Eliminate Node::dispatchGenericEvent.

[Dimitri Glazkov (Google)]

Eliminate Node::dispatchGenericEvent.

[Dimitri Glazkov (Google)]

Created attachment 86814 [details]
Patch

[Dimitri Glazkov (Google)]

Created attachment 86816 [details]
Removed unnecessary extra ref and stale comment.

[Dimitri Glazkov (Google)]

Comment on attachment 86816 [details]
Removed unnecessary extra ref and stale comment.

commit queue should do it, for safety.

[Darin Adler]

Comment on attachment 86816 [details]
Removed unnecessary extra ref and stale comment.

View in context: https://bugs.webkit.org/attachment.cgi?id=86816&action=review

Wow, this is a really great thing to do. Who on earth every understood the difference between dispatchGenericEvent and dispatchEvent!?

> Source/WebCore/dom/Node.cpp:2699
> +    RefPtr<FrameView> view = document()->view();

Since this is never used, I think this should be named protectView instead of just view. I also think it makes sense to put this with the other protect before doing any work with the event.

> Source/WebCore/svg/SVGElementInstance.cpp:142
>      SVGElement* element = shadowTreeElement();
>      if (!element)
>          return false;
>  
> -    RefPtr<FrameView> view = element->document()->view();
> -    return element->dispatchGenericEvent(event.release());
> +    return element->dispatchEvent(event);

I can’t resist pointing out the way I like to write these:

    SVGElement* element = shadowTreeElement();
    return element && element->dispatchEvent(event);

But I will set commit-queue+ anyway, because it probably is clearer the way it’s written here anyway.

[WebKit Commit Bot]

Comment on attachment 86816 [details]
Removed unnecessary extra ref and stale comment.

Clearing flags on attachment: 86816

Committed r81932: <http://trac.webkit.org/changeset/81932>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.