3 // require_once('Tokenizer.php');
7 # Processes wiki markup
9 # There are two main entry points into the Parser class: parse() and preSaveTransform().
10 # The parse() function produces HTML output, preSaveTransform() produces altered wiki markup.
13 # objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
15 # NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
17 # settings: $wgUseTex*, $wgUseCategoryMagic*, $wgUseDynamicDates*, $wgInterwikiMagic*,
18 # $wgNamespacesWithSubpages, $wgLanguageCode, $wgAllowExternalImages*,
21 # * only within ParserOptions
24 #----------------------------------------
25 # Variable substitution O(N^2) attack
26 #-----------------------------------------
27 # Without countermeasures, it would be possible to attack the parser by saving a page
28 # filled with a large number of inclusions of large pages. The size of the generated
29 # page would be proportional to the square of the input size. Hence, we limit the number
30 # of inclusions of any given page, thus bringing any attack back to O(N).
33 define( "MAX_INCLUDE_REPEAT", 20 );
34 define( "MAX_INCLUDE_SIZE", 1000000 ); // 1 Million
36 # Allowed values for $mOutputType
37 define( "OT_HTML", 1 );
38 define( "OT_WIKI", 2 );
39 define( "OT_MSG", 3 );
41 # string parameter for extractTags which will cause it
42 # to strip HTML comments in addition to regular
43 # <XML>-style tags. This should not be anything we
44 # may want to use in wikisyntax
45 define( 'STRIP_COMMENTS', 'HTMLCommentStrip' );
47 # prefix for escaping, used in two functions at least
48 define( 'UNIQ_PREFIX', 'NaodW29');
51 # Constants needed for external link processing
53 define( 'URL_PROTOCOLS', 'http|https|ftp|irc|gopher|news|mailto' );
54 define( 'HTTP_PROTOCOLS', 'http|https' );
55 # Everything except bracket, space, or control characters
56 define( 'EXT_LINK_URL_CLASS', '[^]\\x00-\\x20\\x7F]' );
57 define( 'INVERSE_EXT_LINK_URL_CLASS', '[\]\\x00-\\x20\\x7F]' );
59 define( 'EXT_LINK_TEXT_CLASS', '[^\]\\x00-\\x1F\\x7F]' );
60 define( 'EXT_IMAGE_FNAME_CLASS', '[A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF]' );
61 define( 'EXT_IMAGE_EXTENSIONS', 'gif|png|jpg|jpeg' );
62 define( 'EXT_LINK_BRACKETED', '/\[(('.URL_PROTOCOLS
.'):'.EXT_LINK_URL_CLASS
.'+) *('.EXT_LINK_TEXT_CLASS
.'*?)\]/S' );
63 define( 'EXT_IMAGE_REGEX',
64 '/^('.HTTP_PROTOCOLS
.':)'. # Protocol
65 '('.EXT_LINK_URL_CLASS
.'+)\\/'. # Hostname and path
66 '('.EXT_IMAGE_FNAME_CLASS
.'+)\\.((?i)'.EXT_IMAGE_EXTENSIONS
.')$/S' # Filename
74 # Cleared with clearState():
75 var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
76 var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
79 var $mOptions, $mTitle, $mOutputType,
80 $mTemplates, // cache of already loaded templates, avoids
81 // multiple SQL queries for the same string
82 $mTemplatePath; // stores an unsorted hash of all the templates already loaded
83 // in this path. Used for loop detection.
86 $this->mTemplates
= array();
87 $this->mTemplatePath
= array();
88 $this->mTagHooks
= array();
92 function clearState() {
93 $this->mOutput
= new ParserOutput
;
94 $this->mAutonumber
= 0;
95 $this->mLastSection
= "";
96 $this->mDTopen
= false;
97 $this->mVariables
= false;
98 $this->mIncludeCount
= array();
99 $this->mStripState
= array();
100 $this->mArgStack
= array();
101 $this->mInPre
= false;
104 # First pass--just handle <nowiki> sections, pass the rest off
105 # to internalParse() which does all the real work.
107 # Returns a ParserOutput
109 function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
111 $fname = "Parser::parse";
112 wfProfileIn( $fname );
118 $this->mOptions
= $options;
119 $this->mTitle
=& $title;
120 $this->mOutputType
= OT_HTML
;
123 $text = $this->strip( $text, $this->mStripState
);
124 $text = $this->internalParse( $text, $linestart );
125 $text = $this->unstrip( $text, $this->mStripState
);
126 # Clean up special characters, only run once, next-to-last before doBlockLevels
129 # french spaces, last one Guillemet-left
130 # only if there is something before the space
131 '/(.) (\\?|:|;|!|\\302\\273)/i' => '\\1 \\2',
132 # french spaces, Guillemet-right
133 "/(\\302\\253) /i"=>"\\1 ",
134 '/<hr *>/i' => '<hr />',
135 '/<br *>/i' => '<br />',
136 '/<center *>/i' => '<div class="center">',
137 '/<\\/center *>/i' => '</div>',
138 # Clean up spare ampersands; note that we probably ought to be
139 # more careful about named entities.
140 '/&(?!:amp;|#[Xx][0-9A-fa-f]+;|#[0-9]+;|[a-zA-Z0-9]+;)/' => '&'
142 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
145 # french spaces, last one Guillemet-left
146 '/ (\\?|:|;|!|\\302\\273)/i' => ' \\1',
147 # french spaces, Guillemet-right
148 '/(\\302\\253) /i' => '\\1 ',
149 '/([^> ]+(0(1|3|9);)[^< ]*)/i' => '<span class="diacrit">\\1</span>',
150 '/<center *>/i' => '<div class="center">',
151 '/<\\/center *>/i' => '</div>'
153 $text = preg_replace( array_keys($fixtags), array_values($fixtags), $text );
156 $text = $this->doBlockLevels( $text, $linestart );
157 $text = $this->unstripNoWiki( $text, $this->mStripState
);
159 $text = $this->tidy($text);
161 $this->mOutput
->setText( $text );
162 wfProfileOut( $fname );
163 return $this->mOutput
;
166 /* static */ function getRandomString() {
167 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
170 # Replaces all occurrences of <$tag>content</$tag> in the text
171 # with a random marker and returns the new text. the output parameter
172 # $content will be an associative array filled with data on the form
173 # $unique_marker => content.
175 # If $content is already set, the additional entries will be appended
177 # If $tag is set to STRIP_COMMENTS, the function will extract
178 # <!-- HTML comments -->
180 /* static */ function extractTags($tag, $text, &$content, $uniq_prefix = ""){
181 $rnd = $uniq_prefix . '-' . $tag . Parser
::getRandomString();
188 while ( '' != $text ) {
189 if($tag==STRIP_COMMENTS
) {
190 $p = preg_split( '/<!--/i', $text, 2 );
192 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
195 if ( ( count( $p ) < 2 ) ||
( '' == $p[1] ) ) {
198 if($tag==STRIP_COMMENTS
) {
199 $q = preg_split( '/-->/i', $p[1], 2 );
201 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
203 $marker = $rnd . sprintf('%08X', $n++
);
204 $content[$marker] = $q[0];
205 $stripped .= $marker;
212 # Strips and renders <nowiki>, <pre>, <math>, <hiero>
213 # If $render is set, performs necessary rendering operations on plugins
214 # Returns the text, and fills an array with data needed in unstrip()
215 # If the $state is already a valid strip state, it adds to the state
217 # When $stripcomments is set, HTML comments <!-- like this -->
218 # will be stripped in addition to other tags. This is important
219 # for section editing, where these comments cause confusion when
220 # counting the sections in the wikisource
221 function strip( $text, &$state, $stripcomments = false ) {
222 $render = ($this->mOutputType
== OT_HTML
);
223 $html_content = array();
224 $nowiki_content = array();
225 $math_content = array();
226 $pre_content = array();
227 $comment_content = array();
228 $ext_content = array();
230 # Replace any instances of the placeholders
231 $uniq_prefix = UNIQ_PREFIX
;
232 #$text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
237 $text = Parser
::extractTags('html', $text, $html_content, $uniq_prefix);
238 foreach( $html_content as $marker => $content ) {
240 # Raw and unchecked for validity.
241 $html_content[$marker] = $content;
243 $html_content[$marker] = "<html>$content</html>";
249 $text = Parser
::extractTags('nowiki', $text, $nowiki_content, $uniq_prefix);
250 foreach( $nowiki_content as $marker => $content ) {
252 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
254 $nowiki_content[$marker] = "<nowiki>$content</nowiki>";
259 $text = Parser
::extractTags('math', $text, $math_content, $uniq_prefix);
260 foreach( $math_content as $marker => $content ){
262 if( $this->mOptions
->getUseTeX() ) {
263 $math_content[$marker] = renderMath( $content );
265 $math_content[$marker] = "<math>$content<math>";
268 $math_content[$marker] = "<math>$content</math>";
273 $text = Parser
::extractTags('pre', $text, $pre_content, $uniq_prefix);
274 foreach( $pre_content as $marker => $content ){
276 $pre_content[$marker] = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>';
278 $pre_content[$marker] = "<pre>$content</pre>";
284 $text = Parser
::extractTags(STRIP_COMMENTS
, $text, $comment_content, $uniq_prefix);
285 foreach( $comment_content as $marker => $content ){
286 $comment_content[$marker] = "<!--$content-->";
291 foreach ( $this->mTagHooks
as $tag => $callback ) {
292 $ext_contents[$tag] = array();
293 $text = Parser
::extractTags( $tag, $text, $ext_content[$tag], $uniq_prefix );
294 foreach( $ext_content[$tag] as $marker => $content ) {
296 $ext_content[$tag][$marker] = $callback( $content );
298 $ext_content[$tag][$marker] = "<$tag>$content</$tag>";
303 # Merge state with the pre-existing state, if there is one
305 $state['html'] = $state['html'] +
$html_content;
306 $state['nowiki'] = $state['nowiki'] +
$nowiki_content;
307 $state['math'] = $state['math'] +
$math_content;
308 $state['pre'] = $state['pre'] +
$pre_content;
309 $state['comment'] = $state['comment'] +
$comment_content;
311 foreach( $ext_content as $tag => $array ) {
312 if ( array_key_exists( $tag, $state ) ) {
313 $state[$tag] = $state[$tag] +
$array;
318 'html' => $html_content,
319 'nowiki' => $nowiki_content,
320 'math' => $math_content,
321 'pre' => $pre_content,
322 'comment' => $comment_content,
328 # always call unstripNoWiki() after this one
329 function unstrip( $text, &$state ) {
330 # Must expand in reverse order, otherwise nested tags will be corrupted
331 $contentDict = end( $state );
332 for ( $contentDict = end( $state ); $contentDict !== false; $contentDict = prev( $state ) ) {
333 if( key($state) != 'nowiki' && key($state) != 'html') {
334 for ( $content = end( $contentDict ); $content !== false; $content = prev( $contentDict ) ) {
335 $text = str_replace( key( $contentDict ), $content, $text );
342 # always call this after unstrip() to preserve the order
343 function unstripNoWiki( $text, &$state ) {
344 # Must expand in reverse order, otherwise nested tags will be corrupted
345 for ( $content = end($state['nowiki']); $content !== false; $content = prev( $state['nowiki'] ) ) {
346 $text = str_replace( key( $state['nowiki'] ), $content, $text );
351 for ( $content = end($state['html']); $content !== false; $content = prev( $state['html'] ) ) {
352 $text = str_replace( key( $state['html'] ), $content, $text );
359 # Add an item to the strip state
360 # Returns the unique tag which must be inserted into the stripped text
361 # The tag will be replaced with the original text in unstrip()
363 function insertStripItem( $text, &$state ) {
364 $rnd = UNIQ_PREFIX
. '-item' . Parser
::getRandomString();
373 $state['item'][$rnd] = $text;
378 # generate a list of subcategories and pages for a category
379 # depending on wfMsg("usenewcategorypage") it either calls the new
380 # or the old code. The new code will not work properly for some
381 # languages due to sorting issues, so they might want to turn it
383 function categoryMagic() {
384 $msg = wfMsg('usenewcategorypage');
385 if ( '0' == @$msg[0] )
387 return $this->oldCategoryMagic();
389 return $this->newCategoryMagic();
393 # This method generates the list of subcategories and pages for a category
394 function oldCategoryMagic () {
396 $fname = 'Parser::oldCategoryMagic';
398 if ( !$this->mOptions
->getUseCategoryMagic() ) return ; # Doesn't use categories at all
400 if ( $this->mTitle
->getNamespace() != NS_CATEGORY
) return "" ; # This ain't a category page
402 $r = "<br style=\"clear:both;\"/>\n";
405 $sk =& $this->mOptions
->getSkin() ;
407 $articles = array() ;
408 $children = array() ;
410 $id = $this->mTitle
->getArticleID() ;
413 $dbr =& wfGetDB( DB_SLAVE
);
414 $cur = $dbr->tableName( 'cur' );
415 $categorylinks = $dbr->tableName( 'categorylinks' );
417 $t = $dbr->strencode( $this->mTitle
->getDBKey() );
418 $sql = "SELECT DISTINCT cur_title,cur_namespace FROM $cur,$categorylinks " .
419 "WHERE cl_to='$t' AND cl_from=cur_id ORDER BY cl_sortkey" ;
420 $res = $dbr->query( $sql, $fname ) ;
421 while ( $x = $dbr->fetchObject ( $res ) ) $data[] = $x ;
423 # For all pages that link to this category
424 foreach ( $data AS $x )
426 $t = $wgLang->getNsText ( $x->cur_namespace
) ;
427 if ( $t != "" ) $t .= ":" ;
428 $t .= $x->cur_title
;
430 if ( $x->cur_namespace
== NS_CATEGORY
) {
431 array_push ( $children , $sk->makeLink ( $t ) ) ; # Subcategory
433 array_push ( $articles , $sk->makeLink ( $t ) ) ; # Page in this category
436 $dbr->freeResult ( $res ) ;
438 # Showing subcategories
439 if ( count ( $children ) > 0 ) {
440 $r .= '<h2>'.wfMsg('subcategories')."</h2>\n" ;
441 $r .= implode ( ', ' , $children ) ;
444 # Showing pages in this category
445 if ( count ( $articles ) > 0 ) {
446 $ti = $this->mTitle
->getText() ;
447 $h = wfMsg( 'category_header', $ti );
448 $r .= "<h2>{$h}</h2>\n" ;
449 $r .= implode ( ', ' , $articles ) ;
457 function newCategoryMagic () {
459 if ( !$this->mOptions
->getUseCategoryMagic() ) return ; # Doesn't use categories at all
461 if ( $this->mTitle
->getNamespace() != NS_CATEGORY
) return '' ; # This ain't a category page
463 $r = "<br style=\"clear:both;\"/>\n";
466 $sk =& $this->mOptions
->getSkin() ;
468 $articles = array() ;
469 $articles_start_char = array();
470 $children = array() ;
471 $children_start_char = array();
473 $id = $this->mTitle
->getArticleID() ;
476 $dbr =& wfGetDB( DB_SLAVE
);
477 $cur = $dbr->tableName( 'cur' );
478 $categorylinks = $dbr->tableName( 'categorylinks' );
480 $t = $dbr->strencode( $this->mTitle
->getDBKey() );
481 $sql = "SELECT DISTINCT cur_title,cur_namespace,cl_sortkey FROM " .
482 "$cur,$categorylinks WHERE cl_to='$t' AND cl_from=cur_id ORDER BY cl_sortkey" ;
483 $res = $dbr->query ( $sql ) ;
484 while ( $x = $dbr->fetchObject ( $res ) )
486 $t = $ns = $wgLang->getNsText ( $x->cur_namespace
) ;
487 if ( $t != '' ) $t .= ':' ;
488 $t .= $x->cur_title
;
490 if ( $x->cur_namespace
== NS_CATEGORY
) {
491 $ctitle = str_replace( '_',' ',$x->cur_title
);
492 array_push ( $children, $sk->makeKnownLink ( $t, $ctitle ) ) ; # Subcategory
494 // If there's a link from Category:A to Category:B, the sortkey of the resulting
495 // entry in the categorylinks table is Category:A, not A, which it SHOULD be.
496 // Workaround: If sortkey == "Category:".$title, than use $title for sorting,
497 // else use sortkey...
498 if ( ($ns.":".$ctitle) == $x->cl_sortkey
) {
499 array_push ( $children_start_char, $wgLang->firstChar( $x->cur_title
) );
501 array_push ( $children_start_char, $wgLang->firstChar( $x->cl_sortkey
) ) ;
504 array_push ( $articles , $sk->makeKnownLink ( $t ) ) ; # Page in this category
505 array_push ( $articles_start_char, $wgLang->firstChar( $x->cl_sortkey
) ) ;
508 $dbr->freeResult ( $res ) ;
510 $ti = $this->mTitle
->getText() ;
512 # Don't show subcategories section if there are none.
513 if ( count ( $children ) > 0 )
515 # Showing subcategories
516 $r .= '<h2>' . wfMsg( 'subcategories' ) . "</h2>\n";
518 $numchild = count( $children );
520 $r .= wfMsg( 'subcategorycount1', 1 );
522 $r .= wfMsg( 'subcategorycount' , $numchild );
526 if ( count ( $children ) > 6 ) {
528 // divide list into three equal chunks
529 $chunk = (int) (count ( $children ) / 3);
531 // get and display header
532 $r .= '<table width="100%"><tr valign="top">';
537 // loop through the chunks
538 for($startChunk = 0, $endChunk = $chunk, $chunkIndex = 0;
540 $chunkIndex++
, $startChunk = $endChunk, $endChunk +
= $chunk +
1)
544 // output all subcategories to category
545 for ($index = $startChunk ;
546 $index < $endChunk && $index < count($children);
549 // check for change of starting letter or begging of chunk
550 if ( ($children_start_char[$index] != $children_start_char[$index - 1])
551 ||
($index == $startChunk) )
553 $r .= "</ul><h3>{$children_start_char[$index]}</h3>\n<ul>";
556 $r .= "<li>{$children[$index]}</li>";
562 $r .= '</tr></table>';
564 // for short lists of subcategories to category.
566 $r .= "<h3>{$children_start_char[0]}</h3>\n";
567 $r .= '<ul><li>'.$children[0].'</li>';
568 for ($index = 1; $index < count($children); $index++
)
570 if ($children_start_char[$index] != $children_start_char[$index - 1])
572 $r .= "</ul><h3>{$children_start_char[$index]}</h3>\n<ul>";
575 $r .= "<li>{$children[$index]}</li>";
579 } # END of if ( count($children) > 0 )
581 $r .= '<h2>' . wfMsg( 'category_header', $ti ) . "</h2>\n";
583 $numart = count( $articles );
585 $r .= wfMsg( 'categoryarticlecount1', 1 );
587 $r .= wfMsg( 'categoryarticlecount' , $numart );
591 # Showing articles in this category
592 if ( count ( $articles ) > 6) {
593 $ti = $this->mTitle
->getText() ;
595 // divide list into three equal chunks
596 $chunk = (int) (count ( $articles ) / 3);
598 // get and display header
599 $r .= '<table width="100%"><tr valign="top">';
601 // loop through the chunks
602 for($startChunk = 0, $endChunk = $chunk, $chunkIndex = 0;
604 $chunkIndex++
, $startChunk = $endChunk, $endChunk +
= $chunk +
1)
609 // output all articles in category
610 for ($index = $startChunk ;
611 $index < $endChunk && $index < count($articles);
614 // check for change of starting letter or begging of chunk
615 if ( ($articles_start_char[$index] != $articles_start_char[$index - 1])
616 ||
($index == $startChunk) )
618 $r .= "</ul><h3>{$articles_start_char[$index]}</h3>\n<ul>";
621 $r .= "<li>{$articles[$index]}</li>";
627 $r .= '</tr></table>';
628 } elseif ( count ( $articles ) > 0) {
629 // for short lists of articles in categories.
630 $ti = $this->mTitle
->getText() ;
632 $r .= '<h3>'.$articles_start_char[0]."</h3>\n";
633 $r .= '<ul><li>'.$articles[0].'</li>';
634 for ($index = 1; $index < count($articles); $index++
)
636 if ($articles_start_char[$index] != $articles_start_char[$index - 1])
638 $r .= "</ul><h3>{$articles_start_char[$index]}</h3>\n<ul>";
641 $r .= "<li>{$articles[$index]}</li>";
650 # Return allowed HTML attributes
651 function getHTMLattrs () {
652 $htmlattrs = array( # Allowed attributes--no scripting, etc.
653 'title', 'align', 'lang', 'dir', 'width', 'height',
654 'bgcolor', 'clear', /* BR */ 'noshade', /* HR */
655 'cite', /* BLOCKQUOTE, Q */ 'size', 'face', 'color',
656 /* FONT */ 'type', 'start', 'value', 'compact',
657 /* For various lists, mostly deprecated but safe */
658 'summary', 'width', 'border', 'frame', 'rules',
659 'cellspacing', 'cellpadding', 'valign', 'char',
660 'charoff', 'colgroup', 'col', 'span', 'abbr', 'axis',
661 'headers', 'scope', 'rowspan', 'colspan', /* Tables */
662 'id', 'class', 'name', 'style' /* For CSS */
667 # Remove non approved attributes and javascript in css
668 function fixTagAttributes ( $t ) {
669 if ( trim ( $t ) == '' ) return '' ; # Saves runtime ;-)
670 $htmlattrs = $this->getHTMLattrs() ;
672 # Strip non-approved attributes from the tag
674 '/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e',
675 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
678 $t = str_replace ( "<></>" , "" , $t ) ; # This should fix bug 980557
680 # Strip javascript "expression" from stylesheets. Brute force approach:
681 # If anythin offensive is found, all attributes of the HTML tag are dropped
684 '/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is',
685 wfMungeToUtf8( $t ) ) )
693 # interface with html tidy, used if $wgUseTidy = true
694 function tidy ( $text ) {
695 global $wgTidyConf, $wgTidyBin, $wgTidyOpts;
696 global $wgInputEncoding, $wgOutputEncoding;
697 $fname = 'Parser::tidy';
698 wfProfileIn( $fname );
701 switch(strtoupper($wgOutputEncoding)) {
703 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)?
' -latin1':' -raw';
706 $wgTidyOpts .= ($wgInputEncoding == $wgOutputEncoding)?
' -utf8':' -raw';
709 $wgTidyOpts .= ' -raw';
712 $wrappedtext = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"'.
713 ' "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html>'.
714 '<head><title>test</title></head><body>'.$text.'</body></html>';
715 $descriptorspec = array(
716 0 => array('pipe', 'r'),
717 1 => array('pipe', 'w'),
718 2 => array('file', '/dev/null', 'a')
720 $process = proc_open("$wgTidyBin -config $wgTidyConf $wgTidyOpts", $descriptorspec, $pipes);
721 if (is_resource($process)) {
722 fwrite($pipes[0], $wrappedtext);
724 while (!feof($pipes[1])) {
725 $cleansource .= fgets($pipes[1], 1024);
728 $return_value = proc_close($process);
731 wfProfileOut( $fname );
733 if( $cleansource == '' && $text != '') {
734 wfDebug( "Tidy error detected!\n" );
735 return $text . "\n<!-- Tidy found serious XHTML errors -->\n";
741 # parse the wiki syntax used to render tables
742 function doTableStuff ( $t ) {
743 $fname = 'Parser::doTableStuff';
744 wfProfileIn( $fname );
746 $t = explode ( "\n" , $t ) ;
747 $td = array () ; # Is currently a td tag open?
748 $ltd = array () ; # Was it TD or TH?
749 $tr = array () ; # Is currently a tr tag open?
750 $ltr = array () ; # tr attributes
751 $indent_level = 0; # indent level of the table
752 foreach ( $t AS $k => $x )
755 $fc = substr ( $x , 0 , 1 ) ;
756 if ( preg_match( '/^(:*)\{\|(.*)$/', $x, $matches ) )
758 $indent_level = strlen( $matches[1] );
760 str_repeat( "<dl><dd>", $indent_level ) .
761 "<table " . $this->fixTagAttributes ( $matches[2] ) . '>' ;
762 array_push ( $td , false ) ;
763 array_push ( $ltd , '' ) ;
764 array_push ( $tr , false ) ;
765 array_push ( $ltr , '' ) ;
767 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
768 else if ( '|}' == substr ( $x , 0 , 2 ) )
771 $l = array_pop ( $ltd ) ;
772 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
773 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
775 $t[$k] = $z . str_repeat( "</dd></dl>", $indent_level );
777 else if ( '|-' == substr ( $x , 0 , 2 ) ) # Allows for |---------------
779 $x = substr ( $x , 1 ) ;
780 while ( $x != '' && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
782 $l = array_pop ( $ltd ) ;
783 if ( array_pop ( $tr ) ) $z = '</tr>' . $z ;
784 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
787 array_push ( $tr , false ) ;
788 array_push ( $td , false ) ;
789 array_push ( $ltd , '' ) ;
790 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
792 else if ( '|' == $fc ||
'!' == $fc ||
'|+' == substr ( $x , 0 , 2 ) ) # Caption
794 if ( '|+' == substr ( $x , 0 , 2 ) )
797 $x = substr ( $x , 1 ) ;
799 $after = substr ( $x , 1 ) ;
800 if ( $fc == '!' ) $after = str_replace ( '!!' , '||' , $after ) ;
801 $after = explode ( '||' , $after ) ;
803 foreach ( $after AS $theline )
808 $tra = array_pop ( $ltr ) ;
809 if ( !array_pop ( $tr ) ) $z = "<tr {$tra}>\n" ;
810 array_push ( $tr , true ) ;
811 array_push ( $ltr , '' ) ;
814 $l = array_pop ( $ltd ) ;
815 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
816 if ( $fc == '|' ) $l = 'td' ;
817 else if ( $fc == '!' ) $l = 'th' ;
818 else if ( $fc == '+' ) $l = 'caption' ;
820 array_push ( $ltd , $l ) ;
821 $y = explode ( '|' , $theline , 2 ) ;
822 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
823 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
825 array_push ( $td , true ) ;
830 # Closing open td, tr && table
831 while ( count ( $td ) > 0 )
833 if ( array_pop ( $td ) ) $t[] = '</td>' ;
834 if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
838 $t = implode ( "\n" , $t ) ;
839 # $t = $this->removeHTMLtags( $t );
840 wfProfileOut( $fname );
844 # Parses the text and adds the result to the strip state
845 # Returns the strip tag
846 function stripParse( $text, $newline, $args )
848 $text = $this->strip( $text, $this->mStripState
);
849 $text = $this->internalParse( $text, (bool)$newline, $args, false );
850 return $newline.$this->insertStripItem( $text, $this->mStripState
);
853 function internalParse( $text, $linestart, $args = array(), $isMain=true ) {
854 $fname = 'Parser::internalParse';
855 wfProfileIn( $fname );
857 $text = $this->removeHTMLtags( $text );
858 $text = $this->replaceVariables( $text, $args );
860 $text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );
862 $text = $this->doHeadings( $text );
863 if($this->mOptions
->getUseDynamicDates()) {
864 global $wgDateFormatter;
865 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
867 $text = $this->doAllQuotes( $text );
868 $text = $this->replaceExternalLinks( $text );
869 $text = $this->doMagicLinks( $text );
870 $text = $this->replaceInternalLinks ( $text );
871 $text = $this->replaceInternalLinks ( $text );
873 $text = $this->unstrip( $text, $this->mStripState
);
874 $text = $this->unstripNoWiki( $text, $this->mStripState
);
876 $text = $this->doTableStuff( $text );
877 $text = $this->formatHeadings( $text, $isMain );
878 $sk =& $this->mOptions
->getSkin();
879 $text = $sk->transformContent( $text );
881 if ( $isMain && !isset ( $this->categoryMagicDone
) ) {
882 $text .= $this->categoryMagic () ;
883 $this->categoryMagicDone
= true ;
886 wfProfileOut( $fname );
890 /* private */ function &doMagicLinks( &$text ) {
891 $text = $this->magicISBN( $text );
892 $text = $this->magicGEO( $text );
893 $text = $this->magicRFC( $text );
897 # Parse ^^ tokens and return html
898 /* private */ function doExponent ( $text )
900 $fname = 'Parser::doExponent';
901 wfProfileIn( $fname);
902 $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
903 wfProfileOut( $fname);
907 # Parse headers and return html
908 /* private */ function doHeadings( $text ) {
909 $fname = 'Parser::doHeadings';
910 wfProfileIn( $fname );
911 for ( $i = 6; $i >= 1; --$i ) {
912 $h = substr( '======', 0, $i );
913 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
914 "<h{$i}>\\1</h{$i}>\\2", $text );
916 wfProfileOut( $fname );
920 /* private */ function doAllQuotes( $text ) {
921 $fname = 'Parser::doAllQuotes';
922 wfProfileIn( $fname );
924 $lines = explode( "\n", $text );
925 foreach ( $lines as $line ) {
926 $outtext .= $this->doQuotes ( $line ) . "\n";
928 $outtext = substr($outtext, 0,-1);
929 wfProfileOut( $fname );
933 /* private */ function doQuotes( $text ) {
934 $arr = preg_split ("/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
935 if (count ($arr) == 1)
939 # First, do some preliminary work. This may shift some apostrophes from
940 # being mark-up to being text. It also counts the number of occurrences
941 # of bold and italics mark-ups.
949 # If there are ever four apostrophes, assume the first is supposed to
950 # be text, and the remaining three constitute mark-up for bold text.
951 if (strlen ($arr[$i]) == 4)
956 # If there are more than 5 apostrophes in a row, assume they're all
957 # text except for the last 5.
958 else if (strlen ($arr[$i]) > 5)
960 $arr[$i-1] .= str_repeat ("'", strlen ($arr[$i]) - 5);
963 # Count the number of occurrences of bold and italics mark-ups.
964 # We are not counting sequences of five apostrophes.
965 if (strlen ($arr[$i]) == 2) $numitalics++
; else
966 if (strlen ($arr[$i]) == 3) $numbold++
; else
967 if (strlen ($arr[$i]) == 5) { $numitalics++
; $numbold++
; }
972 # If there is an odd number of both bold and italics, it is likely
973 # that one of the bold ones was meant to be an apostrophe followed
974 # by italics. Which one we cannot know for certain, but it is more
975 # likely to be one that has a single-letter word before it.
976 if (($numbold %
2 == 1) && ($numitalics %
2 == 1))
979 $firstsingleletterword = -1;
980 $firstmultiletterword = -1;
984 if (($i %
2 == 1) and (strlen ($r) == 3))
986 $x1 = substr ($arr[$i-1], -1);
987 $x2 = substr ($arr[$i-1], -2, 1);
989 if ($firstspace == -1) $firstspace = $i;
990 } else if ($x2 == " ") {
991 if ($firstsingleletterword == -1) $firstsingleletterword = $i;
993 if ($firstmultiletterword == -1) $firstmultiletterword = $i;
999 # If there is a single-letter word, use it!
1000 if ($firstsingleletterword > -1)
1002 $arr [ $firstsingleletterword ] = "''";
1003 $arr [ $firstsingleletterword-1 ] .= "'";
1005 # If not, but there's a multi-letter word, use that one.
1006 else if ($firstmultiletterword > -1)
1008 $arr [ $firstmultiletterword ] = "''";
1009 $arr [ $firstmultiletterword-1 ] .= "'";
1011 # ... otherwise use the first one that has neither.
1012 # (notice that it is possible for all three to be -1 if, for example,
1013 # there is only one pentuple-apostrophe in the line)
1014 else if ($firstspace > -1)
1016 $arr [ $firstspace ] = "''";
1017 $arr [ $firstspace-1 ] .= "'";
1021 # Now let's actually convert our apostrophic mush to HTML!
1026 foreach ($arr as $r)
1030 if ($state == 'both')
1037 if (strlen ($r) == 2)
1040 { $output .= "</em>"; $state = ''; }
1041 else if ($state == 'strongem')
1042 { $output .= "</em>"; $state = 'strong'; }
1043 else if ($state == 'emstrong')
1044 { $output .= "</strong></em><strong>"; $state = 'strong'; }
1045 else if ($state == 'both')
1046 { $output .= "<strong><em>{$buffer}</em>"; $state = 'strong'; }
1047 else # $state can be 'strong' or ''
1048 { $output .= "<em>"; $state .= 'em'; }
1050 else if (strlen ($r) == 3)
1052 if ($state == 'strong')
1053 { $output .= "</strong>"; $state = ''; }
1054 else if ($state == 'strongem')
1055 { $output .= "</em></strong><em>"; $state = 'em'; }
1056 else if ($state == 'emstrong')
1057 { $output .= "</strong>"; $state = 'em'; }
1058 else if ($state == 'both')
1059 { $output .= "<em><strong>{$buffer}</strong>"; $state = 'em'; }
1060 else # $state can be 'em' or ''
1061 { $output .= "<strong>"; $state .= 'strong'; }
1063 else if (strlen ($r) == 5)
1065 if ($state == 'strong')
1066 { $output .= "</strong><em>"; $state = 'em'; }
1067 else if ($state == 'em')
1068 { $output .= "</em><strong>"; $state = 'strong'; }
1069 else if ($state == 'strongem')
1070 { $output .= "</em></strong>"; $state = ''; }
1071 else if ($state == 'emstrong')
1072 { $output .= "</strong></em>"; $state = ''; }
1073 else if ($state == 'both')
1074 { $output .= "<em><strong>{$buffer}</strong></em>"; $state = ''; }
1075 else # ($state == '')
1076 { $buffer = ''; $state = 'both'; }
1081 # Now close all remaining tags. Notice that the order is important.
1082 if ($state == 'strong' ||
$state == 'emstrong')
1083 $output .= "</strong>";
1084 if ($state == 'em' ||
$state == 'strongem' ||
$state == 'emstrong')
1086 if ($state == 'strongem')
1087 $output .= "</strong>";
1088 if ($state == 'both')
1089 $output .= "<strong><em>{$buffer}</em></strong>";
1094 # Note: we have to do external links before the internal ones,
1095 # and otherwise take great care in the order of things here, so
1096 # that we don't end up interpreting some URLs twice.
1098 /* private */ function replaceExternalLinks( $text ) {
1099 $fname = 'Parser::replaceExternalLinks';
1100 wfProfileIn( $fname );
1102 $sk =& $this->mOptions
->getSkin();
1103 $linktrail = wfMsg('linktrail');
1104 $bits = preg_split( EXT_LINK_BRACKETED
, $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1106 $s = $this->replaceFreeExternalLinks( array_shift( $bits ) );
1109 while ( $i<count( $bits ) ) {
1111 $protocol = $bits[$i++
];
1112 $text = $bits[$i++
];
1113 $trail = $bits[$i++
];
1115 # If the link text is an image URL, replace it with an <img> tag
1116 # This happened by accident in the original parser, but some people used it extensively
1117 $img = $this->maybeMakeImageLink( $text );
1118 if ( $img !== false ) {
1124 # No link text, e.g. [http://domain.tld/some.link]
1125 if ( $text == '' ) {
1126 # Autonumber if allowed
1127 if ( strpos( HTTP_PROTOCOLS
, $protocol ) !== false ) {
1128 $text = "[" . ++
$this->mAutonumber
. "]";
1130 # Otherwise just use the URL
1131 $text = htmlspecialchars( $url );
1134 # Have link text, e.g. [http://domain.tld/some.link text]s
1136 if ( preg_match( $linktrail, $trail, $m2 ) ) {
1142 $encUrl = htmlspecialchars( $url );
1143 # Bit in parentheses showing the URL for the printable version
1144 if( $url == $text ||
preg_match( "!$protocol://" . preg_quote( $text, "/" ) . "/?$!", $url ) ) {
1147 # Expand the URL for printable version
1148 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $encUrl ) . "</i>)</span>";
1151 # Process the trail (i.e. everything after this link up until start of the next link),
1152 # replacing any non-bracketed links
1153 $trail = $this->replaceFreeExternalLinks( $trail );
1155 $la = $sk->getExternalLinkAttributes( $url, $text );
1157 # Use the encoded URL
1158 # This means that users can paste URLs directly into the text
1159 # Funny characters like ö aren't valid in URLs anyway
1160 # This was changed in August 2004
1161 $s .= "<a href=\"{$url}\" {$la}>{$text}</a>{$dtrail}{$paren}{$trail}";
1164 wfProfileOut( $fname );
1168 # Replace anything that looks like a URL with a link
1169 function replaceFreeExternalLinks( $text ) {
1170 $bits = preg_split( '/((?:'.URL_PROTOCOLS
.'):)/', $text, -1, PREG_SPLIT_DELIM_CAPTURE
);
1171 $s = array_shift( $bits );
1174 $sk =& $this->mOptions
->getSkin();
1176 while ( $i < count( $bits ) ){
1177 $protocol = $bits[$i++
];
1178 $remainder = $bits[$i++
];
1180 if ( preg_match( '/^('.EXT_LINK_URL_CLASS
.'+)(.*)$/s', $remainder, $m ) ) {
1181 # Found some characters after the protocol that look promising
1182 $url = $protocol . $m[1];
1185 # Move trailing punctuation to $trail
1187 # If there is no left bracket, then consider right brackets fair game too
1188 if ( strpos( $url, '(' ) === false ) {
1192 $numSepChars = strspn( strrev( $url ), $sep );
1193 if ( $numSepChars ) {
1194 $trail = substr( $url, -$numSepChars ) . $trail;
1195 $url = substr( $url, 0, -$numSepChars );
1198 # Replace & from obsolete syntax with &
1199 $url = str_replace( '&', '&', $url );
1201 # Is this an external image?
1202 $text = $this->maybeMakeImageLink( $url );
1203 if ( $text === false ) {
1204 # Not an image, make a link
1205 $text = $sk->makeExternalLink( $url, $url );
1207 $s .= $text . $trail;
1209 $s .= $protocol . $remainder;
1215 function maybeMakeImageLink( $url ) {
1216 $sk =& $this->mOptions
->getSkin();
1218 if ( $this->mOptions
->getAllowExternalImages() ) {
1219 if ( preg_match( EXT_IMAGE_REGEX
, $url ) ) {
1221 $text = $sk->makeImage( htmlspecialchars( $url ) );
1227 /* private */ function replaceInternalLinks( $s ) {
1228 global $wgLang, $wgLinkCache;
1229 global $wgNamespacesWithSubpages, $wgLanguageCode;
1230 static $fname = 'Parser::replaceInternalLinks' ;
1231 wfProfileIn( $fname );
1233 wfProfileIn( $fname.'-setup' );
1235 # the % is needed to support urlencoded titles as well
1236 if ( !$tc ) { $tc = Title
::legalChars() . '#%'; }
1237 $sk =& $this->mOptions
->getSkin();
1239 $redirect = MagicWord
::get ( MAG_REDIRECT
) ;
1241 $a = explode( '[[', ' ' . $s );
1242 $s = array_shift( $a );
1243 $s = substr( $s, 1 );
1245 # Match a link having the form [[namespace:link|alternate]]trail
1247 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
1248 # Match the end of a line for a word that's not followed by whitespace,
1249 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1250 static $e2 = '/^(.*?)([a-zA-Z\x80-\xff]+)$/sD';
1252 $useLinkPrefixExtension = $wgLang->linkPrefixExtension();
1253 # Special and Media are pseudo-namespaces; no pages actually exist in them
1255 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
1257 if ( $useLinkPrefixExtension ) {
1258 if ( preg_match( $e2, $s, $m ) ) {
1259 $first_prefix = $m[2];
1262 $first_prefix = false;
1268 wfProfileOut( $fname.'-setup' );
1270 foreach ( $a as $line ) {
1271 wfProfileIn( $fname.'-prefixhandling' );
1272 if ( $useLinkPrefixExtension ) {
1273 if ( preg_match( $e2, $s, $m ) ) {
1281 $prefix = $first_prefix;
1282 $first_prefix = false;
1285 wfProfileOut( $fname.'-prefixhandling' );
1287 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1289 # fix up urlencoded title texts
1290 if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
1292 } else { # Invalid form; output directly
1293 $s .= $prefix . '[[' . $line ;
1297 /* Valid link forms:
1299 :Foobar -- override special treatment of prefix (images, language links)
1300 /Foobar -- convert to CurrentPage/Foobar
1301 /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1303 $c = substr($m[1],0,1);
1304 $noforce = ($c != ':');
1305 if( $c == '/' ) { # subpage
1306 if(substr($m[1],-1,1)=='/') { # / at end means we don't want the slash to be shown
1307 $m[1]=substr($m[1],1,strlen($m[1])-2);
1310 $noslash=substr($m[1],1);
1312 if(!empty($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()])) { # subpages allowed here
1313 $link = $this->mTitle
->getPrefixedText(). '/' . trim($noslash);
1316 } # this might be changed for ugliness reasons
1318 $link = $noslash; # no subpage allowed, use standard link
1320 } elseif( $noforce ) { # no subpage
1323 $link = substr( $m[1], 1 );
1325 $wasblank = ( '' == $text );
1329 $nt = Title
::newFromText( $link );
1331 $s .= $prefix . '[[' . $line;
1334 $ns = $nt->getNamespace();
1335 $iw = $nt->getInterWiki();
1337 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
1338 array_push( $this->mOutput
->mLanguageLinks
, $nt->getFullText() );
1339 $tmp = $prefix . $trail ;
1340 $s .= (trim($tmp) == '')?
'': $tmp;
1343 if ( $ns == NS_IMAGE
) {
1344 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1345 $wgLinkCache->addImageLinkObj( $nt );
1348 if ( $ns == NS_CATEGORY
) {
1349 $t = $nt->getText() ;
1350 $nnt = Title
::newFromText ( Namespace::getCanonicalName(NS_CATEGORY
).":".$t ) ;
1352 $wgLinkCache->suspend(); # Don't save in links/brokenlinks
1353 $pPLC=$sk->postParseLinkColour();
1354 $sk->postParseLinkColour( false );
1355 $t = $sk->makeLinkObj( $nnt, $t, '', '' , $prefix );
1356 $sk->postParseLinkColour( $pPLC );
1357 $wgLinkCache->resume();
1359 $sortkey = $wasblank ?
$this->mTitle
->getPrefixedText() : $text;
1360 $wgLinkCache->addCategoryLinkObj( $nt, $sortkey );
1361 $this->mOutput
->mCategoryLinks
[] = $t ;
1362 $s .= $prefix . $trail ;
1366 if( ( $nt->getPrefixedText() == $this->mTitle
->getPrefixedText() ) &&
1367 ( strpos( $link, '#' ) == FALSE ) ) {
1368 # Self-links are handled specially; generally de-link and change to bold.
1369 $s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
1373 if( $ns == NS_MEDIA
) {
1374 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1375 $wgLinkCache->addImageLinkObj( $nt );
1377 } elseif( $ns == NS_SPECIAL
) {
1378 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, '', $trail );
1381 $s .= $sk->makeLinkObj( $nt, $text, '', $trail, $prefix );
1383 wfProfileOut( $fname );
1387 # Some functions here used by doBlockLevels()
1389 /* private */ function closeParagraph() {
1391 if ( '' != $this->mLastSection
) {
1392 $result = '</' . $this->mLastSection
. ">\n";
1394 $this->mInPre
= false;
1395 $this->mLastSection
= '';
1398 # getCommon() returns the length of the longest common substring
1399 # of both arguments, starting at the beginning of both.
1401 /* private */ function getCommon( $st1, $st2 ) {
1402 $fl = strlen( $st1 );
1403 $shorter = strlen( $st2 );
1404 if ( $fl < $shorter ) { $shorter = $fl; }
1406 for ( $i = 0; $i < $shorter; ++
$i ) {
1407 if ( $st1{$i} != $st2{$i} ) { break; }
1411 # These next three functions open, continue, and close the list
1412 # element appropriate to the prefix character passed into them.
1414 /* private */ function openList( $char )
1416 $result = $this->closeParagraph();
1418 if ( '*' == $char ) { $result .= '<ul><li>'; }
1419 else if ( '#' == $char ) { $result .= '<ol><li>'; }
1420 else if ( ':' == $char ) { $result .= '<dl><dd>'; }
1421 else if ( ';' == $char ) {
1422 $result .= '<dl><dt>';
1423 $this->mDTopen
= true;
1425 else { $result = '<!-- ERR 1 -->'; }
1430 /* private */ function nextItem( $char ) {
1431 if ( '*' == $char ||
'#' == $char ) { return '</li><li>'; }
1432 else if ( ':' == $char ||
';' == $char ) {
1434 if ( $this->mDTopen
) { $close = '</dt>'; }
1435 if ( ';' == $char ) {
1436 $this->mDTopen
= true;
1437 return $close . '<dt>';
1439 $this->mDTopen
= false;
1440 return $close . '<dd>';
1443 return '<!-- ERR 2 -->';
1446 /* private */function closeList( $char ) {
1447 if ( '*' == $char ) { $text = '</li></ul>'; }
1448 else if ( '#' == $char ) { $text = '</li></ol>'; }
1449 else if ( ':' == $char ) {
1450 if ( $this->mDTopen
) {
1451 $this->mDTopen
= false;
1452 $text = '</dt></dl>';
1454 $text = '</dd></dl>';
1457 else { return '<!-- ERR 3 -->'; }
1461 /* private */ function doBlockLevels( $text, $linestart ) {
1462 $fname = 'Parser::doBlockLevels';
1463 wfProfileIn( $fname );
1465 # Parsing through the text line by line. The main thing
1466 # happening here is handling of block-level elements p, pre,
1467 # and making lists from lines starting with * # : etc.
1469 $textLines = explode( "\n", $text );
1471 $lastPrefix = $output = $lastLine = '';
1472 $this->mDTopen
= $inBlockElem = false;
1474 $paragraphStack = false;
1476 if ( !$linestart ) {
1477 $output .= array_shift( $textLines );
1479 foreach ( $textLines as $oLine ) {
1480 $lastPrefixLength = strlen( $lastPrefix );
1481 $preCloseMatch = preg_match("/<\\/pre/i", $oLine );
1482 $preOpenMatch = preg_match("/<pre/i", $oLine );
1483 if ( !$this->mInPre
) {
1484 # Multiple prefixes may abut each other for nested lists.
1485 $prefixLength = strspn( $oLine, '*#:;' );
1486 $pref = substr( $oLine, 0, $prefixLength );
1489 $pref2 = str_replace( ';', ':', $pref );
1490 $t = substr( $oLine, $prefixLength );
1491 $this->mInPre
= !empty($preOpenMatch);
1493 # Don't interpret any other prefixes in preformatted text
1495 $pref = $pref2 = '';
1500 if( $prefixLength && 0 == strcmp( $lastPrefix, $pref2 ) ) {
1501 # Same as the last item, so no need to deal with nesting or opening stuff
1502 $output .= $this->nextItem( substr( $pref, -1 ) );
1503 $paragraphStack = false;
1505 if ( ";" == substr( $pref, -1 ) ) {
1506 # The one nasty exception: definition lists work like this:
1507 # ; title : definition text
1508 # So we check for : in the remainder text to split up the
1509 # title and definition, without b0rking links.
1510 # FIXME: This is not foolproof. Something better in Tokenizer might help.
1511 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1513 $output .= $term . $this->nextItem( ':' );
1517 } elseif( $prefixLength ||
$lastPrefixLength ) {
1518 # Either open or close a level...
1519 $commonPrefixLength = $this->getCommon( $pref, $lastPrefix );
1520 $paragraphStack = false;
1522 while( $commonPrefixLength < $lastPrefixLength ) {
1523 $output .= $this->closeList( $lastPrefix{$lastPrefixLength-1} );
1524 --$lastPrefixLength;
1526 if ( $prefixLength <= $commonPrefixLength && $commonPrefixLength > 0 ) {
1527 $output .= $this->nextItem( $pref{$commonPrefixLength-1} );
1529 while ( $prefixLength > $commonPrefixLength ) {
1530 $char = substr( $pref, $commonPrefixLength, 1 );
1531 $output .= $this->openList( $char );
1533 if ( ';' == $char ) {
1534 # FIXME: This is dupe of code above
1535 if( preg_match( '/^(.*?(?:\s| )):(.*)$/', $t, $match ) ) {
1537 $output .= $term . $this->nextItem( ":" );
1541 ++
$commonPrefixLength;
1543 $lastPrefix = $pref2;
1545 if( 0 == $prefixLength ) {
1546 # No prefix (not in list)--go to paragraph mode
1547 $uniq_prefix = UNIQ_PREFIX
;
1548 // XXX: use a stack for nestable elements like span, table and div
1549 $openmatch = preg_match('/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|<p|<ul|<li|<\\/tr|<\\/td|<\\/th)/i', $t );
1550 $closematch = preg_match(
1551 '/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|'.
1552 '<td|<th|<div|<\\/div|<hr|<\\/pre|<\\/p|'.$uniq_prefix.'-pre|<\\/li|<\\/ul)/i', $t );
1553 if ( $openmatch or $closematch ) {
1554 $paragraphStack = false;
1555 $output .= $this->closeParagraph();
1556 if($preOpenMatch and !$preCloseMatch) {
1557 $this->mInPre
= true;
1559 if ( $closematch ) {
1560 $inBlockElem = false;
1562 $inBlockElem = true;
1564 } else if ( !$inBlockElem && !$this->mInPre
) {
1565 if ( " " == $t{0} and ( $this->mLastSection
== 'pre' or trim($t) != '' ) ) {
1567 if ($this->mLastSection
!= 'pre') {
1568 $paragraphStack = false;
1569 $output .= $this->closeParagraph().'<pre>';
1570 $this->mLastSection
= 'pre';
1574 if ( '' == trim($t) ) {
1575 if ( $paragraphStack ) {
1576 $output .= $paragraphStack.'<br />';
1577 $paragraphStack = false;
1578 $this->mLastSection
= 'p';
1580 if ($this->mLastSection
!= 'p' ) {
1581 $output .= $this->closeParagraph();
1582 $this->mLastSection
= '';
1583 $paragraphStack = '<p>';
1585 $paragraphStack = '</p><p>';
1589 if ( $paragraphStack ) {
1590 $output .= $paragraphStack;
1591 $paragraphStack = false;
1592 $this->mLastSection
= 'p';
1593 } else if ($this->mLastSection
!= 'p') {
1594 $output .= $this->closeParagraph().'<p>';
1595 $this->mLastSection
= 'p';
1601 if ($paragraphStack === false) {
1605 while ( $prefixLength ) {
1606 $output .= $this->closeList( $pref2{$prefixLength-1} );
1609 if ( '' != $this->mLastSection
) {
1610 $output .= '</' . $this->mLastSection
. '>';
1611 $this->mLastSection
= '';
1614 wfProfileOut( $fname );
1618 # Return value of a magic variable (like PAGENAME)
1619 function getVariableValue( $index ) {
1620 global $wgLang, $wgSitename, $wgServer;
1623 case MAG_CURRENTMONTH
:
1624 return $wgLang->formatNum( date( 'm' ) );
1625 case MAG_CURRENTMONTHNAME
:
1626 return $wgLang->getMonthName( date('n') );
1627 case MAG_CURRENTMONTHNAMEGEN
:
1628 return $wgLang->getMonthNameGen( date('n') );
1629 case MAG_CURRENTDAY
:
1630 return $wgLang->formatNum( date('j') );
1632 return $this->mTitle
->getText();
1634 return $this->mTitle
->getPartialURL();
1636 # return Namespace::getCanonicalName($this->mTitle->getNamespace());
1637 return $wgLang->getNsText($this->mTitle
->getNamespace()); // Patch by Dori
1638 case MAG_CURRENTDAYNAME
:
1639 return $wgLang->getWeekdayName( date('w')+
1 );
1640 case MAG_CURRENTYEAR
:
1641 return $wgLang->formatNum( date( 'Y' ) );
1642 case MAG_CURRENTTIME
:
1643 return $wgLang->time( wfTimestampNow(), false );
1644 case MAG_NUMBEROFARTICLES
:
1645 return $wgLang->formatNum( wfNumberOfArticles() );
1655 # initialise the magic variables (like CURRENTMONTHNAME)
1656 function initialiseVariables() {
1657 global $wgVariableIDs;
1658 $this->mVariables
= array();
1659 foreach ( $wgVariableIDs as $id ) {
1660 $mw =& MagicWord
::get( $id );
1661 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1665 /* private */ function replaceVariables( $text, $args = array() ) {
1666 global $wgLang, $wgScript, $wgArticlePath;
1668 # Prevent too big inclusions
1669 if(strlen($text)> MAX_INCLUDE_SIZE
)
1672 $fname = 'Parser::replaceVariables';
1673 wfProfileIn( $fname );
1676 $titleChars = Title
::legalChars();
1677 $nonBraceChars = str_replace( array( '{', '}' ), array( '', '' ), $titleChars );
1679 # This function is called recursively. To keep track of arguments we need a stack:
1680 array_push( $this->mArgStack
, $args );
1682 # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
1683 $GLOBALS['wgCurParser'] =& $this;
1686 if ( $this->mOutputType
== OT_HTML
) {
1687 # Variable substitution
1688 $text = preg_replace_callback( "/{{([$nonBraceChars]*?)}}/", 'wfVariableSubstitution', $text );
1690 # Argument substitution
1691 $text = preg_replace_callback( "/(\\n?){{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
1693 # Template substitution
1694 $regex = '/(\\n?){{(['.$nonBraceChars.']*)(\\|.*?|)}}/s';
1695 $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
1697 array_pop( $this->mArgStack
);
1699 wfProfileOut( $fname );
1703 function variableSubstitution( $matches ) {
1704 if ( !$this->mVariables
) {
1705 $this->initialiseVariables();
1707 if ( array_key_exists( $matches[1], $this->mVariables
) ) {
1708 $text = $this->mVariables
[$matches[1]];
1709 $this->mOutput
->mContainsOldMagic
= true;
1711 $text = $matches[0];
1716 # Split template arguments
1717 function getTemplateArgs( $argsString ) {
1718 if ( $argsString === '' ) {
1722 $args = explode( '|', substr( $argsString, 1 ) );
1724 # If any of the arguments contains a '[[' but no ']]', it needs to be
1725 # merged with the next arg because the '|' character between belongs
1726 # to the link syntax and not the template parameter syntax.
1727 $argc = count($args);
1729 for ( $i = 0; $i < $argc-1; $i++
) {
1730 if ( substr_count ( $args[$i], "[[" ) != substr_count ( $args[$i], "]]" ) ) {
1731 $args[$i] .= "|".$args[$i+
1];
1732 array_splice($args, $i+
1, 1);
1741 function braceSubstitution( $matches ) {
1742 global $wgLinkCache, $wgLang;
1743 $fname = 'Parser::braceSubstitution';
1750 # $newline is an optional newline character before the braces
1751 # $part1 is the bit before the first |, and must contain only title characters
1752 # $args is a list of arguments, starting from index 0, not including $part1
1754 $newline = $matches[1];
1755 $part1 = $matches[2];
1756 # If the third subpattern matched anything, it will start with |
1758 $args = $this->getTemplateArgs($matches[3]);
1759 $argc = count( $args );
1762 if ( strpos( $matches[0], '{{{' ) !== false ) {
1763 $text = $matches[0];
1770 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1771 if ( $mwSubst->matchStartAndRemove( $part1 ) ) {
1772 if ( $this->mOutputType
!= OT_WIKI
) {
1773 # Invalid SUBST not replaced at PST time
1774 # Return without further processing
1775 $text = $matches[0];
1779 } elseif ( $this->mOutputType
== OT_WIKI
) {
1780 # SUBST not found in PST pass, do nothing
1781 $text = $matches[0];
1786 # MSG, MSGNW and INT
1789 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1790 if ( $mwMsgnw->matchStartAndRemove( $part1 ) ) {
1793 # Remove obsolete MSG:
1794 $mwMsg =& MagicWord
::get( MAG_MSG
);
1795 $mwMsg->matchStartAndRemove( $part1 );
1798 # Check if it is an internal message
1799 $mwInt =& MagicWord
::get( MAG_INT
);
1800 if ( $mwInt->matchStartAndRemove( $part1 ) ) {
1801 if ( $this->incrementIncludeCount( 'int:'.$part1 ) ) {
1802 $text = wfMsgReal( $part1, $args, true );
1810 # Check for NS: (namespace expansion)
1811 $mwNs = MagicWord
::get( MAG_NS
);
1812 if ( $mwNs->matchStartAndRemove( $part1 ) ) {
1813 if ( intval( $part1 ) ) {
1814 $text = $wgLang->getNsText( intval( $part1 ) );
1817 $index = Namespace::getCanonicalIndex( strtolower( $part1 ) );
1818 if ( !is_null( $index ) ) {
1819 $text = $wgLang->getNsText( $index );
1826 # LOCALURL and LOCALURLE
1828 $mwLocal = MagicWord
::get( MAG_LOCALURL
);
1829 $mwLocalE = MagicWord
::get( MAG_LOCALURLE
);
1831 if ( $mwLocal->matchStartAndRemove( $part1 ) ) {
1832 $func = 'getLocalURL';
1833 } elseif ( $mwLocalE->matchStartAndRemove( $part1 ) ) {
1834 $func = 'escapeLocalURL';
1839 if ( $func !== '' ) {
1840 $title = Title
::newFromText( $part1 );
1841 if ( !is_null( $title ) ) {
1843 $text = $title->$func( $args[0] );
1845 $text = $title->$func();
1852 # Internal variables
1853 if ( !$this->mVariables
) {
1854 $this->initialiseVariables();
1856 if ( !$found && array_key_exists( $part1, $this->mVariables
) ) {
1857 $text = $this->mVariables
[$part1];
1859 $this->mOutput
->mContainsOldMagic
= true;
1862 # Template table test
1864 # Did we encounter this template already? If yes, it is in the cache
1865 # and we need to check for loops.
1866 if ( isset( $this->mTemplates
[$part1] ) ) {
1867 # Infinite loop test
1868 if ( isset( $this->mTemplatePath
[$part1] ) ) {
1872 # set $text to cached message.
1873 $text = $this->mTemplates
[$part1];
1877 # Load from database
1879 $title = Title
::newFromText( $part1, NS_TEMPLATE
);
1880 if ( !is_null( $title ) && !$title->isExternal() ) {
1881 # Check for excessive inclusion
1882 $dbk = $title->getPrefixedDBkey();
1883 if ( $this->incrementIncludeCount( $dbk ) ) {
1884 # This should never be reached.
1885 $article = new Article( $title );
1886 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1887 if ( $articleContent !== false ) {
1889 $text = $articleContent;
1894 # If the title is valid but undisplayable, make a link to it
1895 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1896 $text = '[[' . $title->getPrefixedText() . ']]';
1900 # Template cache array insertion
1901 $this->mTemplates
[$part1] = $text;
1905 # Recursive parsing, escaping and link table handling
1906 # Only for HTML output
1907 if ( $nowiki && $found && $this->mOutputType
== OT_HTML
) {
1908 $text = wfEscapeWikiText( $text );
1909 } elseif ( $this->mOutputType
== OT_HTML
&& $found && !$noparse) {
1910 # Clean up argument array
1911 $assocArgs = array();
1913 foreach( $args as $arg ) {
1914 $eqpos = strpos( $arg, '=' );
1915 if ( $eqpos === false ) {
1916 $assocArgs[$index++
] = $arg;
1918 $name = trim( substr( $arg, 0, $eqpos ) );
1919 $value = trim( substr( $arg, $eqpos+
1 ) );
1920 if ( $value === false ) {
1923 if ( $name !== false ) {
1924 $assocArgs[$name] = $value;
1929 # Do not enter included links in link table
1930 if ( !is_null( $title ) ) {
1931 $wgLinkCache->suspend();
1934 # Add a new element to the templace recursion path
1935 $this->mTemplatePath
[$part1] = 1;
1937 $text = $this->stripParse( $text, $newline, $assocArgs );
1939 # Resume the link cache and register the inclusion as a link
1940 if ( !is_null( $title ) ) {
1941 $wgLinkCache->resume();
1942 $wgLinkCache->addLinkObj( $title );
1945 # Empties the template path
1946 $this->mTemplatePath
= array();
1955 # Triple brace replacement -- used for template arguments
1956 function argSubstitution( $matches ) {
1957 $newline = $matches[1];
1958 $arg = trim( $matches[2] );
1959 $text = $matches[0];
1960 $inputArgs = end( $this->mArgStack
);
1962 if ( array_key_exists( $arg, $inputArgs ) ) {
1963 $text = $this->stripParse( $inputArgs[$arg], $newline, array() );
1969 # Returns true if the function is allowed to include this entity
1970 function incrementIncludeCount( $dbk ) {
1971 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1972 $this->mIncludeCount
[$dbk] = 0;
1974 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1982 # Cleans up HTML, removes dangerous tags and attributes
1983 /* private */ function removeHTMLtags( $text ) {
1984 global $wgUseTidy, $wgUserHtml;
1985 $fname = 'Parser::removeHTMLtags';
1986 wfProfileIn( $fname );
1989 $htmlpairs = array( # Tags that must be closed
1990 'b', 'del', 'i', 'ins', 'u', 'font', 'big', 'small', 'sub', 'sup', 'h1',
1991 'h2', 'h3', 'h4', 'h5', 'h6', 'cite', 'code', 'em', 's',
1992 'strike', 'strong', 'tt', 'var', 'div', 'center',
1993 'blockquote', 'ol', 'ul', 'dl', 'table', 'caption', 'pre',
1994 'ruby', 'rt' , 'rb' , 'rp', 'p'
1996 $htmlsingle = array(
1997 'br', 'hr', 'li', 'dt', 'dd'
1999 $htmlnest = array( # Tags that can be nested--??
2000 'table', 'tr', 'td', 'th', 'div', 'blockquote', 'ol', 'ul',
2001 'dl', 'font', 'big', 'small', 'sub', 'sup'
2003 $tabletags = array( # Can only appear inside table
2007 $htmlpairs = array();
2008 $htmlsingle = array();
2009 $htmlnest = array();
2010 $tabletags = array();
2013 $htmlsingle = array_merge( $tabletags, $htmlsingle );
2014 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
2016 $htmlattrs = $this->getHTMLattrs () ;
2018 # Remove HTML comments
2019 $text = preg_replace( '/(\\n *<!--.*--> *(?=\\n)|<!--.*-->)/sU', '$2', $text );
2021 $bits = explode( '<', $text );
2022 $text = array_shift( $bits );
2024 $tagstack = array(); $tablestack = array();
2025 foreach ( $bits as $x ) {
2026 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
2027 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2029 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2030 error_reporting( $prev );
2033 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2037 if ( ! in_array( $t, $htmlsingle ) &&
2038 ( $ot = @array_pop
( $tagstack ) ) != $t ) {
2039 @array_push
( $tagstack, $ot );
2042 if ( $t == 'table' ) {
2043 $tagstack = array_pop( $tablestack );
2048 # Keep track for later
2049 if ( in_array( $t, $tabletags ) &&
2050 ! in_array( 'table', $tagstack ) ) {
2052 } else if ( in_array( $t, $tagstack ) &&
2053 ! in_array ( $t , $htmlnest ) ) {
2055 } else if ( ! in_array( $t, $htmlsingle ) ) {
2056 if ( $t == 'table' ) {
2057 array_push( $tablestack, $tagstack );
2058 $tagstack = array();
2060 array_push( $tagstack, $t );
2062 # Strip non-approved attributes from the tag
2063 $newparams = $this->fixTagAttributes($params);
2067 $rest = str_replace( '>', '>', $rest );
2068 $text .= "<$slash$t $newparams$brace$rest";
2072 $text .= '<' . str_replace( '>', '>', $x);
2074 # Close off any remaining tags
2075 while ( is_array( $tagstack ) && ($t = array_pop( $tagstack )) ) {
2077 if ( $t == 'table' ) { $tagstack = array_pop( $tablestack ); }
2080 # this might be possible using tidy itself
2081 foreach ( $bits as $x ) {
2082 preg_match( '/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/',
2084 @list
( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
2085 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
2086 $newparams = $this->fixTagAttributes($params);
2087 $rest = str_replace( '>', '>', $rest );
2088 $text .= "<$slash$t $newparams$brace$rest";
2090 $text .= '<' . str_replace( '>', '>', $x);
2094 wfProfileOut( $fname );
2101 * This function accomplishes several tasks:
2102 * 1) Auto-number headings if that option is enabled
2103 * 2) Add an [edit] link to sections for logged in users who have enabled the option
2104 * 3) Add a Table of contents on the top for users who have enabled the option
2105 * 4) Auto-anchor headings
2107 * It loops through all headlines, collects the necessary data, then splits up the
2108 * string and re-inserts the newly formatted headlines.
2112 /* private */ function formatHeadings( $text, $isMain=true ) {
2113 global $wgInputEncoding, $wgMaxTocLevel;
2115 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
2116 $doShowToc = $this->mOptions
->getShowToc();
2117 $forceTocHere = false;
2118 if( !$this->mTitle
->userCanEdit() ) {
2120 $rightClickHack = 0;
2122 $showEditLink = $this->mOptions
->getEditSection();
2123 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
2126 # Inhibit editsection links if requested in the page
2127 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
2128 if( $esw->matchAndRemove( $text ) ) {
2131 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
2133 $mw =& MagicWord
::get( MAG_NOTOC
);
2134 if( $mw->matchAndRemove( $text ) ) {
2138 # never add the TOC to the Main Page. This is an entry page that should not
2139 # be more than 1-2 screens large anyway
2140 if( $this->mTitle
->getPrefixedText() == wfMsg('mainpage') ) {
2144 # Get all headlines for numbering them and adding funky stuff like [edit]
2145 # links - this is for later, but we need the number of headlines right now
2146 $numMatches = preg_match_all( '/<H([1-6])(.*?' . '>)(.*?)<\/H[1-6]>/i', $text, $matches );
2148 # if there are fewer than 4 headlines in the article, do not show TOC
2149 if( $numMatches < 4 ) {
2153 # if the string __TOC__ (not case-sensitive) occurs in the HTML,
2154 # override above conditions and always show TOC at that place
2155 $mw =& MagicWord
::get( MAG_TOC
);
2156 if ($mw->match( $text ) ) {
2158 $forceTocHere = true;
2160 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
2161 # override above conditions and always show TOC above first header
2162 $mw =& MagicWord
::get( MAG_FORCETOC
);
2163 if ($mw->matchAndRemove( $text ) ) {
2170 # We need this to perform operations on the HTML
2171 $sk =& $this->mOptions
->getSkin();
2176 # Ugh .. the TOC should have neat indentation levels which can be
2177 # passed to the skin functions. These are determined here
2182 $sublevelCount = array();
2185 foreach( $matches[3] as $headline ) {
2188 $prevlevel = $level;
2190 $level = $matches[1][$headlineCount];
2191 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
2192 # reset when we enter a new level
2193 $sublevelCount[$level] = 0;
2194 $toc .= $sk->tocIndent( $level - $prevlevel );
2195 $toclevel +
= $level - $prevlevel;
2197 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
2198 # reset when we step back a level
2199 $sublevelCount[$level+
1]=0;
2200 $toc .= $sk->tocUnindent( $prevlevel - $level );
2201 $toclevel -= $prevlevel - $level;
2203 # count number of headlines for each level
2204 @$sublevelCount[$level]++
;
2205 if( $doNumberHeadings ||
$doShowToc ) {
2207 for( $i = 1; $i <= $level; $i++
) {
2208 if( !empty( $sublevelCount[$i] ) ) {
2212 $numbering .= $sublevelCount[$i];
2218 # The canonized header is a version of the header text safe to use for links
2219 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
2220 $canonized_headline = $this->unstrip( $headline, $this->mStripState
);
2221 $canonized_headline = $this->unstripNoWiki( $headline, $this->mStripState
);
2224 $canonized_headline = preg_replace( '/<.*?' . '>/','',$canonized_headline );
2225 $tocline = trim( $canonized_headline );
2226 $canonized_headline = urlencode( do_html_entity_decode( str_replace(' ', '_', $tocline), ENT_COMPAT
, $wgInputEncoding ) );
2227 $replacearray = array(
2231 $canonized_headline = str_replace(array_keys($replacearray),array_values($replacearray),$canonized_headline);
2232 $refer[$headlineCount] = $canonized_headline;
2234 # count how many in assoc. array so we can track dupes in anchors
2235 @$refers[$canonized_headline]++
;
2236 $refcount[$headlineCount]=$refers[$canonized_headline];
2238 # Prepend the number to the heading text
2240 if( $doNumberHeadings ||
$doShowToc ) {
2241 $tocline = $numbering . ' ' . $tocline;
2243 # Don't number the heading if it is the only one (looks silly)
2244 if( $doNumberHeadings && count( $matches[3] ) > 1) {
2245 # the two are different if the line contains a link
2246 $headline=$numbering . ' ' . $headline;
2250 # Create the anchor for linking from the TOC to the section
2251 $anchor = $canonized_headline;
2252 if($refcount[$headlineCount] > 1 ) {
2253 $anchor .= '_' . $refcount[$headlineCount];
2255 if( $doShowToc && ( !isset($wgMaxTocLevel) ||
$toclevel<$wgMaxTocLevel ) ) {
2256 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
2258 if( $showEditLink ) {
2259 if ( empty( $head[$headlineCount] ) ) {
2260 $head[$headlineCount] = '';
2262 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+
1);
2265 # Add the edit section span
2266 if( $rightClickHack ) {
2267 $headline = $sk->editSectionScript($headlineCount+
1,$headline);
2270 # give headline the correct <h#> tag
2271 @$head[$headlineCount] .= "<a name=\"$anchor\"></a><h".$level.$matches[2][$headlineCount] .$headline."</h".$level.">";
2277 $toclines = $headlineCount;
2278 $toc .= $sk->tocUnindent( $toclevel );
2279 $toc = $sk->tocTable( $toc );
2282 # split up and insert constructed headlines
2284 $blocks = preg_split( '/<H[1-6].*?' . '>.*?<\/H[1-6]>/i', $text );
2287 foreach( $blocks as $block ) {
2288 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
2289 # This is the [edit] link that appears for the top block of text when
2290 # section editing is enabled
2292 # Disabled because it broke block formatting
2293 # For example, a bullet point in the top line
2294 # $full .= $sk->editSectionLink(0);
2297 if( $doShowToc && !$i && $isMain && !$forceTocHere) {
2298 # Top anchor now in skin
2302 if( !empty( $head[$i] ) ) {
2308 $mw =& MagicWord
::get( MAG_TOC
);
2309 return $mw->replace( $toc, $full );
2315 # Return an HTML link for the "ISBN 123456" text
2316 /* private */ function magicISBN( $text ) {
2318 $fname = 'Parser::magicISBN';
2319 wfProfileIn( $fname );
2321 $a = split( 'ISBN ', " $text" );
2322 if ( count ( $a ) < 2 ) {
2323 wfProfileOut( $fname );
2326 $text = substr( array_shift( $a ), 1);
2327 $valid = '0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ';
2329 foreach ( $a as $x ) {
2330 $isbn = $blank = '' ;
2331 while ( ' ' == $x{0} ) {
2333 $x = substr( $x, 1 );
2335 while ( strstr( $valid, $x{0} ) != false ) {
2337 $x = substr( $x, 1 );
2339 $num = str_replace( '-', '', $isbn );
2340 $num = str_replace( ' ', '', $num );
2343 $text .= "ISBN $blank$x";
2345 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Booksources' );
2346 $text .= '<a href="' .
2347 $titleObj->escapeLocalUrl( "isbn={$num}" ) .
2348 "\" class=\"internal\">ISBN $isbn</a>";
2352 wfProfileOut( $fname );
2356 # Return an HTML link for the "GEO ..." text
2357 /* private */ function magicGEO( $text ) {
2358 global $wgLang, $wgUseGeoMode;
2359 if ( !isset ( $wgUseGeoMode ) ||
!$wgUseGeoMode ) return $text ;
2360 $fname = 'Parser::magicGEO';
2361 wfProfileIn( $fname );
2363 # These next five lines are only for the ~35000 U.S. Census Rambot pages...
2364 $directions = array ( "N" => "North" , "S" => "South" , "E" => "East" , "W" => "West" ) ;
2365 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2366 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['N']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2367 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['W']}/" , "(GEO +\$1.\$2.\$3:-\$4.\$5.\$6)" , $text ) ;
2368 $text = preg_replace ( "/(\d+)°(\d+)'(\d+)\" {$directions['S']}, (\d+)°(\d+)'(\d+)\" {$directions['E']}/" , "(GEO +\$1.\$2.\$3:+\$4.\$5.\$6)" , $text ) ;
2370 $a = split( 'GEO ', " $text" );
2371 if ( count ( $a ) < 2 ) {
2372 wfProfileOut( $fname );
2375 $text = substr( array_shift( $a ), 1);
2376 $valid = '0123456789.+-:';
2378 foreach ( $a as $x ) {
2379 $geo = $blank = '' ;
2380 while ( ' ' == $x{0} ) {
2382 $x = substr( $x, 1 );
2384 while ( strstr( $valid, $x{0} ) != false ) {
2386 $x = substr( $x, 1 );
2388 $num = str_replace( '+', '', $geo );
2389 $num = str_replace( ' ', '', $num );
2391 if ( '' == $num ||
count ( explode ( ":" , $num , 3 ) ) < 2 ) {
2392 $text .= "GEO $blank$x";
2394 $titleObj = Title
::makeTitle( NS_SPECIAL
, 'Geo' );
2395 $text .= '<a href="' .
2396 $titleObj->escapeLocalUrl( "coordinates={$num}" ) .
2397 "\" class=\"internal\">GEO $geo</a>";
2401 wfProfileOut( $fname );
2405 # Return an HTML link for the "RFC 1234" text
2406 /* private */ function magicRFC( $text ) {
2409 $a = split( 'RFC ', ' '.$text );
2410 if ( count ( $a ) < 2 ) return $text;
2411 $text = substr( array_shift( $a ), 1);
2412 $valid = '0123456789';
2414 foreach ( $a as $x ) {
2415 $rfc = $blank = '' ;
2416 while ( ' ' == $x{0} ) {
2418 $x = substr( $x, 1 );
2420 while ( strstr( $valid, $x{0} ) != false ) {
2422 $x = substr( $x, 1 );
2426 $text .= "RFC $blank$x";
2428 $url = wfmsg( 'rfcurl' );
2429 $url = str_replace( '$1', $rfc, $url);
2430 $sk =& $this->mOptions
->getSkin();
2431 $la = $sk->getExternalLinkAttributes( $url, "RFC {$rfc}" );
2432 $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
2438 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true ) {
2439 $this->mOptions
= $options;
2440 $this->mTitle
=& $title;
2441 $this->mOutputType
= OT_WIKI
;
2443 if ( $clearState ) {
2444 $this->clearState();
2447 $stripState = false;
2451 $text = str_replace(array_keys($pairs), array_values($pairs), $text);
2455 "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
2456 "/<br *?>/i" => "<br />",
2458 $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
2460 $text = $this->strip( $text, $stripState, false );
2461 $text = $this->pstPass2( $text, $user );
2462 $text = $this->unstrip( $text, $stripState );
2463 $text = $this->unstripNoWiki( $text, $stripState );
2467 /* private */ function pstPass2( $text, &$user ) {
2468 global $wgLang, $wgLocaltimezone, $wgCurParser;
2470 # Variable replacement
2471 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
2472 $text = $this->replaceVariables( $text );
2476 $n = $user->getName();
2477 $k = $user->getOption( 'nickname' );
2478 if ( '' == $k ) { $k = $n; }
2479 if(isset($wgLocaltimezone)) {
2480 $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
2482 /* Note: this is an ugly timezone hack for the European wikis */
2483 $d = $wgLang->timeanddate( date( 'YmdHis' ), false ) .
2484 ' (' . date( 'T' ) . ')';
2485 if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
2487 $text = preg_replace( '/~~~~~/', $d, $text );
2488 $text = preg_replace( '/~~~~/', '[[' . $wgLang->getNsText( NS_USER
) . ":$n|$k]] $d", $text );
2489 $text = preg_replace( '/~~~/', '[[' . $wgLang->getNsText( NS_USER
) . ":$n|$k]]", $text );
2491 # Context links: [[|name]] and [[name (context)|]]
2493 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
2494 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
2495 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
2496 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
2498 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
2499 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
2500 $p3 = "/\[\[($namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]]
2501 $p4 = "/\[\[($namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/";
2502 # [[ns:page (cont)|]]
2504 $t = $this->mTitle
->getText();
2505 if ( preg_match( $conpat, $t, $m ) ) {
2508 $text = preg_replace( $p4, '[[\\1:\\2 (\\3)|\\2]]', $text );
2509 $text = preg_replace( $p1, '[[\\1 (\\2)|\\1]]', $text );
2510 $text = preg_replace( $p3, '[[\\1:\\2|\\2]]', $text );
2512 if ( '' == $context ) {
2513 $text = preg_replace( $p2, '[[\\1]]', $text );
2515 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
2519 $mw =& MagicWord::get( MAG_SUBST );
2520 $wgCurParser = $this->fork();
2521 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
2522 $this->merge( $wgCurParser );
2525 # Trim trailing whitespace
2526 # MAG_END (__END__) tag allows for trailing
2527 # whitespace to be deliberately included
2528 $text = rtrim( $text );
2529 $mw =& MagicWord
::get( MAG_END
);
2530 $mw->matchAndRemove( $text );
2535 # Set up some variables which are usually set up in parse()
2536 # so that an external function can call some class members with confidence
2537 function startExternalParse( &$title, $options, $outputType, $clearState = true ) {
2538 $this->mTitle
=& $title;
2539 $this->mOptions
= $options;
2540 $this->mOutputType
= $outputType;
2541 if ( $clearState ) {
2542 $this->clearState();
2546 function transformMsg( $text, $options ) {
2548 static $executing = false;
2550 # Guard against infinite recursion
2556 $this->mTitle
= $wgTitle;
2557 $this->mOptions
= $options;
2558 $this->mOutputType
= OT_MSG
;
2559 $this->clearState();
2560 $text = $this->replaceVariables( $text );
2566 # Create an HTML-style tag, e.g. <yourtag>special text</yourtag>
2567 # Callback will be called with the text within
2568 # Transform and return the text within
2569 function setHook( $tag, $callback ) {
2570 $oldVal = @$this->mTagHooks
[$tag];
2571 $this->mTagHooks
[$tag] = $callback;
2578 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
2579 var $mCacheTime; # Used in ParserCache
2581 function ParserOutput( $text = "", $languageLinks = array(), $categoryLinks = array(),
2582 $containsOldMagic = false )
2584 $this->mText
= $text;
2585 $this->mLanguageLinks
= $languageLinks;
2586 $this->mCategoryLinks
= $categoryLinks;
2587 $this->mContainsOldMagic
= $containsOldMagic;
2588 $this->mCacheTime
= "";
2591 function getText() { return $this->mText
; }
2592 function getLanguageLinks() { return $this->mLanguageLinks
; }
2593 function getCategoryLinks() { return $this->mCategoryLinks
; }
2594 function getCacheTime() { return $this->mCacheTime
; }
2595 function containsOldMagic() { return $this->mContainsOldMagic
; }
2596 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
2597 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
2598 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
2599 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
2600 function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime
, $t ); }
2602 function merge( $other ) {
2603 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
2604 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
2605 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
2612 # All variables are private
2613 var $mUseTeX; # Use texvc to expand <math> tags
2614 var $mUseCategoryMagic; # Treat [[Category:xxxx]] tags specially
2615 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
2616 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
2617 var $mAllowExternalImages; # Allow external images inline
2618 var $mSkin; # Reference to the preferred skin
2619 var $mDateFormat; # Date format index
2620 var $mEditSection; # Create "edit section" links
2621 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
2622 var $mNumberHeadings; # Automatically number headings
2623 var $mShowToc; # Show table of contents
2625 function getUseTeX() { return $this->mUseTeX
; }
2626 function getUseCategoryMagic() { return $this->mUseCategoryMagic
; }
2627 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
2628 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
2629 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
2630 function getSkin() { return $this->mSkin
; }
2631 function getDateFormat() { return $this->mDateFormat
; }
2632 function getEditSection() { return $this->mEditSection
; }
2633 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
2634 function getNumberHeadings() { return $this->mNumberHeadings
; }
2635 function getShowToc() { return $this->mShowToc
; }
2637 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
2638 function setUseCategoryMagic( $x ) { return wfSetVar( $this->mUseCategoryMagic
, $x ); }
2639 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
2640 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
2641 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
2642 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
2643 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
2644 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
2645 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
2646 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
2648 function setSkin( &$x ) { $this->mSkin
=& $x; }
2650 /* static */ function newFromUser( &$user ) {
2651 $popts = new ParserOptions
;
2652 $popts->initialiseFromUser( $user );
2656 function initialiseFromUser( &$userInput ) {
2657 global $wgUseTeX, $wgUseCategoryMagic, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
2659 if ( !$userInput ) {
2661 $user->setLoaded( true );
2663 $user =& $userInput;
2666 $this->mUseTeX
= $wgUseTeX;
2667 $this->mUseCategoryMagic
= $wgUseCategoryMagic;
2668 $this->mUseDynamicDates
= $wgUseDynamicDates;
2669 $this->mInterwikiMagic
= $wgInterwikiMagic;
2670 $this->mAllowExternalImages
= $wgAllowExternalImages;
2671 $this->mSkin
=& $user->getSkin();
2672 $this->mDateFormat
= $user->getOption( 'date' );
2673 $this->mEditSection
= $user->getOption( 'editsection' );
2674 $this->mEditSectionOnRightClick
= $user->getOption( 'editsectiononrightclick' );
2675 $this->mNumberHeadings
= $user->getOption( 'numberheadings' );
2676 $this->mShowToc
= $user->getOption( 'showtoc' );
2682 # Regex callbacks, used in Parser::replaceVariables
2683 function wfBraceSubstitution( $matches )
2685 global $wgCurParser;
2686 return $wgCurParser->braceSubstitution( $matches );
2689 function wfArgSubstitution( $matches )
2691 global $wgCurParser;
2692 return $wgCurParser->argSubstitution( $matches );
2695 function wfVariableSubstitution( $matches )
2697 global $wgCurParser;
2698 return $wgCurParser->variableSubstitution( $matches );
2701 function wfNumberOfArticles()
2703 global $wgNumberOfArticles;
2706 return $wgNumberOfArticles;
2709 /* private */ function wfLoadSiteStats()
2711 global $wgNumberOfArticles, $wgTotalViews, $wgTotalEdits;
2712 $fname = 'wfLoadSiteStats';
2714 if ( -1 != $wgNumberOfArticles ) return;
2715 $dbr =& wfGetDB( DB_SLAVE
);
2716 $s = $dbr->getArray( 'site_stats',
2717 array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
2718 array( 'ss_row_id' => 1 ), $fname
2721 if ( $s === false ) {
2724 $wgTotalViews = $s->ss_total_views
;
2725 $wgTotalEdits = $s->ss_total_edits
;
2726 $wgNumberOfArticles = $s->ss_good_articles
;
2730 function wfEscapeHTMLTagsOnly( $in ) {
2732 array( '"', '>', '<' ),
2733 array( '"', '>', '<' ),