4 * File for Parser and related classes
11 * Variable substitution O(N^2) attack
13 * Without countermeasures, it would be possible to attack the parser by saving
14 * a page filled with a large number of inclusions of large pages. The size of
15 * the generated page would be proportional to the square of the input size.
16 * Hence, we limit the number of inclusions of any given page, thus bringing any
17 * attack back to O(N).
19 define( 'MAX_INCLUDE_REPEAT', 100 );
20 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
22 # Allowed values for $mOutputType
23 define( 'OT_HTML', 1 );
24 define( 'OT_WIKI', 2 );
25 define( 'OT_MSG' , 3 );
27 # string parameter for extractTags which will cause it
28 # to strip HTML comments in addition to regular
29 # <XML>-style tags. This should not be anything we
30 # may want to use in wikisyntax
31 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
33 # prefix for escaping, used in two functions at least
34 define( 'UNIQ_PREFIX', 'NaodW29');
36 # Constants needed for external link processing
37 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
38 define( 'HTTP_PROTOCOLS', 'http|https' );
39 # Everything except bracket, space, or control characters
40 define( 'EXT_LINK_URL_CLASS', '[^]\\x00-\\x20\\x7F]' );
41 define( 'INVERSE_EXT_LINK_URL_CLASS', '[\]\\x00-\\x20\\x7F]' );
43 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
44 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
45 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
46 define( 'EXT_LINK_BRACKETED', '/\[(('.URL_PROTOCOLS
.'):'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
47 define( 'EXT_IMAGE_REGEX',
48 '/^('.HTTP_PROTOCOLS
.':)'. # Protocol
49 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
50 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
56 * Processes wiki markup
59 * There are three main entry points into the Parser class:
61 * produces HTML output
63 * produces altered wiki markup.
65 * performs brace substitution on MediaWiki messages
68 * objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
70 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
73 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
74 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
77 * * only within ParserOptions
90 # Cleared with clearState():
91 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
92 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
95 var $mOptions, $mTitle, $mOutputType,
96 $mTemplates, // cache of already loaded templates, avoids
97 // multiple SQL queries for the same string
98 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
99 // in this path. Used for loop detection.
109 $this->mTemplates
= array();
110 $this->mTemplatePath
= array();
111 $this->mTagHooks
= array();
120 function clearState() {
121 $this->mOutput
= new ParserOutput
;
122 $this->mAutonumber
= 0;
123 $this->mLastSection
= "";
124 $this->mDTopen
= false;
125 $this->mVariables
= false;
126 $this->mIncludeCount
= array();
127 $this->mStripState
= array();
128 $this->mArgStack
= array();
129 $this->mInPre
= false;
133 * First pass--just handle <nowiki> sections, pass the rest off
134 * to internalParse() which does all the real work.
137 * @return ParserOutput a ParserOutput
139 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
141 $fname = 'Parser::parse';
142 wfProfileIn( $fname );
148 $this->mOptions
= $options;
149 $this->mTitle
=& $title;
150 $this->mOutputType
= OT_HTML
;
153 $text = $this->strip( $text, $this->mStripState
);
154 $text = $this->internalParse( $text, $linestart );
155 $text = $this->unstrip( $text, $this->mStripState
);
156 # Clean up special characters, only run once, next-to-last before doBlockLevels
159 # french spaces, last one Guillemet-left
160 # only if there is something before the space
161 '/(.) (?=\\?|:|;|!|\\302\\273)/i' => '\\1 \\2',
162 # french spaces, Guillemet-right
163 "/(\\302\\253) /i"=>"\\1 ",
164 '/<hr *>/i' => '<hr />',
165 '/<br *>/i' => '<br />',
166 '/<center *>/i' => '<div class="center">',
167 '/<\\/center *>/i' => '</div>',
168 # Clean up spare ampersands; note that we probably ought to be
169 # more careful about named entities.
170 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
172 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
175 # french spaces, last one Guillemet-left
176 '/ (\\?|:|;|!|\\302\\273)/i' => ' \\1',
177 # french spaces, Guillemet-right
178 '/(\\302\\253) /i' => '\\1 ',
179 '/<center *>/i' => '<div class="center">',
180 '/<\\/center *>/i' => '</div>'
182 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
185 $text = $this->doBlockLevels( $text, $linestart );
186 $text = $this->unstripNoWiki( $text, $this->mStripState
);
187 $this->mOutput
->setText( $text );
188 wfProfileOut( $fname );
189 return $this->mOutput
;
193 * Get a random string
198 function getRandomString() {
199 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
203 * Replaces all occurrences of <$tag>content</$tag> in the text
204 * with a random marker and returns the new text. the output parameter
205 * $content will be an associative array filled with data on the form
206 * $unique_marker => content.
208 * If $content is already set, the additional entries will be appended
209 * If $tag is set to STRIP_COMMENTS, the function will extract
210 * <!-- HTML comments -->
215 function extractTags($tag, $text, &$content, $uniq_prefix = ''){
216 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
223 while ( '' != $text ) {
224 if($tag==STRIP_COMMENTS
) {
225 $p = preg_split( '/<!--/i', $text, 2 );
227 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
230 if ( ( count( $p ) < 2 ) ||
( '' == $p[1] ) ) {
233 if($tag==STRIP_COMMENTS
) {
234 $q = preg_split( '/-->/i', $p[1], 2 );
236 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
238 $marker = $rnd . sprintf('%08X', $n++
);
239 $content[$marker] = $q[0];
240 $stripped .= $marker;
248 * Strips and renders nowiki, pre, math, hiero
249 * If $render is set, performs necessary rendering operations on plugins
250 * Returns the text, and fills an array with data needed in unstrip()
251 * If the $state is already a valid strip state, it adds to the state
253 * @param bool $stripcomments when set, HTML comments <!-- like this -->
254 * will be stripped in addition to other tags. This is important
255 * for section editing, where these comments cause confusion when
256 * counting the sections in the wikisource
260 function strip( $text, &$state, $stripcomments = false ) {
261 $render = ($this->mOutputType
== OT_HTML
);
262 $html_content = array();
263 $nowiki_content = array();
264 $math_content = array();
265 $pre_content = array();
266 $comment_content = array();
267 $ext_content = array();
269 # Replace any instances of the placeholders
270 $uniq_prefix = UNIQ_PREFIX
;
271 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
274 global $wgRawHtml, $wgWhitelistEdit;
275 if( $wgRawHtml && $wgWhitelistEdit ) {
276 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
277 foreach( $html_content as $marker => $content ) {
279 # Raw and unchecked for validity.
280 $html_content[$marker] = $content;
282 $html_content[$marker] = '<html>'.$content.'</html>';
288 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
289 foreach( $nowiki_content as $marker => $content ) {
291 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
293 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
298 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
299 foreach( $math_content as $marker => $content ){
301 if( $this->mOptions
->getUseTeX() ) {
302 $math_content[$marker] = renderMath( $content );
304 $math_content[$marker] = '<math>'.$content.'<math>';
307 $math_content[$marker] = '<math>'.$content.'</math>';
312 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
313 foreach( $pre_content as $marker => $content ){
315 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
317 $pre_content[$marker] = '<pre>'.$content.'</pre>';
323 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
324 foreach( $comment_content as $marker => $content ){
325 $comment_content[$marker] = '<!--'.$content.'-->';
330 foreach ( $this->mTagHooks
as $tag => $callback ) {
331 $ext_contents[$tag] = array();
332 $text = Parser
::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
333 foreach( $ext_content[$tag] as $marker => $content ) {
335 $ext_content[$tag][$marker] = $callback( $content );
337 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
342 # Merge state with the pre-existing state, if there is one
344 $state['html'] = $state['html'] +
$html_content;
345 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
346 $state['math'] = $state['math'] +
$math_content;
347 $state['pre'] = $state['pre'] +
$pre_content;
348 $state['comment'] = $state['comment'] +
$comment_content;
350 foreach( $ext_content as $tag => $array ) {
351 if ( array_key_exists( $tag, $state ) ) {
352 $state[$tag] = $state[$tag] +
$array;
357 'html' => $html_content,
358 'nowiki' => $nowiki_content,
359 'math' => $math_content,
360 'pre' => $pre_content,
361 'comment' => $comment_content,
368 * restores pre, math, and heiro removed by strip()
370 * always call unstripNoWiki() after this one
373 function unstrip( $text, &$state ) {
374 # Must expand in reverse order, otherwise nested tags will be corrupted
375 $contentDict = end( $state );
376 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
377 if( key($state) != 'nowiki' && key($state) != 'html') {
378 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
379 $text = str_replace( key( $contentDict ), $content, $text );
388 * always call this after unstrip() to preserve the order
392 function unstripNoWiki( $text, &$state ) {
393 # Must expand in reverse order, otherwise nested tags will be corrupted
394 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
395 $text = str_replace( key( $state['nowiki'] ), $content, $text );
400 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
401 $text = str_replace( key( $state['html'] ), $content, $text );
409 * Add an item to the strip state
410 * Returns the unique tag which must be inserted into the stripped text
411 * The tag will be replaced with the original text in unstrip()
415 function insertStripItem( $text, &$state ) {
416 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
425 $state['item'][$rnd] = $text;
430 * Return allowed HTML attributes
434 function getHTMLattrs () {
435 $htmlattrs = array( # Allowed attributes--no scripting, etc.
436 'title', 'align', 'lang', 'dir', 'width', 'height',
437 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
438 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
439 /* FONT */ 'type', 'start', 'value', 'compact',
440 /* For various lists, mostly deprecated but safe */
441 'summary', 'width', 'border', 'frame', 'rules',
442 'cellspacing', 'cellpadding', 'valign', 'char',
443 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
444 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
445 'id', 'class', 'name', 'style' /* For CSS */
451 * Remove non approved attributes and javascript in css
455 function fixTagAttributes ( $t ) {
456 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
457 $htmlattrs = $this->getHTMLattrs() ;
459 # Strip non-approved attributes from the tag
461 '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
462 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
465 $t = str_replace ( '<></>' , '' , $t ) ; # This should fix bug 980557
467 # Strip javascript "expression" from stylesheets. Brute force approach:
468 # If anythin offensive is found, all attributes of the HTML tag are dropped
471 '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
472 wfMungeToUtf8( $t ) ) )
481 * interface with html tidy, used if $wgUseTidy = true
486 function tidy ( $text ) {
487 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
488 global $wgInputEncoding, $wgOutputEncoding;
489 $fname = 'Parser::tidy';
490 wfProfileIn( $fname );
494 switch(strtoupper($wgOutputEncoding)) {
496 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -latin1':' -raw';
499 $opts .= ($wgInputEncoding == $wgOutputEncoding)?
' -utf8':' -raw';
505 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
506 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
507 '<head><title>test</title></head><body>'.$text.'</body></html>';
508 $descriptorspec = array(
509 0 => array('pipe', 'r'),
510 1 => array('pipe', 'w'),
511 2 => array('file', '/dev/null', 'a')
513 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
514 if (is_resource($process)) {
515 fwrite($pipes[0], $wrappedtext);
517 while (!feof($pipes[1])) {
518 $cleansource .= fgets($pipes[1], 1024);
521 $return_value = proc_close($process);
524 wfProfileOut( $fname );
526 if( $cleansource == '' && $text != '') {
527 wfDebug( "Tidy error detected!\n" );
528 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
535 * parse the wiki syntax used to render tables
539 function doTableStuff ( $t ) {
540 $fname = 'Parser::doTableStuff';
541 wfProfileIn( $fname );
543 $t = explode ( "\n" , $t ) ;
544 $td = array () ; # Is currently a td tag open?
545 $ltd = array () ; # Was it TD or TH?
546 $tr = array () ; # Is currently a tr tag open?
547 $ltr = array () ; # tr attributes
548 $indent_level = 0; # indent level of the table
549 foreach ( $t AS $k => $x )
552 $fc = substr ( $x , 0 , 1 ) ;
553 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
554 $indent_level = strlen( $matches[1] );
556 str_repeat( '<dl><dd>', $indent_level ) .
557 '<table ' . $this->fixTagAttributes ( $matches[2] ) . '>' ;
558 array_push ( $td , false ) ;
559 array_push ( $ltd , '' ) ;
560 array_push ( $tr , false ) ;
561 array_push ( $ltr , '' ) ;
563 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
564 else if ( '|}' == substr ( $x , 0 , 2 ) ) {
566 $l = array_pop ( $ltd ) ;
567 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
568 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
570 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
572 else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
573 $x = substr ( $x , 1 ) ;
574 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
576 $l = array_pop ( $ltd ) ;
577 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
578 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
581 array_push ( $tr , false ) ;
582 array_push ( $td , false ) ;
583 array_push ( $ltd , '' ) ;
584 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
586 else if ( '|' == $fc ||
'!' == $fc ||
'|+' == substr ( $x , 0 , 2 ) ) { # Caption
587 if ( '|+' == substr ( $x , 0 , 2 ) ) {
589 $x = substr ( $x , 1 ) ;
591 $after = substr ( $x , 1 ) ;
592 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
593 $after = explode ( '||' , $after ) ;
595 foreach ( $after AS $theline )
600 $tra = array_pop ( $ltr ) ;
601 if ( !array_pop ( $tr ) ) $z = '<tr '.$tra.">\n" ;
602 array_push ( $tr , true ) ;
603 array_push ( $ltr , '' ) ;
606 $l = array_pop ( $ltd ) ;
607 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
608 if ( $fc == '|' ) $l = 'td' ;
609 else if ( $fc == '!' ) $l = 'th' ;
610 else if ( $fc == '+' ) $l = 'caption' ;
612 array_push ( $ltd , $l ) ;
613 $y = explode ( '|' , $theline , 2 ) ;
614 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
615 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
617 array_push ( $td , true ) ;
622 # Closing open td, tr && table
623 while ( count ( $td ) > 0 )
625 if ( array_pop ( $td ) ) $t[] = '</td>' ;
626 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
630 $t = implode ( "\n" , $t ) ;
631 # $t = $this->removeHTMLtags( $t );
632 wfProfileOut( $fname );
637 * Helper function for parse() that transforms wiki markup into
638 * HTML. Only called for $mOutputType == OT_HTML.
642 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
645 $fname = 'Parser::internalParse';
646 wfProfileIn( $fname );
648 $text = $this->removeHTMLtags( $text );
649 $text = $this->replaceVariables( $text, $args );
651 $text = $wgContLang->convert($text);
653 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
655 $text = $this->doHeadings( $text );
656 if($this->mOptions
->getUseDynamicDates()) {
657 global $wgDateFormatter;
658 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
660 $text = $this->doAllQuotes( $text );
661 $text = $this->doMagicLinks( $text );
662 $text = $this->replaceInternalLinks ( $text );
663 # Another call to replace links and images inside captions of images
664 $text = $this->replaceInternalLinks ( $text );
665 $text = $this->replaceExternalLinks( $text );
666 $text = $this->doTableStuff( $text );
667 $text = $this->formatHeadings( $text, $isMain );
668 $sk =& $this->mOptions
->getSkin();
669 $text = $sk->transformContent( $text );
671 wfProfileOut( $fname );
676 * Replace special strings like "ISBN xxx" and "RFC xxx" with
677 * magic external links.
681 function &doMagicLinks( &$text ) {
682 global $wgUseGeoMode;
683 $text = $this->magicISBN( $text );
684 if ( isset( $wgUseGeoMode ) && $wgUseGeoMode ) {
685 $text = $this->magicGEO( $text );
687 $text = $this->magicRFC( $text );
692 * Parse ^^ tokens and return html
696 function doExponent ( $text ) {
697 $fname = 'Parser::doExponent';
698 wfProfileIn( $fname);
699 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
700 wfProfileOut( $fname);
705 * Parse headers and return html
709 function doHeadings( $text ) {
710 $fname = 'Parser::doHeadings';
711 wfProfileIn( $fname );
712 for ( $i = 6; $i >= 1; --$i ) {
713 $h = substr( '======', 0, $i );
714 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
715 "<h{$i}>\\1</h{$i}>\\2", $text );
717 wfProfileOut( $fname );
722 * Replace single quotes with HTML markup
724 * @return string the altered text
726 function doAllQuotes( $text ) {
727 $fname = 'Parser::doAllQuotes';
728 wfProfileIn( $fname );
730 $lines = explode( "\n", $text );
731 foreach ( $lines as $line ) {
732 $outtext .= $this->doQuotes ( $line ) . "\n";
734 $outtext = substr($outtext, 0,-1);
735 wfProfileOut( $fname );
740 * Helper function for doAllQuotes()
743 function doQuotes( $text ) {
744 $arr = preg_split ("/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
745 if (count ($arr) == 1)
749 # First, do some preliminary work. This may shift some apostrophes from
750 # being mark-up to being text. It also counts the number of occurrences
751 # of bold and italics mark-ups.
759 # If there are ever four apostrophes, assume the first is supposed to
760 # be text, and the remaining three constitute mark-up for bold text.
761 if (strlen ($arr[$i]) == 4)
766 # If there are more than 5 apostrophes in a row, assume they're all
767 # text except for the last 5.
768 else if (strlen ($arr[$i]) > 5)
770 $arr[$i-1] .= str_repeat ("'", strlen ($arr[$i]) - 5);
773 # Count the number of occurrences of bold and italics mark-ups.
774 # We are not counting sequences of five apostrophes.
775 if (strlen ($arr[$i]) == 2) $numitalics++
; else
776 if (strlen ($arr[$i]) == 3) $numbold++
; else
777 if (strlen ($arr[$i]) == 5) { $numitalics++
; $numbold++
; }
782 # If there is an odd number of both bold and italics, it is likely
783 # that one of the bold ones was meant to be an apostrophe followed
784 # by italics. Which one we cannot know for certain, but it is more
785 # likely to be one that has a single-letter word before it.
786 if (($numbold %
2 == 1) && ($numitalics %
2 == 1))
789 $firstsingleletterword = -1;
790 $firstmultiletterword = -1;
794 if (($i %
2 == 1) and (strlen ($r) == 3))
796 $x1 = substr ($arr[$i-1], -1);
797 $x2 = substr ($arr[$i-1], -2, 1);
799 if ($firstspace == -1) $firstspace = $i;
800 } else if ($x2 == ' ') {
801 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
803 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
809 # If there is a single-letter word, use it!
810 if ($firstsingleletterword > -1)
812 $arr [ $firstsingleletterword ] = "''";
813 $arr [ $firstsingleletterword-1 ] .= "'";
815 # If not, but there's a multi-letter word, use that one.
816 else if ($firstmultiletterword > -1)
818 $arr [ $firstmultiletterword ] = "''";
819 $arr [ $firstmultiletterword-1 ] .= "'";
821 # ... otherwise use the first one that has neither.
822 # (notice that it is possible for all three to be -1 if, for example,
823 # there is only one pentuple-apostrophe in the line)
824 else if ($firstspace > -1)
826 $arr [ $firstspace ] = "''";
827 $arr [ $firstspace-1 ] .= "'";
831 # Now let's actually convert our apostrophic mush to HTML!
840 if ($state == 'both')
847 if (strlen ($r) == 2)
850 { $output .= '</i>'; $state = ''; }
851 else if ($state == 'bi')
852 { $output .= '</i>'; $state = 'b'; }
853 else if ($state == 'ib')
854 { $output .= '</b></i><b>'; $state = 'b'; }
855 else if ($state == 'both')
856 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
857 else # $state can be 'b' or ''
858 { $output .= '<i>'; $state .= 'i'; }
860 else if (strlen ($r) == 3)
863 { $output .= '</b>'; $state = ''; }
864 else if ($state == 'bi')
865 { $output .= '</i></b><i>'; $state = 'i'; }
866 else if ($state == 'ib')
867 { $output .= '</b>'; $state = 'i'; }
868 else if ($state == 'both')
869 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
870 else # $state can be 'i' or ''
871 { $output .= '<b>'; $state .= 'b'; }
873 else if (strlen ($r) == 5)
876 { $output .= '</b><i>'; $state = 'i'; }
877 else if ($state == 'i')
878 { $output .= '</i><b>'; $state = 'b'; }
879 else if ($state == 'bi')
880 { $output .= '</i></b>'; $state = ''; }
881 else if ($state == 'ib')
882 { $output .= '</b></i>'; $state = ''; }
883 else if ($state == 'both')
884 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
885 else # ($state == '')
886 { $buffer = ''; $state = 'both'; }
891 # Now close all remaining tags. Notice that the order is important.
892 if ($state == 'b' ||
$state == 'ib')
894 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
898 if ($state == 'both')
899 $output .= '<b><i>'.$buffer.'</i></b>';
905 * Replace external links
907 * Note: we have to do external links before the internal ones,
908 * and otherwise take great care in the order of things here, so
909 * that we don't end up interpreting some URLs twice.
913 function replaceExternalLinks( $text ) {
914 $fname = 'Parser::replaceExternalLinks';
915 wfProfileIn( $fname );
917 $sk =& $this->mOptions
->getSkin();
918 $linktrail = wfMsgForContent('linktrail');
919 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
921 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
924 while ( $i<count( $bits ) ) {
926 $protocol = $bits[$i++
];
928 $trail = $bits[$i++
];
930 # If the link text is an image URL, replace it with an <img> tag
931 # This happened by accident in the original parser, but some people used it extensively
932 $img = $this->maybeMakeImageLink( $text );
933 if ( $img !== false ) {
939 # No link text, e.g. [http://domain.tld/some.link]
941 # Autonumber if allowed
942 if ( strpos( HTTP_PROTOCOLS
, $protocol ) !== false ) {
943 $text = '[' . ++
$this->mAutonumber
. ']';
945 # Otherwise just use the URL
946 $text = htmlspecialchars( $url );
949 # Have link text, e.g. [http://domain.tld/some.link text]s
951 if ( preg_match( $linktrail, $trail, $m2 ) ) {
957 $encUrl = htmlspecialchars( $url );
958 # Bit in parentheses showing the URL for the printable version
959 if( $url == $text ||
preg_match( "!$protocol://" . preg_quote( $text, '/' ) . "/?$!", $url ) ) {
962 # Expand the URL for printable version
963 if ( ! $sk->suppressUrlExpansion() ) {
964 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $encUrl ) . "</i>)</span>";
970 # Process the trail (i.e. everything after this link up until start of the next link),
971 # replacing any non-bracketed links
972 $trail = $this->replaceFreeExternalLinks( $trail );
974 $la = $sk->getExternalLinkAttributes( $url, $text );
976 # Use the encoded URL
977 # This means that users can paste URLs directly into the text
978 # Funny characters like ö aren't valid in URLs anyway
979 # This was changed in August 2004
980 $s .= "<a href=\"{$url}\"{$la}>{$text}</a>{$dtrail}{$paren}{$trail}";
983 wfProfileOut( $fname );
988 * Replace anything that looks like a URL with a link
991 function replaceFreeExternalLinks( $text ) {
992 $bits = preg_split( '/((?:'.URL_PROTOCOLS
.'):)/', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
993 $s = array_shift( $bits );
996 $sk =& $this->mOptions
->getSkin();
998 while ( $i < count( $bits ) ){
999 $protocol = $bits[$i++
];
1000 $remainder = $bits[$i++
];
1002 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1003 # Found some characters after the protocol that look promising
1004 $url = $protocol . $m[1];
1007 # Move trailing punctuation to $trail
1009 # If there is no left bracket, then consider right brackets fair game too
1010 if ( strpos( $url, '(' ) === false ) {
1014 $numSepChars = strspn( strrev( $url ), $sep );
1015 if ( $numSepChars ) {
1016 $trail = substr( $url, -$numSepChars ) . $trail;
1017 $url = substr( $url, 0, -$numSepChars );
1020 # Replace & from obsolete syntax with &
1021 $url = str_replace( '&', '&', $url );
1023 # Is this an external image?
1024 $text = $this->maybeMakeImageLink( $url );
1025 if ( $text === false ) {
1026 # Not an image, make a link
1027 $text = $sk->makeExternalLink( $url, $url );
1029 $s .= $text . $trail;
1031 $s .= $protocol . $remainder;
1038 * make an image if it's allowed
1041 function maybeMakeImageLink( $url ) {
1042 $sk =& $this->mOptions
->getSkin();
1044 if ( $this->mOptions
->getAllowExternalImages() ) {
1045 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1047 $text = $sk->makeImage( htmlspecialchars( $url ) );
1054 * Process [[ ]] wikilinks
1058 function replaceInternalLinks( $s ) {
1059 global $wgLang, $wgContLang, $wgLinkCache;
1060 global $wgNamespacesWithSubpages;
1061 static $fname = 'Parser::replaceInternalLinks' ;
1062 wfProfileIn( $fname );
1064 wfProfileIn( $fname.'-setup' );
1066 # the % is needed to support urlencoded titles as well
1067 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1068 $sk =& $this->mOptions
->getSkin();
1070 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1072 $a = explode( '[[', ' ' . $s );
1073 $s = array_shift( $a );
1074 $s = substr( $s, 1 );
1076 # Match a link having the form [[namespace:link|alternate]]trail
1078 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1079 # Match the end of a line for a word that's not followed by whitespace,
1080 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1081 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1083 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1084 # Special and Media are pseudo-namespaces; no pages actually exist in them
1086 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1088 if ( $useLinkPrefixExtension ) {
1089 if ( preg_match( $e2, $s, $m ) ) {
1090 $first_prefix = $m[2];
1093 $first_prefix = false;
1099 wfProfileOut( $fname.'-setup' );
1101 # start procedeeding each line
1102 foreach ( $a as $line ) {
1103 wfProfileIn( $fname.'-prefixhandling' );
1104 if ( $useLinkPrefixExtension ) {
1105 if ( preg_match( $e2, $s, $m ) ) {
1113 $prefix = $first_prefix;
1114 $first_prefix = false;
1117 wfProfileOut( $fname.'-prefixhandling' );
1119 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1121 # fix up urlencoded title texts
1122 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1124 } else { # Invalid form; output directly
1125 $s .= $prefix . '[[' . $line ;
1129 # Don't allow internal links to pages containing
1130 # PROTO: where PROTO is a valid URL protocol; these
1131 # should be external links.
1132 if (preg_match('/((?:'.URL_PROTOCOLS
.'):)/', $m[1])) {
1133 $s .= $prefix . '[[' . $line ;
1139 # :Foobar -- override special treatment of prefix (images, language links)
1140 # /Foobar -- convert to CurrentPage/Foobar
1141 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1143 # Look at the first character
1144 $c = substr($m[1],0,1);
1145 $noforce = ($c != ':');
1149 # / at end means we don't want the slash to be shown
1150 if(substr($m[1],-1,1)=='/') {
1151 $m[1]=substr($m[1],1,strlen($m[1])-2);
1154 $noslash=substr($m[1],1);
1157 # Some namespaces don't allow subpages
1158 if(!empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()])) {
1159 # subpages allowed here
1160 $link = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1163 } # this might be changed for ugliness reasons
1165 # no subpage allowed, use standard link
1169 } elseif( $noforce ) { # no subpage
1172 # We don't want to keep the first character
1173 $link = substr( $m[1], 1 );
1176 $wasblank = ( '' == $text );
1177 if( $wasblank ) $text = $link;
1179 $nt = Title
::newFromText( $link );
1181 $s .= $prefix . '[[' . $line;
1185 $ns = $nt->getNamespace();
1186 $iw = $nt->getInterWiki();
1188 # Link not escaped by : , create the various objects
1192 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1193 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1194 $tmp = $prefix . $trail ;
1195 $s .= (trim($tmp) == '')?
'': $tmp;
1199 if ( $ns == NS_IMAGE
) {
1200 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1201 $wgLinkCache->addImageLinkObj( $nt );
1205 if ( $ns == NS_CATEGORY
) {
1206 $t = $nt->getText() ;
1207 $nnt = Title
::newFromText ( Namespace::getCanonicalName(NS_CATEGORY
).':'.$t ) ;
1209 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1210 $pPLC=$sk->postParseLinkColour();
1211 $sk->postParseLinkColour( false );
1212 $t = $sk->makeLinkObj( $nnt, $t, '', '' , $prefix );
1213 $sk->postParseLinkColour( $pPLC );
1214 $wgLinkCache->resume();
1217 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1218 $sortkey = $this->mTitle
->getText();
1220 $sortkey = $this->mTitle
->getPrefixedText();
1225 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1226 $this->mOutput
->mCategoryLinks
[] = $t ;
1227 $s .= $prefix . $trail ;
1232 if( ( $nt->getPrefixedText() === $this->mTitle
->getPrefixedText() ) &&
1233 ( strpos( $link, '#' ) === FALSE ) ) {
1234 # Self-links are handled specially; generally de-link and change to bold.
1235 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1239 if( $ns == NS_MEDIA
) {
1240 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1241 $wgLinkCache->addImageLinkObj( $nt );
1243 } elseif( $ns == NS_SPECIAL
) {
1244 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1247 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1249 wfProfileOut( $fname );
1254 * Used by doBlockLevels()
1257 /* private */ function closeParagraph() {
1259 if ( '' != $this->mLastSection
) {
1260 $result = '</' . $this->mLastSection
. ">\n";
1262 $this->mInPre
= false;
1263 $this->mLastSection
= '';
1266 # getCommon() returns the length of the longest common substring
1267 # of both arguments, starting at the beginning of both.
1269 /* private */ function getCommon( $st1, $st2 ) {
1270 $fl = strlen( $st1 );
1271 $shorter = strlen( $st2 );
1272 if ( $fl < $shorter ) { $shorter = $fl; }
1274 for ( $i = 0; $i < $shorter; ++
$i ) {
1275 if ( $st1{$i} != $st2{$i} ) { break; }
1279 # These next three functions open, continue, and close the list
1280 # element appropriate to the prefix character passed into them.
1282 /* private */ function openList( $char ) {
1283 $result = $this->closeParagraph();
1285 if ( '*' == $char ) { $result .= '<ul><li>'; }
1286 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1287 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1288 else if ( ';' == $char ) {
1289 $result .= '<dl><dt>';
1290 $this->mDTopen
= true;
1292 else { $result = '<!-- ERR 1 -->'; }
1297 /* private */ function nextItem( $char ) {
1298 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1299 else if ( ':' == $char ||
';' == $char ) {
1301 if ( $this->mDTopen
) { $close = '</dt>'; }
1302 if ( ';' == $char ) {
1303 $this->mDTopen
= true;
1304 return $close . '<dt>';
1306 $this->mDTopen
= false;
1307 return $close . '<dd>';
1310 return '<!-- ERR 2 -->';
1313 /* private */ function closeList( $char ) {
1314 if ( '*' == $char ) { $text = '</li></ul>'; }
1315 else if ( '#' == $char ) { $text = '</li></ol>'; }
1316 else if ( ':' == $char ) {
1317 if ( $this->mDTopen
) {
1318 $this->mDTopen
= false;
1319 $text = '</dt></dl>';
1321 $text = '</dd></dl>';
1324 else { return '<!-- ERR 3 -->'; }
1330 * Make lists from lines starting with ':', '*', '#', etc.
1333 * @return string the lists rendered as HTML
1335 function doBlockLevels( $text, $linestart ) {
1336 $fname = 'Parser::doBlockLevels';
1337 wfProfileIn( $fname );
1339 # Parsing through the text line by line. The main thing
1340 # happening here is handling of block-level elements p, pre,
1341 # and making lists from lines starting with * # : etc.
1343 $textLines = explode( "\n", $text );
1345 $lastPrefix = $output = $lastLine = '';
1346 $this->mDTopen
= $inBlockElem = false;
1348 $paragraphStack = false;
1350 if ( !$linestart ) {
1351 $output .= array_shift( $textLines );
1353 foreach ( $textLines as $oLine ) {
1354 $lastPrefixLength = strlen( $lastPrefix );
1355 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1356 $preOpenMatch = preg_match('/<pre/i', $oLine );
1357 if ( !$this->mInPre
) {
1358 # Multiple prefixes may abut each other for nested lists.
1359 $prefixLength = strspn( $oLine, '*#:;' );
1360 $pref = substr( $oLine, 0, $prefixLength );
1363 $pref2 = str_replace( ';', ':', $pref );
1364 $t = substr( $oLine, $prefixLength );
1365 $this->mInPre
= !empty($preOpenMatch);
1367 # Don't interpret any other prefixes in preformatted text
1369 $pref = $pref2 = '';
1374 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1375 # Same as the last item, so no need to deal with nesting or opening stuff
1376 $output .= $this->nextItem( substr( $pref, -1 ) );
1377 $paragraphStack = false;
1379 if ( substr( $pref, -1 ) == ';') {
1380 # The one nasty exception: definition lists work like this:
1381 # ; title : definition text
1382 # So we check for : in the remainder text to split up the
1383 # title and definition, without b0rking links.
1384 # FIXME: This is not foolproof. Something better in Tokenizer might help.
1385 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1387 $output .= $term . $this->nextItem( ':' );
1391 } elseif( $prefixLength ||
$lastPrefixLength ) {
1392 # Either open or close a level...
1393 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1394 $paragraphStack = false;
1396 while( $commonPrefixLength < $lastPrefixLength ) {
1397 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1398 --$lastPrefixLength;
1400 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1401 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1403 while ( $prefixLength > $commonPrefixLength ) {
1404 $char = substr( $pref, $commonPrefixLength, 1 );
1405 $output .= $this->openList( $char );
1407 if ( ';' == $char ) {
1408 # FIXME: This is dupe of code above
1409 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1411 $output .= $term . $this->nextItem( ':' );
1415 ++
$commonPrefixLength;
1417 $lastPrefix = $pref2;
1419 if( 0 == $prefixLength ) {
1420 # No prefix (not in list)--go to paragraph mode
1421 $uniq_prefix = UNIQ_PREFIX
;
1422 // XXX: use a stack for nestable elements like span, table and div
1423 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1424 $closematch = preg_match(
1425 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1426 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1427 if ( $openmatch or $closematch ) {
1428 $paragraphStack = false;
1429 $output .= $this->closeParagraph();
1430 if($preOpenMatch and !$preCloseMatch) {
1431 $this->mInPre
= true;
1433 if ( $closematch ) {
1434 $inBlockElem = false;
1436 $inBlockElem = true;
1438 } else if ( !$inBlockElem && !$this->mInPre
) {
1439 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1441 if ($this->mLastSection
!= 'pre') {
1442 $paragraphStack = false;
1443 $output .= $this->closeParagraph().'<pre>';
1444 $this->mLastSection
= 'pre';
1446 $t = substr( $t, 1 );
1449 if ( '' == trim($t) ) {
1450 if ( $paragraphStack ) {
1451 $output .= $paragraphStack.'<br />';
1452 $paragraphStack = false;
1453 $this->mLastSection
= 'p';
1455 if ($this->mLastSection
!= 'p' ) {
1456 $output .= $this->closeParagraph();
1457 $this->mLastSection
= '';
1458 $paragraphStack = '<p>';
1460 $paragraphStack = '</p><p>';
1464 if ( $paragraphStack ) {
1465 $output .= $paragraphStack;
1466 $paragraphStack = false;
1467 $this->mLastSection
= 'p';
1468 } else if ($this->mLastSection
!= 'p') {
1469 $output .= $this->closeParagraph().'<p>';
1470 $this->mLastSection
= 'p';
1476 if ($paragraphStack === false) {
1480 while ( $prefixLength ) {
1481 $output .= $this->closeList( $pref2{$prefixLength-1} );
1484 if ( '' != $this->mLastSection
) {
1485 $output .= '</' . $this->mLastSection
. '>';
1486 $this->mLastSection
= '';
1489 wfProfileOut( $fname );
1494 * Return value of a magic variable (like PAGENAME)
1498 function getVariableValue( $index ) {
1499 global $wgContLang, $wgSitename, $wgServer;
1502 case MAG_CURRENTMONTH
:
1503 return $wgContLang->formatNum( date( 'm' ) );
1504 case MAG_CURRENTMONTHNAME
:
1505 return $wgContLang->getMonthName( date('n') );
1506 case MAG_CURRENTMONTHNAMEGEN
:
1507 return $wgContLang->getMonthNameGen( date('n') );
1508 case MAG_CURRENTDAY
:
1509 return $wgContLang->formatNum( date('j') );
1511 return $this->mTitle
->getText();
1513 return $this->mTitle
->getPartialURL();
1515 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1516 return $wgContLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1517 case MAG_CURRENTDAYNAME
:
1518 return $wgContLang->getWeekdayName( date('w')+
1 );
1519 case MAG_CURRENTYEAR
:
1520 return $wgContLang->formatNum( date( 'Y' ) );
1521 case MAG_CURRENTTIME
:
1522 return $wgContLang->time( wfTimestampNow(), false );
1523 case MAG_NUMBEROFARTICLES
:
1524 return $wgContLang->formatNum( wfNumberOfArticles() );
1535 * initialise the magic variables (like CURRENTMONTHNAME)
1539 function initialiseVariables() {
1540 global $wgVariableIDs;
1541 $this->mVariables
= array();
1542 foreach ( $wgVariableIDs as $id ) {
1543 $mw =& MagicWord
::get( $id );
1544 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1549 * Replace magic variables, templates, and template arguments
1550 * with the appropriate text. Templates are substituted recursively,
1551 * taking care to avoid infinite loops.
1553 * Note that the substitution depends on value of $mOutputType:
1554 * OT_WIKI: only {{subst:}} templates
1555 * OT_MSG: only magic variables
1556 * OT_HTML: all templates and magic variables
1558 * @param string $tex The text to transform
1559 * @param array $args Key-value pairs representing template parameters to substitute
1562 function replaceVariables( $text, $args = array() ) {
1563 global $wgLang, $wgScript, $wgArticlePath;
1565 # Prevent too big inclusions
1566 if(strlen($text)> MAX_INCLUDE_SIZE
)
1569 $fname = 'Parser::replaceVariables';
1570 wfProfileIn( $fname );
1572 $titleChars = Title
::legalChars();
1574 # This function is called recursively. To keep track of arguments we need a stack:
1575 array_push( $this->mArgStack
, $args );
1577 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1578 $GLOBALS['wgCurParser'] =& $this;
1580 # Variable substitution
1581 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", 'wfVariableSubstitution', $text );
1583 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1584 # Argument substitution
1585 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1587 # Template substitution
1588 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1589 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1591 array_pop( $this->mArgStack
);
1593 wfProfileOut( $fname );
1598 * Replace magic variables
1601 function variableSubstitution( $matches ) {
1602 if ( !$this->mVariables
) {
1603 $this->initialiseVariables();
1606 if ( $this->mOutputType
== OT_WIKI
) {
1607 # Do only magic variables prefixed by SUBST
1608 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1609 if (!$mwSubst->matchStartAndRemove( $matches[1] ))
1611 # Note that if we don't substitute the variable below,
1612 # we don't remove the {{subst:}} magic word, in case
1613 # it is a template rather than a magic variable.
1615 if ( !$skip && array_key_exists( $matches[1], $this->mVariables
) ) {
1616 $text = $this->mVariables
[$matches[1]];
1617 $this->mOutput
->mContainsOldMagic
= true;
1619 $text = $matches[0];
1624 # Split template arguments
1625 function getTemplateArgs( $argsString ) {
1626 if ( $argsString === '' ) {
1630 $args = explode( '|', substr( $argsString, 1 ) );
1632 # If any of the arguments contains a '[[' but no ']]', it needs to be
1633 # merged with the next arg because the '|' character between belongs
1634 # to the link syntax and not the template parameter syntax.
1635 $argc = count($args);
1637 for ( $i = 0; $i < $argc-1; $i++
) {
1638 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
1639 $args[$i] .= '|'.$args[$i+
1];
1640 array_splice($args, $i+
1, 1);
1650 * Return the text of a template, after recursively
1651 * replacing any variables or templates within the template.
1653 * @param array $matches The parts of the template
1654 * $matches[1]: the title, i.e. the part before the |
1655 * $matches[2]: the parameters (including a leading |), if any
1656 * @return string the text of the template
1659 function braceSubstitution( $matches ) {
1660 global $wgLinkCache, $wgContLang;
1661 $fname = 'Parser::braceSubstitution';
1668 # Need to know if the template comes at the start of a line,
1669 # to treat the beginning of the template like the beginning
1670 # of a line for tables and block-level elements.
1671 $linestart = $matches[1];
1673 # $part1 is the bit before the first |, and must contain only title characters
1674 # $args is a list of arguments, starting from index 0, not including $part1
1676 $part1 = $matches[2];
1677 # If the third subpattern matched anything, it will start with |
1679 $args = $this->getTemplateArgs($matches[3]);
1680 $argc = count( $args );
1682 # Don't parse {{{}}} because that's only for template arguments
1683 if ( $linestart === '{' ) {
1684 $text = $matches[0];
1691 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1692 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
1693 # One of two possibilities is true:
1694 # 1) Found SUBST but not in the PST phase
1695 # 2) Didn't find SUBST and in the PST phase
1696 # In either case, return without further processing
1697 $text = $matches[0];
1703 # MSG, MSGNW and INT
1706 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1707 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1710 # Remove obsolete MSG:
1711 $mwMsg =& MagicWord
::get( MAG_MSG
);
1712 $mwMsg->matchStartAndRemove( $part1 );
1715 # Check if it is an internal message
1716 $mwInt =& MagicWord
::get( MAG_INT
);
1717 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1718 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1719 $text = $linestart . wfMsgReal( $part1, $args, true );
1727 # Check for NS: (namespace expansion)
1728 $mwNs = MagicWord
::get( MAG_NS
);
1729 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1730 if ( intval( $part1 ) ) {
1731 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
1734 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1735 if ( !is_null( $index ) ) {
1736 $text = $linestart . $wgContLang->getNsText( $index );
1743 # LOCALURL and LOCALURLE
1745 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1746 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1748 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1749 $func = 'getLocalURL';
1750 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1751 $func = 'escapeLocalURL';
1756 if ( $func !== '' ) {
1757 $title = Title
::newFromText( $part1 );
1758 if ( !is_null( $title ) ) {
1760 $text = $linestart . $title->$func( $args[0] );
1762 $text = $linestart . $title->$func();
1770 if ( !$found && $argc == 1 ) {
1771 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
1772 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
1773 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
1778 # Template table test
1780 # Did we encounter this template already? If yes, it is in the cache
1781 # and we need to check for loops.
1782 if ( !$found && isset( $this->mTemplates
[$part1] ) ) {
1783 # set $text to cached message.
1784 $text = $linestart . $this->mTemplates
[$part1];
1787 # Infinite loop test
1788 if ( isset( $this->mTemplatePath
[$part1] ) ) {
1791 $text .= '<!-- WARNING: template loop detected -->';
1795 # Load from database
1796 $itcamefromthedatabase = false;
1798 $title = Title
::newFromText( $part1, NS_TEMPLATE
);
1799 if ( !is_null( $title ) && !$title->isExternal() ) {
1800 # Check for excessive inclusion
1801 $dbk = $title->getPrefixedDBkey();
1802 if ( $this->incrementIncludeCount( $dbk ) ) {
1803 # This should never be reached.
1804 $article = new Article( $title );
1805 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1806 if ( $articleContent !== false ) {
1808 $text = $linestart . $articleContent;
1809 $itcamefromthedatabase = true;
1813 # If the title is valid but undisplayable, make a link to it
1814 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1815 $text = $linestart . '[['.$title->getPrefixedText().']]';
1819 # Template cache array insertion
1820 $this->mTemplates
[$part1] = $text;
1824 # Recursive parsing, escaping and link table handling
1825 # Only for HTML output
1826 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1827 $text = wfEscapeWikiText( $text );
1828 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
1829 # Clean up argument array
1830 $assocArgs = array();
1832 foreach( $args as $arg ) {
1833 $eqpos = strpos( $arg, '=' );
1834 if ( $eqpos === false ) {
1835 $assocArgs[$index++
] = $arg;
1837 $name = trim( substr( $arg, 0, $eqpos ) );
1838 $value = trim( substr( $arg, $eqpos+
1 ) );
1839 if ( $value === false ) {
1842 if ( $name !== false ) {
1843 $assocArgs[$name] = $value;
1848 # Add a new element to the templace recursion path
1849 $this->mTemplatePath
[$part1] = 1;
1851 $text = $this->strip( $text, $this->mStripState
);
1852 $text = $this->removeHTMLtags( $text );
1853 $text = $this->replaceVariables( $text, $assocArgs );
1855 # Resume the link cache and register the inclusion as a link
1856 if ( $this->mOutputType
== OT_HTML
&& !is_null( $title ) ) {
1857 $wgLinkCache->addLinkObj( $title );
1860 # If the template begins with a table or block-level
1861 # element, it should be treated as beginning a new line.
1862 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
1863 $text = "\n" . $text;
1867 # Empties the template path
1868 $this->mTemplatePath
= array();
1872 # replace ==section headers==
1873 # XXX this needs to go away once we have a better parser.
1874 if ( $this->mOutputType
!= OT_WIKI
&& $itcamefromthedatabase ) {
1875 if( !is_null( $title ) )
1876 $encodedname = base64_encode($title->getPrefixedDBkey());
1878 $encodedname = base64_encode("");
1879 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
1880 PREG_SPLIT_DELIM_CAPTURE
);
1883 for( $i = 0; $i < count($m); $i +
= 2 ) {
1885 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
1887 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
1891 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
1892 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
1893 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
1900 # Empties the template path
1901 $this->mTemplatePath
= array();
1910 * Triple brace replacement -- used for template arguments
1913 function argSubstitution( $matches ) {
1914 $arg = trim( $matches[1] );
1915 $text = $matches[0];
1916 $inputArgs = end( $this->mArgStack
);
1918 if ( array_key_exists( $arg, $inputArgs ) ) {
1919 $text = $inputArgs[$arg];
1926 * Returns true if the function is allowed to include this entity
1929 function incrementIncludeCount( $dbk ) {
1930 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1931 $this->mIncludeCount
[$dbk] = 0;
1933 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1942 * Cleans up HTML, removes dangerous tags and attributes, and
1943 * removes HTML comments
1946 function removeHTMLtags( $text ) {
1947 global $wgUseTidy, $wgUserHtml;
1948 $fname = 'Parser::removeHTMLtags';
1949 wfProfileIn( $fname );
1952 $htmlpairs = array( # Tags that must be closed
1953 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
1954 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
1955 'strike', 'strong', 'tt', 'var', 'div', 'center',
1956 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
1957 'ruby', 'rt' , 'rb' , 'rp', 'p'
1959 $htmlsingle = array(
1960 'br', 'hr', 'li', 'dt', 'dd'
1962 $htmlnest = array( # Tags that can be nested--??
1963 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
1964 'dl', 'font', 'big', 'small', 'sub', 'sup'
1966 $tabletags = array( # Can only appear inside table
1970 $htmlpairs = array();
1971 $htmlsingle = array();
1972 $htmlnest = array();
1973 $tabletags = array();
1976 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1977 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1979 $htmlattrs = $this->getHTMLattrs () ;
1981 # Remove HTML comments
1982 $text = $this->removeHTMLcomments( $text );
1984 $bits = explode( '<', $text );
1985 $text = array_shift( $bits );
1987 $tagstack = array(); $tablestack = array();
1988 foreach ( $bits as $x ) {
1989 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
1990 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
1992 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1993 error_reporting( $prev );
1996 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2000 if ( ! in_array( $t, $htmlsingle ) &&
2001 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
2002 @array_push
( $tagstack, $ot );
2005 if ( $t == 'table' ) {
2006 $tagstack = array_pop( $tablestack );
2011 # Keep track for later
2012 if ( in_array( $t, $tabletags ) &&
2013 ! in_array( 'table', $tagstack ) ) {
2015 } else if ( in_array( $t, $tagstack ) &&
2016 ! in_array ( $t , $htmlnest ) ) {
2018 } else if ( ! in_array( $t, $htmlsingle ) ) {
2019 if ( $t == 'table' ) {
2020 array_push( $tablestack, $tagstack );
2021 $tagstack = array();
2023 array_push( $tagstack, $t );
2025 # Strip non-approved attributes from the tag
2026 $newparams = $this->fixTagAttributes($params);
2030 $rest = str_replace( '>', '>', $rest );
2031 $text .= "<$slash$t $newparams$brace$rest";
2035 $text .= '<' . str_replace( '>', '>', $x);
2037 # Close off any remaining tags
2038 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2040 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2043 # this might be possible using tidy itself
2044 foreach ( $bits as $x ) {
2045 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2047 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2048 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2049 $newparams = $this->fixTagAttributes($params);
2050 $rest = str_replace( '>', '>', $rest );
2051 $text .= "<$slash$t $newparams$brace$rest";
2053 $text .= '<' . str_replace( '>', '>', $x);
2057 wfProfileOut( $fname );
2062 * Remove '<!--', '-->', and everything between.
2063 * To avoid leaving blank lines, when a comment is both preceded
2064 * and followed by a newline (ignoring spaces), trim leading and
2065 * trailing spaces and one of the newlines.
2069 function removeHTMLcomments( $text ) {
2070 $fname='Parser::removeHTMLcomments';
2071 wfProfileIn( $fname );
2072 while (($start = strpos($text, '<!--')) !== false) {
2073 $end = strpos($text, '-->', $start +
4);
2074 if ($end === false) {
2075 # Unterminated comment; bail out
2081 # Trim space and newline if the comment is both
2082 # preceded and followed by a newline
2083 $spaceStart = max($start - 1, 0);
2084 $spaceLen = $end - $spaceStart;
2085 while (substr($text, $spaceStart, 1) === ' ' && $spaceStart > 0) {
2089 while (substr($text, $spaceStart +
$spaceLen, 1) === ' ')
2091 if (substr($text, $spaceStart, 1) === "\n" and substr($text, $spaceStart +
$spaceLen, 1) === "\n") {
2092 # Remove the comment, leading and trailing
2093 # spaces, and leave only one newline.
2094 $text = substr_replace($text, "\n", $spaceStart, $spaceLen +
1);
2097 # Remove just the comment.
2098 $text = substr_replace($text, '', $start, $end - $start);
2101 wfProfileOut( $fname );
2106 * This function accomplishes several tasks:
2107 * 1) Auto-number headings if that option is enabled
2108 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2109 * 3) Add a Table of contents on the top for users who have enabled the option
2110 * 4) Auto-anchor headings
2112 * It loops through all headlines, collects the necessary data, then splits up the
2113 * string and re-inserts the newly formatted headlines.
2116 /* private */ function formatHeadings( $text, $isMain=true ) {
2117 global $wgInputEncoding, $wgMaxTocLevel, $wgContLang, $wgLinkHolders;
2119 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2120 $doShowToc = $this->mOptions
->getShowToc();
2121 $forceTocHere = false;
2122 if( !$this->mTitle
->userCanEdit() ) {
2124 $rightClickHack = 0;
2126 $showEditLink = $this->mOptions
->getEditSection();
2127 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
2130 # Inhibit editsection links if requested in the page
2131 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2132 if( $esw->matchAndRemove( $text ) ) {
2135 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2137 $mw =& MagicWord
::get( MAG_NOTOC
);
2138 if( $mw->matchAndRemove( $text ) ) {
2142 # never add the TOC to the Main Page. This is an entry page that should not
2143 # be more than 1-2 screens large anyway
2144 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
2148 # Get all headlines for numbering them and adding funky stuff like [edit]
2149 # links - this is for later, but we need the number of headlines right now
2150 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
2152 # if there are fewer than 4 headlines in the article, do not show TOC
2153 if( $numMatches < 4 ) {
2157 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2158 # override above conditions and always show TOC at that place
2159 $mw =& MagicWord
::get( MAG_TOC
);
2160 if ($mw->match( $text ) ) {
2162 $forceTocHere = true;
2164 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2165 # override above conditions and always show TOC above first header
2166 $mw =& MagicWord
::get( MAG_FORCETOC
);
2167 if ($mw->matchAndRemove( $text ) ) {
2174 # We need this to perform operations on the HTML
2175 $sk =& $this->mOptions
->getSkin();
2179 $sectionCount = 0; # headlineCount excluding template sections
2181 # Ugh .. the TOC should have neat indentation levels which can be
2182 # passed to the skin functions. These are determined here
2187 $sublevelCount = array();
2190 foreach( $matches[3] as $headline ) {
2192 $templatetitle = "";
2193 $templatesection = 0;
2195 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2197 $templatetitle = base64_decode($mat[1]);
2198 $templatesection = 1 +
(int)base64_decode($mat[2]);
2199 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2204 $prevlevel = $level;
2206 $level = $matches[1][$headlineCount];
2207 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2208 # reset when we enter a new level
2209 $sublevelCount[$level] = 0;
2210 $toc .= $sk->tocIndent( $level - $prevlevel );
2211 $toclevel +
= $level - $prevlevel;
2213 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2214 # reset when we step back a level
2215 $sublevelCount[$level+
1]=0;
2216 $toc .= $sk->tocUnindent( $prevlevel - $level );
2217 $toclevel -= $prevlevel - $level;
2219 # count number of headlines for each level
2220 @$sublevelCount[$level]++
;
2221 if( $doNumberHeadings ||
$doShowToc ) {
2223 for( $i = 1; $i <= $level; $i++
) {
2224 if( !empty( $sublevelCount[$i] ) ) {
2228 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2234 # The canonized header is a version of the header text safe to use for links
2235 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2236 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2237 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2239 # Remove link placeholders by the link text.
2240 # <!--LINK number-->
2242 # link text with suffix
2243 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2244 "\$wgLinkHolders['texts'][\$1]",
2245 $canonized_headline );
2248 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2249 $tocline = trim( $canonized_headline );
2250 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2251 $replacearray = array(
2255 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2256 $refer[$headlineCount] = $canonized_headline;
2258 # count how many in assoc. array so we can track dupes in anchors
2259 @$refers[$canonized_headline]++
;
2260 $refcount[$headlineCount]=$refers[$canonized_headline];
2262 # Prepend the number to the heading text
2264 if( $doNumberHeadings ||
$doShowToc ) {
2265 $tocline = $numbering . ' ' . $tocline;
2267 # Don't number the heading if it is the only one (looks silly)
2268 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2269 # the two are different if the line contains a link
2270 $headline=$numbering . ' ' . $headline;
2274 # Create the anchor for linking from the TOC to the section
2275 $anchor = $canonized_headline;
2276 if($refcount[$headlineCount] > 1 ) {
2277 $anchor .= '_' . $refcount[$headlineCount];
2279 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2280 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2282 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2283 if ( empty( $head[$headlineCount] ) ) {
2284 $head[$headlineCount] = '';
2287 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2289 $head[$headlineCount] .= $sk->editSectionLink($sectionCount+
1);
2292 # Add the edit section span
2293 if( $rightClickHack ) {
2295 $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
2297 $headline = $sk->editSectionScript($sectionCount+
1,$headline);
2300 # give headline the correct <h#> tag
2301 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2309 $toclines = $headlineCount;
2310 $toc .= $sk->tocUnindent( $toclevel );
2311 $toc = $sk->tocTable( $toc );
2314 # split up and insert constructed headlines
2316 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2319 foreach( $blocks as $block ) {
2320 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2321 # This is the [edit] link that appears for the top block of text when
2322 # section editing is enabled
2324 # Disabled because it broke block formatting
2325 # For example, a bullet point in the top line
2326 # $full .= $sk->editSectionLink(0);
2329 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2330 # Top anchor now in skin
2334 if( !empty( $head[$i] ) ) {
2340 $mw =& MagicWord
::get( MAG_TOC
);
2341 return $mw->replace( $toc, $full );
2348 * Return an HTML link for the "ISBN 123456" text
2351 function magicISBN( $text ) {
2353 $fname = 'Parser::magicISBN';
2354 wfProfileIn( $fname );
2356 $a = split( 'ISBN ', ' '.$text );
2357 if ( count ( $a ) < 2 ) {
2358 wfProfileOut( $fname );
2361 $text = substr( array_shift( $a ), 1);
2362 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2364 foreach ( $a as $x ) {
2365 $isbn = $blank = '' ;
2366 while ( ' ' == $x{0} ) {
2368 $x = substr( $x, 1 );
2370 if ( $x == '' ) { # blank isbn
2371 $text .= "ISBN $blank";
2374 while ( strstr( $valid, $x{0} ) != false ) {
2376 $x = substr( $x, 1 );
2378 $num = str_replace( '-', '', $isbn );
2379 $num = str_replace( ' ', '', $num );
2382 $text .= "ISBN $blank$x";
2384 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2385 $text .= '<a href="' .
2386 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2387 "\" class=\"internal\">ISBN $isbn</a>";
2391 wfProfileOut( $fname );
2396 * Return an HTML link for the "GEO ..." text
2399 function magicGEO( $text ) {
2400 global $wgLang, $wgUseGeoMode;
2401 $fname = 'Parser::magicGEO';
2402 wfProfileIn( $fname );
2404 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2405 $directions = array ( 'N' => 'North' , 'S' => 'South' , 'E' => 'East' , 'W' => 'West' ) ;
2406 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2407 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2408 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2409 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2411 $a = split( 'GEO ', ' '.$text );
2412 if ( count ( $a ) < 2 ) {
2413 wfProfileOut( $fname );
2416 $text = substr( array_shift( $a ), 1);
2417 $valid = '0123456789.+-:';
2419 foreach ( $a as $x ) {
2420 $geo = $blank = '' ;
2421 while ( ' ' == $x{0} ) {
2423 $x = substr( $x, 1 );
2425 while ( strstr( $valid, $x{0} ) != false ) {
2427 $x = substr( $x, 1 );
2429 $num = str_replace( '+', '', $geo );
2430 $num = str_replace( ' ', '', $num );
2432 if ( '' == $num ||
count ( explode ( ':' , $num , 3 ) ) < 2 ) {
2433 $text .= "GEO $blank$x";
2435 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2436 $text .= '<a href="' .
2437 $titleObj->escapeLocalUrl( 'coordinates='.$num ) .
2438 "\" class=\"internal\">GEO $geo</a>";
2442 wfProfileOut( $fname );
2447 * Return an HTML link for the "RFC 1234" text
2449 * @param string $text text to be processed
2451 function magicRFC( $text ) {
2454 $valid = '0123456789';
2457 $a = split( 'RFC ', ' '.$text );
2458 if ( count ( $a ) < 2 ) return $text;
2459 $text = substr( array_shift( $a ), 1);
2461 /* Check if RFC keyword is preceed by [[.
2462 * This test is made here cause of the array_shift above
2463 * that prevent the test to be done in the foreach.
2465 if(substr($text, -2) == '[[') { $internal = true; }
2467 foreach ( $a as $x ) {
2468 /* token might be empty if we have RFC RFC 1234 */
2474 $rfc = $blank = '' ;
2476 /** remove and save whitespaces in $blank */
2477 while ( $x{0} == ' ' ) {
2479 $x = substr( $x, 1 );
2482 /** remove and save the rfc number in $rfc */
2483 while ( strstr( $valid, $x{0} ) != false ) {
2485 $x = substr( $x, 1 );
2489 /* call back stripped spaces*/
2490 $text .= "RFC $blank$x";
2491 } elseif( $internal) {
2493 $text .= "RFC $rfc$x";
2495 /* build the external link*/
2496 $url = wfmsg( 'rfcurl' );
2497 $url = str_replace( '$1', $rfc, $url);
2498 $sk =& $this->mOptions
->getSkin();
2499 $la = $sk->getExternalLinkAttributes( $url, 'RFC '.$rfc );
2500 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2503 /* Check if the next RFC keyword is preceed by [[ */
2504 $internal = (substr($x,-2) == '[[');
2510 * Transform wiki markup when saving a page by doing \r\n -> \n
2511 * conversion, substitting signatures, {{subst:}} templates, etc.
2513 * @param string $text the text to transform
2514 * @param Title &$title the Title object for the current article
2515 * @param User &$user the User object describing the current user
2516 * @param ParserOptions $options parsing options
2517 * @param bool $clearState whether to clear the parser state first
2518 * @return string the altered wiki markup
2521 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2522 $this->mOptions
= $options;
2523 $this->mTitle
=& $title;
2524 $this->mOutputType
= OT_WIKI
;
2526 if ( $clearState ) {
2527 $this->clearState();
2530 $stripState = false;
2534 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2538 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2539 "/<br *?>/i" => "<br />",
2541 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2543 $text = $this->strip( $text, $stripState, false );
2544 $text = $this->pstPass2( $text, $user );
2545 $text = $this->unstrip( $text, $stripState );
2546 $text = $this->unstripNoWiki( $text, $stripState );
2551 * Pre-save transform helper function
2554 function pstPass2( $text, &$user ) {
2555 global $wgLang, $wgContLang, $wgLocaltimezone, $wgCurParser;
2557 # Variable replacement
2558 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2559 $text = $this->replaceVariables( $text );
2563 $n = $user->getName();
2564 $k = $user->getOption( 'nickname' );
2565 if ( '' == $k ) { $k = $n; }
2566 if(isset($wgLocaltimezone)) {
2567 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2569 /* Note: this is an ugly timezone hack for the European wikis */
2570 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
2571 ' (' . date( 'T' ) . ')';
2572 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2574 $text = preg_replace( '/~~~~~/', $d, $text );
2575 $text = preg_replace( '/~~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]] $d", $text );
2576 $text = preg_replace( '/~~~/', '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]]", $text );
2578 # Context links: [[|name]] and [[name (context)|]]
2580 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2581 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2582 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2583 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2585 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2586 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2587 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2588 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2590 $t = $this->mTitle
->getText();
2591 if ( preg_match( $conpat, $t, $m ) ) {
2594 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2595 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2596 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2598 if ( '' == $context ) {
2599 $text = preg_replace( $p2, '[[\\1]]', $text );
2601 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2604 # Trim trailing whitespace
2605 # MAG_END (__END__) tag allows for trailing
2606 # whitespace to be deliberately included
2607 $text = rtrim( $text );
2608 $mw =& MagicWord
::get( MAG_END
);
2609 $mw->matchAndRemove( $text );
2615 * Set up some variables which are usually set up in parse()
2616 * so that an external function can call some class members with confidence
2619 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2620 $this->mTitle
=& $title;
2621 $this->mOptions
= $options;
2622 $this->mOutputType
= $outputType;
2623 if ( $clearState ) {
2624 $this->clearState();
2629 * Transform a MediaWiki message by replacing magic variables.
2631 * @param string $text the text to transform
2632 * @param ParserOptions $options options
2633 * @return string the text with variables substituted
2636 function transformMsg( $text, $options ) {
2638 static $executing = false;
2640 # Guard against infinite recursion
2646 $this->mTitle
= $wgTitle;
2647 $this->mOptions
= $options;
2648 $this->mOutputType
= OT_MSG
;
2649 $this->clearState();
2650 $text = $this->replaceVariables( $text );
2657 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2658 * Callback will be called with the text within
2659 * Transform and return the text within
2662 function setHook( $tag, $callback ) {
2663 $oldVal = @$this->mTagHooks
[$tag];
2664 $this->mTagHooks
[$tag] = $callback;
2671 * @package MediaWiki
2675 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2676 var $mCacheTime; # Used in ParserCache
2678 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
2679 $containsOldMagic = false )
2681 $this->mText
= $text;
2682 $this->mLanguageLinks
= $languageLinks;
2683 $this->mCategoryLinks
= $categoryLinks;
2684 $this->mContainsOldMagic
= $containsOldMagic;
2685 $this->mCacheTime
= '';
2688 function getText() { return $this->mText
; }
2689 function getLanguageLinks() { return $this->mLanguageLinks
; }
2690 function getCategoryLinks() { return $this->mCategoryLinks
; }
2691 function getCacheTime() { return $this->mCacheTime
; }
2692 function containsOldMagic() { return $this->mContainsOldMagic
; }
2693 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2694 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2695 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2696 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2697 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
2699 function merge( $other ) {
2700 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2701 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2702 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2708 * Set options of the Parser
2710 * @package MediaWiki
2714 # All variables are private
2715 var $mUseTeX; # Use texvc to expand <math> tags
2716 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2717 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2718 var $mAllowExternalImages; # Allow external images inline
2719 var $mSkin; # Reference to the preferred skin
2720 var $mDateFormat; # Date format index
2721 var $mEditSection; # Create "edit section" links
2722 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2723 var $mNumberHeadings; # Automatically number headings
2724 var $mShowToc; # Show table of contents
2726 function getUseTeX() { return $this->mUseTeX
; }
2727 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2728 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2729 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2730 function getSkin() { return $this->mSkin
; }
2731 function getDateFormat() { return $this->mDateFormat
; }
2732 function getEditSection() { return $this->mEditSection
; }
2733 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
2734 function getNumberHeadings() { return $this->mNumberHeadings
; }
2735 function getShowToc() { return $this->mShowToc
; }
2737 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2738 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2739 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2740 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2741 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2742 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
2743 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
2744 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
2745 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
2747 function setSkin( &$x ) { $this->mSkin
=& $x; }
2749 # Get parser options
2750 /* static */ function newFromUser( &$user ) {
2751 $popts = new ParserOptions
;
2752 $popts->initialiseFromUser( $user );
2757 function initialiseFromUser( &$userInput ) {
2758 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2760 $fname = 'ParserOptions::initialiseFromUser';
2761 wfProfileIn( $fname );
2762 if ( !$userInput ) {
2764 $user->setLoaded( true );
2766 $user =& $userInput;
2769 $this->mUseTeX
= $wgUseTeX;
2770 $this->mUseDynamicDates
= $wgUseDynamicDates;
2771 $this->mInterwikiMagic
= $wgInterwikiMagic;
2772 $this->mAllowExternalImages
= $wgAllowExternalImages;
2773 wfProfileIn( $fname.'-skin' );
2774 $this->mSkin
=& $user->getSkin();
2775 wfProfileOut( $fname.'-skin' );
2776 $this->mDateFormat
= $user->getOption( 'date' );
2777 $this->mEditSection
= $user->getOption( 'editsection' );
2778 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
2779 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
2780 $this->mShowToc
= $user->getOption( 'showtoc' );
2781 wfProfileOut( $fname );
2787 # Regex callbacks, used in Parser::replaceVariables
2788 function wfBraceSubstitution( $matches ) {
2789 global $wgCurParser;
2790 return $wgCurParser->braceSubstitution( $matches );
2793 function wfArgSubstitution( $matches ) {
2794 global $wgCurParser;
2795 return $wgCurParser->argSubstitution( $matches );
2798 function wfVariableSubstitution( $matches ) {
2799 global $wgCurParser;
2800 return $wgCurParser->variableSubstitution( $matches );
2804 * Return the total number of articles
2806 function wfNumberOfArticles() {
2807 global $wgNumberOfArticles;
2810 return $wgNumberOfArticles;
2814 * Get various statistics from the database
2817 function wfLoadSiteStats() {
2818 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
2819 $fname = 'wfLoadSiteStats';
2821 if ( -1 != $wgNumberOfArticles ) return;
2822 $dbr =& wfGetDB( DB_SLAVE
);
2823 $s = $dbr->getArray( 'site_stats',
2824 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
2825 array( 'ss_row_id' => 1 ), $fname
2828 if ( $s === false ) {
2831 $wgTotalViews = $s->ss_total_views
;
2832 $wgTotalEdits = $s->ss_total_edits
;
2833 $wgNumberOfArticles = $s->ss_good_articles
;
2837 function wfEscapeHTMLTagsOnly( $in ) {
2839 array( '"', '>', '<' ),
2840 array( '"', '>', '<' ),