Wikia code/includes/parser/Preprocessor DOM.php
From MediaWiki.org
< Wikia code | includes | parser
--- D:\Programming\SVN\mediawiki\branches\REL1_16\phase3\includes\parser\Preprocessor_DOM.php 2011-07-18 22:30:54.902343800 +0100 +++ D:\Programming\SVN\wikia\trunk\includes\parser\Preprocessor_DOM.php 2011-08-17 15:28:13.351562500 +0100 @@ -69,6 +69,21 @@ $xml = false; $cacheable = strlen( $text ) > $wgPreprocessorCacheThreshold; + + global $wgCategorySelectEnabled; + if(!empty($wgCategorySelectEnabled)) { + $cacheable = false; + } + + # RTE (Rich Text Editor) - begin + # @author: Inez Korczyński + # Disable preprocessor cache for RTE mode + global $wgRTETemplateParams; + if(!empty($wgRTETemplateParams) || !empty($wgRTEParserEnabled)) { + $cacheable = false; + } + # RTE - end + if ( $cacheable ) { wfProfileIn( __METHOD__.'-cacheable' ); @@ -119,6 +134,7 @@ } function preprocessToXml( $text, $flags = 0 ) { + global $wgCategorySelectEnabled; wfProfileIn( __METHOD__ ); $rules = array( '{' => array( @@ -154,6 +170,33 @@ $ignoredElements = array( 'includeonly' ); $xmlishElements[] = 'includeonly'; } + + # RTE (Rich Text Editor) - begin + # @author: Inez Korczyński + global $wgRTEParserEnabled, $wgRTETemplateParams; + if(!empty($wgRTEParserEnabled)) { + $rules['['] = array( + 'end' => ']', + 'names' => array( 1=> 'external', 2 => null ), + 'min' => 1, + 'max' => 2 + ); + $RTE_flags = $flags; + $ignoredTags = $ignoredElements = array(); + } + if(!empty($wgRTETemplateParams)) { + $rules['{']['min'] = 3; + $rules['{']['names'] = array(3 => 'tplarg'); + } + # RTE - end + + //CategorySelect + if (!empty($wgCategorySelectEnabled)) { + $ignoredTags = $ignoredElements = array(); + $xmlishElements[] = 'noinclude'; + $xmlishElements[] = 'onlyinclude'; + } + $xmlishRegex = implode( '|', array_merge( $xmlishElements, $ignoredTags ) ); // Use "A" modifier (anchored) instead of "^", because ^ doesn't work with an offset @@ -174,6 +217,12 @@ $noMoreGT = false; # True if there are no more greater-than (>) signs right of $i $findOnlyinclude = $enableOnlyinclude; # True to ignore all input up to the next <onlyinclude> $fakeLineStart = true; # Do a line-start run without outputting an LF character + $openAt = $closeAt = array(); # CategorySelect + + # RTE (Rich Text Editor) - begin + # @author: Inez Korczyński + $openAt = $closeAt = array(); + # RTE - end while ( true ) { //$this->memCheck(); @@ -389,6 +438,15 @@ // Note that the attr element contains the whitespace between name and attribute, // this is necessary for precise reconstruction during pre-save transform. '<attr>' . htmlspecialchars( $attr ) . '</attr>'; + + # RTE (Rich Text Editor) - begin + # @author: Inez Korczyński + if(!empty($wgRTEParserEnabled)) { + $accum .= '<inner>' . RTEMarker::generate(RTEMarker::EXT_WIKITEXT, RTEData::put('wikitext', substr( $text, $tagStartPos, $i - $tagStartPos ))) . '</inner>'; + $inner = null; + } + # RTE - end + if ( $inner !== null ) { $accum .= '<inner>' . htmlspecialchars( $inner ) . '</inner>'; } @@ -505,7 +563,19 @@ # Add literal brace(s) $accum .= htmlspecialchars( str_repeat( $curChar, $count ) ); } + + # Wysiwyg + if($flags == 0 && ($wgCategorySelectEnabled) && $count == 2 && $curChar == "{") { + $openAt[] = $i; + } + $i += $count; + # RTE (Rich Text Editor) - begin + # @author: Inez Korczyński + if(!empty($wgRTEParserEnabled) && $RTE_flags === 0 && ($count == 2 || $count == 3) && $curChar == "{") { + $openAt[] = $i; + } + # RTE - end } elseif ( $found == 'close' ) { @@ -540,9 +610,22 @@ continue; } $name = $rule['names'][$matchingCount]; - if ( $name === null ) { + # RTE (Rich Text Editor) - begin + # @author: Inez Korczyński + if ( $name === null || $name == 'external' || (!empty($wgRTEParserEnabled) && $name == 'external')) { // No element, just literal text $element = $piece->breakSyntax( $matchingCount ) . str_repeat( $rule['end'], $matchingCount ); + + if(!empty($wgRTEParserEnabled)) { + if($name === null) { + $dataIdx = RTEData::put('wikitext', $element); + $element = '[[' . RTEMarker::generate(RTEMarker::INTERNAL_WIKITEXT, $dataIdx) . substr($element, 2); + } else { + $dataIdx = RTEData::put('wikitext', $element); + $element .= RTEMarker::generate(RTEMarker::EXTERNAL_WIKITEXT, $dataIdx); + } + } + # RTE - end } else { # Create XML element # Note: $parts is already XML, does not need to be encoded further @@ -558,8 +641,22 @@ $attr = ''; } + # RTE (Rich Text Editor) - begin + # @author: Inez Korczyński + if(!empty($wgRTEParserEnabled) && $RTE_flags === 0 && ($count == 2 || $count == 3) && $curChar == '}') { + $closeAt[] = $i; + if(count($closeAt) == count($openAt)) { + $openIdx = $openAt[0]; + $closeIdx = $closeAt[count($closeAt)-1]; + $openAt = $closeAt = array(); + $attr .= ' _rte_wikitextidx="'.RTEData::put('wikitext', substr($text, $openIdx-$count, $closeIdx-$openIdx+2*$count)).'"'; + } + } + # RTE - end + $element = "<$name$attr>"; $element .= "<title>$title</title>"; + $argIndex = 1; foreach ( $parts as $partIndex => $part ) { if ( isset( $part->eqpos ) ) { @@ -899,7 +996,16 @@ $iteratorStack = array( false, $root ); $indexStack = array( 0, 0 ); + $RTEext_1 = false; + $RTEext_2 = false; + while ( count( $iteratorStack ) > 1 ) { + + if($RTEext_1) { + $RTEext_1 = false; + $RTEext_2 = true; + } + $level = count( $outStack ) - 1; $iteratorNode =& $iteratorStack[ $level ]; $out =& $outStack[$level]; @@ -944,8 +1050,21 @@ $newIterator = $contextNode; } elseif ( $contextNode instanceof DOMNode ) { if ( $contextNode->nodeType == XML_TEXT_NODE ) { + + # RTE (Rich Text Editor) - begin + global $wgRTEParserEnabled; + if(!empty($wgRTEParserEnabled)) { + if($RTEext_2) { + if(strpos($contextNode->nodeValue, 'table') !== false) { + RTE::$edgeCases[] = 'COMPLEX.11'; + } + } + } + # RTE - end + $out .= $contextNode->nodeValue; } elseif ( $contextNode->nodeName == 'template' ) { + # Double-brace expansion $xpath = new DOMXPath( $contextNode->ownerDocument ); $titles = $xpath->query( 'title', $contextNode ); @@ -951,6 +1070,18 @@ $titles = $xpath->query( 'title', $contextNode ); $title = $titles->item( 0 ); $parts = $xpath->query( 'part', $contextNode ); + + # RTE (Rich Text Editor) - begin + # @author: Inez Korczyński + global $wgRTEParserEnabled; + if(!empty($wgRTEParserEnabled)) { + $dataIdx = RTEData::put('placeholder', array( + 'type' => 'double-brackets', + 'wikitextIdx' => $contextNode->getAttribute('_rte_wikitextidx'), + 'lineStart' => $contextNode->getAttribute('lineStart'), + 'title' => $title->textContent)); + $out .= RTEMarker::generate(RTEMarker::PLACEHOLDER, $dataIdx); + } else { if ( $flags & self::NO_TEMPLATES ) { $newIterator = $this->virtualBracketedImplode( '{{', '|', '}}', $title, $parts ); } else { @@ -966,12 +1097,28 @@ $out .= $ret['text']; } } + } + # RTE - end + } elseif ( $contextNode->nodeName == 'tplarg' ) { # Triple-brace expansion $xpath = new DOMXPath( $contextNode->ownerDocument ); $titles = $xpath->query( 'title', $contextNode ); $title = $titles->item( 0 ); $parts = $xpath->query( 'part', $contextNode ); + + # RTE (Rich Text Editor) - begin + # @author: Wladyslaw Bodzek + global $wgRTEParserEnabled; + if ( !empty($wgRTEParserEnabled) ) { + //var_dump($contextNode->getAttribute('_rte_wikitextidx')); + $dataIdx = RTEData::put('placeholder', array( + 'type' => 'tplarg', + 'wikitextIdx' => $contextNode->getAttribute('_rte_wikitextidx'), + 'lineStart' => $contextNode->getAttribute('lineStart'), + 'title' => $title->textContent)); + $out .= RTEMarker::generate(RTEMarker::PLACEHOLDER, $dataIdx); + } else { if ( $flags & self::NO_ARGS ) { $newIterator = $this->virtualBracketedImplode( '{{{', '|', '}}}', $title, $parts ); } else { @@ -985,6 +1132,8 @@ $out .= $ret['text']; } } + } + # RTE - end } elseif ( $contextNode->nodeName == 'comment' ) { # HTML-style comment # Remove it in HTML, pre+remove and STRIP_COMMENTS modes @@ -992,8 +1141,32 @@ || ( $this->parser->ot['pre'] && $this->parser->mOptions->getRemoveComments() ) || ( $flags & self::STRIP_COMMENTS ) ) { + # RTE (Rich Text Editor) - begin + # @author: Inez Korczyński + global $wgRTEParserEnabled; + if(!empty($wgRTEParserEnabled)) { + if(strlen($out) === 0 || substr($out, -1) == "\n") { + if(substr($contextNode->textContent, -1) == "\n") { + $add = "\n"; + $text = substr($contextNode->textContent, 0, -1); + } else { + $add = ""; + $text = $contextNode->textContent; + } + $dataIdx = RTEData::put('placeholder', array( + 'type' => 'comment', + 'wikitext' => $text)); + $out .= RTEMarker::generate(RTEMarker::PLACEHOLDER, $dataIdx) . $add; + } else { + RTE::$edgeCases[] = 'COMMENT'; + $out .= ''; + } + } else { $out .= ''; } + # RTE - end + + } # Add a strip marker in PST mode so that pstPass2() can run some old-fashioned regexes on the result # Not in RECOVER_COMMENTS mode (extractSections) though elseif ( $this->parser->ot['wiki'] && ! ( $flags & self::RECOVER_COMMENTS ) ) { @@ -1027,6 +1200,7 @@ 'close' => $closes->length > 0 ? new PPNode_DOM( $closes->item( 0 ) ) : null, ); $out .= $this->parser->extensionSubstitution( $params, $this ); + $RTEext_1 = true; } elseif ( $contextNode->nodeName == 'h' ) { # Heading $s = $this->expand( $contextNode->childNodes, $flags ); @@ -1074,6 +1248,7 @@ $level--; } } + $RTEext_2 = false; } --$expansionDepth; wfProfileOut( __METHOD__ );
