Capitalization fix in memcached setting
[lhc/web/wiklou.git] / includes / Parser.php
index f97bffc..67faaa7 100644 (file)
@@ -67,7 +67,7 @@ define( 'EXT_IMAGE_REGEX',
  *   performs brace substitution on MediaWiki messages
  *
  * Globals used:
- *    objects:   $wgLang, $wgDateFormatter, $wgLinkCache, $wgCurParser
+ *    objects:   $wgLang, $wgDateFormatter, $wgLinkCache
  *
  * NOT $wgArticle, $wgUser or $wgTitle. Keep them away!
  *
@@ -277,6 +277,7 @@ class Parser
                $pre_content = array();
                $comment_content = array();
                $ext_content = array();
+               $gallery_content = array();
 
                # Replace any instances of the placeholders
                $uniq_prefix = UNIQ_PREFIX;
@@ -330,6 +331,17 @@ class Parser
                        }
                }
 
+               # gallery
+               $text = Parser::extractTags('gallery', $text, $gallery_content, $uniq_prefix);
+               foreach( $gallery_content as $marker => $content ) {
+                       require_once( 'ImageGallery.php' );
+                       if ( $render ) {
+                               $gallery_content[$marker] = Parser::renderImageGallery( $content );
+                       } else {
+                               $gallery_content[$marker] = '<gallery>'.$content.'</gallery>';
+                       }
+               }
+
                # Comments
                if($stripcomments) {
                        $text = Parser::extractTags(STRIP_COMMENTS, $text, $comment_content, $uniq_prefix);
@@ -358,6 +370,7 @@ class Parser
                        $state['math'] = $state['math'] + $math_content;
                        $state['pre'] = $state['pre'] + $pre_content;
                        $state['comment'] = $state['comment'] + $comment_content;
+                       $state['gallery'] = $state['gallery'] + $gallery_content;
 
                        foreach( $ext_content as $tag => $array ) {
                                if ( array_key_exists( $tag, $state ) ) {
@@ -371,13 +384,14 @@ class Parser
                          'math' => $math_content,
                          'pre' => $pre_content,
                          'comment' => $comment_content,
+                         'gallery' => $gallery_content,
                        ) + $ext_content;
                }
                return $text;
        }
 
        /**
-        * restores pre, math, and heiro removed by strip()
+        * restores pre, math, and hiero removed by strip()
         *
         * always call unstripNoWiki() after this one
         * @access private
@@ -680,7 +694,12 @@ class Parser
                }
                $text = $this->doAllQuotes( $text );
                $text = $this->replaceInternalLinks ( $text );
-               $text = $this->replaceExternalLinks( $text );
+               $text = $this->replaceExternalLinks( $text );           
+               
+               # replaceInternalLinks may sometimes leave behind
+               # absolute URLs, which have to be masked to hide them from replaceExternalLinks
+               $text = str_replace("http-noparse://","http://",$text);
+               
                $text = $this->doMagicLinks( $text );
                $text = $this->doTableStuff( $text );
                $text = $this->formatHeadings( $text, $isMain );
@@ -703,7 +722,8 @@ class Parser
                if ( isset( $wgUseGeoMode ) && $wgUseGeoMode ) {
                        $text = $this->magicGEO( $text );
                }
-               $text = $this->magicRFC( $text );
+               $text = $this->magicRFC( $text, 'RFC ', 'rfcurl' );
+               $text = $this->magicRFC( $text, 'PMID ', 'pubmedurl' );
                return $text;
        }
 
@@ -712,11 +732,11 @@ class Parser
         *
         * @access private
         */
-       function doExponent ( $text ) {
+       function doExponent( $text ) {
                $fname = 'Parser::doExponent';
-               wfProfileIn( $fname);
+               wfProfileIn( $fname );
                $text = preg_replace('/\^\^(.*)\^\^/','<small><sup>\\1</sup></small>', $text);
-               wfProfileOut( $fname);
+               wfProfileOut( $fname );
                return $text;
        }
 
@@ -760,8 +780,8 @@ class Parser
         * @access private
         */
        function doQuotes( $text ) {
-               $arr = preg_split ("/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE);
-               if (count ($arr) == 1)
+               $arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE );
+               if ( count( $arr ) == 1 )
                        return $text;
                else
                {
@@ -771,29 +791,29 @@ class Parser
                        $i = 0;
                        $numbold = 0;
                        $numitalics = 0;
-                       foreach ($arr as $r)
+                       foreach ( $arr as $r )
                        {
-                               if (($i % 2) == 1)
+                               if ( ( $i % 2 ) == 1 )
                                {
                                        # If there are ever four apostrophes, assume the first is supposed to
                                        # be text, and the remaining three constitute mark-up for bold text.
-                                       if (strlen ($arr[$i]) == 4)
+                                       if ( strlen( $arr[$i] ) == 4 )
                                        {
                                                $arr[$i-1] .= "'";
                                                $arr[$i] = "'''";
                                        }
                                        # If there are more than 5 apostrophes in a row, assume they're all
                                        # text except for the last 5.
-                                       else if (strlen ($arr[$i]) > 5)
+                                       else if ( strlen( $arr[$i] ) > 5 )
                                        {
-                                               $arr[$i-1] .= str_repeat ("'", strlen ($arr[$i]) - 5);
+                                               $arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
                                                $arr[$i] = "'''''";
                                        }
                                        # Count the number of occurrences of bold and italics mark-ups.
                                        # We are not counting sequences of five apostrophes.
-                                       if (strlen ($arr[$i]) == 2) $numitalics++;  else
-                                       if (strlen ($arr[$i]) == 3) $numbold++;     else
-                                       if (strlen ($arr[$i]) == 5) { $numitalics++; $numbold++; }
+                                       if ( strlen( $arr[$i] ) == 2 ) $numitalics++;  else
+                                       if ( strlen( $arr[$i] ) == 3 ) $numbold++;     else
+                                       if ( strlen( $arr[$i] ) == 5 ) { $numitalics++; $numbold++; }
                                }
                                $i++;
                        }
@@ -802,15 +822,15 @@ class Parser
                        # that one of the bold ones was meant to be an apostrophe followed
                        # by italics. Which one we cannot know for certain, but it is more
                        # likely to be one that has a single-letter word before it.
-                       if (($numbold % 2 == 1) && ($numitalics % 2 == 1))
+                       if ( ( $numbold % 2 == 1 ) && ( $numitalics % 2 == 1 ) )
                        {
                                $i = 0;
                                $firstsingleletterword = -1;
                                $firstmultiletterword = -1;
                                $firstspace = -1;
-                               foreach ($arr as $r)
+                               foreach ( $arr as $r )
                                {
-                                       if (($i % 2 == 1) and (strlen ($r) == 3))
+                                       if ( ( $i % 2 == 1 ) and ( strlen( $r ) == 3 ) )
                                        {
                                                $x1 = substr ($arr[$i-1], -1);
                                                $x2 = substr ($arr[$i-1], -2, 1);
@@ -923,9 +943,8 @@ class Parser
        /**
         * Replace external links
         *
-        * Note: we have to do external links before the internal ones,
-        * and otherwise take great care in the order of things here, so
-        * that we don't end up interpreting some URLs twice.
+        * Note: this is all very hackish and the order of execution matters a lot.
+        * Make sure to run maintenance/parserTests.php if you change this code.
         *
         * @access private
         */
@@ -1093,10 +1112,9 @@ class Parser
 
        function replaceInternalLinks( $s ) {
                global $wgLang, $wgContLang, $wgLinkCache;
+               global $wgDisableLangConversion;
                static $fname = 'Parser::replaceInternalLinks' ;
-               # use a counter to prevent too much unknown links from
-               # being checked for different language variants.
-               static $convertCount;
+
                wfProfileIn( $fname );
 
                wfProfileIn( $fname.'-setup' );
@@ -1145,6 +1163,7 @@ class Parser
 
                wfProfileOut( $fname.'-setup' );
 
+               $checkVariantLink = sizeof($wgContLang->getVariants())>1;
                # Loop for each link
                for ($k = 0; isset( $a[$k] ); $k++) {
                        $line = $a[$k];
@@ -1184,7 +1203,7 @@ class Parser
                        # Don't allow internal links to pages containing
                        # PROTO: where PROTO is a valid URL protocol; these
                        # should be external links.
-                       if (preg_match('/((?:'.URL_PROTOCOLS.'):)/', $m[1])) {
+                       if (preg_match('/^((?:'.URL_PROTOCOLS.'):)/', $m[1])) {
                                $s .= $prefix . '[[' . $line ;
                                continue;
                        }
@@ -1204,29 +1223,11 @@ class Parser
                                continue;
                        }
 
-                       //check other language variants of the link
-                       //if the article does not exist
-                       global $wgContLang;
-                       $variants = $wgContLang->getVariants();
-
-                       if(sizeof($variants) > 1 && $convertCount < 200) {
-                               $varnt = false; 
-                               if($nt->getArticleID() == 0) {
-                                       foreach ( $variants as $v ) {
-                                               if($v == $wgContLang->getPreferredVariant())
-                                                       continue;
-                                               $convertCount ++;
-                                               $varlink = $wgContLang->autoConvert($link, $v);
-                                               $varnt = Title::newFromText($varlink);
-                                               if($varnt && $varnt->getArticleID()>0) {
-                                                       break;
-                                               }
-                                       }
-                               }
-                               if($varnt && $varnt->getArticleID()>0) {
-                                       $nt = $varnt;
-                                       $link = $varlink;
-                               }
+                       #check other language variants of the link
+                       #if the article does not exist
+                       if( $nt->getArticleID() == 0
+                               && $checkVariantLink ) {
+                               $wgContLang->findVariantLink($link, $nt);
                        }
 
                        $ns = $nt->getNamespace();
@@ -1331,7 +1332,7 @@ class Parser
 
                        # Special and Media are pseudo-namespaces; no pages actually exist in them
                        if( $ns == NS_MEDIA ) {
-                               $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text ) . $trail;
+                               $s .= $prefix . $sk->makeMediaLinkObj( $nt, $text, true ) . $trail;
                                $wgLinkCache->addImageLinkObj( $nt );
                                continue;
                        } elseif( $ns == NS_SPECIAL ) {
@@ -1756,19 +1757,16 @@ class Parser
                # This function is called recursively. To keep track of arguments we need a stack:
                array_push( $this->mArgStack, $args );
 
-               # PHP global rebinding syntax is a bit weird, need to use the GLOBALS array
-               $GLOBALS['wgCurParser'] =& $this;
-
                # Variable substitution
-               $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", 'wfVariableSubstitution', $text );
+               $text = preg_replace_callback( "/{{([$titleChars]*?)}}/", array( &$this, 'variableSubstitution' ), $text );
                
                if ( $this->mOutputType == OT_HTML || $this->mOutputType == OT_WIKI ) {
                        # Argument substitution
-                       $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", 'wfArgSubstitution', $text );
+                       $text = preg_replace_callback( "/{{{([$titleChars]*?)}}}/", array( &$this, 'argSubstitution' ), $text );
                }
                # Template substitution
                $regex = '/(\\n|{)?{{(['.$titleChars.']*)(\\|.*?|)}}/s';
-               $text = preg_replace_callback( $regex, 'wfBraceSubstitution', $text );
+               $text = preg_replace_callback( $regex, array( &$this, 'braceSubstitution' ), $text );
 
                array_pop( $this->mArgStack );
 
@@ -2481,7 +2479,7 @@ class Parser
                                if( $istemplate )
                                        $headline = $sk->editSectionScriptForOther($templatetitle, $templatesection, $headline);
                                else
-                                       $headline = $sk->editSectionScript($this->title, $sectionCount+1,$headline);
+                                       $headline = $sk->editSectionScript($this->mTitle, $sectionCount+1,$headline);
                        }
 
                        # give headline the correct <h#> tag
@@ -2635,30 +2633,34 @@ class Parser
         * @access private
         * @param string $text text to be processed
         */
-       function magicRFC( $text ) {
+       function magicRFC( $text, $keyword='RFC ', $urlmsg='rfcurl'  ) {
                global $wgLang;
                
                $valid = '0123456789';
                $internal = false;
 
-               $a = split( 'RFC ', ' '.$text );
-               if ( count ( $a ) < 2 ) return $text;
+               $a = split( $keyword, ' '.$text );
+               if ( count ( $a ) < 2 ) {
+                       return $text;
+               }
                $text = substr( array_shift( $a ), 1);
                
-               /* Check if RFC keyword is preceed by [[.
+               /* Check if keyword is preceed by [[.
                 * This test is made here cause of the array_shift above
                 * that prevent the test to be done in the foreach.
                 */
-               if(substr($text, -2) == '[[') { $internal = true; }
+               if ( substr( $text, -2 ) == '[[' ) {
+                       $internal = true;
+               }
 
                foreach ( $a as $x ) {
                        /* token might be empty if we have RFC RFC 1234 */
-                       if($x=='') {
-                               $text.='RFC ';
+                       if ( $x=='' ) {
+                               $text.=$keyword;
                                continue;
                                }
 
-                       $rfc = $blank = '' ;
+                       $id = $blank = '' ;
 
                        /** remove and save whitespaces in $blank */
                        while ( $x{0} == ' ' ) {
@@ -2666,29 +2668,29 @@ class Parser
                                $x = substr( $x, 1 );
                        }
 
-                       /** remove and save the rfc number in $rfc */
+                       /** remove and save the rfc number in $id */
                        while ( strstr( $valid, $x{0} ) != false ) {
-                               $rfc .= $x{0};
+                               $id .= $x{0};
                                $x = substr( $x, 1 );
                        }
 
-                       if ( $rfc == '') {
+                       if ( $id == '' ) {
                                /* call back stripped spaces*/
-                               $text .= "RFC $blank$x";
-                       } elseif( $internal) {
+                               $text .= $keyword.$blank.$x;
+                       } elseif( $internal ) {
                                /* normal link */
-                               $text .= "RFC $rfc$x";
+                               $text .= $keyword.$id.$x;
                        } else {
                                /* build the external link*/
-                               $url = wfmsg( 'rfcurl' );
-                               $url = str_replace( '$1', $rfc, $url);
+                               $url = wfmsg( $urlmsg );
+                               $url = str_replace( '$1', $id, $url);
                                $sk =& $this->mOptions->getSkin();
-                               $la = $sk->getExternalLinkAttributes( $url, 'RFC '.$rfc );
-                               $text .= "<a href='{$url}'{$la}>RFC {$rfc}</a>{$x}";
+                               $la = $sk->getExternalLinkAttributes( $url, $keyword.$id );
+                               $text .= "<a href='{$url}'{$la}>{$keyword}{$id}</a>{$x}";
                        }
                        
                        /* Check if the next RFC keyword is preceed by [[ */
-                       $internal = (substr($x,-2) == '[[');
+                       $internal = ( substr($x,-2) == '[[' );
                }
                return $text;
        }
@@ -2718,15 +2720,7 @@ class Parser
                $pairs = array(
                        "\r\n" => "\n",
                        );
-               $text = str_replace(array_keys($pairs), array_values($pairs), $text);
-               // now with regexes
-               /*
-               $pairs = array(
-                       "/<br.+(clear|break)=[\"']?(all|both)[\"']?\\/?>/i" => '<br style="clear:both;"/>',
-                       "/<br *?>/i" => "<br />",
-               );
-               $text = preg_replace(array_keys($pairs), array_values($pairs), $text);
-               */
+               $text = str_replace( array_keys( $pairs ), array_values( $pairs ), $text );
                $text = $this->strip( $text, $stripState, false );
                $text = $this->pstPass2( $text, $user );
                $text = $this->unstrip( $text, $stripState );
@@ -2739,7 +2733,7 @@ class Parser
         * @access private
         */
        function pstPass2( $text, &$user ) {
-               global $wgLang, $wgContLang, $wgLocaltimezone, $wgCurParser;
+               global $wgLang, $wgContLang, $wgLocaltimezone;
 
                # Variable replacement
                # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
@@ -2750,15 +2744,18 @@ class Parser
                $n = $user->getName();
                $k = $user->getOption( 'nickname' );
                if ( '' == $k ) { $k = $n; }
-               if(isset($wgLocaltimezone)) {
-                       $oldtz = getenv('TZ'); putenv('TZ='.$wgLocaltimezone);
+               if ( isset( $wgLocaltimezone ) ) {
+                       $oldtz = getenv( 'TZ' );
+                       putenv( 'TZ='.$wgLocaltimezone );
                }
                /* Note: this is an ugly timezone hack for the European wikis */
                $d = $wgContLang->timeanddate( date( 'YmdHis' ), false ) .
                  ' (' . date( 'T' ) . ')';
-               if(isset($wgLocaltimezone)) putenv('TZ='.$oldtzs);
+               if ( isset( $wgLocaltimezone ) ) {
+                       putenv( 'TZ='.$oldtzs );
+               }
 
-               $text = preg_replace( '/~~~~~/', $d, $text );
+               $text = preg_replace( '/~~~~~~/', $d, $text );
                $text = preg_replace( '/~~~~/', '[[' . $wgContLang->getNsText( NS_USER ) . ":$n|$k]] $d", $text );
                $text = preg_replace( '/~~~/', '[[' . $wgContLang->getNsText( NS_USER ) . ":$n|$k]]", $text );
 
@@ -2863,6 +2860,8 @@ class Parser
         */
        function replaceLinkHolders( &$text, $options = 0 ) {
                global $wgUser, $wgLinkCache, $wgUseOldExistenceCheck, $wgLinkHolders;
+               global $wgInterwikiLinkHolders;
+               global $outputReplace;
                
                if ( $wgUseOldExistenceCheck ) {
                        return array();
@@ -2955,7 +2954,6 @@ class Parser
                        
                        # Construct search and replace arrays
                        wfProfileIn( $fname.'-construct' );
-                       global $outputReplace;
                        $outputReplace = array();
                        foreach ( $wgLinkHolders['namespaces'] as $key => $ns ) {
                                $pdbk = $pdbks[$key];
@@ -2987,20 +2985,57 @@ class Parser
                                "outputReplaceMatches",
                                $text);
                        wfProfileOut( $fname.'-replace' );
+               }
 
+               if ( !empty( $wgInterwikiLinkHolders ) ) {
                        wfProfileIn( $fname.'-interwiki' );
-                       global $wgInterwikiLinkHolders;
                        $outputReplace = $wgInterwikiLinkHolders;
                        $text = preg_replace_callback(
                                '/<!--IWLINK (.*?)-->/',
                                "outputReplaceMatches",
-                               $text);
+                               $text );
                        wfProfileOut( $fname.'-interwiki' );
                }
 
                wfProfileOut( $fname );
                return $colours;
        }
+
+       /**
+        * Renders an image gallery from a text with one line per image.
+        * text labels may be given by using |-style alternative text. E.g.
+        *   Image:one.jpg|The number "1"
+        *   Image:tree.jpg|A tree
+        * given as text will return the HTML of a gallery with two images,
+        * labeled 'The number "1"' and
+        * 'A tree'.
+        */
+       function renderImageGallery( $text ) {
+               global $wgLinkCache;
+               $ig = new ImageGallery();
+               $ig->setShowBytes( false );
+               $ig->setShowFilename( false );
+               $lines = explode( "\n", $text );
+
+               foreach ( $lines as $line ) {
+                       # match lines like these:
+                       # Image:someimage.jpg|This is some image
+                       preg_match( "/^([^|]+)(\\|(.*))?$/", $line, $matches );
+                       # Skip empty lines
+                       if ( count( $matches ) == 0 ) {
+                               continue;
+                       }
+                       $nt = Title::newFromURL( $matches[1] );
+                       if ( isset( $matches[3] ) ) {
+                               $label = $matches[3];
+                       } else {
+                               $label = '';
+                       }
+                       $ig->add( Image::newFromTitle( $nt ), $label );
+                       $wgLinkCache->addImageLinkObj( $nt );
+               }
+               return $ig->toHTML();
+       }
 }
 
 /**
@@ -3022,16 +3057,16 @@ class ParserOutput
                $this->mCacheTime = '';
        }
 
-       function getText() { return $this->mText; }
-       function getLanguageLinks() { return $this->mLanguageLinks; }
-       function getCategoryLinks() { return $this->mCategoryLinks; }
-       function getCacheTime() { return $this->mCacheTime; }
-       function containsOldMagic() { return $this->mContainsOldMagic; }
-       function setText( $text ) { return wfSetVar( $this->mText, $text ); }
-       function setLanguageLinks( $ll ) { return wfSetVar( $this->mLanguageLinks, $ll ); }
-       function setCategoryLinks( $cl ) { return wfSetVar( $this->mCategoryLinks, $cl ); }
+       function getText()                   { return $this->mText; }
+       function getLanguageLinks()          { return $this->mLanguageLinks; }
+       function getCategoryLinks()          { return $this->mCategoryLinks; }
+       function getCacheTime()              { return $this->mCacheTime; }
+       function containsOldMagic()          { return $this->mContainsOldMagic; }
+       function setText( $text )            { return wfSetVar( $this->mText, $text ); }
+       function setLanguageLinks( $ll )     { return wfSetVar( $this->mLanguageLinks, $ll ); }
+       function setCategoryLinks( $cl )     { return wfSetVar( $this->mCategoryLinks, $cl ); }
        function setContainsOldMagic( $com ) { return wfSetVar( $this->mContainsOldMagic, $com ); }
-       function setCacheTime( $t ) { return wfSetVar( $this->mCacheTime, $t ); }
+       function setCacheTime( $t )          { return wfSetVar( $this->mCacheTime, $t ); }
 
        function merge( $other ) {
                $this->mLanguageLinks = array_merge( $this->mLanguageLinks, $other->mLanguageLinks );
@@ -3125,28 +3160,11 @@ class ParserOptions
  * Callback function used by Parser::replaceLinkHolders()
  * to substitute link placeholders.
  */
-function &outputReplaceMatches($matches) {
+function &outputReplaceMatches( $matches ) {
        global $outputReplace;
        return $outputReplace[$matches[1]];
 }
 
-
-# Regex callbacks, used in Parser::replaceVariables
-function wfBraceSubstitution( $matches ) {
-       global $wgCurParser;
-       return $wgCurParser->braceSubstitution( $matches );
-}
-
-function wfArgSubstitution( $matches ) {
-       global $wgCurParser;
-       return $wgCurParser->argSubstitution( $matches );
-}
-
-function wfVariableSubstitution( $matches ) {
-       global $wgCurParser;
-       return $wgCurParser->variableSubstitution( $matches );
-}
-
 /**
  * Return the total number of articles
  */
@@ -3167,7 +3185,7 @@ function wfLoadSiteStats() {
 
        if ( -1 != $wgNumberOfArticles ) return;
        $dbr =& wfGetDB( DB_SLAVE );
-       $s = $dbr->getArray( 'site_stats',
+       $s = $dbr->selectRow( 'site_stats',
                array( 'ss_total_views', 'ss_total_edits', 'ss_good_articles' ),
                array( 'ss_row_id' => 1 ), $fname
        );