Version: (using KDE KDE 3.4.2) Installed from: SuSE RPMs In an XML document created via XMLLoad "<a><b>first</b><b> </b></a>" should result in something like: a b #text "first" b #text " " but the second text node is nonexistant. I think the bug is in http://websvn.kde.org/trunk/KDE/kdelibs/khtml/xml/xml_tokenizer.cpp XMLHandler::characters(...) A small test can be found under: http://www.kawo1.rwth-aachen.de/~lurchi/js_dom_test/ws_only_node.html
SVN commit 514945 by orlovich: Remove hack that swallowed lots of text nodes, we shouldn't do this for XML BUG:110426 M +0 -5 xml_tokenizer.cpp --- branches/KDE/3.5/kdelibs/khtml/xml/xml_tokenizer.cpp #514944:514945 @@ -232,11 +232,6 @@ bool XMLHandler::characters( const QString& ch ) { - //this is needed for xhtml parsing. otherwise we try to attach - //"\n\t" to html, head and other nodes which don't accept textchildren - if ( ch.stripWhiteSpace().isEmpty() ) - return true; - if (currentNode()->nodeType() == Node::TEXT_NODE || currentNode()->nodeType() == Node::CDATA_SECTION_NODE || enterText()) {