bruns added a comment.
In D16488#351584 <https://phabricator.kde.org/D16488#351584>, @astippich wrote: > I guess you plan to convert the existing extractors to use this? Yes, thats the plan > The only concern I might have that now two iterations over the node are required when this is used in e.g. odfextractor, isn't it? The QDomDocument can be reused, so parsing is only done once. Iteration needs to be done twice, but this is only a minor issue as the metadata documents are **really** small (~10 nodes, 1kByte). For the office2007extractor it actually is a win, as it currently iterates the tree once per property. REPOSITORY R286 KFileMetaData REVISION DETAIL https://phabricator.kde.org/D16488 To: bruns, #frameworks, astippich Cc: kde-frameworks-devel, #baloo, ashaposhnikov, michaelh, astippich, spoorun, ngraham, bruns, abrahams