Bug 41160 – Spatial Navigation: make elements in inner frames nested more than 1 level deep focusable

• Status: CLOSED
• Resolution: FIXED
• Component: Accessibility

[Antonio Gomes]

Spatial navigation does not work fine traversing nested iframe content as in [1].

[1] http://www.w3schools.com/TAGS/tryit.asp?filename=tryhtml_iframe

patch coming ...

[Antonio Gomes]

Created attachment 59656 [details]
patch v1

[Antonio Gomes]

(In reply to comment #1)
> Created an attachment (id=59656) [details]
> patch v1

@Simon, Kenneth: could any of you help with reviewing this patch?

[Simon Fraser (smfr)]

Comment on attachment 59656 [details]
patch v1

> +bool checkDescendenceRecursively(Document* baseDocument, Node* node)

Check for what? I think this method name could be improved.

[Antonio Gomes]

Thank you for the review.

(In reply to comment #3)
> (From update of attachment 59656 [details])
> > +bool checkDescendenceRecursively(Document* baseDocument, Node* node)
> 
> Check for what? I think this method name could be improved.

This method works like Node::isDescendantOf, but for nested documents.

Basically, it checks if a given node is descendant of a given baseDocument. If it is not, it goes recursively up in the document chain checking if the current document is a descendant of baseDocument.

Would you have a better name to suggest based on the description?

[Simon Fraser (smfr)]

isDeepDescendantOf?

[Antonio Gomes]

Created attachment 60056 [details]
patch v2 - s/checkDescendanceRecursively/isDeepDescendantOf/

As per Simon suggestion, naming the added method as isDeepDescendantOf/

[Antonio Gomes]

Created attachment 60062 [details]
(r62179 - r=smfr) patch v2.1 - right patch

[Simon Fraser (smfr)]

Comment on attachment 60062 [details]
(r62179 - r=smfr) patch v2.1 - right patch

Since this addition is a function, not a method, the name should be more like isNodeDeepDescendentOfDocument(), or perhaps documentTreeContainsNode().

r=me either way.

[Antonio Gomes]

Comment on attachment 60062 [details]
(r62179 - r=smfr) patch v2.1 - right patch

Clearing flags on attachment: 60062

Committed r62179: <http://trac.webkit.org/changeset/62179>

[Antonio Gomes]

(In reply to comment #8)
> (From update of attachment 60062 [details])
> Since this addition is a function, not a method, the name should be more like isNodeDeepDescendentOfDocument(), or perhaps documentTreeContainsNode().
> 
> r=me either way.

I used the former. Thank you!

[Simon Hausmann]

Revision r62179 cherry-picked into qtwebkit-2.0 with commit 18375caeaf6b2cf99c45e1b225e23c6028e1dd68