946119 - XML fragment parsing algorithm doesn't use the context element's default namespace URI

Status: RESOLVED INVALID

(Core :: DOM: HTML Parser, defect)

Description

[Ryosuke Niwa]

Attachment: test.html

Right now, Gecko doesn't use the context element's namespace as the default namespace in the XHTML fragment parsing algorithm
when xmlns is not set as specified in http://www.whatwg.org/specs/web-apps/current-work/multipage/the-xhtml-syntax.html#parsing-xhtml-fragments

See the attached test case.

Also see https://bugs.webkit.org/show_bug.cgi?id=125132.

Comment 1

[William Chen [:wchen]]

The attached test looks broken. It gives me the error: "contextElement.body is undefined". contextElement is a HTMLBodyElement. It looks like you were trying to get the element created by innerHTML. After fixing this up the test passes for me. A quick glance at the code looks like we are doing the right thing with namespaces.

It also looks like the test FAIL was committed as an expected result in webkit.

rniwa: Was the test fail intentional?

Comment 2

[Ryosuke Niwa]

Oops, you're right. There is a bug in the code.