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).
19 define( 'MAX_INCLUDE_REPEAT', 100 );
20 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
22 # Allowed values for $mOutputType
23 define( 'OT_HTML', 1 );
24 define( 'OT_WIKI', 2 );
25 define( 'OT_MSG' , 3 );
27 # string parameter for extractTags which will cause it
28 # to strip HTML comments in addition to regular
29 # <XML>-style tags. This should not be anything we
30 # may want to use in wikisyntax
31 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
33 # prefix for escaping, used in two functions at least
34 define( 'UNIQ_PREFIX', 'NaodW29');
36 # Constants needed for external link processing
37 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
38 define( 'HTTP_PROTOCOLS', 'http|https' );
39 # Everything except bracket, space, or control characters
40 define( 'EXT_LINK_URL_CLASS', '[^]\\x00-\\x20\\x7F]' );
41 define( 'INVERSE_EXT_LINK_URL_CLASS', '[\]\\x00-\\x20\\x7F]' );
43 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
44 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
45 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
46 define( 'EXT_LINK_BRACKETED', '/\[(('.URL_PROTOCOLS
.'):'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
47 define( 'EXT_IMAGE_REGEX',
48 '/^('.HTTP_PROTOCOLS
.':)'. # Protocol
49 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
50 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
56 * Processes wiki markup
59 * There are three main entry points into the Parser class:
61 * produces HTML output
63 * produces altered wiki markup.
65 * performs brace substitution on MediaWiki messages
68 * objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
70 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
73 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
74 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
77 * * only within ParserOptions
90 # Cleared with clearState():
91 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
92 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
95 var $mOptions, $mTitle, $mOutputType,
96 $mTemplates, // cache of already loaded templates, avoids
97 // multiple SQL queries for the same string
98 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
99 // in this path. Used for loop detection.
109 $this->mTemplates
= array();
110 $this->mTemplatePath
= array();
111 $this->mTagHooks
= array();
120 function clearState() {
121 $this->mOutput
= new ParserOutput
;
122 $this->mAutonumber
= 0;
123 $this->mLastSection
= "";
124 $this->mDTopen
= false;
125 $this->mVariables
= false;
126 $this->mIncludeCount
= array();
127 $this->mStripState
= array();
128 $this->mArgStack
= array();
129 $this->mInPre
= false;
133 * First pass--just handle <nowiki> sections, pass the rest off
134 * to internalParse() which does all the real work.
137 * @return ParserOutput a ParserOutput
139 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
141 $fname = 'Parser::parse';
142 wfProfileIn( $fname );
148 $this->mOptions
= $options;
149 $this->mTitle
=& $title;
150 $this->mOutputType
= OT_HTML
;
153 $text = $this->strip( $text, $this->mStripState
);
154 $text = $this->internalParse( $text, $linestart );
155 $text = $this->unstrip( $text, $this->mStripState
);
156 # Clean up special characters, only run once, next-to-last before doBlockLevels
159 # french spaces, last one Guillemet-left
160 # only if there is something before the space
161 '/(.) (?=\\?|:|;|!|\\302\\273)/i' => '\\1 \\2',
162 # french spaces, Guillemet-right
163 "/(\\302\\253) /i"=>"\\1 ",
164 '/<hr *>/i' => '<hr />',
165 '/<br *>/i' => '<br />',
166 '/<center *>/i' => '<div class="center">',
167 '/<\\/center *>/i' => '</div>',
168 # Clean up spare ampersands; note that we probably ought to be
169 # more careful about named entities.
170 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
172 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
175 # french spaces, last one Guillemet-left
176 '/ (\\?|:|;|!|\\302\\273)/i' => ' \\1',
177 # french spaces, Guillemet-right
178 '/(\\302\\253) /i' => '\\1 ',
179 '/<center *>/i' => '<div class="center">',
180 '/<\\/center *>/i' => '</div>'
182 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
185 $text = $this->doBlockLevels( $text, $linestart );
186 $text = $this->unstripNoWiki( $text, $this->mStripState
);
187 $this->mOutput
->setText( $text );
188 wfProfileOut( $fname );
189 return $this->mOutput
;
193 * Get a random string
198 function getRandomString() {
199 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
203 * Replaces all occurrences of <$tag>content</$tag> in the text
204 * with a random marker and returns the new text. the output parameter
205 * $content will be an associative array filled with data on the form
206 * $unique_marker => content.
208 * If $content is already set, the additional entries will be appended
209 * If $tag is set to STRIP_COMMENTS, the function will extract
210 * <!-- HTML comments -->
215 function extractTags($tag, $text, &$content, $uniq_prefix = ''){
216 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
223 while ( '' != $text ) {
224 if($tag==STRIP_COMMENTS
) {
225 $p = preg_split( '/<!--/i', $text, 2 );
227 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
230 if ( ( count( $p ) < 2 ) ||
( '' == $p[1] ) ) {
233 if($tag==STRIP_COMMENTS
) {
234 $q = preg_split( '/-->/i', $p[1], 2 );
236 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
238 $marker = $rnd . sprintf('%08X', $n++
);
239 $content[$marker] = $q[0];
240 $stripped .= $marker;
248 * Strips and renders nowiki, pre, math, hiero
249 * If $render is set, performs necessary rendering operations on plugins
250 * Returns the text, and fills an array with data needed in unstrip()
251 * If the $state is already a valid strip state, it adds to the state
253 * @param bool $stripcomments when set, HTML comments <!-- like this -->
254 * will be stripped in addition to other tags. This is important
255 * for section editing, where these comments cause confusion when
256 * counting the sections in the wikisource
260 function strip( $text, &$state, $stripcomments = false ) {
261 $render = ($this->mOutputType
== OT_HTML
);
262 $html_content = array();
263 $nowiki_content = array();
264 $math_content = array();
265 $pre_content = array();
266 $comment_content = array();
267 $ext_content = array();
269 # Replace any instances of the placeholders
270 $uniq_prefix = UNIQ_PREFIX
;
271 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
274 global $wgRawHtml, $wgWhitelistEdit;
275 if( $wgRawHtml && $wgWhitelistEdit ) {
276 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
277 foreach( $html_content as $marker => $content ) {
279 # Raw and unchecked for validity.
280 $html_content[$marker] = $content;
282 $html_content[$marker] = '<html>'.$content.'</html>';
288 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
289 foreach( $nowiki_content as $marker => $content ) {
291 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
293 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
298 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
299 foreach( $math_content as $marker => $content ){
301 if( $this->mOptions
->getUseTeX() ) {
302 $math_content[$marker] = renderMath( $content );
304 $math_content[$marker] = '<math>'.$content.'<math>';
307 $math_content[$marker] = '<math>'.$content.'</math>';
312 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
313 foreach( $pre_content as $marker => $content ){
315 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
317 $pre_content[$marker] = '<pre>'.$content.'</pre>';
323 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
324 foreach( $comment_content as $marker => $content ){
325 $comment_content[$marker] = '<!--'.$content.'-->';
330 foreach ( $this->mTagHooks
as $tag => $callback ) {
331 $ext_contents[$tag] = array();
332 $text = Parser
::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
333 foreach( $ext_content[$tag] as $marker => $content ) {
335 $ext_content[$tag][$marker] = $callback( $content );
337 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
342 # Merge state with the pre-existing state, if there is one
344 $state['html'] = $state['html'] +
$html_content;
345 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
346 $state['math'] = $state['math'] +
$math_content;
347 $state['pre'] = $state['pre'] +
$pre_content;
348 $state['comment'] = $state['comment'] +
$comment_content;
350 foreach( $ext_content as $tag => $array ) {
351 if ( array_key_exists( $tag, $state ) ) {
352 $state[$tag] = $state[$tag] +
$array;
357 'html' => $html_content,
358 'nowiki' => $nowiki_content,
359 'math' => $math_content,
360 'pre' => $pre_content,
361 'comment' => $comment_content,
368 * restores pre, math, and heiro removed by strip()
370 * always call unstripNoWiki() after this one
373 function unstrip( $text, &$state ) {
374 # Must expand in reverse order, otherwise nested tags will be corrupted
375 $contentDict = end( $state );
376 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
377 if( key($state) != 'nowiki' && key($state) != 'html') {
378 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
379 $text = str_replace( key( $contentDict ), $content, $text );
388 * always call this after unstrip() to preserve the order
392 function unstripNoWiki( $text, &$state ) {
393 # Must expand in reverse order, otherwise nested tags will be corrupted
394 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
395 $text = str_replace( key( $state['nowiki'] ), $content, $text );
400 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
401 $text = str_replace( key( $state['html'] ), $content, $text );
409 * Add an item to the strip state
410 * Returns the unique tag which must be inserted into the stripped text
411 * The tag will be replaced with the original text in unstrip()
415 function insertStripItem( $text, &$state ) {
416 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
425 $state['item'][$rnd] = $text;
430 * Return allowed HTML attributes
434 function getHTMLattrs () {
435 $htmlattrs = array( # Allowed attributes--no scripting, etc.
436 'title', 'align', 'lang', 'dir', 'width', 'height',
437 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
438 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
439 /* FONT */ 'type', 'start', 'value', 'compact',
440 /* For various lists, mostly deprecated but safe */
441 'summary', 'width', 'border', 'frame', 'rules',
442 'cellspacing', 'cellpadding', 'valign', 'char',
443 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
444 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
445 'id', 'class', 'name', 'style' /* For CSS */
451 * Remove non approved attributes and javascript in css
455 function fixTagAttributes ( $t ) {
456 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
457 $htmlattrs = $this->getHTMLattrs() ;
459 # Strip non-approved attributes from the tag
461 '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
462 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
465 $t = str_replace ( '<></>' , '' , $t ) ; # This should fix bug 980557
467 # Strip javascript "expression" from stylesheets. Brute force approach:
468 # If anythin offensive is found, all attributes of the HTML tag are dropped
471 '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
472 wfMungeToUtf8( $t ) ) )
481 * interface with html tidy, used if $wgUseTidy = true
486 function tidy ( $text ) {
487 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
488 global $wgInputEncoding, $wgOutputEncoding;
489 $fname = 'Parser::tidy';
490 wfProfileIn( $fname );
494 switch(strtoupper($wgOutputEncoding)) {
496 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -latin1':' -raw';
499 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -utf8':' -raw';
505 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
506 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
507 '<head><title>test</title></head><body>'.$text.'</body></html>';
508 $descriptorspec = array(
509 0 => array('pipe', 'r'),
510 1 => array('pipe', 'w'),
511 2 => array('file', '/dev/null', 'a')
513 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
514 if (is_resource($process)) {
515 fwrite($pipes[0], $wrappedtext);
517 while (!feof($pipes[1])) {
518 $cleansource .= fgets($pipes[1], 1024);
521 $return_value = proc_close($process);
524 wfProfileOut( $fname );
526 if( $cleansource == '' && $text != '') {
527 wfDebug( "Tidy error detected!\n" );
528 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
535 * parse the wiki syntax used to render tables
539 function doTableStuff ( $t ) {
540 $fname = 'Parser::doTableStuff';
541 wfProfileIn( $fname );
543 $t = explode ( "\n" , $t ) ;
544 $td = array () ; # Is currently a td tag open?
545 $ltd = array () ; # Was it TD or TH?
546 $tr = array () ; # Is currently a tr tag open?
547 $ltr = array () ; # tr attributes
548 $indent_level = 0; # indent level of the table
549 foreach ( $t AS $k => $x )
552 $fc = substr ( $x , 0 , 1 ) ;
553 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
554 $indent_level = strlen( $matches[1] );
556 str_repeat( '<dl><dd>', $indent_level ) .
557 '<table ' . $this->fixTagAttributes ( $matches[2] ) . '>' ;
558 array_push ( $td , false ) ;
559 array_push ( $ltd , '' ) ;
560 array_push ( $tr , false ) ;
561 array_push ( $ltr , '' ) ;
563 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
564 else if ( '|}' == substr ( $x , 0 , 2 ) ) {
566 $l = array_pop ( $ltd ) ;
567 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
568 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
570 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
572 else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
573 $x = substr ( $x , 1 ) ;
574 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
576 $l = array_pop ( $ltd ) ;
577 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
578 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
581 array_push ( $tr , false ) ;
582 array_push ( $td , false ) ;
583 array_push ( $ltd , '' ) ;
584 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
586 else if ( '|' == $fc ||
'!' == $fc ||
'|+' == substr ( $x , 0 , 2 ) ) { # Caption
587 if ( '|+' == substr ( $x , 0 , 2 ) ) {
589 $x = substr ( $x , 1 ) ;
591 $after = substr ( $x , 1 ) ;
592 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
593 $after = explode ( '||' , $after ) ;
595 foreach ( $after AS $theline )
600 $tra = array_pop ( $ltr ) ;
601 if ( !array_pop ( $tr ) ) $z = '<tr '.$tra.">\n" ;
602 array_push ( $tr , true ) ;
603 array_push ( $ltr , '' ) ;
606 $l = array_pop ( $ltd ) ;
607 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
608 if ( $fc == '|' ) $l = 'td' ;
609 else if ( $fc == '!' ) $l = 'th' ;
610 else if ( $fc == '+' ) $l = 'caption' ;
612 array_push ( $ltd , $l ) ;
613 $y = explode ( '|' , $theline , 2 ) ;
614 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
615 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
617 array_push ( $td , true ) ;
622 # Closing open td, tr && table
623 while ( count ( $td ) > 0 )
625 if ( array_pop ( $td ) ) $t[] = '</td>' ;
626 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
630 $t = implode ( "\n" , $t ) ;
631 # $t = $this->removeHTMLtags( $t );
632 wfProfileOut( $fname );
637 * Helper function for parse() that transforms wiki markup into
638 * HTML. Only called for $mOutputType == OT_HTML.
642 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
645 $fname = 'Parser::internalParse';
646 wfProfileIn( $fname );
648 $text = $this->removeHTMLtags( $text );
649 $text = $this->replaceVariables( $text, $args );
651 $text = $wgContLang->convert($text);
653 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
655 $text = $this->doHeadings( $text );
656 if($this->mOptions
->getUseDynamicDates()) {
657 global $wgDateFormatter;
658 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
660 $text = $this->doAllQuotes( $text );
661 $text = $this->replaceInternalLinks ( $text );
662 # Another call to replace links and images inside captions of images
663 $text = $this->replaceInternalLinks ( $text );
664 $text = $this->replaceExternalLinks( $text );
665 $text = $this->doMagicLinks( $text );
666 $text = $this->doTableStuff( $text );
667 $text = $this->formatHeadings( $text, $isMain );
668 $sk =& $this->mOptions
->getSkin();
669 $text = $sk->transformContent( $text );
671 wfProfileOut( $fname );
676 * Replace special strings like "ISBN xxx" and "RFC xxx" with
677 * magic external links.
681 function &doMagicLinks( &$text ) {
682 global $wgUseGeoMode;
683 $text = $this->magicISBN( $text );
684 if ( isset( $wgUseGeoMode ) && $wgUseGeoMode ) {
685 $text = $this->magicGEO( $text );
687 $text = $this->magicRFC( $text );
692 * Parse ^^ tokens and return html
696 function doExponent ( $text ) {
697 $fname = 'Parser::doExponent';
698 wfProfileIn( $fname);
699 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
700 wfProfileOut( $fname);
705 * Parse headers and return html
709 function doHeadings( $text ) {
710 $fname = 'Parser::doHeadings';
711 wfProfileIn( $fname );
712 for ( $i = 6; $i >= 1; --$i ) {
713 $h = substr( '======', 0, $i );
714 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
715 "<h{$i}>\\1</h{$i}>\\2", $text );
717 wfProfileOut( $fname );
722 * Replace single quotes with HTML markup
724 * @return string the altered text
726 function doAllQuotes( $text ) {
727 $fname = 'Parser::doAllQuotes';
728 wfProfileIn( $fname );
730 $lines = explode( "\n", $text );
731 foreach ( $lines as $line ) {
732 $outtext .= $this->doQuotes ( $line ) . "\n";
734 $outtext = substr($outtext, 0,-1);
735 wfProfileOut( $fname );
740 * Helper function for doAllQuotes()
743 function doQuotes( $text ) {
744 $arr = preg_split ("/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
745 if (count ($arr) == 1)
749 # First, do some preliminary work. This may shift some apostrophes from
750 # being mark-up to being text. It also counts the number of occurrences
751 # of bold and italics mark-ups.
759 # If there are ever four apostrophes, assume the first is supposed to
760 # be text, and the remaining three constitute mark-up for bold text.
761 if (strlen ($arr[$i]) == 4)
766 # If there are more than 5 apostrophes in a row, assume they're all
767 # text except for the last 5.
768 else if (strlen ($arr[$i]) > 5)
770 $arr[$i-1] .= str_repeat ("'", strlen ($arr[$i]) - 5);
773 # Count the number of occurrences of bold and italics mark-ups.
774 # We are not counting sequences of five apostrophes.
775 if (strlen ($arr[$i]) == 2) $numitalics++
; else
776 if (strlen ($arr[$i]) == 3) $numbold++
; else
777 if (strlen ($arr[$i]) == 5) { $numitalics++
; $numbold++
; }
782 # If there is an odd number of both bold and italics, it is likely
783 # that one of the bold ones was meant to be an apostrophe followed
784 # by italics. Which one we cannot know for certain, but it is more
785 # likely to be one that has a single-letter word before it.
786 if (($numbold %
2 == 1) && ($numitalics %
2 == 1))
789 $firstsingleletterword = -1;
790 $firstmultiletterword = -1;
794 if (($i %
2 == 1) and (strlen ($r) == 3))
796 $x1 = substr ($arr[$i-1], -1);
797 $x2 = substr ($arr[$i-1], -2, 1);
799 if ($firstspace == -1) $firstspace = $i;
800 } else if ($x2 == ' ') {
801 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
803 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
809 # If there is a single-letter word, use it!
810 if ($firstsingleletterword > -1)
812 $arr [ $firstsingleletterword ] = "''";
813 $arr [ $firstsingleletterword-1 ] .= "'";
815 # If not, but there's a multi-letter word, use that one.
816 else if ($firstmultiletterword > -1)
818 $arr [ $firstmultiletterword ] = "''";
819 $arr [ $firstmultiletterword-1 ] .= "'";
821 # ... otherwise use the first one that has neither.
822 # (notice that it is possible for all three to be -1 if, for example,
823 # there is only one pentuple-apostrophe in the line)
824 else if ($firstspace > -1)
826 $arr [ $firstspace ] = "''";
827 $arr [ $firstspace-1 ] .= "'";
831 # Now let's actually convert our apostrophic mush to HTML!
840 if ($state == 'both')
847 if (strlen ($r) == 2)
850 { $output .= '</i>'; $state = ''; }
851 else if ($state == 'bi')
852 { $output .= '</i>'; $state = 'b'; }
853 else if ($state == 'ib')
854 { $output .= '</b></i><b>'; $state = 'b'; }
855 else if ($state == 'both')
856 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
857 else # $state can be 'b' or ''
858 { $output .= '<i>'; $state .= 'i'; }
860 else if (strlen ($r) == 3)
863 { $output .= '</b>'; $state = ''; }
864 else if ($state == 'bi')
865 { $output .= '</i></b><i>'; $state = 'i'; }
866 else if ($state == 'ib')
867 { $output .= '</b>'; $state = 'i'; }
868 else if ($state == 'both')
869 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
870 else # $state can be 'i' or ''
871 { $output .= '<b>'; $state .= 'b'; }
873 else if (strlen ($r) == 5)
876 { $output .= '</b><i>'; $state = 'i'; }
877 else if ($state == 'i')
878 { $output .= '</i><b>'; $state = 'b'; }
879 else if ($state == 'bi')
880 { $output .= '</i></b>'; $state = ''; }
881 else if ($state == 'ib')
882 { $output .= '</b></i>'; $state = ''; }
883 else if ($state == 'both')
884 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
885 else # ($state == '')
886 { $buffer = ''; $state = 'both'; }
891 # Now close all remaining tags. Notice that the order is important.
892 if ($state == 'b' ||
$state == 'ib')
894 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
898 if ($state == 'both')
899 $output .= '<b><i>'.$buffer.'</i></b>';
905 * Replace external links
907 * Note: we have to do external links before the internal ones,
908 * and otherwise take great care in the order of things here, so
909 * that we don't end up interpreting some URLs twice.
913 function replaceExternalLinks( $text ) {
914 $fname = 'Parser::replaceExternalLinks';
915 wfProfileIn( $fname );
917 $sk =& $this->mOptions
->getSkin();
918 $linktrail = wfMsgForContent('linktrail');
919 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
921 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
924 while ( $i<count( $bits ) ) {
926 $protocol = $bits[$i++
];
928 $trail = $bits[$i++
];
930 # If the link text is an image URL, replace it with an <img> tag
931 # This happened by accident in the original parser, but some people used it extensively
932 $img = $this->maybeMakeImageLink( $text );
933 if ( $img !== false ) {
939 # No link text, e.g. [http://domain.tld/some.link]
941 # Autonumber if allowed
942 if ( strpos( HTTP_PROTOCOLS
, $protocol ) !== false ) {
943 $text = '[' . ++
$this->mAutonumber
. ']';
945 # Otherwise just use the URL
946 $text = htmlspecialchars( $url );
949 # Have link text, e.g. [http://domain.tld/some.link text]s
951 if ( preg_match( $linktrail, $trail, $m2 ) ) {
957 $encUrl = htmlspecialchars( $url );
958 # Bit in parentheses showing the URL for the printable version
959 if( $url == $text ||
preg_match( "!$protocol://" . preg_quote( $text, '/' ) . "/?$!", $url ) ) {
962 # Expand the URL for printable version
963 if ( ! $sk->suppressUrlExpansion() ) {
964 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $encUrl ) . "</i>)</span>";
970 # Process the trail (i.e. everything after this link up until start of the next link),
971 # replacing any non-bracketed links
972 $trail = $this->replaceFreeExternalLinks( $trail );
974 $la = $sk->getExternalLinkAttributes( $url, $text );
976 # Use the encoded URL
977 # This means that users can paste URLs directly into the text
978 # Funny characters like ö aren't valid in URLs anyway
979 # This was changed in August 2004
980 $s .= "<a href=\"{$url}\"{$la}>{$text}</a>{$dtrail}{$paren}{$trail}";
983 wfProfileOut( $fname );
988 * Replace anything that looks like a URL with a link
991 function replaceFreeExternalLinks( $text ) {
992 $bits = preg_split( '/((?:'.URL_PROTOCOLS
.'):)/', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
993 $s = array_shift( $bits );
996 $sk =& $this->mOptions
->getSkin();
998 while ( $i < count( $bits ) ){
999 $protocol = $bits[$i++
];
1000 $remainder = $bits[$i++
];
1002 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1003 # Found some characters after the protocol that look promising
1004 $url = $protocol . $m[1];
1007 # Move trailing punctuation to $trail
1009 # If there is no left bracket, then consider right brackets fair game too
1010 if ( strpos( $url, '(' ) === false ) {
1014 $numSepChars = strspn( strrev( $url ), $sep );
1015 if ( $numSepChars ) {
1016 $trail = substr( $url, -$numSepChars ) . $trail;
1017 $url = substr( $url, 0, -$numSepChars );
1020 # Replace & from obsolete syntax with &
1021 $url = str_replace( '&', '&', $url );
1023 # Is this an external image?
1024 $text = $this->maybeMakeImageLink( $url );
1025 if ( $text === false ) {
1026 # Not an image, make a link
1027 $text = $sk->makeExternalLink( $url, $url );
1029 $s .= $text . $trail;
1031 $s .= $protocol . $remainder;
1038 * make an image if it's allowed
1041 function maybeMakeImageLink( $url ) {
1042 $sk =& $this->mOptions
->getSkin();
1044 if ( $this->mOptions
->getAllowExternalImages() ) {
1045 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1047 $text = $sk->makeImage( htmlspecialchars( $url ) );
1054 * Process [[ ]] wikilinks
1058 function replaceInternalLinks( $s ) {
1059 global $wgLang, $wgContLang, $wgLinkCache;
1060 static $fname = 'Parser::replaceInternalLinks' ;
1061 wfProfileIn( $fname );
1063 wfProfileIn( $fname.'-setup' );
1065 # the % is needed to support urlencoded titles as well
1066 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1067 $sk =& $this->mOptions
->getSkin();
1069 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1071 $a = explode( '[[', ' ' . $s );
1072 $s = array_shift( $a );
1073 $s = substr( $s, 1 );
1075 # Match a link having the form [[namespace:link|alternate]]trail
1077 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1078 # Match the end of a line for a word that's not followed by whitespace,
1079 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1080 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1082 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1083 # Special and Media are pseudo-namespaces; no pages actually exist in them
1085 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1087 if ( $useLinkPrefixExtension ) {
1088 if ( preg_match( $e2, $s, $m ) ) {
1089 $first_prefix = $m[2];
1092 $first_prefix = false;
1098 wfProfileOut( $fname.'-setup' );
1100 # start procedeeding each line
1101 foreach ( $a as $line ) {
1102 wfProfileIn( $fname.'-prefixhandling' );
1103 if ( $useLinkPrefixExtension ) {
1104 if ( preg_match( $e2, $s, $m ) ) {
1112 $prefix = $first_prefix;
1113 $first_prefix = false;
1116 wfProfileOut( $fname.'-prefixhandling' );
1118 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1120 # fix up urlencoded title texts
1121 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1123 } else { # Invalid form; output directly
1124 $s .= $prefix . '[[' . $line ;
1128 # Don't allow internal links to pages containing
1129 # PROTO: where PROTO is a valid URL protocol; these
1130 # should be external links.
1131 if (preg_match('/((?:'.URL_PROTOCOLS
.'):)/', $m[1])) {
1132 $s .= $prefix . '[[' . $line ;
1136 # Make subpage if necessary
1137 $link = $this->maybeDoSubpageLink( $m[1], $text );
1139 $noforce = (substr($m[1], 0, 1) != ':');
1141 # Strip off leading ':'
1142 $link = substr($link, 1);
1145 $wasblank = ( '' == $text );
1146 if( $wasblank ) $text = $link;
1148 $nt = Title
::newFromText( $link );
1150 $s .= $prefix . '[[' . $line;
1154 $ns = $nt->getNamespace();
1155 $iw = $nt->getInterWiki();
1157 # Link not escaped by : , create the various objects
1161 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1162 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1163 $tmp = $prefix . $trail ;
1164 $s .= (trim($tmp) == '')?
'': $tmp;
1168 if ( $ns == NS_IMAGE
) {
1169 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1170 $wgLinkCache->addImageLinkObj( $nt );
1174 if ( $ns == NS_CATEGORY
) {
1175 $t = $nt->getText() ;
1176 $nnt = Title
::newFromText ( Namespace::getCanonicalName(NS_CATEGORY
).':'.$t ) ;
1178 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1179 $pPLC=$sk->postParseLinkColour();
1180 $sk->postParseLinkColour( false );
1181 $t = $sk->makeLinkObj( $nnt, $t, '', '' , $prefix );
1182 $sk->postParseLinkColour( $pPLC );
1183 $wgLinkCache->resume();
1186 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1187 $sortkey = $this->mTitle
->getText();
1189 $sortkey = $this->mTitle
->getPrefixedText();
1194 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1195 $this->mOutput
->mCategoryLinks
[] = $t ;
1196 $s .= $prefix . $trail ;
1201 if( ( $nt->getPrefixedText() === $this->mTitle
->getPrefixedText() ) &&
1202 ( strpos( $link, '#' ) === FALSE ) ) {
1203 # Self-links are handled specially; generally de-link and change to bold.
1204 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1208 if( $ns == NS_MEDIA
) {
1209 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1210 $wgLinkCache->addImageLinkObj( $nt );
1212 } elseif( $ns == NS_SPECIAL
) {
1213 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1216 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1218 wfProfileOut( $fname );
1223 * Handle link to subpage if necessary
1224 * @param $target string the source of the link
1225 * @param &$text the link text, modified as necessary
1226 * @return string the full name of the link
1229 function maybeDoSubpageLink($target, &$text) {
1232 # :Foobar -- override special treatment of prefix (images, language links)
1233 # /Foobar -- convert to CurrentPage/Foobar
1234 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1235 global $wgNamespacesWithSubpages;
1237 $fname = 'Parser::maybeDoSubpageLink';
1238 wfProfileIn( $fname );
1239 # Look at the first character
1240 if( $target{0} == '/' ) {
1241 # / at end means we don't want the slash to be shown
1242 if(substr($target,-1,1)=='/') {
1243 $target=substr($target,1,-1);
1246 $noslash=substr($target,1);
1249 # Some namespaces don't allow subpages
1250 if(!empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()])) {
1251 # subpages allowed here
1252 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1253 if( '' === $text ) {
1255 } # this might be changed for ugliness reasons
1257 # no subpage allowed, use standard link
1265 wfProfileOut( $fname );
1270 * Used by doBlockLevels()
1273 /* private */ function closeParagraph() {
1275 if ( '' != $this->mLastSection
) {
1276 $result = '</' . $this->mLastSection
. ">\n";
1278 $this->mInPre
= false;
1279 $this->mLastSection
= '';
1282 # getCommon() returns the length of the longest common substring
1283 # of both arguments, starting at the beginning of both.
1285 /* private */ function getCommon( $st1, $st2 ) {
1286 $fl = strlen( $st1 );
1287 $shorter = strlen( $st2 );
1288 if ( $fl < $shorter ) { $shorter = $fl; }
1290 for ( $i = 0; $i < $shorter; ++
$i ) {
1291 if ( $st1{$i} != $st2{$i} ) { break; }
1295 # These next three functions open, continue, and close the list
1296 # element appropriate to the prefix character passed into them.
1298 /* private */ function openList( $char ) {
1299 $result = $this->closeParagraph();
1301 if ( '*' == $char ) { $result .= '<ul><li>'; }
1302 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1303 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1304 else if ( ';' == $char ) {
1305 $result .= '<dl><dt>';
1306 $this->mDTopen
= true;
1308 else { $result = '<!-- ERR 1 -->'; }
1313 /* private */ function nextItem( $char ) {
1314 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1315 else if ( ':' == $char ||
';' == $char ) {
1317 if ( $this->mDTopen
) { $close = '</dt>'; }
1318 if ( ';' == $char ) {
1319 $this->mDTopen
= true;
1320 return $close . '<dt>';
1322 $this->mDTopen
= false;
1323 return $close . '<dd>';
1326 return '<!-- ERR 2 -->';
1329 /* private */ function closeList( $char ) {
1330 if ( '*' == $char ) { $text = '</li></ul>'; }
1331 else if ( '#' == $char ) { $text = '</li></ol>'; }
1332 else if ( ':' == $char ) {
1333 if ( $this->mDTopen
) {
1334 $this->mDTopen
= false;
1335 $text = '</dt></dl>';
1337 $text = '</dd></dl>';
1340 else { return '<!-- ERR 3 -->'; }
1346 * Make lists from lines starting with ':', '*', '#', etc.
1349 * @return string the lists rendered as HTML
1351 function doBlockLevels( $text, $linestart ) {
1352 $fname = 'Parser::doBlockLevels';
1353 wfProfileIn( $fname );
1355 # Parsing through the text line by line. The main thing
1356 # happening here is handling of block-level elements p, pre,
1357 # and making lists from lines starting with * # : etc.
1359 $textLines = explode( "\n", $text );
1361 $lastPrefix = $output = $lastLine = '';
1362 $this->mDTopen
= $inBlockElem = false;
1364 $paragraphStack = false;
1366 if ( !$linestart ) {
1367 $output .= array_shift( $textLines );
1369 foreach ( $textLines as $oLine ) {
1370 $lastPrefixLength = strlen( $lastPrefix );
1371 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1372 $preOpenMatch = preg_match('/<pre/i', $oLine );
1373 if ( !$this->mInPre
) {
1374 # Multiple prefixes may abut each other for nested lists.
1375 $prefixLength = strspn( $oLine, '*#:;' );
1376 $pref = substr( $oLine, 0, $prefixLength );
1379 $pref2 = str_replace( ';', ':', $pref );
1380 $t = substr( $oLine, $prefixLength );
1381 $this->mInPre
= !empty($preOpenMatch);
1383 # Don't interpret any other prefixes in preformatted text
1385 $pref = $pref2 = '';
1390 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1391 # Same as the last item, so no need to deal with nesting or opening stuff
1392 $output .= $this->nextItem( substr( $pref, -1 ) );
1393 $paragraphStack = false;
1395 if ( substr( $pref, -1 ) == ';') {
1396 # The one nasty exception: definition lists work like this:
1397 # ; title : definition text
1398 # So we check for : in the remainder text to split up the
1399 # title and definition, without b0rking links.
1400 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1402 $output .= $term . $this->nextItem( ':' );
1405 } elseif( $prefixLength ||
$lastPrefixLength ) {
1406 # Either open or close a level...
1407 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1408 $paragraphStack = false;
1410 while( $commonPrefixLength < $lastPrefixLength ) {
1411 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1412 --$lastPrefixLength;
1414 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1415 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1417 while ( $prefixLength > $commonPrefixLength ) {
1418 $char = substr( $pref, $commonPrefixLength, 1 );
1419 $output .= $this->openList( $char );
1421 if ( ';' == $char ) {
1422 # FIXME: This is dupe of code above
1423 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1425 $output .= $term . $this->nextItem( ':' );
1428 ++
$commonPrefixLength;
1430 $lastPrefix = $pref2;
1432 if( 0 == $prefixLength ) {
1433 # No prefix (not in list)--go to paragraph mode
1434 $uniq_prefix = UNIQ_PREFIX
;
1435 // XXX: use a stack for nestable elements like span, table and div
1436 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1437 $closematch = preg_match(
1438 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1439 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1440 if ( $openmatch or $closematch ) {
1441 $paragraphStack = false;
1442 $output .= $this->closeParagraph();
1443 if($preOpenMatch and !$preCloseMatch) {
1444 $this->mInPre
= true;
1446 if ( $closematch ) {
1447 $inBlockElem = false;
1449 $inBlockElem = true;
1451 } else if ( !$inBlockElem && !$this->mInPre
) {
1452 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1454 if ($this->mLastSection
!= 'pre') {
1455 $paragraphStack = false;
1456 $output .= $this->closeParagraph().'<pre>';
1457 $this->mLastSection
= 'pre';
1459 $t = substr( $t, 1 );
1462 if ( '' == trim($t) ) {
1463 if ( $paragraphStack ) {
1464 $output .= $paragraphStack.'<br />';
1465 $paragraphStack = false;
1466 $this->mLastSection
= 'p';
1468 if ($this->mLastSection
!= 'p' ) {
1469 $output .= $this->closeParagraph();
1470 $this->mLastSection
= '';
1471 $paragraphStack = '<p>';
1473 $paragraphStack = '</p><p>';
1477 if ( $paragraphStack ) {
1478 $output .= $paragraphStack;
1479 $paragraphStack = false;
1480 $this->mLastSection
= 'p';
1481 } else if ($this->mLastSection
!= 'p') {
1482 $output .= $this->closeParagraph().'<p>';
1483 $this->mLastSection
= 'p';
1489 if ($paragraphStack === false) {
1493 while ( $prefixLength ) {
1494 $output .= $this->closeList( $pref2{$prefixLength-1} );
1497 if ( '' != $this->mLastSection
) {
1498 $output .= '</' . $this->mLastSection
. '>';
1499 $this->mLastSection
= '';
1502 wfProfileOut( $fname );
1507 * Split up a string on ':', ignoring any occurences inside
1508 * <a>..</a> or <span>...</span>
1509 * @param $str string the string to split
1510 * @param &$before string set to everything before the ':'
1511 * @param &$after string set to everything after the ':'
1512 * return string the position of the ':', or false if none found
1514 function findColonNoLinks($str, &$before, &$after) {
1515 # I wonder if we should make this count all tags, not just <a>
1516 # and <span>. That would prevent us from matching a ':' that
1517 # comes in the middle of italics other such formatting....
1519 $fname = 'Parser::findColonNoLinks';
1520 wfProfileIn( $fname );
1523 $colon = strpos($str, ':', $pos);
1525 if ($colon !== false) {
1526 $before = substr($str, 0, $colon);
1527 $after = substr($str, $colon +
1);
1529 # Skip any ':' within <a> or <span> pairs
1530 $a = substr_count($before, '<a');
1531 $s = substr_count($before, '<span');
1532 $ca = substr_count($before, '</a>');
1533 $cs = substr_count($before, '</span>');
1535 if ($a <= $ca and $s <= $cs) {
1536 # Tags are balanced before ':'; ok
1541 } while ($colon !== false);
1542 wfProfileOut( $fname );
1547 * Return value of a magic variable (like PAGENAME)
1551 function getVariableValue( $index ) {
1552 global $wgContLang, $wgSitename, $wgServer;
1555 case MAG_CURRENTMONTH
:
1556 return $wgContLang->formatNum( date( 'm' ) );
1557 case MAG_CURRENTMONTHNAME
:
1558 return $wgContLang->getMonthName( date('n') );
1559 case MAG_CURRENTMONTHNAMEGEN
:
1560 return $wgContLang->getMonthNameGen( date('n') );
1561 case MAG_CURRENTDAY
:
1562 return $wgContLang->formatNum( date('j') );
1564 return $this->mTitle
->getText();
1566 return $this->mTitle
->getPartialURL();
1568 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1569 return $wgContLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1570 case MAG_CURRENTDAYNAME
:
1571 return $wgContLang->getWeekdayName( date('w')+
1 );
1572 case MAG_CURRENTYEAR
:
1573 return $wgContLang->formatNum( date( 'Y' ) );
1574 case MAG_CURRENTTIME
:
1575 return $wgContLang->time( wfTimestampNow(), false );
1576 case MAG_NUMBEROFARTICLES
:
1577 return $wgContLang->formatNum( wfNumberOfArticles() );
1588 * initialise the magic variables (like CURRENTMONTHNAME)
1592 function initialiseVariables() {
1593 $fname = 'Parser::initialiseVariables';
1594 wfProfileIn( $fname );
1595 global $wgVariableIDs;
1596 $this->mVariables
= array();
1597 foreach ( $wgVariableIDs as $id ) {
1598 $mw =& MagicWord
::get( $id );
1599 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1601 wfProfileOut( $fname );
1605 * Replace magic variables, templates, and template arguments
1606 * with the appropriate text. Templates are substituted recursively,
1607 * taking care to avoid infinite loops.
1609 * Note that the substitution depends on value of $mOutputType:
1610 * OT_WIKI: only {{subst:}} templates
1611 * OT_MSG: only magic variables
1612 * OT_HTML: all templates and magic variables
1614 * @param string $tex The text to transform
1615 * @param array $args Key-value pairs representing template parameters to substitute
1618 function replaceVariables( $text, $args = array() ) {
1619 global $wgLang, $wgScript, $wgArticlePath;
1621 # Prevent too big inclusions
1622 if(strlen($text)> MAX_INCLUDE_SIZE
)
1625 $fname = 'Parser::replaceVariables';
1626 wfProfileIn( $fname );
1628 $titleChars = Title
::legalChars();
1630 # This function is called recursively. To keep track of arguments we need a stack:
1631 array_push( $this->mArgStack
, $args );
1633 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1634 $GLOBALS['wgCurParser'] =& $this;
1636 # Variable substitution
1637 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", 'wfVariableSubstitution', $text );
1639 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1640 # Argument substitution
1641 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1643 # Template substitution
1644 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1645 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1647 array_pop( $this->mArgStack
);
1649 wfProfileOut( $fname );
1654 * Replace magic variables
1657 function variableSubstitution( $matches ) {
1658 if ( !$this->mVariables
) {
1659 $this->initialiseVariables();
1662 if ( $this->mOutputType
== OT_WIKI
) {
1663 # Do only magic variables prefixed by SUBST
1664 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1665 if (!$mwSubst->matchStartAndRemove( $matches[1] ))
1667 # Note that if we don't substitute the variable below,
1668 # we don't remove the {{subst:}} magic word, in case
1669 # it is a template rather than a magic variable.
1671 if ( !$skip && array_key_exists( $matches[1], $this->mVariables
) ) {
1672 $text = $this->mVariables
[$matches[1]];
1673 $this->mOutput
->mContainsOldMagic
= true;
1675 $text = $matches[0];
1680 # Split template arguments
1681 function getTemplateArgs( $argsString ) {
1682 if ( $argsString === '' ) {
1686 $args = explode( '|', substr( $argsString, 1 ) );
1688 # If any of the arguments contains a '[[' but no ']]', it needs to be
1689 # merged with the next arg because the '|' character between belongs
1690 # to the link syntax and not the template parameter syntax.
1691 $argc = count($args);
1693 for ( $i = 0; $i < $argc-1; $i++
) {
1694 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1695 $args[$i] .= '|'.$args[$i+
1];
1696 array_splice($args, $i+
1, 1);
1706 * Return the text of a template, after recursively
1707 * replacing any variables or templates within the template.
1709 * @param array $matches The parts of the template
1710 * $matches[1]: the title, i.e. the part before the |
1711 * $matches[2]: the parameters (including a leading |), if any
1712 * @return string the text of the template
1715 function braceSubstitution( $matches ) {
1716 global $wgLinkCache, $wgContLang;
1717 $fname = 'Parser::braceSubstitution';
1724 # Need to know if the template comes at the start of a line,
1725 # to treat the beginning of the template like the beginning
1726 # of a line for tables and block-level elements.
1727 $linestart = $matches[1];
1729 # $part1 is the bit before the first |, and must contain only title characters
1730 # $args is a list of arguments, starting from index 0, not including $part1
1732 $part1 = $matches[2];
1733 # If the third subpattern matched anything, it will start with |
1735 $args = $this->getTemplateArgs($matches[3]);
1736 $argc = count( $args );
1738 # Don't parse {{{}}} because that's only for template arguments
1739 if ( $linestart === '{' ) {
1740 $text = $matches[0];
1747 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1748 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
1749 # One of two possibilities is true:
1750 # 1) Found SUBST but not in the PST phase
1751 # 2) Didn't find SUBST and in the PST phase
1752 # In either case, return without further processing
1753 $text = $matches[0];
1759 # MSG, MSGNW and INT
1762 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1763 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1766 # Remove obsolete MSG:
1767 $mwMsg =& MagicWord
::get( MAG_MSG
);
1768 $mwMsg->matchStartAndRemove( $part1 );
1771 # Check if it is an internal message
1772 $mwInt =& MagicWord
::get( MAG_INT
);
1773 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1774 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1775 $text = $linestart . wfMsgReal( $part1, $args, true );
1783 # Check for NS: (namespace expansion)
1784 $mwNs = MagicWord
::get( MAG_NS
);
1785 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1786 if ( intval( $part1 ) ) {
1787 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1790 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1791 if ( !is_null( $index ) ) {
1792 $text = $linestart . $wgContLang->getNsText( $index );
1799 # LOCALURL and LOCALURLE
1801 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1802 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1804 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1805 $func = 'getLocalURL';
1806 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1807 $func = 'escapeLocalURL';
1812 if ( $func !== '' ) {
1813 $title = Title
::newFromText( $part1 );
1814 if ( !is_null( $title ) ) {
1816 $text = $linestart . $title->$func( $args[0] );
1818 $text = $linestart . $title->$func();
1826 if ( !$found && $argc == 1 ) {
1827 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
1828 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1829 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
1834 # Template table test
1836 # Did we encounter this template already? If yes, it is in the cache
1837 # and we need to check for loops.
1838 if ( !$found && isset( $this->mTemplates
[$part1] ) ) {
1839 # set $text to cached message.
1840 $text = $linestart . $this->mTemplates
[$part1];
1843 # Infinite loop test
1844 if ( isset( $this->mTemplatePath
[$part1] ) ) {
1847 $text .= '<!-- WARNING: template loop detected -->';
1851 # Load from database
1852 $itcamefromthedatabase = false;
1855 $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
1856 if ($subpage !== '') {
1857 $ns = $this->mTitle
->getNamespace();
1859 $title = Title
::newFromText( $part1, $ns );
1860 if ( !is_null( $title ) && !$title->isExternal() ) {
1861 # Check for excessive inclusion
1862 $dbk = $title->getPrefixedDBkey();
1863 if ( $this->incrementIncludeCount( $dbk ) ) {
1864 # This should never be reached.
1865 $article = new Article( $title );
1866 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1867 if ( $articleContent !== false ) {
1869 $text = $linestart . $articleContent;
1870 $itcamefromthedatabase = true;
1874 # If the title is valid but undisplayable, make a link to it
1875 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1876 $text = $linestart . '[['.$title->getPrefixedText().']]';
1880 # Template cache array insertion
1881 $this->mTemplates
[$part1] = $text;
1885 # Recursive parsing, escaping and link table handling
1886 # Only for HTML output
1887 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1888 $text = wfEscapeWikiText( $text );
1889 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
1890 # Clean up argument array
1891 $assocArgs = array();
1893 foreach( $args as $arg ) {
1894 $eqpos = strpos( $arg, '=' );
1895 if ( $eqpos === false ) {
1896 $assocArgs[$index++
] = $arg;
1898 $name = trim( substr( $arg, 0, $eqpos ) );
1899 $value = trim( substr( $arg, $eqpos+
1 ) );
1900 if ( $value === false ) {
1903 if ( $name !== false ) {
1904 $assocArgs[$name] = $value;
1909 # Add a new element to the templace recursion path
1910 $this->mTemplatePath
[$part1] = 1;
1912 $text = $this->strip( $text, $this->mStripState
);
1913 $text = $this->removeHTMLtags( $text );
1914 $text = $this->replaceVariables( $text, $assocArgs );
1916 # Resume the link cache and register the inclusion as a link
1917 if ( $this->mOutputType
== OT_HTML
&& !is_null( $title ) ) {
1918 $wgLinkCache->addLinkObj( $title );
1921 # If the template begins with a table or block-level
1922 # element, it should be treated as beginning a new line.
1923 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
1924 $text = "\n" . $text;
1928 # Empties the template path
1929 $this->mTemplatePath
= array();
1933 # replace ==section headers==
1934 # XXX this needs to go away once we have a better parser.
1935 if ( $this->mOutputType
!= OT_WIKI
&& $itcamefromthedatabase ) {
1936 if( !is_null( $title ) )
1937 $encodedname = base64_encode($title->getPrefixedDBkey());
1939 $encodedname = base64_encode("");
1940 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
1941 PREG_SPLIT_DELIM_CAPTURE
);
1944 for( $i = 0; $i < count($m); $i +
= 2 ) {
1946 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
1948 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
1952 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
1953 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
1954 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
1961 # Empties the template path
1962 $this->mTemplatePath
= array();
1971 * Triple brace replacement -- used for template arguments
1974 function argSubstitution( $matches ) {
1975 $arg = trim( $matches[1] );
1976 $text = $matches[0];
1977 $inputArgs = end( $this->mArgStack
);
1979 if ( array_key_exists( $arg, $inputArgs ) ) {
1980 $text = $inputArgs[$arg];
1987 * Returns true if the function is allowed to include this entity
1990 function incrementIncludeCount( $dbk ) {
1991 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1992 $this->mIncludeCount
[$dbk] = 0;
1994 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
2003 * Cleans up HTML, removes dangerous tags and attributes, and
2004 * removes HTML comments
2007 function removeHTMLtags( $text ) {
2008 global $wgUseTidy, $wgUserHtml;
2009 $fname = 'Parser::removeHTMLtags';
2010 wfProfileIn( $fname );
2013 $htmlpairs = array( # Tags that must be closed
2014 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
2015 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
2016 'strike', 'strong', 'tt', 'var', 'div', 'center',
2017 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
2018 'ruby', 'rt' , 'rb' , 'rp', 'p'
2020 $htmlsingle = array(
2021 'br', 'hr', 'li', 'dt', 'dd'
2023 $htmlnest = array( # Tags that can be nested--??
2024 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
2025 'dl', 'font', 'big', 'small', 'sub', 'sup'
2027 $tabletags = array( # Can only appear inside table
2031 $htmlpairs = array();
2032 $htmlsingle = array();
2033 $htmlnest = array();
2034 $tabletags = array();
2037 $htmlsingle = array_merge( $tabletags, $htmlsingle );
2038 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
2040 $htmlattrs = $this->getHTMLattrs () ;
2042 # Remove HTML comments
2043 $text = $this->removeHTMLcomments( $text );
2045 $bits = explode( '<', $text );
2046 $text = array_shift( $bits );
2048 $tagstack = array(); $tablestack = array();
2049 foreach ( $bits as $x ) {
2050 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
2051 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2053 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2054 error_reporting( $prev );
2057 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2061 if ( ! in_array( $t, $htmlsingle ) &&
2062 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
2063 @array_push
( $tagstack, $ot );
2066 if ( $t == 'table' ) {
2067 $tagstack = array_pop( $tablestack );
2072 # Keep track for later
2073 if ( in_array( $t, $tabletags ) &&
2074 ! in_array( 'table', $tagstack ) ) {
2076 } else if ( in_array( $t, $tagstack ) &&
2077 ! in_array ( $t , $htmlnest ) ) {
2079 } else if ( ! in_array( $t, $htmlsingle ) ) {
2080 if ( $t == 'table' ) {
2081 array_push( $tablestack, $tagstack );
2082 $tagstack = array();
2084 array_push( $tagstack, $t );
2086 # Strip non-approved attributes from the tag
2087 $newparams = $this->fixTagAttributes($params);
2091 $rest = str_replace( '>', '>', $rest );
2092 $text .= "<$slash$t $newparams$brace$rest";
2096 $text .= '<' . str_replace( '>', '>', $x);
2098 # Close off any remaining tags
2099 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2101 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2104 # this might be possible using tidy itself
2105 foreach ( $bits as $x ) {
2106 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2108 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2109 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2110 $newparams = $this->fixTagAttributes($params);
2111 $rest = str_replace( '>', '>', $rest );
2112 $text .= "<$slash$t $newparams$brace$rest";
2114 $text .= '<' . str_replace( '>', '>', $x);
2118 wfProfileOut( $fname );
2123 * Remove '<!--', '-->', and everything between.
2124 * To avoid leaving blank lines, when a comment is both preceded
2125 * and followed by a newline (ignoring spaces), trim leading and
2126 * trailing spaces and one of the newlines.
2130 function removeHTMLcomments( $text ) {
2131 $fname='Parser::removeHTMLcomments';
2132 wfProfileIn( $fname );
2133 while (($start = strpos($text, '<!--')) !== false) {
2134 $end = strpos($text, '-->', $start +
4);
2135 if ($end === false) {
2136 # Unterminated comment; bail out
2142 # Trim space and newline if the comment is both
2143 # preceded and followed by a newline
2144 $spaceStart = max($start - 1, 0);
2145 $spaceLen = $end - $spaceStart;
2146 while (substr($text, $spaceStart, 1) === ' ' && $spaceStart > 0) {
2150 while (substr($text, $spaceStart +
$spaceLen, 1) === ' ')
2152 if (substr($text, $spaceStart, 1) === "\n" and substr($text, $spaceStart +
$spaceLen, 1) === "\n") {
2153 # Remove the comment, leading and trailing
2154 # spaces, and leave only one newline.
2155 $text = substr_replace($text, "\n", $spaceStart, $spaceLen +
1);
2158 # Remove just the comment.
2159 $text = substr_replace($text, '', $start, $end - $start);
2162 wfProfileOut( $fname );
2167 * This function accomplishes several tasks:
2168 * 1) Auto-number headings if that option is enabled
2169 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2170 * 3) Add a Table of contents on the top for users who have enabled the option
2171 * 4) Auto-anchor headings
2173 * It loops through all headlines, collects the necessary data, then splits up the
2174 * string and re-inserts the newly formatted headlines.
2177 /* private */ function formatHeadings( $text, $isMain=true ) {
2178 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders;
2180 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2181 $doShowToc = $this->mOptions
->getShowToc();
2182 $forceTocHere = false;
2183 if( !$this->mTitle
->userCanEdit() ) {
2185 $rightClickHack = 0;
2187 $showEditLink = $this->mOptions
->getEditSection();
2188 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
2191 # Inhibit editsection links if requested in the page
2192 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2193 if( $esw->matchAndRemove( $text ) ) {
2196 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2198 $mw =& MagicWord
::get( MAG_NOTOC
);
2199 if( $mw->matchAndRemove( $text ) ) {
2203 # never add the TOC to the Main Page. This is an entry page that should not
2204 # be more than 1-2 screens large anyway
2205 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
2209 # Get all headlines for numbering them and adding funky stuff like [edit]
2210 # links - this is for later, but we need the number of headlines right now
2211 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
2213 # if there are fewer than 4 headlines in the article, do not show TOC
2214 if( $numMatches < 4 ) {
2218 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2219 # override above conditions and always show TOC at that place
2220 $mw =& MagicWord
::get( MAG_TOC
);
2221 if ($mw->match( $text ) ) {
2223 $forceTocHere = true;
2225 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2226 # override above conditions and always show TOC above first header
2227 $mw =& MagicWord
::get( MAG_FORCETOC
);
2228 if ($mw->matchAndRemove( $text ) ) {
2235 # We need this to perform operations on the HTML
2236 $sk =& $this->mOptions
->getSkin();
2240 $sectionCount = 0; # headlineCount excluding template sections
2242 # Ugh .. the TOC should have neat indentation levels which can be
2243 # passed to the skin functions. These are determined here
2248 $sublevelCount = array();
2251 foreach( $matches[3] as $headline ) {
2253 $templatetitle = "";
2254 $templatesection = 0;
2256 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2258 $templatetitle = base64_decode($mat[1]);
2259 $templatesection = 1 +
(int)base64_decode($mat[2]);
2260 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2265 $prevlevel = $level;
2267 $level = $matches[1][$headlineCount];
2268 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2269 # reset when we enter a new level
2270 $sublevelCount[$level] = 0;
2271 $toc .= $sk->tocIndent( $level - $prevlevel );
2272 $toclevel +
= $level - $prevlevel;
2274 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2275 # reset when we step back a level
2276 $sublevelCount[$level+
1]=0;
2277 $toc .= $sk->tocUnindent( $prevlevel - $level );
2278 $toclevel -= $prevlevel - $level;
2280 # count number of headlines for each level
2281 @$sublevelCount[$level]++
;
2282 if( $doNumberHeadings ||
$doShowToc ) {
2284 for( $i = 1; $i <= $level; $i++
) {
2285 if( !empty( $sublevelCount[$i] ) ) {
2289 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2295 # The canonized header is a version of the header text safe to use for links
2296 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2297 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2298 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2300 # Remove link placeholders by the link text.
2301 # <!--LINK number-->
2303 # link text with suffix
2304 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2305 "\$wgLinkHolders['texts'][\$1]",
2306 $canonized_headline );
2309 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2310 $tocline = trim( $canonized_headline );
2311 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2312 $replacearray = array(
2316 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2317 $refer[$headlineCount] = $canonized_headline;
2319 # count how many in assoc. array so we can track dupes in anchors
2320 @$refers[$canonized_headline]++
;
2321 $refcount[$headlineCount]=$refers[$canonized_headline];
2323 # Prepend the number to the heading text
2325 if( $doNumberHeadings ||
$doShowToc ) {
2326 $tocline = $numbering . ' ' . $tocline;
2328 # Don't number the heading if it is the only one (looks silly)
2329 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2330 # the two are different if the line contains a link
2331 $headline=$numbering . ' ' . $headline;
2335 # Create the anchor for linking from the TOC to the section
2336 $anchor = $canonized_headline;
2337 if($refcount[$headlineCount] > 1 ) {
2338 $anchor .= '_' . $refcount[$headlineCount];
2340 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2341 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2343 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2344 if ( empty( $head[$headlineCount] ) ) {
2345 $head[$headlineCount] = '';
2348 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2350 $head[$headlineCount] .= $sk->editSectionLink($sectionCount+
1);
2353 # Add the edit section span
2354 if( $rightClickHack ) {
2356 $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
2358 $headline = $sk->editSectionScript($sectionCount+
1,$headline);
2361 # give headline the correct <h#> tag
2362 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2370 $toclines = $headlineCount;
2371 $toc .= $sk->tocUnindent( $toclevel );
2372 $toc = $sk->tocTable( $toc );
2375 # split up and insert constructed headlines
2377 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2380 foreach( $blocks as $block ) {
2381 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2382 # This is the [edit] link that appears for the top block of text when
2383 # section editing is enabled
2385 # Disabled because it broke block formatting
2386 # For example, a bullet point in the top line
2387 # $full .= $sk->editSectionLink(0);
2390 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2391 # Top anchor now in skin
2395 if( !empty( $head[$i] ) ) {
2401 $mw =& MagicWord
::get( MAG_TOC
);
2402 return $mw->replace( $toc, $full );
2409 * Return an HTML link for the "ISBN 123456" text
2412 function magicISBN( $text ) {
2414 $fname = 'Parser::magicISBN';
2415 wfProfileIn( $fname );
2417 $a = split( 'ISBN ', ' '.$text );
2418 if ( count ( $a ) < 2 ) {
2419 wfProfileOut( $fname );
2422 $text = substr( array_shift( $a ), 1);
2423 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2425 foreach ( $a as $x ) {
2426 $isbn = $blank = '' ;
2427 while ( ' ' == $x{0} ) {
2429 $x = substr( $x, 1 );
2431 if ( $x == '' ) { # blank isbn
2432 $text .= "ISBN $blank";
2435 while ( strstr( $valid, $x{0} ) != false ) {
2437 $x = substr( $x, 1 );
2439 $num = str_replace( '-', '', $isbn );
2440 $num = str_replace( ' ', '', $num );
2443 $text .= "ISBN $blank$x";
2445 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2446 $text .= '<a href="' .
2447 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2448 "\" class=\"internal\">ISBN $isbn</a>";
2452 wfProfileOut( $fname );
2457 * Return an HTML link for the "GEO ..." text
2460 function magicGEO( $text ) {
2461 global $wgLang, $wgUseGeoMode;
2462 $fname = 'Parser::magicGEO';
2463 wfProfileIn( $fname );
2465 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2466 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2467 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2468 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2469 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2470 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2472 $a = split( 'GEO ', ' '.$text );
2473 if ( count ( $a ) < 2 ) {
2474 wfProfileOut( $fname );
2477 $text = substr( array_shift( $a ), 1);
2478 $valid = '0123456789.+-:';
2480 foreach ( $a as $x ) {
2481 $geo = $blank = '' ;
2482 while ( ' ' == $x{0} ) {
2484 $x = substr( $x, 1 );
2486 while ( strstr( $valid, $x{0} ) != false ) {
2488 $x = substr( $x, 1 );
2490 $num = str_replace( '+', '', $geo );
2491 $num = str_replace( ' ', '', $num );
2493 if ( '' == $num ||
count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2494 $text .= "GEO $blank$x";
2496 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2497 $text .= '<a href="' .
2498 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2499 "\" class=\"internal\">GEO $geo</a>";
2503 wfProfileOut( $fname );
2508 * Return an HTML link for the "RFC 1234" text
2510 * @param string $text text to be processed
2512 function magicRFC( $text ) {
2515 $valid = '0123456789';
2518 $a = split( 'RFC ', ' '.$text );
2519 if ( count ( $a ) < 2 ) return $text;
2520 $text = substr( array_shift( $a ), 1);
2522 /* Check if RFC keyword is preceed by [[.
2523 * This test is made here cause of the array_shift above
2524 * that prevent the test to be done in the foreach.
2526 if(substr($text, -2) == '[[') { $internal = true; }
2528 foreach ( $a as $x ) {
2529 /* token might be empty if we have RFC RFC 1234 */
2535 $rfc = $blank = '' ;
2537 /** remove and save whitespaces in $blank */
2538 while ( $x{0} == ' ' ) {
2540 $x = substr( $x, 1 );
2543 /** remove and save the rfc number in $rfc */
2544 while ( strstr( $valid, $x{0} ) != false ) {
2546 $x = substr( $x, 1 );
2550 /* call back stripped spaces*/
2551 $text .= "RFC $blank$x";
2552 } elseif( $internal) {
2554 $text .= "RFC $rfc$x";
2556 /* build the external link*/
2557 $url = wfmsg( 'rfcurl' );
2558 $url = str_replace( '$1', $rfc, $url);
2559 $sk =& $this->mOptions
->getSkin();
2560 $la = $sk->getExternalLinkAttributes( $url, 'RFC '.$rfc );
2561 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2564 /* Check if the next RFC keyword is preceed by [[ */
2565 $internal = (substr($x,-2) == '[[');
2571 * Transform wiki markup when saving a page by doing \r\n -> \n
2572 * conversion, substitting signatures, {{subst:}} templates, etc.
2574 * @param string $text the text to transform
2575 * @param Title &$title the Title object for the current article
2576 * @param User &$user the User object describing the current user
2577 * @param ParserOptions $options parsing options
2578 * @param bool $clearState whether to clear the parser state first
2579 * @return string the altered wiki markup
2582 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2583 $this->mOptions
= $options;
2584 $this->mTitle
=& $title;
2585 $this->mOutputType
= OT_WIKI
;
2587 if ( $clearState ) {
2588 $this->clearState();
2591 $stripState = false;
2595 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2599 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2600 "/<br *?>/i" => "<br />",
2602 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2604 $text = $this->strip( $text, $stripState, false );
2605 $text = $this->pstPass2( $text, $user );
2606 $text = $this->unstrip( $text, $stripState );
2607 $text = $this->unstripNoWiki( $text, $stripState );
2612 * Pre-save transform helper function
2615 function pstPass2( $text, &$user ) {
2616 global $wgLang, $wgContLang, $wgLocaltimezone, $wgCurParser;
2618 # Variable replacement
2619 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2620 $text = $this->replaceVariables( $text );
2624 $n = $user->getName();
2625 $k = $user->getOption( 'nickname' );
2626 if ( '' == $k ) { $k = $n; }
2627 if(isset($wgLocaltimezone)) {
2628 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2630 /* Note: this is an ugly timezone hack for the European wikis */
2631 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
2632 ' (' . date( 'T' ) . ')';
2633 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2635 $text = preg_replace( '/~~~~~/', $d, $text );
2636 $text = preg_replace( '/~~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]] $d", $text );
2637 $text = preg_replace( '/~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]]", $text );
2639 # Context links: [[|name]] and [[name (context)|]]
2641 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2642 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2643 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2644 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2646 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2647 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2648 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2649 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2651 $t = $this->mTitle
->getText();
2652 if ( preg_match( $conpat, $t, $m ) ) {
2655 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2656 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2657 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2659 if ( '' == $context ) {
2660 $text = preg_replace( $p2, '[[\\1]]', $text );
2662 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2665 # Trim trailing whitespace
2666 # MAG_END (__END__) tag allows for trailing
2667 # whitespace to be deliberately included
2668 $text = rtrim( $text );
2669 $mw =& MagicWord
::get( MAG_END
);
2670 $mw->matchAndRemove( $text );
2676 * Set up some variables which are usually set up in parse()
2677 * so that an external function can call some class members with confidence
2680 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2681 $this->mTitle
=& $title;
2682 $this->mOptions
= $options;
2683 $this->mOutputType
= $outputType;
2684 if ( $clearState ) {
2685 $this->clearState();
2690 * Transform a MediaWiki message by replacing magic variables.
2692 * @param string $text the text to transform
2693 * @param ParserOptions $options options
2694 * @return string the text with variables substituted
2697 function transformMsg( $text, $options ) {
2699 static $executing = false;
2701 # Guard against infinite recursion
2707 $this->mTitle
= $wgTitle;
2708 $this->mOptions
= $options;
2709 $this->mOutputType
= OT_MSG
;
2710 $this->clearState();
2711 $text = $this->replaceVariables( $text );
2718 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2719 * Callback will be called with the text within
2720 * Transform and return the text within
2723 function setHook( $tag, $callback ) {
2724 $oldVal = @$this->mTagHooks
[$tag];
2725 $this->mTagHooks
[$tag] = $callback;
2732 * @package MediaWiki
2736 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2737 var $mCacheTime; # Used in ParserCache
2739 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
2740 $containsOldMagic = false )
2742 $this->mText
= $text;
2743 $this->mLanguageLinks
= $languageLinks;
2744 $this->mCategoryLinks
= $categoryLinks;
2745 $this->mContainsOldMagic
= $containsOldMagic;
2746 $this->mCacheTime
= '';
2749 function getText() { return $this->mText
; }
2750 function getLanguageLinks() { return $this->mLanguageLinks
; }
2751 function getCategoryLinks() { return $this->mCategoryLinks
; }
2752 function getCacheTime() { return $this->mCacheTime
; }
2753 function containsOldMagic() { return $this->mContainsOldMagic
; }
2754 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2755 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2756 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2757 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2758 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
2760 function merge( $other ) {
2761 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2762 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2763 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2769 * Set options of the Parser
2771 * @package MediaWiki
2775 # All variables are private
2776 var $mUseTeX; # Use texvc to expand <math> tags
2777 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2778 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2779 var $mAllowExternalImages; # Allow external images inline
2780 var $mSkin; # Reference to the preferred skin
2781 var $mDateFormat; # Date format index
2782 var $mEditSection; # Create "edit section" links
2783 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2784 var $mNumberHeadings; # Automatically number headings
2785 var $mShowToc; # Show table of contents
2787 function getUseTeX() { return $this->mUseTeX
; }
2788 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2789 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2790 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2791 function getSkin() { return $this->mSkin
; }
2792 function getDateFormat() { return $this->mDateFormat
; }
2793 function getEditSection() { return $this->mEditSection
; }
2794 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
2795 function getNumberHeadings() { return $this->mNumberHeadings
; }
2796 function getShowToc() { return $this->mShowToc
; }
2798 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2799 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2800 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2801 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2802 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2803 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
2804 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
2805 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
2806 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
2808 function setSkin( &$x ) { $this->mSkin
=& $x; }
2810 # Get parser options
2811 /* static */ function newFromUser( &$user ) {
2812 $popts = new ParserOptions
;
2813 $popts->initialiseFromUser( $user );
2818 function initialiseFromUser( &$userInput ) {
2819 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2821 $fname = 'ParserOptions::initialiseFromUser';
2822 wfProfileIn( $fname );
2823 if ( !$userInput ) {
2825 $user->setLoaded( true );
2827 $user =& $userInput;
2830 $this->mUseTeX
= $wgUseTeX;
2831 $this->mUseDynamicDates
= $wgUseDynamicDates;
2832 $this->mInterwikiMagic
= $wgInterwikiMagic;
2833 $this->mAllowExternalImages
= $wgAllowExternalImages;
2834 wfProfileIn( $fname.'-skin' );
2835 $this->mSkin
=& $user->getSkin();
2836 wfProfileOut( $fname.'-skin' );
2837 $this->mDateFormat
= $user->getOption( 'date' );
2838 $this->mEditSection
= $user->getOption( 'editsection' );
2839 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
2840 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
2841 $this->mShowToc
= $user->getOption( 'showtoc' );
2842 wfProfileOut( $fname );
2848 # Regex callbacks, used in Parser::replaceVariables
2849 function wfBraceSubstitution( $matches ) {
2850 global $wgCurParser;
2851 return $wgCurParser->braceSubstitution( $matches );
2854 function wfArgSubstitution( $matches ) {
2855 global $wgCurParser;
2856 return $wgCurParser->argSubstitution( $matches );
2859 function wfVariableSubstitution( $matches ) {
2860 global $wgCurParser;
2861 return $wgCurParser->variableSubstitution( $matches );
2865 * Return the total number of articles
2867 function wfNumberOfArticles() {
2868 global $wgNumberOfArticles;
2871 return $wgNumberOfArticles;
2875 * Get various statistics from the database
2878 function wfLoadSiteStats() {
2879 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
2880 $fname = 'wfLoadSiteStats';
2882 if ( -1 != $wgNumberOfArticles ) return;
2883 $dbr =& wfGetDB( DB_SLAVE
);
2884 $s = $dbr->getArray( 'site_stats',
2885 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
2886 array( 'ss_row_id' => 1 ), $fname
2889 if ( $s === false ) {
2892 $wgTotalViews = $s->ss_total_views
;
2893 $wgTotalEdits = $s->ss_total_edits
;
2894 $wgNumberOfArticles = $s->ss_good_articles
;
2898 function wfEscapeHTMLTagsOnly( $in ) {
2900 array( '"', '>', '<' ),
2901 array( '"', '>', '<' ),