3 // require_once('Tokenizer.php');
6 * File for Parser and related classes
13 * Variable substitution O(N^2) attack
15 * Without countermeasures, it would be possible to attack the parser by saving
16 * a page filled with a large number of inclusions of large pages. The size of
17 * the generated page would be proportional to the square of the input size.
18 * Hence, we limit the number of inclusions of any given page, thus bringing any
19 * attack back to O(N).
21 define( 'MAX_INCLUDE_REPEAT', 100 );
22 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
24 # Allowed values for $mOutputType
25 define( 'OT_HTML', 1 );
26 define( 'OT_WIKI', 2 );
27 define( 'OT_MSG' , 3 );
29 # string parameter for extractTags which will cause it
30 # to strip HTML comments in addition to regular
31 # <XML>-style tags. This should not be anything we
32 # may want to use in wikisyntax
33 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
35 # prefix for escaping, used in two functions at least
36 define( 'UNIQ_PREFIX', 'NaodW29');
38 # Constants needed for external link processing
39 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
40 define( 'HTTP_PROTOCOLS', 'http|https' );
41 # Everything except bracket, space, or control characters
42 define( 'EXT_LINK_URL_CLASS', '[^]\\x00-\\x20\\x7F]' );
43 define( 'INVERSE_EXT_LINK_URL_CLASS', '[\]\\x00-\\x20\\x7F]' );
45 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
46 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
47 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
48 define( 'EXT_LINK_BRACKETED', '/\[(('.URL_PROTOCOLS
.'):'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
49 define( 'EXT_IMAGE_REGEX',
50 '/^('.HTTP_PROTOCOLS
.':)'. # Protocol
51 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
52 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
58 * Processes wiki markup
61 * There are three main entry points into the Parser class:
63 * produces HTML output
65 * produces altered wiki markup.
67 * performs brace substitution on MediaWiki messages
70 * objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
72 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
75 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
76 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
79 * * only within ParserOptions
92 # Cleared with clearState():
93 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
94 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
97 var $mOptions, $mTitle, $mOutputType,
98 $mTemplates, // cache of already loaded templates, avoids
99 // multiple SQL queries for the same string
100 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
101 // in this path. Used for loop detection.
111 $this->mTemplates
= array();
112 $this->mTemplatePath
= array();
113 $this->mTagHooks
= array();
122 function clearState() {
123 $this->mOutput
= new ParserOutput
;
124 $this->mAutonumber
= 0;
125 $this->mLastSection
= "";
126 $this->mDTopen
= false;
127 $this->mVariables
= false;
128 $this->mIncludeCount
= array();
129 $this->mStripState
= array();
130 $this->mArgStack
= array();
131 $this->mInPre
= false;
135 * First pass--just handle <nowiki> sections, pass the rest off
136 * to internalParse() which does all the real work.
139 * @return ParserOutput a ParserOutput
141 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
143 $fname = 'Parser::parse';
144 wfProfileIn( $fname );
150 $this->mOptions
= $options;
151 $this->mTitle
=& $title;
152 $this->mOutputType
= OT_HTML
;
155 $text = $this->strip( $text, $this->mStripState
);
156 $text = $this->internalParse( $text, $linestart );
157 $text = $this->unstrip( $text, $this->mStripState
);
158 # Clean up special characters, only run once, next-to-last before doBlockLevels
161 # french spaces, last one Guillemet-left
162 # only if there is something before the space
163 '/(.) (?=\\?|:|;|!|\\302\\273)/i' => '\\1 \\2',
164 # french spaces, Guillemet-right
165 "/(\\302\\253) /i"=>"\\1 ",
166 '/<hr *>/i' => '<hr />',
167 '/<br *>/i' => '<br />',
168 '/<center *>/i' => '<div class="center">',
169 '/<\\/center *>/i' => '</div>',
170 # Clean up spare ampersands; note that we probably ought to be
171 # more careful about named entities.
172 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
174 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
177 # french spaces, last one Guillemet-left
178 '/ (\\?|:|;|!|\\302\\273)/i' => ' \\1',
179 # french spaces, Guillemet-right
180 '/(\\302\\253) /i' => '\\1 ',
181 '/<center *>/i' => '<div class="center">',
182 '/<\\/center *>/i' => '</div>'
184 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
187 $text = $this->doBlockLevels( $text, $linestart );
188 $text = $this->unstripNoWiki( $text, $this->mStripState
);
190 $text = $this->tidy($text);
192 $this->mOutput
->setText( $text );
193 wfProfileOut( $fname );
194 return $this->mOutput
;
198 * Get a random string
203 function getRandomString() {
204 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
208 * Replaces all occurrences of <$tag>content</$tag> in the text
209 * with a random marker and returns the new text. the output parameter
210 * $content will be an associative array filled with data on the form
211 * $unique_marker => content.
213 * If $content is already set, the additional entries will be appended
214 * If $tag is set to STRIP_COMMENTS, the function will extract
215 * <!-- HTML comments -->
220 function extractTags($tag, $text, &$content, $uniq_prefix = ''){
221 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
228 while ( '' != $text ) {
229 if($tag==STRIP_COMMENTS
) {
230 $p = preg_split( '/<!--/i', $text, 2 );
232 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
235 if ( ( count( $p ) < 2 ) ||
( '' == $p[1] ) ) {
238 if($tag==STRIP_COMMENTS
) {
239 $q = preg_split( '/-->/i', $p[1], 2 );
241 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
243 $marker = $rnd . sprintf('%08X', $n++
);
244 $content[$marker] = $q[0];
245 $stripped .= $marker;
253 * Strips and renders nowiki, pre, math, hiero
254 * If $render is set, performs necessary rendering operations on plugins
255 * Returns the text, and fills an array with data needed in unstrip()
256 * If the $state is already a valid strip state, it adds to the state
258 * @param bool $stripcomments when set, HTML comments <!-- like this -->
259 * will be stripped in addition to other tags. This is important
260 * for section editing, where these comments cause confusion when
261 * counting the sections in the wikisource
265 function strip( $text, &$state, $stripcomments = false ) {
266 $render = ($this->mOutputType
== OT_HTML
);
267 $html_content = array();
268 $nowiki_content = array();
269 $math_content = array();
270 $pre_content = array();
271 $comment_content = array();
272 $ext_content = array();
274 # Replace any instances of the placeholders
275 $uniq_prefix = UNIQ_PREFIX
;
276 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
279 global $wgRawHtml, $wgWhitelistEdit;
280 if( $wgRawHtml && $wgWhitelistEdit ) {
281 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
282 foreach( $html_content as $marker => $content ) {
284 # Raw and unchecked for validity.
285 $html_content[$marker] = $content;
287 $html_content[$marker] = '<html>'.$content.'</html>';
293 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
294 foreach( $nowiki_content as $marker => $content ) {
296 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
298 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
303 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
304 foreach( $math_content as $marker => $content ){
306 if( $this->mOptions
->getUseTeX() ) {
307 $math_content[$marker] = renderMath( $content );
309 $math_content[$marker] = '<math>'.$content.'<math>';
312 $math_content[$marker] = '<math>'.$content.'</math>';
317 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
318 foreach( $pre_content as $marker => $content ){
320 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
322 $pre_content[$marker] = '<pre>'.$content.'</pre>';
328 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
329 foreach( $comment_content as $marker => $content ){
330 $comment_content[$marker] = '<!--'.$content.'-->';
335 foreach ( $this->mTagHooks
as $tag => $callback ) {
336 $ext_contents[$tag] = array();
337 $text = Parser
::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
338 foreach( $ext_content[$tag] as $marker => $content ) {
340 $ext_content[$tag][$marker] = $callback( $content );
342 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
347 # Merge state with the pre-existing state, if there is one
349 $state['html'] = $state['html'] +
$html_content;
350 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
351 $state['math'] = $state['math'] +
$math_content;
352 $state['pre'] = $state['pre'] +
$pre_content;
353 $state['comment'] = $state['comment'] +
$comment_content;
355 foreach( $ext_content as $tag => $array ) {
356 if ( array_key_exists( $tag, $state ) ) {
357 $state[$tag] = $state[$tag] +
$array;
362 'html' => $html_content,
363 'nowiki' => $nowiki_content,
364 'math' => $math_content,
365 'pre' => $pre_content,
366 'comment' => $comment_content,
373 * restores pre, math, and heiro removed by strip()
375 * always call unstripNoWiki() after this one
378 function unstrip( $text, &$state ) {
379 # Must expand in reverse order, otherwise nested tags will be corrupted
380 $contentDict = end( $state );
381 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
382 if( key($state) != 'nowiki' && key($state) != 'html') {
383 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
384 $text = str_replace( key( $contentDict ), $content, $text );
393 * always call this after unstrip() to preserve the order
397 function unstripNoWiki( $text, &$state ) {
398 # Must expand in reverse order, otherwise nested tags will be corrupted
399 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
400 $text = str_replace( key( $state['nowiki'] ), $content, $text );
405 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
406 $text = str_replace( key( $state['html'] ), $content, $text );
414 * Add an item to the strip state
415 * Returns the unique tag which must be inserted into the stripped text
416 * The tag will be replaced with the original text in unstrip()
420 function insertStripItem( $text, &$state ) {
421 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
430 $state['item'][$rnd] = $text;
435 * Return allowed HTML attributes
439 function getHTMLattrs () {
440 $htmlattrs = array( # Allowed attributes--no scripting, etc.
441 'title', 'align', 'lang', 'dir', 'width', 'height',
442 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
443 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
444 /* FONT */ 'type', 'start', 'value', 'compact',
445 /* For various lists, mostly deprecated but safe */
446 'summary', 'width', 'border', 'frame', 'rules',
447 'cellspacing', 'cellpadding', 'valign', 'char',
448 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
449 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
450 'id', 'class', 'name', 'style' /* For CSS */
456 * Remove non approved attributes and javascript in css
460 function fixTagAttributes ( $t ) {
461 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
462 $htmlattrs = $this->getHTMLattrs() ;
464 # Strip non-approved attributes from the tag
466 '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
467 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
470 $t = str_replace ( '<></>' , '' , $t ) ; # This should fix bug 980557
472 # Strip javascript "expression" from stylesheets. Brute force approach:
473 # If anythin offensive is found, all attributes of the HTML tag are dropped
476 '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
477 wfMungeToUtf8( $t ) ) )
486 * interface with html tidy, used if $wgUseTidy = true
490 function tidy ( $text ) {
491 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
492 global $wgInputEncoding, $wgOutputEncoding;
493 $fname = 'Parser::tidy';
494 wfProfileIn( $fname );
497 switch(strtoupper($wgOutputEncoding)) {
499 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)?
' -latin1':' -raw';
502 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)?
' -utf8':' -raw';
505 $wgTidyOpts .= ' -raw';
508 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
509 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
510 '<head><title>test</title></head><body>'.$text.'</body></html>';
511 $descriptorspec = array(
512 0 => array('pipe', 'r'),
513 1 => array('pipe', 'w'),
514 2 => array('file', '/dev/null', 'a')
516 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts", $descriptorspec, $pipes);
517 if (is_resource($process)) {
518 fwrite($pipes[0], $wrappedtext);
520 while (!feof($pipes[1])) {
521 $cleansource .= fgets($pipes[1], 1024);
524 $return_value = proc_close($process);
527 wfProfileOut( $fname );
529 if( $cleansource == '' && $text != '') {
530 wfDebug( "Tidy error detected!\n" );
531 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
538 * parse the wiki syntax used to render tables
542 function doTableStuff ( $t ) {
543 $fname = 'Parser::doTableStuff';
544 wfProfileIn( $fname );
546 $t = explode ( "\n" , $t ) ;
547 $td = array () ; # Is currently a td tag open?
548 $ltd = array () ; # Was it TD or TH?
549 $tr = array () ; # Is currently a tr tag open?
550 $ltr = array () ; # tr attributes
551 $indent_level = 0; # indent level of the table
552 foreach ( $t AS $k => $x )
555 $fc = substr ( $x , 0 , 1 ) ;
556 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
557 $indent_level = strlen( $matches[1] );
559 str_repeat( '<dl><dd>', $indent_level ) .
560 '<table ' . $this->fixTagAttributes ( $matches[2] ) . '>' ;
561 array_push ( $td , false ) ;
562 array_push ( $ltd , '' ) ;
563 array_push ( $tr , false ) ;
564 array_push ( $ltr , '' ) ;
566 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
567 else if ( '|}' == substr ( $x , 0 , 2 ) ) {
569 $l = array_pop ( $ltd ) ;
570 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
571 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
573 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
575 else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
576 $x = substr ( $x , 1 ) ;
577 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
579 $l = array_pop ( $ltd ) ;
580 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
581 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
584 array_push ( $tr , false ) ;
585 array_push ( $td , false ) ;
586 array_push ( $ltd , '' ) ;
587 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
589 else if ( '|' == $fc ||
'!' == $fc ||
'|+' == substr ( $x , 0 , 2 ) ) { # Caption
590 if ( '|+' == substr ( $x , 0 , 2 ) ) {
592 $x = substr ( $x , 1 ) ;
594 $after = substr ( $x , 1 ) ;
595 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
596 $after = explode ( '||' , $after ) ;
598 foreach ( $after AS $theline )
603 $tra = array_pop ( $ltr ) ;
604 if ( !array_pop ( $tr ) ) $z = '<tr '.$tra.">\n" ;
605 array_push ( $tr , true ) ;
606 array_push ( $ltr , '' ) ;
609 $l = array_pop ( $ltd ) ;
610 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
611 if ( $fc == '|' ) $l = 'td' ;
612 else if ( $fc == '!' ) $l = 'th' ;
613 else if ( $fc == '+' ) $l = 'caption' ;
615 array_push ( $ltd , $l ) ;
616 $y = explode ( '|' , $theline , 2 ) ;
617 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
618 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
620 array_push ( $td , true ) ;
625 # Closing open td, tr && table
626 while ( count ( $td ) > 0 )
628 if ( array_pop ( $td ) ) $t[] = '</td>' ;
629 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
633 $t = implode ( "\n" , $t ) ;
634 # $t = $this->removeHTMLtags( $t );
635 wfProfileOut( $fname );
640 * Helper function for parse() that transforms wiki markup into
641 * HTML. Only called for $mOutputType == OT_HTML.
645 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
648 $fname = 'Parser::internalParse';
649 wfProfileIn( $fname );
651 $text = $this->removeHTMLtags( $text );
652 $text = $this->replaceVariables( $text, $args );
654 $text = $wgLang->convert($text);
656 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
658 $text = $this->doHeadings( $text );
659 if($this->mOptions
->getUseDynamicDates()) {
660 global $wgDateFormatter;
661 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
663 $text = $this->doAllQuotes( $text );
664 $text = $this->replaceExternalLinks( $text );
665 $text = $this->doMagicLinks( $text );
666 $text = $this->replaceInternalLinks ( $text );
667 # Another call to replace links and images inside captions of images
668 $text = $this->replaceInternalLinks ( $text );
670 $text = $this->unstrip( $text, $this->mStripState
);
671 $text = $this->unstripNoWiki( $text, $this->mStripState
);
673 $text = $this->doTableStuff( $text );
674 $text = $this->formatHeadings( $text, $isMain );
675 $sk =& $this->mOptions
->getSkin();
676 $text = $sk->transformContent( $text );
678 wfProfileOut( $fname );
683 * Replace special strings like "ISBN xxx" and "RFC xxx" with
684 * magic external links.
688 function &doMagicLinks( &$text ) {
689 global $wgUseGeoMode;
690 $text = $this->magicISBN( $text );
691 if ( isset( $wgUseGeoMode ) && $wgUseGeoMode ) {
692 $text = $this->magicGEO( $text );
694 $text = $this->magicRFC( $text );
699 * Parse ^^ tokens and return html
703 function doExponent ( $text ) {
704 $fname = 'Parser::doExponent';
705 wfProfileIn( $fname);
706 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
707 wfProfileOut( $fname);
712 * Parse headers and return html
716 function doHeadings( $text ) {
717 $fname = 'Parser::doHeadings';
718 wfProfileIn( $fname );
719 for ( $i = 6; $i >= 1; --$i ) {
720 $h = substr( '======', 0, $i );
721 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
722 "<h{$i}>\\1</h{$i}>\\2", $text );
724 wfProfileOut( $fname );
729 * Replace single quotes with HTML markup
731 * @return string the altered text
733 function doAllQuotes( $text ) {
734 $fname = 'Parser::doAllQuotes';
735 wfProfileIn( $fname );
737 $lines = explode( "\n", $text );
738 foreach ( $lines as $line ) {
739 $outtext .= $this->doQuotes ( $line ) . "\n";
741 $outtext = substr($outtext, 0,-1);
742 wfProfileOut( $fname );
747 * Helper function for doAllQuotes()
750 function doQuotes( $text ) {
751 $arr = preg_split ("/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
752 if (count ($arr) == 1)
756 # First, do some preliminary work. This may shift some apostrophes from
757 # being mark-up to being text. It also counts the number of occurrences
758 # of bold and italics mark-ups.
766 # If there are ever four apostrophes, assume the first is supposed to
767 # be text, and the remaining three constitute mark-up for bold text.
768 if (strlen ($arr[$i]) == 4)
773 # If there are more than 5 apostrophes in a row, assume they're all
774 # text except for the last 5.
775 else if (strlen ($arr[$i]) > 5)
777 $arr[$i-1] .= str_repeat ("'", strlen ($arr[$i]) - 5);
780 # Count the number of occurrences of bold and italics mark-ups.
781 # We are not counting sequences of five apostrophes.
782 if (strlen ($arr[$i]) == 2) $numitalics++
; else
783 if (strlen ($arr[$i]) == 3) $numbold++
; else
784 if (strlen ($arr[$i]) == 5) { $numitalics++
; $numbold++
; }
789 # If there is an odd number of both bold and italics, it is likely
790 # that one of the bold ones was meant to be an apostrophe followed
791 # by italics. Which one we cannot know for certain, but it is more
792 # likely to be one that has a single-letter word before it.
793 if (($numbold %
2 == 1) && ($numitalics %
2 == 1))
796 $firstsingleletterword = -1;
797 $firstmultiletterword = -1;
801 if (($i %
2 == 1) and (strlen ($r) == 3))
803 $x1 = substr ($arr[$i-1], -1);
804 $x2 = substr ($arr[$i-1], -2, 1);
806 if ($firstspace == -1) $firstspace = $i;
807 } else if ($x2 == ' ') {
808 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
810 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
816 # If there is a single-letter word, use it!
817 if ($firstsingleletterword > -1)
819 $arr [ $firstsingleletterword ] = "''";
820 $arr [ $firstsingleletterword-1 ] .= "'";
822 # If not, but there's a multi-letter word, use that one.
823 else if ($firstmultiletterword > -1)
825 $arr [ $firstmultiletterword ] = "''";
826 $arr [ $firstmultiletterword-1 ] .= "'";
828 # ... otherwise use the first one that has neither.
829 # (notice that it is possible for all three to be -1 if, for example,
830 # there is only one pentuple-apostrophe in the line)
831 else if ($firstspace > -1)
833 $arr [ $firstspace ] = "''";
834 $arr [ $firstspace-1 ] .= "'";
838 # Now let's actually convert our apostrophic mush to HTML!
847 if ($state == 'both')
854 if (strlen ($r) == 2)
857 { $output .= '</i>'; $state = ''; }
858 else if ($state == 'bi')
859 { $output .= '</i>'; $state = 'b'; }
860 else if ($state == 'ib')
861 { $output .= '</b></i><b>'; $state = 'b'; }
862 else if ($state == 'both')
863 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
864 else # $state can be 'b' or ''
865 { $output .= '<i>'; $state .= 'i'; }
867 else if (strlen ($r) == 3)
870 { $output .= '</b>'; $state = ''; }
871 else if ($state == 'bi')
872 { $output .= '</i></b><i>'; $state = 'i'; }
873 else if ($state == 'ib')
874 { $output .= '</b>'; $state = 'i'; }
875 else if ($state == 'both')
876 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
877 else # $state can be 'i' or ''
878 { $output .= '<b>'; $state .= 'b'; }
880 else if (strlen ($r) == 5)
883 { $output .= '</b><i>'; $state = 'i'; }
884 else if ($state == 'i')
885 { $output .= '</i><b>'; $state = 'b'; }
886 else if ($state == 'bi')
887 { $output .= '</i></b>'; $state = ''; }
888 else if ($state == 'ib')
889 { $output .= '</b></i>'; $state = ''; }
890 else if ($state == 'both')
891 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
892 else # ($state == '')
893 { $buffer = ''; $state = 'both'; }
898 # Now close all remaining tags. Notice that the order is important.
899 if ($state == 'b' ||
$state == 'ib')
901 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
905 if ($state == 'both')
906 $output .= '<b><i>'.$buffer.'</i></b>';
912 * Replace external links
914 * Note: we have to do external links before the internal ones,
915 * and otherwise take great care in the order of things here, so
916 * that we don't end up interpreting some URLs twice.
920 function replaceExternalLinks( $text ) {
921 $fname = 'Parser::replaceExternalLinks';
922 wfProfileIn( $fname );
924 $sk =& $this->mOptions
->getSkin();
925 $linktrail = wfMsg('linktrail');
926 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
928 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
931 while ( $i<count( $bits ) ) {
933 $protocol = $bits[$i++
];
935 $trail = $bits[$i++
];
937 # If the link text is an image URL, replace it with an <img> tag
938 # This happened by accident in the original parser, but some people used it extensively
939 $img = $this->maybeMakeImageLink( $text );
940 if ( $img !== false ) {
946 # No link text, e.g. [http://domain.tld/some.link]
948 # Autonumber if allowed
949 if ( strpos( HTTP_PROTOCOLS
, $protocol ) !== false ) {
950 $text = '[' . ++
$this->mAutonumber
. ']';
952 # Otherwise just use the URL
953 $text = htmlspecialchars( $url );
956 # Have link text, e.g. [http://domain.tld/some.link text]s
958 if ( preg_match( $linktrail, $trail, $m2 ) ) {
964 $encUrl = htmlspecialchars( $url );
965 # Bit in parentheses showing the URL for the printable version
966 if( $url == $text ||
preg_match( "!$protocol://" . preg_quote( $text, '/' ) . "/?$!", $url ) ) {
969 # Expand the URL for printable version
970 if ( ! $sk->suppressUrlExpansion() ) {
971 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $encUrl ) . "</i>)</span>";
977 # Process the trail (i.e. everything after this link up until start of the next link),
978 # replacing any non-bracketed links
979 $trail = $this->replaceFreeExternalLinks( $trail );
981 $la = $sk->getExternalLinkAttributes( $url, $text );
983 # Use the encoded URL
984 # This means that users can paste URLs directly into the text
985 # Funny characters like ö aren't valid in URLs anyway
986 # This was changed in August 2004
987 $s .= "<a href=\"{$url}\" {$la}>{$text}</a>{$dtrail}{$paren}{$trail}";
990 wfProfileOut( $fname );
995 * Replace anything that looks like a URL with a link
998 function replaceFreeExternalLinks( $text ) {
999 $bits = preg_split( '/((?:'.URL_PROTOCOLS
.'):)/', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1000 $s = array_shift( $bits );
1003 $sk =& $this->mOptions
->getSkin();
1005 while ( $i < count( $bits ) ){
1006 $protocol = $bits[$i++
];
1007 $remainder = $bits[$i++
];
1009 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1010 # Found some characters after the protocol that look promising
1011 $url = $protocol . $m[1];
1014 # Move trailing punctuation to $trail
1016 # If there is no left bracket, then consider right brackets fair game too
1017 if ( strpos( $url, '(' ) === false ) {
1021 $numSepChars = strspn( strrev( $url ), $sep );
1022 if ( $numSepChars ) {
1023 $trail = substr( $url, -$numSepChars ) . $trail;
1024 $url = substr( $url, 0, -$numSepChars );
1027 # Replace & from obsolete syntax with &
1028 $url = str_replace( '&', '&', $url );
1030 # Is this an external image?
1031 $text = $this->maybeMakeImageLink( $url );
1032 if ( $text === false ) {
1033 # Not an image, make a link
1034 $text = $sk->makeExternalLink( $url, $url );
1036 $s .= $text . $trail;
1038 $s .= $protocol . $remainder;
1045 * make an image if it's allowed
1048 function maybeMakeImageLink( $url ) {
1049 $sk =& $this->mOptions
->getSkin();
1051 if ( $this->mOptions
->getAllowExternalImages() ) {
1052 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1054 $text = $sk->makeImage( htmlspecialchars( $url ) );
1061 * Process [[ ]] wikilinks
1065 function replaceInternalLinks( $s ) {
1066 global $wgLang, $wgLinkCache;
1067 global $wgNamespacesWithSubpages;
1068 static $fname = 'Parser::replaceInternalLinks' ;
1069 wfProfileIn( $fname );
1071 wfProfileIn( $fname.'-setup' );
1073 # the % is needed to support urlencoded titles as well
1074 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1075 $sk =& $this->mOptions
->getSkin();
1077 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1079 $a = explode( '[[', ' ' . $s );
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 the end of a line for a word that's not followed by whitespace,
1087 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1088 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1090 $useLinkPrefixExtension = $wgLang->linkPrefixExtension();
1091 # Special and Media are pseudo-namespaces; no pages actually exist in them
1093 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1095 if ( $useLinkPrefixExtension ) {
1096 if ( preg_match( $e2, $s, $m ) ) {
1097 $first_prefix = $m[2];
1100 $first_prefix = false;
1106 wfProfileOut( $fname.'-setup' );
1108 # start procedeeding each line
1109 foreach ( $a as $line ) {
1110 wfProfileIn( $fname.'-prefixhandling' );
1111 if ( $useLinkPrefixExtension ) {
1112 if ( preg_match( $e2, $s, $m ) ) {
1120 $prefix = $first_prefix;
1121 $first_prefix = false;
1124 wfProfileOut( $fname.'-prefixhandling' );
1126 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1128 # fix up urlencoded title texts
1129 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1131 } else { # Invalid form; output directly
1132 $s .= $prefix . '[[' . $line ;
1138 # :Foobar -- override special treatment of prefix (images, language links)
1139 # /Foobar -- convert to CurrentPage/Foobar
1140 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1142 # Look at the first character
1143 $c = substr($m[1],0,1);
1144 $noforce = ($c != ':');
1148 # / at end means we don't want the slash to be shown
1149 if(substr($m[1],-1,1)=='/') {
1150 $m[1]=substr($m[1],1,strlen($m[1])-2);
1153 $noslash=substr($m[1],1);
1156 # Some namespaces don't allow subpages
1157 if(!empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()])) {
1158 # subpages allowed here
1159 $link = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1162 } # this might be changed for ugliness reasons
1164 # no subpage allowed, use standard link
1168 } elseif( $noforce ) { # no subpage
1171 # We don't want to keep the first character
1172 $link = substr( $m[1], 1 );
1175 $wasblank = ( '' == $text );
1176 if( $wasblank ) $text = $link;
1178 $nt = Title
::newFromText( $link );
1180 $s .= $prefix . '[[' . $line;
1184 $ns = $nt->getNamespace();
1185 $iw = $nt->getInterWiki();
1187 # Link not escaped by : , create the various objects
1191 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
1192 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1193 $tmp = $prefix . $trail ;
1194 $s .= (trim($tmp) == '')?
'': $tmp;
1198 if ( $ns == NS_IMAGE
) {
1199 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1200 $wgLinkCache->addImageLinkObj( $nt );
1204 if ( $ns == NS_CATEGORY
) {
1205 $t = $nt->getText() ;
1206 $nnt = Title
::newFromText ( Namespace::getCanonicalName(NS_CATEGORY
).':'.$t ) ;
1208 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1209 $pPLC=$sk->postParseLinkColour();
1210 $sk->postParseLinkColour( false );
1211 $t = $sk->makeLinkObj( $nnt, $t, '', '' , $prefix );
1212 $sk->postParseLinkColour( $pPLC );
1213 $wgLinkCache->resume();
1216 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1217 $sortkey = $this->mTitle
->getText();
1219 $sortkey = $this->mTitle
->getPrefixedText();
1224 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1225 $this->mOutput
->mCategoryLinks
[] = $t ;
1226 $s .= $prefix . $trail ;
1231 if( ( $nt->getPrefixedText() === $this->mTitle
->getPrefixedText() ) &&
1232 ( strpos( $link, '#' ) === FALSE ) ) {
1233 # Self-links are handled specially; generally de-link and change to bold.
1234 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1238 if( $ns == NS_MEDIA
) {
1239 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1240 $wgLinkCache->addImageLinkObj( $nt );
1242 } elseif( $ns == NS_SPECIAL
) {
1243 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1246 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1248 wfProfileOut( $fname );
1253 * Used by doBlockLevels()
1256 /* private */ function closeParagraph() {
1258 if ( '' != $this->mLastSection
) {
1259 $result = '</' . $this->mLastSection
. ">\n";
1261 $this->mInPre
= false;
1262 $this->mLastSection
= '';
1265 # getCommon() returns the length of the longest common substring
1266 # of both arguments, starting at the beginning of both.
1268 /* private */ function getCommon( $st1, $st2 ) {
1269 $fl = strlen( $st1 );
1270 $shorter = strlen( $st2 );
1271 if ( $fl < $shorter ) { $shorter = $fl; }
1273 for ( $i = 0; $i < $shorter; ++
$i ) {
1274 if ( $st1{$i} != $st2{$i} ) { break; }
1278 # These next three functions open, continue, and close the list
1279 # element appropriate to the prefix character passed into them.
1281 /* private */ function openList( $char ) {
1282 $result = $this->closeParagraph();
1284 if ( '*' == $char ) { $result .= '<ul><li>'; }
1285 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1286 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1287 else if ( ';' == $char ) {
1288 $result .= '<dl><dt>';
1289 $this->mDTopen
= true;
1291 else { $result = '<!-- ERR 1 -->'; }
1296 /* private */ function nextItem( $char ) {
1297 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1298 else if ( ':' == $char ||
';' == $char ) {
1300 if ( $this->mDTopen
) { $close = '</dt>'; }
1301 if ( ';' == $char ) {
1302 $this->mDTopen
= true;
1303 return $close . '<dt>';
1305 $this->mDTopen
= false;
1306 return $close . '<dd>';
1309 return '<!-- ERR 2 -->';
1312 /* private */ function closeList( $char ) {
1313 if ( '*' == $char ) { $text = '</li></ul>'; }
1314 else if ( '#' == $char ) { $text = '</li></ol>'; }
1315 else if ( ':' == $char ) {
1316 if ( $this->mDTopen
) {
1317 $this->mDTopen
= false;
1318 $text = '</dt></dl>';
1320 $text = '</dd></dl>';
1323 else { return '<!-- ERR 3 -->'; }
1329 * Make lists from lines starting with ':', '*', '#', etc.
1332 * @return string the lists rendered as HTML
1334 function doBlockLevels( $text, $linestart ) {
1335 $fname = 'Parser::doBlockLevels';
1336 wfProfileIn( $fname );
1338 # Parsing through the text line by line. The main thing
1339 # happening here is handling of block-level elements p, pre,
1340 # and making lists from lines starting with * # : etc.
1342 $textLines = explode( "\n", $text );
1344 $lastPrefix = $output = $lastLine = '';
1345 $this->mDTopen
= $inBlockElem = false;
1347 $paragraphStack = false;
1349 if ( !$linestart ) {
1350 $output .= array_shift( $textLines );
1352 foreach ( $textLines as $oLine ) {
1353 $lastPrefixLength = strlen( $lastPrefix );
1354 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1355 $preOpenMatch = preg_match('/<pre/i', $oLine );
1356 if ( !$this->mInPre
) {
1357 # Multiple prefixes may abut each other for nested lists.
1358 $prefixLength = strspn( $oLine, '*#:;' );
1359 $pref = substr( $oLine, 0, $prefixLength );
1362 $pref2 = str_replace( ';', ':', $pref );
1363 $t = substr( $oLine, $prefixLength );
1364 $this->mInPre
= !empty($preOpenMatch);
1366 # Don't interpret any other prefixes in preformatted text
1368 $pref = $pref2 = '';
1373 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1374 # Same as the last item, so no need to deal with nesting or opening stuff
1375 $output .= $this->nextItem( substr( $pref, -1 ) );
1376 $paragraphStack = false;
1378 if ( substr( $pref, -1 ) == ';') {
1379 # The one nasty exception: definition lists work like this:
1380 # ; title : definition text
1381 # So we check for : in the remainder text to split up the
1382 # title and definition, without b0rking links.
1383 # FIXME: This is not foolproof. Something better in Tokenizer might help.
1384 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1386 $output .= $term . $this->nextItem( ':' );
1390 } elseif( $prefixLength ||
$lastPrefixLength ) {
1391 # Either open or close a level...
1392 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1393 $paragraphStack = false;
1395 while( $commonPrefixLength < $lastPrefixLength ) {
1396 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1397 --$lastPrefixLength;
1399 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1400 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1402 while ( $prefixLength > $commonPrefixLength ) {
1403 $char = substr( $pref, $commonPrefixLength, 1 );
1404 $output .= $this->openList( $char );
1406 if ( ';' == $char ) {
1407 # FIXME: This is dupe of code above
1408 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1410 $output .= $term . $this->nextItem( ':' );
1414 ++
$commonPrefixLength;
1416 $lastPrefix = $pref2;
1418 if( 0 == $prefixLength ) {
1419 # No prefix (not in list)--go to paragraph mode
1420 $uniq_prefix = UNIQ_PREFIX
;
1421 // XXX: use a stack for nestable elements like span, table and div
1422 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1423 $closematch = preg_match(
1424 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1425 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1426 if ( $openmatch or $closematch ) {
1427 $paragraphStack = false;
1428 $output .= $this->closeParagraph();
1429 if($preOpenMatch and !$preCloseMatch) {
1430 $this->mInPre
= true;
1432 if ( $closematch ) {
1433 $inBlockElem = false;
1435 $inBlockElem = true;
1437 } else if ( !$inBlockElem && !$this->mInPre
) {
1438 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1440 if ($this->mLastSection
!= 'pre') {
1441 $paragraphStack = false;
1442 $output .= $this->closeParagraph().'<pre>';
1443 $this->mLastSection
= 'pre';
1445 $t = substr( $t, 1 );
1448 if ( '' == trim($t) ) {
1449 if ( $paragraphStack ) {
1450 $output .= $paragraphStack.'<br />';
1451 $paragraphStack = false;
1452 $this->mLastSection
= 'p';
1454 if ($this->mLastSection
!= 'p' ) {
1455 $output .= $this->closeParagraph();
1456 $this->mLastSection
= '';
1457 $paragraphStack = '<p>';
1459 $paragraphStack = '</p><p>';
1463 if ( $paragraphStack ) {
1464 $output .= $paragraphStack;
1465 $paragraphStack = false;
1466 $this->mLastSection
= 'p';
1467 } else if ($this->mLastSection
!= 'p') {
1468 $output .= $this->closeParagraph().'<p>';
1469 $this->mLastSection
= 'p';
1475 if ($paragraphStack === false) {
1479 while ( $prefixLength ) {
1480 $output .= $this->closeList( $pref2{$prefixLength-1} );
1483 if ( '' != $this->mLastSection
) {
1484 $output .= '</' . $this->mLastSection
. '>';
1485 $this->mLastSection
= '';
1488 wfProfileOut( $fname );
1493 * Return value of a magic variable (like PAGENAME)
1497 function getVariableValue( $index ) {
1498 global $wgLang, $wgSitename, $wgServer;
1501 case MAG_CURRENTMONTH
:
1502 return $wgLang->formatNum( date( 'm' ) );
1503 case MAG_CURRENTMONTHNAME
:
1504 return $wgLang->getMonthName( date('n') );
1505 case MAG_CURRENTMONTHNAMEGEN
:
1506 return $wgLang->getMonthNameGen( date('n') );
1507 case MAG_CURRENTDAY
:
1508 return $wgLang->formatNum( date('j') );
1510 return $this->mTitle
->getText();
1512 return $this->mTitle
->getPartialURL();
1514 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1515 return $wgLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1516 case MAG_CURRENTDAYNAME
:
1517 return $wgLang->getWeekdayName( date('w')+
1 );
1518 case MAG_CURRENTYEAR
:
1519 return $wgLang->formatNum( date( 'Y' ) );
1520 case MAG_CURRENTTIME
:
1521 return $wgLang->time( wfTimestampNow(), false );
1522 case MAG_NUMBEROFARTICLES
:
1523 return $wgLang->formatNum( wfNumberOfArticles() );
1534 * initialise the magic variables (like CURRENTMONTHNAME)
1538 function initialiseVariables() {
1539 global $wgVariableIDs;
1540 $this->mVariables
= array();
1541 foreach ( $wgVariableIDs as $id ) {
1542 $mw =& MagicWord
::get( $id );
1543 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1548 * Replace magic variables, templates, and template arguments
1549 * with the appropriate text. Templates are substituted recursively,
1550 * taking care to avoid infinite loops.
1552 * Note that the substitution depends on value of $mOutputType:
1553 * OT_WIKI: only {{subst:}} templates
1554 * OT_MSG: only magic variables
1555 * OT_HTML: all templates and magic variables
1557 * @param string $tex The text to transform
1558 * @param array $args Key-value pairs representing template parameters to substitute
1561 function replaceVariables( $text, $args = array() ) {
1562 global $wgLang, $wgScript, $wgArticlePath;
1564 # Prevent too big inclusions
1565 if(strlen($text)> MAX_INCLUDE_SIZE
)
1568 $fname = 'Parser::replaceVariables';
1569 wfProfileIn( $fname );
1571 $titleChars = Title
::legalChars();
1573 # This function is called recursively. To keep track of arguments we need a stack:
1574 array_push( $this->mArgStack
, $args );
1576 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1577 $GLOBALS['wgCurParser'] =& $this;
1579 # Variable substitution
1580 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", 'wfVariableSubstitution', $text );
1582 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1583 # Argument substitution
1584 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1586 # Template substitution
1587 $regex = '/{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1588 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1590 array_pop( $this->mArgStack
);
1592 wfProfileOut( $fname );
1597 * Replace magic variables
1600 function variableSubstitution( $matches ) {
1601 if ( !$this->mVariables
) {
1602 $this->initialiseVariables();
1605 if ( $this->mOutputType
== OT_WIKI
) {
1606 # Do only magic variables prefixed by SUBST
1607 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1608 if (!$mwSubst->matchStartAndRemove( $matches[1] ))
1610 # Note that if we don't substitute the variable below,
1611 # we don't remove the {{subst:}} magic word, in case
1612 # it is a template rather than a magic variable.
1614 if ( !$skip && array_key_exists( $matches[1], $this->mVariables
) ) {
1615 $text = $this->mVariables
[$matches[1]];
1616 $this->mOutput
->mContainsOldMagic
= true;
1618 $text = $matches[0];
1623 # Split template arguments
1624 function getTemplateArgs( $argsString ) {
1625 if ( $argsString === '' ) {
1629 $args = explode( '|', substr( $argsString, 1 ) );
1631 # If any of the arguments contains a '[[' but no ']]', it needs to be
1632 # merged with the next arg because the '|' character between belongs
1633 # to the link syntax and not the template parameter syntax.
1634 $argc = count($args);
1636 for ( $i = 0; $i < $argc-1; $i++
) {
1637 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1638 $args[$i] .= '|'.$args[$i+
1];
1639 array_splice($args, $i+
1, 1);
1649 * Return the text of a template, after recursively
1650 * replacing any variables or templates within the template.
1652 * @param array $matches The parts of the template
1653 * $matches[1]: the title, i.e. the part before the |
1654 * $matches[2]: the parameters (including a leading |), if any
1655 * @return string the text of the template
1658 function braceSubstitution( $matches ) {
1659 global $wgLinkCache, $wgLang;
1660 $fname = 'Parser::braceSubstitution';
1664 $itcamefromthedatabase = false;
1668 # $part1 is the bit before the first |, and must contain only title characters
1669 # $args is a list of arguments, starting from index 0, not including $part1
1671 $part1 = $matches[1];
1672 # If the second subpattern matched anything, it will start with |
1674 $args = $this->getTemplateArgs($matches[2]);
1675 $argc = count( $args );
1678 if ( strpos( $matches[0], '{{{' ) !== false ) {
1679 $text = $matches[0];
1686 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1687 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
1688 # One of two possibilities is true:
1689 # 1) Found SUBST but not in the PST phase
1690 # 2) Didn't find SUBST and in the PST phase
1691 # In either case, return without further processing
1692 $text = $matches[0];
1698 # MSG, MSGNW and INT
1701 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1702 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1705 # Remove obsolete MSG:
1706 $mwMsg =& MagicWord
::get( MAG_MSG
);
1707 $mwMsg->matchStartAndRemove( $part1 );
1710 # Check if it is an internal message
1711 $mwInt =& MagicWord
::get( MAG_INT
);
1712 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1713 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1714 $text = wfMsgReal( $part1, $args, true );
1722 # Check for NS: (namespace expansion)
1723 $mwNs = MagicWord
::get( MAG_NS
);
1724 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1725 if ( intval( $part1 ) ) {
1726 $text = $wgLang->getNsText( intval( $part1 ) );
1729 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1730 if ( !is_null( $index ) ) {
1731 $text = $wgLang->getNsText( $index );
1738 # LOCALURL and LOCALURLE
1740 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1741 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1743 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1744 $func = 'getLocalURL';
1745 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1746 $func = 'escapeLocalURL';
1751 if ( $func !== '' ) {
1752 $title = Title
::newFromText( $part1 );
1753 if ( !is_null( $title ) ) {
1755 $text = $title->$func( $args[0] );
1757 $text = $title->$func();
1765 if ( !$found && $argc == 1 ) {
1766 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
1767 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1768 $text = $wgLang->convertGrammar( $args[0], $part1 );
1773 # Template table test
1775 # Did we encounter this template already? If yes, it is in the cache
1776 # and we need to check for loops.
1777 if ( isset( $this->mTemplates
[$part1] ) ) {
1778 # Infinite loop test
1779 if ( isset( $this->mTemplatePath
[$part1] ) ) {
1783 # set $text to cached message.
1784 $text = $this->mTemplates
[$part1];
1788 # Load from database
1790 $title = Title
::newFromText( $part1, NS_TEMPLATE
);
1791 if ( !is_null( $title ) && !$title->isExternal() ) {
1792 # Check for excessive inclusion
1793 $dbk = $title->getPrefixedDBkey();
1794 if ( $this->incrementIncludeCount( $dbk ) ) {
1795 # This should never be reached.
1796 $article = new Article( $title );
1797 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1798 if ( $articleContent !== false ) {
1800 $text = $articleContent;
1801 $itcamefromthedatabase = true;
1805 # If the title is valid but undisplayable, make a link to it
1806 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1807 $text = '[['.$title->getPrefixedText().']]';
1811 # Template cache array insertion
1812 $this->mTemplates
[$part1] = $text;
1816 # Recursive parsing, escaping and link table handling
1817 # Only for HTML output
1818 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1819 $text = wfEscapeWikiText( $text );
1820 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
1821 # Clean up argument array
1822 $assocArgs = array();
1824 foreach( $args as $arg ) {
1825 $eqpos = strpos( $arg, '=' );
1826 if ( $eqpos === false ) {
1827 $assocArgs[$index++
] = $arg;
1829 $name = trim( substr( $arg, 0, $eqpos ) );
1830 $value = trim( substr( $arg, $eqpos+
1 ) );
1831 if ( $value === false ) {
1834 if ( $name !== false ) {
1835 $assocArgs[$name] = $value;
1840 # Add a new element to the templace recursion path
1841 $this->mTemplatePath
[$part1] = 1;
1843 $text = $this->strip( $text, $this->mStripState
);
1844 $text = $this->removeHTMLtags( $text );
1845 $text = $this->replaceVariables( $text, $assocArgs );
1847 # Resume the link cache and register the inclusion as a link
1848 if ( $mOutputType == OT_HTML
&& !is_null( $title ) ) {
1849 $wgLinkCache->addLinkObj( $title );
1853 # Empties the template path
1854 $this->mTemplatePath
= array();
1858 # replace ==section headers==
1859 # XXX this needs to go away once we have a better parser.
1860 if ( $this->mOutputType
!= OT_WIKI
&& $itcamefromthedatabase ) {
1861 if( !is_null( $title ) )
1862 $encodedname = base64_encode($title->getPrefixedDBkey());
1864 $encodedname = base64_encode("");
1865 $matches = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
1866 PREG_SPLIT_DELIM_CAPTURE
);
1869 for( $i = 0; $i < count($matches); $i +
= 2 ) {
1870 $text .= $matches[$i];
1871 if (!isset($matches[$i +
1]) ||
$matches[$i +
1] == "") continue;
1872 $hl = $matches[$i +
1];
1873 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
1877 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
1878 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
1879 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
1889 * Triple brace replacement -- used for template arguments
1892 function argSubstitution( $matches ) {
1893 $arg = trim( $matches[1] );
1894 $text = $matches[0];
1895 $inputArgs = end( $this->mArgStack
);
1897 if ( array_key_exists( $arg, $inputArgs ) ) {
1898 $text = $inputArgs[$arg];
1905 * Returns true if the function is allowed to include this entity
1908 function incrementIncludeCount( $dbk ) {
1909 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1910 $this->mIncludeCount
[$dbk] = 0;
1912 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1921 * Cleans up HTML, removes dangerous tags and attributes, and
1922 * removes HTML comments
1925 function removeHTMLtags( $text ) {
1926 global $wgUseTidy, $wgUserHtml;
1927 $fname = 'Parser::removeHTMLtags';
1928 wfProfileIn( $fname );
1931 $htmlpairs = array( # Tags that must be closed
1932 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
1933 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
1934 'strike', 'strong', 'tt', 'var', 'div', 'center',
1935 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
1936 'ruby', 'rt' , 'rb' , 'rp', 'p'
1938 $htmlsingle = array(
1939 'br', 'hr', 'li', 'dt', 'dd'
1941 $htmlnest = array( # Tags that can be nested--??
1942 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
1943 'dl', 'font', 'big', 'small', 'sub', 'sup'
1945 $tabletags = array( # Can only appear inside table
1949 $htmlpairs = array();
1950 $htmlsingle = array();
1951 $htmlnest = array();
1952 $tabletags = array();
1955 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1956 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1958 $htmlattrs = $this->getHTMLattrs () ;
1960 # Remove HTML comments
1961 $text = $this->removeHTMLcomments( $text );
1963 $bits = explode( '<', $text );
1964 $text = array_shift( $bits );
1966 $tagstack = array(); $tablestack = array();
1967 foreach ( $bits as $x ) {
1968 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
1969 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1971 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1972 error_reporting( $prev );
1975 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1979 if ( ! in_array( $t, $htmlsingle ) &&
1980 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
1981 @array_push
( $tagstack, $ot );
1984 if ( $t == 'table' ) {
1985 $tagstack = array_pop( $tablestack );
1990 # Keep track for later
1991 if ( in_array( $t, $tabletags ) &&
1992 ! in_array( 'table', $tagstack ) ) {
1994 } else if ( in_array( $t, $tagstack ) &&
1995 ! in_array ( $t , $htmlnest ) ) {
1997 } else if ( ! in_array( $t, $htmlsingle ) ) {
1998 if ( $t == 'table' ) {
1999 array_push( $tablestack, $tagstack );
2000 $tagstack = array();
2002 array_push( $tagstack, $t );
2004 # Strip non-approved attributes from the tag
2005 $newparams = $this->fixTagAttributes($params);
2009 $rest = str_replace( '>', '>', $rest );
2010 $text .= "<$slash$t $newparams$brace$rest";
2014 $text .= '<' . str_replace( '>', '>', $x);
2016 # Close off any remaining tags
2017 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2019 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2022 # this might be possible using tidy itself
2023 foreach ( $bits as $x ) {
2024 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2026 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2027 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2028 $newparams = $this->fixTagAttributes($params);
2029 $rest = str_replace( '>', '>', $rest );
2030 $text .= "<$slash$t $newparams$brace$rest";
2032 $text .= '<' . str_replace( '>', '>', $x);
2036 wfProfileOut( $fname );
2041 * Remove '<!--', '-->', and everything between.
2042 * To avoid leaving blank lines, when a comment is both preceded
2043 * and followed by a newline (ignoring spaces), trim leading and
2044 * trailing spaces and one of the newlines.
2048 function removeHTMLcomments( $text ) {
2049 $fname='Parser::removeHTMLcomments';
2050 wfProfileIn( $fname );
2051 while (($start = strpos($text, '<!--')) !== false) {
2052 $end = strpos($text, '-->', $start +
4);
2053 if ($end === false) {
2054 # Unterminated comment; bail out
2060 # Trim space and newline if the comment is both
2061 # preceded and followed by a newline
2062 $spaceStart = max($start - 1, 0);
2063 $spaceLen = $end - $spaceStart;
2064 while (substr($text, $spaceStart, 1) === ' ' && $spaceStart > 0) {
2068 while (substr($text, $spaceStart +
$spaceLen, 1) === ' ')
2070 if (substr($text, $spaceStart, 1) === "\n" and substr($text, $spaceStart +
$spaceLen, 1) === "\n") {
2071 # Remove the comment, leading and trailing
2072 # spaces, and leave only one newline.
2073 $text = substr_replace($text, "\n", $spaceStart, $spaceLen +
1);
2076 # Remove just the comment.
2077 $text = substr_replace($text, '', $start, $end - $start);
2080 wfProfileOut( $fname );
2085 * This function accomplishes several tasks:
2086 * 1) Auto-number headings if that option is enabled
2087 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2088 * 3) Add a Table of contents on the top for users who have enabled the option
2089 * 4) Auto-anchor headings
2091 * It loops through all headlines, collects the necessary data, then splits up the
2092 * string and re-inserts the newly formatted headlines.
2095 /* private */ function formatHeadings( $text, $isMain=true ) {
2096 global $wgInputEncoding, $wgMaxTocLevel, $wgLang, $wgLinkHolders;
2098 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2099 $doShowToc = $this->mOptions
->getShowToc();
2100 $forceTocHere = false;
2101 if( !$this->mTitle
->userCanEdit() ) {
2103 $rightClickHack = 0;
2105 $showEditLink = $this->mOptions
->getEditSection();
2106 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
2109 # Inhibit editsection links if requested in the page
2110 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2111 if( $esw->matchAndRemove( $text ) ) {
2114 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2116 $mw =& MagicWord
::get( MAG_NOTOC
);
2117 if( $mw->matchAndRemove( $text ) ) {
2121 # never add the TOC to the Main Page. This is an entry page that should not
2122 # be more than 1-2 screens large anyway
2123 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
2127 # Get all headlines for numbering them and adding funky stuff like [edit]
2128 # links - this is for later, but we need the number of headlines right now
2129 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
2131 # if there are fewer than 4 headlines in the article, do not show TOC
2132 if( $numMatches < 4 ) {
2136 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2137 # override above conditions and always show TOC at that place
2138 $mw =& MagicWord
::get( MAG_TOC
);
2139 if ($mw->match( $text ) ) {
2141 $forceTocHere = true;
2143 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2144 # override above conditions and always show TOC above first header
2145 $mw =& MagicWord
::get( MAG_FORCETOC
);
2146 if ($mw->matchAndRemove( $text ) ) {
2153 # We need this to perform operations on the HTML
2154 $sk =& $this->mOptions
->getSkin();
2158 $sectionCount = 0; # headlineCount excluding template sections
2160 # Ugh .. the TOC should have neat indentation levels which can be
2161 # passed to the skin functions. These are determined here
2166 $sublevelCount = array();
2169 foreach( $matches[3] as $headline ) {
2171 $templatetitle = "";
2172 $templatesection = 0;
2174 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2176 $templatetitle = base64_decode($mat[1]);
2177 $templatesection = 1 +
(int)base64_decode($mat[2]);
2178 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2183 $prevlevel = $level;
2185 $level = $matches[1][$headlineCount];
2186 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2187 # reset when we enter a new level
2188 $sublevelCount[$level] = 0;
2189 $toc .= $sk->tocIndent( $level - $prevlevel );
2190 $toclevel +
= $level - $prevlevel;
2192 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2193 # reset when we step back a level
2194 $sublevelCount[$level+
1]=0;
2195 $toc .= $sk->tocUnindent( $prevlevel - $level );
2196 $toclevel -= $prevlevel - $level;
2198 # count number of headlines for each level
2199 @$sublevelCount[$level]++
;
2200 if( $doNumberHeadings ||
$doShowToc ) {
2202 for( $i = 1; $i <= $level; $i++
) {
2203 if( !empty( $sublevelCount[$i] ) ) {
2207 $numbering .= $wgLang->formatNum( $sublevelCount[$i] );
2213 # The canonized header is a version of the header text safe to use for links
2214 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2215 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2216 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2218 # Remove link placeholders by the link text.
2219 # <!--LINK number-->
2221 # link text with suffix
2222 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2223 "\$wgLinkHolders['texts'][\$1]",
2224 $canonized_headline );
2227 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2228 $tocline = trim( $canonized_headline );
2229 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2230 $replacearray = array(
2234 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2235 $refer[$headlineCount] = $canonized_headline;
2237 # count how many in assoc. array so we can track dupes in anchors
2238 @$refers[$canonized_headline]++
;
2239 $refcount[$headlineCount]=$refers[$canonized_headline];
2241 # Prepend the number to the heading text
2243 if( $doNumberHeadings ||
$doShowToc ) {
2244 $tocline = $numbering . ' ' . $tocline;
2246 # Don't number the heading if it is the only one (looks silly)
2247 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2248 # the two are different if the line contains a link
2249 $headline=$numbering . ' ' . $headline;
2253 # Create the anchor for linking from the TOC to the section
2254 $anchor = $canonized_headline;
2255 if($refcount[$headlineCount] > 1 ) {
2256 $anchor .= '_' . $refcount[$headlineCount];
2258 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2259 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2261 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2262 if ( empty( $head[$headlineCount] ) ) {
2263 $head[$headlineCount] = '';
2266 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2268 $head[$headlineCount] .= $sk->editSectionLink($sectionCount+
1);
2271 # Add the edit section span
2272 if( $rightClickHack ) {
2274 $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
2276 $headline = $sk->editSectionScript($sectionCount+
1,$headline);
2279 # give headline the correct <h#> tag
2280 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2288 $toclines = $headlineCount;
2289 $toc .= $sk->tocUnindent( $toclevel );
2290 $toc = $sk->tocTable( $toc );
2293 # split up and insert constructed headlines
2295 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2298 foreach( $blocks as $block ) {
2299 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2300 # This is the [edit] link that appears for the top block of text when
2301 # section editing is enabled
2303 # Disabled because it broke block formatting
2304 # For example, a bullet point in the top line
2305 # $full .= $sk->editSectionLink(0);
2308 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2309 # Top anchor now in skin
2313 if( !empty( $head[$i] ) ) {
2319 $mw =& MagicWord
::get( MAG_TOC
);
2320 return $mw->replace( $toc, $full );
2327 * Return an HTML link for the "ISBN 123456" text
2330 function magicISBN( $text ) {
2332 $fname = 'Parser::magicISBN';
2333 wfProfileIn( $fname );
2335 $a = split( 'ISBN ', ' '.$text );
2336 if ( count ( $a ) < 2 ) {
2337 wfProfileOut( $fname );
2340 $text = substr( array_shift( $a ), 1);
2341 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2343 foreach ( $a as $x ) {
2344 $isbn = $blank = '' ;
2345 while ( ' ' == $x{0} ) {
2347 $x = substr( $x, 1 );
2349 if ( $x == '' ) { # blank isbn
2350 $text .= "ISBN $blank";
2353 while ( strstr( $valid, $x{0} ) != false ) {
2355 $x = substr( $x, 1 );
2357 $num = str_replace( '-', '', $isbn );
2358 $num = str_replace( ' ', '', $num );
2361 $text .= "ISBN $blank$x";
2363 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2364 $text .= '<a href="' .
2365 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2366 "\" class=\"internal\">ISBN $isbn</a>";
2370 wfProfileOut( $fname );
2375 * Return an HTML link for the "GEO ..." text
2378 function magicGEO( $text ) {
2379 global $wgLang, $wgUseGeoMode;
2380 $fname = 'Parser::magicGEO';
2381 wfProfileIn( $fname );
2383 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2384 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2385 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2386 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2387 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2388 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2390 $a = split( 'GEO ', ' '.$text );
2391 if ( count ( $a ) < 2 ) {
2392 wfProfileOut( $fname );
2395 $text = substr( array_shift( $a ), 1);
2396 $valid = '0123456789.+-:';
2398 foreach ( $a as $x ) {
2399 $geo = $blank = '' ;
2400 while ( ' ' == $x{0} ) {
2402 $x = substr( $x, 1 );
2404 while ( strstr( $valid, $x{0} ) != false ) {
2406 $x = substr( $x, 1 );
2408 $num = str_replace( '+', '', $geo );
2409 $num = str_replace( ' ', '', $num );
2411 if ( '' == $num ||
count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2412 $text .= "GEO $blank$x";
2414 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2415 $text .= '<a href="' .
2416 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2417 "\" class=\"internal\">GEO $geo</a>";
2421 wfProfileOut( $fname );
2426 * Return an HTML link for the "RFC 1234" text
2428 * @param string $text text to be processed
2430 function magicRFC( $text ) {
2433 $valid = '0123456789';
2436 $a = split( 'RFC ', ' '.$text );
2437 if ( count ( $a ) < 2 ) return $text;
2438 $text = substr( array_shift( $a ), 1);
2440 /* Check if RFC keyword is preceed by [[.
2441 * This test is made here cause of the array_shift above
2442 * that prevent the test to be done in the foreach.
2444 if(substr($text, -2) == '[[') { $internal = true; }
2446 foreach ( $a as $x ) {
2447 /* token might be empty if we have RFC RFC 1234 */
2453 $rfc = $blank = '' ;
2455 /** remove and save whitespaces in $blank */
2456 while ( $x{0} == ' ' ) {
2458 $x = substr( $x, 1 );
2461 /** remove and save the rfc number in $rfc */
2462 while ( strstr( $valid, $x{0} ) != false ) {
2464 $x = substr( $x, 1 );
2468 /* call back stripped spaces*/
2469 $text .= "RFC $blank$x";
2470 } elseif( $internal) {
2472 $text .= "RFC $rfc$x";
2474 /* build the external link*/
2475 $url = wfmsg( 'rfcurl' );
2476 $url = str_replace( '$1', $rfc, $url);
2477 $sk =& $this->mOptions
->getSkin();
2478 $la = $sk->getExternalLinkAttributes( $url, 'RFC '.$rfc );
2479 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2482 /* Check if the next RFC keyword is preceed by [[ */
2483 $internal = (substr($x,-2) == '[[');
2489 * Transform wiki markup when saving a page by doing \r\n -> \n
2490 * conversion, substitting signatures, {{subst:}} templates, etc.
2492 * @param string $text the text to transform
2493 * @param Title &$title the Title object for the current article
2494 * @param User &$user the User object describing the current user
2495 * @param ParserOptions $options parsing options
2496 * @param bool $clearState whether to clear the parser state first
2497 * @return string the altered wiki markup
2500 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2501 $this->mOptions
= $options;
2502 $this->mTitle
=& $title;
2503 $this->mOutputType
= OT_WIKI
;
2505 if ( $clearState ) {
2506 $this->clearState();
2509 $stripState = false;
2513 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2517 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2518 "/<br *?>/i" => "<br />",
2520 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2522 $text = $this->strip( $text, $stripState, false );
2523 $text = $this->pstPass2( $text, $user );
2524 $text = $this->unstrip( $text, $stripState );
2525 $text = $this->unstripNoWiki( $text, $stripState );
2530 * Pre-save transform helper function
2533 function pstPass2( $text, &$user ) {
2534 global $wgLang, $wgLocaltimezone, $wgCurParser;
2536 # Variable replacement
2537 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2538 $text = $this->replaceVariables( $text );
2542 $n = $user->getName();
2543 $k = $user->getOption( 'nickname' );
2544 if ( '' == $k ) { $k = $n; }
2545 if(isset($wgLocaltimezone)) {
2546 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2548 /* Note: this is an ugly timezone hack for the European wikis */
2549 $d = $wgLang->timeanddate( date( 'YmdHis' ), false ) .
2550 ' (' . date( 'T' ) . ')';
2551 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2553 $text = preg_replace( '/~~~~~/', $d, $text );
2554 $text = preg_replace( '/~~~~/', '[[' . $wgLang->getNsText( NS_USER
) . ":$n|$k]] $d", $text );
2555 $text = preg_replace( '/~~~/', '[[' . $wgLang->getNsText( NS_USER
) . ":$n|$k]]", $text );
2557 # Context links: [[|name]] and [[name (context)|]]
2559 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2560 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2561 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2562 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2564 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2565 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2566 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2567 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2569 $t = $this->mTitle
->getText();
2570 if ( preg_match( $conpat, $t, $m ) ) {
2573 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2574 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2575 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2577 if ( '' == $context ) {
2578 $text = preg_replace( $p2, '[[\\1]]', $text );
2580 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2583 # Trim trailing whitespace
2584 # MAG_END (__END__) tag allows for trailing
2585 # whitespace to be deliberately included
2586 $text = rtrim( $text );
2587 $mw =& MagicWord
::get( MAG_END
);
2588 $mw->matchAndRemove( $text );
2594 * Set up some variables which are usually set up in parse()
2595 * so that an external function can call some class members with confidence
2598 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2599 $this->mTitle
=& $title;
2600 $this->mOptions
= $options;
2601 $this->mOutputType
= $outputType;
2602 if ( $clearState ) {
2603 $this->clearState();
2608 * Transform a MediaWiki message by replacing magic variables.
2610 * @param string $text the text to transform
2611 * @param ParserOptions $options options
2612 * @return string the text with variables substituted
2615 function transformMsg( $text, $options ) {
2617 static $executing = false;
2619 # Guard against infinite recursion
2625 $this->mTitle
= $wgTitle;
2626 $this->mOptions
= $options;
2627 $this->mOutputType
= OT_MSG
;
2628 $this->clearState();
2629 $text = $this->replaceVariables( $text );
2636 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2637 * Callback will be called with the text within
2638 * Transform and return the text within
2641 function setHook( $tag, $callback ) {
2642 $oldVal = @$this->mTagHooks
[$tag];
2643 $this->mTagHooks
[$tag] = $callback;
2650 * @package MediaWiki
2654 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2655 var $mCacheTime; # Used in ParserCache
2657 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
2658 $containsOldMagic = false )
2660 $this->mText
= $text;
2661 $this->mLanguageLinks
= $languageLinks;
2662 $this->mCategoryLinks
= $categoryLinks;
2663 $this->mContainsOldMagic
= $containsOldMagic;
2664 $this->mCacheTime
= '';
2667 function getText() { return $this->mText
; }
2668 function getLanguageLinks() { return $this->mLanguageLinks
; }
2669 function getCategoryLinks() { return $this->mCategoryLinks
; }
2670 function getCacheTime() { return $this->mCacheTime
; }
2671 function containsOldMagic() { return $this->mContainsOldMagic
; }
2672 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2673 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2674 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2675 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2676 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
2678 function merge( $other ) {
2679 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2680 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2681 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2687 * Set options of the Parser
2689 * @package MediaWiki
2693 # All variables are private
2694 var $mUseTeX; # Use texvc to expand <math> tags
2695 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2696 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2697 var $mAllowExternalImages; # Allow external images inline
2698 var $mSkin; # Reference to the preferred skin
2699 var $mDateFormat; # Date format index
2700 var $mEditSection; # Create "edit section" links
2701 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2702 var $mNumberHeadings; # Automatically number headings
2703 var $mShowToc; # Show table of contents
2705 function getUseTeX() { return $this->mUseTeX
; }
2706 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2707 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2708 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2709 function getSkin() { return $this->mSkin
; }
2710 function getDateFormat() { return $this->mDateFormat
; }
2711 function getEditSection() { return $this->mEditSection
; }
2712 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
2713 function getNumberHeadings() { return $this->mNumberHeadings
; }
2714 function getShowToc() { return $this->mShowToc
; }
2716 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2717 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2718 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2719 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2720 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2721 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
2722 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
2723 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
2724 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
2726 function setSkin( &$x ) { $this->mSkin
=& $x; }
2728 # Get parser options
2729 /* static */ function newFromUser( &$user ) {
2730 $popts = new ParserOptions
;
2731 $popts->initialiseFromUser( $user );
2736 function initialiseFromUser( &$userInput ) {
2737 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2739 $fname = 'ParserOptions::initialiseFromUser';
2740 wfProfileIn( $fname );
2741 if ( !$userInput ) {
2743 $user->setLoaded( true );
2745 $user =& $userInput;
2748 $this->mUseTeX
= $wgUseTeX;
2749 $this->mUseDynamicDates
= $wgUseDynamicDates;
2750 $this->mInterwikiMagic
= $wgInterwikiMagic;
2751 $this->mAllowExternalImages
= $wgAllowExternalImages;
2752 wfProfileIn( $fname.'-skin' );
2753 $this->mSkin
=& $user->getSkin();
2754 wfProfileOut( $fname.'-skin' );
2755 $this->mDateFormat
= $user->getOption( 'date' );
2756 $this->mEditSection
= $user->getOption( 'editsection' );
2757 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
2758 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
2759 $this->mShowToc
= $user->getOption( 'showtoc' );
2760 wfProfileOut( $fname );
2766 # Regex callbacks, used in Parser::replaceVariables
2767 function wfBraceSubstitution( $matches ) {
2768 global $wgCurParser;
2769 return $wgCurParser->braceSubstitution( $matches );
2772 function wfArgSubstitution( $matches ) {
2773 global $wgCurParser;
2774 return $wgCurParser->argSubstitution( $matches );
2777 function wfVariableSubstitution( $matches ) {
2778 global $wgCurParser;
2779 return $wgCurParser->variableSubstitution( $matches );
2783 * Return the total number of articles
2785 function wfNumberOfArticles() {
2786 global $wgNumberOfArticles;
2789 return $wgNumberOfArticles;
2793 * Get various statistics from the database
2796 function wfLoadSiteStats() {
2797 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
2798 $fname = 'wfLoadSiteStats';
2800 if ( -1 != $wgNumberOfArticles ) return;
2801 $dbr =& wfGetDB( DB_SLAVE
);
2802 $s = $dbr->getArray( 'site_stats',
2803 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
2804 array( 'ss_row_id' => 1 ), $fname
2807 if ( $s === false ) {
2810 $wgTotalViews = $s->ss_total_views
;
2811 $wgTotalEdits = $s->ss_total_edits
;
2812 $wgNumberOfArticles = $s->ss_good_articles
;
2816 function wfEscapeHTMLTagsOnly( $in ) {
2818 array( '"', '>', '<' ),
2819 array( '"', '>', '<' ),