If we import a bunch of w3c reftests in the future, it sounds reasonable and unavoidable to use manifest file, assuming the manifest file is auto-generated by w3c's build process.
But I'd like to leave an option to developers to write reftests more casually without worrying about maintaing the manifest file. At the same time, I don't want to increase the number of ways to write/run tests anymore. So that would be great that we have the best of both worlds somehow. On Fri, Nov 4, 2011 at 1:39 PM, Ryosuke Niwa <rn...@webkit.org> wrote: > On Fri, Nov 4, 2011 at 1:32 PM, Hayato Ito <hay...@chromium.org> wrote: > >> Could you clarify why we have to need to modify DRT if we have Link >> Element approach? >> I guess one of the reasons is the performance. It might be better that we >> use DRT to parse and extract reference links from HTML since parsing HTML >> using Python might take unacceptable time and might be inaccurate. >> Is there any other reasons we should modify DRT to support Link Element >> approach? >> > > That's the primary reason. I can see that there are more than 100,000 .xht > files just in css2.1 test suite, and the number is growing. If we include > css3, and other w3c test suites, we'll end up having tens of thousands of > tests, if not millions. > > In addition, link element approach doesn't scale well in that there's no > mandate as to how reference files are named, so we'll end up needlessly > parsing all reference files as well depending on the order they appear in > the directory. > > - Ryosuke > > -- Hayato
_______________________________________________ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev