[Webkit-unassigned] [Bug 67277] REGRESSION (r94132): broke fast/loader/location-port.html on GTK

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=67277





--- Comment #14 from Devdatta Deshpande <webkit.devdatta at gmail.com>  2011-11-02 01:31:03 PST ---
(In reply to comment #13)
> (From update of attachment 111957 [details])
> I think it would be better to handle this in soupURIToKURL by looking up the port part of the SoupURI nad setting it on the new KURL.

The port part of SoupURI is by default 0, i.e. even if a port is not specified.

For example, consider following two URLs:
1. file:///:0/home/code/WebKit/LayoutTests/fast/loader/location-port.html
2. file:///home/code/WebKit/LayoutTests/fast/loader/location-port.html

In both the cases soup_uri_to_string would return URL #2 and port as 0. 
Whereas, for case #1 KURL will return port as 0 and KURL::hasPort would return true and for case #2 port is 0 and KURL::hasPort returns false.

Thus I don't think we can figure out if a port is set to 0 from SoupURI. So if we try to handle it in soupURIToKURL, test case #2 fails in location-port.html
i.e. shouldBe('internalFrame.contentWindow.location.port == ""', true);

As mentioned earlier, we should not be setting m_url again in ResourceRequest::updateFromSoupMessage as it is already filled. It is ok to do so in ResourceResponse.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.