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]' );
42 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
43 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
44 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
45 define( 'EXT_LINK_BRACKETED', '/\[(('.URL_PROTOCOLS
.'):'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
46 define( 'EXT_IMAGE_REGEX',
47 '/^('.HTTP_PROTOCOLS
.':)'. # Protocol
48 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
49 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
55 * Processes wiki markup
58 * There are three main entry points into the Parser class:
60 * produces HTML output
62 * produces altered wiki markup.
64 * performs brace substitution on MediaWiki messages
67 * objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
69 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
72 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
73 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
76 * * only within ParserOptions
89 # Cleared with clearState():
90 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
91 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
94 var $mOptions, $mTitle, $mOutputType,
95 $mTemplates, // cache of already loaded templates, avoids
96 // multiple SQL queries for the same string
97 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
98 // in this path. Used for loop detection.
108 $this->mTemplates
= array();
109 $this->mTemplatePath
= array();
110 $this->mTagHooks
= array();
119 function clearState() {
120 $this->mOutput
= new ParserOutput
;
121 $this->mAutonumber
= 0;
122 $this->mLastSection
= "";
123 $this->mDTopen
= false;
124 $this->mVariables
= false;
125 $this->mIncludeCount
= array();
126 $this->mStripState
= array();
127 $this->mArgStack
= array();
128 $this->mInPre
= false;
132 * First pass--just handle <nowiki> sections, pass the rest off
133 * to internalParse() which does all the real work.
136 * @return ParserOutput a ParserOutput
138 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
140 $fname = 'Parser::parse';
141 wfProfileIn( $fname );
147 $this->mOptions
= $options;
148 $this->mTitle
=& $title;
149 $this->mOutputType
= OT_HTML
;
152 $text = $this->strip( $text, $this->mStripState
);
153 $text = $this->internalParse( $text, $linestart );
154 $text = $this->unstrip( $text, $this->mStripState
);
155 # Clean up special characters, only run once, next-to-last before doBlockLevels
158 # french spaces, last one Guillemet-left
159 # only if there is something before the space
160 '/(.) (?=\\?|:|;|!|\\302\\273)/i' => '\\1 \\2',
161 # french spaces, Guillemet-right
162 "/(\\302\\253) /i"=>"\\1 ",
163 '/<hr *>/i' => '<hr />',
164 '/<br *>/i' => '<br />',
165 '/<center *>/i' => '<div class="center">',
166 '/<\\/center *>/i' => '</div>',
167 # Clean up spare ampersands; note that we probably ought to be
168 # more careful about named entities.
169 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
171 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
174 # french spaces, last one Guillemet-left
175 '/ (\\?|:|;|!|\\302\\273)/i' => ' \\1',
176 # french spaces, Guillemet-right
177 '/(\\302\\253) /i' => '\\1 ',
178 '/<center *>/i' => '<div class="center">',
179 '/<\\/center *>/i' => '</div>'
181 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
184 $text = $this->doBlockLevels( $text, $linestart );
185 $text = $this->unstripNoWiki( $text, $this->mStripState
);
186 $this->mOutput
->setText( $text );
187 wfProfileOut( $fname );
188 return $this->mOutput
;
192 * Get a random string
197 function getRandomString() {
198 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
202 * Replaces all occurrences of <$tag>content</$tag> in the text
203 * with a random marker and returns the new text. the output parameter
204 * $content will be an associative array filled with data on the form
205 * $unique_marker => content.
207 * If $content is already set, the additional entries will be appended
208 * If $tag is set to STRIP_COMMENTS, the function will extract
209 * <!-- HTML comments -->
214 function extractTags($tag, $text, &$content, $uniq_prefix = ''){
215 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
222 while ( '' != $text ) {
223 if($tag==STRIP_COMMENTS
) {
224 $p = preg_split( '/<!--/i', $text, 2 );
226 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
229 if ( ( count( $p ) < 2 ) ||
( '' == $p[1] ) ) {
232 if($tag==STRIP_COMMENTS
) {
233 $q = preg_split( '/-->/i', $p[1], 2 );
235 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
237 $marker = $rnd . sprintf('%08X', $n++
);
238 $content[$marker] = $q[0];
239 $stripped .= $marker;
247 * Strips and renders nowiki, pre, math, hiero
248 * If $render is set, performs necessary rendering operations on plugins
249 * Returns the text, and fills an array with data needed in unstrip()
250 * If the $state is already a valid strip state, it adds to the state
252 * @param bool $stripcomments when set, HTML comments <!-- like this -->
253 * will be stripped in addition to other tags. This is important
254 * for section editing, where these comments cause confusion when
255 * counting the sections in the wikisource
259 function strip( $text, &$state, $stripcomments = false ) {
260 $render = ($this->mOutputType
== OT_HTML
);
261 $html_content = array();
262 $nowiki_content = array();
263 $math_content = array();
264 $pre_content = array();
265 $comment_content = array();
266 $ext_content = array();
268 # Replace any instances of the placeholders
269 $uniq_prefix = UNIQ_PREFIX
;
270 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
273 global $wgRawHtml, $wgWhitelistEdit;
274 if( $wgRawHtml && $wgWhitelistEdit ) {
275 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
276 foreach( $html_content as $marker => $content ) {
278 # Raw and unchecked for validity.
279 $html_content[$marker] = $content;
281 $html_content[$marker] = '<html>'.$content.'</html>';
287 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
288 foreach( $nowiki_content as $marker => $content ) {
290 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
292 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
297 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
298 foreach( $math_content as $marker => $content ){
300 if( $this->mOptions
->getUseTeX() ) {
301 $math_content[$marker] = renderMath( $content );
303 $math_content[$marker] = '<math>'.$content.'<math>';
306 $math_content[$marker] = '<math>'.$content.'</math>';
311 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
312 foreach( $pre_content as $marker => $content ){
314 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
316 $pre_content[$marker] = '<pre>'.$content.'</pre>';
322 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
323 foreach( $comment_content as $marker => $content ){
324 $comment_content[$marker] = '<!--'.$content.'-->';
329 foreach ( $this->mTagHooks
as $tag => $callback ) {
330 $ext_contents[$tag] = array();
331 $text = Parser
::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
332 foreach( $ext_content[$tag] as $marker => $content ) {
334 $ext_content[$tag][$marker] = $callback( $content );
336 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
341 # Merge state with the pre-existing state, if there is one
343 $state['html'] = $state['html'] +
$html_content;
344 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
345 $state['math'] = $state['math'] +
$math_content;
346 $state['pre'] = $state['pre'] +
$pre_content;
347 $state['comment'] = $state['comment'] +
$comment_content;
349 foreach( $ext_content as $tag => $array ) {
350 if ( array_key_exists( $tag, $state ) ) {
351 $state[$tag] = $state[$tag] +
$array;
356 'html' => $html_content,
357 'nowiki' => $nowiki_content,
358 'math' => $math_content,
359 'pre' => $pre_content,
360 'comment' => $comment_content,
367 * restores pre, math, and heiro removed by strip()
369 * always call unstripNoWiki() after this one
372 function unstrip( $text, &$state ) {
373 # Must expand in reverse order, otherwise nested tags will be corrupted
374 $contentDict = end( $state );
375 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
376 if( key($state) != 'nowiki' && key($state) != 'html') {
377 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
378 $text = str_replace( key( $contentDict ), $content, $text );
387 * always call this after unstrip() to preserve the order
391 function unstripNoWiki( $text, &$state ) {
392 # Must expand in reverse order, otherwise nested tags will be corrupted
393 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
394 $text = str_replace( key( $state['nowiki'] ), $content, $text );
399 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
400 $text = str_replace( key( $state['html'] ), $content, $text );
408 * Add an item to the strip state
409 * Returns the unique tag which must be inserted into the stripped text
410 * The tag will be replaced with the original text in unstrip()
414 function insertStripItem( $text, &$state ) {
415 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
424 $state['item'][$rnd] = $text;
429 * Return allowed HTML attributes
433 function getHTMLattrs () {
434 $htmlattrs = array( # Allowed attributes--no scripting, etc.
435 'title', 'align', 'lang', 'dir', 'width', 'height',
436 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
437 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
438 /* FONT */ 'type', 'start', 'value', 'compact',
439 /* For various lists, mostly deprecated but safe */
440 'summary', 'width', 'border', 'frame', 'rules',
441 'cellspacing', 'cellpadding', 'valign', 'char',
442 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
443 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
444 'id', 'class', 'name', 'style' /* For CSS */
450 * Remove non approved attributes and javascript in css
454 function fixTagAttributes ( $t ) {
455 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
456 $htmlattrs = $this->getHTMLattrs() ;
458 # Strip non-approved attributes from the tag
460 '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
461 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
464 $t = str_replace ( '<></>' , '' , $t ) ; # This should fix bug 980557
466 # Strip javascript "expression" from stylesheets. Brute force approach:
467 # If anythin offensive is found, all attributes of the HTML tag are dropped
470 '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
471 wfMungeToUtf8( $t ) ) )
480 * interface with html tidy, used if $wgUseTidy = true
485 function tidy ( $text ) {
486 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
487 global $wgInputEncoding, $wgOutputEncoding;
488 $fname = 'Parser::tidy';
489 wfProfileIn( $fname );
493 switch(strtoupper($wgOutputEncoding)) {
495 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -latin1':' -raw';
498 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -utf8':' -raw';
504 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
505 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
506 '<head><title>test</title></head><body>'.$text.'</body></html>';
507 $descriptorspec = array(
508 0 => array('pipe', 'r'),
509 1 => array('pipe', 'w'),
510 2 => array('file', '/dev/null', 'a')
512 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
513 if (is_resource($process)) {
514 fwrite($pipes[0], $wrappedtext);
516 while (!feof($pipes[1])) {
517 $cleansource .= fgets($pipes[1], 1024);
520 $return_value = proc_close($process);
523 wfProfileOut( $fname );
525 if( $cleansource == '' && $text != '') {
526 wfDebug( "Tidy error detected!\n" );
527 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
534 * parse the wiki syntax used to render tables
538 function doTableStuff ( $t ) {
539 $fname = 'Parser::doTableStuff';
540 wfProfileIn( $fname );
542 $t = explode ( "\n" , $t ) ;
543 $td = array () ; # Is currently a td tag open?
544 $ltd = array () ; # Was it TD or TH?
545 $tr = array () ; # Is currently a tr tag open?
546 $ltr = array () ; # tr attributes
547 $indent_level = 0; # indent level of the table
548 foreach ( $t AS $k => $x )
551 $fc = substr ( $x , 0 , 1 ) ;
552 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
553 $indent_level = strlen( $matches[1] );
555 str_repeat( '<dl><dd>', $indent_level ) .
556 '<table ' . $this->fixTagAttributes ( $matches[2] ) . '>' ;
557 array_push ( $td , false ) ;
558 array_push ( $ltd , '' ) ;
559 array_push ( $tr , false ) ;
560 array_push ( $ltr , '' ) ;
562 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
563 else if ( '|}' == substr ( $x , 0 , 2 ) ) {
565 $l = array_pop ( $ltd ) ;
566 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
567 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
569 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
571 else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
572 $x = substr ( $x , 1 ) ;
573 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
575 $l = array_pop ( $ltd ) ;
576 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
577 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
580 array_push ( $tr , false ) ;
581 array_push ( $td , false ) ;
582 array_push ( $ltd , '' ) ;
583 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
585 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 ) ;
596 # Loop through each table cell
597 foreach ( $after AS $theline )
602 $tra = array_pop ( $ltr ) ;
603 if ( !array_pop ( $tr ) ) $z = '<tr '.$tra.">\n" ;
604 array_push ( $tr , true ) ;
605 array_push ( $ltr , '' ) ;
608 $l = array_pop ( $ltd ) ;
609 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
610 if ( $fc == '|' ) $l = 'td' ;
611 else if ( $fc == '!' ) $l = 'th' ;
612 else if ( $fc == '+' ) $l = 'caption' ;
614 array_push ( $ltd , $l ) ;
617 $y = explode ( '|' , $theline , 2 ) ;
618 # Note that a '|' inside an invalid link should not
619 # be mistaken as delimiting cell parameters
620 if ( strpos( $y[0], '[[' ) !== false ) {
621 $y = array ($theline);
623 if ( count ( $y ) == 1 )
624 $y = "{$z}<{$l}>{$y[0]}" ;
625 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
627 array_push ( $td , true ) ;
632 # Closing open td, tr && table
633 while ( count ( $td ) > 0 )
635 if ( array_pop ( $td ) ) $t[] = '</td>' ;
636 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
640 $t = implode ( "\n" , $t ) ;
641 # $t = $this->removeHTMLtags( $t );
642 wfProfileOut( $fname );
647 * Helper function for parse() that transforms wiki markup into
648 * HTML. Only called for $mOutputType == OT_HTML.
652 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
655 $fname = 'Parser::internalParse';
656 wfProfileIn( $fname );
658 $text = $this->removeHTMLtags( $text );
659 $text = $this->replaceVariables( $text, $args );
661 $text = $wgContLang->convert($text);
663 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
665 $text = $this->doHeadings( $text );
666 if($this->mOptions
->getUseDynamicDates()) {
667 global $wgDateFormatter;
668 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
670 $text = $this->doAllQuotes( $text );
671 $text = $this->replaceInternalLinks ( $text );
672 $text = $this->replaceExternalLinks( $text );
673 $text = $this->doMagicLinks( $text );
674 $text = $this->doTableStuff( $text );
675 $text = $this->formatHeadings( $text, $isMain );
676 $sk =& $this->mOptions
->getSkin();
677 $text = $sk->transformContent( $text );
679 wfProfileOut( $fname );
684 * Replace special strings like "ISBN xxx" and "RFC xxx" with
685 * magic external links.
689 function &doMagicLinks( &$text ) {
690 global $wgUseGeoMode;
691 $text = $this->magicISBN( $text );
692 if ( isset( $wgUseGeoMode ) && $wgUseGeoMode ) {
693 $text = $this->magicGEO( $text );
695 $text = $this->magicRFC( $text );
700 * Parse ^^ tokens and return html
704 function doExponent ( $text ) {
705 $fname = 'Parser::doExponent';
706 wfProfileIn( $fname);
707 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
708 wfProfileOut( $fname);
713 * Parse headers and return html
717 function doHeadings( $text ) {
718 $fname = 'Parser::doHeadings';
719 wfProfileIn( $fname );
720 for ( $i = 6; $i >= 1; --$i ) {
721 $h = substr( '======', 0, $i );
722 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
723 "<h{$i}>\\1</h{$i}>\\2", $text );
725 wfProfileOut( $fname );
730 * Replace single quotes with HTML markup
732 * @return string the altered text
734 function doAllQuotes( $text ) {
735 $fname = 'Parser::doAllQuotes';
736 wfProfileIn( $fname );
738 $lines = explode( "\n", $text );
739 foreach ( $lines as $line ) {
740 $outtext .= $this->doQuotes ( $line ) . "\n";
742 $outtext = substr($outtext, 0,-1);
743 wfProfileOut( $fname );
748 * Helper function for doAllQuotes()
751 function doQuotes( $text ) {
752 $arr = preg_split ("/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
753 if (count ($arr) == 1)
757 # First, do some preliminary work. This may shift some apostrophes from
758 # being mark-up to being text. It also counts the number of occurrences
759 # of bold and italics mark-ups.
767 # If there are ever four apostrophes, assume the first is supposed to
768 # be text, and the remaining three constitute mark-up for bold text.
769 if (strlen ($arr[$i]) == 4)
774 # If there are more than 5 apostrophes in a row, assume they're all
775 # text except for the last 5.
776 else if (strlen ($arr[$i]) > 5)
778 $arr[$i-1] .= str_repeat ("'", strlen ($arr[$i]) - 5);
781 # Count the number of occurrences of bold and italics mark-ups.
782 # We are not counting sequences of five apostrophes.
783 if (strlen ($arr[$i]) == 2) $numitalics++
; else
784 if (strlen ($arr[$i]) == 3) $numbold++
; else
785 if (strlen ($arr[$i]) == 5) { $numitalics++
; $numbold++
; }
790 # If there is an odd number of both bold and italics, it is likely
791 # that one of the bold ones was meant to be an apostrophe followed
792 # by italics. Which one we cannot know for certain, but it is more
793 # likely to be one that has a single-letter word before it.
794 if (($numbold %
2 == 1) && ($numitalics %
2 == 1))
797 $firstsingleletterword = -1;
798 $firstmultiletterword = -1;
802 if (($i %
2 == 1) and (strlen ($r) == 3))
804 $x1 = substr ($arr[$i-1], -1);
805 $x2 = substr ($arr[$i-1], -2, 1);
807 if ($firstspace == -1) $firstspace = $i;
808 } else if ($x2 == ' ') {
809 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
811 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
817 # If there is a single-letter word, use it!
818 if ($firstsingleletterword > -1)
820 $arr [ $firstsingleletterword ] = "''";
821 $arr [ $firstsingleletterword-1 ] .= "'";
823 # If not, but there's a multi-letter word, use that one.
824 else if ($firstmultiletterword > -1)
826 $arr [ $firstmultiletterword ] = "''";
827 $arr [ $firstmultiletterword-1 ] .= "'";
829 # ... otherwise use the first one that has neither.
830 # (notice that it is possible for all three to be -1 if, for example,
831 # there is only one pentuple-apostrophe in the line)
832 else if ($firstspace > -1)
834 $arr [ $firstspace ] = "''";
835 $arr [ $firstspace-1 ] .= "'";
839 # Now let's actually convert our apostrophic mush to HTML!
848 if ($state == 'both')
855 if (strlen ($r) == 2)
858 { $output .= '</i>'; $state = ''; }
859 else if ($state == 'bi')
860 { $output .= '</i>'; $state = 'b'; }
861 else if ($state == 'ib')
862 { $output .= '</b></i><b>'; $state = 'b'; }
863 else if ($state == 'both')
864 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
865 else # $state can be 'b' or ''
866 { $output .= '<i>'; $state .= 'i'; }
868 else if (strlen ($r) == 3)
871 { $output .= '</b>'; $state = ''; }
872 else if ($state == 'bi')
873 { $output .= '</i></b><i>'; $state = 'i'; }
874 else if ($state == 'ib')
875 { $output .= '</b>'; $state = 'i'; }
876 else if ($state == 'both')
877 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
878 else # $state can be 'i' or ''
879 { $output .= '<b>'; $state .= 'b'; }
881 else if (strlen ($r) == 5)
884 { $output .= '</b><i>'; $state = 'i'; }
885 else if ($state == 'i')
886 { $output .= '</i><b>'; $state = 'b'; }
887 else if ($state == 'bi')
888 { $output .= '</i></b>'; $state = ''; }
889 else if ($state == 'ib')
890 { $output .= '</b></i>'; $state = ''; }
891 else if ($state == 'both')
892 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
893 else # ($state == '')
894 { $buffer = ''; $state = 'both'; }
899 # Now close all remaining tags. Notice that the order is important.
900 if ($state == 'b' ||
$state == 'ib')
902 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
906 if ($state == 'both')
907 $output .= '<b><i>'.$buffer.'</i></b>';
913 * Replace external links
915 * Note: we have to do external links before the internal ones,
916 * and otherwise take great care in the order of things here, so
917 * that we don't end up interpreting some URLs twice.
921 function replaceExternalLinks( $text ) {
922 $fname = 'Parser::replaceExternalLinks';
923 wfProfileIn( $fname );
925 $sk =& $this->mOptions
->getSkin();
926 $linktrail = wfMsgForContent('linktrail');
927 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
929 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
932 while ( $i<count( $bits ) ) {
934 $protocol = $bits[$i++
];
936 $trail = $bits[$i++
];
938 # If the link text is an image URL, replace it with an <img> tag
939 # This happened by accident in the original parser, but some people used it extensively
940 $img = $this->maybeMakeImageLink( $text );
941 if ( $img !== false ) {
947 # No link text, e.g. [http://domain.tld/some.link]
949 # Autonumber if allowed
950 if ( strpos( HTTP_PROTOCOLS
, $protocol ) !== false ) {
951 $text = '[' . ++
$this->mAutonumber
. ']';
953 # Otherwise just use the URL
954 $text = htmlspecialchars( $url );
957 # Have link text, e.g. [http://domain.tld/some.link text]s
959 if ( preg_match( $linktrail, $trail, $m2 ) ) {
965 $encUrl = htmlspecialchars( $url );
966 # Bit in parentheses showing the URL for the printable version
967 if( $url == $text ||
preg_match( "!$protocol://" . preg_quote( $text, '/' ) . "/?$!", $url ) ) {
970 # Expand the URL for printable version
971 if ( ! $sk->suppressUrlExpansion() ) {
972 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $encUrl ) . "</i>)</span>";
978 # Process the trail (i.e. everything after this link up until start of the next link),
979 # replacing any non-bracketed links
980 $trail = $this->replaceFreeExternalLinks( $trail );
982 # Use the encoded URL
983 # This means that users can paste URLs directly into the text
984 # Funny characters like ö aren't valid in URLs anyway
985 # This was changed in August 2004
986 $s .= $sk->makeExternalLink( $url, $text, false ) . $dtrail. $paren . $trail;
989 wfProfileOut( $fname );
994 * Replace anything that looks like a URL with a link
997 function replaceFreeExternalLinks( $text ) {
998 $bits = preg_split( '/((?:'.URL_PROTOCOLS
.'):)/', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
999 $s = array_shift( $bits );
1002 $sk =& $this->mOptions
->getSkin();
1004 while ( $i < count( $bits ) ){
1005 $protocol = $bits[$i++
];
1006 $remainder = $bits[$i++
];
1008 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1009 # Found some characters after the protocol that look promising
1010 $url = $protocol . $m[1];
1013 # Move trailing punctuation to $trail
1015 # If there is no left bracket, then consider right brackets fair game too
1016 if ( strpos( $url, '(' ) === false ) {
1020 $numSepChars = strspn( strrev( $url ), $sep );
1021 if ( $numSepChars ) {
1022 $trail = substr( $url, -$numSepChars ) . $trail;
1023 $url = substr( $url, 0, -$numSepChars );
1026 # Replace & from obsolete syntax with &
1027 $url = str_replace( '&', '&', $url );
1029 # Is this an external image?
1030 $text = $this->maybeMakeImageLink( $url );
1031 if ( $text === false ) {
1032 # Not an image, make a link
1033 $text = $sk->makeExternalLink( $url, $url );
1035 $s .= $text . $trail;
1037 $s .= $protocol . $remainder;
1044 * make an image if it's allowed
1047 function maybeMakeImageLink( $url ) {
1048 $sk =& $this->mOptions
->getSkin();
1050 if ( $this->mOptions
->getAllowExternalImages() ) {
1051 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1053 $text = $sk->makeImage( htmlspecialchars( $url ) );
1060 * Process [[ ]] wikilinks
1064 function replaceInternalLinks( $s ) {
1065 global $wgLang, $wgContLang, $wgLinkCache;
1066 static $fname = 'Parser::replaceInternalLinks' ;
1067 wfProfileIn( $fname );
1069 wfProfileIn( $fname.'-setup' );
1071 # the % is needed to support urlencoded titles as well
1072 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1073 $sk =& $this->mOptions
->getSkin();
1075 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1077 #split the entire text string on occurences of [[
1078 $a = explode( '[[', ' ' . $s );
1079 #get the first element (all text up to first [[), and remove the space we added
1080 $s = array_shift( $a );
1081 $s = substr( $s, 1 );
1083 # Match a link having the form [[namespace:link|alternate]]trail
1085 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1086 # Match cases where there is no "]]", which might still be images
1087 static $e1_img = FALSE;
1088 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1089 # Match the end of a line for a word that's not followed by whitespace,
1090 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1091 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1093 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1095 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1097 if ( $useLinkPrefixExtension ) {
1098 if ( preg_match( $e2, $s, $m ) ) {
1099 $first_prefix = $m[2];
1102 $first_prefix = false;
1108 wfProfileOut( $fname.'-setup' );
1110 # Loop for each link
1111 for ($k = 0; isset( $a[$k] ); $k++
) {
1113 wfProfileIn( $fname.'-prefixhandling' );
1114 if ( $useLinkPrefixExtension ) {
1115 if ( preg_match( $e2, $s, $m ) ) {
1123 $prefix = $first_prefix;
1124 $first_prefix = false;
1127 wfProfileOut( $fname.'-prefixhandling' );
1129 $might_be_img = false;
1131 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1133 # fix up urlencoded title texts
1134 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1136 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1137 $might_be_img = true;
1139 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1141 } else { # Invalid form; output directly
1142 $s .= $prefix . '[[' . $line ;
1146 # Don't allow internal links to pages containing
1147 # PROTO: where PROTO is a valid URL protocol; these
1148 # should be external links.
1149 if (preg_match('/((?:'.URL_PROTOCOLS
.'):)/', $m[1])) {
1150 $s .= $prefix . '[[' . $line ;
1154 # Make subpage if necessary
1155 $link = $this->maybeDoSubpageLink( $m[1], $text );
1157 $noforce = (substr($m[1], 0, 1) != ':');
1159 # Strip off leading ':'
1160 $link = substr($link, 1);
1163 $nt = Title
::newFromText( $link );
1165 $s .= $prefix . '[[' . $line;
1169 //check other language variants of the link
1170 //if the article does not exist
1172 $variants = $wgContLang->getVariants();
1173 if(sizeof($variants) > 1) {
1175 if($nt->getArticleID() == 0) {
1176 foreach ( $variants as $v ) {
1177 if($v == $wgContLang->getPreferredVariant())
1179 $varlink = $wgContLang->autoConvert($link, $v);
1180 $varnt = Title
::newFromText($varlink);
1181 if($varnt && $varnt->getArticleID()>0) {
1186 if($varnt && $varnt->getArticleID()>0) {
1192 $ns = $nt->getNamespace();
1193 $iw = $nt->getInterWiki();
1195 if ($might_be_img) { # if this is actually an invalid link
1196 if ($ns == NS_IMAGE
&& $noforce) { #but might be an image
1198 while (isset ($a[$k+
1]) ) {
1199 #look at the next 'line' to see if we can close it there
1200 $next_line = array_shift(array_splice( $a, $k +
1, 1) );
1201 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1202 # the first ]] closes the inner link, the second the image
1204 $text .= '[[' . $m[1];
1207 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1208 #if there's exactly one ]] that's fine, we'll keep looking
1209 $text .= '[[' . $m[0];
1211 #if $next_line is invalid too, we need look no further
1212 $text .= '[[' . $next_line;
1217 # we couldn't find the end of this imageLink, so output it raw
1218 $s .= $prefix . '[[' . $link . '|' . $text;
1219 # note: no $trail, because without an end, there *is* no trail
1222 } else { #it's not an image, so output it raw
1223 $s .= $prefix . '[[' . $link . '|' . $text;
1224 # note: no $trail, because without an end, there *is* no trail
1229 $wasblank = ( '' == $text );
1230 if( $wasblank ) $text = $link;
1233 # Link not escaped by : , create the various objects
1237 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1238 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1239 $tmp = $prefix . $trail ;
1240 $s .= (trim($tmp) == '')?
'': $tmp;
1244 if ( $ns == NS_IMAGE
) {
1245 # recursively parse links inside the image caption
1246 # actually, this will parse them in any other parameters, too,
1247 # but it might be hard to fix that, and it doesn't matter ATM
1248 $text = $this->replaceExternalLinks($text);
1249 $text = $this->replaceInternalLinks($text);
1251 # replace the image with a link-holder so that replaceExternalLinks() can't mess with it
1252 $s .= $prefix . $this->insertStripItem( $sk->makeImageLinkObj( $nt, $text ), $this->mStripState
) . $trail;
1253 $wgLinkCache->addImageLinkObj( $nt );
1257 if ( $ns == NS_CATEGORY
) {
1258 $t = $nt->getText() ;
1260 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1261 $pPLC=$sk->postParseLinkColour();
1262 $sk->postParseLinkColour( false );
1263 $t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
1264 $sk->postParseLinkColour( $pPLC );
1265 $wgLinkCache->resume();
1268 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1269 $sortkey = $this->mTitle
->getText();
1271 $sortkey = $this->mTitle
->getPrefixedText();
1276 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1277 $this->mOutput
->mCategoryLinks
[] = $t ;
1278 $s .= $prefix . $trail ;
1283 if( ( $nt->getPrefixedText() === $this->mTitle
->getPrefixedText() ) &&
1284 ( strpos( $link, '#' ) === FALSE ) ) {
1285 # Self-links are handled specially; generally de-link and change to bold.
1286 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1290 # Special and Media are pseudo-namespaces; no pages actually exist in them
1291 if( $ns == NS_MEDIA
) {
1292 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1293 $wgLinkCache->addImageLinkObj( $nt );
1295 } elseif( $ns == NS_SPECIAL
) {
1296 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1299 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1301 wfProfileOut( $fname );
1306 * Handle link to subpage if necessary
1307 * @param $target string the source of the link
1308 * @param &$text the link text, modified as necessary
1309 * @return string the full name of the link
1312 function maybeDoSubpageLink($target, &$text) {
1315 # :Foobar -- override special treatment of prefix (images, language links)
1316 # /Foobar -- convert to CurrentPage/Foobar
1317 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1318 global $wgNamespacesWithSubpages;
1320 $fname = 'Parser::maybeDoSubpageLink';
1321 wfProfileIn( $fname );
1322 # Look at the first character
1323 if( $target{0} == '/' ) {
1324 # / at end means we don't want the slash to be shown
1325 if(substr($target,-1,1)=='/') {
1326 $target=substr($target,1,-1);
1329 $noslash=substr($target,1);
1332 # Some namespaces don't allow subpages
1333 if(!empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()])) {
1334 # subpages allowed here
1335 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1336 if( '' === $text ) {
1338 } # this might be changed for ugliness reasons
1340 # no subpage allowed, use standard link
1348 wfProfileOut( $fname );
1353 * Used by doBlockLevels()
1356 /* private */ function closeParagraph() {
1358 if ( '' != $this->mLastSection
) {
1359 $result = '</' . $this->mLastSection
. ">\n";
1361 $this->mInPre
= false;
1362 $this->mLastSection
= '';
1365 # getCommon() returns the length of the longest common substring
1366 # of both arguments, starting at the beginning of both.
1368 /* private */ function getCommon( $st1, $st2 ) {
1369 $fl = strlen( $st1 );
1370 $shorter = strlen( $st2 );
1371 if ( $fl < $shorter ) { $shorter = $fl; }
1373 for ( $i = 0; $i < $shorter; ++
$i ) {
1374 if ( $st1{$i} != $st2{$i} ) { break; }
1378 # These next three functions open, continue, and close the list
1379 # element appropriate to the prefix character passed into them.
1381 /* private */ function openList( $char ) {
1382 $result = $this->closeParagraph();
1384 if ( '*' == $char ) { $result .= '<ul><li>'; }
1385 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1386 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1387 else if ( ';' == $char ) {
1388 $result .= '<dl><dt>';
1389 $this->mDTopen
= true;
1391 else { $result = '<!-- ERR 1 -->'; }
1396 /* private */ function nextItem( $char ) {
1397 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1398 else if ( ':' == $char ||
';' == $char ) {
1400 if ( $this->mDTopen
) { $close = '</dt>'; }
1401 if ( ';' == $char ) {
1402 $this->mDTopen
= true;
1403 return $close . '<dt>';
1405 $this->mDTopen
= false;
1406 return $close . '<dd>';
1409 return '<!-- ERR 2 -->';
1412 /* private */ function closeList( $char ) {
1413 if ( '*' == $char ) { $text = '</li></ul>'; }
1414 else if ( '#' == $char ) { $text = '</li></ol>'; }
1415 else if ( ':' == $char ) {
1416 if ( $this->mDTopen
) {
1417 $this->mDTopen
= false;
1418 $text = '</dt></dl>';
1420 $text = '</dd></dl>';
1423 else { return '<!-- ERR 3 -->'; }
1429 * Make lists from lines starting with ':', '*', '#', etc.
1432 * @return string the lists rendered as HTML
1434 function doBlockLevels( $text, $linestart ) {
1435 $fname = 'Parser::doBlockLevels';
1436 wfProfileIn( $fname );
1438 # Parsing through the text line by line. The main thing
1439 # happening here is handling of block-level elements p, pre,
1440 # and making lists from lines starting with * # : etc.
1442 $textLines = explode( "\n", $text );
1444 $lastPrefix = $output = $lastLine = '';
1445 $this->mDTopen
= $inBlockElem = false;
1447 $paragraphStack = false;
1449 if ( !$linestart ) {
1450 $output .= array_shift( $textLines );
1452 foreach ( $textLines as $oLine ) {
1453 $lastPrefixLength = strlen( $lastPrefix );
1454 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1455 $preOpenMatch = preg_match('/<pre/i', $oLine );
1456 if ( !$this->mInPre
) {
1457 # Multiple prefixes may abut each other for nested lists.
1458 $prefixLength = strspn( $oLine, '*#:;' );
1459 $pref = substr( $oLine, 0, $prefixLength );
1462 $pref2 = str_replace( ';', ':', $pref );
1463 $t = substr( $oLine, $prefixLength );
1464 $this->mInPre
= !empty($preOpenMatch);
1466 # Don't interpret any other prefixes in preformatted text
1468 $pref = $pref2 = '';
1473 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1474 # Same as the last item, so no need to deal with nesting or opening stuff
1475 $output .= $this->nextItem( substr( $pref, -1 ) );
1476 $paragraphStack = false;
1478 if ( substr( $pref, -1 ) == ';') {
1479 # The one nasty exception: definition lists work like this:
1480 # ; title : definition text
1481 # So we check for : in the remainder text to split up the
1482 # title and definition, without b0rking links.
1483 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1485 $output .= $term . $this->nextItem( ':' );
1488 } elseif( $prefixLength ||
$lastPrefixLength ) {
1489 # Either open or close a level...
1490 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1491 $paragraphStack = false;
1493 while( $commonPrefixLength < $lastPrefixLength ) {
1494 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1495 --$lastPrefixLength;
1497 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1498 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1500 while ( $prefixLength > $commonPrefixLength ) {
1501 $char = substr( $pref, $commonPrefixLength, 1 );
1502 $output .= $this->openList( $char );
1504 if ( ';' == $char ) {
1505 # FIXME: This is dupe of code above
1506 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1508 $output .= $term . $this->nextItem( ':' );
1511 ++
$commonPrefixLength;
1513 $lastPrefix = $pref2;
1515 if( 0 == $prefixLength ) {
1516 # No prefix (not in list)--go to paragraph mode
1517 $uniq_prefix = UNIQ_PREFIX
;
1518 // XXX: use a stack for nestable elements like span, table and div
1519 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1520 $closematch = preg_match(
1521 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1522 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1523 if ( $openmatch or $closematch ) {
1524 $paragraphStack = false;
1525 $output .= $this->closeParagraph();
1526 if($preOpenMatch and !$preCloseMatch) {
1527 $this->mInPre
= true;
1529 if ( $closematch ) {
1530 $inBlockElem = false;
1532 $inBlockElem = true;
1534 } else if ( !$inBlockElem && !$this->mInPre
) {
1535 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1537 if ($this->mLastSection
!= 'pre') {
1538 $paragraphStack = false;
1539 $output .= $this->closeParagraph().'<pre>';
1540 $this->mLastSection
= 'pre';
1542 $t = substr( $t, 1 );
1545 if ( '' == trim($t) ) {
1546 if ( $paragraphStack ) {
1547 $output .= $paragraphStack.'<br />';
1548 $paragraphStack = false;
1549 $this->mLastSection
= 'p';
1551 if ($this->mLastSection
!= 'p' ) {
1552 $output .= $this->closeParagraph();
1553 $this->mLastSection
= '';
1554 $paragraphStack = '<p>';
1556 $paragraphStack = '</p><p>';
1560 if ( $paragraphStack ) {
1561 $output .= $paragraphStack;
1562 $paragraphStack = false;
1563 $this->mLastSection
= 'p';
1564 } else if ($this->mLastSection
!= 'p') {
1565 $output .= $this->closeParagraph().'<p>';
1566 $this->mLastSection
= 'p';
1572 if ($paragraphStack === false) {
1576 while ( $prefixLength ) {
1577 $output .= $this->closeList( $pref2{$prefixLength-1} );
1580 if ( '' != $this->mLastSection
) {
1581 $output .= '</' . $this->mLastSection
. '>';
1582 $this->mLastSection
= '';
1585 wfProfileOut( $fname );
1590 * Split up a string on ':', ignoring any occurences inside
1591 * <a>..</a> or <span>...</span>
1592 * @param $str string the string to split
1593 * @param &$before string set to everything before the ':'
1594 * @param &$after string set to everything after the ':'
1595 * return string the position of the ':', or false if none found
1597 function findColonNoLinks($str, &$before, &$after) {
1598 # I wonder if we should make this count all tags, not just <a>
1599 # and <span>. That would prevent us from matching a ':' that
1600 # comes in the middle of italics other such formatting....
1602 $fname = 'Parser::findColonNoLinks';
1603 wfProfileIn( $fname );
1606 $colon = strpos($str, ':', $pos);
1608 if ($colon !== false) {
1609 $before = substr($str, 0, $colon);
1610 $after = substr($str, $colon +
1);
1612 # Skip any ':' within <a> or <span> pairs
1613 $a = substr_count($before, '<a');
1614 $s = substr_count($before, '<span');
1615 $ca = substr_count($before, '</a>');
1616 $cs = substr_count($before, '</span>');
1618 if ($a <= $ca and $s <= $cs) {
1619 # Tags are balanced before ':'; ok
1624 } while ($colon !== false);
1625 wfProfileOut( $fname );
1630 * Return value of a magic variable (like PAGENAME)
1634 function getVariableValue( $index ) {
1635 global $wgContLang, $wgSitename, $wgServer;
1638 case MAG_CURRENTMONTH
:
1639 return $wgContLang->formatNum( date( 'm' ) );
1640 case MAG_CURRENTMONTHNAME
:
1641 return $wgContLang->getMonthName( date('n') );
1642 case MAG_CURRENTMONTHNAMEGEN
:
1643 return $wgContLang->getMonthNameGen( date('n') );
1644 case MAG_CURRENTDAY
:
1645 return $wgContLang->formatNum( date('j') );
1647 return $this->mTitle
->getText();
1649 return $this->mTitle
->getPartialURL();
1651 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1652 return $wgContLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1653 case MAG_CURRENTDAYNAME
:
1654 return $wgContLang->getWeekdayName( date('w')+
1 );
1655 case MAG_CURRENTYEAR
:
1656 return $wgContLang->formatNum( date( 'Y' ) );
1657 case MAG_CURRENTTIME
:
1658 return $wgContLang->time( wfTimestampNow(), false );
1659 case MAG_NUMBEROFARTICLES
:
1660 return $wgContLang->formatNum( wfNumberOfArticles() );
1671 * initialise the magic variables (like CURRENTMONTHNAME)
1675 function initialiseVariables() {
1676 $fname = 'Parser::initialiseVariables';
1677 wfProfileIn( $fname );
1678 global $wgVariableIDs;
1679 $this->mVariables
= array();
1680 foreach ( $wgVariableIDs as $id ) {
1681 $mw =& MagicWord
::get( $id );
1682 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1684 wfProfileOut( $fname );
1688 * Replace magic variables, templates, and template arguments
1689 * with the appropriate text. Templates are substituted recursively,
1690 * taking care to avoid infinite loops.
1692 * Note that the substitution depends on value of $mOutputType:
1693 * OT_WIKI: only {{subst:}} templates
1694 * OT_MSG: only magic variables
1695 * OT_HTML: all templates and magic variables
1697 * @param string $tex The text to transform
1698 * @param array $args Key-value pairs representing template parameters to substitute
1701 function replaceVariables( $text, $args = array() ) {
1702 global $wgLang, $wgScript, $wgArticlePath;
1704 # Prevent too big inclusions
1705 if(strlen($text)> MAX_INCLUDE_SIZE
)
1708 $fname = 'Parser::replaceVariables';
1709 wfProfileIn( $fname );
1711 $titleChars = Title
::legalChars();
1713 # This function is called recursively. To keep track of arguments we need a stack:
1714 array_push( $this->mArgStack
, $args );
1716 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1717 $GLOBALS['wgCurParser'] =& $this;
1719 # Variable substitution
1720 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", 'wfVariableSubstitution', $text );
1722 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1723 # Argument substitution
1724 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1726 # Template substitution
1727 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1728 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1730 array_pop( $this->mArgStack
);
1732 wfProfileOut( $fname );
1737 * Replace magic variables
1740 function variableSubstitution( $matches ) {
1741 if ( !$this->mVariables
) {
1742 $this->initialiseVariables();
1745 if ( $this->mOutputType
== OT_WIKI
) {
1746 # Do only magic variables prefixed by SUBST
1747 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1748 if (!$mwSubst->matchStartAndRemove( $matches[1] ))
1750 # Note that if we don't substitute the variable below,
1751 # we don't remove the {{subst:}} magic word, in case
1752 # it is a template rather than a magic variable.
1754 if ( !$skip && array_key_exists( $matches[1], $this->mVariables
) ) {
1755 $text = $this->mVariables
[$matches[1]];
1756 $this->mOutput
->mContainsOldMagic
= true;
1758 $text = $matches[0];
1763 # Split template arguments
1764 function getTemplateArgs( $argsString ) {
1765 if ( $argsString === '' ) {
1769 $args = explode( '|', substr( $argsString, 1 ) );
1771 # If any of the arguments contains a '[[' but no ']]', it needs to be
1772 # merged with the next arg because the '|' character between belongs
1773 # to the link syntax and not the template parameter syntax.
1774 $argc = count($args);
1776 for ( $i = 0; $i < $argc-1; $i++
) {
1777 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1778 $args[$i] .= '|'.$args[$i+
1];
1779 array_splice($args, $i+
1, 1);
1789 * Return the text of a template, after recursively
1790 * replacing any variables or templates within the template.
1792 * @param array $matches The parts of the template
1793 * $matches[1]: the title, i.e. the part before the |
1794 * $matches[2]: the parameters (including a leading |), if any
1795 * @return string the text of the template
1798 function braceSubstitution( $matches ) {
1799 global $wgLinkCache, $wgContLang;
1800 $fname = 'Parser::braceSubstitution';
1807 # Need to know if the template comes at the start of a line,
1808 # to treat the beginning of the template like the beginning
1809 # of a line for tables and block-level elements.
1810 $linestart = $matches[1];
1812 # $part1 is the bit before the first |, and must contain only title characters
1813 # $args is a list of arguments, starting from index 0, not including $part1
1815 $part1 = $matches[2];
1816 # If the third subpattern matched anything, it will start with |
1818 $args = $this->getTemplateArgs($matches[3]);
1819 $argc = count( $args );
1821 # Don't parse {{{}}} because that's only for template arguments
1822 if ( $linestart === '{' ) {
1823 $text = $matches[0];
1830 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1831 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
1832 # One of two possibilities is true:
1833 # 1) Found SUBST but not in the PST phase
1834 # 2) Didn't find SUBST and in the PST phase
1835 # In either case, return without further processing
1836 $text = $matches[0];
1842 # MSG, MSGNW and INT
1845 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1846 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1849 # Remove obsolete MSG:
1850 $mwMsg =& MagicWord
::get( MAG_MSG
);
1851 $mwMsg->matchStartAndRemove( $part1 );
1854 # Check if it is an internal message
1855 $mwInt =& MagicWord
::get( MAG_INT
);
1856 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1857 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1858 $text = $linestart . wfMsgReal( $part1, $args, true );
1866 # Check for NS: (namespace expansion)
1867 $mwNs = MagicWord
::get( MAG_NS
);
1868 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1869 if ( intval( $part1 ) ) {
1870 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1873 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1874 if ( !is_null( $index ) ) {
1875 $text = $linestart . $wgContLang->getNsText( $index );
1882 # LOCALURL and LOCALURLE
1884 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1885 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1887 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1888 $func = 'getLocalURL';
1889 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1890 $func = 'escapeLocalURL';
1895 if ( $func !== '' ) {
1896 $title = Title
::newFromText( $part1 );
1897 if ( !is_null( $title ) ) {
1899 $text = $linestart . $title->$func( $args[0] );
1901 $text = $linestart . $title->$func();
1909 if ( !$found && $argc == 1 ) {
1910 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
1911 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1912 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
1917 # Template table test
1919 # Did we encounter this template already? If yes, it is in the cache
1920 # and we need to check for loops.
1921 if ( !$found && isset( $this->mTemplates
[$part1] ) ) {
1922 # set $text to cached message.
1923 $text = $linestart . $this->mTemplates
[$part1];
1926 # Infinite loop test
1927 if ( isset( $this->mTemplatePath
[$part1] ) ) {
1930 $text .= '<!-- WARNING: template loop detected -->';
1934 # Load from database
1935 $itcamefromthedatabase = false;
1938 $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
1939 if ($subpage !== '') {
1940 $ns = $this->mTitle
->getNamespace();
1942 $title = Title
::newFromText( $part1, $ns );
1943 if ( !is_null( $title ) && !$title->isExternal() ) {
1944 # Check for excessive inclusion
1945 $dbk = $title->getPrefixedDBkey();
1946 if ( $this->incrementIncludeCount( $dbk ) ) {
1947 # This should never be reached.
1948 $article = new Article( $title );
1949 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1950 if ( $articleContent !== false ) {
1952 $text = $linestart . $articleContent;
1953 $itcamefromthedatabase = true;
1957 # If the title is valid but undisplayable, make a link to it
1958 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1959 $text = $linestart . '[['.$title->getPrefixedText().']]';
1963 # Template cache array insertion
1964 $this->mTemplates
[$part1] = $text;
1968 # Recursive parsing, escaping and link table handling
1969 # Only for HTML output
1970 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1971 $text = wfEscapeWikiText( $text );
1972 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
1973 # Clean up argument array
1974 $assocArgs = array();
1976 foreach( $args as $arg ) {
1977 $eqpos = strpos( $arg, '=' );
1978 if ( $eqpos === false ) {
1979 $assocArgs[$index++
] = $arg;
1981 $name = trim( substr( $arg, 0, $eqpos ) );
1982 $value = trim( substr( $arg, $eqpos+
1 ) );
1983 if ( $value === false ) {
1986 if ( $name !== false ) {
1987 $assocArgs[$name] = $value;
1992 # Add a new element to the templace recursion path
1993 $this->mTemplatePath
[$part1] = 1;
1995 $text = $this->strip( $text, $this->mStripState
);
1996 $text = $this->removeHTMLtags( $text );
1997 $text = $this->replaceVariables( $text, $assocArgs );
1999 # Resume the link cache and register the inclusion as a link
2000 if ( $this->mOutputType
== OT_HTML
&& !is_null( $title ) ) {
2001 $wgLinkCache->addLinkObj( $title );
2004 # If the template begins with a table or block-level
2005 # element, it should be treated as beginning a new line.
2006 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
2007 $text = "\n" . $text;
2011 # Empties the template path
2012 $this->mTemplatePath
= array();
2016 # replace ==section headers==
2017 # XXX this needs to go away once we have a better parser.
2018 if ( $this->mOutputType
!= OT_WIKI
&& $itcamefromthedatabase ) {
2019 if( !is_null( $title ) )
2020 $encodedname = base64_encode($title->getPrefixedDBkey());
2022 $encodedname = base64_encode("");
2023 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
2024 PREG_SPLIT_DELIM_CAPTURE
);
2027 for( $i = 0; $i < count($m); $i +
= 2 ) {
2029 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
2031 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
2035 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
2036 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
2037 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
2044 # Empties the template path
2045 $this->mTemplatePath
= array();
2054 * Triple brace replacement -- used for template arguments
2057 function argSubstitution( $matches ) {
2058 $arg = trim( $matches[1] );
2059 $text = $matches[0];
2060 $inputArgs = end( $this->mArgStack
);
2062 if ( array_key_exists( $arg, $inputArgs ) ) {
2063 $text = $inputArgs[$arg];
2070 * Returns true if the function is allowed to include this entity
2073 function incrementIncludeCount( $dbk ) {
2074 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
2075 $this->mIncludeCount
[$dbk] = 0;
2077 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
2086 * Cleans up HTML, removes dangerous tags and attributes, and
2087 * removes HTML comments
2090 function removeHTMLtags( $text ) {
2091 global $wgUseTidy, $wgUserHtml;
2092 $fname = 'Parser::removeHTMLtags';
2093 wfProfileIn( $fname );
2096 $htmlpairs = array( # Tags that must be closed
2097 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
2098 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
2099 'strike', 'strong', 'tt', 'var', 'div', 'center',
2100 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
2101 'ruby', 'rt' , 'rb' , 'rp', 'p'
2103 $htmlsingle = array(
2104 'br', 'hr', 'li', 'dt', 'dd'
2106 $htmlnest = array( # Tags that can be nested--??
2107 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
2108 'dl', 'font', 'big', 'small', 'sub', 'sup'
2110 $tabletags = array( # Can only appear inside table
2114 $htmlpairs = array();
2115 $htmlsingle = array();
2116 $htmlnest = array();
2117 $tabletags = array();
2120 $htmlsingle = array_merge( $tabletags, $htmlsingle );
2121 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
2123 $htmlattrs = $this->getHTMLattrs () ;
2125 # Remove HTML comments
2126 $text = $this->removeHTMLcomments( $text );
2128 $bits = explode( '<', $text );
2129 $text = array_shift( $bits );
2131 $tagstack = array(); $tablestack = array();
2132 foreach ( $bits as $x ) {
2133 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
2134 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2136 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2137 error_reporting( $prev );
2140 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2144 if ( ! in_array( $t, $htmlsingle ) &&
2145 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
2146 @array_push
( $tagstack, $ot );
2149 if ( $t == 'table' ) {
2150 $tagstack = array_pop( $tablestack );
2155 # Keep track for later
2156 if ( in_array( $t, $tabletags ) &&
2157 ! in_array( 'table', $tagstack ) ) {
2159 } else if ( in_array( $t, $tagstack ) &&
2160 ! in_array ( $t , $htmlnest ) ) {
2162 } else if ( ! in_array( $t, $htmlsingle ) ) {
2163 if ( $t == 'table' ) {
2164 array_push( $tablestack, $tagstack );
2165 $tagstack = array();
2167 array_push( $tagstack, $t );
2169 # Strip non-approved attributes from the tag
2170 $newparams = $this->fixTagAttributes($params);
2174 $rest = str_replace( '>', '>', $rest );
2175 $text .= "<$slash$t $newparams$brace$rest";
2179 $text .= '<' . str_replace( '>', '>', $x);
2181 # Close off any remaining tags
2182 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2184 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2187 # this might be possible using tidy itself
2188 foreach ( $bits as $x ) {
2189 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2191 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2192 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2193 $newparams = $this->fixTagAttributes($params);
2194 $rest = str_replace( '>', '>', $rest );
2195 $text .= "<$slash$t $newparams$brace$rest";
2197 $text .= '<' . str_replace( '>', '>', $x);
2201 wfProfileOut( $fname );
2206 * Remove '<!--', '-->', and everything between.
2207 * To avoid leaving blank lines, when a comment is both preceded
2208 * and followed by a newline (ignoring spaces), trim leading and
2209 * trailing spaces and one of the newlines.
2213 function removeHTMLcomments( $text ) {
2214 $fname='Parser::removeHTMLcomments';
2215 wfProfileIn( $fname );
2216 while (($start = strpos($text, '<!--')) !== false) {
2217 $end = strpos($text, '-->', $start +
4);
2218 if ($end === false) {
2219 # Unterminated comment; bail out
2225 # Trim space and newline if the comment is both
2226 # preceded and followed by a newline
2227 $spaceStart = max($start - 1, 0);
2228 $spaceLen = $end - $spaceStart;
2229 while (substr($text, $spaceStart, 1) === ' ' && $spaceStart > 0) {
2233 while (substr($text, $spaceStart +
$spaceLen, 1) === ' ')
2235 if (substr($text, $spaceStart, 1) === "\n" and substr($text, $spaceStart +
$spaceLen, 1) === "\n") {
2236 # Remove the comment, leading and trailing
2237 # spaces, and leave only one newline.
2238 $text = substr_replace($text, "\n", $spaceStart, $spaceLen +
1);
2241 # Remove just the comment.
2242 $text = substr_replace($text, '', $start, $end - $start);
2245 wfProfileOut( $fname );
2250 * This function accomplishes several tasks:
2251 * 1) Auto-number headings if that option is enabled
2252 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2253 * 3) Add a Table of contents on the top for users who have enabled the option
2254 * 4) Auto-anchor headings
2256 * It loops through all headlines, collects the necessary data, then splits up the
2257 * string and re-inserts the newly formatted headlines.
2260 /* private */ function formatHeadings( $text, $isMain=true ) {
2261 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders;
2263 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2264 $doShowToc = $this->mOptions
->getShowToc();
2265 $forceTocHere = false;
2266 if( !$this->mTitle
->userCanEdit() ) {
2268 $rightClickHack = 0;
2270 $showEditLink = $this->mOptions
->getEditSection();
2271 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
2274 # Inhibit editsection links if requested in the page
2275 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2276 if( $esw->matchAndRemove( $text ) ) {
2279 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2281 $mw =& MagicWord
::get( MAG_NOTOC
);
2282 if( $mw->matchAndRemove( $text ) ) {
2286 # never add the TOC to the Main Page. This is an entry page that should not
2287 # be more than 1-2 screens large anyway
2288 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
2292 # Get all headlines for numbering them and adding funky stuff like [edit]
2293 # links - this is for later, but we need the number of headlines right now
2294 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
2296 # if there are fewer than 4 headlines in the article, do not show TOC
2297 if( $numMatches < 4 ) {
2301 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2302 # override above conditions and always show TOC at that place
2303 $mw =& MagicWord
::get( MAG_TOC
);
2304 if ($mw->match( $text ) ) {
2306 $forceTocHere = true;
2308 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2309 # override above conditions and always show TOC above first header
2310 $mw =& MagicWord
::get( MAG_FORCETOC
);
2311 if ($mw->matchAndRemove( $text ) ) {
2318 # We need this to perform operations on the HTML
2319 $sk =& $this->mOptions
->getSkin();
2323 $sectionCount = 0; # headlineCount excluding template sections
2325 # Ugh .. the TOC should have neat indentation levels which can be
2326 # passed to the skin functions. These are determined here
2331 $sublevelCount = array();
2334 foreach( $matches[3] as $headline ) {
2336 $templatetitle = "";
2337 $templatesection = 0;
2339 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2341 $templatetitle = base64_decode($mat[1]);
2342 $templatesection = 1 +
(int)base64_decode($mat[2]);
2343 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2348 $prevlevel = $level;
2350 $level = $matches[1][$headlineCount];
2351 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2352 # reset when we enter a new level
2353 $sublevelCount[$level] = 0;
2354 $toc .= $sk->tocIndent( $level - $prevlevel );
2355 $toclevel +
= $level - $prevlevel;
2357 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2358 # reset when we step back a level
2359 $sublevelCount[$level+
1]=0;
2360 $toc .= $sk->tocUnindent( $prevlevel - $level );
2361 $toclevel -= $prevlevel - $level;
2363 # count number of headlines for each level
2364 @$sublevelCount[$level]++
;
2365 if( $doNumberHeadings ||
$doShowToc ) {
2367 for( $i = 1; $i <= $level; $i++
) {
2368 if( !empty( $sublevelCount[$i] ) ) {
2372 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2378 # The canonized header is a version of the header text safe to use for links
2379 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2380 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2381 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2383 # Remove link placeholders by the link text.
2384 # <!--LINK number-->
2386 # link text with suffix
2387 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2388 "\$wgLinkHolders['texts'][\$1]",
2389 $canonized_headline );
2392 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2393 $tocline = trim( $canonized_headline );
2394 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2395 $replacearray = array(
2399 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2400 $refer[$headlineCount] = $canonized_headline;
2402 # count how many in assoc. array so we can track dupes in anchors
2403 @$refers[$canonized_headline]++
;
2404 $refcount[$headlineCount]=$refers[$canonized_headline];
2406 # Prepend the number to the heading text
2408 if( $doNumberHeadings ||
$doShowToc ) {
2409 $tocline = $numbering . ' ' . $tocline;
2411 # Don't number the heading if it is the only one (looks silly)
2412 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2413 # the two are different if the line contains a link
2414 $headline=$numbering . ' ' . $headline;
2418 # Create the anchor for linking from the TOC to the section
2419 $anchor = $canonized_headline;
2420 if($refcount[$headlineCount] > 1 ) {
2421 $anchor .= '_' . $refcount[$headlineCount];
2423 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2424 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2426 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2427 if ( empty( $head[$headlineCount] ) ) {
2428 $head[$headlineCount] = '';
2431 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2433 $head[$headlineCount] .= $sk->editSectionLink($sectionCount+
1);
2436 # Add the edit section span
2437 if( $rightClickHack ) {
2439 $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
2441 $headline = $sk->editSectionScript($sectionCount+
1,$headline);
2444 # give headline the correct <h#> tag
2445 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2453 $toclines = $headlineCount;
2454 $toc .= $sk->tocUnindent( $toclevel );
2455 $toc = $sk->tocTable( $toc );
2458 # split up and insert constructed headlines
2460 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2463 foreach( $blocks as $block ) {
2464 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2465 # This is the [edit] link that appears for the top block of text when
2466 # section editing is enabled
2468 # Disabled because it broke block formatting
2469 # For example, a bullet point in the top line
2470 # $full .= $sk->editSectionLink(0);
2473 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2474 # Top anchor now in skin
2478 if( !empty( $head[$i] ) ) {
2484 $mw =& MagicWord
::get( MAG_TOC
);
2485 return $mw->replace( $toc, $full );
2492 * Return an HTML link for the "ISBN 123456" text
2495 function magicISBN( $text ) {
2497 $fname = 'Parser::magicISBN';
2498 wfProfileIn( $fname );
2500 $a = split( 'ISBN ', ' '.$text );
2501 if ( count ( $a ) < 2 ) {
2502 wfProfileOut( $fname );
2505 $text = substr( array_shift( $a ), 1);
2506 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2508 foreach ( $a as $x ) {
2509 $isbn = $blank = '' ;
2510 while ( ' ' == $x{0} ) {
2512 $x = substr( $x, 1 );
2514 if ( $x == '' ) { # blank isbn
2515 $text .= "ISBN $blank";
2518 while ( strstr( $valid, $x{0} ) != false ) {
2520 $x = substr( $x, 1 );
2522 $num = str_replace( '-', '', $isbn );
2523 $num = str_replace( ' ', '', $num );
2526 $text .= "ISBN $blank$x";
2528 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2529 $text .= '<a href="' .
2530 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2531 "\" class=\"internal\">ISBN $isbn</a>";
2535 wfProfileOut( $fname );
2540 * Return an HTML link for the "GEO ..." text
2543 function magicGEO( $text ) {
2544 global $wgLang, $wgUseGeoMode;
2545 $fname = 'Parser::magicGEO';
2546 wfProfileIn( $fname );
2548 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2549 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2550 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2551 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2552 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2553 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2555 $a = split( 'GEO ', ' '.$text );
2556 if ( count ( $a ) < 2 ) {
2557 wfProfileOut( $fname );
2560 $text = substr( array_shift( $a ), 1);
2561 $valid = '0123456789.+-:';
2563 foreach ( $a as $x ) {
2564 $geo = $blank = '' ;
2565 while ( ' ' == $x{0} ) {
2567 $x = substr( $x, 1 );
2569 while ( strstr( $valid, $x{0} ) != false ) {
2571 $x = substr( $x, 1 );
2573 $num = str_replace( '+', '', $geo );
2574 $num = str_replace( ' ', '', $num );
2576 if ( '' == $num ||
count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2577 $text .= "GEO $blank$x";
2579 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2580 $text .= '<a href="' .
2581 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2582 "\" class=\"internal\">GEO $geo</a>";
2586 wfProfileOut( $fname );
2591 * Return an HTML link for the "RFC 1234" text
2593 * @param string $text text to be processed
2595 function magicRFC( $text ) {
2598 $valid = '0123456789';
2601 $a = split( 'RFC ', ' '.$text );
2602 if ( count ( $a ) < 2 ) return $text;
2603 $text = substr( array_shift( $a ), 1);
2605 /* Check if RFC keyword is preceed by [[.
2606 * This test is made here cause of the array_shift above
2607 * that prevent the test to be done in the foreach.
2609 if(substr($text, -2) == '[[') { $internal = true; }
2611 foreach ( $a as $x ) {
2612 /* token might be empty if we have RFC RFC 1234 */
2618 $rfc = $blank = '' ;
2620 /** remove and save whitespaces in $blank */
2621 while ( $x{0} == ' ' ) {
2623 $x = substr( $x, 1 );
2626 /** remove and save the rfc number in $rfc */
2627 while ( strstr( $valid, $x{0} ) != false ) {
2629 $x = substr( $x, 1 );
2633 /* call back stripped spaces*/
2634 $text .= "RFC $blank$x";
2635 } elseif( $internal) {
2637 $text .= "RFC $rfc$x";
2639 /* build the external link*/
2640 $url = wfmsg( 'rfcurl' );
2641 $url = str_replace( '$1', $rfc, $url);
2642 $sk =& $this->mOptions
->getSkin();
2643 $la = $sk->getExternalLinkAttributes( $url, 'RFC '.$rfc );
2644 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2647 /* Check if the next RFC keyword is preceed by [[ */
2648 $internal = (substr($x,-2) == '[[');
2654 * Transform wiki markup when saving a page by doing \r\n -> \n
2655 * conversion, substitting signatures, {{subst:}} templates, etc.
2657 * @param string $text the text to transform
2658 * @param Title &$title the Title object for the current article
2659 * @param User &$user the User object describing the current user
2660 * @param ParserOptions $options parsing options
2661 * @param bool $clearState whether to clear the parser state first
2662 * @return string the altered wiki markup
2665 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2666 $this->mOptions
= $options;
2667 $this->mTitle
=& $title;
2668 $this->mOutputType
= OT_WIKI
;
2670 if ( $clearState ) {
2671 $this->clearState();
2674 $stripState = false;
2678 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2682 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2683 "/<br *?>/i" => "<br />",
2685 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2687 $text = $this->strip( $text, $stripState, false );
2688 $text = $this->pstPass2( $text, $user );
2689 $text = $this->unstrip( $text, $stripState );
2690 $text = $this->unstripNoWiki( $text, $stripState );
2695 * Pre-save transform helper function
2698 function pstPass2( $text, &$user ) {
2699 global $wgLang, $wgContLang, $wgLocaltimezone, $wgCurParser;
2701 # Variable replacement
2702 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2703 $text = $this->replaceVariables( $text );
2707 $n = $user->getName();
2708 $k = $user->getOption( 'nickname' );
2709 if ( '' == $k ) { $k = $n; }
2710 if(isset($wgLocaltimezone)) {
2711 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2713 /* Note: this is an ugly timezone hack for the European wikis */
2714 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
2715 ' (' . date( 'T' ) . ')';
2716 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2718 $text = preg_replace( '/~~~~~/', $d, $text );
2719 $text = preg_replace( '/~~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]] $d", $text );
2720 $text = preg_replace( '/~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]]", $text );
2722 # Context links: [[|name]] and [[name (context)|]]
2724 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2725 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2726 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2727 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2729 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2730 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2731 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2732 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2734 $t = $this->mTitle
->getText();
2735 if ( preg_match( $conpat, $t, $m ) ) {
2738 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2739 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2740 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2742 if ( '' == $context ) {
2743 $text = preg_replace( $p2, '[[\\1]]', $text );
2745 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2748 # Trim trailing whitespace
2749 # MAG_END (__END__) tag allows for trailing
2750 # whitespace to be deliberately included
2751 $text = rtrim( $text );
2752 $mw =& MagicWord
::get( MAG_END
);
2753 $mw->matchAndRemove( $text );
2759 * Set up some variables which are usually set up in parse()
2760 * so that an external function can call some class members with confidence
2763 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2764 $this->mTitle
=& $title;
2765 $this->mOptions
= $options;
2766 $this->mOutputType
= $outputType;
2767 if ( $clearState ) {
2768 $this->clearState();
2773 * Transform a MediaWiki message by replacing magic variables.
2775 * @param string $text the text to transform
2776 * @param ParserOptions $options options
2777 * @return string the text with variables substituted
2780 function transformMsg( $text, $options ) {
2782 static $executing = false;
2784 # Guard against infinite recursion
2790 $this->mTitle
= $wgTitle;
2791 $this->mOptions
= $options;
2792 $this->mOutputType
= OT_MSG
;
2793 $this->clearState();
2794 $text = $this->replaceVariables( $text );
2801 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2802 * Callback will be called with the text within
2803 * Transform and return the text within
2806 function setHook( $tag, $callback ) {
2807 $oldVal = @$this->mTagHooks
[$tag];
2808 $this->mTagHooks
[$tag] = $callback;
2815 * @package MediaWiki
2819 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2820 var $mCacheTime; # Used in ParserCache
2822 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
2823 $containsOldMagic = false )
2825 $this->mText
= $text;
2826 $this->mLanguageLinks
= $languageLinks;
2827 $this->mCategoryLinks
= $categoryLinks;
2828 $this->mContainsOldMagic
= $containsOldMagic;
2829 $this->mCacheTime
= '';
2832 function getText() { return $this->mText
; }
2833 function getLanguageLinks() { return $this->mLanguageLinks
; }
2834 function getCategoryLinks() { return $this->mCategoryLinks
; }
2835 function getCacheTime() { return $this->mCacheTime
; }
2836 function containsOldMagic() { return $this->mContainsOldMagic
; }
2837 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2838 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2839 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2840 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2841 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
2843 function merge( $other ) {
2844 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2845 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2846 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2852 * Set options of the Parser
2854 * @package MediaWiki
2858 # All variables are private
2859 var $mUseTeX; # Use texvc to expand <math> tags
2860 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2861 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2862 var $mAllowExternalImages; # Allow external images inline
2863 var $mSkin; # Reference to the preferred skin
2864 var $mDateFormat; # Date format index
2865 var $mEditSection; # Create "edit section" links
2866 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2867 var $mNumberHeadings; # Automatically number headings
2868 var $mShowToc; # Show table of contents
2870 function getUseTeX() { return $this->mUseTeX
; }
2871 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2872 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2873 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2874 function getSkin() { return $this->mSkin
; }
2875 function getDateFormat() { return $this->mDateFormat
; }
2876 function getEditSection() { return $this->mEditSection
; }
2877 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
2878 function getNumberHeadings() { return $this->mNumberHeadings
; }
2879 function getShowToc() { return $this->mShowToc
; }
2881 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2882 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2883 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2884 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2885 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2886 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
2887 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
2888 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
2889 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
2891 function setSkin( &$x ) { $this->mSkin
=& $x; }
2893 # Get parser options
2894 /* static */ function newFromUser( &$user ) {
2895 $popts = new ParserOptions
;
2896 $popts->initialiseFromUser( $user );
2901 function initialiseFromUser( &$userInput ) {
2902 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2904 $fname = 'ParserOptions::initialiseFromUser';
2905 wfProfileIn( $fname );
2906 if ( !$userInput ) {
2908 $user->setLoaded( true );
2910 $user =& $userInput;
2913 $this->mUseTeX
= $wgUseTeX;
2914 $this->mUseDynamicDates
= $wgUseDynamicDates;
2915 $this->mInterwikiMagic
= $wgInterwikiMagic;
2916 $this->mAllowExternalImages
= $wgAllowExternalImages;
2917 wfProfileIn( $fname.'-skin' );
2918 $this->mSkin
=& $user->getSkin();
2919 wfProfileOut( $fname.'-skin' );
2920 $this->mDateFormat
= $user->getOption( 'date' );
2921 $this->mEditSection
= $user->getOption( 'editsection' );
2922 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
2923 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
2924 $this->mShowToc
= $user->getOption( 'showtoc' );
2925 wfProfileOut( $fname );
2931 # Regex callbacks, used in Parser::replaceVariables
2932 function wfBraceSubstitution( $matches ) {
2933 global $wgCurParser;
2934 return $wgCurParser->braceSubstitution( $matches );
2937 function wfArgSubstitution( $matches ) {
2938 global $wgCurParser;
2939 return $wgCurParser->argSubstitution( $matches );
2942 function wfVariableSubstitution( $matches ) {
2943 global $wgCurParser;
2944 return $wgCurParser->variableSubstitution( $matches );
2948 * Return the total number of articles
2950 function wfNumberOfArticles() {
2951 global $wgNumberOfArticles;
2954 return $wgNumberOfArticles;
2958 * Get various statistics from the database
2961 function wfLoadSiteStats() {
2962 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
2963 $fname = 'wfLoadSiteStats';
2965 if ( -1 != $wgNumberOfArticles ) return;
2966 $dbr =& wfGetDB( DB_SLAVE
);
2967 $s = $dbr->getArray( 'site_stats',
2968 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
2969 array( 'ss_row_id' => 1 ), $fname
2972 if ( $s === false ) {
2975 $wgTotalViews = $s->ss_total_views
;
2976 $wgTotalEdits = $s->ss_total_edits
;
2977 $wgNumberOfArticles = $s->ss_good_articles
;
2981 function wfEscapeHTMLTagsOnly( $in ) {
2983 array( '"', '>', '<' ),
2984 array( '"', '>', '<' ),