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 static $fname = 'Parser::replaceInternalLinks' ;
1097 # use a counter to prevent too much unknown links from
1098 # being checked for different language variants.
1099 static $convertCount;
1100 wfProfileIn( $fname );
1102 wfProfileIn( $fname.'-setup' );
1104 # the % is needed to support urlencoded titles as well
1105 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1106 $sk =& $this->mOptions
->getSkin();
1108 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1110 #split the entire text string on occurences of [[
1111 $a = explode( '[[', ' ' . $s );
1112 #get the first element (all text up to first [[), and remove the space we added
1113 $s = array_shift( $a );
1114 $s = substr( $s, 1 );
1116 # Match a link having the form [[namespace:link|alternate]]trail
1118 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1119 # Match cases where there is no "]]", which might still be images
1120 static $e1_img = FALSE;
1121 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1122 # Match the end of a line for a word that's not followed by whitespace,
1123 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1124 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1126 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1128 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1130 if ( $useLinkPrefixExtension ) {
1131 if ( preg_match( $e2, $s, $m ) ) {
1132 $first_prefix = $m[2];
1135 $first_prefix = false;
1141 wfProfileOut( $fname.'-setup' );
1143 # Loop for each link
1144 for ($k = 0; isset( $a[$k] ); $k++
) {
1146 wfProfileIn( $fname.'-prefixhandling' );
1147 if ( $useLinkPrefixExtension ) {
1148 if ( preg_match( $e2, $s, $m ) ) {
1156 $prefix = $first_prefix;
1157 $first_prefix = false;
1160 wfProfileOut( $fname.'-prefixhandling' );
1162 $might_be_img = false;
1164 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1166 # fix up urlencoded title texts
1167 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1169 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1170 $might_be_img = true;
1172 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1174 } else { # Invalid form; output directly
1175 $s .= $prefix . '[[' . $line ;
1179 # Don't allow internal links to pages containing
1180 # PROTO: where PROTO is a valid URL protocol; these
1181 # should be external links.
1182 if (preg_match('/((?:'.URL_PROTOCOLS
.'):)/', $m[1])) {
1183 $s .= $prefix . '[[' . $line ;
1187 # Make subpage if necessary
1188 $link = $this->maybeDoSubpageLink( $m[1], $text );
1190 $noforce = (substr($m[1], 0, 1) != ':');
1192 # Strip off leading ':'
1193 $link = substr($link, 1);
1196 $nt = Title
::newFromText( $this->unstripNoWiki($link, $this->mStripState
) );
1198 $s .= $prefix . '[[' . $line;
1202 //check other language variants of the link
1203 //if the article does not exist
1205 $variants = $wgContLang->getVariants();
1207 if(sizeof($variants) > 1 && $convertCount < 200) {
1209 if($nt->getArticleID() == 0) {
1210 foreach ( $variants as $v ) {
1211 if($v == $wgContLang->getPreferredVariant())
1214 $varlink = $wgContLang->autoConvert($link, $v);
1215 $varnt = Title
::newFromText($varlink);
1216 if($varnt && $varnt->getArticleID()>0) {
1221 if($varnt && $varnt->getArticleID()>0) {
1227 $ns = $nt->getNamespace();
1228 $iw = $nt->getInterWiki();
1230 if ($might_be_img) { # if this is actually an invalid link
1231 if ($ns == NS_IMAGE
&& $noforce) { #but might be an image
1233 while (isset ($a[$k+
1]) ) {
1234 #look at the next 'line' to see if we can close it there
1235 $next_line = array_shift(array_splice( $a, $k +
1, 1) );
1236 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1237 # the first ]] closes the inner link, the second the image
1239 $text .= '[[' . $m[1];
1242 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1243 #if there's exactly one ]] that's fine, we'll keep looking
1244 $text .= '[[' . $m[0];
1246 #if $next_line is invalid too, we need look no further
1247 $text .= '[[' . $next_line;
1252 # we couldn't find the end of this imageLink, so output it raw
1253 #but don't ignore what might be perfectly normal links in the text we've examined
1254 $text = $this->replaceInternalLinks($text);
1255 $s .= $prefix . '[[' . $link . '|' . $text;
1256 # note: no $trail, because without an end, there *is* no trail
1259 } else { #it's not an image, so output it raw
1260 $s .= $prefix . '[[' . $link . '|' . $text;
1261 # note: no $trail, because without an end, there *is* no trail
1266 $wasblank = ( '' == $text );
1267 if( $wasblank ) $text = $link;
1270 # Link not escaped by : , create the various objects
1274 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1275 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1276 $tmp = $prefix . $trail ;
1277 $s .= (trim($tmp) == '')?
'': $tmp;
1281 if ( $ns == NS_IMAGE
) {
1282 # recursively parse links inside the image caption
1283 # actually, this will parse them in any other parameters, too,
1284 # but it might be hard to fix that, and it doesn't matter ATM
1285 $text = $this->replaceExternalLinks($text);
1286 $text = $this->replaceInternalLinks($text);
1288 # replace the image with a link-holder so that replaceExternalLinks() can't mess with it
1289 $s .= $prefix . $this->insertStripItem( $sk->makeImageLinkObj( $nt, $text ), $this->mStripState
) . $trail;
1290 $wgLinkCache->addImageLinkObj( $nt );
1294 if ( $ns == NS_CATEGORY
) {
1295 $t = $nt->getText() ;
1297 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1298 $pPLC=$sk->postParseLinkColour();
1299 $sk->postParseLinkColour( false );
1300 $t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
1301 $sk->postParseLinkColour( $pPLC );
1302 $wgLinkCache->resume();
1305 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1306 $sortkey = $this->mTitle
->getText();
1308 $sortkey = $this->mTitle
->getPrefixedText();
1313 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1314 $this->mOutput
->mCategoryLinks
[] = $t ;
1315 $s .= $prefix . $trail ;
1320 $text = $wgContLang->convert($text);
1322 if( ( $nt->getPrefixedText() === $this->mTitle
->getPrefixedText() ) &&
1323 ( $nt->getFragment() === '' ) ) {
1324 # Self-links are handled specially; generally de-link and change to bold.
1325 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1329 # Special and Media are pseudo-namespaces; no pages actually exist in them
1330 if( $ns == NS_MEDIA
) {
1331 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1332 $wgLinkCache->addImageLinkObj( $nt );
1334 } elseif( $ns == NS_SPECIAL
) {
1335 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1338 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1340 wfProfileOut( $fname );
1345 * Handle link to subpage if necessary
1346 * @param string $target the source of the link
1347 * @param string &$text the link text, modified as necessary
1348 * @return string the full name of the link
1351 function maybeDoSubpageLink($target, &$text) {
1354 # :Foobar -- override special treatment of prefix (images, language links)
1355 # /Foobar -- convert to CurrentPage/Foobar
1356 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1357 global $wgNamespacesWithSubpages;
1359 $fname = 'Parser::maybeDoSubpageLink';
1360 wfProfileIn( $fname );
1361 # Look at the first character
1362 if( $target{0} == '/' ) {
1363 # / at end means we don't want the slash to be shown
1364 if(substr($target,-1,1)=='/') {
1365 $target=substr($target,1,-1);
1368 $noslash=substr($target,1);
1371 # Some namespaces don't allow subpages
1372 if(!empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()])) {
1373 # subpages allowed here
1374 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1375 if( '' === $text ) {
1377 } # this might be changed for ugliness reasons
1379 # no subpage allowed, use standard link
1387 wfProfileOut( $fname );
1392 * Used by doBlockLevels()
1395 /* private */ function closeParagraph() {
1397 if ( '' != $this->mLastSection
) {
1398 $result = '</' . $this->mLastSection
. ">\n";
1400 $this->mInPre
= false;
1401 $this->mLastSection
= '';
1404 # getCommon() returns the length of the longest common substring
1405 # of both arguments, starting at the beginning of both.
1407 /* private */ function getCommon( $st1, $st2 ) {
1408 $fl = strlen( $st1 );
1409 $shorter = strlen( $st2 );
1410 if ( $fl < $shorter ) { $shorter = $fl; }
1412 for ( $i = 0; $i < $shorter; ++
$i ) {
1413 if ( $st1{$i} != $st2{$i} ) { break; }
1417 # These next three functions open, continue, and close the list
1418 # element appropriate to the prefix character passed into them.
1420 /* private */ function openList( $char ) {
1421 $result = $this->closeParagraph();
1423 if ( '*' == $char ) { $result .= '<ul><li>'; }
1424 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1425 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1426 else if ( ';' == $char ) {
1427 $result .= '<dl><dt>';
1428 $this->mDTopen
= true;
1430 else { $result = '<!-- ERR 1 -->'; }
1435 /* private */ function nextItem( $char ) {
1436 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1437 else if ( ':' == $char ||
';' == $char ) {
1439 if ( $this->mDTopen
) { $close = '</dt>'; }
1440 if ( ';' == $char ) {
1441 $this->mDTopen
= true;
1442 return $close . '<dt>';
1444 $this->mDTopen
= false;
1445 return $close . '<dd>';
1448 return '<!-- ERR 2 -->';
1451 /* private */ function closeList( $char ) {
1452 if ( '*' == $char ) { $text = '</li></ul>'; }
1453 else if ( '#' == $char ) { $text = '</li></ol>'; }
1454 else if ( ':' == $char ) {
1455 if ( $this->mDTopen
) {
1456 $this->mDTopen
= false;
1457 $text = '</dt></dl>';
1459 $text = '</dd></dl>';
1462 else { return '<!-- ERR 3 -->'; }
1468 * Make lists from lines starting with ':', '*', '#', etc.
1471 * @return string the lists rendered as HTML
1473 function doBlockLevels( $text, $linestart ) {
1474 $fname = 'Parser::doBlockLevels';
1475 wfProfileIn( $fname );
1477 # Parsing through the text line by line. The main thing
1478 # happening here is handling of block-level elements p, pre,
1479 # and making lists from lines starting with * # : etc.
1481 $textLines = explode( "\n", $text );
1483 $lastPrefix = $output = $lastLine = '';
1484 $this->mDTopen
= $inBlockElem = false;
1486 $paragraphStack = false;
1488 if ( !$linestart ) {
1489 $output .= array_shift( $textLines );
1491 foreach ( $textLines as $oLine ) {
1492 $lastPrefixLength = strlen( $lastPrefix );
1493 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1494 $preOpenMatch = preg_match('/<pre/i', $oLine );
1495 if ( !$this->mInPre
) {
1496 # Multiple prefixes may abut each other for nested lists.
1497 $prefixLength = strspn( $oLine, '*#:;' );
1498 $pref = substr( $oLine, 0, $prefixLength );
1501 $pref2 = str_replace( ';', ':', $pref );
1502 $t = substr( $oLine, $prefixLength );
1503 $this->mInPre
= !empty($preOpenMatch);
1505 # Don't interpret any other prefixes in preformatted text
1507 $pref = $pref2 = '';
1512 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1513 # Same as the last item, so no need to deal with nesting or opening stuff
1514 $output .= $this->nextItem( substr( $pref, -1 ) );
1515 $paragraphStack = false;
1517 if ( substr( $pref, -1 ) == ';') {
1518 # The one nasty exception: definition lists work like this:
1519 # ; title : definition text
1520 # So we check for : in the remainder text to split up the
1521 # title and definition, without b0rking links.
1522 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1524 $output .= $term . $this->nextItem( ':' );
1527 } elseif( $prefixLength ||
$lastPrefixLength ) {
1528 # Either open or close a level...
1529 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1530 $paragraphStack = false;
1532 while( $commonPrefixLength < $lastPrefixLength ) {
1533 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1534 --$lastPrefixLength;
1536 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1537 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1539 while ( $prefixLength > $commonPrefixLength ) {
1540 $char = substr( $pref, $commonPrefixLength, 1 );
1541 $output .= $this->openList( $char );
1543 if ( ';' == $char ) {
1544 # FIXME: This is dupe of code above
1545 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1547 $output .= $term . $this->nextItem( ':' );
1550 ++
$commonPrefixLength;
1552 $lastPrefix = $pref2;
1554 if( 0 == $prefixLength ) {
1555 # No prefix (not in list)--go to paragraph mode
1556 $uniq_prefix = UNIQ_PREFIX
;
1557 // XXX: use a stack for nestable elements like span, table and div
1558 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1559 $closematch = preg_match(
1560 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1561 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1562 if ( $openmatch or $closematch ) {
1563 $paragraphStack = false;
1564 $output .= $this->closeParagraph();
1565 if($preOpenMatch and !$preCloseMatch) {
1566 $this->mInPre
= true;
1568 if ( $closematch ) {
1569 $inBlockElem = false;
1571 $inBlockElem = true;
1573 } else if ( !$inBlockElem && !$this->mInPre
) {
1574 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1576 if ($this->mLastSection
!= 'pre') {
1577 $paragraphStack = false;
1578 $output .= $this->closeParagraph().'<pre>';
1579 $this->mLastSection
= 'pre';
1581 $t = substr( $t, 1 );
1584 if ( '' == trim($t) ) {
1585 if ( $paragraphStack ) {
1586 $output .= $paragraphStack.'<br />';
1587 $paragraphStack = false;
1588 $this->mLastSection
= 'p';
1590 if ($this->mLastSection
!= 'p' ) {
1591 $output .= $this->closeParagraph();
1592 $this->mLastSection
= '';
1593 $paragraphStack = '<p>';
1595 $paragraphStack = '</p><p>';
1599 if ( $paragraphStack ) {
1600 $output .= $paragraphStack;
1601 $paragraphStack = false;
1602 $this->mLastSection
= 'p';
1603 } else if ($this->mLastSection
!= 'p') {
1604 $output .= $this->closeParagraph().'<p>';
1605 $this->mLastSection
= 'p';
1611 if ($paragraphStack === false) {
1615 while ( $prefixLength ) {
1616 $output .= $this->closeList( $pref2{$prefixLength-1} );
1619 if ( '' != $this->mLastSection
) {
1620 $output .= '</' . $this->mLastSection
. '>';
1621 $this->mLastSection
= '';
1624 wfProfileOut( $fname );
1629 * Split up a string on ':', ignoring any occurences inside
1630 * <a>..</a> or <span>...</span>
1631 * @param string $str the string to split
1632 * @param string &$before set to everything before the ':'
1633 * @param string &$after set to everything after the ':'
1634 * return string the position of the ':', or false if none found
1636 function findColonNoLinks($str, &$before, &$after) {
1637 # I wonder if we should make this count all tags, not just <a>
1638 # and <span>. That would prevent us from matching a ':' that
1639 # comes in the middle of italics other such formatting....
1641 $fname = 'Parser::findColonNoLinks';
1642 wfProfileIn( $fname );
1645 $colon = strpos($str, ':', $pos);
1647 if ($colon !== false) {
1648 $before = substr($str, 0, $colon);
1649 $after = substr($str, $colon +
1);
1651 # Skip any ':' within <a> or <span> pairs
1652 $a = substr_count($before, '<a');
1653 $s = substr_count($before, '<span');
1654 $ca = substr_count($before, '</a>');
1655 $cs = substr_count($before, '</span>');
1657 if ($a <= $ca and $s <= $cs) {
1658 # Tags are balanced before ':'; ok
1663 } while ($colon !== false);
1664 wfProfileOut( $fname );
1669 * Return value of a magic variable (like PAGENAME)
1673 function getVariableValue( $index ) {
1674 global $wgContLang, $wgSitename, $wgServer;
1677 case MAG_CURRENTMONTH
:
1678 return $wgContLang->formatNum( date( 'm' ) );
1679 case MAG_CURRENTMONTHNAME
:
1680 return $wgContLang->getMonthName( date('n') );
1681 case MAG_CURRENTMONTHNAMEGEN
:
1682 return $wgContLang->getMonthNameGen( date('n') );
1683 case MAG_CURRENTDAY
:
1684 return $wgContLang->formatNum( date('j') );
1686 return $this->mTitle
->getText();
1688 return $this->mTitle
->getPartialURL();
1690 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1691 return $wgContLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1692 case MAG_CURRENTDAYNAME
:
1693 return $wgContLang->getWeekdayName( date('w')+
1 );
1694 case MAG_CURRENTYEAR
:
1695 return $wgContLang->formatNum( date( 'Y' ) );
1696 case MAG_CURRENTTIME
:
1697 return $wgContLang->time( wfTimestampNow(), false );
1698 case MAG_NUMBEROFARTICLES
:
1699 return $wgContLang->formatNum( wfNumberOfArticles() );
1710 * initialise the magic variables (like CURRENTMONTHNAME)
1714 function initialiseVariables() {
1715 $fname = 'Parser::initialiseVariables';
1716 wfProfileIn( $fname );
1717 global $wgVariableIDs;
1718 $this->mVariables
= array();
1719 foreach ( $wgVariableIDs as $id ) {
1720 $mw =& MagicWord
::get( $id );
1721 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1723 wfProfileOut( $fname );
1727 * Replace magic variables, templates, and template arguments
1728 * with the appropriate text. Templates are substituted recursively,
1729 * taking care to avoid infinite loops.
1731 * Note that the substitution depends on value of $mOutputType:
1732 * OT_WIKI: only {{subst:}} templates
1733 * OT_MSG: only magic variables
1734 * OT_HTML: all templates and magic variables
1736 * @param string $tex The text to transform
1737 * @param array $args Key-value pairs representing template parameters to substitute
1740 function replaceVariables( $text, $args = array() ) {
1741 global $wgLang, $wgScript, $wgArticlePath;
1743 # Prevent too big inclusions
1744 if(strlen($text)> MAX_INCLUDE_SIZE
)
1747 $fname = 'Parser::replaceVariables';
1748 wfProfileIn( $fname );
1750 $titleChars = Title
::legalChars();
1752 # This function is called recursively. To keep track of arguments we need a stack:
1753 array_push( $this->mArgStack
, $args );
1755 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1756 $GLOBALS['wgCurParser'] =& $this;
1758 # Variable substitution
1759 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", 'wfVariableSubstitution', $text );
1761 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1762 # Argument substitution
1763 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1765 # Template substitution
1766 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1767 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1769 array_pop( $this->mArgStack
);
1771 wfProfileOut( $fname );
1776 * Replace magic variables
1779 function variableSubstitution( $matches ) {
1780 if ( !$this->mVariables
) {
1781 $this->initialiseVariables();
1784 if ( $this->mOutputType
== OT_WIKI
) {
1785 # Do only magic variables prefixed by SUBST
1786 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1787 if (!$mwSubst->matchStartAndRemove( $matches[1] ))
1789 # Note that if we don't substitute the variable below,
1790 # we don't remove the {{subst:}} magic word, in case
1791 # it is a template rather than a magic variable.
1793 if ( !$skip && array_key_exists( $matches[1], $this->mVariables
) ) {
1794 $text = $this->mVariables
[$matches[1]];
1795 $this->mOutput
->mContainsOldMagic
= true;
1797 $text = $matches[0];
1802 # Split template arguments
1803 function getTemplateArgs( $argsString ) {
1804 if ( $argsString === '' ) {
1808 $args = explode( '|', substr( $argsString, 1 ) );
1810 # If any of the arguments contains a '[[' but no ']]', it needs to be
1811 # merged with the next arg because the '|' character between belongs
1812 # to the link syntax and not the template parameter syntax.
1813 $argc = count($args);
1815 for ( $i = 0; $i < $argc-1; $i++
) {
1816 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1817 $args[$i] .= '|'.$args[$i+
1];
1818 array_splice($args, $i+
1, 1);
1828 * Return the text of a template, after recursively
1829 * replacing any variables or templates within the template.
1831 * @param array $matches The parts of the template
1832 * $matches[1]: the title, i.e. the part before the |
1833 * $matches[2]: the parameters (including a leading |), if any
1834 * @return string the text of the template
1837 function braceSubstitution( $matches ) {
1838 global $wgLinkCache, $wgContLang;
1839 $fname = 'Parser::braceSubstitution';
1846 # Need to know if the template comes at the start of a line,
1847 # to treat the beginning of the template like the beginning
1848 # of a line for tables and block-level elements.
1849 $linestart = $matches[1];
1851 # $part1 is the bit before the first |, and must contain only title characters
1852 # $args is a list of arguments, starting from index 0, not including $part1
1854 $part1 = $matches[2];
1855 # If the third subpattern matched anything, it will start with |
1857 $args = $this->getTemplateArgs($matches[3]);
1858 $argc = count( $args );
1860 # Don't parse {{{}}} because that's only for template arguments
1861 if ( $linestart === '{' ) {
1862 $text = $matches[0];
1869 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1870 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
1871 # One of two possibilities is true:
1872 # 1) Found SUBST but not in the PST phase
1873 # 2) Didn't find SUBST and in the PST phase
1874 # In either case, return without further processing
1875 $text = $matches[0];
1881 # MSG, MSGNW and INT
1884 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1885 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1888 # Remove obsolete MSG:
1889 $mwMsg =& MagicWord
::get( MAG_MSG
);
1890 $mwMsg->matchStartAndRemove( $part1 );
1893 # Check if it is an internal message
1894 $mwInt =& MagicWord
::get( MAG_INT
);
1895 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1896 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1897 $text = $linestart . wfMsgReal( $part1, $args, true );
1905 # Check for NS: (namespace expansion)
1906 $mwNs = MagicWord
::get( MAG_NS
);
1907 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1908 if ( intval( $part1 ) ) {
1909 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1912 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1913 if ( !is_null( $index ) ) {
1914 $text = $linestart . $wgContLang->getNsText( $index );
1921 # LOCALURL and LOCALURLE
1923 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1924 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1926 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1927 $func = 'getLocalURL';
1928 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1929 $func = 'escapeLocalURL';
1934 if ( $func !== '' ) {
1935 $title = Title
::newFromText( $part1 );
1936 if ( !is_null( $title ) ) {
1938 $text = $linestart . $title->$func( $args[0] );
1940 $text = $linestart . $title->$func();
1948 if ( !$found && $argc == 1 ) {
1949 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
1950 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1951 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
1956 # Template table test
1958 # Did we encounter this template already? If yes, it is in the cache
1959 # and we need to check for loops.
1960 if ( !$found && isset( $this->mTemplates
[$part1] ) ) {
1961 # set $text to cached message.
1962 $text = $linestart . $this->mTemplates
[$part1];
1965 # Infinite loop test
1966 if ( isset( $this->mTemplatePath
[$part1] ) ) {
1969 $text .= '<!-- WARNING: template loop detected -->';
1973 # Load from database
1974 $itcamefromthedatabase = false;
1977 $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
1978 if ($subpage !== '') {
1979 $ns = $this->mTitle
->getNamespace();
1981 $title = Title
::newFromText( $part1, $ns );
1982 if ( !is_null( $title ) && !$title->isExternal() ) {
1983 # Check for excessive inclusion
1984 $dbk = $title->getPrefixedDBkey();
1985 if ( $this->incrementIncludeCount( $dbk ) ) {
1986 # This should never be reached.
1987 $article = new Article( $title );
1988 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1989 if ( $articleContent !== false ) {
1991 $text = $linestart . $articleContent;
1992 $itcamefromthedatabase = true;
1996 # If the title is valid but undisplayable, make a link to it
1997 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1998 $text = $linestart . '[['.$title->getPrefixedText().']]';
2002 # Template cache array insertion
2003 $this->mTemplates
[$part1] = $text;
2007 # Recursive parsing, escaping and link table handling
2008 # Only for HTML output
2009 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
2010 $text = wfEscapeWikiText( $text );
2011 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
2012 # Clean up argument array
2013 $assocArgs = array();
2015 foreach( $args as $arg ) {
2016 $eqpos = strpos( $arg, '=' );
2017 if ( $eqpos === false ) {
2018 $assocArgs[$index++
] = $arg;
2020 $name = trim( substr( $arg, 0, $eqpos ) );
2021 $value = trim( substr( $arg, $eqpos+
1 ) );
2022 if ( $value === false ) {
2025 if ( $name !== false ) {
2026 $assocArgs[$name] = $value;
2031 # Add a new element to the templace recursion path
2032 $this->mTemplatePath
[$part1] = 1;
2034 $text = $this->strip( $text, $this->mStripState
);
2035 $text = $this->removeHTMLtags( $text );
2036 $text = $this->replaceVariables( $text, $assocArgs );
2038 # Resume the link cache and register the inclusion as a link
2039 if ( $this->mOutputType
== OT_HTML
&& !is_null( $title ) ) {
2040 $wgLinkCache->addLinkObj( $title );
2043 # If the template begins with a table or block-level
2044 # element, it should be treated as beginning a new line.
2045 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
2046 $text = "\n" . $text;
2050 # Empties the template path
2051 $this->mTemplatePath
= array();
2055 # replace ==section headers==
2056 # XXX this needs to go away once we have a better parser.
2057 if ( $this->mOutputType
!= OT_WIKI
&& $itcamefromthedatabase ) {
2058 if( !is_null( $title ) )
2059 $encodedname = base64_encode($title->getPrefixedDBkey());
2061 $encodedname = base64_encode("");
2062 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
2063 PREG_SPLIT_DELIM_CAPTURE
);
2066 for( $i = 0; $i < count($m); $i +
= 2 ) {
2068 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
2070 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
2074 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
2075 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
2076 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
2083 # Empties the template path
2084 $this->mTemplatePath
= array();
2093 * Triple brace replacement -- used for template arguments
2096 function argSubstitution( $matches ) {
2097 $arg = trim( $matches[1] );
2098 $text = $matches[0];
2099 $inputArgs = end( $this->mArgStack
);
2101 if ( array_key_exists( $arg, $inputArgs ) ) {
2102 $text = $inputArgs[$arg];
2109 * Returns true if the function is allowed to include this entity
2112 function incrementIncludeCount( $dbk ) {
2113 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
2114 $this->mIncludeCount
[$dbk] = 0;
2116 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
2125 * Cleans up HTML, removes dangerous tags and attributes, and
2126 * removes HTML comments
2129 function removeHTMLtags( $text ) {
2130 global $wgUseTidy, $wgUserHtml;
2131 $fname = 'Parser::removeHTMLtags';
2132 wfProfileIn( $fname );
2135 $htmlpairs = array( # Tags that must be closed
2136 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
2137 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
2138 'strike', 'strong', 'tt', 'var', 'div', 'center',
2139 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
2140 'ruby', 'rt' , 'rb' , 'rp', 'p'
2142 $htmlsingle = array(
2143 'br', 'hr', 'li', 'dt', 'dd'
2145 $htmlnest = array( # Tags that can be nested--??
2146 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
2147 'dl', 'font', 'big', 'small', 'sub', 'sup'
2149 $tabletags = array( # Can only appear inside table
2153 $htmlpairs = array();
2154 $htmlsingle = array();
2155 $htmlnest = array();
2156 $tabletags = array();
2159 $htmlsingle = array_merge( $tabletags, $htmlsingle );
2160 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
2162 $htmlattrs = $this->getHTMLattrs () ;
2164 # Remove HTML comments
2165 $text = $this->removeHTMLcomments( $text );
2167 $bits = explode( '<', $text );
2168 $text = array_shift( $bits );
2170 $tagstack = array(); $tablestack = array();
2171 foreach ( $bits as $x ) {
2172 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
2173 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2175 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2176 error_reporting( $prev );
2179 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2183 if ( ! in_array( $t, $htmlsingle ) &&
2184 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
2185 @array_push
( $tagstack, $ot );
2188 if ( $t == 'table' ) {
2189 $tagstack = array_pop( $tablestack );
2194 # Keep track for later
2195 if ( in_array( $t, $tabletags ) &&
2196 ! in_array( 'table', $tagstack ) ) {
2198 } else if ( in_array( $t, $tagstack ) &&
2199 ! in_array ( $t , $htmlnest ) ) {
2201 } else if ( ! in_array( $t, $htmlsingle ) ) {
2202 if ( $t == 'table' ) {
2203 array_push( $tablestack, $tagstack );
2204 $tagstack = array();
2206 array_push( $tagstack, $t );
2208 # Strip non-approved attributes from the tag
2209 $newparams = $this->fixTagAttributes($params);
2213 $rest = str_replace( '>', '>', $rest );
2214 $text .= "<$slash$t $newparams$brace$rest";
2218 $text .= '<' . str_replace( '>', '>', $x);
2220 # Close off any remaining tags
2221 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2223 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2226 # this might be possible using tidy itself
2227 foreach ( $bits as $x ) {
2228 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2230 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2231 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2232 $newparams = $this->fixTagAttributes($params);
2233 $rest = str_replace( '>', '>', $rest );
2234 $text .= "<$slash$t $newparams$brace$rest";
2236 $text .= '<' . str_replace( '>', '>', $x);
2240 wfProfileOut( $fname );
2245 * Remove '<!--', '-->', and everything between.
2246 * To avoid leaving blank lines, when a comment is both preceded
2247 * and followed by a newline (ignoring spaces), trim leading and
2248 * trailing spaces and one of the newlines.
2252 function removeHTMLcomments( $text ) {
2253 $fname='Parser::removeHTMLcomments';
2254 wfProfileIn( $fname );
2255 while (($start = strpos($text, '<!--')) !== false) {
2256 $end = strpos($text, '-->', $start +
4);
2257 if ($end === false) {
2258 # Unterminated comment; bail out
2264 # Trim space and newline if the comment is both
2265 # preceded and followed by a newline
2266 $spaceStart = max($start - 1, 0);
2267 $spaceLen = $end - $spaceStart;
2268 while (substr($text, $spaceStart, 1) === ' ' && $spaceStart > 0) {
2272 while (substr($text, $spaceStart +
$spaceLen, 1) === ' ')
2274 if (substr($text, $spaceStart, 1) === "\n" and substr($text, $spaceStart +
$spaceLen, 1) === "\n") {
2275 # Remove the comment, leading and trailing
2276 # spaces, and leave only one newline.
2277 $text = substr_replace($text, "\n", $spaceStart, $spaceLen +
1);
2280 # Remove just the comment.
2281 $text = substr_replace($text, '', $start, $end - $start);
2284 wfProfileOut( $fname );
2289 * This function accomplishes several tasks:
2290 * 1) Auto-number headings if that option is enabled
2291 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2292 * 3) Add a Table of contents on the top for users who have enabled the option
2293 * 4) Auto-anchor headings
2295 * It loops through all headlines, collects the necessary data, then splits up the
2296 * string and re-inserts the newly formatted headlines.
2299 /* private */ function formatHeadings( $text, $isMain=true ) {
2300 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders;
2302 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2303 $doShowToc = $this->mOptions
->getShowToc();
2304 $forceTocHere = false;
2305 if( !$this->mTitle
->userCanEdit() ) {
2307 $rightClickHack = 0;
2309 $showEditLink = $this->mOptions
->getEditSection();
2310 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
2313 # Inhibit editsection links if requested in the page
2314 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2315 if( $esw->matchAndRemove( $text ) ) {
2318 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2320 $mw =& MagicWord
::get( MAG_NOTOC
);
2321 if( $mw->matchAndRemove( $text ) ) {
2325 # never add the TOC to the Main Page. This is an entry page that should not
2326 # be more than 1-2 screens large anyway
2327 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
2331 # Get all headlines for numbering them and adding funky stuff like [edit]
2332 # links - this is for later, but we need the number of headlines right now
2333 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
2335 # if there are fewer than 4 headlines in the article, do not show TOC
2336 if( $numMatches < 4 ) {
2340 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2341 # override above conditions and always show TOC at that place
2342 $mw =& MagicWord
::get( MAG_TOC
);
2343 if ($mw->match( $text ) ) {
2345 $forceTocHere = true;
2347 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2348 # override above conditions and always show TOC above first header
2349 $mw =& MagicWord
::get( MAG_FORCETOC
);
2350 if ($mw->matchAndRemove( $text ) ) {
2357 # We need this to perform operations on the HTML
2358 $sk =& $this->mOptions
->getSkin();
2362 $sectionCount = 0; # headlineCount excluding template sections
2364 # Ugh .. the TOC should have neat indentation levels which can be
2365 # passed to the skin functions. These are determined here
2370 $sublevelCount = array();
2373 foreach( $matches[3] as $headline ) {
2375 $templatetitle = "";
2376 $templatesection = 0;
2378 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2380 $templatetitle = base64_decode($mat[1]);
2381 $templatesection = 1 +
(int)base64_decode($mat[2]);
2382 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2387 $prevlevel = $level;
2389 $level = $matches[1][$headlineCount];
2390 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2391 # reset when we enter a new level
2392 $sublevelCount[$level] = 0;
2393 $toc .= $sk->tocIndent( $level - $prevlevel );
2394 $toclevel +
= $level - $prevlevel;
2396 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2397 # reset when we step back a level
2398 $sublevelCount[$level+
1]=0;
2399 $toc .= $sk->tocUnindent( $prevlevel - $level );
2400 $toclevel -= $prevlevel - $level;
2402 # count number of headlines for each level
2403 @$sublevelCount[$level]++
;
2404 if( $doNumberHeadings ||
$doShowToc ) {
2406 for( $i = 1; $i <= $level; $i++
) {
2407 if( !empty( $sublevelCount[$i] ) ) {
2411 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2417 # The canonized header is a version of the header text safe to use for links
2418 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2419 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2420 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2422 # Remove link placeholders by the link text.
2423 # <!--LINK number-->
2425 # link text with suffix
2426 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2427 "\$wgLinkHolders['texts'][\$1]",
2428 $canonized_headline );
2431 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2432 $tocline = trim( $canonized_headline );
2433 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2434 $replacearray = array(
2438 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2439 $refer[$headlineCount] = $canonized_headline;
2441 # count how many in assoc. array so we can track dupes in anchors
2442 @$refers[$canonized_headline]++
;
2443 $refcount[$headlineCount]=$refers[$canonized_headline];
2445 # Prepend the number to the heading text
2447 if( $doNumberHeadings ||
$doShowToc ) {
2448 $tocline = $numbering . ' ' . $tocline;
2450 # Don't number the heading if it is the only one (looks silly)
2451 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2452 # the two are different if the line contains a link
2453 $headline=$numbering . ' ' . $headline;
2457 # Create the anchor for linking from the TOC to the section
2458 $anchor = $canonized_headline;
2459 if($refcount[$headlineCount] > 1 ) {
2460 $anchor .= '_' . $refcount[$headlineCount];
2462 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2463 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2465 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2466 if ( empty( $head[$headlineCount] ) ) {
2467 $head[$headlineCount] = '';
2470 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2472 $head[$headlineCount] .= $sk->editSectionLink($this->mTitle
, $sectionCount+
1);
2475 # Add the edit section span
2476 if( $rightClickHack ) {
2478 $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
2480 $headline = $sk->editSectionScript($this->title
, $sectionCount+
1,$headline);
2483 # give headline the correct <h#> tag
2484 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2492 $toclines = $headlineCount;
2493 $toc .= $sk->tocUnindent( $toclevel );
2494 $toc = $sk->tocTable( $toc );
2497 # split up and insert constructed headlines
2499 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2502 foreach( $blocks as $block ) {
2503 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2504 # This is the [edit] link that appears for the top block of text when
2505 # section editing is enabled
2507 # Disabled because it broke block formatting
2508 # For example, a bullet point in the top line
2509 # $full .= $sk->editSectionLink(0);
2512 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2513 # Top anchor now in skin
2517 if( !empty( $head[$i] ) ) {
2523 $mw =& MagicWord
::get( MAG_TOC
);
2524 return $mw->replace( $toc, $full );
2531 * Return an HTML link for the "ISBN 123456" text
2534 function magicISBN( $text ) {
2536 $fname = 'Parser::magicISBN';
2537 wfProfileIn( $fname );
2539 $a = split( 'ISBN ', ' '.$text );
2540 if ( count ( $a ) < 2 ) {
2541 wfProfileOut( $fname );
2544 $text = substr( array_shift( $a ), 1);
2545 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2547 foreach ( $a as $x ) {
2548 $isbn = $blank = '' ;
2549 while ( ' ' == $x{0} ) {
2551 $x = substr( $x, 1 );
2553 if ( $x == '' ) { # blank isbn
2554 $text .= "ISBN $blank";
2557 while ( strstr( $valid, $x{0} ) != false ) {
2559 $x = substr( $x, 1 );
2561 $num = str_replace( '-', '', $isbn );
2562 $num = str_replace( ' ', '', $num );
2565 $text .= "ISBN $blank$x";
2567 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2568 $text .= '<a href="' .
2569 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2570 "\" class=\"internal\">ISBN $isbn</a>";
2574 wfProfileOut( $fname );
2579 * Return an HTML link for the "GEO ..." text
2582 function magicGEO( $text ) {
2583 global $wgLang, $wgUseGeoMode;
2584 $fname = 'Parser::magicGEO';
2585 wfProfileIn( $fname );
2587 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2588 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2589 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2590 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2591 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2592 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2594 $a = split( 'GEO ', ' '.$text );
2595 if ( count ( $a ) < 2 ) {
2596 wfProfileOut( $fname );
2599 $text = substr( array_shift( $a ), 1);
2600 $valid = '0123456789.+-:';
2602 foreach ( $a as $x ) {
2603 $geo = $blank = '' ;
2604 while ( ' ' == $x{0} ) {
2606 $x = substr( $x, 1 );
2608 while ( strstr( $valid, $x{0} ) != false ) {
2610 $x = substr( $x, 1 );
2612 $num = str_replace( '+', '', $geo );
2613 $num = str_replace( ' ', '', $num );
2615 if ( '' == $num ||
count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2616 $text .= "GEO $blank$x";
2618 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2619 $text .= '<a href="' .
2620 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2621 "\" class=\"internal\">GEO $geo</a>";
2625 wfProfileOut( $fname );
2630 * Return an HTML link for the "RFC 1234" text
2632 * @param string $text text to be processed
2634 function magicRFC( $text ) {
2637 $valid = '0123456789';
2640 $a = split( 'RFC ', ' '.$text );
2641 if ( count ( $a ) < 2 ) return $text;
2642 $text = substr( array_shift( $a ), 1);
2644 /* Check if RFC keyword is preceed by [[.
2645 * This test is made here cause of the array_shift above
2646 * that prevent the test to be done in the foreach.
2648 if(substr($text, -2) == '[[') { $internal = true; }
2650 foreach ( $a as $x ) {
2651 /* token might be empty if we have RFC RFC 1234 */
2657 $rfc = $blank = '' ;
2659 /** remove and save whitespaces in $blank */
2660 while ( $x{0} == ' ' ) {
2662 $x = substr( $x, 1 );
2665 /** remove and save the rfc number in $rfc */
2666 while ( strstr( $valid, $x{0} ) != false ) {
2668 $x = substr( $x, 1 );
2672 /* call back stripped spaces*/
2673 $text .= "RFC $blank$x";
2674 } elseif( $internal) {
2676 $text .= "RFC $rfc$x";
2678 /* build the external link*/
2679 $url = wfmsg( 'rfcurl' );
2680 $url = str_replace( '$1', $rfc, $url);
2681 $sk =& $this->mOptions
->getSkin();
2682 $la = $sk->getExternalLinkAttributes( $url, 'RFC '.$rfc );
2683 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2686 /* Check if the next RFC keyword is preceed by [[ */
2687 $internal = (substr($x,-2) == '[[');
2693 * Transform wiki markup when saving a page by doing \r\n -> \n
2694 * conversion, substitting signatures, {{subst:}} templates, etc.
2696 * @param string $text the text to transform
2697 * @param Title &$title the Title object for the current article
2698 * @param User &$user the User object describing the current user
2699 * @param ParserOptions $options parsing options
2700 * @param bool $clearState whether to clear the parser state first
2701 * @return string the altered wiki markup
2704 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2705 $this->mOptions
= $options;
2706 $this->mTitle
=& $title;
2707 $this->mOutputType
= OT_WIKI
;
2709 if ( $clearState ) {
2710 $this->clearState();
2713 $stripState = false;
2717 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2721 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2722 "/<br *?>/i" => "<br />",
2724 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2726 $text = $this->strip( $text, $stripState, false );
2727 $text = $this->pstPass2( $text, $user );
2728 $text = $this->unstrip( $text, $stripState );
2729 $text = $this->unstripNoWiki( $text, $stripState );
2734 * Pre-save transform helper function
2737 function pstPass2( $text, &$user ) {
2738 global $wgLang, $wgContLang, $wgLocaltimezone, $wgCurParser;
2740 # Variable replacement
2741 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2742 $text = $this->replaceVariables( $text );
2746 $n = $user->getName();
2747 $k = $user->getOption( 'nickname' );
2748 if ( '' == $k ) { $k = $n; }
2749 if(isset($wgLocaltimezone)) {
2750 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2752 /* Note: this is an ugly timezone hack for the European wikis */
2753 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
2754 ' (' . date( 'T' ) . ')';
2755 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2757 $text = preg_replace( '/~~~~~/', $d, $text );
2758 $text = preg_replace( '/~~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]] $d", $text );
2759 $text = preg_replace( '/~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]]", $text );
2761 # Context links: [[|name]] and [[name (context)|]]
2763 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2764 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2765 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2766 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2768 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2769 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2770 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2771 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2773 $t = $this->mTitle
->getText();
2774 if ( preg_match( $conpat, $t, $m ) ) {
2777 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2778 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2779 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2781 if ( '' == $context ) {
2782 $text = preg_replace( $p2, '[[\\1]]', $text );
2784 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2787 # Trim trailing whitespace
2788 # MAG_END (__END__) tag allows for trailing
2789 # whitespace to be deliberately included
2790 $text = rtrim( $text );
2791 $mw =& MagicWord
::get( MAG_END
);
2792 $mw->matchAndRemove( $text );
2798 * Set up some variables which are usually set up in parse()
2799 * so that an external function can call some class members with confidence
2802 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2803 $this->mTitle
=& $title;
2804 $this->mOptions
= $options;
2805 $this->mOutputType
= $outputType;
2806 if ( $clearState ) {
2807 $this->clearState();
2812 * Transform a MediaWiki message by replacing magic variables.
2814 * @param string $text the text to transform
2815 * @param ParserOptions $options options
2816 * @return string the text with variables substituted
2819 function transformMsg( $text, $options ) {
2821 static $executing = false;
2823 # Guard against infinite recursion
2829 $this->mTitle
= $wgTitle;
2830 $this->mOptions
= $options;
2831 $this->mOutputType
= OT_MSG
;
2832 $this->clearState();
2833 $text = $this->replaceVariables( $text );
2840 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2841 * Callback will be called with the text within
2842 * Transform and return the text within
2845 function setHook( $tag, $callback ) {
2846 $oldVal = @$this->mTagHooks
[$tag];
2847 $this->mTagHooks
[$tag] = $callback;
2852 * Replace <!--LINK--> link placeholders with actual links, in the buffer
2853 * Placeholders created in Skin::makeLinkObj()
2854 * Returns an array of links found, indexed by PDBK:
2858 * $options is a bit field, RLH_FOR_UPDATE to select for update
2860 function replaceLinkHolders( &$text, $options = 0 ) {
2861 global $wgUser, $wgLinkCache, $wgUseOldExistenceCheck, $wgLinkHolders;
2863 if ( $wgUseOldExistenceCheck ) {
2867 $fname = 'Parser::replaceLinkHolders';
2868 wfProfileIn( $fname );
2873 #if ( !empty( $tmpLinks[0] ) ) { #TODO
2874 if ( !empty( $wgLinkHolders['namespaces'] ) ) {
2875 wfProfileIn( $fname.'-check' );
2876 $dbr =& wfGetDB( DB_SLAVE
);
2877 $cur = $dbr->tableName( 'cur' );
2878 $sk = $wgUser->getSkin();
2879 $threshold = $wgUser->getOption('stubthreshold');
2882 asort( $wgLinkHolders['namespaces'] );
2886 foreach ( $wgLinkHolders['namespaces'] as $key => $val ) {
2888 $title = $wgLinkHolders['titles'][$key];
2890 # Skip invalid entries.
2891 # Result will be ugly, but prevents crash.
2892 if ( is_null( $title ) ) {
2895 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
2897 # Check if it's in the link cache already
2898 if ( $wgLinkCache->getGoodLinkID( $pdbk ) ) {
2899 $colours[$pdbk] = 1;
2900 } elseif ( $wgLinkCache->isBadLink( $pdbk ) ) {
2901 $colours[$pdbk] = 0;
2903 # Not in the link cache, add it to the query
2904 if ( !isset( $current ) ) {
2906 $query = "SELECT cur_id, cur_namespace, cur_title";
2907 if ( $threshold > 0 ) {
2908 $query .= ", LENGTH(cur_text) AS cur_len, cur_is_redirect";
2910 $query .= " FROM $cur WHERE (cur_namespace=$val AND cur_title IN(";
2911 } elseif ( $current != $val ) {
2913 $query .= ")) OR (cur_namespace=$val AND cur_title IN(";
2918 $query .= $dbr->addQuotes( $wgLinkHolders['dbkeys'][$key] );
2923 if ( $options & RLH_FOR_UPDATE
) {
2924 $query .= ' FOR UPDATE';
2927 $res = $dbr->query( $query, $fname );
2929 # Fetch data and form into an associative array
2930 # non-existent = broken
2933 while ( $s = $dbr->fetchObject($res) ) {
2934 $title = Title
::makeTitle( $s->cur_namespace
, $s->cur_title
);
2935 $pdbk = $title->getPrefixedDBkey();
2936 $wgLinkCache->addGoodLink( $s->cur_id
, $pdbk );
2938 if ( $threshold > 0 ) {
2939 $size = $s->cur_len
;
2940 if ( $s->cur_is_redirect ||
$s->cur_namespace
!= 0 ||
$length < $threshold ) {
2941 $colours[$pdbk] = 1;
2943 $colours[$pdbk] = 2;
2946 $colours[$pdbk] = 1;
2950 wfProfileOut( $fname.'-check' );
2952 # Construct search and replace arrays
2953 wfProfileIn( $fname.'-construct' );
2954 global $outputReplace;
2955 $outputReplace = array();
2956 foreach ( $wgLinkHolders['namespaces'] as $key => $ns ) {
2957 $pdbk = $pdbks[$key];
2958 $searchkey = '<!--LINK '.$key.'-->';
2959 $title = $wgLinkHolders['titles'][$key];
2960 if ( empty( $colours[$pdbk] ) ) {
2961 $wgLinkCache->addBadLink( $pdbk );
2962 $colours[$pdbk] = 0;
2963 $outputReplace[$searchkey] = $sk->makeBrokenLinkObj( $title,
2964 $wgLinkHolders['texts'][$key],
2965 $wgLinkHolders['queries'][$key] );
2966 } elseif ( $colours[$pdbk] == 1 ) {
2967 $outputReplace[$searchkey] = $sk->makeKnownLinkObj( $title,
2968 $wgLinkHolders['texts'][$key],
2969 $wgLinkHolders['queries'][$key] );
2970 } elseif ( $colours[$pdbk] == 2 ) {
2971 $outputReplace[$searchkey] = $sk->makeStubLinkObj( $title,
2972 $wgLinkHolders['texts'][$key],
2973 $wgLinkHolders['queries'][$key] );
2976 wfProfileOut( $fname.'-construct' );
2979 wfProfileIn( $fname.'-replace' );
2981 $text = preg_replace_callback(
2982 '/(<!--LINK .*?-->)/',
2983 "outputReplaceMatches",
2985 wfProfileOut( $fname.'-replace' );
2987 wfProfileIn( $fname.'-interwiki' );
2988 global $wgInterwikiLinkHolders;
2989 $outputReplace = $wgInterwikiLinkHolders;
2990 $text = preg_replace_callback(
2991 '/<!--IWLINK (.*?)-->/',
2992 "outputReplaceMatches",
2994 wfProfileOut( $fname.'-interwiki' );
2997 wfProfileOut( $fname );
3004 * @package MediaWiki
3008 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
3009 var $mCacheTime; # Used in ParserCache
3011 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
3012 $containsOldMagic = false )
3014 $this->mText
= $text;
3015 $this->mLanguageLinks
= $languageLinks;
3016 $this->mCategoryLinks
= $categoryLinks;
3017 $this->mContainsOldMagic
= $containsOldMagic;
3018 $this->mCacheTime
= '';
3021 function getText() { return $this->mText
; }
3022 function getLanguageLinks() { return $this->mLanguageLinks
; }
3023 function getCategoryLinks() { return $this->mCategoryLinks
; }
3024 function getCacheTime() { return $this->mCacheTime
; }
3025 function containsOldMagic() { return $this->mContainsOldMagic
; }
3026 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
3027 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
3028 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
3029 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
3030 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
3032 function merge( $other ) {
3033 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
3034 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
3035 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
3041 * Set options of the Parser
3043 * @package MediaWiki
3047 # All variables are private
3048 var $mUseTeX; # Use texvc to expand <math> tags
3049 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
3050 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
3051 var $mAllowExternalImages; # Allow external images inline
3052 var $mSkin; # Reference to the preferred skin
3053 var $mDateFormat; # Date format index
3054 var $mEditSection; # Create "edit section" links
3055 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
3056 var $mNumberHeadings; # Automatically number headings
3057 var $mShowToc; # Show table of contents
3059 function getUseTeX() { return $this->mUseTeX
; }
3060 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
3061 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
3062 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
3063 function getSkin() { return $this->mSkin
; }
3064 function getDateFormat() { return $this->mDateFormat
; }
3065 function getEditSection() { return $this->mEditSection
; }
3066 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
3067 function getNumberHeadings() { return $this->mNumberHeadings
; }
3068 function getShowToc() { return $this->mShowToc
; }
3070 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
3071 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
3072 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
3073 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
3074 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
3075 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
3076 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
3077 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
3078 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
3080 function setSkin( &$x ) { $this->mSkin
=& $x; }
3082 # Get parser options
3083 /* static */ function newFromUser( &$user ) {
3084 $popts = new ParserOptions
;
3085 $popts->initialiseFromUser( $user );
3090 function initialiseFromUser( &$userInput ) {
3091 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
3093 $fname = 'ParserOptions::initialiseFromUser';
3094 wfProfileIn( $fname );
3095 if ( !$userInput ) {
3097 $user->setLoaded( true );
3099 $user =& $userInput;
3102 $this->mUseTeX
= $wgUseTeX;
3103 $this->mUseDynamicDates
= $wgUseDynamicDates;
3104 $this->mInterwikiMagic
= $wgInterwikiMagic;
3105 $this->mAllowExternalImages
= $wgAllowExternalImages;
3106 wfProfileIn( $fname.'-skin' );
3107 $this->mSkin
=& $user->getSkin();
3108 wfProfileOut( $fname.'-skin' );
3109 $this->mDateFormat
= $user->getOption( 'date' );
3110 $this->mEditSection
= $user->getOption( 'editsection' );
3111 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
3112 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
3113 $this->mShowToc
= $user->getOption( 'showtoc' );
3114 wfProfileOut( $fname );
3121 * Callback function used by Parser::replaceLinkHolders()
3122 * to substitute link placeholders.
3124 function &outputReplaceMatches($matches) {
3125 global $outputReplace;
3126 return $outputReplace[$matches[1]];
3130 # Regex callbacks, used in Parser::replaceVariables
3131 function wfBraceSubstitution( $matches ) {
3132 global $wgCurParser;
3133 return $wgCurParser->braceSubstitution( $matches );
3136 function wfArgSubstitution( $matches ) {
3137 global $wgCurParser;
3138 return $wgCurParser->argSubstitution( $matches );
3141 function wfVariableSubstitution( $matches ) {
3142 global $wgCurParser;
3143 return $wgCurParser->variableSubstitution( $matches );
3147 * Return the total number of articles
3149 function wfNumberOfArticles() {
3150 global $wgNumberOfArticles;
3153 return $wgNumberOfArticles;
3157 * Get various statistics from the database
3160 function wfLoadSiteStats() {
3161 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
3162 $fname = 'wfLoadSiteStats';
3164 if ( -1 != $wgNumberOfArticles ) return;
3165 $dbr =& wfGetDB( DB_SLAVE
);
3166 $s = $dbr->getArray( 'site_stats',
3167 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
3168 array( 'ss_row_id' => 1 ), $fname
3171 if ( $s === false ) {
3174 $wgTotalViews = $s->ss_total_views
;
3175 $wgTotalEdits = $s->ss_total_edits
;
3176 $wgNumberOfArticles = $s->ss_good_articles
;
3180 function wfEscapeHTMLTagsOnly( $in ) {
3182 array( '"', '>', '<' ),
3183 array( '"', '>', '<' ),