605465 - Write tests checking that there is no placeholder frame/node when placeholder attribute isn't set

Status: RESOLVED WONTFIX

(Core :: Layout: Form Controls, defect)

Description

[Mounir Lamouri (:mounir)]

Tests wrote in bug 553097 have been broken by bug 604386. So, we have to found another way to check that if possible.

Comment 1

[Mounir Lamouri (:mounir)]

As Alexander said on IRC, it would be better to access nsIContent::GetChildren from a XPCOM interface instead of relying on a11y API.

Comment 2

[(no longer active)]

(In reply to comment #1)
> As Alexander said on IRC, it would be better to access nsIContent::GetChildren
> from a XPCOM interface instead of relying on a11y API.

nsIContent methods are not exposed through XPCOM.

I wonder if we can apply a UA stylesheet to a document setting the placeholder DIV's background color to green or something, and then reftest it against another document to make sure that the style has not been resolved?

One possible way to do that is this method <http://mxr.mozilla.org/mozilla-central/source/editor/idl/nsIEditorStyleSheets.idl#87>, although I'm not sure if it would work with non-editable documents.  But there should be nothing preventing us from making the document editable if needed.

Comment 3

[Robert O'Callahan (:roc) (email my personal email if necessary)]

I don't think we should stress out here. Bug 553097 is basically a performance optimization. Those are often hard to test directly. If it's really hard to test this one directly, that's OK, we should just not bother. I certainly wouldn't want to add new API just to test this.

Comment 4

[(no longer active)]

(In reply to comment #3)
> I don't think we should stress out here. Bug 553097 is basically a performance
> optimization. Those are often hard to test directly. If it's really hard to
> test this one directly, that's OK, we should just not bother. I certainly
> wouldn't want to add new API just to test this.

Makes sense to me.

Comment 5

[Mounir Lamouri (:mounir)]

Ok, closing this then.