2010-07-01 Adam Barth <abarth@webkit.org>
authorabarth@webkit.org <abarth@webkit.org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Thu, 1 Jul 2010 07:39:52 +0000 (07:39 +0000)
committerabarth@webkit.org <abarth@webkit.org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Thu, 1 Jul 2010 07:39:52 +0000 (07:39 +0000)
commitc1f1ba66d472a953c974e6e3956e05dc51b83f83
tree841022b5338c93c98b37c6d2b27c1dcb24fd0111
parent85ff4ecd486d1916d436074ab5b91390ef91785e
2010-07-01  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        BogusCommentState should come in from the cold
        https://bugs.webkit.org/show_bug.cgi?id=41439

        Test progression => :)

        * html5lib/runner-expected.txt:
        * html5lib/webkit-resumer-expected.txt:
2010-07-01  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        BogusCommentState should come in from the cold
        https://bugs.webkit.org/show_bug.cgi?id=41439

        The BogusCommentState has always been wrong.  The proximate issue is
        that it didn't handle resuming correctly when parsing a partial input
        stream.  Now that we have EOF working properly, we can actually
        implement this state correctly.

        We need to distinguish when we enter this state from when we continue
        in this state.  We could do that with a branch for each character, but
        it seemed easier to split the state in two, even though that leaves us
        with one more state in our tokenizer than we have in the HTML5 spec.

        * html/HTMLTokenizer.cpp:
        (WebCore::HTMLTokenizer::nextToken):
        * html/HTMLTokenizer.h:
        (WebCore::HTMLTokenizer::):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@62233 268f45cc-cd09-0410-ab3c-d52691b4dbfc
LayoutTests/ChangeLog
LayoutTests/html5lib/runner-expected.txt
LayoutTests/html5lib/webkit-resumer-expected.txt
WebCore/ChangeLog
WebCore/html/HTMLTokenizer.cpp
WebCore/html/HTMLTokenizer.h