From 84041c20999a5bc32bfdd13612543fff0d738c97 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bartosz=20Dziewo=C5=84ski?= Date: Tue, 1 Dec 2015 21:11:43 +0100 Subject: [PATCH] Attribute-aware StripState Change-Id: Id37d6e8d60bab9d787cc6dcb7606ce49f51a1c83 --- includes/parser/Parser.php | 4 ++++ includes/parser/StripState.php | 28 ++++++++++++++++++++++++++++ 2 files changed, 32 insertions(+) diff --git a/includes/parser/Parser.php b/includes/parser/Parser.php index cfbf0b4..6d4f7f5 100644 --- a/includes/parser/Parser.php +++ b/includes/parser/Parser.php @@ -1288,6 +1288,8 @@ class Parser { * @return string */ private function internalParseHalfParsed( $text, $isMain = true, $linestart = true ) { + $this->mStripState->attrAware = true; + $text = $this->mStripState->unstripGeneral( $text ); if ( $isMain ) { @@ -1372,6 +1374,8 @@ class Parser { if ( $isMain ) { Hooks::run( 'ParserAfterTidy', array( &$this, &$text ) ); } + + $this->mStripState->attrAware = false; return $text; } diff --git a/includes/parser/StripState.php b/includes/parser/StripState.php index b11dc8c..0859cd2 100644 --- a/includes/parser/StripState.php +++ b/includes/parser/StripState.php @@ -33,6 +33,8 @@ class StripState { protected $tempType, $tempMergePrefix; protected $circularRefGuard; protected $recursionLevel = 0; + + public $attrAware = false; const UNSTRIP_RECURSION_LIMIT = 20; @@ -121,10 +123,36 @@ class StripState { if ( !count( $this->data[$type] ) ) { return $text; } + if ( !$text ) { + return $text; + } $oldType = $this->tempType; $this->tempType = $type; + + // var_dump($text); + + if ( $this->attrAware ) { + // Step 1. Replace strip markers inside attributes, treating their values as plain text. + $dom = new DomDocument; + $dom->loadHTML( "$text" ); + foreach ($dom->childNodes as $item) { + if ($item->nodeType == XML_PI_NODE) { + $dom->removeChild($item); // remove hack + } + } + $dom->encoding = 'UTF-8'; // insert proper + + $xpath = new DOMXPath( $dom ); + foreach ( $xpath->query( '//@*' ) as $attrNode ) { + $attrNode->value = preg_replace_callback( $this->regex, array( $this, 'unstripCallback' ), $attrNode->value ); + } + $text = preg_replace('/^/', '', str_replace( array('', '', '', ''), array('', '', '', ''), $dom->saveHTML())); + } + + // Step 2. Replace strip markers inside tag content, treating their values as HTML. $text = preg_replace_callback( $this->regex, array( $this, 'unstripCallback' ), $text ); + $this->tempType = $oldType; return $text; } -- 1.9.5.msysgit.0