3 * File for Parser and related classes
10 require_once( 'Sanitizer.php' );
11 require_once( 'HttpFunctions.php' );
14 * Update this version number when the ParserOutput format
15 * changes in an incompatible way, so the parser cache
16 * can automatically discard old data.
18 define( 'MW_PARSER_VERSION', '1.5.0' );
21 * Variable substitution O(N^2) attack
23 * Without countermeasures, it would be possible to attack the parser by saving
24 * a page filled with a large number of inclusions of large pages. The size of
25 * the generated page would be proportional to the square of the input size.
26 * Hence, we limit the number of inclusions of any given page, thus bringing any
27 * attack back to O(N).
30 define( 'MAX_INCLUDE_REPEAT', 100 );
31 define( 'MAX_INCLUDE_SIZE', 1000000 ); // 1 Million
33 define( 'RLH_FOR_UPDATE', 1 );
35 # Allowed values for $mOutputType
36 define( 'OT_HTML', 1 );
37 define( 'OT_WIKI', 2 );
38 define( 'OT_MSG' , 3 );
40 # string parameter for extractTags which will cause it
41 # to strip HTML comments in addition to regular
42 # <XML>-style tags. This should not be anything we
43 # may want to use in wikisyntax
44 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
46 # prefix for escaping, used in two functions at least
47 define( 'UNIQ_PREFIX', 'NaodW29');
49 # Constants needed for external link processing
50 define( 'URL_PROTOCOLS', 'http:\/\/|https:\/\/|ftp:\/\/|irc:\/\/|gopher:\/\/|news:|mailto:' );
51 define( 'HTTP_PROTOCOLS', 'http:\/\/|https:\/\/' );
52 # Everything except bracket, space, or control characters
53 define( 'EXT_LINK_URL_CLASS', '[^]<>"\\x00-\\x20\\x7F]' );
55 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
56 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
57 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
58 define( 'EXT_LINK_BRACKETED', '/\[(\b('.URL_PROTOCOLS
.')'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
59 define( 'EXT_IMAGE_REGEX',
60 '/^('.HTTP_PROTOCOLS
.')'. # Protocol
61 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
62 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
68 * Processes wiki markup
71 * There are three main entry points into the Parser class:
73 * produces HTML output
75 * produces altered wiki markup.
77 * performs brace substitution on MediaWiki messages
80 * objects: $wgLang, $wgLinkCache
82 * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
85 * $wgUseTex*, $wgUseDynamicDates*, $wgInterwikiMagic*,
86 * $wgNamespacesWithSubpages, $wgAllowExternalImages*,
87 * $wgLocaltimezone, $wgAllowSpecialInclusion*
89 * * only within ParserOptions
102 # Cleared with clearState():
103 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
104 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
105 var $mInterwikiLinkHolders, $mLinkHolders;
108 var $mOptions, $mTitle, $mOutputType,
109 $mTemplates, // cache of already loaded templates, avoids
110 // multiple SQL queries for the same string
111 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
112 // in this path. Used for loop detection.
114 var $mIWTransData = array();
125 $this->mTemplates
= array();
126 $this->mTemplatePath
= array();
127 $this->mTagHooks
= array();
136 function clearState() {
137 $this->mOutput
= new ParserOutput
;
138 $this->mAutonumber
= 0;
139 $this->mLastSection
= '';
140 $this->mDTopen
= false;
141 $this->mVariables
= false;
142 $this->mIncludeCount
= array();
143 $this->mStripState
= array();
144 $this->mArgStack
= array();
145 $this->mInPre
= false;
146 $this->mInterwikiLinkHolders
= array(
150 $this->mLinkHolders
= array(
151 'namespaces' => array(),
153 'queries' => array(),
160 * First pass--just handle <nowiki> sections, pass the rest off
161 * to internalParse() which does all the real work.
164 * @param string $text Text we want to parse
165 * @param Title &$title A title object
166 * @param array $options
167 * @param boolean $linestart
168 * @param boolean $clearState
169 * @return ParserOutput a ParserOutput
171 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
172 global $wgUseTidy, $wgContLang;
173 $fname = 'Parser::parse';
174 wfProfileIn( $fname );
180 $this->mOptions
= $options;
181 $this->mTitle
=& $title;
182 $this->mOutputType
= OT_HTML
;
184 $this->mStripState
= NULL;
186 //$text = $this->strip( $text, $this->mStripState );
187 // VOODOO MAGIC FIX! Sometimes the above segfaults in PHP5.
188 $x =& $this->mStripState
;
190 wfRunHooks( 'ParserBeforeStrip', array( &$this, &$text, &$x ) );
191 $text = $this->strip( $text, $x );
192 wfRunHooks( 'ParserAfterStrip', array( &$this, &$text, &$x ) );
194 $text = $this->internalParse( $text );
196 $text = $this->unstrip( $text, $this->mStripState
);
198 # Clean up special characters, only run once, next-to-last before doBlockLevels
200 # french spaces, last one Guillemet-left
201 # only if there is something before the space
202 '/(.) (?=\\?|:|;|!|\\302\\273)/' => '\\1 \\2',
203 # french spaces, Guillemet-right
204 '/(\\302\\253) /' => '\\1 ',
205 '/<center *>/i' => '<div class="center">',
206 '/<\\/center *>/i' => '</div>',
208 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
211 $text = $this->doBlockLevels( $text, $linestart );
213 $this->replaceLinkHolders( $text );
215 # the position of the convert() call should not be changed. it
216 # assumes that the links are all replaces and the only thing left
217 # is the <nowiki> mark.
218 $text = $wgContLang->convert($text);
219 $this->mOutput
->setTitleText($wgContLang->getParsedTitle());
221 $text = $this->unstripNoWiki( $text, $this->mStripState
);
223 wfRunHooks( 'ParserBeforeTidy', array( &$this, &$text ) );
225 $text = Sanitizer
::normalizeCharReferences( $text );
228 $text = Parser
::tidy($text);
231 wfRunHooks( 'ParserAfterTidy', array( &$this, &$text ) );
233 $this->mOutput
->setText( $text );
234 wfProfileOut( $fname );
235 return $this->mOutput
;
239 * Get a random string
244 function getRandomString() {
245 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
249 * Replaces all occurrences of <$tag>content</$tag> in the text
250 * with a random marker and returns the new text. the output parameter
251 * $content will be an associative array filled with data on the form
252 * $unique_marker => content.
254 * If $content is already set, the additional entries will be appended
255 * If $tag is set to STRIP_COMMENTS, the function will extract
256 * <!-- HTML comments -->
261 function extractTagsAndParams($tag, $text, &$content, &$tags, &$params, $uniq_prefix = ''){
262 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
277 if( $tag == STRIP_COMMENTS
) {
281 $start = "/<$tag(\\s+[^>]*|\\s*)>/i";
282 $end = "/<\\/$tag\\s*>/i";
285 while ( '' != $text ) {
286 $p = preg_split( $start, $text, 2, PREG_SPLIT_DELIM_CAPTURE
);
288 if( count( $p ) < 3 ) {
294 $marker = $rnd . sprintf('%08X', $n++
);
295 $stripped .= $marker;
297 $tags[$marker] = "<$tag$attributes>";
298 $params[$marker] = Sanitizer
::decodeTagAttributes( $attributes );
300 $q = preg_split( $end, $inside, 2 );
301 $content[$marker] = $q[0];
302 if( count( $q ) < 1 ) {
303 # No end tag -- let it run out to the end of the text.
313 * Wrapper function for extractTagsAndParams
314 * for cases where $tags and $params isn't needed
315 * i.e. where tags will never have params, like <nowiki>
320 function extractTags( $tag, $text, &$content, $uniq_prefix = '' ) {
321 $dummy_tags = array();
322 $dummy_params = array();
324 return Parser
::extractTagsAndParams( $tag, $text, $content,
325 $dummy_tags, $dummy_params, $uniq_prefix );
329 * Strips and renders nowiki, pre, math, hiero
330 * If $render is set, performs necessary rendering operations on plugins
331 * Returns the text, and fills an array with data needed in unstrip()
332 * If the $state is already a valid strip state, it adds to the state
334 * @param bool $stripcomments when set, HTML comments <!-- like this -->
335 * will be stripped in addition to other tags. This is important
336 * for section editing, where these comments cause confusion when
337 * counting the sections in the wikisource
341 function strip( $text, &$state, $stripcomments = false ) {
342 $render = ($this->mOutputType
== OT_HTML
);
343 $html_content = array();
344 $nowiki_content = array();
345 $math_content = array();
346 $pre_content = array();
347 $comment_content = array();
348 $ext_content = array();
350 $ext_params = array();
351 $gallery_content = array();
353 # Replace any instances of the placeholders
354 $uniq_prefix = UNIQ_PREFIX
;
355 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
360 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
361 foreach( $html_content as $marker => $content ) {
363 # Raw and unchecked for validity.
364 $html_content[$marker] = $content;
366 $html_content[$marker] = '<html>'.$content.'</html>';
372 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
373 foreach( $nowiki_content as $marker => $content ) {
375 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
377 $nowiki_content[$marker] = '<nowiki>'.$content.'</nowiki>';
382 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
383 foreach( $math_content as $marker => $content ){
385 if( $this->mOptions
->getUseTeX() ) {
386 $math_content[$marker] = renderMath( $content );
388 $math_content[$marker] = '<math>'.$content.'<math>';
391 $math_content[$marker] = '<math>'.$content.'</math>';
396 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
397 foreach( $pre_content as $marker => $content ){
399 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
401 $pre_content[$marker] = '<pre>'.$content.'</pre>';
406 $text = Parser
::extractTags('gallery', $text, $gallery_content, $uniq_prefix);
407 foreach( $gallery_content as $marker => $content ) {
408 require_once( 'ImageGallery.php' );
410 $gallery_content[$marker] = Parser
::renderImageGallery( $content );
412 $gallery_content[$marker] = '<gallery>'.$content.'</gallery>';
418 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
419 foreach( $comment_content as $marker => $content ){
420 $comment_content[$marker] = '<!--'.$content.'-->';
425 foreach ( $this->mTagHooks
as $tag => $callback ) {
426 $ext_content[$tag] = array();
427 $text = Parser
::extractTagsAndParams( $tag, $text, $ext_content[$tag],
428 $ext_tags[$tag], $ext_params[$tag], $uniq_prefix );
429 foreach( $ext_content[$tag] as $marker => $content ) {
430 $full_tag = $ext_tags[$tag][$marker];
431 $params = $ext_params[$tag][$marker];
433 $ext_content[$tag][$marker] = $callback( $content, $params );
435 $ext_content[$tag][$marker] = "$full_tag$content</$tag>";
440 # Merge state with the pre-existing state, if there is one
442 $state['html'] = $state['html'] +
$html_content;
443 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
444 $state['math'] = $state['math'] +
$math_content;
445 $state['pre'] = $state['pre'] +
$pre_content;
446 $state['comment'] = $state['comment'] +
$comment_content;
447 $state['gallery'] = $state['gallery'] +
$gallery_content;
449 foreach( $ext_content as $tag => $array ) {
450 if ( array_key_exists( $tag, $state ) ) {
451 $state[$tag] = $state[$tag] +
$array;
456 'html' => $html_content,
457 'nowiki' => $nowiki_content,
458 'math' => $math_content,
459 'pre' => $pre_content,
460 'comment' => $comment_content,
461 'gallery' => $gallery_content,
468 * restores pre, math, and hiero removed by strip()
470 * always call unstripNoWiki() after this one
473 function unstrip( $text, &$state ) {
474 # Must expand in reverse order, otherwise nested tags will be corrupted
475 $contentDict = end( $state );
476 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
477 if( key($state) != 'nowiki' && key($state) != 'html') {
478 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
479 $text = str_replace( key( $contentDict ), $content, $text );
488 * always call this after unstrip() to preserve the order
492 function unstripNoWiki( $text, &$state ) {
493 # Must expand in reverse order, otherwise nested tags will be corrupted
494 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
495 $text = str_replace( key( $state['nowiki'] ), $content, $text );
500 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
501 $text = str_replace( key( $state['html'] ), $content, $text );
509 * Add an item to the strip state
510 * Returns the unique tag which must be inserted into the stripped text
511 * The tag will be replaced with the original text in unstrip()
515 function insertStripItem( $text, &$state ) {
516 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
523 'comment' => array(),
524 'gallery' => array(),
527 $state['item'][$rnd] = $text;
532 * Interface with html tidy, used if $wgUseTidy = true.
533 * If tidy isn't able to correct the markup, the original will be
534 * returned in all its glory with a warning comment appended.
536 * Either the external tidy program or the in-process tidy extension
537 * will be used depending on availability. Override the default
538 * $wgTidyInternal setting to disable the internal if it's not working.
540 * @param string $text Hideous HTML input
541 * @return string Corrected HTML output
545 function tidy( $text ) {
546 global $wgTidyInternal;
547 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
548 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
549 '<head><title>test</title></head><body>'.$text.'</body></html>';
550 if( $wgTidyInternal ) {
551 $correctedtext = Parser
::internalTidy( $wrappedtext );
553 $correctedtext = Parser
::externalTidy( $wrappedtext );
555 if( is_null( $correctedtext ) ) {
556 wfDebug( "Tidy error detected!\n" );
557 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
559 return $correctedtext;
563 * Spawn an external HTML tidy process and get corrected markup back from it.
568 function externalTidy( $text ) {
569 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
570 $fname = 'Parser::externalTidy';
571 wfProfileIn( $fname );
576 $descriptorspec = array(
577 0 => array('pipe', 'r'),
578 1 => array('pipe', 'w'),
579 2 => array('file', '/dev/null', 'a')
582 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts$opts", $descriptorspec, $pipes);
583 if (is_resource($process)) {
584 fwrite($pipes[0], $text);
586 while (!feof($pipes[1])) {
587 $cleansource .= fgets($pipes[1], 1024);
590 proc_close($process);
593 wfProfileOut( $fname );
595 if( $cleansource == '' && $text != '') {
596 // Some kind of error happened, so we couldn't get the corrected text.
597 // Just give up; we'll use the source text and append a warning.
605 * Use the HTML tidy PECL extension to use the tidy library in-process,
606 * saving the overhead of spawning a new process. Currently written to
607 * the PHP 4.3.x version of the extension, may not work on PHP 5.
609 * 'pear install tidy' should be able to compile the extension module.
614 function internalTidy( $text ) {
616 $fname = 'Parser::internalTidy';
617 wfProfileIn( $fname );
619 tidy_load_config( $wgTidyConf );
620 tidy_set_encoding( 'utf8' );
621 tidy_parse_string( $text );
623 if( tidy_get_status() == 2 ) {
624 // 2 is magic number for fatal error
625 // http://www.php.net/manual/en/function.tidy-get-status.php
628 $cleansource = tidy_get_output();
630 wfProfileOut( $fname );
635 * parse the wiki syntax used to render tables
639 function doTableStuff ( $t ) {
640 $fname = 'Parser::doTableStuff';
641 wfProfileIn( $fname );
643 $t = explode ( "\n" , $t ) ;
644 $td = array () ; # Is currently a td tag open?
645 $ltd = array () ; # Was it TD or TH?
646 $tr = array () ; # Is currently a tr tag open?
647 $ltr = array () ; # tr attributes
648 $indent_level = 0; # indent level of the table
649 foreach ( $t AS $k => $x )
652 $fc = substr ( $x , 0 , 1 ) ;
653 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) ) {
654 $indent_level = strlen( $matches[1] );
655 $t[$k] = str_repeat( '<dl><dd>', $indent_level ) .
656 '<table' . Sanitizer
::fixTagAttributes ( $matches[2], 'table' ) . '>' ;
657 array_push ( $td , false ) ;
658 array_push ( $ltd , '' ) ;
659 array_push ( $tr , false ) ;
660 array_push ( $ltr , '' ) ;
662 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
663 else if ( '|}' == substr ( $x , 0 , 2 ) ) {
664 $z = "</table>" . substr ( $x , 2);
665 $l = array_pop ( $ltd ) ;
666 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
667 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
669 $t[$k] = $z . str_repeat( '</dd></dl>', $indent_level );
671 else if ( '|-' == substr ( $x , 0 , 2 ) ) { # Allows for |---------------
672 $x = substr ( $x , 1 ) ;
673 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
675 $l = array_pop ( $ltd ) ;
676 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
677 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
680 array_push ( $tr , false ) ;
681 array_push ( $td , false ) ;
682 array_push ( $ltd , '' ) ;
683 array_push ( $ltr , Sanitizer
::fixTagAttributes ( $x, 'tr' ) ) ;
685 else if ( '|' == $fc ||
'!' == $fc ||
'|+' == substr ( $x , 0 , 2 ) ) { # Caption
687 if ( '|+' == substr ( $x , 0 , 2 ) ) {
689 $x = substr ( $x , 1 ) ;
691 $after = substr ( $x , 1 ) ;
692 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
693 $after = explode ( '||' , $after ) ;
696 # Loop through each table cell
697 foreach ( $after AS $theline )
702 $tra = array_pop ( $ltr ) ;
703 if ( !array_pop ( $tr ) ) $z = '<tr'.$tra.">\n" ;
704 array_push ( $tr , true ) ;
705 array_push ( $ltr , '' ) ;
708 $l = array_pop ( $ltd ) ;
709 if ( array_pop ( $td ) ) $z = '</'.$l.'>' . $z ;
710 if ( $fc == '|' ) $l = 'td' ;
711 else if ( $fc == '!' ) $l = 'th' ;
712 else if ( $fc == '+' ) $l = 'caption' ;
714 array_push ( $ltd , $l ) ;
717 $y = explode ( '|' , $theline , 2 ) ;
718 # Note that a '|' inside an invalid link should not
719 # be mistaken as delimiting cell parameters
720 if ( strpos( $y[0], '[[' ) !== false ) {
721 $y = array ($theline);
723 if ( count ( $y ) == 1 )
724 $y = "{$z}<{$l}>{$y[0]}" ;
725 else $y = $y = "{$z}<{$l}".Sanitizer
::fixTagAttributes($y[0], $l).">{$y[1]}" ;
727 array_push ( $td , true ) ;
732 # Closing open td, tr && table
733 while ( count ( $td ) > 0 )
735 if ( array_pop ( $td ) ) $t[] = '</td>' ;
736 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
740 $t = implode ( "\n" , $t ) ;
741 wfProfileOut( $fname );
746 * Helper function for parse() that transforms wiki markup into
747 * HTML. Only called for $mOutputType == OT_HTML.
751 function internalParse( $text ) {
755 $fname = 'Parser::internalParse';
756 wfProfileIn( $fname );
758 $text = Sanitizer
::removeHTMLtags( $text, array( &$this, 'replaceVariables' ) );
759 $text = $this->replaceVariables( $text, $args );
761 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
763 $text = $this->doHeadings( $text );
764 if($this->mOptions
->getUseDynamicDates()) {
765 $df =& DateFormatter
::getInstance();
766 $text = $df->reformat( $this->mOptions
->getDateFormat(), $text );
768 $text = $this->doAllQuotes( $text );
769 $text = $this->replaceInternalLinks( $text );
770 $text = $this->replaceExternalLinks( $text );
772 # replaceInternalLinks may sometimes leave behind
773 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
774 $text = str_replace("http-noparse://","http://",$text);
776 $text = $this->doMagicLinks( $text );
777 $text = $this->doTableStuff( $text );
778 $text = $this->formatHeadings( $text, $isMain );
780 $regex = '/<!--IW_TRANSCLUDE (\d+)-->/';
781 $text = preg_replace_callback($regex, array(&$this, 'scarySubstitution'), $text);
783 wfProfileOut( $fname );
787 function scarySubstitution($matches) {
788 # return "[[".$matches[0]."]]";
789 return $this->mIWTransData
[(int)$matches[0]];
793 * Replace special strings like "ISBN xxx" and "RFC xxx" with
794 * magic external links.
798 function &doMagicLinks( &$text ) {
799 $text = $this->magicISBN( $text );
800 $text = $this->magicRFC( $text, 'RFC ', 'rfcurl' );
801 $text = $this->magicRFC( $text, 'PMID ', 'pubmedurl' );
806 * Parse ^^ tokens and return html
810 function doExponent( $text ) {
811 $fname = 'Parser::doExponent';
812 wfProfileIn( $fname );
813 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
814 wfProfileOut( $fname );
819 * Parse headers and return html
823 function doHeadings( $text ) {
824 $fname = 'Parser::doHeadings';
825 wfProfileIn( $fname );
826 for ( $i = 6; $i >= 1; --$i ) {
827 $h = substr( '======', 0, $i );
828 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
829 "<h{$i}>\\1</h{$i}>\\2", $text );
831 wfProfileOut( $fname );
836 * Replace single quotes with HTML markup
838 * @return string the altered text
840 function doAllQuotes( $text ) {
841 $fname = 'Parser::doAllQuotes';
842 wfProfileIn( $fname );
844 $lines = explode( "\n", $text );
845 foreach ( $lines as $line ) {
846 $outtext .= $this->doQuotes ( $line ) . "\n";
848 $outtext = substr($outtext, 0,-1);
849 wfProfileOut( $fname );
854 * Helper function for doAllQuotes()
857 function doQuotes( $text ) {
858 $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
859 if ( count( $arr ) == 1 )
863 # First, do some preliminary work. This may shift some apostrophes from
864 # being mark-up to being text. It also counts the number of occurrences
865 # of bold and italics mark-ups.
869 foreach ( $arr as $r )
871 if ( ( $i %
2 ) == 1 )
873 # If there are ever four apostrophes, assume the first is supposed to
874 # be text, and the remaining three constitute mark-up for bold text.
875 if ( strlen( $arr[$i] ) == 4 )
880 # If there are more than 5 apostrophes in a row, assume they're all
881 # text except for the last 5.
882 else if ( strlen( $arr[$i] ) > 5 )
884 $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
887 # Count the number of occurrences of bold and italics mark-ups.
888 # We are not counting sequences of five apostrophes.
889 if ( strlen( $arr[$i] ) == 2 ) $numitalics++
; else
890 if ( strlen( $arr[$i] ) == 3 ) $numbold++
; else
891 if ( strlen( $arr[$i] ) == 5 ) { $numitalics++
; $numbold++
; }
896 # If there is an odd number of both bold and italics, it is likely
897 # that one of the bold ones was meant to be an apostrophe followed
898 # by italics. Which one we cannot know for certain, but it is more
899 # likely to be one that has a single-letter word before it.
900 if ( ( $numbold %
2 == 1 ) && ( $numitalics %
2 == 1 ) )
903 $firstsingleletterword = -1;
904 $firstmultiletterword = -1;
906 foreach ( $arr as $r )
908 if ( ( $i %
2 == 1 ) and ( strlen( $r ) == 3 ) )
910 $x1 = substr ($arr[$i-1], -1);
911 $x2 = substr ($arr[$i-1], -2, 1);
913 if ($firstspace == -1) $firstspace = $i;
914 } else if ($x2 == ' ') {
915 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
917 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
923 # If there is a single-letter word, use it!
924 if ($firstsingleletterword > -1)
926 $arr [ $firstsingleletterword ] = "''";
927 $arr [ $firstsingleletterword-1 ] .= "'";
929 # If not, but there's a multi-letter word, use that one.
930 else if ($firstmultiletterword > -1)
932 $arr [ $firstmultiletterword ] = "''";
933 $arr [ $firstmultiletterword-1 ] .= "'";
935 # ... otherwise use the first one that has neither.
936 # (notice that it is possible for all three to be -1 if, for example,
937 # there is only one pentuple-apostrophe in the line)
938 else if ($firstspace > -1)
940 $arr [ $firstspace ] = "''";
941 $arr [ $firstspace-1 ] .= "'";
945 # Now let's actually convert our apostrophic mush to HTML!
954 if ($state == 'both')
961 if (strlen ($r) == 2)
964 { $output .= '</i>'; $state = ''; }
965 else if ($state == 'bi')
966 { $output .= '</i>'; $state = 'b'; }
967 else if ($state == 'ib')
968 { $output .= '</b></i><b>'; $state = 'b'; }
969 else if ($state == 'both')
970 { $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
971 else # $state can be 'b' or ''
972 { $output .= '<i>'; $state .= 'i'; }
974 else if (strlen ($r) == 3)
977 { $output .= '</b>'; $state = ''; }
978 else if ($state == 'bi')
979 { $output .= '</i></b><i>'; $state = 'i'; }
980 else if ($state == 'ib')
981 { $output .= '</b>'; $state = 'i'; }
982 else if ($state == 'both')
983 { $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
984 else # $state can be 'i' or ''
985 { $output .= '<b>'; $state .= 'b'; }
987 else if (strlen ($r) == 5)
990 { $output .= '</b><i>'; $state = 'i'; }
991 else if ($state == 'i')
992 { $output .= '</i><b>'; $state = 'b'; }
993 else if ($state == 'bi')
994 { $output .= '</i></b>'; $state = ''; }
995 else if ($state == 'ib')
996 { $output .= '</b></i>'; $state = ''; }
997 else if ($state == 'both')
998 { $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
999 else # ($state == '')
1000 { $buffer = ''; $state = 'both'; }
1005 # Now close all remaining tags. Notice that the order is important.
1006 if ($state == 'b' ||
$state == 'ib')
1008 if ($state == 'i' ||
$state == 'bi' ||
$state == 'ib')
1012 if ($state == 'both')
1013 $output .= '<b><i>'.$buffer.'</i></b>';
1019 * Replace external links
1021 * Note: this is all very hackish and the order of execution matters a lot.
1022 * Make sure to run maintenance/parserTests.php if you change this code.
1026 function replaceExternalLinks( $text ) {
1028 $fname = 'Parser::replaceExternalLinks';
1029 wfProfileIn( $fname );
1031 $sk =& $this->mOptions
->getSkin();
1033 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1035 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
1038 while ( $i<count( $bits ) ) {
1040 $protocol = $bits[$i++
];
1041 $text = $bits[$i++
];
1042 $trail = $bits[$i++
];
1044 # The characters '<' and '>' (which were escaped by
1045 # removeHTMLtags()) should not be included in
1046 # URLs, per RFC 2396.
1047 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1048 $text = substr($url, $m2[0][1]) . ' ' . $text;
1049 $url = substr($url, 0, $m2[0][1]);
1052 # If the link text is an image URL, replace it with an <img> tag
1053 # This happened by accident in the original parser, but some people used it extensively
1054 $img = $this->maybeMakeExternalImage( $text );
1055 if ( $img !== false ) {
1061 # Set linktype for CSS - if URL==text, link is essentially free
1062 $linktype = ($text == $url) ?
'free' : 'text';
1064 # No link text, e.g. [http://domain.tld/some.link]
1065 if ( $text == '' ) {
1066 # Autonumber if allowed
1067 if ( strpos( HTTP_PROTOCOLS
, str_replace('/','\/', $protocol) ) !== false ) {
1068 $text = '[' . ++
$this->mAutonumber
. ']';
1069 $linktype = 'autonumber';
1071 # Otherwise just use the URL
1072 $text = htmlspecialchars( $url );
1076 # Have link text, e.g. [http://domain.tld/some.link text]s
1078 list( $dtrail, $trail ) = Linker
::splitTrail( $trail );
1081 $text = $wgContLang->markNoConversion($text);
1083 # Replace & from obsolete syntax with &.
1084 # All HTML entities will be escaped by makeExternalLink()
1085 # or maybeMakeExternalImage()
1086 $url = str_replace( '&', '&', $url );
1088 # Process the trail (i.e. everything after this link up until start of the next link),
1089 # replacing any non-bracketed links
1090 $trail = $this->replaceFreeExternalLinks( $trail );
1093 # Use the encoded URL
1094 # This means that users can paste URLs directly into the text
1095 # Funny characters like ö aren't valid in URLs anyway
1096 # This was changed in August 2004
1097 $s .= $sk->makeExternalLink( $url, $text, false, $linktype ) . $dtrail . $trail;
1100 wfProfileOut( $fname );
1105 * Replace anything that looks like a URL with a link
1108 function replaceFreeExternalLinks( $text ) {
1110 $fname = 'Parser::replaceFreeExternalLinks';
1111 wfProfileIn( $fname );
1113 $bits = preg_split( '/(\b(?:'.URL_PROTOCOLS
.'))/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1114 $s = array_shift( $bits );
1117 $sk =& $this->mOptions
->getSkin();
1119 while ( $i < count( $bits ) ){
1120 $protocol = $bits[$i++
];
1121 $remainder = $bits[$i++
];
1123 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1124 # Found some characters after the protocol that look promising
1125 $url = $protocol . $m[1];
1128 # The characters '<' and '>' (which were escaped by
1129 # removeHTMLtags()) should not be included in
1130 # URLs, per RFC 2396.
1131 if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE
)) {
1132 $trail = substr($url, $m2[0][1]) . $trail;
1133 $url = substr($url, 0, $m2[0][1]);
1136 # Move trailing punctuation to $trail
1138 # If there is no left bracket, then consider right brackets fair game too
1139 if ( strpos( $url, '(' ) === false ) {
1143 $numSepChars = strspn( strrev( $url ), $sep );
1144 if ( $numSepChars ) {
1145 $trail = substr( $url, -$numSepChars ) . $trail;
1146 $url = substr( $url, 0, -$numSepChars );
1149 # Replace & from obsolete syntax with &.
1150 # All HTML entities will be escaped by makeExternalLink()
1151 # or maybeMakeExternalImage()
1152 $url = str_replace( '&', '&', $url );
1154 # Is this an external image?
1155 $text = $this->maybeMakeExternalImage( $url );
1156 if ( $text === false ) {
1157 # Not an image, make a link
1158 $text = $sk->makeExternalLink( $url, $wgContLang->markNoConversion($url), true, 'free' );
1160 $s .= $text . $trail;
1162 $s .= $protocol . $remainder;
1170 * make an image if it's allowed
1173 function maybeMakeExternalImage( $url ) {
1174 $sk =& $this->mOptions
->getSkin();
1176 if ( $this->mOptions
->getAllowExternalImages() ) {
1177 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1179 $text = $sk->makeExternalImage( htmlspecialchars( $url ) );
1186 * Process [[ ]] wikilinks
1190 function replaceInternalLinks( $s ) {
1191 global $wgContLang, $wgLinkCache;
1192 static $fname = 'Parser::replaceInternalLinks' ;
1194 wfProfileIn( $fname );
1196 wfProfileIn( $fname.'-setup' );
1198 # the % is needed to support urlencoded titles as well
1199 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1201 $sk =& $this->mOptions
->getSkin();
1203 #split the entire text string on occurences of [[
1204 $a = explode( '[[', ' ' . $s );
1205 #get the first element (all text up to first [[), and remove the space we added
1206 $s = array_shift( $a );
1207 $s = substr( $s, 1 );
1209 # Match a link having the form [[namespace:link|alternate]]trail
1211 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
1212 # Match cases where there is no "]]", which might still be images
1213 static $e1_img = FALSE;
1214 if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
1215 # Match the end of a line for a word that's not followed by whitespace,
1216 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1217 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1219 $useLinkPrefixExtension = $wgContLang->linkPrefixExtension();
1221 if( is_null( $this->mTitle
) ) {
1222 wfDebugDieBacktrace( 'nooo' );
1224 $nottalk = !$this->mTitle
->isTalkPage();
1226 if ( $useLinkPrefixExtension ) {
1227 if ( preg_match( $e2, $s, $m ) ) {
1228 $first_prefix = $m[2];
1231 $first_prefix = false;
1237 $selflink = $this->mTitle
->getPrefixedText();
1238 wfProfileOut( $fname.'-setup' );
1240 $checkVariantLink = sizeof($wgContLang->getVariants())>1;
1241 $useSubpages = $this->areSubpagesAllowed();
1243 # Loop for each link
1244 for ($k = 0; isset( $a[$k] ); $k++
) {
1246 if ( $useLinkPrefixExtension ) {
1247 wfProfileIn( $fname.'-prefixhandling' );
1248 if ( preg_match( $e2, $s, $m ) ) {
1256 $prefix = $first_prefix;
1257 $first_prefix = false;
1259 wfProfileOut( $fname.'-prefixhandling' );
1262 $might_be_img = false;
1264 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1266 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
1267 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
1268 # the real problem is with the $e1 regex
1271 # Still some problems for cases where the ] is meant to be outside punctuation,
1272 # and no image is in sight. See bug 2095.
1274 if( $text !== '' && preg_match( "/^\](.*)/s", $m[3], $n ) ) {
1275 $text .= ']'; # so that replaceExternalLinks($text) works later
1278 # fix up urlencoded title texts
1279 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1281 } elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
1282 $might_be_img = true;
1284 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1286 } else { # Invalid form; output directly
1287 $s .= $prefix . '[[' . $line ;
1291 # Don't allow internal links to pages containing
1292 # PROTO: where PROTO is a valid URL protocol; these
1293 # should be external links.
1294 if (preg_match('/^(\b(?:'.URL_PROTOCOLS
.'))/', $m[1])) {
1295 $s .= $prefix . '[[' . $line ;
1299 # Make subpage if necessary
1300 if( $useSubpages ) {
1301 $link = $this->maybeDoSubpageLink( $m[1], $text );
1306 $noforce = (substr($m[1], 0, 1) != ':');
1308 # Strip off leading ':'
1309 $link = substr($link, 1);
1312 $nt =& Title
::newFromText( $this->unstripNoWiki($link, $this->mStripState
) );
1314 $s .= $prefix . '[[' . $line;
1318 #check other language variants of the link
1319 #if the article does not exist
1320 if( $checkVariantLink
1321 && $nt->getArticleID() == 0 ) {
1322 $wgContLang->findVariantLink($link, $nt);
1325 $ns = $nt->getNamespace();
1326 $iw = $nt->getInterWiki();
1328 if ($might_be_img) { # if this is actually an invalid link
1329 if ($ns == NS_IMAGE
&& $noforce) { #but might be an image
1331 while (isset ($a[$k+
1]) ) {
1332 #look at the next 'line' to see if we can close it there
1333 $next_line = array_shift(array_splice( $a, $k +
1, 1) );
1334 if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
1335 # the first ]] closes the inner link, the second the image
1337 $text .= '[[' . $m[1];
1340 } elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
1341 #if there's exactly one ]] that's fine, we'll keep looking
1342 $text .= '[[' . $m[0];
1344 #if $next_line is invalid too, we need look no further
1345 $text .= '[[' . $next_line;
1350 # we couldn't find the end of this imageLink, so output it raw
1351 #but don't ignore what might be perfectly normal links in the text we've examined
1352 $text = $this->replaceInternalLinks($text);
1353 $s .= $prefix . '[[' . $link . '|' . $text;
1354 # note: no $trail, because without an end, there *is* no trail
1357 } else { #it's not an image, so output it raw
1358 $s .= $prefix . '[[' . $link . '|' . $text;
1359 # note: no $trail, because without an end, there *is* no trail
1364 $wasblank = ( '' == $text );
1365 if( $wasblank ) $text = $link;
1368 # Link not escaped by : , create the various objects
1372 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
1373 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1374 $s = rtrim($s . "\n");
1375 $s .= trim($prefix . $trail, "\n") == '' ?
'': $prefix . $trail;
1379 if ( $ns == NS_IMAGE
) {
1380 wfProfileIn( "$fname-image" );
1381 if ( !wfIsBadImage( $nt->getDBkey() ) ) {
1382 # recursively parse links inside the image caption
1383 # actually, this will parse them in any other parameters, too,
1384 # but it might be hard to fix that, and it doesn't matter ATM
1385 $text = $this->replaceExternalLinks($text);
1386 $text = $this->replaceInternalLinks($text);
1388 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
1389 $s .= $prefix . str_replace('http://', 'http-noparse://', $this->makeImage( $nt, $text ) ) . $trail;
1390 $wgLinkCache->addImageLinkObj( $nt );
1392 wfProfileOut( "$fname-image" );
1395 wfProfileOut( "$fname-image" );
1399 if ( $ns == NS_CATEGORY
) {
1400 wfProfileIn( "$fname-category" );
1401 $t = $wgContLang->convertHtml( $nt->getText() );
1402 $s = rtrim($s . "\n"); # bug 87
1404 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1405 $t = $sk->makeLinkObj( $nt, $t, '', '' , $prefix );
1406 $wgLinkCache->resume();
1409 if ( $this->mTitle
->getNamespace() == NS_CATEGORY
) {
1410 $sortkey = $this->mTitle
->getText();
1412 $sortkey = $this->mTitle
->getPrefixedText();
1417 $sortkey = $wgContLang->convertCategoryKey( $sortkey );
1418 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1419 $this->mOutput
->addCategoryLink( $t );
1422 * Strip the whitespace Category links produce, see bug 87
1423 * @todo We might want to use trim($tmp, "\n") here.
1425 $s .= trim($prefix . $trail, "\n") == '' ?
'': $prefix . $trail;
1427 wfProfileOut( "$fname-category" );
1432 if( ( $nt->getPrefixedText() === $selflink ) &&
1433 ( $nt->getFragment() === '' ) ) {
1434 # Self-links are handled specially; generally de-link and change to bold.
1435 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1439 # Special and Media are pseudo-namespaces; no pages actually exist in them
1440 if( $ns == NS_MEDIA
) {
1441 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text, true ) . $trail;
1442 $wgLinkCache->addImageLinkObj( $nt );
1444 } elseif( $ns == NS_SPECIAL
) {
1445 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1448 if( !$nt->isExternal() && $nt->isAlwaysKnown() ) {
1450 * Skip lookups for special pages and self-links.
1451 * External interwiki links are not included here because
1452 * the HTTP urls would break output in the next parse step;
1453 * they will have placeholders kept.
1455 $s .= $sk->makeKnownLinkObj( $nt, $text, '', $trail, $prefix );
1458 * Add a link placeholder
1459 * Later, this will be replaced by a real link, after the existence or
1460 * non-existence of all the links is known
1462 $s .= $this->makeLinkHolder( $nt, $text, '', $trail, $prefix );
1465 wfProfileOut( $fname );
1470 * Make a link placeholder. The text returned can be later resolved to a real link with
1471 * replaceLinkHolders(). This is done for two reasons: firstly to avoid further
1472 * parsing of interwiki links, and secondly to allow all extistence checks and
1473 * article length checks (for stub links) to be bundled into a single query.
1476 function makeLinkHolder( &$nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
1477 if ( ! is_object($nt) ) {
1479 $retVal = "<!-- ERROR -->{$prefix}{$text}{$trail}";
1481 # Separate the link trail from the rest of the link
1482 list( $inside, $trail ) = Linker
::splitTrail( $trail );
1484 if ( $nt->isExternal() ) {
1485 $nr = array_push( $this->mInterwikiLinkHolders
['texts'], $prefix.$text.$inside );
1486 $this->mInterwikiLinkHolders
['titles'][] = $nt;
1487 $retVal = '<!--IWLINK '. ($nr-1) ."-->{$trail}";
1489 $nr = array_push( $this->mLinkHolders
['namespaces'], $nt->getNamespace() );
1490 $this->mLinkHolders
['dbkeys'][] = $nt->getDBkey();
1491 $this->mLinkHolders
['queries'][] = $query;
1492 $this->mLinkHolders
['texts'][] = $prefix.$text.$inside;
1493 $this->mLinkHolders
['titles'][] = $nt;
1495 $retVal = '<!--LINK '. ($nr-1) ."-->{$trail}";
1502 * Return true if subpage links should be expanded on this page.
1505 function areSubpagesAllowed() {
1506 # Some namespaces don't allow subpages
1507 global $wgNamespacesWithSubpages;
1508 return !empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]);
1512 * Handle link to subpage if necessary
1513 * @param string $target the source of the link
1514 * @param string &$text the link text, modified as necessary
1515 * @return string the full name of the link
1518 function maybeDoSubpageLink($target, &$text) {
1521 # :Foobar -- override special treatment of prefix (images, language links)
1522 # /Foobar -- convert to CurrentPage/Foobar
1523 # /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1524 # ../ -- convert to CurrentPage, from CurrentPage/CurrentSubPage
1525 # ../Foobar -- convert to CurrentPage/Foobar, from CurrentPage/CurrentSubPage
1527 $fname = 'Parser::maybeDoSubpageLink';
1528 wfProfileIn( $fname );
1529 $ret = $target; # default return value is no change
1531 # Some namespaces don't allow subpages,
1532 # so only perform processing if subpages are allowed
1533 if( $this->areSubpagesAllowed() ) {
1534 # Look at the first character
1535 if( $target != '' && $target{0} == '/' ) {
1536 # / at end means we don't want the slash to be shown
1537 if( substr( $target, -1, 1 ) == '/' ) {
1538 $target = substr( $target, 1, -1 );
1541 $noslash = substr( $target, 1 );
1544 $ret = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1545 if( '' === $text ) {
1547 } # this might be changed for ugliness reasons
1549 # check for .. subpage backlinks
1551 $nodotdot = $target;
1552 while( strncmp( $nodotdot, "../", 3 ) == 0 ) {
1554 $nodotdot = substr( $nodotdot, 3 );
1556 if($dotdotcount > 0) {
1557 $exploded = explode( '/', $this->mTitle
->GetPrefixedText() );
1558 if( count( $exploded ) > $dotdotcount ) { # not allowed to go below top level page
1559 $ret = implode( '/', array_slice( $exploded, 0, -$dotdotcount ) );
1560 # / at the end means don't show full path
1561 if( substr( $nodotdot, -1, 1 ) == '/' ) {
1562 $nodotdot = substr( $nodotdot, 0, -1 );
1563 if( '' === $text ) {
1567 $nodotdot = trim( $nodotdot );
1568 if( $nodotdot != '' ) {
1569 $ret .= '/' . $nodotdot;
1576 wfProfileOut( $fname );
1581 * Used by doBlockLevels()
1584 /* private */ function closeParagraph() {
1586 if ( '' != $this->mLastSection
) {
1587 $result = '</' . $this->mLastSection
. ">\n";
1589 $this->mInPre
= false;
1590 $this->mLastSection
= '';
1593 # getCommon() returns the length of the longest common substring
1594 # of both arguments, starting at the beginning of both.
1596 /* private */ function getCommon( $st1, $st2 ) {
1597 $fl = strlen( $st1 );
1598 $shorter = strlen( $st2 );
1599 if ( $fl < $shorter ) { $shorter = $fl; }
1601 for ( $i = 0; $i < $shorter; ++
$i ) {
1602 if ( $st1{$i} != $st2{$i} ) { break; }
1606 # These next three functions open, continue, and close the list
1607 # element appropriate to the prefix character passed into them.
1609 /* private */ function openList( $char ) {
1610 $result = $this->closeParagraph();
1612 if ( '*' == $char ) { $result .= '<ul><li>'; }
1613 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1614 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1615 else if ( ';' == $char ) {
1616 $result .= '<dl><dt>';
1617 $this->mDTopen
= true;
1619 else { $result = '<!-- ERR 1 -->'; }
1624 /* private */ function nextItem( $char ) {
1625 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1626 else if ( ':' == $char ||
';' == $char ) {
1628 if ( $this->mDTopen
) { $close = '</dt>'; }
1629 if ( ';' == $char ) {
1630 $this->mDTopen
= true;
1631 return $close . '<dt>';
1633 $this->mDTopen
= false;
1634 return $close . '<dd>';
1637 return '<!-- ERR 2 -->';
1640 /* private */ function closeList( $char ) {
1641 if ( '*' == $char ) { $text = '</li></ul>'; }
1642 else if ( '#' == $char ) { $text = '</li></ol>'; }
1643 else if ( ':' == $char ) {
1644 if ( $this->mDTopen
) {
1645 $this->mDTopen
= false;
1646 $text = '</dt></dl>';
1648 $text = '</dd></dl>';
1651 else { return '<!-- ERR 3 -->'; }
1657 * Make lists from lines starting with ':', '*', '#', etc.
1660 * @return string the lists rendered as HTML
1662 function doBlockLevels( $text, $linestart ) {
1663 $fname = 'Parser::doBlockLevels';
1664 wfProfileIn( $fname );
1666 # Parsing through the text line by line. The main thing
1667 # happening here is handling of block-level elements p, pre,
1668 # and making lists from lines starting with * # : etc.
1670 $textLines = explode( "\n", $text );
1672 $lastPrefix = $output = '';
1673 $this->mDTopen
= $inBlockElem = false;
1675 $paragraphStack = false;
1677 if ( !$linestart ) {
1678 $output .= array_shift( $textLines );
1680 foreach ( $textLines as $oLine ) {
1681 $lastPrefixLength = strlen( $lastPrefix );
1682 $preCloseMatch = preg_match('/<\\/pre/i', $oLine );
1683 $preOpenMatch = preg_match('/<pre/i', $oLine );
1684 if ( !$this->mInPre
) {
1685 # Multiple prefixes may abut each other for nested lists.
1686 $prefixLength = strspn( $oLine, '*#:;' );
1687 $pref = substr( $oLine, 0, $prefixLength );
1690 $pref2 = str_replace( ';', ':', $pref );
1691 $t = substr( $oLine, $prefixLength );
1692 $this->mInPre
= !empty($preOpenMatch);
1694 # Don't interpret any other prefixes in preformatted text
1696 $pref = $pref2 = '';
1701 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1702 # Same as the last item, so no need to deal with nesting or opening stuff
1703 $output .= $this->nextItem( substr( $pref, -1 ) );
1704 $paragraphStack = false;
1706 if ( substr( $pref, -1 ) == ';') {
1707 # The one nasty exception: definition lists work like this:
1708 # ; title : definition text
1709 # So we check for : in the remainder text to split up the
1710 # title and definition, without b0rking links.
1712 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1714 $output .= $term . $this->nextItem( ':' );
1717 } elseif( $prefixLength ||
$lastPrefixLength ) {
1718 # Either open or close a level...
1719 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1720 $paragraphStack = false;
1722 while( $commonPrefixLength < $lastPrefixLength ) {
1723 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1724 --$lastPrefixLength;
1726 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1727 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1729 while ( $prefixLength > $commonPrefixLength ) {
1730 $char = substr( $pref, $commonPrefixLength, 1 );
1731 $output .= $this->openList( $char );
1733 if ( ';' == $char ) {
1734 # FIXME: This is dupe of code above
1735 if ($this->findColonNoLinks($t, $term, $t2) !== false) {
1737 $output .= $term . $this->nextItem( ':' );
1740 ++
$commonPrefixLength;
1742 $lastPrefix = $pref2;
1744 if( 0 == $prefixLength ) {
1745 wfProfileIn( "$fname-paragraph" );
1746 # No prefix (not in list)--go to paragraph mode
1747 $uniq_prefix = UNIQ_PREFIX
;
1748 // XXX: use a stack for nestable elements like span, table and div
1749 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/iS', $t );
1750 $closematch = preg_match(
1751 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1752 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/iS', $t );
1753 if ( $openmatch or $closematch ) {
1754 $paragraphStack = false;
1755 $output .= $this->closeParagraph();
1756 if($preOpenMatch and !$preCloseMatch) {
1757 $this->mInPre
= true;
1759 if ( $closematch ) {
1760 $inBlockElem = false;
1762 $inBlockElem = true;
1764 } else if ( !$inBlockElem && !$this->mInPre
) {
1765 if ( ' ' == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1767 if ($this->mLastSection
!= 'pre') {
1768 $paragraphStack = false;
1769 $output .= $this->closeParagraph().'<pre>';
1770 $this->mLastSection
= 'pre';
1772 $t = substr( $t, 1 );
1775 if ( '' == trim($t) ) {
1776 if ( $paragraphStack ) {
1777 $output .= $paragraphStack.'<br />';
1778 $paragraphStack = false;
1779 $this->mLastSection
= 'p';
1781 if ($this->mLastSection
!= 'p' ) {
1782 $output .= $this->closeParagraph();
1783 $this->mLastSection
= '';
1784 $paragraphStack = '<p>';
1786 $paragraphStack = '</p><p>';
1790 if ( $paragraphStack ) {
1791 $output .= $paragraphStack;
1792 $paragraphStack = false;
1793 $this->mLastSection
= 'p';
1794 } else if ($this->mLastSection
!= 'p') {
1795 $output .= $this->closeParagraph().'<p>';
1796 $this->mLastSection
= 'p';
1801 wfProfileOut( "$fname-paragraph" );
1803 if ($paragraphStack === false) {
1807 while ( $prefixLength ) {
1808 $output .= $this->closeList( $pref2{$prefixLength-1} );
1811 if ( '' != $this->mLastSection
) {
1812 $output .= '</' . $this->mLastSection
. '>';
1813 $this->mLastSection
= '';
1816 wfProfileOut( $fname );
1821 * Split up a string on ':', ignoring any occurences inside
1822 * <a>..</a> or <span>...</span>
1823 * @param string $str the string to split
1824 * @param string &$before set to everything before the ':'
1825 * @param string &$after set to everything after the ':'
1826 * return string the position of the ':', or false if none found
1828 function findColonNoLinks($str, &$before, &$after) {
1829 # I wonder if we should make this count all tags, not just <a>
1830 # and <span>. That would prevent us from matching a ':' that
1831 # comes in the middle of italics other such formatting....
1833 $fname = 'Parser::findColonNoLinks';
1834 wfProfileIn( $fname );
1837 $colon = strpos($str, ':', $pos);
1839 if ($colon !== false) {
1840 $before = substr($str, 0, $colon);
1841 $after = substr($str, $colon +
1);
1843 # Skip any ':' within <a> or <span> pairs
1844 $a = substr_count($before, '<a');
1845 $s = substr_count($before, '<span');
1846 $ca = substr_count($before, '</a>');
1847 $cs = substr_count($before, '</span>');
1849 if ($a <= $ca and $s <= $cs) {
1850 # Tags are balanced before ':'; ok
1855 } while ($colon !== false);
1856 wfProfileOut( $fname );
1861 * Return value of a magic variable (like PAGENAME)
1865 function getVariableValue( $index ) {
1866 global $wgContLang, $wgSitename, $wgServer, $wgServerName, $wgArticle, $wgScriptPath;
1869 * Some of these require message or data lookups and can be
1870 * expensive to check many times.
1872 static $varCache = array();
1873 if( isset( $varCache[$index] ) ) return $varCache[$index];
1876 case MAG_CURRENTMONTH
:
1877 return $varCache[$index] = $wgContLang->formatNum( date( 'm' ) );
1878 case MAG_CURRENTMONTHNAME
:
1879 return $varCache[$index] = $wgContLang->getMonthName( date('n') );
1880 case MAG_CURRENTMONTHNAMEGEN
:
1881 return $varCache[$index] = $wgContLang->getMonthNameGen( date('n') );
1882 case MAG_CURRENTMONTHABBREV
:
1883 return $varCache[$index] = $wgContLang->getMonthAbbreviation( date('n') );
1884 case MAG_CURRENTDAY
:
1885 return $varCache[$index] = $wgContLang->formatNum( date('j') );
1887 return $this->mTitle
->getText();
1889 return $this->mTitle
->getPartialURL();
1890 case MAG_REVISIONID
:
1891 return $wgArticle->getRevIdFetched();
1893 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1894 return $wgContLang->getNsText($this->mTitle
->getNamespace()); # Patch by Dori
1895 case MAG_CURRENTDAYNAME
:
1896 return $varCache[$index] = $wgContLang->getWeekdayName( date('w')+
1 );
1897 case MAG_CURRENTYEAR
:
1898 return $varCache[$index] = $wgContLang->formatNum( date( 'Y' ), true );
1899 case MAG_CURRENTTIME
:
1900 return $varCache[$index] = $wgContLang->time( wfTimestampNow(), false );
1901 case MAG_CURRENTWEEK
:
1902 return $varCache[$index] = $wgContLang->formatNum( date('W') );
1903 case MAG_CURRENTDOW
:
1904 return $varCache[$index] = $wgContLang->formatNum( date('w') );
1905 case MAG_NUMBEROFARTICLES
:
1906 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfArticles() );
1907 case MAG_NUMBEROFFILES
:
1908 return $varCache[$index] = $wgContLang->formatNum( wfNumberOfFiles() );
1913 case MAG_SERVERNAME
:
1914 return $wgServerName;
1915 case MAG_SCRIPTPATH
:
1916 return $wgScriptPath;
1923 * initialise the magic variables (like CURRENTMONTHNAME)
1927 function initialiseVariables() {
1928 $fname = 'Parser::initialiseVariables';
1929 wfProfileIn( $fname );
1930 global $wgVariableIDs;
1931 $this->mVariables
= array();
1932 foreach ( $wgVariableIDs as $id ) {
1933 $mw =& MagicWord
::get( $id );
1934 $mw->addToArray( $this->mVariables
, $id );
1936 wfProfileOut( $fname );
1940 * Replace magic variables, templates, and template arguments
1941 * with the appropriate text. Templates are substituted recursively,
1942 * taking care to avoid infinite loops.
1944 * Note that the substitution depends on value of $mOutputType:
1945 * OT_WIKI: only {{subst:}} templates
1946 * OT_MSG: only magic variables
1947 * OT_HTML: all templates and magic variables
1949 * @param string $tex The text to transform
1950 * @param array $args Key-value pairs representing template parameters to substitute
1953 function replaceVariables( $text, $args = array() ) {
1955 # Prevent too big inclusions
1956 if( strlen( $text ) > MAX_INCLUDE_SIZE
) {
1960 $fname = 'Parser::replaceVariables';
1961 wfProfileIn( $fname );
1963 $titleChars = Title
::legalChars();
1965 # This function is called recursively. To keep track of arguments we need a stack:
1966 array_push( $this->mArgStack
, $args );
1968 # Variable substitution
1969 $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", array( &$this, 'variableSubstitution' ), $text );
1971 if ( $this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) {
1972 # Argument substitution
1973 $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", array( &$this, 'argSubstitution' ), $text );
1975 # Template substitution
1976 $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
1977 $text = preg_replace_callback( $regex, array( &$this, 'braceSubstitution' ), $text );
1979 array_pop( $this->mArgStack
);
1981 wfProfileOut( $fname );
1986 * Replace magic variables
1989 function variableSubstitution( $matches ) {
1990 $fname = 'parser::variableSubstitution';
1991 $varname = $matches[1];
1992 wfProfileIn( $fname );
1993 if ( !$this->mVariables
) {
1994 $this->initialiseVariables();
1997 if ( $this->mOutputType
== OT_WIKI
) {
1998 # Do only magic variables prefixed by SUBST
1999 $mwSubst =& MagicWord
::get( MAG_SUBST
);
2000 if (!$mwSubst->matchStartAndRemove( $varname ))
2002 # Note that if we don't substitute the variable below,
2003 # we don't remove the {{subst:}} magic word, in case
2004 # it is a template rather than a magic variable.
2006 if ( !$skip && array_key_exists( $varname, $this->mVariables
) ) {
2007 $id = $this->mVariables
[$varname];
2008 $text = $this->getVariableValue( $id );
2009 $this->mOutput
->mContainsOldMagic
= true;
2011 $text = $matches[0];
2013 wfProfileOut( $fname );
2017 # Split template arguments
2018 function getTemplateArgs( $argsString ) {
2019 if ( $argsString === '' ) {
2023 $args = explode( '|', substr( $argsString, 1 ) );
2025 # If any of the arguments contains a '[[' but no ']]', it needs to be
2026 # merged with the next arg because the '|' character between belongs
2027 # to the link syntax and not the template parameter syntax.
2028 $argc = count($args);
2030 for ( $i = 0; $i < $argc-1; $i++
) {
2031 if ( substr_count ( $args[$i], '[[' ) != substr_count ( $args[$i], ']]' ) ) {
2032 $args[$i] .= '|'.$args[$i+
1];
2033 array_splice($args, $i+
1, 1);
2043 * Return the text of a template, after recursively
2044 * replacing any variables or templates within the template.
2046 * @param array $matches The parts of the template
2047 * $matches[1]: the title, i.e. the part before the |
2048 * $matches[2]: the parameters (including a leading |), if any
2049 * @return string the text of the template
2052 function braceSubstitution( $matches ) {
2053 global $wgLinkCache, $wgContLang;
2054 $fname = 'Parser::braceSubstitution';
2055 wfProfileIn( $fname );
2063 # Need to know if the template comes at the start of a line,
2064 # to treat the beginning of the template like the beginning
2065 # of a line for tables and block-level elements.
2066 $linestart = $matches[1];
2068 # $part1 is the bit before the first |, and must contain only title characters
2069 # $args is a list of arguments, starting from index 0, not including $part1
2071 $part1 = $matches[2];
2072 # If the third subpattern matched anything, it will start with |
2074 $args = $this->getTemplateArgs($matches[3]);
2075 $argc = count( $args );
2077 # Don't parse {{{}}} because that's only for template arguments
2078 if ( $linestart === '{' ) {
2079 $text = $matches[0];
2086 $mwSubst =& MagicWord
::get( MAG_SUBST
);
2087 if ( $mwSubst->matchStartAndRemove( $part1 ) xor ($this->mOutputType
== OT_WIKI
) ) {
2088 # One of two possibilities is true:
2089 # 1) Found SUBST but not in the PST phase
2090 # 2) Didn't find SUBST and in the PST phase
2091 # In either case, return without further processing
2092 $text = $matches[0];
2098 # MSG, MSGNW and INT
2101 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
2102 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
2105 # Remove obsolete MSG:
2106 $mwMsg =& MagicWord
::get( MAG_MSG
);
2107 $mwMsg->matchStartAndRemove( $part1 );
2110 # Check if it is an internal message
2111 $mwInt =& MagicWord
::get( MAG_INT
);
2112 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
2113 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
2114 $text = $linestart . wfMsgReal( $part1, $args, true );
2122 # Check for NS: (namespace expansion)
2123 $mwNs = MagicWord
::get( MAG_NS
);
2124 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
2125 if ( intval( $part1 ) ) {
2126 $text = $linestart . $wgContLang->getNsText( intval( $part1 ) );
2129 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
2130 if ( !is_null( $index ) ) {
2131 $text = $linestart . $wgContLang->getNsText( $index );
2138 # LOCALURL and LOCALURLE
2140 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
2141 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
2143 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
2144 $func = 'getLocalURL';
2145 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
2146 $func = 'escapeLocalURL';
2151 if ( $func !== '' ) {
2152 $title = Title
::newFromText( $part1 );
2153 if ( !is_null( $title ) ) {
2155 $text = $linestart . $title->$func( $args[0] );
2157 $text = $linestart . $title->$func();
2165 if ( !$found && $argc == 1 ) {
2166 $mwGrammar =& MagicWord
::get( MAG_GRAMMAR
);
2167 if ( $mwGrammar->matchStartAndRemove( $part1 ) ) {
2168 $text = $linestart . $wgContLang->convertGrammar( $args[0], $part1 );
2173 # Template table test
2175 # Did we encounter this template already? If yes, it is in the cache
2176 # and we need to check for loops.
2177 if ( !$found && isset( $this->mTemplates
[$part1] ) ) {
2180 # Infinite loop test
2181 if ( isset( $this->mTemplatePath
[$part1] ) ) {
2184 $text = $linestart .
2186 '<!-- WARNING: template loop detected -->';
2187 wfDebug( "$fname: template loop broken at '$part1'\n" );
2189 # set $text to cached message.
2190 $text = $linestart . $this->mTemplates
[$part1];
2194 # Load from database
2195 $replaceHeadings = false;
2197 $lastPathLevel = $this->mTemplatePath
;
2200 $part1 = $this->maybeDoSubpageLink( $part1, $subpage='' );
2201 if ($subpage !== '') {
2202 $ns = $this->mTitle
->getNamespace();
2204 $title = Title
::newFromText( $part1, $ns );
2207 $interwiki = Title
::getInterwikiLink($title->getInterwiki());
2208 if ($interwiki != '' && $title->isTrans()) {
2209 return $this->scarytransclude($title, $interwiki);
2213 if ( !is_null( $title ) && !$title->isExternal() ) {
2214 # Check for excessive inclusion
2215 $dbk = $title->getPrefixedDBkey();
2216 if ( $this->incrementIncludeCount( $dbk ) ) {
2217 if ( $title->getNamespace() == NS_SPECIAL
&& $this->mOptions
->getAllowSpecialInclusion() ) {
2218 # Capture special page output
2219 $text = SpecialPage
::capturePath( $title );
2220 if ( $text && !is_object( $text ) ) {
2224 $this->mOutput
->setCacheTime( -1 );
2227 $article = new Article( $title );
2228 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
2229 if ( $articleContent !== false ) {
2231 $text = $articleContent;
2232 $replaceHeadings = true;
2237 # If the title is valid but undisplayable, make a link to it
2238 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
2239 $text = '[['.$title->getPrefixedText().']]';
2243 # Template cache array insertion
2245 $this->mTemplates
[$part1] = $text;
2246 $text = $linestart . $text;
2251 # Recursive parsing, escaping and link table handling
2252 # Only for HTML output
2253 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
2254 $text = wfEscapeWikiText( $text );
2255 } elseif ( ($this->mOutputType
== OT_HTML ||
$this->mOutputType
== OT_WIKI
) && $found && !$noparse) {
2256 # Clean up argument array
2257 $assocArgs = array();
2259 foreach( $args as $arg ) {
2260 $eqpos = strpos( $arg, '=' );
2261 if ( $eqpos === false ) {
2262 $assocArgs[$index++
] = $arg;
2264 $name = trim( substr( $arg, 0, $eqpos ) );
2265 $value = trim( substr( $arg, $eqpos+
1 ) );
2266 if ( $value === false ) {
2269 if ( $name !== false ) {
2270 $assocArgs[$name] = $value;
2275 # Add a new element to the templace recursion path
2276 $this->mTemplatePath
[$part1] = 1;
2278 if( $this->mOutputType
== OT_HTML
) {
2279 $text = $this->strip( $text, $this->mStripState
);
2280 $text = Sanitizer
::removeHTMLtags( $text, array( &$this, 'replaceVariables' ), $assocArgs );
2282 $text = $this->replaceVariables( $text, $assocArgs );
2284 # Resume the link cache and register the inclusion as a link
2285 if ( $this->mOutputType
== OT_HTML
&& !is_null( $title ) ) {
2286 $wgLinkCache->addLinkObj( $title );
2289 # If the template begins with a table or block-level
2290 # element, it should be treated as beginning a new line.
2291 if ($linestart !== '\n' && preg_match('/^({\\||:|;|#|\*)/', $text)) {
2292 $text = "\n" . $text;
2295 # Prune lower levels off the recursion check path
2296 $this->mTemplatePath
= $lastPathLevel;
2299 wfProfileOut( $fname );
2303 # Replace raw HTML by a placeholder
2304 # Add a blank line preceding, to prevent it from mucking up
2305 # immediately preceding headings
2306 $text = "\n\n" . $this->insertStripItem( $text, $this->mStripState
);
2308 # replace ==section headers==
2309 # XXX this needs to go away once we have a better parser.
2310 if ( $this->mOutputType
!= OT_WIKI
&& $replaceHeadings ) {
2311 if( !is_null( $title ) )
2312 $encodedname = base64_encode($title->getPrefixedDBkey());
2314 $encodedname = base64_encode("");
2315 $m = preg_split('/(^={1,6}.*?={1,6}\s*?$)/m', $text, -1,
2316 PREG_SPLIT_DELIM_CAPTURE
);
2319 for( $i = 0; $i < count($m); $i +
= 2 ) {
2321 if (!isset($m[$i +
1]) ||
$m[$i +
1] == "") continue;
2323 if( strstr($hl, "<!--MWTEMPLATESECTION") ) {
2327 preg_match('/^(={1,6})(.*?)(={1,6})\s*?$/m', $hl, $m2);
2328 $text .= $m2[1] . $m2[2] . "<!--MWTEMPLATESECTION="
2329 . $encodedname . "&" . base64_encode("$nsec") . "-->" . $m2[3];
2337 # Prune lower levels off the recursion check path
2338 $this->mTemplatePath
= $lastPathLevel;
2341 wfProfileOut( $fname );
2344 wfProfileOut( $fname );
2350 * Translude an interwiki link.
2352 function scarytransclude($title, $interwiki) {
2353 global $wgEnableScaryTranscluding;
2355 if (!$wgEnableScaryTranscluding)
2356 return wfMsg('scarytranscludedisabled');
2358 $articlename = "Template:" . $title->getDBkey();
2359 $url = str_replace('$1', urlencode($articlename), $interwiki);
2360 if (strlen($url) > 255)
2361 return wfMsg('scarytranscludetoolong');
2362 $text = $this->fetchScaryTemplateMaybeFromCache($url);
2363 $this->mIWTransData
[] = $text;
2364 return "<!--IW_TRANSCLUDE ".(count($this->mIWTransData
) - 1)."-->";
2367 function fetchScaryTemplateMaybeFromCache($url) {
2368 $dbr = wfGetDB(DB_SLAVE
);
2369 $obj = $dbr->selectRow('transcache', array('tc_time', 'tc_contents'),
2370 array('tc_url' => $url));
2372 $time = $obj->tc_time
;
2373 $text = $obj->tc_contents
;
2374 if ($time && $time < (time() +
(60*60))) {
2379 $text = wfGetHTTP($url . '?action=render');
2381 return wfMsg('scarytranscludefailed', $url);
2383 $dbw = wfGetDB(DB_MASTER
);
2384 $dbw->replace('transcache', array(), array(
2386 'tc_time' => time(),
2387 'tc_contents' => $text));
2393 * Triple brace replacement -- used for template arguments
2396 function argSubstitution( $matches ) {
2397 $arg = trim( $matches[1] );
2398 $text = $matches[0];
2399 $inputArgs = end( $this->mArgStack
);
2401 if ( array_key_exists( $arg, $inputArgs ) ) {
2402 $text = $inputArgs[$arg];
2409 * Returns true if the function is allowed to include this entity
2412 function incrementIncludeCount( $dbk ) {
2413 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
2414 $this->mIncludeCount
[$dbk] = 0;
2416 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
2424 * This function accomplishes several tasks:
2425 * 1) Auto-number headings if that option is enabled
2426 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2427 * 3) Add a Table of contents on the top for users who have enabled the option
2428 * 4) Auto-anchor headings
2430 * It loops through all headlines, collects the necessary data, then splits up the
2431 * string and re-inserts the newly formatted headlines.
2433 * @param string $text
2434 * @param boolean $isMain
2437 function formatHeadings( $text, $isMain=true ) {
2438 global $wgMaxTocLevel, $wgContLang, $wgLinkHolders, $wgInterwikiLinkHolders;
2440 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2442 $forceTocHere = false;
2443 if( !$this->mTitle
->userCanEdit() ) {
2446 $showEditLink = $this->mOptions
->getEditSection();
2449 # Inhibit editsection links if requested in the page
2450 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2451 if( $esw->matchAndRemove( $text ) ) {
2454 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2456 $mw =& MagicWord
::get( MAG_NOTOC
);
2457 if( $mw->matchAndRemove( $text ) ) {
2461 # Get all headlines for numbering them and adding funky stuff like [edit]
2462 # links - this is for later, but we need the number of headlines right now
2463 $numMatches = preg_match_all( '/<H([1-6])(.*?'.'>)(.*?)<\/H[1-6] *>/i', $text, $matches );
2465 # if there are fewer than 4 headlines in the article, do not show TOC
2466 if( $numMatches < 4 ) {
2470 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2471 # override above conditions and always show TOC at that place
2473 $mw =& MagicWord
::get( MAG_TOC
);
2474 if($mw->match( $text ) ) {
2476 $forceTocHere = true;
2478 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2479 # override above conditions and always show TOC above first header
2480 $mw =& MagicWord
::get( MAG_FORCETOC
);
2481 if ($mw->matchAndRemove( $text ) ) {
2486 # Never ever show TOC if no headers
2487 if( $numMatches < 1 ) {
2491 # We need this to perform operations on the HTML
2492 $sk =& $this->mOptions
->getSkin();
2496 $sectionCount = 0; # headlineCount excluding template sections
2498 # Ugh .. the TOC should have neat indentation levels which can be
2499 # passed to the skin functions. These are determined here
2503 $sublevelCount = array();
2504 $levelCount = array();
2511 foreach( $matches[3] as $headline ) {
2513 $templatetitle = '';
2514 $templatesection = 0;
2517 if (preg_match("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", $headline, $mat)) {
2519 $templatetitle = base64_decode($mat[1]);
2520 $templatesection = 1 +
(int)base64_decode($mat[2]);
2521 $headline = preg_replace("/<!--MWTEMPLATESECTION=([^&]+)&([^_]+)-->/", "", $headline);
2525 $prevlevel = $level;
2526 $prevtoclevel = $toclevel;
2528 $level = $matches[1][$headlineCount];
2530 if( $doNumberHeadings ||
$doShowToc ) {
2532 if ( $level > $prevlevel ) {
2533 # Increase TOC level
2535 $sublevelCount[$toclevel] = 0;
2536 $toc .= $sk->tocIndent();
2538 elseif ( $level < $prevlevel && $toclevel > 1 ) {
2539 # Decrease TOC level, find level to jump to
2541 if ( $toclevel == 2 && $level <= $levelCount[1] ) {
2542 # Can only go down to level 1
2545 for ($i = $toclevel; $i > 0; $i--) {
2546 if ( $levelCount[$i] == $level ) {
2547 # Found last matching level
2551 elseif ( $levelCount[$i] < $level ) {
2552 # Found first matching level below current level
2559 $toc .= $sk->tocUnindent( $prevtoclevel - $toclevel );
2562 # No change in level, end TOC line
2563 $toc .= $sk->tocLineEnd();
2566 $levelCount[$toclevel] = $level;
2568 # count number of headlines for each level
2569 @$sublevelCount[$toclevel]++
;
2571 for( $i = 1; $i <= $toclevel; $i++
) {
2572 if( !empty( $sublevelCount[$i] ) ) {
2576 $numbering .= $wgContLang->formatNum( $sublevelCount[$i] );
2582 # The canonized header is a version of the header text safe to use for links
2583 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2584 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2585 $canonized_headline = $this->unstripNoWiki( $canonized_headline, $this->mStripState
);
2587 # Remove link placeholders by the link text.
2588 # <!--LINK number-->
2590 # link text with suffix
2591 $canonized_headline = preg_replace( '/<!--LINK ([0-9]*)-->/e',
2592 "\$this->mLinkHolders['texts'][\$1]",
2593 $canonized_headline );
2594 $canonized_headline = preg_replace( '/<!--IWLINK ([0-9]*)-->/e',
2595 "\$this->mInterwikiLinkHolders['texts'][\$1]",
2596 $canonized_headline );
2599 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2600 $tocline = trim( $canonized_headline );
2601 $canonized_headline = urlencode( Sanitizer
::decodeCharReferences( str_replace(' ', '_', $tocline) ) );
2602 $replacearray = array(
2606 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2607 $refers[$headlineCount] = $canonized_headline;
2609 # count how many in assoc. array so we can track dupes in anchors
2610 @$refers[$canonized_headline]++
;
2611 $refcount[$headlineCount]=$refers[$canonized_headline];
2613 # Don't number the heading if it is the only one (looks silly)
2614 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2615 # the two are different if the line contains a link
2616 $headline=$numbering . ' ' . $headline;
2619 # Create the anchor for linking from the TOC to the section
2620 $anchor = $canonized_headline;
2621 if($refcount[$headlineCount] > 1 ) {
2622 $anchor .= '_' . $refcount[$headlineCount];
2624 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2625 $toc .= $sk->tocLine($anchor, $tocline, $numbering, $toclevel);
2627 if( $showEditLink && ( !$istemplate ||
$templatetitle !== "" ) ) {
2628 if ( empty( $head[$headlineCount] ) ) {
2629 $head[$headlineCount] = '';
2632 $head[$headlineCount] .= $sk->editSectionLinkForOther($templatetitle, $templatesection);
2634 $head[$headlineCount] .= $sk->editSectionLink($this->mTitle
, $sectionCount+
1);
2637 # give headline the correct <h#> tag
2638 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline.'</h'.$level.'>';
2646 $toc .= $sk->tocUnindent( $toclevel - 1 );
2647 $toc = $sk->tocList( $toc );
2650 # split up and insert constructed headlines
2652 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2655 foreach( $blocks as $block ) {
2656 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2657 # This is the [edit] link that appears for the top block of text when
2658 # section editing is enabled
2660 # Disabled because it broke block formatting
2661 # For example, a bullet point in the top line
2662 # $full .= $sk->editSectionLink(0);
2665 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2666 # Top anchor now in skin
2670 if( !empty( $head[$i] ) ) {
2676 $mw =& MagicWord
::get( MAG_TOC
);
2677 return $mw->replace( $toc, $full );
2684 * Return an HTML link for the "ISBN 123456" text
2687 function magicISBN( $text ) {
2688 $fname = 'Parser::magicISBN';
2689 wfProfileIn( $fname );
2691 $a = split( 'ISBN ', ' '.$text );
2692 if ( count ( $a ) < 2 ) {
2693 wfProfileOut( $fname );
2696 $text = substr( array_shift( $a ), 1);
2697 $valid = '0123456789-Xx';
2699 foreach ( $a as $x ) {
2700 $isbn = $blank = '' ;
2701 while ( ' ' == $x{0} ) {
2703 $x = substr( $x, 1 );
2705 if ( $x == '' ) { # blank isbn
2706 $text .= "ISBN $blank";
2709 while ( strstr( $valid, $x{0} ) != false ) {
2711 $x = substr( $x, 1 );
2713 $num = str_replace( '-', '', $isbn );
2714 $num = str_replace( ' ', '', $num );
2715 $num = str_replace( 'x', 'X', $num );
2718 $text .= "ISBN $blank$x";
2720 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2721 $text .= '<a href="' .
2722 $titleObj->escapeLocalUrl( 'isbn='.$num ) .
2723 "\" class=\"internal\">ISBN $isbn</a>";
2727 wfProfileOut( $fname );
2732 * Return an HTML link for the "RFC 1234" text
2735 * @param string $text Text to be processed
2736 * @param string $keyword Magic keyword to use (default RFC)
2737 * @param string $urlmsg Interface message to use (default rfcurl)
2740 function magicRFC( $text, $keyword='RFC ', $urlmsg='rfcurl' ) {
2742 $valid = '0123456789';
2745 $a = split( $keyword, ' '.$text );
2746 if ( count ( $a ) < 2 ) {
2749 $text = substr( array_shift( $a ), 1);
2751 /* Check if keyword is preceed by [[.
2752 * This test is made here cause of the array_shift above
2753 * that prevent the test to be done in the foreach.
2755 if ( substr( $text, -2 ) == '[[' ) {
2759 foreach ( $a as $x ) {
2760 /* token might be empty if we have RFC RFC 1234 */
2768 /** remove and save whitespaces in $blank */
2769 while ( $x{0} == ' ' ) {
2771 $x = substr( $x, 1 );
2774 /** remove and save the rfc number in $id */
2775 while ( strstr( $valid, $x{0} ) != false ) {
2777 $x = substr( $x, 1 );
2781 /* call back stripped spaces*/
2782 $text .= $keyword.$blank.$x;
2783 } elseif( $internal ) {
2785 $text .= $keyword.$id.$x;
2787 /* build the external link*/
2788 $url = wfMsg( $urlmsg, $id);
2789 $sk =& $this->mOptions
->getSkin();
2790 $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
2791 $text .= "<a href='{$url}'{$la}>{$keyword}{$id}</a>{$x}";
2794 /* Check if the next RFC keyword is preceed by [[ */
2795 $internal = ( substr($x,-2) == '[[' );
2801 * Transform wiki markup when saving a page by doing \r\n -> \n
2802 * conversion, substitting signatures, {{subst:}} templates, etc.
2804 * @param string $text the text to transform
2805 * @param Title &$title the Title object for the current article
2806 * @param User &$user the User object describing the current user
2807 * @param ParserOptions $options parsing options
2808 * @param bool $clearState whether to clear the parser state first
2809 * @return string the altered wiki markup
2812 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2813 $this->mOptions
= $options;
2814 $this->mTitle
=& $title;
2815 $this->mOutputType
= OT_WIKI
;
2817 if ( $clearState ) {
2818 $this->clearState();
2821 $stripState = false;
2825 $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
2826 $text = $this->strip( $text, $stripState, false );
2827 $text = $this->pstPass2( $text, $user );
2828 $text = $this->unstrip( $text, $stripState );
2829 $text = $this->unstripNoWiki( $text, $stripState );
2834 * Pre-save transform helper function
2837 function pstPass2( $text, &$user ) {
2838 global $wgContLang, $wgLocaltimezone;
2840 # Variable replacement
2841 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2842 $text = $this->replaceVariables( $text );
2846 $n = $user->getName();
2847 $k = $user->getOption( 'nickname' );
2848 if ( '' == $k ) { $k = $n; }
2849 if ( isset( $wgLocaltimezone ) ) {
2850 $oldtz = getenv( 'TZ' );
2851 putenv( 'TZ='.$wgLocaltimezone );
2854 /* Note: This is the timestamp saved as hardcoded wikitext to
2855 * the database, we use $wgContLang here in order to give
2856 * everyone the same signiture and use the default one rather
2857 * than the one selected in each users preferences.
2859 $d = $wgContLang->timeanddate( date( 'YmdHis' ), false, false) .
2860 ' (' . date( 'T' ) . ')';
2861 if ( isset( $wgLocaltimezone ) ) {
2862 putenv( 'TZ='.$oldtz );
2865 if( $user->getOption( 'fancysig' ) ) {
2868 $sigText = '[[' . $wgContLang->getNsText( NS_USER
) . ":$n|$k]]";
2870 $text = preg_replace( '/~~~~~/', $d, $text );
2871 $text = preg_replace( '/~~~~/', "$sigText $d", $text );
2872 $text = preg_replace( '/~~~/', $sigText, $text );
2874 # Context links: [[|name]] and [[name (context)|]]
2876 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2877 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2878 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2879 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2881 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2882 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2883 $p3 = "/\[\[(:*$namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]] and [[:namespace:page|]]
2884 $p4 = "/\[\[(:*$namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/"; # [[ns:page (cont)|]] and [[:ns:page (cont)|]]
2886 $t = $this->mTitle
->getText();
2887 if ( preg_match( $conpat, $t, $m ) ) {
2890 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2891 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2892 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2894 if ( '' == $context ) {
2895 $text = preg_replace( $p2, '[[\\1]]', $text );
2897 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2900 # Trim trailing whitespace
2901 # MAG_END (__END__) tag allows for trailing
2902 # whitespace to be deliberately included
2903 $text = rtrim( $text );
2904 $mw =& MagicWord
::get( MAG_END
);
2905 $mw->matchAndRemove( $text );
2911 * Set up some variables which are usually set up in parse()
2912 * so that an external function can call some class members with confidence
2915 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2916 $this->mTitle
=& $title;
2917 $this->mOptions
= $options;
2918 $this->mOutputType
= $outputType;
2919 if ( $clearState ) {
2920 $this->clearState();
2925 * Transform a MediaWiki message by replacing magic variables.
2927 * @param string $text the text to transform
2928 * @param ParserOptions $options options
2929 * @return string the text with variables substituted
2932 function transformMsg( $text, $options ) {
2934 static $executing = false;
2936 # Guard against infinite recursion
2942 $this->mTitle
= $wgTitle;
2943 $this->mOptions
= $options;
2944 $this->mOutputType
= OT_MSG
;
2945 $this->clearState();
2946 $text = $this->replaceVariables( $text );
2953 * Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2954 * Callback will be called with the text within
2955 * Transform and return the text within
2958 function setHook( $tag, $callback ) {
2959 $oldVal = @$this->mTagHooks
[$tag];
2960 $this->mTagHooks
[$tag] = $callback;
2965 * Replace <!--LINK--> link placeholders with actual links, in the buffer
2966 * Placeholders created in Skin::makeLinkObj()
2967 * Returns an array of links found, indexed by PDBK:
2971 * $options is a bit field, RLH_FOR_UPDATE to select for update
2973 function replaceLinkHolders( &$text, $options = 0 ) {
2974 global $wgUser, $wgLinkCache;
2975 global $wgOutputReplace;
2977 $fname = 'Parser::replaceLinkHolders';
2978 wfProfileIn( $fname );
2982 $sk = $this->mOptions
->getSkin();
2984 if ( !empty( $this->mLinkHolders
['namespaces'] ) ) {
2985 wfProfileIn( $fname.'-check' );
2986 $dbr =& wfGetDB( DB_SLAVE
);
2987 $page = $dbr->tableName( 'page' );
2988 $threshold = $wgUser->getOption('stubthreshold');
2991 asort( $this->mLinkHolders
['namespaces'] );
2995 foreach ( $this->mLinkHolders
['namespaces'] as $key => $val ) {
2997 $title = $this->mLinkHolders
['titles'][$key];
2999 # Skip invalid entries.
3000 # Result will be ugly, but prevents crash.
3001 if ( is_null( $title ) ) {
3004 $pdbk = $pdbks[$key] = $title->getPrefixedDBkey();
3006 # Check if it's in the link cache already
3007 if ( $wgLinkCache->getGoodLinkID( $pdbk ) ) {
3008 $colours[$pdbk] = 1;
3009 } elseif ( $wgLinkCache->isBadLink( $pdbk ) ) {
3010 $colours[$pdbk] = 0;
3012 # Not in the link cache, add it to the query
3013 if ( !isset( $current ) ) {
3015 $query = "SELECT page_id, page_namespace, page_title";
3016 if ( $threshold > 0 ) {
3017 $query .= ', page_len, page_is_redirect';
3019 $query .= " FROM $page WHERE (page_namespace=$val AND page_title IN(";
3020 } elseif ( $current != $val ) {
3022 $query .= ")) OR (page_namespace=$val AND page_title IN(";
3027 $query .= $dbr->addQuotes( $this->mLinkHolders
['dbkeys'][$key] );
3032 if ( $options & RLH_FOR_UPDATE
) {
3033 $query .= ' FOR UPDATE';
3036 $res = $dbr->query( $query, $fname );
3038 # Fetch data and form into an associative array
3039 # non-existent = broken
3042 while ( $s = $dbr->fetchObject($res) ) {
3043 $title = Title
::makeTitle( $s->page_namespace
, $s->page_title
);
3044 $pdbk = $title->getPrefixedDBkey();
3045 $wgLinkCache->addGoodLinkObj( $s->page_id
, $title );
3047 if ( $threshold > 0 ) {
3048 $size = $s->page_len
;
3049 if ( $s->page_is_redirect ||
$s->page_namespace
!= 0 ||
$size >= $threshold ) {
3050 $colours[$pdbk] = 1;
3052 $colours[$pdbk] = 2;
3055 $colours[$pdbk] = 1;
3059 wfProfileOut( $fname.'-check' );
3061 # Construct search and replace arrays
3062 wfProfileIn( $fname.'-construct' );
3063 $wgOutputReplace = array();
3064 foreach ( $this->mLinkHolders
['namespaces'] as $key => $ns ) {
3065 $pdbk = $pdbks[$key];
3066 $searchkey = "<!--LINK $key-->";
3067 $title = $this->mLinkHolders
['titles'][$key];
3068 if ( empty( $colours[$pdbk] ) ) {
3069 $wgLinkCache->addBadLinkObj( $title );
3070 $colours[$pdbk] = 0;
3071 $wgOutputReplace[$searchkey] = $sk->makeBrokenLinkObj( $title,
3072 $this->mLinkHolders
['texts'][$key],
3073 $this->mLinkHolders
['queries'][$key] );
3074 } elseif ( $colours[$pdbk] == 1 ) {
3075 $wgOutputReplace[$searchkey] = $sk->makeKnownLinkObj( $title,
3076 $this->mLinkHolders
['texts'][$key],
3077 $this->mLinkHolders
['queries'][$key] );
3078 } elseif ( $colours[$pdbk] == 2 ) {
3079 $wgOutputReplace[$searchkey] = $sk->makeStubLinkObj( $title,
3080 $this->mLinkHolders
['texts'][$key],
3081 $this->mLinkHolders
['queries'][$key] );
3084 wfProfileOut( $fname.'-construct' );
3087 wfProfileIn( $fname.'-replace' );
3089 $text = preg_replace_callback(
3090 '/(<!--LINK .*?-->)/',
3091 "wfOutputReplaceMatches",
3094 wfProfileOut( $fname.'-replace' );
3097 # Now process interwiki link holders
3098 # This is quite a bit simpler than internal links
3099 if ( !empty( $this->mInterwikiLinkHolders
['texts'] ) ) {
3100 wfProfileIn( $fname.'-interwiki' );
3101 # Make interwiki link HTML
3102 $wgOutputReplace = array();
3103 foreach( $this->mInterwikiLinkHolders
['texts'] as $key => $link ) {
3104 $title = $this->mInterwikiLinkHolders
['titles'][$key];
3105 $wgOutputReplace[$key] = $sk->makeLinkObj( $title, $link );
3108 $text = preg_replace_callback(
3109 '/<!--IWLINK (.*?)-->/',
3110 "wfOutputReplaceMatches",
3112 wfProfileOut( $fname.'-interwiki' );
3115 wfProfileOut( $fname );
3120 * Replace <!--LINK--> link placeholders with plain text of links
3121 * (not HTML-formatted).
3122 * @param string $text
3125 function replaceLinkHoldersText( $text ) {
3126 global $wgUser, $wgLinkCache;
3127 global $wgOutputReplace;
3129 $fname = 'Parser::replaceLinkHoldersText';
3130 wfProfileIn( $fname );
3132 $text = preg_replace_callback(
3133 '/<!--(LINK|IWLINK) (.*?)-->/',
3134 array( &$this, 'replaceLinkHoldersTextCallback' ),
3137 wfProfileOut( $fname );
3142 * @param array $matches
3146 function replaceLinkHoldersTextCallback( $matches ) {
3147 $type = $matches[1];
3149 if( $type == 'LINK' ) {
3150 if( isset( $this->mLinkHolders
['texts'][$key] ) ) {
3151 return $this->mLinkHolders
['texts'][$key];
3153 } elseif( $type == 'IWLINK' ) {
3154 if( isset( $this->mInterwikiLinkHolders
['texts'][$key] ) ) {
3155 return $this->mInterwikiLinkHolders
['texts'][$key];
3162 * Renders an image gallery from a text with one line per image.
3163 * text labels may be given by using |-style alternative text. E.g.
3164 * Image:one.jpg|The number "1"
3165 * Image:tree.jpg|A tree
3166 * given as text will return the HTML of a gallery with two images,
3167 * labeled 'The number "1"' and
3172 function renderImageGallery( $text ) {
3174 global $wgUser, $wgTitle;
3175 $parserOptions = ParserOptions
::newFromUser( $wgUser );
3176 $localParser = new Parser();
3178 global $wgLinkCache;
3179 $ig = new ImageGallery();
3180 $ig->setShowBytes( false );
3181 $ig->setShowFilename( false );
3182 $lines = explode( "\n", $text );
3184 foreach ( $lines as $line ) {
3185 # match lines like these:
3186 # Image:someimage.jpg|This is some image
3187 preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
3189 if ( count( $matches ) == 0 ) {
3192 $nt = Title
::newFromURL( $matches[1] );
3193 if( is_null( $nt ) ) {
3194 # Bogus title. Ignore these so we don't bomb out later.
3197 if ( isset( $matches[3] ) ) {
3198 $label = $matches[3];
3203 $html = $localParser->parse( $label , $wgTitle, $parserOptions );
3204 $html = $html->mText
;
3206 $ig->add( new Image( $nt ), $html );
3207 $wgLinkCache->addImageLinkObj( $nt );
3209 return $ig->toHTML();
3213 * Parse image options text and use it to make an image
3215 function makeImage( &$nt, $options ) {
3216 global $wgContLang, $wgUseImageResize;
3217 global $wgUser, $wgThumbLimits;
3221 # Check if the options text is of the form "options|alt text"
3223 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
3224 # * left no resizing, just left align. label is used for alt= only
3225 # * right same, but right aligned
3226 # * none same, but not aligned
3227 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
3228 # * center center the image
3229 # * framed Keep original image size, no magnify-button.
3231 $part = explode( '|', $options);
3233 $mwThumb =& MagicWord
::get( MAG_IMG_THUMBNAIL
);
3234 $mwLeft =& MagicWord
::get( MAG_IMG_LEFT
);
3235 $mwRight =& MagicWord
::get( MAG_IMG_RIGHT
);
3236 $mwNone =& MagicWord
::get( MAG_IMG_NONE
);
3237 $mwWidth =& MagicWord
::get( MAG_IMG_WIDTH
);
3238 $mwCenter =& MagicWord
::get( MAG_IMG_CENTER
);
3239 $mwFramed =& MagicWord
::get( MAG_IMG_FRAMED
);
3242 $width = $height = $framed = $thumb = false;
3243 $manual_thumb = '' ;
3245 foreach( $part as $key => $val ) {
3246 $val_parts = explode ( '=' , $val , 2 ) ;
3247 $left_part = array_shift ( $val_parts ) ;
3248 if ( $wgUseImageResize && ! is_null( $mwThumb->matchVariableStartToEnd($val) ) ) {
3250 } elseif ( $wgUseImageResize && count ( $val_parts ) == 1 && ! is_null( $mwThumb->matchVariableStartToEnd($left_part) ) ) {
3251 # use manually specified thumbnail
3253 $manual_thumb = array_shift ( $val_parts ) ;
3254 } elseif ( ! is_null( $mwRight->matchVariableStartToEnd($val) ) ) {
3255 # remember to set an alignment, don't render immediately
3257 } elseif ( ! is_null( $mwLeft->matchVariableStartToEnd($val) ) ) {
3258 # remember to set an alignment, don't render immediately
3260 } elseif ( ! is_null( $mwCenter->matchVariableStartToEnd($val) ) ) {
3261 # remember to set an alignment, don't render immediately
3263 } elseif ( ! is_null( $mwNone->matchVariableStartToEnd($val) ) ) {
3264 # remember to set an alignment, don't render immediately
3266 } elseif ( $wgUseImageResize && ! is_null( $match = $mwWidth->matchVariableStartToEnd($val) ) ) {
3267 wfDebug( "MAG_IMG_WIDTH match: $match\n" );
3268 # $match is the image width in pixels
3269 if ( preg_match( '/^([0-9]*)x([0-9]*)$/', $match, $m ) ) {
3270 $width = intval( $m[1] );
3271 $height = intval( $m[2] );
3273 $width = intval($match);
3275 } elseif ( ! is_null( $mwFramed->matchVariableStartToEnd($val) ) ) {
3281 # Strip bad stuff out of the alt text
3282 $alt = $this->replaceLinkHoldersText( $caption );
3283 $alt = Sanitizer
::stripAllTags( $alt );
3285 # Linker does the rest
3286 $sk =& $this->mOptions
->getSkin();
3287 return $sk->makeImageLinkObj( $nt, $caption, $alt, $align, $width, $height, $framed, $thumb, $manual_thumb );
3293 * @package MediaWiki
3297 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
3298 var $mCacheTime; # Timestamp on this article, or -1 for uncacheable. Used in ParserCache.
3299 var $mVersion; # Compatibility check
3300 var $mTitleText; # title text of the chosen language variant
3302 function ParserOutput( $text = '', $languageLinks = array(), $categoryLinks = array(),
3303 $containsOldMagic = false, $titletext = '' )
3305 $this->mText
= $text;
3306 $this->mLanguageLinks
= $languageLinks;
3307 $this->mCategoryLinks
= $categoryLinks;
3308 $this->mContainsOldMagic
= $containsOldMagic;
3309 $this->mCacheTime
= '';
3310 $this->mVersion
= MW_PARSER_VERSION
;
3311 $this->mTitleText
= $titletext;
3314 function getText() { return $this->mText
; }
3315 function getLanguageLinks() { return $this->mLanguageLinks
; }
3316 function getCategoryLinks() { return array_keys( $this->mCategoryLinks
); }
3317 function getCacheTime() { return $this->mCacheTime
; }
3318 function getTitleText() { return $this->mTitleText
; }
3319 function containsOldMagic() { return $this->mContainsOldMagic
; }
3320 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
3321 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
3322 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
3323 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
3324 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
3325 function setTitleText( $t ) { return wfSetVar ($this->mTitleText
, $t); }
3327 function addCategoryLink( $c ) { $this->mCategoryLinks
[$c] = 1; }
3329 function merge( $other ) {
3330 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
3331 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
3332 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
3336 * Return true if this cached output object predates the global or
3337 * per-article cache invalidation timestamps, or if it comes from
3338 * an incompatible older version.
3340 * @param string $touched the affected article's last touched timestamp
3344 function expired( $touched ) {
3345 global $wgCacheEpoch;
3346 return $this->getCacheTime() == -1 ||
// parser says it's uncacheable
3347 $this->getCacheTime() <= $touched ||
3348 $this->getCacheTime() <= $wgCacheEpoch ||
3349 !isset( $this->mVersion
) ||
3350 version_compare( $this->mVersion
, MW_PARSER_VERSION
, "lt" );
3355 * Set options of the Parser
3357 * @package MediaWiki
3361 # All variables are private
3362 var $mUseTeX; # Use texvc to expand <math> tags
3363 var $mUseDynamicDates; # Use DateFormatter to format dates
3364 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
3365 var $mAllowExternalImages; # Allow external images inline
3366 var $mSkin; # Reference to the preferred skin
3367 var $mDateFormat; # Date format index
3368 var $mEditSection; # Create "edit section" links
3369 var $mNumberHeadings; # Automatically number headings
3370 var $mAllowSpecialInclusion; # Allow inclusion of special pages
3372 function getUseTeX() { return $this->mUseTeX
; }
3373 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
3374 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
3375 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
3376 function getSkin() { return $this->mSkin
; }
3377 function getDateFormat() { return $this->mDateFormat
; }
3378 function getEditSection() { return $this->mEditSection
; }
3379 function getNumberHeadings() { return $this->mNumberHeadings
; }
3380 function getAllowSpecialInclusion() { return $this->mAllowSpecialInclusion
; }
3383 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
3384 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
3385 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
3386 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
3387 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
3388 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
3389 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
3390 function setAllowSpecialInclusion( $x ) { return wfSetVar( $this->mAllowSpecialInclusion
, $x ); }
3392 function setSkin( &$x ) { $this->mSkin
=& $x; }
3394 function ParserOptions() {
3396 $this->initialiseFromUser( $wgUser );
3400 * Get parser options
3403 function newFromUser( &$user ) {
3404 $popts = new ParserOptions
;
3405 $popts->initialiseFromUser( $user );
3409 /** Get user options */
3410 function initialiseFromUser( &$userInput ) {
3411 global $wgUseTeX, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages,
3412 $wgAllowSpecialInclusion;
3413 $fname = 'ParserOptions::initialiseFromUser';
3414 wfProfileIn( $fname );
3415 if ( !$userInput ) {
3417 $user->setLoaded( true );
3419 $user =& $userInput;
3422 $this->mUseTeX
= $wgUseTeX;
3423 $this->mUseDynamicDates
= $wgUseDynamicDates;
3424 $this->mInterwikiMagic
= $wgInterwikiMagic;
3425 $this->mAllowExternalImages
= $wgAllowExternalImages;
3426 wfProfileIn( $fname.'-skin' );
3427 $this->mSkin
=& $user->getSkin();
3428 wfProfileOut( $fname.'-skin' );
3429 $this->mDateFormat
= $user->getOption( 'date' );
3430 $this->mEditSection
= true;
3431 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
3432 $this->mAllowSpecialInclusion
= $wgAllowSpecialInclusion;
3433 wfProfileOut( $fname );
3438 * Callback function used by Parser::replaceLinkHolders()
3439 * to substitute link placeholders.
3441 function &wfOutputReplaceMatches( $matches ) {
3442 global $wgOutputReplace;
3443 return $wgOutputReplace[$matches[1]];
3447 * Return the total number of articles
3449 function wfNumberOfArticles() {
3450 global $wgNumberOfArticles;
3453 return $wgNumberOfArticles;
3457 * Return the number of files
3459 function wfNumberOfFiles() {
3460 $fname = 'Parser::wfNumberOfFiles';
3462 wfProfileIn( $fname );
3463 $dbr =& wfGetDB( DB_SLAVE
);
3464 $res = $dbr->selectField('image', 'COUNT(*)', array(), $fname );
3465 wfProfileOut( $fname );
3471 * Get various statistics from the database
3474 function wfLoadSiteStats() {
3475 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
3476 $fname = 'wfLoadSiteStats';
3478 if ( -1 != $wgNumberOfArticles ) return;
3479 $dbr =& wfGetDB( DB_SLAVE
);
3480 $s = $dbr->selectRow( 'site_stats',
3481 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
3482 array( 'ss_row_id' => 1 ), $fname
3485 if ( $s === false ) {
3488 $wgTotalViews = $s->ss_total_views
;
3489 $wgTotalEdits = $s->ss_total_edits
;
3490 $wgNumberOfArticles = $s->ss_good_articles
;
3496 * Basicly replacing " > and < with HTML entities ( ", >, <)
3498 * @param string $in Text that might contain HTML tags
3499 * @return string Escaped string
3501 function wfEscapeHTMLTagsOnly( $in ) {
3503 array( '"', '>', '<' ),
3504 array( '"', '>', '<' ),