Move the above named methods from WebCore::Frame to WebCore::Chrome.
Created attachment 26859 [details] Carry out the move. The statusbar methods were left in Frame. They require more research, e.g. if the per frame statusbar text needs to be safed.
Comment on attachment 26859 [details] Carry out the move. Looks fine, long live refactoring, r=me :-)
Created attachment 27496 [details] Carry out the move Add null check for Frame::page in the FrameLoaderClient. shouldClose will be true when we have no page or chrome()->shouldClose will return true. This should resemble the current state with Frame::shouldClose.
Created attachment 27497 [details] Move stuff to Chrome.. Unmangled version?
Comment on attachment 27497 [details] Move stuff to Chrome.. My intent here for these functions was to have the top-frame check be in DOMWindow, and the functions in Page not take a Frame*. I don't think it makes sense to "close" a non-main Frame. Could you make that change?
(In reply to comment #5) > (From update of attachment 27497 [details] [review]) > My intent here for these functions was to have the top-frame check be in > DOMWindow, and the functions in Page not take a Frame*. I don't think it makes > sense to "close" a non-main Frame. Could you make that change? > Sure.
Comment on attachment 27497 [details] Move stuff to Chrome.. > + Like the many other Chrome methods take a Frame* pointer. Update the > + FrameLoader and the DOMWindow call site. In the FrameLoader it is assumed > + that the page is around... This last sentence doesn't seem necessary anymore -- there's no code in FrameLoader in this patch that assumes there's a Page. You need to patch WebView::shouldClose in WebKit/win/WebView.cpp as well. r=me
Comment on attachment 27497 [details] Move stuff to Chrome.. I'm going to change this back to review- since I've requested a change.
Comment on attachment 26859 [details] Carry out the move. Clearing review flag to remove it from the commit queue.