3 include_once('Tokenizer.php');
5 if( $GLOBALS['wgUseWikiHiero'] ){
6 include_once('wikihiero.php');
11 # Processes wiki markup
13 # There are two main entry points into the Parser class: parse() and preSaveTransform().
14 # The parse() function produces HTML output, preSaveTransform() produces altered wiki markup.
17 # objects: $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
19 # NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
21 # settings: $wgUseTex*, $wgUseCategoryMagic*, $wgUseDynamicDates*, $wgInterwikiMagic*,
22 # $wgNamespacesWithSubpages, $wgLanguageCode, $wgAllowExternalImages*,
25 # * only within ParserOptions
28 #----------------------------------------
29 # Variable substitution O(N^2) attack
30 #-----------------------------------------
31 # Without countermeasures, it would be possible to attack the parser by saving a page
32 # filled with a large number of inclusions of large pages. The size of the generated
33 # page would be proportional to the square of the input size. Hence, we limit the number
34 # of inclusions of any given page, thus bringing any attack back to O(N).
36 define( "MAX_INCLUDE_REPEAT", 5 );
38 # Recursion depth of variable/inclusion evaluation
39 define( "MAX_INCLUDE_PASSES", 3 );
41 # Allowed values for $mOutputType
42 define( "OT_HTML", 1 );
43 define( "OT_WIKI", 2 );
47 # Cleared with clearState():
48 var $mOutput, $mAutonumber, $mLastSection, $mDTopen, $mStripState = array();
49 var $mVariables, $mIncludeCount;
52 var $mOptions, $mTitle, $mOutputType;
61 $this->mOutput
= new ParserOutput
;
62 $this->mAutonumber
= 0;
63 $this->mLastSection
= "";
64 $this->mDTopen
= false;
65 $this->mVariables
= false;
66 $this->mIncludeCount
= array();
67 $this->mStripState
= array();
70 # First pass--just handle <nowiki> sections, pass the rest off
71 # to doWikiPass2() which does all the real work.
73 # Returns a ParserOutput
75 function parse( $text, &$title, $options, $linestart = true, $clearState = true )
77 $fname = "Parser::parse";
78 wfProfileIn( $fname );
84 $this->mOptions
= $options;
85 $this->mTitle
=& $title;
86 $this->mOutputType
= OT_HTML
;
89 $text = $this->strip( $text, $this->mStripState
);
90 $text = $this->doWikiPass2( $text, $linestart );
91 $text = $this->unstrip( $text, $this->mStripState
);
93 $this->mOutput
->setText( $text );
94 wfProfileOut( $fname );
95 return $this->mOutput
;
98 /* static */ function getRandomString()
100 return dechex(mt_rand(0, 0x7fffffff)) . dechex(mt_rand(0, 0x7fffffff));
103 # Replaces all occurences of <$tag>content</$tag> in the text
104 # with a random marker and returns the new text. the output parameter
105 # $content will be an associative array filled with data on the form
106 # $unique_marker => content.
108 /* static */ function extractTags($tag, $text, &$content, $uniq_prefix = ""){
110 $rnd = $uniq_prefix . Parser
::getRandomString();
115 while ( "" != $text ) {
116 $p = preg_split( "/<\\s*$tag\\s*>/i", $text, 2 );
118 if ( ( count( $p ) < 2 ) ||
( "" == $p[1] ) ) {
121 $q = preg_split( "/<\\/\\s*$tag\\s*>/i", $p[1], 2 );
122 $marker = $rnd . sprintf("%08X", $n++
);
123 $content[$marker] = $q[0];
124 $stripped .= $marker;
131 # Strips <nowiki>, <pre> and <math>
132 # Returns the text, and fills an array with data needed in unstrip()
134 function strip( $text, &$state )
136 $render = ($this->mOutputType
== OT_HTML
);
137 $nowiki_content = array();
138 $hiero_content = array();
139 $math_content = array();
140 $pre_content = array();
142 # Replace any instances of the placeholders
143 $uniq_prefix = "NaodW29";
144 $text = str_replace( $uniq_prefix, wfHtmlEscapeFirst( $uniq_prefix ), $text );
146 $text = Parser
::extractTags("nowiki", $text, $nowiki_content, $uniq_prefix);
147 foreach( $nowiki_content as $marker => $content ){
149 $nowiki_content[$marker] = wfEscapeHTMLTagsOnly( $content );
151 $nowiki_content[$marker] = "<nowiki>$content</nowiki>";
155 if( $GLOBALS['wgUseWikiHiero'] ){
156 $text = Parser
::extractTags("hiero", $text, $hiero_content, $uniq_prefix);
157 foreach( $hiero_content as $marker => $content ){
159 $hiero_content[$marker] = WikiHiero( $content, WH_MODE_HTML
);
161 $hiero_content[$marker] = "<hiero>$content</hiero>";
166 if( $this->mOptions
->getUseTeX() ){
167 $text = Parser
::extractTags("math", $text, $math_content, $uniq_prefix);
168 foreach( $math_content as $marker => $content ){
170 $math_content[$marker] = renderMath( $content );
172 $math_content[$marker] = "<math>$content</math>";
177 $text = Parser
::extractTags("pre", $text, $pre_content, $uniq_prefix);
178 foreach( $pre_content as $marker => $content ){
180 $pre_content[$marker] = "<pre>" . wfEscapeHTMLTagsOnly( $content ) . "</pre>";
182 $pre_content[$marker] = "<pre>$content</pre>";
186 # Must expand in reverse order, otherwise nested tags will be corrupted
187 $state = array( $pre_content, $math_content, $hiero_content, $nowiki_content );
191 function unstrip( $text, &$state )
193 foreach( $state as $content_dict ){
194 foreach( $content_dict as $marker => $content ){
195 $text = str_replace( $marker, $content, $text );
201 function categoryMagic ()
203 global $wgLang , $wgUser ;
204 if ( !$this->mOptions
->getUseCategoryMagic() ) return ;
205 $id = $this->mTitle
->getArticleID() ;
206 $cat = $wgLang->ucfirst ( wfMsg ( "category" ) ) ;
207 $ti = $this->mTitle
->getText() ;
208 $ti = explode ( ":" , $ti , 2 ) ;
209 if ( $cat != $ti[0] ) return "" ;
210 $r = "<br break='all'/>\n" ;
212 $articles = array() ;
213 $parents = array () ;
214 $children = array() ;
217 # $sk =& $this->mGetSkin();
218 $sk =& $wgUser->getSkin() ;
221 $sql1 = "SELECT DISTINCT cur_title,cur_namespace FROM cur,links WHERE l_to={$id} AND l_from=cur_id";
222 $sql2 = "SELECT DISTINCT cur_title,cur_namespace FROM cur,brokenlinks WHERE bl_to={$id} AND bl_from=cur_id" ;
224 $res = wfQuery ( $sql1, DB_READ
) ;
225 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
227 $res = wfQuery ( $sql2, DB_READ
) ;
228 while ( $x = wfFetchObject ( $res ) ) $data[] = $x ;
231 foreach ( $data AS $x )
233 $t = $wgLang->getNsText ( $x->cur_namespace
) ;
234 if ( $t != "" ) $t .= ":" ;
235 $t .= $x->cur_title
;
237 $y = explode ( ":" , $t , 2 ) ;
238 if ( count ( $y ) == 2 && $y[0] == $cat ) {
239 array_push ( $children , $sk->makeLink ( $t , $y[1] ) ) ;
241 array_push ( $articles , $sk->makeLink ( $t ) ) ;
244 wfFreeResult ( $res ) ;
247 if ( count ( $children ) > 0 )
249 asort ( $children ) ;
250 $r .= "<h2>".wfMsg("subcategories")."</h2>\n" ;
251 $r .= implode ( ", " , $children ) ;
255 if ( count ( $articles ) > 0 )
257 asort ( $articles ) ;
258 $h = wfMsg( "category_header", $ti[1] );
259 $r .= "<h2>{$h}</h2>\n" ;
260 $r .= implode ( ", " , $articles ) ;
267 function getHTMLattrs ()
269 $htmlattrs = array( # Allowed attributes--no scripting, etc.
270 "title", "align", "lang", "dir", "width", "height",
271 "bgcolor", "clear", /* BR */ "noshade", /* HR */
272 "cite", /* BLOCKQUOTE, Q */ "size", "face", "color",
273 /* FONT */ "type", "start", "value", "compact",
274 /* For various lists, mostly deprecated but safe */
275 "summary", "width", "border", "frame", "rules",
276 "cellspacing", "cellpadding", "valign", "char",
277 "charoff", "colgroup", "col", "span", "abbr", "axis",
278 "headers", "scope", "rowspan", "colspan", /* Tables */
279 "id", "class", "name", "style" /* For CSS */
284 function fixTagAttributes ( $t )
286 if ( trim ( $t ) == "" ) return "" ; # Saves runtime ;-)
287 $htmlattrs = $this->getHTMLattrs() ;
289 # Strip non-approved attributes from the tag
291 "/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e",
292 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
294 # Strip javascript "expression" from stylesheets. Brute force approach:
295 # If anythin offensive is found, all attributes of the HTML tag are dropped
298 "/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is",
299 wfMungeToUtf8( $t ) ) )
307 function doTableStuff ( $t )
309 $t = explode ( "\n" , $t ) ;
310 $td = array () ; # Is currently a td tag open?
311 $ltd = array () ; # Was it TD or TH?
312 $tr = array () ; # Is currently a tr tag open?
313 $ltr = array () ; # tr attributes
314 foreach ( $t AS $k => $x )
317 $fc = substr ( $x , 0 , 1 ) ;
318 if ( "{|" == substr ( $x , 0 , 2 ) )
320 $t[$k] = "<table " . $this->fixTagAttributes ( substr ( $x , 3 ) ) . ">" ;
321 array_push ( $td , false ) ;
322 array_push ( $ltd , "" ) ;
323 array_push ( $tr , false ) ;
324 array_push ( $ltr , "" ) ;
326 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
327 else if ( "|}" == substr ( $x , 0 , 2 ) )
330 $l = array_pop ( $ltd ) ;
331 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
332 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
336 /* else if ( "|_" == substr ( $x , 0 , 2 ) ) # Caption
338 $z = trim ( substr ( $x , 2 ) ) ;
339 $t[$k] = "<caption>{$z}</caption>\n" ;
341 else if ( "|-" == substr ( $x , 0 , 2 ) ) # Allows for |---------------
343 $x = substr ( $x , 1 ) ;
344 while ( $x != "" && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
346 $l = array_pop ( $ltd ) ;
347 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
348 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
351 array_push ( $tr , false ) ;
352 array_push ( $td , false ) ;
353 array_push ( $ltd , "" ) ;
354 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
356 else if ( "|" == $fc ||
"!" == $fc ||
"|+" == substr ( $x , 0 , 2 ) ) # Caption
358 if ( "|+" == substr ( $x , 0 , 2 ) )
361 $x = substr ( $x , 1 ) ;
363 $after = substr ( $x , 1 ) ;
364 if ( $fc == "!" ) $after = str_replace ( "!!" , "||" , $after ) ;
365 $after = explode ( "||" , $after ) ;
367 foreach ( $after AS $theline )
372 $tra = array_pop ( $ltr ) ;
373 if ( !array_pop ( $tr ) ) $z = "<tr {$tra}>\n" ;
374 array_push ( $tr , true ) ;
375 array_push ( $ltr , "" ) ;
378 $l = array_pop ( $ltd ) ;
379 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
380 if ( $fc == "|" ) $l = "TD" ;
381 else if ( $fc == "!" ) $l = "TH" ;
382 else if ( $fc == "+" ) $l = "CAPTION" ;
384 array_push ( $ltd , $l ) ;
385 $y = explode ( "|" , $theline , 2 ) ;
386 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
387 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
389 array_push ( $td , true ) ;
394 # Closing open td, tr && table
395 while ( count ( $td ) > 0 )
397 if ( array_pop ( $td ) ) $t[] = "</td>" ;
398 if ( array_pop ( $tr ) ) $t[] = "</tr>" ;
402 $t = implode ( "\n" , $t ) ;
403 # $t = $this->removeHTMLtags( $t );
407 # Well, OK, it's actually about 14 passes. But since all the
408 # hard lifting is done inside PHP's regex code, it probably
409 # wouldn't speed things up much to add a real parser.
411 function doWikiPass2( $text, $linestart )
413 $fname = "Parser::doWikiPass2";
414 wfProfileIn( $fname );
416 $text = $this->removeHTMLtags( $text );
417 $text = $this->replaceVariables( $text );
419 # $text = preg_replace( "/(^|\n)-----*/", "\\1<hr>", $text );
420 $text = str_replace ( "<HR>", "<hr/>", $text );
422 $text = $this->doHeadings( $text );
423 $text = $this->doBlockLevels( $text, $linestart );
425 if($this->mOptions
->getUseDynamicDates()) {
426 global $wgDateFormatter;
427 $text = $wgDateFormatter->reformat( $this->mOptions
->getDateFormat(), $text );
430 $text = $this->replaceExternalLinks( $text );
431 $text = $this->doTokenizedParser ( $text );
432 $text = $this->doTableStuff ( $text ) ;
434 $text = $this->formatHeadings( $text );
436 $sk =& $this->mOptions
->getSkin();
437 $text = $sk->transformContent( $text );
438 $text .= $this->categoryMagic () ;
440 wfProfileOut( $fname );
445 /* private */ function doHeadings( $text )
447 for ( $i = 6; $i >= 1; --$i ) {
448 $h = substr( "======", 0, $i );
449 $text = preg_replace( "/^{$h}(.+){$h}(\\s|$)/m",
450 "<h{$i}>\\1</h{$i}>\\2", $text );
455 # Note: we have to do external links before the internal ones,
456 # and otherwise take great care in the order of things here, so
457 # that we don't end up interpreting some URLs twice.
459 /* private */ function replaceExternalLinks( $text )
461 $fname = "Parser::replaceExternalLinks";
462 wfProfileIn( $fname );
463 $text = $this->subReplaceExternalLinks( $text, "http", true );
464 $text = $this->subReplaceExternalLinks( $text, "https", true );
465 $text = $this->subReplaceExternalLinks( $text, "ftp", false );
466 $text = $this->subReplaceExternalLinks( $text, "irc", false );
467 $text = $this->subReplaceExternalLinks( $text, "gopher", false );
468 $text = $this->subReplaceExternalLinks( $text, "news", false );
469 $text = $this->subReplaceExternalLinks( $text, "mailto", false );
470 wfProfileOut( $fname );
474 /* private */ function subReplaceExternalLinks( $s, $protocol, $autonumber )
476 $unique = "4jzAfzB8hNvf4sqyO9Edd8pSmk9rE2in0Tgw3";
477 $uc = "A-Za-z0-9_\\/~%\\-+&*#?!=()@\\x80-\\xFF";
479 # this is the list of separators that should be ignored if they
480 # are the last character of an URL but that should be included
481 # if they occur within the URL, e.g. "go to www.foo.com, where .."
482 # in this case, the last comma should not become part of the URL,
483 # but in "www.foo.com/123,2342,32.htm" it should.
485 $fnc = "A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF";
486 $images = "gif|png|jpg|jpeg";
488 # PLEASE NOTE: The curly braces { } are not part of the regex,
489 # they are interpreted as part of the string (used to tell PHP
490 # that the content of the string should be inserted there).
491 $e1 = "/(^|[^\\[])({$protocol}:)([{$uc}{$sep}]+)\\/([{$fnc}]+)\\." .
492 "((?i){$images})([^{$uc}]|$)/";
494 $e2 = "/(^|[^\\[])({$protocol}:)(([".$uc."]|[".$sep."][".$uc."])+)([^". $uc . $sep. "]|[".$sep."]|$)/";
495 $sk =& $this->mOptions
->getSkin();
497 if ( $autonumber and $this->mOptions
->getAllowExternalImages() ) { # Use img tags only for HTTP urls
498 $s = preg_replace( $e1, "\\1" . $sk->makeImage( "{$unique}:\\3" .
499 "/\\4.\\5", "\\4.\\5" ) . "\\6", $s );
501 $s = preg_replace( $e2, "\\1" . "<a href=\"{$unique}:\\3\"" .
502 $sk->getExternalLinkAttributes( "{$unique}:\\3", wfEscapeHTML(
503 "{$unique}:\\3" ) ) . ">" . wfEscapeHTML( "{$unique}:\\3" ) .
505 $s = str_replace( $unique, $protocol, $s );
507 $a = explode( "[{$protocol}:", " " . $s );
508 $s = array_shift( $a );
509 $s = substr( $s, 1 );
511 $e1 = "/^([{$uc}"."{$sep}]+)](.*)\$/sD";
512 $e2 = "/^([{$uc}"."{$sep}]+)\\s+([^\\]]+)](.*)\$/sD";
514 foreach ( $a as $line ) {
515 if ( preg_match( $e1, $line, $m ) ) {
516 $link = "{$protocol}:{$m[1]}";
518 if ( $autonumber ) { $text = "[" . ++
$this->mAutonumber
. "]"; }
519 else { $text = wfEscapeHTML( $link ); }
520 } else if ( preg_match( $e2, $line, $m ) ) {
521 $link = "{$protocol}:{$m[1]}";
525 $s .= "[{$protocol}:" . $line;
528 if( $link == $text ||
preg_match( "!$protocol://" . preg_quote( $text, "/" ) . "/?$!", $link ) ) {
531 # Expand the URL for printable version
532 $paren = "<span class='urlexpansion'> (<i>" . htmlspecialchars ( $link ) . "</i>)</span>";
534 $la = $sk->getExternalLinkAttributes( $link, $text );
535 $s .= "<a href='{$link}'{$la}>{$text}</a>{$paren}{$trail}";
541 /* private */ function handle3Quotes( &$state, $token )
543 if ( $state["strong"] ) {
544 if ( $state["em"] && $state["em"] > $state["strong"] )
546 # ''' lala ''lala '''
547 $s = "</em></strong><em>";
551 $state["strong"] = FALSE;
554 $state["strong"] = $token["pos"];
559 /* private */ function handle2Quotes( &$state, $token )
561 if ( $state["em"] ) {
562 if ( $state["strong"] && $state["strong"] > $state["em"] )
564 # ''lala'''lala'' ....'''
565 $s = "</strong></em><strong>";
569 $state["em"] = FALSE;
572 $state["em"] = $token["pos"];
577 /* private */ function handle5Quotes( &$state, $token )
580 if ( $state["em"] && $state["strong"] ) {
581 if ( $state["em"] < $state["strong"] ) {
582 $s .= "</strong></em>";
584 $s .= "</em></strong>";
586 $state["strong"] = $state["em"] = FALSE;
587 } elseif ( $state["em"] ) {
588 $s .= "</em><strong>";
589 $state["em"] = FALSE;
590 $state["strong"] = $token["pos"];
591 } elseif ( $state["strong"] ) {
592 $s .= "</strong><em>";
593 $state["strong"] = FALSE;
594 $state["em"] = $token["pos"];
595 } else { # not $em and not $strong
596 $s .= "<strong><em>";
597 $state["strong"] = $state["em"] = $token["pos"];
602 /* private */ function doTokenizedParser( $str )
604 global $wgLang; # for language specific parser hook
606 $tokenizer=Tokenizer
::newFromString( $str );
607 $tokenStack = array();
610 $state["em"] = FALSE;
611 $state["strong"] = FALSE;
615 # The tokenizer splits the text into tokens and returns them one by one.
616 # Every call to the tokenizer returns a new token.
617 while ( $token = $tokenizer->nextToken() )
620 switch ( $token["type"] )
623 # simple text with no further markup
624 $txt = $token["text"];
627 # remember the tag opened with 3 [
631 # FIXME : Treat orphaned open tags (stack not empty when text is over)
633 array_push( $tokenStack, $token );
640 # get text from stack, glue it together, and call the code to handle a
643 if ( count( $tokenStack ) == 0 )
645 # stack empty. Found a ]] without an opening [[
649 $lastToken = array_pop( $tokenStack );
650 while ( !(($lastToken["type"] == "[[[") or ($lastToken["type"] == "[[")) )
652 if( !empty( $lastToken["text"] ) ) {
653 $linkText = $lastToken["text"] . $linkText;
655 $lastToken = array_pop( $tokenStack );
658 $txt = $linkText ."]]";
660 if( isset( $lastToken["text"] ) ) {
661 $prefix = $lastToken["text"];
665 $nextToken = $tokenizer->previewToken();
666 if ( $nextToken["type"] == "text" )
668 # Preview just looks at it. Now we have to fetch it.
669 $nextToken = $tokenizer->nextToken();
670 $txt .= $nextToken["text"];
672 $txt = $this->handleInternalLink( $txt, $prefix );
674 # did the tag start with 3 [ ?
676 # show the first as text
682 $tagIsOpen = (count( $tokenStack ) != 0);
688 # This and the three next ones handle quotes
689 $txt = $this->handle3Quotes( $state, $token );
692 $txt = $this->handle2Quotes( $state, $token );
695 $txt = $this->handle5Quotes( $state, $token );
705 $txt = $this->doMagicRFC( $tokenizer );
712 $txt = $this->doMagicISBN( $tokenizer );
716 # Call language specific Hook.
717 $txt = $wgLang->processToken( $token, $tokenStack );
718 if ( NULL == $txt ) {
719 # An unkown token. Highlight.
720 $txt = "<font color=\"#FF0000\"><b>".$token["type"]."</b></font>";
721 $txt .= "<font color=\"#FFFF00\"><b>".$token["text"]."</b></font>";
725 # If we're parsing the interior of a link, don't append the interior to $s,
726 # but push it to the stack so it can be processed when a ]] token is found.
727 if ( $tagIsOpen && $txt != "" ) {
728 $token["type"] = "text";
729 $token["text"] = $txt;
730 array_push( $tokenStack, $token );
735 if ( count( $tokenStack ) != 0 )
737 # still objects on stack. opened [[ tag without closing ]] tag.
739 while ( $lastToken = array_pop( $tokenStack ) )
741 if ( $lastToken["type"] == "text" )
743 $txt = $lastToken["text"] . $txt;
745 $txt = $lastToken["type"] . $txt;
753 /* private */ function handleInternalLink( $line, $prefix )
755 global $wgLang, $wgLinkCache;
756 global $wgNamespacesWithSubpages, $wgLanguageCode;
757 static $fname = "Parser::handleInternalLink" ;
758 wfProfileIn( $fname );
760 wfProfileIn( "$fname-setup" );
762 if ( !$tc ) { $tc = Title
::legalChars() . "#"; }
763 $sk =& $this->mOptions
->getSkin();
765 # Match a link having the form [[namespace:link|alternate]]trail
767 if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD"; }
768 # Match the end of a line for a word that's not followed by whitespace,
769 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
770 #$e2 = "/^(.*)\\b(\\w+)\$/suD";
771 #$e2 = "/^(.*\\s)(\\S+)\$/suD";
772 static $e2 = '/^(.*\s)([a-zA-Z\x80-\xff]+)$/sD';
775 # Special and Media are pseudo-namespaces; no pages actually exist in them
776 static $image = FALSE;
777 static $special = FALSE;
778 static $media = FALSE;
779 static $category = FALSE;
780 if ( !$image ) { $image = Namespace::getImage(); }
781 if ( !$special ) { $special = Namespace::getSpecial(); }
782 if ( !$media ) { $media = Namespace::getMedia(); }
783 if ( !$category ) { $category = wfMsg ( "category" ) ; }
785 $nottalk = !Namespace::isTalk( $this->mTitle
->getNamespace() );
787 wfProfileOut( "$fname-setup" );
790 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
793 } else { # Invalid form; output directly
794 $s .= $prefix . "[[" . $line ;
800 :Foobar -- override special treatment of prefix (images, language links)
801 /Foobar -- convert to CurrentPage/Foobar
802 /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
804 $c = substr($m[1],0,1);
805 $noforce = ($c != ":");
806 if( $c == "/" ) { # subpage
807 if(substr($m[1],-1,1)=="/") { # / at end means we don't want the slash to be shown
808 $m[1]=substr($m[1],1,strlen($m[1])-2);
811 $noslash=substr($m[1],1);
813 if($wgNamespacesWithSubpages[$this->mTitle
->getNamespace()]) { # subpages allowed here
814 $link = $this->mTitle
->getPrefixedText(). "/" . trim($noslash);
817 } # this might be changed for ugliness reasons
819 $link = $noslash; # no subpage allowed, use standard link
821 } elseif( $noforce ) { # no subpage
824 $link = substr( $m[1], 1 );
829 $nt = Title
::newFromText( $link );
831 $s .= $prefix . "[[" . $line;
834 $ns = $nt->getNamespace();
835 $iw = $nt->getInterWiki();
837 if( $iw && $this->mOptions
->getInterwikiMagic() && $nottalk && $wgLang->getLanguageName( $iw ) ) {
838 array_push( $this->mOutput
->mLanguageLinks
, $nt->getPrefixedText() );
839 $s .= $prefix . $trail;
842 if( $ns == $image ) {
843 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
844 $wgLinkCache->addImageLinkObj( $nt );
848 if( ( $nt->getPrefixedText() == $this->mTitle
->getPrefixedText() ) &&
849 ( strpos( $link, "#" ) == FALSE ) ) {
850 $s .= $prefix . "<strong>" . $text . "</strong>" . $trail;
855 $catns = strtoupper ( $nt->getDBkey () ) ;
856 $catns = explode ( ":" , $catns ) ;
857 if ( count ( $catns ) > 1 ) $catns = array_shift ( $catns ) ;
859 if ( $catns == strtoupper($category) && $this->mOptions
->getUseCategoryMagic() ) {
860 $t = explode ( ":" , $nt->getText() ) ;
862 $t = implode ( ":" , $t ) ;
863 $t = $wgLang->ucFirst ( $t ) ;
864 $nnt = Title
::newFromText ( $category.":".$t ) ;
865 $t = $sk->makeLinkObj( $nnt, $t, "", $trail , $prefix );
866 $this->mOutput
->mCategoryLinks
[] = $t ;
867 $s .= $prefix . $trail ;
871 if( $ns == $media ) {
872 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
873 $wgLinkCache->addImageLinkObj( $nt );
875 } elseif( $ns == $special ) {
876 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, "", $trail );
879 $s .= $sk->makeLinkObj( $nt, $text, "", $trail , $prefix );
881 wfProfileOut( $fname );
885 # Some functions here used by doBlockLevels()
887 /* private */ function closeParagraph()
890 if ( 0 != strcmp( "", $this->mLastSection
) ) {
891 $result = "</" . $this->mLastSection
. ">";
893 $this->mLastSection
= "";
896 # getCommon() returns the length of the longest common substring
897 # of both arguments, starting at the beginning of both.
899 /* private */ function getCommon( $st1, $st2 )
901 $fl = strlen( $st1 );
902 $shorter = strlen( $st2 );
903 if ( $fl < $shorter ) { $shorter = $fl; }
905 for ( $i = 0; $i < $shorter; ++
$i ) {
906 if ( $st1{$i} != $st2{$i} ) { break; }
910 # These next three functions open, continue, and close the list
911 # element appropriate to the prefix character passed into them.
913 /* private */ function openList( $char )
915 $result = $this->closeParagraph();
917 if ( "*" == $char ) { $result .= "<ul><li>"; }
918 else if ( "#" == $char ) { $result .= "<ol><li>"; }
919 else if ( ":" == $char ) { $result .= "<dl><dd>"; }
920 else if ( ";" == $char ) {
921 $result .= "<dl><dt>";
922 $this->mDTopen
= true;
924 else { $result = "<!-- ERR 1 -->"; }
929 /* private */ function nextItem( $char )
931 if ( "*" == $char ||
"#" == $char ) { return "</li><li>"; }
932 else if ( ":" == $char ||
";" == $char ) {
934 if ( $this->mDTopen
) { $close = "</dt>"; }
935 if ( ";" == $char ) {
936 $this->mDTopen
= true;
937 return $close . "<dt>";
939 $this->mDTopen
= false;
940 return $close . "<dd>";
943 return "<!-- ERR 2 -->";
946 /* private */function closeList( $char )
948 if ( "*" == $char ) { $text = "</li></ul>"; }
949 else if ( "#" == $char ) { $text = "</li></ol>"; }
950 else if ( ":" == $char ) {
951 if ( $this->mDTopen
) {
952 $this->mDTopen
= false;
953 $text = "</dt></dl>";
955 $text = "</dd></dl>";
958 else { return "<!-- ERR 3 -->"; }
962 /* private */ function doBlockLevels( $text, $linestart )
964 $fname = "Parser::doBlockLevels";
965 wfProfileIn( $fname );
966 # Parsing through the text line by line. The main thing
967 # happening here is handling of block-level elements p, pre,
968 # and making lists from lines starting with * # : etc.
970 $a = explode( "\n", $text );
971 $text = $lastPref = "";
972 $this->mDTopen
= $inBlockElem = false;
974 if ( ! $linestart ) { $text .= array_shift( $a ); }
975 foreach ( $a as $t ) {
976 if ( "" != $text ) { $text .= "\n"; }
979 $opl = strlen( $lastPref );
980 $npl = strspn( $t, "*#:;" );
981 $pref = substr( $t, 0, $npl );
982 $pref2 = str_replace( ";", ":", $pref );
983 $t = substr( $t, $npl );
985 if ( 0 != $npl && 0 == strcmp( $lastPref, $pref2 ) ) {
986 $text .= $this->nextItem( substr( $pref, -1 ) );
988 if ( ";" == substr( $pref, -1 ) ) {
989 $cpos = strpos( $t, ":" );
990 if ( ! ( false === $cpos ) ) {
991 $term = substr( $t, 0, $cpos );
992 $text .= $term . $this->nextItem( ":" );
993 $t = substr( $t, $cpos +
1 );
996 } else if (0 != $npl ||
0 != $opl) {
997 $cpl = $this->getCommon( $pref, $lastPref );
999 while ( $cpl < $opl ) {
1000 $text .= $this->closeList( $lastPref{$opl-1} );
1003 if ( $npl <= $cpl && $cpl > 0 ) {
1004 $text .= $this->nextItem( $pref{$cpl-1} );
1006 while ( $npl > $cpl ) {
1007 $char = substr( $pref, $cpl, 1 );
1008 $text .= $this->openList( $char );
1010 if ( ";" == $char ) {
1011 $cpos = strpos( $t, ":" );
1012 if ( ! ( false === $cpos ) ) {
1013 $term = substr( $t, 0, $cpos );
1014 $text .= $term . $this->nextItem( ":" );
1015 $t = substr( $t, $cpos +
1 );
1022 if ( 0 == $npl ) { # No prefix--go to paragraph mode
1024 "/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6|<p)/i", $t ) ) {
1025 $text .= $this->closeParagraph();
1026 $inBlockElem = true;
1028 if ( ! $inBlockElem ) {
1029 if ( " " == $t{0} ) {
1030 $newSection = "pre";
1031 # $t = wfEscapeHTML( $t );
1033 else { $newSection = "p"; }
1035 if ( 0 == strcmp( "", trim( $oLine ) ) ) {
1036 $text .= $this->closeParagraph();
1037 $text .= "<" . $newSection . ">";
1038 } else if ( 0 != strcmp( $this->mLastSection
,
1040 $text .= $this->closeParagraph();
1041 if ( 0 != strcmp( "p", $newSection ) ) {
1042 $text .= "<" . $newSection . ">";
1045 $this->mLastSection
= $newSection;
1047 if ( $inBlockElem &&
1048 preg_match( "/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6|<\\/p)/i", $t ) ) {
1049 $inBlockElem = false;
1055 $text .= $this->closeList( $pref2{$npl-1} );
1058 if ( "" != $this->mLastSection
) {
1059 if ( "p" != $this->mLastSection
) {
1060 $text .= "</" . $this->mLastSection
. ">";
1062 $this->mLastSection
= "";
1064 wfProfileOut( $fname );
1068 function getVariableValue( $index ) {
1072 case MAG_CURRENTMONTH
:
1074 case MAG_CURRENTMONTHNAME
:
1075 return $wgLang->getMonthName( date("n") );
1076 case MAG_CURRENTMONTHNAMEGEN
:
1077 return $wgLang->getMonthNameGen( date("n") );
1078 case MAG_CURRENTDAY
:
1080 case MAG_CURRENTDAYNAME
:
1081 return $wgLang->getWeekdayName( date("w")+
1 );
1082 case MAG_CURRENTYEAR
:
1084 case MAG_CURRENTTIME
:
1085 return $wgLang->time( wfTimestampNow(), false );
1086 case MAG_NUMBEROFARTICLES
:
1087 return wfNumberOfArticles();
1093 function initialiseVariables()
1095 global $wgVariableIDs;
1096 $this->mVariables
= array();
1097 foreach ( $wgVariableIDs as $id ) {
1098 $mw =& MagicWord
::get( $id );
1099 $mw->addToArray( $this->mVariables
, $this->getVariableValue( $id ) );
1103 /* private */ function replaceVariables( $text )
1105 global $wgLang, $wgCurParser;
1106 global $wgScript, $wgArticlePath;
1108 $fname = "Parser::replaceVariables";
1109 wfProfileIn( $fname );
1112 if ( !$this->mVariables
) {
1113 $this->initialiseVariables();
1115 $titleChars = Title
::legalChars();
1116 $regex = "/{{([$titleChars]*?)}}/s";
1118 # "Recursive" variable expansion: run it through a couple of passes
1119 for ( $i=0; $i<MAX_INCLUDE_REPEAT
&& !$bail; $i++
) {
1122 # It's impossible to rebind a global in PHP
1123 # Instead, we run the substitution on a copy, then merge the changed fields back in
1124 $wgCurParser = $this->fork();
1126 $text = preg_replace_callback( $regex, "wfBraceSubstitution", $text );
1127 if ( $oldText == $text ) {
1130 $this->merge( $wgCurParser );
1136 # Returns a copy of this object except with various variables cleared
1137 # This copy can be re-merged with the parent after operations on the copy
1141 $copy->mOutput
= new ParserOutput
;
1145 # Merges a copy split off with fork()
1146 function merge( &$copy )
1148 $this->mOutput
->merge( $copy->mOutput
);
1150 # Merge include throttling arrays
1151 foreach( $copy->mIncludeCount
as $dbk => $count ) {
1152 if ( array_key_exists( $dbk, $this->mIncludeCount
) ) {
1153 $this->mIncludeCount
[$dbk] +
= $count;
1155 $this->mIncludeCount
[$dbk] = $count;
1160 function braceSubstitution( $matches )
1162 global $wgLinkCache;
1163 $fname = "Parser::braceSubstitution";
1167 $text = $matches[1];
1170 $mwSubst =& MagicWord
::get( MAG_SUBST
);
1171 if ( $mwSubst->matchStartAndRemove( $text ) ) {
1172 if ( $this->mOutputType
== OT_HTML
) {
1173 # Invalid SUBST not replaced at PST time
1174 # Return without further processing
1175 $text = $matches[0];
1178 } elseif ( $this->mOutputType
== OT_WIKI
) {
1179 # SUBST not found in PST pass, do nothing
1180 $text = $matches[0];
1187 $mwMsgnw =& MagicWord
::get( MAG_MSGNW
);
1188 if ( $mwMsgnw->matchStartAndRemove( $text ) ) {
1191 # Remove obsolete MSG:
1192 $mwMsg =& MagicWord
::get( MAG_MSG
);
1193 $mwMsg->matchStartAndRemove( $text );
1196 # Check if it is an internal message
1197 $mwInt =& MagicWord
::get( MAG_INT
);
1198 if ( $mwInt->matchStartAndRemove( $text ) ) {
1199 $text = wfMsg( $text );
1204 # Check for a match against internal variables
1205 if ( !$found && array_key_exists( $text, $this->mVariables
) ) {
1206 $text = $this->mVariables
[$text];
1208 $this->mOutput
->mContainsOldMagic
= true;
1211 # Load from database
1213 $title = Title
::newFromText( $text, NS_TEMPLATE
);
1214 if ( !is_null( $text ) && !$title->isExternal() ) {
1215 # Check for excessive inclusion
1216 $dbk = $title->getPrefixedDBkey();
1217 if ( !array_key_exists( $dbk, $this->mIncludeCount
) ) {
1218 $this->mIncludeCount
[$dbk] = 0;
1220 if ( ++
$this->mIncludeCount
[$dbk] <= MAX_INCLUDE_REPEAT
) {
1221 $article = new Article( $title );
1222 $articleContent = $article->getContentWithoutUsingSoManyDamnGlobals();
1223 if ( $articleContent !== false ) {
1225 $text = $articleContent;
1227 # Escaping and link table handling
1228 # Not required for preSaveTransform()
1229 if ( $this->mOutputType
== OT_HTML
) {
1231 $text = wfEscapeWikiText( $text );
1233 $text = $this->removeHTMLtags( $text );
1235 $wgLinkCache->suspend();
1236 $text = $this->doTokenizedParser( $text );
1237 $wgLinkCache->resume();
1238 $wgLinkCache->addLinkObj( $title );
1244 # If the title is valid but undisplayable, make a link to it
1245 if ( $this->mOutputType
== OT_HTML
&& !$found ) {
1246 $text = "[[" . $title->getPrefixedText() . "]]";
1259 # Cleans up HTML, removes dangerous tags and attributes
1260 /* private */ function removeHTMLtags( $text )
1262 $fname = "Parser::removeHTMLtags";
1263 wfProfileIn( $fname );
1264 $htmlpairs = array( # Tags that must be closed
1265 "b", "i", "u", "font", "big", "small", "sub", "sup", "h1",
1266 "h2", "h3", "h4", "h5", "h6", "cite", "code", "em", "s",
1267 "strike", "strong", "tt", "var", "div", "center",
1268 "blockquote", "ol", "ul", "dl", "table", "caption", "pre",
1269 "ruby", "rt" , "rb" , "rp", "p"
1271 $htmlsingle = array(
1272 "br", "hr", "li", "dt", "dd", "hr/"
1274 $htmlnest = array( # Tags that can be nested--??
1275 "table", "tr", "td", "th", "div", "blockquote", "ol", "ul",
1276 "dl", "font", "big", "small", "sub", "sup"
1278 $tabletags = array( # Can only appear inside table
1282 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1283 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1285 $htmlattrs = $this->getHTMLattrs () ;
1287 # Remove HTML comments
1288 $text = preg_replace( "/<!--.*-->/sU", "", $text );
1290 $bits = explode( "<", $text );
1291 $text = array_shift( $bits );
1292 $tagstack = array(); $tablestack = array();
1294 foreach ( $bits as $x ) {
1295 $prev = error_reporting( E_ALL
& ~
( E_NOTICE | E_WARNING
) );
1296 preg_match( "/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/",
1298 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1299 error_reporting( $prev );
1302 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1306 if ( ! in_array( $t, $htmlsingle ) &&
1307 ( $ot = array_pop( $tagstack ) ) != $t ) {
1308 array_push( $tagstack, $ot );
1311 if ( $t == "table" ) {
1312 $tagstack = array_pop( $tablestack );
1317 # Keep track for later
1318 if ( in_array( $t, $tabletags ) &&
1319 ! in_array( "table", $tagstack ) ) {
1321 } else if ( in_array( $t, $tagstack ) &&
1322 ! in_array ( $t , $htmlnest ) ) {
1324 } else if ( ! in_array( $t, $htmlsingle ) ) {
1325 if ( $t == "table" ) {
1326 array_push( $tablestack, $tagstack );
1327 $tagstack = array();
1329 array_push( $tagstack, $t );
1331 # Strip non-approved attributes from the tag
1332 $newparams = $this->fixTagAttributes($params);
1336 $rest = str_replace( ">", ">", $rest );
1337 $text .= "<$slash$t $newparams$brace$rest";
1341 $text .= "<" . str_replace( ">", ">", $x);
1343 # Close off any remaining tags
1344 while ( $t = array_pop( $tagstack ) ) {
1346 if ( $t == "table" ) { $tagstack = array_pop( $tablestack ); }
1348 wfProfileOut( $fname );
1354 * This function accomplishes several tasks:
1355 * 1) Auto-number headings if that option is enabled
1356 * 2) Add an [edit] link to sections for logged in users who have enabled the option
1357 * 3) Add a Table of contents on the top for users who have enabled the option
1358 * 4) Auto-anchor headings
1360 * It loops through all headlines, collects the necessary data, then splits up the
1361 * string and re-inserts the newly formatted headlines.
1365 /* private */ function formatHeadings( $text )
1367 $doNumberHeadings = $this->mOptions
->getNumberHeadings();
1368 $doShowToc = $this->mOptions
->getShowToc();
1369 if( !$this->mTitle
->userCanEdit() ) {
1371 $rightClickHack = 0;
1373 $showEditLink = $this->mOptions
->getEditSection();
1374 $rightClickHack = $this->mOptions
->getEditSectionOnRightClick();
1377 # Inhibit editsection links if requested in the page
1378 $esw =& MagicWord
::get( MAG_NOEDITSECTION
);
1379 if( $esw->matchAndRemove( $text ) ) {
1382 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1384 $mw =& MagicWord
::get( MAG_NOTOC
);
1385 if( $mw->matchAndRemove( $text ) ) {
1389 # never add the TOC to the Main Page. This is an entry page that should not
1390 # be more than 1-2 screens large anyway
1391 if( $this->mTitle
->getPrefixedText() == wfMsg("mainpage") ) {
1395 # Get all headlines for numbering them and adding funky stuff like [edit]
1396 # links - this is for later, but we need the number of headlines right now
1397 $numMatches = preg_match_all( "/<H([1-6])(.*?" . ">)(.*?)<\/H[1-6]>/i", $text, $matches );
1399 # if there are fewer than 4 headlines in the article, do not show TOC
1400 if( $numMatches < 4 ) {
1404 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
1405 # override above conditions and always show TOC
1406 $mw =& MagicWord
::get( MAG_FORCETOC
);
1407 if ($mw->matchAndRemove( $text ) ) {
1412 # We need this to perform operations on the HTML
1413 $sk =& $this->mOptions
->getSkin();
1418 # Ugh .. the TOC should have neat indentation levels which can be
1419 # passed to the skin functions. These are determined here
1424 $sublevelCount = array();
1427 foreach( $matches[3] as $headline ) {
1430 $prevlevel = $level;
1432 $level = $matches[1][$headlineCount];
1433 if( ( $doNumberHeadings ||
$doShowToc ) && $prevlevel && $level > $prevlevel ) {
1434 # reset when we enter a new level
1435 $sublevelCount[$level] = 0;
1436 $toc .= $sk->tocIndent( $level - $prevlevel );
1437 $toclevel +
= $level - $prevlevel;
1439 if( ( $doNumberHeadings ||
$doShowToc ) && $level < $prevlevel ) {
1440 # reset when we step back a level
1441 $sublevelCount[$level+
1]=0;
1442 $toc .= $sk->tocUnindent( $prevlevel - $level );
1443 $toclevel -= $prevlevel - $level;
1445 # count number of headlines for each level
1446 @$sublevelCount[$level]++
;
1447 if( $doNumberHeadings ||
$doShowToc ) {
1449 for( $i = 1; $i <= $level; $i++
) {
1450 if( !empty( $sublevelCount[$i] ) ) {
1454 $numbering .= $sublevelCount[$i];
1460 # The canonized header is a version of the header text safe to use for links
1461 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
1462 $canonized_headline = Parser
::unstrip( $headline, $this->mStripState
);
1465 $canonized_headline = preg_replace( "/<.*?" . ">/","",$canonized_headline );
1467 $tocline = trim( $canonized_headline );
1468 $canonized_headline = str_replace( '"', "", $canonized_headline );
1469 $canonized_headline = str_replace( " ", "_", trim( $canonized_headline) );
1470 $refer[$headlineCount] = $canonized_headline;
1472 # count how many in assoc. array so we can track dupes in anchors
1473 @$refers[$canonized_headline]++
;
1474 $refcount[$headlineCount]=$refers[$canonized_headline];
1476 # Prepend the number to the heading text
1478 if( $doNumberHeadings ||
$doShowToc ) {
1479 $tocline = $numbering . " " . $tocline;
1481 # Don't number the heading if it is the only one (looks silly)
1482 if( $doNumberHeadings && count( $matches[3] ) > 1) {
1483 # the two are different if the line contains a link
1484 $headline=$numbering . " " . $headline;
1488 # Create the anchor for linking from the TOC to the section
1489 $anchor = $canonized_headline;
1490 if($refcount[$headlineCount] > 1 ) {
1491 $anchor .= "_" . $refcount[$headlineCount];
1494 $toc .= $sk->tocLine($anchor,$tocline,$toclevel);
1496 if( $showEditLink ) {
1497 if ( empty( $head[$headlineCount] ) ) {
1498 $head[$headlineCount] = "";
1500 $head[$headlineCount] .= $sk->editSectionLink($headlineCount+
1);
1504 # the headline might have a link
1505 if( preg_match( "/(.*)<a(.*)/", $headline, $headlinematches ) ) {
1506 # if so give an anchor name to the already existent link
1507 $headline = $headlinematches[1]
1508 . "<a name=\"$anchor\" " . $headlinematches[2];
1510 # else create an anchor link for the headline
1511 $headline = "<a name=\"$anchor\">$headline</a>";
1514 # give headline the correct <h#> tag
1515 @$head[$headlineCount] .= "<h".$level.$matches[2][$headlineCount] .$headline."</h".$level.">";
1517 # Add the edit section link
1518 if( $rightClickHack ) {
1519 $head[$headlineCount] = $sk->editSectionScript($headlineCount+
1,$head[$headlineCount]);
1526 $toclines = $headlineCount;
1527 $toc .= $sk->tocUnindent( $toclevel );
1528 $toc = $sk->tocTable( $toc );
1531 # split up and insert constructed headlines
1533 $blocks = preg_split( "/<H[1-6].*?" . ">.*?<\/H[1-6]>/i", $text );
1536 foreach( $blocks as $block ) {
1537 if( $showEditLink && $headlineCount > 0 && $i == 0 && $block != "\n" ) {
1538 # This is the [edit] link that appears for the top block of text when
1539 # section editing is enabled
1540 $full .= $sk->editSectionLink(0);
1543 if( $doShowToc && !$i) {
1544 # Let's add a top anchor just in case we want to link to the top of the page
1545 $full = "<a name=\"top\"></a>".$full.$toc;
1548 if( !empty( $head[$i] ) ) {
1557 /* private */ function doMagicISBN( &$tokenizer )
1561 # Check whether next token is a text token
1562 # If yes, fetch it and convert the text into a
1563 # Special::BookSources link
1564 $token = $tokenizer->previewToken();
1565 while ( $token["type"] == "" )
1567 $tokenizer->nextToken();
1568 $token = $tokenizer->previewToken();
1570 if ( $token["type"] == "text" )
1572 $token = $tokenizer->nextToken();
1573 $x = $token["text"];
1574 $valid = "0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ";
1576 $isbn = $blank = "" ;
1577 while ( " " == $x{0} ) {
1579 $x = substr( $x, 1 );
1581 while ( strstr( $valid, $x{0} ) != false ) {
1583 $x = substr( $x, 1 );
1585 $num = str_replace( "-", "", $isbn );
1586 $num = str_replace( " ", "", $num );
1589 $text = "ISBN $blank$x";
1591 $titleObj = Title
::makeTitle( NS_SPECIAL
, "Booksources" );
1592 $text = "<a href=\"" .
1593 $titleObj->escapeLocalUrl( "isbn={$num}" ) .
1594 "\" class=\"internal\">ISBN $isbn</a>";
1602 /* private */ function doMagicRFC( &$tokenizer )
1606 # Check whether next token is a text token
1607 # If yes, fetch it and convert the text into a
1608 # link to an RFC source
1609 $token = $tokenizer->previewToken();
1610 while ( $token["type"] == "" )
1612 $tokenizer->nextToken();
1613 $token = $tokenizer->previewToken();
1615 if ( $token["type"] == "text" )
1617 $token = $tokenizer->nextToken();
1618 $x = $token["text"];
1619 $valid = "0123456789";
1621 $rfc = $blank = "" ;
1622 while ( " " == $x{0} ) {
1624 $x = substr( $x, 1 );
1626 while ( strstr( $valid, $x{0} ) != false ) {
1628 $x = substr( $x, 1 );
1632 $text .= "RFC $blank$x";
1634 $url = wfmsg( "rfcurl" );
1635 $url = str_replace( "$1", $rfc, $url);
1636 $sk =& $this->mOptions
->getSkin();
1637 $la = $sk->getExternalLinkAttributes( $url, "RFC {$rfc}" );
1638 $text = "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
1646 function preSaveTransform( $text, &$title, &$user, $options, $clearState = true )
1648 $this->mOptions
= $options;
1649 $this->mTitle
=& $title;
1650 $this->mOutputType
= OT_WIKI
;
1652 if ( $clearState ) {
1653 $this->clearState();
1656 $stripState = false;
1657 $text = str_replace("\r\n", "\n", $text);
1658 $text = $this->strip( $text, $stripState, false );
1659 $text = $this->pstPass2( $text, $user );
1660 $text = $this->unstrip( $text, $stripState );
1664 /* private */ function pstPass2( $text, &$user )
1666 global $wgLang, $wgLocaltimezone, $wgCurParser;
1668 # Variable replacement
1669 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
1670 $text = $this->replaceVariables( $text );
1674 $n = $user->getName();
1675 $k = $user->getOption( "nickname" );
1676 if ( "" == $k ) { $k = $n; }
1677 if(isset($wgLocaltimezone)) {
1678 $oldtz = getenv("TZ"); putenv("TZ=$wgLocaltimezone");
1680 /* Note: this is an ugly timezone hack for the European wikis */
1681 $d = $wgLang->timeanddate( date( "YmdHis" ), false ) .
1682 " (" . date( "T" ) . ")";
1683 if(isset($wgLocaltimezone)) putenv("TZ=$oldtz");
1685 $text = preg_replace( "/~~~~~/", $d, $text );
1686 $text = preg_replace( "/~~~~/", "[[" . $wgLang->getNsText(
1687 Namespace::getUser() ) . ":$n|$k]] $d", $text );
1688 $text = preg_replace( "/~~~/", "[[" . $wgLang->getNsText(
1689 Namespace::getUser() ) . ":$n|$k]]", $text );
1691 # Context links: [[|name]] and [[name (context)|]]
1693 $tc = "[&;%\\-,.\\(\\)' _0-9A-Za-z\\/:\\x80-\\xff]";
1694 $np = "[&;%\\-,.' _0-9A-Za-z\\/:\\x80-\\xff]"; # No parens
1695 $namespacechar = '[ _0-9A-Za-z\x80-\xff]'; # Namespaces can use non-ascii!
1696 $conpat = "/^({$np}+) \\(({$tc}+)\\)$/";
1698 $p1 = "/\[\[({$np}+) \\(({$np}+)\\)\\|]]/"; # [[page (context)|]]
1699 $p2 = "/\[\[\\|({$tc}+)]]/"; # [[|page]]
1700 $p3 = "/\[\[($namespacechar+):({$np}+)\\|]]/"; # [[namespace:page|]]
1701 $p4 = "/\[\[($namespacechar+):({$np}+) \\(({$np}+)\\)\\|]]/";
1702 # [[ns:page (cont)|]]
1704 $t = $this->mTitle
->getText();
1705 if ( preg_match( $conpat, $t, $m ) ) {
1708 $text = preg_replace( $p4, "[[\\1:\\2 (\\3)|\\2]]", $text );
1709 $text = preg_replace( $p1, "[[\\1 (\\2)|\\1]]", $text );
1710 $text = preg_replace( $p3, "[[\\1:\\2|\\2]]", $text );
1712 if ( "" == $context ) {
1713 $text = preg_replace( $p2, "[[\\1]]", $text );
1715 $text = preg_replace( $p2, "[[\\1 ({$context})|\\1]]", $text );
1719 $mw =& MagicWord::get( MAG_SUBST );
1720 $wgCurParser = $this->fork();
1721 $text = $mw->substituteCallback( $text, "wfBraceSubstitution" );
1722 $this->merge( $wgCurParser );
1725 # Trim trailing whitespace
1726 # MAG_END (__END__) tag allows for trailing
1727 # whitespace to be deliberately included
1728 $text = rtrim( $text );
1729 $mw =& MagicWord
::get( MAG_END
);
1730 $mw->matchAndRemove( $text );
1735 # Set up some variables which are usually set up in parse()
1736 # so that an external function can call some class members with confidence
1737 function startExternalParse( &$title, $options, $outputType, $clearState = true )
1739 $this->mTitle
=& $title;
1740 $this->mOptions
= $options;
1741 $this->mOutputType
= $outputType;
1742 if ( $clearState ) {
1743 $this->clearState();
1750 var $mText, $mLanguageLinks, $mCategoryLinks, $mContainsOldMagic;
1752 function ParserOutput( $text = "", $languageLinks = array(), $categoryLinks = array(),
1753 $containsOldMagic = false )
1755 $this->mText
= $text;
1756 $this->mLanguageLinks
= $languageLinks;
1757 $this->mCategoryLinks
= $categoryLinks;
1758 $this->mContainsOldMagic
= $containsOldMagic;
1761 function getText() { return $this->mText
; }
1762 function getLanguageLinks() { return $this->mLanguageLinks
; }
1763 function getCategoryLinks() { return $this->mCategoryLinks
; }
1764 function containsOldMagic() { return $this->mContainsOldMagic
; }
1765 function setText( $text ) { return wfSetVar( $this->mText
, $text ); }
1766 function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks
, $ll ); }
1767 function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks
, $cl ); }
1768 function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic
, $com ); }
1770 function merge( $other ) {
1771 $this->mLanguageLinks
= array_merge( $this->mLanguageLinks
, $other->mLanguageLinks
);
1772 $this->mCategoryLinks
= array_merge( $this->mCategoryLinks
, $this->mLanguageLinks
);
1773 $this->mContainsOldMagic
= $this->mContainsOldMagic ||
$other->mContainsOldMagic
;
1780 # All variables are private
1781 var $mUseTeX; # Use texvc to expand <math> tags
1782 var $mUseCategoryMagic; # Treat [[Category:xxxx]] tags specially
1783 var $mUseDynamicDates; # Use $wgDateFormatter to format dates
1784 var $mInterwikiMagic; # Interlanguage links are removed and returned in an array
1785 var $mAllowExternalImages; # Allow external images inline
1786 var $mSkin; # Reference to the preferred skin
1787 var $mDateFormat; # Date format index
1788 var $mEditSection; # Create "edit section" links
1789 var $mEditSectionOnRightClick; # Generate JavaScript to edit section on right click
1790 var $mNumberHeadings; # Automatically number headings
1791 var $mShowToc; # Show table of contents
1793 function getUseTeX() { return $this->mUseTeX
; }
1794 function getUseCategoryMagic() { return $this->mUseCategoryMagic
; }
1795 function getUseDynamicDates() { return $this->mUseDynamicDates
; }
1796 function getInterwikiMagic() { return $this->mInterwikiMagic
; }
1797 function getAllowExternalImages() { return $this->mAllowExternalImages
; }
1798 function getSkin() { return $this->mSkin
; }
1799 function getDateFormat() { return $this->mDateFormat
; }
1800 function getEditSection() { return $this->mEditSection
; }
1801 function getEditSectionOnRightClick() { return $this->mEditSectionOnRightClick
; }
1802 function getNumberHeadings() { return $this->mNumberHeadings
; }
1803 function getShowToc() { return $this->mShowToc
; }
1805 function setUseTeX( $x ) { return wfSetVar( $this->mUseTeX
, $x ); }
1806 function setUseCategoryMagic( $x ) { return wfSetVar( $this->mUseCategoryMagic
, $x ); }
1807 function setUseDynamicDates( $x ) { return wfSetVar( $this->mUseDynamicDates
, $x ); }
1808 function setInterwikiMagic( $x ) { return wfSetVar( $this->mInterwikiMagic
, $x ); }
1809 function setAllowExternalImages( $x ) { return wfSetVar( $this->mAllowExternalImages
, $x ); }
1810 function setSkin( $x ) { return wfSetRef( $this->mSkin
, $x ); }
1811 function setDateFormat( $x ) { return wfSetVar( $this->mDateFormat
, $x ); }
1812 function setEditSection( $x ) { return wfSetVar( $this->mEditSection
, $x ); }
1813 function setEditSectionOnRightClick( $x ) { return wfSetVar( $this->mEditSectionOnRightClick
, $x ); }
1814 function setNumberHeadings( $x ) { return wfSetVar( $this->mNumberHeadings
, $x ); }
1815 function setShowToc( $x ) { return wfSetVar( $this->mShowToc
, $x ); }
1817 /* static */ function newFromUser( &$user )
1819 $popts = new ParserOptions
;
1820 $popts->initialiseFromUser( &$user );
1824 function initialiseFromUser( &$userInput )
1826 global $wgUseTeX, $wgUseCategoryMagic, $wgUseDynamicDates, $wgInterwikiMagic, $wgAllowExternalImages;
1828 if ( !$userInput ) {
1831 $user =& $userInput;
1834 $this->mUseTeX
= $wgUseTeX;
1835 $this->mUseCategoryMagic
= $wgUseCategoryMagic;
1836 $this->mUseDynamicDates
= $wgUseDynamicDates;
1837 $this->mInterwikiMagic
= $wgInterwikiMagic;
1838 $this->mAllowExternalImages
= $wgAllowExternalImages;
1839 $this->mSkin
=& $user->getSkin();
1840 $this->mDateFormat
= $user->getOption( "date" );
1841 $this->mEditSection
= $user->getOption( "editsection" );
1842 $this->mEditSectionOnRightClick
= $user->getOption( "editsectiononrightclick" );
1843 $this->mNumberHeadings
= $user->getOption( "numberheadings" );
1844 $this->mShowToc
= $user->getOption( "showtoc" );
1850 # Regex callbacks, used in Parser::replaceVariables
1851 function wfBraceSubstitution( $matches )
1853 global $wgCurParser;
1854 return $wgCurParser->braceSubstitution( $matches );