[Webkit-unassigned] [Bug 150740] Link preview doesn't work on XHTML pages with Content-Type header as `application/xhtml+xml`

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Sun Nov 1 11:44:34 PST 2015


https://bugs.webkit.org/show_bug.cgi?id=150740

--- Comment #8 from Beth Dakin <bdakin at apple.com> ---
(In reply to comment #7)
> (In reply to comment #6)
> > Is it really correct to just check the tag name, or is this one of those
> > cases where XML namespaces should be taken into account?
> 
> I was planning to comment on that but had lost track of this bug. Thanks for
> drawing my attention back here! As Dan implies, this new code is incorrect
> in the presence of non-HTML elements that happen to have the same local name
> as HTML elements.
> 
> This clickableElementName field is apparently a copy of the DOM nodeName
> attribute, shipped across processes. We can’t use nodeName alone to check if
> something is a particular HTML element.
> 
> If our intent is to use this attribute to check if something is a particular
> HTML element, then I suggest we replace this code:
> 
>     info.clickableElementName = hitNode->nodeName();
> 
> With something more like this:
> 
>     if (is<HTMLElement>(*hitNode))
>         info.clickableElementName =
> downcast<HTMLElement>(*hitNode).localName();
> 
> This will always be lowercase, and so there will be no need to fold case on
> the UI process side.
> 
> If it happens that we want to use this for SVG elements as well, and there
> is also no practical concern about collision between HTML and SVG local
> names (I think there are very few such collisions), then it could instead be
> this:
> 
>     if (is<HTMLElement>(*hitNode) || is<SVGElement>(*hitNode))
>         info.clickableElementName = downcast<Element>(*hitNode).localName();

Thanks Darin and Dan! I will fix this.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20151101/2309a0f7/attachment.html>


More information about the webkit-unassigned mailing list