This will make it easy to merge asynchronous path and synchronous path of spellcheck.
Created attachment 118325 [details]
Comment on attachment 118325 [details]
View in context: https://bugs.webkit.org/attachment.cgi?id=118325&action=review
Basically looks fine. let's have one iteration.
> + request->setSequence(++m_lastRequestSequence);
Let us make sure that published id never conflict against the magic number.
I know this will never happen in reality, but preventing it explicitly gives us more confidence.
> + Element* m_rootEditableElement;
Let's make this RefPtr to prevent possible dangling pointer access.
> + static PassRefPtr<SpellCheckRequest> createRequest(TextCheckingTypeMask, PassRefPtr<Range> checkingRange, PassRefPtr<Range> paragraphRange);
SpellCheckerRequest::create() is our standard way if there is no special reason.
Created attachment 118340 [details]
Comment on attachment 118340 [details]
Clearing flags on attachment: 118340
Committed r102328: <http://trac.webkit.org/changeset/102328>
All reviewed patches have been landed. Closing bug.