Bug 22532
Summary: | [Mac] 307 redirect resets method to GET | ||
---|---|---|---|
Product: | WebKit | Reporter: | Adam Barth <abarth> |
Component: | Forms | Assignee: | Nobody <webkit-unassigned> |
Status: | RESOLVED INVALID | ||
Severity: | Normal | CC: | ap, beidson, collinj, julian.reschke, mrowe, sam |
Priority: | P2 | ||
Version: | 528+ (Nightly build) | ||
Hardware: | All | ||
OS: | All | ||
URL: | http://crypto.stanford.edu/~collinj/test/redirection/ |
Adam Barth
According to the HTTP RFC at http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html, an HTTP 307 redirect should keep the current HTTP method. However, Safari resets the method to GET. I haven't debugged this enough to know whether this is a WebKit bug or a CFNetwork bug.
Test case:
http://crypto.stanford.edu/~collinj/test/redirection/
Attachments | ||
---|---|---|
Add attachment proposed patch, testcase, etc. |
Alexey Proskuryakov
See also: bug 11118, bug 11119.
Darin Fisher (:fishd, Google)
The redirect request is constructed by the ResourceHandle implementation, so this is likely a bug with the underlying network library.
Julian Reschke
Note that this is still the case in Safari 5.0.2, although it may be limited to synchronous XHR.
Adam Barth
This line of code appears to do this work:
http://trac.webkit.org/browser/trunk/Source/WebCore/platform/network/mac/ResourceHandleMac.mm?rev=90834#L695
Perhaps we're not hitting that line in synchronous requests?
Alexey Proskuryakov
> The redirect request is constructed by the ResourceHandle implementation, so this is likely a bug with the underlying network library.
Yes, that's how we see it too.
<rdar://problem/10859411>/<rdar://problem/13625208>.