Bug 70491 – Add mechanisms to notify WebCore of showing and dismissing context menu

• Status: NEW
• Component: WebKit Misc.

[chandra shekar vallala]

Port specific changes for bug https://bugs.webkit.org/show_bug.cgi?id=69137

[Ryosuke Niwa]

Sorry I was probably not clear. What I meant is that you need to file blockers and fix them first, not the other way around.

[chandra shekar vallala]

Created attachment 111964 [details]
patch for WebCore changes.

Updated the WebCore changes ..

[Ryosuke Niwa]

Comment on attachment 111964 [details]
patch for WebCore changes.

View in context: https://bugs.webkit.org/attachment.cgi?id=111964&action=review

> Source/WebCore/page/EventHandler.cpp:2264
> +    m_frame->selection()->setCaretBlinkingSuspended(true);

This patch shouldn't be adding a bug fix. Instead, you should add an empty stub functions that are called by each port.

[chandra shekar vallala]

Created attachment 112334 [details]
updated patch with chromium + gtk port changes.

Patch includes changes for chromium and gtk port changes to notify EventHandler about context menu show & hide actions.
I don't have the infrastructure at my side to make the changes for other ports, so it would be great if someone can add the patches for those ports in the bug.

[WebKit Review Bot]

Please wait for approval from [e-mail] before submitting because this patch contains changes to the Chromium public API.

[Ryosuke Niwa]

Comment on attachment 112334 [details]
updated patch with chromium + gtk port changes.

View in context: https://bugs.webkit.org/attachment.cgi?id=112334&action=review

I'd say r- due to various nits. How about win, mac, qt, etc...?

> Source/WebCore/ChangeLog:7
> +        Reviewed by NOBODY (OOPS!).

This line should appear before the description and blank lines before and after.

> Source/WebCore/ChangeLog:9
> +        No new tests. (OOPS!)

Please remove this line or explain why this change can't be tested.

> Source/WebKit/chromium/ChangeLog:8
> +        Reviewed by NOBODY (OOPS!).

Again this line should appear above the description and a blank line.

> Source/WebKit/chromium/public/WebView.h:328
> +    virtual void didCloseContextMenu() { }

Shouldn't this be a pure virtual function?

> Source/WebKit/gtk/ChangeLog:8
> +        Reviewed by NOBODY (OOPS!).

Ditto.

> Source/WebKit/gtk/webkit/webkitwebview.cpp:320
> +    HitTestResult result = controller->hitTestResult();
> +
> +    Node* node = result.innerNonSharedNode();
> +    if (!node)
> +        return;

Why do we need to hit-test?

> Source/WebKit/gtk/webkit/webkitwebview.cpp:321
> +    sleep(1000);

Where did this sleep(1000); come from?

[Ryosuke Niwa]

Comment on attachment 112334 [details]
updated patch with chromium + gtk port changes.

View in context: https://bugs.webkit.org/attachment.cgi?id=112334&action=review

Also, I'm wondering

>> Source/WebKit/gtk/webkit/webkitwebview.cpp:320
>> +        return;
> 
> Why do we need to hit-test?

Ah, now I see why you need a hit testing here to access frame. I would move this code into ContextMenuController if I were you.

[Chang Shu]

Comment on attachment 112334 [details]
updated patch with chromium + gtk port changes.

View in context: https://bugs.webkit.org/attachment.cgi?id=112334&action=review

> Source/WebCore/page/EventHandler.cpp:2269
> +void EventHandler::willShowContextMenu()
> +{
> +}
> +
> +void EventHandler::respondToContextMenuDismissal()
> +{
> +}
> +

What do these blank functions do?

[Chang Shu]

(In reply to comment #8)
> (From update of attachment 112334 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=112334&action=review
> 
> > Source/WebCore/page/EventHandler.cpp:2269
> > +void EventHandler::willShowContextMenu()
> > +{
> > +}
> > +
> > +void EventHandler::respondToContextMenuDismissal()
> > +{
> > +}
> > +
> 
> What do these blank functions do?

Ryosuke told me they will be implemented in a follow-up patch. So I think it's good to put notImplemented() there.