[webkit-reviews] review denied: [Bug 14757] REGRESSION: HTMLTokenizer::processingData implementation is incorrect : [Attachment 15709] patch with assertion fix

[bugzilla-daemon at webkit.org]

Brady Eidson <beidson at apple.com> has denied Matt Perry <mpComplete at gmail.com>'s
request for review:
Bug 14757: REGRESSION: HTMLTokenizer::processingData implementation is
incorrect
http://bugs.webkit.org/show_bug.cgi?id=14757

Attachment 15709: patch with assertion fix
http://bugs.webkit.org/attachment.cgi?id=15709&action=edit

------- Additional Comments from Brady Eidson <beidson at apple.com>
I think this fix is pretty darned good.  
r- for one issue.
You replace the stopForUserCancel() call with a stopAllLoaders() and
scheduleCheckLoadComplete() call.
It is not obvious to me that those two lines replace the behavior of
"stopForUserCancell()" but just do it asynchronously.
I think it'd be better for future maintainability if stopForUserCancel() had an
asynchronous flag added, and that method decides whether to do the load
complete check now, or later.