Remove ancient, unused testsuite
[mediawiki.git] / includes / OutputPage.php
blob9c798787ded0a572f164d62e153ee39f69d15e10
1 <?php
2 # See design.doc
4 if($wgUseTeX) include_once( "Math.php" );
6 class OutputPage {
7 var $mHeaders, $mCookies, $mMetatags, $mKeywords;
8 var $mLinktags, $mPagetitle, $mBodytext, $mDebugtext;
9 var $mHTMLtitle, $mRobotpolicy, $mIsarticle, $mPrintable;
10 var $mSubtitle, $mRedirect, $mAutonumber, $mHeadtext;
11 var $mLastModified, $mCategoryLinks;
13 var $mDTopen, $mLastSection; # Used for processing DL, PRE
14 var $mLanguageLinks, $mSupressQuickbar;
15 var $mOnloadHandler;
16 var $mDoNothing;
17 var $mContainsOldMagic, $mContainsNewMagic;
18 var $mIsArticleRelated;
20 function OutputPage()
22 $this->mHeaders = $this->mCookies = $this->mMetatags =
23 $this->mKeywords = $this->mLinktags = array();
24 $this->mHTMLtitle = $this->mPagetitle = $this->mBodytext =
25 $this->mLastSection = $this->mRedirect = $this->mLastModified =
26 $this->mSubtitle = $this->mDebugtext = $this->mRobotpolicy =
27 $this->mOnloadHandler = "";
28 $this->mIsArticleRelated = $this->mIsarticle = $this->mPrintable = true;
29 $this->mSupressQuickbar = $this->mDTopen = $this->mPrintable = false;
30 $this->mLanguageLinks = array();
31 $this->mCategoryLinks = array() ;
32 $this->mAutonumber = 0;
33 $this->mDoNothing = false;
34 $this->mContainsOldMagic = $this->mContainsNewMagic = 0;
37 function addHeader( $name, $val ) { array_push( $this->mHeaders, "$name: $val" ) ; }
38 function addCookie( $name, $val ) { array_push( $this->mCookies, array( $name, $val ) ); }
39 function redirect( $url ) { $this->mRedirect = $url; }
41 # To add an http-equiv meta tag, precede the name with "http:"
42 function addMeta( $name, $val ) { array_push( $this->mMetatags, array( $name, $val ) ); }
43 function addKeyword( $text ) { array_push( $this->mKeywords, $text ); }
44 function addLink( $rel, $rev, $target ) { array_push( $this->mLinktags, array( $rel, $rev, $target ) ); }
46 # checkLastModified tells the client to use the client-cached page if
47 # possible. If sucessful, the OutputPage is disabled so that
48 # any future call to OutputPage->output() have no effect. The method
49 # returns true iff cache-ok headers was sent.
50 function checkLastModified ( $timestamp )
52 global $wgLang, $wgCachePages, $wgUser;
53 if( !$wgCachePages ) {
54 wfDebug( "CACHE DISABLED\n", false );
55 return;
57 if( preg_match( '/MSIE ([1-4]|5\.0)/', $_SERVER["HTTP_USER_AGENT"] ) ) {
58 # IE 5.0 has probs with our caching
59 wfDebug( "-- bad client, not caching\n", false );
60 return;
62 if( $wgUser->getOption( "nocache" ) ) {
63 wfDebug( "USER DISABLED CACHE\n", false );
64 return;
67 $lastmod = gmdate( "D, j M Y H:i:s", wfTimestamp2Unix(
68 max( $timestamp, $wgUser->mTouched ) ) ) . " GMT";
70 if( !empty( $_SERVER["HTTP_IF_MODIFIED_SINCE"] ) ) {
71 # IE sends sizes after the date like this:
72 # Wed, 20 Aug 2003 06:51:19 GMT; length=5202
73 # this breaks strtotime().
74 $modsince = preg_replace( '/;.*$/', '', $_SERVER["HTTP_IF_MODIFIED_SINCE"] );
75 $ismodsince = wfUnix2Timestamp( strtotime( $modsince ) );
76 wfDebug( "-- client send If-Modified-Since: " . $modsince . "\n", false );
77 wfDebug( "-- we might send Last-Modified : $lastmod\n", false );
79 if( ($ismodsince >= $timestamp ) and $wgUser->validateCache( $ismodsince ) ) {
80 # Make sure you're in a place you can leave when you call us!
81 header( "HTTP/1.0 304 Not Modified" );
82 $this->mLastModified = $lastmod;
83 $this->sendCacheControl();
84 wfDebug( "CACHED client: $ismodsince ; user: $wgUser->mTouched ; page: $timestamp\n", false );
85 $this->disable();
86 return true;
87 } else {
88 wfDebug( "READY client: $ismodsince ; user: $wgUser->mTouched ; page: $timestamp\n", false );
89 $this->mLastModified = $lastmod;
91 } else {
92 wfDebug( "We're confused.\n", false );
93 $this->mLastModified = $lastmod;
97 function setRobotpolicy( $str ) { $this->mRobotpolicy = $str; }
98 function setHTMLtitle( $name ) { $this->mHTMLtitle = $name; }
99 function setPageTitle( $name ) { $this->mPagetitle = $name; }
100 function getPageTitle() { return $this->mPagetitle; }
101 function setSubtitle( $str ) { $this->mSubtitle = $str; }
102 function getSubtitle() { return $this->mSubtitle; }
103 function isArticle() { return $this->mIsarticle; }
104 function setPrintable() { $this->mPrintable = true; }
105 function isPrintable() { return $this->mPrintable; }
106 function setOnloadHandler( $js ) { $this->mOnloadHandler = $js; }
107 function getOnloadHandler() { return $this->mOnloadHandler; }
108 function disable() { $this->mDoNothing = true; }
110 function setArticleRelated( $v )
112 $this->mIsArticleRelated = $v;
113 if ( !$v ) {
114 $this->mIsarticle = false;
117 function setArticleFlag( $v ) {
118 $this->mIsarticle = $v;
119 if ( $v ) {
120 $this->mIsArticleRelated = $v;
124 function isArticleRelated()
126 return $this->mIsArticleRelated;
129 function getLanguageLinks() {
130 global $wgTitle, $wgLanguageCode;
131 global $wgDBconnection, $wgDBname;
132 return $this->mLanguageLinks;
134 function supressQuickbar() { $this->mSupressQuickbar = true; }
135 function isQuickbarSupressed() { return $this->mSupressQuickbar; }
137 function addHTML( $text ) { $this->mBodytext .= $text; }
138 function addHeadtext( $text ) { $this->mHeadtext .= $text; }
139 function debug( $text ) { $this->mDebugtext .= $text; }
141 # First pass--just handle <nowiki> sections, pass the rest off
142 # to doWikiPass2() which does all the real work.
144 function addWikiText( $text, $linestart = true )
146 global $wgUseTeX, $wgArticle, $wgUser, $action;
147 $fname = "OutputPage::addWikiText";
148 wfProfileIn( $fname );
149 $unique = "3iyZiyA7iMwg5rhxP0Dcc9oTnj8qD1jm1Sfv4";
150 $unique2 = "4LIQ9nXtiYFPCSfitVwDw7EYwQlL4GeeQ7qSO";
151 $unique3 = "fPaA8gDfdLBqzj68Yjg9Hil3qEF8JGO0uszIp";
152 $nwlist = array();
153 $nwsecs = 0;
154 $mathlist = array();
155 $mathsecs = 0;
156 $prelist = array ();
157 $presecs = 0;
158 $stripped = "";
159 $stripped2 = "";
160 $stripped3 = "";
162 # Replace any instances of the placeholders
163 $text = str_replace( $unique, wfHtmlEscapeFirst( $unique ), $text );
164 $text = str_replace( $unique2, wfHtmlEscapeFirst( $unique2 ), $text );
165 $text = str_replace( $unique3, wfHtmlEscapeFirst( $unique3 ), $text );
167 global $wgEnableParserCache;
168 $use_parser_cache =
169 $wgEnableParserCache && $action == "view" &&
170 intval($wgUser->getOption( "stubthreshold" )) == 0 &&
171 isset($wgArticle) && $wgArticle->getID() > 0;
173 if( $use_parser_cache ){
174 if( $this->fillFromParserCache() ){
175 wfProfileOut( $fname );
176 return;
180 while ( "" != $text ) {
181 $p = preg_split( "/<\\s*nowiki\\s*>/i", $text, 2 );
182 $stripped .= $p[0];
183 if ( ( count( $p ) < 2 ) || ( "" == $p[1] ) ) { $text = ""; }
184 else {
185 $q = preg_split( "/<\\/\\s*nowiki\\s*>/i", $p[1], 2 );
186 ++$nwsecs;
187 $nwlist[$nwsecs] = wfEscapeHTMLTagsOnly($q[0]);
188 $stripped .= $unique . $nwsecs . "s";
189 $text = $q[1];
193 if( $wgUseTeX ) {
194 while ( "" != $stripped ) {
195 $p = preg_split( "/<\\s*math\\s*>/i", $stripped, 2 );
196 $stripped2 .= $p[0];
197 if ( ( count( $p ) < 2 ) || ( "" == $p[1] ) ) { $stripped = ""; }
198 else {
199 $q = preg_split( "/<\\/\\s*math\\s*>/i", $p[1], 2 );
200 ++$mathsecs;
201 $mathlist[$mathsecs] = renderMath($q[0]);
202 $stripped2 .= $unique2 . $mathsecs . "s";
203 $stripped = $q[1];
206 } else {
207 $stripped2 = $stripped;
210 while ( "" != $stripped2 ) {
211 $p = preg_split( "/<\\s*pre\\s*>/i", $stripped2, 2 );
212 $stripped3 .= $p[0];
213 if ( ( count( $p ) < 2 ) || ( "" == $p[1] ) ) { $stripped2 = ""; }
214 else {
215 $q = preg_split( "/<\\/\\s*pre\\s*>/i", $p[1], 2 );
216 ++$presecs;
217 $prelist[$presecs] = "<pre>". wfEscapeHTMLTagsOnly($q[0]). "</pre>\n";
218 $stripped3 .= $unique3 . $presecs . "s";
219 $stripped2 = $q[1];
223 $text = $this->doWikiPass2( $stripped3, $linestart );
225 $specialChars = array("\\", "$");
226 $escapedChars = array("\\\\", "\\$");
227 for ( $i = 1; $i <= $presecs; ++$i ) {
228 $text = preg_replace( "/{$unique3}{$i}s/", str_replace( $specialChars,
229 $escapedChars, $prelist[$i] ), $text );
232 for ( $i = 1; $i <= $mathsecs; ++$i ) {
233 $text = preg_replace( "/{$unique2}{$i}s/", str_replace( $specialChars,
234 $escapedChars, $mathlist[$i] ), $text );
237 for ( $i = 1; $i <= $nwsecs; ++$i ) {
238 $text = preg_replace( "/{$unique}{$i}s/", str_replace( $specialChars,
239 $escapedChars, $nwlist[$i] ), $text );
241 $this->addHTML( $text );
243 if($use_parser_cache ){
244 $this->saveParserCache( $text );
246 wfProfileOut( $fname );
249 # Set the maximum cache time on the Squid in seconds
250 function setSquidMaxage( $maxage ) {
251 global $wgSquidMaxage;
252 $wgSquidMaxage = $maxage;
255 function sendCacheControl() {
256 global $wgUseSquid, $wgUseESI, $wgSquidMaxage;
257 # FIXME: This header may cause trouble with some versions of Internet Explorer
258 header( "Vary: Accept-Encoding, Cookie" );
259 if( $this->mLastModified != "" ) {
260 if( $wgUseSquid && ! isset( $_COOKIE[ini_get( "session.name") ] ) &&
261 ! $this->isPrintable() )
263 if ( $wgUseESI ) {
264 # We'll purge the proxy cache explicitly, but require end user agents
265 # to revalidate against the proxy on each visit.
266 # Surrogate-Control controls our Squid, Cache-Control downstream caches
267 wfDebug( "** proxy caching with ESI; {$this->mLastModified} **\n", false );
268 # start with a shorter timeout for initial testing
269 # header( 'Surrogate-Control: max-age=2678400+2678400, content="ESI/1.0"');
270 header( 'Surrogate-Control: max-age='.$wgSquidMaxage.'+'.$wgSquidMaxage.', content="ESI/1.0"');
271 header( 'Cache-Control: s-maxage=0, must-revalidate, max-age=0' );
272 } else {
273 # We'll purge the proxy cache for anons explicitly, but require end user agents
274 # to revalidate against the proxy on each visit.
275 # IMPORTANT! The Squid needs to replace the Cache-Control header with
276 # Cache-Control: s-maxage=0, must-revalidate, max-age=0
277 wfDebug( "** local proxy caching; {$this->mLastModified} **\n", false );
278 # start with a shorter timeout for initial testing
279 # header( "Cache-Control: s-maxage=2678400, must-revalidate, max-age=0" );
280 header( 'Cache-Control: s-maxage='.$wgSquidMaxage.', must-revalidate, max-age=0' );
282 } else {
283 # We do want clients to cache if they can, but they *must* check for updates
284 # on revisiting the page.
285 wfDebug( "** private caching; {$this->mLastModified} **\n", false );
286 header( "Expires: -1" );
287 header( "Cache-Control: private, must-revalidate, max-age=0" );
289 header( "Last-modified: {$this->mLastModified}" );
290 } else {
291 wfDebug( "** no caching **\n", false );
292 header( "Expires: -1" );
293 header( "Cache-Control: no-cache" );
294 header( "Pragma: no-cache" );
295 header( "Last-modified: " . gmdate( "D, j M Y H:i:s" ) . " GMT" );
299 # Finally, all the text has been munged and accumulated into
300 # the object, let's actually output it:
302 function output()
304 global $wgUser, $wgLang, $wgDebugComments, $wgCookieExpiration;
305 global $wgInputEncoding, $wgOutputEncoding, $wgLanguageCode;
306 if( $this->mDoNothing ){
307 return;
309 $fname = "OutputPage::output";
310 wfProfileIn( $fname );
312 $sk = $wgUser->getSkin();
314 $this->sendCacheControl();
316 header( "Content-type: text/html; charset={$wgOutputEncoding}" );
317 header( "Content-language: {$wgLanguageCode}" );
319 if ( "" != $this->mRedirect ) {
320 if( substr( $this->mRedirect, 0, 4 ) != "http" ) {
321 # Standards require redirect URLs to be absolute
322 global $wgServer;
323 $this->mRedirect = $wgServer . $this->mRedirect;
325 header( "Location: {$this->mRedirect}" );
326 return;
329 $exp = time() + $wgCookieExpiration;
330 foreach( $this->mCookies as $name => $val ) {
331 setcookie( $name, $val, $exp, "/" );
334 $sk->outputPage( $this );
335 # flush();
338 function out( $ins )
340 global $wgInputEncoding, $wgOutputEncoding, $wgLang;
341 if ( 0 == strcmp( $wgInputEncoding, $wgOutputEncoding ) ) {
342 $outs = $ins;
343 } else {
344 $outs = $wgLang->iconv( $wgInputEncoding, $wgOutputEncoding, $ins );
345 if ( false === $outs ) { $outs = $ins; }
347 print $outs;
350 function setEncodings()
352 global $wgInputEncoding, $wgOutputEncoding;
353 global $wgUser, $wgLang;
355 $wgInputEncoding = strtolower( $wgInputEncoding );
357 if( $wgUser->getOption( 'altencoding' ) ) {
358 $wgLang->setAltEncoding();
359 return;
362 if ( empty( $_SERVER['HTTP_ACCEPT_CHARSET'] ) ) {
363 $wgOutputEncoding = strtolower( $wgOutputEncoding );
364 return;
368 # This code is unused anyway!
369 # Commenting out. --bv 2003-11-15
371 $a = explode( ",", $_SERVER['HTTP_ACCEPT_CHARSET'] );
372 $best = 0.0;
373 $bestset = "*";
375 foreach ( $a as $s ) {
376 if ( preg_match( "/(.*);q=(.*)/", $s, $m ) ) {
377 $set = $m[1];
378 $q = (float)($m[2]);
379 } else {
380 $set = $s;
381 $q = 1.0;
383 if ( $q > $best ) {
384 $bestset = $set;
385 $best = $q;
388 #if ( "*" == $bestset ) { $bestset = "iso-8859-1"; }
389 if ( "*" == $bestset ) { $bestset = $wgOutputEncoding; }
390 $wgOutputEncoding = strtolower( $bestset );
392 # Disable for now
395 $wgOutputEncoding = $wgInputEncoding;
398 # Returns a HTML comment with the elapsed time since request.
399 # This method has no side effects.
400 function reportTime()
402 global $wgRequestTime;
404 $now = wfTime();
405 list( $usec, $sec ) = explode( " ", $wgRequestTime );
406 $start = (float)$sec + (float)$usec;
407 $elapsed = $now - $start;
408 $com = sprintf( "<!-- Time since request: %01.2f secs. -->",
409 $elapsed );
410 return $com;
413 # Note: these arguments are keys into wfMsg(), not text!
415 function errorpage( $title, $msg )
417 global $wgTitle;
419 $this->mDebugtext .= "Original title: " .
420 $wgTitle->getPrefixedText() . "\n";
421 $this->setHTMLTitle( wfMsg( "errorpagetitle" ) );
422 $this->setPageTitle( wfMsg( $title ) );
423 $this->setRobotpolicy( "noindex,nofollow" );
424 $this->setArticleRelated( false );
426 $this->mBodytext = "";
427 $this->addHTML( "<p>" . wfMsg( $msg ) . "\n" );
428 $this->returnToMain( false );
430 $this->output();
431 wfAbruptExit();
434 function sysopRequired()
436 global $wgUser;
438 $this->setHTMLTitle( wfMsg( "errorpagetitle" ) );
439 $this->setPageTitle( wfMsg( "sysoptitle" ) );
440 $this->setRobotpolicy( "noindex,nofollow" );
441 $this->setArticleRelated( false );
442 $this->mBodytext = "";
444 $sk = $wgUser->getSkin();
445 $ap = $sk->makeKnownLink( wfMsg( "administrators" ), "" );
446 $this->addHTML( wfMsg( "sysoptext", $ap ) );
447 $this->returnToMain();
450 function developerRequired()
452 global $wgUser;
454 $this->setHTMLTitle( wfMsg( "errorpagetitle" ) );
455 $this->setPageTitle( wfMsg( "developertitle" ) );
456 $this->setRobotpolicy( "noindex,nofollow" );
457 $this->setArticleRelated( false );
458 $this->mBodytext = "";
460 $sk = $wgUser->getSkin();
461 $ap = $sk->makeKnownLink( wfMsg( "administrators" ), "" );
462 $this->addHTML( wfMsg( "developertext", $ap ) );
463 $this->returnToMain();
466 function databaseError( $fname, &$conn )
468 global $wgUser, $wgCommandLineMode;
470 $this->setPageTitle( wfMsgNoDB( "databaseerror" ) );
471 $this->setRobotpolicy( "noindex,nofollow" );
472 $this->setArticleRelated( false );
474 if ( $wgCommandLineMode ) {
475 $msg = wfMsgNoDB( "dberrortextcl" );
476 } else {
477 $msg = wfMsgNoDB( "dberrortext" );
480 $msg = str_replace( "$1", htmlspecialchars( $conn->lastQuery() ), $msg );
481 $msg = str_replace( "$2", htmlspecialchars( $fname ), $msg );
482 $msg = str_replace( "$3", $conn->lastErrno(), $msg );
483 $msg = str_replace( "$4", htmlspecialchars( $conn->lastError() ), $msg );
485 if ( $wgCommandLineMode || !is_object( $wgUser )) {
486 print "$msg\n";
487 wfAbruptExit();
489 $sk = $wgUser->getSkin();
490 $shlink = $sk->makeKnownLink( wfMsgNoDB( "searchhelppage" ),
491 wfMsgNoDB( "searchingwikipedia" ) );
492 $msg = str_replace( "$5", $shlink, $msg );
493 $this->mBodytext = $msg;
494 $this->output();
495 wfAbruptExit();
498 function readOnlyPage( $source = "", $protected = false )
500 global $wgUser, $wgReadOnlyFile;
502 $this->setRobotpolicy( "noindex,nofollow" );
503 $this->setArticleRelated( false );
505 if( $protected ) {
506 $this->setPageTitle( wfMsg( "viewsource" ) );
507 $this->addWikiText( wfMsg( "protectedtext" ) );
508 } else {
509 $this->setPageTitle( wfMsg( "readonly" ) );
510 $reason = file_get_contents( $wgReadOnlyFile );
511 $this->addHTML( wfMsg( "readonlytext", $reason ) );
514 if($source) {
515 $rows = $wgUser->getOption( "rows" );
516 $cols = $wgUser->getOption( "cols" );
517 $text .= "</p>\n<textarea cols='$cols' rows='$rows' readonly>" .
518 htmlspecialchars( $source ) . "\n</textarea>";
519 $this->addHTML( $text );
522 $this->returnToMain( false );
525 function fatalError( $message )
527 $this->setPageTitle( wfMsg( "internalerror" ) );
528 $this->setRobotpolicy( "noindex,nofollow" );
529 $this->setArticleRelated( false );
531 $this->mBodytext = $message;
532 $this->output();
533 wfAbruptExit();
536 function unexpectedValueError( $name, $val )
538 $this->fatalError( wfMsg( "unexpected", $name, $val ) );
541 function fileCopyError( $old, $new )
543 $this->fatalError( wfMsg( "filecopyerror", $old, $new ) );
546 function fileRenameError( $old, $new )
548 $this->fatalError( wfMsg( "filerenameerror", $old, $new ) );
551 function fileDeleteError( $name )
553 $this->fatalError( wfMsg( "filedeleteerror", $name ) );
556 function fileNotFoundError( $name )
558 $this->fatalError( wfMsg( "filenotfound", $name ) );
561 function returnToMain( $auto = true )
563 global $wgUser, $wgOut, $returnto;
565 $sk = $wgUser->getSkin();
566 if ( "" == $returnto ) {
567 $returnto = wfMsg( "mainpage" );
569 $link = $sk->makeKnownLink( $returnto, "" );
571 $r = wfMsg( "returnto", $link );
572 if ( $auto ) {
573 $wgOut->addMeta( "http:Refresh", "10;url=" .
574 wfLocalUrlE( wfUrlencode( $returnto ) ) );
576 $wgOut->addHTML( "\n<p>$r\n" );
580 function categoryMagic ()
582 global $wgTitle , $wgUseCategoryMagic ;
583 if ( !isset ( $wgUseCategoryMagic ) || !$wgUseCategoryMagic ) return ;
584 $id = $wgTitle->getArticleID() ;
585 $cat = ucfirst ( wfMsg ( "category" ) ) ;
586 $ti = $wgTitle->getText() ;
587 $ti = explode ( ":" , $ti , 2 ) ;
588 if ( $cat != $ti[0] ) return "" ;
589 $r = "<br break=all>\n" ;
591 $articles = array() ;
592 $parents = array () ;
593 $children = array() ;
596 global $wgUser ;
597 $sk = $wgUser->getSkin() ;
598 $sql = "SELECT l_from FROM links WHERE l_to={$id}" ;
599 $res = wfQuery ( $sql, DB_READ ) ;
600 while ( $x = wfFetchObject ( $res ) )
602 # $t = new Title ;
603 # $t->newFromDBkey ( $x->l_from ) ;
604 # $t = $t->getText() ;
605 $t = $x->l_from ;
606 $y = explode ( ":" , $t , 2 ) ;
607 if ( count ( $y ) == 2 && $y[0] == $cat ) {
608 array_push ( $children , $sk->makeLink ( $t , $y[1] ) ) ;
609 } else {
610 array_push ( $articles , $sk->makeLink ( $t ) ) ;
613 wfFreeResult ( $res ) ;
615 # Children
616 if ( count ( $children ) > 0 )
618 asort ( $children ) ;
619 $r .= "<h2>".wfMsg("subcategories")."</h2>\n" ;
620 $r .= implode ( ", " , $children ) ;
623 # Articles
624 if ( count ( $articles ) > 0 )
626 asort ( $articles ) ;
627 $h = wfMsg( "category_header", $ti[1] );
628 $r .= "<h2>{$h}</h2>\n" ;
629 $r .= implode ( ", " , $articles ) ;
633 return $r ;
636 function getHTMLattrs ()
638 $htmlattrs = array( # Allowed attributes--no scripting, etc.
639 "title", "align", "lang", "dir", "width", "height",
640 "bgcolor", "clear", /* BR */ "noshade", /* HR */
641 "cite", /* BLOCKQUOTE, Q */ "size", "face", "color",
642 /* FONT */ "type", "start", "value", "compact",
643 /* For various lists, mostly deprecated but safe */
644 "summary", "width", "border", "frame", "rules",
645 "cellspacing", "cellpadding", "valign", "char",
646 "charoff", "colgroup", "col", "span", "abbr", "axis",
647 "headers", "scope", "rowspan", "colspan", /* Tables */
648 "id", "class", "name", "style" /* For CSS */
650 return $htmlattrs ;
653 function fixTagAttributes ( $t )
655 if ( trim ( $t ) == "" ) return "" ; # Saves runtime ;-)
656 $htmlattrs = $this->getHTMLattrs() ;
658 # Strip non-approved attributes from the tag
659 $t = preg_replace(
660 "/(\\w+)(\\s*=\\s*([^\\s\">]+|\"[^\">]*\"))?/e",
661 "(in_array(strtolower(\"\$1\"),\$htmlattrs)?(\"\$1\".((\"x\$3\" != \"x\")?\"=\$3\":'')):'')",
662 $t);
663 # Strip javascript "expression" from stylesheets. Brute force approach:
664 # If anythin offensive is found, all attributes of the HTML tag are dropped
666 if( preg_match(
667 "/style\\s*=.*(expression|tps*:\/\/|url\\s*\().*/is",
668 wfMungeToUtf8( $t ) ) )
670 $t="";
673 return trim ( $t ) ;
676 function doTableStuff ( $t )
678 $t = explode ( "\n" , $t ) ;
679 $td = array () ; # Is currently a td tag open?
680 $ltd = array () ; # Was it TD or TH?
681 $tr = array () ; # Is currently a tr tag open?
682 $ltr = array () ; # tr attributes
683 foreach ( $t AS $k => $x )
685 $x = rtrim ( $x ) ;
686 $fc = substr ( $x , 0 , 1 ) ;
687 if ( "{|" == substr ( $x , 0 , 2 ) )
689 $t[$k] = "<table " . $this->fixTagAttributes ( substr ( $x , 3 ) ) . ">" ;
690 array_push ( $td , false ) ;
691 array_push ( $ltd , "" ) ;
692 array_push ( $tr , false ) ;
693 array_push ( $ltr , "" ) ;
695 else if ( count ( $td ) == 0 ) { } # Don't do any of the following
696 else if ( "|}" == substr ( $x , 0 , 2 ) )
698 $z = "</table>\n" ;
699 $l = array_pop ( $ltd ) ;
700 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
701 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
702 array_pop ( $ltr ) ;
703 $t[$k] = $z ;
705 /* else if ( "|_" == substr ( $x , 0 , 2 ) ) # Caption
707 $z = trim ( substr ( $x , 2 ) ) ;
708 $t[$k] = "<caption>{$z}</caption>\n" ;
710 else if ( "|-" == substr ( $x , 0 , 2 ) ) # Allows for |---------------
712 $x = substr ( $x , 1 ) ;
713 while ( $x != "" && substr ( $x , 0 , 1 ) == '-' ) $x = substr ( $x , 1 ) ;
714 $z = "" ;
715 $l = array_pop ( $ltd ) ;
716 if ( array_pop ( $tr ) ) $z = "</tr>" . $z ;
717 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
718 array_pop ( $ltr ) ;
719 $t[$k] = $z ;
720 array_push ( $tr , false ) ;
721 array_push ( $td , false ) ;
722 array_push ( $ltd , "" ) ;
723 array_push ( $ltr , $this->fixTagAttributes ( $x ) ) ;
725 else if ( "|" == $fc || "!" == $fc || "|+" == substr ( $x , 0 , 2 ) ) # Caption
727 if ( "|+" == substr ( $x , 0 , 2 ) )
729 $fc = "+" ;
730 $x = substr ( $x , 1 ) ;
732 $after = substr ( $x , 1 ) ;
733 if ( $fc == "!" ) $after = str_replace ( "!!" , "||" , $after ) ;
734 $after = explode ( "||" , $after ) ;
735 $t[$k] = "" ;
736 foreach ( $after AS $theline )
738 $z = "" ;
739 if ( $fc != "+" )
741 $tra = array_pop ( $ltr ) ;
742 if ( !array_pop ( $tr ) ) $z = "<tr {$tra}>\n" ;
743 array_push ( $tr , true ) ;
744 array_push ( $ltr , "" ) ;
747 $l = array_pop ( $ltd ) ;
748 if ( array_pop ( $td ) ) $z = "</{$l}>" . $z ;
749 if ( $fc == "|" ) $l = "TD" ;
750 else if ( $fc == "!" ) $l = "TH" ;
751 else if ( $fc == "+" ) $l = "CAPTION" ;
752 else $l = "" ;
753 array_push ( $ltd , $l ) ;
754 $y = explode ( "|" , $theline , 2 ) ;
755 if ( count ( $y ) == 1 ) $y = "{$z}<{$l}>{$y[0]}" ;
756 else $y = $y = "{$z}<{$l} ".$this->fixTagAttributes($y[0]).">{$y[1]}" ;
757 $t[$k] .= $y ;
758 array_push ( $td , true ) ;
763 # Closing open td, tr && table
764 while ( count ( $td ) > 0 )
766 if ( array_pop ( $td ) ) $t[] = "</td>" ;
767 if ( array_pop ( $tr ) ) $t[] = "</tr>" ;
768 $t[] = "</table>" ;
771 $t = implode ( "\n" , $t ) ;
772 # $t = $this->removeHTMLtags( $t );
773 return $t ;
776 # Well, OK, it's actually about 14 passes. But since all the
777 # hard lifting is done inside PHP's regex code, it probably
778 # wouldn't speed things up much to add a real parser.
780 function doWikiPass2( $text, $linestart )
782 global $wgUser, $wgLang, $wgUseDynamicDates;
783 $fname = "OutputPage::doWikiPass2";
784 wfProfileIn( $fname );
786 $text = $this->removeHTMLtags( $text );
787 $text = $this->replaceVariables( $text );
789 $text = preg_replace( "/(^|\n)-----*/", "\\1<hr>", $text );
790 $text = str_replace ( "<HR>", "<hr>", $text );
792 $text = $this->doAllQuotes( $text );
793 $text = $this->doHeadings( $text );
794 $text = $this->doBlockLevels( $text, $linestart );
796 if($wgUseDynamicDates) {
797 global $wgDateFormatter;
798 $text = $wgDateFormatter->reformat( $wgUser->getOption("date"), $text );
801 $text = $this->replaceExternalLinks( $text );
802 $text = $this->replaceInternalLinks ( $text );
803 $text = $this->doTableStuff ( $text ) ;
805 $text = $this->magicISBN( $text );
806 $text = $this->magicRFC( $text );
807 $text = $this->formatHeadings( $text );
809 $sk = $wgUser->getSkin();
810 $text = $sk->transformContent( $text );
811 $text .= $this->categoryMagic () ;
813 wfProfileOut( $fname );
814 return $text;
817 /* private */ function doAllQuotes( $text )
819 $outtext = "";
820 $lines = explode( "\r\n", $text );
821 foreach ( $lines as $line ) {
822 $outtext .= $this->doQuotes ( "", $line, "" ) . "\r\n";
824 return $outtext;
827 /* private */ function doQuotes( $pre, $text, $mode )
829 if ( preg_match( "/^(.*)''(.*)$/sU", $text, $m ) ) {
830 $m1_strong = ($m[1] == "") ? "" : "<strong>{$m[1]}</strong>";
831 $m1_em = ($m[1] == "") ? "" : "<em>{$m[1]}</em>";
832 if ( substr ($m[2], 0, 1) == "'" ) {
833 $m[2] = substr ($m[2], 1);
834 if ($mode == "em") {
835 return $this->doQuotes ( $m[1], $m[2], ($m[1] == "") ? "both" : "emstrong" );
836 } else if ($mode == "strong") {
837 return $m1_strong . $this->doQuotes ( "", $m[2], "" );
838 } else if (($mode == "emstrong") || ($mode == "both")) {
839 return $this->doQuotes ( "", $pre.$m1_strong.$m[2], "em" );
840 } else if ($mode == "strongem") {
841 return "<strong>{$pre}{$m1_em}</strong>" . $this->doQuotes ( "", $m[2], "em" );
842 } else {
843 return $m[1] . $this->doQuotes ( "", $m[2], "strong" );
845 } else {
846 if ($mode == "strong") {
847 return $this->doQuotes ( $m[1], $m[2], ($m[1] == "") ? "both" : "strongem" );
848 } else if ($mode == "em") {
849 return $m1_em . $this->doQuotes ( "", $m[2], "" );
850 } else if ($mode == "emstrong") {
851 return "<em>{$pre}{$m1_strong}</em>" . $this->doQuotes ( "", $m[2], "strong" );
852 } else if (($mode == "strongem") || ($mode == "both")) {
853 return $this->doQuotes ( "", $pre.$m1_em.$m[2], "strong" );
854 } else {
855 return $m[1] . $this->doQuotes ( "", $m[2], "em" );
858 } else {
859 $text_strong = ($text == "") ? "" : "<strong>{$text}</strong>";
860 $text_em = ($text == "") ? "" : "<em>{$text}</em>";
861 if ($mode == "") {
862 return $pre . $text;
863 } else if ($mode == "em") {
864 return $pre . $text_em;
865 } else if ($mode == "strong") {
866 return $pre . $text_strong;
867 } else if ($mode == "strongem") {
868 return (($pre == "") && ($text == "")) ? "" : "<strong>{$pre}{$text_em}</strong>";
869 } else {
870 return (($pre == "") && ($text == "")) ? "" : "<em>{$pre}{$text_strong}</em>";
875 /* private */ function doHeadings( $text )
877 for ( $i = 6; $i >= 1; --$i ) {
878 $h = substr( "======", 0, $i );
879 $text = preg_replace( "/^{$h}([^=]+){$h}(\\s|$)/m",
880 "<h{$i}>\\1</h{$i}>\\2", $text );
882 return $text;
885 # Note: we have to do external links before the internal ones,
886 # and otherwise take great care in the order of things here, so
887 # that we don't end up interpreting some URLs twice.
889 /* private */ function replaceExternalLinks( $text )
891 $fname = "OutputPage::replaceExternalLinks";
892 wfProfileIn( $fname );
893 $text = $this->subReplaceExternalLinks( $text, "http", true );
894 $text = $this->subReplaceExternalLinks( $text, "https", true );
895 $text = $this->subReplaceExternalLinks( $text, "ftp", false );
896 $text = $this->subReplaceExternalLinks( $text, "irc", false );
897 $text = $this->subReplaceExternalLinks( $text, "gopher", false );
898 $text = $this->subReplaceExternalLinks( $text, "news", false );
899 $text = $this->subReplaceExternalLinks( $text, "mailto", false );
900 wfProfileOut( $fname );
901 return $text;
904 /* private */ function subReplaceExternalLinks( $s, $protocol, $autonumber )
906 global $wgUser, $printable;
907 global $wgAllowExternalImages;
910 $unique = "4jzAfzB8hNvf4sqyO9Edd8pSmk9rE2in0Tgw3";
911 $uc = "A-Za-z0-9_\\/~%\\-+&*#?!=()@\\x80-\\xFF";
913 # this is the list of separators that should be ignored if they
914 # are the last character of an URL but that should be included
915 # if they occur within the URL, e.g. "go to www.foo.com, where .."
916 # in this case, the last comma should not become part of the URL,
917 # but in "www.foo.com/123,2342,32.htm" it should.
918 $sep = ",;\.:";
919 $fnc = "A-Za-z0-9_.,~%\\-+&;#*?!=()@\\x80-\\xFF";
920 $images = "gif|png|jpg|jpeg";
922 # PLEASE NOTE: The curly braces { } are not part of the regex,
923 # they are interpreted as part of the string (used to tell PHP
924 # that the content of the string should be inserted there).
925 $e1 = "/(^|[^\\[])({$protocol}:)([{$uc}{$sep}]+)\\/([{$fnc}]+)\\." .
926 "((?i){$images})([^{$uc}]|$)/";
928 $e2 = "/(^|[^\\[])({$protocol}:)(([".$uc."]|[".$sep."][".$uc."])+)([^". $uc . $sep. "]|[".$sep."]|$)/";
929 $sk = $wgUser->getSkin();
931 if ( $autonumber and $wgAllowExternalImages) { # Use img tags only for HTTP urls
932 $s = preg_replace( $e1, "\\1" . $sk->makeImage( "{$unique}:\\3" .
933 "/\\4.\\5", "\\4.\\5" ) . "\\6", $s );
935 $s = preg_replace( $e2, "\\1" . "<a href=\"{$unique}:\\3\"" .
936 $sk->getExternalLinkAttributes( "{$unique}:\\3", wfEscapeHTML(
937 "{$unique}:\\3" ) ) . ">" . wfEscapeHTML( "{$unique}:\\3" ) .
938 "</a>\\5", $s );
939 $s = str_replace( $unique, $protocol, $s );
941 $a = explode( "[{$protocol}:", " " . $s );
942 $s = array_shift( $a );
943 $s = substr( $s, 1 );
945 $e1 = "/^([{$uc}"."{$sep}]+)](.*)\$/sD";
946 $e2 = "/^([{$uc}"."{$sep}]+)\\s+([^\\]]+)](.*)\$/sD";
948 foreach ( $a as $line ) {
949 if ( preg_match( $e1, $line, $m ) ) {
950 $link = "{$protocol}:{$m[1]}";
951 $trail = $m[2];
952 if ( $autonumber ) { $text = "[" . ++$this->mAutonumber . "]"; }
953 else { $text = wfEscapeHTML( $link ); }
954 } else if ( preg_match( $e2, $line, $m ) ) {
955 $link = "{$protocol}:{$m[1]}";
956 $text = $m[2];
957 $trail = $m[3];
958 } else {
959 $s .= "[{$protocol}:" . $line;
960 continue;
962 if ( $printable == "yes") $paren = " (<i>" . htmlspecialchars ( $link ) . "</i>)";
963 else $paren = "";
964 $la = $sk->getExternalLinkAttributes( $link, $text );
965 $s .= "<a href='{$link}'{$la}>{$text}</a>{$paren}{$trail}";
968 return $s;
971 /* private */ function replaceInternalLinks( $s )
973 global $wgTitle, $wgUser, $wgLang;
974 global $wgLinkCache, $wgInterwikiMagic, $wgUseCategoryMagic;
975 global $wgNamespacesWithSubpages, $wgLanguageCode;
976 wfProfileIn( $fname = "OutputPage::replaceInternalLinks" );
978 wfProfileIn( "$fname-setup" );
979 $tc = Title::legalChars() . "#";
980 $sk = $wgUser->getSkin();
982 $a = explode( "[[", " " . $s );
983 $s = array_shift( $a );
984 $s = substr( $s, 1 );
986 # Match a link having the form [[namespace:link|alternate]]trail
987 $e1 = "/^([{$tc}]+)(?:\\|([^]]+))?]](.*)\$/sD";
988 # Match the end of a line for a word that's not followed by whitespace,
989 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
990 #$e2 = "/^(.*)\\b(\\w+)\$/suD";
991 #$e2 = "/^(.*\\s)(\\S+)\$/suD";
992 $e2 = '/^(.*\s)([a-zA-Z\x80-\xff]+)$/sD';
995 # Special and Media are pseudo-namespaces; no pages actually exist in them
996 $image = Namespace::getImage();
997 $special = Namespace::getSpecial();
998 $media = Namespace::getMedia();
999 $nottalk = !Namespace::isTalk( $wgTitle->getNamespace() );
1001 if ( $wgLang->linkPrefixExtension() && preg_match( $e2, $s, $m ) ) {
1002 $new_prefix = $m[2];
1003 $s = $m[1];
1004 } else {
1005 $new_prefix="";
1008 wfProfileOut( "$fname-setup" );
1010 foreach ( $a as $line ) {
1011 $prefix = $new_prefix;
1012 if ( $wgUseLinkPrefixCombination && preg_match( $e2, $line, $m ) ) {
1013 $new_prefix = $m[2];
1014 $line = $m[1];
1015 } else {
1016 $new_prefix = "";
1018 if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
1019 $text = $m[2];
1020 $trail = $m[3];
1021 } else { # Invalid form; output directly
1022 $s .= $prefix . "[[" . $line ;
1023 continue;
1026 /* Valid link forms:
1027 Foobar -- normal
1028 :Foobar -- override special treatment of prefix (images, language links)
1029 /Foobar -- convert to CurrentPage/Foobar
1030 /Foobar/ -- convert to CurrentPage/Foobar, strip the initial / from text
1032 $c = substr($m[1],0,1);
1033 $noforce = ($c != ":");
1034 if( $c == "/" ) { # subpage
1035 if(substr($m[1],-1,1)=="/") { # / at end means we don't want the slash to be shown
1036 $m[1]=substr($m[1],1,strlen($m[1])-2);
1037 $noslash=$m[1];
1038 } else {
1039 $noslash=substr($m[1],1);
1041 if($wgNamespacesWithSubpages[$wgTitle->getNamespace()]) { # subpages allowed here
1042 $link = $wgTitle->getPrefixedText(). "/" . trim($noslash);
1043 if( "" == $text ) {
1044 $text= $m[1];
1045 } # this might be changed for ugliness reasons
1046 } else {
1047 $link = $noslash; # no subpage allowed, use standard link
1049 } elseif( $noforce ) { # no subpage
1050 $link = $m[1];
1051 } else {
1052 $link = substr( $m[1], 1 );
1054 if( "" == $text )
1055 $text = $link;
1057 $nt = Title::newFromText( $link );
1058 if( !$nt ) {
1059 $s .= $prefix . "[[" . $line;
1060 continue;
1062 $ns = $nt->getNamespace();
1063 $iw = $nt->getInterWiki();
1064 if( $noforce ) {
1065 if( $iw && $wgInterwikiMagic && $nottalk && $wgLang->getLanguageName( $iw ) ) {
1066 array_push( $this->mLanguageLinks, $nt->getPrefixedText() );
1067 $s .= $prefix . $trail;
1068 continue;
1070 if( $ns == $image ) {
1071 $s .= $prefix . $sk->makeImageLinkObj( $nt, $text ) . $trail;
1072 $wgLinkCache->addImageLinkObj( $nt );
1073 continue;
1076 if( ( $nt->getPrefixedText() == $wgTitle->getPrefixedText() ) &&
1077 ( strpos( $link, "#" ) == FALSE ) ) {
1078 $s .= $prefix . "<strong>" . $text . "</strong>" . $trail;
1079 continue;
1081 if( $ns == $media ) {
1082 $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
1083 $wgLinkCache->addImageLinkObj( $nt );
1084 continue;
1085 } elseif( $ns == $special ) {
1086 $s .= $prefix . $sk->makeKnownLinkObj( $nt, $text, "", $trail );
1087 continue;
1089 $s .= $sk->makeLinkObj( $nt, $text, "", $trail , $prefix );
1091 wfProfileOut( $fname );
1092 return $s;
1095 # Some functions here used by doBlockLevels()
1097 /* private */ function closeParagraph()
1099 $result = "";
1100 if ( 0 != strcmp( "p", $this->mLastSection ) &&
1101 0 != strcmp( "", $this->mLastSection ) ) {
1102 $result = "</" . $this->mLastSection . ">";
1104 $this->mLastSection = "";
1105 return $result."\n";
1107 # getCommon() returns the length of the longest common substring
1108 # of both arguments, starting at the beginning of both.
1110 /* private */ function getCommon( $st1, $st2 )
1112 $fl = strlen( $st1 );
1113 $shorter = strlen( $st2 );
1114 if ( $fl < $shorter ) { $shorter = $fl; }
1116 for ( $i = 0; $i < $shorter; ++$i ) {
1117 if ( $st1{$i} != $st2{$i} ) { break; }
1119 return $i;
1121 # These next three functions open, continue, and close the list
1122 # element appropriate to the prefix character passed into them.
1124 /* private */ function openList( $char )
1126 $result = $this->closeParagraph();
1128 if ( "*" == $char ) { $result .= "<ul><li>"; }
1129 else if ( "#" == $char ) { $result .= "<ol><li>"; }
1130 else if ( ":" == $char ) { $result .= "<dl><dd>"; }
1131 else if ( ";" == $char ) {
1132 $result .= "<dl><dt>";
1133 $this->mDTopen = true;
1135 else { $result = "<!-- ERR 1 -->"; }
1137 return $result;
1140 /* private */ function nextItem( $char )
1142 if ( "*" == $char || "#" == $char ) { return "</li><li>"; }
1143 else if ( ":" == $char || ";" == $char ) {
1144 $close = "</dd>";
1145 if ( $this->mDTopen ) { $close = "</dt>"; }
1146 if ( ";" == $char ) {
1147 $this->mDTopen = true;
1148 return $close . "<dt>";
1149 } else {
1150 $this->mDTopen = false;
1151 return $close . "<dd>";
1154 return "<!-- ERR 2 -->";
1157 /* private */function closeList( $char )
1159 if ( "*" == $char ) { $text = "</li></ul>"; }
1160 else if ( "#" == $char ) { $text = "</li></ol>"; }
1161 else if ( ":" == $char ) {
1162 if ( $this->mDTopen ) {
1163 $this->mDTopen = false;
1164 $text = "</dt></dl>";
1165 } else {
1166 $text = "</dd></dl>";
1169 else { return "<!-- ERR 3 -->"; }
1170 return $text."\n";
1173 /* private */ function doBlockLevels( $text, $linestart )
1175 $fname = "OutputPage::doBlockLevels";
1176 wfProfileIn( $fname );
1177 # Parsing through the text line by line. The main thing
1178 # happening here is handling of block-level elements p, pre,
1179 # and making lists from lines starting with * # : etc.
1181 $a = explode( "\n", $text );
1182 $text = $lastPref = "";
1183 $this->mDTopen = $inBlockElem = false;
1185 if ( ! $linestart ) { $text .= array_shift( $a ); }
1186 foreach ( $a as $t ) {
1187 if ( "" != $text ) { $text .= "\n"; }
1189 $oLine = $t;
1190 $opl = strlen( $lastPref );
1191 $npl = strspn( $t, "*#:;" );
1192 $pref = substr( $t, 0, $npl );
1193 $pref2 = str_replace( ";", ":", $pref );
1194 $t = substr( $t, $npl );
1196 if ( 0 != $npl && 0 == strcmp( $lastPref, $pref2 ) ) {
1197 $text .= $this->nextItem( substr( $pref, -1 ) );
1199 if ( ";" == substr( $pref, -1 ) ) {
1200 $cpos = strpos( $t, ":" );
1201 if ( ! ( false === $cpos ) ) {
1202 $term = substr( $t, 0, $cpos );
1203 $text .= $term . $this->nextItem( ":" );
1204 $t = substr( $t, $cpos + 1 );
1207 } else if (0 != $npl || 0 != $opl) {
1208 $cpl = $this->getCommon( $pref, $lastPref );
1210 while ( $cpl < $opl ) {
1211 $text .= $this->closeList( $lastPref{$opl-1} );
1212 --$opl;
1214 if ( $npl <= $cpl && $cpl > 0 ) {
1215 $text .= $this->nextItem( $pref{$cpl-1} );
1217 while ( $npl > $cpl ) {
1218 $char = substr( $pref, $cpl, 1 );
1219 $text .= $this->openList( $char );
1221 if ( ";" == $char ) {
1222 $cpos = strpos( $t, ":" );
1223 if ( ! ( false === $cpos ) ) {
1224 $term = substr( $t, 0, $cpos );
1225 $text .= $term . $this->nextItem( ":" );
1226 $t = substr( $t, $cpos + 1 );
1229 ++$cpl;
1231 $lastPref = $pref2;
1233 if ( 0 == $npl ) { # No prefix--go to paragraph mode
1234 if ( preg_match(
1235 "/(<table|<blockquote|<h1|<h2|<h3|<h4|<h5|<h6)/i", $t ) ) {
1236 $text .= $this->closeParagraph();
1237 $inBlockElem = true;
1239 if ( ! $inBlockElem ) {
1240 if ( " " == $t{0} ) {
1241 $newSection = "pre";
1242 # $t = wfEscapeHTML( $t );
1244 else { $newSection = "p"; }
1246 if ( 0 == strcmp( "", trim( $oLine ) ) ) {
1247 $text .= $this->closeParagraph();
1248 $text .= "<" . $newSection . ">";
1249 } else if ( 0 != strcmp( $this->mLastSection,
1250 $newSection ) ) {
1251 $text .= $this->closeParagraph();
1252 if ( 0 != strcmp( "p", $newSection ) ) {
1253 $text .= "<" . $newSection . ">";
1256 $this->mLastSection = $newSection;
1258 if ( $inBlockElem &&
1259 preg_match( "/(<\\/table|<\\/blockquote|<\\/h1|<\\/h2|<\\/h3|<\\/h4|<\\/h5|<\\/h6)/i", $t ) ) {
1260 $inBlockElem = false;
1263 $text .= $t;
1265 while ( $npl ) {
1266 $text .= $this->closeList( $pref2{$npl-1} );
1267 --$npl;
1269 if ( "" != $this->mLastSection ) {
1270 if ( "p" != $this->mLastSection ) {
1271 $text .= "</" . $this->mLastSection . ">";
1273 $this->mLastSection = "";
1275 wfProfileOut( $fname );
1276 return $text;
1279 /* private */ function replaceVariables( $text )
1281 global $wgLang, $wgCurOut;
1282 $fname = "OutputPage::replaceVariables";
1283 wfProfileIn( $fname );
1285 $magic = array();
1287 # Basic variables
1288 # See Language.php for the definition of each magic word
1289 # As with sigs, this uses the server's local time -- ensure
1290 # this is appropriate for your audience!
1292 $magic[MAG_CURRENTMONTH] = date( "m" );
1293 $magic[MAG_CURRENTMONTHNAME] = $wgLang->getMonthName( date("n") );
1294 $magic[MAG_CURRENTMONTHNAMEGEN] = $wgLang->getMonthNameGen( date("n") );
1295 $magic[MAG_CURRENTDAY] = date("j");
1296 $magic[MAG_CURRENTDAYNAME] = $wgLang->getWeekdayName( date("w")+1 );
1297 $magic[MAG_CURRENTYEAR] = date( "Y" );
1298 $magic[MAG_CURRENTTIME] = $wgLang->time( wfTimestampNow(), false );
1300 $this->mContainsOldMagic += MagicWord::replaceMultiple($magic, $text, $text);
1302 $mw =& MagicWord::get( MAG_NUMBEROFARTICLES );
1303 if ( $mw->match( $text ) ) {
1304 $v = wfNumberOfArticles();
1305 $text = $mw->replace( $v, $text );
1306 if( $mw->getWasModified() ) { $this->mContainsOldMagic++; }
1309 # "Variables" with an additional parameter e.g. {{MSG:wikipedia}}
1310 # The callbacks are at the bottom of this file
1311 $wgCurOut = $this;
1312 $mw =& MagicWord::get( MAG_MSG );
1313 $text = $mw->substituteCallback( $text, "wfReplaceMsgVar" );
1314 if( $mw->getWasModified() ) { $this->mContainsNewMagic++; }
1316 $mw =& MagicWord::get( MAG_MSGNW );
1317 $text = $mw->substituteCallback( $text, "wfReplaceMsgnwVar" );
1318 if( $mw->getWasModified() ) { $this->mContainsNewMagic++; }
1320 wfProfileOut( $fname );
1321 return $text;
1324 # Cleans up HTML, removes dangerous tags and attributes
1325 /* private */ function removeHTMLtags( $text )
1327 $fname = "OutputPage::removeHTMLtags";
1328 wfProfileIn( $fname );
1329 $htmlpairs = array( # Tags that must be closed
1330 "b", "i", "u", "font", "big", "small", "sub", "sup", "h1",
1331 "h2", "h3", "h4", "h5", "h6", "cite", "code", "em", "s",
1332 "strike", "strong", "tt", "var", "div", "center",
1333 "blockquote", "ol", "ul", "dl", "table", "caption", "pre",
1334 "ruby", "rt" , "rb" , "rp"
1336 $htmlsingle = array(
1337 "br", "p", "hr", "li", "dt", "dd"
1339 $htmlnest = array( # Tags that can be nested--??
1340 "table", "tr", "td", "th", "div", "blockquote", "ol", "ul",
1341 "dl", "font", "big", "small", "sub", "sup"
1343 $tabletags = array( # Can only appear inside table
1344 "td", "th", "tr"
1347 $htmlsingle = array_merge( $tabletags, $htmlsingle );
1348 $htmlelements = array_merge( $htmlsingle, $htmlpairs );
1350 $htmlattrs = $this->getHTMLattrs () ;
1352 # Remove HTML comments
1353 $text = preg_replace( "/<!--.*-->/sU", "", $text );
1355 $bits = explode( "<", $text );
1356 $text = array_shift( $bits );
1357 $tagstack = array(); $tablestack = array();
1359 foreach ( $bits as $x ) {
1360 $prev = error_reporting( E_ALL & ~( E_NOTICE | E_WARNING ) );
1361 preg_match( "/^(\\/?)(\\w+)([^>]*)(\\/{0,1}>)([^<]*)$/",
1362 $x, $regs );
1363 list( $qbar, $slash, $t, $params, $brace, $rest ) = $regs;
1364 error_reporting( $prev );
1366 $badtag = 0 ;
1367 if ( in_array( $t = strtolower( $t ), $htmlelements ) ) {
1368 # Check our stack
1369 if ( $slash ) {
1370 # Closing a tag...
1371 if ( ! in_array( $t, $htmlsingle ) &&
1372 ( $ot = array_pop( $tagstack ) ) != $t ) {
1373 array_push( $tagstack, $ot );
1374 $badtag = 1;
1375 } else {
1376 if ( $t == "table" ) {
1377 $tagstack = array_pop( $tablestack );
1379 $newparams = "";
1381 } else {
1382 # Keep track for later
1383 if ( in_array( $t, $tabletags ) &&
1384 ! in_array( "table", $tagstack ) ) {
1385 $badtag = 1;
1386 } else if ( in_array( $t, $tagstack ) &&
1387 ! in_array ( $t , $htmlnest ) ) {
1388 $badtag = 1 ;
1389 } else if ( ! in_array( $t, $htmlsingle ) ) {
1390 if ( $t == "table" ) {
1391 array_push( $tablestack, $tagstack );
1392 $tagstack = array();
1394 array_push( $tagstack, $t );
1396 # Strip non-approved attributes from the tag
1397 $newparams = $this->fixTagAttributes($params);
1400 if ( ! $badtag ) {
1401 $rest = str_replace( ">", "&gt;", $rest );
1402 $text .= "<$slash$t $newparams$brace$rest";
1403 continue;
1406 $text .= "&lt;" . str_replace( ">", "&gt;", $x);
1408 # Close off any remaining tags
1409 while ( $t = array_pop( $tagstack ) ) {
1410 $text .= "</$t>\n";
1411 if ( $t == "table" ) { $tagstack = array_pop( $tablestack ); }
1413 wfProfileOut( $fname );
1414 return $text;
1419 * This function accomplishes several tasks:
1420 * 1) Auto-number headings if that option is enabled
1421 * 2) Add an [edit] link to sections for logged in users who have enabled the option
1422 * 3) Add a Table of contents on the top for users who have enabled the option
1423 * 4) Auto-anchor headings
1425 * It loops through all headlines, collects the necessary data, then splits up the
1426 * string and re-inserts the newly formatted headlines.
1428 * */
1429 /* private */ function formatHeadings( $text )
1431 global $wgUser,$wgArticle,$wgTitle,$wpPreview;
1432 $nh=$wgUser->getOption( "numberheadings" );
1433 $st=$wgUser->getOption( "showtoc" );
1434 if(!$wgTitle->userCanEdit()) {
1435 $es=0;
1436 $esr=0;
1437 } else {
1438 $es=$wgUser->getID() && $wgUser->getOption( "editsection" );
1439 $esr=$wgUser->getID() && $wgUser->getOption( "editsectiononrightclick" );
1442 # Inhibit editsection links if requested in the page
1443 $esw =& MagicWord::get( MAG_NOEDITSECTION );
1444 if ($esw->matchAndRemove( $text )) {
1445 $es=0;
1447 # if the string __NOTOC__ (not case-sensitive) occurs in the HTML,
1448 # do not add TOC
1449 $mw =& MagicWord::get( MAG_NOTOC );
1450 if ($mw->matchAndRemove( $text ))
1452 $st = 0;
1455 # never add the TOC to the Main Page. This is an entry page that should not
1456 # be more than 1-2 screens large anyway
1457 if($wgTitle->getPrefixedText()==wfMsg("mainpage")) {$st=0;}
1459 # We need this to perform operations on the HTML
1460 $sk=$wgUser->getSkin();
1462 # Get all headlines for numbering them and adding funky stuff like [edit]
1463 # links
1464 preg_match_all("/<H([1-6])(.*?>)(.*?)<\/H[1-6]>/i",$text,$matches);
1466 # headline counter
1467 $c=0;
1469 # Ugh .. the TOC should have neat indentation levels which can be
1470 # passed to the skin functions. These are determined here
1471 foreach($matches[3] as $headline) {
1472 if($level) { $prevlevel=$level;}
1473 $level=$matches[1][$c];
1474 if(($nh||$st) && $prevlevel && $level>$prevlevel) {
1476 $h[$level]=0; // reset when we enter a new level
1477 $toc.=$sk->tocIndent($level-$prevlevel);
1478 $toclevel+=$level-$prevlevel;
1481 if(($nh||$st) && $level<$prevlevel) {
1482 $h[$level+1]=0; // reset when we step back a level
1483 $toc.=$sk->tocUnindent($prevlevel-$level);
1484 $toclevel-=$prevlevel-$level;
1487 $h[$level]++; // count number of headlines for each level
1489 if($nh||$st) {
1490 for($i=1;$i<=$level;$i++) {
1491 if($h[$i]) {
1492 if($dot) {$numbering.=".";}
1493 $numbering.=$h[$i];
1494 $dot=1;
1499 // The canonized header is a version of the header text safe to use for links
1501 $canonized_headline=preg_replace("/<.*?>/","",$headline); // strip out HTML
1502 $tocline = trim( $canonized_headline );
1503 $canonized_headline=str_replace('"',"",$canonized_headline);
1504 $canonized_headline=str_replace(" ","_",trim($canonized_headline));
1505 $refer[$c]=$canonized_headline;
1506 $refers[$canonized_headline]++; // count how many in assoc. array so we can track dupes in anchors
1507 $refcount[$c]=$refers[$canonized_headline];
1509 // Prepend the number to the heading text
1511 if($nh||$st) {
1512 $tocline=$numbering ." ". $tocline;
1514 // Don't number the heading if it is the only one (looks silly)
1515 if($nh && count($matches[3]) > 1) {
1516 $headline=$numbering . " " . $headline; // the two are different if the line contains a link
1520 // Create the anchor for linking from the TOC to the section
1522 $anchor=$canonized_headline;
1523 if($refcount[$c]>1) {$anchor.="_".$refcount[$c];}
1524 if($st) {
1525 $toc.=$sk->tocLine($anchor,$tocline,$toclevel);
1527 if($es && !isset($wpPreview)) {
1528 $head[$c].=$sk->editSectionLink($c+1);
1531 // Put it all together
1533 $head[$c].="<h".$level.$matches[2][$c]
1534 ."<a name=\"".$anchor."\">"
1535 .$headline
1536 ."</a>"
1537 ."</h".$level.">";
1539 // Add the edit section link
1541 if($esr && !isset($wpPreview)) {
1542 $head[$c]=$sk->editSectionScript($c+1,$head[$c]);
1545 $numbering="";
1546 $c++;
1547 $dot=0;
1550 if($st) {
1551 $toclines=$c;
1552 $toc.=$sk->tocUnindent($toclevel);
1553 $toc=$sk->tocTable($toc);
1556 // split up and insert constructed headlines
1558 $blocks=preg_split("/<H[1-6].*?>.*?<\/H[1-6]>/i",$text);
1559 $i=0;
1561 foreach($blocks as $block) {
1562 if(($es) && !isset($wpPreview) && $c>0 && $i==0) {
1563 # This is the [edit] link that appears for the top block of text when
1564 # section editing is enabled
1565 $full.=$sk->editSectionLink(0);
1567 $full.=$block;
1568 if($st && $toclines>3 && !$i) {
1569 # Let's add a top anchor just in case we want to link to the top of the page
1570 $full="<a name=\"top\"></a>".$full.$toc;
1573 $full.=$head[$i];
1574 $i++;
1577 return $full;
1580 /* private */ function magicISBN( $text )
1582 global $wgLang;
1584 $a = split( "ISBN ", " $text" );
1585 if ( count ( $a ) < 2 ) return $text;
1586 $text = substr( array_shift( $a ), 1);
1587 $valid = "0123456789-ABCDEFGHIJKLMNOPQRSTUVWXYZ";
1589 foreach ( $a as $x ) {
1590 $isbn = $blank = "" ;
1591 while ( " " == $x{0} ) {
1592 $blank .= " ";
1593 $x = substr( $x, 1 );
1595 while ( strstr( $valid, $x{0} ) != false ) {
1596 $isbn .= $x{0};
1597 $x = substr( $x, 1 );
1599 $num = str_replace( "-", "", $isbn );
1600 $num = str_replace( " ", "", $num );
1602 if ( "" == $num ) {
1603 $text .= "ISBN $blank$x";
1604 } else {
1605 $text .= "<a href=\"" . wfLocalUrlE( $wgLang->specialPage(
1606 "Booksources"), "isbn={$num}" ) . "\" class=\"internal\">ISBN $isbn</a>";
1607 $text .= $x;
1610 return $text;
1613 /* private */ function magicRFC( $text )
1615 return $text;
1618 /* private */ function headElement()
1620 global $wgDocType, $wgDTD, $wgUser, $wgLanguageCode, $wgOutputEncoding, $wgLang;
1622 $ret = "<!DOCTYPE HTML PUBLIC \"$wgDocType\"\n \"$wgDTD\">\n";
1624 if ( "" == $this->mHTMLtitle ) {
1625 $this->mHTMLtitle = $this->mPagetitle;
1627 $rtl = $wgLang->isRTL() ? " dir='RTL'" : "";
1628 $ret .= "<html lang=\"$wgLanguageCode\"$rtl><head><title>{$this->mHTMLtitle}</title>\n";
1629 array_push( $this->mMetatags, array( "http:Content-type", "text/html; charset={$wgOutputEncoding}" ) );
1630 foreach ( $this->mMetatags as $tag ) {
1631 if ( 0 == strcasecmp( "http:", substr( $tag[0], 0, 5 ) ) ) {
1632 $a = "http-equiv";
1633 $tag[0] = substr( $tag[0], 5 );
1634 } else {
1635 $a = "name";
1637 $ret .= "<meta $a=\"{$tag[0]}\" content=\"{$tag[1]}\">\n";
1639 $p = $this->mRobotpolicy;
1640 if ( "" == $p ) { $p = "index,follow"; }
1641 $ret .= "<meta name=\"robots\" content=\"$p\">\n";
1643 if ( count( $this->mKeywords ) > 0 ) {
1644 $ret .= "<meta name=\"keywords\" content=\"" .
1645 implode( ",", $this->mKeywords ) . "\">\n";
1647 foreach ( $this->mLinktags as $tag ) {
1648 $ret .= "<link ";
1649 if ( "" != $tag[0] ) { $ret .= "rel=\"{$tag[0]}\" "; }
1650 if ( "" != $tag[1] ) { $ret .= "rev=\"{$tag[1]}\" "; }
1651 $ret .= "href=\"{$tag[2]}\">\n";
1653 $sk = $wgUser->getSkin();
1654 $ret .= $sk->getHeadScripts();
1655 $ret .= $sk->getUserStyles();
1657 $ret .= "</head>\n";
1658 return $ret;
1661 /* private */ function fillFromParserCache(){
1662 global $wgUser, $wgArticle;
1663 $hash = $wgUser->getPageRenderingHash();
1664 $pageid = intval( $wgArticle->getID() );
1665 $res = wfQuery("SELECT pc_data FROM parsercache WHERE pc_pageid = {$pageid} ".
1666 " AND pc_prefhash = '{$hash}' AND pc_expire > NOW()", DB_WRITE);
1667 $row = wfFetchObject ( $res );
1668 if( $row ){
1669 $data = unserialize( gzuncompress($row->pc_data) );
1670 $this->addHTML( $data['html'] );
1671 $this->mLanguageLinks = $data['mLanguageLinks'];
1672 $this->mCategoryLinks = $data['mCategoryLinks'];
1673 wfProfileOut( $fname );
1674 return true;
1675 } else {
1676 return false;
1680 /* private */ function saveParserCache( $text ){
1681 global $wgUser, $wgArticle;
1682 $hash = $wgUser->getPageRenderingHash();
1683 $pageid = intval( $wgArticle->getID() );
1684 $title = wfStrencode( $wgArticle->mTitle->getPrefixedDBKey() );
1685 $data = array();
1686 $data['html'] = $text;
1687 $data['mLanguageLinks'] = $this->mLanguageLinks;
1688 $data['mCategoryLinks'] = $this->mCategoryLinks;
1689 $ser = addslashes( gzcompress( serialize( $data ) ) );
1690 if( $this->mContainsOldMagic ){
1691 $expire = "1 HOUR";
1692 } else {
1693 $expire = "7 DAY";
1696 wfQuery("REPLACE INTO parsercache (pc_prefhash,pc_pageid,pc_title,pc_data, pc_expire) ".
1697 "VALUES('{$hash}', {$pageid}, '{$title}', '{$ser}', ".
1698 "DATE_ADD(NOW(), INTERVAL {$expire}))", DB_WRITE);
1700 if( rand() % 50 == 0 ){ // more efficient to just do it sometimes
1701 $this->purgeParserCache();
1705 /* static private */ function purgeParserCache(){
1706 wfQuery("DELETE FROM parsercache WHERE pc_expire < NOW() LIMIT 250", DB_WRITE);
1709 /* static */ function parsercacheClearLinksTo( $pid ){
1710 $pid = intval( $pid );
1711 wfQuery("DELETE parsercache FROM parsercache,links ".
1712 "WHERE pc_title=links.l_from AND l_to={$pid}", DB_WRITE);
1713 wfQuery("DELETE FROM parsercache WHERE pc_pageid='{$pid}'", DB_WRITE);
1716 # $title is a prefixed db title, for example like Title->getPrefixedDBkey() returns.
1717 /* static */ function parsercacheClearBrokenLinksTo( $title ){
1718 $title = wfStrencode( $title );
1719 wfQuery("DELETE parsercache FROM parsercache,brokenlinks ".
1720 "WHERE pc_pageid=bl_from AND bl_to='{$title}'", DB_WRITE);
1723 # $pid is a page id
1724 /* static */ function parsercacheClearPage( $pid, $namespace ){
1725 $pid = intval( $pid );
1726 if( $namespace == NS_MEDIAWIKI ){
1727 OutputPage::parsercacheClearLinksTo( $pid );
1728 } else {
1729 wfQuery("DELETE FROM parsercache WHERE pc_pageid='{$pid}'", DB_WRITE);
1734 # Regex callbacks, used in OutputPage::replaceVariables
1736 # Just get rid of the dangerous stuff
1737 # Necessary because replaceVariables is called after removeHTMLtags,
1738 # and message text can come from any user
1739 function wfReplaceMsgVar( $matches ) {
1740 global $wgCurOut, $wgLinkCache;
1741 $text = $wgCurOut->removeHTMLtags( wfMsg( $matches[1] ) );
1742 $wgLinkCache->suspend();
1743 $text = $wgCurOut->replaceInternalLinks( $text );
1744 $wgLinkCache->resume();
1745 $wgLinkCache->addLinkObj( Title::makeTitle( NS_MEDIAWIKI, $matches[1] ) );
1746 return $text;
1749 # Effective <nowiki></nowiki>
1750 # Not real <nowiki> because this is called after nowiki sections are processed
1751 function wfReplaceMsgnwVar( $matches ) {
1752 global $wgCurOut, $wgLinkCache;
1753 $text = wfEscapeWikiText( wfMsg( $matches[1] ) );
1754 $wgLinkCache->addLinkObj( Title::makeTitle( NS_MEDIAWIKI, $matches[1] ) );
1755 return $text;