( !$this->mInPre
) {
1483 # Multiple prefixes may abut each other for nested lists.
1484 $prefixLength = strspn( $oLine, '*#:;' );
1485 $pref = substr( $oLine, 0, $prefixLength );
1488 $pref2 = str_replace( ';', ':', $pref );
1489 $t = substr( $oLine, $prefixLength );
1490 $this->mInPre
= !empty($preOpenMatch);
1492 # Don't interpret any other prefixes in preformatted text
1494 $pref = $pref2 = '';
1499 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1500 # Same as the last item, so no need to deal with nesting or opening stuff
1501 $output .= $this->nextItem( substr( $pref, -1 ) );
1502 $paragraphStack = false;
1504 if ( substr( $pref, -1 ) == ';') {
1505 # The one nasty exception: definition lists work like this:
1506 # ; title : definition text
1507 # So we check for : in the remainder text to split up the
1508 # title and definition, without b0rking links.
1509 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1511 $output .= $term . $this->nextItem( ':' );
1514 } elseif( $prefixLength ||
$lastPrefixLength ) {
1515 # Either open or close a level...
1516 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1517 $paragraphStack = false;
1519 while( $commonPrefixLength < $lastPrefixLength ) {
1520 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1521 --$lastPrefixLength;
1523 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1524 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1526 while ( $prefixLength > $commonPrefixLength ) {
1527 $char = substr( $pref, $commonPrefixLength, 1 );
1528 $output .= $this->openList( $char );
1530 if ( ';' == $char ) {
1531 # FIXME: This is dupe of code above
1532 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1534 $output .= $term . $this->nextItem( ':' );
1537 ++
$commonPrefixLength;
1539 $lastPrefix = $pref2;
1541 if( 0 == $prefixLength ) {
1542 # No prefix (not in list)--go to paragraph mode
1543 $uniq_prefix = UNIQ_PREFIX
;
1544 // XXX: use a stack for nestable elements like span, table and div
1545 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1546 $closematch = preg_match(
1547 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1548 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1549 if ( $openmatch or $closematch ) {
1550 $paragraphStack = false;
1551 $output .= $this->closeParagraph();
1552 if($preOpenMatch and !$preCloseMatch) {
1553 $this->mInPre
= true;
1555 if ( $closematch ) {
1556 $inBlockElem = false;
1558 $inBlockElem = true;
1560 } else if ( !$inBlockElem && !$this->mInPre
) {
1561 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1563 if ($this->mLastSection
!= 'pre') {
1564 $paragraphStack = false;
1565 $output .= $this->closeParagraph().'<pre>';
1566 $this->mLastSection
= 'pre';
1568 $t = substr( $t, 1 );
1571 if ( '' == trim($t) ) {
1572 if ( $paragraphStack ) {
1573 $output .= $paragraphStack.'<br />';
1574 $paragraphStack = false;
1575 $this->mLastSection
= 'p';
1577 if ($this->mLastSection
!= 'p' ) {
1578 $output .= $this->closeParagraph();
1579 $this->mLastSection
= '';
1580 $paragraphStack = '<p>';
1582 $paragraphStack = '</p><p>';
1586 if ( $paragraphStack ) {
1587 $output .= $paragraphStack;
1588 $paragraphStack = false;
1589 $this->mLastSection
= 'p';
1590 } else if ($this->mLastSection
!= 'p') {
1591 $output .= $this->closeParagraph().'<p>';
1592 $this->mLastSection
= 'p';
1598 if ($paragraphStack === false) {
1602 while ( $prefixLength ) {
1603 $output .= $this->closeList( $pref2{$prefixLength-1} );
1606 if ( '' != $this->mLastSection
) {
1607 $output .= '</' . $this->mLastSection
. '>';
1608 $this->mLastSection
= '';
1611 wfProfileOut( $fname );
1616 * Split up a string on ':', ignoring any occurences inside
1617 * <a>..</a> or <span>...</span>
1618 * @param string $str the string to split
1619 * @param string &$before set to everything before the ':'
1620 * @param string &$after set to everything after the ':'
1621 * return string the position of the ':', or false if none found
1623 function findColonNoLinks($str, &$before, &$after) {
1624 # I wonder if we should make this count all tags, not just <a>
1625 # and <span>. That would prevent us from matching a ':' that
1626 # comes in the middle of italics other such formatting....
1628 $fname = 'Parser::findColonNoLinks';
1629 wfProfileIn( $fname );
1632 $colon = strpos($str, ':', $pos);
1634 if ($colon !== false) {
1635 $before = substr($str, 0, $colon);
1636 $after = substr($str, $colon +
1);
1638 # Skip any ':' within <a> or <span> pairs
1639 $a = substr_count($before, '<a');
1640 $s = substr_count($before, '<span');
1641 $ca = substr_count($before, '</a>');
1642 $cs = substr_count($before, '</span>');
1644 if ($a <= $ca and $s <= $cs) {
1645 # Tags are balanced before ':'; ok
1650 } while ($colon !== false);
1651 wfProfileOut( $fname );
1656 * Return value of a magic variable (like PAGENAME)
1660 function getVariableValue( $index ) {
1661 global $wgContLang, $wgSitename, $wgServer;
1664 case MAG_CURRENTMONTH
:
1665 return $wgContLang->formatNum( date( 'm' ) );
1666 case MAG_CURRENTMONTHNAME
:
1667 return $wgContLang->getMonthName( date('n') );
1668 case MAG_CURRENTMONTHNAMEGEN
:
1669 return $wgContLang->getMonthNameGen( date('n') );
1670 case MAG_CURRENTDAY
:
1671 return $wgContLang->formatNum( date('j') );
1673 return $this->mTitle
->getText();
1675 return $this->mTitle
->getPartialURL();
1677 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1678 return $wgContLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1679 case MAG_CURRENTDAYNAME
:
1680 return $wgContLang->getWeekdayName( date('w')+
1 );
1681 case MAG_CURRENTYEAR
:
1682 return $wgContLang->formatNum( date( 'Y' ) );
1683 case MAG_CURRENTTIME
:
1684 return $wgContLang->time( wfTimestampNow(), false );
1685 case MAG_NUMBEROFARTICLES
:
1686 return $wgContLang->formatNum( wfNumberOfArticles() );
1697 * initialise the magic variables (like CURRENTMONTHNAME)
1701 function initialiseVariables() {
1702 $fname = 'Parser::initialiseVariables';
1703 wfProfileIn( $fname );
1704 global $wgVariableIDs;
1705 $this->mVariables
= array();
1706 foreach ( $wgVariableIDs as $id ) {
1707 $mw =& MagicWord
::get( $id );
1708 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1710 wfProfileOut( $fname );
1714 * Replace magic variables, templates, and template arguments
1715 * with the appropriate text. Templates are substituted recursively,
1716 * taking care to avoid infinite loops.
1718 * Note that the substitution depends on value of $mOutputType:
1719 * OT_WIKI: only {{subst:}} templates
1720 * OT_MSG: only magic variables
1721 * OT_HTML: all templates and magic variables
1723 * @param string $tex The text to transform
1724 * @param array $args Key-value pairs representing template parameters to substitute
1727 function replaceVariables( $text, $args = array() ) {
1728 global $wgLang, $wgScript, $wgArticlePath;
1730 # Prevent too big inclusions
1731 if(strlen($text)> MAX_INCLUDE_SIZE
)
1734 $fname = 'Parser::replaceVariables';
1735 wfProfileIn( $fname );
1737 $titleChars = Title
::legalChars();
1739 # This function is called recursively. To keep track of arguments we need a stack:
1740 array_push( $this->mArgStack
, $args );
1742 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1743 $GLOBALS['wgCurParser'] =& $this;
1745 # Variable substitution
1746 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", 'wfVariableSubstitution', $text );
1748 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1749 # Argument substitution
1750 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1752 # Template substitution
1753 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1754 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1756 array_pop( $this->mArgStack
);
1758 wfProfileOut( $fname );
1763 * Replace magic variables
1766 function variableSubstitution( $matches ) {
1767 if ( !$this->mVariables
) {
1768 $this->initialiseVariables();
1771 if ( $this->mOutputType
== OT_WIKI
) {
1772 # Do only magic variables prefixed by SUBST
1773 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1774 if (!$mwSubst->matchStartAndRemove( $matches[1] ))
1776 # Note that if we don't substitute the variable below,
1777 # we don't remove the {{subst:}} magic word, in case
1778 # it is a template rather than a magic variable.
1780 if ( !$skip && array_key_exists( $matches[1], $this->mVariables
) ) {
1781 $text = $this->mVariables
[$matches[1]];
1782 $this->mOutput
->mContainsOldMagic
= true;
1784 $text = $matches[0];
1789 # Split template arguments
1790 function getTemplateArgs( $argsString ) {
1791 if ( $argsString === '' ) {
1795 $args = explode( '|', substr( $argsString, 1 ) );
1797 # If any of the arguments contains a '[[' but no ']]', it needs to be
1798 # merged with the next arg because the '|' character between belongs
1799 # to the link syntax and not the template parameter syntax.
1800 $argc = count($args);
1802 for ( $i = 0; $i < $argc-1; $i++
) {
1803 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1804 $args[$i] .= '|'.$args[$i+
1];
1805 array_splice($args, $i+
1, 1);
1815 * Return the text of a template, after recursively
1816 * replacing any variables or templates within the template.
1818 * @param array $matches The parts of the template
1819 * $matches[1]: the title, i.e. the part before the |
1820 * $matches[2]: the parameters (including a leading |), if any
1821 * @return string the text of the template
1824 function braceSubstitution( $matches ) {
1825 global $wgLinkCache, $wgContLang;
1826 $fname = 'Parser::braceSubstitution';
1833 # Need to know if the template comes at the start of a line,
1834 # to treat the beginning of the template like the beginning
1835 # of a line for tables and block-level elements.
1836 $linestart = $matches[1];
1838 # $part1 is the bit before the first |, and must contain only title characters
1839 # $args is a list of arguments, starting from index 0, not including $part1
1841 $part1 = $matches[2];
1842 # If the third subpattern matched anything, it will start with |
1844 $args = $this->getTemplateArgs($matches[3]);
1845 $argc = count( $args );
1847 # Don't parse {{{}}} because that's only for template arguments
1848 if ( $linestart === '{' ) {
1849 $text = $matches[0];
1856 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1857 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
1858 # One of two possibilities is true:
1859 # 1) Found SUBST but not in the PST phase
1860 # 2) Didn't find SUBST and in the PST phase
1861 # In either case, return without further processing
1862 $text = $matches[0];
1868 # MSG, MSGNW and INT
1871 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1872 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1875 # Remove obsolete MSG:
1876 $mwMsg =& MagicWord
::get( MAG_MSG
);
1877 $mwMsg->matchStartAndRemove( $part1 );
1880 # Check if it is an internal message
1881 $mwInt =& MagicWord
::get( MAG_INT
);
1882 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1883 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1884 $text = $linestart . wfMsgReal( $part1, $args, true );
1892 # Check for NS: (namespace expansion)
1893 $mwNs = MagicWord
::get( MAG_NS
);
1894 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1895 if ( intval( $part1 ) ) {
1896 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1899 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1900 if ( !is_null( $index ) ) {
1901 $text = $linestart . $wgContLang->getNsText( $index );
1908 # LOCALURL and LOCALURLE
1910 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1911 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1913 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1914 $func = 'getLocalURL';
1915 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1916 $func = 'escapeLocalURL';
1921 if ( $func !== '' ) {
1922 $title = Title
::newFromText( $part1 );
1923 if ( !is_null( $title ) ) {
1925 $text = $linestart . $title->$func( $args[0] );
1927 $text = $linestart . $title->$func();
1935 if ( !$found && $argc == 1 ) {
1936 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
1937 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1938 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
1943 # Template table test
1945 # Did we encounter this template already? If yes, it is in the cache
1946 # and we need to check for loops.
1947 if ( !$found && isset( $this->mTemplates
[$part1] ) ) {
1948 # set $text to cached message.
1949 $text = $linestart . $this->mTemplates
[$part1];
1952 # Infinite loop test
1953 if ( isset( $this->mTemplatePath
[$part1] ) ) {
1956 $text .= '<!-- WARNING: template loop detected -->';
1960 # Load from database
1961 $itcamefromthedatabase = false;
1964 $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
1965 if ($subpage !== '') {
1966 $ns = $this->mTitle
->getNamespace();
1968 $title = Title
::newFromText( $part1, $ns );
1969 if ( !is_null( $title ) && !$title->isExternal() ) {
1970 # Check for excessive inclusion
1971 $dbk = $title->getPrefixedDBkey();
1972 if ( $this->incrementIncludeCount( $dbk ) ) {
1973 # This should never be reached.
1974 $article = new Article( $title );
1975 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1976 if ( $articleContent !== false ) {
1978 $text = $linestart . $articleContent;
1979 $itcamefromthedatabase = true;
1983 # If the title is valid but undisplayable, make a link to it
1984 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1985 $text = $linestart . '[['.$title->getPrefixedText().']]';
1989 # Template cache array insertion
1990 $this->mTemplates
[$part1] = $text;
1994 # Recursive parsing, escaping and link table handling
1995 # Only for HTML output
1996 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1997 $text = wfEscapeWikiText( $text );
1998 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
1999 # Clean up argument array
2000 $assocArgs = array();
2002 foreach( $args as $arg ) {
2003 $eqpos = strpos( $arg, '=' );
2004 if ( $eqpos === false ) {
2005 $assocArgs[$index++
] = $arg;
2007 $name = trim( substr( $arg, 0, $eqpos ) );
2008 $value = trim( substr( $arg, $eqpos+
1 ) );
2009 if ( $value === false ) {
2012 if ( $name !== false ) {
2013 $assocArgs[$name] = $value;
2018 # Add a new element to the templace recursion path
2019 $this->mTemplatePath
[$part1] = 1;
2021 $text = $this->strip( $text, $this->mStripState
);
2022 $text = $this->removeHTMLtags( $text );
2023 $text = $this->replaceVariables( $text, $assocArgs );
2025 # Resume the link cache and register the inclusion as a link
2026 if ( $this->mOutputType
== OT_HTML
&& !is_null( $title ) ) {
2027 $wgLinkCache->addLinkObj( $title );
2030 # If the template begins with a table or block-level
2031 # element, it should be treated as beginning a new line.
2032 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
2033 $text = "\n" . $text;
2037 # Empties the template path
2038 $this->mTemplatePath
= array();
2042 # replace ==section headers==
2043 # XXX this needs to go away once we have a better parser.
2044 if ( $this->mOutputType
!= OT_WIKI
&& $itcamefromthedatabase ) {
2045 if( !is_null( $title ) )
2046 $encodedname = base64_encode($title->getPrefixedDBkey());
2048 $encodedname = base64_encode("");
2049 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
2050 PREG_SPLIT_DELIM_CAPTURE
);
2053 for( $i = 0; $i < count($m); $i +
= 2 ) {
2055 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
2057 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
2061 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
2062 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
2063 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
2070 # Empties the template path
2071 $this->mTemplatePath
= array();
2080 * Triple brace replacement -- used for template arguments
2083 function argSubstitution( $matches ) {
2084 $arg = trim( $matches[1] );
2085 $text = $matches[0];
2086 $inputArgs = end( $this->mArgStack
);
2088 if ( array_key_exists( $arg, $inputArgs ) ) {
2089 $text = $inputArgs[$arg];
2096 * Returns true if the function is allowed to include this entity
2099 function incrementIncludeCount( $dbk ) {
2100 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
2101 $this->mIncludeCount
[$dbk] = 0;
2103 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
2112 * Cleans up HTML, removes dangerous tags and attributes, and
2113 * removes HTML comments
2116 function removeHTMLtags( $text ) {
2117 global $wgUseTidy, $wgUserHtml;
2118 $fname = 'Parser::removeHTMLtags';
2119 wfProfileIn( $fname );
2122 $htmlpairs = array( # Tags that must be closed
2123 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
2124 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
2125 'strike', 'strong', 'tt', 'var', 'div', 'center',
2126 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
2127 'ruby', 'rt' , 'rb' , 'rp', 'p'
2129 $htmlsingle = array(
2130 'br', 'hr', 'li', 'dt', 'dd'
2132 $htmlnest = array( # Tags that can be nested--??
2133 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
2134 'dl', 'font', 'big', 'small', 'sub', 'sup'
2136 $tabletags = array( # Can only appear inside table
2140 $htmlpairs = array();
2141 $htmlsingle = array();
2142 $htmlnest = array();
2143 $tabletags = array();
2146 $htmlsingle = array_merge( $tabletags, $htmlsingle );
2147 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
2149 $htmlattrs = $this->getHTMLattrs () ;
2151 # Remove HTML comments
2152 $text = $this->removeHTMLcomments( $text );
2154 $bits = explode( '<', $text );
2155 $text = array_shift( $bits );
2157 $tagstack = array(); $tablestack = array();
2158 foreach ( $bits as $x ) {
2159 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
2160 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2162 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2163 error_reporting( $prev );
2166 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2170 if ( ! in_array( $t, $htmlsingle ) &&
2171 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
2172 @array_push
( $tagstack, $ot );
2175 if ( $t == 'table' ) {
2176 $tagstack = array_pop( $tablestack );
2181 # Keep track for later
2182 if ( in_array( $t, $tabletags ) &&
2183 ! in_array( 'table', $tagstack ) ) {
2185 } else if ( in_array( $t, $tagstack ) &&
2186 ! in_array ( $t , $htmlnest ) ) {
2188 } else if ( ! in_array( $t, $htmlsingle ) ) {
2189 if ( $t == 'table' ) {
2190 array_push( $tablestack, $tagstack );
2191 $tagstack = array();
2193 array_push( $tagstack, $t );
2195 # Strip non-approved attributes from the tag
2196 $newparams = $this->fixTagAttributes($params);
2200 $rest = str_replace( '>', '>', $rest );
2201 $text .= "<$slash$t $newparams$brace$rest";
2205 $text .= '<' . str_replace( '>', '>', $x);
2207 # Close off any remaining tags
2208 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2210 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2213 # this might be possible using tidy itself
2214 foreach ( $bits as $x ) {
2215 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2217 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2218 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2219 $newparams = $this->fixTagAttributes($params);
2220 $rest = str_replace( '>', '>', $rest );
2221 $text .= "<$slash$t $newparams$brace$rest";
2223 $text .= '<' . str_replace( '>', '>', $x);
2227 wfProfileOut( $fname );
2232 * Remove '<!--', '-->', and everything between.
2233 * To avoid leaving blank lines, when a comment is both preceded
2234 * and followed by a newline (ignoring spaces), trim leading and
2235 * trailing spaces and one of the newlines.
2239 function removeHTMLcomments( $text ) {
2240 $fname='Parser::removeHTMLcomments';
2241 wfProfileIn( $fname );
2242 while (($start = strpos($text, '<!--')) !== false) {
2243 $end = strpos($text, '-->', $start +
4);
2244 if ($end === false) {
2245 # Unterminated comment; bail out
2251 # Trim space and newline if the comment is both
2252 # preceded and followed by a newline
2253 $spaceStart = max($start - 1, 0);
2254 $spaceLen = $end - $spaceStart;
2255 while (substr($text, $spaceStart, 1) === ' ' && $spaceStart > 0) {
2259 while (substr($text, $spaceStart +
$spaceLen, 1) === ' ')
2261 if (substr($text, $spaceStart, 1) === "\n" and substr($text, $spaceStart +
$spaceLen, 1) === "\n") {
2262 # Remove the comment, leading and trailing
2263 # spaces, and leave only one newline.
2264 $text = substr_replace($text, "\n", $spaceStart, $spaceLen +
1);
2267 # Remove just the comment.
2268 $text = substr_replace($text, '', $start, $end - $start);
2271 wfProfileOut( $fname );
2276 * This function accomplishes several tasks:
2277 * 1) Auto-number headings if that option is enabled
2278 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2279 * 3) Add a Table of contents on the top for users who have enabled the option
2280 * 4) Auto-anchor headings
2282 * It loops through all headlines, collects the necessary data, then splits up the
2283 * string and re-inserts the newly formatted headlines.
2286 /* private */ function formatHeadings( $text, $isMain=true ) {
2287 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders;
2289 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2290 $doShowToc = $this->mOptions
->getShowToc();
2291 $forceTocHere = false;
2292 if( !$this->mTitle
->userCanEdit() ) {
2294 $rightClickHack = 0;
2296 $showEditLink = $this->mOptions
->getEditSection();
2297 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
2300 # Inhibit editsection links if requested in the page
2301 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2302 if( $esw->matchAndRemove( $text ) ) {
2305 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2307 $mw =& MagicWord
::get( MAG_NOTOC
);
2308 if( $mw->matchAndRemove( $text ) ) {
2312 # never add the TOC to the Main Page. This is an entry page that should not
2313 # be more than 1-2 screens large anyway
2314 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
2318 # Get all headlines for numbering them and adding funky stuff like [edit]
2319 # links - this is for later, but we need the number of headlines right now
2320 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
2322 # if there are fewer than 4 headlines in the article, do not show TOC
2323 if( $numMatches < 4 ) {
2327 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2328 # override above conditions and always show TOC at that place
2329 $mw =& MagicWord
::get( MAG_TOC
);
2330 if ($mw->match( $text ) ) {
2332 $forceTocHere = true;
2334 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2335 # override above conditions and always show TOC above first header
2336 $mw =& MagicWord
::get( MAG_FORCETOC
);
2337 if ($mw->matchAndRemove( $text ) ) {
2344 # We need this to perform operations on the HTML
2345 $sk =& $this->mOptions
->getSkin();
2349 $sectionCount = 0; # headlineCount excluding template sections
2351 # Ugh .. the TOC should have neat indentation levels which can be
2352 # passed to the skin functions. These are determined here
2357 $sublevelCount = array();
2360 foreach( $matches[3] as $headline ) {
2362 $templatetitle = "";
2363 $templatesection = 0;
2365 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2367 $templatetitle = base64_decode($mat[1]);
2368 $templatesection = 1 +
(int)base64_decode($mat[2]);
2369 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2374 $prevlevel = $level;
2376 $level = $matches[1][$headlineCount];
2377 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2378 # reset when we enter a new level
2379 $sublevelCount[$level] = 0;
2380 $toc .= $sk->tocIndent( $level - $prevlevel );
2381 $toclevel +
= $level - $prevlevel;
2383 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2384 # reset when we step back a level
2385 $sublevelCount[$level+
1]=0;
2386 $toc .= $sk->tocUnindent( $prevlevel - $level );
2387 $toclevel -= $prevlevel - $level;
2389 # count number of headlines for each level
2390 @$sublevelCount[$level]++
;
2391 if( $doNumberHeadings ||
$doShowToc ) {
2393 for( $i = 1; $i <= $level; $i++
) {
2394 if( !empty( $sublevelCount[$i] ) ) {
2398 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2404 # The canonized header is a version of the header text safe to use for links
2405 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2406 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2407 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2409 # Remove link placeholders by the link text.
2410 # <!--LINK number-->
2412 # link text with suffix
2413 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2414 "\$wgLinkHolders['texts'][\$1]",
2415 $canonized_headline );
2418 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2419 $tocline = trim( $canonized_headline );
2420 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2421 $replacearray = array(
2425 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2426 $refer[$headlineCount] = $canonized_headline;
2428 # count how many in assoc. array so we can track dupes in anchors
2429 @$refers[$canonized_headline]++
;
2430 $refcount[$headlineCount]=$refers[$canonized_headline];
2432 # Prepend the number to the heading text
2434 if( $doNumberHeadings ||
$doShowToc ) {
2435 $tocline = $numbering . ' ' . $tocline;
2437 # Don't number the heading if it is the only one (looks silly)
2438 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2439 # the two are different if the line contains a link
2440 $headline=$numbering . ' ' . $headline;
2444 # Create the anchor for linking from the TOC to the section
2445 $anchor = $canonized_headline;
2446 if($refcount[$headlineCount] > 1 ) {
2447 $anchor .= '_' . $refcount[$headlineCount];
2449 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2450 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2452 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2453 if ( empty( $head[$headlineCount] ) ) {
2454 $head[$headlineCount] = '';
2457 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2459 $head[$headlineCount] .= $sk->editSectionLink($this->mTitle
, $sectionCount+
1);
2462 # Add the edit section span
2463 if( $rightClickHack ) {
2465 $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
2467 $headline = $sk->editSectionScript($this->title
, $sectionCount+
1,$headline);
2470 # give headline the correct <h#> tag
2471 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2479 $toclines = $headlineCount;
2480 $toc .= $sk->tocUnindent( $toclevel );
2481 $toc = $sk->tocTable( $toc );
2484 # split up and insert constructed headlines
2486 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2489 foreach( $blocks as $block ) {
2490 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2491 # This is the [edit] link that appears for the top block of text when
2492 # section editing is enabled
2494 # Disabled because it broke block formatting
2495 # For example, a bullet point in the top line
2496 # $full .= $sk->editSectionLink(0);
2499 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2500 # Top anchor now in skin
2504 if( !empty( $head[$i] ) ) {
2510 $mw =& MagicWord
::get( MAG_TOC
);
2511 return $mw->replace( $toc, $full );
2518 * Return an HTML link for the "ISBN 123456" text
2521 function magicISBN( $text ) {
2523 $fname = 'Parser::magicISBN';
2524 wfProfileIn( $fname );
2526 $a = split( 'ISBN ', ' '.$text );
2527 if ( count ( $a ) < 2 ) {
2528 wfProfileOut( $fname );
2531 $text = substr( array_shift( $a ), 1);
2532 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2534 foreach ( $a as $x ) {
2535 $isbn = $blank = '' ;
2536 while ( ' ' == $x{0} ) {
2538 $x = substr( $x, 1 );
2540 if ( $x == '' ) { # blank isbn
2541 $text .= "ISBN $blank";
2544 while ( strstr( $valid, $x{0} ) != false ) {
2546 $x = substr( $x, 1 );
2548 $num = str_replace( '-', '', $isbn );
2549 $num = str_replace( ' ', '', $num );
2552 $text .= "ISBN $blank$x";
2554 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2555 $text .= '<a href="' .
2556 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2557 "\" class=\"internal\">ISBN $isbn</a>";
2561 wfProfileOut( $fname );
2566 * Return an HTML link for the "GEO ..." text
2569 function magicGEO( $text ) {
2570 global $wgLang, $wgUseGeoMode;
2571 $fname = 'Parser::magicGEO';
2572 wfProfileIn( $fname );
2574 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2575 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2576 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2577 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2578 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2579 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2581 $a = split( 'GEO ', ' '.$text );
2582 if ( count ( $a ) < 2 ) {
2583 wfProfileOut( $fname );
2586 $text = substr( array_shift( $a ), 1);
2587 $valid = '0123456789.+-:';
2589 foreach ( $a as $x ) {
2590 $geo = $blank = '' ;
2591 while ( ' ' == $x{0} ) {
2593 $x = substr( $x, 1 );
2595 while ( strstr( $valid, $x{0} ) != false ) {
2597 $x = substr( $x, 1 );
2599 $num = str_replace( '+', '', $geo );
2600 $num = str_replace( ' ', '', $num );
2602 if ( '' == $num ||
count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2603 $text .= "GEO $blank$x";
2605 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2606 $text .= '<a href="' .
2607 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2608 "\" class=\"internal\">GEO $geo</a>";
2612 wfProfileOut( $fname );
2617 * Return an HTML link for the "RFC 1234" text
2619 * @param string $text text to be processed
2621 function magicRFC( $text ) {
2624 $valid = '0123456789';
2627 $a = split( 'RFC ', ' '.$text );
2628 if ( count ( $a ) < 2 ) return $text;
2629 $text = substr( array_shift( $a ), 1);
2631 /* Check if RFC keyword is preceed by [[.
2632 * This test is made here cause of the array_shift above
2633 * that prevent the test to be done in the foreach.
2635 if(substr($text, -2) == '[[') { $internal = true; }
2637 foreach ( $a as $x ) {
2638 /* token might be empty if we have RFC RFC 1234 */
2644 $rfc = $blank = '' ;
2646 /** remove and save whitespaces in $blank */
2647 while ( $x{0} == ' ' ) {
2649 $x = substr( $x, 1 );
2652 /** remove and save the rfc number in $rfc */
2653 while ( strstr( $valid, $x{0} ) != false ) {
2655 $x = substr( $x, 1 );
2659 /* call back stripped spaces*/
2660 $text .= "RFC $blank$x";
2661 } elseif( $internal) {
2663 $text .= "RFC $rfc$x";
2665 /* build the external link*/
2666 $url = wfmsg( 'rfcurl' );
2667 $url = str_replace( '$1', $rfc, $url);
2668 $sk =& $this->mOptions
->getSkin();
2669 $la = $sk->getExternalLinkAttributes( $url, 'RFC '.$rfc );
2670 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2673 /* Check if the next RFC keyword is preceed by [[ */
2674 $internal = (substr($x,-2) == '[[');
2680 * Transform wiki markup when saving a page by doing \r\n -> \n
2681 * conversion, substitting signatures, {{subst:}} templates, etc.
2683 * @param string $text the text to transform
2684 * @param Title &$title the Title object for the current article
2685 * @param User &$user the User object describing the current user
2686 * @param ParserOptions $options parsing options
2687 * @param bool $clearState whether to clear the parser state first
2688 * @return string the altered wiki markup
2691 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2692 $this->mOptions
= $options;
2693 $this->mTitle
=& $title;
2694 $this->mOutputType
= OT_WIKI
;
2696 if ( $clearState ) {
2697 $this->clearState();
2700 $stripState = false;
2704 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2708 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2709 "/<br *?>/i" => "<br />",
2711 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2713 $text = $this->strip( $text, $stripState, false );
2714 $text = $this->pstPass2( $text, $user );
2715 $text = $this->unstrip( $text, $stripState );
2716 $text = $this->unstripNoWiki( $text, $stripState );
2721 * Pre-save transform helper function
2724 function pstPass2( $text, &$user ) {
2725 global $wgLang, $wgContLang, $wgLocaltimezone, $wgCurParser;
2727 # Variable replacement
2728 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2729 $text = $this->replaceVariables( $text );
2733 $n = $user->getName();
2734 $k = $user->getOption( 'nickname' );
2735 if ( '' == $k ) { $k = $n; }
2736 if(isset($wgLocaltimezone)) {
2737 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2739 /* Note: this is an ugly timezone hack for the European wikis */
2740 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
2741 ' (' . date( 'T' ) . ')';
2742 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2744 $text = preg_replace( '/~~~~~/', $d, $text );
2745 $text = preg_replace( '/~~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]] $d", $text );
2746 $text = preg_replace( '/~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]]", $text );
2748 # Context links: [[|name]] and [[name (context)|]]
2750 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2751 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2752 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2753 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2755 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2756 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2757 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2758 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2760 $t = $this->mTitle
->getText();
2761 if ( preg_match( $conpat, $t, $m ) ) {
2764 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2765 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2766 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2768 if ( '' == $context ) {
2769 $text = preg_replace( $p2, '[[\\1]]', $text );
2771 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2774 # Trim trailing whitespace
2775 # MAG_END (__END__) tag allows for trailing
2776 # whitespace to be deliberately included
2777 $text = rtrim( $text );
2778 $mw =& MagicWord
::get( MAG_END
);
2779 $mw->matchAndRemove( $text );
2785 * Set up some variables which are usually set up in parse()
2786 * so that an external function can call some class members with confidence
2789 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2790 $this->mTitle
=& $title;
2791 $this->mOptions
= $options;
2792 $this->mOutputType
= $outputType;
2793 if ( $clearState ) {
2794 $this->clearState();
2799 * Transform a MediaWiki message by replacing magic variables.
2801 * @param string $text the text to transform
2802 * @param ParserOptions $options options
2803 * @return string the text with variables substituted
2806 function transformMsg( $text, $options ) {
2808 static $executing = false;
2810 # Guard against infinite recursion
2816 $this->mTitle
= $wgTitle;
2817 $this->mOptions
= $options;
2818 $this->mOutputType
= OT_MSG
;
2819 $this->clearState();
2820 $text = $this->replaceVariables( $text );
2827 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2828 * Callback will be called with the text within
2829 * Transform and return the text within
2832 function setHook( $tag, $callback ) {
2833 $oldVal = @$this->mTagHooks
[$tag];
2834 $this->mTagHooks
[$tag] = $callback;
2839 * Replace <!--LINK--> link placeholders with actual links, in the buffer
2840 * Placeholders created in Skin::makeLinkObj()
2841 * Returns an array of links found, indexed by PDBK:
2845 * $options is a bit field, RLH_FOR_UPDATE to select for update
2847 function replaceLinkHolders( &$text, $options = 0 ) {
2848 global $wgUser, $wgLinkCache, $wgUseOldExistenceCheck, $wgLinkHolders;
2850 if ( $wgUseOldExistenceCheck ) {
2854 $fname = 'Parser::replaceLinkHolders';
2855 wfProfileIn( $fname );
2860 #if ( !empty( $tmpLinks[0] ) ) { #TODO
2861 if ( !empty( $wgLinkHolders['namespaces'] ) ) {
2862 wfProfileIn( $fname.'-check' );
2863 $dbr =& wfGetDB( DB_SLAVE
);
2864 $cur = $dbr->tableName( 'cur' );
2865 $sk = $wgUser->getSkin();
2866 $threshold = $wgUser->getOption('stubthreshold');
2869 asort( $wgLinkHolders['namespaces'] );
2873 foreach ( $wgLinkHolders['namespaces'] as $key => $val ) {
2875 $title = $wgLinkHolders['titles'][$key];
2877 # Skip invalid entries.
2878 # Result will be ugly, but prevents crash.
2879 if ( is_null( $title ) ) {
2882 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
2884 # Check if it's in the link cache already
2885 if ( $wgLinkCache->getGoodLinkID( $pdbk ) ) {
2886 $colours[$pdbk] = 1;
2887 } elseif ( $wgLinkCache->isBadLink( $pdbk ) ) {
2888 $colours[$pdbk] = 0;
2890 # Not in the link cache, add it to the query
2891 if ( !isset( $current ) ) {
2893 $query = "SELECT cur_id, cur_namespace, cur_title";
2894 if ( $threshold > 0 ) {
2895 $query .= ", LENGTH(cur_text) AS cur_len, cur_is_redirect";
2897 $query .= " FROM $cur WHERE (cur_namespace=$val AND cur_title IN(";
2898 } elseif ( $current != $val ) {
2900 $query .= ")) OR (cur_namespace=$val AND cur_title IN(";
2905 $query .= $dbr->addQuotes( $wgLinkHolders['dbkeys'][$key] );
2910 if ( $options & RLH_FOR_UPDATE
) {
2911 $query .= ' FOR UPDATE';
2914 $res = $dbr->query( $query, $fname );
2916 # Fetch data and form into an associative array
2917 # non-existent = broken
2920 while ( $s = $dbr->fetchObject($res) ) {
2921 $title = Title
::makeTitle( $s->cur_namespace
, $s->cur_title
);
2922 $pdbk = $title->getPrefixedDBkey();
2923 $wgLinkCache->addGoodLink( $s->cur_id
, $pdbk );
2925 if ( $threshold > 0 ) {
2926 $size = $s->cur_len
;
2927 if ( $s->cur_is_redirect ||
$s->cur_namespace
!= 0 ||
$length < $threshold ) {
2928 $colours[$pdbk] = 1;
2930 $colours[$pdbk] = 2;
2933 $colours[$pdbk] = 1;
2937 wfProfileOut( $fname.'-check' );
2939 # Construct search and replace arrays
2940 wfProfileIn( $fname.'-construct' );
2941 global $outputReplace;
2942 $outputReplace = array();
2943 foreach ( $wgLinkHolders['namespaces'] as $key => $ns ) {
2944 $pdbk = $pdbks[$key];
2945 $searchkey = '<!--LINK '.$key.'-->';
2946 $title = $wgLinkHolders['titles'][$key];
2947 if ( empty( $colours[$pdbk] ) ) {
2948 $wgLinkCache->addBadLink( $pdbk );
2949 $colours[$pdbk] = 0;
2950 $outputReplace[$searchkey] = $sk->makeBrokenLinkObj( $title,
2951 $wgLinkHolders['texts'][$key],
2952 $wgLinkHolders['queries'][$key] );
2953 } elseif ( $colours[$pdbk] == 1 ) {
2954 $outputReplace[$searchkey] = $sk->makeKnownLinkObj( $title,
2955 $wgLinkHolders['texts'][$key],
2956 $wgLinkHolders['queries'][$key] );
2957 } elseif ( $colours[$pdbk] == 2 ) {
2958 $outputReplace[$searchkey] = $sk->makeStubLinkObj( $title,
2959 $wgLinkHolders['texts'][$key],
2960 $wgLinkHolders['queries'][$key] );
2963 wfProfileOut( $fname.'-construct' );
2966 wfProfileIn( $fname.'-replace' );
2968 $text = preg_replace_callback(
2969 '/(<!--LINK .*?-->)/',
2970 "outputReplaceMatches",
2972 wfProfileOut( $fname.'-replace' );
2974 wfProfileIn( $fname.'-interwiki' );
2975 global $wgInterwikiLinkHolders;
2976 $outputReplace = $wgInterwikiLinkHolders;
2977 $text = preg_replace_callback(
2978 '/<!--IWLINK (.*?)-->/',
2979 "outputReplaceMatches",
2981 wfProfileOut( $fname.'-interwiki' );
2984 wfProfileOut( $fname );
2991 * @package MediaWiki
2995 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2996 var $mCacheTime; # Used in ParserCache
2998 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
2999 $containsOldMagic = false )
3001 $this->mText
= $text;
3002 $this->mLanguageLinks
= $languageLinks;
3003 $this->mCategoryLinks
= $categoryLinks;
3004 $this->mContainsOldMagic
= $containsOldMagic;
3005 $this->mCacheTime
= '';
3008 function getText() { return $this->mText
; }
3009 function getLanguageLinks() { return $this->mLanguageLinks
; }
3010 function getCategoryLinks() { return $this->mCategoryLinks
; }
3011 function getCacheTime() { return $this->mCacheTime
; }
3012 function containsOldMagic() { return $this->mContainsOldMagic
; }
3013 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
3014 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
3015 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
3016 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
3017 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
3019 function merge( $other ) {
3020 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
3021 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
3022 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
3028 * Set options of the Parser
3030 * @package MediaWiki
3034 # All variables are private
3035 var $mUseTeX; # Use texvc to expand <math> tags
3036 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
3037 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
3038 var $mAllowExternalImages; # Allow external images inline
3039 var $mSkin; # Reference to the preferred skin
3040 var $mDateFormat; # Date format index
3041 var $mEditSection; # Create "edit section" links
3042 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
3043 var $mNumberHeadings; # Automatically number headings
3044 var $mShowToc; # Show table of contents
3046 function getUseTeX() { return $this->mUseTeX
; }
3047 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
3048 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
3049 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
3050 function getSkin() { return $this->mSkin
; }
3051 function getDateFormat() { return $this->mDateFormat
; }
3052 function getEditSection() { return $this->mEditSection
; }
3053 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
3054 function getNumberHeadings() { return $this->mNumberHeadings
; }
3055 function getShowToc() { return $this->mShowToc
; }
3057 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
3058 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
3059 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
3060 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
3061 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
3062 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
3063 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
3064 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
3065 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
3067 function setSkin( &$x ) { $this->mSkin
=& $x; }
3069 # Get parser options
3070 /* static */ function newFromUser( &$user ) {
3071 $popts = new ParserOptions
;
3072 $popts->initialiseFromUser( $user );
3077 function initialiseFromUser( &$userInput ) {
3078 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
3080 $fname = 'ParserOptions::initialiseFromUser';
3081 wfProfileIn( $fname );
3082 if ( !$userInput ) {
3084 $user->setLoaded( true );
3086 $user =& $userInput;
3089 $this->mUseTeX
= $wgUseTeX;
3090 $this->mUseDynamicDates
= $wgUseDynamicDates;
3091 $this->mInterwikiMagic
= $wgInterwikiMagic;
3092 $this->mAllowExternalImages
= $wgAllowExternalImages;
3093 wfProfileIn( $fname.'-skin' );
3094 $this->mSkin
=& $user->getSkin();
3095 wfProfileOut( $fname.'-skin' );
3096 $this->mDateFormat
= $user->getOption( 'date' );
3097 $this->mEditSection
= $user->getOption( 'editsection' );
3098 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
3099 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
3100 $this->mShowToc
= $user->getOption( 'showtoc' );
3101 wfProfileOut( $fname );
3108 * Callback function used by Parser::replaceLinkHolders()
3109 * to substitute link placeholders.
3111 function &outputReplaceMatches($matches) {
3112 global $outputReplace;
3113 return $outputReplace[$matches[1]];
3117 # Regex callbacks, used in Parser::replaceVariables
3118 function wfBraceSubstitution( $matches ) {
3119 global $wgCurParser;
3120 return $wgCurParser->braceSubstitution( $matches );
3123 function wfArgSubstitution( $matches ) {
3124 global $wgCurParser;
3125 return $wgCurParser->argSubstitution( $matches );
3128 function wfVariableSubstitution( $matches ) {
3129 global $wgCurParser;
3130 return $wgCurParser->variableSubstitution( $matches );
3134 * Return the total number of articles
3136 function wfNumberOfArticles() {
3137 global $wgNumberOfArticles;
3140 return $wgNumberOfArticles;
3144 * Get various statistics from the database
3147 function wfLoadSiteStats() {
3148 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
3149 $fname = 'wfLoadSiteStats';
3151 if ( -1 != $wgNumberOfArticles ) return;
3152 $dbr =& wfGetDB( DB_SLAVE
);
3153 $s = $dbr->getArray( 'site_stats',
3154 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
3155 array( 'ss_row_id' => 1 ), $fname
3158 if ( $s === false ) {
3161 $wgTotalViews = $s->ss_total_views
;
3162 $wgTotalEdits = $s->ss_total_edits
;
3163 $wgNumberOfArticles = $s->ss_good_articles
;
3167 function wfEscapeHTMLTagsOnly( $in ) {
3169 array( '"', '>', '<' ),
3170 array( '"', '>', '<' ),